Development Geology Reference Manual
edited by
Diana Morton-Thompson Arnold M. Woods
AAPG Methods in Exploration Series, No. 10
Published by The American Association of Petroleum Geologists Tulsa, Oklahoma, U.S.A. 74101
Copyright © 1992 by The American Association of Petroleum Geologists All Rights Reserved Printed in the U.S.A.
Published April 1993; Second printing (with minor revisions) May 1997; Third printing August 1999
ISBN: 0-89181-660-7 A A P G grants permission for a single photocopy of an item f r o m this publication for personal use. Authorization for additional copies of items f r o m this publication for personal or internal u s e is g r a n t e d b y A A P G p r o v i d e d that the base fee of $15.00 p e r copy is p a i d directly to the Copyright Clearance Center, 222 Rosewood Drive, Danvers, Massachusetts 01923 (phone: 978/750-8400). Fees are subject to change. A n y f o r m of electronic or digital scanning or other digital transformation of portions of this publication into computer-readable and/or transmittable form for personal or corporate use requires special permission from, and is subject to fee charges by, the AAPG.
Association Editor: Susan Longacre Science Director: Gary D. Howell Publications Manager: Cathleen P. Williams Special Projects Editor: Anne H. Thomas Production: Kathy A. and Dana M. Walker, Editorial Technologies, Renton, WA
THE AMERICAN ASSOCIATION OF PETROLEUM GEOLOGISTS (AAPG) DOES NOT ENDORSE OR R E C O M M E N D ANY PRODUCTS OR SERVICES THAT MAY BE CITED, USED OR DISCUSSED IN AAPG PUBLICATIONS OR IN PRESENTATIONS AT EVENTS ASSOCIATED WITH THE AAPG.
C o v e r p h o t o b y Joel M. Dexter, Illinois State Geological Survey. Gas flare in the Salem oil field, Illinois Basin, south-central Illinois. This field has p r o d u c e d approximately 400 million barrels of oil f r o m Devonian and Mississippian sandstones and carbonate rocks.
This and other AAPG publications are available from:
The AAPG Bookstore P.O. Box 979 Tulsa, OK 74101-0979 Telephone: +1-918- 584-2555 or 1-800-364-AAPG (USA) Fax: +1-918-560-2652 or 1-800- 898-2274 (USA) www.aapg.org
Geological Society Publishing House Unit 7, Brassmill Enterprise Centre Brassmill Lane, Bath, U.K. BAl 3JN Tel +44-1225-445046 Fax +44-1225-442836 www.geolsoc.org.uk
Australian Mineral Foundation AMF Bookshop 63 Conyngham Street Glenside, South Australia 5065 Australia Tel. +61-8-8379-0444 Fax +61-8-8379-4634 www.amf.com.au/amf
Affiliated East-West Press Private Ltd. G-I/16 Ansari Road DaryaGanj N e w Delhi 110 002 India Tel +91 11 3279113 Fax +91 11 3260538 e-mail: aewp.newdel@aworld.net
Foreword
Whether evaluating offshore wells in an international arena or managing a domestic waterflood, a development geologist is often called on to bridge the gap between geologists, geophysicists, geochemists, petrophysicists, and drilling, production, and reservoir engineers. This requires a broad knowledge base of specific techniques and technologies, as well as the ability to integrate and communicate multidisciplinary data. These skills are rarely learned in a college classroom; instead they are learned on the job.
With this in mind, the AAPG Development Geology Reference Manual was designed to guide both newcomers and more experienced hands through a spectrum of concepts, technologies, and methods that encompass the day-to-day work of a development geologist. The Manual is not intended to be a definitive work on any one topic, but is a handy desktop or field reference where key facts on many topics are presented in a succinct, easy-to-use format. The Manual provides a great deal of specific, practical information, including
• Definitions of technical terms • Explanations of multidisciplinary techniques and technologies • Introductions to methodologies that can be used to solve problems and aid the reader
in performing immediate tasks • Cross references to other chapters in the Manual where additional information is given
on a particular topic • Common abbreviations and conversion formulas • Lists of References Cited (given at the end of each of the ten parts) that can serve as a
starting point for additional literature searches.
The loose-leaf, three-ring binder format was chosen so that pertinent chapters can be taken out and used in the field or in discussions with colleagues down the hall. This format also allows room for personal notes and for addition of updates as the various technical areas continue to expand and evolve.
The Manual comprises ten parts that fall into four broad categories. The first group consists of Parts 1 and 2, which are related to activities that usually precede drilling or reservoir development. In Part 1, Land and Leasing, James Tinkler discusses land and leasing practices, including objectives and procedures for acquiring acreage tracts. In Part 2, Economics and Risk Assessment, Pete Rose and Robert Thompson discuss fundamental economic considerations related to oil and gas exploitation.
The second group consists of three parts that focus on wellsite equipment and data collection and analyses. Part 3, Wellsite Methods, is edited by Arnold Woods, Byram Reed, and Diana Morton-Thompson and contains chapters on drilling and evaluation equipment and on wellsite data collection and interpretation. Part 4 on Wireline Methods, edited by Mark Alberty, focuses on borehole evaluation, including common logging equipment, and interpretation methods. Part 5 on Laboratory Methods, edited by Frank Ethridge, concentrates on laboratory analytical methods and data interpretation.
The third broad group comprises three parts that cover integration and interpretation of data for reservoir description. Part 6 on Geological Methods, edited by Roger Slatt, contains chapters on techniques and approaches that can be used to evaluate a variety of reservoir types, as well as material on statistical analysis of geological data. Geophysical Methods, Part 7, is edited by Peter Duncan and includes information on seismic and other remote data acquisition, processing, and interpretation. Integrated Computer Methods, Part 8, edited by Brian Shaw, explains how to use computers to quantify and present reservoir data effectively from a geological and geophysical perspective.
The final g r o u p consists of two parts that encompass engineering techniques and technologies commonly encountered by development geologists. Part 9 on Production Engineering Methods, edited by Steve Holditch, includes chapters on completion, stimulation, and production procedures and on various reservoir testing and evaluation methods. Part 10 on Reservoir Engineering Methods, edited by E. G. (Skip) Rhodes, discusses fundamental techniques for understanding, quantifying, and modeling reservoir properties.
The chapters included in this comprehensive manual were solicited from a wide spectrum of authors. There are approximately 125 individual authors and editors that represent many major corporations, independents, consultants, service companies, state agencies, and
iii
iv
Foreword
universities. Although care was taken to edit papers for greater clarity and to coordinate the topics as much as possible, the reader will find a variety of writing styles and content levels in the different parts. If you have questions about the information presented in a particular chapter, we encourage you to contact the author(s) of the chapter directly.
ACKNOWLEDGMENTS
The compilation editors would like to acknowledge our employers during the period this Manual was in preparation and thank them for their support on this project: Chevron U.S.A. and ARCO (D. M.-T.) and Conoco (A. W.). We would also like to thank the Part Editors, outside reviewers, and individual authors for their hard work. Without their dedication, this Manual would not have been possible. We also wish to thank the AAPG Committee on Development Geology for moral support during this process. We extend our special thanks and appreciation to Cathleen Williams, AAPG Publications Manager (alias "the Velvet Hammer"), and Anne Thomas, AAPG Special Projects Editor, for all of their help, friendship, and patience. We would also like to thank Kathy Walker at Editorial Technologies for her work on design, layout, copyediting, and production management.
It is our hope that the Development Geology Reference Manual will serve as a useful tool for geoscientists and other industry workers who are undertaking new tasks or perfecting old skills. We also hope that it will help promote better interdisciplinary communication by introducing fundamental concepts on a variety of topics. This should lead to the generation of new ideas and approaches for solving reservoir problems.
Diana Morton-Thompson Arnold M. Woods
Contents
Foreword
i"
Aboutthe Editors
ix
Common Oil Field Abbreviations
Jt
Conversion Chart
xiv
Acknowledgment for Contributions
xv
Part 1. Land and Leasing
1
Edited by James C. Tinkler
Introduction
3
Functions of a Petroleum Landman
.4
Land Description and Maps
5
Determining Owners of Oil and Gas Interests, and Methods of Conveyance
9
Nature of the Oil and Gas Lease
13
Oil and Gas Contracts
16
References Cited
19
Part 2. Economics and Risk Assessment
21
Edited by Peter R. Rose and Robert S. Thompson
Introduction
23
Fundamental Economic Equations for Oil and Gas Property Evaluation
24
Uncertainties Impacting Reserves, Revenue, and Costs
25
Expected Value and Chance of Success
30
The Time Value of Money
35
Building a Cash Row Model
38
About Taxes
-43
Key Economic Parameters
.47
Dealing with Risk Aversion
52
Economics of Property Acquisitions
54
References Cited
56
Part 3. Wellsite Methods
57
Edited by Arnold M. Woods, Byram Reed, and Diana Morton-Thompson
Introduction
59
WellPlanning
60
Land Rigs
-62
OffshoreRigs
.65
Rig Personnel
67
WellsiteSafety
.68
Wellbore Trajectory
.71
DrUling Fluid
76
Pressure Detection
79
Fishing
33
Drilling Problems
87
Measurement While Drilling
88
Rate of Penetration
.91
Wellsite Math
.93
Mudlogging Equipment, Services, and Personnel
98
Mudlogging: TheMudlog
101
Mudlogging: Drill Cuttings Analysis
104
Mudlogging: Gas Extraction and Monitoring
106
Show Evaluation
109
Conventional Coring
115
Sidewall Coring
119
Core Orientation
122
Core Handling
125
Core Alteration and Preservation
127
Drill Stem Testing
131
References Cited
138
v
vi
Contents
Part 4. Wireline Methods
141
Edited by Mark W. Alberty
Introduction
143
Basic Open Hole Tools
144
Basic Tool Table
150
Basic Cased Hole Tools
151
Wireline Formation Testers
154
Dipmeters
158
Borehole Imaging Devices
163
Preprocessing of Logging Data
167
Determination of Water Resistivity
170
Quick-Look Lithology from Logs
174
Standard Interpretation
180
Difficult Lithologies
186
Formation Evaluation of Naturally Fractured Reservoirs
192
References Cited
194
Part 5. Laboratory Methods
195
Edited by Frank Ethridge
Introduction
197
Core Description
198
Overview of Routine Core Analysis
-201
Porosity
-204
Permeability
-210
Core-Log Transformations and Porosity-Permeability Relationships
214
Wettability
-218
Capillary Pressure
-221
Relative Permeability
226
Paleontology
229
Thin Section Analysis
233
SEM, XRD7 CL, and XF Methods
237
Oil and Condensate Analysis
241
Oilfield Water Analysis
247
Rock-Water Reaction: Formation Damage
249
References Cited
253
Part 6. Geological Methods
259
Edited by Roger M. Slatt
Introduction
261
Lithofacies and Environmental Analysis of Clastic Depositional Systems
263
Carbonate Reservoir Models: Fades, Diagenesis, and Flow Characterization
269
Reservoir Quality
275
Geological Heterogeneities
278
Row Units for Reservoir Characterization
282
Effective Pay Determination
286
Conversion of Well Log Data to Subsurface Stratigraphic
and Structural Information
289
Subsurface Maps
294
Geological Cross Sections
300
Fluid Contacts
305
Evaluating Stratigraphically Complex Fields
311
Evaluating Diagenetically Complex Reservoirs
314
Evaluating Tight Gas Reservoirs
321
Evaluating Fractured Reservoirs
-326
Evaluating Structurally Complex Reservoirs
331
Statistics Overview
339
Correlation and Regression Analysis
343
Multivariate Data Analysis
345
Monte Carlo and Stochastic Simulation Methods
348
References Cited
350
Contents
vii
Part 7. Geophysical Methods Edited by Peter Duncan
Introduction Seismic Data Acquisition on Land Marine Seismic Data Acquisition Basic Seismic Processing Seismic Migration Displaying Seismic Data Seismic Interpretation Mapping with Two-Dimensional Seismic Data Three-Dimensional Seismic Method Vertical and Lateral Seismic Resolution and Attenuation Synthetic Seismograms Forward Modeling of Seismic Data Seismic Inversion Amplitude Versus Offset (AVO) Analysis Checkshots and Vertical Seismic Profiles Cross-Borehole Tomography in Development Geology Full Waveform Acoustic Logging The Gravity Method Borehole Gravity Magnetics Electrical Methods References Cited
355
357 358 361 364 372 377 .........379 381 385 388 390 392 395 398 401 404 409 411 413 -415 -417 420
Part 8. Integrated Computer Methods
423
Edited by Brian Shaw
Introduction
425
Introduction to Contouring Geological Data with a Computer
426
Using and Improving Surface Models Built by Computer
431
Log Analysis Applications
-441
A Development Geology Workstation
.447
Two-Dimensional Geophysical Workstation Interpretation: Generic Problems
and Solutions
451
References Cited
455
Part 9. Production Engineering Methods
457
Edited by Stephen A. Holditch
Introduction
459
Production Histories
460
Well Completions
463
Stimulation
469
Production Testing
474
Pressure Transient Testing
477
Surface Production Equipment
482
Artificial Lift
485
Production Logging
488
Production Problems
492
Workovers
496
References Cited
-500
Part 10. Reservoir Engineering Methods
501
Edited by E. G. Rhodes
Introduction
-503
Petroleum Reservoir Ruid Properties
504
Fundamentals of Fluid Flow
508
viii
Contents
Part 10 (continued)
Reserves Estimation
-513
Drive Mechanisms and Recovery
518
Waterflooding
-523
Enhanced Oil Recovery
527
Reservoir Modeling for Simulation Purposes
531
Conducting a Reservoir Simulation Study: An Overview
536
References Cited
-541
Index
543
About the Editors
Diana Morton-Thompson is a consulting geologist in both the energy and environmental fields. She has a M.A. in geology from The University of Texas at Austin, where she also worked at the Texas Bureau of Economic Geology. She has ten years of diversified foreign and domestic experience with Chevron U.S.A. and ARCO Research and Technical Service, performing applied research, through the integration of geological, geophysical, geochemical, petrophysical, and engineering data, on reservoir characterization and management. This work has focused particularly on fades analysis of modern and ancient depositional systems; facies effects on reservoir quality, distribution, continuity, and behavior; and how hydrocarbon recoveries can be subsequently enhanced. Diana is now with Wilkins & Wheaton, Inc., in Kalamazoo, Michigan. She is currently involved in corporate training; m u l t i d i s c i p l i n a r y project m a n a g e m e n t and team b u i l d i n g ; and integration of multidisciplinary data to describe surface and subsurface soil-rock-fluid systems for reservoir and property evaluation and management, environmental site characterization and remediation, and environmental regulatory compliance auditing. Diana has been a member of the AAPG Committee on Development Geology since 1987, and in 1991 was appointed Chairman of that committee.
Arnold M. Woods received his B.A. in geology from San Francisco State University in 1976. Following graduation, he worked for two years as a mudlogging engineer and unit manager in California, Alaska, and Canada. After receiving his M.A. in geology from The University of Texas at Austin in 1981, he was employed by Phillips Petroleum as a development geologist, successfully working shallow and deep oil and gas plays in the Anadarko basin. He started with Conoco in 1986 as a member of the International Operations Group and worked exploration and development projects in Egypt, Tunisia, and Angola. Since 1988, he has been assigned to the Casper Division office, where he is presently evaluating EOR potential and performing detailed reservoir studies for fields in the Williston, Powder River, Bighorn, and North Park basins. His interests span a range from clastic sequence stratigraphy to dinosaur extinction. He is an active member of the Wyoming Geological Association and of AAPG and has been a member of the AAPG Committee on Development Geology since 1988.
Common Oil Field Abbreviations
aa abd abdt acdz ACT addn AE amt anhy alys ann AOF API appx arg ARO att avg AW B/ BA (W) BC bd(d) BDO bent BF (H) BHL ВНР BHT bl bldg bledg blk blkt BLO BLOTBR BLW BN (O) (W) B(O) (W) BOE ВОР ВОР (D) (H) BP brkn brn BS BSUW BPV brg B(S)W BTM BU brk CAOFP carb CC CFGP (D) (H) CG
eg
as above abandon(ed) abundant acidize(d) automatic custody transfer addition asphalt emulsion amount anhydrite analysis annulus absolute open flow American Petroleum Institute approximate(ly) argillaceous at a rate of attempt(ed) average acid water base barrels acid (water) barrels condensate bed (ded) barrels diesel oil bentonite barrels fluid per (hour) bottom hole location bottom hole pressure bottom hole temperature black building bleeding block blanket barrels load oil barrels load oil to be recovered barrels load water barrels new (oil) (water) barrels (oil (water) blow out equipment blow out preventer barrels oil per (day) (hour) bridge plug broken brown basic sediment black sulfur water back pressure valve bearings barrels (salt) water bottom build up broke or break calculated absolute open flow potential carbonaceous casing collar cubic feet gas per (day) (hour) corrected gravity coarse grained
X
CH chl cht CIBP CIP CIRC CK CJPF Cl cmt CO CO2 comm compl cond CONDS cong conn cont coord corr corro CP cplg cr CSG CSPK
CtC
CTR CUSH
CVgS
CX (In) d D&A DC DD dec deer deg delr detr detr dev DF DHC dia direc displ DIST dk DM dn dns DO DOC dol(o) DP dpn DRL
choke channel chert cast iron bridge plug cement in place circulate choke casing jets per foot chloride cement cleaned out carbon dioxide communicate(d) complete(d) condition or conduct condensate conglomerate connection continue coordinates correct(ion) corrosion casing pressure coupling core casing casing packer contact cable tool rig cushion cavings coarse crystalline day dry and abandoned drill collar drilled deeper decide or decision decrease degree deliver determine detrital deviat(ed) (ion) derrick floor or drilling fluid dry hole contribution diameter direction displace(d) distillate dark drilling mud down dense drilled out drilled out cement dolomite drill pipe deepen drill
DST DWT elec ELEV emul equiv est estab eval F FARO FB FBHP fed FEL FER FFP
%
FHH FL fluor FNL F/O Fm. foss FP fr frac FSH FSIP FSL FTP FWL fx (In) FG GB GCDM GCR GGE GIDP GIH GIP GL glauc GOC GOR Gr. GRD grn
gry GSO GTS
gty gyp H2S hd HFO HH hi HOCM HO&GCM horiz
drill stem test dead weight tester electric elevation emulsion equivalent estimate(d) establish(ed) evaluate(d) flowed flowed at rate of fair blow flowing bottom hole pressure federal feet from east line fluid energy rate final flowing pressure fine grain(ed) final hydrostatic head fluid or fluid level fluorescent(ce) feet from north line farmout Formation fossiliferous flowing pressure or free point from fracture fish final shut-in pressure feet from south line flowing tubing pressure feet from west line finely crystalline full gauge good blow gas cut drilling mud gas-condensate ratio gauge gas in drill pipe going in hole gas in pipe ground level glauconitic gas-oil contact gas-oil ratio Group ground green gray good show of oil gas to surface gravity gypsum hydrogen sulfide hard hole full of oil hydrostatic head high heavily oil cut mud heavily oil and gas cut mud horizontal
HP hr IF(P) IHH immed imp incl incr inh inj int insp INTFP IP(F) (P) INTSIP ISIP inst IG J&A JK JT JKB KB KO LCM LD LFL Hth Htho Im In LNR LO Ioc LPG Is Ise LT LTS Iv max Mb. MC MCA md. med MG MI MIE min. min MLU MN MO MORT MT MTD MTS MW mx (In) NEP NIP
Abbreviations
xi
hydrostatic pressure hour initial flow(ing) (pressure) initial hydrostatic head immediate impression inclusions or including increase inhibitor or inhibit inject interval inspect intermediate flowing pressure initial potential (flowing) (pumping) intermediate shut-in pressure initial shut-in pressure install in gauge junked and abandoned junk joint junk basket keHy bushing kicked off lost circulation material land low fluid level Hthology Hthographic lime line liner load oil locate or location Hquified petroleum gas limestone lease lower or light tubing long tubing string leave maximum member mud cut mud cleanout agent milHdarcies medium mud gas move in move in equipment minute minimum mud logging unit midnight move moving off rotary tools middle tubing measured total depth mud to surface muddy water medium crystalHne net effective pay nipple or nipple up
xii
Abbreviations
NOJV NR NS NSOFC
N2 O&GCM OB OCM OD OE OEG OH OIP ON ool opn opr orig OS OStn OTD OTS owe OWDD OWWO OOG P P&A P&F P&P par PB PBTD pet PD perf perm
Pf PKR P.L. pld
Pig pmp POOH POP por poss pot ppm PPP prep press prob proc prod PU qtz R&T R&P R rbds RBP
non-operated joint venture(s) no report no show no show, odor, fluorescence, cut nitrogen oil and gas cut mud off bottom oil cut mud outside diameter open end oil equivalent gas open hole oil in place overnight oolitic open operator original overshot oilstain old total depth oil to surface oil-water contact oil well drilled deeper old well worked over out of gauge pumped plug and abandon pump and flow porosity and permeability paraffin plug back plug back total depth percent per day perforate(d) permeability per foot packer pipe line pulled pulling pump pulled (pulling) out of hole put on pump porosity possible potential parts per million pinpoint porosity prepare pressure probable or problem procedure produce pulling unit quartz rods and tubing rods and pump radioactivity red beds retrievable bridge plug
RD rd rec recmd recv rel repl repr req retr RGE RIH RM rng rmv RO RPM rpt RR rr Rt rtnr RU Rw S&F SAB sal sat scat SCF sd sec sep SFP (S)SG sh SI(C) (P) SIBHP silic SITB sh SLM slt(st) SN (S)SO SOCM spl SPM SPD spt sqz
SS
St
STBOIP std stk stn STS SULW surv surf SW
rig down red recover recommend receive released replace repair request retrieve range running (ran) in hole ream running remove reversed out revolutions per minute report rotary rig rare true resistivity retainer rig up water resistivity swab and flow strong air blow salinity saturate(d) scattered standard cubic feet sand section separator surface flowing pressure (slight) show gas shale shut-in (casing) (pressure) shut-in bottom hole pressure siliceous shut-in tubing pressure slightly steel line measurement silt(stone) seating nipple (slight) show oil slightly oil cut mud sample strokes per minute spud spot squeeze sandstone state stock tank barrels oil in place standard or stand staked stain short tubing string sulfur water survey surface salt water
swb SWCM
SX
т/
ТА tbg TD temp TJPF Т.О. TOC TP tr trt trmt tst TST TSTM TVD TWP
U
UT
V
vac vel
swab salt water cut mud sack top temporarily abandoned tubing total depth temperature tubing jet(s) per foot tool open top of cement tubing pressure trace treat treatment test true stratigraphic thickness too small to measure true vertical depth township unit upper tubing very vacuum velocity
vert vis vol vole vug w/ w/o WB WC WCM WDW Wgt wh WH WI W.I. wk WL WO (C) (O) WS wtr WW WP xln yell zn
Abbreviations
xiii
vertical viscosity volume volcanic vuggy with without water base water cushion water cut mud water disposal well weight white wash water injection working interest weak wire line waiting on (cement) (orders) whipstock water water well wash pipe crystalline yellow zone
Conversion Charta
To Convert:
Linear units inches (in.) feet (ft) miles (mi) nautical miles (nmi) meters (m) centimeters (cm) kilometers (km) kilometers (km)
Area units square feet (ft2) square miles (mi2) acres (ac) square meters (m2) square kilometers (km2) hectares (ha)
Volume units cubic feet (ft3) cubic meters (m3) cubic meters (m) metric tons (MT) thousand cubic feet (mcf) million cubic feet (mmcf) billionb cubic feet (bcf) trillion cubic feet (tcf) gallons (gal) liters (L) liters (L) barrels (bbl) barrels (bbl) barrels (bbl)
Mass units pounds (lb) kilograms (kg) short tons (ton) metric tons (MT)
Pressure units
pound-force per square inch (psi) pound-force per square inch (psi) kilopascals (kPa) kilopascals (kPa) bars (bar) bars (bar) Energy units
British thermal units (Btu)
kilojoules (kJ)
To Convert:
Temperature degrees CeIsius(0C) degrees Celsius (0C) degrees Fahrenheit (0F) kelvins (K)
To:
centimeters (cm) meters (m) kilometers (km) kilometers (km) feet (ft) inches (in.) miles (mi) nautical miles (nmi)
square meters (m2) square kilometers (km2) hectares (ha) square feet (ft2) square miles (mi2) acres (ac)
cubic meters (m3) cubic feet (ft3) barrels (bbl) barrels (bbl) cubic meters (m3) thousand cubic meters (thousand m3) million cubic meters (million m3) billion'3 cubic meters (billion m3) liters (L) gallons (gal) barrels (bbl) cubic meters (m3) metric tons (MT) liters (L)
kilograms (kg) pounds (lb) metric tons (MT) short tons (ton)
kilopascals (kPa) bars (bar) pound-force per square inch (psi) bars (bar) pound-force per square inch (psi) kilopascals (kPa)
kilojoules (kJ) British thermal units (Btu)
To:
degrees Fahrenheit(T) kelvins (K) degrees Celsius (0C) degrees Celsius (0C)
Miscellaneous
1 °F/100 ft = 1.8°C/100 m 1 °C/km = 0.055°F/100 ft =2.9°F/mi
million year interval = m.y. million years before present = Ma billion years before present = Ga
Multiply By:
2.54 0.305 1.609 1.852 3.281 0.394 0.621 0.540
0.093 2.590 0.405 10.764 0.386 2.471
0.028 35.315
6.290 7.34 (approx.) 28.317 28.317 28.317 28.317 3.785 0.264 0.0063 0.159 0.14 (approx.) 158.987
0.454 2.205 0.907 1.102
6.895 0.069 0.145 0.01 14.504
100
1.055 0.948
Use Formula:
(°C x 9/5) + 32 0 C + 273.15 (°F - 32) x 5/9 K-273.15
aFor more information on units and conversions, see The Sl Metric System of Units and SPE Metric Standard (1982), published by the Societyof Petroleum Engineers, 39 p., b 1 billion = 109.
xiv
AAPG wishes to thank the following for their generous contributions
to the
Development Geology Reference Manual
APTECH Associates ARCO Exploration and Production Technology
Douglas C. Bleakly Parke A. Dickey
Arnold Orange/Arnold Orange Associates Halliburton Energy Services
Contributions are applied against the production costs of the publication, thus directly reducing the book's purchase price and
increasing its availability to a greater audience.
XV
Partl
LAND AND LEASING
Contents
• Introduction • Functions of a Petroleum L a n d m a n • Land Description and Maps • Determining O w n e r s of Oil a n d Gas Interests, a n d
M e t h o d s of Conveyance • N a t u r e of the Oil and Gas Lease • Oil and Gas Contracts • ReferencesCited
edited by
James C. Tinkler
University of Houston-Downtown Houston, Texas, U.S.A.
Introduction
The purpose of Part 1, Land and Leasing, is to provide the user with the basic fundamentals of the land and leasing functions that are generically and legally essential to acquiring, maintaining, and disposing of oil and gas interests underlying publicly and privately owned lands in the United States.
To achieve this purpose, this part of the Manual will provide information relating to the following:
• General facts and the functions and activities of the typical petroleum landman
• Understanding land descriptions and maps • Determining owners of oil and gas interests and
methods of conveyance • The nature and negotiation of an oil and gas lease
(OGL) • The nature and negotiation of an oil and gas contract • Sources of additional information (see References Cited
at the end of Part 1)
James С. Tinkler
University of Houston-Downtown Houston, Texas, U.S.A.
ACKNOWLEDGMENTS
My special thanks and grateful appreciation go to William B. Beall and Paul F. Neilson7 who kindly devoted their time, experience, a n d s u g g e s t i o n s to r e v i e w this p a r t of the Manual; to Gail Evans, Chair Person, Business Management and Administrative Services Department, the University of Houston-Downtown, whose support and encouragement greatly aided my effort; to my many compatriot landmen from all over the United States; to the American Association of Petroleum Landmen; and finally, to my wife Cecilia, son Dan, a n d d a u g h t e r Susan, all of w h o m f u r n i s h e d the supportive environment for this effort.
Functions of a Petroleum Landman
James C. Tinkler
University of Houston-Downtown Houston, Texas, U.S.A.
GENERAL FACTS
The United States and the individual states have followed a legal system based on Roman law, which allows private ownership of oil and gas interests. Nevertheless, the federal and state governments control the leasing of about one-third of the oil and gas interests underlying the onshore landmass of the United States (estimated at 2.2 billion acres). They also control 100% (97% federal and 3% state) of the offshore areas of the United States (estimated at 885.6 million acres). In additional, 54 million acres of Indian lands are subject to U.S. government leasing control, and leasing rights to 44 million acres in Alaska are controlled by Alaskan natives (Mineral Management Service, 1989; Petroleum Independent, 1991).
FUNCTIONS OF A TYPICAL PETROLEUM LANDMAN
The functions of a petroleum landman, who should be a highly skilled negotiator, vary from company to company. In the case of an i n d e p e n d e n t l a n d m a n , the functions vary depending upon his or her area of specialization. Generally, these functions can be divided into three (sometimes overlapping) phases (Burke, 1983; Kimball, 1982; American Association of Petroleum Landmen, 1984):
• Acquisition • Maintenance • Disposition
Acquisition Phase
The acquisition phase involves the following functions:
1. The landman must obtain the rights to explore and produce oil and gas interests underlying lands deemed prospective for exploratory or developmental purposes (including producing properties). These rights may be obtained by purchasing oil and gas leases, permits, options, contracts (farm-outs), lease trades, well trades, joint venture agreements, creation of partnerships, a n d / o r acquisition of entities.
2. Preliminary to obtaining such rights, the landman must determine the availability of the desired acreage by (a) examining the public records, (b) reviewing notices of lease sales, (c) making local inquiries, and (d) observing the activity (if any) of competitors. Also, costs must be
determined along with the terms and willingness of the owners to either lease or trade their interests.
Upon supporting the recommendations to acquire interests, obtaining the funding, and setting the terms of the trade(s), the next functions in the acquisition phase include t h e f o l l o w i n g : (1) select p e r s o n n e l ; (2) m a k e d e t a i l e d ownership checks; (3) negotiate with the interest owners on costs and terms; (4) cause or p r e p a r e the necessary documents; (5) pay the seller either by cash, check, or customer's draft, or by contract to perform the drilling or services promised; (6) obtain execution and recording of the affected documents; (7) process the documents, financial records, and data necessary to maintain the interests acquired; and finally, (8) perform the necessary "due diligence" or title curative.
Maintenance Phase
The maintenance phase during the ownership period involves lease administration related to the following functions:
1. Processing payments of delay rentals, royalties, and shut-in royalties, as well as landowner relation activities
2. Title curative work involving ownership changes affecting various payments; obtaining pooling and unitization agreements; confirming good or defensible title prior to drilling by obtaining documents for title examination; and preparing and getting execution of curative documents to satisfy the "requirements" made by title examiners (attorneys) in their title opinions
3. Promotional a n d / o r contract negotiation and preparation with outsiders or investors on deals such as farm-outs, farm-ins, bottom hole contributions, seismic options, easements, joint operating agreements, bidding agreements, unitization, and other contracts
4. Keeping track of lease and contract obligations (receivable and payable), and maintaining land statistical records
Disposition Phase
The disposition phase involves the following functions:
1. Release or surrender of interests in leases or contracts as they expire
2. Farm-out, sale, or disposal of oil and gas interests 3. Keeping land records for federal and state tax purposes
4
Land Description and Maps
James C. Tinkler
University of Houston-Downtown Houston, Texas, U.S.A.
INTRODUCTION
Any person involved in the exploration, development, and production of oil and gas is usually exposed on a daily basis to property descriptions and oil and gas maps. Fundamental to property descriptions and maps is the principle that every tract of land or point on the earth is unique and distinguishable from any other tract or point— that is, if they are described in a legally sufficient manner.
The maps discussed in this part of the Manual are limited to those that are typically called "land" maps in the oil and gas industry. Generally, land maps are planimetric, that is, they reflect a flat or horizontal surface that show the shape, dimensions, and extent of a given area or tract of land. In contrast to this, altametric (commonly called topographic) maps, additionally reflect the variations in ground surfaces with respect to sea level.
Any point on earth can be identified in terms of latitude (east-west lines parallel to one another, with a base line located on the Equator), longitude (north-south circular lines all intersecting at the true north and south poles), altitude, and time.
Several commercial map companies have revolutionized map making and map sources for use in the oil and gas industry by using modern computer and satellite positioning devices and technology, all tied to base points that are defined, identified, and delineated by the U.S. Coast and Geodetic Survey.
LAND MAPS
The land ownership map is one of the most valuable tools the landman uses because it synthesizes a majority of the information from the official records and files. The land map should serve as an overview of the activities in the area or region, along with the current lease status.
The following excerpts from a paper by Horace E. Rowald (1984, p. 1-13, chap. II) provide a s u m m a r y of pertinent information related to land descriptions and maps used in the onshore United States (see also Stamper, 1973).
[quoted material begins here] [T]here is certain basic information contained..., such as
1. Property identification by township, range, section, survey, or block
2. Namesof lessees a n d / o r operators 3. Status of leases—expiration date or held by production 4. Surface land ownership 5. Mineralownership 6. Well data (i.e., producing wells, dry holes, total depths
drilled, operator's name, and occasionally, unit outlines and topographic information)
7. N a m e s o f r o a d s 8. Namesofstreams 9. Namesoflakes
Information . . . as symbols and abbreviations is explained by a legend.
Another commonly used map is the individual tract or unit survey plat. This is prepared by . . . [a] civil engineer f r o m an "on the g r o u n d " survey of a particular tract in connection with a drilling or development program. Common procedure is to establish on the ground boundary lines, as called for in the property description of the lease or deed, through a survey to determine the lines of use and occupancy (established by fences or other definite features). Variances between the property description and the lines of use and occupancy are reflected by the written report.
A set of symbols and abbreviations developed through the U.S. Geological Survey . . . are used to identify specifically defined features.
PROPERTY DESCRIPTIONS
An accurate p r o p e r t y description is one of the most important aspects of the conveyance. A defective description is one of the most frequent instances of title failure. In order for title to pass, the property description . . . must describe the parcel of land so it can be identified and located on the ground.
The Rectangular System of Surveys, adopted by the U.S. G o v e r n m e n t , g e n e r a l l y covers all l a n d s w e s t of the Mississippi except Texas, all states north of the Ohio River, and Alabama, Florida, and Mississippi.
[Editor's note: The systems used in the excepted states to describe land include sectionalized descriptions developed by each individual state, metes and bounds descriptions, or the use of lot, block, and subdivision descriptions.]
For example, 17 states, including the original 13 colonies, used a sectionalized basis to describe their land, in which . . . townships six miles square were laid out, but grants within the township varied in size and shape, and do not conform with the U.S. Government's system.
Land grants were made by France, Spain, and Mexico in southwestern states such as Texas, Louisiana, and New Mexico prior to these lands becoming a part of the United States. These descriptions generally cover large blocks of land and are described by metes and bounds. Over 30,000,000 acres of public domain in Texas were granted by the state to railroad companies.... The basic unit of area used in laying out, locating, and surveying these lands was the section, which, for the most part, was a square mile. These sections frequently were assembled into blocks, with a number assigned to each section within the block and a number assigned to the block
5
6
PART 1—LAND AND LEASING
U.S. RECTANGULAR SYSTEM
The rectangular survey system . . . sets up what is known as "initial points/' a position which is determined by accurate field astronomical methods. From these . . . points, lines known as principal meridians and base lines are extended. The principal meridian line runs north and south conforming to the true meridian and extending from the initial point or monument. Regular township corners are established at intervals of 480 chains (6 miles) along this line.
Figure 1 is an example of a U.S. Congressional Township. A standard Congressional Township contains 36 sections with 23,040 acres of land. Normally, sections (nos.) 16 and 36 were . . . set aside . . . for school purposes. Figure 2 is an example of a 640-acre section and how it is subdivided and described in smaller components.
Due to the curvature of the earth, it is necessary to make corrections in the parallels and meridians. . . . This is accomplished . . . usually at intervals [of] 24 miles east and west of the principal meridians. This method produces the rectangular sections of 640 acres each, with the sections along the north and west boundaries of the township absorbing the deficiency or excess in the measurements and convergence.
An important point is that patents from the United States use descriptions based upon the surveys, and they bind all parties . . . as to the b o u n d a r i e s of land conveyed. The government can re-resurvey . . . and change the acreage. The location of the lines m a y n o t be c h a n g e d , even w h e n surveying error is known.
Reading and applying descriptions from the rectangular survey system is made easier by following the whole description backwards.
[Editor's note: For example, the southwest quarter (SW/4) of the northeast q u a r t e r ( N E / 4 ) of Section 24 w o u l d be analyzed by first looking at the N E / 4 (160 acres) then going to the SW/4 within the NE/4, identifying 40 acres. A correct way to describe this acreage would be as follows:
Township 12 North, Range 10 West 14th Principal Meridian (T-12-N, R-IO-W,
14th PM) Section Twenty-four (24): Southwest quarter (SW/4) of the Northeast Quarter (NE/4).]
The system of rectangular surveys makes the proper identification of lands less complicated.... H o w e v e r , . . . the placement of commas to show separate tracts is extremely important. Thus, the N / 2 SW/4 is an 80-acre tract, while the N / 2 , SW/4 comprises 480 acres.
METES AND BOUNDS DESCRIPTIONS
The metes and bounds description involves a method of describing a parcel of land by calls (course) and distances from an established beginning point and following boundary lines, either provided or given with terminal points and angles, back to the point of beginning.
A simplified version of a metes and bounds description would be as follows:
36 31 32 33 34 35 36 31
1
6
5
4
3
2
1
6
12
7
8
9 10 11 12 7
13 18 17 16 15 14 13 18
24 19 20 21 22 23 24 19
25 30 29 28 27 26 25 30
36 31 32 33 34 35 36 31
1
6
5
4
3
(a)
2
1
6
T4N Z <
Q T3N OC Ш T2N
BA SE
T1N
LlhJE
R4W R3W R2W R1W R1E R2E R3E R4E
T1S -< J
O- T2S O Z
OC T3S O-
T4S (b)
Figure 1. (a) Official plat of township sectionized and numbered, with adjoining sections, (b) Official plan of numbering Congressional Townships from meridians and baselines.
Beginning at the . . . Chestnut tree, which is the established northeast corner of t h e . . . Revere Plantation; THENCE N 40 deg E 200' to a stake for corner; THENCE S 30 deg E 250' to a stake for corner; THENCE S 20 deg W 150' to a stake for corner; THENCE N 45 deg W 290' to the place of beginning.
Using a 360 degree compass, shown in Figure 3, . . . the plot of the above metes and bounds property description
160 ACRES
80 ACRES
80 ACRES
Land Description and Maps
7
N
40 ACRES
40 ACRES
40 ACRES
40 ACRES
20
20
ACRES ACRES
20 ACRES
20 ACRES
10
10
ACRES ACRES
10
10
10
ACRES ACRES ACRES
Figure 2. An acre section and its subdivision into smaller components.
Figure 3. A circle divided into 360 degrees, showing typical compass bearings.
use[s] the following procedure:
1. Place the center of the compass at the point of beginning of the line of course, designated as Point A in Figure 4a, with the north and south line of the compass coinciding with the line selected to be the north and south line of the map. [Editor's note: A course is the combination of the direction and length of any particular line, such as "north 40 deg east, 200 ft."]
2. Then lay off 40 degrees from north toward the east and draw a line through this point. With the scale, measure off the first distance, which is 200 feet, and which will be Point B.
3. . . . The process is repeated using the data for the second course... and for the remaining calls and distances, as shown in Figure 4b.
Metes and bounds descriptions are often lengthy and subject to error in the calls and distances. . . . Many use a "deed reference" . . . . An example of a referenced description is as follows:
Being 136.00 acres of land, more or less, a part of the M . . . Russell Survey, A-12,... County, Texas, and being the same lands described in a warranty deed dated October 14,1970, from ... to .. . , recorded in Volume 126 page 583 of the Deed Records of County, Texas, to which deed reference is here made for a more complete description of said lands.
. . . The word describe rather than the word conveyed should be used . . . because the . . . deed may convey . . . an interest different from what the parties intended.
SECTIONALIZED DESCRIPTIONS
Sectionalized descriptions are used mostly in the states that established surveying systems of describing property prior to the establishment of the . . . [U.S.] rectangular survey system. The following is an e x a m p l e . . . :
Being 640.00 acres of land, more or less, Section (10) in Block (2) of the Katy and Brazos Railroad Company Survey, A-63, in ... County, Texas.
LOT AND BLOCK DESCRIPTIONS
Lot and block descriptions usually refer to the lot number, block number, and subdivision plat number under which the subdivision is recorded. . . . An example . . . Lot 26, Block 74, of the t o w n of Red Spring in M i d l a n d C o u n t y , Texas, according to the Plat recorded in Volume 63, page 273, of the Plat Records of Midland County, Texas.
[quoted material ends here]
OFFSHORE DESCRIPTIONS
As for the maps and descriptions used offshore United States, areas subject to state control use systems developed by each affected state that are tied to the onshore land system generally used in that state.
Maps and descriptions affecting federally controlled offshore interests are based on a series of maps designated by the Mineral Management Service (1984) as "Outer Continental Shelf Leasing Maps" and "OCS Official Protraction Diagrams." Each map has an identifier name, such as 'Texas Map 8" or "NH 16-B" and, in most cases, a name such as
8
PART 1—LAND AND LEASING
N
(a)
S
I
I
"Sabine Pass Area" or "Destin Dome Area." Each OCS map is divided into blocks that are generally
three miles square (5760 acres), except those portions in offshore Louisiana lying in water depths less than 200 meters or along state or federal jurisdictional boundaries. The shallow water offshore Louisiana tracts are generally in the shape of a square, containing 5000 acres. Boundary line tracts (between states or federal areas) vary in size and usually contain less than 5000 or 5760 acres, respectively. In some frontier OCS areas, the Mineral Management Service has used larger tracts based on the metric system.
Because water depths are important to the geological and economic evaluations and the potential of offshore leases, the Mineral Management Service also has bathymetric maps available for use by the industry.
Here is an example of a legal description set out in the Mineral Management Service (1984) OCS procedures guide for use in federal waters:
All of Block 397, Eugene Island Area, South Addition, OCS Leasing Map, Louisiana Map No. 4A and all of Block 4, Green Canyon, OCS Official Protraction Diagram, NS15-3.
Another example is
That p o r t i o n of Block 870, OCS Official Protraction Diagram, Mobile N H 16-4, which is more than three geographical miles seaward from the low water line off the coast of Mississippi and/or Alabama.
(b)
Figure 4. (a) Example showing the procedure for plotting the metes and bounds property description, (b) The completed plot of the metes and bounds property description.
Determining Owners of Oil and Gas Interests, and Methods of Conveyance
James c Tinkler
^SST
INTRODUCTION
Ownership of oil and gas interests, whether privately or publicly o w n e d , is reflected in the public records of the various states or, in the case of interests owned by the federal government, in the offices of the various federal agencies, principally within the Department of the Interior.
Fundamental to determining ownership is the necessity of understanding the rudiments of real property principles and laws because in most states, oil and gas interests are classified as real property, and if not, they are treated in a manner that achieves the same or a similar result.
PRINCIPLES AND TERMS OF PROPERTY LAW AFFECTING OIL AND GAS INTERESTS
A person who owns land in fee simple owns, among all real property rights, all of the oil and gas interests underlying the land. Lesser interests in real property can be carved out of the fee simple estate on a undivided basis (e.g., 1 / 2 or 1 /4). Also, a party can dispose of their rights to the surface (surface fee), to all or part of the minerals (mineral fee) underlying the land, to the oil and gas underlying the land, and to the royalties from oil a n d / o r gas production. The oil a n d / o r gas production payments and/or the interests can be pledged as collateral to loans or mortgages or may be subject to Hens.
By the same token, all of these types of interest may be limited as to term, horizon, purpose, or amount of money. The most prevalent example of a term oil and gas interest is an oil and gas lease, which creates in the lessee a leasehold estate commonly referred to in the oil and gas industry as a working or operating interest. The rights granted under an oil and gas lease to a lessee may vary from lease to lease. An overriding royalty is generally a cost-free interest carved out of the lessee's leasehold estate.
Among other types of term interests are life estates for the life of a persons (life tenants) and remainder estates for future interests maturing to the remainderman upon the death of the life tenant. Generally, both the life tenant and remainderman must execute an oil and gas lease for it to be effective since both interests are affected.
The possibility of reverter is another type of contingent future interest. This occurs when a party (who, with its heirs, successors, or assigns, retain the possibility of reverter) transfers an interest in land to a party upon the condition that the land be used for a specified purpose, such as a road, park, or church. If the specified use is discontinued, then the land reverts and is returned to the those holding the rights of the transferer.
CONVEYANCE OR TRANSFER OF INTERESTS IN LAND
An owner of property may generally dispose of any or all of the rights of ownership by sale, gift, devise, or descent. Transfers of property may also be the result of judicial decree or of equitable interests created by contracts.
1. Sales or gifts may be concluded by documents such as the following: a. Deeds (e.g., warranty, quit claim, easement, permit, foreclosure, tax, mineral, and royalty deeds) b. Leases (e.g., oil and gas, coal, uranium, and surface leases) c. Assignments or partial assignments of leases, overriding royalty, production payments, and so on. d. Equitable interests created by contracts such as seismic options, joint venture agreements, joint operating agreements, farm-outs, pooling or unit agreements, and partnership agreements
All of the documents must be in writing to be vaUd against the claims of innocent purchasers for value. Generally the law requires that they be registered or recorded in the public records of the county or parish where the land lies. 2. Transfers by devise or descent are affected by the laws relating to inheritance of property interests from deceased persons. Such transfers can be accomplished by a will or, in the absence of a will, by laws of descent and distribution of the various states. A party who dies without leaving a will is said to have died intestate, and the beneficiaries or heirs of the deceased person may have the estate administered in the appropriate judicial forum. A person who dies leaving a will is said to have died testate, and the will is probated in the appropriate judicial forum. In some cases, wills are not probated and heirs do not seek court sanctioned administration. In those cases, most states recognize the transfer of interests, but some form of proof of heirship may be necessary. This proof can be offered in several different ways, such as by filing the unprobated will in the pubUc records or by obtaining affidavits of death and heirship from people who knew the deceased person's family history.
3. Transfers by judicial decree may occur as the result of court action involving such things as divorce, foreclosures, contested wills or heirship, partition, taxation, levies on property for collection of money judgments, adverse possession, title disputes, and eminent domain.
9
10
PART 1—LAND AND LEASING
ASCERTAINING OWNERSHIP OF PROPERTY INTERESTS
Private and State Lands
Each state has a registration system, usually on a county basis, open to the public whose principal functions include (1) safekeeping of documents and (2) providing a search system that allows the general public to ascertain the current owners property rights (and other ancillary rights), along with the recorded history of all transactions affecting a given tract of land. These recording offices usually have such names as County Recorder, County Clerk, County Registrar, or Parish Clerk. In addition, since judicial decrees may reflect property rights not revealed in the county records, searches in the official clerical records of the courts (both at the state and f e d e r a l level c o u r t s of original jurisdiction) m a y be appropriate.
The various systems may be indexed in several different ways. For example, alphabetical listings of those giving u p rights in property (e.g., grantors, lessors, assignors, mortgagors, and trustors) and of those receiving property rights (e.g., grantees, lessees, assignees, mortgagees, and trustees) may be keyed to the property descriptions involved in the transactions.
Some states (mostly those using the rectangular system) use a plat or property description system in which each document is entered in chronological order on a page or pages specifically devoted to a given tract of land. Many recording offices use combinations of these two systems. Separate index books or systems (many of which have been computerized in the more active counties) may be set up by category, for example, transactions affecting real property, personal property, mineral interests, oil and gas interests, mortgages and liens, or judicial actions.
In some cases, the search of the recording offices may be aided by searches of other public offices such as the tax assessor or collector whose records may reflect the owners or agents of the owners whose property interests are subject to ad valorem taxes. But note that in many states, owners of minerals are not taxed, nor does the taxpayer necessarily own all of the interests in the affected lands.
In addition to the use of the public records to ascertain ownership, commercial entities, primarily abstract companies, will make title searches identifying the current owners or will prepare abstracts of title that include copies of all the recorded documents affecting the particular property described for the period of time so requested. These commercial abstract companies have created index systems from the public records that generally simplify the search for ownership.
Commercial land ownership maps are also available to aid in the identification of owners of property interests.
Federal Lands
The responsibility for management of most federal lands falls within the purview of the U.S. Department of the Interior and its subordinate agencies. The Mineral Management Service (MMS) has the responsibility of m a n a g i n g and maintaining the records of ownership on lands and minerals underlying the federal portion of the outer continental shelf (OCS) of the Atlantic and Pacific oceans, the Gulf of Mexico,
Table 1. Costs and Revenue Interests in Example 1
Oily Co., lessee
Costs (%) 100.00000
Revenue (%) 87.50000
I.S. Lucky, lessor #1
0
(3/4 of 1/8 royalty)
9.37500
I.S. Rich, lessor #2 (1/4 of 1/8 royalty)
TOTALS
0 100.00000
3.12500 100.00000
and the various seas adjacent to Alaska and Hawaii. The Bureau of Land Management (BLM) has a similar
responsibility for most federal onshore land and minerals. The U.S. Forest Service, Bureau of Indian Affairs, the Defense Department, and other federal agency departments may have overlapping or separate responsibilities with reference to ownership of lands within their jurisdiction.
Just as in the case of private lands, commercial entities such as abstract companies, map companies, and computer statistical reporting companies may have information available ascertaining ownership.
DETERMINING SPECIFIC OWNERSHIP COST AND REVENUE INTERESTS
Since the discovery of America by Columbus, literally billions of transactions have occurred affecting the ownership of property rights. As a result, the norm is to find that a given tract may have multiple owners of varying fractional interests in various property rights. In the case of oil and gas interests, varied ownership of the "working interest" (a cost-bearing interest) in a well or lease is common, as is that of the revenue interest owners, which are those who own non-cost bearing interests such as royalty, overriding royalty, net profits, production payments, and other working interest shares of revenue.
Because all calculations of ownership involve the use of the n u m b e r of acres o w n e d by a p a r t y , u n d e r s t a n d i n g the meaning of the terms gross acres and net acres is essential. Gross acres refers to the number of surface acres lying within the b o u n d a r i e s of a g i v e n tract of l a n d . A net acre is determined by multiplying the fractional interest owned by a party in the given tract of land times the number of gross acres in the tract. For example, if John Doe o w n s a 1 / 2 interest in a tract of land consisting of 100 surface acres, Mr. Doe would own 100 gross and 50 net acres. If a lessee called XYZ Co. bought an oil and gas lease from Mr. Doe on that tract, the company would have acquired a lease covering 100 gross and 50 net acres.
The following examples demonstrate the calculation of cost and revenue interests under various circumstances.
Example 1
Assume the Oily Co., lessee, obtains oil and gas leases from I. M. Lucky, lessor, who owns a 3 / 4 interest in a 20-acre
Determining Owners of Oil and Gas Interests, and Methods of Conveyance
11
Table 2. Costs and Revenue Interests of Example 2
Gassy Co. 20-ac tract (20/320)3 300-ac tract (300/3203 x 5/8)
GASSY CO. TOTALS
OPM Co. 300-ac tract (300/3203 x 3/8) Oily Co. ORII 20-ac tract I.M. Lucky, lessor, 20-ac tract I.S. Rich, lessor, 20-ac tract B. A. Winner, lessor, 5/8 of 300-ac tract Gotta Fortune, lessor, 3/8 of 300-ac tract Fast Buck, 1/2 of Fortune's royalty
GRAND TOTALS FOR UNIT
Costs (%)
= 06.25000 = 58.59375 = 64.84375
= 35.15625 0 0 0 0 0 0
100.00000
Multiplication Factors (%)
x [100 - (12.5 + 5)]b x (100 - 20)b
x (100 - 20)b 6.253 x 0.05 6.253 x 3/4 x 1/8 6.253 x 1/4 x 1/8 93.753 x 1/5 x 5/8 93.753 x 1/5 x 1/2x3/8 93.753 x 1/5 x 1/2 x 3/8
aEach tract has a unit or tract participation factor equal to the number of acres in each tract within a unit divided by the total acres within the unit. b100% less lease burdens plus royalties and overriding royalty.
Revenue (%)
= 05.1562500 = 46.8750000
52.0312500
= 28.1250000 = 00.3125000 = 00.5859375 = 00.1953125 = 11.7187500 = 03.5156250 = 03.5156250
100.0000000
tract of land, and from I. S. Rich, lessor, who owns the other 1 / 4 interest in the tract. Mr. Lucky and Mr. Rich each reserve a 1/8 royalty (cost-free) interest. Oily Co. drills and completes an oil well at a depth above 3000 ft on the leased 20-acre tract. The costs and revenue interests are shown in Table 1.
Example 2
Assume the Oily Co. sold the deep rights to the leases in the 20-acre tract to the Gassy Co. with the Oily Co. retaining a 5% overriding royalty interest (ORRI). The Gassy Co. obtained an oil and gas lease covering a 5 / 8 interest from B. A. Winner, lessor, who retained a 1 / 5 royalty on the 300-acre tract adjoining the 20-acre tract. The OPM Co. owned a lease obtained from Gotta Fortune, lessor, on the other 3/8 interest in the 300-acre tract. The Fortune lease provided for a 1/5 royalty,but Gotta had sold 1 / 2 of her royalty to Fast Buck. All four leases authorized pooling, which allowed the Gassy Co. and the OPM Co. to pool their interests in the 20-acre leases with the 300-acre leases, creating a 320-acre unit on which the companies successfully drilled and completed a gas well on the unit. The costs and revenue interests of the parties in this unit are shown in Table 2.
The period of time covered by title examination may vary depending upon circumstances (previous opinions may be available) and the title risks the examiner's client is willing to assume. Many examiner's opinions commence with "sovereignty of the soil," the date that the property rights were severed or conveyed by a sovereign government to a private party. This is necessary because all private property rights in the United States exist only if a proper grant or patent is obtained from a sovereign government (English, Spanish, French, Mexican, and Republic of Texas grants made prior to U.S. sovereignty are recognized).
When the title examiner discovers defects of title, they are noted in the opinion and title requirements, or instructions on how to cure these defects, are specified (called title curative). The examiner may also make comments and advisory statements in the opinion to further explain the rights, obligations, and liabilities that may have been known, but not thoroughly understood, by the examiner's client.
Title Opinions
In the oil and gas industry, the types of title opinions obtained on a tract of land examined are generally classified as follows.
TITLE OPINIONS AND TITLE CURATIVE
When a party acquires real property interests of value, they usually obtain a title opinion from a title examining attorney. The p u r p o s e of the opinion is to obtain a title experts' advice as to whether the interest acquired or to be acquired is as represented and whether it is free of title defects or claims of others that could reduce their interests, impair their rights, or create liabilities an acquiring party is unwilling to assume. The title examiner's opinions are based upon examination of the documents filed in the public records affecting the particular property interests under examination and the examiner's knowledge of the legal implications of those transactions.
Original Title Opinions
Original title opinions are the first opinions obtained. They may be limited to opinions on rights, obligations, and liabilities acquired pursuant to an oil and gas lease—as o p p o s e d to other opinions r e n d e r e d for the p u r p o s e of reflecting upon the rights of the ownership of the surface, non-oil and gas mineral ownership, entitlements to production revenue or royalties, easements, and so on.
Supplemental Title Opinions
These opinions are the second and subsequent opinions to the original or drilling opinions. They are obtained to (1) u p d a t e the period of time covered since rendition of the original opinion, (2) reflect upon the effects of title curative
12
PART 1—LAND AND LEASING
submitted to meet opinion requirements, or (3) both.
Drilling Opinions
Drilling opinions may be original opinions, but they are expressly rendered for the p u r p o s e of ensuring that the property rights essential to drilling and ownership of any of the oil and gas obtained from a well are vested in the drilling parties. These opinions will cover at least the drillsite tract.
Division Order Opinions
These opinions are rendered, after production of oil and gas has been obtained, for the purpose of determining who is entitled to production and/or revenue obtained from a given well or unit.
Special Opinions
D u r i n g the life of an oil and gas lease, special circumstances may arise in which a client lessee may need legal advice about the payment of delay rentals, shut-in gas payments, transfers of lessor's interests, and so on. The most commonly obtained opinions of this type are the rental and shut-in gas royalty opinions.
Title Curative
In the oil and gas industry, the amount of title curative effort is substantial and varied. The most common title curative documents requested by examiners include the
following: • Land surveys by civil engineers to delineate lease and unit lines • Affidavits, which are sworn statements of facts by knowledgeable and credible persons as to heirship, death, possession or nonpossession of land, development or nondevelopment of oil and gas rights, identities of persons, and so on • Ratifications, or confirmation of previous actions • Tax certificates, which provide evidence of payment of taxes • Correction instruments to correct mistakes in deeds, leases, and so on • Subordinations, which are usually obtained by a lessee from a lessor's mortgagee or lien holder so that the lessee's rights are not subject to foreclosure in the event the lessor defaults on the loans • Judicial action, which may involve guardianship, disputed heirships, adverse possession, fraud, declaratory judgments, probating wills, foreclosures, and title disputes arising from a multitude of sources • Regulatory activity, which involves conforming to the rules, regulations, and requirements of local, state, and federal regulatory bodies • Miscellaneous requirements, including obtaining new leases, obtaining consents to pooling or unitization, amending contracts, and providing corporate resolutions reflecting the authority of persons to act on behalf of the corporation
Nature of the Oil and Gas Lease
James C. Tinkler
University of Houston-Downtown Houston, Texas, U.S.A.
GENERAL TERMS
An oil and gas lease is an instrument executed by the o w n e r or h o l d e r , the lessor, of the r i g h t s of execution (executive rights) who conveys leasehold rights as to oil and gas and such other mineral interests as are to be included in the property described, to a lessee, for a fixed or determinable period of time. In some cases, top leases are obtained on lands already leased to others, with the top lease only becoming effective, at the option of the lessee, u p o n the termination of the preexisting lease.
Generally, the typical private oil and gas lease provides for the lessee to obtain the rights incidental to exploration, drilling, developing, producing, and disposing of the oil, gas, and associated hydrocarbons underlying the leased premises. In consideration for a lease, the lessee usually pays the lessor a negotiated or competitive signature bonus at the time the lease is executed. In some cases, the lessee, in Heu of or in addition to the bonus, may agree to perform certain services, such as drilling and seismic and/or geological surveys. Also, the lessee provides the lessor with a cost-free royalty interest that is at least 1 / 8 of the production.
The terms of an oil and gas lease spelHng out the rights, duties, and obligations of the parties may vary substantially depending upon the marketplace, the negotiating abiHties of the parties, and/or the statutory terms mandated on public lands by governmental agencies. Generally, most oil and gas leases contain the foHowing provisions and topics.
Date and Parties to the Lease
The date shown on the lease is the commencement date of the lease term. Some states require that the addresses of parties be shown on the face of the lease.
Consideration
Nominal consideration is nearly always specified. A few states require the actual consideration to be shown.
Granting Clause
This clause contains words of grant and sets out the rights the lessee is acquiring under the lease.
Description Clause
The description clause describes the land covered by the lease.
Mother Hubbard Clause
This clause is intended to include within the lease description small contiguous tracts of land owned by the lessee, which due to minor discrepancies, may have been omitted or not covered by the lease description.
Estimate of Acreage Clause
Because delay rental payments are based on the amount of acreage in a given tract and because at the time a lease is executed an exact amount is not always determinable, this clause allows the rentals to be based on the estimated amount.
Habendum (or Term) Clause
This clause specifies the primary term of the lease and usually provides for an extension of the primary term in the event that production in paying quantities is had from the lease. The Habendum clause is usually affected by other clauses in the lease, such as the pooling, rental, dry hole and cessation of production, nonforfeiture, and force majeure clauses, as weH as other special clauses (e.g., Pugh clause and continuous drilling clauses).
Royalty Clause
The Royalty clause generally provides for several things. First, it provides for the lessor to receive from the lessee a cost-free payment for its fractional share of any oil and gas production obtained from the lease, valued at the wellhead. Second, it allows the lessor the right to take its share of production in-kind. Third, it provides for the lessee to maintain the lease in force during periods when gas wells must be shut-in for lack of a market, by making specified payments periodically to the royalty owners (called shut-in gas royalty). Fourth, it gives the lessee free use of the oil, gas, and water (except from the wells of others) in its operations on the lease. FinaHy, it allows the lessor to receive royalty for other minerals such as sulfur.
Pooling Clause
This clause provides for pooling all or part of the leased acreage with other acreage for the purpose of drilling a well on the pooled or unitized area of a specified size (and in some cases, as limited in shape and horizons) in the interest of conservation and prevention of waste, as permitted by the appropriate governmental regulatory agency. The effect of this provision is to eliminate the drilling of unnecessary wells and to perpetuate the lease by allowing off-the-lease-premise unit well(s) to maintain the lease in force. Royalty is paid pro rata on an acreage basis (that is, the n u m b e r of acres contributed by the lease divided by the number of acres in the pooled unit).
Delay Rental Clause
This clause sets out the following conditions: (1) an amount of money that may be paid periodically (usually annually) to the lessor by the lessee to defer drilling operations during the period; (2) a depository bank for deHvery of the payments; (3) the methods and delivery of the
13
14
PART 1—LAND AND LEASING
p a y m e n t ; and (4) the m e t h o d s for h a n d l i n g changes in ownership and designating a different depository bank. It also ensures that the shut-in gas royalty payment procedures conform to those provided for delay rental payment and that the lessee may release a portion of the lease and reduce its rentals and obligations accordingly. Unless operations or production of the type provided for in the lease occur to relieve the rental payment obligation, the nonpayment of rentals will cause the lease to terminate. Note that in some cases, usually where small tracts or interests are leased, the lessee purchases a paid-up lease, avoiding the administrative need for rental payments.
Dry Hole Clause and Cessation of Production Clause
These clauses generally set out three conditions. First, in the absence of production or after drilling a dry hole, they establish what type of activity (drilling, reworking, or rental payment resumption) by the lessee must occur to maintain a lease in force during and after the end of the primary term of the lease. Second, they state when and how rental payments may be resumed during the primary term. Third, they set out the obligations of the lessee to protect the lease from drainage by offset wells.
owns less than a full interest in the lease, rental and/or royalty payments may be reduced accordingly.
Force Majeure Clause
This clause protects the lessee from losing the lease or from incurring damages when natural catastrophes or other named significant events occur that are beyond the control of the lessee.
Other Special or Unusual Clauses
One typical special clause limits the period of time and the a m o u n t of leased acreage the lessee m a y k e e p a r o u n d producing wells or units after the end of the primary term, which in some jurisdictions are called Pugh or Freestone rider clauses. Another example of a special clause is one that defines the costs that can be charged and the calculations to be used in determining the royalty owner's share of revenue from plant processing production for value-enhanced products, called plant product clauses. Another special clause limits the lessee's access to the lease d u r i n g p e r i o d s of environmental sensitivity, and yet another specifies well or work obligations.
Environmental and Property Obligation Clause
This clause sets the obligations of the lessee to (1) remove its property from the leased lands in a timely manner; (2) restore the surface to its original condition, also in a timely manner; (3) bury its pipelines to a specified depth; (4) restrict the location of wells and use of the surface; (5) pay damages for destruction of crops, trees, animals, or structures; and finally, (6) take care of environmental obligations, if any.
Assignment, Changes in Ownership, and Lease Divisibility Clause
This clause sets out the rights and obligations of the parties as to the assignability of their interests in the lease, and in the event of changes in ownership, it specifies the impact of the changes u p o n the obligations of the parties and how the changes are to be administered.
Nonforfeiture Clause or Breach of Obligations and Development Clause
This clause usually provides (1) that the lessor, within a specified time period, must notify the lessee of any breaches by the lessee of lease obligations, allowing the lessee time to remedy the breach, and (2) that the lease must be developed in a reasonably prudent manner in accordance with that permitted by regulatory authority as to acreage spacing for oil and gas wells. Alternatively, the clause may provide for a specific spacing requirement.
Warranty of Title Clause
The warranty of title clause generally provides (1) that the lessor warrants and agrees to defend title to the leased interests; (2) that the lessee may pay taxes or the liens owed by the lessor to prevent foreclosure of the lessee's rights (and the lessee, if production is established, may recover such costs from the lessor's share); and (3) that in the event the lessor
PRIVATE VERSUS PUBLIC SECTOR LEASING
Private sector oil and gas leasing usually requires intensive title searches on relatively small tracts and negotiation of sometimes complex terms with multiple owners of interests underlying the same tract of land.
Public sector leasing generally involves relatively large tracts of land leased f r o m a single governmental agency whose lease terms are not negotiable since they are fixed by law. The consideration paid is usually determined by competitive bidding. Operationally, governmental oversight on public sector leases is considerably more involved than those on private leases.
Leasing of State Lands
The statutory terms of state leases vary from state to state and with time as market conditions change. For specific information about each state's leasing terms and requirements, one should contact the state leasing agent. Table 1 lists the most active state leasing agents.
Leasing of Federal Onshore Lands
The Oil and Gas Leasing Reform Act of 1987, and the regulations issued in 1988 to implement the Act, materially changed requirements for the leasing of federal onshore lands. The changes implemented require that any federal land offered for leasing by any federal agency must first be offered for competitive, oral bonus bidding at sales to be held at least quarterly by the Bureau of Land Management state office. The statutory minimum bid is $2.00 per acre, and the royalty must be no less than 12.5%. Rental rates must be no less than $1.50 per acre for the first 5 years and $2.00 thereafter, and the maximum term is 10 years. Lands that receive no bids at an offering for a period of two years may be leased to the first qualified applicant on the same terms
Table 1. State Leasing Agents for State Lands
State
Alaska Arizona Arkansas California Colorado Idaho Illinois Indiana Kansas Louisiana Michigan Mississippi Montana Nebraska New Mexico New York North Dakota Ohio Oklahoma Pennsylvania South Dakota Texas Utah West Virginia Wyoming
Source: American Association of Petroleum Landmen (1990).
Nature of the Oil and Gas Lease
15
Leasing Agent
Commissioner of Public Lands State Land Commissioner Department of Commerce State Land Commissioner Board of Land Commissioner Board of Land Commissioner Department of Mines and Minerals Oil and Gas Division, Department of Conservation Department of Interior State Mineral Board Department of Mineral Resources Mineral Lease Commission Board of Land Commissioner Board of Education, Lands and Funds Commissioner of Public Lands Bureau of Surplus Real Property State Land Department Department of Natural Resources Commissioner of Land Office Department of Forests (and others) Commissioner of School and Public Lands General Land Office State Land Board Department of Natural Resources State Land Commissioner
specified for competitive leasing. This system is called the "over-the-counter" noncompetitive leasing method.
The Reform Act did away with the old noncompetitive "simultaneous filing" or "lottery" system and the competitive KGS (known geological structure) lease sales.
Leasing of Federal Offshore Lands (Outer Continental Shelf)
The procedures for the leasing of federal offshore lands are the responsibility of the Mineral M a n a g e m e n t Service, Department of the Interior (see Mineral Management Service, 1984). These procedures are generally the same for each of the MMS designated regional areas, including the Gulf of Mexico, Alaska, Pacific, and Atlantic regions.
Federal OCS leases are generally granted on the basis of sealed competitive bonus bids calling for at least 1/6 royalties, $3.00 per acre rentals, and a 5-year term (10 years for deep water tracts). Sales are offered periodically in accordance with the Secretary of Interior's Tentative Milestones for 5-Year Offshore Leasing Schedule, which are revised from time to time. It takes about three years from the time the MMS makes a "call for information" (that is, it seeks information on the
desirability of a sale for a given area) until an actual lease sale occurs (Gossett, 1984).
Indian and Alaskan Native Claims Lands
The p r o c e d u r e s for the leasing of I n d i a n l a n d s are g e n e r a l l y p r o v i d e d for in Title 25, C o d e of Federal Regulations, and involve either oral or sealed bidding and/or negotiation and approval of the lease terms by the affected Indian tribe (in the case of unallotted or tribal lands) or by the individual Indian (in the case of allotted lands). In both cases, the Bureau of Indian Affairs, Department of the Interior, supervises and must approve the lease terms.
As to the leasing of Alaskan Native lands, approximately 39.2 million acres may be leased from 12 different regional native corporations. An additional 4 million acres may be leased from the various village corporations. These subsurface rights were obtained by the Indian corporations (created under state laws) pursuant to the Alaskan Native Claims Settlement Act of 1971. As of 1989, only about 60% of the regional corporations lands have been conveyed to them by the United States. Approximately 80% of the village lands have been conveyed.
Oil and Gas Contracts
James C. Tinkler
University of Houston-Downtown Houston, Texas, U.S.A.
INTRODUCTION
Because of the diversity of o w n e r s h i p of oil a n d gas interests a n d / o r the need to share economic risks, the oil and gas industry has utilized a number of different contractual arrangements. The most common types of contracts used are farm-outs-farm-ins, or well trade agreements, and joint operating agreements.
FARM-OUTS AND FARM-INS (WELL TRADES)
When the owner (farmor) of an oil and gas working interest agrees to assign an interest in a lease (called the farm-out area) to another p a r t y (farmee) in consideration of the f a r m e e drilling a well or wells (farm-out wells) on the farm-out area, the farmor is said to have made a farm-out and the farmee has made a farm-in. Sometimes the farmee may be required to do more than drill a well, including performing geological and seismic studies or paying a cash consideration for past costs incurred by the farmor.
These farm-out agreements are usually accomplished in a nonrecordable form of letter agreement that typically contains provisions relating to the following:
The marketing of farm-outs and their negotiation and preparation require many skills. The terms of a farm-out deal vary with the market conditions of the times. Promotion is an art in itself. It involves allowing the farmor to receive more than what costs would have been if 100% of risks associated with the farmor's eventual interest after farm-out had been paid for by the farmor.
During recent years, the term "third for a quarter" has been the basis for promotion of many farm-out deals. In these deals, the farmor attempts to recover all or as much of its past costs as the market will bear, along with the costs of the drilling of a well (to casing point, to dry hole, or through production facilities), reserving for the farmor as a back-in a percentage of the working interest (25% in "third for a quarter" deals) after the farmee has recovered the costs of the promotion (called after payout). For example, if a farmor owned 100% of the farm-out area and had land, geological and seismic studies, and estimated dry hole farm-out well costs of $300,000, the farmee, using a ratio of 3 to 4, would pay 100% of those costs for a 75% interest. A party paying 1 / 3 of the costs on the same promoted basis would pay $100,000 for a 25% working interest.
JOINT OPERATING AGREEMENTS
1. Names of the parties and the effective date of the agreement
2. Description of the leases and lands to be farmed-out 3. The location, well objective depth, commencement date,
and geological requirements of the farm-out wells 4. Substitute or lost hole well provisions in the event the
initial farm-out wells are lost because of drilling problems 5. Earning requirements of the various possibilities, such as a dry hole, a producer, a producer at any depth, more than one well, continuous drilling, and so on 6. Rights retained by the farmor, including working interest, overriding royalty, net profits, or combinations of these interests 7. Tender of wells or leases before abandonment and / o r surrender 8. Negotiation and setting out of the terms of the joint operating agreement if the farmor retains a right to a working interest 9. Obligations for rental a n d / o r royalty payments in the event of production 10. Liabilities of the parties and insurance provisions 11. Obligations to tender interests to the farmor if the farmee obtains extensions or renewals of farm-out leases 12. General clauses related to notices and information furnished to the parties, audits, marketing of production, access to the wells, and gas processing
Anytime two or more owners of working interests decide to share the risk of drilling, development, or operations related to the production of oil and gas, they enter into what the industry calls a joint operating agreement (JOA) or, simply, an operating agreement. The JOA generally provides for one of the parties to act as the operator for the parties on the joint area covered by the JOA. It also specifies the operation for which the JOA was formed (the drilling of a well) and how costs and revenues will be shared, determined, and accounted for. In addition, it provides for each party's rights to the production obtained and sets out how leases will be acquired, maintained, transferred, and disposed.
Most JOAs are predicated on the basis that the operator will not profit from its management of the joint interests. Except in an emergency, it must obtain authorization from the other parties (the nonoperators) to spend money for the joint account. Also, except in certain limited circumstances, no party may prevent another party from proceeding with operations that it desires to undertake at its own cost, risk, and expense. In these cases, if less than all the parties to the JOA proceed with a project on their own and in the event production is obtained from these sole expense or sole account operations, the consenting parties who took the risk for the project are allowed to recover from the nonconsenting party's share of production 100% of the costs incurred on behalf of the n o n c o n s e n t i n g p a r t y plus a s u b s t a n t i a l a d d i t i o n a l percentage, usually several hundred percentage points depending upon the risks of the project. The percentage is
16
Oil and Gas Contracts
17
higher for exploratory wells than for development wells. Additional subjects covered by a JOA include the
following:
1. Handling of title examination and the effect of loss or failure of title upon a party's interest
2. Designation, resignation, and removal of an operator 3. An operator's rights, duties, and liabilities 4. Providing for the initial project, usually a test well's
objective depth, commencement date, location, and abandonment procedures 5. Expenditvires and liabilities of the parties, including liens and payment defaults; payment and accounting requirements; limitations on expenditures to drill, deepen, rework, and plug back; and other operations 6. Handling of rentals, shut-in payments, and minimum royalties 7. Taxes 8. Insurance 9. Internal Revenue Service elections 10. Claims and law suits against the parties 11. T e r m o f t h e J O A 12. Acquisition, maintenance, or transfer of interests 13. Other provisions, such as notices, force majeure, designation of areas of mutual interest, taking of production, gas balancing, preferential rights to purchase interests offered for sale by any party to the JOA, and compliance with laws and regulations
One of the more important parts of a JOA is the accounting schedule, which usually appears as an exhibit to and becomes a part of the JOA. This exhibit consists of five or six pages of fine print in a form developed by the Council of Petroleum Accountants Society, hence, it is called the COPAS form. The form, which is revised periodically, spells out the specific accounting methods that the operator must use to account for expenses and revenues.
Onshore JOAs used today stem from work done by the American Association of Petroleum Landmen (AAPL) to create a standard form to simplify and facilitate the negotiation of JOAs with equitable results for all the parties concerned. Revision of AAPL Form 610 w a s last accomplished in 1989. Offshore JOAs in present use vary from party to party, but are similar in format to the onshore JOA. The American Petroleum Institute, who first created a model form Offshore Operating Agreement in December 1984, is presently attempting to standardize the offshore JOA.
The principal differences between the onshore and offshore agreements are in the areas related to penalties (which are higher offshore than onshore because of the cost and risk) for nonconsent operations and to the number of decision points for consent or nonconsent on future high cost operations. In addition, many nomenclature changes are needed to reflect the different operational activities occasioned by an ocean e n v i r o n m e n t . Also, because of intense federal and state regulation, other factors complicate the offshore agreements, such as environmental control, compliance with federally mandated nondiscriminatory practices, and the different provisions needed to handle potential catastrophes affecting insurance and liability protection.
OTHER AGREEMENTS
There are a variety of other special agreements used in oil and gas exploration and development activities.
Well Support Agreements
The three types of well support agreements are dry hole contribution, bottom hole contribution, and acreage contribution.
1. A dry hole contribution is used by drilling parties to obtain money contributions from parties whose working interest leases located near the well to be supported will benefit from the drilling results. Dry hole contributions are paid (usually an agreed upon amount based on footage drilled) only in the event that the drilling results in a dry hole drilled to the depth specified. The party paying the contribution is entitled to all of the well data.
2. A bottom hole contribution is similar to a dry hole contribution except that the agreed upon money contribution is paid whether the well is completed as a producer or abandoned as a dry hole.
3. An acreage contribution is similar to a dry or bottom hole contribution except that the nondrilling party agrees to contribute all or part of the leases located near the support well rather than money.
Joint Exploration and Development Agreements
These agreements or ventures arise from situations in which two or more parties pool their divided or undivided interests to share the costs and risks of either exploration or development or both. Typically, geological, seismic, and/or petroleum engineering studies, surveys, or evaluations are requisites to the agreements. Also, the typical venture involves large areas of mutual interest involving potential future lease acquisitions. Some of the participants may pay a disproportionate share of the costs of the venture for a chance to participate. These transactions may be very complex.
Bidding Agreements
Bidding agreements commonly involve frontier or offshore areas where unleased public sector oil and gas interests will possibly become desirable to a group of companies who may wish to share the high bid costs and bid as a group. The group may have been formed as a result of joint exploration a n d / o r development activities, or it may simply be a case of where a financial party desires to bid with a more knowledgeable industry partner or venturer. These agreements may be extremely complex as to methodology in determining what to bid, with whom, and at what time, as well as in the preparation process for a competitive lease sale. The formulas for participation after a sale may also be complex. Federal and state antitrust laws and other laws pertaining to penalties for collusion further complicate the processes.
18
PART 1—LAND AND LEASING
Purchase or Acquisition Agreements
Purchase agreements arise when two or more parties agree to share in the f u t u r e p u r c h a s e of either exploratory or producing oil and gas interests. These agreements usually spell out the subject matter to be considered for purchase; the interests of the parties; how prepurchase and after purchase costs, if different, will be borne; how revenues will be shared if one or more of the parties is entitled to a disproportionate share; and all of the operating provisions to be invoked upon purchase of the interests.
Seismic Option Agreements
Seismic option agreements result from a party obtaining the right to purchase oil and gas interests, conditioned upon the results of a new seismic survey a n d / o r evaluation of existing seismic. Sometimes a cash consideration must be paid for the option.
Lease Exchange Agreements
Lease exchange agreements involve situations in which two or more parties exchange rights and interests in an oil and gas lease in one geographic area for rights and interests in another area.
Partl References Cited
American Association of Petroleum Landmen, 1984, AAPL Comprehensive Land Practices, Chap. I—XII.
American Association of Petroleum Landmen, 1990, AAPL Directory and Guidebook: Fort Worth, TX. AAPL.
Burke,}., 1983, Petroleum Lands and Leasing: Tulsa, OK, Pen Well Books, 176 p.
Gossett, R. A., 1984, Offshore leasing: American Association of Petroleum Landmen Comprehensive Land Practices, Chap. XII, p. 35-40.
Independent Producers of Association of America, 1991, The oil and gas producing in your state, 1990-91: IPAA, Sept., 120 p.
References Cited
19
Kimball, W. C., 1982, Fee Land Determination (Record Checker's Bible): Midland, TX, Midland Enterprises, 65 p.
Mineral Management Service, 1984, Oil and Gas Leasing Procedures Guidelines, Gulf of Mexico Region: MMS, Department of the Interior, 188 p.
Mineral Management Service, 1989, Bulletin No. 381: MMS, Department of the Interior, 20 p.
Rowald, H. E., 1984, Land maps and property descriptions: American Association of Landmen Comprehensive Land Practices, Chap. II, p. 1-13.
Stamper, F. A., 1973, A Handbook of Texas Abstractors and Title Men: Austin, TX, Texas Land Title Assoc., p. 1-321.
Part 2
ECONOMICS AND RISK A. Sn Sn ESn Sn M_ ENTT_
edited by
Peter R Rose
Telegraph Exploration, Inc. Austin, Texas, U.S.A.
Contents
• Introduction • Fundamental Economic Equations for Oil and Gas
Property Evaluation • Uncertainties Impacting Reserves, Revenue, and
Costs • Expected Value a n d Chance of Success • TheTimeValueofMoney • Building a Cash Flow Model • AboutTaxes • Key Economic Parameters • Dealing with Risk Aversion • Economics of Property Acquisitions • ReferencesCited
Introduction
Part 2 comprises nine main topics, arranged to lead the reader logically along a chain of linked considerations of economics and risk assessment. We start with the prediction of geological reserves and chance of success and continue t h r o u g h the construction of a cash flow model of the anticipated producing property that considers the time value of money and petroleum taxes. Next we proceed to an assessment of various economic yardsticks used to measure the economic potential of projects, then on to a discussion of risk aversion. We conclude with the economics of property acquisitions.
Most chapters in this part of the Manual show a single author indicating primary responsibility; however, both authors Rose and Thompson contributed to each of the nine chapters. The chapters written by Rose were reviewed by E. S. Capen, Consultant, Dallas, Texas, and R. E. Megill, Consultant (Retired), Kingwood, Texas. Thompson's chapters were reviewed by J. D. Wright, Questa Engineering Corp., Golden, Colorado.
Peter R. Rose
Telegraph Exploration, Inc. Austin, Texas, U.S.A.
Tlie reader is reminded that each chapter is only a general summary that will be quickly exhausted by the avid student, who should go without delay to the References Cited list at the end of Part 2.
Acknowledgments
Much of Part 2 reflects viewpoints and materials from the ongoing AAPG School "Evaluating and Managing Petroleum Risk," which has been team taught since 1984 by E. C. Capen, R. E. Megill, and P. R. Rose. This work is also drawn from the professional association and publications of J. D. Wright and co-author R. S. Thompson. We are greatly indebted to these friends and professional associates for their unselfish contributions to this paper as well as for their thoughtful reviews. Any errors, however, are solely our own responsibility. Thanks also go to Elizabeth Huebner and Barbara Wiley for excellent and expert manuscript preparation and drafting.
Fundamental Economic Equations for Oil and Gas Property Evaluation
Peter R Rose
PROFIT OR LOSS EQUATION
The viability of any business venture can be expressed as the difference (either profit or loss) between revenue and costs. In other words,
Profit or (Loss) = Revenue - Costs
(1)
In modern business, this simplistic equation becomes complicated by various forms of taxes and tax provisions, such as depreciation and allowable depletion calculations. Such complications—plus the accounting procedures they have engendered—obscure the financial performance of most businesses.
However, there is a second reason why most business ventures today are not so simplistic: money is often invested months before revenues begin to be generated and years before profits begin to be realized. Profits are not received as lump sums or even in predictable installments. Also, maintenance costs are incurred repeatedly during a project's lifetime. So the time value of money invested and received must be taken into account.
EQUATION FOR EVALUATING A PRODUCING PROPERTY
The purchase of any oil and gas producing property is a complex business venture. The basic economic equation for evaluating a producing property is as follows:
After-tax profit or (loss), expressed as present value of the
cumulative net cash flow stream
= [(Net revenue interest x Reserves x Wellhead price)
- Wellliead taxes - Operating costs
- Federal income taxes - Investments]
(2)
Several important observations can be made about Equation (2). First, the owners of the producing property u s u a l l y p a y 100% of the costs b u t receive a r e d u c e d proportion—ordinarily from about 70% to 87.5%—of the revenue from production. This reduced proportion is the net revenue interest (NRI). The remainder goes to the royalty owners—generally the landowner. Second, the equation expresses the profit or (loss) as if it were a " l u m p s u m " payment, whereas it is actually received over a long period of time, a net cash flow stream combining production decline, price fluctuations, expenses (including taxes), and inflation. Third, to consider the time value of money, the net cash flows
are expressed as a discounted cash flow stream, so the entire venture can be compared to current alternative financial investments. Wherever a dollar value is expressed as a present value (PV), it means that the value has been discounted to reflect the time value of money.
Uncertainty attends every item in Equation (2) except the net revenue interest. These uncertainties are diverse, relating to geology, engineering, law, politics, economics, and Acts of God. It is the special responsibility of the geotechnical p r o f e s s i o n a l to e s t i m a t e the m a g n i t u d e of reserves, production rates, and costs; to reduce the level of uncertainty as much as possible through sound scientific and technological judgment (and investigation, where warranted); and to convey estimates—as well as uncertainty levels—to management accurately and consistently. Otherwise, management's investment decisions may be misguided and imprudent. Thus, the financial responsibilities and consequences of geotechnical predictions and estimates are enormous.
EXPECTED NET PRESENT VALUE EQUATION FOR DRILLING VENTURES
The financial value of any proposed oil or gas drilling venture can be evaluated by assuming a successful project (Equation 2) and by adding one additional important consideration: the chance of success or failure. This leads to the expected value of the venture, as shown by Equation (3):
Expected net present value
= Chance of success (after-tax net present value)
- Chance of failure (after-tax dry hole cost
+ associated geotechnical and lease costs)
(3)
Thus, expected net present value (ENPV) represents the riskweighted value of a proposed drilling venture. Assuming accurate and consistent perception of both reserves and chance of success, ENPV represents the probabilistic value of each venture and thus becomes a primary tool for decisionmaking and program forecasting. The ENPV is the average value that could be expected if the venture or similar ventures could be repeated many times. Some ventures will result in successes and some will result in failures, but on the average, we expect to make the expected net present value.
The next questions (covered in the next two chapters) concern methods for estimating (1) reserves, rates, and costs, and for estimating (2) chances of success and failure.
24
Uncertainties Impacting Reserves, Revenue, and Costs
Peter R Rose
Telegraph Exploration, Inc. Austin, Texas, U.S.A.
RANGES AND PROBABILITIES
Each of the elements impacting the profit of an oil and gas property—reserves, production rates, costs and wellhead prices, and interest rates—is a prediction made under uncertainty. This fact alone means that "single-number estimates," which are deterministic projections, are usually inadequate for economic assessment of drilling ventures.
The problem is how to express our technical uncertainties realistically and in a form by which they can be utilized in economic equations and formulae. The most common convention in use today involves the formulation of a range of a n t i c i p a t e d v a l u e s for a given p a r a m e t e r , w i t h probabilities—ordinarily 10%, 50%, and 90%—assigned to the values that comprise the range. For example, the geologist may think there is only a 10% chance that the anticipated pay zone will be less than 8 ft thick, 50% sure that it will be less than 12 ft thick, and 90% sure that it will be less than 18 ft thick. The same procedure can be applied to any parameter, including drainage area, production rate, decline rate, and wellhead prices.
However, such estimates cannot be "pulled out of the air"! They must rely on objective considerations of all relevant data, especially maps, cross-sections, geophysical data, borehole log interpretations, analogous producing patterns, and other factors. Moreover, the geotechnical professional must arrive at a final distribution by "shaping it," that is, making trial estimates, examining the implications of various values in the distribution, comparing it with analog data, and adjusting it repeatedly until finally becoming comfortable with the estimates.
BIASES IN ESTIMATING
Unfortunately, a number of psychological biases exist, many of which are described by Tversky and Kahneman (1981), that tend to produce inconsistencies whenever we estimate under uncertainty (Table 1). For the development geologist, three such biases are especially dangerous:
1. Overconfidence, which leads to excessively narrow ranges. People naturally tend to set predictive ranges that typically correspond to a confidence significantly lower than the ranges they think they are setting (Capen, 1976).
2. Conservatism, which leads to underestimates because professionals, fearing criticism, may feel it is worse to overestimate a project than to underestimate it (Rose, 1987).
3. Motivation, which leads to overestimates because of career or economic self-interest on the part of the professional in "selling the deal" (Rose, 1987).
The use of multiple independent estimators utilizing the same database, will go far toward reducing these various biases and producing accurate and consistent estimates. Also, systematic record keeping of geotechnical predictions versus outcomes and periodic review of such "before and after" data has proved to be helpful.
LOGNORMALITY AND LOG PROBABILITY PAPER
Most geological and production parameters are not distributed according to a symmetrical or normal distribution, that is, they do not f o r m a bell-shaped f r e q u e n c y curve. Instead, they tend to produce a frequency distribution skewed to the right, so that there are many small values and only a few large ones. Such patterns approximate a lognormal distributer, and they arise from multiplication of several factors to produce one geological parameter (Kaufman, 1963; Capen, 1984; Megill, 1984). Good examples include field sizes, production rates of wells in a field, porosity-feet (фh) of reservoirs, and effective well drainage area.
Here it is important to remind the reader that in a lognormal frequency distribution, the mode (or most likely point) is positioned to the left, at the peak of the curve. The median (or 50% point) lies in the middle, separating the area under the curve into two equal parts, whereas the mean (or average) lies to the right of the median (see Figure 1). We shall be concerned mostly with the median and the mean in our estimates and calculations, generally discouraging use of the mode, as will be explained later.
In combination with the cumulative probability curve, lognormality provides us with a very useful and powerful predictive tool. Accordingly, it is important to utilize (and understand) log probability paper. Although several forms are commercially available, the three-cycle type in which the probabilities extend from 0.01% to 99.99% is recommended (see Figure 2).
Suppose you want to plot a distribution of field sizes from a basin or trend containing n number of fields.
1. Prior to plotting the data, first arrange all n values in increasing order so that the final curve will be read as "cumulative probability of occurrence equal to or less than " To distribute all n values properly through the entire probability range, utilize the fractile approach in which the incremental spacing of values is determined by dividing n + 1 into 100. For example, in a distribution where n = 6, each value will be plotted at successive increments of 14.3%—14.3%, 28.6%, 42.9%, and so on. The procedure espoused here associates the 90% point
25
26
PART 2—ECONOMICS AND RISK ASSESSMENT
Table 1. Biases Affecting Judgments Under Uncertainty
Type of Bias Overconfidence Representativeness Availability
Anchoring
Unrecognized limits Motivation Conservatism
Common Example
Estimators are much less accurate than they think they are. Analog based on small sample size may not be truly analogous. Recent or spectacular examples are more prone to be cited, regardless of their real
frequency in nature. In estimating, a low starting point leads to a lower final estimate, and a high starting point
leads to a higher final estimate. Geologists forecasting future discoveries may disregard nongeological factors. Prospectors exaggerate the magnitude of reserves in order to sell the deal. The feeling that overestimating a project is worse than underestimating it.
Modified from Tversky and Kahneman (1981) and Rose (1987).
^907c) with larger sizes and values, which are usually less likely. However, this is only a convention. Some workers switch P90% and P10% so that the high percentage (P90%) is associated with the lower value and the low percentage (P10%) goes with the higher value. Either procedure is acceptable, but you must be consistent. 4. Any distribution that approximates lognormal will plot as an approximately straight line on log probability paper. The slope of the line is an expression of the variance. When the log probability paper is oriented so that the probability axis is vertical, as in this article, distributions expressing relative certainty (that is, small variance) have steep slopes, whereas those reflecting great uncertainty have more gentle slopes, often spanning three cycles or more on the log scale.
5. When formulating a distribution, the geologist or engineer should work informally, shaping or sketching possible distributions on the log probability paper, moving back and forth among maps, sections, logs, and data sheets, and considering the implications of many probability levels. After many adjustments, the final curve or line will capture the perceived probability distributions.
6. It is hard to overemphasize the power of independent multiple estimates. If it is possible to obtain and average three or four independent estimates derived from the same data set, the accuracy of the final estimate will nearly always be substantially improved.
7. A useful way to begin is to select an absolute upper maximum possible value and plot it at the 99% position. Then select an absolute lower minimum possible value—some small but finite value—and plot it at the 1 % position. Connect them with a straight line. Now, what value is associated with P500/o? Is it reasonable? What about P90% and P10%? Are they reasonable high-side and low-side values? If not, move the line so that the P90% and P10% values are acceptable. Continue to adjust the curve until you are comfortable with all of its implications. The P99% and P1% plots are merely conventions to help you get started with the estimating process. It is expected that you will adjust these values as you work.
8. Remember that the absolute maximum and minimum values are not P90% and P10%!
9. Although the values associated with any conventional probabilities (P90%, P50%/ and P10%) can be read directly from the log probability paper, neither the mode nor the mean is so apparent. These parameters must be calculated.
10. The mean (or average) is the single best numerical representation of the distribution and is determined approximately by Swanson s Rule of approximation (Megill, 1988) as follows:
Mean = 0.3(P90% value) + 0.4(P50% value)
+ 0.3(P10% value)
(1)
11. Must all distributions be lognormal? A good working rule here is to assume lognormality, but be willing to modify or adjust the distribution if there is legitimate evidence (not just wishes or guesses) supporting it.
COMBINING DISTRIBUTIONS
Distributions can be combined using Monte Carlo or Latin hypercube simulation. These are routinely performed using available software and can involve almost any desired number of parameters.
However, a graphical shortcut is readily available for combining any three lognormally distributed factors. This procedure is quick and can be performed manually. The example shown in Table 2 uses area of drainage, average net pay, and hydrocarbon recovery factor. Other variables important in development evaluations include initial flow rate (IP), decline rate, wellhead prices, and costs.
Note that multiplying the three P90% values for area, net pay, and hydrocarbon recovery does not yield a P90% value for reserves; in fact, it gives a product corresponding to 98.7%! Similarly, multiplying the three P10% values gives a reserves product that corresponds to P13%, not P10%.
However, we can put this to work. First, plot the P98 7% reserves product at P98 7% on the log probability paper and the P13% reserves product at P13%. Draw a line connecting them. The median value should lie on or near the line at P50%. Now, derive the values associated with P90% and P10% from the new reserves line and use them to solve for mean reserves using Swanson's Rule. Table 2 illustrates the calculations, and Figure 2 shows the graphical procedure. (As a reality check, you can also determine the mean values for area, net pay, and
Uncertainties Impacting Reserves, Revenue, and Costs
27
MODE ("-MOST LIKELY) MEDIAN (=50% POINT) / MEAN (= AVERAGE)
Figure 1. Location of mode, median, and mean shown schematically on a Iognormal frequency distribution.
hydrogen recovery using Swanson's Rule, and then multiply them to yield a mean reserves figure that should approximate the previously calculated mean.)
This procedure works only when combining three distributions. But an analogous approach for combining two distributions utilizes the values of P92 7% and P7 3 % (in place of P98 7% and P13%). This allows you to combine any reasonable number of variables into a final probability distribution by serial multiplications. Both approaches produce a mean that is accurate within about 3%—which is far more accurate than our precision in estimating such geotechnical parameters!
FIELD SIZE AND WELL SIZE DISTRIBUTIONS
Within geologically consistent entities, such as basins, trends, or plays, populations of analogous oil and gas fields typically occur in lognormal distributions. This pattern becomes useful in several ways:
ACCURACY LEVELS IN GEOTECHNICAL PREDICTIONS
Geologists, geophysicists, and engineers think they are much more accurate than they really are. In exploration forecasting, a goal of 0.5x to 2x for area of accumulation, net pay, and hydrocarbon recovery factor is about as accurate as nature will allow us to estimate. In other words, 80% of our predictions should be within an envelope of one-half to twice that of reality. Reserves predictions may vary more because of the m u l t i p l i c i t y effect; p e r h a p s 0.2x to 5x is a m o r e reasonable range to expect here. Such variances can be portrayed on log probability paper.
Geologists working on development projects should do somewhat better than this, however, and perhaps a general range of 0.8x to 1.25x of actuality is expectable for predictions based on geological parameters and reservoir performance. Correspondingly, reserves predictions should fall within the 0.67x to 1.5x envelope for development wells, at the 80% confidence level.
Professionals interested in improving their performance in estimating will find that keeping records on predictions versus actuals will help them gradually improve their geotechnical forecasts (see Rose, 1987).
OTHER UNCERTAINTIES IN THE ECONOMIC EQUATION
The remaining uncertainties involved in evaluating drilling ventures center around three main elements: reservoir performance, prices and costs, and taxes and regulatory costs.
Reservoir Performance
An important uncertainty is reservoir performance, especially initial production rates, cumulative and ultimate production, and production decline rates. Field IPs and cumulative production tend to be lognormally distributed, whereas variations above or below typical decline rates seem to follow a more normal distribution.
1. Field size distributions (FSDs) drawn on log probability paper serve as useful reality checks (Megill, 1988) against which prospects can be compared for purposes of perspective and likelihood.
2. If realistic estimates can be made of minimum economic field size, FSDs can be used to determine the chance of commerciality, given that a discovery is indeed made. For example, if 20% of all fields in a producing trend are too small to return enough revenues to cover the costs to find, develop, and operate them (plus a reasonable profit on the investment), then we can expect one out of five new field discoveries in that trend to have found a noncommercial field.
3. Commonly, wells within a field also show lognormal distribution, especially with regard to initial production rates and projected ultimate recoveries. Cumulative log probability plots of such parameters have great predictive value for the development geologist.
Prices and Costs
Prices and costs can be significant, especially wellhead prices, operating costs, inflation rates, and interest rates. Although many companies routinely project future increases in prices and costs in their economic analyses, the historical record indicates this practice may be misleading and unreliable. It is misleading because some marginal drilling ventures may become "commercial" only through such price escalation forecasts, and it is unreliable because the practice requires two forecasts to be made (price and cost escalation plus inflation rates), and the industry record at predicting either is demonstrably abysmal.
One procedure is to hold all wellhead prices and operating costs constant, assuming that they will rise or fall together. In addition, many firms now recognize that, by present standards, the historical price of oil has oscillated since 1880 between about $11 and $22 per barrel, (except for the 1978-1984 and 1990-1991 spikes), and they set their price predictions in that range. However, when constant prices and
28
PART 2—ECONOMICS AND RISK ASSESSMENT
I5OOObbI
10,000 bbl
IOO5OOObbI
I5OOO5OOObbI
Reserves
Figure 2. Worksheet showing graphical method of combining distributions to derive the mean reserves on three-cycle log probability paper.
Uncertainties Impacting Reserves, Revenue, and Costs
29
Table 2. Calculation of Means (see Figure 2)
P10o/ovalue
P50o/o value
P 9 0 O 7 o value Mean3 (Mz)
Area
20 ac 32 ac 50 ac 33.8 ас
Net Pay
x 7ft x 12ft x 20 ft x 12.9 ft
Approximated by Swanson's Rule.
HC Recovery
x 100 bbl/ac-ft x 190 bbl/ac-ft x 350 bbl/ac-ft x 211 bbl/ac-ft
Reserves
= 14,000 bbl = P 1 . 3 % = 72,960 bbl = P 5 0 % = 350,000 bbl = P 9 8 . 7 % = 92,000 bbl
Derived
P 1 0 O 7 o = 28,000 P 5 0 O 7 o = 72,960 P 9 0 O 7 o = 180,000
Mza = 91,584 bbl
costs are a s s u m e d and the inflation c o m p o n e n t of the d i s c o u n t rate is o m i t t e d (leaving a d i s c o u n t rate of approximately 3-4%), the project analyst is assuming that any price or cost changes that do occur will exactly offset any inflation or loss of purchasing power. If all the assumptions are correct, the result would be a constant purchasing power net present value.
Variations in oil prices would be expected to follow a lognormal statistical distribution, given a free world market— many small price fluctuations and only a few very large ones. H o w e v e r , p r e d i c t i n g variations t e n d s to be a n e a r - t e r m preoccupation, which is of limited value in a business that is as long-term as the oil business. Predicting trends would indeed be useful if we could do it—but the historical evidence says we cannot.
Predictive accuracy of cost forecasts for exploration ventures typically lies within the 0.5x to 2.Ox range (80% of predictions are within one-half to twice the actual costs incurred). Because lower uncertainties ordinarily attend development projects, cost forecasts should be more accurate, perhaps within a range of about 0.80x to 1.25x.
Estimators who deliberately overestimate such costs to
prevent cost overruns actually impose unnecessary burdens on project economics by making projects appear to be more expensive to carry out than they actually are. This may cause the firm to rule out many prospects and thereby suffer the opportunity loss that each rejected prospect represents. Ideally, such predictions should be objective so that cost overruns and underruns balance at the end of each drilling year.
Taxes and Regulatory Costs
Taxes and regulatory costs, which show substantial variation, can also be expressed as ranges. Commonly, however, the effect of such governmental regulatory activity is to delay project performance, thus reducing profitability because of the time value of money. There is a clear tendency for operators to underestimate both the number and duration of such delays. (Capen, 1976). It is also possible that future investments and operating costs will increase as a result of future regulatory activity. (For more details on taxes, see the chapter entitled "About Taxes.")
Expected Value and Chance of Success
Peter R. Rose
Telegraph Exploration, Inc. Austin, Texas, U.S.A.
INTRODUCTION
Economic analysis of contemplated oil and gas ventures must be carried out on the assumption that the project is successful, with "success" often being expressed as one of several levels of profitability based on the various ranges in geotechnical and economic parameters that impact project commerciality. However, many exploration ventures do not succeed, and not all development wells and projects succeed either, so the consequences of such failure must be considered in appraising the economic merit of a proposed development venture.
Accordingly, the expected value (EV) of any venture can be expressed as follows:
EV = Probabilitysuccess(Project present value)
- Probabilityfailure(Costs of failure
(1)
EXPECTED VALUE CONCEPT
Imagine that you have the opportunity to participate in a simple game in which you are asked to correctly call the toss of a fair coin. If your call is correct, you will win $20,000; if it is incorrect, you will win nothing.
If you w e r e able to play such a g a m e "for free," the expected value of each trial would be +$10,000. If you had to pay $10,000 each time you played, the EV would be zero, so that, statistically, you then would be "trading dollars." If you were willing to invest $8,000 in one trial of this game, the EV would be +$2,000 (Table 1). In this example, there are only two possible outcomes and you are betting on one trial. It is important to emphasize that in oil and gas exploration, there are many possible outcomes. Furthermore, the concept of expected value as a decision criterion requires repeated trials. The expected value is the average profit per decision assuming repeated trials are made.
Faced with choosing among several options, the decision rule is to select the option having the highest EV. Remember, one alternative is to invest in a risk-free project having some minimum return (net present value = 0 discounted at a riskfree interest rate). Obviously, when operators choose to participate in ventures having negative expected values, they are "betting against the House."
In exploratory ventures, cost of failure usually includes dry hole cost, cost of lease bonuses of the condemned leases, and some G & G costs. For development ventures, some substantial additional capital investments may also occur, plus expense items that will have to be written off as well— expenditures that were needed to determine the viability of the project, such as several completed wells, equipment, materials, and supplies.
Newendorp (1975) presents the subject of expected value
very thoroughly for the reader who wishes additional background.
PROBABILITY OF GEOLOGICAL SUCCESS
Every expression in the expected value equation requires a responsible geotechnical estimate. We have already addressed procedures for estimating reserves (the primary component of revenue) and costs (both of success as well as failure) in the previous chapter ("Uncertainties Impacting Reserves, Revenue, and Costs"). Here we address the problem of assessing the chance of geological success.
For exploratory prospects (including shallow pool, deeper pool, and extension wildcats, commonly managed by development geologists), the recommended procedure is for the geotechnical professional to express his or her confidence i n d e p e n d e n t l y in four critical geological aspects of any prospect:
1. What is the probability (or confidence) that reservoir rock is present, of sufficient porosity and permeability to be productive, and in some minimal thickness and extent sufficient to contain detectable (i.e., measurable) quantities of mobile hydrocarbons, or to tempt a prudent onshore domestic operator to attempt a completion?
One approach is to estimate the minimum required flow rate and relate this flow rate to thickness and permeability. In any case, what we seek is the geologist's confidence in the existence of at least a minimal reservoir—thickness, extent, porosity, and effective permeability. Under this approach, encountering a wet, commercial-quality sandstone would not be a failure in the reservoir category, but rather in one of the other categories, such as an unexpected structural low, an absence of hydrocarbon charge, or a leaky trap. However, the presence of a 1-ft-thick tight siltstone where a 10-ft-thick porous sandstone objective had been predicted would be a reservoir failure!
2. What is the probability (or confidence) that the geological structure of the reservoir objective is, in reality, essentially as represented on maps and cross sections?
It is important to note here that we do not require an actual "structure," such as a domal anticline or a fault closure, only that prospect maps and sections accurately depict the s t r u c t u r a l c o n f i g u r a t i o n . For e x a m p l e , if only r e g u l a r monoclinal south dip is required in the case of a stratigraphic trap prospect, then the geologist should express confidence— as a probabilistic estimate—that the structure in the vicinity of the prospect actually is indeed regular monoclinal south dip.
30
Table 1. Expected Value Examples (Coin Toss)
Trial Free trial
Outcome
Correct call Incorrect call
Consequence - Cost
+$20,000 - 0 = 0- 0=
$10,000 trial
Correctcall Incorrectcall
+$20,000 - $10,000 = 0-$10,000 =
$8,000 trial
Correctcall Incorrectcall
+$20,000 - $8,000 = 0-$8,000=
Expected Value and Chance of Success
31
Profit/Loss x Probability = +$20,000 x 0.5 = 0x0.5=
+$10,000x0.5 = -$10,000x 0.5=
+$12,000 x 0 . 5 = -$8,000x0.5=
Risked Result +$10,000 0
EV = +$10,000
+$5,000 -$5.000
EV=
0
+$6,000 -$4.000
EV= +$2,000
If the map shows an antithetic fault closure, then what is the probability that such a structural configuration will actually turn out to be present?
This geological chance factor is formulated to apply to stratigraphic as well as structural traps, and in tacit acknowledgment that the structural map is ordinarily the single most important map involved in most prospects and many development projects. Also, structural "busts" are a common reason for dry holes (Rose, 1987).
The geological structure chance factor, in combination with the reservoir requirement, focuses on the geometry of the envisioned oil or gas accumulation and on the volumes of fluids necessary to sustain a production test or prudent drill stem test.
3. What is the probability (or confidence) that hydrocarbons are present in the subsurface geological environment such that the prospect has had access to them in some quantity to provide at least some modicum of hydrocarbon charge?
sealing faults). Fluid viscosity, bed thickness, differential permeability, and fault history all influence the seal question. The second question is about timing, as noted in item #3 above: if the trapping configuration came into being after migration occurred, then the gate has been shut only after the horse got out. The third question has to do with preservation f r o m s u b s e q u e n t f r e s h w a t e r flushing or d e g r a d a t i o n of reservoired hydrocarbons. As used here, the term trap has no i m p l i c a t i o n s of g e o m e t r y or c o n f i g u r a t i o n — o n l y of containment and sealing. The troublesome issue of "fill-up" (best represented as a percentage) falls into this category. For most development wells, the sealed trap requirement has been satisfied.
The voice of experience warns you that for exploration projects, do not use probabilities of 1.0—you simply cannot be that sure! "Absolute certainty" is 0.9 or 0.95.
The next step is to multiply the decimal fractions representing the four geological chance factors, which produces the probability of geological success, or Psg.
This geological chance factor deals with such questions as the volumetric adequacy of petroleum source rocks, the generation of oil a n d / o r gas, the migrational pathways to the site of the prospect, and the concentration of hydrocarbons in the reservoir fluid (hydrocarbon saturation of at least 50% is required). The question of timing is not addressed here. In most frontier basins, the hydrocarbon charge issue is very important. In established basins and producing trends, however, its significance tends to be slightly diminished. Obviously, for development projects, the hydrocarbon charge requirement has ordinarily been satisfied.
4. What is the probability (or confidence) that a sealed trap exists, based on the lithologic combinations and structural configurations depicted, and that the trapping configuration was already formed when hydrocarbons were migrating into the area of the prospect?
Here we address three questions. First is the idea of the sealing capability between reservoir and top seals, seat seals, and lateral seals (whether formed by stratigraphic contrasts or
PROBABILITY OF COMPLETION
The key question for most development geologists is, "What is the probability that this well will be completed?" Thus, the probability of success, or Ps, is really the probability of completion. The probability of geological success (as defined and derived above) can be made to approximate the probability of completion (or the probability of success) by two linked measures:
1. Minimum but finite dimensions are required for all reserves parameters, such as area, net pay, and hydrocarbon recovery factor. The concept here is that a small but finite volume of oil or natural gas, and some minimum reservoir thickness and quality, must be present for an accumulation even to be detected by an operator. In other words, the lower limit of an accumulation thus defined is substantially larger than 1 bbl of oil!
2. The four geological chance factors are defined so as to include the concept of the practical lower limit, that is, a modicum of porosity, permeability, and thickness of
32
PART 2—ECONOMICS AND RISK ASSESSMENT
reservoir rock; a closure sufficient to contain an accumulation large enough to sustain a production test (or even a prudent drill stem test); and a hydrocarbon charge and sealing capability sufficient for at least 50% hydrogen saturation.
The probability of success and—by subtracting it from 1.0—its derived counterpart, the probability of failure, or Pf, are the expressions required to calculate the expected value of an exploratory drilling venture. A further modification is necessary for development wells and projects.
When compared with exploration drilling statistics, such as those reported annually by the AAPGs Committee on Statistics of Drilling (CSD), Ps was approximately equivalent to their definition of success, that is, that the subject well was completed and did produce some hydrocarbons. This does not mean that the venture was profitable. In fact, this definition of success contains at least four possible outcomes:
1. The well was completed as the discovery well for a profitable exploratory project (a commercial success).
2. The well was completed because anticipated future production would return a profit on the cost of completing and operating, but not on the full exploratory costs, which are thus viewed as sunk and not recoverable (an incremental success). Ordinarily, no more wells would be drilled on the property by the operator, assuming that the well did not provide other new encouragement.
3. The well was completed as an incremental success, but subsequent performance was inadequate to sustain even operating costs, resulting in abandonment a short time later.
4. Tlie well was completed for "business reasons," that is, to hold a lease position or to satisfy a contractual or regulatory obligation.
Contemporaneous drilling statistics serve to put all this into proper perspective, as shown in Table 2, which reports 1988 results by different classes of wells.
PROBABILITY OF COMMERCIAL SUCCESS
For exploration ventures, the recommended method to assess the chance of commercial success is to first identify the minimum field size associated with your firm's definition of the threshold of commerciality, and then to determine what proportion of such fields occur in the natural population of counterpart accumulations in the subject trend, play, or basin. This requires the geologist or engineer to construct a field size distribution, as previously discussed.
Example
For a given extension project having a predicted mean reserve size of 1,500,000 ВОЕ, the geologist has concluded that the p r o b a b i l i t y of r e s e r v o i r rock is 0.9, the s t r u c t u r a l probability is 0.8, and the probability of hydrocarbon charge is 0.9. The chief geological risk concerns whether a key fault will or will not seal, and the geologist assesses this as a 50/50 proposition. Thus, the perceived chance of geological success is
Table 2. 1988 Success Rates, United States and Canada
Well Class
New pool wildcats Deeper pool wildcats Shallower pool wildcats Outpost (extension) wildcats New field wildcats
All exploratory wells All development wells
Success Rate
U.S.
Canada
0.53
0.48
0.15
0.54
0.62
None reported
0.42
0.68
0.14
0.30
0.30
0.56
0.79
0.85
Source: AAPG Committee on Statistics of Drilling
32%. However, construction of a field-size distribution for 20 analogous fault-separated fields in the trend reveals that only 3 / 4 of t h e m are larger than 200,000 bbl, w h i c h is the minimum economic field size in this trend for your firm. Therefore, the chance of commercial success is 0.75 x 0.32 = 0.24. Calculated in this way, 0.24 represents the chance of finding a field of 200,000 bbl or larger.
APPLICATIONS TO DEVELOPMENT PROJECTS
In development projects, just as in exploration projects, the geological chance factors must be derived from the study of maps, cross sections, and well data. They cannot be "pulled out of the air." However, there is an important difference: the fact that development is contemplated at all implies that a petroleum accumulation exists, so the hydrocarbon charge and seal/trap requirements have generally been satisfied. The only remaining geological risks have to do with (1) structural variations that may depress the reservoir below the oil-water contact and (2) stratigraphic variations affecting both thickness and quality of the reservoir section.
Although individual development wells have a high probability of success, some development dry holes are drilled. Naturally, the proportion of development dry holes will vary a c c o r d i n g to the geological characteristics of individual fields and trends. Nevertheless, this failure rate is significant and must be anticipated in (1) constructing the cash flow model of the development project; (2) determining the expected net present value of each development well; and (3) anticipating the chance of failure of the entire development project, particularly if it is a small one involving only a few wells.
Moreover, since development wells should generate sufficient production revenues to pay out in less than about 3 years, most operators will not purposefully continue to drill development wells that are only incrementally commercial. Even so, it is still true that many development wells are completed each year that return only enough to pay for completion and operating costs, not the cost to drill them. N e v e r t h e l e s s , the p r o b a b i l i t y of success as a p p l i e d to development projects should always be the probability of commercial success, which for most development wells should generally be 60-80%.
For enhanced oil recovery projects, the geologist or engineer is well advised to anticipate ranges in final project or
Expected Value and Chance of Success
33
= M$400K
5($400K) = (-) $200 K EV= M $ 2 0 0 K
I
! (+)$295K ,
I (-)$ 33K [EV=j+)$262K_|
<qX
! .67(+$440K)
^
L=W$295K J
33 ($100K) = (-) $33K
oov^>-8(+$800K-$100K-$25K) = -8(+$675K) = (+)$540K
^•2(-$400K-$100K=H$100K EV = (+)$440K
Figure 1. Development well decision tree problem.
process efficiency when constructing future scenarios. In addition, project commerciality may be severely impacted by future negative (or positive) trends in nongeological factors, such as costs, wellhead prices, transportation problems, and time delays. Thus, these contingencies should also be anticipated and expressed probabilistically. Therefore, the probability of commercial success for a development project should have three components: (1) geotechnical, (2) operational, and (3) economic. A key parameter here is the minimum acceptable economic threshold for project performance.
Accordingly, the chance of project success becomes the probability of achieving the minimum acceptable return (or higher). This is related to the probability of finding at least some m i n i m u m reserves capable of p r o d u c i n g at some minimum rate. At reserves greater than this minimum, the project will be commercial. Thus, the calculation of expected value for development projects is
ENPVcommercial = (Probability of commercial success)
x (Net present value of mean project outcomes
above commercial minimum)
(2)
- (Probability of commercial failure)
x (Net present value of mean project outcomes
below commercial minimum, including dry holes)
where ENPV = expected net present value.
DECISION TREE ANALYSIS
The expected value concept also has important applicability in the analysis of complex a n d sequential decisions. Here the basic idea is to "map out" the sequence of events, indicating decision and chance nodes as follows:
на: Decision node
Chance mode
Probabilities must be assigned to all possible outcomes emanating from chance nodes, but not to branches from decision nodes. Risked values are assigned to each foreseen decision. Decision trees are constructed from left to right and solved from right to left.
Theoretically, the basic decision rule is to always choose the branch having the highest expected value. In practice, however, capital-constrained companies often select a less desirable option due to capital requirements of a higher EV alternative.
Figure 1 shows a decision tree for a simple development well problem. The problem here is whether to drill a well at the edge of a developing oil field or to shoot a seismic line first to try to determine whether the location may be structurally
34
PART 2—ECONOMICS AND RISK ASSESSMENT
low and wet and thus move the drill site to the most favorable location. There is also uncertainty about reservoir quality, which cannot be resolved without drilling. The costs in this example are as follows: cost of seismic ($100,000), cost of deferring production to allow time to shoot, process, and interpret seismic ($25,000), cost of dry hole ($400,000), and mean present value of producing well ($800,000, including drilling costs). An analysis of this p r o b l e m s h o w s two options:
Option A: If you drill without shooting seismic, your staff sees two possible outcomes: (1) a 50% chance of a profitable producing well or (2) a 50% chance of a dry hole.
Option B: If you first shoot a seismic line, your staff sees the following possible outcomes: (1) a 33% chance of
getting a negative seismic result (that is, a structural low) and aborting the project or (2) a 67% chance of getting an encouraging seismic result and therefore drilling the well in the most favorable location. Positive seismic results would be expected to change the odds to an 80% chance of a profitable producer and a 20% chance of a dry hole.
The solution to the problem shown in Figure 1 is option B. It is the preferred choice because it has the higher estimated present value ($262,000 versus $200,000). Thus, the value (or profit) to the project of the additional seismic is +$62,000. Note, however, that if the proposed seismic line costs $200,000, option A becomes the preferred choice ($200,000 versus $161,000), so it would be more cost effective to drill rather than to shoot.
The Time Value of Money
Robert s Thompson
Colorado School of Mines Petroleum Engineering Department Golden, Colorado, U.S.A
FUTURE VALUE
When money is borrowed for a period of time, rent (or interest) for the use of the money must be paid in addition to repayment of the amount borrowed. Thus, money has a time value. In oil and gas property evaluation, two equations with respect to time value are useful: the future value of a lump sum investment and the present value of a l u m p sum of money received in the future. The concept being applied is one of equivalence, in which the period interest rate is used to calculate this equivalency. In the case when interest is compounded annually, interest earned during each year earns interest in future years. For example, $1.00 invested today at 10% interest compounded annually will be worth $1.10 a year from now, $1.21 two years from now, and $1.33 three years from now. Tlie following equation represents the f u t u r e v a l u e of a l u m p s u m i n v e s t m e n t c o m p o u n d e d annually:
F - P(1 + Of
(1)
where
F = a future lump sum of money P = a present lump sum of money i = nominal annual interest rate t - years
In Equation (1), the amount F received t years from now is equivalent to the present day investment P if the interest rate is i per year.
PRESENT VALUE
This same equation can be rearranged to solve for the equivalence (or present value) of a future sum of money (such as a project net cash flow) received some time in the future. For example, a dollar that we expect to receive one, two, and three years hence is worth today $0,909, $0,826, and $0,751, respectively, if the time value of money is 10% per year c o m p o u n d e d annually. Equation (2) expresses this principle of present value:
P = F / ( l + i)f
(2)
The present value concept is important in petroleum economics because we need to know how to place a value on cash flows to be received from production in future years. This concept is demonstrated in Figure 1. In oil and gas property evaluation, profit is measured in terms of net cash inflows and net cash outflows. In Figure 1, the horizontal line represents time, the vertical arrows above the horizontal line
represent net cash inflows, and the vertical arrows below the horizontal line represent net cash outflows. The profit is usually measured for increments of one year. One exception is the time 0 profit period. Time 0 is the instant in time when the first significant expenditure is made.
When each of the future net cash flows are discounted to time 0 using Equation (2), the resulting net cash flow is called the net present value and is equivalent to the project cash flows at the assumed discount rate. The equivalent time 0 net cash flow is also shown in Figure 1.
DISCOUNT RATES
Table 1 shows present value factors at different interest rates (or discount rates) in future years. Some corporations use a discount rate approximately equal to the corporate cost of capital, or the present inflation rate, plus an additional 3-4% (which represents "real" bank interest). At the time this paper was written, this gave an 8-9% corporate discount rate. During this time, however, most U.S. companies were using a discount rate of 12-15%, and international firms were using 15-18%.
Some corporations advocate calculating an average reinvestment opportunity rate based on past performance and using this discount rate. When the historical record is analyzed (sometimes called a post audit), the analysis can be done on both a dollars-of-the-day basis and in terms of constant purchasing power dollars. It is not an easy task to put all the project cash flows in terms of constant purchasing power for the b a s k e t of goods" the corporation purchases. The concept of using the future price or cost increases of the capital goods purchased by the corporation as an index for loss of purchasing power is discussed in a paper by Krasts and Henkel (1977), in which they apply the concept to discounted cash flow rate of return (DCFROR) calculations. When the post audit average constant purchasing power rate of return is used as the discount rate in net present value calculations, the effect is that projects are compared assuming treasury growth. The project cash flows must also be in terms of constant purchasing power.
This brief discussion demonstrates the diverse approaches that are used to estimate the proper discount rate. This topic deserves more research. For example, should multiple discount rates be used—one discount rate for high risk exploration projects versus a lower discount rate for lower risk development projects? At issue here is how to best deal with risk. A good case can be made for using expected values (probabilistic approach) to account for risk and using the discount rate to account for the time value of money ( T h o m p s o n a n d W r i g h t , 1992). This method is strongly recommended by the author.
35
36
PART 2—ECONOMICS AND RISK ASSESSMENT
Cash Flow Time Diagram
Net Cash I n f l o w s Net Cash O u t f l o w
D
3
A
Proj ect Cash F Iows
E q u i v a l e n t T i me 0 P r o j e c t Cash Flow
0
T I me
The "Rule of 72" is a useful rule of thumb that allows us to estimate the d o u b l i n g time or rate of any p r o p o s e d investment. To find the doubling time of a sum invested at any compounded interest rate, divide the interest rate into 72. For example, at 12% c o m p o u n d e d annual interest, the investment will approximately double in six years.
More detailed discussions of the time value of money can be found in Thompson and Wright (1985).
Figure 1. Comparison of project cash flows and equivalent present value.
The Time Value of Money
37
Table 1. PresentVaIueFactors
Years Hence
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
1%
0.990 0.980 0.971 0.961 0.951 0.942 0.933 0.923 0.914 0.905 0.896 0.887 0.879 0.870 0.861 0.853 0.844 0.836 0.828 0.820 0.811 0.803 0.795 0.788 0.780 0.772 0.764 0.757 0.749 0.742 0.735 0.727 0.720 0.713 0.706 0.699 0.692 0.685 0.678 0.672 0.665 0.658 0.652 0.645 0.639 0.633 0.626 0.620 0.614 0.608
3%
0.971 0.943 0.915 0.888 0.863 0.837 0.813 0.789 0.766 0.744 0.722 0.701 0.681 0.661 0.642 0.623 0.605 0.587 0.570 0.554 0.538 0.522 0.507 0.492 0.478 0.464 0.450 0.437 0.424 0.412 0.400 0.388 0.377 0.366 0.355 0.345 0.335 0.325 0.316 0.307 0.298 0.289 0.281 0.272 0.264 0.257 0.249 0.242 0.235 0.228
4%
0.962 0.925 0.889 0.855 0.822 0.790 0.760 0.731 0.703 0.676 0.650 0.625 0.601 0.577 0.555 0.534 0.513 0.494 0.475 0.456 0.439 0.422 0.406 0.390 0.375 0.361 0.347 0.333 0.321 0.308 0.296 0.285 0.274 0.264 0.253 0.244 0.234 0.225 0.217 0.208 0.200 0.193 0.185 0.178 0.171 0.165 0.158 0.152 0.146 0.141
5%
0.952 0.907 0.864 0.823 0.784 0.746 0.711 0.677 0.645 0.614 0.585 0.557 0.530 0.505 0.481 0.458 0.436 0.416 0.396 0.377 0.359 0.342 0.326 0.310 0.295 0.281 0.268 0.255 0.243 0.231 0.220 0.210 0.200 0.190 0.181 0.173 0.164 0.157 0.149 0.142 0.135 0.129 0.123 0.117 0.111 0.106 0.101 0.096 0.092 0.087
6%
0.943 0.890 0.840 0.792 0.747 0.705 0.665 0.627 0.592 0.558 0.527 0.497 0.469 0.442 0.417 0.394 0.371 0.350 0.331 0.312 0.294 0.278 0.262 0.247 0.233 0.220 0.207 0.196 0.185 0.174 0.164 0.155 0.146 0.138 0.130 0.123 0.116 0.109 0.103 0.097 0.092 0.087 0.082 0.077 0.073 0.069 0.065 0.061 0.058 0.054
8%
0.926 0.857 0.794 0.735 0.681 0.630 0.583 0.540 0.500 0.463 0.429 0.397 0.368 0.340 0.315 0.292 0.270 0.250 0.232 0.215 0.199 0.184 0.170 0.158 0.146 0.135 0.125 0.116 0.107 0.099 0.092 0.085 0.079 0.073 0.068 0.063 0.058 0.054 0.050 0.046 0.043 0.039 0.037 0.034 0.031 0.029 0.027 0.025 0.023 0.021
10%
0.909 0.826 0.751 0.683 0.621 0.564 0.513 0.467 0.424 0.386 0.350 0.319 0.290 0.263 0.239 0.218 0.198 0.180 0.164 0.149 0.135 0.123 0.112 0.102 0.092 0.084 0.076 0.069 0.063 0.057 0.052 0.047 0.043 0.039 0.036 0.032 0.029 0.027 0.024 0.022 0.020 0.018 0.017 0.015 0.014 0.012 0.011 0.010 0.009 0.009
12%
0.893 0.797 0.712 0.636 0.567 0.507 0.452 0.404 0.361 0.322 0.287 0.257 0.229 0.205 0.183 0.163 0.146 0.130 0.116 0.104 0.093 0.083 0.074 0.066 0.059 0.053 0.047 0.042 0.037 0.033 0.030 0.027 0.024 0.021 0.019 0.017 0.015 0.013 0.012 0.011 0.010 0.009 0.008 0.007 0.006 0.005 0.005 0.004 0.004 0.003
15%
0.870 0.756 0.658 0.572 0.497 0.432 0.376 0.327 0.284 0.247 0.215 0.187 0.163 0.141 0.123 0.107 0.093 0.081 0.070 0.061 0.053 0.046 0.040 0.035 0.030 0.026 0.023 0.020 0.017 0.015 0.013 0.011 0.010 0.009 0.008 0.007 0.006 0.005 0.004 0.004 0.003 0.003 0.002 0.002 0.002 0.002 0.001 0.001 0.001 0.001
20%
0.833 0.694 0.579 0.482 0.402 0.335 0.279 0.233 0.194 0.162 0.135 0.112 0.093 0.078 0.065 0.054 0.045 0.038 0.031 0.026 0.022 0.018 0.015 0.013 0.010 0.009 0.007 0.006 0.005 0.004 0.004 0.003 0.002 0.002 0.002 0.001 0.001 0.001 0.001 0.001 0.001
30%
0.769 0.592 0.455 0.350 0.269 0.207 0.159 0.123 0.094 0.073 0.056 0.043 0.033 0.025 0.020 0.015 0.012 0.009 0.007 0.005 0.004 0.003 0.002 0.002 0.001 0.001 0.001 0.001
40%
0.714 0.510 0.364 0.260 0.186 0.133 0.095 0.068 0.048 0.035 0.025 0.018 0.013 0.009 0.006 0.005 0.003 0.002 0.002 0.001 0.001 0.001
50%
0.667 0.444 0.296 0.198 0.132 0.088 0.059 0.039 0.026 0.017 0.012 0.008 0.005 0.003 0.002 0.002 0.001 0.001 0.000
Building a Cash Flow Model
Robert S. Thompson
Colorado School of Mines Petroleum Engineering Department Golden, Colorado, U.S.A.
DATA REQUIREMENTS
The basis of economic evaluation of any proposed drilling venture—a new field, pool, or just a single well—is the cash flow model of investments, expenses, taxes, and wellhead revenues involved with the project. The values for the parameters in this model must come from a geotechnical analysis (including maps, cross sections, and reservoir analysis) of the anticipated n e w field or well and from geotechnical estimates of area, ultimate recoverable reserves, and projected well production schedules. Here are the general data that are required:
1. All front end costs—leases, geology and geophysics (G & G), overhead, exploration drilling, and completion costs
2. Projected dry hole costs 3. Ultimate recoverable reserves (including secondary
recovery) 4. Field area (and thus number of producing wells), unless
it's a single-well project 5. Typical well production schedules, including initial
production, decline rate, and producing life to abandonment 6. All development costs, classified for tax calculations as either tangible expenditures or intangible drilling and development costs (IDCs), and scheduled by the month and year 7. Annual operating cost per well 8. Wellliead prices, wellhead taxes, and transportation costs 9. Abandonment costs 10. Annual federal income taxes. T h e m o d e l c a n b e adapted for the analysis of international prospects by changing this item in the cash flow model to "outsider's take." Simply apply the tax rules for that country. (For more information on calculating U.S. income taxes, see the chapter on "About Taxes.") 11. Anticipated price or cost escalation schedule, if any (the example shown in Table 1 uses the constant dollar concept, thus no price or cost escalation schedule is provided). Inflation (loss of purchasing power) is also assumed to be zero. 12. Net revenue interest (NRI) of lease (1.00 - royalty interest) and company share of working interest 13. Company discount rate 14. Anticipated incremental income tax rate
The procedure is basically to plan out the life of the field, project, or well from the first expenditure through abandonment, assigning costs and revenues to events and dates.
CALCULATIONS IN THE CASH FLOW MODEL
The net cash flow (NCF) for each assumed time period, including time 0, can be determined using the following equation.
After-tax NCF = (Net revenue interest x Production
x Wellhead price) - Wellhead taxes
- Operating costs - Federal income taxes
-Investments
(1)
Some of the economic parameters presented later utilize after-tax net operating income (NOI) as a measure of profit. After-tax NOI is defined as follows:
After-tax NOI = (Net revenue interest x Production x Wellhead price) - Wellhead taxes
- Operating costs - Federal income taxes
(2)
After the revenue and expenditure schedule has been determined, we can now calculate cash income taxes for our project. Finally, once the cash income taxes have been calculated for each year, the cash flow time diagram can be prepared and we are ready to calculate the net present value for our venture. (For an example and explanation of a cash flow time diagram, see the previous chapter on "The Time Value of Money.")
For now, let's assume we are provided the cash income tax number, so we are ready to look at an example problem for a single development well. The assumptions for this example problem and a completed worksheet are presented in Table 1. A completed cash flow time diagram is shown in Table 2 along with the equivalent net present value calculation.
The same steps also apply to a multiwell project. Field development projects are constructed by combining individual well models in a realistic time frame. The income tax calculation must be done on a total project basis since oil and gas taxation applies to the total property. An example of a multiwell field extension project is shown in Table 3. Since the project has a longer life than the example development well, the results are summarized in a slightly different format. Table 4 presents the production, investment, and tax assumptions for the multiwell extension project.
POINTS TO REMEMBER
Here are a few final points that are important to remember concerning cash flow models:
1. Cash flow analysis is the third step in evaluating a proposed (or existing) petroleum property; it occurs after estimating (a) the reserves, rates, and costs and (b) the chance of success.
38
Building a Cash Flow Model
39
Table 1. Cash Flow Model for a Development Well
Year
0
Gross Oil Prod MBO Gross Gas Prod MMCF Working Interest Net Revenue Interest Oil Price $/bbl
Gas Price $/MCF
0.000 0.000 1.000 0.875 18.000
1.500
Gross Income, $M
0.000
- Operating costs, $M 0.000
- Sev.Adv. Tax, $M
0.000
NOI BFIT, $M
0.000
- Cash Taxes, $M
0.000
NOI AFIT, $M
0.000
- Investment, $M
1375.000
NCF AFIT, $M
-1375.000
1991
96.066 48.033
1.000 0.875 18.000 1.500
1576.075 24.000 126.086
1425.989 66.891
1359.098
1359.098
Cash Flow Worksheet
1992
1993
1994
1995
ASSUMED FACTS
52.722 26.361
1.000 0.875 18.000
28.934 14.467
1.000 0.875 18.000
1.500
1.500
15.880 7.940 1.000 0.875
18.000
1.500
8.715 4.357 1.000 0.875 18.000
1.500
CALCULATIONS
864.968 24.000 69.197
771.771 193.309 578.462
474.705 24.000 37.976
412.728 98.278
314.450
260.523 24.000 20.842
215.682 47.305 168.376
142.978 24.000 11.438 107.540 20.163 87.377
1996
4.783 2.391 1.000 0.875 18.000 1.500
78.468 24.000
6.277 48.191
3.274 44.916
578.462 314.450 168.376 87.377 44.916
1997 1998
2.625 1.312 1.000 0.875 18.000
1.500
0.168 0.084 1.000 0.87 18.000
1.500
43.064 24.000
3.445 15.619 -3.798 19.417
2.752 2.158 0.220 0.373 -4.422 4.795
19.417 4.795
Source: After Thompson and Wright (1992).
Assumed Facts: Independent Producer and Royalty Owner status, therefore eligible for percentage depletion NRI = 0.875, Wellhead tax on oil and gas revenue is 8%, annual operating cost is $24,000, incremental tax rate is 34%, oil price is $18/bbl, and gas price is $1.50/MCF Assumed Time 0 investments made on 1 -1 -91: Lease Bonus and G&G (depletable basis for tax calculation) = $125,000 IDC's (100 % expensed for tax calculation) = $950,000 Tangible expenditures (depreciable basis for tax calculation) = $300,000 Estimated dry hole cost if the well is unsuccessful is $750,000 (After-tax = $750,000 x (1 - 0.34) = $495,000)
2. Do not carry out cash flow analysis of "risked reserves"—the cash flow model is built on the success case!
3. Net cash flow is the sum of outlays and inflows. 4. Any investment is a purchase of anticipated future
annual cash flows. 5. A permanent alternative to any petroleum venture is to
put the investment capital "in the bank" (that is, alternative safe investments), where it will earn regular interest at the corporate rate. 6. If prices and costs are assumed to be in terms of constant purchasing power, then the discount component should only include the real interest component. The inflation component should not be included. If the prices and costs are escalated, then the discount rate selected should include the real interest rate and the inflation component. 7. Higher discount rates tend to favor shorter term and lower dollar volume projects (in preference to longer term and higher dollar volume projects), whereas lower discount rates allow substandard projects that may be a drag on corporate earnings. Either excess is deleterious, but the excessively high discount rate is clearly more harmful (Capen, pers. comm., 1990). 8. Some firms use mid-year discounting (rather than endof-year discounting) as being more realistic (see Megill, 1988). Some firms use continuous rather than annual discounting.
9. Although the final cumulative net present value can only be determined by projecting the cash flow model out through the full life of the field, the final few years will typically represent only a small fraction of its worth. Ordinarily, a field production model of about 15 years will be adequate for most purposes, except in the case of very large fields or in cases of "late" enhanced oil recovery (EOR) projects on older fields.
10. The present value of most projects will decrease as successively higher corporate discount rates are utilized. The exception would be an acceleration project (see Thompson and Wright, 1985). The discount rate at which the present value is zero is called the discounted cashflow rate of return (DCFROR) (for more information on DCFROR, see the chapter on "Key Economic Parameters").
11. All figures and estimates should be objective. You should neither purposefully overestimate (to sell the deal) nor underestimate (to be conservative and thereby protect yourself from being wrong). Be professional— give it your best shot!
12. It is a good idea to make several cash flow "cases" using different assumptions for reserves, number of wells, initial potentials (IPs), and decline rates. This is easy to do using modern software. Such sensitivity analyses give the decision maker a better idea of the range of possibilities for project outcomes. However, one shortcoming of many sensitivity analyses is that no
40
PART 2—ECONOMICS AND RISK ASSESSMENT
Table 2. Comparison of Project Cash Flows and Equivalent Present Value for Example Development Well (in $ Thousands)
13 59
I i
5 78
J
P r o j e c t Cash F I ows
3 4
l
16 8
8 7 45
к
19
5
01
2
34
5t 6 * 7 8*
1375
Net Present Value Calculation
Time
Cash Flow
0
$;-1375.000
1
1359.098
2
578.462
3
314.450
4
168.376
5
87.377
6
44.916
7
19.417
8
4.795
Project NPV
Discount Factor
1.0000 0.9615 0.9246 0.8890 0.8548 0.8219 0.7903 0.7599 0.7307
Present Value
$-1375.0 1306.8 534.8 279.5 143.9 71.8 35.5 14.8 3.5
$1015.6
1,015.6
E q u i v a l e n t Time 0 Cash Flow
0
1
2
3
4
5678
probability of occurrence can be assigned to a given case. As a result, the decision maker has an idea of the range of possible outcomes, but no sense of the chance of occurrence of any one outcome! Fortunately, this deficiency is readily correctable by using probabilistic ranges for key variables and Monte Carlo simulation to combine such variables. (For more information on ranges and probabilities, see the chapter on "Uncertainties Impacting Reserves, Revenue, and Costs.") 13. A final step from such sensitivity analyses is the identification of critical threshold values necessary for the project to be commercial. In particular, requisite values for net pay, porosity, and initial production rate may be crucial in helping the well site geologist or engineer to make critical decisions on testing, stimulation, completion, or abandonment. These predetermined values should accompany the geologist or engineer to the well site. 14. There are many commercial computer programs available for both mainframe and microcomputer hardware that carry out cash flow modeling routinely and quickly, allowing many different values to be input for the significant parameters. Use of such software greatly simplifies the process of cash flow modeling once the procedure is fully understood. But for those who want to test their understanding, construct your model from scratch on a spreadsheet. It is not that hard and may prove to be an invaluable learning exercise.
Table 3. Cash Flow Model for Example Multiwell Extension Project
Year
0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Totals
Year
0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
Totals
Gross Oil
Production (MbbI)
Gross Gas Production (MMSCF)
Gas-Oil Ratio
(SCF/STB)
XYZ Oil Co. Net Oil
Production (MbbI)
XYZ Oil Co. Net Gas
Production (MMSCF)
Oil Price ($/bbl)
Gas Price ($MCF)
XYZ Oil Co. Oil
Income ($M)
XYZ Oil Co. Gas
Income ($M)
XYZ Oil Co. Gross Income ($M)
96.066
48.033
500
84.057
42.029
18.00
148.787
74.394
500
130.189
65.094
18.00
177.722
88.861
500
155.507
77.753
18.00
193.601
96.801
500
169.401
84.701
18.00
106.251
53.125
500
92.969
46.485
18.00
58.312
29.156
500
51.023
25.511
18.00
32.002
16.001
500
28.002
14.001
18.00
16.290
8.145
500
14.254
7.127
18.00
7.575
3.788
500
6.628
3.314
18.00
2.793
1.396
500
2.444
1.222
18.00
0.168
0.084
500
0.147
0.073
18.00
839.566
419.783
500
734.621
367.310
Operating Costs ($M)
State + Local Tax
($M)
Net Oper. Inc BFIT
($M)
24.000 48.000 72.000 96.000 96.000 96.000 96.000 74.158 50.158 26.158
2.158
680.632
126.086 195.283 233.260 254.102 139.454
76.534 42.003 21.381
9.943 3.665 0.220
1101.931
1425.989 2197.760 2610.488 2826.170 1507.721
784.141 387.031 171.723
64.183 15.993
0.374
11991.572
Total Cash Invest ($M)
1375.000 3250.000 1350.000 1250.000
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
7225.000
XYZ Oil Company
Net Cash
Fed. Inc. Net Cash
Flow BFIT
Tax
Flow AFIT
($M)
($M)
($M)
-1375.000 -1824.011
847.760 1360.488 2826.170 1507.721
784.141 387.031 171.723
64.183 15.993
0.374
-727.504 210.240 322.798 717.869 365.818 168.643 59.271 12.891 1.826 0.000 0.000
-1375.000 -1096.507
637.520 1037.690 2108.301 1141.903
615.497 327.760 158.833
62.357 15.993
0.374
4766.572 1131.851 3634.721
1.50
1513.032
63.043
1576.075
1.50
2343.402
97.642
2441.043
1.50
2799.118
116.630
2915.748
1.50
3049.221
127.051
3176.271
1.50
1673.448
69.727
1743.175
1.50
918.408
38.267
956.675
1.50
504.033
21.001
525.034
1.50
256.572
10.690
267.262
1.50
119.313
4.971
124.284
1.50
43.983
1.833
45.816
1.50
2.642
0.110
2.752
Cum. NCF AFIT ($M)
-1375.000 -2471.507 -1833.987
-796.297 1312.004 2453.907 3069.404 3397.165 3555.997 3618.355 3634.347 3634.721
Disc. NCF AFIT @ 4%
($M)
-1375.000 -1054.333
589.423 922.503 1802.184 938.561 486.437 249.071 116.057
43.811 10.804
0.243
2729.760
Cum. Disc. NCF AFIT @ 4% ($M)
-1375.000 -2429.333 -1839.911
-917.408 884.776 1823.337 2309.774 2558.845 2674.902 2718.714 2729.518 2729.760
Disc. Invest @ 4% ($M)
1375.000 3125.000 1248.151 1111.245
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
6859.396
42
PART 2—ECONOMICS AND RISK ASSESSMENT
Table 4. Assumptions for Example Multiwell Extension Project
Independent Producer and Royalty Owner status, therefore eligible for percentage depletion
NRI = 0.875
Wellhead tax on oil and gas revenue is 8%
Annual operating cost is $24,000/well. Note in last four years the operating costs are not a multiple of $24,000. This is because the typical well produces only a fraction of a year in the eighth year.
Incremental tax rate is 34%
Oil price is $18.00/bbl Gas price is $1.50/MCF
Cost of Failure is assumed to be the after tax cost of 2 dry holes. ($750,000 x (1 - 0.34) x 2 = $990,000)
Assumed Investment Schedule: Time 0 investments made on 1-1-91: Lease Bonus and G&G = $125,000 IDC's (100% expensed for tax calculation) = $950,000 (for one completion) Tangible expenditures (depreciable basis for tax calculation) = $300,000
Other investments made during 1991: Lease Bonus and G&G = $200,000 IDC's (100% expensed for tax calculation) = $950,000 (for one completion) DHCs (100% expensed for tax calculation) = $1,500,000 (for two dry holes) Tangible expenditures (depreciable basis for tax calculation) = $600,000
Investments made during 1992: Lease Bonus and G&G = $100,000 IDC's (100% expensed for tax calculation) = $950,000 (for one completion) Tangible expenditures (depreciable basis for tax calculation) = $300,000
Investments made during 1993: IDC's (100% expensed for tax calculation) = $950,000 (for one completion) Tangible expenditures (depreciable basis for tax calculation) = $300,000
Assumed Production Schedule: The production forecast for the typical well in the Example Development Well was used for the Multiwell Extension Project. A typical well was assumed to be placed on production at the beginning of each of the years 1991,1992, 1993, and 1994.
About Taxes
Robert S. Thompson
Colorado School of Mines Petroleum Engineering Department Golden, Colorado, U.S.A.
INTRODUCTION
Oil and gas taxes are of two general types: wellhead or production taxes, which are paid w h e n the oil and gas are produced regardless of the amount of profit (or loss), and income taxes, which are paid on the amount of profit (called taxable income) defined by the IRS. Wellhead taxes in most states average about 5 to 10% of the revenue, and the current corporate income tax rate varies from 15% to 39% (see Table 1). Wellhead taxes include excise, ad valorem, and severance taxes.
In federal income taxation, the fundamental concept being applied is the tax free return of invested capital. Congress legislates the definition of profit and thus when the investor will receive the tax f r e e r e t u r n of his i n v e s t m e n t . Expenditures in which the tax payer receives immediate tax free return are called expensed, whereas expenditures in w h i c h the tax p a y e r s p r e a d s the tax free r e t u r n of his investment over several years are said to be capitalized. The timing of the tax deduction for the capitalized expenditures are determined by a set of rules for depreciation, depletion, and amortization.
Oil and gas taxation is one of the most difficult areas of taxation. We attempt here to provide only a foundation of knowledge by presenting basic oil and gas terminology as it applies to taxation and a detailed tax model (Thompson and Wright, 1985,1991).
OIL AND GAS TAXATION TERMINOLOGY
Several important terms are used in oil and gas taxation practices of which the reader should be aware.
Leasehold Costs
Leasehold costs are e x p e n d i t u r e s associated with the acquisition of an economic interest in a natural resource that are deemed to benefit future tax periods. Accordingly, leasehold costs are capitalized, and the tax free return of that expenditure are recovered through the allowable depletion calculation. The depletion deduction was created by Congress to give a means of returning the invested capital to the investor tax free as the resource is depleted. There are two m e t h o d s of calculating d e p l e t i o n : cost d e p l e t i o n and percentage depletion. (See the chapter on "Determining O w n e r s of Oil a n d Gas Interests, a n d M e t h o d s of Conveyance" in Part 1 for more information on lease ownership costs and revenue.)
Lease Bonus
The payment made by the lessee to the lessor for the right to explore for oil and gas and develop the property is called a
lease bonus. The lease bonus payment made by the lessee is capitalized by the lessee and recovered through the allowable depletion calculation. (See the chapter on "Nature of the Oil and Gas Lease" in Part 1 for more on leases.)
Geological and Geophysical (G & G) Costs
Geological and geophysical (G & G) costs are expenditures for geological studies and geophysical work such as seismic surveys. If the expenditures result in the acquisition and retention of a property, the expenditures are capitalized and the tax free return of the investment is determined from the allowable depletion calculation. If the G & G costs do not lead to the acquisition or retention of a property, then these costs can be expensed in the taxable year the property is relinquished.
Lease and Well Equipment
Lease and well equipment includes such items as the wellhead, flow lines, separators, and other equipment necessary to operate the property as well as the labor to install the equipment. Expenditures for tangible items such as these are capitalized, a n d the tax free r e t u r n of the expenditure is recovered through depreciation calculations. O t h e r e x a m p l e s of tangible items i n c l u d e s u r f a c e and production casing, even though the casing is cemented in the well and has no apparent salvage value. Upon abandonment, any undepreciated amount would be written off. The depreciation rules are always changing. As a result of the 1986 Tax Reform Act, the depreciation system is called the Modified Accelerated Cost Recovery System (MARCS) (see Table 1).
Intangible Drilling and Development Costs (IDCs)
Special tax treatment for expenditures classified as intangible drilling and development costs, or IDCs, is available to the tax payer. U.S. Treasury Regulation 1.612-4 states that "all expenditures made by an operator for wages, fuel, repairs, hauling,..., incident to and necessary for the drilling of wells and the preparation of the well for the production of oil and gas" are IDCs. A well is considered to be prepared for production when the wellhead is installed. Table 1 shows the current tax treatment of IDCs. Again, tangible items such as surface casing and production casing are classified as tangible e x p e n d i t u r e s , and the tax free r e t u r n of the expenditure is recovered through depreciation.
Lease Operating Expenses
Lease operating expenses are expenditures incurred in the day-to-day activities of the production operations. Basically, costs incurred during the current tax period that were spent
43
44
PART 2—ECONOMICS AND RISK ASSESSMENT
Table 1. Federal Income Tax Schedules and Rules
Corporate Income Tax Rates
Taxable Income
Not over $50,000 Over $50,000, less than $75,000 Over $75,000, less than $100,000 Over $100,000, less than $335,000 Over $335,000
Rate (%)
15 25 34 39 34
Depreciation Schedule, 7 Year MARCS
Year Placed in Service
1 2 3 4 5 6 7 8
Total
Recovery
0.1428 0.2449 0.1749 0.1249 0.0893 0.0893 0.0893 0.0446
1.0000
Tax Treatment of IDC's
IDC for an Integrated Producer: May elect to expense 70% of IDCs in year incurred and amortize the 30% over 60 months for domestic wells (begins in month costs are incurred). Must capitalize foreign IDCs. May recover over 10-year period using straight-line amortization or may add to depletable basis.
IDC for an Independent Producer and Royalty Owner: May elect to expense 100% of IDCs in the year incurred for domestic wells. Foreign wells are treated the same as the integrated producer.
Taxable Income x Effective Tax Rate Less: Tax Credits Plus: Adjustment Due to Minimum Tax Calculation Equals: Cash Taxes
Figure 1. Federal income tax model for oil and gas transactions. Amounts are net to your working and revenue interest. Calculations are made each tax year. (After Thompson and Wright, 1985.)
Income Taxation. As you can see f r o m Table 1, the IPRO receives preferential tax treatment.
CALCULATING AFTER-TAX NET CASH FLOW
The recommended approach to calculate after-tax net cash flow (NCF) is to use Equation (1) in the previous chapter on "Building a Cash Row Model," which is
in an effort to generate revenues during the current tax period are treated as current period expenses. Direct labor, power costs, and maintenance costs are some examples.
CURRENT TAX TREATMENT
The last significant overhaul of oil and gas taxation is the result of the 1986 Tax R e f o r m Act. An analysis a n d comparison of the tax rules prior to and after the 1986 Tax Reform Act are presented by Thompson (1987). Table 1 summarizes the current tax treatment for the integrated producer (typically, the major oil companies) and the independent producer and royalty owner (IPRO). The tax definitions for an IPRO and an integrated producer are presented in Ernst and Young's (1990) Oil and Gas Federal
After-tax NCF = (Net revenue interest x Production
x Wellhead price) - Wellhead taxes
- Operating costs - Federal income taxes
-Investments
(1)
All transactions in the equation are cash items, one of w h i c h is cash income taxes. The s e p a r a t i o n of the tax calculation from the NCF calculation is recommended because of the many complications in oil and gas taxation. Instead of c o m b i n i n g the N C F calculation and the tax calculation, the federal income tax model for oil and gas transactions (Figure 1) should be used to calculate the yearly taxes for the property, taking into consideration the appropriate tax treatment of each of the transactions. Once a cash tax liability (negative tax = tax savings) is calculated, this tax amount is subtracted, as shown in Equation (1).
Table 2. Income Tax Calculations for Example Development Wella
Detailed Tax Worksheet
Year
Gross Inc. (GI) - Operating Costs - Sev.Adv.Tax - IDC Expense - IDC Amortization - Depreciation
Taxable Income Before Depletion
- Depl. Allowance Taxable Income Cash Tax Before Credits
and Minimum Tax - Tax Credit + Minimum Tax
Final Cash Tax
1991 ($M) 1576.075 24.000 126.086 950.000 0.000 42.840
433.149 236.411 196.738
66.891
66.891
1992 ($M)
864.968 24.000 69.197 0.000 0.000 73.470
698.301 129.745 568.556
193.309
1993 ($M)
474.705 24.000 37.976 0.000 0.000 52.470
360.258 71.206
289.053
98.278
193.309
98.278
1994 ($M) 260.523 24.000 20.842 0.000 0.000 37.470
178.212 39.079
139.133
47.305
47.305
aSee Table 1 in chapter on "Building a Cash Flow Model." Source: AfterThompson and Wright (1992).
1995 ($M) 142.978 24.000 11.438 0.000 0.000 26.790
80.750 21.447 59.303
20.163
20.163
About Taxes
45
1996 ($M)
78.468 24.000
6.277 0.000 0.000 26.790
1997 ($M)
43.064 24.000
3.445 0.000 0.000 26.790
1998 ($M)
2.752 2.158 0.220 0.000 0.000 13.380
21.401 11.770
9.630
-11.171 -13.007 0.000 0.000
-11.171 -13.007
3.274 -3.798 -4.422
3.274 -3.798 -4.422
It sounds simple and it actually is, once some experience is gained. Two worksheets are provided to help keep the numbers straight (Thompson and Wright, 1992). Tables 2 and 3 are the completed worksheets for the example development well in which the producer is an independent producer and royalty owner eligible for percentage depletion. Table 4 is the federal income tax calculations for the example multiwell extension project.
The case for an integrated producer is an easier case since the producer can only take cost depletion.
Calculate Percentage Depletion
Calculate Cost Depletion
CALCULATING ALLOWABLE DEPLETION
Determining the allowable depletion deduction is probably the most difficult calculation. Figure 2 is intended to help with this calculation. As shown in Figure 2, allowable depletion is the greater of cost depletion or p e r c e n t a g e depletion. Cost depletion is calculated by taking the remaining depletable basis (unrecovered G & G costs and lease bonus) and multiplying by the fraction of the remaining reserves produced during the year (production during the year d i v i d e d by reserves at the b e g i n n i n g of the year). All producers are eligible for cost depletion. Independent producers and royalty owners are also eligible for percentage depletion. Percentage depletion is the lesser of 15% of gross income or 100% of taxable income before depletion from the property. Prior to January 1, 1991, the Taxable Income limitation was 50% for each property. This change is the result of the Revenue Reconciliation Act of 1990. This recent change also demonstrates the "dynamics" of tax rules and the importance of seeking professional advice in this area. An example of another complication in the tax law is the 65% of taxable income limitation from all sources (not just limited to the producing property). The 65% taxable income limit from all sources is difficult to apply to single project economics and is ignored in the example problems presented.
Figure 2. Depletion allowance calculation. Note that remaining reserves (U) are end of year. In cost depletion calculation, S = sales during year. (After Thompson and Wright, 1985.)
46
PART 2—ECONOMICS AND RISK ASSESSMENT
Table 3. Depreciation, Depletion, and Amortization Calculations for Example Development Well
Tax Depreciation, Depletion and Amortization Worksheet
Year
1991
1992
1993
1994
1995
Capitalized IDCs1 $M Amortization rate
IDC Amortization, $M
0.000 0.20
0.000
AMORTIZATION
0.20 0.000
0.20 0.000
0.20 0.000
0.20 0.000
Depr. basis (tangibles), $M 7-year MACRS rates
Depreciation Expense, $M
300.000 0.1428
42.840
DEPRECIATION
0.2449 73.470
0.1749 0.1249 0.0893 52.470 37.470 26.790
Depletion basis, $M Cost depletion, $M
15% Gl 100% TIBD Percentage depletion, $M Lessor of 15% Gl or 100% TIBD
125.000 57.211 236.411 433.149
236.411
Allowable Depletion, $M Greater Cost or %
236.411
Gross oil production (MBO) 96.066
Oil res. beg. of period
209.892
0.000 0.000 129.745 698.301
129.745
129.745 52.722
113.826
DEPLETION
0.000 0.000 71.206 360.258
0.000 0.000 39.079 178.212
71.206 39.079
71.206
28.934 61.104
39.079
15.880 32.170
0.000 0.000 21.447 80.750
21.447
21.447 8.715
16.290
aSee Table 1 in chapter on "Building a Cash Flow Model." Source: After Thompson and Wright (1992).
1996
1997
1998
0.0893 26.790
0.0893 0.044 26.790 13.380
0.000 0.000 11.770 21.401
0.000 0.000 6.460 -11.171
0.000 0.000 0.413 -13.007
11.770 -11.171 -13.007
11.770
4.783 7.575
0.000
2.625 2.793
0.000
0.168 0.168
Table 4. Federal Income Tax Calculations for Example Multiwell Extension Project
Year
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
Gross Income
($M)
1576.075 2441.043 2915.748 3176.271 1743.175
956.675 525.034 267.262 124.284
45.816 2.752
Operat. Costs ($M)
24.000 48.000 72.000 96.000 96.000 96.000 96.000 74.158 50.158 26.158
2.158
SEV&ADV Tax ($M)
126.086 195.283 233.260 254.102 139.454
76.534 42.003 21.381
9.943 3.665 0.220
IDCs DHCs ($M)
DEPR ($M)
3400.000 950.000 950.000
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
128.520 263.250 273.720 238.350 170.310 144.630 133.950
93.720 40.170 13.380
0.000
5300.000 1500.000
XYZ Oil Co
Cost
TIBD Depletion
($M)
($M)
Percent Depletion
($M)
Allowable Depletion
($M)
Taxable Income
($M)
-2102.531 984.510
1386.768 2587.820 1337.411
639.511 253.081
78.003 24.013
2.613 0.374
37.187 77.608
6.472
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
-2102.531 366.156 437.362 476.441 261.476 143.501 78.755 40.089 18.643 2.613 0.374
37.187 366.156 437.362 476.441 261.476 143.501
78.755 40.089 18.643
2.613 0-374
-2139.718 618.353 949.406
2111.379 1075.935
496.009 174.326
37.914 5.371
0.000 0.000
1862.597
Tax Credits
($M)
Cash Taxes ($M)
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
-727.504 210.240 322.798 717.869 365.818 168.643 59.271 12.891 1.826
0.000 0.000
1131.851
Key Economic Parameters
Peter R. Rose
Telegraph Exploration, Inc. Austin, Texas, U.S.A.
Robert S. Thompson
Colorado School of Mines Petroleum Engineering Department Golden, Colorado, U.S.A.
INTRODUCTION
Each of the many economic parameters used in assessing and comparing oil and gas ventures has utility in measuring s o m e d e s i r e d aspect of the p r o p o s e d o p p o r t u n i t y . Unfortunately, no single parameter "does it all," but some are better than others. For a thorough treatment of this subject, see Capen et al (1976), Newendorp (1975), Megill (1988), and Thompson and Wright (1985). The "yardsticks" discussed below represent the more commonly used economic measures. The key economic parameters for the example development well and the example multiwell extension project, given in the chapter on "Building a Cash Flow Model," are summarized here in Tables 1 and 2, respectively.
Annual budgeting is carried out once the drilling program is chosen by selecting projects that rank high with regard to whichever economic yardsticks best fit the organization's needs.
NET PRESENT VALUE
Net present value (NPV), or net present worth, is based on the concept of equivalence discussed in the chapter on "The Time Value of Money." The net present value is equivalent to the future cash flows at the assumed discount rate. It is the fundamental parameter to express value of a project assuming success, and it measures the cumulative cash worth of the venture above the corporate discount rate. Ordinarily, it is based upon the mean reserves case. For the development well in Table 1 of the chapter on "Building a Cash Flow Model," the NPV at 4% = $1,015,600. For the extension project (see Table 3 of the same chapter), the NPV at 4% = $2,729,760.
DISCOUNTED CASH FLOW RATE OF RETURN
Discounted cash flow rate of return (DCFROR), or internal rate of return, is calculated by a trial-and-error method. In this method, different discount rates—and thus present value (PV) factors—are inserted in the cash flow model to yield a series of NPVs, beginning with a discount rate of zero (equals undiscounted net cash flow stream). For project cash flows, such as our two examples, the cumulative NPV of a project will decline with successively higher discount rates (Figure 1). The point at which the declining curve intersects the zero present value line corresponds to the DCFROR of the project.
For complex projects involving subsequent enhanced oil recovery additions, the resulting dual rate problem may generate more than one DCFROR.
If treasury growth (Capen, 1976) is the goal of the firm, selecting projects based on DCFROR will not necessarily result in the best selection of projects. Thompson and Wright (1985) discuss the use of DCFROR as a decision criterion and the reinvestment assumption.
It is the authors' opinion that DCFROR should not be used as a "risking measure"—this measure has nothing whatever to do with project risk! Some firms think (incorrectly) that by setting high DCFROR hurdle rates, they are selecting the better projects. Ideally, such hurdle rates should reflect the c u r r e n t real e a r n i n g p e r f o r m a n c e of the firm. For a discussion of the use of high hurdle rates to account for risk, see Thompson and Wright (1992). Excessively high DCFROR hurdles in fact tend to favor short-term, lower reserve, high profit projects (which, upon project completion, the company has a hard time replacing) at the expense of long-term, larger reserve projects—the kind of projects that build corporations. DCFROR is only recommended as a minimum hurdle that all proposed projects must clear to be considered.
The present value profile for the two example cases (Figure 1) s h o w s graphically that the DCFROR for the development well is approximately 45% and that the DCFROR for the extension project is approximately 28%. These two DCFRORs demonstrate an important point about development well economics and exploration economics. Development wells are evaluated on an incremental basis. If over the long term the firm is to generate sufficient cash flow from production to sustain a continuing exploratory program, the DCFROR for the development well must be greater than the DCFROR for the exploration project. This is true because the development well must (1) earn enough to pay for itself, (2) earn a satisfactory return as an investment, and (3) provide additional earnings proportionally equivalent to at least the true cost of the exploratory effort required to discover it (including exploratory dry holes).
Finally, in capital budgeting, management must remember that exploration and prospect generation are longterm commitments. The capital commitment for exploration should not be turned on and off from year to year. Considerable lead time is required for geological and geophysical studies. O n e m e a s u r e of the success of an exploratory p r o g r a m is by a historical analysis of past exploration activities including the actual development
47
48
PART 2—ECONOMICS AND RISK ASSESSMENT
Table 1. Summary of Key Economic Parameters for Example Development Well
Economic Parameter
NPV at 4% DCFROR (%) Cumulative undiscounted NCF (after-tax) Undiscounted payout (years) Undiscounted profit to investment ratio Investment efficiency (4% discount rate) Discounted profit to investment ratio (4% discount ratio) Expected net present value (4% discount rate)
Ps = 0.80, Pf = 0.20
Value
$1,015,600 44.9 $1,201,893 1.0 $1,201,893/1,375,000 = 0.87 $1,015,600/1,375,000 =0.74 $1,015,600/1,375,000 =0.74
0.80 x 1,015,600 - 0.20 x 495,000 = $713,480
Table 2. Summary of Key Economic Parameters for Example Multiwell Extension Project
Economic Parameter
NPV at 4% DCFROR (%) Cumulative undiscounted NCF (after-tax) Undiscounted payout (years) Undiscounted profit to investment ratio Investment efficiency (4% discount rate) Discounted profit to investment ratio (4% discount ratio) Expected net present value (4% discount rate)
Ps = 0.70, Pf = 0.30
Value
$2,729,760 28.0 $3,634,721 3.4 $3,634,721/7,225,000 = 0.50 $2,729,760/2,429,333 = 1.12 $2,729,760/6,859,396 = 0.40
0.70 x 2,729,760 - 0.30 x 990,000 = $1,613,832
expenditures. The result should be a DCFROR greater than the minimum hurdle rate. On the other hand, a historical analysis of the development costs and subsequent production should result in higher historical DCFRORs than the total exploration program since the exploratory costs would be left out in this analysis.
PAYOUT
The payout is the length of time required for the venture to generate income sufficient to equal capital investment and e x p e n s e s (see Figure 2). This m e a s u r e is of g r e a t e r importance to small investors, who are concerned about liquidity and risk exposure. It can be calculated using constant purchasing power dollars. In cases where a loan payment is required, dollars of the day should be used. Its major drawbacks are that it doesn't consider cash flows after payout occurs, nor does it address any aspect of investment performance. Although payout implicitly touches on financial risk and exposure (which translates to "when do I get my money back?!"), it does not address chance of success in any way.
Megill (1988) points out a useful rule of thumb: a rough, reciprocal relationship exists between payout and DCFROR. A project that pays out in 3 years will have a DCFROR of about 33%; one that pays out in 4 years will have a DCFROR of about 25%.
Individual development wells should pay out in less than 3 years. The development well in Table 1 in the chapter on "Building a Cash Flow Model" pays out in approximately 1
year. Most exploration projects should pay out in about 4 to 8 years, except for very large projects in difficult areas. The development well example has a 1.0 year payout, whereas the extension project has a 3.4 year payout.
Discounted payout, in which both the investment stream and the revenue stream are discounted, is probably a slightly more useful measure, but is not widely employed. The extension project has a discounted payout of approximately 3.5 years.
MAXIMUM NEGATIVE CASH FLOW
Maximum negative cash flow (MNCF) is an i m p o r t a n t measure because it expresses the greatest cumulative out-ofpocket expense—that is, the greatest cash "exposure" in any project—and thus is useful in budgeting, planning, and project comparison in firms that are cash constrained. It is derived from the cash flow model by expressing the net of investments and costs against early revenues. Thus, it is the turnaround spot on the cumulative net cash flow stream. It is not a discounted number. The undiscounted MNCF for the extension project is $2,471,507, as shown in Figure 2 (see Table 3 in the chapter on "Building a Cash Flow Model" in Part 2).
Like payout, MNCF may be relatively more important to risk-averse smaller investors. It does not address chance of project success. For individual development wells, such as our example problem, MNCF is not particularly useful since there is rarely much overlap in time between capital expenses and production revenue.
4000
4000 2000
Key Economic Parameters
49
Un d lscoun td NCF"I 8 3635
Dlecoun Ie NCF'e 8 2730
UDleevIelIopment
-2000
-4000
MNCF = - 8 2429 (Discounted) MNCF = - 8 2471 (Undiscounted)
20
40
60
80
D l s c o u n L R o Ie ( %)
Figure 1. Present value profile and determination of DCFROR for example development well and example multiwell extension project.
-6000
Figure 2. Undiscounted and discounted cumulative net cash flow streams for example multiwell extension project.
1976) and is defined as
UNDISCOUNTED PROFIT TO INVESTMENT RATIO
Undiscounted profit to investment ratio (P/I) measures the magnitude of cash flow with respect to investment, but it does not address the time frame in which the profits are received. Neither does it express the magnitude of the venture or any aspect of risk. Profit can be defined as the net operating income (NOl) or as the net cash flow (NCF). If profit is defined as NCF, the minimum acceptable ratio is 0.0, whereas if it is defined as NOI, the minimum acceptable ratio is 1.0. The difference between the two ratios for a project is always 1.0. If someone uses profit to investment ratios, ask them what their definition of profit is. Also, projects need to be ranked and compared on a consistent basis.
The undiscounted profit to investment ratio (with profit as NCF) for the development well and the extension project in the chapter on "Building a Cash Flow Model" are $1,201,893/$1,375,000 = 0.87 and $3,634,721/$7,225,000 = 0.50, respectively.
INVESTMENT EFFICIENCY
Several economic parameters can be used to measure investment efficiency. When capital is limited, these yardsticks permit projects to be ranked from high to low until the available capital is exhausted. The investment efficiency ratios a r e indicators of t h e projects profit p e r dollar of investment. The two parameters presented here are very similar but do differ slightly. Primarily, they differ in our perception of how the budgeting process actually takes place. Regardless of the economic yardstick used, y o u must be consistent and compare all projects with the same yardstick. The first ratio is called the investment efficiency (Capen et al.,
Investment efficiency Cumulative net present value
PV of maximum negative cash flow
(1)
The second parameter is called the discounted profit to investment ratio and is defined as follows:
Discounted profit to investment ratio
Cumulative net present value
PV of all investments
(2)
Note that the numerator is the same for both yardsticks. However, the investment efficiency parameter uses only the present value of the early negative NCFs in the denominator. The present value of the MNCF is defined in this case as the greatest cumulative discounted out-of-pocket expense. The discounted profit to investment ratio keeps all investments separate a n d uses the cumulative present value of each i n v e s t m e n t as t h e d e n o m i n a t o r . The significance of this difference has not been rigorously tested to our knowledge. It is believed, however, that the investment efficiency can be best applied when the investor believes that the project itself finances some or all of future project investments, in contrast to the case when it is felt that all future investments must go through the budgeting process and compete with other projects for investment capital. In the case of smaller projects that generate cash rapidly, the investment efficiency parameter may be the best one. In cases where major capital expenditures occur over several years, however, the profit to investment ratio may be more representative of the actual process. Regardless of w h i c h p a r a m e t e r y o u u s e , be consistent!
Although these are highly recommended yardsticks, they do have some limitations since they do not express the magnitude of cash flows or the chance of success. Table 3
50
PART 2—ECONOMICS AND RISK ASSESSMENT
Table 3. Comparison of Discounted Profit to Investment Ratio and Investment Efficiency Ratio
Time
0 1 2 3
NOI
600 800 1100
Invest.
1000 1000 200
0
NCF
-1000 -400 600 1100
Neg. NCFs
-1000 -400
PV Factor @ 4%
1.0000 0.9615 0.9246 0.8890 NPV
PV of Neg. NCFs
-1000.0 -384.6 0.0 0.0
-1384.6
PV of Invest.
1000.0 961.5 184.9
0.0 2146.4
PV of NCFs
-1000.0 -384.6 554.8 977.9 148.1
Cum. Disc. NCF (2)4%
-1000.0 -1384.6
-829.8 148.1
Investment efficiency ratio: PV of net cash flow
Investment efficiency = PV of maximum negative cash flow
•
\ r
1 •••
I1
'400
600
1 -1 • • i i
t
2
3
InvestmentefficiencJ y=
^ ^ =0.11 1384.6
Discounted profit to investment ratio:
P PV of net cash flow PV of investments
O r -IOOO
6 DO i i
1 1 r -IOOO
a DO J L
11 • • i
r 2 2(DO
P 148.1 = 0.07
I 2146.4
shows a simple example demonstrating the difference in the two economic yardsticks.
The investment efficiency for our example development well is $1,015,600/$1,375,000 = 0.74, whereas the investment efficiency for the extension project is $2,729,760/$2,429,333 = 1.12. Both ratios are based on a 4% discount rate.
EXPECTED NET PRESENT VALUE AND ENPV TO EXPECTED INVESTMENT RATIO
The expected net present value (ENPV) and the expected net present value to expected investment ratio are t w o highly recommended economic parameters for comparing and ranking projects, primarily because these concepts take into account the uncertainty and help in predicting outcomes of program inventories. They are particularly useful as exploration measures and less so for development projects.
The ENPV considers the net after-tax monetary value of the venture over the full life of the project (gross revenues minus capital investments and costs) discounted at the corporate discount rate. It incorporates current views on future wellhead prices, costs, and inflation rates and also takes into account the probabilities of success and failure as well as the cost of failure. Thus, it is a "risked" value and is the amount the company could expect to make on average if this prospect (or ones like it) could be undertaken many times.
If the development well modeled in Table 1 in the chapter
on "Building a Cash R o w Model" had been assigned an 80% chance of success by the geologist, the calculated ENPV (aftertax) at 4% would be $713,480 (0.80 x $1,015,600 - 0.20 x $495,000). Since this is a linear function, a plot of ENPV versus the probability of success results in a straight line. Assume two points (Ps = 1 and 0) to define the line. The intersection where the ENPV is zero will give the breakover point for the probability of success. Figure 3 demonstrates this concept. The breakover probability of success for the development well is approximately 33%.
If the extension project had been assigned a 70% chance of success by the geologist, the calculated ENPV (after-tax) at 4% would be $1,613,832 (0.70 x $2,729,760 - 0.30 x $990,000). The breakover probability of success for the extension project is approximately 27% (see Figure 3).
The expected net present value is useful in comparing large or complex ventures, as well as projects having different discovery probabilities and reserves potential. The parameter is useful in prospect inventories and for program planning a n d b u d g e t i n g if e s t i m a t e s of r e s e r v e s a n d d i s c o v e r y probability are reliable. The expected net present value to expected investment ratio is also useful in capital investment decisions when capital is limited. N e w e n d o r p (1975) demonstrates the use of this investment efficiency ratio.
Again, a final cautionary note: do not carry out economic analyses—that is, cash flow models—on expected (or "risked") reserves. The risking is done only on the net present value of the "success case"!
Figure 3. Expected value profile plot. Expected value is plotted versus probability of success example for development well and multiwell extension project.
Key Economic Parameters
Dealing with Risk Aversion
Peter R. Rose
Telegraph Exploration, Inc. Austin, Texas, U.S.A.
INTRODUCTION
Please refer back to the coin-tossing game described at the b e g i n n i n g of the c h a p t e r on " U n c e r t a i n t i e s I m p a c t i n g Reserves, Revenue, and Costs" (in Part 2). Most people would not be willing to pay $10,000 for one chance to play the game. Many would be willing to risk perhaps $1,000-5,000. Some risk-prone or affluent players might pay as much as $9,500, whereas some highly conservative or nonaffluent players might not be willing to hazard even a small loss on the gamble.
This illustrates the subtle, variable, but powerful human attribute called risk aversion. It is the proportion by which the expected value (EV) of a v e n t u r e is d i s c o u n t e d by the investor. Stated simply, risk aversion expresses the common h u m a n reaction to risk propositions, that the displeasure associated with losing a certain amount of money exceeds the pleasure associated with winning the same amount, that is, "it hurts worse to lose than it feels good to win." Although many psychological forces influence risk aversion, the most important ones have to do with the size of the bankroll (or budget), the magnitude of potential gain (or loss), and the chance of success or failure. Risk aversion makes small and large firms respond differently to the same venture.
People are usually inconsistent when making risk decisions, and their inconsistencies can affect development decisions as well as exploration decisions. Table 1 lists the more common risk biases. For example, some production departments in major companies are so risk averse that they transfer development wells deemed to have less than a 90% chance of success to the exploration department as so-called in-field wildcats! The result is to retard reserve development significantly because such wells must then compete with legitimate large-reserve exploratory projects, and the margins of developing oil fields do not receive timely evaluation.
DEVELOPMENT VERSUS EXPLORATION RISK AVERSION
In contrast to exploration ventures, which are commonly perceived to be "high risk" propositions, development and enhanced oil recovery (EOR) projects are sometimes seen as "low risk" ventures by management. Since all projects— exploration and development—are competing for limited funds, the risk assessment needs to be consistent and reliable for all projects so a fair comparison can be made. Often development (including EOR) projects require significant capital expenditures in relation to anticipated revenues. In these cases especially, it is important to assess the uncertainties involved correctly. For example, the uncertainties in process efficiency, development time and costs, and future prices are commonly greater than we care to admit, and they may negatively impact project profitability
and the chance of commercial success. The geologist and the engineer must team up to improve the characterization of the r e s e r v o i r and the a s s e s s m e n t of the u n c e r t a i n t y in development and EOR projects. Geologists likely have more experience in estimating uncertainty, whereas the engineer is likely better at quantifying a process. The strengths of the two disciplines must be combined. This will hopefully minimize any bias that m a y exist in assessing the riskiness of development and EOR projects compared to exploration projects.
Properly done, all projects should be put on the same playing field by good, unbiased risk analysis. If this is done, then all project cash flows can be discounted at the same discount rate (the real rate of return) and projects can be compared on the basis of the expected net present value. We are probably a long way away from achieving this goal, but we need to head in this direction. The problem is simply that for EOR and development projects, the risks are different, and nongeological risks often go unrecognized and/or underestimated.
COMMON INDUSTRY MEASURES TO REDUCE RISK
Equations are available to quantify the risk-adjusted value (RAV) of any given venture to a given firm, if the firm's risk aversion (or preference function) is known. (Cozzolino, 1977), Grayson (1960), and Walls (1989) provide methods for determining it. The risk-adjusted value (RAV) is calculated as follows:
RAV = — ln[p x e~r(R~C) +(l-p)x e~'c ]
(1)
where
R - gross reward (in millions of dollars) C = cost (in millions of dollars) p - probability of success r = risk-aversion function (in millionths)
Note that this equation combines both expected value and risk aversion (after Cozzolino, 1977). This equation is especially useful to small companies seeking the proper investment share among several ventures in a program.
Arps and Arps (1974) provided a graphical method for determining whether the risk of a given exploration venture is acceptable, if the cost, reward, and chance of success are k n o w n and if the firm has chosen a probabilistic "safety factor" (that is, 95%) for its overall program. The Arps and Arps method is not applicable to most development projects, except for very small firms.
However, the petroleum industry has traditionally dealt with risk aversion through more pragmatic business methods,
52
Dealing with Risk Aversion
53
Table 1. Biases Affecting Risk Decisions
Type of Bias Framing effects Existence of prior account Maintaining a consistent frame of reference Probability of success
Wrong action versus inaction Number of people making the decision Workload and venture size Personal familiarity
Common Example
Decision makers will take a greater gamble to avoid a loss than to make an equal gain
Decision makers are more inclined to take a risk at the beginning of a project than later in the project's life
Decision makers are more likely to invest during a run of good fortune and less likely to invest during a run of bad fortune
A venture having a perceived high chance of success is preferred over a second venture having a low chance of success, even though the expected value of the second venture is clearly superior
Managers avoid criticism by not making a decision rather than taking action that could result in the same loss
Groups are more prone to take risks than are individuals Large volume ventures are preferred over smaller ones, especially when
decision makers are busy The "comfort bias": decision makers are more risk prone in deals or
environments with which they have had past good experience
Source: Modified from Tversky and Kahneman (1981) and Rose (1987).
including the following:
1. Farming-out leased acreage in exchange for a drilling commitment
2. Making bottom-hole or dry-hole contributions on nearby competitor acreage
3. Obtaining a share of a given venture at an especially favorable price
4. Acquiring a legal option allowing subsequent enlargement of interest in the event of early success
5. Promoting one's partners as a result of obtaining an early favorable lease position, having unique technical capabilities, or their legal or financial disadvantage
6. Reducing one's share in the venture 7. "Going nonconsent," which means not participating in
an early part of a venture only to come back in later under a penalty, usually severe
RISK AVERSION: PRACTICAL CONSIDERATIONS
Risk aversion is a universal—but not uniform—attribute of human beings. Although the phenomenon is rooted in concern for safety, its cost—in keeping the firm out of highrisk, high potential ventures—is not generally acknowledged or even recognized. Companies that are strongly risk averse pay a high price for their conservatism. Here are some points to
remember: 1. Geotechnical personnel must operate using the risk preferences of the management or client, not their own personal risk preferences. For example, if a firm is seeking only near-certain development well opportunities, the development geologist should be cautious in generating risky edge wells or new zone ventures. 2. Constant use of the expected value concept as a benchmark is of considerable help. 3. Risk aversion commonly operates when decision makers fear criticism or loss of position. Here the most common consequence is that the firm will not participate in highly imaginative or controversial projects, many of which have substantial reserve potential, preferring instead those that are "orthodox" in concept and execution—and commonly mediocre in results! 4. Hidden hurdles—overly conservative parameters within the economic evaluation process and ordinarily not readily perceived by geotechnical professionals—tend to screen out many projects from consideration, often for the wrong reasons and often counterproductively. Good examples of hidden hurdles include (a) unrealistically high DCFROR or corporate discount rates, (b) overly cautious cost estimates, and (c) overly pessimistic wellhead price forecasts.
5. Remember the common biases that cause decision makers to behave inconsistently.
Economics of Property Acquisitions
Peter R. Rose
Telegraph Exploration, Inc. Austin, Texas, U.S.A.
ESTIMATING PROPERTY VALUE
The basic concepts and principles discussed in Part 2 are as applicable to evaluating producing properties for purchase as they are to oil and gas drilling ventures or development projects. As before, it is assumed that the data and estimates are well founded. For property acquisitions, however, even more importance is placed on rate, cost, and price estimates, as well as on process efficiencies if subsequent enhanced oil recovery and in-fill p r o g r a m s are c o n t e m p l a t e d . If the property is also perceived as having exploration potential, geological uncertainties (reserves and chance of success) must be considered as well for that component of the property value.
A common procedure for evaluating such proposed acquisitions is to establish the present value of the future cash flow stream. This should be based on very careful geotechnical and operational evaluation and discounted at the purchaser's corporate discount rate. Frequently, the next step is to recalculate the present value of the p r o p e r t y at an artificial discount rate—perhaps 50 to 100% higher— and attempt to acquire the property for that value. This is an e x a m p l e of the p r a g m a t i c use of d i s c o u n t e d cash flow analysis and the discount rate as a "risking measure." The logic is that if the p r o p e r t y is w o r t h , say, $2 million discounted at 10% and $1 million discounted at 18%, then purchase at about the $1 million figure should give an adequate "cushion" to protect the buyer from unanticipated negative surprises in the future, while still returning at least 10% on the purchase price. Although such methodology is commonly employed, it should be recognized as pragmatic in the extreme, inasmuch as no relationship exists between the discount rate and chance of commercial success.
An alternative approach is to use the expected value concept. After making very careful probabilistic assessments of remaining reserves, rates, costs, and prices, project present value is estimated at the firm's chosen discount rate at probabilistic levels (P90%/ P50%, and P10%). Then the chance of minimal acceptable commercial profitability is used to define the chance of commercial success and the chance of failure (as described in the chapter on "Expected Value and Chance of Success"). This allows determination of the expected value of the proposed acquisition and facilitates equitable comparison and capital allocation among prospects that are competing for consideration by the firm. Be sure to include the purchase price in the expected value calculations.
This approach also allows the decision maker to assess such proposed acquisitions in terms of the chance of returning various profit levels (and also various loss levels). Finally, use of different possible purchase prices in the expected value calculation is useful in identifying the appropriate bid amount.
Remaining reserves are calculated as follows:
Reservesremaining = (Annual production - Annual production
rate at abandonment) Annual decline rate
This equation is a "quick and dirty" method only, allowing rapid estimation of remaining reserves of a producing well, lease, or field, and it can be useful in early assessments and negotiations. A comprehensive coverage of p r o d u c t i o n forecasting and estimating reserves from production data is presented by Thompson and Wright (1985).
BID STRATEGY
Sales of producing properties are often well advertised, attracting several potential purchasers, and are commonly carried out by sealed bidding. It is important to recognize that sealed bid sales of oil and gas properties contain several inherent pitfalls:
1. Because reserves and producing rates are lognormally distributed, independent present value estimates of the same property will also tend to be lognormally distributed. This phenomenon is well documented for offshore lease sales (Capen et al., 1971; Megill, 1984). Accordingly, there will tend to be a larger numerical differential between the first and second bids than between the lowest and second-lowest bids. This leads naturally to large overbids or "leaving money on the table" as an inherent byproduct of the mathematics.
2. Geotechnical forecasts of reserves, producing rates, and so on are estimates made under uncertainty. Hence, both overestimates as well as underestimates are likely to occur in a group of independent estimates of the same property. Final bid levels are most influenced by estimates of reserves and rates. In sealed bidding, the property goes to the highest bidder. Accordingly, there is a marked tendency for winners of sealed bid sales to have overestimated the present value of the property— usually by overestimating reserves or rates. This is called the winner's curse. Moreover, because of lognormality, the amount of money left on the table by the winner is frequently substantial. This pattern is most significant for exploration bidding, but it should still be taken into account in sealed bidding for producing properties.
3. The result is commonly that acquisitions are substantially less profitable than the purchaser has anticipated. And when the previously discussed unanticipated geotechnical, process, and economic risks
54
Economics of Property Acquisitions
55
Table 1. Comparison of Calculated Bid Levels
Assumptions
Expected present value calculation Bid strategy method Recalculation of expected present
value (including recommended bid) Conventional method (using PV as
a risking measure)
Corporate discount rate = 10% Minimum commercial PV @ 10% = $1,000,000 Chance of success (achieve minimum commercial PV or more) = 80% Mean present value of all success scenarios @ 10% = $2,000,000 Chance of failure (achieve less than minimum commercial PV) = 20% Mean present value of all failure scenarios @ 10% = -$400,000 Mean present value of all success scenarios @ 18% = $1,000,000 Mean present value of all success scenarios @ 20% = $800,000 0.8($2,000,000) + 0.2(-$400,000) = +$1,520,000 = EPV10O/o
$1,520,000 x 0.5 = $760,000 = Recommended bid
0.8($2,000,000 - $760,000) + 0.2(-$400,000 - $760,000) = $760,000 = EPV10o/o
PV @ 10% = $2,000,000 PV @ 18% = $1,000,000 = conventional bid to allow for risk = $1,000,000 PV @ 20% = $800,000 = more prudent bid to guard against overestimating
= $800,000
are factored in, the danger is doubled. Technical assessors must never forget that the central goal is to make a sound profit on the purchase. The appropriate mind set to operate from is, "If we cannot acquire this property at our price, we do not want it!"
How can the prudent purchaser guard against overbidding? Megill (1984) shows that the average overbid in offshore continental shelf sales is about twice the size of the second bid, suggesting that a 50% reduction of calculated expected present value is an appropriate reduction. (Here it is important to note that the term expected present value as used here includes all geotechnical costs, but not the purchase price, which is what we are attempting to fix.) Capen et al. (1971) go farther, suggesting that for exploratory ventures having great uncertainty, sealed bonus bids should be reduced to 35% to 20% of expected present value, depending on the anticipated numbers of participants.
However, less uncertainty ordinarily attends producing properties, so perhaps such sealed bids could be reduced by 25% to 50%. Table 1 shows sample calculations and company bid levels determined by two methods: (1) a 50% reduction of the expected value (not including bid) versus (2) conventional procedures utilizing present values at artificially elevated
discount rates, as described at the beginning of this chapter. The problem with the latter approach is that it fails to
protect against substantial overestimates, and thus overbids, which are by no means uncommon! Accordingly, if some form of the artificial present value method is used and there is significant uncertainty as to ultimate recoverable reserves or if serious financial consequences would arise from a substantial overestimate, it is recommended that the artificially elevated discount rate can be as much as two times higher then the firm's actual corporate discount rate (Table 1).
A common reaction to these recommended large reductions in bid levels is, "We will never win a property with such low bids." But abundant experience demonstrates that this is simply not correct, assuming multiple bidding opportunities and exposures. Also, the reader is reminded that the objective is not just to acquire producing properties; the goal is to make a profit! It is interesting to note that the prevailing opinions a m o n g recent b u y e r s and sellers of producing properties sold by sealed bids is that the sellers are most often more satisfied than the buyers.
P e r h a p s this explains the proliferation since 1990 of auction sales of U.S. producing propeties, which minimize the winner's curse.
56
PART 2—ECONOMICS AND RISK ASSESSMENT
Part 2 References Cited
Arps7 J. Jv and J. L. and Arps, 1974, Prudent risk-taking: Journal of Petroleum Technology, v. 26, p. 711-715.
Capen, E. C., R. V. Clapp, and W. M. Campbell, 1971, Competitive bidding in high-risk situations: Journal of Petroleum Technology, v. 23, p. 641-653.
Capen, E. C., R. V. Clapp, and W. W. Phelps, 1976, Growth rate—a rate-of-return measure of investment efficiency: Journal of Petroleum Technology, v. 28, p. 531-543.
Capen, E. C., 1976, The difficulty of assessing uncertainty: Journal of Petroleum Technology, v. 28, p. 843-850.
Capen, E. C., 1984, Why lognormal? in E. C.Capen, R. E. Megill, and P. R. Rose, ed., Prospect Evaluation: AAPG Course Notes: Tulsa, OK, AAPG, 8 p.
Cozzolino, J. M., 1977, Management of oil and gas exploration risk: West Berlin, NJ, Cozzolino Associates.
Ernst and Young, 1990, Oil and gas federal income taxation, p. 158-163.
Grayson, C. J., 1960, Decisions under uncertainty: Cambridge, MA, Harvard University, Division of Research, Graduate School of Business Administration, 402 p.
Kaufman, G., 1963, Statistical decision and related techniques in oil and gas exploration: Englewood Cliffs, NJ, PrenticeHall, 307 p.
Krasts, A., and T. Henkel, 1977, Effect of inflation on discounted cash flow rates of return: Managerial Planning, N o v / D e c , p. 21-26.
Megill, R. E., 1984, An introduction to risk analysis, 2nd ed.: Tulsa, OK, PennWell Books, 274 p.
Megill, R. E., 1988, An introduction to exploration economics, 3rd ed.: Tulsa, OK, PennWell Books, 238 p.
Newendorp, P. D., 1975, Decision analysis for petroleum exploration: Tulsa, OK, PennWell Books, 668 p.
Rose, P. R., 1987, Dealing with risk and uncertainty in exploration—how can we improve?: AAPG Bulletin, v. 71, n. 1, p. 1-16.
Thompson, R. S., 1987, Impact of the new tax law on internal cash flow generation: Dallas, TX, 1987 SPE Hydrocarbon and Economics Symposium, SPE Paper 16309, p. 161-172.
Thompson, R. S., and J. D. Wright, 1985, Oil property evaluation, 2nd ed.: Golden, CO, Thompson-Wright Associates, 212 p.
Thompson, R. S., and J. D. Wright, 1992, Oil and gas property evaluation, 3rd ed.: Golden, CO, Thompson-Wright Associates, in prep.
Tversky, A., and D. Kahneman, 1981, The framing of decisions and the psychology of choice: Science, v. 211, p. 453-458.
Walls, M. R., 1989, Assessing the corporate utility function—a model for the oil and gas exploration firm: South Texas Geological Society Bulletin, Dec., p. 13-27.
Part3
WELLSITE METHODS
edited by
Arnold M. Woods
Conoco Inc. Casper, Wyoming, U.S.A.
Byram Reed
BP Exploration Bogota, Colombia
Diana Morton-Thompson
Consultant Kalamazoo, Michigan, U.S.A.
Contents
• Introduction • Well Planning • Land Rigs • Offshore Rigs • RigPersonnel • WellsiteSafety • WellboreTrajectory • DrillingFluid • Pressure Detection • Fishing • Drilling Problems • Measurement While Drilling • RateofPenetration • Wellsite Math • Mudlogging Equipment, Services, and Personnel • Mudlogging: TheMudlog • Mudlogging: Drill Cuttings Analysis • Mudlogging: Gas Extraction and Monitoring • Show Evaluation • ConventionalCoring • Sidewall Coring • CoreOrientation • Core Handling • Core Alteration and Preservation • DrillStemTesting • ReferencesCited
Introduction
Arnold M. Woods
Conoco Inc. Casper, Wyoming, U.S.A..
Byram Reed
BP Exploration Bogota, Colombia
Diana Morton-Thompson1
Consultant Kalamazoo, Michigan, U.S.A.
The purpose of this part of the Manual is to introduce the geoscientist to wellsite equipment and evaluation procedures that will be encountered while on location. The first half of Part 3 focuses on equipment and fundamental drilling operations, while the second half focuses on wellsite formation evaluation and acquisition of geological data.
Part 3, Wellsite Methods, begins with a discussion by Arnold W o o d s and Byram Reed of factors that m u s t be considered when planning to drill a well. Reed also provides overviews of onshore and offshore rigs and rig personnel. Elmo Eltzroth then d e s c r i b e s s o m e of the safety considerations peculiar to drilling operations. This is followed by a chapter on methods for establishing wellbore trajectory (vertical, deviated, and horizontal) by Curtis Cheatham and a chapter on drilling fluid systems by David Young. Parke Dickey discusses the problems associated with drilling in over- and underpressure conditions, and Arnold Woods covers commonly used fishing tools and procedures. A chapter on other common drilling problems is given by PhylHs Loose.
The second half of Part 3 begins with a chapter on measurement while drilling (MWD) techniques by Mike Medeiros, followed by Scott Boone's chapter on the use of rate of penetration (ROP) as an evaluation tool. Wellsite math used in various on-site calculations is outlined and explained by Greg Dunn. Mudlogging is then covered by several papers: a description of the equipment, services, and personnel, as well as the mudlog itself, by Alun Whittaker; drill cuttings analysis by Whittaker and Diana MortonThompson; and gas extraction and monitoring, also by Whittaker. Show evaluation is discussed by Paul Daniels, David Finnell, a n d William A n d e r s o n . T e c h n i q u e s of conventional coring (including sleeved, sponge, and pressure core) and sidewall coring (percussion and rotary) are described by Lee Whitebay, and core orientation is discussed by Douglas Bleakly. Core handling procedures are covered
by Byram Reed, and core alteration and preservation techniques are described by Caroline Bajsarowicz. Part 3 concludes with an overview of drill stem testing techniques and pitfalls by Ingrid Borah.
Acknowledgments
We would like to thank all the authors and the outside reviewers (Mike Taylor, BP, and Ed Banaszek, Exlog) for their contributions and their patience while this part was assembled. The individual authors would also like to acknowledge their respective companies for permission to work on and publish the various papers in Part 3:
William Anderson Carolina J. Bajsarowicz Douglas Bleakly Scott Boone lngrid Borah Curtis Cheatham Paul Daniels, Jr. Parke Dickey Greg Dunn EIrno Eltzroth
David Finnell Phyllis Loose Mike Medeiros Diana Morton-Thompson
Byram Reed Lee Whitebay Alun Whittaker Arnold M. Woods David Young
Epoch Well logging, Inc. BP Exploration Versar IDL Logging Conoco Sperry-Sun Consultant Consultant Tecltnical Drilling Services Michigan Department of
Natural Resources Epoch Well Logging, Inc. Conoco SheU Chevron U.S.A. and
ARCO Research BP Exploration Conoco Consultant Conoco BP Research
1Formerly with Chevron U.S.A. and ARCO Research. 59
Well Planning
Arnold M. Woods
Conoco Inc. Casper, Wyoming, U.S.A.
Byram Reed
BP Exploration Bogota, Colombia
INTRODUCTION
Planning a well is an iterative process between geoscience and engineering staffs that involves frequent accessing of several databases and clear communication. Three basic areas need to be examined in planning a well:
1. What target(s) will be evaluated? 2. How will the well(s) be drilled to reach those target(s)? 3. How will the target zone(s) be evaluated?
For the development geologist, well planning includes v e r i f y i n g n u m e r o u s p o i n t s u n d e r a v a r i e t y of g e n e r a l headings. This chapter provides a reasonably thorough checklist of those areas that should be investigated when planning a well.
WELLSITE DATA PACKAGE
A data package should be prepared for use on the wellsite c o n t a i n i n g all of the d a t a n e e d e d for correlation a n d evaluation. This includes maps, offset well logs, the well prognosis, and any other data that the geologist believes may be useful. Only copies (not the originals) of these data should be taken to the field.
WELL PLANNING CHECKLIST
Lease Status • Ownership • Limitations (e.g., oil rights down to a certain formation; gas rights below a certain level) • Nearby acreage availability (to extend play) • Well permit filed with state? • Survey (lease lines, ground elevation) • "Legal" well location
Geological and Economic Justification • Anticipated pay zone(s) • Anticipated pay type(s) (e.g., oil, gas, CO2) • Economic reserves • Economic scenario(s)
Geological Description for Each Zone • Name • Age • Depth
• Thickness • Structure • Lithology • Secondary and exotic minerals • Porosity and permeability
Environmental Concerns
• Weather • Topography (e.g., is proposed location in a streambed?) • Environmental Impact Statement (EIS) needed? • Cultural hazards and considerations • Accessibility • Safety hazards (training needed?)
Drilling Program
• Straight, slanted, or horizontal well • Anticipated total depth (TD) in measured depth (MD)
and TVD (true vertical depth) • H2S and other gases • Over- and underpressured zones • Casingpoints • Drillingproblems
• Sloughingshales • Swellingshales • Saltbeds • Faults • Highanglebeds • Lost circulation zones (e.g., fractures, excessive mud
weight)
Communications for Operators and Partners
• Contact list with addresses and phone numbers • Reporting requirements for partners
Data Collection
• Mud logging • Crew size and services to be provided • Number of sets of each type of sample - wet - dry - geochemical - paleontological • Pressurized unit • Hot wire or flame ionization detection (FID) chromatograph • Bitdullingplot • Overpressureplot • Rate of penetration (ROP), weight on bit (WOB),
60
pump pressure (PP)7 revolutions per minute (RPM), and bit data • Mud data • Pit volume monitors • Specialanalyses - ion tracing - radioactive monitoring - shale densities - d and dc exponent - hydrogen sulfide (H2S) detectors • CRT displays on rig floor, company representative's office • Electriclogging • Big, regular, or slim hole tools • Temperature requirements • Fresh, salt, or oil mud tools • Tool combinations and order of logging runs (usually gamma ray, deep resistivity, porosity, sonic, pad-type resistivity, dipmeter, checkshot survey, and sidewall cores) • Backup tools • Calibration, both at surface (before and after logging) and downhole • Repeat and overlap sections • Scales • Check shot survey points (determined after first logging run) • Bottom hole temperature (BHT) with each tool • Maximum logging speeds • Magnetic tape requirements (e.g., 800 or 1600 bpi) • Display format • Ruid samples • Repeat formation test (RFT) and drill stem test (DST) - Interval(s) - Packer seats - Sampling method • Coring • Conventional, wireline, and/or sidewall - Formations automatically cored? - Core on show? - Core handling procedures
Well Planning
61
Data Distribution
• Types of data • Drilling records (e.g., bit record, geolograph charts) • Mudlogs and mudlog records • Electriclogs • Core and/or cuttings (wet and/or dry) • Fluid samples • Pressurecharts • Laboratoryanalyses - cuttings - cores - biostratigraphic - geochemical - fluid properties
• Number of copies of each type of data for • Divisionoffice • Headquartersoffice • Partners • Federal and state agencies • Wellsitecopies • Geologist and geophysicist • Drillingengineer • Reservoirengineer
Sources of Data for Well Planning
• Offsetwells • wireline logs • mudlogs • daily drilling reports • velocity surveys • bit records • scout tickets • paleontology and geochemistry reports • production data
• State agencies (e.g., Railroad Commission in Texas) • Industry libraries (e.g., Petroleum Iiiformation [PI],
Dwights) • Servicecompanies • Literature search
Land Rigs
Byram Reed
BP Exploration Bogota, Colombia
INTRODUCTION
The parts of the rig can be grouped into five systems (Figure 1):
• Power • Rotating • Hoisting • Circulating • Control and measurement
POWER SYSTEM
Power is provided to the rig by diesel engines, diesel-electric engines, or in some cases, butane engines. Power is transferred from the engines to the different rig systems by belts, chains, and drive shafts on a mechanical rig, or by generated DC electrical power on an electric rig. Power is distributed to the rotary table and mud pumps while drilling and to the drawworks when tripping.
ROTATING SYSTEM
The rotating system consists of the rotary table and the drill stem (kelly, drill string, and bit). The rotary table is a square hole in the derrick floor with a rotary bushing that is used to turn the kelly bushing and kelly. The kelly is a square or hexagonal length of pipe that is screwed on the drill pipe and used to convey the rotary movement to the drill string and bit. The drill string refers to the combination of drill pipe, collars, and other bottom hole assembly components. (For more information on bottom hole assemblies, see chapter on "Wellbore Trajectory" in Part 3.) Attached to the end of the drill collars is the bit. The bit does the actual grinding or cutting of the rock. The style of bit used is dependent on the rock type and drilling conditions. Common bit types include drag, tri-cone, insert, PDC, and diamond.
The hole is drilled by adding joints or lengths of drill pipe to the end of the kelly. When pipe is added, the hoisting system is used to pick up the kelly so that it hangs from the derrick above the rotary table. Tongs (large pipe wrenches) or chains are used to unscrew the kelly from the previous joint of pipe. The kelly is then screwed into a new joint of pipe that has been temporarily stored in the mouse hole, a cased opening in the rig floor. The kelly and the new joint of pipe are then screwed on the previous pipe and lowered into the hole so that drilling can resume. When all of the pipe is pulled out of the hole, it is referred to as a trip. Pulling out part of the drill string, then returning to drilling, is called a short or wiper trip. Such trips are performed to verify that the drill string can move through a recently drilled potentially troublesome section of the borehole.
HOISTING SYSTEM
The hoisting system includes the parts of the rig that are used to raise the drill stem. The hoisting gear parts include the drawworks, crown block, and traveling block. The drawworks is a large winch on which the drill line spools. The drill line is wire rope that is strung between the crown block (a pully located at the top of the derrick), the traveling block, and the drawworks. The drill line can be strung in multiples for a total of of 4, 6, 8,10, or 12 lines. More lines means more lifting capacity but a slower running speed. The drill line needs to be "cut and slipped" at periodic ton-miles to distribute the line wear and stress. A weight indicator is attached to the drill line so that the driller can measure the drill string, slackoff, and pickup weights. This information helps determine the amount of hole friction and the correct amount of weight to put on the bit. The d r a w w o r k s also transfers power to make up and break out the drill string via the tongs.
The derrick supports the crown block and provides a place to stack pipe that is pulled out of the hole. Tlie depth rating of the derrick is related to the size of the rig. The height of the derrick is commonly referred to in multiples of pipe joints (a joint of pipe is approximately 30-ft long). Rigs that can stack double joints of drill pipe are called double derricks, and those that stack three joints are called treble derricks. Part way up the derrick are the monkey board and pipe fingers. The derrickman handles the top end of each stand of pipe from the monkey board during trips. The pipe is racked in the finger boards and tied off to keep it from falling. The derrick substructure, the platform under the derrick, is rated by set back capacity; that is, the weight of the drill string stacked in the derrick plus the weight of casing that can be lifted. The height of the s u b s t r u c t u r e is dictated by the height of the b l o w o u t preventers. The top of the substructure is called the derrick floor. This is the primary working area of the rig.
The catwalk is the deck located to the side of the derrick floor and between the pipe racks. Joints of drill pipe and casing are rolled from the pipe racks where they are stored on the catwalk and hoisted up through the slide and V-door. The catwalk is also the primary location to process core and assemble wireline logging tools.
CIRCULATING SYSTEM
Circulation of drilling fluid (mud) serves several functions on a rig, including cooling the bit, providing hole stability, and aiding in formation evaluation. (For more information on drilling fluid, see the chapter on "Drilling Fluid" in Part 3, and for more on how the circulating system aids in formation evaluation, see the chapter on "Mudlogging: Drill Cuttings Analysis" also in Part 3.)
62
Figure 1. Components of a typical land rig: (1) crown block, (2) mast, (3) monkey board, (4) traveling block, (5) hook, (6) swivel, (7) elevators, (8) kelly, (9) kelly bushing, (10) master
bushing, (11) mouse hole, (12) rathole, (13) backup tongs, (14) makeup tongs, (15) drawworks, (16) weight indicator, (17) driller's console, (18) dog house, (19) rotary hose, (20)
accumulator unit, (21) pipe ramp, (22) pipe rack, (23) substructure, (24) mud return line, (25) shale shaker, (26) choke manifold, (27) mud-gas separator, (28) degasser, (29) reserve pit,
cS
(30) mud pits, (31) desilter, (32) desander, (33) centrifuge, (34) mud pumps, (35) dry mud components storage, (36) water storage, (37) engines and generators, and (38) blowout
preventor stack. (From IHRDC.)
64
PART 3—WELLS1TE METHODS
Drilling fluid is circulated by the mud pumps. The volume of mud being pumped is measured by the stroke counters, and the rate of movement is recorded by the stand pipe pressure. The stand pipe connects the mud pumps to the kelly hose. The kelly hose is connected to the swivel on top of the kelly. Mud is pumped down the drill string through the bit and up the annulus or 'Ъаск side" (the space between the drill pipe and the borehole). Returning mud flows down the flowline into a surge tank (possum belly) and across the shale shakers. Shale shakers are vibrating screening devices that are designed to shake so as to separate out the drill cuttings from the mud. The shale shakers are the first place that drill cuttings can be examined and where the gas is extracted from the mud (Figure 2). After going through the shake shakers the mud passes through a series of tanks or pits where the finer solids are removed via desanders, desilters and centrifuges, and the mud properties are adjusted. Pits are named for their function (e.g., shale pit, settling pit, volume pit, mixing pit, and suction pit). The mud pumps are charged from the suction pit. Excess mud can also be diverted from the metal mud pits into a large, plastic lined reserve pit located to the side of the rig.
CONTROL AND MEASUREMENT SYSTEM
The blowout preventers (BOPs) are the major component of the control system on a rig, and they are the last line of defence against a blowout. The BOPs are bolted to the wellhead and are not removed until the well is completed and production equipment is installed. BOPs usually have at least four sections:
• Aiuiulars are large, liard rubber slips that fit around any sized pipe. The size flexibility, however, is at the expense of the pressure rating.
• Pipe ranis are metal donut-shaped sealing mechanisms that fit only a specific sized pipe but at a high pressure rating.
• Blind or shear rams are edged high-carbon steel sealing mechanisms that can cut pipe and close the hole completely.
• Crossover spools are metal junctions where the choke and kill lines attach.
Additional control equipment includes the kill line, chokc
Figure 2. Circulating system of a rig. (From Whittaker, 1985.)
line, and flare line. The kill line is used to p u m p mud into the annulus at the crossover spool in the event that heavier mud is needed to control wellbore pressures. The choke line also helps control wellbore pressures by allowing drilling mud to circulate through a choke manifold (which is a set of backpressure valves). The mud and gas can then be sent down the flare line for disposal or burning in the reserve pit.
Monitoring and measurement of the basic functions of the rig are usually made from the driller's console located on the derrick floor. From this console, the driller can monitor equipment, distribute power, change gears, and oversee crew activities. Additional measurement equipment and records are located in the dog house or service shed. The dog house is located to the side of the rig floor and contains the geolograph, a device that makes a time-based chart record of several rig functions, including the following:
• Kelly height or rate of penetration (ROP) (in ft/hr, m/hr, or min/ft)
• Depth (in ft or m) • Pump pressure (in psi) • String weight (in thousands of lb) • Rotary speed (in rpm) • Rotary torque (in ft-lb) • Stroke rate (in strokes/min) • Circulation rate (in gal/min)
Offshore Rigs
Byram Reed
BP Exploration Bogota, Colombia
INTRODUCTION
Offshore rigs are similar to land rigs but with several additional features to adapt them to the marine environment (see previous chapter on "Land Rigs"). Those features include
• Heliport • Living quarters • Cranes • Risers
The heliport, also known as the helipad, is a large deck area that is placed high and to the side of offshore rigs. It is an important feature since helicopters are often the primary means of transportation. The living quarters usually comprise bedrooms, a dining hall, a recreation room, office space, and an infirmary. Escape boats are usually located near the living quarters.
Cranes are used to move equipment and material from work boats onto the rig and to shift the loads around on the rig. Most rigs have more than one crane to ensure that all areas are accessible. A riser is used to extend the wellhead from the mudline to the surface. On platforms and jackup rigs, the blowout preventors (BOPs) are mounted above sea level. On floaters, the BOPs are mounted on the seafloor.
TYPES OF OFFSHORE RIGS
The v a r i o u s types of o f f s h o r e rigs include barges, submersibles, platforms, jackups, and floaters (the latter of which include semisubmersibles and drill ships).
beams. After all the wells are drilled, the rig and quarters are removed from the platform. Smaller platforms use a jackup rig to drill the wells.
Jackups
Jackups are similar to platforms except that the support legs are not permanently attached to the seafloor (Figure 2). The weight of the rig is sufficient to keep it on location. The rig's legs can be jacked down to drill and jacked up to move to a new location. When under tow, a floatation hull buoys the jackup. The derrick is cantilevered over the rear to fit over preset risers if necessary.
Floaters
Offshore rigs that are not attached to or resting on the ocean bottom are called floaters. These rigs can drill in water depths deeper than jackups or platforms can. They have several special features to facilitate this:
• They are held on location by anchors or dynamic positioning.
• The drill string and riser are isolated from wave motion by motion compensators.
• The wellheads and BOPs are on the ocean bottom and are connected to the rig by a riser to allow circulation of drilling mud.
Barges
A barge rig is designed to work in shallow water (less than 20 ft deep). The rig is floated to the drillsite, and the lower hull is sunk to rest on the sea bottom. The large surface area of the lower hull keeps the rig from sinking into the soft mud and provides a stable drilling platform.
Submersibles
A submersible rig is a barge that is designed to work in deeper water (to 50 ft deep). It has extensions that allow it to raise its upper hull above the water level.
Platforms
Platforms use a jacket (a steel tubular framework anchored to the ocean bottom) to support the surface production equipment, living quarters, and drilling rig (Figure 1). Multiple directional wells are drilled from the platform by using a rig with a movable substructure. The rig is positioned over preset wellheads by jacking across on skid
Figure 1. Fixed production platforms. (From Whittaker, 1985.)
65
66
PART 3—WELLS1TE METHODS
DRILLING SLOT
(a)
Figure 2. Jackup rigs. (From Whittaker, 1985.)
There are two categories of floaters: semisubmersibles and drill ships.
Semisubmersibles
Semisubmersibles (also called semi's) are usually anchored in place (Figure 3). Although a few semi's are self-propelled, most require towing. Because floaters are subject to wave motion, their drilling apparatus is located in the center where wave motion is minimal. Semi's are flooded to a drilling draft where the lower pontoons are below the active wave base, thereby stabilizing the motion.
Drill Ships
The drilling apparatus on a drill ship is mounted in the center of the ship over a moon pool, which is a reinforced hole in the bottom of the ship through which the drill string is raised and lowered (Figure 4). The ship can be turned into the oncoming wind or currents for better stability, and it can operate in water too deep for anchors.
(b)
Figure 3. Semisubmersible rigs, (a) Pontoon type semisubmersible. (b) Twin hull semisubmersible. (From Whittaker, 1985.)
Figure 4. Dynamic positioning drill ship. (From Whittaker, 1985.)
A
Rig Personnel
Byram Reed
BP Exploration Bogota, Colombia
BASIC RIG CREW
The type and number of rig personnel is related to the size and complexity of the rig. At the simplest level, a basic rig includes a toolpusher, a driller, a derrickman, a motorman, and a floorhand.
The toolpusher is the senior manager at the rig and is responsible for personnel, parts, and performance. The driller reports directly to the toolpusher and is responsible for the active drilling phase, trips, and the safety of the floor hands. Reporting to the driller are the derrickman, motorman (and lead tong), and floorhand.
The derrickman handles the top end of each stand of pipe during trips. When not tripping pipe, the derrickman is in charge of the circulating system. The motorman keeps the engines and other parts of the rig in working order and functions as the lead tong during trips. The floorhand handles the backup tong in trips and is apprenticed to the senior crew members.
Although the toolpusher stays at the rig 24 hr a day, the crew(s) usually commute and work 8- or 12-hr shifts or tours.
TOOLPUSHER DERRICKMAN
COMPANY MAN
MATERIALS MANAGER
ELECTRICIAN
CRANE OPERATOR
ROUSTABOUT CHIEF
ROUSTABOUT
HOUSEKEEPING
FIRST MATE
DYNAMIC POSITION
ADDITIONAL PERSONNEL
On large, complex rigs, additional personnel are required. A representive from the client company called the drilling foreman or company man is typically present on the rig. Tlie company representive is in charge of all drilling operations at the wellsite, including safety.
On offshore and remote land locations, rig personnel stay at the wellsite. Additional personnel that provide services such as welding and electrical work are also housed at the wellsite, along with personnel to house, feed, and care for the crew. An example of the general organization required to run a complex operation is given in Figure 1. Each of the positions shown represents at least two people (assuming 12hr tours), with a five- to six-person catering crew and additional seamen on offshore rigs.
CEMENTER
MUD ENGINEER
DIRECTIONAL
Figure 1. Organizational chart for a dynamically positioned drill ship.
Wellsite Safety
Elmore Eltzroth
Michigan Department of Natural Resources Geological Survey Division Lansing, Michigan, U.S.A.
INTRODUCTION
Disregard for safety at the wellsite can be the cause of injury or death. The wellsite is a hostile environment, often either very cold or very hot, wet, and slippery. Hazards include heavy equipment being moved around from all directions, heights, and industrial activities such as welding, the presence of flammable fluids, poisonous gases, machinery, vehicles, and noise.
PLANNING FOR SAFETY
Working efficiently and effectively in this environment calls for being a w a r e of the s u r r o u n d i n g s and p l a n n i n g accordingly. Before entering the drilling site, ask yourself the following questions:
1. How is the site laid out? 2. What are the current and potential hazards? 3. What protective equipment is needed? 4. What are the warning signals and emergency
procedures at this site? 5. What are the escape routes? 6. What training and survival courses are needed and
available?
GOVERNMENT SAFETY REGULATIONS
Safety at a wellsite is governed by federal and state regulations, as well as by the operator's in-house codes. Federal regulations are established in the Safety and Health Standards set forth by the Occupational Safety and Health Administration (OSHA, 1983). In many states, these regulations are administered by state agencies. State regulatory agencies can also have their own specific standards that are more stringent than OSHA requirements. A prudent operator should be aware of all levels of safety regulations.
RIG SAFETY EQUIPMENT
A drilling site is equipped with various safety features, including
• First aid kits • Fire extinguishers • Emergency air supplies • Warning sirens • Gas monitors • Wind indicators
The location of this equipment should be noted prior to an
emergency since emergency conditions can be noisy and can be made more dangerous by limited visibility and confusion.
OFFSHORE SAFETY
The offshore environment poses special safety issues. Transportation to and from the rig and location of the rig are issues that require special training in helicopter and boat safety and in procedures on how to get on and off the rig. When arriving at an offshore location, a safety briefing is given that includes the rules and alarms followed on the rig as well as the escape boat assignments. Each rig should have a weekly safety drill. This practice may include having personnel donning life jackets, air packs, or survival suits and assembling at the escape boat. Some drills include boarding the escape capsule and starting the engine.
PERSONAL SAFETY EQUIPMENT
Personal safety equipment is usually determined by the employer and must be in compliance with state and federal regulations. Required equipment includes the following:
• Hard hat • Safety shoes • Safety glasses • Hearing protection
The following are recommended:
• Well-fitted, protective clothing
• CPR and first aid training
Do not have the following:
• Long shoe strings • Floppy gloves • Neckties or fringes • Jewelry • Long, loose hair • A bad attitude In general, the minimal personal safety equipment includes a hard hat, steel-toed boots, and eye and ear protection (Michigan Department of Labor, 1989). In some cases, it may also include gas monitors and supplied air respirators.
Head Protection
Hard hats provide some protection from small falling objects—but only if worn. Plastic hard hats are generally
68
Wellsite Safety
69
used because they provide protection from electrical shock. Metal hard hats are used only in situations where heat or chemical reactions can deteriorate the plastic. A wool liner can be attached inside the crown of the hat for cold weather work.
Foot Protection
Steel-toed boots can provide a high degree of protection from dropped or rolling objects. Steel-toed boots are available in a variety of materials and styles. Generally a good leather boot with a nonskid sole is preferred. Boots can be made of waterproof materials and can be insulated if needed. There are even steeltoed "tennis shoes" available.
Eye Protection
Eye protection is either in the form of goggles or safety glasses. Safety glasses can be provided as prescription glasses and can be mounted in respirator face pieces if necessary. Slip-on side panels are available for side protection of the eyes and should be worn at all times when on any rig.
Ear Protection
Ear protection is recommended in areas where noise levels exceed 85 decibels (Michigan Department of Public Health, 1989). Two types of ear protection are available:
1. Protective muffs mounted on a hard hat 2. Small plugs made of soft plastic that are inserted into
the ear canal
Either style should be rated to reduce the noise level below 85 decibels.
Gas Monitors
Two types of wellsite gas monitors are used:
Figure 1. Typical self-contained breathing apparatus (SCBA). (Photo courtesy of Scott Aviation.)
1. Fixed monitors are installed at the wellsite in locations where gas might be expected to break out or accumulate (e.g., bell nipple, shale shaker, mud mixer, and rig floor).
2. Portable personal monitors are worn at any location where fixed monitors are not available and a toxic gas hazard exists. Personal monitors continuously monitor gas concentrations and provide a sound and light warning.
Air Respirators
Air respirators require specific training before they can be used in life-threatening environments. There are two general types of air respirators designed for personal use:
1. The 5-minute air tank with a face mask is designed to provide emergency escape from the site
2. The more familiar self-contained breathing apparatus (SCBA or air pack) is designed for short-term work in hazardous environments (Figure 1).
In using either system, a good face mask fit is essential. Because facial hair tends to break the seal, beards are often not
permitted on the wellsite. On drilling rigs where known hazards from poisonous gas exist, air lines are installed and workers plug their individual air hoses into the manifold system.
HAZARDS FROM FLAMMABLE GASES AND FLUIDS
Flammable gases and fluids are a prime concern on a drilling site, and smoking is not advised anywhere except in open or designated areas. Carbon dioxide, nitrogen, hydrogen sulfide, or any other gas can be a killer in a confined space if it displaces oxygen. Typical confined spaces at a wellsite include the rig cellar, mud tanks, and mud pits.
Hydrogen Sulfide
A special concern at many drilling locations is hydrogen sulfide. Hydrogen sulfide (H2S) is a colorless gas that has a "rotten egg" odor at concentrations below 1 ppm (part per million). At 10 to 20 ppm, protective steps against long-term
70
PART 3—WELLS1TE METHODS
exposure need to be taken to prevent worker discomfort, such as eye irritation. At concentrations above 20 ppm, the gas deadens the sense of smell and headaches or nausea may develop. At 600 ppm (or less for some people), the sense of smell is immediately paralyzed, breathing stops, and without immediate resuscitation, death follows (American National Standards Institute, 1972; American Petroleum Institute, 1974).
When entering a site with known or suspected H2S present, conditions should be determined immediately from site personnel or warning signs. When a well is drilling with a risk of H2S release, it is important to be continually aware of the wind direction, site layout, safety equipment, and the various means of escape.
Hydrogen sulfide is a heavier than air and will flow along the ground surface and collect in low spots. Clothes can also absorb H2S dissolved in water or oil. As the material dries or
warms (as in a heated vehicle, room, or dryer), the gas can be released, creating a dangerous situation.
OTHER COMMON HAZARDS
Additional common hazards can come from a number of sources:
• WildHfe • Slipping on water, ice, drilling muds, or lubricants • Falling from rig ladders • Tripping over the numerous pieces of equipment lying
around the rig
The old sailor's adage of "one hand for the ship and one for yourself" is good advice at any wellsite.
Wellbore Trajectory
Curtis Cheatham
Sperry-Sun Drilling Services Houston, Texas, U.S.A.
INTRODUCTION
Wellbore trajectory is controlled by the type of bottom hole assembly used and the weight on the bit. The bottom hole assembly, or BHA, is that portion of the drill string closest to the drill bit. It consists of several components:
• Heavy-weight drill pipe, which has the same outer diameter as regular drill pipe but with thicker walls for greater weight, used as a transition between drill collars and drill pipe.
• Drill collars, which are heavy, large diameter pipe located above the bit and below the heavy wall and used to apply weight to the bit.
• Stabilizers, which are short drill collars with larger diameter blades that are used to control contact with the borehole wall.
• Subs, which are devices used to connect various parts of the BHA.
There are two basic types of wellbore trajectories:
• Vertical or straight • Directional, including both deviated and horizontal
wellbore trajectories
VERTICAL OR STRAIGHT WELLBORE
A vertical hole is called a "straight" hole. However, some minor deviation from vertical often occurs naturally. This is related to formation properties, such as dip angle and hardness, and to other factors, such as the BHA, the bit design, and the weight on the bit.
Two types of BHAs are commonly used to drill a vertical hole: slick and pendulum. A slick BHA consists of a drill bit, drill collars, heavy-weight drill pipe, and regular drill pipe. The name slick is related to the absence of stabilizers. Slick BHAs have limited application due to their high potential for becoming differentially stuck. Square or spiral collars can be used in conjunction with slick BHAs when differential sticking is known to occur. In addition, slick BHAs can be r u n w h e n there is a risk of losing the BHA d u e to hole problems.
A pendulum BHA is p r o b a b l y the m o s t o f t e n u s e d assembly for drilling a vertical hole. A pendulum BHA is similar to a slick BHA, but contains one or more stabilizers (Figure 1). The closest stabilizer to the bit acts as a pendulum point. Gravity tends to force the bit to the "low side" of the hole, decreasing hole angle. Pendulum BHAs are run at a high RPM rate and a low weight-on-bit (WOB) rate in areas where deviation needs to be minimized.
DIRECTIONAL WELLBORE
Directional drilling refers to any method employed to hit a p r e d e t e r m i n e d s u b s u r f a c e target. O n e a p p l i c a t i o n of directional drilling is the development of offshore fields. Field development costs are reduced by directionally drilling many wells from one (or more) platforms. Other applications for directional drilling include the following:
• Building a surface location away from the bottom hole location to avoid cultural or topographic restrictions
• Sidetracking around a "fish" or lost open hole • Sidetracking out of casing for recompletions or
collapsed casing • Drilling a relief well to kill a blowout • Increasing contact between the reservoir and the
wellbore, (e.g., horizontal drilling) To hit a subsurface target, control must be exercised on both the angle of hole inclination from vertical (the drift or angle) and the azimuth angle (the direction). Wellbores have a tendency to move from left to right as the hole is drilled. This phenomenon, known as "walking to the right," is presumably due to right-hand rotation of the bit and drill string and is affected by inclination angle, rotary speed, weight on the bit, formation dip and strike, and bit design. Most directional wells are oriented to the left of the direction
Figure 1. Using a pendulum bottom hole assembly to drop angle.
71
72
PART 3—WELLS1TE METHODS
\ Target
ЛI \ ^ ^\ •II»\\\\ I \
II»» \\\\
\ \ ^ ^ T a r g e t Azimuth
»I> \\\
» \
\
\
\
\
\ \г
v v -
%\ \
*
\
\
\
\
\
\
\
\
\
4X \
\
\
\
\
Lead Angles^ Л
1
IN
Surface Location
к ^t
.Surlace Location
- * — Vertical
True Vertical Depth
^ M к ^fc-*
Kick OH Point Anglo Build Section End ol Build
^^^ ^^^
Angle Hold ("Tangent") y f Section
H
Horizontal Displacement ("Kick")
Total Depth H
Figure 3. Directional well design example: build and hold.
Figure 2. Right-hand walk and lead angle.
of the target azimuth by an amount known as the lead angle (Figure 2). By compensating for right-hand walk in this fashion, the wellbore is allowed to move naturally to the right, forming an arc into the target.
"Kicking Off" a Directional Well
In directional wells, the wellbore will be deviated at a preselected depth known as the kick-off point. An example of a directional well plan called "build and hold" is shown in Figure 3. A borehole inclination of at least 15° is desirable since it is harder to maintain directional control in holes with shallower deviation angles (Adams, 1985). However, wells with higher deviation angles can present operational problems (such as running wireline logs to total depth).
Whipstock Method
The oldest method of kicking off or deviating a wellbore uses an open hole whipstock (Figure 4), which is a casing joint w i t h an u p w a r d - t a p e r i n g w e d g e cut out of one side. Although this operation is time consuming, it is still occasionally used to sidetrack around fish or abandoned open holes. Whipstocks are also used to sidetrack out of casing by milling a "window" and deflecting the wellbore using a mud motor.
Jet Bit
Another kicking-off method uses a jet bit. The bit has one large nozzle that erodes a pocket from the hole bottom in the desired trajectory (Figure 5). Weight is applied to the bit while it is rotated into the pocket. This procedure is repeated until the desired trajectory is achieved. The disadvantage of jetting is that it is highly dependent on formation hardness. Some formations are too hard to be hydraulically eroded, and some soft formations erode too quickly, making it difficult to jet in the desired trajectory.
Bottom Hole Assemblages Used to Kick Off Wells
Two types of BHAs are used to kick off wells:
1. Rotary BHAs, in which the power to turn the bit is supplied by the rotary table
2. BHAs that use a downhole motor to provide bit power
There are two different downhole power sources: mud turbines and positive displacement mud motors. Both systems use the hydraulic energy of the mud to rotate the bit. Mud motors combined with a bend in the BHA are used to drill the well directionally. The bend is located in the motor housing (bent housing motor) (Figure 6a) or in a short sub (bent sub) directly behind the motor (Figure 6b). The purpose of the bend is to tilt the bit axis relative to the hole axis.
To change course, drilling stops, the bend in the BHA is oriented to the new borehole trajectory and the bit drills
Wellbore Trajectory
73
Figure 5. Kicking off with a jet bit.
Figure 4. Using an open hole whipstock for sidetracking.
ahead. This procedure is called sliding because the entire BHA above the motor is moving downhole without rotating.
This system originally had several limitations:
• Short bit life • Low power motors with low reliability • No means to monitor the progress of the wellbore
trajectory continually
Current technology has simplified directional drilling based on improvements in the following:
• Polycrystalline diamond compact (PDC) bits. • Improved motor design. • Reliable measurement while drilling (MWD) tools and
systems. This system allows the directional driller to monitor the azimuth and inclination of the borehole continously near the bit and make changes as required (see the chapter on "Measurement While Drilling" in Part 3). • Downhole adjustable bent subs, called steerable subs.
Steerable subs can be reset by changing the pump pressure. This changes the angle of the bent sub from straight to +1 The main advantage of steerable systems is that, after achieving the desired wellbore deflection, it is possible to continue drilling without tripping. If needed, changes in wellbore trajectory can be made at any time in very gradual steps, which reduces the probability of severe dog legs. The main disadvantage of steerable systems is that they are more expensive than other deflection systems.
Drilling a Directional Well
After the well has been kicked off, drilling proceeds either by maintaining the wellbore trajectory or by altering it as required to hit the target(s). Three types of rotary BHAs can be used in a directional well to drill ahead or to hold, drop, or build the inclination angle:
• Drill ahead—After the well has been kicked off, the entire drill string is rotated to drill ahead while maintaining the trajectory. An undergauge stabilizer is often included on the motor to reduce the tendency to drop the angle. Further corrections in trajectory are made by orienting and sliding.
• Hold angle—These require BHAs that are called packed because they contain many stabilizers. This tends to limit changes hi wellbore trajectory (Figure 7a).
• Drop angle—The pendulum BHA, discussed earlier, is also used for dropping the angle (see Figure 1).
• Build angle—This requires a near bit stabilizer to act as a fulcrum point. The bend in the drill collars above the near bit stabilizer causes the bit axis to tilt relative to the hole axis (Figure 7b).
HORIZONTAL WELLBORE
A horizontal well is a special type of directional well. At a predetermined depth, the well is kicked off and, the angle is built to a 90° inclination. The major reasons for horizontal drilling are
• Intersecting many vertical fractures in a single wellbore
74
PART 3—WELLS1TE METHODS
Measuroment-Whilo DrIIIIna Tool
Drill Collar —
Sncond StabiIiZFir
10' Pony Collar
(a)
Extonaion Sub
Near Bit Stabilizer
Near Bit
!
Stabilizer
Near Bit Stabilizer
Bit
Figure 6. (a) SteerabIe bottom hole assembly, (b) Kicking off with a bent sub and straight mud motor.
• Increasing production rate in low permeability formations
• Reducing water and gas coning problems • Increased ultimate recovery • Faster payout
(a)
(b)
Figure 7. (a) Packed bottom hole assembly, (b) Fulcrum effect for build angle.
Problems with horizontal wells include additional well costs and difficulties with formation evaluation, completion, and workover services.
Horizontal wells are classified as long, medium, or short radius, depending on the build rate from vertical to horizontal (Figure 8). As the build rate increases, the radius of curvature of the wellbore trajectory decreases. Long radius wells have smaller build rates and therefore reach a 90° inclination over a longer horizontal distance than short radius wells.
Long Radius
Long radius design methods are used primarily for achieving extended reach from platforms and in applications where large horizontal displacement is desired. These wells are really just conventional directional wells with final hole inclinations of 90°. Build rates typically range from 2° to 6° per 100 ft. More than 4000 ft of horizontal section can be drilled after reaching a 90° inclination.
BHAs used to provide wellbore trajectories for this type of well are similar to those used for conventional directional drilling. Steerable systems with bent housing motors are generally used for both the build and the horizontal sections. Slick BHAs with downhole motors are sometimes used to drill the horizontal portion of the well.
Short radius 1 - 3 ° per ft
Medium radius 8°-20° per 100 ft
Long radius 2 -6° per 100 ft
Figure 8. Build rates for classification of horizontal wells.
Medium Radius
Medium radius horizontal wells use similar methods and equipment developed for long radius wells, although sometimes in a slightly different fashion. Build rates range from 8° to 20° per 100 ft, and the maximum horizontal section drilled is presently greater than 4200 ft (Franco, 1990). When build rates of greater than approximately 15° per 100 ft are needed, BHAs with a double bend are used. A typical double bend BHA consists of a bent housing motor and a bent sub above the motor. Generally, this configuration is used only in a sliding mode since rotation can cause premature downhole failures due to the large cyclic bending stresses in the BHA.
Wellbore Trajectory
75
Short Radius
Short radius horizontal wells require different equipment than long or medium radius wells. Articulated ("wiggly") drill collars, a whipstock with retrievable packer assembly, and other special equipment are used. Build rates of 1° to 3° per ft are needed to change from vertical to horizontal within a 30- to 90-ft window. Typical horizontal sections extend 200 to 400 ft, with a record reach of more than 1200 ft. In special situations, pliable BHAs are used. These BHAs limit the reach of short radius wells but allow multiple drain holes to be drilled from the same vertical hole.
DrillingFluid
DaVidBYoung
BP Research Houston, Texas, U.S.A.
PURPOSE OF FLUIDS
An essential element of drilling a well is the drilling fluid or mud. Drilling fluids serve a number of functions:
• Removal of cuttings from the bottom of the hole • Suspend cuttings and weight material • Transport cuttings and gas to the surface • Cool and lubricate the bit and drill string • Add bouyancy to the drill string • Control subsurface pressures
The most important feature of any drilling fluid (or mud) system is that the interaction between the mud and the drilled formations must have a minimal effect on the mechanical properties of the formation. This is essential to maintaining an open hole and successfully completing the drilling operation.
PROPERTIES OF FLUIDS
The large number of functions performed by the drilling fluid require that some minimum properties of the fluids be maintained. The measurement of these properties gives the m u d engineer a "status report" of the fluid and how it is reacting with the formation and the subsurface environment. Tlie most critical of the properties are density, viscosity, fluid loss control, and chemical composition.
Density
The correct drilling fluid density is dependent on the subsurface formation pressures. Strong, competent formations can be drilled with a density less than 1.0, but overpressured shales and high pressure formations may require a fluid with specific gravities approaching 2.4. The density can be adjusted with soluble salts or by addition of solids, termed weight material (for example, barite is added to the mud to increase the density). Density values can be expressed as one of the following:
• ppg = pounds per gallon (United States) • S.G. = specific gravity (dimensionless) (international) • psi/ft = pounds per square inch per foot (uncommon) • pcf = pounds per cubic foot (California)
Table 1 summarizes how these different measurements of mud density compare with one another.
Viscosity
The flow properties of the mud depend on the depth of the hole and the annular viscosities. In the upper hole, water may be sufficient, but at greater depths more viscous fluids may be
required. Deep wells, directional wells, high penetration rates, high mud weights, and high temperature gradients create conditions requiring close attention to the flow properties. The viscosity can be adjusted upward with polymers or clay material or adjusted downward with chemical thinners or water.
Fluid Loss Control
The fluid loss gives a relative indication of how the mud is controlling loss of the base fluid into the formation. This becomes important when porous formations, particularly those containing oil or gas, are drilled. In porous formations, the drilling fluid may penetrate the rock and cause formation damage. (However, a low fluid loss does not always ensure minimal formation damage.) There are many types of fluid loss additives, such as bentonite, that can be used in the mud to help mitigate this problem.
Chemical Composition
Drilling fluids are two-phase compounds: a fluid and solid phase. The character of the fluid phase is determined by chemically analyzing the concentrations of calcium, chlorides, hydroxols, bicarbonate and carbonate ions, sodium, potassium, and nitrates. The character of the solid phase is tested to determine solids concentration, specific densities, and particle sizes. The primary means of controlling solids are by removal via shale shakers, desanders, desilters, and/or dilution.
TYPES OF FLUIDS
Drilling fluids include three main types: water-based muds, oil-based muds, and air. Air drilling fluids, such as mist, foams, and stiff foams, are used in only very specific, limited applications.
Water-Based Muds
Water-based drilling fluids are the most commonly used of the mud systems. They are generally less expensive and less difficult to maintain than oil muds, and in some special types of systems, they are almost as shale inhibitive. However, inevitably the action of drilling the hole in a consolidated formation relieves stress. If a water-based fluid is used, the water will tend to enter the formation and change the mechanical properties of the rock. These changes may be enough to cause formation damage and borehole instability. These damaging effects can be minimized by using an inhibited water-based fluid. The inhibited water-based systems cannot totally prevent water wetting of the rock pores, but they can minimize it.
Water-based muds fall into two basic categories: dispersed and nondispersed muds.
76
DrilHngFluid
77
Table 1. Mud Density Measurements Comparison
ppg
8.0 8.335 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0 16.67 17.0 18.0 19.0 19.23 20.0
S.G.
0.96 1.0 1.08 1.20 1.32 1.44 1.56 1.68 1.80 1.92 2.0 2.04 2.16 2.28 2.31 2.40
psi/ft
0.416 0.433 0.468 0.520 0.572 0.624 0.676 0.728 0.780 0.832 0.867 0.884 0.936 0.988 0.9999 1.040
pcf
59.84 62.35 67.32 74.80 82.28 89.76 97.24 104.72 112.20 119.68 124.69 127.16 134.64 142.12 143.84 149.60
Dispersed Muds
These muds have a chemical dispersant added to the system which is used to deflocculate mud solids. Most of the chemical dispersants in use (such as lignite and lignosulfonate) are acidic and require an alkaline environment in which to function properly. Of all the water-based muds, h i g h p H m u d s are the m o s t tolerant of solids a n d contamination. They are, without a doubt, the least difficult of the water muds to maintain. Clay (bentonite) is used as a viscosifier and fluid loss agent. Dispersants are use to permit enough clay into the system to control fluid losses. Caustic soda (NaOH) is used for pH control, and the density is adjusted with weight materials.
Dispersed muds can be broken into two smaller categories: calcium-based and seawater muds.
• Calcium-Based Mud—Calcium-based m u d systems maintain a desired amount of calcium in the water phase. The calcium concentration can be maintained by using gypsum (CaSO4) or lime [Ca(OH)2]. These muds are more inhibitive and can tolerate cement and anhydrite contamination better than a freshwaterdispersed fluid. However, their thermal limitation is somewhat reduced.
• Seawater Mud—In seawater muds, the upper limit for conventional dispersed fluids to function efficiently is 20,000 m g / L chlorides (which is the salinity of seawater). The cost for this type of system is slightly higher than that of a freshwater system. However, in offshore environments, this cost is offset by allowing muds to be run using native seawater rather than transporting in freshwater.
Nondispersed Muds
A basic difference between dispersed and nondispersed muds is the lack of dispersants. Nondispersed drilling muds do not require an elevated pH. By not having a dispersant present, they are less tolerant of solids and contamination.
The m a j o r i t y of the fluid loss control a n d viscosity is maintained via polymers, and these products are very susceptible to contamination from the formation, produced gases, and fluids.
Oil-Based Muds
Oil-based muds were developed to prevent water from entering the pore spaces and causing formation damage. There are several advantages and disadvantages of this type of mud system. The advantages include the following:
• Shale inhibition—In highly smectitic or "gumbo" shales, the borehole maintains stability and cuttings samples are generally intact.
• Reduction of torque and drag problems—Since oil is the continuous phase, the borehole and the tubulars are wetted with a lubricating fluid. This is a distinct advantage in deviated wellbores.
• Thermal stability—Oil-based muds have shown stability in wells, with BHTs of 585°F.
• Resistance to chemical contamination—Carbonate, evaporite, and salt formations do not adversely affect the properties of an oil mud. CO2 and H2S can easily be removed with the addition of lime (CaCO3).
Disadvantages of oil-based m u d systems include the following:
• High initial cost—The oil fraction alone of a barrel of oil mud may cost $40-70 per barrel. This is considerably higher than most water-based muds at any weight.
• Slow rates of penetration—Oil m u d s historically have had lower rates of penetration as compared to water-based muds.
• Pollution control—Most areas where oil muds are used have environmental restrictions. Rig modifications may be necessary to contain possible spills, to clean up oil mud cuttings, and to handle whole mud without dumping.
• Disposal—Oil mud cuttings may have to be cleaned up before dumping. Some regulatory agencies require cuttings be sent to a designated disposal area.
• Kick detection—H2S, CO2 and CH4 are soluble in oil muds. If gas enters the wellbore, it can go into solution under pressure. As the gas moves up the wellbore, it can break out of solution at the bubble point and rapidly evacuate the hole, blowing the mud with it.
• Formation evaluation—Some wireline logs should not be run in oil-based muds. Also, additional steps are needed to remove oil coatings from cuttings before they are described. (For more information on wireline tool compatibility with drilling fluid composition, see chapter on "Basic Tool Table" in Part 4, and for more on removing oil coatings from cuttings, see "Mudlogging: Drill Cuttings Analysis" in Part 3.)
Oil-based muds contain three phases: oil, brine, and solids phase.
78
PART 3—WELLS1TE METHODS
Oil Phase
The oil phase is the continuous phase in which everything else in the system is mixed. The oil can be diesel, mineral oil, or one of the new types of synthetic oils.
Brine Phase
The brine phase is present in the system as a high concentration salt solution that is emulsified into the base oil. Usually a solution of calcium chloride is used because it gives a greater flexibility in adjusting the concentration of the salts. This p h a s e is d i f f i c u l t to c o n t r o l b e c a u s e , if the salt concentration nears saturation, the emulsifiers and oil-wetting compounds precipitate.
Solids Phase
The solids phase includes the weight material, viscosifiers, and fluid loss reducers. A primary requirement for this phase is that it remain oil wet. Compounds exclusively developed for this purpose are included in the oil mud make-up. If the solid phase ever becomes water wet, the system is said to have "flipped" and the consequences are severe and operationally expensive. The system will separate into two phases: solid and liquid. The solid phase will pack and plug the wellbore, necessitating remedial drilling.
Air Drilling
Under a restricted set of conditions, air can be used as the drilling fluid when drilling through formations having little or no permeability to water. Although classified as "air" drilling, several types of gasses are actually used.
Dry Air
Air is compressed and pumped down the drill pipe at 500-800 ft3/min (cfm). The returned air is blown out the "blooie" line to a pit designed to retain the dust and cuttings. Dry air is preferred for fast drilling in dry, hard rock conditions with no water influx.
Mist
Mist drilling follows the same format as dry air drilling, but brine water is injected into the air stream. This is the method of choice when drilling wet formations with minimal water influx. The brine mist is injected to minimize reaction of the formation with a freshwater influx.
Foam
Foam drilling follows the same format as mist drilling, but with a foaming agent introduced into the mist stream. Foam is preferred when drilling stable formations that may have a moderate influx of water.
Pressure Detection
Parke A. Dickey
Consultant Owasso, Oklahoma, U.S.A.
INTRODUCTION
Normal reservoir pressure is the pressure in the reservoir fluids necessary to sustain a column of water to the surface (Fertl, 1976). Normal pressures range between 0.43 and 0.50 psi/ft. Normal drilling muds weigh about 9 ppg (pounds per gallon) and exert a bottom hole pressure of approximately 0.47 psi/ft of depth.
By convention in the petroleum industry, overpressure refers to pressures higher than normal that require heavy drilling mud to keep formation fluids from entering the borehole. Pressures lower than normal are called subnormal.
OVERPRESSURED RESERVOIRS
Drilling Problems with Overpressured Reservoirs
Notable effects of overpressured reservoirs that are costly include the following:
• Blowouts—uncontrolled production of formation fluids • Caving—high pore pressure in low permeability rocks
causes them to stress relieve or "cave" into the borehole • Stuck pipe—the drill pipe adheres to the side of the borehole due to the swelling (stress relief) of the borehole walls behind the bit • Lost circulation—by raising the mud weight to control the formation pressure at the bit, the formation may rupture. The mud will then run out into a cavity of its own making.
When drilling in areas where overpressured zones are k n o w n to occur, it is necessary to be a w a r e of both the pressure of the fluids in the pores and the pressure at which the formations will fracture. It is not enough simply to drill with heavier mud to prevent blowouts. If the m u d is too heavy, the formation will rupture and lost circulation will result. It is usually impossible to determine these critical pressures in a new area in advance of drilling.
Geological Cause of Overpressured Reservoirs
There is considerable debate and literature on the causes of o v e r p r e s s u r e . This section outlines s o m e c o m m o n explanations.
Arrested Compaction of Shale
The most commonly accepted cause of overpressure is arrested compaction of shale. Compaction requires the expulsion of pore water. When clays first start to compact, they are quite p e r m e a b l e and most of the water m o v e s upward. As compaction continues, however, the clay flakes become parallel, reducing vertical permeability. Sands and
silts compact less than clays and shales and can maintain permeability to greater depths.
As long as there is a silty or sandy bed within a few feet of the shales, the shales continue on a normal compaction trend. However, if no sandy beds are present, the water remains in the shale pores. As additional overburden is deposited, the shale then has to sustain all or part of the additional weight. This results in high pressure in the shale pore water. If there is a small, isolated sand body enclosed by the shale, whatever fluid it contains (water, oil, or gas), will share the same pressure. The fact that overpressures have been maintained for hundreds of millions of years over small vertical intervals indicates that the permeability of the enclosing shales can be virtually zero.
The distribution of reservoirs and o v e r p r e s s u r i n g is strongly controlled by the depositional environment (Figure 1). Overpressured reservoirs are commonly found where there are thick deposits of shaly sediments.
Aquathermal Effects
Aquathermal effects also cause overpressure. The temperature increases as sediment is buried, causing an increase in the volume of water. This in turn results in an i n c r e a s e in p r e s s u r e if the s e d i m e n t is sealed by an impermeable layer (Barker, 1972). For example, if a shale is totally sealed and there is no dilation to increase the pore volume, and if the geothermal gradient is 250C per 1000 m, then the pressure increase is about 1.8 psi per ft. This is more than the increase in weight of the overburden. Consequently, this a q u a t h e r m a l p r e s s u r i n g will cause an increase of pressure up to the pressure at which the rocks fracture (Figure 2).
Pressure data from some U.S. Gulf coast wells suggest that the aquathermal effect is important.
Tectonic Phenomena
Tectonic phenomena also produce overpressures. In the Gulf of Alaska, fluid pore pressures up to 0.85 psi per ft were found due to horizontal compressive stress in the rocks. In Western Alberta, large thicknesses of Paleozoic carbonates have been thrust over soft Cretaceous shales, resulting in overpressuring of the lenticular oil-bearing sandstones that extend under the overthrust (such as Leafland and Pembina).
Thermal Cracking of Organic matter
The thermal cracking of organic matter may cause an increase in the volume of fluids, which would in turn cause an increase in pressure.
Other Causes
Other suggested causes of overpressuring include the loss of water in smectite clay as it changes to illite, osmotic pressures due to salinity variations in the water, and effects of cementation.
79
80
PART 3—WELLS1TE METHODS
North
South
Figure 1. Schematic diagram showing the location of abnormal pressures in southern Louisiana. The continental and deltaic facies contains sandy beds. The neritic (nearshore) facies contains a few silty and sandy beds that connect laterally to the deltaic facies. The outer shelf facies contains almost no sandy beds, and the pore fluids cannot escape. The growth faults are seals that stop the lateral flow of pore water toward the neritic facies. (From Dickey et al., 1968.)
Prediction of Overpressure Before Drilling: Seismic Detection
As shales compact, the velocity of seismic waves increases so that seismic wave velocity normally increases with depth. If the shales have not been able to compact, the seismic velocity will be less.
Interval velocity can be determined from the surface by the c o m m o n d e p t h point m e t h o d of seismic acquisition. If interval velocity increases normally with depth and then decreases, it is possible that a zone of overpressure exists. This method of predicting the depth to an overpressured zone has been widely used, especially in the offshore Gulf coast of the United States.
Detection of Overpressure While Drilling
Drilling Rate
When drilling shales, the drilling rate normally decreases with depth as the shales become more compact. If the drilling rate increases, it can be inferred that an overpressured zone is being encountered. The rate increases because bottom hole conditions change from overbalanced to underbalanced.
Because lithology, rotary speed, and weight on the bit also affect d r i l l i n g rate, a c o r r e c t e d d r i l l i n g r a t e called "d exponent" should be used (Equation 1). This method can be one of the most reliable indicators that the drill is penetrating a zone of abnormally high pressure.
d exponent
(1)
where R = rate of penetration in feet per hour, N = rotary speed in revolutions per minute, W = weight on bit in pounds, and D = hole diameter in inches (Jorden and Shirley, 1966).
Figure 2. Common patterns of increasing pressure with depth. In the case illustrated by line A, the pressure increases normally to a certain depth, then increases abruptly to almost the weight of the overburden, which it then parallels. In the case of line B, the increase of pressure above normal follows the aquathermal gradient (constant water density) and then follows the fracture gradient. (After Barker and Horsfeld, 1982.)
Mud Tank Level
A rising level of mud in the tanks indicates that more mud is coming out of the hole than is going in. This is called a "kick." This happens because formation fluids are entering the hole and the well is threatening to blow out. The situation is extremely serious, and proper steps must be taken to get the gas, oil, or water out of the hole. The most common method is to close the blowout preventers and stop the pumps. After a few minutes, the pressure at the top of the drill pipe will equal the pressure in the formation minus the weight of the column of mud. This is the excess pressure that must be balanced by increasing the mud weight. The pumps are then started to circulate the extraneous fluid out of the hole. The drill pipe pressure is carefully controlled with the choke. If the equilibrium drill pipe pressure is exceeded, the well may lose circulation, and if it is too low, the well will blow out.
RESERVOIR FPG.
Pressure Detection
81
EQUIVALENT MUO WEIGHT,
Figure 4. Shale resistivity parameter, resistivity of normally pressured shale divided by observed resistivity of abnormally pressured shale, plotted against formation pressure gradient (FPG) and equivalent mud weight. (From Hottman and Johnson, 1965; by permission of SPE.)
Figure 3. Electric logs of two wells offshore Louisiana. Well A had normal pressure. Well B, 2000 ft away and across a growth fault, showed a sudden decrease in resistivity of shale (increase in conductivity) at about 11,100 ft. Shortly thereafter, the well showed indications of an impending blowout. (After Wallace, 1965.)
Delayed Indications
Several other indications of overpressure may occur after the lag time necessary for the mud to return to the surface from the bottom of the hole. They are much less reliable than the drilling rate, but can be monitored by mudlogging equipment. These indicators include the following:
• Drilling mud gas—Gas in the drilling mud often increases because methane is dissolved in the pore water of some overpressured shales. As the cuttings and cavings come up the hole, the gas escapes and can be detected in the mud. Gas in the mud is also caused by oil- or gas-bearing formations and by organic-rich shales.
• Shale density—Undercompacted shales, characteristic of overpressured zones, have a lower density (because of abnormally high porosity) than normal shales at a given depth. The density of shale cuttings can be measured
by several methods. Also, the shape of drill cuttings from undercompacted shales may be different than those from normally compacted shales. • Temperature—There may be an increase in the temperature of the mud returns. Although it has been widely claimed that the geothermal gradient is higher in overpressured shales because of their abnormally high porosity and lower thermal conductivity, a doubling of shale porosity from 10 to 20% should cause a decrease in conductivity of only about 1 % (with a correspondingly small increase in geothermal gradient). Thus, the increase in temperature is probably due to faster drilling and increased cavings in undercompacted shales.
Detection of Overpressure with Well Logs
Undercompacted shales associated with overpressured zones have a much lower electrical resistivity than normally compacted shales (Figure 3). According to the Archie formula, doubling the porosity of a shale from 10 to 20% should cause its resistivity to drop to one-fourth. As a result, it is possible to d e t e r m i n e a c c u r a t e l y the d e g r e e of undercompaction of a shale from its resistivity and to estimate the pore pressure (Figure 4) (Hottman and Iohnson,1965).
Because undercompacted shale has slow seismic velocity and low density, a high pressure zone can also be identified from sonic and density logs (Magara, 1978).
Note that as is always the case with well logs, there are pitfalls in interpretation, and the local geology and hole conditions must be taken into account.
82
PART 3—WELLS1TE METHODS
SUBNORMALLY PRESSURED RESERVOIRS
Subnormal reservoir pressures, that is, pressures less than 0.43 psi per ft of depth, are very common (Dickey and Cox, 1977). The cause of abnormally low reservoir pressures is not well understood. If a reservoir containing either gas or oil is isolated and then subjected to uplift and erosion, the removal of overburden causes an elastic rebound of the solids and an increase in v o l u m e of the pores. The elastic dilation of sandstones is about 7 x IO"6 volumes per psi. Water expands only 3 x IO"6 volumes per psi, so that the pressure of the pore water in the aquifer and the enclosing shales will drop, possibly sucking some of the water out of the aquifer. Most low pressure reservoirs are in areas where there has been uplift and erosion since the sediments forming the reservoir were deposited and lithified.
Drilling Problems with Subnormally Pressured Reservoirs
Much less attention has been paid to subnormally pressured reservoirs than to overpressured reservoirs. This is
probably because there are fewer spectacular drilling problems associated with subnormal pressures and underpressures. However, problems exist that can be serious.
If the reservoir pressure is much lower than the pressure in the drilling fluid, severe formation damage can occur. The drilling mud filtrate penetrates the reservoir, causing swelling and migration of clays, which may plug the pore throats. Even a little water in the hole can kill a low pressure producing gas well. The water is drawn into the pores by capillarity and ruins the relative permeability to gas. In the case of low pressure gas sandstone reservoirs, it is desirable to set casing at the top of the reservoir interval and drill with gas, salt water, or oil-based mud to minimize formation damage.
Also, if the gas reservoir has a low pressure, there may be no indication of gas on the mudlog. The logs of many abandoned dry holes should be reexamined to look for bypassed gas zones.
Fishing
Arnold M.Woods
Conoco Inc. Casper, Wyoming, U.S.A.
REASONS FOR FISHING
Fishing (in the oilfield sense) refers to the recovery of unwanted material left in the wellbore. Numerous situations can occur that require fishing:
• Items dropped into the hole from the rig floor. • Failure of surface equipment, especially pumps, which
allows the hole to cave in and stick the drill string. • Differential sticking to a permeable formation
(Figure 1). • Key seating, where a slot is worn into the wall of the
well (Figure 2). • Twist-offs resulting from a stuck drill string being
rotated until the pipe shears above the sticking point. • Loss of portions of downhole equipment, such as
stabilizer fragments and bit cones. • Drill string failure from such causes as metal fatigue,
H2S embrittlement, differential sticking, and overstressed drilling assembly, causing a twist-off. • Pulling the drill string apart when trying to free stuck pipe. • Bridging off by swelling formations or a collapsed hole.
When material is left in the hole due to these or other factors, a decision must be made whether to try to recover the material, to sidetrack around the original material, to abandon the well, or to attempt to complete in a shallower zone. Because all but the first of these options are costly, and the last is often undesirable, an attempt to recover the fish is almost always made.
TYPES OF FISHING TOOLS
Many specialized tools have been developed to address specific types of recoveries. In general, fishing tools fall into two categories: those used to recover small items (junk) and those used to recover larger items such as drill pipe.
Common Junk Fishing Tools
Magnet. A magnet is used to recover small metal pieces from the hole, such as bit cones (Figure 3).
Junk or Boot Basket. A "junk" or "boot" basket is run just above the bit and catches small pieces of junk thrown up by turbulence. It is often part of the BHA, and it is almost always run before a core barrel or diamond bit to prevent damage (Figure 4).
Figure 1. Differential sticking. (Courtesy of Exploration Logging, 1979.)
Figure 2. Key seating. (From Short, 1981; courtesy of PennWeII Books.)
83
84
PART 3—WELLS1TE METHODS
^—\Г?Г Magnet
-Iir—Junk B a s k e t
Figure 3. Magnet.
Metal Junk
Figure 4. Junk basket.
Poor Boy Junk Baaketr;
Ж
Barrel
Catchers
Figure 5. Poor boy junk basket. (Courtesy of Exploration Logging, 1979.)
Mill ahoa
Figure 6. Core type basket.
Poor Boy Junk Basket. Often made at the wellsite, a "poor boy" junk basket consists of a small section of casing with beveled "fingers" cut into the end. The basket is lowered on the drill string where it cuts a small core. More weight is applied to the tool, bending the fingers in and snaring the core containing the junk (Figure 5).
Core Type Basket. A core type basket consists of a barrel with a mill shoe and two sets of internal catchers. The basket is slowly rotated onto the junk in the hole where it cuts a small core. The catchers hold the core while the tool is retrieved (Figure 6).
Fishing
85
У \
Drill String
Waehover Pipe Fish Cutting Shoe
Figure 7. (a) Tapered mill, (b) Flat mill.
Figure 8. Washover pipe.
Tools Commonly Used To Retrieve Drill Pipe and Logging Tools
Jars. Jars are hydraulic or mechanical tools that provide a sudden, powerful upward (oil jar) or downward (bumper jar) stroke on the drill string. A typical jarring set up would include a catching tool, bumper jar, oil jar, drill collars, and drill string. Drilling jars (specially strengthened jars) are commonly run in the drill string in case the drill string becomes stuck. In a fishing job, the jars are run just above the fishing tool.
Qrapplet
Mill. Before some fish can be recovered, it is necessary to "dress" the top, that is, to grind the uneven top to a shape that the fishing tool can grasp, or to clear and open the inside of the fish. To do this, either a flat or a tapered mill bit is used (Figure 7). The mill bit has tungsten carbide abrasive surfaces that can trim the fish to the desired shape.
Flth
Washover. Washover equipment consists of casing with a cutting shoe. It is run over the fish to clear away debris before the fishing tool is run (Figure 8). Washing over is often done to free key-seated or differentially stuck pipe.
Figure 9. Overshot.
Overshot. The most common and widely used tool, the overshot consists of a large diameter, open mouth pipe with a tapered tap or die, or, more commonly, with a set of slips that grab the outside of the fish when the assembly is lifted and prevent its release (Figure 9).
Pipe Spear. Like an overshot, a pipe spear is used to recover large tubulars, but it instead attaches to the inside of the fish via a set of slips that are extended after the spear is inserted (Figure 10).
Wireline Spear. A wireline spear is a simple spear with barbs used to hook parted wireline (Figure 11).
FREEING STUCK PIPE
Before a stuck pipe becomes a fishing job, several procedures can be used to free it. These include the following:
1. Jarring on the stuck pipe until it loosens 2. Placing oil or chemicals (such as surfactants) in the hole
to reduce intersurface tension 3. Displacing part of the mud with a lighter fluid to
reduce hydrostatic pressure, which is particularly effective for differentially stuck pipe
86
PART 3—WELLS1TE METHODS
Figure 10. Pipe spear.
Drilling Problems
Phyllis Loose
Conoco Inc. Casper, Wyoming, U.S.A.
INTRODUCTION
Drilling problems include any difficulties encountered while drilling a well. The most common drilling problems are the creation of doglegs and key seats, hole instability, lost circulation, and excessive bottom hole temperatures.
DOGLEGS
A dogleg is usually defined as any deviation greater than 3° per 100 ft, and it occurs when a sharp change of direction is taken in the wellbore. Typically, a dogleg is caused by a change in the dip of the formation or by a change in the weight applied to the bit. Severe doglegs can result in stuck casing, drill pipe failure, and inability to run casing to total depth. If casing is successfully r u n through the dogleg, excessive wear on production equipment can occur. The use of properly placed stabilizers, large diameter drill collars, and the proper weight applied to the bit will minimize the formation of doglegs.
KEY SEATS
Key seats usually form as a result of dog legs. A key seat is formed when a channel or groove is cut in the side of the hole, parallel to the axis of the hole. The drill pipe dragging action through the sharp bend in a doglog creates the groove in the side of the wellbore. Key seats can be prevented by not creating doglegs.
HOLE INSTABILITY
Hole instability occurs when encountered formations flow, slough, or swell. The most unstable formations are shales and salt beds.
Instability can result from the following phenomena:
1. Overburdenpressure 2. Earth movement forces 3. Porepressure 4. Water absorption, swelling, or dispersion
Instability occurs when the relief of overburden pressure exceeds the yield strength of the drilled formation, resulting in flow of the formation (plastic flow). Abnormally high pore pressures can cause blowouts when present in highly permeability formations. If the pressure differential between the wall of the hole and the fluid in the hole is large, the formation may slough off. Structural stresses can also cause hole instability.
Problems associated with hole instability include the following:
1. Ineffectiveholecleaning 2. Stuckpipe 3. Bridgesandfill-up 4. Wellboreeiilargement 5. Increased mud volume 6. Increased cost 7. Poorcementjobs 8. Loggingdifficulties
Hole instability can usually be controlled by use of the proper drilling fluid.
LOST CIRCULATION
Lost circulation is the complete or partial loss of drilling mud into a formation. The loss occurs when the total pressure exerted against the formation exceeds the total pressure of the formation. Formations succeptible to lost circulation have the following characteristics:
1. Cavernousandopenfissured 2. Very coarse, permeable, and shallow, such as loose
gravel 3. Naturally or intrinsically fractured 4. Easilyfractured 5. Underpressured or depleted
Lost circulation results in increased mud expense and may cause subsurface blowouts. The proper drilling fluids and use of lost circulation material will minimize lost circulation.
BOTTOM HOLE TEMPERATURES
Extremely high bottom hole temperatures can occur in deep wellbores or in areas of abnormally high geothermal gradient. These excessively high bottom hole temperatures (greater than 250 °C) can cause drilling problems because of the accelerated thickening of water-based drilling fluids. The increase in viscosity and density of the drilling can may result in the following:
1. Reductioninpenetrationrates 2. Lostcirculation 3. The well being swabbed when drill pipe is pulled 4. Stucktools
This problem can be mitigated by using oil-based muds.
87
Measurement While Drilling
Michael F. Medeiros
Shell Offshore, Inc. New Orleans, Louisiana, U.S.A.
INTRODUCTION
Measurement while drilling (MWD) technology provides downhole evaluation of formation gamma ray, resistivity, and porosity while drilling is in progress. Mechanical parameters such as the following are also measured and recorded for well surveillance:
• Well deviation and azimuth • Rate of penetration (ROP) • Weight and torque on bit
DATA ACQUISITION
Downhole MWD hardware consists of sensors built into a drill collar positioned near the bit (Figure 1). Electrical energy for the system is provided by a battery pack or generated by a small turbine.
In a battery pack MWD system, information is recorded and stored downhole in the microprocessor. The data are retrieved when the MWD collar is brought to the surface and are transferred to the computer in the logging unit for additional processing.
In a typical turbine-powered "real time" MWD system, data are sent directly to the surface by mud telemetry, which utilizes the c o l u m n of fluid i n s i d e the drill p i p e as a transmission line for digital acoustic signals. Downhole measurements recorded by the sensors are transmitted through the mud as positive or negative pressure pulses or as a continuous, fixed-frequency pressure wave.
The mud telemetry signals are detected with pressure transducers in the standpipe. The digital signals are then recorded by a computer. Data are converted to engineering units and processed to generate depth- or time-based output.
drilling are commonly being used to supplement or replace wireline logs for formation evaluation and geological correlation in high risk or high cost operations.
Formation Evaluation
Formation evaluation using MWD technology has several advantages over conventional wireline log methods:
• Because MWD measurements are made soon after the interval has been penetrated by the bit, the effects of invasion fluids are reduced. By minimizing the effects of invasion, a more reliable evaluation of critical
DATA DISPLAY
At the wellsite, information is displayed on video monitors located in the logging unit or at other locations around the rig. Real time MWD data can also be transmitted via telephone modem to a remote location where drilling progress can be monitored by the development geologist, engineering personnel, and management.
Hardcopy directional plots and well logs can be produced at specified scales and annotated with mechanical and geological information (Figures 2 and 3).
APPLICATIONS
Measurement while drilling technology has become an important tool for reservoir evaluation in the past 10 years. Gamma ray, resistivity, and porosity logs obtained during
Figure 1. Location of MWD hardware (not drawn to scale). (From Anadrill, 1988.)
88
МММ IUY API
Measurement While Drilling
89
FORMATION LOGGING SERVICE (ROTARY MODE)
MVUflED RtSSTWITY
ANNULAR TtMPtRAIUS
SHORT MMMAL ROlSTivnY
DMfNHOU TOMUt
Figure 2. HardcopypIotofMWDdata. (From Anadrill, 1988.)
formation properties can be achieved. • In directionally drilled wells, where hole angles can
deviate as much as 80° from vertical, wireline logs are often difficult and very costly to obtain. In these situations, MWD data may provide the only permanent record of the borehole. MWD logs also provide "insurance" in the event a well must be abandoned due to mechanical reasons. • The electrode spacing and slower penetration rate of the MWD tool provide a larger number of samples per foot. The increased sample density often results in better resolution of thin beds, particularly with electromagnetic propagation resistivity tools. Because
the logging speed is dependent on the rate of penetration, drilling can be controlled through zones of interest to obtain maximum resolution.
Geological Correlation
Prior to MWD technology, drilling time plots or rate of penetration (ROP) plots were used for geological correlation while drilling was in progress. These plots can be difficult to use, particularly in complex areas, because penetration rate can be controlled by mechanical parameters. With a MWD gamma ray or resistivity log, correlation with offset wells is much more reliable.
Measurement while drilling logs have been used
90
PART 3—WELLS1TE
METHODS
MWD TIME Inlerfed Lnhology Irom
GR FORS. FORM. Resislivity
LAG TIME Verilied Lilhology
from Cuttings Description
FORS
I
1001
FORM
L-J.___2
REAL TIME Lilhology
Segmentation Irom MEL
SL SHALY COMPACT SL SHALY
8L SHALY COMPACT
SHALE SANO
MW=14.2, m = 6 2 PV= 15 Y P * 1 2 pH = 11. WL= 17 CL=950.
Sand SOL=5%.
SL SHALY
Shale
COMPACT
SLTY SHALE
COMPACT
SLTY
SLTY SL SHALY SL SHALY COMPACT
SLTY
SS: It—dk gy, SAND" al arg. f - m a d gr,
p artd. ang, calc. w cmt. occ mica, occ tr hvy oil on SHALEl aurf, Int dk fluor
SL SHALY
COMPACT COMPACT
CONG: It bm, prad uncona. era gm. mod art. a bang, gld fluor. atmg dk yal-brn cut
8 LTY
SHALE
Figure 3. Composite Iithological log from MWD data. (From Schlumberger, 1989.)
successfully for geological correlation by providing the following functions:
1. Determination of coring points—Prediction of lithology ahead of the bit is critical when selecting prospective intervals for coring. With MWD log data, the geologist can obtain a more reliable correlation and can select core points with more confidence. This reduces the risk of missing critical formation tops or coring intervals that are not essential for geological or reservoir evaluation.
2. Selection of optimum casing and total well depths— Accurate and timely selection of casing and total well depth become extremely important when the possibility of penetrating abnormally pressured zones exists. This condition often requires extra rig time to circulate and condition the drilling mud or, in the worst case, results in drill pipe sticking and expensive fishing operations (see the chapter on "Fishing" in Part 3).
3. Determination of kick-off point(s) for sidetrack wells—The decision to sidetrack or redrill a well to a more favorable location can be made immediately if MWD data show the objective section to be poorly developed, faulted out, or drained. This often results in considerable cost savings by preventing unnecessary drilling below the objective interval and by eliminating the need for wireline logs.
4. Detertnination of kick-off point(s) for horizontal wells— MWD logs are used for correlation with offset data and for directional planning in complex horizontal wells. With real time information obtained near the bit, borehole deviation and azimuth are continuously monitored and corrected to obtain the optimum angle of penetration. This helps ensure that the horizontal wellbore stays on target and within the desired reservoir interval.
5. Aid in "steering" highly deviated and horizontal wells— Using a real time MWD system, the geologist and field personnel can monitor the well path as the objectives are penetrated. The geologist is therefore better equipped to make important cost-saving decisions and provide management with the most recent interpretation and well strategy.
Rate of Penetration
Scott Boone
IDL Mudlogging Houston, Texas, U.S.A.
INTRODUCTION
Rate of penetration (ROP) logs are recorded at the wellsite as the well is being drilled. The ROP can be expressed as either distance drilled per unit of time (feet per hour) or time per distance drilled (minutes per foot).
accurate depth record. The depth can be checked by periodically strapping out of the hole (each stand is measured with a steel tape as it is pulled or tripped out of the hole) and making a depth correction to the geolograph if necessary.
Computers can measure depth and ROP with a high degree of accuracy and sensitivity. However, it must be
MEASUREMENT TECHNIQUES
Rate of penetration is calculated by measuring the length of time required to drill 1 ft of depth. This is typically done by reading the chart on the geolograph.
The geolograph, or drilling recorder, mechanically monitors depth and records drilling parameters in time. These parameters are recorded on a paper chart, graduated in minutes, that is wrapped around a drum. The drum rotates one revolution in 8, 12, or 24 hr. To record depth, a small cable is run from the geolograph to the top of the kelly via a pulley on the crown of the derrick (see the chapter on "Land Rigs" in Part 3). Kelly height can then be measured and directly related to bit depth. As each foot is drilled, an ink pen on the geolograph places a small mark on the chart. Every 5 ft the pen places a larger mark on the chart (Figure 1).
Other more advanced monitoring techniques utilize computers and digital encoders to monitor depth. These systems are typically stand-alone. The digital encoder or transducer is attached to a part of the rig that moves in proportion to the movement of the drill string. Common attachment points are the drill line, drawworks drum, or crown sheaves.
ERRORS AND SENSITIVITY
With standard equipment, most errors in ROP are related to the operation of the geolograph. When a connection is made, the driller should disengage the geolograph recorder before picking u p a new joint of pipe. The recorder must then be reengaged when drilling resumes. Unless this operation is performed properly, the rate of penetration just prior to or after a connection can be questionable. In addition, high winds can whip the geolograph recording cable, causing extra footage to be recorded. Tliis results in an apparent increase in ROP. It is not unknown for the drilling hands to "gig" or pull on the recording cable, causing a similar apparent increase in ROP.
Other sources of error include
• Holefill • Pipe stretch • Sticking pipe
All of these errors must be b a c k e d out" to maintain an
Figure 1. Example of a geolograph chart.
91
92
PART 3—WELLS1TE METHODS
remembered that virtually all methods of measuring ROP depend upon surface responses to the performance of the drill bit. Because the drill string stretches in proportion to the weight on bit (WOB), highly sensitive ROP logs can appear erratic as a result of the driller slacking off on the brake. For this reason, computer logs are frequently averaged to reflect the ROP and drillability of the formation.
APPLICATIONS
Rate of penetration logs serve as historical records of drilling performance and can aid in optimizing drilling operation and formation evaluation.
GR 160
fs*
I — —
<!
tr-
-4 4; —J
GEOLOGRAPH CHART
SANDSTONE ЙЭУ SHALE
Bit Selection
Hard, compacted formations drill slow and consistent. Soft formations drill fast and have erratic ROP.
Mud Weight (Pore Pressure)
The drilling rate of shale is very sensitive to the pressure balance between the shale and wellbore. Shales drilled underbalanced are characterized by fast drilling, high gas readings, and large cuttings. Shales drilled overbalanced are characterized by slow drilling, low gas readings, and small cuttings. Recognizing this relationship can aid in the selection of a mud weight to optimize drilling speed.
Correlation
The ROP log is the first source of data that can be used to correlate to nearby wells. This is done by comparing ROP to gamma ray or SP curves (Figure 2). This correlation can help determine structural and stratigraphic position and is used to predict when the well will reach a zone of interest.
Lithology
Shales generally drill slower than sandstones or carbonates, and as a result, ROP logs tend to reflect indirectly the lithology of the formation.
Relative Porosity
Porosity in sandstone is qualitatively inferred by observing the ROP in shale and comparing that to the ROP in a known sandstone interval. Porosity in carbonates is inferred by comparing the relatively slow ROP in rocks having low matrix porosity to rocks having higher porosity and faster ROP. This comparison requires that the prevailing drilling conditions (mud weight, bit condition, WOB, and rotary speed) are consistent.
at
Y X =2
Ш
Figure 2. Example of correlating ROP with gamma ray curves.
NORMALIZED RATE OF PENETRATION (NROP)
When comparing ROP curves from different wells, a normalized curve (NROP) must be made for each well. Actual recorded values are corrected back (or normalized) to base values for each bit run using the following equation:
NROP = (WOB„ - M)(RPMn) x L(dPb„ x Qn) x (WOB0 - M)(RPM0) x L(dPb0 x Q0)
where n = normal value 0 = observed value M = threshold weight, or the WOB needed to cause formation failure, generally taken from tables
RPM = bit rotating speed, in rotations per minute L = rotary exponent, taken from tables
dPb = pressure drop at bit, in psi; the pump pressure needed to push mud through the bit nozzels at a particular flow rate
Q = flow rate in gallons per minute of mud through the system
Wellsite Math
Greg Dunn
Technical Drilling Services Oklahoma City, Oklahoma, U.S.A.
INTRODUCTION
A few simple mathematical formulas are used to make key calculations during the drilling of a well. The results from these calculations increase knowledge of well behavior and facilitate communication with drilling personnel. Definitions of the mathematical variables and abbreviations used in this chapter are given in Table 1.
HYDROSTATIC PRESSURE IN LIQUID COLUMNS
Mud weight is the key pressure gauge used in drilling since it responds directly to formation pressure. The hydrostatic pressure at any point in a wellbore filled with liquid drilling fluid (mud) under static conditions (that is, not circulating and/or not moving the drill string up or down) is a function of only two variables:
• The height of the liquid column (depth) • The density of the liquid (mud weight)
The total volume of liquid or shape of the hole have no influence on hydrostatic pressure, but the height of the liquid column (depth) must be measured in the same direction as the force due to gravity, that is, true vertical. This consideration is important in highly deviated and horizontal wellbores. The hydrostatic pressure (P) at any depth in a wellbore is calculated as follows:
P = 0.052 x MW x D
Example: If a 10,000-ft-deep hole contains drilling fluid w i t h a w e i g h t of 11.5 p p g ( p o u n d s p e r gallon), the hydrostatic pressure at the bottom of the hole is
P = 0.052 x 11.5 x 10,000 = 5980 psi
Mud Weight
If the equation for hydrostatic pressure is solved for mud weight (MW), then the equivalent mud weight can be calculated, which has several important wellsite applications. Solving for MW, the equation becomes
MW = P/(0.052 x D)
Example: Assume a formation at 10,000 ft has a known hydrostatic pressure of 6292 psi. The mud weight needed to drill this formation balanced" (hydrostatic pressure equal to formation pressure) is calculated as follows:
MW = 6292/(0.052 x 10,000) = 12.1 ppg
Kill Weight of Mud
In a well control situation after a kick has been taken, it is necessary to increase the mud weight to balance the formation pressure. The kill weight of the m u d can be calculated by observing the stabilized shut-in drill pipe pressure. The shut-in drill pipe pressure records the excess formation pressure over the hydrostatic pressure of the drilling fluid in the hole. Shut-in drill pipe pressure is converted to an equivalent mud weight. This weight is added to the mud weight currently in the hole to obtain the kill weight needed. The kill weight of the mud can also be converted back to pounds per square inch to obtain an estimation of the formation pressure.
Example: Assume a well has taken a kick at 10,000 ft while drilling with 11.5-ppg mud. The well is shut-in, and the drill pipe pressure reads 312 psi. The excess in pounds per gallon mud weight equivalent is calculated as follows:
MW = 312/(0.052 x 10,000) = 0.6 ppg
Adding the calculated excess of 0.6 p p g to the current m u d weight of 11.5 p p g results in 12.1 p p g necessary to control the well. The mud weight is typically increased beyond the calculated kill weight to allow for the negative pressure (swab) exerted on the hole when tripping the drill string.
Formation Fracture Pressure
The formation fracture pressure (FFP) can be determined by the use of a leak-off or formation integrity test. In this test, the well is shut-in and drilling fluid is pumped slowly into the well. Wellbore pressure increases linearly up to the point where the formation starts to take drilling fluid. The pressure at this point is the leak-off pressure and is used to estimate formation fracture pressure. Like shut-in drill pipe pressure, leak-off pressure is added to the pressure of the mud in the hole.
Example: Assume a leak-off test conducted in a 10,000-ft wellbore containing 11.5-ppg mud with a leak-off pressure of 1040 psi. The formation fracture pressure is estimated as
MW = 1040/(0.052 x 10,000) = 2.0 ppg
The formation fracture pressure is equivalent to an 11.5 + 2.0 = 13.5 ppg mud weight, or
FFP = 0.052 x 13.5 x 10,000 = 7020 psi
Therefore, loss of circulation m a y be a p r o b l e m in this w e l l b o r e if m u d w e i g h t s exceed a b o u t 13.5 p p g . This procedure is sufficiently accurate for quick field calculations,
93
94
PART 3—WELLS1TE METHODS
Table 1. Definitions of Terms and Abbreviations Used in This Chapter
Term
P MW D Vbpf Dl Ds Vh Va
Vd
Vc W od id Dt Dd Ls D1 Dr AV GPM Dh An J1 . . . Jn JNV THhp Pp JNPL % Hhpb BHhp Hhp/in.2 Ob JIF FFP T% Hhpb
Definition
Hydrostatic pressure at depth d Mud weight Vertical depth Volume Larger diameter Smaller diameter Hole volume; the volume of the open or cased hole Annular volume; the volume of the annulus, the area between
the outside of the drillpipe or collar and the open or cased hole Displacement volume; the volume displaced by the steel volume
of drillpipe or collar. Displacement is the volume between the outside diameter and inside diameter of a drillpipe or collar. Capacity volume; the volume contained inside a drill pipe or collar Weight of collars or casing Outside diameter of pipe Inside diameter of pipe Triplex mud pump output Duplex mud pump output Length of pump stroke Diameterofpumpliner Diameter of pump rod (duplex only) Annular velocity Mud pump output Hole diameter Jet nozzle area Size of jet nozzles, "32nd" omitted Jet nozzle velocity Total hydraulic horsepower Mud pump pressure Jet nozzle pressure loss Percentage of total horsepower expended at bit Hydraulic horsepower at the bit Hydraulic horsepower per square inch area of bit Bit diameter Jet impact force Formation fracture pressure Total percentage hydraulic horsepower at the bit
Unit
psi ib/gal ft bbl/ft in. in. bbl/ft
bbl/ft
bbl/ft bbl/ft lb/ft in. in. bbl/stroke bbl/stroke in. in. in. ft/min gal/min in. in.2 32nd in. ft/sec hp psi psi % hp hp/in.2 in. Ib psi %
but it slightly overestimates the formation fracture pressures because the effective compressibility of the drilling fluid and frictional pressure losses are not accounted for.
WELLBORE VOLUMES
All important wellbore volumes are calculated with a single formula:
Vbpf= (D12-Ds2)/1029.4
Volumes are reported in units of barrels and can be calculated for an in-gauge hole by using this equation to determine volume in barrels per foot, then multiplying that value by the length of the section of hole in feet. (Note: washouts and thick mud cake can substantially alter hole volume.)
Example: If a 10,000-ft-deep hole with an 8.0-in.-diameter contained pipe with an outside diameter of 5.0 in. and an inside diameter of 4.0 in., the volume calculations are as follows:
Vh = [(8.02)/1029.4] x 10,000 = 621.7 bbl
Va = [(8.02-5.02)/1029.4] x 10,000 = 378.9 bbl
Vd = [(5.02 - 4.02)/1029.4] x 10,000 = 87.4 bbl
Vc = [(4.02)/1029.4] X 10,000 = 155.4 bbl
Although this solution is correct mathematically (that is, the s u m of a n n u l a r volume, displacement, and capacity equals the hole volume), the basic formula results in an
WellsiteMath
95
underestimate for displacement of drill pipe. Since drill pipe (not collars) has tool joints that are welded onto the pipe, calculation of drill pipe displacement involves an additional step to account for the additional volume displaced by the tool joints. For most calculations with commonly used sizes and weights of drill pipe (4.5 and 5 in.), the addition of the quantity 0.001 to the barrel per foot volume calculation for displacement accounts for the additional tool joint volume with sufficient accuracy. Using the data from the previous example, we obtain the following the displacement calculation:
Vd = ([5.02 - 4.02)/1029.4] + 0.001) x 10,000
= 97.4 bbl
Notice that when drill pipe tool joints are considered, the displacement is 10.0 bbl greater. Again, this additional step is not necessary when calculating the displacement of collars since collars do not have welded-on tool joints.
When volume calculations are made at the wellsite, it is important to identify each hole geometry in the wellbore. This is best accomplished by drawing a schematic diagram of the wellbore and identifying all casing, open hole, drill pipe, and collar dimensions. Several handbooks have extensive tables listing pipe and collar dimensions (Baker Service Tools, 1985). Calculations must then be made for each annular displacement capacity and hole volume and then summed for a total.
stroke length and 6-in. liners installed is
Dt =(11 x 62)/4118 = 0.0962 bbl/stroke
This is a theoretical calculation based on the pump working at 100% efficiency. Mud p u m p s are rarely 100% efficient, and this calculation is multiplied by some volumetric efficiency factor, typically 0.90 to 0.99 for triplex pumps and 0.80 to 0.90 for duplex p u m p s . The actual efficiency of a pump is determined by measuring the suction pit, isolating the suction pit so that mud is pumped out of that pit only for a certain n u m b e r of p u m p strokes, then calculating the volume p u m p e d out of the pit and dividing by the total number of p u m p strokes.
In addition to barrels per stroke, mud pump output can also be expressed in gallons per stroke, barrels per minute, or gallons per minute. Once barrel per stroke is determined, the remaining parameters are easily calculated as follows:
Bbl/min = Bbls/stroke x Strokes/ min
Gal/stroke = Bbls/stroke x 42
Gal/min = Gal/stroke x Strokes/min
Example: If the triplex p u m p previously discussed was operating at 95% volumetric efficiency at 98 strokes/min, the following calculations can be made:
WEIGHT OF COLLARS OR CASING
The weight (W) of collars or casing, in pounds per foot, is estimated by
W = 2.67 x (od2 - id2)
Often the outside diameter (od) and the weight of the collars are the only dimensions known. In this case, the previous equation can be solved for the inside diameter (id) of the collars:
id = [od2 - (W/2.67)]1^2
Example: To estimate the inside diameter of drill collars having an outside diameter of 6.5 in. and weighing 96 lb/ft, the following calculation is made:
id = [6.52- (96/2.67)]1/2 = 2.5 in.
MUD PUMP OUTPUT
The theoretical pump outputs for triplex and duplex type pumps are calculated by the following formulas:
Dt = (Ls XD12)/4118 Dd = Ls x [(2 x D12)-Dr2 ]/6176
Example: The output of a triplex mud p u m p with an 11-in.
Bbl/stroke = 0.0962 x 0.95 = 0.0914 bbl/stroke
Bbl/min = 0.9814 x 98 = 8.96 bbl/min
Gal/stroke = 0.0914 x 42 = 3.84 gal/stroke
Gal/min = 3.84 x 98 = 376 gal/min
CIRCULATION OR LAG TIME
Bottoms-up circulation, or lag time, is the time for samples or gas created at the bit to arrive at the surface (via the drilling fluid) for examination. This calculation is dependent on the volume and rate.
After calculating wellbore volumes and mud pump output, several parameters for the circulating time of the drilling fluid can be estimated. The most routinely used of these parameters are
• surface-to-bit pump strokes, • surface-to-bit time, • bit-to-surface pump strokes (lag strokes), and • bit-to-surface time (lag time).
Because drilling fluid is pumped down the inside of the drill string to the bit, surface-to-bit (s-to-b) calculations are made as follows:
S-to-b strokes = (Total capacity in bbl)/(pump output in bbl/ stroke)
96
PART 3—WELLSITE METHODS
S-to-b min = (S-to-b strokes)/(pump rate in strokes/min)
Drilling fluid is then pumped back up the hole to the surface via the annulus, therefore bit-to-surface (b-to-s) or lag calculations are made as follows:
B-to-s strokes = (Total annular volume in bbl)/(pump output in bbl/stroke)
B-to-s min = (B-to-s strokes)/(pump rate in strokes/min)
Example: Using the previously calculated data, we can calculate circulating time as follows:
S-to-b strokes = (155.4 bbl)/(0.0914-in. bbl/stroke) = 1700 strokes
S-to-b min = (1700 strokes)/(98 strokes/min) = 17.35 min
B-to-s strokes = (378.9 bbl)/(0.0914-in. bbl/stroke) = 4146 strokes
B-to-s min = (4146 strokes)/(98 strokes/min) = 42.30 min
Lag time can also be measured at the wellsite by placing calcium carbide in the pipe when a connection is made. The carbide reacts with the water in the mud and forms acetylene gas that is readily detected by the chromatograph. The stroke counter is reset when the carbide is "dropped" down the pipe, and the number of strokes needed for the acetylene return is noted. The actual and calculated lags are then compared, and an interpretation is made.
Comparison of measured and calculated lag time and a knowledge of v o l u m e and m u d p u m p calculations have several other problem-solving applications at the wellsite. These include the location of a washout (crack or hole) in the drill string, where to spot oil when the drill string is stuck, and the density of the intruding kick fluid.
BIT HYDRAULICS
Hydraulics is a branch of science that deals with practical applications of fluids in motion. Bit hydraulics is important because hole cleaning and drilling efficency are directly affected by horsepower expended at the bit.
Annular Velocity
Annular velocity is the average speed at which the drilling fluid is moving back up the annular space as the well is circulated. Although the mud pump output remains constant, annular velocities vary at different points in the wellbore due to changes in pipe, collar, and hole sizes. Annular velocity (AV) can be calculated as
AV = (24.5 x GPM)/(Dh2 - od2)
Example: If mud is circulated at 400 gal/min in an 8.5-in. hole containing a 4.5-in. drill pipe and 6.5-in. collars, the annular velocities are
DriU pipe: AV = (24.5 x 400)/(8.52 - 4.52) = 188.5 f t / m i n
Collars: AV = (24.5 x 400)/(8.52 - 6.52) = 326.7 ft/min
Jet Nozzle Area
A conventional rotary drilling bit typically has two to four jet nozzles installed to impart a jetting action on the mud to clean the bottom of the hole. Nozzle size is variable and is measured in 32nds of an inch. Note that when jet nozzle sizes are given, the "32nd" is usually omitted. Thus, a bit with "three 13's" installed means the bit has three 13/32-in. nozzles installed. The total area (An) of the jet nozzles is calculated as
An
=
0.000767
+
(J12
+
J
2 2
+
J„2)
Example: The area of jet nozzles for a bit with three 13/32in. nozzles installed is
An = 0.000767 x (132 + 132 + 132) = 0.3889 in 2
Jet Nozzle Velocity
Jet nozzle velocity (JNV) is the velocity of the mud exiting the jet nozzles of the bit and is estimated as
JNV = (0.32086 x GPM)/An
Example: The jet nozzle velocity for a bit with three 13's jets installed and a circulation rate of 400 GPM is
JNV = (0.32086 x 400)/0.3889 = 330 ft/sec
Total Hydraulic Horsepower
The total hydraulic horsepower (THhp) available for drilling hydraulics is defined by the circulation rate and pressure of the m u d pump. Total hydraulic horsepower is calculated as
THhp = (Pp x GPM)/1714
Example: The total hydraulic horsepower available if circulating at 400 gal/ min with a p u m p pressure of 2000 psi is
THhp = (2000 x 400)/1714 = 467 hp
Jet Nozzle Pressure Loss
Pump pressure is the total pressure expended throughout the circulating system's surface equipment (such as the standpipe, kelly hose, and kelly), drill string bore, jet nozzles, and annulus. Only the pressure expended through the jet nozzles accomplishes useful work for drilling. The remaining pressure losses are referred to as parasitic pressure losses. Jet nozzle pressure loss is estimated as follows:
JNPL = (MW x GPM2)/(10,858 x An2)
Example: The pressure lost through three 13's jet nozzles while circulating a 12.0-ppg mud at 400 gal/min is
JNPL = (12 x 4002)/(10,858 x 0.38892) = 1169 psi
Hydraulic Horsepower at the Bit
The hydraulic horsepower at the bit (BHhp) is calculated as for total hydraulic horsepower (THhp), but the mud pump pressure (Pp) is replaced by the jet nozzle pressure loss (JNPL):
BHhp = (1169 x 400)/1714 = 273 hp
The percentage of the total hydraulic horsepower expended at the bit is an important parameter to determine and is calculated in two ways:
% Hhpb - (BHhp/THhp) x 100
or
T% Hhpb = (JNPL/Pp) x 100
Example: Using the data from the previous examples, we can calculate the percentage hydraulic horsepower at the bit as follows:
% Hhpb = (273/467) x 100 = 58%
Wellsite Math
97
or
% Hhpb = (1169/2000) x 100 = 58%
Another key hydraulic parameter that indicates the m a g n i t u d e of p o w e r e x p e n d e d at the bit is h y d r a u l i c horsepower per square inch of bit area. It is calculated by
Hhp/in.2 = BHhp/(0.7854 x Ob2)
Example: Using the previous data with an 8.5-in. bit, we can calculate the hydraulic horsepower per square inch of bit area as
H h p / i n 2 = 273/(0.7854 x 8.52) = 4.8 hp/in.2
Jet Impact Force One of the theories of "optimum" drilling hydraulics holds
that the cleaning action of the bit on the hole bottom is maximized by maximizing the jet impact force. Jet impact force (JIF) is estimated by
JIF = 0.000516 x MW x GPM x JNV
Example: Using the previous data, we can calculate the jet impact force as follows:
JIF = 0.000516 x 12 x 400 x 330 = 817 lb
Mudlogging: Equipment, Services, and Personnel
Alun Whittaker
Sanders and Whittaker Vallejo, California, U.S.A.
INTRODUCTION
Mudlogging, sometimes called hydrocarbon logging or physical formation logging, involves the monitoring and recording of a variety of data related to the borehole and to the drilling process. Mudlogging incorporates gas analysis and cuttings data with drilling information to build a continuous formation evaluation record as the well is being drilled. Equipment and services for mudlogging can range from simple monitoring to integrated computer modeling of the wellsite and borehole environment.
made of all logged data. • Mud test kit—This kit contains equipment to perform
routine drilling fluid analyses and titrations. • Sample processing equipmen—This equipment is used for
catching, sieving, preparing, and testing lithology, mineralogy, and oil content in cuttings (Figures 3 and 4). • Optical equipment—A stereomicroscope is equipped for use with natural and ultraviolet light. • Log production equipment—A typewriter, plotter, and copy system are needed for log reproduction.
EQUIPMENT
Standard Equipment
The m i n i m u m c o m p l e m e n t of e q u i p m e n t for basic m u d l o g g i n g is listed here. Figure 1 shows m u c h of this equipment in a typical mudlogger's work area.
Additional equipment may be needed to perform special on-site analyses.
• Gas extraction—An automatic gas trap is located in the return mudstream for continuous gas extraction (see chapter on "Mudlogging: Gas Extraction and Monitoring" in Part 3).
• Vacuum system—A vacuum pump, pressure regulators, flow meters, filters, and dryers maintain uniform gas sample flow rate and composition from the gas trap and blender. This system includes a pressure and flow alarm and purge system to detect vacuum system blockage or leaks.
• Total gas detector—Hot wire, flame ionization, or infrared detectors are acceptable for total gas detection.
• Chromatograph—An automatic gas chromatograph is equipped with a flame ionization detector (FID) or infrared detector capable of accepting samples from either the gas trap or blender (Figures 1 and 2).
• Gas calibration—A gas calibration system injects known single and mixed gases into the total gas detector and chromatograph in a range of concentrations for calibration.
• Blender—A blender agitates and grinds a discrete sample of mud and cuttings and extracts the released gas for measurement.
• Depth and rate of penetration monitor—A sensor independent of the drilling system is capable of detecting changes in depth and compensating for rig motion offshore (Figure 1).
• Pump rate counter—A p u m p stroke counter is needed for each mud pump
• Mud pit level monitor—Continuous monitoring of mud pit level(s) and rig motion-compensated volume is necessary
• Chart recording—A continuous analog chart recording is
Figure 1. Standard mudlogger's work area. (A) Chromatograph. (B) Chromatograph chart. (C) Back-up chromatograph. (D) Depth and time monitoring equipment. (E) Power panel. (Photo courtesy of Long Co. Technical Service, Casper, WY.)
98
Mudlogging: Equipment, Services, and Personnel
99
Figure 2. Flame ionization detection (FID) chromatograph. (Photo courtesy of Long Co. Technical Service, Casper, WY.)
Figure 3. General layout of mudlogging unit. (A) Sample examination area. (B) Mudlogger's work area. (C) Sample preparation area. (Photo courtesy of Long Co. Technical Service, Casper, WY.)
Figure 4. Sample examination area. (A) Binocular microscope. (B) Ultraviolet box. (C) Radioactivity metering device. (Photo courtesy of Long Co. Technical Service, Casper, WY.)
Figure 5. Standard mudlogging unit exterior. (Photo courtesy of Long Co. Technical Service, Casper, WY.)
Logging Unit Equipment
The equipment used in mudlogging is housed in a portable "unit" or small trailer (Figure 5). Most mudlogging units have similar physical layouts. Electronic instruments are typically housed at one end of the unit, while a sample handling area is along one wall and a work area along another wall (Figure 3). Sleeping and eating facilities may also be housed at the far end of the unit.
For extremely hazardous locations, mudlogging units are constructed of steel plate, pressurized to prevent gas entry, and equipped with insulating fireproof and explosion-proof components.
SERVICES
Standard Services
Standard mudlogging services include the following:
• Monitoring rate of penetration (including factors that can affect ROP response, such as changes in weight on bit, revolutions per minute, and pump pressure) (see chapter on "Rate of Penetration" in Part 3)
• Monitoring of drilling activity • Monitoring of gas (see chapter on "Mudlogging: Gas
Extraction and Monitoring" in Part 3) • Drill cuttings analysis (including Hthology breakdown
and observed versus interpreted lithology) (see chapter
100
PART 3—WELLSITE METHODS
on "Mudlogging: Drill Cuttings Analysis" in Part 3) • Show evaluation (factoring in data such as ROP,
lithology, cut type, and intensity) (see chapter on "Show Evaluation" in Part 3) • Safety considerations, such as checking for flow and sudden gas increases
Specialized Services
In addition to the standard services, certain specialized mudlogging services can be provided:
• Pressure detection • Geochemical analyses (such as cation exchange
capacity, calcimetry, and pyroanalysis) • Petrophysical analysis • Measurements while drilling (MWD) • Communications
PERSONNEL
Standard mudlogging is generally performed by a twoperson crew that works 12-hr rotating shifts, ensuring continuous monitoring of wellsite activities. The on-tour logger catches, prepares, and analyses the drill cuttings and records this information along with other data from the instrument panel and rig drilling records.
During periods of rapid drilling or when multiple special analyses are run, an additional logger or sample catcher is often added to each tour.
In some mature exploitation areas, companies will hire one mudlogging contractor to monitor several wells at a time. In such cases, rig crews are required to catch cuttings samples for later analysis by the logger. These samples are often caught as time allows and are not as useful as cuttings that have been systematically caught and lagged (see chapter on "Wellsite Math" in Part 3).
Mudlogging: The Mudlog
AlunWhillaker
Sanders and Whittaker Vallejo, California, U.S.A.
INTRODUCTION
Information gathered by the mudlogging crew is first recorded on a data sheet (Figure 1) and then plotted onto a mudlog form (Figure 2). Each mudlogger is responsible for plotting the data acquired during Iiis or her tour. In this way, the mudlog becomes an accurate and up-to-date record of nearly all activity related to the drilling of the well.
It is very important that a mudlog be recorded on a standard format using standardized plotting and reporting procedures. This is necessary to allow valid and convenient comparison of log data from well to well and to prevent confusion or misunderstanding over the meaning of a log trace or annotation.
material that arrives at the surface as drilling takes place (see chapter on "Mudlogging: Drill Cuttings Analysis" in Part 3). Information is commonly presented as lithological symbols, each representing 10% of the total sample volume. This is commonly referred to as a percentage log.
Some mudlog formats have an interpreted lithology column adjacent to the percent lithology. This interpreted lithology log attempts to take into account factors that might affect cuttings recovery, such as sloughing and trips.
Additional reservoir information can also be recorded in track 2. In the example log shown in Figure 2, a separate column is included for qualitative porosity estimation.
Track 3: Depth Annotation
LAYOUT OF A MUDLOG
Figure 2 is an example of the layout and information content of a standard m u d log. The general usage of the tracks (columns) on a mudlog are outlined here.
Log Headings
The log heading identifies mudlog and records parameters used in preparing the log. The various headings are as follows:
• Operator and well name • Location • Rig name and type • Elevations • Dates and times • Mudlogging company and crew • Other mudlogging services • Casing sizes and depths • Mud type and characteristics • Mudlog symbols and abbreviations • Gas calibration information
Track 1: Rate of Penetration (ROP)
This track gives a relative indication of rock strength and porosity. Track 1 includes the following information:
• ROP and normalized ROP (NROP) (see chapter on "Rate of Penetration" in Part 3)
• Bit number, type, and diameter • Number and size of jets • Bit run (time per feet drilled) • Bit grade (wear and damage) • Weight on bit • Rotary speed • Mud (standpipe pressure and flow rate)
Track 2: Cuttings Lithology
Track 2 identifies and records the lithology of the cuttings
Depths are noted in track 3 using the traditional scales (in the United States) of 2 in. to 100 ft (1:600) and 5 in. to 100 ft (1:240). This track also includes the location (depth or interval) of the following:
• Conventional cores • Sidewall cores • Drillstemtests(DSTs) • Wireline formation tests
Tracks 4 and 5: Hydrocarbon Evaluation
These two tracks consist of the following information:
• Total combustible hydrocarbons from gas trap, mud, cuttings blender tests, and nonformational borehole gas events such as trip gas, carbide lag checks, and connection gas are recorded.
• Hydrocarbon chromatograph with gaseous alkanes (methane, ethane, propane, butanes, and pentanes) are plotted as individual tracks. Chromatograph calibration checks are also included, providing they do not obscure well data.
• Oil evaluation includes subjective and qualitative evaluation of oil staining and solvent-cut oil. (Many mudloggers include these data in the lithology track.) This evaluation can also include qualitative analysis of oil fluorimetry, refractometry, and geochemistry.
Track 6: Geological and Lithological Evaluation
Track 6 is subdivided into the following categories:
• Mineralogy and geochemical analyses • Interpreted lithology log • Lithological descriptions • Comments, which include special analyses, reports of
borehole conditions, and reports of drilling events that have an effect on overall log interpretation (such as loss of returns and power failure)
101
102
PART 3—WELLSITE METHODS
I 0 LNGTH
Ii
«
Ii 1
S и ^( |1
I IS §
8 S S
а СП 5
f r (СкГПH
*
S
о
S
tлm
—
—Ньl иlIEОiг
I а
§•:*8са
D
О
:
SГ5SOSOS но- Л>и \\сLl ь & | н 5 S - 1и
I
5
3
гOI O,9 ££
CCI iОlI t ~
8 о
О с а а а рр
С E
ее I
й И en cn %
US f$ii.
8* СQZП S
So 2 ?
mi
22
Ю
«*Я<->
35 |о
Isf-I
ШОС Э»-
5 I= I I l l i U 8*
QООfОc® QОС2Sи iDu о8 **
о >
\ I
ir
Ii
I I
С
I
I
-
!
I
I
I
I 5
I
« у
—I
5 ->cr>5ГZL
— —
Щ 1
ф2ЙГ5С ээас: 2 Г 5 S •о
8 I
.
I WFl I COMPANY
I I
I I
<0D)
ST(ГяЛZ3О
1
Ос ) '5О))
о
тз
3 E
со
ч—
О
Q)
О.
EгXо
ш
DО)
Mudlogging: The Mudlog
103
Generic Logging, Inc.
I.AT—Logtfd Atlcr Trip
CO—Circulated Oul
NO-No Returns
PH
DST-DriII Stem Test NB-New Bit
Mud
DaU
DRILLING RATF.
Mm. / Ft
mmmm
I ITHOI.OGY
Sa%mpleIns
DOLOMITE _ SH.TSTONE_. ANHYDRITE _
Depth & Racord
COMPANY. WELL SEC_ COUNTY.
SUPERVISOR.
UNIT
PROFfTABLE ОД.
BOOMER
CHROMA IOGRAPH
MUD GAS ANALYSIS
Div 5QUNITS
Propane) Д?У. Div. JOO UAITS
Ethanev
Div IOOUNITS
O
Methane у 2i54 niv IQO UNITS
LH HOLOGICAl. DESCRIPTION
and other Pertinent DuU
MililUbdTltfftFF
Il H
л i.j ,1 S
f 4 Ht h
Sl i-
/ • •л \
Kl
I-
(Я I-
1
1уi-i
W и I
U'ili Slli Al \ V)1 г Il/Г,
-
мир»'
WT 8.8 VIS 4 4 WLIO.I PH95 CUiOO
-.JA
•—
,s ш AAJi 1Шй I Л1 HWШ
IM ( Kl INlIS Va Ilr im/LirI IV. i, WI TIS • IWi .V i/i lit
IVnс, (I"!Iir 11 IXI V 1YYIAiill
7*T К N ппмл (К •Ti'I' PSIwiYulIit
—
tv -4
--
— -
-
-
...
>
- '—
—
— -—
-
-
\
.
у I
./
I I...
T| 1F N SRI
-
ас
B
Г
MNl RJIj Hl
Jl- I ; -I
г
t
L.
J
T - ——
г Ь •ь 1 >f W
-
—
" I
/ 1
" f Г /
SH-DKGY-MGY MOO FRM-FRM BLKY RTHY SL CALC T R - N CARB SL SLTY
MOD F R M - F R M
BLKY RTHY BRIT
—
CALC SL CARB
SLTY IP TR X L N -
-
RHMB CALC
I ITl L/1-ч VJI1VJI ОП1\ MOD F R M - F R M
BLKY RTHY BRIT
CALC S L S L T Y -
SDY IP INCR W /
DPTH SL BENT LT
GY S F T S B W X Y
Ffr
гЧ-
s a p FRI-MODFRV
F-MGR SBANG-
SBRD P SRT CLY FlJ
SL ARG TR GLAU
SLCALC P-FR 0
FR-G SPTY YEL
FLOR STMG CRSH CUT V FNT BRN STN NOIL H-INTBD THRUM - D K GY MOD FRN BLKY RTHY BRIT
BRLTKHYY-FSIBSBPLRTrriTr R
CARB TR BENTW/1
Ь М INTBD SS STRC
WH-LTGY VF-F
IV
GR CLY F L TT NSF
BLKY-SBPLTY RTHY BRIT TR CARB TRBENT W/ SM I N T B D SS STRG W H - L T GY V F - F GR CLY FL TT NSFC
Figure 2. Example of a standard mudlogging form. Note that although efforts are continually being made to standardize mudlog formats, present and previous mudlogs may differ in outline from this example.
Mudlogging: Drill Cuttings Analysis
Alun Whittaker
Sanders and Whittaker Vallejo, California, U.S.A.
Diana Morton-Thompson1
Consultant Kalamazoo, Michigan, U.S.A.
INTRODUCTION
Cuttings are the small pieces of rock that are chipped away by the bit while a well is being drilled. The rock fragments are transported by the mudstream from the bit to the surface where they can be "caught" and analyzed (see chapter on "Land Rigs" in Part 3).
Drill cuttings are important because they are often the only physical lithological data that are recovered from a well. These data are used as supportive information for subsequent show evaluation, correlation of lithology to wireline logs, and other special geological, geophysical, or engineering analyses (see chapter on "Show Evaluation" in Part 3). Some of the activities that utilize data obtained from drill cuttings are summarized as follows:
• Show evaluation • Reservoir and lithological description • Geological correlation and formation identification • Verification of wireline log response
CATCHING REPRESENTATIVE SAMPLES
Sample Interval
Both the sample interval and the number and types of sample sets are usually determined at the start of the well (see chapter on "Well Planning" in Part 3). Typically, drill cuttings are caught or collected as composite samples that reflect the various lithologies drilled over a 10-ft interval. The interval length can vary depending on the drilling rate and the detail needed for reservoir and lithological descriptions. When the drill rate is low, it is important to collect a sample that is representative of the whole interval drilled and not just the final few feet. It is also important to collect representative samples when drill rates are excessively high and it is physically difficult for the mudlogger to collect and analyze all of the samples accurately. In this situation, the sample interval can be increased or an additional mudlogger can be assigned to the tour. Also, it is not unreasonable to request that the drill rate be slowed so that better samples can be obtained and used to evaluate prospective reservoir zones and casing or kick-off points.
When a drilling break or change in background gas occurs, supplemental samples should also be caught to aid in show evaluation.
Sampling Location
Drill cuttings are typically collected from the rig shale shakers, a sluice box in the possum belly, or the desander/desilter outlets (see "Land Rigs" chapter).
Shale Shaker Most rigs have one or more shaker screens for separating
the cuttings from the mud. As returning mud travels across the screens, the mud fluid falls through the mesh, while solids travel to the end of the screen. A board or catching box is placed at the foot of the screen to catch the composite sample and prevent the samples from falling into the mudpit or onto the ground.
When a "double deck" shaker is used, cuttings on both the upper and lower screens should be sampled.
Possum Belly Another method of catching drill cuttings is to place a
bucket or metal box in the possum belly near the entrance of the flow line. This method is thought to help prevent the disaggregation of sandstone and conglomerate cuttings.
Desander/Desilter Unconsohdated samples can be taken from the desilter or
desander outlets if they use the underflow from the shakers and are not supplied from a pit. Samples from the desander contain both formation particles and solids related to the drilling fluid. It is the responsibility of the mudlogger to discriminate between the two.
SAMPLE LAG
Because there is a time difference between when the rock is first broken away from the formation and when the sample is caught at the surface, all cuttings samples should be "lagged" or adjusted to the proper depth (see chapter on "Wellsite Math" in Part 3). This process is critical in showing evaluation and Hthological correlation.
SAMPLE PREPARATION
Once the composite sample of drill cuttings is retrieved from the mud system, it is typically split or subdivided into a bulk, unwashed wet-cut sample and a washed and sieved drycut sample.
1Formerly with Chevron U.S.A. and ARCO Research. 104
Mudlogging: Drill Cuttings Analysis
105
Wet-Cut Sample
A small portion of the unwashed wet-cut sample is set aside and used for blender gas analysis (see chapter on "Mudlogging: Gas Extraction and Monitoring" in Part 3). Although this sample should not be rigorously washed, discretion can be used in lightly rinsing the sample to remove surface drilling m u d film. The remainder of the sample is packaged for later analysis and archiving.
Dry-Cut Sample
These samples are washed and sieved before they are analyzed. Consolidated cuttings are cleaned by washing or hosing the sample in a container of water to remove the mud film. Washing of poorly consolidated sandstones and shales is more difficult and requires several precautions. Clays and shales are often soft and tend to be washed away. In samples with swelling clays, it is often helpful to wash the sample in f o r m a t i o n w a t e r or a 2% s o l u t i o n of KCl to p r e v e n t defloculation.
Cuttings from wells with oil-based muds are usually more representative of the formation than cuttings from waterbased mud because the oil emulsion prevents sloughing and dispersion of clays and shales. However, these samples cannot be cleaned by washing in water alone. It is usually necessary to wash the cuttings in a detergent solution to remove the drilling fluid. In extreme cases, it may be necessary to wash the cuttings first with a nonfluorescent solvent and then wash them in a detergent solution. Use of a solvent is not advisable unless absolutely necessary because of the risk of removing any oil staining present. Solvents and material contaminated by solvents should be properly d i s p o s e d of in a c c o r d a n c e w i t h safety a n d h a z a r d o u s materials procedures.
After washing the cuttings to remove any drilling mud, the samples are typically washed through a 5-mm sieve. Particles greater than 5 mm usually contain a high percentage of lost circulation material and cavings. This material should be cursorily examined and discarded.
BASIC SAMPLE ANALYSIS
A split of the washed and sieved cuttings is examined wet under a binocular microscope. From this sample, the mudlogger estimates the percentages of the various sample constituents and records the following information:
• Rock type and lithological composition • Color • Hardness (induration) • Grain size • Grain shape • Sorting • Luster • Cementation or matrix • Sedimentary structures • Porosity
• Hydrocarbon show - Stain - Odor - Fluorescence - Cut - Gas (total and petroleum vapor) (see chapter on "Show Evaluation" in Part 3)
The quality of the description is directly related to the quality of the cuttings. Factors that adversely affect cuttings quality are summarized as follows:
• Excessive weight on bit, which results in cuttings being ground into a fine powder
• Insufficient mud viscosity, which results in cuttings not being transported to the surface; can be a significant problem in highly deviated and horizontal wells
• Improper mud chemistry, which results in (1) a high percentage of cavings that can mask the true drilled Uthology, (2) loss of soluble minerals such as gypsum and salt and some shales, and (3) a high percentage of contamination by cement, lost circulation material (LCM), and metal.
Drill cuttings samples can be laid out in small partitioned trays so that vertical changes in rock properties are easier to observe and interpret. The first occurrence of a specific lithology reflects the highest possible position of that bed. However, because the borehole may tend to cave or slough during drilling, a lithology may continue to be present in samples from deeper drilled depths.
SAMPLE PACKAGING
Wet-cut samples are packaged in cloth bags and labeled with the well name, operator, API number or well location, and sample depth. Dry-cut samples are either dried in the air or under heat lamps. Each dried sample is then placed in a small paper envelope which is labeled. All labeling should be done with a dark, waterproof marker because nonwaterproof markers will inevitably smudge and streak, and pencil marks will fade with time.
The packaged samples may be sent to the operator at regular intervals during the drilling of the well or sent en masse at the end of the well. Sample sets are usually required for all partners and may be required for the appropriate state and/or federal agencies, depending on the location.
Samples for Special Analysis
In general, samples used in special analyses should not be artificially dried or stored near intense heat (such as in the "dog house" near the generators). This is because the heat causes more of the Hght hydrocarbons to volatilize, which will bias subsequent analyses. If precision is required, the bulk wet-cut samples should be placed in glass jars and covered with formation water (or a compatible fluid). Bactericide should be added and the containers sealed and labeled.
Mudlogging: Gas Extraction and Monitoring
Alun Whittaker
Sanders and Whittaker Consultants Vallejo, California, U.S.A.
INTRODUCTION
The monitoring of gas—both types and amount—is one of the most critical jobs is mudlogging. Early detection of a gas influx will give rig crews time to respond to toxic gases or a potential blowout condition. Accurate gas data recorded during drilling is of great value in proper reservoir evaluation and may even pinpoint potentially overlooked producing zones.
GAS EXTRACTION METHODS
Samples of gas m u s t generally be extracted f r o m the drilling mud by the gas trap. The gas trap is a metal box immersed in the shale shaker possum belly, preferably in a location of maximum mudflow rate (Figure 1). Ports in the lower part of the trap allow mud to enter and leave the trap. An agitator provides both pumping and degassing of mud passing through the trap. Gas evolved from the mud is mixed with ambient air in the upper part of the trap and is then drawn through a vacuum line to the logging unit for analysis (Figure 2).
Inside the logging unit, the gas-air mixture is drawn through a sample chamber containing a heated platinum filament forming one arm of a Wheatstone bridge circuit. Catalytic combustion of hydrocarbon gases at the filament (the so-called hot wire) further heats the filament, causing an increase in its electrical resistance. This change in resistance unbalances the bridge, causing an electrical current to flow in proportion to the resistance change and a resulting concentration of gas. The current is displayed by deflection on a milliammeter or may be used to scale the meter or recorder in terms of percentage or "gas units" by calibrating
the hot wire periodically with a gas-air mixture of known concentration. A second hot wire detector, set at a lower t e m p e r a t u r e , a l l o w s d i s c r i m i n a t i o n of total gas (all c o m b u s t i b l e h y d r o c a r b o n gases) f r o m petroleum vapors (combustible hydrocarbon gases other than methane). This aids in identifying the types of hydrocarbons in place (that is, oil or gas).
EXTRACTION EFFICIENCY
The efficiency of the device can vary from 30 to 70% and is somewhat affected by drilling practices. Pump rate, mud level, and rheology influence flow rates through the trap and are factors in the efficiency of the trap (see chapter on "Show Evaluation" in Part 3). Mud and ambient air temperature around the trap and vacuum line affect the relative efficiency with which light and heavy hydrocarbons are extracted and retained in the gas phase. Tliis latter effect is most noticeable in areas of high diurnal temperature.
Numerous mechanical factors can affect trap efficiency:
• Revolutions per minute and surface area of impeller blades
• Vacuum strength • Ditch or trap blocked by cuttings • Air mixture
Extraction efficiency is also affected by the following factors:
• Drilled volume—The volume of gas seen in the drilling mud depends upon the volume of formation cut by the bit. Rate of penetration (ROP) increases will increase
Figure 1. Mud system. (From Whittaker, 1991.)
Figure 2. Location of gas trap in possum belly. (From Whittaker, 1991.)
106
Mudlogging: Gas Extraction and Monitoring
107
A /
I
I
mirrnHcs
I
(шкпюпита) I
•итйис
J !-«urine I raonw
I
I I
ICRRfUOl
J CTHMC J ЛТИЯМС J MR J
CWJHUtTIOK MJTOfIAriC МТССААТЮМ
I
«г
I
^ЯЕШизвд.
PtHTAKS I
I
(UHMTtROCTnTtD) I
coHcacTimnoM Ctrm ы a * ) - M IM(M • CaakrtllM - И ~ • Cs ,«Ic. COHCtMTRATIOU ( f f m Im Mr) - P<>k Atm • CdDriIiM
- Г С н " • » ' ) • Cj ,.Ic.
Figure 3. Gas chromatograph system. (From Whittaker, 1991.)
Figure 4. Peak heights from chromatographic analysis. (From Whittaker, 1991.)
the volume of the formation cut per unit of mud pumped and therefore the volume of gas liberated. However, ROP increases may reflect increasing porosity and, at constant gas saturation, increasing gas Uberated. Comparing ROP, cuttings lithology, and all three total hydrocarbon curves may help in interpretation. • Crushed volume—Tlie crushed volume of formation is recovered at the surface in the flowing mudstream. If the mudflow rate is increased, the cuttings are dispersed in a larger volume of mud, and the percentage of gas in mud is reduced. • Permeability—If permeability is excellent or nonexistent, then the mudflow rate will have no effect on mud and cuttings blender test results. However, with moderate permeability, the mudflow rate and the resulting lag time will have some effect on the amount of gas that is allowed to expand and escape from the cuttings into the mud. • Pressure and temperature—As the cuttings are recovered at the surface, both pressure and temperature decline. With declining pressure, the gas expands and escapes from solution in oil or water and all three fluids are displaced from the cuttings. Conversely, declining temperature causes heavier alkane vapors to condense, decreasing the total fluid volume and the proportion of gas in the mud and cuttings (see chapter on "Show Evaluation" in Part 3).
All of these factors combine to m o d i f y the type a n d quantity of gases contained in the cuttings and m u d and extracted from them at the surface. The influx of fluids from the borehole wall and cavings will also have an effect. None of these effects are sufficient alone or in combination to invalidate the total hydrocarbon log as an indicator of gasfilled porosity in the formation. It does mean, however, that care must be taken when comparing total hydrocarbon indicators with rate of penetration.
An alternative to the conventional gas trap is the steam or
vacuum mud still. In this device, a small sample of drilling mud is collected at the ditch and distilled under vacuum. The method provides a relatively high and uniform extraction efficiency for all hydrocarbons. It is, however, a timeconsuming manual process.
GAS CHROMATOGRAPHY
Most logging units contain a gas chromatograph (see chapter on "Mudlogging: Equipment, Services, and Personnel" in Part 3). The hydrocarbon gas chromatograph is probably the most accurate and most consistent of the data recorded on the mudlog.
In gas chromatography, a fixed volume gas sample is carried through a separating column by a carrier gas, usually air (Figure 3). The column contains liquid solvent surface or a fine molecular sieve solid. By difference in gas solubility or by differential diffusion, the gas mixture becomes separated by molecular weight into its components, the lightest traveling quickly through the column and the heaviest more slowly. Common alkanes detected within a reasonable time frame include
• Methane(Cl) • Ethane (C2) • Propane (C3) • Isobutane (C4) • Pentane (C5)
The chromatograph does not provide a continuous analysis but rather processes batch samples separated by a number of minutes. The gasses are usually presented as units based on peak height (Figure 4). They can also be noted in parts per million, percentage, mole fraction, or British thermal units. Integration of the area under the curve gives a better r e p r e s e n t a t i o n of the p r o p o r t i o n s b u t requires a m o r e expensive device.
108
PART 3—WELLSITE METHODS
Calibration
Chromatographs are calibrated by injecting a known volume of test gas into an injection port. The test gas consists of an accurately m e a s u r e d mixture of m e t h a n e , ethane, propane, isobutane, and normal butane. As the test gas cycles through the chromatograph, curve peaks are noted for each
component gas. Formation gas readings can then be compared to the cycle peaks to derive wellbore gas values. Calibration should be done routinely on a daily basis and additional checks should be performed shortly in advance of penetrating critical zones. The Society of Professional Well Log Analysts (Kennedy, 1984) has published standards for the calibrating of total gas and chromatographs detectors.
Show Evaluation
Paul A. Daniels, Jr.
Kalamazoo, Michigan, U.S.A.
David B. Finnell
Epoch Well Logging, Inc. Bakersfield, California, U.S.A.
William J. Anderson
Epoch Well Logging, Inc. Ventura, California, U.S.A.
INTRODUCTION
Show evaluation at the wellsite is important because it represents the first, and sometimes only, opportunity to assess the potential economic viability of a particular well or prospect. Decisions of considerable economic impact are made based on show evaluation results, including formation testing, setting pipe, participation elections, and lease acquisition or relinquishment.
Mudlogs are the most useful wellsite evaluation tool due to the integration of lithology, rate of penetration (ROP), gas recordings, and oil description (see the chapter on "Mudlogging: The Mudlog" in Part 3).
Wellsite show evaluation relies on the following:
• Detection of formation gas or oil • Detection of hydrocarbons in drill cuttings • Knowledge of drilling and wellsite activities • Geological knowledge of the interpreter
Figure 1, a show evaluation form, summarizes some of the criteria commonly used to evaluate a show. Note, however, that the presence or absence of a recordable show does not absolutely determine hydrocarbon producibility. For example, zones of good porosity and permeability that yield no show may still be viable hydrocarbon-bearing reservoirs. Conversely, poor porosity-permeability zones that have gas increases may be nonproducible. For the most effective formation evaluation, wellsite shows of all types should be integrated with the results of postdrilling logging and testing.
GAS DETECTION
A gas show is a recorded increase in gases above a baseline a m o u n t indicative of the h y d r o c a r b o n potential of the formation. This increase is assumed to be independent of any borehole drilling or circulation process. Table 1 summarizes the nomenclature of gas that can result from the drilling process. The presence of these gases should not be confused with a "show."
The analysis of a gas show begins with the detection of hydrocarbon gases that are the result of drilling a specific interval. The amounts and compositions of these gases are recorded on gas detection and chromatograph equipment. The simplest and most widely utilized gas detection device is
the hot wire. This type of detector can provide readings of the total gas present (in units or percentage) as long as the gas concentrations are low enough that the equipment does not become "saturated" (on most hot wire equipment, gas percentages must be below a certain maximum for accurate readings).
Flame ionization gas detector (FID) equipment operates on a different principle and does not saturate. An FID detector can also provide chromatographic analysis of the gases. Chromatograph recordings normally distinguish the a m o u n t s of m e t h a n e (Cl), e t h a n e (C2), p r o p a n e (C3), isobutane (IC4), normal butane (NC4), isopentane (IC5), and normal pentane (NC5). The greater the amount and p e r c e n t a g e of a gas s h o w r e p r e s e n t e d by the heavier hydrocarbons (particularly C4 and C5), the greater the potential for oil production from that zone (Ferrie et al., 1981). However, the presence of anomalous, noncombustible gases such as nitrogen or carbon dioxide can attenuate hydrocarbon gas curves.
Ratios of the different gases are often used to do the following:
• Support show identification • Cross-check lag time calculations • Discriminate among different show zones • Indicate possible "richness trends" • Identify fluid content • Provide stiatigraphic correlation
Various gas ratios can be used depending on the data available. The most common gas ratios used are those with the most separation (C5/C1) and those with the heaviest composition (C4/C1 or C5/C1). Because gas ratio analysis is empirical in nature, it can sometimes prove inconclusive. However, the following "rules of t h u m b " can be useful (Exploration Logging, Inc., 1985):
• Zones with a high Cl value may represent dry gas, coal, biogenic gas, or a water wet zone.
• Wet gas zones commonly have a Cl /СЗ ratio that is higher than the Cl /C4 ratio.
• Nonproductive zones tend to have a ratio trend where subsequent values are lower than preceding values.
109
110
PART 3—WELLSITE METHODS
SHOW EVALUATION FORM
Well name and operator WeB location: K.B7G.L elevation:
(tUm.\
Date (local): T m e (local arrVpm);
DRILLING OPERATIONS/WELLSITE ACTIVITIES
P R E S E N T S T A T U S (depth, drilling, circulating, testing, logging etc.):
SHOW FORMATION:
SHOW DEPTHftop-bottomk
MUD: type: DATA weirtt:
addtives:
viscosity: chloridesfbefore*:
R O P (ft.-m./min.-hr.): before:
djrinq:
LIVE OIL DESCRIPTION(color, amount):
Dits:
tanks:
remarks:
tap:
water loss/filtrate: fehcwV after mudstream:
GAS
Total: C1: C2: C3 IC4: NC4: IC5: NC5:
Backoround
HOT WIRE & FID (units/percent)
Before
CXiina
After
BLENDER
CUTTINGS
LITHOLOGY & POROSITYfdescription and amount):
VISIBLE STAIN:
Cuttings Condition (un)washed dried crushed acidized
none:
Color very light brown light brown medium brown dark brown black
present (circle attributes listed below):
Distribution I%) trace ( %) spotty ( %) scattered ( %) patchy ( %) even/uniform ( %)
Show Evaluation
111
FLUORESCENCE: Cuttings Condition (un)washed dried crushed acidized
CUT:
Cuttings
Condition
(un)washed dried crushed acidized
Cuttings
Condition
(un)washed dried crushed acidized
ODOR: Cuttings
Condition
(un)washed dried crushed acidized
none:_
present (rirrie attributes listed below):
Color
pale yellow brown orange goldyellow white green blue
Irtenaty
very weak (transparent) moderate (translucent) moderate-strong strong (opaque)
Distribution (%)
trace ( %) pinpoint ( %) spotty ( %) patchy ( %) even ( %)
none:
present (rJrrJft attributes listed below):
,
Color
pale yellow yellow yellow/orange light brown dark brown
Normal Light Type
residual streaming
Rate slow moderate fast instant
Intensitv pale dull medium bright
Color
white yellow yellow/orange orange blue
Ultraviolet (UV) Light Type
residual streaming
Rate
slow moderate fast instant
Intensitv
pale dull medium bright
none:
present (rirrJe attribiites listed below): ,
Intensity
faint fair strong
REMARKS & RECOMMENDATIONS:
Interpreter Figure 1. Show evaluation form. (Copyright © 1992 by Paul A. Daniels, Jr., and Diana Morton-Thompson. Used with permission.)
112
PART 3—WELLSITE METHODS
Table 1. Types of Gas That Result from the Drilling Process3
Type of Gas "Zero" gas
Background gas
Liberated gas Connection gas Produced gas Trip gas Recycled gas
aThese are not "shows.'
Description
Gas present in the mud circulating system when the bit is off-bottom and there is no vertical movement of the drill string. This reading results from the liberation of gases from the mud system or from the recycling of previously encountered gases in the wellbore. Although a "zero" gas value will constantly vary, it acts as a starting point for evaluating any subsequent formation gas shows.
Gas that reflects the geological character of a consistent lithology. Background gas readings incorporate gas contributions due to the formation, but also those included as zero gas. Gas from the formation is due to the crushing of the rock as it is being drilled and typically has a low volume. These gas readings are plotted on the mudlog as background gas and represent the relative baseline against which all other gas shows are compared.
Gas that is produced by the drilling process due to the crushing of the rock formation by the drill bit. Formation gas that enters the wellbore while drilling and circulation are halted to make a connection.
For this condition to occur, the contributing formation must be underbalanced by the mud system at some point within the borehole. Formation gas that enters the wellbore while drilling and circulating. This gas represents an underbalanced formation and, if left alone, will cause a blowout. Formation gas that enters the wellbore when the drill string is being "tripped" or pulled out of the wellbore. The contributing formation must be underbalanced at some point within the wellbore; such underbalance is due to the "swabbing" effect caused by pulling the drill string out of the hole. Gas that has been previously contributed to the borehole and not completely removed from the mud circulation system by surface equipment (such as a gas trap or degasser). Such gases that remain in the mud system are pumped back down the borehole to be subsequently re-recorded by the gas detection equipment. This recycled gas can usually be recognized because the "show" will be detected one full circulation cycle later than originally encountered and will appear more diffuse in character.
Factors Affecting Gas Detection
Factors that affect the quality or presence of gas shows include mud weight and wellbore flushing, operation of the surface mud system, and accuracy of calculated lag time (see chapter on "Mudlogging: Gas Extraction and Monitoring" in Part 3).
Mud Weight and Wellbore Flushing
Near wellbore flushing occurs when the pressure or weight of the mud column exceeds the fluid entry pressure of the formation (for information on calculating mud weight, see chapter on "Wellsite Math" in Part 3). This flushing by mud filtrate occurs above and ahead of the bit and is a function of time. If the mud system is overbalanced, gas shows can be reduced or totally suppressed. Even in a carefully balanced mud system where the fluid loss is minimized and the radius of wellbore flushing is small, problems can still occur in evaluating gas s h o w quality if the rock's petrophysical properties are not considered. In zones of low effective porosity, even relatively small volumes of filtrate loss may result in deep invasion profiles. This causes a zone with a good gas show when drilled to recover only mud filtrate or to appear water saturated when later tested or electric logged. In zones of liigh effective porosity and permeability, the rocks will initially be flushed, then return to their native state soon after drilling, with little or no gas liberated. This causes a zone with minimal gas show when drilled to appear productive on electric logs or when later tested. Low
permeability overpressured zones will not flush and will give high gas readings.
Information that can aid in the interpretation of flushed anomalies includes the following (Exploration Logging, Inc., 1985):
• Pump pressure • Jet nozzle size(s) of the bit • Mud rheology (plastic viscosity and yield point) • Mud weight and effective circulation density • Formation balance gradient (mud weight required to
equalize formation pressure) • Mud filtrate (water loss) amount • Lithology descriptions of
- visual porosity (absolute, effective) - porosity description (type, size, distribution) - cementation (type, degree, mineralogy)
Surface Mud System
Flowline. A high degree of degassing takes place in the conductor pipe and flowline. Loss of gas from the mud to the atmosphere also occurs extensively in the flowline (Exploration Logging, Inc., 1985), particularly when
• The flowline is not filled with mud • Changes in flowline slope promote turbulence • Sections of flowline are open to the atmosphere • The flowline enters the possum belly above mud level
Show Evaluation
113
Table 2. Fluorescence of Common Minerals and Artificial Materials
Mineral or Material
Fluorescence Color
Minerals3
Dolomite, magnesian limestones
Yellow, yellowish brown to dark brown
Aragonite and calcareous mudstones
Chalky limestones
Yellow-white to pale brown Purple
Foliated shales Anhydrite Pyrite
Tan to grayish brown Blue to mid-gray Mustard yellow to greenish
brown
Diesel fuel
Pipe dope Oil-based mudb
Artificial Materials Dull brown Bright blue Varies
aModified from Exploration Logging, Inc. (1985).
bSampIes of oil-based mud and other petroleum products used around the wellsite should be routinely sampled and examined under UV light to avoid potential confusion with hydrocarbon shows from the rocks.
In extreme situations, the amount of gas lost from the mud system prior to analysis can be so great that any resultant shows are significantly biased.
Gas Trap and Agitator. This equipment is usually in or around where the flowline enters the possum belly. If the equipment is not operational (plugged orifices or lines, not powered-up, etc.), insufficient gas may reach gas analysis equipment to allow for an accurate analysis (see chapter on "Mudlogging: Gas Extraction and Monitoring" in Part 3).
Gas Analysis System. If the detection and analytical equipment are not properly and constantly maintained and calibrated, inaccurate gas detection and analyses can result. Because there are numerous ways that the drilling, the gasgathering system, and equipment wear can affect this analytical equipment, a complete system check and calibration is recommended on each tour.
Lag Time
It is of the utmost importance to know the exact depth that all drill cuttings samples and gas are coming from within the borehole (for information on calculating lag time, see the chapter on "Wellsite Math" in Part 3).
CUTTINGS EVALUATION
In many wells, drill cuttings collected may represent the only subsurface data available for geological interpretation. After a detailed lithology description, cuttings are analyzed for hydrocarbon indications (see the chapter on "Mudlogging: Drill Cuttings Analysis" in Part 3). Traces of gas and oil in the cuttings represent formation hydrocarbons that have not been flushed by the drilling fluid. Gas in cuttings is analyzed by
grinding a measured a m o u n t (approximately 100 mg) of unwashed cuttings in a blender, with any liberated gases analyzed by the standard gas detection system. This analysis is often divided into two components: total gas, comprising all combustible gasses; and petroleum vapors, comprising C2 through C5. This type of analysis can indicate the amount and composition of gases in the formation, even if the larger rock pores are flushed.
Evaluation of oil in cuttings is performed on unwashed and washed bulk cuttings and on individual grains. Evaluation includes visual inspection and analysis using a microscope and ultraviolet (UV) box. Oil shows are described by their physical properties of visual stain, fluorescence, cut, and odor. Care must be taken always to evaluate hydrocarbon shows in cuttings with respect to their petrophysical properties (see review by Swanson, 1981).
Visual Stain
Staining of the drill cuttings by oil is an indication that hydrocarbons have been in the formation at some point in time. The lack of sample staining, however, does not prove that a reservoir lacks producible hydrocarbons. The amount and distribution of staining is a function of the reservoir porosity and permeability. Stain color can be related to oil gravity, with darker staining indicating heavier hydrocarbons. If a stained sample does not fluoresce or cut, then this indicator is classified as thermally "dead oil" and is not considered a show. Staining is described in terms of its color, d i s t r i b u t i o n , p e r c e n t a g e of s a m p l e s t a i n e d , a n d fluorescence (if any).
Fluorescence
Fluorescence refers to the color of the drill cuttings under UV light of various wavelengths. A lack of fluorescence, however, does not prove the absence of hydrocarbons in the z o n e of interest. C a r e m u s t be t a k e n to d i s t i n g u i s h hydrocarbon fluorescence from natural minerals or artificial materials (Table 2). Ruorescence is described in terms of its color, intensity, distribution, and percentage of s a m p l e fluorescing.
Cut Fluorescence
A cut is the oil liberated from drill cuttings when a solvent is added. A common solvent used for inducing cuts is chlorothene; others include acetone, petroleum ether, alcohol, hot water, and acid. Note that because of its toxicity, carbon tetrachloride should not be used. Most solvents are flammable, and great care must be taken to handle these materials safely. A cut is performed while viewing the rock samples under both normal and UV light. Solvent cuts allow deductions to be made regarding oil mobility and reservoir permeability. A cut is described in terms of its natural color, fluorescence color, "liberation" rate and intensity, and residue. All suspected hydrocarbon-bearing intervals should be tested for cut fluorescence. This is because there may be a positive cut fluorescence test when other hydrocarbon detection methods fail.
114
PART 3—WELLSITE METHODS
Odor
The odor of hydrocarbons may be present even in the absence of any other hydrocarbon indicators. This condition is most noticeable during the sample drying process, when lighter hydrocarbons are driven off. Odor is described as faint, fair, or strong (indicative of heavier hydrocarbons). Keep in mind that methane through butane have no odor.
Factors Affecting Cuttings Evaluation
Attention to differential pressure (over- or underbalanced), ROP, hole size and condition, contaminants, recycling, etc. is necessary for proper cuttings evaluation. Also, as there is no substitute for representative cuttings samples that are correctly correlated to depth, the importance of an accurate lag time cannot be overemphasized.
DRILLING OPERATIONS AND WELLSITE ACTIVITIES
Wellsite evaluation of hydrocarbon shows requires clear and constant communication with vendor and drilling company personnel.
Drilling activities that represent shows include
• Raring gas
• Oil in the pits, tanks, or mud system
Activities that may not represent a show include • A drilling break or increase in the ROP (see chapter
on "Rate of Penetration" in Part 3) • An increase in the mud tank level
Activities that are not a show but can be mistaken for a show include
• Rig maintenance (lubrication, washing diesel into the mud pits)
• Mud additives • Lowering mud weight
Activities that can lead to overlooked shows include
• Drilling with a constant ROP (controlled drilling) • Overbalanced mud system • High mud filtrate loss • Losing circulation • Bypassing the shaker • Electrical power fluctuations
Conventional Coring
Leewhitebay
Conoco Inc. Ponca City, Oklahoma, U.S.A.
INTRODUCTION
This c h a p t e r p r e s e n t s an o v e r v i e w of coring tools, including guidelines for selecting coring tools for specific applications. Coring provides the only means of obtaining high quality samples for the direct measurement of rock and reservoir properties. Cores provide both geological and engineering information, and their analysis ultimately leads to a profitable field development.
Many coring systems exist. The system used depends on the objectives for the coring program and on the physical constraints of both the formation and the drilling location. A remote location might require a ready-made core-handling and preservation system such as a disposable inner barrel. An exploration program might profit by employing a wireline core barrel to core and drill alternately without the expense of tripping pipe. A m a t u r e field might need a sponge or pressure core to help pin down remaining reserves. These three situations describe three coring programs, each with different objectives and constraints and each with a different "best" way to get the job done.
CONVENTIONAL CORING SYSTEMS
Designed to recover core from consolidated formations, conventional coring systems consist of an inner core barrel with a core catcher, suspended by a swivel assembly inside of an outer core barrel that is attached to the drill string, and a core bit (Figure 1). As mentioned previously, the final selection of a particular system depends upon the formation, location, and objectives of the coring program.
Types of inner barrels are summarized in Table 1. Core diameter can vary from 1.75 to 5.25 in. Core length can range from 1.5 ft for short radius horizontal wells to greater than 400 ft for vertical holes.
The core catcher, the device that holds the core in the barrel, is tailored to the type of inner barrel and lithology expected. Table 2 lists core catchers by their common names and usages. In some cases, multiple catchers are used. Friable sandstone interbeddded with shale might require both slip and flapper type catchers. Full-closure catchers, run primarily to ensure success when coring unconsolidated sand, also incorporate split ring or slip type catchers to improve core recovery in the event that coring ends in hard rock.
Salety Joint
Bearing Assembly
Orop Ball
Core Barrel Stabilizer Inner Barrel
SPECIAL CORING
Heavy Duty Conventional Core Barrels
A heavy duty core barrel should be considered when cutting long lengths of relatively homogeneous formations or
Figure 1. Conventional core barrel. (CourtesyofEastman Christensen, Technical Data Sheet C-105.)
115
116
PART 3—WELLSITE METHODS
Table 1. Conventional Coring Systems
Inner Barrel
Mild steel Mild steel High strength steel
Core Length (ft)
30-120 1.5
120->400
Fiberglass
30-90
Aluminum Steel with a plastic liner Steel with a fiberglass liner Steel with a steel liner
30-90 30
30
30
Special Features
Ready-made core preservation system; high temperature applications Designed for short radius coring Stronger barrel, includes additional inner and outer core barrel
stabilization Ready-made core preservation system; used for consolidated and
unconsolidated formations Ready-made core preservation system; high temperature applications Ready-made core preservation system; maximum temperature 180°F;
reduces core diameter by 0.5 in. Ready-made core preservation system; maximum temperature 250T;
reduces core diameter by 0.5 in. Ready-made core preservation system; maximum temperature 350°F;
reduces core diameter by 0.5 in.
when anticipating higher than normal torque loads. This system can also be especially attractive when rig time is the largest component of the coring expense.
The precursor of today's heavy duty core barrels is the marine core barrel. This tool was developed to be stronger than conventional systems for use in offshore settings. The marine barrel increases the margin of safety against tool failure, but is restricted to cutting a 3-in.-diameter core.
Today's special heavy duty core barrels have been developed to core harder than normal formations and to cut extended length cores. These tools are designed to cut cores up to 5.25 in. in diameter. Heavy duty threads allow more torque to be applied to the bit and improve the margin of safety against tool failure.
Disposable Inner Core Barrels
Disposable inner core barrels have improved both core recovery and core quality. Aluminum or fiberglass inner core barrels can be substituted for conventional steel inner barrels. The most common type of disposable inner core barrel is the fiberglass barrel. The low coefficient of friction of the fiberglass is thought to improve core recovery by reducing the resistance to core entering the barrel. A more quantifiable benefit comes in the form of a ready-made core-handling and core preservation system that can be used in remote locations or with delicate cores. (For example, fractured or unconsolidated core material tends to be kept intact as the inner barrel is laid down and cut into sections for shipping.) Aluminum inner barrels are recommended when high temperatures (>150°F) are expected.
Core Barrel Liners
PVC plastic, ABS plastic, and aluminum have all been used as inner core barrel liners. The liner slips inside a steel inner core barrel, reducing the effective inner diameter of the core barrel by approximately 0.5 in. The liner simplifies core handling, especially for friable or unconsolidated core, and serves as a core preservation system. Liners are less expensive than disposable inner core barrels. However, besides reducing the diameter of the core, they restrict the
length of the core to 30 ft. Attempts have been made to use core barrel liners to cut
naturally fractured rock. Unfortunately, shards of core often cut into the relatively soft liner material, causing the core barrel to jam.
Sponge-Lined Coring System
The sponge-lined coring system was developed to improve the accuracy of core-based oil saturation data. This is very important information to have when evaluating the merits of a secondary or tertiary oil recovery program.
The sponge coring system consists of a conventional core barrel fitted with a series of sponge-lined aluminum inserts (Figure 2). Tlie system cuts a relatively small core, 2.5 or 3.25 in. wide and 30 ft long. The sponge liner catches oil that "bleeds" out of the rock as the core is pulled (tripped) from the hole. The oil-wet sponge holds oil tightly, while allowing water and gas to move through the sponge and out vent holes drilled in the aluminum liner. Sponge and conventional cores may be cut one after another without making special trips to open the hole, so it is possible to "spot" a sponge core and still acquire larger diameter cores for other analyses.
Full-Closure Coring Systems
Full-closure coring systems were developed for coring unconsolidated formations. They use disposable inner barrels or inner barrel liners and a special core-catching system to retrieve soft cores successfully. Currently, full-closure coring systems cut 30 ft of 3- to 5-in.-diameter core. Systems are being developed to cut larger diameter and longer cores.
The difference between full-closure and other coring systems is that the core catcher is not exposed during coring. This allows the inner core barrel to slip over the soft core with a minimum of disturbance. After coring, a ball is pumped downhole activating the core catcher and sealing off the bottom of the barrel (Figure 3).
The hidden core catcher is both the major asset and major liability of this coring system. Since the core catcher is not exposed during coring, coming off bottom is likely to result in lost core. If the catcher is not activated after coring, the core
Table 2. CoreCatchers
Type Split ring or spring Collet Slip
Dog or flapper
Basket
Full closure
Recommended Usage
Consolidated formations Where formation characteristics are
unknown Consolidated formations, normally
run with flapper catcher or with orientation knives Consolidated, fractured, and unconsolidated formations where geology is unknown Unconsolidated formations, normally run with another core catcher type Friable to unconsolidated formations to provide full closure
Aluminum Liner Sponge Sleeve Centralizer Inner Core Barrel Outer Core Barrel Stabilizer
Conventional Coring
117
Inner Barrel Sub
will be lost during tripping. Against these risks, the benefits are a longer length and larger diameter core than is possible with a rubber sleeve core barrel and a less disturbed core than one cut with a conventional coring system.
Core Bit
Rubber Sleeve Core Barrel
Using a rubber sleeve core barrel is a reliable way to recover core from unconsolidated, conglomeritic, and/or hard fractured formations. The rubber sleeve barrel is unique in that the top of the inner barrel does not move relative to the core during coring. The outer barrel is drilled down around a column of rock that is progressively encased in the rubber sleeve. The system works best from fixed drilling platforms, but it can be operated from floating rigs if rig movement is minimal. The rubber sleeve barrel cuts a 20-ft length of 3-in.diameter core.
Pressure Coring
Pressure core barrels are designed to retrieve cores maintained at reservoir pressure. Accepted as the best method for obtaining core-based oil saturation information, pressure cores also retain reservoir gases. Current tools are designed to cut u p to a 20-ft length of 2.5-in.-diameter core and contain 10,000 psi.
Pressure core barrels are sophisticated tools requiring an on-site facility to service the barrel and handle the pressurized cores. Core handling includes displacing the drilling fluid from the inner core barrel with a nonfreezing gel and then freezing the entire inner barrel assembly in dry ice.
When planning a pressure coring operation, it is important to acknowledge that flushing ahead of the coring bit may take place. Thus, one should minimize overbalance, use a low
Figure 2. Sponge-lined coring system. (Courtesy of DBS, a Baroid Company.)
fluid loss drilling fluid, and when possible, employ the new low-invasion coring bit technology to reduce core flushing.
Wireline Core Barrel
The wireline core barrel allows drilling or coring without pulling the drill string. This reduces drilling costs by eliminating the need to trip pipe. The maximum size of a wireline core is 30 ft long and 2.75 in. wide. The most commonly used tools produce a 2.36-in.-diameter core. For applications in which small diameter samples are adequate, this is a cost effective method.
The wireline core barrel is designed so that the inner core barrel and bearing assembly can be dropped through the drill string, locked in place for coring, and retrieved by wireline. Sequential cores may be cut or a diamond drilling plug can be dropped into place to drill ahead. The drilling plug may also be retrieved by wireline for inspection or to resume coring. During planning, it is important to check that the inner diameter of the drill pipe is large enough to allow the tool to pass through.
118
PART 3—WELLSITE
METHODS
Figure 3. Full-closure core catcher. (From Whitebay, 1986.)
Sidewall Coring
Leewhitebay
Conoco Inc. Ponca City, Oklahoma, U.S.A.
INTRODUCTION
Sidewall coring systems have been developed to obtain core samples after a well has been drilled and logged. The tools can be precisely positioned in zones of interest using gamma logs or SP logs as guides. Percussion or rotary drilled samples provide small bits of formation material, suitable for geological and engineering studies. Other systems such as the new Sidetrak Coring System® cut a more conventional type of core angled off to the side of the hole. Sidewall cores are an excellent cost effective way to increase knowledge of formations. However, sidewall cores should not be used in lieu of whole cores since the discontinuous sampling could lead to misinterpretation of the geological sequence.
i
л\ =i
S. P RING
PERCUSSION SIDEWALL CORING
Most sidewall cores are obtained by percussion sidewall coring systems (Figure 1). These tools shoot hollow, retrievable, cylindrical bullets 1 in. wide by 1.75 in. long into the borehole wall. The tool (gun) can be combined in multiples of approximately 30 bullets with 120 shots a general maximum. The gun is lowered to the desired depth, then individual bullets are electrically fired from the surface. The bullets remain connected to the gun by wires, and movement of the gun pulls the bullets from the borehole wall.
Different bullet "core barrel" designs are available for unconsolidated, soft, and medium to hard formations. Therefore, it is wise to have more than one type of core barrel on location until acceptable core recovery can be demonstrated. Figure 1 illustrates a typical percussion sidewall coring tool.
The advantages of this coring system are speed, low cost, and the precise ability to sample zones of interest after open hole logs have been run. The disadvantage is that the bullet usually alters the formation by shattering harder rock or compressing softer sediments, thereby reducing the quantitative value of the sidewall core analysis data. Also, percussion sidewall core recovery tends to be low in very hard or fractured rock.
•SWITCH SECTION
t=?
GUN BODY
CORE BARRELS
ROTARY SIDEWALL CORING
The rotary, or drilled sidewall coring tool, was developed to recover sidewall core samples without the shattering impact of the percussion system. Suitable for hard to friable rock, the rotary sidewall corer uses a diamond-tipped drill to cut individual plugs from the sidewall. The samples are broken off and pulled from the sidewall by the core drill. The drill is then retracted into the body of the tool where the samples are deposited. The tool is moved to a new sample location after d e p o s i t i n g each sample. A " g u n " of 30
cY
-STABILIZER
Figure 1. Percussion sidewall coring system. (Courtesy of Halliburton Logging Services, Inc.)
119
120
PART 3—WELLSITE METHODS
TOOL POSITIONING
SHOE
HYDRAULIC DRILL MOTOR DRILL ACTUATOR PLATE DRILL BIT FLEXIBLE HYDRAULIC LINE CORE STORAGE TUBE
CORE RECOVERY INDICATOR
Figure 2. Rotary sidewall coring tool. (Courtesy of Halliburton Logging Services, Inc.)
Figure 3. Process of Sidetrak Coring System®. (Courtesy of Foothills Diamond Coring, Calgary, Canada.)
samples, each 0.9375 in. wide by 1.75 in. long, can be taken during one trip.
The advantage of the rotary sidewall coring system is that it produces samples suitable for quantitative core analysis. The disadvantages are high cost per sample and longer time per recovered core. Rotary sidewall core recovery is low in unconsolidated formations. Figure 2 shows how a rotary sidewall coring tool operates.
SIDETRAK CORING SYSTEM
A new Sidetrak Coring System® is just coming on the market and merits discussion for two reasons. First, the system is designed to acquire a larger, more continuous core sample from a drilled and logged wellbore than is possible with existing sidewall coring tools. Second, the emergence of a new tool confirms there is still room for improvement in the area of acquiring high quality, low cost core samples.
A very close cousin to a conventional core barrel, the Sidetrak Coring System is designed to cut u p to 10 ft of 2.5-in. core. The tool is attached to a conventional drill string and is lowered into the zone of interest. An integral arm pushes the core barrel against one side of the wellbore, resulting in a slightly sidetracked core. This tool is still in the development stage, but it shows promise and should be available in the future. Figure 3 shows a prototype of the n e w Sidetrak Coring System.
WELLSITE HANDLING OF SIDEWALL CORES
Sidewall core handling procedures are always influenced by the objectives of the coring program, but a few items are constant for every job. First, unconsolidated samples must be stabilized prior to shipping. Second, time is of the essence, especially when saturations are to be determined. Finally, there can never be too much information on a sample container or data sheet.
Stabilization
Unconsolidated samples, especially those designated for detailed petrographic or petrophysical studies, must be stabilized to prevent deterioration during shipping. Stabilization can take the form of freezing, jacketing the samples with aluminum foil or lead sleeves, or leaving the samples in the bullets. Freezing samples is usually a good method for stabilizing clean oil or gas sands, but is a less attractive choice as the clay and water content increases.
Time Factor
The time factor is important when evaluating all formations. The time between extraction of the sample from the bullet and packaging should be minimized to prevent the cores from drying out. The faster a sample is examined, wrapped, and sealed either in a glass jar or a ProtecCore®
sleeve, the more accurate the fluid saturation analysis will be. Dictating the core descriptions into a portable tape recorder can speed up this work and make more complete descriptions.
Procedure
A typical sidewall core handling operation begins by examining the gun to determine if all the bullets had fired.
Sidezvall Coring
121
Next, the cores are extracted from the bullets, described, and wrapped in plastic wrap and aluminum foil. The wrapped cores should be sealed in glass jars or ProtecCore sleeves as soon as possible. The sample containers should be labeled with the sample depth, operator's name, and well name. The detailed sample description should be on a separate sheet that accompanies the samples.
Core Orientation
Douglas C. Bleakly
Versar Inc. Alameda, California, U.S.A.
INTRODUCTION
Core orientation is the process by which the original in situ position or orientation of a core cylinder is determined. Typically, a mark, groove, or line is placed on the surface of the core and the in situ azimuth of the marking is determined with respect to geographic north.
Cores are oriented to facilitate measurement of directional properties in the rock. Most routinely, orientation is used to measure large scale features such as bedding, cross-bedding, fractures, flow textures, and stylohtes (Pettijohn et al., 1973). In recent years, oriented core has been used to establish the directions of downhole stress and strain fields (Teufel et al., 1984; Smith et al., 1984; Lacy, 1984).
Core orientation techniques fall broadly into two categories: mechanical and core-based.
MECHANICAL ORIENTATION TECHNIQUES
The industry standard for many years has been a mechanical orientation technique utilizing a specially designed core barrel and nonmagnetic drill collar with a compass, camera, battery pack, and timer (Rowley et al., 1971). A typical assembly is shown in Figure 1.
The compass, camera, and timer system is mounted at the top of the inner core barrel, inside a nonmagnetic collar. The battery-driven timer is set prior to running in the hole to expose a frame of film automatically at intervals of 1 to 8 min. Each f r a m e r e c o r d s the c o m p a s s a n d the p o s i t i o n of a reference mark. As the coring assembly is made up, a lug on the compass is aligned with a reference scribe (one of usually three knives set in a scribe shoe at the base of the inner core barrel) (Figure 2). As the core is cut and enters the mouth of the barrel, grooves are cut in the surface of the core by the scribe knives. Angles between the scribe knives vary between coring companies, but the reference scribe is usually offset from the secondary knives by oblique angles on the order of 130° to 150°. A preferred arrangement is an asymmetric scribe shoe in which the angles from the reference scribe to each of the secondary scribes differs (Figure 2). Such an arrangement allows up-down directions on the core to be distinguished easily.
At agreed upon depth intervals, pumps and rotation are shut d o w n for a period of several minutes to allow for a vibration-free film frame to be snapped. The process is repeated at intervals until the core barrel is tripped out. At the surface, the film is retrieved and developed. The azimuth of the compass lug, and hence of the reference groove, is then determined for the depths at which vibration-free shots were taken. Results can be available at the wellsite within hours of core retrieval. A recent refinement of this system uses a modified measurement while drilling (MWD) unit to replace the camera and timer within the nonmagnetic collar. The unit
Figure 1. Typical assembly for mechanically orienting core.
122
Core Orientation
123
Specimen reference line
Figure 2. Scribe knife arrangement.
is powered by a battery pack and records azimuth data continuously. The recorded data are retrieved from the unit when the barrel is tripped out. This system has the advantage that it is not necessary to shut down pumps and rotation to record data.
Figure 3. Paleomagnetic plug sample.
CORE-BASED ORIENTATION TECHNIQUES
Core-based techniques rely at least in part on data derived from the core for orientation. If bedding planes and fractures can be discerned in the core, it may be possible to use dipmeter logs or data from borehole televiewers to determine the orientation of the core (see chapters on "Dipmeters" and "Borehole Imaging Devices" in Part 4). Another approach to core-based orientation involves analysis of naturally existing paleomagnetic signals in rock to determine their relationship to present-day geographic north.
Paleomagnetic Core Orientation
The paleomagnetic technique is based on the fact that nearly all rocks, including quartzite, chert, and chalk, contain at least trace amounts of magnetic minerals such as magnetite and hematite. These minerals act as miniature compasses and "lock-in" the earth's ambient magnetic field. A primary signal is imprinted near the time of deposition; over time, one or more secondary signals may be overlain on the primary signal. In specially equipped laboratories, the signals in rock samples can be separated, interpreted, and used to orient core to present-day geographic north.
Paleomagnetic core orientation requires that a number of l-in.-diameter plug samples be collected from the core at the surface (Figure 3). The n u m b e r of samples required to achieve a good statistical result is dependent on the rock type and other factors. Plugs are then sent to a laboratory for analysis, and core orientation results are typically returned in one to four weeks.
No core-based orientation technique can be considered to give real-time results. In each case, recovered core must be analyzed or sampled. That data, combined with laboratory or borehole information (wireline or televiewer runs) may not be available for days or months.
ADVANTAGES AND DISADVANTAGES OF CORE ORIENTATION TECHNIQUES
Numerous studies have been done to determine the accuracy of different orientation techniques. Nelson et al. (1987) provides a particularly comprehensive review of error sources for mechanical orientation a n d c o m p a r i s o n s of accuracy with other techniques. Core orientation techniques are capable of achieving, u n d e r o p t i m u m c o n d i t i o n s , accuracies of <5°. Sources and severity of errors depend on many factors, some of which are listed here:
• At high latitudes, most techniques are prone to failure for similar reasons: the inclination of the ambient earth's field makes an accurate measurement of azimuth difficult.
• High angle deviation of a borehole may exacerbate or reduce errors, depending on the direction of deviation. In holes approaching horizontal, orientation of cores to geographic north loses its meaning; up-down core directions then become important.
• During mechanical core orientation, film or batteries can fail under high temperature conditions (generally >240 T) or film may run out if downhole time is extended. Tandem systems are recommended to reduce the chance of failure.
• Scribe knives may break or be blunted in extremely hard formations; soft or unconsolidated formations may be unsuitable for scribing because the scribe knife may not leave a permanent core groove.
Dipmeter and Borehole Televiewer
Dipmeters and borehole televiewers must detect formation contrast to be effective orientation tools. Formations that have apparent bedding dips at right angles or parallel to the core axis and limited changes in bedding attitude over the length
124
PART 3—WELLSITE METHODS
of the core m a y not be a p p r o p r i a t e for o r i e n t a t i o n by dipmeter or televiewer. If fractures are absent or abundant, or only at right angles to the core axis, results may also be unusable for orientation. Studies by Hocker et al. (1990) offer some indication of the limits of resolution possible using one contemporary dipmeter tool.
Mechanical Core Orientation
Mechanical core orientation relies on a premeasured relationship being maintained between the compass lug at the top of the core barrel and the reference scribe at the bottom of the core barrel. Depending on barrel length, there are one or more makeup joints between the two components subject to torquing during coring rotation. Furthermore, the grooves in the core tend to precess along the length of the core. Hence, orientation accuracy in part depends on the ability to recognize and quantify these effects. Nelson et al. (1987) and Bleakly et al. (1985b) describe graphic and overlay techniques as well as wellsite procedures that can be used to recognize and partially correct for such problems.
Paleomagnetic Core Orientation
Paleomagnetic orientation can be used in hard and soft formations and has been successfully applied under very high temperature conditions in geothermal wells. The technique may not be appropriate in uniformly very coarse grain sands, conglomerates, or brecciated formations. Orientation accuracy is in part dependent on the quantity of magnetic material present in samples; comparatively more samples must be taken from intervals exhibiting very weak magnetic signals. Under some circumstances, collecting a statistically significant n u m b e r of samples may entail removal of an
appreciable portion of the available core.
CORE ORIENTATION QUALITY
For best results, core orientation technique(s) should be selected on the basis of the operating conditions and the ultimate applications of the results.
Proper core handling can enhance orientation quality regardless of the techniques selected (Nelson et al., 1987; Bleakly et al., 1985b) (also see the chapter on "Core Handling" in Part 3). A p r i m a r y goal of handling procedures is to reconstruct the core as accurately and completely as possible immediately after recovery from the barrel. It is recommended that an asymmetric scribe shoe be used on all core runs to aid reconstruction.
At the well site, cores should be carefully fitted together into continuous intervals, which are defined as lengths of core in which each piece fits into the next. Continuous intervals can vary in length from a few centimeters to 10 m or more. Intervals are nomally broken by drill string connection points, core rubble zones, core spinoffs, and the tops and bottoms of core runs.
Each continuous interval should be marked with a continuous line or lines parallel to the axis of the core. A right angle straight edge can be used to ensure that this master orientation line (MOL) (Bleakly et al., 1985a) is parallel to the core's long axis. Placement of an MOL on each interval can be critical to interpretation of orientation results after the core leaves the wellsite. If mechanical orientation has been done, orientation results can be compared against measurements taken on the core, using the MOL as a baseline (Bleakly et al., 1985b). For core-based orientation techniques, the MOL can serve as the orientation reference line.
Core Handling
Byram Reed
BP Exploration Bogota, Colombia
INTRODUCTION
Core handling is a critical phase in the core acquisition process. There are common techniques to handling both "hard" rock and "soft" sediment cores. All handling procedures are aimed at ensuring proper labeling, minimizing damage, and carefully transporting for analysis. In all cases, one should have all equipment prepared before the core is pulled. Every effort should be made to prevent the core from weathering or changing it's fluid contents. Speed is important when handling hard rock to prevent alteration of reservoir and character fluids. In contrast, patience is paramount when handling soft cores to prevent mechanical damage.
Labeling is probably the most important and visible feature in wellsite core handling. Mud should be wiped off the core or inner barrel to get a good marking surface. Do not wash the core. Washing changes the fluid content or wettability and impacts the subsequent analysis.
All cores should have orientation marks. The general convention is to use two different color markers and draw parallel lines on the core or inner barrel. Each operator has preferences for colors and arrangements, such as red and black with red on the right. Up arrows are marked on every segment of the core to ensure proper orientation.
HARD ROCK CORE HANDLING
Hard rock cores represent a major workload for the wellsite geologist. The core must be unloaded from the core barrel on the rig floor and taken elsewhere for processing. Safety in the extraction process is p a r a m o u n t . Heed the following rules:
• Hold a safety meeting with all floor personnel and make sure everyone will work at your direction and speed.
• Everyone on the floor should wear eye protection. • Make sure the driller has a good view of the operation. • Only one person gives directions to the driller. • Never put anything (such as hands or feet) under the
core. • Always run the core barrel back to the floor before
releasing the core. Stand the barrel on the floor and raise the barrel. This prevents the core from dropping and shattering on the floor. An alternative is to extrude the core on the catwalk. • Never raise the core barrel more than 18 in. before breaking off. • Bag all core rubble. • Support core to keep it from breaking while moving to core boxes. • Thank all the floor hands for a job well done.
Move core in core boxes marked with a box number and "top" and "bottom" on each box. Guarantee that core is placed in proper orientation in the boxes. Scribes should be used on hard rock cores to both orient the core and help with reassembly (see the chapter on "Core Orientation" in Part 3). If the catwalk or pipe rack is the location of choice for reassembly, two joints of pipe tied together make an ideal brace for handling.
After the core is laid out, the following should be done:
• Wipe the core clean; do not wash it. • Use the scribe lines to reassemble the core; put the
primary groove up. • Draw the twin orienting lines on the core; put the
primary line in the scribe groove. • Mark each segment of the core with up arrows. • Determine where missing footage is located and space
it out appropriately. Bag rubble and place it with the core. If the depth interval of missing core is unknown, place it at the bottom of the core. • Mark the core with depths. Proper marking is critical to tying core footage to wireline measurements. Mark the depth on each segment of the core.
Note that not all core is lost from the catcher at the bottom. Compare the rate of penetration and lithology of the core. Shale sections will often break up when entering the catcher and be lost. Spin marks are generally seen in the shale sections and may represent lost segments. When taking multiple cores, examine the top segment for additional catcher marks that represent recovery of a post from the last core. Be sure to include this when depth marking the core.
SOFT SEDIMENT HANDLING
Soft sediment core represents an especially difficult handling problem. Soft implies that the sediment is not cemented and therefore has little internal strength. Disturbing the grain orientations will negate the ability to measure porosity, permeability, and other reservoir properties accurately. The sediment can be disturbed in several ways:
• Gas expansion while tripping out of the hole • Slumping caused by rotating or jarring the core barrel
while tripping • Slumping caused by rough handling of the core barrel
at the surface • Slumping caused by vibration in transport • Shearing caused by flexing the inner barrel
Soft sediment cannot be removed from a steel barrel without disrupting it. Therefore, a disposable inner barrel
125
126
PART 3—WELLSITE METHODS
made of fiberglass, PVC, or aluminium can be used. The inner barrel is generally mated to a specific core catcher design (see chapter on "Conventional Coring" in Part 3).
Well-designed equipment cannot alleviate the problems caused by poor handling. Since the core is liidden in the inner barrel, these problems are not noticed unless the core is CAT scanned. Scanning will determine the extent of damage and is an excellent aid to picking analysis plug points. Based on experience, the following steps are recommended to prevent damage:
Use a perforated inner barrel to allow for gas expansion. Do not rotate the drill string while tripping out. Set the slips softly to prevent jarring of the core. Stop 100 ft below the rotary table to let the core degas. Use a brace for the inner barrel to prevent it from flexing while laying it down. Do not rotate, bump, or jar the inner barrel while laying down the core or cutting it into sections. Freeze the core to prevent disruption during transport.
Core Alteration and Preservation
Caroline J. Bajsarowicz
BP Exploration Aberdeen, United Kingdom
INTRODUCTION
Considerable resources are invested in core analysis programs designed to furnish information on geological and petrophysical rock properties and on engineering and completion data (Keelan, 1985). The economic implications of the accuracy and credibility of the data obtained from these analyses can be significant, especially in equity determinations. It is important to obtain data that relate as closely as possible to virgin reservoir conditions. Thus, alteration of the core during recovery, wellsite handling, shipment, and storage must be minimized.
CORE ALTERATION DURING RECOVERY
Changes in the core and fluid content during coring are unavoidable. However, changes can be minimized by understanding the processes that affect the core during recovery. Cores can be damaged during recovery by
• Filtrate invasion • Fluid expansion and expulsion • Physical damage to the rock
Filtrate Invasion
During core acquisition and retrieval, the mud filtrate often invades the core. Invasion can displace over half of the native fluid, which can change the in situ fluid saturations in the core. Invasion can also alter rock properties through interaction with the core minerals and fluids. For example, the filtrate may cause clays either to swell or to shrink.
The a m o u n t of native fluid displaced by m u d filtrate depends on the rate of bit penetration, permeability of the formation, viscosity and compressibility of the native fluid and the filtrate, mud cake permeability, pressure differential and relative permeability of the formation to the mud filtrate, and core diameter (Basan et al., 1988; American Petroleum Institute, 1960).
Filtrate invasion can be minimized several ways (Basan et al., 1988; Keelan and Donohue, 1985):
• Select a bit that directs the drilling fluid away from the core rather than toward it.
• Increase the coring speed. The faster the core enters the core barrel, the less time there is for invasion to occur.
• Establish a low pressure differential between the drilling fluid and the reservoir.
• Optimize the fluid loss properties of the drilling fluid. • Increase the diameter of the core cut to increase the
area of uninvaded central core.
Evaluation of fluid invasion can be tested by doping the coring fluid with a suitable tracer and then checking the
tracer concentration in the fluids extracted from the recovered core. The effect of invasion on fluid saturations is measured using "plug and donut" analysis.
Fluid Expansion and Expulsion
As the core barrel is brought to the surface, the core and fluids are subjected to a reduction in pressure and temperature from reservoir to atmospheric conditions. Only minor changes occur to the rock matrix. However, the fluids undergo substantial changes in volume. Oil releases gas from solution, resulting in shrinkage of the oil. The gas dissolved in the oil and water expands and escapes from the core, leading to expulsion of the fluids. These phenomena result in surface saturations that are different from those downhole (American Petroleum Institute, 1960; Keelan and Donohue, 1985). The magnitude of saturation changes that can occur during coring and recovery with water-based and oil-based coring fluids are illustrated in Figure 1.
A pressure coring tool is designed to maintain reservoir pressure in the core by enclosing the core in a pressurized chamber before it is brought to the surface. This helps prevent the fluid changes that occur with expansion and expulsion. Saturation measurements from pressure cores are much more accurate than those from conventional core. However, they are still not 100% accurate, as pressure cores can still be subject to flushing during the coring process. A sponge core liner system can also help minimize errors in saturation measurements due to fluid expansion and expulsion by the retention of the expulsed formation fluids in a sponge or foam lining, (see the chapter on "Conventional Coring" in Part 3).
Physical Damage
Petrophysical properties can be altered when the rock is damaged during the coring process. Physical damage to the core can occur in many ways:
• Fractures induced due to stress relief or jarring of the core barrel during retrieval
• Disaggregation and fracturing of unconsolidated sediments
• Crushed grains due to the high impact from percussion sidewall coring
CORE ALTERATION DURING WELLSITE HANDLING
Although changes in the core and its fluid content during coring are unavoidable, it is important to minimize any further damage to the core during wellsite handling which would make the core even less representative of the reservoir. The time a core is exposed to the atmosphere and the drilling
127
128
PART 3—WELLSITE METHODS
Saturation Conditions
Flushing with Water Base Mud
Flushing with Oil Base Mud
Core at A Surface A
Trip to Surface Gas Expands
12% 40% 48%
40% 30% 30%
Solution gas expulses oif and water while oil
shrinks
I
Solution gas expulses oil while oil shrinks
After Coring in Barrel at Reservoir
Pressure
Filtrate Invasion During Coring
30%
70%
70%
30%
Filtrate water flushes out oil
Filtrate oil exchanges with pore oil
Saturations in Reservoir
70%
30%
(a)
Oil
Gas I I Water Q
Saturation Conditions
Flushing with Water Base Mud
Flushing with Oil Base Mud
Core at Surface
Trip to Surface Gas Expands
After Coring in Barrel at Reservoir
Pressure
Filtrate Invasion During Coring
1% 49%
50%
40% 30% 30%
Pore gas expulses water Pore gas and solution
while gas expands and gas expulse oil while
may condense
oil shrinks
30%
70%
50% 20% 30%
Filtrate water flushes out gas
Filtrate oil flushes pore gas
JStelA
I
inXeservdir J
70%
30%
(b)
Oil Щ
Gas I ^ J Water £ j
Figure 1. Typical fluid contents from reservoir to surface, (a) Oil-productive formation, (b) Gas-productive formation. (Courtesy of Core Laboratories, a Division of Western Atlas International.)
fluid during wellsite handling will affect subsequent core analysis measurements.
Depending on atmospheric conditions, exposure of cores for even a short period of time can cause significant loss of water and light hydrocarbon fractions. Tests show exposure for even 30 min can result in 10 to 25% loss in w a t e r (American Petroleum Institute, 1960). To prevent saturation changes, the time the core is exposed to the atmosphere should be minimized (see chapter on "Core Handling" in Part 3 for additional information on core handling techniques).
CORE PRESERVATION DURING SHIPMENT AND STORAGE
Core preservation is an attempt to maintain a core during shipment and storage in the same condition it was in when the core was originally removed from the core barrel.
Core preservation techniques should keep the core in correct sequence, prevent breakage during shipment and storage (which is very important for soft or poorly consolidated cores), minimize core alteration, and preserve the volume and distribution of the core fluids.
Problems that core preservation methods must address include the following:
• Dehydration and salt precipitation • Oxidation • Redistribution of fluids • Evaporation and condensation • Hydrocarbon deposition • Clay collapse • Bacterial growth
Preservation should be quick in order to minimize exposure time. Head space in preservation materials should be small to reduce the a m o u n t of air in the package and decrease evaporation and condensation losses. Porous materials that can affect saturations should not be used in the preservation package. Temperature fluctuations that can cause problems with evaporation and condensation of core fluids should also be minimized.
METHODS OF CORE PRESERVATION
Ideally, all core should be preserved. The method of preservation and packaging of cores varies depending upon the type of core (consolidated versus unconsolidated), the core analysis measurements required, and the length of time the core is stored before testing.
Core preservation methods are typically either "dry" or "wet." Dry methods enclose the core in a material that prevents evaporation of formation fluids. Wet methods of preservation involve submerging the core in a brine or other fluid that preserves core wettability (Basan et al., 1988). A variety of dry and wet preservation methods used by the industry are summarized in Table 1. Note that none of these methods provide an ideal solution to core preservation.
Dry Core Preservation Methods
Air-Tight Metal Cans
Metal cans are excellent vapor barriers, but they can react with water. Consequently, canned cores should be prewrapped to prevent moisture loss. Prewrapping also minimizes headspace in the container, preventing movement
Core Alteration and Preservation
129
Table 1. Summary of Dry and Wet Core Preservation Methods3
Method Dry
Alternatives
• Sealing in air tight metal cans • Sealing in rubber, plastic, aluminum, steel, or
fiberglass tubes • Sealing in plastic bags • Wrapping in plastic wrap and aluminum foil
and coating with wax or plastic • Sealing in laminated, heat-sealable packages • Freezing with dry ice
Wet
• Sealing in anaerobic jars or polycarbonate,
steel, glass, or PVC containers with brine,
oil, or other fluids
aFrom American Petroleum Institute (1960) and Basan et al. (1988).
of the core in the container and reducing evaporation and condensation losses. The prewrap should be inert so that it does not react with the formation fluids, and it should be nonporous so as not to affect core saturations (American Petroleum Institute, 1960).
Core Sleeves, Liners and Barrels
Rubber, plastic and aluminum sleeves, fiberglass liners, and pressure (steel) core barrels can be cut into suitable lengths a n d then c a p p e d for storage of core. This preservation technique provides some protection to the core during surface handling, particularly for fractured and unconsolidated cores.
Except for a l u m i n u m a n d steel tubes, n o n e of these materials are effective vapor barriers. Therefore, this preservation method should be used for temporary storage only. Excess m u d should be drained out of the tubes to minimize exposure of the core to the drilling fluid. However, leaving the space filled with air can result in evaporation and dehydration of the core.
Plastic Bags
The simplest preservation method is to wrap the core in plastic or heat-sealed plastic bags. Plastic is not an absolute oxygen or water vapor barrier; it only reduces the rate of evaporation. Studies by Auman (1989) show that cores in heat-sealed plastic bags lose 6% of their water in 10 days. One pinhole more than triples water loss. Cores wrapped in plastic lose about 30% of their water in 3 days.
Hot Wax or Strippable Plastic
Coating cores with hot wax or strippable plastic is a widely used preservation method that involves wrapping the core in plastic wrap and aluminum foil and then dipping the core in paraffin or a plastic sealant. The steps to preserve a core using this method are as follows:
1. Wrap the core in several layers of plastic wrap or film to prevent fluids in the core from contacting the outer wrapping of aluminium foil. Of the commercially available food wraps, Saran Wrap® has been found to
Table 2. Chemical Reactivity of Barex and Saran Wrap3
Reactive Liquid
Heptane Cyclohexane Gasoline Benzene Toluene Alaskan Crude Artie diesel Oil-phase drilling mud
Weight Lose After 30 Days Exposure at 100°F (%)
Barex
Saran Wrap
1.2
3.1
0.1
2.0
0.1
2.0
1.1
2.3
0.2
1.9
0.3
2.3
0.4
8.2
0.6
1.4
aFrom Hunt and Cobb (1988); courtesy of SPE.
be the least reactive with formation fluids. However, the wrap has been found to degrade with some hydrocarbon compositions (Table 2) (Hunt and Cobb, 1988). Barex® film, which is relatively inert against organic solvents and corrosive fluids, can be used, but it is inflexible and difficult to wrap around core (Hunt and Cobb, 1988). 2. Then wrap the core in two or three layers of heavy duty aluminum foil. The edges should be crimped. The aluminum foil acts as a vapor barrier (Table 3). 3. Double dip the wrapped core in melted wax or plastic. String should be used to dip the core, not wire, because wire can rip the aluminum foil. The string should be cut off and the ends also dipped in the wax or plastic. This wax or plastic coating protects the core and the aluminum foil during shipping and storage.
Note that wax and plastic are permeable and do not serve as barriers to oxygen or water vapor. However, CoreSeal® is relatively impermeable to water vapor (Table 3) (Bajsarowicz, unpubl. data), as are several common polymers (Table 4) (Hunt and Cobb, 1988).
Barrier Foil Laminate
The most common barrier foil laminate is ProtecCore. This laminate consists of aluminum foil—the major moisture and oxygen barrier—between several layers of bonded plastic. The innermost layer, Barex, is inert and heat sealable. The outer two plastic layers, polyethylene and polyester, provide strength and rigidity (Hunt and Cobb, 1988). The properties of the various components of ProtecCore are given in Table 4. The steps to preserve a core using this method are as follows:
1. Wrap the core in three or four layers of Barex film to prevent the core from puncturing the ProtecCore laminated material.
2. Slip the prewrapped core into the ProtecCore laminate tube. One end of the package is sealed with a heat sealer. The air space within the package is minimized by flattening it out as much as possible before heat sealing the other end of the package. To reduce free space around the core even further, a small hole can be left in a corner and a vacuum pulled on the package.
130
PART 3—WELLSITE METHODS
Table 3. Transmissivity of Seal and Wrap Materials
B-60 wax3 CoreSeal®3 Aluminum foilb Saran Wrap®b
Oxygen (cm3-mil/100 in.^D-atm)
3015 Too high to measure
0 1.52
aFrom unpublished analyses by C. Bajsarowicz; courtesy of BP Exploration. bFrom Hunt and Cobb (1988); courtesy of SPE.
Water Vapor (g-mil/100 in.2-D-atm)
122 2-13
0 0.18
Carbon Dioxide (cm3-mil/100 in 2-D-atm)
—
—
0 1.0
Table 4. Permeation Rates for Various Polymers3
Barex® 210 Resin Polyvinyl chloride Polyester Low-density polyethylene High-density polyethylene Polystyrene Polypropylene
Oxygen (cm3-mil/100 in.2-D-atm)
0.4 4 4.5 500 200 350 85
aFrom Hunt and Cobb (1988); courtesy of SPE.
Water Vapor (g-mil/100 in 2-D-atm)
3.5 2 2.2 1.0 0.4 6.5 0.26
Carbon Dioxide (cm3-mil/100 in.2-D-atm)
0.8 20 20 1900 500 900 300
Tliis pulls the packaging down tight on the core and removes most of the air. The hole is then heat sealed (Whitebay, 1986). 3. For protection during shipment, the individual pieces should be wrapped in pads or bubble wrap.
This preservation method provides a good vapor barrier, superior to the hot wax or strippable plastic method (Auman, 1989; Hunt and Cobb, 1988). However, the packaging is fragile and can be easily ripped or punctured and is subject to pinholes and cracks (Basan et al., 1988; Whitebay, 1986). Careful handling can minimize this type of damage.
Freezing with Dry Ice
Freezing core is often done to minimize the loss of the volatile hydrocarbons, to preserve the fabric and structure of unconsolidated cores, and to immobilize the fluids in pressure cores (Torsaeter, 1985). The most common method of freezing core is with dry ice. However, light hydrocarbon fractions are not maintained at dry ice temperatures (-78.5 °C), thus liquid nitrogen temperatures (-195.8°C) are required.
The exact effects of f r e e z i n g on the rock a n d its petrophysical properties are still u n k n o w n . A variety of studies examining the effects of freezing on porosity and permeability show contradictory results (Wisenbaker, 1947; Lebeaux, 1952; Kelton, 1953; Torsaeter, 1985). The freezing process may affect the rock structure due to ice formation and m a y a f f e c t the w e t t a b i l i t y d u e to p r e c i p i t a t i o n of hydrocarbons onto pore surfaces. To minimize damage by ice, the cores should be frozen quickly to reduce ice crystal growth.
Sublimation from the core surface must be prevented
during storage. One method is to quick freeze a layer of brine on the surface of the core. The brine does not enter the frozen core; subsequent sublimation comes from this layer, not the core.
Wet Core Preservation Methods
Cores can be preserved by submerging them in jars of deoxygenated formation brine or diesel. A bactericide, generally formaldehyde, is added to prevent bacterial growth during storage. The jars are closed and the system purged with nitrogen. This system inhibits most oxidation (Basan et al, 1988).
Polycarbonate or anaerobic jars are the most commonly used containers. Other containers that can be used are made of steel, PVC, or glass. Caution must be exercised when using steel because it can rust in the presence of water. PVC containers are not optimal because they permit diffusion of water and oxygen (Basan et al., 1988). Glass containers are excellent preservation containers, but they are difficult to use in the field without breaking.
The wet method of core storage is often used when the core analysis program requires maintaining wettability. There is still some debate about which fluid should be used in the containers. Wet preservation cannot be used when cores are cut to evaluate interstitial water, to measure fluid levels, or to interpret gas, oil, or water production. This is because exposure of the core to a fluid results in imbibition of that fluid and alteration of saturations.
Preserving and storing core with a wet system has a high cost and requires regular maintenance. Each jar must be purged with nitrogen every two weeks.
Drill Stem Testing
Ingrid Borah
Conoco Inc. Casper, Wyoming, U.S.A.
INTRODUCTION
A drill stem test (DST) is a temporary completion of a wellbore that provides information on whether or not to complete the well. The zone in question is sealed off from the rest of the wellbore by packers, and the formations' pressure and fluids are measured. Data obtained from a DST include the following:
• fluid samples • reservoir pressure (P*) • formation properties, including permeability (к), skin
(S), and radius of investigation (rp • productivity estimates, including flow rate (Q) • hydrodynamic information
This chapter focuses on nonflowing DSTs (NFDSTs). In a NFDST, fluid does not flow to the surface and a stabilized flow rate is not obtained (also, the well can flow to the surface but die or be shut-in before steady-state rates are achieved). Analysis of flowing DSTs is less complicated because flow rates can be measured throughout the test (Earlougher, 1977; Erdle, 1984; Matthews and Russell, 1967).
PLANNING
The key to successful testing depends upon planning and teamwork between the geoscientist and the engineer. Potential pay zones should be identified before drilling commences so that the drilling program can be designed to a c c o m m o d a t e the test. If offset data are available, the magnitude of porosity, permeability, and reservoir pressure should be identified. Knowledge of zonal mineralogy may prevent excessive damage by drilling fluids and should be used in designing the mud program. The anticipated reservoir properties are used to design the test string and test times so that the best, most useable data can be obtained.
Safety
Running a DST is one of the most dangerous jobs in the oil field because the well is essentially uncontrolled during the test. All fire fighting equipment and the blowout preventers should be inspected and tested before starting a DST. Hydrogen sulfide (H2S) equipment should be on hand if anticipated conditions are sour. No test should be initiated at night or during an electrical storm. No smoking should be allowed on the drill floor or near any flowlines or surface test equipment.
Packer Seats
The best time to run a DST is just after drilling through a potential pay zone, when exposure to damaging fluids is minimal and the hole is in its best condition for a good packer seat. Packer seats are generally located in competent sandstones or carbonates above the test interval or both above and below it. The packer seat is picked by the geoscientist or mudlogger using cuttings and the rate of penetration (ROP) as guides. Soft formations, characterized by a fast ROP, usually do not make good packer seats because the packer has trouble seating or gripping the side of the borehole. This results in poor pressure seals, or in extreme cases, the packer may slide down the hole.
Length of Test Interval
The length of the test interval should be short so that less mud will be displaced from the rathole (the portion of the open hole below the bottom packer) into the drill string.
Length of Test Time
The length of flow and shut-in periods during the test are critical to obtaining good reservoir data. The dual flow, dual shut-in test is most commonly used. The initial flow period of 3-5 min removes the "supercharge" effect of mud filtrate near wellbore. The first b u i l d u p is run for 60 min to determine a valid P* (reservoir pressure). The second flow period is used to collect a fluid sample and create a pressure disturbance at a distance beyond any damaged zone. The duration of the final flow period may be anywhere between 60 and 120 min, depending upon the time available for the test and the final buildup. The final buildup is used to evaluate reservoir transmissibility, damage, and radius of investigation and should be at least as long as the final flow period. It is preferable (daylight permitting) to run the final build u p three times as long as the final flow period to ensure that good pressure transient data is recorded.
Mud System
If a drill stem test is anticipated, low fluid loss mud will prevent excessive leakoff into the target zone and doping the mud with nitrates will distinguish filtrate from recovered formation water.
TEST TOOLS
A drill stem test string consists of packers, a downhole shut-rn valve, a safety joint, and pressure gauges (Figure 1). The bottom packer and blanked off gauge are shown as an "add on" to a straddle test.
131
132
Part 3—Wellsite Methods
Drill pipe
Drill collars
O Reverse circulating sub
Drill collars
X over sub
0 Recorder above closing tool
ШClosing tool
b Bypass
0 Recorder (inside)
Jars
Safety joint
Safety seal
Packer
Packer
Perforated sub
0 Recorder (outside)
ОоOОOОO Perforated sub
X over
Drill collars
X over WОоОоО Perforated sub
For straddle test: Packer-bull plug or blanked off pressure gauge
Figure 1. Drill stem test tool string.
Packers Compression set packers are generally more reliable than
inflatable packers because they can withstand more differential pressure between the annulus and the drill string. The number of packers depend upon experience and test type (conventional, straddle, or hookwall) (Johnson-Flopetrol, 1980). Figure 2 illustrates other types of test strings.
Packer selection is also determined by the need for a cushion. A cushion consists of water or gas and is run for the following reasons:
1. To prevent drill string collapse during deep tests or when high mud weights are used
2. To prevent excessive differential pressure across the packer(s) during the flow periods
3. To prevent high differential pressure across the sand face in unconsolidated formations, which will result in sand flow
4. To prevent corrosion of the drill string from corrosive gases such as H2S or CO2
Cushions also provide back pressure on the formation, which inhibits flow into the test string. If considerable damage or low permeability is expected, the cushion should be small.
Pressure Gauges
A m i n i m u m of three (mechanical, electronic, or a combination) pressure and temperature recorders should be run on a conventional test and four on a straddle test. Selection depends on how accurate the data need to be.
One gauge should always be run inside the drill string above the closing tool. This gauge measures the hydrostatic head of fluid produced into the drill pipe and is critical to evaluating the volumes of fluids produced during the test. It also indicates drill string leakage during the test.
T w o g a u g e s s h o u l d be r u n below the closing tool to measure pressure during the flow and shut-in periods. Two are needed to verify that they are reading within their calibration ranges and to provide a backup in case one fails.
A blanked off gauge must be run on a straddle test to verify that the bottom packers were holding. In most cases of straddle test failure, it is the bottom packers that fail.
ANALYZING THE DST
Once the drill stem test is concluded, mechanical validity and fluid recovery must be verified.
DST Chart Analysis
Descriptions of valid DST charts are shown in Figures 3,4, 5,6,7, and 8. (See figure captions for discussion.)
Fluid Recovery (Nonflowing Test)
On an NFDST, the volume of fluid produced from the formation is contained in the drill string. A fluid level must be determined to calculate the volume recovered. If the fluid is highly gas-cut, a straight volume calculation will be inaccurate. Therefore, samples should be collected at regular intervals while reversing to a clean tank on location. Although it is common to reverse to a pit, the amount of fluid recovered cannot be determined if this is done. Error in measuring fluid recovery often makes the difference between an economic disaster or a success. Also, grindouts (centrifuge) must be performed on each sample to determine the percentage of oil, water, and solids. Resistivity, chloride content, and nitrate content of p r o d u c e d water and the specific gravity of all phases should be measured.
Estimating Flow Rates (Nonflowing Test)
During the flow period on a DST, the flow rate is not constant. A flow rate can be calculated using pressure data from the gauge above the shut-in tool. By dividing the time
Д 6
7
-я.о.,
Drillpipe
Drill Collars Pump-Out Reverse Tool Drill Collars (1 Stand) Break-Off Plug Reverse Tool Drill Collars (1 Stand) Bar-Catcher Multi-Flow Evaluator (MFE) Pressure Recorder
(Inside Reading) TR Hydraulic Jars Rotary Pump
Multi-Stage Relief Valve Safety Joint Upper Inflate Packer Blank Spacer Pipe
or Drill Collars Lower Inflate Packer
Deflate Drag Spring Tool
Pressure Recorder (Inflate Pressure) Bullnose
Drill Stem Testing
133
Remainder of Test String as per MFE Open Hole String
TR Hydraulic Jar Safety Joint Upper Straddle Bypass Sub Safety Seal Upper Packer Perforated Anchor
Lower Straddle Bypass Sub
Pressure Recorder (Outside Reading) Pressure Recorder (Inside Reading) Drill Collars or Blank Pipe
Lower Packer
Selective Zone Anchor
(b)
Figure 2. Straddle drill stem test tool string, (a) Typical MFE inflate open hole string (straddle, off-bottom), (b) Typical MFE straddle string (open hole, conventional, off-bottom). (From Flopetrol-Johnston, 1987.)
scale into discrete increments and recording the pressure, the data can be transposed to a position in the drill pipe. Since the v o l u m e of the drill string and the c h a n g i n g fluid composition are known, instantaneous flow rates can be calculated as shown in Table 1.
Other methods include the following:
1. Superposition uses the rate schedule in analyzing the pressure data. The technique is described in Earlougher (1977) and in Matthews and Russell (1967).
2. Tlte Odeh-Selig method (modified superposition) is fairly rigorous (Odeh and Selig, 1963). It is only valid if the shut-in period is 1.5 times greater than the flowing period.
3. The modified Horner method uses effective flow time and a flow rate calculated from the end of the second flow period. The effective flow time is calculated by
dividing the total fluid recovery in barrels by the flow rate in barrels per minute, which is then converted to barrels per day. 4. The simplified Horner method converts total recovery and the flow time of the test to an average daily rate. The flow time using this method is the time the tool was open. Although this method is commonly used, it is incorrect to assume that this rate reflects stabilized producing rates. This method should only be used as a last resort.
INTERPRETATION
The most important parameters in DST interpretation are the radius of investigation (ROI) and the observed wellbore storage constant (WBSC). A short ROI, c o m b i n e d with knowledge of drilling fluid properties (fluid loss or amount of overbalance), may indicate that the calculated permeability is
134
Part 3—Wellsite Methods
H- — к _ —T i me
t
P
r
e
S S U
r
\ e
Figure 3. Perfect chart. Gauge located inside and above the closing tool. (A) Add cushion/run in hole; (B) initial flow period; (C) initial shut-in period; (D) final flow period; (E) final shut-in period; and (F) pulling out of hole.
Figure 4. Collar leak. Gauge located inside and above the closing tool. Chart indicates increasing pressure during running in hole and shut-in periods
Figure 5. Fluid loss from drill pipe. Gauge located inside and above the closing tool. Bleeder valve on drill string left open during shut-in periods.
Figure 6. Perfect chart. Gauges inside above and outside below the closing tool. Pressure transient analysis done from these gauges. (A) Run in hole, gauge measuring hydrostatic pressure of mud column; (B) initial flow period; (C) initial buildup; (D) final flow period; (E) final buildup; and (F) release packer and pulling out of hole.
Figure 7. Perfect chart. Blanked off gauge below the bottom packer on a straddle test. (A) running in hole; (B) initial flow period; (C) initial buildup; (D) final flow period; (E) final buildup; and (F) pulling out of hole.
Figure 8. Bottom packer failure. Blanked off gauge below bottom packer on a straddle test. When the bottom packer fails, the pressure gauge will read some flow and buildup data but will not replicate gauges run above the bottom packer because of a restricted flow area around the packer elements. (Types of gauge failures are described in Flopetrol-Johnston, 1980.)
Table 1. Calculation of Instaneous Flow Rates
dP/dT (psi/min) (^-P0WWo) ( P 2 - P 1 M f e - Z1)
aCan be estimated from fluid samples while pulling out of hole. bDPV = drill pipe volume (bbl/ft).
Fluid Gradient3 (psi/ft)
Gi G2 Gn
Drill Stem Testing
135
Flow Rate (BPD)
(Col 1 -5- Col 2) x 1440 XDPVb (Col 1 -5- Col 2) x 1440 x DPV (Col 1 - Col 2) x 1440 x DPV
indicative of only the damaged zone. Usually, a skin factor close to zero is calculated under these conditions, along with a low formation transmissibility (kh/\i). One way to "see" past this damaged zone is to rerun the drill stem test with longer flow and shut-in periods.
If fractures are present, the measured WBSC may be much higher than the calculated WBSC. A high transmissibility and a negative skin will be computed under these conditions. A negative skin implies stimulation, which cannot occur during n o r m a l drilling o p e r a t i o n s . If an interval is f r a c t u r e d , recoveries and calculated flow rates can be much greater than expected future production.
Calculating Wellbore Storage Capacity
To calculate the wellbore storage capacity (WBSC), plot log (Pws - Pwf) versus log (ts - tw{) (terms defined in Table 2). Because of storage effects, the early portion of the data will have a unit slope. As these effects diminish, data points on the log-log plot fall below the unit slope line and approach the slowly curving line for no wellbore storage. The measured WBSC (WBSCmens) is calculated using one point from the unit slope line in the following equation:
WBSCr
_ QlastBo a ^ 1440AP
(See Table 2 for explanation of mathematical variables.) The calculated WBSC (WBSCcalc) is computed as follows:
W
B
S
CCca
ik l
c
=-^-
Pwf
Calculating Static Reservoir Pressure
Reservoir pressure (P*) is calculated by extrapolating the pressure data (from the first buildup on a DST) on the Horner (or superposition) plot to an infinite shut-in time (Figure 9). This pressure provides a guide for selecting the slope of the second buildup Horner plot. If the second buildup slope extrapolates to a pressure significantly less than P*, depletion might be suspected. To see true depletion, the reservoir would have to be very small. Although depletion is possible in rare cases, identifying it is usually a result of poor test design and analysis.
A method for constructing the Horner plot is outlined as follows. After determining the effective producing time and producing rate, tabulate time and pressure for each buildup
Figure 9. Horner plot.
period. Make plots using Cartesian paper because this makes it easier to expand the plot (see Figure 9). Now plot P versus log[(£ + At)/At] for buildup #1 (the first buildup following the first flow period). Note that t in this equation is equal to the length of the initial flow period and that At is the time since the start of the buildup period.
Extrapolate this curve to P* at log[(f + At)/At] = 0. P* provides the "guiding light" for determining the proper slope found using buildup #2. Now plot P versus log[(f + At)/At] for buildup #2. Note here that this t is the effective producing time calculated from one of the methods outlined earlier.
Extrapolate the straight-line portion of the data to P*. Use a data band within the accuracy of the gauge to make sure you choose the correct slope. In other words, if the gauge is accurate within ±5 psig, then the data band should be 10 psi wide on the graph. A data band is extremely important in cases where m < 50 psi/cycle. Calculate m using two points: (X1, y-[) and (x2, y2)/ where
m -- У2-У1 X2-X1
Now calculate the formation transmissibility as follows:
kh _ 162.6CjB0
JLL
M
173
Part 3—Wellsite Methods
Table 2. Nomenclature of Mathematical Variables
riable Symbol
p
' WS
Pwf to
^wf tS ^last At AP
4h h
MQ P* 'p ave m к t
Ф ct
rW
kh/\i
re Bo S WBSCca,c WBSCmeas G1, G2, Gn P0, P1, P2, Pn
tV *2> tn
Explanation
Pressure during the shut-in Final flowing pressure Time during shut-in Time at last flowing pressure Shut in time Last flow rate Time difference Pressure difference Volume of the rathole below the packer Sand thickness Fluid viscosity Producing rate Reservoir pressure determined from buildup #1 Average flowing pressure Slope from Horner plot Permeability Flow time (in skin equation) Porosity System compressibility Wellbore radius Transmissibility Estimated drainage radius Formation volume factor Skin term Calculated wellbore storage constant Measured wellbore storage constant Gradients of individual fluid samples Flowing pressures at times t0, Ц, t2, tn Flowing times
Units
psi psia minutes minutes minutes stb/day minutes psi bbl ft cp stb/day psia psia psi/cycle md hr decimal psi"1 ft md-ft/cp ft bbl/stb (dimensionless) bbl/psi bbl/psi psi/ft psi minutes
N o t e t h a t k/\i can be c a l c u l a t e d d i r e c t l y f r o m the transmissibility term if the thickness, h, is known. Also, if a good value for |i is known, the formation capacity, kh, can be determined.
Calculating Skin Factor
The skin factor (S) can be calculated as follows:
(^-•Pave)
S = 1.151
-log
kt + 3.23
W /
1.05x10 - 3 kt ~ 2 V <t>]ict
where t is the buildup time in hours and all other terms have been previously defined. The ROI calculated from a DST should not be used to identify faults or boundaries. If such items are in question, they should be determined through further transient testing.
Using the pseudo-steady-state radial flow equation, the potential of the tested interval can be estimated as follows:
If fluid properties are unavailable, skin can be calculated assuming that the log term in the previous equation is equal to 7.5. In most cases, the skin will either be positive or close to zero. A negative skin greater than -1 should be viewed with caution since the well has not been stimulated.
Calculating Radius of Investigation
The radius of investigation (ROI) is important because it helps determine whether or not the test saw beyond wellbore damage. It is defined as
c] = 0.00708' k h '
(P*-Pwf )
r 1N
V r' w 2
/
Note that Pwf in this case is the assumed drawdown by pumping or flowing. It is not the same flowing pressure seen on the DST. Also, the skin factor S here may be the skin expected after cleanup or stimulation.
Table 3. Example 1: Impact of Flow Rates on Reserve Parameters
Method
Superposition Odeh-Selig Simplified Horner
Calculated Flow rate (bbl/day)
1313 192 454 1426
Producing Time (hrs)
1 2.9 1
Permeability (md)
6.6 9.6 13.1
Skin
-4.5 -3.6 -4.1
Drill Stem Testing
137
Reservoir
ROI
Pressure
(ft)
(Psia)
37
2405
—
—
53
2405
COMMON PITFALLS IN DST TESTING AND ANALYSIS
All too often, DSTs are run for fluid recovery instead of for reservoir data. Flowing test times are long and shut-in times short. Producing rates are calculated using the simplified Horner method and can be overly optimistic. Even in DSTs in which the test times are run as recommended, incorrect estimation of fluid recovery can lead to incorrectly calculated flow rates used in reservoir calculations. Care should be taken to analyze the test. For example, a recovery of 50 ft of oil-cut mud might indicate that a zone was too damaged to produce and not that it was tight!
EXAMPLES OF PRACTICAL DST ANALYSIS
Example 1: Calculation of Flow Rates
Well: Location: Test Interval:
TestResults:
Number 26-1 Roosevelt County, Utah 5910-6011 ft Green River Formation Highly gas-cut oil flowed to surface. Stabilized rate was not obtained. Estimated recovery (drill string volume reversed to pit) was 54 bbl of oil and 5 bbl of mud. Test was analyzed using the three methods outlined in Table 3, which shows calculated reservoir parameters.
As can be seen in Table 3, flow rates vary widely depending upon the method used. The calculated steadystate producing rate was 181 BOPD using the superposition method. The actual producing rate for this well was 120 BOPD. The negative skin indicates a fracture system. The ROI was far enough away from the wellbore to see true reservoir permeability. Despite the excellent DST recovery, this well had projected reserves of 50,000 bbl of oil, which did not pay for drilling and completion costs.
Example 2: Incorrect Estimation of Recovery Without Verifying Recovery
Well: Location: Test Interval:
Number 17-1 Williston Basin 8549-8639 ft Duperow
Test Results:
Pipe was pulled to the fluid in the drill pipe. The DST report indicated a recovery of 1575 ft of highly oil and gas-cut water at a 60% oil-cut (11.5 bbl oil and 7.7 bbl water according to pipe measurements). Total flow time on the test was 1 hour. The well was completed, fractured, and plugged after testing less than 30 BFPD. This well should have been plugged after the DST. Pressure data from the gauge above the closing tool indicated that during the flow periods, the fluid pressure in the drill string increased by 115 psig. An oil gravity of 42° API was recorded from the sample chamber. A gradient of 0.389 p s i / f t was calculated for the fluid mixture in the drill pipe . A fluid column of 295 ft of un-gascut fluid was recovered in the drill string. This translated into an actual recovery of 1.8 bbl of fluid (1.08 bbl oil and 0.72 bbl water).
Tlie fact that the drilling fluid was highly gas-cut led to an e r r o n e o u s estimate of fluid recovery and u n n e c e s s a r y investment. It is always important to verify recovery using gauge data and field measured fluid densities.
Example 3: Basing Potential Productivity on Recovery
Wells: Location:
Test Results:
34 and 65 Sussex Field Johnson County, Wyoming These two wells were drilled in the early 1950s and tested the Shannon sand at approximately 4600 ft. The test tools were open for about 1 hour and shut-in for 30 minutes. In both cases, less than 50 ft of oil-cut mud was recovered. Both wells were plugged on the basis of recovery.
Several years later, four offsets were drilled and had initial potentials greater than 200 BOPD after being hydraulically fractured in the same zone. In wells 34 and 65, the formation had been too damaged to produce during the DST. Both wells would have been productive after stimulation had the pressure data been analyzed on these tests.
138
PART 3—WELLSITE METHODS
Part 3 References Cited
American National Standards Institute, 1972, Acceptable concentrations of hydrogen sulfide: ANSI Report No. Z37.2, New York.
American Petroleum Institute, 1960, API recommended practices for core-analysis procedure: API Report No. 40, Dallas, TX, 55 p.
American Petroleum Institute, 1974, API recommended practices for safe drilling of wells containing hydrogen sulfide: API Report No. 49, Dallas, TX, l i p .
Auman, J. B., 1989, A laboratory evaluation of core preservation materials: SPE Formation Evaluation, v. 3, n. 4, p. 691-695.
Anadrill, 1988, Measurement Wliile Drilling-Formation Logging System service brochure, 2000(A-4/89).
Baker Service Tools, 1985, Technical information for the oil & gas specialist.
Barker, C., 1972, Aquathermal pressuring—role of temperature in development of abnormal pressure zones: AAPG Bulletin, v. 56, n. 10, p. 2068-2071.
Barker, C., and B. Horsfeld, 1982, Mechanical versus thermal cause of abnormally high pore pressures in shales— discussion: AAPG Bulletin, v. 66, n. 1, p. 99-100.
Basan, P., J. R. Hook, and K. Hughes, 1988, Measuring porosity, saturation, and permeability from cores: The Technical Review, v. 36, n. 4, p. 22-36.
Bleakly, D. C., D. R. Van Alstine, and D. R. Packer, 1985a, Controlling errors minimizes risk and cost in core orientation in technology: Oil and Gas Journal, v. 83, n. 48, p. 103-110.
Bleakly, D. C., D. R. Van Alstine, and D. R. Packer, 1985b, How to evaluate orientation data, quality control in technology: Oil and Gas Journal, v. 83, n. 49, p. 46-54.
Dickey, P. A., and W. C. Cox, 1977, Oil and gas reservoirs with subnormal pressures: AAPG Bulletin, v. 61, n. 12, p. 2134-2142.
Dickey, P. A., C. R. Shriram, and W. R. Paine, 1968, Abnormal pressures in deep wells of southwestern Louisiana: Science, May 10, v. 160, p. 609-615.
Earlougher, R. C, 1977, Advances in well test analysis: SPE Monograph Vol. 5, Society of Petroleum Engineers, New York, p. 90-103.
Erdle, J. C., 1984, Current drillstem testing practices—design, conduct, and interpretation: Society of Petroleum Engineers Paper No. 13182, p 1-20.
Exploration Logging, Inc., 1979, Field geologists training guide—an introduction to oilfield geology, mudlogging, and formation evaluation: Sacramento, CA, p. 4-52 to 4-57.
Exploration Logging, Inc., 1985, Mud Logging: Principles and Interpretations. Boston, MA, IHRDC, 92 p.
Ferrie, G. H., В. O. Pixler, and S. Allen, 1981, Well-site formation evaluation by analysis of hydrocarbon ratios: 83rd Annual General Meeting of the Canadian Institute of Mining and Metallurgy, Paper 81-32-20.
Fertl, W. H., 1976, Abnormal formation pressures: New York, Elsevier Scientific Publishing Company, 382 p.
Flopetrol-Johnston, 1980, Drill-stem testing manual: p 1#31#28.
Flopetrol-Johnston, 1987, Downhole testing services brochure, p. DTS/M-28[5-87],
Franco, A., 1990, "Hot play" gets hotter in the Austin Chalk: Drilling Contractor, June/July, p. 55.
Hocker, С., К. M. Eastwood, J. C. Herweijer, and J. T. Adams, 1990, Use of dipmeter data in clastic sedimentological studies: AAPG Bulletin, v. 74, n. 2, p. 105-118.
Hottman, C. E., and R. K. Johnson, 1965, Estimation of formation pressures from log-derived shale properties: Journal of Petroleum Technology, v. 17, p. 717-723.
Hunt, P. K., and S. L. Cobb, 1988, Core preservation with a laminated, heat-sealed package: SPE Formation Evaluation, v. 3, n. 4, p. 691-695.
Jorden, J. R., and O. J. Shirley, 1966, Application of drilling performance to overpressure detection: Journal of Petroleum Technology, v. 18, p. 1387-1394.
Keelan, D. K., 1985, Coring Part 1—Why it's done: World Oil, v. 200, n. 4, p. 83-90.
Keelan, D. K., and D. A. T. Donohue, 1985, Core analysis: Boston, MA, IHRDC Video Library for Exploration and Production Specialists, п. PE405,186 p.
Kelton, F. C., 1953, Effect of quick-freezing versus saturation of oil well cores: Petroleum Transactions, AIME, v. 198, p. 312-314.
Kennedy, K. F., ed., 1984, Hydrocarbon well logging recommended practice: Society of Professional Well Log Analysts.
Lacy, L. L., 1984, Comparison of hydraulic fracture orientation techniques: Society of Petroleum Engineers Paper No. 13225,12 p.
Lebeaux, J. M., 1952, Some effects of quick-freezing upon the permeability and porosity of oil well cores: Journal of Petroleum Technology, v. 4, n. 11, p. 19-20.
Magara, K., 1978, Compaction and fluid migration: New York, Elsevier Scientific Publishing Company, 319 p.
Matthews, C. S.,and D. G Russell, 1967, Pressure buildup and flow tests in Wells: SPE Monograph Vol. 1, Society of Petroleum Engineers, New York, p. 84-91.
Michigan Department of Labor, 1989, Personal protective equipment, Part 33: Safety Standards Division Occupational Safety Standards for General Industry, Lansing, MI.
Michigan Department of Public Health, 1989, Occupational Health Division, Occupational Health Standards, Lansing, MI.
Nelson, R. A., L. C. Lenox, and B. J. Ward, 1987, Oriented core—its use, error, and uncertainty: AAPG Bulletin, v. 71, n. 4, p. 357-367.
Occupational Safety and Health Administration, 1983, Code of Federal Regulations, Part 1910 General Industry, Title 29: U.S. Department of Labor, Safety and Health Standards, Washington, D.C.
Odeh, A. S., and P. Selig, 1963, Pressure buildup analysis, variable-rate case: Journal of Petroleum Technology (July), Trans., AIME, v. 228, p. 790-794.
Pettijohn, F. J., P. E. Potter, and R. Siever, 1973, Sand and sandstone: New York, Springer-Verlag, 618 p.
Rowley, D. S., C. A. Burk, T. Manual, and W. F. Kempe, 1971, Oriented cores: Cliristensen Diamond Products Paper, 15 p.
Schlumberger, 1989, Prospector logging while drilling brochure, SMP 5109.
Short, J. A., 1981, Fishing and casing repair: Tulsa, OK, PennWell Books, 365 p.
Smith, M. B., N.-K. Ren, G. G. Sorrells, and L. W. Teufel, 1985, A comprehensive fracture diagnostics experiment, Part II,
Comparison of seven fracture azimuth measurements: Society of Petroleum Engineers Paper No. 13894,16 p. Swanson, R. G., 1981, Sample examination manual: Tulsa, OK, AAPG Methods in Exploration Series, 35 p. Teufel, L. W., C. M. Hart, A. R. Sattler, and J. A. Clark, 1984, Determination of hydraulic fracture azimuth by geophysical, geological, and oriented core methods at the multi-well experiment site, Rifle, Colorado: Sandia National Laboratories Paper SAND 84-0380, Society of Petroleum Engineers Paper No. 13226,15 p.
Torsaeter, O., 1985, The effect of freezing of slightly consolidated cores: SPE Paper 14300,60th Annual Technical Conference and Exhibition, Las Vegas, NV, Sept. 22-25.
References Cited
139
Wallace, W. E., Abnormal subsurface pressures measured from conductivity or resistivity logs: The Log Analyst, v. 6, p. 26-38.
Whitebay, L. E., 1986, Improved coring and core-handling procedures for the unconsolidated sands of the Green Canyon area, Gulf of Mexico: SPE Paper 15385,61st Annual Technical Conference and Exhibition, New Orleans, LA, Oct. 5-8.
Whittaker, A., ed., 1985, Field geologists training guide: Boston, MA, IHRDC, 291 p.
Whittaker, A., 1991, Mud logging handbook: Englewood Cliffs, NJ, Prentice-Hall.
Wisenbaker, J. D., 1947, Quick freezing of cores preserves fluid contents: Oil Weekly, v. 124, n. 9, p. 42-46.
Part 4
WIRELINE METHODS
Contents
• Introduction • Basic Open Hole Tools • BasicToolTable • Basic Cased Hole Tools • Wireline Formation Testers • Dipmeters • Borehole Imaging Devices • Preprocessing of Logging Data • Determination of Water Resistivity • Quick-Look Lithology from Logs • Standard Interpretation • Difficult Lithologies • Formation Evaluation of Naturally Fractured
Reservoirs • ReferencesCited
edited by
Mark W. Alberty
BP Exploration Houston, Texas, U.S.A.
Introduction
Mark W. Alberty
BP Exploration Houston, Texas, U.S.A.
Wireline logs provide a survey of the formations drilled by the bit. These recordings enable geoscientists and engineers to determine reservoir characteristics such as lithology, porosity, fluid saturations, pressure, formation dip, hydrocarbon type, and their associated depth. Logs are an extremely i m p o r t a n t element in the characterization of subsurface formations. However, logs are not capable by themselves of providing full and perfectly accurate reservoir characterization. The best characterizations occur when logs are combined with cores and their associated analysis, mudlogs, measurement while drilling (MWD) data, seismic data, well tests, analysis of cuttings, and production tests. The characterization of reservoir properties from logs only is c o m m o n l y called the science of log analysis. The characterization of reservoir properties from the analysis of all these measurements is commonly called petrophysics.
Part 4 of the Manual focuses primarily on the logging tools, logs, and their associated analysis. The first three chapters address basic open hole and cased hole logging tools and their uses, limitations, and advantages (Alberty). The next three chapters discuss speciality tools and their interpretation, in particular, the formation tester (Smolen), the dipmeter (Goetz), and imaging devices (Luthi). These tool discussions are followed by a series of chapters on the interpretation of logging measurements. The series includes preprocessing (Patchett), determination of water resistivity (Rw) (Peveraro), lithology (Hancock), standard interpretation (Alberty), difficult lithologies (Hashmy and Alberty), and fractured reservoirs (Augilera).
The variety of wireline measurements are affected by the environment in which the log is recorded. Borehole size, mud properties, and invaded zone can all influence the measurements. These environmental factors can significantly alter the apparent responses of the logs, frequently leading to erroneous analysis. Two steps can be taken to minimize the environmental perturbations. First, collect the logging data
u n d e r c o n d i t i o n s that m i n i m i z e the i n f l u e n c e of the environment. Second, be sure to correct the logs for the residual influence of the environment before analysis. Never assume that corrections after the fact can compensate for running logs under less than ideal conditions. Minimizing this influence of the environment at the time of collection can be accomplished by doing the following: optimize mud properties, use appropriate centralizers and standoffs, combine only those tools that require the same positioning within the borehole (e.g., centralized, decentralized, or stood off), and select the types of tools most o p t i m u m for the expected formation properties and hole conditions. Corrections after the acquisition requires an understanding of the correction required, the order in which they may best be applied, accurate knowledge of the borehole environment, and identification of the appropriate borehole corrections for the particular tool used.
Analysis of logs is almost always a problem of more u n k n o w n s than measurements. The "art" of this science comes from knowing which assumptions can best be made when and causing the least amount of uncertainty in the answer. Usually, these assumptions are best made when an analyst is experienced in a given geological area. The best analysts continue to expand their knowledge as they analyze more and more logs. Beware when you think you have conquered the art of log analysis.
ACKNOWLEDGMENTS
Special thanks to the authors: Joe Goetz, Stefan Luthi, James Smolen, Roberto Peveraro, Jay Patchett, Khaled Hashmy, Nigel Hancock, and Roberto Augilera. Thanks also to George Coates, Chuck Konen, and Zaki Bassouini for their contributions. The review editors for this section were Cary Purdy and Klialed Hashmy. Thank you.
143
Basic Open Hole Tools
Mark w Alberty
BP Exploration Houston, Texas, U.S.A.
PURPOSE AND TYPES
Gamma Ray
Open hole logging devices are used to characterize subsurface formations. Common formation attributes that may be characterized include
1. Storage capacity of the formation, which normally includes porosity and fluid saturations Fluid properties, which include density, gas to oil ratio, API gravity, water resistivity and salinity, temperature, and pressure
3. Geological setting, which may include structural or stratigraphic dip, facies characteristics, and reservoir heterogeneities
The basic open hole wireline logging devices can be divided into four general groups, as shown in Table 1. The correlation and lithology devices are used primarily to correlate between wells and to discriminate reservoir from n o n r e s e r v o i r rocks. The resistivity devices are u s e d to determine formation resistivity at varying distances from the wellbore, which is used for correlation and the determination of water saturation. The lithology and porosity devices are used to d e t e r m i n e both lithology a n d porosity. A variety of auxiliary tools a r e u s e d to m a k e special l o g g i n g measurements. (For more on tool specifications, see the chapter on "Basic Tool Table" in Part 4.)
CORRELATION AND LITHOLOGY
Correlation devices are used to identify common formations between wells and to distinguish potential reservoir rocks from nonreservoir rocks. These devices make use of three different physical p h e n o m e n a : s p o n t a n e o u s potential, gamma rays, and photoelectric effect. Table 2 shows the resolution and applications of the correlation devices.
Gamma rays tools measure the natural radioactivity of the formation. This radioactivity is emitted primarily from potassium in the structure of clay minerals, radioactive salts in the formation waters, radioactive salts bound to the charged surfaces of clay minerals, potassium associated with feldspars, and radioactive minerals associated with igneous rocks and rock fragments. The gamma ray response is used for correlation of formations between wells and for estimating volume shale and/or volume clay minerals.
An advanced version of the gamma ray tool, called the spectral gamma ray, breaks d o w n or segments the detected gamma rays by their different energies using spectral analysis techniques. These segments correspond to the radioactive families of potassium, u r a n i u m , and thorium. U r a n i u m frequently occurs as a precipitated salt deposited in a formation from waters having flown through that formation. When this occurs, the uranium counts disguise radioactivity due to mineralogy. The use of the spectral tool allows the removal of gamma ray counts caused by uranium, typically permitting more accurate use of the remaining gamma rays for determining lithology, volume shale, or volume clay. In some local areas, ratios of potassium to thorium have been successfully used to determine some clay types. However, this clay typing has not proven particularly universal and should be attempted with much caution.
Typical presentations of gamma ray measurements are shown in the logs in both Figures 1 and 2. (For information on the cased hole gamma ray tool, see the chapter on "Basic Cased Hole Tools" in Part 4.)
Table 1. Basic Open Hole Tools
Spontaneous Potential
Spontaneous potential (SP) is a natural voltage or electrical potential that arises due to differences in the ionic activities (relative saltiness) of the drilling m u d and the formation waters. This potential can be used to correlate formations between wells, to indicate permeability, and to estimate formation water resistivity. No SP occurs when oil-based mud is used in the borehole. Hydrocarbons and shaliness in the formation suppress the SP. The magnitude of the SP decreases as the resistivity of the mud filtrate and formation waters approach a common resistivity. The direction of SP deflection reverses as the ratio of the resistivity of the mud filtrate (Rmf) to that of the formation water (Kw) reaches 1.0 or more. If there is no contrast in the mud filtrate and formation water salinities, there is no measurable SP. A typical presentation of SP is shown on the left of the log in Figure 1.
Jm Correlation and lithology Resistivity
Porosity and Lithology
Auxiliary
Devices
Spontaneous potential Gamma ray Photoelectric effect
Induction Laterolog Microresistivity
Density Compensated Neutron Sonic
Photoelectric effect
Caliper Formation Tester Dipmeter Borehole Televiewer
144
Basic Open Hole Tools
145
Table 2. Resolution and Applications of Correlation and Lithology Measurements
Vertical
Tool
Resolution
SpontaneouspotentiaI(SP) 6-10 ft
Gammaray
2 ft
Spectral gamma ray
3 ft
Photoelectrical effect (Pe)
2 in.
Radius of Investigation
N/A
12 in. 16 in. 2 in.
Applications Well-to-well correlation,
estimate Rw, and
indicate permeability Well-to-well correlation and
estimate Vsh
Well-to-well correlation and estimate Vsh
Identify lithology and well-to-well correlation
Limitations Does not work in oil-based mud and
Rmi and Rw must contrast
Sensitive to hole size changes
Sensitive to hole size changes
Does not work in barite mud, is a pad device, and uses a radioactive source
Photoelectric Effect
The photoelectric effect, or Pe, measures a formation's ability to absorb gamma rays. The absorptive abilities of formations vary with lithology. The photoelectric absorption is recorded as a supplementary measurement to the formation density measurement, using common detectors and radioactive sources. Since this measurement is part of the density measurement, the tool is a pad contact tool and is subject to borehole wall rugosity. The measurement is not valid in muds weighted with barite. The recording can be used both for correlation of formations between wells and for determining lithology. A typical presentation of Pe is shown in the log in Figure 2.
RESISTIVITY
Resistivity tools are primarily used for correlation and to determine the volume of the pore space saturated with water. Resistivity tools can be divided into three characteristic types: induction, laterolog, and microresistivity tools. The three types each have their individual applications, advantages, and limitations, which are summarized in Table 3.
Induction
Induction tools use electromagnetic coils to establish magnetic fields that excite current flow in the formation, which in turn excites secondary magnetic fields and current flow in receiver coils in the tool. This principle of exciting magnetic fields allows induction tools to measure resistivity without the requirement of a direct electrical connection to the formation. This feature permits the tool to be used in nonconductive muds. Different transmitter and receiver arrays allow focusing of the m e a s u r e m e n t for different vertical resolution and depths of investigation. A typical presentation of a dual induction log is shown in Figure 1.
formation, which is normally provided by the drilling mud. This characteristic does not allow this measurement to be m a d e in oil-based m u d s . The focusing of the laterolog measurement is accomplished through the placement of the electrodes. Generally, laterologs exhibit very good vertical resolution. Because the measured currents must pass through the drilling mud and the flushed zone to enter the unaltered formation, laterolog measurements are usually unfavorably influenced by nonconductive mud and mud filtrate. The presentation of the d u a l laterolog is very similar to the presentation of the dual induction shown in Figure 1.
The deep laterolog measurement current is returned to the earth's surface to ensure deep investigation and to minimize the influence of resistive beds. However, the surface return can give rise to anomalously high resistivity readings for tens of feet below massive, extensive, highly resistive beds. This phenomenon is known as the Groningen effect.
Microresistivity
Microresistivit]/ devices are used to estimate the resistivity of the flushed zone immediately adjacent to the borehole. The devices are of the p a d contact t y p e to e n s u r e that the investigation is very shallow and to minimize the influence of changing hole sizes and tool position within the borehole. This shallow investigation can result in mudcake being a significant influence. Hole size and mudcake corrections are commonly required. Like laterologs, these devices require a direct electrical contact with the formation. For this reason, microresistivity devices cannot be used in oil-based muds.
Formation resistivity is typically profiled with three resistivity measurements of different depths of investigation to characterize the influence of the invading mud filtrate upon apparent formation resistivity. This characterization permits the influence of the flushed zone to be separated from the reading of the deep device for a more accurate determination of the true formation resistivity (Rt).
Laterologs
The laterolog device measures the voltage and current magnitudes associated with a series of current electrodes m o u n t e d on the surface of the logging sonde. These measurements require direct electrical contact with the
POROSITY
Each of the porosity tools—density, compensated neutron, sonic, and photoelectrical effect—can be used to estimate porosity when lithology and fluid properties are known.
146
PART 4— WIRELINE METHODS
Dual Induction/Sonic Log
Compensated Neutron/Litho-Density Log
PJBttPl 5/£i>
[
-0.050
0.4500110000 .
PEF < > 0.0
0.3000
DPHU
)
0.3000
PfHJA
i
."PHI <
>
Jtrti UJ _ .i
0 .0 10.000 -0.100 -0.100
Figure 1. A typical log showing SP, gamma ray, dual induction, and sonic measurements. (Courtesy of Schlumberger, 1983.)
Figure 2. A typical log showing density, compensated neutron, Pe, gamma ray, and caliper measurements. (Courtesy of Schlumberger, 1983.)
Table 3. Resolution and Applications of Resistivity Devices
Tool
Dual induction Deep Medium Shallow3
Phasor induction Deep Medium Shallow3
Vertical Resolution
7ft 5ft 2.5 ft
3ft 3ft 2.5 ft
Radius of Investigation
50 in. 28 in. 16 in.
65 in. 40 in. 16 in.
High resolution induction
Deep
Medium Shallow3
2.5 ft 2.5 ft 2.5 ft
95 in. 60 in. 16 in.
AIT*
4 ft, 2 ft, 1 ft 10 in., 20 in., 30 in.,
60 in, 90 in.
Dual Iaterolog Deep Shallow Microresistivity
ARI*
2ft
45 in.
2ft
16 in.
(See below)
8 in.
4ft
Micro SFL
2-3 in.
1 - 4 in.
Microlaterolog
2 in.
4 in.
Microlog
2 - 4 in.
1-2 in.
aShallow measurements do not work in oil-based muds. bThe ohm-meter (Q-m) is a unit of measurement of resistance.
'Mark of Schlumberger.
Basic Open Hole Tools
147
Applications
Limitations13
Estimate Rv Rxo, and Di in relatively fresh and oil mud systems
Estimate Rv Rxo, and Di in relatively fresh and oil mud systems; reduced shoulder effects
Not recommended; Res > 200 Q-m or Rmi/Rw < 2.5
Not recommended; Res > 250 Q-m or Rm1,R\N < 2 5
Estimate Rv Rxo, and Dl in relatively fresh and oil mud systems; reduced shoulder effects
Estimate Rv Rxo, and Di in fresh and oil mud systems; reduced shoulder and rugosity effects
Not recommended; Res > 250 Q-m or RJRvj <2.5
Not recommended; Res > 200 Q-m or flmf/flw<2.5
Estimate Rv Rxo, and Di in relatively salty mud
Estimate Rv Rxo, and Di in relatively salty mud, locate fractures
Permeability and moved hydrocarbon indicator; estimate Rxo
Permeability and moved hydrocarbon indicator; estimate Rxo
Permeability and moved hydrocarbon indicator; estimate Rxo
Not recommended; RmiZRw > 2.5; Does not work in oil-based mud
Resistivity range 0.2 to 100,000 Q-m
No oil-based muds
No oil-based muds
No oil-based muds
(Methods for estimating porosity from these devices individually are described in the chapter on "Standard Interpretation" in Part 4.) When both porosity and Hthology are u n k n o w n , two or more of the devices can be used together to determine both porosity and lithology. (The most common methods for determining both porosity and lithology are described in the chapter on "Lithology from Logs" in Part 4.) Table 4 shows the resolution and applications of porosity devices.
Density
The density tool measures the apparent density of the formation and then measures the n u m b e r of lower energy gamma rays returning to the detectors. The detectors and source are mounted in a pad that is forced against the borehole wall. The measurement attempts to correct
automatically for the effects of m u d c a k e and minor hole rugosity. The measurement is sensitive to significant borehole wall rugosity and pad standoff, which cause the tool to read too low of a density. A typical presentation of the density (as well as several other parameters) is shown in the log in Figure 2.
Compensated Neutron
Compensated neutron devices measure the hydrogen index of the formation using a radioactive neutron source that bombards the formation with fast-moving neutrons. Neutrons collide with atoms of the formation, transferring their energy through these collisions. The most efficient transfer of energy occurs with hydrogen atoms because the mass of hydrogen is approximately the same as the mass of a n e u t r o n . T w o d e t e c t o r s c o u n t the n u m b e r of
148
PART 4— WIRELINE METHODS
Table 4. Resolution and Application of Porosity and Lithology Devices
Tool
Compensated density Compensated neutron
Vertical Resolution
18 in. 2ft
Radius of Investigation
8 in. 10 in.
Applications
Estimate porosity Estimate porosity and
identify presence of gas
I PL* (Integrated Porosity Lithology)
Sonic
FWS (monopole)
Dipole sonic CMR* (Combinable
Magnetic Resonance) Photoelectrical effect (Pe)
1 ft
2ft 4ft 4ft 6 in. 2 in.
—
Typically 6 in.
Typically 6 in.
Typically 1I i i0 IIiIn. 1 in. 2 in.
Estimate porosity and identify presence of gas, thin bed evaluation, shaly sand evaluation
Measure compressional velocity and estimate porosity
Measure compressional and shear velocities and estimate porosity
Measure shear velocity
Porosity, pore size distribution, permeability
Identify lithology and correlation
Limitations
Pad contact device Needs environmental
corrections; sensitive to standoff from wall Needs environmental corrections; sensitive to standoff from wall
Sensitive to compressibility
Cannot measure shear velocity when shear velocity > mud velocity
—
Minimum 6.5 inch wellbore
Does not work in barite mud and pad contact tool
4Mark of Schlumberger.
deenergized (thermal) neutrons returning from the formation. The ratio of the detector count rates is primarily related to the hydrogen index or the apparent water-filled porosity.
The source and detectors are mounted in a mandrel that, ideally, is pressed against the borehole to minimize the influence of the high apparent porosity of the borehole. This measurement is very sensitive to tool standoff, hole size, temperature, and salinity. Environmental corrections are highly recommended before attempting to interpret results . Gas has a very low hydrogen index compared to water, which causes the tool to report abnormally low porosities in gasbearing formations. When used in conjunction with density measurements, gas-bearing intervals are often easy to identify. A typical presentation of a compensated neutron measurement is shown in the log in Figure 2.
Sonic
Sonic devices measure the velocity of various acoustic waves, most notably compressional, shear, and Stoneley waves. The velocity of the waves is a function of the elastic properties and the density of the formation. Logs normally present the inverse of velocity, called the interval transit time or delta t (At). A number of empirical relationships have been developed to relate compressional velocity to porosity (which are explained in the chapter on "Standard Interpretation" in Part 4).
Two versions of the compressional sonic device are available: the compensated sonic and the full waveform sonic (FWS). The full waveform sonic contains an array of receivers that are used to determine both compressional and shear velocities. Sonics are available in a variety of transmitter-toreceiver spacings from 3 to 12 ft or more. The longer spacings
are capable of investigating deeper into the formation. Both the conventional sonic and the full waveform sonic devices are used to measure compressional velocity. A typical presentation of compressional sonic measurements is shown in the log in Figure 1.
Shear velocities are used to determine mechanical properties of the formations and to determine Poisson's ratio for use in interpreting seismic data. Shear velocities can be determined from the FWS (monopole), the dipole sonic, or the quadrupole sonic. The monopole sonic is not able to measure shear velocities when the shear velocity of the formation is slower than the compressional velocity of the mud. Mud interval transit times are typically in the 190 psec/ft range. When this condition is not met, no shear energy is refracted toward the receivers, making shear velocity measurements impossible. The dipole overcomes this limitation by directly exciting shear flexural energy in the formation regardless of the mud velocities.
The CMR* (MIRL of NuMar) is the latest generation of nuclear magnetic resonance devices. Originally used only for cased holes, the tool is now a pad device for use in open holes. It can be used to estimate not only gross porosity, but pore size distribution and permeability as well.
Photoelectric Effect
The photoelectric effect is used for lithology determination, and its measurement is identical to that described in the correlation subsection. Knowledge of lithology significantly improves the accuracy of interpretation of all the porosity measurements.
Table 5. Resolution and Applications of Auxiliary Devices
Tool
Calipers Formation testers
Vertical Resolution
N/A 0.5 in.
Dipmeters
0.4 in.
Formation microscanner
0.2 in.
Radius of Investigation
N/A N/A
1 in.
1 in.
Televiewer
0.5 in.
0 in.
Formation Microlmager
0.2 in.
0 in.
Basic Open Hole Tools
149
Applications
Determine borehole diameter Measure formation pressures and
recover formation fluid samples Structural dip, stratigraphic
dips, and hole geometry Structural dip, stratigraphic
dips, formation images, and hole geometry Structural dip, stratigraphic dips, formation images, and hole geometry As for Televiewer
AUXILIARY TOOLS
A wide variety of auxiliary wireline tools exist for solving special problems. The more commonly encountered devices are summarized in Table 5. (For more information on other auxiliary tools, see the chapters on "Wireline Formation Testers/' "Dipmeters," and "Borehole Imaging Devices" in Part 4.)
Calipers
Calipers come in a w i d e variety of types, the most common being one-arm, two-arm, three-arm, four-arm, and six-arm. One-arm calipers use the mandrel of the logging device as one side of the caliper and an arm extending out from the body of the sonde as the other. This technique is commonly used in density measurements. This configuration typically measures the long axis of an elliptical borehole. The measurement is unable to characterize accurately hole size changes less than the tool length on the "tool" side of the measurement.
The two-arm device has two caliper arms extending in opposite directions from the tool mandrel. This configuration typically measures the long axis of an elliptical borehole. This
device is able to characterize small changes in hole size on both sides of the borehole. In deviated boreholes, the arm on the low side of the borehole will often collapse under the weight of the logging tool.
The three-arm device is normally a spring-loaded caliper with all three arms ganged to operate in unison. This causes the mandrel of the logging device to center in the borehole; thus, it is typically used with sonic tools for centralization. This device reports the shortest of the three measurements and frequently collapses under tool weight in deviated boreholes.
The four-arm device is typically used on a dipmeter. The most common configuration has opposite arms ganged to work together so that the tool mandrel is centralized in the borehole. One set of arms typically reports the long axis of an elliptical borehole, while the other set reports the short axis. Dipmeters usually have operator-adjustable pad pressure, which increases arm pressure to lift the mandrel of the tool in deviated boreholes. If insufficient pad pressure is applied, the calipers will underreport hole size in deviated boreholes. If excessive pad pressure is applied, the tool will display irregular tool motion as the tool stops and goes with cable tension.
Basic Tool Table
Mark W. Alberty
BP Exploration Houston, Texas, U.S.A.
Logging tools are generally designed for operation under limited borehole conditions. A minimum hole size is the consequence of maximum tool diameter and pad curvatures, while maximum hole size is established by signal strength and caliper arm lengths. Mud types can affect signal transmission. Hole position affects signal strength and mud or borehole effects. Table 1 provides general operating limitations for the standard logging tools. Service companies have specially designed or modified tools that may allow extension of o p e r a t i o n ranges. C o n s u l t y o u r local representative if hole conditions are not as recommended in Table 1.
Computerized surface systems and cable communication systems have made tool combinations virtually unlimited.
H o w e v e r , the combining of different tools into a single logging run may be limited by more than the physical ability to hook them together. Some devices are designed to operate excentered, some centered, and some stood off from the borehole wall. Tool positioning is important in ensuring valid environmental corrections. Table 1 includes optimum hole positions for each device. Caution should be used in combining a tool designed to be excentered, such as the neutron, with one designed to be centered, such as the sonic. The environmental effects upon the measurement may be uncorrectable. Also note that the maximum and minimum hole sizes are general recommendations only. Some loggmg devices are modified for larger and smaller holes. This should be discussed with your local logging company representative.
Table 1. Basic Tool Table3
Minimum Maximum
Hole
Hole
Tool
Size (in.) Size (in.) F
SP
_
_
л/
Gamma ray
6
20
V
Spectral GR
6
20
V
Induction
4.75
20
V
Laterolog
4.5
20
X
Microresistivity
6
16
V
Density I PL*
6
16
V
6
16
V
Photoelectrical
6
Neutron
6
Sonic
6
16
V
16
V
20
V
FWS (monopole) 6
20
V
Dipole Dipmeter
6
14
V
6
22
V
Formation tester 6
FMS
6
16
V
22
V
Televiewer
6
14
V
FMI
6
21
V
Pulsed neutron 2
CMR*
6.5
12
V
20
V
Mud Typeb
SBк о
V V V X V V V V V V NR V X VX V V V V X V V V X V V V V V V V V VX VV V V V V V V V V V V V V V V V V V л/ л/ M V V V V V V V X V V yl V V V VX V V V V V л/ V V
Preferred Hole
Recommended
Positions
Logging Speed
Excentered Stand-Off (in.) Centered
(ft/hr)
V
V
V
NA
V
-
-
<1800
V
NR
-
<900
-
1.5
-
<3600
-
X
л/
<3600
-
-
V
<3600
V
0-2.0
X
<1800
V
0-2.0
X
1800
V
0-2.0
X
<1800
V
X
X
<1800
-
-
V
<3600
X
X
V
<1800
-
X
V
<1400
X
X
V
<3000
NA
NA
NA
NA
X
X
V
<1800
X
X
л/
<1200
X
X
V
<1800
V
-
-
<1800
V
-
-
1800
aSymbols: V = good conditions, X = unsuitable conditions, - = marginally acceptable conditions, NR = not recommended, NA = not applicable, M = with special modifications. bMud types: F = freshwater (low salt), S = high salt, B = barite, K = high potassium salt, O = oil-based. 'Mark of Schlumberger
150
Basic Cased Hole Tools
Mark w Alberty
BP Exploration Houston, Texas, U.S.A.
PURPOSE OF MEASUREMENTS
Basic formation evaluation measurements are made through casing for three general reasons:
1. To supplement measurements taken in open hole. It may be necessary to supplement open hole measurements because die well conditions may preclude the possibility of reliably or safely making the necessary open measurements or because insufficient or inappropriate open hole measurements were taken over particular zones.
2. To monitor the changes information properties that have occurred since the casing was set. During the life of a well, changes in saturation of the pore space by oil, gas, and/or water may be induced by sustained production. When these changes occur, evaluation of the nature of the changes may be necessary to design more effective hydrocarbon recovery strategies.
3. To provide a depth reference between open hole and cased hole measurements and services.
BASIC TOOLS
The basic cased hole tools can be divided into three general groups: correlation, saturation, and porosity. The correlation device is used to correlate cased hole measurements with open hole measurements and to estimate shale volume. The saturation device is used to determine water saturation when porosity and water salinity are known. The porosity devices are used to estimate porosity when lithology is known. Table 1 lists the types of devices used for these purposes, and Table 2 gives the resolutions and limitations of the various devices. (For more details on tool specifications, see the chapter on "Basic Tool Table" in Part 4.)
Gamma Ray Tool
The gamma ray measurement responds to naturally occurring gamma rays from the formation. These gamma rays are able to penetrate steel casing. This permits the gamma ray to be used in a cased hole for correlation with open hole logs, for the discrimination of sands and shales, and for the calculation of shale volume. Its use is essentially identical to its use in an open hole with the exception of minor environmental corrections needed for the influence of the steel casing and cement. A common problem encountered in the cased hole use of the gamma ray device is scaling of radioactive salts in casing. When produced water containing dissolved radioactive salts enters the casing, the encountered drop in pressure may cause the salts to precipitate from the waters and deposit on the casing near the perforations. These salts will normally dominate the gamma
ray response near the perforations, making the gamma ray useless in those intervals for both correlation to open hole or the estimation of shale volume. If the salts are predominantly uranium, their influence can be removed through the use of the spectral gamma ray. However, if the salts are in part potassium, their influence cannot normally be corrected through the use of the spectral gamma ray. The gamma ray is usually run in combination with a collar locator to provide a depth reference for mechanical cased hole services. (For information on the open hole gamma ray tool, see the chapter on "Basic Open Hole Tools" in Part 4.)
Compensated Neutron Tool
The compensated neutron measurement is little affected by the presence of steel casing. This p e r m i t s the compensated neutron to be used in cased hole to estimate porosity when lithology is known. However, the presence of gas in the formation will cause the compensated neutron to underestimate porosity significantly. (For more information on the compensated neutron tool, see chapter on "Basic Open Hole Tools" in Part 4.)
Pulsed Neutron Tool
Pulsed neutron devices are electronic devices that generate pulses of high energy neutrons. These high energy neutrons bombard the formation, losing energy as they collide with atoms of the rock. Eventually, the neutrons lose so much energy that they are captured (generally by chlorine that exists as part of the salt dissolved in the formation waters). When a neutron is captured, a gamma ray is emitted. The detectors in the pulsed neutron tool are designed to measure these "capture" gamma rays, thus, a "capture cross section" of the formation through casing can be determined. This formation property allows one to estimate the water saturation when porosity and formation water salinity are known. A typical pulsed neutron log is shown in Figure 1.
The gamma rays detected by the pulsed neutron devices can also be processed in a manner similar to the compensated neutron and provide a very similar estimation of porosity.
Table 1. Application and Types of Basic Cased Hole Tools
Application Correlation
Saturation Porosity
Type of Device
Gamma ray Spectral gamma ray
Pulsed neutron
Compensated neutron Pulsed neutron
151
152
PART 4— WIRELINE METHODS
Table 2. Resolutions and Limitations of Cased Hole Logging Devices
Tool Gamma ray Spectral gamma ray Pulsed neutron
Compensated neutron
Vertical Resolution
2ft
3ft
2ft
Radius of Investigation
12 in.
16 in.
18 in.
Applications Well-to-well correlation and
estimates Vsh Well-to-well correlation and
estimates Vsh
Determines water saturation and estimates porosity
3ft
10 in.
Estimates porosity
Limitations Affected by radioactive scale
near perforations
—
Does not work in freshwater and severely underestimates porosity in the presence of gas
Severely underestimates porosity in the presence of gas
This estimation of porosity can be reasonably good w h e n the formation water is relatively salty and the formation does not contain significant a m o u n t s of gas. This combined m e a s u r e m e n t of c a p t u r e cross section a n d porosity from this single device allows the pulsed neutron to be used frequently as a single pass cased hole formation evaluation device. As with the compensated neutron, the presence of gas in the formation will cause the pulsed neutron to underestimate porosity significantly.
The p r o d u c t i o n of f l u i d s f r o m a r e s e r v o i r causes changes in the water saturations of producing reservoirs. Cased hole m e a s u r e m e n t s of the changes in f o r m a t i o n saturations are an extremely important diagnostic tool for the development of strategies for complete hydrocarbon recovery. Bypassed zones can be identified by their virgin state saturations. Intervals of water breakthrough can be identified by their high water saturations. Candidate zones for secondary and tertiary recovery techniques can be identified through evaluation of residual hydrocarbon saturation measured in cased hole by pulsed neutron devices.
Techniques to determine the residual oil saturation of reservoirs are most accurate when contrasting fluids can be injected into the formations to displace natural formation waters. This method of logging the formation with native waters, then injecting contrasting waters and logging again, has become known as log-inject-log techniques. These techniques are most easily accomplished when casing is in place to provide a means to inject the fluids. This method provides the pulsed neutron with an
advantage over open hole logging methods in accurately determining the amount of oil available for secondary or tertiary recover methods.
NEW DEVELOPMENTS
As the interest in and need for greater formation evaluation capabilities grow, the industry continues to develop newer and more sophisticated measurements in through-casing applications. Pulsed neutron spectroscopy is a promising method to measure lithology, porosity, and s a t u r a t i o n s t h r o u g h casing, e v e n in the p r e s e n c e of freshwater. Their basic principle lies in activating elements through bombardment by very high energy neutrons. The energy of the activated g a m m a rays are m e a s u r e d and used to perform elemental analysis of the matrix a n d / o r the fluids. Variations are being marketed under the g e n e r a l n a m e s of c a r b o n / o x y g e n logs or g a m m a spectroscopy. The latest generation tool by Schlumberger, the RST* (Reservoir Saturation Tool), combines a carbon/oxygen and TDT system and determines formation oil saturation even in flowing wells. Other companies are testing resistivity devices that measure formation resistivity through casing. The a d v e n t of full waveform acoustics have enabled some vendors to successfully measure the velocity of compressional a n d / o r shear waves in formations behind casing, under appropriate conditions.
Basic Cased Hole Tools
Wireline Formation Testers
,ames 1 Smolen
Consultant Missouri City, Texas, U.S.A.
INTRODUCTION
Formation testers are a class of wireline tools used to measure the downhole pressure of formations. Stationary measurements of formation pressure in an open hole are made at any number of depths during a single trip into the hole. These pressure measurements are useful in determining the following: (1) variations in pressure a m o n g various formations, (2) gradients of fluid pressure within a formation that can indicate fluid content, (3) gas-oil or water-oil contacts, and (4) local permeability. Comparison with initial reservoir pressures in development wells may indicate zonal pressure depletion. Formation testers are also used to retrieve samples containing fluids.
OPERATION OF THE TOOLS
When the tool is being moved downhole, the formation tester is in its retracted configuration. When a depth is selected for a pressure measurement, the formation tester is activated and hydraulically set or pressed against the formation. These conditions are shown schematically in Figure 1.
Two modes of measurement exist: (1) a pretest measurement in which formation pressures are examined and (2) a fluid sample measurement in which formation fluids are physically withdrawn and retrieved to the surface for examination. The sequence of tool movements required to measure formation pressure begins when the formation tester is activated at a selected depth. The sampling system shown in Figure 2 illustrates how this measurement is made. Prior to activating the tool, the pressure gauge measures hydrostatic mud column pressure. Upon activation, the equalizing valve is shut and the packer and probe are pushed against the formation. The tool is now set. Within a few seconds, the pistons of the pretest chambers begin to withdraw, causing fluid from the formation to flow into the tool through the packer and probe assembly. The pretests are done sequentially with a small volume (typically 10 cc) of fluid drawn into chamber #1 over about 15 sec, followed by a similar volume flowing into chamber #2 at a higher flow rate. Upon completion of the pretests, a sample can be taken or the tool can be retracted. During retraction, the equalizing valve is opened and the pretest pistons expel the fluid taken in. The tool is now ready for the next test depth. Different vendors may use variations of this design that incorporate a single pretest chamber of fixed or variable volume.
PRESSURE MEASUREMENT
During the pretest sequence, pressure is monitored with the tool sampling system. The schematic drawing in Figure 3 shows the pressure typically recorded during the pretest
> >
>
(a)
>
(b)
Figure 1. Wireline formation tester, (a) Retracted configuration (tool closed), (b) Set configuration (tool set).
FLOW LINE
PACKER •FILTER PROBE
PRESSURE * GAGE
EQUALIZING VALVE
(to mud column )
SEAL VALVE
(fo lower somple chamber)'
r Q ] CHAMBER-* I
Ц Ю CHAMBERS 2 PRETEST CHAMBER
SEAL VALVE
(to upper sample chamber)
Figure 2. Dual pretest formation tester sampling system. (From Smolen and Litsey, 1979.)
154
FLOW RATE
Wireline Formation Testers
155
TOOL COLLAPSED TOOL MUD COLUMN SETTING PRESSURE
PRETEST 1 ACTIVE
PRETEST 2 ACTIVE
SHUT IN PRESSURE APPROACHING FORMATION
PRESSURE
Figure 3. Pressure recording by a dual pretest formation tester. (From Smolen and Litsey, 1979.)
sequence. In this figure, time is shown as increasing to the right and pressure and flow rate are increasing upward. Initially, and when the tool is being set, the equalizing valve is open and the recorded pressure is hydrostatic mud column pressure. During the drawdown associated with the first pretest, the pressure is observed to drop, stabilize, and further decrease during the second pretest.
When the second pretest is completed, the flow stops and the pressure builds up to formation pressure. Figure 3 illustrates the flow rates associated with the first and second pretests. Actual formation pressure can be taken when the pressure becomes stable or can be estimated from the character of the buildup if it has not yet stabilized when the tool is retracted.
Figure 4 is a typical log recorded by a formation tester that has a single pretest chamber. In the left log track, the solid curve represents the pressure detected by the gauge within the tool sampling system. Time is increasing downward along the log. The hydrostatic mud column, drawdown, and formation pressures are clearly detectable. The sudden increase at about 105 sec corresponds to the tool retraction and opening of the equalizing valve. On the right are four narrow tracks displaying the thousands, hundreds, tens, and units digit of the measured pressure. For example, after 80 sec, the formation pressure is 3927 psi. It is obvious on the units track that the pressure is increasing and actual formation pressure has not yet been reached.
FORMATION PERMEABILITY
Formation permeability has a significant effect on the drawdown response during a pretest. The pretest pressure recordings shown in Figure 5 illustrate typical records for sandstones at 0.1,1,10, and 100 md (millidarcys). While these figures are qualitative, quantitative techniques exist for estimating permeability using both the pressure drawdown and buildup characteristics of the pretest or sample test.
Figure 4. Typical log record of a single pretest formation tester showing both analog and digital log presentations. (From Western Atlas International, 1987; courtesy of Atlas Wireline Services Division of Western Atlas International, Inc.)
PRESSURE APPLICATIONS
Although formation testers can take samples, they are often run solely for the pressure information available from the pretest. Figure 6 shows how such pretest data can be useful. In this figure, the well is shown penetrating a reservoir that has gas, oil, and water intervals. The formation tester is set at numerous depth intervals across this reservoir. The formation pressure recorded by the tool is indicated by an "x," while the hydrostatic mud column pressure is indicated by an "o." The degree of overbalance (that is, the difference between the mud and formation pressures) is clearly visible from the schematic. The fluid gradients are also detectable, and the gas column is readily distinguished from the oil, which is also distinguishable from the water. The location of the gas-oil and water-oil contacts can also be determined from the formation pressure profile.
After a reservoir has been produced, some pressure decline can be expected. Formation testers are frequently run in development or in-fill wells. When compared to initial reservoir pressures, pressure profiles in these wells often show that certain zones may have produced more than n e i g h b o r i n g zones, t h e r e b y i n d i c a t i n g the p r e s e n c e of permeability barriers.
Pressures in zones of injection have also been monitored using the Formation tester. Such a case is shown in Figure 7. In this example, 22 wells were drilled some years after water flooding was begun in a reservoir. Formation tester pressure data from these 22 wells were used to plot a contour map of the formation pressure. This map clearly shows high pressure ridges associated with the b a n k s of injection wells and troughs associated with the producing wells.
156
PART 4— WIRELINE METHODS
h
Qr Ш
-- • -
I'
I
!
i
i
—Ir I
«duu I iuuma
а
•
I I
HIi -
ADUU I iu mo
KJ
...
--
- -
—
—
-• -
— """""
VZ -
— —
—
'
!
—
i
- - г - I
—
-
-
M 1111 III I Ml II A B O U T .I m d
Figure 6. Pretest pressure response measurements across a reservoir. (Courtesy of Schlumberger Well Services, 1981.)
- d
FLUID SAMPLING
—
Lrh::
f
f
... !: . . : TIGHT _e —
— — •
Based on the ability of the fluid to freely fill the pretest chamber, a larger sample of formation fluid can be taken for analysis on the surface. The larger samples can range from 1 to 10 gal or more. Due to m u d filtrate invasion, a large fraction (if not all) of the retrieved fluid may be mud filtrate. Proper analysis of the sample involves discriminating the
i filtrate from the native formation fluids.
Measurements of water resistivity, API gravity, gas to oil ratio, and water chemistry can be performed at the wellsite with prior planning with the vendor. Formation fluid samples can also be maintained at formation pressures and shipped to laboratories for detailed analysis. However, shipment of pressurized samples may require use of special vessels approved by the Department of Transportation, and prior arrangements should be made with the involved vendors.
Figure 5. Pretest pressure response to formation permeability as measured by a dual pretest tool. (From Smolen and Litsey, 1979.)
Wireline Formation Testers
157
Figure 7. Pressure contour map in zone of water injection. (From Western Atlas International, 1987; courtesy of Atlas Wireline Services Division of Western Atlas International, Inc.)
Dipmeters
Joseph F. Goetz
Oil & Gas Consultants International, Inc. Tulsa, Oklahoma, U.S.A.
INTRODUCTION
As long as oil migrates updip, it would seem there is nothing more fundamental in oil exploration than determining which way is up. It is for this basic purpose that the electric wireline technique known as the dipmeter was designed. From these primitive beginnings, however, the d i p m e t e r has evolved to become a device capable of providing a major input into a complete geological description of the formations crossed by a borehole.
DIPMETER DATA ACQUISITION
The determination of dip angle and direction of a planar surface requires the elevation and geographical position of at least three points. Dipmeter tools achieve this result by measuring some sensitive formation parameter by means of three or more identical sensors mounted on caliper arms so as to scan in detail different sides of the borehole wall. A bedding plane crossing the borehole at an angle would generate anomalies at each sensor, and these anomalies
would be recorded at slightly different depths on the surface recording. The relative displacements and the radial and azimuthal positions of each sensor are then used to compute dip relative to the tool. Microresistivity has been the traditional formation parameter logged. Modern dipmeter tools usually carry more than three sensor arms, the latest version being a device with six arms. More measure points provide the advantage of systematic r e d u n d a n c y , which allows the application of statistical error m i n i m i z a t i o n techniques.
For the results to be geographically significant, it is necessary to define the orientation of the tool in space. This involves continuous measurements of the orientation of the electrode array relative to north, its rotation relative to the high side of the hole, and the inclination of the tool axis from vertical. Such navigation data are produced from the output of an a s s e m b l y of t h r e e o r t h o g o n a l l y m o u n t e d magnetometers and a similar array of accelerometers. Figure 1 is a sketch of a four-arm tool illustrating the orientation measurements. Figure 2 is an expanded scale recording of the raw dipmeter data showing the orientation curves, calipers, gamma ray, and correlation curves from a six-arm tool. Note that the curves are responding to apparent bedding features less than 1 in. thick.
THIN BED RESOLUTION
In the detailed d e s c r i p t i o n of r e s e r v o i r s , thin bed resolution by wireline techniques is a vital factor. For example, the evaluation of laminated shaly sands can be best accomplished by a sensitive delineation of sand and shale layers. Rather than running a tool designed exclusively for this purpose, money can be saved if the desired information can be derived from a logging tool routinely run for other purposes. Dipmeter surveys, particularly those with a calibrated resistivity scale, can be used for this purpose. Several descriptive properties can be derived from the dipmeter survey:
• Lamination thickness and regularity • Layering contrast and frequency • Layering continuity • Rasering and load structures
Figure 1. Sketch of a four-arm dipmeter tool illustrating pertinent orientation measurements.
In Figure 3, a raw dipmeter curve sharply distinguishes between sandstone and shale layers thinner than 1 in. By establishing a cutoff line as shown, everything to the right can be identified as sand and everything to the left as shale. A reliable measurement of net sand thickness is thus provided, as well as the bulk volume fraction of shale in laminated form for input into laminated shaly sand saturation equations.
158
Dipmeters
159
PDD1
SIX ARM DIPMETER
-»• M
X830
1855
AccZ
i?
3 Cals
AZ #1
1865
WaterBased Mud
X835
Resistivity
Figure 2. Expanded scale recording of raw dipmeter data from a six-arm tool.
DIP COMPUTATION
The conversion of raw data into useable dip quantities involves three operations:
• Curve correlation and displacement definition • Dip determination • Data manipulation and presentation
It is important to realize that each of these stages usually involves filtering to some extent which adds a perspective factor to the results. Therefore, results will vary somewhat with the computation technique employed. The fact is, at any one depth, more than one true dip exists; which dip is reported is a matter of perspective.
In the curve correlation phase, the relative displacement between any set of two curves is determined by using one of many mathematical cross-correlation procedures similar to those used in seismic processing. Usually this involves crossmultiplying the incremental curve values and integrating over some window of depth. The length of the correlation window
X840
Figure 3. Example of thin bed laminated sandstone-shale resolution by means of a dipmeter correlation curve.
effectively acts as a stacking filter—the longer the window, the stronger the filtering effect.
In the dip determination phase, one has a choice between a geometric solution and a stochastic approach. While a geometric solution works well in a uniquely determined (three-point) situation, it becomes cumbersome in an overdetermined condition such as that found with four-arm and six-arm tools. In these cases, a stochastic or global mapping approach is more effective in that it uses the redundancy to advantage in minimizing errors. Actually, this approach acts like another stage of stacking filtering. It has the added advantage of neatly solving for orientation of the tool in space.
160
PART 4— WIRELINE METHODS
DIPMETER PRESENTATIONS
The most common presentation of dipmeter data is the arrow or tadpole plot, which is a clever two-dimensional representation of a three-dimensional quantity. In this plot, the base of the a r r o w is positioned at the d e p t h of the midpoint of the correlation interval, and the distance from the left-hand margin to the base of the arrow is proportional to the true dip angle as calibrated by the scale shown on the heading. The shaft of the a r r o w points in the d o w n d i p direction with true north being straight up the page. Figure 4 is a standard arrow plot that also carries a correlation gamma ray curve and maximum and minimum caliper values. On the right-hand side is a representation of the inclination angle and direction of the tool, which will usually be similar to the deviation of the borehole. A n u m b e r of other computer generated presentations have been introduced. Many are useful in special situations, but none is a replacement for the arrow plot.
APPLICATIONS OF DIPMETERS
Structural Applications
Superficially, the determination of structural dip from a dipmeter seems simple and straightforward. In practice, it may be tricky. Difficulties arise from questions of scale and p e r s p e c t i v e . Structural dip by d e f i n i t i o n is the d i p of recognizable lithological unit boundaries in the general vicinity of the borehole. H o w e v e r , a s h a r p c h a n g e in lithology, such as from a shale to a sandstone, is the signature of a catastrophic event in geological history. Therefore, such a contact is liable to be highly irregular over the extremely short section exposed by the borehole. On an outcrop, a field geologist w o u l d astutely m e a s u r e the d i p of an eyeball average of the contact. The stringent confines of the borehole offer no such luxury of perspective, and the section of the contact exposed is liable to be highly unrepresentative of the average structural dip. Outcrop perspectives cannot be extrapolated to the borehole, so a different approach is needed.
Because shale is a low energy deposit, it is generally deposited in thin laminations that are parallel, planar, and horizontal at the time of deposition. In the subsurface, then, shale laminations tend to be parallel to one another and parallel to the average of most lithological bed boundaries. In dipmeter computations, there are a number of ways in which the influence of parallel bedding can be accentuated, thus biasing the results toward structural dip. Essentially these are stacking filter effects that can be applied at various stages in the computation chain, such as in the following:
Figure 4. Dip data expressed on a standard arrow plot.
1. Longer correlation intervals (e.g., 10 ft) 2. Correlation interval overlap 3. Global mapping dip computation 4. Clustering techniques 5. Polarplots 6. Vectoraveraging
Given computation approaches tailored for structural applications, structural dip can then be defined by looking for
a consistent trend on the arrow plot. The most repetitive dip should be the structural dip.
The i n t e r p r e t a t i o n of s t r u c t u r a l a n o m a l i e s is best accomplished by comparison to a set of models—the simpler the model, the better. A simple model, such as the one shown in Figure 5 for a normal fault with drag, is adequate to describe the geometry of such a fault. Figure 6 shows a more complicated arrow plot of low angle dips, reducing to a minimum then increasing to a high angle with the azimuth
Dipmeters
161
Figure 5. Simple dip model for the description of a normal fault with drag.
Figure 6. Model of a tilted plunging anticline as it would appear on an arrow plot.
162
PART 4— WIRELINE METHODS
changing continuously with the dip angle. This is the signature of a tilted plunging anticline. A cross-sectional sketch of the anticline can be produced using the rule of interchangability of perspectives in which the horizontal geometry is interpretable from the vertical pattern of dip. (The application of d i p m e t e r data to solving structural problems is covered in the chapter on "Evaluating Structurally Complex Reservoirs" in Part 6.)
Stratigraphic Applications
For stratigraphic applications, comparisons to a set of models is also a valid interpretation approach, but in this case, very simple models do not work well. Sedimentary geology is too complex. In fact, hand-drawn models may not be valid at all. It is preferable to use independently verified field examples, such as the one shown in Figure 7. In this example, the s a n d y s e d i m e n t s are m a d e u p of a s e q u e n c e of interrupted, truncated, and stacked point bars deposited by meandering stream activity. Few complete fining-upward point bar cycles (gravels to sands to shales) are present. Instead, erosional cuts and festoon cross-bedding indicate the start of a new cycle that may or may not be interrupted before the d e p o s i t i o n of l o w a n g l e p l a n a r c u r r e n t b e d d i n g . (Additional characteristics of meandering streams are covered in the chapter on "Lithofacies and Environmental Analysis of Clastic Depositional Systems" in Part 6.) In the general case, a catalog of field examples of dip plots from various known depositional environments is a valuable aid to stratigraphic dipmeter interpretation.
Gamma Ray Depth DipAngIeandDirection 4
(I
AIM
t !><»
Caliper
i <isi
OMNIDIP
(i
Mm
Hi
о
Iii
'A) : « ) >111 M > n i l / « ш и т
f,
M.I*
Ili Itn
Cu rren Bedding
@
tr
J J INlrEF JP"ГЕ D P<DlhJT ЗА Ri
^ « Ia—
Hole Dritt
и 'i i n
-
Planar
9140
Festoon
t> %
?«O
£ ® 3
--
Planar
Icr^
Curved
h' *
(
\® 4 Ь
9160
U
1
T
л
<c TI _
u
i Q
3
5
--
* n
с D1U3
-
I
•
Curved
- а - л о — •o
;
Figure 7. Field example of a detailed dip computation through a sequence of interrupted meandering stream point bars.
Borehole Imaging Devices
Stefan M. Luthi
Schlumberger-Doll Research Ridgefield, Connecticut, U.S.A.
INTRODUCTION
Borehole imaging devices are high resolution wireline tools that produce maps of physical measurements of the rocks exposed on the borehole wall. They are thus twodimensional logs, with depth and azimuth as independent variables.
Three main imaging techniques are presently in use:
• Ultrasonic reflection • Electrical scanning • Optical scanning
Among these, the optical method is omitted from the following discussion because it is confined to boreholes with clear borehole fluids and is rarely used in the petroleum industry.
Table 1 summarizes the main characteristics of the two tools discussed in this chapter. The currently available borehole image methods provide only semiquantitative physical measurements and therefore are not suitable for the estimation of bulk rock properties as is done with wireline logs. Because of the two-dimensional nature of borehole images, the principal information is structural, that is, it is pertinent to the geometrical arrangement of rock layers caused by sedimentary and tectonic forces. This makes borehole images powerful tools in the hands of petroleum geologists and reservoir engineers in determining lithofacies types, folds, faults, and fractures influencing reservoir compartmentalization, extent, and performance.
Borehole images are normally displayed in a depth and
azimuth reference system, which corresponds to the projection of the cylindrical surface of a drill core onto a plane that is split by tradition along magnetic north. In this projection, an inclined planar feature such as a bedding surface, fracture, or fault shows up as a sinusoid. The minimum, or trough, of a sinusoid points in the direction of the dip azimuth, while the amplitude (divided by the borehole diameter) indicates the dip angle.
Most borehole images are digitized, either downhole or uphole, and are thus amenable to digital computer processing such as image enhancement (e.g., filtering, sharpening, and false coloring), higher level image analysis (e.g., feature recognition and extraction), and threedimensional projections. Interactive graphic workstations are useful tools for manipulating and interpreting borehole images.
ULTRASONIC REFLECTION TECHNIQUE: THE BOREHOLE TELEVIEWER
Acoustic borehole televiewers all contain a rotating ultrasonic transducer that emits a pulse and receives the signal reflected by the borehole wall (Figure 1). The transducers are either focused or unfocused and have frequencies in the range of 200 kHz to 1 MHz; they typically perform 3 to 16 rotations per second. Orientation of the transducer is normally provided through a magnetometer. Two types of images can be obtained: one based on the reflected amplitude and one on the two-way traveltime. By convention, darker gray tones are used for lower reflected
Table 1. Summary of Characteristics of Borehole Imaging Devices
Characteristic
Typical logging speed Borehole wall coverage
Borehole Televiewer
800 ft/hr (245 m/hr) Full (360°)
Borehole size Borehole muds Typical Sampling Rates:
Vertical Horizontal Properties influencing measurement
Shortcomings
Main applications
6-12 in. (15-30 cm) Best results in light muds
0.2-0.5 in. (5-12.5 mm) 0.1 in. (2.5 mm) Acoustic impedance contrasts;
borehole geometry Sensitive to borehole rugosity
and drillmarks; sensitive to tool eccentering and borehole ellipticity Fracture detection (orientation); borehole geometry; breakout detection (orientation)
263
Formation MicroScanner
1600 ft/hr (490 m/hr) Partial (depending on number
of pads and runs) 6-18 in. (15-46 cm) Conductive muds only
0.1 in. (2.5 mm) 0.1-0.2 in. (2.5-5 mm) Rock microresistivity
Resolution decreased by pad stand-off; irregular tool speed causes layer misalignment
Bedding types; fracture detection (orientation and width); fault detection (orientation)
164
PART 4— WIRELINE METHODS
Acoustic Signal To Surface
•0
Figure 1. Borehole televiewer technique. (From Zemanek et al., 1969.)
[0.5 m
!Г",' •„';. Л-4!V Vj^ 0.5 m
Figure 2. Drill marks (arrow) on borehole televiewer images.
Figure 3. Borehole televiewer image showing fracture (arrow) crossing the wellbore.
0.5 m
amplitudes and higher two-way traveltimes. The borehole televiewer provides strictly a borehole wall
surface measurement and as such is affected by rugosity, drill marks (Figure 2), borehole ellipticity, and tool eccentering. It is an effective tool when good hole conditions with relatively smooth borehole walls prevail. Reflected amplitude images are successfully used in the characterization of fractured reservoirs (Figure 3) and to a lesser degree for layered reservoir analysis (Figure 4). The transit time images provide an acoustic, full-circumference caliper of the borehole which is used for wellbore volume calculations, breakout analysis (Figure 5), and wellbore rugosity estimates.
ELECTRICAL BOREHOLE SCANNING: THE FORMATION MICROSCANNER1
Electrical b o r e h o l e s c a n n i n g is an extension of the dipmeter technique. In this method, a large number of closely
Figure 4. Borehole televiewer images showing shale layers (dark) intercalated with limestone layers (bright).
spaced electrodes of 0.2-in. d i a m e t e r are m o u n t e d on a conductive pad and pressed against the borehole wall. The amount of current emitted from each electrode is recorded as a function of azimuth and depth (Figure 6). The tool thus produces a microresistivity map. Currently on the market are versions with two and four imaging pads perpendicular to each other. The resulting image stripes are about 2.5 in. wide and are oriented azimuthally by a downhole magnetometer
1Trademark of Schlu mberger.
Borehole Imaging Device
165
Л
0.5 m
л
INSULATING SUB INSULATING "SLEEVE
INCLINOMETER
HYDRAULICS 4 ARM SONDE
6
PAD CONFIGURATION
JL-Jl
i> 7cmt HK 5 m m
O O
O
Figure 5. Borehole televiewer image showing breakouts (dark patches) at NNW and SSE.
Figure 6. Formation MicroScanner technique. (From Ekstrom etal., 1987.)
unit. In a 8.5-in. borehole, the four-pad Formation MicroScanner provides a 45% circumferential coverage. Repeat logging passes can often increase this percentage. By convention, darker gray tones are used for lower resistivities.
Formation MicroScanner images record changes in rock resistivity caused by variations in porosity and clay content of a small rock volume in the vicinity of the borehole wall. Increased pad stand-off due to mudcake buildup on the borehole wall may decrease the spatial resolution, while abrupt changes in tool movement may produce a local misalignment or sawtooth effect in the layers (Figure 7). Important bedding types and surfaces can be identified and measured for their dip and azimuth (Figure 8), as are fractures (Figure 9) and stylolites (Figure 10). Faults are often readily recognized on Formation MicroScanner images because of the offset of rock types across the fault plane.
0.5 m
1|*|48|il,1!,illH|iP! I
S
W
N
Ш Ш Iwiiiiii'l!l.|iif" : l a w * * !
d.m.
Figure 7. Formation MieroScanner images of cross-bedded sequence showing drill marks (d.m.) at WSW and sawtooth effects (s.e.) on some layers.
166
PART 4— WIRELINE METHODS
Figure 8. Formation MicroScanner images showing dipping unconformity at contact (arrow) of Cretaceous clastic rocks with Mississippian carbonates.
Figure 9. Formation MicroScanner images of open conductive fracture (arrow).
Figure 10. Formation MicroScanner images of stylolitic limestone.
Preprocessing of Logging Data
J. G. Patchett
Amoco Production Company Tulsa, Oklahoma, U.S.A.
INTRODUCTION
The preprocessing or data preparation step is potentially the most time-consuming part of a large logging project. It is estimated that 60% to 80% of the project time is spent on editing and other data preparation tasks. These activities are key to a successful logging project.
DATA PREPARATION STEPS
Data preparation is usually conducted in a prescribed order. The order may vary slightly because of personal preference and the nature of available computer programs. The order recommended here is as follows:
1. Move digital log data from logging field tapes or digitized logs to the log processing environment. If paper copies of logs were digitized, all digital log data should be plotted and compared with the original log hard copies.
2. D i s p l a y all logging data. This way the user can become familiar with typical responses for the different lithologies and/or formations in the study area. This is an initial quality control step and is also used as input to the merging of logging runs.
3. Collect all information related to the logs. Locate the headers for all logs, and identify the logging tool model used on each well and logging run. Collect all temperature data and mud properties. Determine the top and bottom of the valid logging measurements for each curve to be utilized for each logging run. The stacking of logging tools can lead to large differences (several tens of feet) in the starting depth of individual measurements.
4. Merge logging r u n s (if required). When merging different logging runs in a single well, the first problem is that the depths may be different (for equivalent positions in the wellbore) from run to run. Depth shifts between runs should be made when necessary. Run overlaps should be noted because they allow the user to compare common-point depth measurements from different logging runs. This comparison of two logging measurements at the same depth aids in quality control.
5. Edit logs to eliminate invalid data. Erroneous data can be recorded in casing at the top of the logging run a n d / o r recorded with the tools setting on bottom before pick up. These data should be deleted (replaced with a missing data flag) on the edited copy of the logging curve. Logging data that are invalid because of environmental conditions (such as hole washouts or gas in the drilling mud) should also be
deleted. This will result in data gaps, but these are preferable to erroneous data. If the data gaps occur within potential reservoirs, a replacement value of all reservoir parameters (such as porosity and water saturation) should be estimated from other logging measurements if at all possible. Logging data can also be incorrect due to incorrectly calibrated or malfunctioning logging tools (see Lang, 1980; Neinast and Knox, 1973; Patchett and Coalson, 1979).
6a. Depth shift all log curves not recorded with the base curve or log. When logging tools are run in sequence, differences always occur in depth from tool to tool and from run to run. Even when the logging tools are run in a single string there are potential depth differences due to differential cable stretch. Stretch can be pronounced when the logging tool string sticks or temporally hangs up in the hole. All logging measurements must be adjusted to a common depth reference before data processing can continue. A depth shift of 3 ft can destroy an otherwise good correlation among logging measurements or between well logs and cores.
All depths should be referenced to what is termed a base log. The base log is selected from a logging tool where strong or forceful tool positioning is not used. Free-moving tools travel through the borehole more smoothly than tools that are pushed with great force against the borehole wall, such as the density log. For this reason, strongly centralized tools are not selected as the base log. A resistivity log (induction or laterolog) is usually selected as the base log. For example, if gamma ray logs are available from both the density tool and induction tool strings, it is wise to select the gamma ray from the induction tool as the base log. The gamma ray from the density curve and all curves recorded with the density are then shifted to match the induction log depths. The base curve should also be selected based upon its expected strong correlation with the curves to be depth matched.
Depth-shifting programs are commonly of two types: (a) automatic depth-shifting programs in which mathematical correlations are made among curves from different tool strings and the shifting is accomplished without user input, or (b) visual correlation programs in which the curves to be shifted are laid beside or on top of the base curve, allowing the user to instruct the program by noting correlative points on each log and calculating the depth offset. With older programs, the correlations can be made by using log prints and the shifts input to the screen or a file. Most programs allow the user to carry or cause the same shift to be performed on other curves recorded on the same tool.
The depth-shifting operation necessarily stretches or shrinks the curve being shifted, thus it should be kept in mind that data are both created and lost in the process. For this reason, subsequent depth shifts (corrections) should start with the original raw logs, not with a previously depthshifted copy.
267
168
PART 4— WIRELINE METHODS
Table 1. Summary of Common Log Environmental Corrections
Measurement Gamma ray
Chart3
H: GEN-7 S: POR-7 W: 3-3
Effects
Mud weight, hole size, and tool position
Guard logs
Spherically focused
Induction
Microlaterolog microspherically focused proximity microguard
Compensated neutron logs
S: HR-8 S: RC0R-2A W: 6-6
S: RCOR-1 W: 6-6
H: R-1E S: RC0R-4A W: 6-2
H: R-11 S: RXO-3 W: 4-1
H: P0R-14C S: P0R-14C W: 5-13
Density logs
S: P0R-15A W: 5-11
Mud resistivity, hole size, and tool position
Mud resistivity, hole size, and tool position
Mud resistivity, hole size, and tool position
Mudcake thickness, Rxo zone resistivity, mudcake resistivity, and hole size
Hole size, temperature, pressure, borehole, and formation salinity; mud weight, mud type, mudcake thickness, and tool position
Hole size, formation density, and mud weight
aH = Halliburton Logging Services, S = Schlumberger Well Services, W = Western Atlas Wireline Services.
Comments Becomes significant as hole
size and mud weight increase; charts also available for potassium, uranium, and thorium curves.
—
—
—
Mudcake thickness is difficult to determine
Charts tend to overcorrect in elliptical boreholes; some or all corrections may have been applied at the time of recording.
Charts tend to overcorrect in elliptical boreholes
6b. Depth shift core data to logs. The depth correlation of log data to core data is frequently characterized by numerous abrupt changes in the amount to be shifted. Every trip with the core barrel is potentially a change in the relative core or log depth, even if continuous cores are taken. Zones flagged as lost core zones often are not where they were interpreted to be. Because of this, a u t o m a t i c d e p t h shift p r o c e d u r e s generally do not work when shifting core data to well log data. An overlay procedure is recommended where the core is segmented by core run and again where missing data occurs within a core. The core is then usually shifted by segments. Segments can separate or overlap. Separation is caused by incomplete or poor core recovery, and overlap can be caused by poor core-handling procedures. Review of the field core description can help clarify some of these problems. (For more information on cores, see the chapters on "Conventional Coring" and "Core Handling" in Part 3 and the chapter on "Core Description" in Part 5.)
A core gamma ray can be a valuable aid in establishing depth correlations between core and logs. Boyle's law core porosity and core grain density can be used to construct a core bulk density curve to correlate with log bulk density to determine the amount of depth shifts required. Core bulk density usually correlates well with the density log because lithologic variations are eliminated, resulting in two similar curves being correlated.
While interpolation is a necessary step in the depth matching of wireline logs, it is highly undesirable w h e n shifting cores. Interpolation should not be done when a core segment is shifted. Also, core data should not be resampled if
least squares correlations are planned for calibrating logs or developing porosity and permeability relationships. Linear r e s a m p l i n g of p e r m e a b i l i t y d e s t r o y s p o r o s i t y and permeability relationships and can make statistical inference incorrect when making core to core or log to core data c o r r e l a t i o n s . It is r e c o m m e n d e d that in a n y of t h e s e correlations the logs be resampled, not the core data.
7. Perform environmental corrections on logs. Well logs are recorded in the hostile borehole environment where borehole size, temperature, pressure, mud properties, and other environmental factors affect logging tool responses. Logging tools are calibrated to make correct measurements only when specific environmental conditions exist (that is, an 8-in. borehole diameter at standard temperature). The purpose of environmental corrections is to correct the logging measurements to these standard conditions. Environmental corrections can be large. Some logging tools work over a much broader range of environmental conditions than others (see Table 1 or service company charts for more details).
The logging service company chart books contain correction charts for most logging tools and instructions for their use. Users of well log data should become familiar with these so that they can either apply the corrections or recognize the environmental conditions where corrections are not significant. Table 1 gives a s u m m a r y of e n v i r o n m e n t a l corrections for common logging tools emphasizing the conditions for which corrections can be large. Table 1 also contains a partial list of recent service company charts used for typical corrections (Welex, 1985; Schlumberger
Start here
i
Correlate logs
i
Read R]LD, Я ) Ш , R s f l in zone of interest
i
Correct H s f l for borehole effects (Chart 1)
i
Correct Я|1М for borehole effects (Chart 2)
I
Correct Я|1М for shoulder effects (Chart 3)
I
Correct RILD for borehole effects (Chart 4)
i
Correct Я ! Ш for shoulder effects (Chart 5)
i
Compute ratios using corrected log values ^SFL//=?ILD' ^ILM^ILD
i
Enter tornado chart and read fVrILD' Rxo/Rv d\ (Chart 6 )
I
Determine R and RVN from the ratios
Preprocessing of Logging Data
169
Data required:
• Logs: • Other:
Dual induction Borehole size Mud resistivity Formation temperature Shoulder bed resistivity Bed thickness Standoff size
Charts required:
Chart 1 Chart 2 Chart 3 Chart 4 Chart 5 Chart 6
Halliburton
FLCOR"3
OR "1
FLCOR"4
FLCOR"6
FLCOR"4
FLCOR"5
flINT"2
Schlumberger
6-6 6-5 — 6-4
6~8-9 6-12,13
Atlas
R"8 R"1E R"2M R"2E R"3A
R-5a
Assumptions: • Step profile of invasion • 8-in. borehole
Figure 1. Determination of true resistivity (Rt) from dual induction.
Educational Services, 1989a, b; Western Atlas International, 1985). Older tools may require the use of older chart books. Slim hole models available for some tools may require their own specific charts. The user must become familiar with log headings to identify tool models so that the correct chart can be selected.
Environmental corrections are often performed serially. For example, Figure 1 is a flowchart illustrating the steps used to determine true resistivity from a dual induction log. It should be noted that shoulder bed effect corrections are
difficult and require complex algorithms (Dyos, 1987). Also, it is possible that not performing all corrections (e.g., invasion correction without borehole corrections) may be worse than making no corrections at all.
The SP curve presents special environmental problems that are not addressed by charts. Since the SP responds in part to the mud filtrate resistivity, no two SP curves will be recorded under exactly the same environment (Bateman, 1985). (For more information on SP logs, see the chapters on "Basic O p e n Hole Tools" and "Determination of Water Resistivity" in Part 4.)
Determination of Water Resistivity
Roberto Peveraro
BP Exploration Houston, Texas, U.S.A.
INTRODUCTION
The most direct way of finding water resistivity (Rw) is to obtain a s a m p l e of f o r m a t i o n w a t e r a n d m e a s u r e its resistivity. However, this is seldom possible in practice, as f o r m a t i o n w a t e r s a m p l e s , if a v a i l a b l e , are i n v a r i a b l y contaminated by mud filtrate. Rw is therefore usually calculated, and there are three methods available for this purpose:
1. SPmethod 2. Archieequation 3. Resistivity ratio method
SP METHOD
Principle
The first step in the interpretation of the SP log is the establishment of "sand" and "shale" lines, as shown in Figure 1. These are arbitrary limits, with the sand lines normally representing the maximum deflection to the left and shale lines representing the maximum deflection to the right (in shales). Deflections to the left of the shale line are regarded as normal or negative, and correspond to porous and permeable zones containing a more saline interstitial water than the drilling mud (Rw < Rm().
If the mud is more saline than the formation water, the SP currents flow in the opposite direction and the corresponding deflection will be to the right of the shale line. Such a deflection is considered to be reversed or positive (Figure 2). If there is no salinity contrast between the m u d and the formation water, no SP currents are generated and no deflection will be observed (SP = 0).
The magnitude of deflection from the shale base line to the maximum deflection developed in a thick, clean, waterbearing sand is referred to as the static SP, or SSP.
Knowledge of the SSP is essential for the derivation of Rw, which is required for the calculation of water saturation in the uninvaded zone. The equation that relates the SSP to measurable quantities, namely, Rw and Rmf, introduces values that are linearly related to their respective chemical activities. These values are referred to as equivalent resistivities and are denoted by Rwe and Kmfe. Thus, the standard equation that relates the SSP to the mud filtrate and uninvaded formation water resistivities is
Equation (1) is used in the quantitative interpretation of the SP log.
Procedure
The procedure to determine Kw from the SP is outlined as follows. The data required include the SP, the invaded zone resistivity, the formation temperature, and the mud and mud filtrate resistivities at the given formation temperature.
1. After selecting on the log the zone in which water resistivity will be determined, the formation temperature is established and the values of mud and mud filtrate resistivity are corrected accordingly.
2. The shale and sand lines are established, the bed boundaries on the SP curve are determined, and maximum SP inflection for that permeable bed are read off the log.
SSP = -K log K1mfe
(1)
Kn
where K is a constant, the value of which is dependent on the formation temperature. Usually,
K = 61 + 0.133T (0F) or 65 + 0.24T (0C)
(2)
Figure 1. Example of a normal SP deflection. (Courtesy of Schlumberger Educational Services, 1987.)
170
Determination of Water Resistivity
171
SP J20. +
Depth (ft)
0
0
—^
8100
Resistivity 50
500
,I 1 i
J—t-70mv I ^
4
5HA
/
/
8150 — г
I )
\
?
I \
)
Figure 2. Example of a reverse SP deflection. (From Dewan, 1983.)
3. If the bed was too thin for the SP to develop fully (beds thinner than 20 ft), the appropriate correction is applied from the charts.
4. Charts supplied by logging companies permit the determination of R m f e / R w e from the SP and, from that, Kwe-
5. Finally, the true value of Rw is derived from its equivalent Rwe value, again using the appropriate charts.
An example of calculating Rw by using the SP method is as follows. Figure 3 shows the log used in this example. The following charts are required:
Chart 1 Chart 2 Chart 3 Chart 4 Chart 5
Schlumberger (1989)
GEN-6 GEN-9 SP-3 SP-1 SP-2
Given: Rm = 2.5 Q-m at 70°F Rmf = 2.0 Q-m at 70°F Hole diameter = 8 in.
Western Atlas (1985)
1-1 • 1-5
2-1 2-2 2-3
Figure 3. Log used in example of calculation of water resistivity from the SP method. (From Western Atlas International, 1985; courtesy of Atlas Wireline Division of Western Atlas.)
Surface temperature = 60 T BHT = 164°F at 10,500 ft
From the log: SP deflection = 70 mV Bed thickness = 24 ft Short normal resistivity = 65 Q-m
Calculations: 1. Bed is from 8114 to 8138 ft. 2. From Chart 1, using BHT = 164°F at 10,500 ft and
surface temperature = 60°F: formation temperature Tf at 8100 ft = 140°F. 3. From Chart 2, using Rm = 2.5 Q-m at 70°F and Kmf = 2.0 Q-m at 70°F: Rm = 1.3 Q-m at 140°F and Rmf = 1.0 Q-m at 140°F. 4. Shale baseline is two divisions left of depth column. 5. Bed thickness is 8138 to 8114 ft = 24 ft. 6. SP deflection = - 7 0 mV. 7. Use Chart 3 to correct the SP. Use short normal resistivity, .Rsn, for invaded-zone resistivity, R1. This gives R[/Rm = R-SN^m = 65/1.3 = 50. ThisSP correction factor is 1.07, giving a corrected SP of 1.07 x -70 = -75 mV. 8. From Chart 4, using SP corrected and Tf: Rmf/Rwe = 9.7. 9. R w e = Rmf/(Rmf/Rwe) = 1.0/9.7 = 0.103 Q-m. 10. From Chart 5, using Rwe and Tf: Rw = 0.103 Q-m at 140°F (no change).
172
PART 4— WIRELINE METHODS
Assumptions and Limitations
It is assumed that all salts in solution in the fluids are NaCl or equivalent and that the beds are clay free. This method does not work for oil-based muds. Also, it does not give correct estimations of Rw in hydrocarbon-bearing zones.
ARCHIE EQUATION
Principle
If a clean, water-bearing zone is present or can be assumed, Rw can be calculated by using the Archie equation. The zone or permeable bed in which water resistivity is determined is selected on the log. This zone must be 100% water saturated and must not contain any clay or shale. The bed must be thick so that the deep investigation resistivity device is not affected by shoulder beds.
The water saturation equation in general form is
a Rw
(3)
Assumptions and Limitations
It is assumed that there are no clay or conductive minerals in the water-bearing zone, that the zone is clean, and that the invasion is shallow enough for resistivity tools to be unaffected by it. This method only works in clean, waterbearing reservoirs. It is typically unreliable in highly fractured or vuggy reservoirs.
RESISTIVITY RATIO METHOD
Principle
In a clean, water-bearing zone, the following relationships can be obtained from application of the Archie equation in the uninvaded and invaded zones:
a Rv
*t
(5)
Г Sv
Rvn — aR mf
(6)
Г Sx
Let's assume for a, m, and n the usual initial values of 1,2, and 2, respectively (although other values for a, m, and n may be
more appropriate). We have
In
such
a
zone, '
SWw
=
SXvrOi
=
1
and
a
and
ф T
are
the
same
in
both
the flushed and uninvaded zones.
Dividing Equation (5) by Equation (6) gives
^t _ *w
(7)
^xo ^mf
Therefore,
tfw = S w 2 Ф 2 R{
(4)
In a water zone, S w = 1, thus Equation (4) becomes
Rw=CP2Rt
(4a)
P o r o s i t y (ф) can be d e r i v e d f r o m logs. (For m o r e information on methods to derive porosity and the Archie equation, see the chapter on "Standard Interpretation" in Part 4. For more on deriving Rt, see the chapter on "Preprocessing of Logging Data," also in Part 4.)
Procedure
The procedure to determine Rw from the Archie Equation is outlined as follows. The data that are required include Rt (or Rq) f r o m deep-reading resistivity tools such as d e e p induction (ILd) or deep laterolog (LLd) and porosity log(s).
1. After correlating logs for depth mismatch, a clean, water-bearing zone is located. If using an induction log, the minimum thickness should be 15 ft. For a laterolog, the minimum thickness should be 4 ft.
2. Read Rq (= Rt) from the resistivity log, and determine the porosity from logs or otherwise.
3. Calculate Rw from Equation (4a).
or
^W - ^mf
(8)
AX0
Rt and Rxo are obtained from deep and shallow resistivity logs, and Rmf is at formation temperature.
Procedure
The procedure for determining Rw from the resistivity ratio method is outlined as follows. The data that are required include Rt from ILd or LLd, Rxo from SFL or another shallow-reading resistivity device, Rmf, and Tf.
1. After correlating logs for depth mismatch, locate a clean, water-bearing zone. If using an induction log, the minimum thickness should be 15 ft. For a laterolog, the minimum thickness should be 4 ft.
2. Read Rt and Rxo. 3. Correct Rmf to the appropriate formation temperature
(Tf) value. 4. Calculate Rw from Equation (8).
Note that an equivalent "quick look" solution can be obtained by an overlay technique using logarithmically scaled Rt and Rxo logs. If one log is placed over the other and shifted so that the two curves coincide in clean, water-bearing zones, the 1 ohm-m line on the Rxo log grid will lie on the Rt grid at a value equal to R w / R m f . Since Rmf is known, Rw can then be
Determination of Water Resistivity
173
RMLL
RLL
л 1 I
10
100
1000
RMLL4C^R
<
^/RLL
/"
\
/ <^
^
S
<
^^
</."
J I
J—
1
RiiiT 045 1
/
.j
RMLL
10
100
10 rLL
100
I
I
I
I
RMLL
I Shiftec
I
Si I
CF
I
I
I
I I
I I
I
I I
I
I
I
I
S I
I
I I
I
I
1000 1000
I I
I I I I
I I
I I I I I I I
Figure 4. Example problem showing determination of Rvil from resistivity ratio.
easily calculated. Figure 4 shows an example of original and shifted logs indicating a Rt/Rmf ratio of 0.45.
Limitations
This method is unreliable in fractured or vuggy reservoirs. It requires a reliable measurement of Rxo. Results can be questionable if the m u d resistivity properties have varied significantly after drilling the zone to be analyzed.
CONCLUDING REMARKS
It is quite common and good practice to determine Rw w i t h m o r e t h a n one m e t h o d , if possible, to check for consistency and reliability. Variations may occur in the Rw values determined by the three methods outlined here. If this is the case, then the values obtained can be averaged. Once determined, Rw is assumed to be constant throughout the formation being evaluated, unless local knowledge or clear SP log amplitude changes indicate otherwise.
Quick-Look Lithology from Logs
N. J. Hancock
Consultant Los Alamitos, California, U.S.A.
INTRODUCTION
Lithological determination from wireline logs is often done by sophisticated computer programs, but basic quick-look i n t e r p r e t a t i o n can be m a d e by visual inspection of appropriate logs.
The best logs for lithological purposes are those that are (1) most influenced by rock properties and (2) least influenced by fluid properties. The most useful of the commonly available logs are
• Gamma ray • Spontaneous potential (SP) • CaHper • Formation density • Photoelectric absorption • Neutron porosity
(For more details on these logs, see the chapter on "Basic Open Hole Tools" in Part 4. Also, the chapter on "Difficult Lithologies" in Part 4 covers logging tool response in sedimentary minerals.)
Borehole imaging tools such as the Formation MicroScanner1 are invaluable for detailed purposes, including bedding character and sedimentary structures, but are much less commonly available. (For more details, see the chapter on "Borehole Imaging Devices" in Part 4.)
GAMMA RAY LOGS
The common radioactive elements—potassium, thorium, and uranium—are normally insignificant in reservoir fluids, whereas they are important components of the rock system, especially of clay minerals. Gamma ray logs are therefore a good indicator of mineralogy.
Lithological Responses
The principal gamma ray responses are as follows:
is often thought of as a grain size profile. Three basic log shapes are recognized: funnel (coarsening upward), cylinder (blocky), and bell (fining upward) (Figure 1). These three shapes can be subdivided into smooth (relatively homogeneous) or serrate (with interbedded thin shales).
Log shapes typically reflect changing depositional energy from liigh (clean, coarser sand) to low (shaly, finer sand). An interpretive jump is usually made from depositional energy to depositional process and hence depositional environment. Often this jump is made without seriously considering the intermediate steps. This can be dangerous. Each of the steps is highly ambiguous and must be augmented by other evidence, such as unit thickness, associated rock types, and overall depositional setting. Typically,
• Funnel shapes imply upward-increasing energy, which may be found in distributary mouth bars, delta lobe fringes, deep sea fans, and other environments.
• Cylinder shapes reflect relatively constant energy levels and can include eolian dunes, low sinuosity distributary channels, and beaches.
• Bell shapes represent waning-current sequences, which can include alluvial point bars, deltaic distributaries, and deep sea fan channels.
In fact, grain size has no effect on gamma ray logs. The log shapes reflect shaliness, that is, clay and mica content of the sand. Because most sands reflect a hydrodynamic equilibrium, clay content does usually correlate (inversely) with grain size. However, in the following examples, clay content and grain size do not correlate, resulting in misleading log shapes:
• Very fine, clean sand above coarser sand may show a cylinder shape.
• Clay clasts concentrated near the base of a channel may give a funnel shape.
• Clay added later due to bioturbation or mechanical infiltration at the top of a gravel may create a bell shape.
Lithology Sandstone (quartz) Limestone Dolomite Shale Organic-rich shale Anhydrite, halite Sylvite (KCI) Coal
Gamma Ray Values (in API units)
15-30 (rarely to 200) 10-40 15-40 (rarely to 200) 60-150 100-250 8-15 350-500 15-150 (any value possible)
Log Shapes
The shape of a gamma ray (or SP) log through a sand body
(For more details on using log shape to interpret depositional environment, see the chapter on "Lithofacies and Environmental Analysis of Clastic Depositional Systems" in Part 6.)
Problems and Exceptions
• Radioactive minerals in sands, especially K-feldspar, zircon, and mica, can raise sand readings as high as adjacent shales. Gamma ray logs may be useless in immature sands derived from basement terranes. However, beach placers rich in zircon may be valuable correlative markers if not mistaken for shale.
• "Hot" dolomite, especially common in the Permian basin
1A trademark of Schlumberger. 174
Qu ick-Look Lithology from Logs
175
SOOQ 5020 5980 6000 6020 8980 9000
... SP GR (API)
CALIPER (IN)
BULK DENSITY (GM/CC)
Figure 1. Characteristic log shapes for different types of sand bodies set in shale, (a) Funnel shape, coarsening upward. Note that this is the shallowest interval, so the shale is least compacted, (b) Cylinder shape, blocky. Note that the SP log is featureless because the borehole salinity is the same as formation salinity, (c) Bell shape, fining upward. Note that coal is present in addition to shale.
in the United States, may have gamma ray values up to 200 API units, resembling shale. • Radioactive (KCl) muds raise the baseline gamma ray zero reading so that apparent values for all rock types are increased, sometimes by about 20 API units. • Evanescent high gamma ray readings in sands, present on one logging run but vanished some weeks later, have been observed especially in steamflood conditions. While remaining enigmatic, these may be due to concentrations of radon in the pore space.
Spectral Gamma Ray Logs
In this enhancement to natural gamma ray logging, the energy levels of incoming gamma rays are counted in a series of energy windows, and an algorithm converts the energy spectrum to count rates for potassium (%), thorium (ppm), and uranium (ppm). Spectral gamma ray logs are most useful in identifying the following:
• Clay minerals. Illite clays are rich in potassium, whereas
smectite and kaolinite contain thorium. The thorium to potassium ratio can distinguish illitic from smectitic shales and so provide a correlation tool. • Organic-rich rocks. In shales, uranium enrichment is usually associated with organic content and can be a tool for identifying oil source beds. Quantitative relationships between uranium and organic content have been reported, but tend to be inconsistent. • Mica sand. Riclily micaceous sands (such as the Rannoch unit of the Brent Sand in the North Sea) appear shaly on gamma ray logs, but can be distinguished because the radiation is all from potassium. • "Hot" dolomite. This type of dolomite can be distinguished from shale because the gamma rays are principally from uranium. The chemical relationship between uranium and the dolomite is unknown. • Natural fractures. Soluble uranium in pore water often precipitates on open fractures, so thin intervals with high uranium count (a "spiky" log) may mark a fractured interval. • Producing zones. As with natural fractures, uranium may precipitate on flowing perforations, so a spectral gamma ray log run after years of production may show which completed intervals are producing and which are not. • Uranium prospecting. Most of the "uranium" signal actually comes from the tenth decay process in the uranium series, the decay of bismuth-214. This is separated in time from the original uranium by halflives in excess of IO9 years, so the relatively soluble uranium may have moved away during the interim even though the log still records its presence.
SPONTANEOUS POTENTIAL (SP) LOGS
Lithological Responses
Shale SP interpretation depends on first recognizing shale, where fairly constant SP readings form a straight "shale baseline" on the log (Figure la). Its actual SP value is not significant.
Sandstone The potential differences around a sand/shale contact deflect the SP from the shale baseline. The deflection is negative for a normal salinity contrast (borehole fresher than formation). Little change occurs within a sand interval, so a clean sand shows a straight-line "sand line" (Figure lc). (For more details on SP shale and sand baselines, see the chapter on "Determination of Water Resistivity" in Part 4.)
Tight Rocks An SP log is of little use in the absence of boundaries between shale beds and permeable beds. In relatively tight rocks (carbonates, evaporites, etc.), the SP wanders aimlessly, with no sharp usable deflections.
Log Shapes
Funnel, cylinder, and bell-shaped motifs resemble those
176
PART 4— WIRELINE METHODS
previously described for gamma ray logs. They are due to the qualitative shaliness indication given by the SP and can therefore be interpreted in a similar way to the gamma ray (except for the following complications).
Salinity Contrast
Contrasting salinity is critical for SP logs. Three scenarios are possible:
• Fresh borehole fluid in a saline formation. Gives "normal" SP.
• Borehole salinity is same as formation. Featureless SP7 very low amplitude, may be a straight line, no obvious relationship to beds (Figure lb).
• Saline borehole in a fresher formation. Gives a reversed SP, where sands show positive deflections from the shale baseline.
Other Problems
In additional to salinity contrasts, other conditions can create problems in interpreting SP logs. For example,
• Baseline shifts. Although the value of the SP shale baseline is not significant, it will shift if formation fluid salinity changes from one bed to another, making the log hard to interpret.
• Manual shifts. On occasion, the logging engineer adjusts the SP log scale to keep it within the track.
• Mud type. Water-based mud (with suitable salinity) is essential. Oil-filled or empty holes have nothing to carry the SP charges.
• Interference. Remanent magnetism within the winching system often ruins SP logs. Look for a sine-form SP whose cycle length is the circumference of the cable drum.
• Hydrocarbons. The SP is generated in water. High hydrocarbon saturation reduces the SP, making sands appear more shaly.
CALIPER LOGS
Property Measured
For lithological purposes, the critical data are caliper readings relative to bit size. There are three scenarios:
Hard, inert rock Soft or brittle rock Permeable rock
Hole in gauge Hole washes out Mudcake builds up
Caliper = bit size Caliper > bit size Caliper < bit size
Well-designed modern mud systems can minimize washouts, making caliper logs less distinctive for lithological purposes.
Lithological Responses
Sandstone
Consolidated sandstone is usually permeable, so expect mudcake to cause a caliper reading that is about 0.5 in. smaller than the bit size. Bed boundaries are often accurately delimited (Figure 1).
Sand Friable, unconsolidated sand may wash out, causing large caliper readings. Look for this problem in young, shallow formations.
Shale Shale frequently spalls into the borehole, especially in the minimum principal stress direction. This leads to elliptical boreholes identifiable with multiple arm calipers, as on a dipmeter.
Coal Medium to high rank coals are often brittle and welljointed. Such joint blocks cave into the borehole (Figure lc) leaving deep washouts as thick as the coal seam (frequently only 1 ft or so). Not all coals behave this way.
Carbonates Carbonates often fail to show mudcake build-up despite good permeability because individual vuggy or moldic pores are too large to trap mud solids. Mudcake builds up on the back walls of such pores, not into the borehole. Sucrosic dolomite is the only carbonate that typically shows mudcake on calipers.
Tight Rocks Tightly cemented beds, such as ironstones, siltstones, and carbonate concretions in sandstones, are hard, inert rocks that remain in gauge.
Anhydrite and Gypsum Anhydrite and g y p s u m frequently remain in gauge if pure, but shaly intervals may be washed out.
Halite and Potash Salts Salt-saturated or oil-based muds may maintain the hole in gauge, but dilute water-based muds result in severe dissolution leading to huge, unoriented washouts.
FORMATION DENSITY LOGS (ALONE)
Property Measured
Measured density is the sum of the rock system density and the pore fluid system density. Density values can therefore be used directly to identify lithology only when the porosity is insignificant. In porous rocks, density must be interpreted in combination with neutron or other porosity logs.
Lithological Responses (Nonporous rocks)
Evaporites Individual evaporitic minerals (such as anhydrite, halite, sylvite, and carnallite) have well-defined densities and generate straight-line density logs with little variation (Figure 2).
GR (API)
NEUTRON POROSITY (%) BULK DENSITY (GM/CC)
. -IS 2.95
Figure 2. Characteristic log signatures for a carbonate and evaporite sequence. Hole conditions are good.
Qu ick-Look Lithology from Logs
177
Lithological Responses
Sandstone Quartz should read 1.7 to 1.8 barns/electron, but most other minerals can raise the value substantially. Because they are usually present, the log is of limited value.
Limestone Clean limestone reads about 5.0 barns/electron (Figure 2).
Dolomite Dolomite should read about 3.0 barns/electron, providing an easy way to distinguish limestone from dolomite (Figure 2) even if gas is present. Note that iron in ferroan dolomite increases readings to resemble limestone.
Shale "Average" shale reads 3-3.5 barns/electron, but values up to 7 or 8 barns/electron can be obtained depending on iron content and accessory minerals. This large range makes the log of limited value.
Coal Coals are variable but always significantly lighter than 2 g/cm3. Thin beds give a pronounced density spike, but may not resolve a true density reading (Figure lc). Note that deep washouts also give low-density spikes.
Ironstone Concentrations of iron minerals such as pyrite and siderite give high densities, often in thin beds, contrasting with surrounding rocks.
Shale Densities of shales vary between 2.2 and 2.65 g / c m 3 or more, increasing with compaction induced by age and depth of burial (Figure 1). Overpressured shales, in which some of the overburden load is borne by pore fluid, are undercompacted and have low densities relative to normally pressured shales at similar depths.
PHOTOELECTRIC ABSORPTION (Fe) LOGS
Property Measured
Photoelectric absorption (Pe), measured by the newer formation density tools, is related to atomic number Z, raised to the 3.6 power (Z3-6). Consequently, very light components (pore fluids) have negligible effect, making the log good for lithology. Unfortunately, heavy elements have an enormous effect. Thus, a few percent of iron masks basic lithological differences, and barite (usually with m u d weights over 10 ppg) makes the log unusable.
NEUTRON POROSITY LOGS (ALONE)
Property Measured
Compensated neutron porosity is primarily the combined h y d r o g e n content of the rock system and the pore fluid system. Lithology can therefore be interpreted directly from neutron values only when porosity is insignificant. In porous rocks, the neutron log must be interpreted in combination with other logs such as formation density.
Lithological Responses (Nonporous Rocks)
Water of Crystallization (Evaporites)
• Gypsum and anhydrite. The typical neutron porosity value in anhydrite (CaSO4) is close to zero, but that in gypsum (CaSO4 • 2H20) is much higher—up to 60%.
• Potash Evaporites. Sylvite is anhydrous with a near-zero neutron porosity, but carnallite (KMgCl3 • 6H20) gives neutron values of 30% to 60%.
Bound Water in Shale
Some water in shales is chemically bound to clay minerals, whereas some occurs in micropores. Both types raise neutron log readings but represent no effective porosity (Figure 1). Shales consequently have high apparent neutron porosity, but values vary among formations. Often 40% is a good shale cutoff limit, but shale values can be as low as 30%. A local cut-off can often be established by calibration, such as from cores.
NEUTRON AND DENSITY LOGS COMBINED
Neutron and density logs each react to both lithology and porosity, so by analyzing the two logs together, one can begin to distinguish lithology from porosity. Neutron and density logs, together with a caliper measurement recorded by the
178
PART 4— WIRELINE METHODS
density tool and a natural gamma ray log, are commonly run as a combination. This is the most powerful of the commonly available log suites for general purpose determination of lithology.
Crossplotting
Logging company chart books all include neutron-density crossplots that are easy to use for clean (nonshaly) reservoir rocks. The plots are entered with a bulk density and an apparent neutron porosity (should be environmentally corrected, but the corrections are usually negligible). A rock type (sandstone, limestone, or dolomite) and a corrected porosity can be read from the crossplot.
Overlay Presentation
Manual crossplotting is tedious. A much faster way to visualize rock type is directly from the overlay presentation in which both neutron and density logs are superimposed in the same log track. To do this, a compatible scale must be used so that the porosity components of both logs exactly overlay. Then any offset (or residual) between the two logs is attributable to lithology or to the presence of gas.
Both tools are generally calibrated in limestone units, so the compatible scale is defined for freshwater-limestone systems, with theoretical limits as follows:
Neutron (p.u.) Density (g/cm3)
All Porosity (H2O) 100 LO
No Porosity (CaCO3) O
271
In practice, porosities over 50% are seldom needed, whereas rocks with densities over 2.71 g / c m 3 are common.
Thus, with slight rounding, the usual compatible scale is
Neutron (p.u.)
45
30
15
0
-15
Density (g/cm3) 1.95 2.20 2.45 2.70 2.95
In high porosity areas with no dolomite, the scale is often slid across to the following range:
Neutron (p.u.)
60
45
30
15
0
Density (g/cm3) 1.70 1.95 2.20 2.45 2.70
On these scales, any offset of neutron and density logs is maintained regardless of porosity. Offsets are due to rock differences in density and neutron-absorbing properties (capture cross section). Ideal relationships for the three main liquid-filled porous rocks are as follows:
Sandstone • Density displaced 0.05 g / c m 3 to the left. • Neutron displaced about 3 p.u. (porosity units) to the
right. • Cross-over is two small-scale divisions on the usual log
grid.
Limestone • Density and neutron overlay exactly.
Dolomite • Density displaced 0.175 g / c m 3 to the right. • Neutron displaced 4-8 p.u. to the left. • Separation is four to six small-scale divisions on the
usual log grid.
Other noncompatible scales are harder to interpret. One is the sandstone scale: the zero neutron reading is aligned with 2.65 g/cm3. Also, the neutron log may, or may not, be calibrated in sandstone units, reducing cross-over in sandstone by about two, or one, scale divisions, respectively.
If the two scales do not have the same a m p l i t u d e (60 neutron porosity units corresponding to a range of 1 g/cm3), lithological interpretation should not be attempted from the overlay plot because log separations then become a function of porosity as well as lithology.
Lithological Responses
Sandstone (Oil or Water Filled)
Clean quartz sandstones give the typical two-division neutron-density cross-over with density to the left of neutron (Figure 1). The a d d i t i o n of s o m e clay ( f o r m i n g shaly sandstone) increases the neutron reading, reducing log crossover or even reversing it to create separation. Check natural gamma ray for evidence of increasing clay.
Heavier components such as mica increase the density, reducing log cross-over or even reversing it to create separation. Check spectral gamma ray to distinguish the following:
• Mica: potassium radiation only. • Zircon (with other heavy minerals): thorium or
uranium radiation. • Siderite, pyrite, etc.: no increased radiation.
Use the shape of the neutron-density cross-over to provide depositional energy in the same way as an SP or gamma ray log (Figure 1). Thus, a "V" shape is a funnel (coarsening upward) and a "A" shape is a bell (fining upward).
Sandstone (Gas-Filled)
Compared to oil- or water-filled sandstone, the neutron log for a gas-filled sandstone reads as much as 10-15 porosity units too low, and the density log may read about 0.05 g / c m 3 too low. Together these effects increase the log cross-over from two to about five scale divisions.
Sandstone (Air-Filled)
Nonhydrocarbon gas in sandstone can give neutron readings close to zero, depending on residual water and humidity in the pore space. Enormous log cross-over results.
Limestone
Clean limestone has no neutron-density separation (Figure 2). When the neutron drifts to higher values, expect the presence of clay. Check the natural gamma ray. In gas-filled limestone, expect cross-over like that described for sandstone, and use a Pe value of 5 to confirm limestone.
Dolomite
Characteristic four to six scale division separation with density to the right of neutron is relatively consistent in clean dolomite (Figure 2). Gas reduces or eliminates the separation; use a Pe value of 3 to confirm dolomite. Locally high natural gamma ray looks like clay, but if neutron-density separation is unchanged, it may be "hot" dolomite (especially in the Permian basin). Check uranium if spectral gamma ray is available.
Shale
Shale shows a log separation with neutron to the left of density, sometimes displaced by a large amount (Figure 1). At times the separation is only three or four scale divisions, which can resemble dolomite. To distinguish shale, check for the following:
• Apparent neutron porosity is too high for the area. Shale neutron readings are often between 30 and 50 porosity units.
• Caliper log shows washouts. • Natural gamma ray is high; consistently high in beds
where neutron is high. If spectral gamma ray is
Qu ick-Look Lithology from Logs
179
available, look for all radioactive elements elevated (contrast only uranium high in "hot" dolomite).
Coal
Neutron and density logs for coal both read similar very high apparent porosities (Figure lc). Coals give prominent deflections that do not resemble anything but severe washouts. (Diatomite has a density of about 1.4 g / c m 3 and a neutron measurement of about 60 porosity units, so crossover is at least seven scale divisions.)
Complex Rock Mixtures
Using neutron and density logs to resolve porosity and lithology allows only a "one-dimensional" view of lithology. Rock mixtures always create ambiguities for this simple quick-look interpretation. Local knowledge of rock types and mixtures to be expected and not to be expected may eliminate ambiguity (for example, do not look for dolomite and evaporites in a temperate, humid delta). Rock sample and mudlog data are invaluable. For complex rock mixtures, more input log data are needed, and computer-processed multidimensional crossplots must be used to determine lithology. In any case, confidence is always increased by using more input data.
Standard Interpretation
MarkAiberty
BP Exploration Honston, Texas, U.S.A.
INTRODUCTION
Standard interpretation is the process of determining volumes of hydrocarbons in place from wireline logs. This process requires four basic steps:
1. Determine the volume of shale. Shaleaffectsthe response of the various logging devices. To interpret the response for porosity or saturation, the volume of shale must be determined.
2. Determine the porosity. Porosity is the fraction of the total rock available for the storage of fluids.
3. Determine the formation water resistivity (Rw). The resistivity of the water (without hydrocarbons) is used to interpret the formation resistivity for saturation.
4. Determine the water saturation (Sw). A resistivity model is interpreted for saturation. This model relates water saturation, porosity, water resistivity, and volume shale.
ESTIMATION OF SHALE VOLUME
The volume of shale (Vsh) is best estimated by logging measurements that respond primarily to shale, in particular, gamma ray and spontaneous potential (SP). The most common methods for estimating shale volumes from gamma ray and SP logs are outlined here. Other measurements can be used under special conditions to estimate shale volumes, such as the resistivity in very high resistivity formations, the compensated neutron in very low porosity formations, or density versus neutron crossplots in known lithologies.
Shale Volume from Gamma Ray
Application
The gamma ray can be an excellent estimator of shale volume in areas with little uranium and where radioactive salts are associated primarily with clay minerals.
Method
The two most frequently used methods of estimating shale volume from the gamma ray log are the linear and the nonlinear estimators. Both methods require determining the g a m m a ray r e s p o n s e (GR) at the d e p t h of interest and determining the response associated with a clean reservoir having no shale (GRcl) and a zone of 100% shale (GRsh).
The linear method is the simplest, but it tends to o v e r e s t i m a t e shale in the i n t e r m e d i a t e r a n g e s of shale volumes. The Unear response equation is
у _ GRsh-GR sh G R s h - G R c l
The nonlinear method begins by estimating the shale volume from the linear equation and then correcting that estimation using a chart as shown in Figure 1. First, one marks the horizontal axis (labeled "radioactivity index") with the linear estimation of shale volume. One then proceeds vertically to the appropriate rocks and then horizontally to read the corrected volume of shale.
Advantages The gamma ray method is very simple, fast, and generally the most reliable. It can be used with the potassium or thorium curves and with the uranium corrected total gamma ray curve from the spectral gamma ray.
Limitations Gamma ray readings must be corrected for hole size first. This method does not work well in areas where radioactivity is not primarily associated with the clays, such as in feldspathic sands.
Shale Volume from Spontaneous Potential (SP)
Application The SP can be an fair estimator of shale volume in areas where mud filtrate and formation water resistivities contrast.
Method The estimation of v o l u m e shale f r o m the SP requires determining the SP response at the depth of interest and determining the response associated with a clean reservoir with no shale (SPcl) and a zone of 100% shale (SPsh). The response equation is
y _ SPsh-SP sh SPsh-SPcl
Advantages
The SP method is very simple and fast.
Limitations This method does not have good vertical resolution. It overestimates shale volume in hydrocarbon-bearing zones, and it is ensitive to the selection of clean reservoir and shale points. It will not work in zones where Rw ~ Rmf or in oilbased muds.
POROSITY ESTIMATION
Any logging device primarily affected by the presence of porosity can be used to estimate porosity (ф). Best results are
180
Standard Interpretation
181
I
09 -
//tj
0.8 -
/ ///
0.7 -
S ///
0.6 -
0.5 0.4 0.3 0.2 -
Linear Scaling
t^r
// /
///
Larionov (Older rocks)—
•' y j *
Stieber
Clavier, et al
••
— Larionov (Tertiary rocks)
0.1 -
0
'
О
0.1
I i i i 1 1— 1 I
0.2 0.3 0.4
0.5 0.6
0.7
0.8
0.9
1
Figure 1. Nonlinear shale volume chart. (From Western Atlas International, 1985; courtesy of Atlas Wireline Services Division of Western Atlas International, Inc.)
generally produced by the density, compensated neutron, and sonic logs. When using a single porosity measuring device such as the density, neutron, or sonic, assumptions are normally required to estimate porosity. These assumptions are the lithology and the fluid properties, which must be d e t e r m i n e d f r o m local knowledge. If multiple porosity measuring devices are recorded, the multiple measurements can frequently be used together to determine porosity and lithology. (Determination of lithology is discussed in the chapter on "Lithology from Logs" in Part 4.)
Porosity from Density
Application
Density is a good method for determining either total or effective porosity in single or multiple mineral, fluid-filled reservoirs. It is extremely useful in combination with core measurements of grain density.
Method
The m e t h o d of estimating porosity f r o m the density requires determining the matrix density (pfl), the density log reading (pb), and the fluid density (pfl at the depth of interest. The matrix density is determined by the lithology. Normally, sandstone is 2.65 g / c m 3 , limestone is 2.71 g / c m 3 , and dolomite is 2.87 g/cm3. The fluid density is dependent upon t h e salinity of w a t e r a n d the d e n s i t y of h y d r o c a r b o n . F r e s h w a t e r has a d e n s i t y of 1.0 g / c m 3 a n d s a l t w a t e r approximately 1.1 g / c m 3 . Hydrocarbon density can vary widely from 0.05 g / c m 3 for gas at low pressures to nearly 1.0 g / c m 3 for certain oils. A typical value for oil is 0.8 g/cm3.
The density response equation is
Advantages
This method can be used to determine porosity accurately when matrix and fluid densities are known. It can be used to determine either total or effective porosity through the selection of appropriate matrix density. Matrix density can be accurately determined from core measurements. Linear volumetric response allows expansion for multiple mineral formations.
Limitations
The shallow investigation of the density normally results in investigating the flushed zone; one must typically use flush zone fluid properties in its interpretation. This method is difficult to use accurately in gas reservoirs due to the difficulty of determining the appropriate fluid density of the flushed zone. One should use apparent matrix and fluid densities adjusted to the relationship of the ratio of twice the atomic number divided by the atomic weight (2Z/A) for freshwater-filled limestone. This method only investigates one side of the borehole (typically the low side).
Porosity from Compensated Neutron
Application
The compensated neutron measurement gives a rough estimate of the total porosity in simple fluid-filled reservoirs. It is very sensitive to gas, and consequently, it is useful as a gas indicator in conjunction with the density measurement. It can also be used with density to determine both lithology and porosity in non-gas bearing zones. It is sensitive to clay content.
Method
The compensated neutron measurement is recorded at the wellsite to provide the correct water-filled porosity in a userspecified matrix, n o r m a l l y l i m e s t o n e or s a n d s t o n e . If recorded on the appropriate matrix, the log values can be corrected for environmental effects and then used as an estimate of water-filled porosity. If the zone of interest contains fluids with lower hydrogen indexes than water, the porosity will be underestimated.
Advantages
A very strong response to gas is useful in identifying gas reservoirs (indicates low porosity) when used in conjunction with density or sonic porosity measurements. A neutron and density combination is very useful in identifying mineralogy. The most practical porosity measurement is through the casing.
Limitations
Since shallow investigation normally results in investigating the flushed zone, one must typically use flush zone fluid properties in interpretation. This method investigates omnidirectionally around the borehole and typically requires considerable environmental corrections. It is sensitive to standoff from the borehole wall and needs effective decentralization for reasonably accurate results.
182
PART 4— WIRELINE METHODS
Porosity from Sonic
Application
The compressional sonic log can be used as a moderately good estimator of porosity in water and, to some degree, oilfilled rocks.
Method
The two most frequently used m e t h o d s of estimating porosity from sonic measurements are the Wyllie time average method and the Raymer-Hunt-Gardner field observation method. Both methods require determining the sonic response (Ati) at the depth of interest and determining the response associated with the matrix (Afma). Typical values for Afma are 55.5 |isec/ft for sandstone, 47.5 |isec/ft for limestone, and 43.5 |isec/ft for dolomite. (A more complete listing of interval traveltimes for common minerals is given in the chapter on "Difficult Lithologies" in Part 4.)
The Wyllie time average method requires an estimate of fluid interval transit time (Affl). Typical values range from 189 |a,sec/ft for saltwater to 204 [isec/ft for freshwater. This method tends to overestimate porosity in uncompacted sandstones and hydrocarbon-bearing reservoirs. Empirical corrections to lessen this error can be implemented using the compaction factor (Cp) and hydrocarbon correction (Ну) terms. Cp describes the influence of pore pressure on the sonic porosity equation. It is normally estimated from comparison of density and apparent sonic porosity or from the sonic response in nearby shale (Cp = Afsh/100.0). Hy is approximate and is set equal to 0.9 for oil and 0.7 for gas. The Wyllie time average equation is
. _ Ar1-Azma Hy
The Raymer-Hunt-Gardner
equation is an e m p i r i c a l
approach developed from a statistical analysis of a database of
sonic measurements and independently determined porosity.
Tliis equation is
=CAh -Afma Ал
The value of C can vary between 0.625 and 0.70, depending upon local conditions. The most widely accepted value is 0.67. A value of 0.60 is recommended for gas reservoirs.
Graphical charts for solving these equations are normally in service company chartbooks.
Advantages
The sonic method is very fast and can be very accurate when parameters are tuned to a particular reservoir. For this reason, the sonic porosity method is commonly used in the appraisal of field development wells.
Limitations
The interpretation of sonic logs is sensitive to formation pressure and rock compressibility, particularly in unconsolidated rocks. Results are variable in hydrocarbonbearing zones due to the uncertainty of the appropriate C or
Atfl for hydrocarbons. The sonic log may measure the interval traveltime of the flushed zone or of the unaltered formation d e p e n d i n g u p o n the d e p t h of invasion a n d the relative velocities of the t w o zones. This can p r o d u c e s o m e uncertainty in selection of a p p r o p r i a t e interpretation parameters.
Porosity from Density and Neutron
Uses
Porosity and lithology can both be determined from s i m u l t a n e o u s u s e of b o t h the d e n s i t y a n d n e u t r o n measurements. (These methodologies are discussed further in the chapters on "Lithology from Logs" and "Difficult Lithologies" in Part 4.) Two primary methods are used.
Method 1
The first method requires determining the lithology that will cause both the density and neutron response equations to produce similar porosity values. One of the easiest methods is to solve these two equations simultaneously in a graphical manner. Service companies provides their customers with chartbooks containing density versus neutron crossplots for their particular tools. An example is shown in Figure 2. First, one should mark the horizontal axis with the environmentally corrected neutron reading in limestone porosity units and the vertical axis with the bulk density. Be sure that the chart and neutron log are on compatible lithology scales (limestone). The point where the two meet represents the porosity and lithology of a clean, water-filled rock. The porosity is determined by interpolating between equal porosity lines connecting the common porosity points on the sandstone, limestone, and dolomite lines. The lithology is determined by interpolating between the labeled lithologies.
Advantages. The density-neutron generally provides the best sensitivity to lithology, porosity, hydrocarbons, and shaliness.
Limitations. Results are d e g r a d e d in rugose holes where the density is unable to maintain proper wall contact. In heterogeneous formations, the density may respond to formation characteristics occurring on only one side of the hole while the neutron responds to the formation on all sides. In additional, the two devices have d i f f e r e n t d e p t h s of investigation. These conditions may lead to inappropriate conclusions concerning lithology and/or porosity.
Note. Other crossplot techniques available are the density-sonic and the density-neutron. The density-sonic is inappropriate for fractured formations and generally has poor sensitivity to lithology. The neutron-sonic has reduced sensitivity to lithology compared to the density-neutron, but it has increased sensitivity to the presence of gas. Both of these techniques are subject to the limitations of measuring porosity with the sonic discussed earlier.
Method 2
An a p p r o x i m a t i o n of the crossplot a n s w e r for clean reservoirs can be made using the following equation:
Jnt +ФП
(APPARENT LIMESTONE POROSITY)
Figure 2. Density-neutron crossplot chart. (Courtesy of Schlumberger Well Services, 1979.)
Standard Interpretation
183
Method
The predominant electrical conductor in subsurface formations is salt water. Most other fluids and minerals normally encountered in reservoirs are nonconductors. The analysis of the actual conductivity (or resistivity) of the rock compared to the conductivity (or resistivity) of the same rock when fully saturated with water serves as the basis for determining water saturation from wireline logs. This fundamental relationship was described by Gus Archie and is expressed by the following equation:
n _ Fг Ri \ w
w~
Rt
where Sw = water saturation
F = formation resistivity factor Rw = resistivity of formation waters Rt = true formation resistivity
n = saturation exponent
The formation resistivity factor, F, describes the tortuosity of the conductivity paths (pore space) in the rock. It can be determined in the laboratory from cores or it can be estimated from the following relationship:
where ф = estimated porosity фы = neutron porosity for the appropriate lithology ф0 = density porosity for the appropriate lithology
Advantages. This method provides quick, reasonable answers even in the presence of gas.
Limitations. This method requires that the porosities be determined for the appropriate lithology. It is only an approximate answer (typically within one or two porosity units). Accuracy decreases with increasing shale content and gas effects.
DETERMINATION OF WATER SATURATION
Water saturation (Sw) is most often determined from the logging m e a s u r e m e n t of resistivity a n d k n o w l e d g e of porosity, water resistivity, and shale volume. The interpretation procedures can be divided into two separate procedures—one for shaly and one for clean reservoirs.
Water Saturation in Clean Reservoirs
Application
The Archie equation is the primary method for interpreting resistivity measurements. Either the induction or the laterolog wireline measurement can be used for formation water saturation in areas where the formation water contains moderate to saturated amounts of dissolved salts and the reservoir rock contains no shale or clay.
w h e r e ф is the porosity of the rock and a and m are determined from local experience. The most common values for a and m are as follows:
In soft formations: In hard formations:
a = 0.81, m = 2.0, or a = 0.62, m = 2.15 a = 1.0, m = 2.0.
In fractured rocks, m tends toward lower values, and in vugular rocks, it tends toward higher values.
The value of the resistivity of the formation waters, Rw, can be determined from the SP in water sands, from the Archie equation applied to nearby water sands, from water samples, or from local experience (see the chapter on "Determination of Water Resistivity" in Part 4). This value of the true resistivity of the formation, Rt, can be determined from analysis of the dual induction or dual laterolog measurements (see the chapter on "Preprocessing of Logging Data" in Part 4).
The saturation exponent, n, describes how the tortuosity of the conductive paths in the formation increases as the water saturation decreases. The value of n can be determined in the laboratory from measurements on cores or from local experience. The most typically used value for n is 2.0.
Advantages
This method is the simplest for interpreting saturation in clean sands. It can be calibrated with cores for increased accuracy.
184
PART 4— WIRELINE METHODS
Limitations
It only works for clean reservoirs. The value for m is very difficult to determine accurately in fractured or vugular reservoirs.
Water Saturation in Shaly Reservoirs
Application
Many variations on the Archie equation have been developed over the years to describe the influence of clay or shale as a conductor in reservoir rocks. These methods simply have an additional clay or shale term included to the basic Archie equation to account for the clay conductivity. Variations of the different equations have arisen to handle the different distributions that shale or clay may take in the rock. (For references that provide guidelines for selecting the appropriate equation for a particular formation, see the chapter on "Difficult Lithologies" in Part 4.)
Example Interpretation
The logs in Figures 3 and 4 are used for this example. Figure 3 is a deep induction log recorded with a shallow resistivity device and SP, while Figure 4 is a compensated density-neutron log with a gamma ray. The well is an unconsolidated sand and shale sequence from the Gulf Coast area. Interval D is a p p a r e n t l y wet b e c a u s e of the low resistivity. Rw can be determined in this interval from either the SP or by the Archie method (see the chapter on "Determination of Water Resistivity" in Part 4). R w was determined here by the Archie method to be 0.026 Q-m at formation temperature. For the purposes of this example, no environmental corrections were made and were assumed to be unnecessary.
Point A is above the gas-oil contact, as indicated by the severe crossover of the density and neutron porosities that were recorded on a sandstone matrix. The density-porosity log reads 35% and the neutron log 11%. Thus, we have
Method
The Simandoux equation is a good general purpose equation that accounts for the influence of shale with regard to water saturation. The Simandoux equation is
1
^w
R1 FxRw(\-VSH)
j ^sn x ^w Rm
where Sw = water saturation
F = formation resistivity factor Rsh = resistivity of shale Vsh = volume of shale Rw = resistivity of formation waters
Rt = true formation resistivity n = saturation exponent
The selection of parameters is similar to that described for clean reservoirs. Note that the equation is quadratic, making it significantly more difficult to solve manually for saturation. Normally, saturations in shaly sands are determined using computers or programmable calculators.
Advantages
This method accounts for shale conductivity in reservoir rocks. Different equations exist for structural, laminated (interbedded), dispersed, and intermixed distribution clays and shales.
Limitations
One must understand the shale distribution to select the appropriate equation. The value of m is very difficult to determine accurately in fractured or vugular reservoirs. In low salinity reservoirs, these equations overestimate water saturation due to a phenomenon called excess clay conductivity; in these cases, a different model (such as dual water or Waxman-Smit) may be more appropriate.
0A = 0.26, or 26%
Point B is in a light oil column as indicated by the slight crossover of the density and neutron porosities that were recorded on a sandstone matrix. The density log reads 30% and the neutron log 17%. Thus,
ф _ ^ п + Ф Ь _ J0-172 +0.302
фв = 0.24, or 24%
Point C is also apparently in a light oil column as indicated by the slight crossover of the density and neutron porosities that were recorded on a sandstone matrix. The density log reads 31% and the neutron log 23%. Thus,
ф = ^Фи+Фр = J 0-232 +0.312"
фс = 0.27, or 27%
The formation resistivity factor can be calculated from these porosities u s i n g one of the r e l a t i o n s h i p s for soft formations:
0.81 0.81 FA = Ф2 = 0.262
0.81 0.81 h = Ф2 = 0.242
0.81 0.81 Fc = Ф2 = 0.272
Saturations can be estimated from the Archie equation since there is little or no volume shale, as indicated by the
Standard Interpretation
185
гооо
SANDSTONE
Figure 3. Induction log for example interpretation.
08400
Figure 4. Density-neutron log for example interpretation.
gamma ray. Rt should be determined after environmental corrections, but is assumed here to be equal to the deep induction resistivity (RILd) due to the high porosity and relatively shallow invasion. In zone A, Rt is 5.0 П-m, in zone B, 13 £2-m, and in zone C, 0.55 Q-m.
Assuming n to be 2.0, Archie's equation can be written and solved for each zone as follows:
This analysis found zone A to be a gas-bearing sand with 26% porosity and 25% water saturation. However, water saturation is most likely overestimated due to the proximity of the massive conductive shale bed. Zone B was found to be a light oil sand with 24% porosity and 17% water saturation. Zone C was found to be a residual light oil sand with 27% porosity and 72% water saturation.
1^wA _ F x R w *t
= 0.25
12.0 0.026 5.0
FxRw R,
= 0.17
14.1-0.026 13.0
FxRw
^wC -
R,t
/11.1-0.026 , 0.55
= 0.72, or 72%
Difficult Lithologies
Khaled Hashmy
Halliburton Logging Services Houston, Texas, U.S.A.
Mark Alberty
BP Exploration Houston, Texas, U.S.A.
INTRODUCTION
The term difficult lithologies, as addressed in this chapter, refers to a formation composed of two or more mineralogies. The presence of two or more minerals significantly increases the difficulty of determining both porosity and lithology from wireline logs. This chapter reviews some common techniques that can be used to solve for lithology and porosity. It also addresses some commonly encountered lithologies and their characteristics relative to log responses.
IDENTIFYING THE OCCURRENCE OF DIFFICULT LITHOLOGIES
The occurrence of difficult lithologies can be identified from the following sources:
1. Local knowledge of formations in the area 2. Cuttings from the well 3. Mudlogs of the well 4. Conventional core analysis 5. Sidewall percussion core analysis 6. Sidewall rotary core analysis 7. Analysis of log responses
The m e t h o d s to d e t e r m i n e the occurrence of difficult lithologies from the first six sources just listed are not covered in this part of the volume (Part 4). (For information on these sources, see the chapters on "Mudlogging: Drill Cuttings Analysis" and "Mudlogging: The Mudlog" in Part 3; also see "Core Description" in Part 5.)
Identifying the occurrence of difficult lithologies from logs can be formidable. Two crossplot techniques are commonly used to identify the occurrence of mineralogies: (1) the M-N crossplot and (2) the MID crossplot.
M-N Crossplot
These values are crossplotted in Figure 1. Binary mixtures of minerals plot along a line connecting the two mineral points. Ternary mixtures of minerals plot in a triangle connecting the three minerals points. Arrows indicate the effects of gas, salt, sulfur, secondary porosity, and shaliness.
MID Crossplot
The MID or matrix z'dentification crossplot uses the apparent volumetric cross section (UMAA) and the apparent matrix grain density (RHOMAA) to identify minerals. U M A A r e p r e s e n t s the projection of the v o l u m e t r i c photoelectric absorption index, U (the product of Pe and electronic density), to the value at zero porosity. RHOMAA results from a mathematical projection of the bulk density of an interval to its value at zero porosity. These projections to zero porosity effectively eliminate variance due to porosity, resulting in a variance mainly due to lithology.
This method requires that an estimate of total porosity be determined first. Typically, this can be done from a density-neutron crossplot (see the chapter on "Standard Interpretation" in Part 4). Using this porosity, an apparent matrix density is determined from the following equation:
RHOMAA = Ploe8 - Ф- Pfl 1.0-ф
UMAA can be determined from the chart in Figure 2. Start by marking the Pe value on the photoelectric portion of the horizontal axis (left side), then go vertically to the bulk density value. Next, move horizontally to the apparent total porosity, and then down to the UMAA value.
Now you can cross plot the RHOMAA and UMAA values on the chart in Figure 3. Binary mixtures of minerals plot along a line connecting the two mineral points. Ternary mixtures of minerals plot in a triangle connecting the three mineral points. Arrows indicate the affects of gas, secondary porosity, salt, barite, and heavy minerals.
The M-N crossplot uses the density, compensated neutron, and compressional sonic logs to identify binary and ternary mixtures of minerals. The terms M and N are defined as follows:
Atfl - Aflog
M =—
— x 0.01
Pb-Pfl
j _ Фл/fl ~0Nlog
Pb-Pfl
TECHNIQUES FOR ANALYZING DIFFICULT LITHOLOGIES
Each logging measurement can be expressed in a response equation that relates the recorded log to the volumetric components of lithology and fluids. (For these response equations expressed in their most basic configuration, see the chapter on "Standard Interpretation" in Part 4.) These basic e q u a t i o n s can be e x p a n d e d to i n c l u d e a n y n u m b e r of mineralogies and fluids.
186
Difficult LitJwlogies
187
GYPSUM
Fresh Waler (0 ppk) p. = 1.0 U| .398 Sail Waler (200!ppk) p,.= i 1 Ur = 1 36
0.9
M
I \
SECONDARY POROSITY
I I
CALCITE,
Cf
D O L O M I T EC,era'
I 23
^<m019--,5500.94f3l/sm/s
SILICA
»
Vmo
=
5,486
m/s
= 18,000 fl/s
0.7 0.6 0.5
0.3
o-"' ANHYDRITE
/
SULFUR —
(C) Schlumberger
APPROX SHALE REGION
POP11F=R=IOEISOqHmM/gc/Mcrr,v5U,I111D==
620AiS/m l89/xs/fj_
• S
P1
P1-1
ALT
=I I
MMgU/mD3,
i,
=
607/is/m
lgm/cc, = 185 fis/ft
N
0.4
0.5
0.6
0.7
0.8
Figure 1. M-N crossplot for mineral identification. (Courtesy of Schlumberger Well Services, 1979.)
All the logging response equations can then be set up for each measurement, such as the density, neutron, and sonic. The unknowns must be less than or equal to the number of equations for a unique solution to be obtained. A paramount fact that must be kept in mind is that the number of variables that are b e i n g c o m p u t e d c a n n o t exceed the n u m b e r of equations. For example, if the density, neutron, and sonic logs are being used, the total number of equations that can be set u p is four—one for each of the measured logs and the fourth for the material balance equation (exemplifying that the sum of all constituents equals 100% of the volume of the rock). Thus, in this case, only four variables can be computed. Assumptions and local knowledge can be used to constrain the problem by reducing the amount of unknown knowledge.
Once these logging relationships are established for the difficult lithology being addressed, a number of methods are available to solve these equations effectively for a solution. The more commonly used methods are
• Graphical crossplots • Linear matrix solutions • Weighted least squares minimization
Graphical Crossplots
A w i d e v a r i e t y of c r o s s p l o t s are used for analysis. Basically, these crossplots are designed for use with two porosity measurements and are all generally published in vendor chartbooks. The three most common crossplots are discussed here.
P„ Photoelectric Cross Section (barns/electron)
Umoa, Apparent Matrix Volumetric Cross Section (barns/cc)
Figure 2. UMAA determination chart. (Courtesy of Schlumberger Educational Services, 1989.)
Density-Compensated Neutron Crossplot
This type of crossplot is used for binary mixtures of sandstone and limestone, limestone and dolomite, dolomite and anhydrite, and sandstone and shale.
Sonic-Compensated Neutron Crossplot
Two versions of the sonic-compensated neutron crossplot exist, one that uses the Wyllie time average equation and the other the Raymer-Hunt-Gardner equation for the sonic porosity relationship. These are used typically for binary mixtures of s a n d s t o n e and limestone or limestone and dolomite. However, the sonic-compensated neutron crossplot is not particularly useful in fractured or vuggy formations.
Density-Sonic Crossplot
Two versions of the density-sonic crossplot also exist, one using the Wyllie time average equation and the other the Raymer-Hunt-Gardner equation for the sonic porosity relationship. These are used typically for binary mixtures of sandstone, limestone, or dolomite with gypsum, trona, salt, or sylvite. This plot is not particularly useful in fractured or vuggy formations nor for sandstone and limestone or limestone-dolomite mixtures.
Linear Matrix Solutions
Linear matrix solutions are almost always performed on a computer. A matrix is developed for each of the unknowns with the response equations used to describe the relationship between each element of the matrix. The computer uses a matrix solver to find the unique solution to the problem. The response equations must be linear for this system to work. This drawback leads to oversimplification of the response equations, which can result in degradation of the answers.
Weighted Least Squares Minimizations
The weighted least squares minimization method is always performed on a computer. A number of commercial
188
PART 4— WIRELINE METHODS
Umaa, Apparent Matrix Volumetric Cross Section (barns/cc)
Figure 3. MID plot for mineral identification. (Courtesy of Schlumberger Educational Services, 1989.)
packages exist, including Global and Elan from Schlumberger, Optima from Western Atlas, ULTRA from Halliburton, and Probe from Scientific Software. Although these software packages are similar in general principle, they differ in individual methods of implementation, flexibility, and assignment of uncertainty.
This methodology uses an algorithm to search the range of possible answers to the problem to find the single answer that best fits the response equations and the measured logs. The equations are not required to be strictly linear, although as the response equation become somewhat nonlinear, the search becomes more difficult and time consuming. An incoherence function is output which is usually a useful indicator of the reliability of the answer (low incoherence is good). These programs are usually very flexible in the input of assumptions and allow virtually any type of mineralogy to be solved.
TYPES OF DIFFICULT LITHOLOGIES
The more commonly encountered difficult lithologies include
• Shaly sandstones • Evaporites • Carbonates • Metamorphic and igneous rocks • Volcanic and tuffaceous rocks
Shaly Sandstones
The interpretation method best suited for shaly sandstones is dependent upon the distribution of shale, the clay type, the mineralogy of the silt fraction, and the resistivity of water within the sandstones. The classic approach is the sand-silt-shale method introduced by Poupon et. al (1971). An approximate correction for a single heavy mineral was provided for in this approach. Silt is considered to be primarily quartz. Volume of clay, volume of silt, and porosity are determined from interpolation of the density-neutron crossplot. Matrix response points are defined for sand and silt, water, and dry clay minerals. A wet clay point is defined on the dry clay minerals-100% water line. A shale point was defined on the quartz-wet clay line. The model can then determine porosity, shale volume, and silt index from interpolation in this framework. Water saturation can be determined using an appropriate shaly sandstone resistivity equation. This method does not adequately address the more complex case of shaly sandstones with variable volumes of feldspar, mica, or carbonate material. This model can be solved using the graphical, linear matrix, or least squares minimization method.
The solution for the complex case of sandstones with feldspar, mica, and carbonate material was resolved after log analysts became comfortable with the new spectral gamma ray (K, Th, and U) and photoelectric (Pe) measurements. The spectral gamma ray log is helpful in sandstones containing potassium feldspars or thorium-bearing clays. The natural gamma ray spectra, Pe, density, and neutron expanded response equations can be combined to solve for porosity and to estimate volumes of calcite, quartz, dolomite, clay, feldspar, anhydrite, and salt. Once porosity is determined, saturation can be estimated from the appropriate shaly sandstone resistivity equation. This model is too complex to address using graphical methods and must be done using the linear matrix or least squares minimization method.
The distribution of shale and the salinity of formation water can significantly affect the resistivity of the formation. A variety of saturation models have been developed over the years to describe the influence of these factors. Some commonly used equations are as follows:
Simandoux:
Q
=
-C^wr
S
2 w
+ SwVshCsJ1
Indonesia: C4 _ -^r-5¢W2 I I CwVpshCsh5Wc2 +^wr2^sтh/2^-сsh
Dual water:
r
_
~
Cw
p
„2 ^W +
(Cbw ~Cw)VshQv
p
^ ^w
^O
' O
Waxman and Smits:
^
^q
Ct = —v-Sw + — T l S w 1
^
For a more complete explanation of water saturation equations and their terms, refer to Worthington (1985) or Patchett and Herrick (1982). The Simandoux and Indonesia equations were designed mainly for relatively salty formation waters and moderate amounts of dispersed clay. The dual water and Waxman and Smits equations were designed for all water salinities and moderate amounts of dispersed clays.
Recommended logs to use for interpreting shaly sandstones are
• EXial laterolog or dual induction • Density • Compensated neutron • Spectral gamma ray
Carbonates and Evaporites
Carbonates and evaporites can complicate the interpretation process through the occurrence of fractures and vugs. The three primary porosity devices—the density, compensated neutron, and sonic—respond differently to the presence of fractures and vugs. The density investigates one side of the borehole, while the compensated neutron averages the formation properties around the borehole. The sonic generally does not respond to the secondary porosity associated with either fractures or vugs. This unequal response to these properties can lead to significant problems in determining lithology from any of the methods, as they generally presume that all the porosity devices see a common porosity and lithology.
Binary mixtures of carbonates or evaporites can be evaluated using graphical crossplot approaches. The density-neutron is generally the most used approach. This method will solve for sandstone-limestone, limestonedolomite, or dolomite-anhydrite mixtures.
More complicated mixtures are best handled using the linear matrix or preferably the weighted least squares
Difficult LitJwlogies
189
minimization technique. Additional minerals can be identified by incorporating the photoelectric, spectral gamma ray, and sonic logs. Some log measurements do not provide any relevant information in a given section. If the K, Th, and U values are uniformly low in the evaporite minerals, these logs h a v e no resolving ability and are of little help in determining the rock constituents. If dolomite contains uranium, the uranium log is diagnostic.
Recommended logs to use for interpreting carbonates and evaporites are
• Dual laterolog or dual induction • Density with photoelectric effect • Compensated neutron • Compensated sonic • Spectral gamma ray
Metamorphic, Igneous, and Volcanic Sequences
The approach to the interpretation of these sequences is similar to that of the carbonate sequences. Evaluation of tuffites and tuffaceous sandstones has been successful using compensated sonic, density, and compensated neutron crossplots on which the minerals form a lithologic triangle of silica, light tuffite, and heavy tuffite. Models can easily be established for igneous and metamorphic sequences using the linear equations or the weighted least squares minimization approaches. The methods require the identification of the appropriate mineral matrix responses. Table 1 lists the log matrix responses of common minerals.
Recommended logs to use for interpreting metamorphic and igneous sequences are
• Dual laterolog or dual induction • Density with photoelectric effect • Compensated neutron • Compensated sonic • Spectral gamma ray
190
PART 4— WIRELINE METHODS
Table 1. Logging Tool Responses in Sedimentary Minerals
Name
Formula
PLOG Ф SNP Фш
L
U
Pe
U
e
TP
GR
g/cc p.u.
p.u. /us/tt /LTS/FT barn/elect barn/cc tarads/m nsec/m API units c.u.
SILICATES
Quartz
SiO7
2.64
/j-Cristobalite Opal (3.5% H?0) Garnet Hornblende
SiO7 SiO7 (H7O) 1?09 Fe3AI? (SiO4)3 Ca7NaMg7Fe7AISi8O27(O1OH)2
2.15 2.13 4.31 3.20
Tourmaline Zircon
NaMg3AI6B3Si6OHOHb ZrSi 0,
3.02 4.50
CARBONATES
Calcite
CaCO3
2.71
Dolomite Ankerite Siderite
CaCO3MgCO3 Ca(Mg1Fe)(CO3)? FeCO3
2.85 2.86 3.89
OXIDATES
Hematite Magnetite Geothite
Fe?03 Fe3O, FeO(OH)
5.18 5.08 4.34
Limonite Gibbsite
FeO(OH)(H7O)2 Os AI(OH)3
3.59 2.49
PHOSPHATES
Hydroxyapatite Ca5(PO4)3OH
3.17
Chlorapatite Fluorapatite
Ca5(PO4)3CI Ca5(PO4)3F
3.18 3.21
Carbonapatite
(Ca5(P04)3)?C03H?0
3.13
FELDSPARS-AIkaIi
Orthoclase
KAISi3O8
2.52
Anorthoelase Microeline
KAISi3O8 KAISi3O8
2.59 2.53
FELDSPARS-Plagioclase
Albite Anorthite
MICAS
NaAISi3O8 CaAI7Si2O8
2.59 2.74
Muscovite Glaueonite Biotite
KAI7(Si3AIO10)(OH)? K7(Mg1Fe)7AI6(Si4O10)3(OH)2 K(Mg,Fe)3(AISi3010(0H)7
2.82 -2.54 — 2.99
Phlogopite
KMg3(AISi3O,0(OH)7
1. 2. 4. 3. 4. 16. - 1.
0 2. 0 5.
4. 3. 50 + 50 i 50 f
5. 1. 1. 5.
2. 2. 2.
-1. -1.
12. - 23. -11.
2.
56.0 88.0
3.
2.
58.
7.
8.
43.8 I 81.5
22.
3.
-1.
49.0 88.4
1. 44.0 72.
1.
12.
47.
11. 9.
60 t 60 I 60 +
42.9 73.
56.9
79.3 102.6
8.
42.
1. 42.
2.
42.
8.
3.
69.
2.
-3.
-2.
49.
85.
2. 45.
20. - 38. -21.
49. 149.
50.8 50.
224. 207.
1.81 1.81 1.75 11.09 5.99 2.14 69.10
5.08 3.14 9.32 14.69
21.48 22.24 19.02 13.00
1.10
5.81 6.06 5.82 5.58
2.86 2.86 2.86
1.68 3.13
2.40 6.37 6.27
4.79 3.89 3.72 47.80 19.17 6.46 311.
4.65
7.2
-
4.26
-
3.52
-
5.03
-
44.91
-
18.12
-
7449.82
-
6.92
13.77
7.5
9.1
-
9.00
6.8
8.7
-
26.65
-
57.14 6.8 7.5 8.8 9.1
-
7.08 4.70 22.18 52.31
111.27
-
112.98
-
82.55
-
46.67 9.9 10.9 1 0 . 5 - 1 1 . 0
-
-
101.37 103.08
85.37 71.12 23.11
18.4 19.27 18.68 17.47
-
9.60
-
130.21
-
8.48
-
9.09
7.21 4 . 4 - 6 . 0 7.41 4 . 4 - 6 . 0 7.24 4 . 4 - 6 . 0
7.0-8.2 7.0 -8.2 7.0-8.2
-220 -220 -220
15.51 15.91 15.58
4.35 4 . 4 - 6 . 0 7.0 8.2
-
8.58 4 . 4 - 6 . 0 7 . 0 - 8 . 2
-
7.47 7.24
6.74 16.24 18.75
6.2-7.9 4.8 6.0
8.3-9.4 7.2-8.1
-270 -275
16.85 24.79 29.83 33.3
Difficult LitJwlogies
191
Table 1 (continued)
Name
Formula
^LOG 4>SH P
^CNL
I
L
Pe
U
€
tp
GR
L
g/cc p.u.
p.u. Ats/ft fXSitt barn/elect barn/cc farads/m nsec/m API units c.u.
CLAYS
Kaolinite Chlorite Illite
Montmorillonite
EVAPORITES
Halite Anhydrite Gypsum Trona Tachydrite Sylvite Carnalite Langbenite Polyhalite Kainite Kieserite Epsomite Bischofite Barite Celestite
SULFIDES
Pyrite Marcasite Pyrrhotite Sphalerite Chalopyrite Galena Sulfur
COALS
Anthracite Bituminous Lignite
AL, Si, O10(OH)B (MgJe1AI)6(Si1AI)4O10(OH)8 K 1 1 5 A U S i 7 6 S.AI115)0?O(OH), (Ca1Na)7(AI1Mg1Fe)4 (Si,AI)80?0(0H)4(H?0)N
NaCI CaSO4 CaSO4(HpO)? Na?C03NaHC03H?0 CaCI?(MgCI?)?(H?0)1? KCI KCIMgCI?(H?0)6 K?S04(MgS04)? K?S04MgS04(CaS04)?(H?0)? MgS0.,KCI(H?0)3 MgS04H?0 MgSO4(HpO)r MgCIp(HpO)6 BaSO4 SrSO4
FeSp FeSp Fe7S8 ZnS CuFeS? PbS S
CH 3 ^ N 009O o?? CH /93N015O 078 CH 84GN OTSO ?U
2.41 2.76 2.52
2.12
2.04 2.98 2.35 2.08 1.66 1.86 1.57 2.82 2.79 2.12 2.59 1.71 1.54 4.09 3.79
4.99 4.87 4.53 3.85 4.07 6.39 2.02
1.47 1.24 1.19
34. 37. 20.
40.
-2. 1.
50* 24. 50 + -2. 41.
1. 14. 40. 38. 50 + 50 + -1. - 1.
-2. 2.
-2. 3. 2. 3. 2.
37. 50 4 47.
37.
1.83
4.44
- 5.8
-8.0
80-130
14.12
52.
6.30
17.38
- 5.8
-8.0
180-250
24 87
30.
3.45
873
- 5.8
-8.0
250- 300
17.58
44.
2.04
404
-5.8
-8.0
150 200
14.12
3.
67.0
2.
50.
60 T 52.
35.
65.
60 + 92.
3.
60 +
2.
25.
60 t
43.
60 +
60 ) 100
2.
•1.
120.
4.65 5.05 3.99 0.71 3.84 8.51 4.09 3.56 4.32 3.50 1.83 1.15 2.59 266.82 55.19
9.45 5.6 6.3
14.93
6.3
9.37
4.1
1.48
6.37
15.83 4 . 6 - 4 . 8
6.42
10.04
12.05
7.42
4.74
1.97
3.99
1091.
209.
7.9-8.4 8.4 6.8
7.2-7.3
500 + -220 -290 -200 -245 -
754.2 12.45 18.5 15.92
406.02 564.57 368.99
24.19 23.70 195.14 13.96 21.48 323.44
6.77 7.90
3. 39.2
62.1
16.97
84.68
-
3.
16.97
82 64
-
-3.
20.55
93.09
-
3.
35.93
138.33 7.8 8.1 9.3 9.5
-
3.
26.72
108.75
-
3.
1631.37 10424.
-
- 3 . 122.
543
10.97
-
90.10 88.12 94.18 25.34 102.13 13.36 20.22
38. 105. 60 t 120. 52. 160.
0.16 0.17 0.20
0.23 0.21 0.24
-
8.65
-
14.30
-
12.79
From SchIumbergerWeII Services (1989).
Formation Evaluation of Naturally Fractured Reservoirs
Roberto Aguilera
Servipetrol Ltd. Calgary, Alberta, Canada
INTRODUCTION
There are many logging tools and many methods that can be used for detection and evaluation of naturally fractured reservoirs (see Aquilera, 1980). However, there are no panaceas, and tools and methods that work well in one reservoir can fail miserably in the next one. (For information on standard logging tools, see the chapter on "Open Hole Tools" in Part 4).
Well logs can be considered as an indirect source of information. Sometimes they are used in efforts to detect where the fractures are located (qualitative analysis) and sometimes they are used to attempt to quantify the degree of fracturing. (For details of other methods used to evaluate naturally fractured reservoirs, see the chapter on "Evaluating Fractured Reservoirs" in Part 6.)
QUALITATIVE ANALYSIS
Sonic Amplitude Log
Sonic amplitude logs are frequently used in attempts to detect fractures. Tlirough laboratory work and experience, it has been found that the compressional wave amplitude is generally more attenuated by vertical and high angle fractures, while the shear wave amplitude seems to be more attenuated by horizontal and low angle fractures. Care must be exercised when using this tool because other features such as a rough borehole, shales, and changes in porosity, lithology, and tool centralization might produce amplitude drops that have nothing to do with the presence of fractures.
H o w e v e r , solid-to-solid contact along the p l a n e of a fracture might reduce the degree of acoustic discontinuity. In this case, the fracture is present but is not detected by the sonic amplitude log.
Variable Intensity Log
Variable intensity logs are presented commercially as a recording of d e p t h versus time after the initiation of an acoustic pulse at the transmitter. Amplitude changes are indicated by a succession of varying shades across the film track. The darkest areas correspond to the largest positive amplitudes, while the lightest areas correspond to the largest negative amplitudes.
When the tool is run through an unfractured section of reasonably constant porosity, lithology, borehole rugosity, and tool centralization , the shades of dark and light colors give the impression of banding. If a fracture is present, drastic breaks in the banding occur.
Sonic Log
In some cases, the presence of natural fractures might produce cycle skipping in sonic logs. Other potential sources of cycle skipping include operational methods of the logging unit and gas in the borehole. Long spacing sonic logs are also useful for locating fractures. This tool operates with transmitter-receiving spacings of 8, 10, or 12 ft and emits acoustic energy that propagates radially through the formation via the mud and back to the receiver via the mud.
Caliper Log
Sometimes the borehole gets enlarged in the presence of natural fractures. Thus, caliper tools (two-, three-, four-, or six-arm) can provide good indications of fractures.
Resistivity Log
All kinds of resistivity logs can be used under the right circumstances for locating natural fractures. For example, induction logs can show drastic increases in resistivity in wells drilled with oil-based muds. In wells drilled with muds of relatively low resistivity, a shallow laterolog will indicate less resistivity than the induction log. In the same way, a microspherical log will show less resistivity than a deep laterolog. In general, all methods involving resistivity logs depend on a marked contrast between shallow and deep resistivities.
Pe Log
In wells drilled with muds that contain barite, the Pe measurement might locate natural fractures. If the barite penetrates the fractures, the Pe curve shows considerable increases.
Borehole Televiewer
The borehole televiewer presents an acoustic picture obtained with a rotating ultrasonic scanner. As such, it can sometimes "see" natural fractures. Some of the limitations include the necessity of a low solids content in the borehole fluid, a very slow constant logging speed, and a near-perfect tool centralization. Fractures wider than 1/32 in. can probably be seen by the borehole televiewer. In some cases, damage may make the fracture appear wider than it actually is. (For more information on the borehole televiewer, see "Borehole Imaging Devices" in Part 4.)
Dipmeter
The dipmeter log has been used in some instances to indicate zones of vertical fracturing and their orientation. In
192
general, the continuous four-pad dipmeter detects vertical fractures in opposite curves, that is, in curves one and three or curves two and four. This type of response depends on invasion of the fracture by a conductive fluid. A twodirectional caliper run with the dipmeter can sometimes show an elliptical hole in front of the fractured interval. (For more on the dipmeter, see the chapter on "Dipmeters" in Part 4.)
Formation MicroScanner Formation MicroScanner1 is an extension of the dipmeter
in which an array of electrodes is used to produce an electrical image of the formation. In the four-pad configuration, each pad contains 16 buttons that cover approximately 44% of an 8in.-diameter borehole. Obtaining a good electrical image depends on invasion of the fracture by a fluid of relatively low resistivity. (For more on this device, see the chapter on "Borehole Imaging Devices" in Part 4.)
Spontaneous Potential (SP)
In some naturally fractured intervals, a hacked SP log develops in which each pip most likely corresponds to a streaming potential effect resulting from mud filtrate into the fractures.
Correction Curve on the Compensated Density Log
The Др (delta rho) curve corrects the density log for the effect of r o u g h borehole and m u d cake. As such, the correction curve might be affected by roughness due to fractures or mud in the fractures even if the hole is in gauge. Unfortunately, the correction detector diameter is only about 2 in., and it can easily miss the fractures in more than 90% of the borehole.
Borehole Gravity Meter
Due to its large radius of investigation, the borehole gravity meter has proven useful in detecting large fractured bodies, especially in highly brittle materials.
1Trademark of Schlumberger.
Naturally Fractured Reservoirs
193
Uranium Index
Since uranium is highly soluble in water, it is commonly contained in groundwater. An increase in uranium content with respect to a shale volume (calculated from sources that are independent of natural formation radioactivity) might be an indication of natural fractures. This method, however, does not distinguish between open and mineralized fractures.
Temperature and Noise Logs
This combination has provided excellent results especially when dealing with underpressured gas reservoirs that are drilled with air. For these conditions, the temperature decreases drastically and the noise log detects the noise made by gas when it is flowing to the wellbore via the fractures. (For more details on these logs, see the chapter on "Production Logging" in Part 9.)
QUANTITATIVE ANALYSIS
Neutron, Density, and Sonic Logs
In this combined method, the assumption is made that neutron and density logs read total porosity while sonic logs read only matrix porosity. The difference is taken as secondary porosity, which might include fractures. These logs are also used in the lithology-density crossplot that allows estimates of lithology variations and secondary porosity.
Dual Porosity Models
Dual porosity models are based on the observation that the cementation exponent of the fractures (mf) should be very close to 1.0. This is much smaller than the cementation exponent of the matrix (тъ). The cementation exponent of the composite system (m) varies between mf and mb. The smaller the degree of fracturing, the closer the value of m is to mb. The larger the degree of fracturing, the closer the value of m is to mf. These models permit quantification of fracture porosity.
194
PART 4— WIRELINE METHODS
Part 4 References Cited
Aguilera, Rv 1980, Naturally fractured reservoirs: Tulsa, OK, PennWell Books.
Bateman, R. M., Open-hole log analysis and formation evaluation: Boston, MA, IHRDC, p. 165-189.
Dewan, J. T., 1983, Essentials of modern open-hole log interpretation: Tulsa, OK, PennWell Books.
Dyos, C. J., 1987, Inversion of induction log data by the method of maximum entropy: SPWLA 28th Annual Logging Symposium Transactions, Paper T.
Ekstrom, M. P., C. Dahan, M.-Y. Chen, P. Lloyd, and D. Rossi, 1987, Formation imaging with microelectrical scanning arrays: The Log Analyst, v. 28, p. 294-306.
Lang, W. H., Jr., 1980, Porosity log calibration: The Log Analyst, v. 21, n. 2.
Neinast, G. S., and С. C. Knox, 1973, Normalization of well log data: SPWLA 14th Annual Logging Symposium Transactions Paper I.
Patchett, J. G., and E. B. Coalson, 1979, The determination of porosity in sandstones and shaly sandstones, Part 1— quality control: SPWLA 20th Annual Logging Symposium Transactions Paper QQ.
Patchett, J. G.,and D. C. Herrick, 1982, A review of saturation models: SPWLA Reprint Volume Shaly Sands, SPWLA.
Poupon, A., W. R. Hoyle, and A. W. Schmidt, 1971, Log analysis in formations with complex lithologies: Journal of Petroleum Technology.
Schlumberger Educational Services, 1987, Schlumberger log interpretation principles/applications: Houston, TX.
ScMumberger Educational Services, 1989, Schlumberger log interpretation charts: Houston, TX.
Schlumberger Educational Services, 1989, Schlumberger log interpretation principles, volume II: Houston, TX
Schlumberger Well Services, 1981, RFT—essentials of pressure test interpretation: Houston, TX, Sclilumberger document M-081022.
Schlumberger Well Services, 1983, Schlumberger Open Hole Services Catalog: Houston, TX.
Schlumberger Well Services, 1986, Schlumberger Production Services Catalog: Houston, TX.
Schlumberger Well Services, 1979, Schlumberger Log Interpretation Charts: Houston, TX.
Smolen, J., and L. R.Litsey, 1979, Formation evaluation using wireline formation tester pressure data: Journal of Petroleum Technology, v. 31, n. 1, p. 25-32.
Welex, 1985, Welex log interpretation charts: Houston, TX. Western Atlas International, 1985, Atlas log interpretation
charts: Houston, TX. Western Atlas International, 1987, Formation multi-tester
(FMT)—principles, theory, and interpretation: Houston, Texas. Worthington, P., 1985, The evolution of shaly-sand concepts in reservoir evaluation: The Log Analyst. Zemanek, J., R. L. Caldwell, E. E. Glenn, S. V. Holcomb, L. J. Norton, and A. J. D. Straus, 1969, The Borehole Televiewer—a new logging concept for fracture location and other types of borehole inspection: Journal of Petroleum Technology, v. 21, p. 762-774.
Part 5
LABORATORY METHODS
Contents
• Introduction • Core Description • Overview of Routine Core Analysis • Porosity • Permeability • Core-Log Transformations and Porosity-
Permeability Relationships • Wettability • CapillaryPressure • Relative Permeability • Paleontology • Thin Section Analysis • SEM7 XRD, CL, and XF Methods • Oil and Condensate Analysis • Oilfield Waters • Rock-Water Reaction: Formation Damage • ReferencesCited
edited by
Frank G. Ethridge
Department of Earth Resources Colorado State University Fort Collins, Colorado, U.S.A.
Introduction
Frank G. Ethridge
Department of Earth Resources Colorado State University Fort Collins, Colorado, U.S.A.
One essential ingredient for proper reservoir description is a data base of reservoir rock and fluid sample analyses. Without direct access to rock and fluid sample data, geologists, geophysicists, engineers, geostatisticians, and managers are forced to rely mostly on remotely accessed information supplied by well logs and seismic reflection profiles. Rock and fluid data provide a means of assessing the quality of other forms of subsurface data. This part of the Manual provides a handy reference guide to standard laboratory methods for analyzing rock and fluid samples from hydrocarbon reservoirs. It is divided into five topic areas:
1. Coredescription 2. Coreanalysis 3. Paleontology 4. Petrographicanalysis 5. Chemistryoffluids
A single chapter that describes methods for examining and recording lithological data from continuous cores (Ethridge) comprises the first topic area. This procedure represents an essential first step in analysis of continuous core and is critical to core-log depth adjustment and to construction of depositional models.
The second topic area begins with a general overview chapter on routine core analysis (Almon). Other chapters on this topic discuss porosity (Cone and Kersey), permeability (Ohen and Kersey), core-log transformations and porosity-permeability relationships (Nagel and Byerley), wettability (Funk), capillary pressure (Vavra, Kaldi, and Sneider), and relative permeability (Hawkins).
A single chapter on paleontology (Scott) describes the use of fossils in dating and correlating strata and in predicting reservoir and source rocks. The topic of p e t r o g r a p h i c analysis includes papers on thin section methods (Houseknecht) and X-ray diffraction, cathodoluminescence, and X-ray fluoroscopy (Thomas). Laboratory methods for analyzing oil and condensate (Henshaw and Kaufman), water (Dickey), and rock-water reactions (Keelan and Amaefule) are covered in the final topic area on the chemistry of fluids.
The m a i n p o i n t of Part 5 is to p r o v i d e a general introduction to commonly used laboratory methods and a list of references as a means of introducing practitioners to this multidisciplinary subject.
Acknowledgments
Special thanks are given to William Keighin (U.S. Geological Survey, Denver, Colorado), Richard Vesell (Davies and Associates, Kingswood, Texas), and Logan McMillan (Consultant, Littleton, Colorado) for their constructive reviews of earlier manuscripts. The authors collectively acknowledge their respective employers for providing permission and time to publish these contributions:
Frank Ethridge
Bill Almon Peter Cone Henry Ohen David Kesey Walter Nagel Keith Byerley Jim Funk Charles Vavra J. G. Kaldi Robert Sneider Jeff Hawkins Robert Scott David Houseknecht
Jack Thomas Paul Henshaw R. L. Kaufman L. W. Slentz Parke Dickey Dare Keelan J. O. Amaefule
Dept. of Earth Resources, Colorado State Univ.
Texaco E & P Core Laboratories Core Laboratories Core Laboratories Marathon Oil Company Marathon Oil Company Texaco E & P ARCO Oil and Gas Company ARCO Oil and Gas Company R. M. Sneider Exploration Conoco, Inc. Amoco Research Center Dept. of Geology, Univ. of
Missouri Amoco Research Center Chevron Canada Resources Chevron Canada Resources Consultant, Sante Fe Springs, CA Consultant, Owasso, OK Core Laboratories Core Laboratories
197
Core Description
Frank G. Ethridge
Department of Earth Resources Colorado State University Fort Collins, Colorado, U.S.A.
INTRODUCTION
Careful examination and recording of information from continuous cores provide critical data for stratigraphic correlation, environmental interpretation, and wireline log calibration. Core plugs p r o v i d e samples for analysis of porosity, p e r m e a b i l i t y , fluid s a t u r a t i o n , a n d a host of compositional and textural studies. Recommendations for the well site a n d l a b o r a t o r y h a n d l i n g of cores as well as sedimentological analyses are described by Siemers et al. (1981) and Miall (1990).
Continuous cores have some advantages over outcrop exposures for environmental interpretation (Weimer and Tillman, 1980). These advantages include the following:
1. Not limited to stratigraphic unit outcrop positions 2. Often provide a more complete section of the
stratigraphic unit 3. Better preservation of contacts between units having
significantly different resistances to weathering 4. Better preservation of delicate primary and soft
sediment deformation structures in shale and siltstone units 5. Better preservation of trace fossils 6. Ability to obtain material for petrographic study below the present groundwater table 7. Allow comparison of lithologic properties with petrophysical properties and wireline log responses.
These advantages, however, are offset by the lack of a threedimensional view and the inability to observe lateral fades changes and large-scale sedimentary features directly.
A graphic log format, which visually expresses a stratigraphic succession, is strongly recommended for describing continuous cores. Such logs should reflect the following:
• Thickness • Grain size • Sedimentary structures • Accessories, such as fossils and diagenetic features • Lithologies and nature of contacts between
different lithologies • Textural maturity • Oil staining • Fracturing • Porosity
FORMAT FOR CORE LOGGING
Recommended formats for graphic logging are given by Bebout and Loucks (1984) and Boyles et al. (1986). A
completed graphic log is shown in Figure 1. Grain size, sedimentary structures, and accessories are shown in the left column. Let's look at how each type of data is handled in the graphic log.
Grain Size
Terrigenous clastic grain sizes and carbonate rock types are recorded as a continuous vertical curve reflecting changes in depositional energy. Note that grain size increases to the left, as is inferred in the case of spontaneous potential (SP) and gamma ray curves. Grain size is determined by comparison with a chart made of sized sedimentary material or a visual pattern such as that produced by Amstrat. A good quality handlens or reflecting light binocular microscope are essential for accurate grain size determination.
Sedimentary Structures
Sedimentary structures are recorded to the right of and just i n s i d e the grain size curve. T h e r e is no single set of standardized symbols for sedimentary structures. Examples are given in Swanson (1981) and Bebout and Loucks (1984). Realistic sketches of observed structures may be more useful t h a n s t a n d a r d symbols. The vertical succession of sedimentary structures should be recorded as accurately as possible because this succession is often the key to successful environmental interpretation. Description, classification, and interpretation of sedimentary structures are discussed by Collinson and Thompson (1989) and Lindholm (1987).
Accessories
Fossils, trace fossils, diagenetic and structural features, and other accessories are shown on the far right side of the left column of a graphic log. Standard symbols for these features are found in Swanson (1981) and Bebout and Loucks (1984).
Thickness
Thicknesses are recorded in the column labeled "Interval." A scale showing sufficient details of sedimentary structures and accessories for inferring environments is necessary. Boyles et al. (1986) suggest that the smallest feature shown should be approximately 0.05 in. (1.3 mm) high. Usually a scale of 1 in. = 5 or 10 ft is appropriate.
Lithology
D o m i n a n t lithologies a n d the n a t u r e of the contact between lithologic units are recorded in the "Rock Type and Contacts" column near the middle of a graphic log (Figure 1). Standardized symbols for illustrating lithologies are available and are presented by Tucker (1982) and Lindholm (1987). If
198
Core Description
199
COMPANY WELL FIELD
GRAIN SIZE, STRUCTURES & ACCESSORIES
Large Foresets
Parallel Laminae
Irregular Laminae Ripples
Planar Cross-Beds
Plane-Beds Trough Cross-Beds Mud Clasts
TRACE FOSSILS
c^
Burrows
A
Roots
POROSITY TYPES
M = Moldic
I = Intergranular
FR = F r a c t u r e
DATE
BY
LOCATION
UNIT/AGE
POR. (OSffli
COMMENTS
Limestone
FR
Chert
Sharp Contact Conglomeratic SS
M
Calcareous SS
Shale Coal or Lignite Siltstone Interbedded Contact Sandstone Gradational Contact Conglomerate
Erosional Contact
FOSSILS
^P Wood
Fusulinids
Brachiopods
Gastropods
Bryozoans
^5 B i v a l v e s
Figure 1. Idealized graphic log. Explanation for some symbols used for sedimentary structures, lithologies, fossils, and contacts are given. Grain size of siliciclastic rocks and Dunham's (1962) classification of limestones are indicated in the column on the left. All sedimentary structures are depicted graphically as accurately as possible. Porosity amounts in percent can be scaled as needed. (Modified from Casey, 1980; Boyles et al., 1986.)
200
PART 5—LABORATORY METHODS
more than one lithology or interbedded lithologies exist within an interval, lithologies should be logged by estimating the p e r c e n t v o l u m e of each lithology a n d n o t i n g the terrigenous clastic lithologies to the left of the chemical rocks. Contacts between lithologies are recognized in core as gradational, interbedded, or sharp. Some sharp contacts are probably erosional and may represent unconformity surfaces. Additional information such as the presence of root structures and early diagenetic cements may aid in determining the true nature of sharp contacts.
Maturity
Textural maturity of sandstone units is recorded in the next column to the right ("Textural Maturity" in Figure 1) and is described following the method outlined by Folk (1974). Detrital clay (matrix) content and the sorting and roundness of grains are all considered in the determination of textural maturity. Because it is difficult to estimate the relative abundance of detrital versus authigenic clay without thin section or scanning electron microscope data, care must be taken in making estimates of detrital clay content. Image charts can be used for estimating sorting and roundness (Figure 2).
Porosity
Identification of porosity type and a qualitative estimate of porosity abundance should be made in the column labeled "Porosity." For carbonate rocks, the porosity classification scheme discussed by Choquette and Pray (1970) is recommended. For siliciclastic rocks, four types of porosity are common:
Intergranular Intragranular or moldic Microporosity Fracture (Pittman, 1979)
Of these, microporosity is the most difficult to recognize with a binocular microscope. The existence of microporosity is suggested by the presence of detrital or authigenic clays in sandstones. Accurate laboratory or thin section determinations of porosity types and percentages should always augment the estimates made during core logging. (For more on porosity classification schemes, see chapter on "Porosity" in Part 5.)
Comments
Information not recorded elsewhere on the form can be included in the "Comments" column on the far right (Figure 1). Such items as color, presence of hydrocarbon residue, and inferred depositional environment can be placed here. Color should be determined using the standard color chart distributed by the Geological Society of America (Goddard et al., 1979).
оЫр-дкядIлiжJяrP
P
Ъбв
DIAMETER RATIO (MILLIMETERS)
PHI STANDARD DEVIATION
0.00
0.35
0.50
1.00
2.00
VERBAL SCALE
very well sorted well sorted
moderately sorted poorly sorted
very poorly sorted
P O о O O HI6H
9 SPHERICITY
§ 0 LOW
SPHERICITY
0 Q 0
VERY ANGULAR
SUB- SUBANGUL AR ROUNOEO
WELLROUNOCD
Figure 2. (a) Comparison chart for sorting and sorting classes. (Modified from Pettijohn et al. 1987.) (b) Comparison chart for roundness and sphericity. (Modified from Powers, 1953.)
ROCK-LOG CALIBRATION
In most basins, few intervals in only a small number of wells are cored. Thus, calibrating of rock information (from continuous cores, closely spaced sidewall cores, or cuttings) to w i r e l i n e log r e s p o n s e s is essential to i n t e r p r e t a t i o n of depositional environments from logs alone. (For details of correlating cores to logs, see the chapters on "Preprocessing of Logging Data" in Part 4 and "Core-Log Transformations and Porosity-Permeability Relationships" in Part 5. For information on interpreting depositional environments from logs, see the chapter on "Lithology from Logs" in Part 4.)
Overview of Routine Core Analysis
William R. Almon
Texaco E&P Technology Division Houston, Texas, U.S.A.
INTRODUCTION
Retrieval and analysis of cores is essential to all phases of the petroleum industry. Cores offer the only opportunity to obtain intact, vertically continuous samples that allow the visual examination of depositional sequences and variations in reservoir character. Properly analyzed cores provide data available from no other source; these data should provide direct evidence of the presence, quantity, distribution, and deliverability of h y d r o c a r b o n s . Cores are essential to understanding the nature of the pore system in the potential reservoir unit. The knowledge gained from cores enhances our ability to predict reservoir performance and to select procedures to maximize profitable hydrocarbon recovery.
TYPES OF CORES
Continuous coring, first attempted in Holland in 1908 (Andersen, 1975), is usually most desirable, although many types of specialized core can be obtained (Park, 1985). Data from continuous core are typically combined with wireline log and formation test data to evaluate productivity. Diagenetically altered sandstones (Luffel and Howard, 1988) and thinly laminated reservoirs (Bradburn and Cheatham, 1988) require laboratory analysis of large diameter cores to evaluate porosity, hydrocarbon saturation, and net pay. (For more on continuous coring, see the chapter on "Conventional Coring" in Part 3.)
An alternative to continuous coring is the retrieval of discrete samples from the wellbore face known as sidewall cores. These samples can p r o v i d e useful details of the lithology, petrology, porosity, permeability, and hydrocarbon content of the formation (Toney and Speiglets, 1985). The analytical results can be used to verify log analysis calculations. Selection of sidewall core points after logging allows selective sampling of specific zones (Craft and Keelan, 1985). (Also see the chapter on "Sidewall Coring" in Part 3.)
SAMPLING TECHNIQUES
A number of generally accepted sampling and analytical techniques have been developed for evaluating core samples. The choice of t e c h n i q u e d e p e n d s on the t y p e of core recovered, its lithology, and the nature of the pore system. Core analysis can be divided into several different types: (1) conventional or plug analysis, (2) whole core analysis, and (3) sidewall core analysis (American Petroleum Institute, 1960).
Conventional or Plug Analysis
Continuous cores can be analyzed by conventional or whole core procedures, but conventional core analysis is most frequently used (Keelan,1985). This procedure employs a
small sample to represent an interval of core and produces acceptable results when the pore system is relatively homogeneous. Conventional core analysis plugs are usually collected once per foot or three to four times per meter (Monicard, 1980). Variations in pore system development or lithology require more frequent sampling. Sample density should be adequate to define net pay, hydrocarbon-water transition zones, contact levels, and formation boundaries. Sampling can be done statistically at the mid-point of each foot or the most representative sample in each foot can be selected (Core Laboratories, undated).
Whole Core Analysis
Whole core analysis examines the complete length of fulldiameter core in the interval being tested and affords the maximum possible sample size. Large samples are mandatory in heterogeneous formations in which most of the porosity and permeability are due to fractures, solution vugs, or erratically developed pore systems. In these cases, the volume of individual pore spaces may be large in relation to the size of c o n v e n t i o n a l core analysis p l u g s a m p l e s (American Petroleum Institute, 1960). A variation of whole core analysis, called full-diameter analysis, utilizes selected l e n g t h s of a core r a t h e r than the entire core (Core Laboratories, undated).
Sidewall Core Analysis
Sidewall core analysis is performed on cores recovered by any of the sidewall coring techniques. Analysis is limited to basic tests for permeability, porosity, and residual fluid saturation (Craft and Keelan, 1985).
Percussion sidewall cores from hard, well-cemented formations are badly altered during the coring process and g e n e r a l l y fail to p r o d u c e suitable m e a s u r e m e n t s of mechanical and petrophysical properties (Craft and Keelan, 1985). Sample alteration may be reduced through the use of sidewall boring or a hydraulic press to collect sidewall core samples (Toney and Speiglets, 1985).
Sidewall coring and analysis produce acceptable results when suitable formations, such as the soft Miocene and Oligocene sandstones found in Gulf Coast and California reservoirs, are sampled in adequate detail. Data quality in wells where only sidewall cores are available can be improved by developing correlations between conventional and sidewall core values. This requires that conventional cores and sidewall cores be collected from the same interval in selected wells (Craft and Keelan, 1985).
SUMMATION OF FLUIDS METHOD
Porosity and fluid saturations are usually determined by the summation of fluids method (Jenkins, 1987). This
201
202
PART 5—LABORATORY METHODS
SJdewalI Sw > Conventional Sw
Sidawail Sw < Convantional Sw
' Л ' 1 ' J,1 1 1,L 1 1 1 ' 1 1 • 1 1 1 1 1 ' • » >
10
20
30
40
50
60
70
Convantlonal Cora Watar Saturation (Parcant)
Figure 1. Comparison of water saturation data (Jenkins, 1987) from sidewall and conventional cores shows that sidewall core values are almost always higher than conventional core values. Sample values from gas condensate zones are 10 to 15% higher, while values from oil zones are 5 to 10% higher.
inexpensive and rapid procedure determines sample bulk volume and fluid volumes in the pore system. Total fluid volume is considered to be the pore volume. These data permit estimation of porosity and individual fluid saturations (American Petroleum Institute, 1960).
Analytical procedures vary slightly with sample type. For conventional core plugs and sidewall cores, gas volume is determined by mercury injection. Oil and water volumes are determined by high temperature distillation. Sidewall core measurements are all carried out on one sample. In conventional core analysis, gas volume is determined on one sample while oil and water volumes are measured on a second sample. Significant variations in pore system quality between samples can cause errors in fluid saturation and porosity values (Monicard, 1980).
Gas v o l u m e of w h o l e core samples is d e t e r m i n e d by evacuating the sample and resaturating with water while m e a s u r i n g the weight of water absorbed. With v u g g y samples, water injected to refill the gas volume may drain from the surface vugs causing errors in the measured gas saturation (Keelan, 1972). The oil and water volumes are determined by retorting the core under vacuum. The recovered water volume corresponds to the sum of the pore water and the gas volumes. Low API gravity oils are only partly recovered by this process, thus recovered oil volume may be low (Keelan, 1972). (For more on determining porosity from core samples, see "Porosity" in Part 5.)
Figure 2. Comparison of saturation data (Jenkins, 1987) indicate that sidewall cores from gas condensate zones may have measured oil saturation values that are 2% higher than conventional core samples. In oil zones the relationship is less clear. The agreement between oil saturation values in sidewall and conventional cores may vary with such oil characteristics as API gravity.
RESIDUAL FLUID SATURATION
At depth, reservoir rock contains some combination of liquid and gaseous hydrocarbons and water. The exact fluids present and their relative abundance depends on the type of reservoir (gas, condensate, or oil) and the degree to which it is charged. The fluid distribution at or near the surface is quite different from that under reservoir conditions. These changes are due to drilling processes, gas expansion, and handling errors (Bass,1987; Keelan, 1985).
Water saturations determined on sidewall cores from gas condensate zones are generally 10 to 15% liigher than values from conventional cores. In oil zones, the water saturation from sidewall cores m a y be 5 to 10% higher relative to conventional core values (Figure 1).
The agreement between sidewall and conventional core residual oil saturations varies with oil characteristics. When oil gravity is in the range of 35° to 40° API, sidewall core oil saturation values are slightly lower than those obtained by conventional core analysis. As oil gravity and viscosity increase, sidewall core oil saturations become 10 to 20% lower than conventional core saturations (Craft and Keelan, 1985). In gas condensate zones, sidewall cores have measured oil saturations that are equal to or a few percent higher than conventional core values (Figure 2).
Overview of Routine Core Analysis
203
POROSITY
Whole core porosity is usually less than conventional plug porosity because there is a strong tendency to sample the more porous zones preferentially. Whole core samples i n c o r p o r a t e tighter p a r t s of the p o r e s y s t e m that are frequently excluded from conventional samples. However, whole core porosity may be higher than that determined from conventional analysis when large solution voids are present or when the core is badly invaded by mud solids.
In samples having a porosity greater than 30%, sidewall core porosity is 1 to 2% lower than conventional analysis porosity. This results from slight compaction that occurs during coring. Medium and low porosity percussion sidewall samples, especially from highly cemented rocks, display porosity that is much too high due to fracturing and grain shattering. The deviation between measured porosity and true porosity becomes greater as the actual porosity decreases. Uncertainty caused by systematic variation in sidewall core porosity relative to plug analysis values can be minimized by d e v e l o p m e n t of correlations b e t w e e n sidewall core and conventional core values (Craft and Keelan, 1985). (For more on porosity, see the chapter on "Porosity" in Part 5.)
PERMEABILITY
Whole core samples may contain vugs and fractures that are excluded from core analysis plugs and thus often yield higher permeabilities. To offset this effect, especially in fractured samples, whole core permeability is measured in two horizontal directions. One measurement (reported as к or Zcmax) is made parallel to the major fracture planes and reflects the influence of the fractures as flow pathways. The second measurement is made perpendicular to the first. This value, reported as k90, reflects matrix permeability and is close to conventional core analysis permeability (Jenkins, 1987).
Whole core permeability can be reduced by as much as 50 to 80% by the invasion of drilling mud solids into the pore system or the build-up of powdered rock on the core surface. The relative reduction in permeability appears to decrease as the actual value decreases. Whole core samples may require sand blasting prior to permeability measurement to deal with the surficial b u i l d u p of p o w d e r e d rock. N o m e t h o d is available to address the permeability reduction caused by drilling mud fines that have penetrated the pore system. These fines may cause whole core permeability to be significantly lower than conventional permeability. Plug
(Kc) Conventional Core PermeaHlityl Miliidarcies
Figure 3. Data compiled from 5300 sidewall core samples indicate that sidewall cores from low permeability formations have an indicated permeability greater than that determined from conventional core analysis. Sidewall cores from formations with more than 20 md permeability consistently have a measured permeability that is lower than that from conventional analysis. (After Craft and Keelan, 1985; data from Jenkins, 1987.)
samples from the center of the core do not suffer from surface p l u g g i n g , a n d the effects of drilling fines invasion is minimized (Keelan, 1972).
Low permeability, hard formations (к < 20 md) usually yield sidewall cores that have significantly enhanced permeability relative to conventional core plugs. The gain results from fracturing and grain shattering induced by bullet impact. Permeability values of percussion sidewall cores t a k e n f r o m h i g h p e r m e a b i l i t y (k > 20 m d ) , f r i a b l e , or unconsolidated sandstones are usually reduced by 60% or more over those measured on conventional core plug (Figure 3). Partial blocking of the pore system by drilling mud solids and by compression and grain movement resulting from bullet impact are responsible (Toney and Speiglets, 1985). (For details of calculating permeability form core samples, see "Permeability" in Part 5.)
Porosity
M. Peter Cone
David G. Kersey
Core Laboratories Division of Western Atlas International, Inc. Houston, Texas, U.S.A.
INTRODUCTION
Porosity determines reservoir storage capacity. It is defined as the ratio of void space, commonly called pore volume, to bulk volume and is reported either as a fraction or a percentage. Almost all hydrocarbon reservoirs are composed of sedimentary rocks in which porosity values generally vary from 10 to 40% in sandstones and from 5 to 25% in carbonates (Coneybeare, 1967; Keelan, 1982).
DEFINITION OF POROSITY TERMS
Discrepancies often exist between laboratory determined porosity values and porosities derived from downhole logs. Some of these discrepancies result from differences inherent in comparing direct measurements of physical properties made on small samples with indirect assessments of averaged properties. Many of these discrepancies, however, can be explained by noting differences in the definition and assessment of porosity (Figure 1).
Total Porosity
Total porosity includes all void space regardless of whether the pores are interconnected or isolated. There is no practical way in the laboratory to measure isolated pore volume routinely on rocks. However, it can be determined by disaggregating the samples. If the disaggregated rocks contain smectite, the technique used to dry the samples can affect porosity values and the oven-dried total porosity will be larger than the humidity-dried total porosity (see next section on Effective Porosity). Total porosity from a density log would equate with the disaggregated oven-dried total porosity from cores. The neutron log, however, would enlarge the definition to include structural hydroxyl chemistry.
Effective Porosity
Oven-dried core analysis porosity includes the void space of all interconnected pores plus the volume of water bound to smectite. In contrast, humidity-dried core analysis porosity includes the void space of all interconnected pores plus the volume of all bound water in excess of the volume of a film of water, two molecules thick, retained by smectite. Keelan (1982) r e p o r t e d that r e m o v a l of this film m a y increase porosity 3.3 porosity points in rocks containing 10% smectite.
PORE TYPES
Basic clastic and carbonate pore types can be identified by integrating data from core descriptions, thin section
petrography, scanning electron microscopy, and capillary pressure tests. These analyses indicate that significant differences exist between clastic and carbonate pore types.
Sandstone Pore Systems
Four basic porosity types can be recognized in sandstones (Pittman, 1979): (1) intergranular (primary), (2) microporosity, (3) dissolution (secondary), and (4) fracture (Figure 2). Intergranular porosity exists as space between detrital grains. Microporosity exists as small pores (less than 2 |im) commonly associated with detrital and authigenic clay minerals. Dissolution porosity is the pore space formed from the partial to complete dissolution of f r a m e w o r k grains a n d / o r cements. Fracture porosity is the void space associated with natural fractures.
Carbonate Pore Systems
In comparison to clastic pore systems, pore types in carbonate rocks are more varied (see the chapter on "Carbonate Reservoir Models" in Part 6). Three basic pore groups can be recognized (Choquette and Pray, 1970): fabric selective, not fabric selective, and fabric selective or not (Table 1 and Figure 3). Seven porosity types (interparticle, intraparticle, intercrystal, moldic, fenestral, fracture, and vugs) are common and volumetrically important.
Although fracture porosity is very common in carbonate rocks, it is generally less than 1% of the bulk volume in both clastic and carbonate reservoirs.
INFLUENCE OF TEXTURAL PARAMETERS ON POROSITY
Primary porosity in clastic and some carbonate rocks (such as oolites) is a function of grain size, packing, shape, sorting, and amount of intergranular matrix and cement (Pettijohn, 1975). In theory, porosity is i n d e p e n d e n t of grain size. Changes in grain size, however, affect grain shape and sorting. Because these variables directly affect porosity, changes in grain size indirectly affect porosity.
The theoretical effects of grain size a n d p a c k i n g on porosity were investigated by Graton and Fraser (1935) who computed the porosity of various packing arrangements of uniform spheres. The theoretical maximum porosity for a cubic packed rock, regardless of the value assigned to grain radius, is 47.6%. Porosity values for other packing arrangements (Figure 4) can be calculated.
The effects of grain s h a p e on p r i m a r y porosity were investigated by Fraser (1935) and Beard and Weyl (1973). In general, porosity decreases as sphericity increases due to tighter packing arrangements associated with spherical
204
Matrix
I I
Quartz
Clay^
Layers J
////
Structural
(OH") Water
Total Porosity Neutron Log
Total Porosity Density Log
Absolute or Total Porosity
Oven Dried Core Analysis Porosity
rShale
Humidity Dried Core Analysis Porosity
Clay Surfaces & Interlayers
Small Pores
Large Pores
Capillary * Water
Hydration or Bound Water
Hydrocarbon Pore Volume
."Irreducible or Immobile Water
Porosity
205
Isolated Pores
Figure 1. Schematic of a pore system relating mineralogy, water content, and porosity assessment. (Notes: *lf sample is completely disaggregated during measurement. **Varies as a function of height above the free water level.) (After Eslinger and Pevear, 1988; modified from Hill et al., 1969.)
grains. Numerous studies (Fraser, 1935; Rogers and Head, 1961; Beard and Weyl, 1973; Pryor, 1973) indicate that porosity generally increases with sorting. Gaither (1953) showed that when two grain sizes are mixed, porosity is reduced until both grain sizes are present in approximately equal amounts.
-Intergranular Dissolution >
Fracture
LABORATORY DETERMINATION OF POROSITY
Sample Preparation
Most porosity analysis techniques require removal of soluble hydrocarbons before sample analysis. Factors influencing sample cleaning include the types of hydrocarbon present, the presence of salts precipitated from pore waters, rock mineralogy, degree of cementation, and time constraints. Different solvents and cleaning techniques can be used to remove hydrocarbons from rocks. Toluene is generally an e f f e c t i v e s o l v e n t for m o s t l i q u i d h y d r o c a r b o n s . If hydrocarbons cannot be removed with toluene, toluene/methanol (azeotrope), chloroform/methanol (azeotrope), methylene chloride or carbon disulfide may be used. Methanol is used to remove salts formed from the evaporation of saline pore waters. Rocks containing gypsum and smectite require special low temperature cleaning techniques to minimize removal of structural and bound water (Keelan, 1971).
Laboratory determination of porosity generally requires dry samples. Most clay-free samples can be dried in an oven (1150C). If clay minerals, especially smectite, are present, humidity drying (45% relative humidity, 63 °C) is required to prevent removal of clay-bound water.
Microporosity
Figure 2. Idealized sandstone porosity system showing four basic pore types: intergranular, microporosity, dissolution, and fracture. (After Pittman, 1979.)
Laboratory Analysis
Various laboratory techniques are available to determine porosity. Sample type, pore types, time constraints, and accuracy requirements are generally used to determine the best analytical technique (Table 2).
Porosity can be determined by measuring two of three variables: pore volume (V ), bulk volume (Vb), or grain volume (Vg). Equations 1, 2, or 3 are then used to compute porosity:
Porosity = —y Pj—
(1)
206
PART 5—LABORATORY METHODS
Table 1. Carbonate Pore Types
Pore Type
Description
Fabric selective Interparticle Intraparticle Intercrystal Moldic
Fenestral Shelter
Growth framework
Porosity between particles Porosity within individual particles or
grains Porosity between crystals Porosity formed by selective removal
of an individual constituent of the rock Pores larger than grain-supported interstices (interparticle) Porosity created by the sheltering effect of large sedimentary particles Porosity created by in-place growth of a carbonate rock framework
Not fabric selective
Fracture
Porosity formed by fracturing
Channel
Markedly elongate pores
Vug
Pores Iargerthan 1/16 mm in
diameter and somewhat equant in
shape
Cavern
Very large channel or vug
Fabric selective or not
Breccia
Interparticle porosity in breccia
Boring
Porosity created by boring organism
Burrow
Porosity created by organism
burrowing
Shrinkage
Porosity produced by sediment
shrinkage
From Choquette and Pray (1970).
Porosity = (2)
V„
Porosity =
V +V V p -I- V g
(3)
Porosity can also be determined by adding (using summation of fluids) the individual ratios of gas volume to bulk volume (Gb), oil volume to bulk volume (Ob), and water volume to bulk volume (Wb) (Equation 4). Thus,
Porosity = Gb + Ob + Wb
(4)
Pore Volume Measurement
Pore volume can be measured directly by resaturating a clean, dry rock with a fluid. Resaturation is done with either gas (Boyle's law method) or liquid (gravitational method).
In the Boyle's law method, helium is used to saturate the sample because it is inert, not easily adsorbed onto mineral surfaces, and (due to its small molecular size) rapidly enters
Interparticle
Intercrystal
%JкШвШ я
ЖШ1ШШ
Growth-Framework
Moldic
FABRIC SELECTIVE
S Vug FABRIC SELECTIVE OR NOT
NOT FABRIC SELECTIVE
к'.^Ц
Boring
та Shrinkage I
Figure 3. Idealized carbonate porosity system showing three basic porosity groups: fabric selective, not fabric selective, and fabric selective or not. (After Choquette and Pray, 1970.)
the micropore system. In the laboratory, the core is often placed in an apparatus consisting of a flexible rubber boot inside a core holder (Figure 5). Pressure is then applied to the outside of the rubber boot to seat it against the sample. Helium from a reference cell at known pressure is then expanded into the pore volume. The new equilibrium pressure in the system is monitored, and the pore volume is calculated from Boyle's law:
PlVr = P2(Vr + V l + Vp)
(5)
where P1 = initial pressure in the reference cell P2 = final pressure in the system Vr = volume of reference cell V l = volume of connecting tubing (line volume) Vp = pore volume of sample
Very accurate m e a s u r e m e n t s of pore v o l u m e can be achieved with a Boyle's law porosimeter if the boot conforms to the sample. Consequently, this technique is not suited for vuggy or fractured rocks or for samples that cannot be trimmed into cylinders.
In the gravitational method, a cleaned and dried sample is first weighed and then immersed in a saturating vessel. The vessel is filled with a saturating liquid and pressured to 2000 psi for a minimum of 24 hours. After the pressure stabilizes, the fully saturated sample is removed from the saturator, immediately rolled on an absorbent material to remove the surface film of saturating fluid, and weighed. Pore volume is calculated from the following equation:
Vp = (Ws-Wd)Zps
(6)
where V = pore volume Wg = weight of sample (100% saturated)
Unit Cell
Porosity
207
Single Square Layer
Cubic-Double Square Layers
ф = 47.6%
Single Rhombic Layer
Orthorhombic ф =39.5%
Rhombohedral ф = 26%
Tetragonal ф = 30.2%
Figure 4. Schematic diagram of packing arrangements for spheres. Porosity values are calculated for cubic (47.6%), orthorhombic (39.5%), rhombohedral (26%), and tetragonal (30.2%) packing. (After Berg, 1970; modified from Graton and Fraser, 1935.)
Wd = weight of sample (dry) ps = density of saturating liquid
This technique is not suited for vuggy, fractured, or very low permeability samples.
Grain Volume Measurement
Grain volume can also be measured by the Boyle's law method. The equipment used to measure grain volume and pore volume are similar with the exception of the sample chamber. The grain volume porosimeter does not confine the sample by means of a rubber boot (Figure 6). To measure grain volume, the sample is placed into a chamber of known volume. Helium, from a reference cell at known pressure, is then expanded into the sample chamber. The equilibrium pressure of the system is monitored and Boyle's Law is used to calculate the grain volume. Therefore,
PlVr = P2W^Vс+ V l - V J
(7)
where P1 = initial pressure in the reference cell P2 = final pressure in the system Vr = volume of reference cell Vc - volume of sample chamber V l = volume of connecting tubing (line volume) V = grain volume of sample (unknown)
This is an excellent method to determine the grain volume regardless of the samples shape or surface characteristics.
Bulk Volume Measurement
Bulk volume can be determined by direct measurement, by fluid displacement, or gravimetrically. Calipers can be used to measure uniform samples directly, and bulk volume is calculated from the measured dimensions. This technique is not suited for noncylindrical samples.
Bulk volume can also be determined by immersing a small sample in a nonwetting fluid. Mercury is generally used as the nonwetting fluid, and the bulk volume is equal to the
208
PART 5—LABORATORY METHODS
Table 2. Comparison of Porosity Determination Methods
Method Resaturation porosity Boyle's law porosity: grain volume determination
Boyle's law porosity: pore volume determination
Summation of fluids porosity Thin section porosity point count P.I.A. porosity
Advantages
Accurate Saturated samples available for further testing Saturation time is dependent on permeability
Very accurate Not sensitive to rock mineralogy Samples can be used for further testing Grain density readily determined Irregularly shaped, fractured, and/or vuggy samples easily measured Rapid technique (after cleaning and drying)
Very accurate Not sensitive to rock mineralogy Porosity can be determined at reservoir stress Permeability can be run on same apparatus to avoid stress hysteresis Rapid technique (after cleaning and drying)
Accurate for most rock types Porosity and saturation determined on sample splits Requires no cleaning or drying Rapid technique
Porosity can be determined on irregular, fractured, and/or vuggy samples Pore types can be identified Grain, cement, matrix, and pore relationships can be established
Porosity can be determined on irregular shapes Pore types can be identified Precise determination of visible porosity
volume of mercury displaced by the sample. The gravimetric determination of bulk volume is similar to the saturation procedure used to determine pore volume. The fully saturated sample is first weighed in air, and reweighed while immersed in the wetting fluid. The bulk volume is calculated from Archimedes' principle. Thus,
Vb = (Ws-Wi)Zps
(8)
where Vb - bulk volume Ws = weight of sample in air (100% saturated) Wi = weight of sample immersed in saturating liquid ps = density of saturating liquid
Summation of Fluids (Retort Porosity) Technique
The summation of fluids method is a r a p i d a n a l y t i c a l technique of determining porosity by using the assumption that the total volume of the oil, water, and gas in a material constitutes the pore volume of that material. The former two values are determined through retorting to the samples at elevated temperature and the latter by direct mercury injection. (For more on summation of fluids, see the chapter on "Overview of Routine Core Analysis" in Part 5.)
Other Techniques
Another technique available for the determination of porosity in addition to those mentioned here is point counting pore space occupied by blue epoxy in thin sections (see the chapter on "Thin Section Analysis" in Part 5). Also,
significant progress has been made recently in the d e v e l o p m e n t of p e t r o g r a p h i c image analysis (PIA) as a technique for porosity determination (Ehrlich, et al., 1984; Gerard et al., 1991). In this process, pore space is delineated from mineralogy using photographic imaging techniques. Taking images from several locations on a thin section allows one to compensate for a three-dimensional parameter from two dimensions.
Both X-ray computerized tomography (CT) and nuclear magnetic resonance (NMR) have applications to determining porosity. This is outside the scope of this discussion but is comprehensively covered in the literature (e.g., Vinegar, 1986; Wellington and Vinegar, 1987).
Effects of Confining Pressure on Porosity
Porosity decreases with increasing net overburden pressure (lithostatic pressure minus pore pressure), and in clastic rocks, stress sensitivity generally increases with increasing clay and decreasing cement content (Amaefule et al., 1988). Because porosity is stress dependent, laboratory measurements should be made at stress conditions whenever possible. These measurements are done with specially designed Boyle's law (pore volume) porosimeters, similar to that shown in Figure 5, which apply hydrostatic stress to the sample. In the reservoir, however, the resolved stress component is uniaxial. Uniaxial stress is less than hydrostatic stress, and consequently, the hydrostatic strain measured in the laboratory should be converted to an equivalent reservoir (uniaxial) strain.
Figure 5. Schematic diagram of a Boyle's law helium porosimeter for pore volume measurement.
Porosity
209
Figure 6. Schematic diagram of Boyle's law helium porosimeter for grain volume measurement.
Permeability
Henry A. Ohen
David G. Kersey
Core Laboratories Division of Western Atlas International, Inc. Houston, Texas, U.S.A.
INTRODUCTION
Permeability is a p r o p e r t y of p o r o u s m e d i a t h a t characterizes the ease with which fluid can flow through the media in response to an applied pressure gradient. It is a m e a s u r e of fluid conductivity of p o r o u s material. This chapter discusses specific issues relating to the factors influencing the accuracy and precision of permeability determination.
FACTORS CONTROLLING PERMEABILITY
Pore Geometry
Permeability is a function of the geometry of the pore structure of the porous media. Permeability is controlled in sandstone by grain size, grain orientation, packing arrangement, cementation, clay content, bedding, and grain size distribution and sorting. In carbonates, permeability is a
THEORETICAL BACKGROUND
The fundamental relationship given by Henry Darcy (1856) is the basis for permeability determination. Darcy's law originates from the interpretation of the results of the flow of water through an experimental apparatus, shown in Figure 1. In this experiment, water was allowed to flow downward through the sand pack contained in an iron cylinder. Manometers located at the input and output ends measured fluid pressures, which were then related to flow rates to obtain the following fundamental Darcy's law:
where cj = water flow rate K = constant of proportionality that is characteristic of the
sand pack A = cross-sectional area of the sand pack Ah = Zz1 - H2 = difference in height between the water levels
in the manometers L = length (cm)
The units in which permeability is typically expressed are the darcy (d) and millidarcy (md). A permeability of 1 d allows the flow of 1 cm3 per second of fluid with 1 cP (centipoise) viscosity through a cross-sectional area of 1 cm2 w h e n a pressure gradient of 1 a t m / c m is applied. This definition unfortunately contains nonconsistent units, as pressure is expressed in atmospheres rather than in fundamental units. Lowman et al. (1972), however, have redefined the darcy unit in the mks system in which square meters represents the standard dimension of permeability. The millidarcy, which is one-thousandth of a darcy, is commonly used in core analysis and oilfield operations.
Figure 1. Modified schematic diagram of Darcy's experimental apparatus. (Modified from Hubbert, 1953.)
210
Permeability
211
Figure 3. Effect of net confining stress on permeability. (After Amaefule et al., 1988.)
with the same hydraulic properties (Amaefule et al., 1988). (For more on porosity, see the chapters on "Porosity" and "Core-Log Transformations and Porosity-Permeability Relationships" in Part 5.)
Figure 2. Relationship among permeability, sorting, and grain size. (From Pettijohn, 1975; after Krumbein and Monk, 1942.)
f u n c t i o n of the d e g r e e of m i n e r a l a l t e r a t i o n (such as dolomitization), porosity development, and fractures. Figure 2 shows the relationship among permeability, sorting, and gram size.
Bedding
Directional and local variations of permeability generally exist in reservoirs. Permeability perpendicular to bedding planes (vertical permeability) is typically lower than horizontal permeability (parallel to the bedding planes).
Porosity
Several attempts have been made in the past to derive a general relationship between porosity and permeability. Prominent among these relationships is the work of Kozeny (1927), which considered the porous media as a bundle of capillary tubes of equal length. Modifications to account for tortuosity of flow p a t h s in the p o r o u s media have been proposed, including the Carman-Kozeny model (1938). Unfortunately, only qualitative results have been obtained using these permeability-porosity relationships because of the complexity of the geometry of the porous media.
Berg (1970) suggested that a better understanding of the properties of the rock that control size, shape, and continuity of the rock is the key to relating fluid flow properties to reservoir rock properties. Qualitatively, it is reasonable to assume that permeability should increase with increase in porosity in unfractured reservoirs without significant diagenetic alterations. In fact, it has been shown that there is a relationship between porosity and permeability within units
Confining Pressure
Permeabihty decreases with increasing confining pressure. Unconsolidated or poorly lithified rock undergoes much greater permeability reduction under confining pressure than well-consolidated rock. As shown in Figure 3, a greater percentage of permeability reduction is typically observed in lower permeability rock than in higher permeability rock. To determine permeability-stress relationships, which are representative of in situ reservoir conditions, permeability measurements should be made on selected samples at a series of confining pressures. Jones (1988) has recently presented a m e t h o d that allows a t w o - p o i n t d e t e r m i n a t i o n of a permeability-stress model that reduces the required number of permeability measurements u n d e r confining stress for permeability-stress prediction.
PERMEABILITY DETERMINATION
Theoretical Basis
Permeability is not measured; it is calculated. Therefore, although it is a rock property and a constant for a given core sample, permeability values can vary according to the model and boundary conditions used in the calculation. To determine permeability, Darcy's law inherently assumes constant fluid properties. Modifications of Darcy's law (Muskat, 1937) take into account differences in viscosity and density. Thus, permeability can be determined with fluids other than water as long as these fluids are nonreactive. Because permeability is often measured with gases, two additional boundary conditions—gas slippage and inertial effects—must be considered.
Gas Slippage and Inertial Effect
When gas is used to determine permeability at low mean pressure, the resistance to flow from drag is very low, resulting in "gas slip conditions." Consequently, permeability
212
PART 5—LABORATORY METHODS
STEADY-STATE
UNSTEADY-STATE (PRESSURE DRAWDOWN)
250
200
a>
^ 150
о
CL
100
50
Time, Seconds
Figure 4. Schematic diagram of (a) steady-state and (b) unsteady-state apparatus.
Figure 5. Typical pressure drawdown plot. (Modified from Jones, 1972.)
calculated from Darcy's law will be too high and must be corrected using the Klinkenberg (1941) model. When gas permeability is corrected for slippage effects at the fluid/pore wall interface, it is called equivalent, nonreactive, liquid permeability or Klinkenberg permeability.
At high flow rates, gas flowing through porous media accelerates at pore throats and decelerates in pore bodies, giving rise to what is called inertial effects. Non-Darcy flow has been described by Forchheimer (1901), who presented modifications.
where ka = air permeability, md pa = atmospheric pressure, atm P1 = upstream pressure, atm P2 = outlet pressure, atm L = length, cm |i = air viscosity, cP qa = gas flow rate at atmospheric pressure, cm3/sec A = cross-sectional area, cm2
and where
Laboratory Methods for Permeability Determination
Liquid and gas permeability can be determined on core samples in the laboratory. However, gas permeability is determined most frequently because sample preparation is simplified and the analytical procedure is fairly rapid. Two methods currently exist for gas permeability determination: steady-state and unsteady-state.
Gas Permeability by Steady-State Method
A simplified schematic d i a g r a m of the steady-state apparatus is shown in Figure 4a. The apparatus includes a pressurized gas cylinder, a Hassler core holder, and a flowmeter. The apparatus is designed to ensure that no restrictions exist in flow lines that could cause a pressure drop between the core face and the pressure gauges. To determine air permeability, a clean, dried core sample is first placed in the core holder and pressure is applied to the rubber sleeve to seal it to the core. Air is then injected at a constant pressure until gas production rate and pressure stabilize. The pressure differential between the two ends of the core and flow rate are recorded for permeability calculation using the integrated form of Darcy's law for a compressible fluid. Thus,
k = 2000palxqaL
" (P1-P2M
Vaa
=^-, A
cm/sec
Measurements are usually made at several gas flow rates to ensure that flow conditions satisfy Darcy's law. In practice, gas permeability is calculated from the slope of the plot of Va versus (P12 - p22)/L, which results in a straight line passing through the origin as long as the conditions for Darcy's flow are maintained. The steady-state method has been the industry standard for many years because it is a convenient technique and the equipment is easy to operate.
Gas Permeability by Unsteady-State Method
Aronofsky (1954) has discussed the theory of transient p e r m e a b i l i t y m e a s u r e m e n t s , a n d the d e v e l o p m e n t of transient state permeameters has been discussed by Wallick and Aronofsky (1954), Champlin (1962), Morris (1953), and Jones (1972).
A schematic diagram of the unsteady-state Klinkenberg permeameter (Jones, 1990) is shown in Figure 4b. The permeameter works on the principle of transient analysis of pressure pulse decay in which Klinkenberg permeability is determined as a function of gas (ideally helium) pressure decay. This equipment consists of a reference cell of known volume that charges the core sample with gas. A downstream valve vents the gas pressure, and pressure change as a function of time is recorded. A typical pressure drawdown
Permeability
213
Table 1. Comparison of Steady-State and Unsteady-State Techniques
Steady-State
Industry standard for 30 years
Unsteady-State
Determines more representative permeability Zf00 instead of Zfair at reservoir conditions
Convenient to use
Enhanced accuracy results from measurement of pressure versus time instead of rate
Permeability is determined at low confining pressures
Measures additional reservoir description parameters: (3 and b
Develops practical link with historical data
Jil = liquid viscosity, cP ^1 = gas flow rate at atmospheric pressure, cm3/sec A = cross-sectional area, cm2 Ap = pressure drop, atm
As a quality control device, absolute permeability calculated for water or oil should agree with the Zc00 calculated for gas permeability.
Liquid Permeability by Unsteady-State Method
A technique based on pulse decay analysis (Amaefule et al., 1986) has been developed recently to determine effective permeability to liquid for low quality reservoir rocks. The authors reviewed computational techniques and experimental protocols for liquid permeability determination. A technique that allows the s i m u l t a n e o u s d e t e r m i n a t i o n of liquid permeability and compressibility was also developed. A detailed discussion of this technique is beyond the scope of this chapter, therefore, interested readers are referred to the paper by Amaefule et al.(1986).
plot (Jones, 1990) is shown in Figure 5. Advantages of the unsteady-state method include the ability to determine simultaneously (from Figure 5) the Klinkenberg permeability (Zcoo), helium slippage factor ((3He), and the inertial coefficient (P). A c o m p a r i s o n of the s t e a d y - s t a t e m e t h o d to the unsteady-state method is presented in Table 1.
Liquid Permeability by Steady-State Method Water or oil permeabilities are determined on core samples
after first obtaining the permeabilities to gas on the extracted dry samples. The sample is saturated with the test fluid, it is placed in a Hassler cell, and pressure is applied to the sleeve. Filtered, gas-free liquid is flowed through the sample, and the rates and pressure drops are measured. Permeability to liquid is calculated as follows:
, _ lOOO^frL
where Zr1 = liquid permeability, md L = length, cm
Permeability Averaging and Uncertainty Determination
It is necessary to average permeability determined for each pay zone to obtain permeability distribution. The most commonly used method to average horizontal permeability is the arithmetic average. Comparison of core permeabilities shows that arithmetic average permeabilities values generally agree with well test permeabilities.
Systematic and/or random errors may affect the accuracy of permeability d e t e r m i n e d f r o m any m e t h o d , w h e t h e r laboratory core or well test analysis. Uncertainty in the models used for permeability determination and input variables can result only in r a n d o m errors if the s a m e analytical technique, equipment calibration, and quality control scenario are considered. Amaefule and Keelan (1989) have shown that random errors can be addressed through stochastic modeling in which uncertainty can be assigned to the independent variables by multiple measurements and statistical calculations. Typically, accuracy of measured permeabilities decline at low and high values and are usually within ±5% (Keelan, 1971).
Core-Log Transformations and Porosity-Permeability Relationships
Walter A. Nagel1
Keith A. Byerley
Marathon Oil Company Littleton, Colorado, U.S.A.
CORE-LOG TRANSFORMATIONS
Core and wireline log analysis provide the means for e v a l u a t i o n of r e s e r v o i r p o t e n t i a l . P r o p e r core-to-log transformations are required to ensure that parameters used for quantitative log analysis are reasonable and that data from both sources are mutually supportive.
Depth Control
Agreement between log depth and core depth is essential. Generally, core depths are adjusted relative to log depths. Core gamma ray is commonly used for depth control and can be of great assistance only if there is sufficient gamma ray activity in the core to provide precise boundaries. Should closely spaced core porosity measurements be available, a graphical approach can be taken. Figure 1 shows a technique in which measured core porosity is plotted together with the bulk density log. Core porosity is shown in a spiked form, with bulk density overlain as the continuous curve. The
intervening no-data points have zero values. This spiked presentation offers a ragged edge that can be registered to the bulk density curve. The ragged edge offers more coherency to the eye than discrete points.
Figure 2 shows the same information after depth adjustments have been made. Depth registration to sharp lithology breaks, which are reflected in both log responses and core, should be used to confirm the shifts as previously determined. In addition, borehole image logs can provide excellent control for core depth and orientation. (For more on core-log depth adjustment, see the chapter on "Preprocessing of Logging Data" in Part 4.)
Spatial Resolution
The spatial r e s o l u t i o n of core i n f o r m a t i o n m u s t be considered relative to that of the formation itself and to the vertical resolution and depth of investigation of the logging responses to be used in the analysis. Core has physical dimensions that can be discretely measured. Should the
Bulk Density— 0 Core —
13700
Bulk Density
13800
0 Core
Figure 1. Plot showing unshifted core porosity in spiked format, together with bulk density log.
Figure 2. Plot showing shifted core porosity, together with bulk density log.
1Retired.
214
Core-Log Transformations and Porosity-Permeability Relationships
215
10000
1000
100
k, md 10
1.0
CORE PLUG REPRESENTS
2.4 in3
o.i
0.01
ф,%
Figure 4. Air permeability versus helium porosity from North Sea sandstone, conglomerate, and shale.
•8" BOREHOLE
Figure 3. Comparison of the volume of formation measured by the bulk density tool relative to that obtained from a core plug.
formation possess attributes that exceed these dimensions, precise prediction of those attributes cannot be expected. An example would be porosity obtained from l-in.-diameter core plugs taken from a conglomeratic reservoir, where the clast size can exceed 6 in. or more. Any attempt to determine an average porosity from conventional core analysis would need to consider the sampling problem. In this case, it may be more appropriate to use information contained in the log responses for obtaining a true measure of porosity.
If one considers the zone of investigation of the formation density tool in an 8-in. borehole (Figure 3), the volume of rock seen by the tool encompasses roughly one quadrant, with a nominal penetration depth of 4 in. and a thickness of 24 in. This equates to more than 900 in.3 of material measured. When this is compared to a 1-in. core plug 3 in. long, the log measures almost 400 times the volume measured by the plug.
Core samples will offer a sound representation of the formation when features of the rock are contained in the sample, as would be the case for a homogeneous sandstone. Even then, however, corrections may need to be applied to the core data, such as accounting for net overburden conditions.
When relating core information to log responses, one should be aware of the scale of measurement being used. It may be that independently derived log results are more accurate than either core analyses alone or transformations developed through core-to-log relationships.
Interpolation and Filtering
Since core measurements are usually taken at irregular spacings, comparisons to regularly spaced log data require some scheme for infilling. This is usually accomplished t h r o u g h linear i n t e r p o l a t i o n of porosity, density, and logarithm of permeability data. Decisions must be made regarding the distance that the interpreter is willing to accept for bridging no-data intervals, with data variability dictating the choice.
Once interpolated, the vertical resolution of the log and core measurements must be made comparable. Filtering the fine resolution of the core data to the coarser resolution of the log response is normally the route taken. The type of core sample—whether plug, sidewall, or whole core—must be balanced with the log measurement, each with its own vertical resolution. Filtering suppresses the detailed information available from core analyses. However, the objective is to move from small sample physical measurements (core) to relatively large sample remote measurements (logs) so as to make meaningful comparisons between the two data sources.
Core-to-Log Comparisons
Comparison of log results with those from core can take several forms, with overlay plots versus depth, crossplots, and histograms used most typically. Garner (1985) and Wilson and Hensel (1978) give several examples of overlay plots.
POROSITY-PERMEABILITY RELATIONSHIPS
Porosity-Permeability Crossplots
Attempts are often made at finding a relationship between permeability (Zc) and porosity (ф) by making a semilog plot of
216
PART 5—LABORATORY METHODS
1000 100 10
k, md 1
0.1
Con glom erate
• df
• •
P*
• 4 , °o
с
/
O
/
I • JZ
A LJTl/СП & ;Ж д
hf д
чЖ A v r \ Pк ЛД
ж
t
ч 4 Sa Tdsto пе
AVERAGE PARTICLE SIZE
* > 100 u.
о 20-100/x /
/—500M-
0.01
0.001
0 2
16
• cgl plugs
о cgl +ss Ии
• cgl +ss
сСЛ. !<лт>
ж cgl+ss . ОС
д ss plugs
10 12 14 16 18 20
Ф,%
IO
20
INTERPARTICLE POROSITY, X
Figure 6. Permeability versus porosity for various size groups in uniformly cemented nonvuggy rocks. (Taken from Lucia, 1983.)
Figure 5. Slip-corrected permeability (run under stress) versus helium porosity from McArthur River Field, Alaska.
these data from core (Figure 4). Such a relationship can permit permeability estimation over intervals where only core porosity or log porosity information is available. Such estimates of permeability based on regression against porosity alone can be extremely tenuous due to large scatter in the data. The method ignores other rock properties that also influence permeability. For carbonate reservoirs that may contain vugs and fractures, there is often no recognizable relationship.
Sometimes distinct Ar—ф trends can be hidden when rocks with different properties are lumped together. If it is possible to isolate core samples that have similar rock properties, apart from porosity, then /г-ф relationships can be more readily observed (Figure 5). Often these statistical relationships are used to establish productive intervals via a porosity cutoff. For data in Figure 5, a m i n i m u m permeability of 1 m d corresponds to a porosity cutoff of 5.3% in conglomerate and 10.7% in sandstone.
It should be pointed out that there is no theoretical justification for expecting a linear trend on semilog plots of к v e r s u s ф. In fact, t h e o f t e n - c i t e d K o z e n y e q u a t i o n (appropriate for describing porous media, such as filter packs, in which the grains are spherical, uniform in size, and unconsolidated) suggests that a log-log plot of к versus ф would be more appropriate.
In general, for a given porosity, the larger the grain size, the higher the permeability. This is illustrated in Figure 6, where porosity and permeability data from nonvuggy
carbonate rocks have been segregated according to grain size. When using porosity from core and logs, differences in
these measurements must be taken into account. In theory, core measurements provide effective porosity because claybound water and nonconnected pores should be excluded in helium flow measurements. However, in practice, core heating (and even air drying) can collapse clays, resulting in a porosity that is higher than the effective porosity. Conversely, porosity obtained from neutron and density logs is total porosity, which includes bound water associated with clays, microporosity, and porosity from fractures and isolated vugs. Differences in these two porosity sources must be reconciled and accounted for when using porosity from logs and к-ф relationships based on core. (For more on porosity types and measurements, see the chapter on "Porosity" in Part 5.)
Permeability measured under simulated formation stress using formation brine is usually best. Routine air permeability data are almost always optimistic, particularly at low permeability. Correction for slip, net confining stress, and any rock-fluid interaction (should it exist) is necessary. The resulting permeability is the so-called absolute permeability, which applies to a single fluid occupying the pore space. When making comparisons with permeability derived from drill stem tests, where usually more than one fluid is present, relative permeability effects need to be considered. (For more on permeability factors and measurements, see the chapter on "Permeability" in Part 5.)
Multiple Regression
In view of the previous discussion, it should be clear that porosity is just one of several rock properties that have an influence on permeability. To the extent that some of these other properties are contained in well log measurements, it
Core-Log Transformations and Porosity-Permeability Relationships
217
seems reasonable to perform a multiple regression against all logs that may have permeability information associated with them. Wendt et al. (1986) and Allen (1979) have used multiple regression, together with log and core data, to predict permeability. It is important to realize that these empirical equations for permeability prediction should always be considered as local in nature. They should only be applied when rock characteristics match those of the control region.
Uncertainty in Predictions
Prediction of permeability from well logs is far from an exact science. This is aggravated by the fact that ground truth is normally assumed to be core measurement—the accuracy of which varies significantly depending on procedure and how well samples represent the actual formation. To put the uncertainty in perspective, prediction of permeability within a factor of two of the true value is considered very good.
Wettability
Jim Funk
Texaco E &P Technology Division Houston, Texas, U.S.A.
INTRODUCTION
Wettability significantly affects a variety of measurements critical to describing oil reservoirs, from residual saturations to resistivity indices. In spite of the importance attributed to wettability, no standardized technique exists that completely and adequately characterizes the phenomenon. The most c o m p l e t e r e v i e w of the subject is given by A n d e r s o n (1986a-f).
Physically, wettability represents a balance of forces that occur at the interface between three phases, one of which is a solid. The equation describing this balance was first developed by Young in 1805 (Adamson, 1982). For an oil, water, and solid system, the equation would be
SOS WS + Oow cos Qc = 0
(1)
where Gos = interfacial energy between oil and solid Ows = interfacial energy between water and solid o()W = interfacial tension between oil and water
Qc = contact angle
Wettability is determined in three ways:
1. Measurement of the contact angle, or the product of the contact angle and the interfacial tension (the last term in Young's equation)
2. Measurement of the relative amounts of oil and water displaced under similar conditions
3. Observations of displacement or surface phenomena associated with the water or oil phase
The first method is the most difficult, while the second is the most common. The third is the most varied in terms of methods and results.
generally made on a polished surface that simulates the reservoir material. For sandstones, glass slides or polished quartz are often used. Polished marble is usually chosen to s i m u l a t e a c a r b o n a t e reservoir. An a d a p t a t i o n of the technique uses a d r o p of liquid confined b e t w e e n two surfaces (Craig, 1971).
Contact angle measurements can be precise, but even for ideal systems, measurements can show significant variation. Variations are related to surface preparation, equilibration of the solid and liquids, and surface roughness.
Utilization of a surface balance provides information on the product of the interfacial tension and the contact angle (Teeters et al., 1989). This technique is quantitative and flexible but subject to some of the same limitations as those of contact angle measurements.
Displacement Techniques
Measurements that use larger samples are usually employed for reservoir materials that show varying composition, mineralogy, and structure. Two procedures most commonly used are the Amott (Amott, 1959) and the U.S. Bureau of Mines (USBM) method (Donaldson et al., 1969).
Amott Method
The Amott method uses four basic measurements:
1. Amount of oil spontaneously imbibed 2. Amount of oil forcibly imbibed 3. Amount of brine spontaneously imbibed 4. Amount of brine forcibly imbibed.
TECHNIQUES
Contact Angle Measurements
If a liquid wets a surface, it tends to spread and cover that surface. Observed on a microscopic scale, the edge of the liquid has a characteristic shape. A knife edge shape indicates wetting, while a beaded edge shape indicates nonwetting. This is shown quantitatively in Figure 1, which shows a drop of water surrounded by oil and contacting a solid surface. If the edge of the d r o p forms an acute angle (0C < 90°), the surface is considered to be water wet. If the edge meets at an angle greater than 90°, the surface is oil wet. (For more on contact angle, see the chapter on "Capillary Pressure" in Part 5.)
Several techniques used to measure this angle are described by Adamson (1982). The measurements are
Figure 1. Wettability of oil, water, and rock system. (AfterRaza et al., 1968.)
218
Wettability
219
ratio from the displacement by water ratio. Figure 2 shows these volumes, and Equation (2) gives the calculation:
Index = BC/BD-DE/DB
W
The r a n g e of values and the c o r r e s p o n d i n g wettability characteristics are listed in Table 1.
USBM Method
The USBM method developed by Donaldson et al. (1969) uses the same types of data, but considers the work required to do the forced displacement. This requires calculating the area under the capillary pressure curve obtained during the forced displacement. Generally, the capillary pressure displacement is done by centrifuging, but other capillary displacement techniques can be used.
Wettability (W) of the sample is determined by comparing the log of the area (A1) under the oil-displacing brine curve with the log of the area (A2) under the brine-displacing oil curve. The USBM index is defined by the following equation:
W=Iog^
(3)
A2
Figure 2. Combined Amott and USBM method. (After Anderson, 1986b; from Sharma and Wunderlich, 1985.)
For spontaneous processes, the sample is submerged in the fluid to be imbibed and the displaced volume of nonimbibed fluid is measured. The forced processes involve either centrifuging the sample in the imbibing fluid or flowing the imbibing fluid through the sample and measuring the volume of the nonimbibed fluid displaced.
The displacement volumes (both spontaneous and total) are m e a s u r e d for both oil a n d water. The ratio of the spontaneously displaced volume to the total displaced volume is calculated for both the oil and the brine phases.
If a s a m p l e s p o n t a n e o u s l y i m b i b e s only b r i n e , it is considered water wet. Similarly, if it imbibes only oil, it is considered oil wet. If the s a m p l e imbibes neither, it is described as neutrally wet.
A modification of the test in general use first prepares the sample by centrifuging it in brine. This is followed by centrifuging in oil to irreducible water saturation. The Amott procedure is then followed, but a combined index—the Amott-Harvey wettability index (Boneau and Clempett, 1977)—is calculated by subtracting the displacement by oil
The Amott and the USBM methods can be combined into a single test (Sharma and Wunderlich, 1985). Tliis combination and the capillary pressure plot used in the USBM calculation, are shown in Figure 2.
Comparison of the Two Methods
Both the Amott and the USBM test are commonly used in the oil industry, but direct comparisons of the two techniques show only minimal correlation (Crocker and Marchin, 1986). The most significant deviations occur near the neutral wettability region. The Amott method is more sensitive in this area and may be a better indicator. The Amott method can also be used to indicate mixed wettability if a sample spontaneously imbibes both oil and water.
Additional Techniques
Several techniques are described in the literature for determining wettability (Anderson, 1986b). Some are modifications of the Amott or the USBM methods, while others represent significant departures from the standard techniques. These vary from microscopic examination of imbibed fluids to measurement of nuclear magnetic resonance (NMR) longitudinal relaxation. Table 2 lists some of these techniques and their observed variables.
220
PART 5—LABORATORY METHODS
Table 1. Approximate Relationship Among Wettability, Contact Angle, and the USBM and Amott Wettability Indexes
Contact angle Minimum Maximum
USBM wettability index Amott wettability index
Displacement by water ratio Displacement by oil ratio Amott-Harvey wettability index
Water Wet
0° 60 to 75° W near 1
Positive Zero 0.3 to 1.0
Neutrally Wet
60 to 75° 105 to 120° W near 0
Zero Zero -0.3 to 0.3
After Anderson (1986b).
Oil Wet
105 to 120° 180° W near-1
Zero Positive -1.0 to -0.3
Table 2. Alternative Wettability Measurement Techniques
Method
Microscopic examination Flotation method Glass slide method Relative permeability method Reservoir logs Nuclear magnetic resonance Dye adsorption
Observation
Visual examination of the fluid surrounding grains Distribution of grains at water/oil interface or air/water interface Displacement of the nonwetting fluid from a glass slide Location and relative magnitudes of Zcro and Zcrw curves Resistivity logs before and after injection of a reverse wetting agent Changes in the longitudinal relaxation time Adsorption of a water soluble dye
Capillary Pressure
Charles L. Vavra
John G. Kaldi
ARCO Oil and Gas Co. Piano, Texas, U.S.A.
Robert M. Sneider
Robert M. Sneider Exploration, Inc. Houston, Texas, U.S.A.
INTRODUCTION
Capillary pressure concepts can be used by geologists, petrophysicists, and petroleum engineers to evaluate the following:
• Reservoir rock quality • Pay versus nonpay • Expected fluid saturations • Seal capacity (thickness of hydrocarbon column a seal
can hold before it leaks) • Depth of the reservoir fluid contacts • Thickness of the transition zone • An approximation of the recovery efficiency during
primary or secondary recovery.
Evaluating capillary pressure of potential reservoir and seal rocks is important because capillarity controls the static distribution of fluids in the reservoir prior to production and remaining hydrocarbons after primary production.
I Oil I I I l
• (a)
N
I i Illli
-Л! Water'
RockVV;/^^
Л
Kb)
ЙШ Mercury ]::®:;:;:
Sv
Figure 1. Effects of interaction of adhesive and cohesive forces on wettability, (a) If adhesive forces are greater than the cohesive forces, the fluid spreads out on the surface and is termed wetting, (b) If cohesive forces exceed adhesive forces, the liquid beads up and is termed nonwetting. The measure of relative wettability is the contact angle (6), which is measured through the denser phase.
CAPILLARY PRESSURE CONCEPTS
Capillary pressure results f r o m interactions of forces acting within and between fluids and their bounding solids. These include both cohesive forces (surface and interfacial tension) and adhesive (liquid-solid) forces. When adhesive forces are greater than cohesive forces, the liquid is said to be wetting (Figure la). When cohesive forces exceed adhesive forces, the liquid is nonwetting (Figure lb). The relative wettability of the fluids is described by the contact angle (0), which is the angle between the solid and the fluid-fluid interface as measured through the denser fluid (Figure 1). (For information on the measurement of wettability, see the chapter on "Wettability" in Part 5.)
If the end of a narrow capillary tube is placed in a wetting fluid, net adhesive forces draw the fluid into the tube (Figure 2). The wetting phase rises in the capillary above the original interface or free surface until adhesive and gravitational forces are balanced. Because the wetting and nonwetting fluids have different densities, they also have different pressure gradients (Figure 2). Capillarу pressure (Pc) is defined as the difference in pressure across the meniscus in the capillary tube. Put another way, capillary pressure is the amount of extra pressure required to force the nonwetting phase to displace the wetting phase in the capillary. Capillary
kuPnw
1 Non-wetting III
Phase
Pw
height (h)
Free Surface
Wetting:; Phase S
Pressure
\ Pc = Pnw - Pw \ Pc \
- Pc = 0- -V- Free
Surface
IWV V-o
-Sl
\w
Vl
k\
Figure 2. The wetting phase rises above the original or free surface in the capillary tube experiment until adhesive and gravitational forces balance. Capillary pressure (Pc) is the difference in pressure measured across the interface in the capillary (Pc = P n w - P w ) - This pressure results from the contrast in pressure gradients caused by the different densities of the nonwetting (pnw) and wetting (pw) phases (right).
221
222
PART 5—LABORATORY METHODS
pressure can be calculated as follows:
Pc = ( P w - P n w ) ^ ' o r 2(7 COS в
where pw = density of the wetting nonwetting fluid P n w = density of the nonwetting fluid g = gravitational constant h = height above the free surface о = interfacial tension 0 = contact angle between the fluids and the capillary tube rc = radius of the capillary
These equations show that capillary pressure increases with greater height above the free surface and with smaller capillary size.
The importance of capillary pressure in reservoir studies is that many reservoir rocks can be approximated by a bundle of capillaries, with formation water being the wetting phase and hydrocarbons the nonwetting phase. As hydrocarbons begin to migrate into a rock, displacing the pore water, the hydrocarbons first enter the pores with the largest pore throats (capillaries), leaving the wetting phase (water) in the pores with smaller throats or in small nooks and crannies (surface roughness). As the hydrocarbon column increases, the height above the surface where Pc = 0, called the free surface or free water level (FWL), becomes greater and the capillary pressure increases, allowing hydrocarbons to enter pores with smaller and smaller throats. This process continues until one of several things occurs: (1) generation or migration ends, (2) the trap reaches its spill point, or (3) capillary pressure is sufficient to force hydrocarbons into the seal (displacement pressure is exceeded), allowing the seal to leak.
Reservoir capillary pressure relationships can be evaluated by using either of the following:
1. Porous plate or centrifuge method, which uses the actual or simulated hydrocarbon-brine system of the reservoir to approximate the wetting properties
2. Mercury injection, which simulates the premigration wetting characteristics of the reservoir
The mercury injection method is recommended to evaluate the initial static distributions of reservoir fluids prior to production because these distributions are typically controlled by the premigration wetting characteristics. Mercury injection is also favored because it is simpler, cheaper, and less time consuming than porous plate or centrifuge methods. In addition, mercury injection can be conducted on cuttings or sidewall samples. Mercury injection capillary pressure data can also be converted to the reservoir f l u i d s y s t e m . A l t h o u g h this i n v o l v e s a n u m b e r of assumptions on wettability and the effects of cleaning and extracting the core are largely ignored, the cost and time benefits typically favor mercury injection.
(% pore volume)
Figure 3. Mercury injection capillary pressure curve terminology.
TERMINOLOGY AND MERCURY INJECTION SYSTEMATICS
Mercury injection capillary pressure data are acquired by injecting mercury into an evacuated, cleaned, and extracted core plug. Mercury injection pressure is increased in a stepwise manner, and the percentage of rock pore volume saturated by mercury at each step is recorded after allowing sufficient time for equilibrium to be reached. The pressure is then plotted against the mercury saturation (Figure 3), resulting in the injection curve (which is also called the drainage curve because the wetting phase is being drained from the sample).
The pressure at which mercury first enters the sample (after the mercury has filled any surface irregularities on the s a m p l e ) is t e r m e d the displacement pressure (Pd)- The percentage of pore v o l u m e saturated by m e r c u r y at the maximum pressure is the maximum saturation (Smax). The unsaturated pore volume at that pressure is the minimum unsaturated pore volume (wmin) (Figure 3). This is sometimes incorrectly referred to as irreducible saturation. This term is inappropriate for the air-mercury system because saturation d e p e n d s on applied pressure and on the duration of the experiment (Wardlaw and Taylor, 1976).
After the maximum pressure is reached, the pressure is reduced in steps and air (the wetting phase) is allowed to imbibe into the sample. The amount of mercury expelled from the sample at each pressure is expressed as a percentage of total pore volume or bulk volume. Again, pressure is plotted against mercury saturation in the withdrawal curve (Figure 3). The volume of pore space still saturated with mercury after pressure is reduced to the minimum is called the residual mercury saturation (SR).
Data from the mercury injection curve can be used to approximate the distribution of pore volume accessible by throats of given effective radii (in cm) using the following
Capillary Pressure
223
Table 1. Pore Geometry Factors Affecting Recovery Efficiency (RE)
Factor
Pore throat heterogeneity Pore body to pore throat
size ratio Coordination number
High RE Low
Low High
Low RE High
High Low
(% pore or bulk volume)
Figure 4. Idealized mercury injection capillary pressure curve shapes. Note that all of the curves have identical displacement pressures and minimum unsaturated pore volumes, but that the saturation profiles would differ dramatically due to differences in pore throat size distributions.
equation:
2ocos0
Withdrawal efficiency in the air-mercury system is controlled exclusively by pore geometry, which can be interpreted from the capillary pressure curves and by direct observation of the rock or an epoxy resin pore cast. Critical pore geometry factors affecting withdrawal efficiency include effective pore throat heterogeneity; the ratio of pore body to pore throat size (rb/rt); and the number of throats connecting each pore (coordination number). Table 1 shows the effect of these parameters on withdrawal efficiency.
Relating withdrawal efficiency in the air-mercury system to recovery efficiency in the hydrocarbon-water system is dependent on properties of the fluids as well as properties of the pore system. Fluid properties that affect recovery include viscosity, density, interfacial tension, wettability, contact angle hysteresis, and rate of displacement (Wardlaw and Cassan, 1978). Nevertheless, for a given range of fluid properties in a water wet reservoir, the same pore geometry factors that contribute to increased mercury withdrawal efficiency also increase hydrocarbon recovery efficiency.
where о = interfacial tension of the air-mercury system (480 dynes/cm) 0 = air/mercury/solid contact angle (140°) Pc = capillary pressure in d y n e s / c m 2 (1 psi = 69035 dynes/cm2)
Note that this equation calculates the radius of a cylindrical capillary tube; however, real pore throats have very complex geometries. Therefore, the calculated values represent the effective radii of the pore throats, which may not equal their actual dimensions. Samples dominated by throats of similar size (well sorted) have broad, flat plateaus (Figure 4). As sorting becomes poorer, the plateau steepens then disappears and the slope of the curve approaches 45° (unsorted).
Data from the mercury withdrawal (air imbibition) curve can provide information regarding the efficiency with which the nonwetting phase can be withdrawn from the pore system. Withdrawal efficiency (We) is defined as the ratio of the mercury saturation in the sample at minimum pressure after pressure is reduced (Sr) to the saturation at maximum p r e s s u r e (SMAX) (Figure 3) (Wardlaw and Taylor, 1976). Because few samples reach 100% mercury saturation at routinely available injection pressures, data are normalized by the following equation:
W
Smax-S* x l Q 0 %
SJmax
RESERVOIR APPLICATIONS
Conversion from Air-Mercury to Brine-Hydrocarbon System
Before mercury injection data can be applied to reservoirs, the data must be converted from the air-mercury system to the b r i n e - h y d r o c a r b o n system of the reservoir using the following relationship:
p
=p
„ qb/hc cosOb/hc
Сь/ьс Ca/m o a / m c o s 0 a / m
where PCb/hc = capillary pressure in the brine-hydrocarbon system of the reservoir Pca = capillary pressure in the air-mercury system a b / h c = interfacial tension of the brine-hydrocarbon system o a / m = interfacial tension of air-mercury system 0 b / h c = contact angle of the brine-hydrocarbon-solid system 0 a / m = contact angle of the air-mercury-solid system
Ideally, values for brine-hydrocarbon contact angle, surface tension, and fluid densities at reservoir temperature and pressure should be determined in the laboratory using actual reservoir fluids. However, these measurements are difficult and expensive, so approximations such as those given in Tables 2 and 3 are commonly used.
224
PART 5—LABORATORY METHODS
Table 2. Typical Contact Angle and Surface Tension Values
System
Air-brine Air-mercury Crude oil-water
Contact Angle (0)a
0° 140°
0°
aContact angle measured on quartz plate during drainage.
Interfacial Tension (o) (dynes/cm)
72 480
35
Height Above the Free Water Level
Having converted the capillary pressure data to the hydrocarbon-brine system of the reservoir and knowing the hydrocarbon and brine densities from laboratory analysis, we can n o w calculate the a m o u n t of h y d r o c a r b o n c o l u m n required (height above the free water level or level where Pc = 0) to attain a pressure of interest:
S
0.433(pb-phc)
where h = height above the free water level (in ft) Pc = hydrocarbon-brine capillary pressure (in psi) pb = specific density of brine at ambient conditions (in gm/cm3)
phc = specific density of hydrocarbons at ambient conditions (in gm/cm3)
0.433 = the gradient of pure water at ambient conditions
This information can be used to compare expected fluid saturations of different rock types at given levels in the reservoir (Figure 5). (For information on the application of capillary pressure data to evaluating fluid contacts, see the chapter on "Huid Contacts" in Part 6.)
Evaluating Seal Capacity
If the capillary pressure data from the suspected seal are available or can be estimated, the maximum hydrocarbon column each rock type could have before the seal begins leaking can be calculated by the following:
h
,
PdS-PdR
max 0.433(pw-phc)
where ^max = height of the hydrocarbon column (in feet)
Table 3. Typical Fluid Density Ranges
Fluid
Gas Oil Brine
Density Range (g/cm3)
0.00073-0.5 0.51-1.00 1.0-1.2
P d s = brine-hydrocarbon displacement pressure of the seal (in psi)
P d R = brine-hydrocarbon displacement pressure of the reservoir rock (in psi)
Additional Applications
Capillary pressure data can also be applied to help distinguish reservoir from nonreservoir and pay from nonpay (see the chapter on "Effective Pay Determination" in Part 6). Several workers have attempted to correlate capillary pressure data and brine or air permeabilities. Purcell (1949, 1950) related capillary pressures empirically to air permeability through the graphical integral of the curve of mercury saturation versus reciprocal capillary pressure squared. Swanson (1981) proposed a simple nomograph whose application improved estimation of brine permeability from capillary pressure measurements on sidewall cores and ditch cuttings.
Another type of mercury test involves injecting mercury to a saturation less than the maximum, withdrawing the mercury to some residual wetting phase saturation, and then reinjecting the mercury. This process, repeated several times to progressively higher maximum pressures, produces hysteresis loops. These loops, wherein mercury is partially withdrawn and then reinjected, can be used to investigate withdrawal efficiency at various initial saturations (Morrow, 1970; Melrose and Brandner, 1974; Wardlaw and Taylor, 1976; Wardlaw and Cassan, 1978; Wardlaw et a l , 1988). Results suggest that the higher the initial saturation of the nonwetting phase, the greater the withdrawal efficiency. Porosimetry uses hysteresis loops to interpret pore body and pore throat size distributions and their spatial arrangement (Dullien and Dhawan, 1974; Wardlaw et al., 1988). Pore sizes have also been evaluated with rate-controlled mercury injection (Yuan and Swanson, 1986).
It should be noted that although these more specialized procedures are quite informative, especially for approximating hydrocarbon recovery efficiencies, they are also relatively labor intensive and expensive when compared to routine mercury injection tests.
^500-
C> D CD
0) ^ :5 Q)
CL ф 250C4 Ш Ll-
— CD CL >
сa о
Capillary Pressure
225
Rock Type
BCD
250
CD
=3 CD
CO ¢/)
T3
CD
CL CD
Л LI-
= CD CL >
CO O
20
40
60
80
1
Wetting Phase Saturation
(% Pore Volume)
25 Water 75 Saturation ioo (% Pore Volume)
Figure 5. Effect of capillary pressure (left) on water saturation (right). At any given height above the free water level, water saturations vary widely among rock types (A-E) due to diffferences in capillarity. For example, at 50 ft above free water level, water saturations vary from 18% (rock type A) to 95% (rock type E). A well drilled into an interbedded sequence of these rock types would show multiple oil-water contacts and a highly irregular vertical saturation profile. Note also the wide transition zone in rock type B caused by poor sorting of the pore throats.
Relative Permeability
,effrey T-Hawkins
Conoco, Inc. Oklahoma City, Oklahoma, U.S.A.
INTRODUCTION
Relative permeabilities quantify the interaction between two or more fluids as they flow through porous media. Relative permeability data are one of the most important parameters for estimating reservoir performance, whether by classic reservoir engineering techniques or by sophisticated reservoir simulation. Heaviside et al. (1983), Hagoort (1984), Honarpour et al. (1986), and Honarpour and Mahmood (1988) all review aspects of relative permeability.
Relative permeability is a saturation-dependent function involving the ratio of effective permeability of a phase to a base permeability. Absolute permeability quantifies a rock's ability to transport a single fluid. Effective permeability quantifies the rock's ability to transport a fluid in the presence of one or more additional fluids. Three base permeabilities are commonly used for relative permeability data (Craig, 1971):
1. Absolute air permeability 2. Effective permeability to oil at irreducible water
saturation 3. Absolute water permeability
(For information on calculating air and liquid-water permeability, see the chapter on "Permeability" in Part 5.)
Experimental relative permeability data yield two intersecting curves that are concave upward when plotted on a linear scale. One curve is the relative permeability of the displaced phase, while the other is the relative permeability of the displacing phase. The curves span a saturation range from residual displacing phase saturation to residual displaced phase saturation. Example relative permeability curves are shown in Figures 1 and 2. Commonly measured relative permeabilities include the following:
Steady-State
In the steady-state method, fluids are injected into the core sample at a fixed ratio until pressure and saturation equilibrium are reached. Relative permeabilities are calculated directly from the flow rate and pressure drop data. Relative permeabilities at other saturations are obtained by adjusting the fluid ratio while holding total injection rate constant.
Unsteady-State
The unsteady-state method involves injecting a single fluid into a core sample at either constant rate or constant pressure while m e a s u r i n g the a m o u n t of fluid displaced and the pressure drop or rate, respectively, with time. Relative permeabilities are then calculated using the m e t h o d of Johnson et al. (1959), a reservoir simulation method (Archer and Wong, 1973), or a regression method (Sigmund and McCaffery, 1979). More recently, Watson et al. (1988) have developed an improved regression-based method that is believed to honor the experimental data more closely.
The actual displacement process in the reservoir is an unsteady-state displacement, thus the unsteady-state method more closely mimics actual reservoir fluid flow. Comparisons have been made in the literature of the two methods with contradictory results. Amaefule and Handy (1982) presented data showing a difference between steady-state and unsteadystate results, while Johnson et al. (1959) showed agreement between the two methods. Certain wetting situations may dictate the use of one technique over the other or a deviation from conventional testing procedures. These issues are discussed further in papers by Braun and Blackwell (1981), Heaviside et al. (1983), Heaviside et al. (1987), and Mohanty and Miller (1988).
1. Water displacing oil from a sample at irreducible water saturation
2. Gas displacing oil from a sample at irreducible water saturation
3. Oil displacing water from a sample at waterflood residual oil saturation
FACTORS AFFECTING MEASUREMENT
The measurement of relative permeability in the laboratory is affected by many parameters, including heterogeneity of the rock sample, test conditions, wettability, and saturation history.
MEASUREMENT METHODS
Two methods of relative permeability measurement are commonly used in the industry: steady-state and unsteadystate. The steady-state method is computationally easier but requires a more rigorous laboratory procedure. The unsteady-state method is easier from a laboratory standpoint, but is more difficult from a computational standpoint. Gas-oil relative permeabilities are normally measured by the unsteady-state procedure.
Sample Selection
Relative permeability analysis methods generally require homogeneous samples. Whole core is preferable for core flooding experiments, including relative permeability measurements. Core plugs can be cut from the whole core (usually the samples are 1.5 in. in diameter and as long as possible), or measurements can be made on the whole core itself. Generally, rotary drilled sidewall cores are inappropriate for relative permeability measurements because the small sample size makes obtaining accurate measure-
226
Relative Permeability
221
100
100
£
t
H-I
d
«
Ы
Oh
H 3
20
40
60
80
100
WATER SATURATION, %
Figure 1. Water displacing oil relative permeability curve for a water wet rock.
20
40
60
80
100
WATER SATURATION, %
Figure 2. Water displacing oil relative permeability curve for an oil wet rock.
ments difficult. Percussion sidewall cores are inappropriate because of changes in rock porosity and permeability during acquisition.
Test Conditions: Temperature and Confining Pressure
Relative permeabilities may be affected by the temperature at which the measurement is made. Edmonson (1965) showed a temperature effect on relative permeability, while Miller and Ramey (1985) saw no temperature effect. The use of live crude oil and reservoir temperature may not affect the shape of the relative permeability curve, but it can definitely affect the residual oil saturation in a waterflood (Hawkins, 1989).
In the reservoir, rock is subject to a net overburden pressure. This net overburden pressure is equal to the weight of the vertical column of rock and fluid m i n u s the pore p r e s s u r e of the rock. N e t o v e r b u r d e n p r e s s u r e a n d permeability have an inverse relationship—the higher the net overburden pressure, the lower the permeability. Relative permeabilities should be measured at a net overburden pressure comparable to the reservoir net overburden pressure.
Wettability
Wettability has been shown to have a dramatic impact on relative permeability (Labastie et al., 1980; Wendel et al., 1985; Anderson, 1986e; Wang, 1986) (see also the chapter on "Wettability" in Part 5). Figure 1 presents typical water displacing oil relative permeability curves for a water wet rock, and Figure 2 is for an oil wet rock. Note that the water relative permeability is higher in an oil wet rock; this is because in an oil wet rock, the water is in the center of the pores. By flowing through the center of the pores, the water
has a less tortuous path, hence a higher effective permeability. Table 1 outlines Craig's (1971) method for characterizing the wettability of rock from relative permeability curves.
Care must be taken when measuring relative permeabilities to ensure the rock has representative wettability. Two methods are currently employed: restoredstate analysis and native-state analysis. Native-state analysis assumes that the rock recovered from the reservoir has the appropriate wettability. Work by Sharma and Wunderlich (1987) and Yan et al. (1988), however, has shown that many drilling mud components significantly alter rock wettability. Thus, even freshly cut cores can have altered wettability and be unsuitable for measurement. A recently developed low invasion coring system (Tibbitts et al., 1990) may prevent invasion of drilling mud and thus alteration of wettability by the drilling mud. If core is recovered with little mud invasion a n d no a l t e r a t i o n of w e t t a b i l i t y , then r e p r e s e n t a t i v e measurements can be made.
Frequently, the core is received with altered wettability and restored-state analysis is required, as indicated by Cuiec (1975, 1977), Wendel et al. (1985), and Gant and Anderson (1986). Prior to restoration, the core is cleaned to a water wet state and then saturated with reservoir brine. The core is then flushed to irreducible water saturation with reservoir crude and then aged at reservoir temperature. The flushing of the core with the crude simulates the initial migration of crude oil into the reservoir. The restoration time varies, but eventually the sample comes to wettability equilibrium and the measurements are then made.
Saturation History
When measuring relative permeabilities, it is important to consider saturation history (Hawkins and Bouchard, 1989). Many researchers have observed relative permeability
228
PART 5—LABORATORY METHODS
Table 1. Characterization ot Wettability of Rock from Relative Permeability Curves
Irreducible water saturation
Saturation at which oil and water relative permeabilities are equal
Relative permeability to water at residual oil saturation
Water Wet >20%
>50% <30%
Oil Wet <15%
<50% >50%
hysteresis, including Braun and Blackwell (1981), Fulcher et al. (1983), and Torabzadeh and Handy (1984). In this instance, hysteresis means that relative permeabilities are different for a waterflood than an oil flood (flooding phase hysteresis). Relative permeabilities are sometimes different the second time a sample is water flooded than the first time it is water flooded (cycle-dependent hysteresis). Thus, laboratory relative permeability measurements should mimic reservoir flooding history. Waterflood relative permeability data should be measured on a first-cycle waterflood.
Paleontology
Robert W. Scott
Amoco Research Center Tulsa, Oklahoma, U.SA.
INTRODUCTION
Paleontology, the study of fossil organisms and their traces, has been used in the exploration for and exploitation of hydrocarbons since the later half of the nineteenth century. Because fossil assemblages change through time, they aid in the prediction of depth to reservoirs, to casing points, and to overpressured zones. Chronostratigraphic assemblages are the basis for correlating strata among wells, across basins, and between basins. Correlation is the method for predicting the lateral continuity and physical and chronostratigraphic equivalency of strata. Fossil assemblages are also evidence of the depositional environments of the drilled strata and thus play a role in predicting the location of reservoirs, source rocks, and pinch-out of porous strata.
AGE AND CORRELATION OF STRATA
Because fossil species evolved through time in genetically related lineages and because extinction events were followed by n e w assemblages of species, fossils provide the best criteria for dividing the geological record into time intervals characterized by the first-appearance datums (FADs) and last-appearance datums (LADs) of key species (Figure 1). Fossil tops and bases may be synchronous in sections where the strata record continuous deposition and where suitable environmental conditions persisted.
Key fossil datums in reference sections provide the basic division of the geological section into systems, series, and stages (Hancock, 1977). However, reference sections may not contain a complete record of sedimentation because the
160-161 FIELD OIL/GAS RESERVOIRS
PALEOECQLOGY
/актну NERITIC J
/VuZVm/'/
Figure 1. Example of paleontological data from a well showing fossil assemblages and interbedded paleoecology. Abbreviations for bioevents (paleotops): Glob flex = Globorotalia flexuosa, Trim A = Trimosina denticulata, Hyal B = Hyalinea balthica, Glob M = Globorotalia miocenica, Glob alt = Globoquadrina altispira. Paleoecological abbreviations: Bathy = bathyal, M = middle, U = upper, O outer, I = inner. (From Armentrout, 1987.)
229
230
PART 5—LABORATORY METHODS
contacts between many strata are unconformable. Therefore, the continuous span of time is divided into eras, epochs, and ages (Figure 2). The intervals of geological time are calibrated to absolute time by means of radiometric or isotopic ages from interbedded or cross-cutting rocks such as volcanic flows and ash beds.
In most wells, the LADs of fossils are the most useful datum planes for subdividing, dating, and correlating the lithostratigraphic section (Figure D because the drilling p r o c e d u r e m a y extend the FADs of fossils by caving of cuttings. However, in certain conditions, the LAD may be o v e r e x t e n d e d by r e w o r k i n g of the s p e c i m e n s above an unconformity, and the FAD may be in older rocks due to contamination from the drilling mud (Poag, 1977). The fossil top may also be depressed (or older) in a given well for a number of reasons: the strata with the uppermost part of the range may be eroded, environmental conditions prevented the species from living there, or the specimens may have dissolved. If a species is not abundant at the top of its range, it may be missed in drilling and sampling.
K n o w i n g the age a n d thickness of the strata enables prediction of depth to reservoir or casing points and depth to maturation of source rocks. For example, casing points are important for engineering decisions when drilling unconsolidated Plio-Pleistocene muds in the Gulf of Mexico and offshore Trinidad. Drilling stops when key fossils are encountered, and casing is set to prevent the hole from collapsing or to control high pressure zones that he deeper.
Fossil a s s e m b l a g e s also d e f i n e the p o s i t i o n of unconformities and the duration of hiatuses and may aid in the recognition of faults and the correlation of strata across faults.
DEPOSITIONAL ENVIRONMENTS AND FACIES ANALYSIS
Most fossil species r e q u i r e d specific c o n d i t i o n s of temperature, substrate, and surrounding medium to carry on their normal activities. When conditions differed from ideal, the species is not present or is not preserved. The types of sedimentary rocks deposited under specific environmental conditions are biofacies and are identified by their specific rock properties (including fossils) (Scholle and Spearing, 1982; Scholle et al., 1983).
Fossils most commonly used in hydrocarbon exploration are microfossils (generally smaller than about 2 mm) because they can be recovered from drill cuttings without much damage to the fossils. Different processing techniques separate calcareous, siliceous, phosphatic, and organic-walled fossils (Kummel and Raup, 1965; Feldman et al., 1989).
The major groups of marine fossils are nannoplankton (sometimes called coccoliths), foraminifera, radiolaria, diatoms, conodonts, ostracodes, palynomorphs (such as organic-walled dinoflagellates), spores and pollen, and various types of megafossils. Nonmarine facies may yield spores and pollen, and lacustrine facies may also contain ostracodes and diatoms. These and other fossils are described in the Encyclopedia of Paleontology (Fairbridge and Jablonski, 1979).
Figure 2. Example of geological time using the stratigraphic column of Cook Inlet basin, Alaska. Geological time table modified after van Eysinga (1975). Tertiary stages are from Wolfe (1977). (From Magoon and Claypool, 1981.)
The distribution of fossils in a particular basin depends on the age of the strata, biogeographic setting, bathymetry, depositional environment, lithology, and diagenetic events. K n o w l e d g e of these factors influences the decisions of whether to search for fossils in the drill cuttings or core and what fossil group to use.
Geological Age
The geological age of strata determines which fossils may be present. Nannoplankton first evolved in the Late Triassic. Planktic foraminifers evolved during the Late Jurassic and were very diverse during the Cretaceous and Tertiary. Diatoms are found in rocks as old as Early Cretaceous, and radiolaria have a spotty occurrence back to the Ordovician. Ostracodes and benthic foraminifers can be found in marine rocks as old as Ordovician. Phosphatic conodonts are locally abundant from Ordovician to Triassic strata. Of the organic-
SOUTH
SMOOT
CUTSHAW
GU. 2
ANHYDRITE
SMITH
^ REQUIENID PACKSTONEy MOL LUSK-MIL I O l I D ORBiroLINID W C K S L - —
CUTSHAW I
^ M^xmivj
Paleontology
231
NORTH
OWENS
FERRY LAKE RODE SSA
ioi^iiia
CO <
>
OC UJ
VERT. SCALE
Figure 3. Correlation of Rodessa Formation Iithofacies at Running Duke Field, Houston County, Texas. Cores are indicated by solid vertical lines. (From Scott, 1990.)
walled fossils, a variety of marine groups and spores are found throughout the Paleozoic. Marine dinoflagellates were common during the Jurassic. Fossil pollen is no older than Early Cretaceous. Megafossils, large benthic foraminifers, and calcareous algae can be found in most marine strata of the Phanerozoic.
Paleogeographic Setting
The g e o g r a p h i c setting of a basin d u r i n g d e p o s i t i o n influences the type of climate, mean annual temperature and rainfall, and oceanic and wind currents. Climatic conditions may exclude certain species from the basin, or they may favor the proliferation and abundance of a species group. Because of significant reorganizations of crustal plates during the Paleozoic, the p r e s e n t position of a basin m a y be quite different from its position during deposition.
Lithology
Lithology is one of the most useful criteria in selecting the t y p e s of fossils u s e d in solving exploration p r o b l e m s . Organic-walled fossils, such as spores, pollen, and dinoflagellates, are found in most argillaceous rocks, but they are rare in clean sandstone and limestone. Calcareous nannoplankton, foraminifers, conodonts, and ostracodes can be recovered from shales, sandstones, and many types of limestone. Siliceous radiolaria and diatoms occur in shale and deep water limestone.
Diagenesis
Where diagenetic conditions become acidic, calcareous fossils are normally dissolved. Deep, hot burial generally destroys most organic-walled fossils and most siliceous fossils. Ironstone, glauconite, a n d p h o s p h a t e casts of calcareous fossils may be recovered by normal processing procedures.
EXPLOITATION APPLICATIONS
Fossil assemblages may provide evidence for local environmental conditions that also influenced deposition of potential reservoirs and source rocks. For example, reefs are f o r m e d by a c o m m u n i t y of interacting species that live t o g e t h e r u n d e r the s a m e conditions. C o m m u n i t i e s of terrestrial organisms normally are transported and concentrated into fluvial or lacustrine deposits and are mixed with aquatic organisms. Marine organisms may also be transported, and species from different communities may be mixed by storms and other currents. (For more information on depositional environments, see the chapters on "Lithofacies a n d E n v i r o n m e n t a l A n a l y s i s of Clastic Depositional Systems" and "Carbonate Reservoir Models: Facies, Diagenesis, and Flow Characteristics" in Part 6.)
Reservoirs
An example of how fossil assemblages affect reservoir quality and distribution can be seen in the Lower Cretaceous
232
PART 5—LABORATORY METHODS
reef complexes of the Running Duke Field, Houston County, Texas. In these reef complexes, reservoir properties are partially controlled by the ancient communities of reefbuilding mollusks, corals, and algae (Figure 3) (Scott, 1990). The reefs built by corals, algae, and rudists formed shoal water areas where waves and other currents deposited oolites and well sorted bioclasts. Diagenetic conditions in this facies formed reservoir porosity. Recognition of the fossils in cuttings, cores, and thin sections complement petrographic and log analyses and aid in the recognition of reservoir quality rocks.
Source Rocks
Deposition, concentration, and preservation of organic matter to form source rocks require special environmental conditions. Fossils aid in the recognition of these unique
settings, which are primarily defined by geochemical and petrological parameters. The Cretaceous Mowry Shale in Wyoming, for example, is an important source rock and contaiiis from 1 to 5.2% organic carbon (Davis et al., 1989). The richest source rock is homogeneous pelagic mudstone that contains layers of siliceous radiolaria, kerogen, and fish debris. This mudstone was deposited in a restricted basin, where the water column was stratified and bottom waters were depleted in oxygen, allowing the preservation of organic matter. Anaerobic conditions are indicated by the presence of shallow water pelagic species and by the virtual absence of bottom-dwelling species or traces of animal activity. Lateral changes in the k i n d s and a b u n d a n c e s of fossils and sedimentary structures also give clues to the oxygen gradient in the basin. Nearshore bottom-dwelling biota are found only at the margins of the Mowry sea.
Thin Section Analysis
David W. Houseknecht
U.S. Geological Survey Alexandria, Virginia, U.S.A.
INTRODUCTION
Petrographic analysis of thin sections made from rocks within the productive interval of an oil or gas field provides unique information regarding reservoir quality, reservoir homogeneity, and in some cases, the potential for formation damage that can be caused by completion and/or stimulation procedures.
SAMPLE COLLECTION
Petrographic analysis can be performed on samples collected from conventional cores, rotary sidewall cores, percussion sidewall cores, or cuttings, but sample quality varies considerably.
Conventional cores are most desirable because samples can be collected from specific intervals of interest determined by a c o m b i n a t i o n of direct o b s e r v a t i o n of p h y s i c a l characteristics (such as facies), measured properties (such as porosity and permeability), and wireline log responses. They are also better because sample damage tends to be minimal. Thin sections are commonly made from porositypermeability plugs cut from conventional cores because petrographic observations can be compared directly to porosity and permeability measurements.
Sidewall cores are less desirable because sampling sites are remotely selected based on log responses and because sample size is small. This introduces the possibility that the sidewall cores are not r e p r e s e n t a t i v e of the interval of interest. Rotary sidewall cores are preferred because they have better recovery in the most porous parts of a reservoir and because they usually display relatively little damage induced by coring. Percussion sidewall coring often recovers little sample, particularly in the most p o r o u s parts of a reservoir, and commonly induces fractures and disrupted fabrics within samples.
Cuttings are the least desirable because they are the least site specific of the s a m p l e t y p e s a n d b e c a u s e they disproportionately represent the least p o r o u s parts of a reservoir. The most porous parts commonly disintegrate into constituent grains. Moreover, it is difficult to evaluate the homogeneity of reservoir characteristics because of the small size of cuttings.
SAMPLE PREPARATION
Once samples have been selected, impregnation and thin sectioning procedures are critical to successful petrographic analysis. Samples are impregnated with low viscosity epoxy introduced while the samples are under a vacuum. After vacuum impregnation, some laboratories apply pressure via an inert gas to force the epoxy into small pores. The epoxy is stained, usually blue, to facilitate observation of porosity once
thin sections have been completed. Epoxy can also be "stained" with fluorescent dye, either during impregnation or after thin sections are completed, to enhance observation of relatively small pores when thin sections are viewed under incident fluorescent light (Ruzyla and Jezek, 1987). Thin sections must be carefully ground to final thickness (usually 30 (im) to avoid fracturing and plucking.
At this stage, samples can be stained for specific minerals if w a r r a n t e d by rock composition and objectives of the petrographic analysis. Common stains are available for calcite, dolomite, Fe2+ carbonate, K-feldspar, and plagioclase. Thin sections are then either covered or left uncovered. The conventional practice of gluing cover slips onto thin sections with either Canada balsam or epoxy is decreasing in popularity because many analyses must be performed on uncovered thin sections. Tliin sections that are not polished can be "covered" with colorless fingernail polish. If thin sections are to be analyzed by cathodoluminescence or microprobe techniques, they must be polished to yield suitable results.
PETROGRAPHIC TECHNIQUES
Basic petrographic analysis is performed in transmitted light using a polarizing microscope. Certain petrographic problems require the use of a cathodoluminescence (CL) microscope (see the chapter on "SEM, XRD, CL, and XF Methods" in Part 5). CL petrography is used to detect cement stratigraphy and original fabrics in recrystallized carbonate rocks (e.g., Dorobek, 1987) and to distinguish between detrital quartz grains and authigenic quartz overgrowths (cement) in sandstones (e.g., see Houseknecht, 1987).
Petrographic analysis of thin sections in either transmitted light or CL can involve either qualitative description or quantitative estimation of rock properties, depending upon objectives of the analysis. Q u a n t i t a t i v e e s t i m a t i o n of composition and porosity types by modal analysis (point counting) is recommended for both carbonate and sandstone reservoir rocks. In addition, quantitative estimation of textural parameters (grain size and sorting) is recommended for sandstones.
Compositional Analysis
Modal analysis involves identification of rock constituents at a n u m b e r of locations, or "points," in a thin section (Galehouse, 1971). The number of points counted per thin section, and therefore the labor intensiveness of the analysis, depends upon the precision required. For most applications, at least 300 points must be counted per thin section to ensure an acceptable level of precision. Data derived from point counting can be used to evaluate quantitatively which rock
233
234
PART 5—LABORATORY METHODS
properties significantly influence reservoir quality and to estimate the abundance of various porosity types that are present.
Computer-based image analysis of petrographic images is e m e r g i n g as a r a p i d m e t h o d of e s t i m a t i n g certain compositional properties of sedimentary rocks (including porosity). This technology will surely become more widely used in coming years.
Textural Analysis
Textures of sandstones can be qualitatively described by using standard images to estimate the size, sorting, sphericity, and r o u n d n e s s of clastic particles. Quantitative textural analysis involves measurement of a certain number of grains (commonly 100 per thin section) so that mean grain size and sorting (standard deviation) can be calculated. Grain size measurements can be performed on a transmitted light microscope equipped with a graduated ocular lens, or they can be done by placing thin sections in a microfiche reader and measuring grains with a ruler. Whichever instrument is used, a glass slide inscribed with a metric scale must be used to determine a conversion factor for converting raw data to millimeters.
Once data are converted to millimeters, it is also recommended that the data be converted to the phi scale, where the grain size in phi = -Iog2 x grain size in mm. Mean grain size and sorting (standard deviation of grain size measurements) are then calculated. Mean grain size can be expressed in either the millimeter or phi scale, but sorting must be expressed in the phi scale to maintain a sorting index that is useful across a wide range of grain sizes.
SANDSTONE RESERVOIRS
Three fundamental properties should be documented petrographically to evaluate reservoir quality, to plan reservoir stimulation procedures, and to help assess the potential for formation damage (in concert with SEM, XRD, and other analyses). These are texture, composition, and porosity.
Texture
The grain size and sorting of clastic sediment exerts a profound influence on reservoir properties. Grain size has little influence on porosity, but increasing grain size results in significantly higher permeability (Beard and Weyl, 1973). Sorting also influences reservoir quality, with better sorted sands displaying higher porosity and permeability (Beard and Weyl, 1973). Even though diagenesis may alter these relationships in reservoir rocks, many sandstone reservoirs retain relative porosity and, especially, permeability patterns that can be explained by grain size and sorting variations controlled by depositional facies.
Composition
The m i n e r a l o g y a n d v o l u m e s of f r a m e w o r k grains, cement, and matrix are essential for reconstructing the diagenetic history of a s a n d s t o n e and for assessing the
potential for formation damage that may be caused by completion and/or stimulation procedures.
Framework Grains
Framework grain compositions commonly influence the diagenetic history and reservoir quality of a sandstone by controlling compaction and chemical interaction with pore fluids. Sandstones that contain ductile lithic fragments such as shale or slate compact at shallower depths than sandstones composed mostly of brittle grains such as quartz and feldspar. Therefore, the preservation of primary intergranular porosity may be dependent upon original framework grain composition and the compactional history of the sandstone.
Chemical interaction between framework grains and pore fluids can either reduce or enhance porosity. Certain grains provide preferential nucleation sites for common cements. For example, the presence or absence of nucleation sites on detrital q u a r t z g r a i n s i n f l u e n c e s the v o l u m e of q u a r t z o v e r g r o w t h s precipitated and therefore the a m o u n t of porosity preserved in some sandstones. Chemical interaction of f r a m e w o r k grains with some pore fluids can enhance porosity by grain dissolution. For example, a sandstone facies containing chemically unstable grains (commonly feldspar or various lithic fragments) might undergo porosity enhancement by dissolution of those grains. In the same formation, less porosity enhancement would occur in a quartz-rich facies containing fewer chemically unstable grains.
Cement
The composition and abundance of cements reflects the geochemical history of a reservoir s a n d s t o n e and may influence evaluation of reservoir stimulation procedures and potential for formation damage. Numerous minerals that commonly occur as cements in sandstones are known to dissolve and thereby contribute to porosity enhancement. However, recognition of cement dissolution porosity is often difficult because cement dissolution pores mimic primary pores, especially if the dissolved cement was restricted to intergranular space.
Matrix
Clay minerals commonly occur as both matrix and cement in reservoir sandstones (Wilson and Pittman, 1977). Detrital matrix can be introduced into sand during or immediately following sedimentation by depositional processes, infiltration, and bioturbation. It can occur as grain coatings, dispersed matrix, laminae, or discrete grains. Authigenic clay cements commonly precipitate as grain coatings, pore fillings, and grain replacements. Virtually any clay mineral can occur in any of these modes, with kaolinite, chlorite, smectite, mixed layer illite-smectite, and illite occurring as common constituents of reservoir sandstones.
The presence of clay minerals of any origin has both direct and indirect effects on reservoir quality. Directly, clay minerals commonly result in lowered permeability because they constrict pore throats and promote higher irreducible water saturation. Indirectly, clay minerals commonly influence diagenetic processes that impact reservoir quality. For example, clay grain coatings in some sandstones have
Thin Section Analysis
235
inhibited the nueleation of quartz overgrowths and thereby contributed to porosity preservation. However, clay grain coatings in other sandstones have promoted intergranular pressure solution and have thereby contributed to porosity destruction.
Clay minerals commonly pose potential formation damage problems. Fines migration can occur regardless of clay mineral composition. Clay swelling in response to certain completion or stimulation fluids locally occurs if smectites or mixed layer clays are present. The presence of iron-bearing clays m a y cause precipitation of iron h y d r o x i d e , which commonly damages permeability, as a by-product of acid stimulation if proper chelating agents are not used. (For more on formation damage, see the chapter on "Rock-Water Reactions: Formation Damage" in Part 5.)
PRIMARY INTERGRANULAR
POROSITY
Porosity
During point counting, porosity is typically categorized into one of four categories: (1) primary intergranular porosity, (2) microporosity associated with clay minerals or other very fine mineral constituents, (3) dissolution porosity, and (4) fracture porosity (Pittman, 1979). Proportions of the first three porosity types can be conveniently displayed on a ternary diagram (Figure 1), which summarizes relative reservoir quality and some of the positive and negative attributes commonly associated with the three porosity types. (For more details on porosity types, see the chapter on "Porosity" in Part 5.)
CARBONATE RESERVOIRS
Petrographic analysis of carbonate reservoirs provides description of depositional facies, reconstruction of diagenetic history, and documentation of the porosity system.
Depositional facies of reservoir rocks can be inferred on a microscale if diagenesis has not obliterated original carbonate textures. Petrographers recognize a spectrum of original textures that range from mostly carbonate mud (low energy environments) through mostly sand-sized or larger carbonate grains (high energy environments). In fact, this spectrum of textures is the basis for the two most commonly used carbonate classifications, those of Folk (1959) and Dunham (1962), summarized in Figure 2. In certain instances, variation in reservoir quality (porosity and permeability) can be e x p l a i n e d on the basis of t e x t u r a l v a r i a t i o n related to distribution of depositional facies within the carbonate reservoir.
Diagenetic h i s t o r y of c a r b o n a t e r e s e r v o i r rocks is important to reconstruct because it influences the volume, size, shape, and distribution of pores. Diagenesis may involve porosity-reducing cementation, porosity-enhancing dissolution, and recrystallization, which may result in either reduction or enhancement of porosity. An important goal of carbonate petrography is to establish the sequence of such
MICROPOROSITY
DISSOLUTION POROSITY
Figure 1. Ternary diagram of nonfracture porosity types in sandstones summarizing the influence of porosity type on reservoir quality. (Modified after Pittman, 1979.)
events, or paragenesis, of the reservoir. Careful reconstruction of reservoir paragenesis can provide a perspective of the porosity system at the time of hydrocarbon accumulation, thereby enhancing the geologist's u n d e r s t a n d i n g of how reserves may be distributed relative to diagenetic facies.
Documentation of the porosity system within a carbonate reservoir provides a clear understanding of the origin and three dimensional distribution of pores. This information is typically collected by classifying individual pores into discrete categories (Choquette and Pray, 1970) and by evaluating the degree to which the various pore types are interconnected. (For more on carbonate porosity types, see Table 1 and Figure 3 in the chapter on "Porosity" in Part 5, as well as the chapter on "Carbonate Reservoir Models: Facies, Diagenesis, and Flow Characterization" in Part 6.)
This analysis results in a conceptualization of the threedimensional pathways that hydrocarbons must follow from their original location in the virgin reservoir to the wellbore. Knowing, for example, that porosity in a particular reservoir is selective to a specific depositional facies would allow a geologist to plan enhanced recovery by siting injection and withdrawal locations on the basis of facies distribution. In contrast, knowing that porosity is mostly not fabric selective (e.g., a combination of fracture and vuggy) would likely result in a very different plan for siting injection and withdrawal locations. D o c u m e n t a t i o n of the p o r o s i t y s y s t e m also provides information that is fundamental to planning optimum reservoir stimulation procedures.
236
PART 5—LABORATORY METHODS
Percent Allochems
OVER
O- I %
2 / 3 LIME MUD
I - IO % 1 0 - 5 0 %
M A T R I X SUBE0UAL OVER 2 / 3 SPAR
OVER
SPAR а S O R T I N G
5 0 % LIME MUD
POOR
SORTING GOOD
CEMENT
ROUNDED 8 ABRADED
Representative MiCRiTE a
Rock Terms
DISMICRITE
FOSSILIFEROUS MICRITE
SPARSE
PACKED
BIOMICRITE BIOMICRITE
POORLY WASHED BI0SPARITE
UNS0RTED BIOSPARITE
SORTED BIOSPARITE
ROUNDED BIOSPARITE
1 9 5 9 Micrite 3 Terminology Dismicrite
Fossiliferous Micrite
Terrigenous Analogues
C I а fI font
Bion
Sondy Claystone
i сr i t e
Biospa
.
rile
^ jr x
Clayey Immature
or Sandstone
Submature Sandstone
Mature
Supermature
Sandstone Sandstone
DEPOSITIONAL TEXTURE RECOGNIZABLE
Original components not bound together during deposition
Contains mud (particles of clay and fine silt size)
Mud-supported
Grain-supported
Less than 10 More than 10 percent grains percent grains
Lacks mud and is
grain-supported
Mudstone Wackestone
Packstone
Grainstone
Original components were bound together during
deposition... as shown by intergrown skeletal matter,
lamination contrary to gravity, or sediment-floored cavities that are roofed over by organic or questionably organic matter and are too
large to be interstices.
Boundstone
DEPOSITIONAL TEXTURE NOT RECOGNIZABLE
Crystalline Carbonate
(Subdivide according to classifications designed
to bear on physical texture or diagenesis.)
Figure 2. Carbonate classification schemes of (a) Folk (1959) and (b) Dunham (1962), both based on textures observed in hand specimen or thin section. In Folk's scheme, the black pattern represents lime mud matrix, the lined pattern represents sparry calcite cement, and the white objects represent various carbonate grains.
SEM, XRD, CL, and XF Methods
J. B. Thomas
Amoco Research Center Tulsa, Oklahotna, U.S.A.
INTRODUCTION
The four methods commonly used for additional core analysis are
• Scanning electron microscopy (SEM) • X-ray diffractometry (XRD) • Cathodoluminescence (CL) • X-ray fluoroscopy (XF)
These methods provide important extensions to thin section analysis and can be tied closely to log response and productivity (see the chapter on "Thin Section Analysis" in Part 5). Table 1 summarizes the relative values of each and their limitations.
SCANNING ELECTRON MICROSCOPY (SEM)
In SEM analysis, samples up to 4 in. in diameter are struck by a focused beam of electrons in a vacuum chamber (Figure 1). The secondary and reflected primary electrons are detected and amplified. However, the sample must be conductive to the beam and generally is coated with a very thin layer of carbon or gold-palladium. A spectrum of Xrays are generated as the beam hits the sample. The Xradiation is resolved into characteristic peaks for various elements from fluorine (atomic number z = 9) and up. The images are recorded by photograph or video.
From the recorded three-dimensional images, the morphology of mineral and nonmineral grains and matrix samples can be documented. Pore roughness and interconnectedness can be evaluated. These data can then be related to reservoir flow parameters such as permeability and log response (Pittman and Thomas, 1979). (For information on how grain and pore morphology affect permeability, see the chapter on "Permeability" in Part 5.)
X-RAY DIFFRACTION (XRP)
The principal advantage of XRD is that a qualitative or semiquantitative evaluation of mineralogy is generated. A fixed wavelength X-ray source such as copper X-ray tubes, which have a 1.54A wavelength, is used to irradiate a p o w d e r e d sample. The incident angle 0 (theta) of the diffracted beam and the intensity are recorded with a counter or tube (Figure 2). If parallel planes of atoms of a crystal are struck at the same angle, coherent (additive) intensity is detected and recorded as a peak. Bragg's Law (see Cullity, 1959) is the basis for determining the characteristic peaks (d spacing) for known minerals and compounds {rik = 2d sin 0). Tables of standards established by the JCPDS (Joint Council of P o w d e r Diffraction Standards) can be consulted for mineral identification, where 0 is the incident angle, A, the X-
ray wavelength and d the spacing between planes of atoms in the crystal.
Nonminerals such as solid hydrocarbons or glass cannot be identified by XRD because they lack sufficient internal structure.
Determination of bulk rock mineralogy is obtained from combined diffraction analysis of bulk powder and oriented m o u n t s . P o w d e r m o u n t s are best for identification of nonplaty minerals. Platy minerals are best analyzed in slurries dried on metal, glass, or ceramic holders. This is especially true for clays with particle sizes <5 |im. Abundances are then determined by measuring peak intensity or half area for the diffraction peaks of the major three to five diffraction peaks of each mineral (Figure 3). A limitation of this m e t h o d is the inability to d e t e r m i n e abundances of mineral species (such as quartz from chert) or polymineral grains (such as granite from separate feldspar and quartz).
CATHODOLUMINESCENCE (CL)
In CL, electrons from a cold cathode discharge tube strike a rock surface in a vacuum chamber (Figure 4). In a strong vacuum, energy imparted to electrons in activator ions within the grain causes luminescence. The principle activator ions are manganese and lead (Machel, 1985). Concentrations need be in the 100 ppm range to affect the grain. Other rare earth elements such as dysprosium are also activators. Ferric iron (+3) is the most common quenching ion. The emitted color, when observed, shows the zonations in activator ion concentrations related to the type of crystallization or thermal histories of the host minerals (Figure 5). Lattice defect structures in quartz are also thought to cause some CL in quartz.
The most f r e q u e n t application of CL is in carbonate diagenesis (e.g., Machel, 1985; Myers, 1978). As has been shown by Sippel (1968), it is also useful in determining paragenesis of siliciclastic rocks. It is particularly useful in interpreting original composition and texture in recrystallized or dolomitized strata.
X-RAY FLUOROSCOPY (XF)
X-ray fluoroscopy (Bouma, 1969) requires a beam of Xrays at low current and widely ranging voltages passing through a sample, which is mounted in front of a phosphorus screen or film. The image produced is a "negative" governed by the X-ray absorption contrasts within the sample. Subtle differences in X-ray absorption accentuate masked structures in the sedimentary rock. In siliciclastic rocks, where mineralogy and thus X-ray absorption is more irregular, the results are better. In carbonate rocks, mineralogy variations
237
238
PART 5—LABORATORY METHODS
Table 1. Special Methods for Core Evaluation
Technique SEM
General Operating Environment
20 kV approx. Vacuum chamber Low M.A. current
Data Output
3-D views of secondary Electron images. X-ray spectra for elements with Z > 10.
Advantages
Limitations
Pore geometry; grain
1. Sample size limited.
morphology; diagenetic
2. High organic content
sequences; microtextures. frequently causes
Ties to some log
short filament life.
response and ф-к analyses. 3. Carbon or Au/Pd
coating generally
needed.
4. Shielding required.
XRD
10-40 kV X-ray wavelength fixed. Low M.A. current
Diffraction patterns or digital files for comparison to "standards" files.
Mineralogy on semiquantitative scale. Best method for determining clay mineralogy. Does not require microscope.
1. Sample prep.powder. 2. Abundances based on
measured intensities and areas.
CL
12-20 kV
Photographs: intensity In carbonates, color
1. High Fe(II)1 Ni(II),
Vacuum chamber
relates to activator.
zoning can be related
Co(II) content
on microscope.
Sensitizer and
to complex Eh-pH
quenches CL.
Also visible on
quencher ion content history.
2. Samples require
microprobe.
or lattice defect.
In clastics, lattice
special epoxy for
Low M.A. current
defect and activator
prep.
ion content used in
3. Color emitted may
provenance and para-
extinguish at high
genesis studies.
Mn concentration.
4. Vacuum leaks quench
CL.
5. Shielding required.
XF
±40 kV
Photoradiographs
Image often defines
1. Voltage increase
General current
video imaging.
hidden character of
decreases contrast
3-200 M.A.
sample, e.g., may
in image.
show directional
2. Shielding required.
porosity and flow bound- 3. Carbonate photos
aries, internal structures
often lack detail
of fossils.
because mineral
variety simpler than
siliciclastics.
4. Samples must be
relatively thin.
are slight and images are often fuzzy. However, stylolite seams often show up well.
There has been much discussion on fluoroscopy applications using fluorescent dyes at low magnification related to
coals and other minerals. Figure 6 shows the difference between irradiated and normal incident light for a sample. Note the short vertical fractures accentuated in the fluoroscopy photograph.
Filament
\K
0
Output CRT
Electron Beam Focus/Deflection Coils
Sample
Detector
SEM1XRD, CL1 and XF Methods
239
Planes of Atoms
©
Amp ifier
© = Angle of incidence
Detector Tube
Xray tube X fixed
n X = 2d Sin© d = n X/2 Sin ©
Figure 1. Schematic drawing of a common scanning electron microscope showing how the sample is "iluminated" by an electron beam and amplified for viewing by the operator.
Figure 2. X-ray diffraction configuration. Knowledge of the wavelength (X) and angle of incidence allows the d spacing to be calculated.
Random Powder Mount
H
1
I—
4
12 20
20 deg.
K = Kaolinite Q = Quartz C = Chlorite I =IIIite P = Pyrite F = Feldspar
5 = Smectite
Oriented
Slide
28
<5 ц т
Approximate Abundances (a) Q = 2/3
l&S = 1/6 C&l = 1/12 F = 1/12 P = tr
(b)I = 1/2
K = 1/5 S&C = 1/3
Figure 3. X-ray diffraction patterns.
Figure 4. Schematic drawing showing how a typical cathodoluminescence system works. Depending on the manufacturer, the location of the cathode tube may differ.
240
PART 5—LABORATORY METHODS
Figure 5. A photomicrograph taken under cathodoluminescence showing concentric zoning in dolomite cement. High Mn+2 dolomite shows up as bright bands and higher Fe+2 dolomite as
dark bands. (Courtesy of W. J. Myers.)
Figure 6. (a) X-ray fluoroscopy slab photograph and (b) plane light slab photograph of a Pennsylvanian sandstone from Oklahoma.
Oil and Condensate Analysis
P. C. Henshaw
Chevron Canada Resources Calgary, Alberta, Canada
R. L. Kaufman
Chevron Overseas Petroleum, Inc. San Ramon, California, U.S.A.
L. W. Slentz
Consultant for Petroleum Testing Services Santa Fe Springs, California, U.S.A.
INTRODUCTION
Data from analysis of oils and condensates can contribute to the effective development and management of petroleum fields. These data also can be essential for equity studies and for environmental or pollution cases. Table 1 provides a quick reference to common questions that can be answered with the use of oil and condensate analyses. However, chemical analysis data, such as provided by gas chromatography, must be integrated with geological, engineering, and wellsite information to make these interpretations (Figure 1).
SAMPLING
The key to useful data depends on proper sampling. Provisions for p r o p e r sampling of oils and condensates should be included in all well proposals, tests, completions, and workover programs. The following list gives when to sample wells for oils:
• When FTs or DSTs recover oil a n d / o r water • After completion, as soon as stream has cleaned up
• When water cut becomes significant • On recompletion, prior to shutting in zone to be
abandoned • After recompletion, as soon as stream has cleaned up • When unexplained production changes occur • Prior to abandonment • Any time a specific problem occurs • When sampling nearby or associated wells
Note that when sampling, always sample the following:
• All streams or multiple completions • Each stream for both hydrocarbons and water • Streams that may have a bearing on the subject well • Pertinent neighboring wells, if sampling is permitted
The expected character of the petroleum (such as API gravity and gas to oil ratio) and reservoir conditions (such as pressure and temperature) must be considered so that representative samples can be obtained. Changes in pressure and flow rate during well testing can alter petroleum composition and physical properties. In addition, samples of drilling mud, load and kill fluids, drilling additives, and
Table 1. Development Geology and Engineering Questions Answered by Oil and Condensate Analyses
Situation Oil on pits
Oil show in core or cuttings Unknown source rock for oil
Poor production test flow rates
Casing leak Reservoir continuity Monitor multizone production
Workover evaluation
Questions Drilling additive or contaminant
or formation fluid Oil quality and source Expected oil or unexpected oil (new play)
Low viscosity (due to oil source or biodegradation) or high wax paraffinic
Producing zone oil or new zone Lateral or vertical continuity Production allocation (well head samples,
low cost) New oil (zone) or same oil
Analysis GC
GC, HPLC, and GCMS Bulk properties, GC, HPLC,
GCMS, and carbon isotopes Bulk properties, GC, and HPLC
GC and carbon isotopes GC GC
GC
241
242
PART 5—LABORATORY METHODS
• DST • RFT • Swab Run • Rev Circ • Well Head
• Acquire Data • Construct
Baseline • Identify Peaks • Measure
Heights
• Identify Peaks for Correlation
• Calculate Ratios
• Star Plots • Cluster
Analysis • Other
Multivariate Techniques
Figure 1. Flow diagram showing the basic steps followed in oil-oil correlations using gas chromatography.
• Reservoir Continuity
• Leaky Casing, Tubing, Packing, Pipeline, etc.
• Commingled Production
• Source • Maturation • Migration
wellsite fuel oils and lubricants should be kept until p e t r o l e u m analyses are c o m p l e t e d . If contamination is discovered or suspected, these samples can be analyzed and compared to recovered "petroleum."
For drilling wells, all recovered fluids should be sampled, including shows, oil on the pits, DSTs, and formation tests. Drill cuttings and core samples may be solvent extracted to obtain liquid samples. For completed wells, each zone test should be sampled before, during, and after workovers, and prior to abandonment. For reservoir continuity studies or production allocation, samples should be collected as necessary.
Ideally, samples should be collected and stored in glass bottles or jars to minimize contamination. Plastic and rubber caps or stoppers should not be used unless they are made of Teflon. Samples captured in pressure bombs should be sent to the lab in the bomb.
If glass bottles are not available at the wellsite, other temporary containers can be used as long as they are clean and can be capped to minimize the loss of gas and gasoline components. These samples can then be transferred to suitable bottles at the lab.
ANALYSIS
M a n y d i f f e r e n t t y p e s of a n a l y s e s are available. To determine the best and most economic techniques, it is essential to know why analyses are to be run. A variety of common analyses are run for both bulk properties and for molecular composition. Procedures established by the American Society for Testing Materials are used unless proprietary techniques supersede them.
Bulk Properties
The most commonly run analyses performed on bulk samples are
• API gravity • Viscosity
• Pour point • Gastooilratio(GOR) • Percentage sulfur • Carbon isotopes
Bulk properties should be measured on all test samples and measured (or estimated) on shows. They are essential for calculation of fluid properties, reserves, and economic value. They also are critical to the design of production methods and refinery processes. (For more on oil and condensate properties, see the chapter on "Petroleum Reservoir Fluid Properties" in Part 10).
Variations in these properties can be used to test exploration models and evaluate development strategies. All of the basic b u l k p r o p e r t i e s are affected by geological mechanisms that produce variations in oil composition (Table 2) (Tissot and Welte, 1978; Hunt, 1979).
Table 2. Mechanisms That Produce Variations in Oil Composition
Before accumulation
Mechanism
1. Source input and diagenesis 2. Maturation processes
• Time • Temperature • Pressure • Lithology 3. Primary migration
After accumulation
1. Secondary migration 2. Maturation processes 3. Alteration processes
• Biodegradation • Water washing • Asphaltene precipitation • Weathering
Oil and Condensate Analysis
243
Normal oil (AP - 35°)
! P-Xylene
yjJL •. I u L w i
n-C15
Pristane Phytane
n-C2o
Illl
t i L-J-A A - L * . . L
х
^1
к
1! Л
1 Л'i
1 ^ l'х
ЛL .
I I
n - C 30
!
l'Ii 1 !!
Biodegraded oil (API - 15°)
^ J
il • I l . ••
Figure 2. Gas chromatograms of a "normal" 35° API gravity oil and a biodegraded oil showing the lose of aliphatics (л-paraffins and isoprenoids) due to biodegradation.
Molecular Composition
The most common analyses done to determine molecular composition are as follows (see Altgelt and Gouw7 1979, and Tissot and Welt, 1978, for a description of these analyses and their applications):
• Gas chromatography (GC) • High performance liquid chromatography (HPLC) • Porphyrin distribution • Gas chromatography-mass spectroscopy (GCMS)
Molecular compositions can be used to categorize oils (such as waxy or paraffinic and aromatic) and to determine the effects of geological processes (such as biodegraded, water washed, or immature). Bulk properties such as API gravity can sometimes be predicted from molecular composition data (Kennicutt and Brooks, 1988). Chromatographic methods are typically used to determine these characteristics.
Gas chromatography (GC) data can indicate the geological mechanisms responsible for changes in composition of an oil. For example, during biodegradation, bacteria preferentially remove the п-paraffins that are prominent features in most chromatograms (Figure 2). Therefore the decrease or absence of n-paraffins is a strong indication that an oil is biodegraded. Other geological processes recognizable from chromatograms are thermal immaturity (odd-even predominance of the nparaffins), water washing (depletion of light aromatics), leaky reservoir seals (loss of light ends), and source characteristics (biomarkers) (Figure 3). Drilling additives and contaminants can also be identified by chromatography (Figure 4).
If p r o p e r s t a n d a r d i z e d p r o c e d u r e s are f o l l o w e d , chromatograms can provide reproducible "fingerprints" of oils. These oil fingerprints can then be used to compare and correlate oils. This technique of molecular characterization is more discriminating than bulk property data (Kaufman et al., 1990). For ease in interpretation, chromatographic data can also be displayed as polar or star plots of hydrocarbon peak ratios (Figure 5).
Gas chromatography should never be used alone to make these interpretations. Supporting analytical data and geological information should be obtained as well. A combination of several processes (that is, multiple sources for oils and/or different thermal maturities) can make interpretation complex.
APPLICATIONS
Correlation of recovered fluids to other oils or source rock extracts can verify exploration models for development of new plays (Tissot and Welte, 1978; Hunt, 1979)
Correlation techniques, especially gas chromatography, have also been used for the following development applications:
• Reservoir continuity studies (Slentz, 1981; Ross and Ames, 1988)
• Identification of producing and nonproducing zones (Maness and Price, 1977)
• Identification and localization of production problems (Kaufman et al., 1990)
244
PART 5—LABORATORY METHODS
P-Xylene
P-C1,
Ф <D C C CO Q
."S-
>
£
Q- Q-
Mn _Vr^ 2 0
1
n - C 30
...
'J, .1 . .'Ja i, ,4. , Iw-JV^'.. OV L.Vv. L — U J> .
Gas Chromatogram
Measured Parameter
Geologic Interpretation
• Odd n-Paraffins ф Even
• n-Paraffins Absent • Gasoline Hydrocarbons Absent
• Low Aromatic Content
• ProminentC25+ n-Paraffins
Immature Biodegradation Water Washed, Poor
Reservoir Seal Water Washed Waxy, Possible
High Pour Point
Figure 3. Gas chromatography can be used to determine geological processes experienced by an oil.
• Allocation of production to specific intervals when production is commingled (Kaufman et al., 1987)
For example, in a formation that has a continuous oil reservoir, the fingerprint of that oil does not change, in other words, a change in a fingerprint reflects a discontinuous reservoir. Therefore, by sampling reservoirs vertically and laterally, it is possible to determine reservoir continuity. Figure 6 shows a schematic drawing of two reservoir sands that actually define three separate reservoirs. Although geologically equivalent, sands A and B are discontinuous and should be produced as separate reservoirs. Sand C is laterally continuous but vertically discontinuous from both sands A and B. The star d i a g r a m s h o w s a representation of the different oil fingerprints. If production from sands A and C in well 1 is commingled, production could be allocated to the individual sands by using the fingerprint differences with a binary mixing model (Kaufman et al. 1987). Figure 7 illustrates this with a simple mixing diagram.
In addition, if the fingerprint of oil f r o m zone A, for example, is not found in the produced oil, this would indicate no p r o d u c t i o n f r o m z o n e A. The r e a s o n for lack of production could be depletion or watering out of a previously productive zone or the presence of a nonproductive or wet zone.
Hydrocarbons From Oil-Based Mud
C10\i C 11C/,1,2013
C14
\ £ 100 C 9 4
Pristane ,Phytane
Biodegraded Oil
10
20
30
40
50
60
70
80
Time (minutes)
Figure 4. Gas chromatogram showing drilling mud contamination to a biodegraded oil.
1000
Oil and Condensate Analysis
245
W U J U a j A/-J4. ^.LJ^LJ-
Ratio No.
Ratio
Peak Height Ratio
Figure 5. (a) Gas chromatographic data can be displayed as chromatograms at different attenuations, (b) as star diagrams of selected peak height ratios, or (c) as data tables of peak height ratios. The star diagram is a polar plot of the peak height ratio data from the table, which shows data from the chromatographic trace for peaks between n-C17 and n-C25.
246
PART 5—LABORATORY METHODS
Sand A
Sand B
Sand C
Figure 6. Gas chromatographic data can be used to determine oil reservoir continuity in conjunction with other methods such as pressure-depth plots. Diagram on right shows a graphical comparison of chromatographic data.
X = Unknown Oil +,©,О = Lab Mixes
100% A 80/20
0% C
60/40 40/60
20/80
0% A
100% C
Figure 7. A binary mixing diagram showing how the relative proportions of an unknown oil can be determined using the two end-member oils (separate zones) and laboratory mixtures.
Oilfield Water Analysis
Parke A Dickey
Professor Emeritus Tulsa, Oklahoma, U.S.A.
INTRODUCTION
The principal reasons for performing chemical analysis of water produced with oil are to
• Determine its source • Determine water resistivity (Rw) • Detect hydrocarbon-related compounds • Trace the path and location of injected water
METHODS
Sampling
A truly representative sample can best be obtained from the flow line. Another sampling method is by drill stem tests, although this water is usually contaminated with filtrate from the drilling mud. (For more details, see the chapter on "Drill Stem Testing" in Part 3.) Additional sampling methods include formation testers such as the formation inverval tester (FIT) and the repeat formation tester (RFT), which due to their ability to hold a limited volume, usually recover only filtrate. (For more on formation testers, see the chapter on "Wireline Formation Testers" in Part 4.)
Analytical Methods
In the past, only the six principal elements were reported. Only five of these were determined by analysis: calcium, magnesium, chloride, alkalinity (usually reported as bicarbonate), and sulfate. Sodium was estimated by difference. Results were reported as parts per million (ppm), but because the methods are volumetric, it is more correct to report those results as milligrams per liter (mg/L). Recently, physical methods such as atomic absorption and spectrometry have made it possible to analyze for the less abundant elements (American Society for Testing Materials,
1990). Some elements, such as barium, are important because they precipitate and plug pores. Others such as iodine and bromine may be economically profitable to recover. (For more on properties of reservoir water, see the chapter on "Petroleum Reservoir Ruid Properties" in Part 10.)
INTERPRETATION
In interpreting water analyses, it is customary to use reacting values (also called equivalent parts per million). The reacting value is the weight of the element in parts per million divided by the atomic or molecular weight and multiplied by the valence. In comparing waters, it is also useful to calculate the milliequivalent percent (Table 1). Because there are usually three principal cations and three anions, the milliequivalent percent can be plotted on a triangular diagram (Figure 1),
In plotting water composition on maps, it is convenient to show the analyses in the form of patterns. One of the most commonly used patterns was devised by Stiff (1951). The cations are plotted to the left on three or four lines, and the anions are plotted to the right. Milliequivalents are usually plotted on a logarithmic scale (Figure 2).
APPLICATIONS
When a well starts to make water, it is necessary to find out where the water is coming from to determine what actions, if any, are needed. Another important reason for sampling and analyzing water is to determine its resistivity (Rw). This value is needed to determine its saturation (Sw) in the producing formation by wireline log analysis. Consequently, some well logging societies have compiled Rw values for different regions.
Water from dry holes is sometimes analyzed for traces of hydrocarbon-related organic compounds, such as organic
Table 1. Example Calculations To Convert Milligrams per Liter to Milliequivalents and Milliequivalent Percent
Element
Na+ Ca2+ Mg2+
Total cations
Factor
0.0435 0.0499 0.0823
mg/L
44,100 11,000
1,500
Water 1 meq
1918 549 123
2590
meq %
74 20
6
100
Mg/L
3040 21 7
Water 2 meq
132.2 1.0 0.6
133.8
ci-
SO42" HCO3-
Total anions
0.0282 0.0208 0.0164
91,800 None 34
2589
100
—
—
0.5
: I
2590
100
3240 407 1870
91.4 8.5
30.7
130.6
247
meq % 99
0.5 0.5 100
70 6.5
23.5 100.0
248
PART 5—LABORATORY METHODS
Na
#J * &
/ ^
• Maracaibo (Cretaceous) о Frio, Texas
\ A Pennsylvania, Devonian Woodbine
\M O k l a h o m a
A A /o
/
^
/ °
Ca
Mg
Figure 1. Triangular plot showing relative amounts of cations in typical oil field brines. Relative amount of sodium changes, but calcium is always about five times magnesium. (After Dickey, 1966.)
acids and benzene. If they are found, it suggests that the formation had an oil accumulation in the vicinity (Zarella,
1967).
Water for subsurface injection should be carefully filtered and analyzed for its chemical composition. Injection water is filtered because it must be free of suspended matter that might plug the rock pores or coat the faces of the grains. Tliis matter might be bacteria or algae and can be mitigated by including bactericides in the water. Harmful matter can also arise from corrosion of the steel pipes, so it is customary to keep dissolved oxygen out of the injection water. (For
Illllll 14 IMItll11I1l II I
Logarithmic
Cl
I I I j IIKI 1111111 I MjMII I Iljllll нсо-
ttltjll I
I I I Illl I 11 jllM I Illllll I Illllll S04
Illllll I Illllll I Illllll I Illllll I I Illllll I Illllll I Illllll I Illllll СОз
Water 1
(BartIesviIIe)
Logarithmic
llll|!l I llll|ll Г
5
ttlljll I illl|li I нMиil]-иH+I Nllj I I I
TTTl Tm
W
WTJJI -H+1TпtTиT II 1I1lllIlllll н с о з
Mg Illljll I lllljll I HlljLH-4-iiiilll 1
II I1l1IMlllil-lw+lI w >I Il1l1llllIlll
Illllll I Illllll I mil III Illllll I I Illllll I Illllll 1 1I1l,lll I Illllll CO3
Water 2
(Dakota)
Figure 2. Stiff (1951) diagrams used to show water compositions on maps.
information on corrosion and scale, see the chapter on "Production Problems" in Part 9.)
Injected water can also cause authigenic clays in the pores to swell and/or migrate. Injected water may react with interstitial water, forming precipitates that also plug the pores and create formation damage. This reaction can be predicted from chemical analysis. (For more on formation damage, see the chapter on "Rock-Water Reactions: Formation Damage" in Part 5.)
Rock-Water Reaction: Formation Damage
Dare Keelan J. O. Amaefule
Core Laboratories Houston, Texas, U.S.A.
INTRODUCTION
Near wellbore permeability reduction, that is, formation damage, can result from the interaction of the reservoir rock with extraneous drilling, completion, stimulation, or enhanced recovery fluids. Rock-fluid reaction can occur from any of the following operations:
1. Mixing of formation water with extraneous brine in which the total cation concentration is below a critical salinity value
2. Displacement of one brine by another whose divalent cation concentration is lower than the level required to prevent clay dispersion
3. Injection or production of fluids at rates exceeding the critical velocity beyond which mobilization of rock fines by mechanical (drag) forces is initiated
Both the critical salinity (Khilar and Fogler, 1983) and critical velocity (Gruesbeck and Collins, 1982) can be determined from laboratory tests. These two damage mechanisms can exist independently or in combination (Gabriel and I n a m d a r , 1983). Therefore, any system of laboratory tests to evaluate only the influence of salinity on flow impairment must be conducted at rates below the critical velocity. Conversely, the critical velocity will vary depending on whether fluids are chemically compatible with the rock.
FACTORS INFLUENCING ROCK-WATER REACTIONS
Mineral fines that contribute to permeability reduction can be composed of clay, quartz, feldspar, or carbonate. Nonclay fines have no significant surface charge, and commercial clay stabilizers will not prevent their migration. Detrital clays forming the rock framework normally have little impact on rock-fluid reactions. Authigenic clays, however, line, fill, or bridge pores. They are exposed to extraneous pore fluids with which they may react (Eslinger and Pevear, 1988). Expanding clays (smectites and mixed layer clays containing smectites) can reduce cross-sectional areas of pore throats and thus permeability. More important, as they expand, they often contribute mobile fines. These fines along with illite and kaolinite (which have lesser to no swelling tendency) are the primary materials that disperse, migrate, bridge, and impair permeability.
PROBLEM PREVENTION AND CORRECTION
Table 1 summarizes potential rock fluid reactions based on knowledge of clays present, damage prevention, and corrective procedures (Kersey, 1986). Prevention is preferred and, when possible, is likely to cost less than correction.
LABORATORY TESTS
Laboratory tests used to assess rock-water reactions can be grouped into three categories:
1. Petrographicanalysis • Thinsections • X-ray diffraction • Scanning electron microscopy (SEM)
2. Salinity-related tests • Liquid permeability • Depthofdamagestudies • Capillarypressure • Watershock • Critical cation concentration
3. Rate-related tests • Criticalvelocity • Clay stabilizer effectiveness • WettabiHtyoffines
Petrographic Analysis
Petrographic analysis indicates the potential for permeability reduction by identifying types, amount, and location of clays a n d other m i n e r a l s . (For details on petrographic methods, see the chapters on "Thin Section Analysis" and "SEM, XRD, CL, and XF Methods" in Part 5.)
Salinity-Related Tests
Salinity-related tests f u r n i s h direct i n d i c a t i o n of rock-water interaction. They allow evaluation of damage induced by drilling, completion, workover, and injection fluids. Figure 1 illustrates results of a laboratory experiment to evaluate the reaction to reservoir rock with a proposed injection brine. Permeability was reduced to 20% of its original value after exposure to 20 pore volumes of proposed injected brine. Good reservoir management requires that injected volumes equal produced volumes; therefore, reduced injectivity results in reduced hydrocarbon production rates.
Capillary (water retentive) properties of rocks are altered by rock-fluid reactions. Capillary curves before and after
249
250
PART 5—LABORATORY METHODS
Table 1. Rock-Fluid Potential Problems, Prevention, and Corrective Action
Potential Problem
Clay swelling (water sensitive)
Contributing Minerals
Smectite Illite-Smectite Chlorite-Smectite Illite Kaolinitea
Damaging Fluids and System
Freshwaterbased fluids
Any water with inadequate concentration of cations
Damage Prevention Oil-based mudd Potassium Ammonium chloride
Calcium Chloridee
Fines movement (rate sensitive)
Kaolinite Illiteb
Chlorite Illite-Smectitec Chlorite-smectitec
Clay size particles
of quartz or other
minerals
High transient pressure
High flow rates
Perforate slightly underbalanced (1000 psi)
Increase well rate slowly
Maximize perforations per ft
Select rate less than critical velocity
Use clay stabilizer1
aDoes not swell but will be dispersed by freshwater. bExposure to fluids devoid of potassium can cause leaching of potassium and subsequent flexing, breaking, or migration of particles. cSmectite may swell and thereby dislodge associated clays. dNot in gas wells as oil will reduce gas relative permeability or where oil contains asphalt (which will precipitate) in contact with diesel base of mud. eDo not use in micaceous reservoirs or where C02 is in solution. tWiII not stabilize clay size quartz particles.
After Kersey (1986).
Damage Correction
Preflush with HCI and NH4CI
HCI/HF acidize
Postflush with NH4CI and clay stabilizer
Preflush if needed with HCI and NH4CI
Acidize with HCI/HF
Postflush with NH4CI and clay stabilizer
exposure to extraneous fluids indicate if rock-fluid reaction has reduced pore sizes. If reduction occurs, retained water saturation is increased (Amaefule and Masuo, 1986). (For more on capillary pressure, see the chapter on "Capillary Pressure" in Part 5.)
When the source and composition of brine to be injected is unknown, water shock tests indicate potential rock damage and present a worse-case scenario. Permeability of the formation rock to a 0.51-M (3 wt. %) NaCl solution is followed by permeability to freshwater. Sensitive rocks will show permeability reduction. Rock-water reaction must be evaluated for all aqueous fluids introduced into the reservoir system. Critical salinity (cation concentration) below which damage occurs, as well as schemes to lower brine concentration stepwise so as to avoid clay damage, can be evaluated in the laboratory.
Rate-Related Tests
The critical interstitial velocity at which permeability reduction due to fines migration is initiated can be determined in laboratory tests. These tests simulate the effect of high f l o w rates that exist near the wellbores of both injection and production wells. Muecke (1979) discusses factors controlling fines movement. These include fluids flowing, fines wettability, and interfacial forces.
Figure 2 illustrates laboratory data that established a
critical flow velocity of 0.3 cm/sec. Beyond this rate, fines were mobilized that bridged at pore throats and resulted in reduced permeability. The p H of the effluent was monitored throughout the test. The constant pH value indicated that no chemical reaction was occurring between the rock and fluid and that viscous forces w e r e the cause of the r e d u c e d permeability.
The critical flow velocity is normally obtained by testing a cylindrical sample, with flow parallel to the linear axis. The linear velocity can be scaled to the radial flow condition existing in the wellbore. The scaled data yield the maximum well flow rate in barrels per day that can be tolerated before fines bridging and loss of production rate occurs (Gorman et al., 1989). These data allow calculation of the radius of the permeability impaired zone and aid in sizing subsequent acid volumes required to clean up the impairment.
Test conditions should mirror the field condition under study. Thus, water injection tests should be made on reservoir rock specimens in which simulated injection brine is flowed in the presence of residual hydrocarbons. Oil or gas production tests should be made by flowing the appropriate hydrocarbon through the rock specimen with interstitial water present. Drag forces are proportional to both rate and viscosity; therefore, flowing fluid viscosities should also model reservoir values.
Changes in p H indicate fluid-fluid or rock-fluid reactions; therefore, monitoring of injection and produced water p H
Rock-Water Reaction: Formation Damage
251
CRITICAL INTERSTITIAL VELOCITY: cm/e
Figure 1. Liquid permeability test indicating permeability reduction due to rock-liquid reaction.
s h o u l d be an integral p a r t of a n y critical velocity determination. In addition, effectiveness of clay stabilizers should be evaluated as an extension of the critical velocity measurement.
CHEMICAL (SALINITY-RELATED) FINES MOBILIZATION
Chemically initiated fines migration occurs when authigenic clays are contacted by fluid that (1) has an inadequate total cation (Na+, K+, NH4"1", Ca2+, Mg2+, Ba2+, and Sr2+) concentration to prevent dispersion of formation clays or (2) contains an inadequate percentage of divalent cations (Ca2+ and Mg2+), even when total cation concentration is high. Clay dispersion is a complex phenomenon dependent on clay type and quantity and the brine composition of both original formation water and extraneous water.
Clay Type
Clay type influences the brine salinity at which clays start to deflocculate. Higher cation exchange capacity clays require higher salinity to prevent clay deflocculation. Smectite, illite, and kaolinite have cation exchange capacity (CEC) values of approximately 100, 20, and 5, respectively, Thus, total brine cation concentration must be higher to prevent smectite dispersion, followed by lesser concentrations for smectite-illite mixed layer clay, illite, and kaolinite clays, respectively (Scheurerman and Bergersen, 1990).
Chlorite exhibits essentially no water sensitivity. It contains iron and it can react with spent HCl acid if the p H rises above 1.0. A pore-plugging ferric hydroxide precipitate with a hydrated gel-like structure can form unless an oxygen scavenger and iron-chelating or iron-sequestering agents are used.
Brine Composition
Cation species and valence, as well as total cation concentration, influence clay swelling and flocculation. Divalent cations are more effective in promoting flocculation than monovalent cations, and small quantities of Ca2+ or Mg2+ have stabilizing effects.
Figure 2. Critical velocity determination with pH monitoring.
Common monovalent ions include K+, NH4+, H+, and Na+. Sodium (Na+) is the least effective ion in promoting clay stabilization and will increase clay sensitivity if the clays are subsequently exposed to freshwater. Illite and mica are p o t a s s i u m (K+^ rich. Their stability requires that K+ be present in injected waters to prevent potassium extraction from illite or friable micaceous sands. Ion extraction can result in dispersable particles (Reed, 1977).
Calcium chloride brine is incompatible with formation or injection brines containing CO32-, HCO3", or SO42-. Consequently, scale inhibitor must be used or a suitable (normally liigher concentration and more costly) substitution of KCl or NH4Cl brines will suffice. Another alternative is to displace the noncompatible water with a slug of KCl or NH4Cl and follow this with CaCl2.
The contact of calcium or p o t a s s i u m fluids with low c o n c e n t r a t i o n , s p e n t H F acid u s e d for s t i m u l a t i o n of sandstones will also form damaging precipitates. NH4Cl can be used to displace spent HF acid several feet into the formation prior to CaCl2 or KCl treatment, thereby preventing acid reaction with Ca2+ or K+. Conversely, HF acid should not be allowed to contact calcium-bearing minerals such as calcite or dolomite. It is necessary to use HCl as a preflush or cation exchange Ca with an NH4Cl preflush. Only underreaming or fracturing will correct calcium fluoride precipitate damage.
Rate of Salinity Change
Abrupt changes in salinity shock smectite, mixed layer illite-smectite and chlorite-smectite and to a lesser extent illite. Shock results in clay swelling and permeability decline. Nonswelling kaolinite can also be sensitized by contact with freshwater, which results in fines migration. A stepwise d i l u t i o n of injection b r i n e r e d u c e s total salinity, while maintaining the ratio of divalent to total cations. This gradual change can prevent clay damage (Jones, 1964).
Cation exchange can strip divalent cations from injection waters. If the resultant divalent ion concentration falls below the level required to keep clays flocculated, major permeability impairment will result. Pretreatment of the near-wellbore region with CaCl2 (or KCl or NH4Cl solutions if Ca is not compatible) can prevent subsequent damage. The pretreatment should be of sufficient volume to reach at least 5 ft into the formation.
252
PART 5—LABORATORY METHODS
Water pH
Negative charges exist on clay surfaces, and kaolinite is weakly cemented. Consequently, a repulsive force between quartz and clay promotes clay dispersion and reduced permeability at normal to high p H values. Effects of pH are intensified in low salinity solutions and are less important in high ionic strength solutions. Values of pH greater than 9.0 also result in silica dissolution, with resultant fines release. High p H also promotes formation of oil-water emulsions that reduce flow rate. Thus, it is best to avoid high p H systems.
Temperature Effects
The rate of permeability impairment has been summarized (Coskuner and Maini, 1988) and shown to decrease with increasing temperatures up to 200 T when brine flows in the presence of oil. It has also been s h o w n that higher temperatures require higher salt content to stabilize clays when only brine flows. Solubility of quartz increases with temperature, and additional fines can be released and mobilized. It is prudent to make evaluation measurements at temperatures expected to exist at operating conditions.
References Cited
253
Part 5 References Cited
Altgelt, K. W., and T. H. Gouw, 1979, Chromatography in Petroleum Analysis: New York, Marcel Dekker.
Adamson, A. W., 1982, Physical Chemistry of Surfaces, 4th ed.: New York, John Wiley and Sons.
Allen, J. R., 1979, Prediction of permeability from logs using multiple regression: Transactions of the Sixth European Formation Evaluation Symposium, London, Paper M.
Amaefule, J. O., and L. L. Handy, 1982, The effect of interfacial tensions on relative oil/water permeabilities of consolidated porous media: Society of Petroleum Engineers Journal, v. 22, p. 371-381.
Amaefule, J. O., and D. K. Keelan, 1989, Stochastic approach to computation of uncertainty in petrophysical properties: SC Conference Paper No. 8907.
Amaefule, J. O., and S. T. Masuo, 1986, Use of capillary pressure data for rapid evaluation of formation damage or stimulation: Society of Petroleum Engineers Paper No. 12475.
Amaefule, J. O., K. Wolfe, J. D. Walls, A. O. Ajufo, and E. Peterson, 1986, Laboratory determination of effective liquid permeability in low-quality reservoir rocks by the pulse decay technique: 56th California SPE Regional Meeting of the Society of Petroleum Engineers, Oakland, CA, April 2-4, SPE 15149, p. 493-502.
Amaefule, J. O., D. K. Keelan, D. G. Kersey, and D. M. Marschall, 1988, Reservoir description—a practical synergistic engineering and geological approach based on analysis of core data: 63rd SPE Annual Technical Conference and Exliibition of the Society of Petroleum Engineers, Houston, TX, October 2-5, SPE 18167.
American Petroleum Institute, 1960, RP-40 recommended practices for core analysis procedures: Dallas, API, 55 p.
American Society for Testing Materials, 1990, Water and environmental technology: ASTM, v. 11.01,612 p., and v. 11.02,878 p.
Amott, E., 1959, Observations relating to the wettability of porous rock: Petroleum Transactions, AIME, v. 216, p. 156162.
Andersen, G., 1975, Coring and Core Analysis Handbook: Tulsa, OK, PennWell Books, 200 p.
Anderson, W. G., 1986a, Wettability literature survey—Part 1, Rock/oil/brine interactions and the effects of core handling on wettability: Journal of Petroleum Technology, v. 38, p. 1125-1144.
Anderson, W. G., 1986b, Wettability literature survey—Part 2, Wettability measurements: Journal of Petroleum Technology, v. 38, p. 1246-1262.
Anderson, W. G., 1986c, Wettability literature survey—Part 3, Effects of wettability on electrical properties of porous media: Journal of Petroleum Technology, v. 38, p. 13711278.
Anderson, W. G., 1986d, Wettability literature survey—Part 4, Effects of wettability on capillary pressure: Journal of Petroleum Technology, v. 38, p. 1283-1300.
Anderson, W. G., 1986e, Wettability literature survey—Part 5, Effects of wettability on relative permeability: Journal of Petroleum Technology, v 38, p. 1453-1468
Anderson, W. G., 1986f, Wettability literature survey—Part 6, Effects of wettability on waterflooding: Journal of Petroleum Technology, v. 38, p. 1605-1622.
Archer, J. S., and S. W. Wong, 1973, Use of a reservoir simulator to interpret laboratory waterflood data: Society of Petroleum Engineers Journal, v. 13, p. 343-347.
Armentrout, J. M., 1987, Integration of biostratigraphy and seismic stratigraphy, Pliocene-Pleistocene, Gulf of Mexico: 8th Annual Research Conference, Gulf Coast Section, Society of Economic Paleontologists and Mineralogists Foundation, p. 6-14.
Aronofsky, J. S , 1954, Effect of gas slip on unsteady flow of gas through porous media: Journal of Applied Physics, v. 25, n. 1, p. 48-53.
Bass, D. M., 1987, Properties of reservoir rocks, in Bradley, H. B., ed., Petroleum Engineering Handbook: Richardson, TX, Society of Petroleum Engineers.
Beard, D. C., and P. K. Weyl, 1973, Influence of texture on porosity and permeability of unconsolidated sand: AAPG Bulletin, v. 57, p. 349-369.
Bebout, D. G., and R. G. Loucks, 1984, Handbook for Logging Carbonate Rocks: The University of Texas at Austin, Bureau of Economic Geology Handbook 5,43 p.
Berg, R. R., 1970, Method for determining permeability from reservoir rock properties: Transactions Gulf Coast Association of Geological Societies, v. 20, p. 303-317.
Boneau, D. F., and R. L. Clempett, 1977, A surfactant system for the oil wet sandstone of the North Burbank unit: Journal of Petroleum Technology, v. 29, p. 501-506.
Bouma, A. H., 1969, Methods for the Study of Sedimentary Structures: New York, John Wiley and Sons, 458 p.
Boyles, J. M., A. J. Scott, and J. M. Rine, 1986, A logging form for graphic descriptions of core and outcrop: Journal of Sedimentary Petrology, v. 56, p. 567-568.
Bradburn, F. R., and C. A. Cheatham, 1988, Improved core recovery in laminated sand shale sequences: Journal of Petroleum Technology, v. 40, p. 1544-1546.
Braun, E. M., and R. J. Blackwell 1981, A steady-state technique for measuring oil-water relative permeability curves at reservoir conditions: 56th Annual Fall Technical Conference and Exhibition of the Society of Petroleum Engineers, San Antonio, TX, Oct. 5-7, SPE 10155.
Carman, P. C., 1937, The determination of specific surface of powders: Journal of the Society of Chemical Industry, v. 57, p. 225-234.
Casey, J. M., 1980, Depositional systems and basin evolution of the Late Paleozoic Taos trough, northern New Mexico: Ph.D. Dissertation, The University of Texas at Austin, 236 p.
Chatzis, I., N. R. Morrow, and H. T. Lim, 1983, Magnitude and detailed structure of residual oil saturation: Society Petroleum Engineers Journal, v. 23, p. 311-326.
Choquette, P. W. and L. C. Pray, 1970, Geological nomenclature and classification of porosity in sedimentary carbonates: AAPG Bulletin, v. 54, p. 207-250.
Collinson, J. D., and D. B. Thompson, 1989, Sedimentary Structures: Boston, MA, Unwin Hyman, 207 p.
Coneybeare, C. E. B., 1967, Influence of compaction on stratigraphic analysis: Canadian Petroleum Geology Bulletin, v. 15, p. 331-345.
Core Laboratories Inc., undated, A course in the fundamentals of core analysis: Houston, TX.
Craft, M., and D. K. Keelan, 1985, Coring, Part 7—analytical aspects of sidewall coring: World Oil, v. 201, p. 77-90.
302
PART 5—LABORATORY METHODS
Craig, F. F., 1971, The reservoir engineering aspects of waterflooding: Dallas, TX, Society of Petroleum Engineers Monograph No. 3,120 p.
Crocker, M. E., and L. M. Marchin, 1986, Evaluation and determination of cross correlations between wettability methods—Status Report Project BE7: Bartlesville, OK, National Institute for Petroleum Energy and Research.
Cuiec, L. E., 1975, Restoration of the natural state of core samples: 50th Annual Meeting of the Society of Petroleum Engineers, Dallas, TX, September 28-October 1, SPE 5634.
Cuiec, L. E., 1977, Study of problems related to the restoration of the natural state of core samples: Journal of Canadian Petroleum Technology, v. 29, p 68-80.
Cullity, B. D., 1959, Elements of X-ray Diffraction: Reading, MA, Addison-Wesley, 514 p.
Darcy, H., 1856, Les Fontaines Publiques de la Ville de Dijon: Paris, Victor Dalmont, p. 590-594.
Davis, H. G., C. W. Byers, and L. M. Pratt, 1989, Depositional mechanisms and organic matter in Mowry Shale (Cretaceous), Wyoming: AAPG Bulletin, v. 73, p. 11031110.
Dickey, P. A., 1966, Patterns of chemical composition of deep subsurface waters: AAPG Bulletin, v. 50, p. 2472-2478.
Donaldson, E. C., R. D. Thomas, and P. B. Lorenz, 1969, Wettability determination and its effect on recovery efficiency: Society of Petroleum Engineers Journal, v. 9, p. 13-20.
Dorobek, S. L., 1987, Petrography, geochemistry, and origin of burial diagenetic facies, Siluro-Devonian Helderberg Group (carbonate rocks), central Appalachians: AAPG Bulletin, v. 71, p. 492-514.
Dullien, F. A. L., and G. K. Dhawan, 1974, Characterization of pore structure by a combination of quantitative photomicrography and mercury porosimetry: Journal Colloid and Interface Science, v. 47, p. 337-349.
Dunham, R. J., 1962, Classification of carbonate rocks according to depositional texture, in Ham, W. E., ed., Classification of Carbonate Rocks: AAPG Memoir 1, p. 108-121.
Edmonson, T. A., 1965, Effect of temperature on waterflooding: Journal of Canadian Petroleum Technology, n. I l l , p. 236-242.
Ehrlich, R., S. K. Kennedy, S. J. Crabtree, and R. C. Crabtree, 1984, Petrographic image analysis, 1. Analysis of reservoir pore complexes: Journal of Sedimentary Petrology, v. 54, n. 4, p. 1365-1378.
Eslinger, E., and D. Pevear, 1988, Clay minerals for petroleum geologists and engineers: Society fo Economic Paleontologists and Mineralogists, Short Course Notes No. 22.
Fairbridge, R. W., and D. Jablonski, eds., 1979, The Encyclopedia of Paleontology: Stroudsburg, PA, Dowden, Hutchinson and Ross, Inc., 886 p.
Feldman, R. M., R. E. Chapman, and J. T. Hannibal, eds., 1989, Paleotechniques: The Paleontological Society Special Publication No. 4,358 p.
Folk, R. L., 1959, Practical petrographic classification of limestones: AAPG Bulletin, v. 43, p. 1-38.
Folk, R. L., 1974, Petrology of Sedimentary Rocks: Austin, TX, Hemphill Publishers, 159 p.
Forchheimer, P. H., 1901, Wasserbewegung durch Boden: Zeitschrift Verein Deutscher Ingenieure, v. 45, n. 50, p. 1781-1788.
Fraser, H. J., 1935, Experimental study of porosity and permeability of clastic sediments: Journal of Geology, v. 43, p. 910-1010.
Fulcher, R.A., T. Ertekin, and C. D. Stahl, 1983, The effect of the capillary number and its constituents on two-phase relative permeability curves: 44th Annual Meeting of the Society of Petroleum Engineers, Denver, CO, September 28-October 1, SPE 12170.
Gabriel, G. A., and G. R. Inamdar, 1983, An experimental investigation of fines migration in porous media: Society of Petroleum Engineers Paper No. 12168.
Gaither, A., 1953, A study of porosity and grain relationships in experimental sands: Journal of Sedimentary Petrology, v. 23, p. 180-195.
Galehouse, J. S., 1971, Point counting, in Carver, R. E., ed., Procedures in Sedimentary Petrology: New York, John Wiley and Sons, p.385-407.
Gant, P. L., and W. G. Anderson, 1986, Core cleaning for restoration of native wettability: Rocky Mountain Regional Meeting of the Society of Petroleum Engineers, Billings, MN, May 18-21, SPE 14875,.
Garner, B. L., 1985, Case history of Cotton Valley Sand log interpretation for a North Louisiana field: Journal of Petroleum Technology, v. 37, n. 11, p. 1995-2005.
Gerard, R. E., C. A. Philipson, F. M. Bellentine, and D. H. Marshall, 1991, Petrographic image analysis, in Polaz, I., and S. K. Sengupta, eds., Automated Pattern Analysis in Petroleum Exploration: New York, Springer-Verlag.
Goddard, E. N., et al., 1979, Rock-Color Chart: Boulder, CO, Geological Society of America.
Gorman, I., C. Balnaves, J. Amaefule, D. Kersey, and D. Manning, 1989, Gravel packing in poorly lithified reservoirs: Laboratory systems approach to aid decisionmaking strategies: Society of Petroleum Engmeers Paper No. 19477.
Graton, L. C., and H. J. Fraser, 1935, Systematic packing of spheres with particular reference to porosity and permeability: Journal of Geology, v. 43, p. 785-909.
Gruesbeck, C., and R. E. Collins, 1982, Entrainment and deposition of fine particles in porous media: Journal of Petroleum Technology, v. 34,847-856.
Hagoort, J., 1984, Measurement of relative permeability for computer modeling/reservoir simulation: Oil and Gas Journal, Feb. 20, p. 62-68.
Hancock, J. M., 1977, The historic development of biostratigraphic correlation, in Kauffman, E. G., and J. E. Hazel, eds., Concepts and Methods of Biostratigraphy: Stroudsburg, PA, Dowden, Hutchinson and Ross, p. 3-22.
Hawkins, J. T., 1989, Comparison of three methods of relative permeability measurement: The Log Analyst, Sept.-Oct., p. 363-371.
Hawkins, J. T., and A. J. Bouchard, 1989, Reservoir engineering implications of capillary pressure and relative permeability hysteresis: 3rd Annual Meeting of the Society of Core Analysts, New Orleans, LA, August 2-3, Paper No. SCA8909,.
Heaviside, J., C. J. J Black, and J. F. Berry, 1983, Fundamentals of relative permeability—experimental and theoretical considerations: 58th Annual Technical Conference and Exhibition of the Society of Petroleum Engineers, San Francisco, CA, Oct. 5-8, SPE 12173.
Heaviside, J., I. J. A. Gamble, and С. E. Brown, 1987, Relative permeability for intermediate wettability reservoir: 62nd Annual Technical Conference of the Society of Petroleum Engineers, Dallas, TX, Sept. 28-30, SPE 16968,.
Hill, H. J., O. Shirley, and G. Klein, 1979, Bound water in shaly sands—its relation to QV and other formation properties: The Log Analyst, v. 20, n. 3, p. 3-19.
References Cited
253
Honarpour, M., L. Koederitz, and A. H. Harvey, 1986, Relative Permeability of Petroleum Reservoirs: Boca Raton, FL, CRC Press
Honarpour, M., and S. M. Mahmood, 1988, Relativepermeability measurements—an overview: Journal of Petroleum Technology, v. 40, p. 963-966.
Houseknecht, D. W., 1987, Assessing the relative importance of compaction processes and cementation to reduction of porosity in sandstones: AAPG Bulletin, v. 71, p. 633-642.
Hunt, J. M., 1979, Petroleum Geochemistry and Geology: San Francisco, CA, W. H. Freeman, 617 p.
Jenkins, R. E., 1987, Typical core analysis of different formations, in Bradley, H. B., ed., Petroleum Engineering Handbook: Richardson, TX, Society of Petroleum Engineers.
Johnson, E. F., D. P. Bossier, and V. O. Naumann, 1959, Calculations of relative permeability from displacement experiments: Petroleum Transactions, AIME, v. 216, p. 370372.
Jones, F. O., Jr., 1964, Influence of chemical composition of water on clay blocking of permeability: Petroleum Transactions, AIME, v. 231, p. 441-446,.
Jones, S. C., 1972, Rapid accurate unsteady-state klinkenberg permeameter: Society of Petroleum Engineers Journal, v. 12, p. 383-397.
Jones, S. C., 1988, Two-point determinations of permeability and PV versus net confining stress: Society of Petroleum Engineer Formation Evaluation, v. 3, p. 235-241.
Kaufman, R. L., A. S. Ahmed, and W. B. Hempkins, 1987, A new technique for the analysis of commingled oils and its application to production allocation calculations: 16th Annual Convention Proceedings of the Indonesian Petroleum Association, p. 247-268.
Kaufman, R. L., A. S. Ahmed, and R. J. Elsinger, 1990, Gas chromatography as a development and production tool, in Geochemistry of Gulf Coast Oils and Gases: Proceedings of the 9th Annual Research Conference, Gulf Coast Section, Society of Economic Paleontologists and Mineralogists, p. 263-282.
Keelan, D. K., 1971, A critical review of core analysis techniques: 22nd Annual Technical Meeting of the Petroleum Society of the Canadian Institute of Mining, Calgary, Banff, Alberta, June 2-5, Paper No. 7612, p. 1-13.
Keelan, D. K., 1972, A critical review of core analysis techniques: Journal of Canadian Petroleum Technology, v. 2, p. 42-55.
Keelan, D. K., 1977, Core analysis, in LeRoy, L. W., and D. O. LeRoy, eds., Subsurface Geology: Golden, CO, Colorado School of Mines Press, p. 114-126.
Keelan, D. K., 1982, Core analysis for aid in reservoir description: Journal of Petroleum Technology, v. 34, p. 2483-2491.
Keelan, D. K., 1985, Coring, Part 1—why it's done: World Oil, v. 200, p. 83-90.
Kennicutt, M. C., and J. M. Brooks, 1988, Surface geochemical exploration studies predict API gravity off California: Oil and Gas Journal: Sept. 12, p. 101-106.
Kersey, D. G., 1986, The role of petrographic analyses in the design of non-damaging drilling, completion, and stimulation programs: Society of Petroleum Engineers Paper No. 14089.
Khilar, K. C., and H. S. Fogler, 1983, Water sensitivity of sandstones: Society of Petroleum Engineers Journal, Feb., p. 55-64.
Klinkenberg, L. J., 1941, The permeability of porous media to liquid and gases, in Drilling and Production Practices:
Washington, D.C., American Petroleum Institute, p. 200211. Kozeny, J.,S, 1927, Uber Kapillare Leitung des Wassers im Boden (Aufstieg, Versickerung und Anwendung auf die Bewasserung): S.-Ber. Wiener Akad. Abt. II a, v. 136, p. 271-306. Krumbein, W. C., and Monk, G. D., 1943, Permeability as a function of the size parameters of unconsolidated sands: American Institute of Mining and Metallurgical Engineers, Technical Publication 1492, l i p . Kummel, B., and David R., eds., 1965, Handbook of paleontological techniques: San Francisco, CA, W. H. Freeman, 852 p. Labastie, A., M. Guy, J. P. Delclaud, and R. Iffly71980, Effect of flow rate and wettability on water-oil relative permeabilities and capillary pressure: Fall Technical Conference of the Society of Petroleum Engineers, Dallas, TX, Sept. 21-24, SPE 9236. Li, Y., and N. C. Wardlaw, 1986, The influence of wettability and critical pore-throat size ratio on snap-off: Journal Colloid and Interface Science, v. 109, p. 461-472. Lindholm, R. C., 1987, A Practical Approach to Sedimentology: Boston, MA, Allen and Unwin, 276 p. Lowman, S. W., et al., 1972, Definition of selected groundwater terms—revisions and conceptual refinements: U.S. Geological Survey Water Supply Paper 1988,21 p. Lucia, F. J., 1983, Petrophysical parameters estimated from visual descriptions of carbonate rocks—a field classification of carbonate pore space: Journal of Petroleum Technology, v. 35, p. 629-637. Luffel, D. L., and W. E. Howard, 1988, ReUabiHty of laboratory measured porosity in tight gas sands: Society of Petroleum Engineers Formation Evaluation, v. 3, p. 705-710. Machel, H-G., 1985, Cathodoluminescence in calcite and dolomite and its chemical interpretation: Geoscience Canada, v. 12, p. 139-147. Magoon, L. B., and G. E. Claypool, 1981, Petroleum geology of Cook Inlet Basin—an exploration model: AAPG BuUetin, v. 65, p. 1043-1061. Maness, M., and J. G. W. Price, 1977, Well formation characterization by residual hydrocarbon analysis: Annual Meeting of the Society of Petroleum Engineers, Denver, CO, Oct. 9-12, SPE 6860. MeHose, J. C., and C. F. Brandner, 1974, Role of capillary forces in determining microscopic displacement efficiency for oil recovery by waterflooding: Journal Canadian Petroleum Technology, v. 13, p. 54-62. Miall, A. D., 1990, Principles of Sedimentary Basin Analysis, 2nd ed.: New York, Springer-Verlag, 490 p. Miller, M. A., and J. J. Ramey Jr., 1985, Effect of temperature on oil/water relative permeabilities of unconsolidated sands: Society of Petroleum Engineers Journal, AIME, Dec., v. 25, p. 279. Mohanty, K. K., and A. E. Miller, 1988, Factors influencing unsteady relative permeability of mixed-wet reservoir rock: 63rd Annual Technical Conference and Exhibition of the Society of Petroleum Engineers, Houston,TX, Oct. 2-5, SPE 18292. Monicard, R. P., 1980, Properties of Reservoir Rocks—Core Analysis: Houston, Gulf Publishing Company, 168 p. Morris, W. L, 1953, Assignor, Philips Petroleum Co. Portable Permeameter: U.S. Patent No. 2,633,015, March 23. Morrow, N. R., 1970, Irreducible wetting phase saturations in porous media: Chemical Engineering Science, v. 25, p. 1799-1815.
304
PART 5—LABORATORY METHODS
Morrow, N. R., 1987, A review of the effects of initial saturation, pore structure and wettability on oil recovery by waterflooding, in Graham and Trotman, eds., North Sea Oil and Gas Reservoirs: The Norwegian Institute of Technology, p. 179-191.
Muecke, T. W., 1979, Fonnation fines and factors controlling their movement in porous media: Journal of Petroleum Technology, v. 31, p. 144-150.
Muskat, M., 1937, Flow of Homogeneous Huids Through Porous Media: New York, McGraw Hill, 287 p.
Myers, W. J., 1978, Carbonate cements—their regional distribution and interpretation in Mississippian limestones of southwestern New Mexico: Sedimentology, v. 25, p. 371399.
Neuman, C. H., 1980, Log core measurements of oil in place, San Joaquin Valley: Journal of Petroleum Technology, v. 32, p. 1309-1315.
Palmer, C., 1911, The geochemical interpretation of water analyses: Geological Society of America Bulletin, v. 479, 31 p.
Park, A., 1985, Coring, Part 2—core barrel types and uses: World Oil, v. 200, p. 83-90.
Pettijohn, F. J., 1975, Sedimentary rocks, 3rd ed.: New York, Harper and Row, p. 628.
Pettijohn, F. J., P. E. Potter, and R. Siever, 1987, Sand and Sandstone, 2nd. ed.: New York, Springer-Verlag, 553 p.
Pickell, J. J., B. F. Swanson, and W. B. Hickman, 1966, The application of air-mercury and oil-air capillary pressure data in the study of pore structure and fluid distribution: Annual Meeting of the Society of Petroleum Engineers, Denver, CO, SPE 1227.
Pittman, E. D., 1979, Porosity, diagenesis, and productive capability of sandstone reservoirs, in Scholle, P. A., and P. R. Schluger, eds., Aspects of Diagenesis: Society Economic Paleontologists and Mineralogists Special Publication 26, p. 159-173.
Pittman, E. D., and J. B. Thomas, 1979, Some applications of scanning electron microscopy to the study of reservoir rock: Journal of Petroleum Technology, v. 31, p. 1375-1380.
Poag, C. W., 1977, Biostratigraphy in Gulf Coast petroleum exploration, in Kauffman, E. G., and J. E. Hazel, eds., Concepts and Methods of Biostratigraphy: Stroudsburg, PA, Dowden, Hutcliinson and Ross, p. 213-234.
Powers, M. C., 1953, A new roundness scale for sedimentary particles: Journal Sedimentary Petrology, v. 23, p. 117-119.
Pryor, W. A., 1973, Permeability-porosity patterns and variations in some Holocene sand bodies: AAPG Bulletin, v. 57, n. 1, p. 162-189.
Purcell, W. R., 1949, Capillary pressures—their measurement using mercury and the calculation of permeability therefrom: Petroleum Transactions, AIME, v. 186, p. 39-48.
Purcell, W. R., 1950, Interpretation of capillary pressure data: Petroleum Transactions, AIME, v. 189, p. 369-371.
Raza, S. H., L. E. Treiber, and D L. Archer, 1968, Wettability of reservoir rocks and its evolution: Producers Monthly, v. 32, p. 2-7.
Reed, M. G., 1977, Formation permeability damage by mica alteration and carbonate dissolution: Journal of Petroleum Technology, Sept., p. 1056-60.
Rogers, J.J., and W. Head, 1961, Relationsliip between porosity, median size and sorting coefficients of synthetic sands: Journal of Sedimentary Petrology, v. 31, p. 467-470.
Ross, L. M., and R. L. Ames, 1988, Stratification of oils in Columbus Basin off Trinidad: Oil and Gas Journal, Sept. 26, p. 72-76.
Ruzyla, K., and D. I. Jezek, 1987, Stahiing method for recognition of pore space in thin and polished sections: Journal of Sedimentary Petrology, v. 57, p. 777-778.
Scheuerman, R. F., and B. M. Bergersen, 1990, Injection-water salinity, formation pretreatment, and well operations fluidselection guidelines: Journal of Petroleum Technology, Sept., p. 836-845.
Scholle, P. A., and D. Spearing, 1982, Sandstone depositional environments: AAPG Memoir 31,410 p.
Scholle, P. A., D. G. Bebout, and C. H. Moore, 1983, Carbonate depositional environments: AAPG Memoir 33,708 p.
Scott, R. W., 1990, Models and stratigraphy of mid-Cretaceous reef communities, Gulf of Mexico: SEPM Concepts in Sedimentology and Paleontology, v. 2,102 p.
Shamra, M. M., and R. W. Wundlerlich, 1985, The alteration of rock properties due to interactions with drilling fluid components: Annual Technical Conference and Exhibition of the Society of Petroleum Engineers, Las Vegas, NV, Sept. 22-25, SPE14302.
Sharma, M. M., and R. W. Wunderlich, 1987, The alteration of rock properites due to interactions with drilling fluid components: Journal of Petroleum Science and Engineering, v. 1, p. 127-143.
Siemers, С. T., R. W. Tillman, and C. R. Williamson, eds., 1981, Deep-water clastic sediments—a core workshop: SEPM Core Workshop, n. 2,416 p.
Sigmund, P. M., and F. B McCaffery, 1979, An improved unsteady-state procedure for determining relative permeability characteristics of heterogeneous porous media: Society of Petroleum Engineers Journal, Feb., p.15-28.
Sippel, R. T., 1968, Sandstone petrology, evidence from luminescence petrography: Journal of Sedimentary Petrology, v. 38, p. 530-554.
Slentz, L. W., 1981, Geochemistry of reservoir fluids as a unique approach to optimum reservoir management: Middle East Oil Teclinical Conference of the Society of Petroleum Engineers, Manama, Baharain, SPE 9582.
Stiff, H. A., 1951, The interpretation of chemical water analyses by means of patterns: Journal of Petroleum Technology, v. 3, p. 15-17
Swanson, B. F., 1981, A simple correlation between permeabilities and mercury capillary pressures: Journal Petroleum Technology, v. 33, p. 2498-2504.
Swanson, R. G., 1981, Sample examination manual:AAPG Methods in Exploration Series, 35 p.
Teeters, D., M. A. Anderson, and D. C. Thomas, 1989, Formation wettability studies that incorporate the Wilhelmy plate technique, in Borchardt, J. K., and T. F. Yen, eds., Enhanced and Production Stimulation: Washington, D.C., American Chemical Society Symposium Series No. 396.
Tibbitts, G. A., M. G. Reed, and M. McCarter, 1990, New coring system reduce filtrate invasion: Society of Core Analysts European Core Analysis Symposium, London, U.K., May 21-23.
Tissot, B. P., and W. H. Welte, 1978, Petroleum Formation and Occurrence: New York, Springer-Verlag, 538 p.
Toney, J. B., and S. L. Speiglets, 1985, Coring, Part 6—sidewall operations: World Oil, v. 201, p.29-36.
Torabzadeh, S. J., and L. L. Handy, 1984, The effect of temperature and interfacial tension on water/oil relative permeabilities of consolidated sands: Fourth Symposium on Enlianced Oil Recovery of the Society of Petroleum Engineers and the Department of Energy, Tulsa, OK, April 15-18, SPE/DOE 12689.
Tucker, M. E., 1982, The Field Description of Sedimentary Rocks: United Kingdom, The Open University Press, 112 p.
van Eysinga, F. W. B., 1975, Geologic Time Table: Amsterdam, Elsevier (chart).
Vinegar, H. J., 1986, X-ray, CT, and NMR imaging of rocks: Journal of Petroleum Technology, v. 38, p. 257-259.
Wallick, G. C., and J. S. Aronofsk, 1954, Effects of gas slip on unsteady flow of gas through porous media— experimental verification.: Transactions of the American Institute of Mining and Engineering, v. 201, p. 322-324.
Wang, F. H. L., 1986, Effect of wettability alteration on wateroil relative permeability, dispersion, and flowable saturation in porous media: Permian Basin Oil and Gas Recovery Conference of the Society of Petroleum Engineers, Midland, TX, March 13-14, SPE 15019.
Wardlaw, N. C., and J. P. Cassan, 1978, Estimation of recovery efficiency by visual observation of pore systems in reservoir rocks: Bulletin of Canadian Petroleum Geology, v. 26, p. 572-585.
Wardlaw, N. C., M. McKellar, and Y. Li, 1988, Pore and throat size distribution determined by mercury porosimetry and by direct observation: Carbonates and Evaporites, v. 3, p. 1-15.
Wardlaw, N. C., and R. P. Taylor, 1976, Mercury capillary pressure curves and the interpretation of pore structures and capillary behavior in reservoir rocks: Bulletin of Canadian Petroleum Geology, v. 24, p. 225-262.
Watson, А. Т., P. C. Richmond, P. D. Kerig, and T. M Tao, 1988, A regression-based method for estimating relative permeabilities from displacement experiments: Society of Petroleum Engineers Reservoir Engineering, Aug., p. 953958.
Weimer, R. J., and R. W. Tillman, 1980, Tectonic influence on deltaic shoreline facies, Fox Hills Sandstone,west-central Denver basin: Colorado School of Mines Professional Contribution, n. 10,131 p.
References Cited
253
Wellington, S. Lv and H. J. Vinegar, 1987, X-ray computerized tomography: Journal of Petroleum Technology, v. 39, n. 8, p. 885-898.
Wendel, D. J., W. G Anderson, and J. D. Meyers, 1985, Restored-state core analysis for the Hutton reservoir: 60th Annual Technical Conference of the Society of Petroleum Engineers, Las Vegas, NV, Sept. 22-25, SPE14298.
Wendt, W. A., S. Sakurai, and P. H. Nelson, 1986, Permeability prediction from well logs using multiple regression, in Lake, L. W., and H. B. Carroll, Jr., eds., Reservoir Characterization: Orlando, FL, Academic Press, 659 p.
Wilson, D. A., and W. M. Hensel, Jr., 1978, Computer log analysis plus core analysis equals improved formation evaluation in West Howard-Glasscock Unit: Journal of Petroleum Technology, v. 30, p. 43-51.
Wilson, W. D., and E. D. Pittman, 1977, Authigenic clays in sandstones: Recognition and influence on reservoir properties and paleoenvironmental analysis: Journal Sedimentary Petrology, v. 47, p. 3-31.
Wolfe, J. A., 1977, Paleogene floras from the Gulf of Alaska region: U.S. Geological Survey Professional Paper 997, p. 108.
Yan, J. N., J. L. Menezes, and M. M. Sharma, 1988, Wettability alteration due to interaction with oil-based mud components: 63rd Annual Technical Conference and Exhibition of the Society of Petroleum Engineers, Houston, TX, Oct., SPE 18162.
Yuan, H. H., and B. F. Swanson, 1986, Resolving pore space characteristics by rate-controlled porosimetry: 5th Symposium on Enhanced Oil Recovery of the Society of Petroleum Engineers and the Department of Energy, April, SPE/DOE 14892,9 p.
Zarella, W. M., et al. 1967, Analysis and significance of hydrocarbons in subsurface brines: Geochimica et Cosmochimica Acta, n. 13, p. 1155-1166.
Part6
GEOLOGICAL METHODS
edited by
Roger M. Slatt
Department of Geology and Geological Colorado School of Mines Golden, Colorado, U.S.A.
Engineering
Contents
• Introduction • Lithofacies and Environmental Analysis of Clastic
Depositional Systems • Carbonate Reservoir Models: Facies, Diagenesis,
and Flow Characterization • ReservoirQuality • Geological Heterogeneities • Flow Units for Reservoir Characterization • Effective Pay Determination • Conversion of Well Log Data to Subsurface
Stratigraphic and Structural Information • SubsurfaceMaps • Geological Cross Sections • HuidContacts • Evaluating Stratigraphically Complex Fields • Evaluating Diagenetically Complex Reservoirs • Evaluating Tight Gas Reservoirs • Evaluating Fractured Reservoirs • Evaluating Structurally Complex Reservoirs • StatisticsOverview • Correlation and Regression Analysis • Multivariate Data Analysis • Monte Carlo and Stochastic Simulation Methods • ReferencesCited
Introduction
Roger M. Slatt1
Department of Geology and Geological Engineering Colorado School of Mines Golden, Colorado, U.S.A.
In recent years, it has become increasingly apparent that proper geological reservoir description can lead to improved reservoir management and increased recovery efficiency of oil and gas reservoirs. Case studies illustrating successful applications of geological reservoir description have been appearing more frequently in technical journals and in technical convention programs. Development geology groups are appearing on more oil and gas company organization charts, and internal training programs include more development geology and reservoir description components. Traditional methods of reservoir description are being improved by more sophisticated geophysical, geostatistical, and computer technologies. Perhaps most significantly, geological advice is more widely solicited by reservoir engineers and management. With increased emphasis in the oil and gas industry upon prudent expenditures and maximization of hydrocarbon recovery from existing fields, it is anticipated that geological reservoir description will be a routine part of reservoir management well into the future.
Thus, a greater n u m b e r of engineers, geophysicists, geostatisticians, geologists, and managers will be exposed to reservoir geological description principles and practice in coming years. Although many of these people will not need to be experts in the area of geological reservoir description, some basic knowledge will be necessary for them to communicate with reservoir geologists. With this in mind, the purpose of Part 6 is to provide a handy desktop reference on standard methods, concepts, terminology, and approaches to geological reservoir description. This part of the Manual is informally divided into three general themes: (1) specific concepts, f u n d a m e n t a l principles, and techniques; (2) applications to the evaluation of different types of reservoirs; and (3) some fundamental statistical procedures that are applicable to reservoir characterization.
The first theme comprises ten chapters. The first two chapters outline the salient characteristics of clastic (Scheihing and Atkinson) and carbonate (Lucia) reservoirs and present basic descriptive techniques. The next chapter (Grier and Marschall) focuses on the various geological controls on reservoir quality (porosity, permeability, saturations, wettability, etc.) and standard methods for reservoir quality assessment. The next two chapters outline a hierarchy of classification of reservoir properties from the single well to the fieldwide scale (Slatt and Galloway) and describe an approach to translating geological and petrophysical properties into reservoir "flow units" (Ebanks et al.). Other
chapters present methods for determining reservoir net pay (Gaynor and Sneider), converting well log data to subsurface stratigraphic and structural information (Boak), constructing subsurface maps (Weissenburger) and cross sections (Boak), and evaluating fluid contacts (Brown).
The second theme comprises five chapters. These chapters present, in part, information summarized in earlier chapters, as well as additional details for evaluating stratigraphically complex reservoirs (Morton-Thompson and Galloway), diagenetically complex reservoirs (Wilson), tight gas reservoirs (Moslow), fractured reservoirs (Nelson), and structurally complex reservoirs (Hossack and McGuinness).
The third theme comprises four chapters. The first chapter is a discussion of common statistics and hypothesis testing (Shaw and Coburn), while the second explains simple correlation and regression (Coburn). The third describes multivariate data analysis (Esbensen and Journel), and the final chapter describes Monte Carlo and stochastic simulation (Journel).
Emphasis throughout Part 6 is on providing a general overview that can serve as a starting point for individuals seeking information on specific technologies and methods for geologically evaluating a reservoir. It is not intended as a definitive work on any topic but rather as a source book of basic information. Key published references are provided at the end of the chapter to g u i d e individuals to in-depth reviews of topics of interest, and individual chapters are extensively cross-referenced.
Acknowledgments
The authors of Part 6 hope that their collection of papers provides a convenient means of increasing communication a m o n g disciplines for the c o m m o n goal of i m p r o v i n g hydrocarbon recovery. Special thanks go to the outside reviewers Robert M. Sneider (of Robert M. Sneider Exploration, H o u s t o n ) and Marcus E. Milling (of the American Geological Institute).
The special editors of the statistics chapters, A. G. Journel and Т. C. Coburn, gratefully acknowledge the support and assistance of R. A. Olea, T. A. Jones, L. W. Lake, T. A. Hewitt, H. H. Haldorsen, G. E. Fogg, and H. H. H a r d y in the compilation of the information contained in the "Statistics Overview" chapter, as well as in the remaining statistics chapters.
1Formerly with ARCO International Oil and Gas Company, Piano, Texas, U.S.A.
261
262
PART 6—GEOLOGICAL METHODS
The individual authors collectively acknowledge their respective employers for providing the time and permission to publish these contributions:
C. D. Atkinson Jeremy Boak Alton Brown T. C. Coburn William J. Ebanks, Jr. K. H. Esbensen William Galloway Gerard C. Gaynor Susan Grier Jake R. Hossack A. G. Journel
ARCO Indonesia, Inc. U.S. Department of Energy ARCO Oil and Gas Marathon Oil Company ARCO Indonesia, Inc. Norwegian Computing Center The Univ. of Texas at Austin Reservoir Geosystems Inc. Core Laboratories BP Exploration Stanford University
F. Jerry Lucia
D. M. Marschall D. В. McGuinness Diana Morton-Thompson Thomas Moslow Ronald Nelson Mark H. Scheihing Brian R. Shaw
Roger M. Slatt Robert M. Sneider
Kenneth W. Weissenburger Michael D. Wilson
The Univ. of Texas Bureau of Economic Geology
Core Laboratories BP Exploration ARCO Oil and Gas Univ. of Alberta Amoco Production Company ARCO Oil and Gas Battelle Pacific Northwest
Laboratories Colorado School of Mines Robert M. Sneider Exploration
Inc. Conoco Inc. Consultant
Lithofacies and Envi. ronmental Analysi.s or.
J
Clastic Depositional Systems
Mark H Scheihing
Christopher D. Atkinson
ARCO Oil and Gas Company
plano'Texas'asA-
INTRODUCTION
The geological a n d reservoir properties of s e d i m e n t a r y rocks d e p e n d u p o n an interplay of tectonics, sea level, s e d i m e n t s u p p l y , p h y s i c a l a n d biological p r o c e s s e s of sediment transport and deposition, and climate. At the basin scale, these processes interact to produce the geometric a r r a n g e m e n t of d i f f e r e n t d e p o s i t i o n a l e n v i r o n m e n t s or systems tracts through time, known as the stratigraphic architecture of the basin (Miall, 1984). At smaller scales, these processes control the external geometry and internal " a n a t o m y " of clastic sediment bodies (see the chapter on "Geological Heterogeneities" in Part 6). It is at this smaller scale that lithofacies a n a l y s i s a n d i n t e r p r e t a t i o n of depositional environments become important for reservoir evaluation.
DATA REQUIREMENTS
Basic data requirements for facies analysis of subsurface rocks are listed in Table 1. Data associated with wells are most often used, but seismic data, particularly threedimensional data, are becoming increasingly important in defining sandstone body geometries (e.g., see Brown, 1986; also see the chapter on "Three-Dimensional Seismic Methods for Reservoir Development" in Part 7). Conventional core is perhaps the most diagnostic for sedimentological interpretation of vertical sequences (see the chapter on "Core Description" in Part 5). However, wireline tools such as
dipmeters and formation imaging devices can provide electrical images suitable for sedimentological interpretation with the added ability to determine paleocurrent directions in appropriate cases (see the chapters on "Dipmeters" and "Borehole Imaging Devices" in Part 4).
ROCK DESCRIPTION
Within any given depositional environment, various physical and biological processes act to transport and deposit sediment. These processes result in various distributions of grain size and sedimentary structures that characterize the deposited sediment. Relating these features back to the processes that produced them is the basic method used by geologists to i n t e r p r e t the d e p o s i t i o n a l e n v i r o n m e n t of sedimentary sequences (Figure 1).
Lithofacies
O n e of the first steps in the facies analysis of a clastic reservoir is the description and interpretation of available conventional core (Siemers and Tillman, 1981). An important result of core description is the s u b d i v i s i o n of cores into lithofacies, defined as subdivisions of a sedimentary sequence based on lithology, grain size, physical and biogenic sedimentary structures, and stratification that bear a direct relationship to the depositional processes that produced them. Lithofacies a n d lithofacies associations (groups of related lithofacies) are the basic units for the interpretation of depositional environments.
Table 1. Types of Data Commonly Used in Facies Analysis of Clastic Rocks
Data Type
Slabbed conventional core Oriented conventional core Core gamma scan Sidewall cores, well cuttings, thin sections Paleontology (micro, macro, trace fossil), palynology Dipmeter log Formation MicroScanner (FMS) log Spontaneous potential log Gamma ray log Sonic log Caliper log Neutron log Density log Repeat formation test (RFT)
Application
Facies, depositional environment Paleocurrent directions Shift to wireline logs Mineralogy, lithology Water depth, depositional environment, time lines Paleocurrent directions, lithofacies Paleocurrent directions, lithofacies Lithology, curve shape analysis Lithology, curve shape analysis Porosity, curve shape analysis Borehole condition, quality control Porosity, lithology (with sonic,density) Lithology (coal), porosity (with neutron) Pressure (sand body connectedness)
263
264
PART 6—GEOLOGICAL METHODS
OBSERVED M SEQUENCE
DEPOSITIONAL PROCESSES
LITHOFACIES pebbly sandstone
LITHOFACIES ASSOCIATIONS
fl ood-slage deposition of silts
and clays and soil formation
flood-stage deposition of sand,
silt and clay
NORTHWEST >75 PERCENT SANDSTONE
- winnowing — erosion—
pebbly sandstone
Figure 1. Sedimentary processes, lithofacies, and Iithofacies associations for a meandering channel sequence. (The vertical sequence is modified from Walker and Cant, 1984.)
Depositional Environments
Interpretation of the environment in which lithofacies were deposited from analysis of cored sequences involves relating the identified lithofacies to the physical and biological processes that produced them. This process-response relationship identifies the specific processes responsible for the sequence and, by inference, the depositional setting in which these processes occurred. The application of the process-response approach relies primarily on depositional m o d e l s constructed t h r o u g h s t u d y of both m o d e r n and ancient analogs.
Depositional models are important for predicting the distribution of permeability and porosity within different reservoir types. These models are never exact matches to a reservoir; rather, they serve as guides to aid in the interpretation of any one reservoir (Walker, 1984). Reservoir properties are generally observed to be correlative with lithofacies types to one degree or another (see the chapter on "Geological Heterogeneities" in Part 6). This reflects the fundamental control on permeability and porosity by grain size, sorting, and spatial distribution of different lithofacies types. Even where rocks have experienced later physical and chemical diagenesis, permeability and porosity relationships are controlled, in large part, by the original sedimentary fabric of the rock.
Wireline Log Calibration and Correlation
Interpretations of depositional e n v i r o n m e n t based on individual well data are transformed into a three-dimensional picture of the reservoir by wireline log correlation and, where possible, by three-dimensional seismic data.
Wireline logs to be used for facies analysis should, whenever possible, always be calibrated by core. This calibration involves (1) shifting core to log depths (see the chapters on "Preprocessing of Logging Data" in Part 4 and "Core-Log Transformations and Porosity-Permeability Relationships" in Part 5) and (2) establishing a relationship between lithofacies associations and curve shape. Core gamma scans, obtained by passing the core through a device
Figure 2. Gamma ray correlation (dip section) of a series of prograding shoreface sandstones. Note the imbricate nature of the sandstone bodies and the "non-layer cake" nature of the correlations.
that m e a s u r e s the n a t u r a l radioactivity of the rock, are particularly useful for shifting cores to logs. The calibration of wireline log shape by core is particularly important for firmly establishing the log response a n d the identity of vertical sequences on these logs.
For reservoirs in which no core is available, wireline log shape must be used to interpret sandstone body type and identify depositional environments. If closely spaced cuttings or sidewall cores are available, these can sometimes aid rock to log calibration. Log shapes are deduced from the expected wireline log response of the different environments combined with a knowledge of the paleogeography of the area in which the field is situated. Wireline log shapes are often described as "upward coarsening," "upward fining," or "blocky." However, log shape as determined from a gamma ray or SP log in siliciclastic rocks is related more to argillaceous content than to grain size. Upward coarsening log patterns exhibit an upward decrease in argillaceous content. Upward fining log patterns exhibit the reverse trend. Blocky or cylindrical log patterns exhibit relatively little vertical variation in argillaceous content and are typical of siliciclastic rocks that have low overall argillaceous content. Various publications and reference charts are available to aid in this practice (e.g., Spearing, 1974; Cant, 1984; Rider, 1986). However, without core control, curve shape analysis is fraught with hazards (e.g., see Snedden, 1987; also see the chapter on "Quick-Look Lithology from Logs" in Part 4).
Correlation sections that will be used for establishing sandstone body geometry should have a depositionally flat datum (such as a bentonite bed, marine shale bed, or laterally persistent limestone). Sections should be oriented parallel a n d p e r p e n d i c u l a r to d e p o s i t i o n a l strike, if k n o w n , a n d represent as straight a line as possible given well density and placement.
The only sedimentologically significant correlation horizons are those that approximate time lines within and between sandstone bodies. This style of correlation requires an u n d e r s t a n d i n g of the succession of d e p o s i t i o n a l environments and intervening unconformable surfaces. It often leads to nonparallel and nonhorizontal correlations. For example, in shoreface systems, time lines denoted by shale or silt breaks between shingled shoreface sheets and lenses are inclined in a seaward (depositional dip) direction (Figure 2).
Lithofacies and Environmental Analysis of Clastic Systems
265
This imbrication does not occur in a strike direction. This style of correlation is especially i m p o r t a n t for reservoir delineation since the large scale (interwell and field) architecture of the sandstone b o d y exerts a control on the m o v e m e n t of fluids through the volume of the reservoir.
Wireline log correlation is an exercise in pattern recognition combined with the geometry suggested by the interpreted depositional environment(s). The depositional interpretation of a reservoir exerts a major impact on the methods and style of correlation and mapping. For example, correlation in fluvial systems applies very different assumptions about sandstone continuity than does correlation in shoreface or shelf systems.
CLASTIC DEPOSITIONAL LITHOFACIES AND ENVIRONMENTS
Clastic depositional environments range from alpine to abyssal settings (Figure 3 a n d Table 2). Detailed reviews of these are given by Galloway and Hobday (1983), Walker (1984), Berg (1986), Reading (1986), Beaumont and Foster (1987), and others. The following review is a cursory s u m m a r y of the origin, lithofacies, geometry, and reservoir properties of major clastic environments and deposits. The reader should be aware that the remarks offered here for each d e p o s i t i o n a l e n v i r o n m e n t are necessarily of a h i g h l y generalized and idealized nature. Siliciclastic reservoirs are typically composed of multiple bodies deposited (and eroded) through time under varying tectonic, sealevel, and climatic conditions. Corresponding geometry, vertical sequence, wireline log character, and reservoir quality trends for a given reservoir may be, and often are, different from the generalized "single environment" models. In addition, subsequent diagenesis (see the chapter on "Evaluating Diagenetically Complex Reservoirs" in Part 6) m a y alter the permeability and porosity structure created by depositional (and erosional) processes. However, it has been often observed that in siliciclastic rocks, diagenesis generally follows depositional fabric (see the chapter on "Geological Heterogeneities" in Part 6). Complex structural patterns can reduce reservoir continuity as well.
Alluvial Fan Deposits
An alluvial fan is a w e d g e of clastic detritus that forms at the base of a mountain front as sediments eroding from the mountains are transported downslope by streams or debris flows and deposited at the base (Figure 3e). The fan-shaped body is generally characterized by a gradation from coarser sediments at the apex to finer sediments at the toe. Alluvial fans are commonly divided into proximal, mid-fan, and distal fan subenvironments.
Vertical sequences through the proximal fan are generally dominated by gravelly deposits with subordinate sandy deposits. Sequences through the mid- and distal fan are increasingly sand dominated. Gamma ray, SP, and resistivity log responses throughout a fan can generally be expected to be blocky to irregular, d e p e n d i n g on the a m o u n t of clay.
Permeability a n d porosity of alluvial fan deposits v a r y greatly as a function of depositional process a n d differential
response to diagenesis. In general, streamflow deposits have greater permeability and porosity than debris and mudflow deposits. Finer grained but better sorted distal fan deposits are highly p e r m e a b l e a n d p o r o u s . Because of increased sorting, mid- and distal parts of the fan probably have better and more predictable reservoir quality than proximal parts. Little is k n o w n of directional permeability within alluvial fan reservoirs, but paleochannels can be expected to act as preferred pathways of flow.
Where alluvial fans prograde into standing bodies of water (that is, oceans or lakes), they are called fan deltas. The distal parts of these fans are generally m u c h better sorted and cleaner as a result of r e w o r k i n g by w a v e a n d / o r tidal processes. Proximal and mid-fan log responses are the same as alluvial fans. Log response in the distal part depends upon the intensity of w a v e and tidal processes and on whether the fan is actively prograding or being transgressed. Typically, distal parts will have an upward-coarsening gamma ray, SP, and resistivity log character. Barring adverse diagenetic effects, permeability can be expected to be much greater in m a r i n e than m o r e proximal parts of the fan delta d u e to increased sorting, destruction of compositionally i m m a t u r e grains, and w i n n o w i n g of fines. Directional permeability trends in distal parts may be different from more proximal locations because of different sand body trends between these different parts of the fan delta.
Braided and Meandering Fluvial Deposits
Downdip from alluvial fans, rivers typically grade first into braided channels then, farther down the alluvial valley toward the coastal plain, into meandering channels. These different channel types can occur in the same river system and produce distinctly different kinds of sandstone bodies.
Braided rivers and braidplains form elongate, tabular, sandy and gravelly deposits c o m p o s e d of b r a i d e d , sand-filled channels and sand and gravel bars (Figure 3c). They typically consist of coarse s a n d and gravel with relatively m i n o r a m o u n t s of clay. Vertical sequences are composed of stacked, upward-fining channel sands and sand and gravel bars. Lateral trends in these deposits are dominated by an overall tabular geometry bounded by floodplain muds with an internally c o m p l e x g e o m e t r y of cross-cutting s a n d s a n d gravels with subordinate mud-rich beds of varying thickness and dimension. Bar and channel deposits are typically elongate in the paleocurrent direction.
Meandering rivers are different in that sand is restricted to a single channel and surrounded by fine-grained sediments (Figure 3d). Sand is concentrated mainly in the channel bottoms and point bars. A vertical sequence through such a channel system frequently has an upward-fining character, starting from the channel lag at the bottom and grading u p w a r d into deposits of the adjacent levee and floodplain. I n d i v i d u a l m e a n d e r belts are built of c r o s s - c u t t i n g a n d stacked individual upward-fining sequences often separated laterally by meander loop cutoffs and clay plugs. Multiple m e a n d e r belts are built b y a b a n d o n m e n t of an entire river segment (avulsion) and by establishment of a n e w section in another position on the floodplain.
Gamma ray, SP, and resistivity logs through braided channel complexes generally have a blocky character,
266
PART 6—GEOLOGICAL METHODS
Figure 3. Models of major depositional environments. The curve on the left shows the SP or gamma ray response and the curve on the right shows the relative grain size profile. The size of the dots next to the vertical profile indicates the relative magnitude of permeability expected in such a sequence. (Parts с and d are from Walker, 1984, and parts f, h, and i are from Galloway and Hobday, 1983.)
whereas individual meandering channels have an upwardfining signature except where stacked and cross-cut, where they may exhibit more complex wireline log signatures.
The upward-fining character of fluvial channels tends to produce sandstone bodies that have their greatest permeability at the base of the body. However, the common stacking and cross-cutting of channels in both braided and meandering river deposits often produces a complex spatial distribution of permeability within the braided or meander belt. Preferred permeability pathways, and consequently, fluid flow, can be expected to follow the paleochannel direction (Qiu Yinan, 1984).
Eolian Deposits
Eolian sands develop in arid settings and commonly form extensive, blanket-like deposits (Figure 3b). Wind transport removes fines and produces rounded and extremely well sorted grains often leading to favorable reservoir quality.
This c o m b i n a t i o n of w i d e s p r e a d occurrence a n d good reservoir properties makes eolian sandstones attractive exploration targets and many hydrocarbon accumulations have been discovered in such deposits (see Ahlbrandt and Fryberger, 1982).
Eolian deposits include dune, interdune, sand sheets (marginal to dune complex), and extradune (noneolian) lateral deposits (Ahlbrandt and Fryberger, 1982). Dune deposits comprise the major sedimentary bodies in eolian successions. All are characterized by large scale cross stratification in which foreset dips range up to 35°. Associated deposits may include those of wadi (fluvial), playa (lacustrine), and sabkha (arid tidal flat) origin.
In the subsurface, eolian sandstones generally comprise thickly bedded sequences with few major interstratified shales. The sequences tend to be uniform and lack discernible coarsening- or fining-upward trends and, thus, exhibit blocky to weakly serrated gamma ray, SP, and resistivity log profiles The well-bedded and high angle cross stratified n a t u r e of
Lithofacies and Environmental Analysis of Clastic Systems
267
Table 2. Major Clastic Depositional Environments
Alluvial sediments Alluvial fans Fan deltas Braidplains Braided rivers Meandering rivers
Lacustrine sediments Lake deltas Freshwater lakes Saline lakes
Eolian sediments Dunes Interdune deposits Sand sheets
Deltaic sediments River-dominated deltas Wave-dominated deltas Tidally dominated deltas
Siliciclastic shoreline sediment Wave-dominated shorelines Beaches Microtidal barrier islands Cheniers Mixed wave-tide influenced shorelines Mesotidal barrier islands Tide-dominated shorelines Tidal flats Estuaries
Shallow siliciclastic seas Tidal sand ridges Sand shoals Sand sheets
Deep marine slope and basin Slope channel and gully deposits Slope canyon deposits Intraslope basin deposits Sand spillover sheets Slope aprons Submarine fans
Glacial sediments Supraglacial Glaciofluvial Glacioeolian Glaciolacustrine Glaciomarine
eolian sandstones promotes reliable results from dipmeter logs. Dune and interdune deposits can often be distinguished and paleowind directions inferred using correctly processed dipmeter data (Lupe and Ahlbrandt71979).
Eolian sandstones generally comprise excellent reservoir intervals but often possess complex porosity and permeability variations. They are commonly anisotropic with regard to the flow of fluids a n d exhibit greater horizontal than vertical permeability because of their pronounced lamination (Weber 1987).
Deltas
Deltaic bodies are generally classified into three major categories or e n d - m e m b e r s on the basis of the d o m i n a n t sediment transport process that influences their facies constituents and external geometries. These three endmembers are as follows:
1. River- or fluvially dominated deltas (such as the Mississippi River delta) are those in which wave and tidal energy is low and river transport processes dominate. Sandstone bodies in these systems tend to form well-developed sand bars at the m o u t h s of distributary channels. River-dominated systems periodically abandon their lower course and begin deposition in an adjacent area resulting in the deposition of sandstone bodies over a fairly large area.
2. Wave-dominated deltas (such as the Nile and Rlione deltas) are those in which wave energy at the coast exceeds either the fluvial or the tidal energy. Wave reworking causes sand to be formed into shore-parallel bodies that are cuspate at distributary mouths.
3. Tidally dominated deltas (such as the Gulf of Papua delta) are those in which the tidal energy exceeds that of either wave or fluvial processes. Sand deposited by the distributaries is reworked by tidal currents into elongate sand ridges that are generally perpendicular to the regional coastline.
Rarely do deltas conform perfectly to these end-members. In general, they are transitional, giving rise to complexity and variability in the geometry and reservoir heterogeneities of resulting sandstone bodies (e.g., Sneider et al., 1978).
Distributary mouth bars and channel deposits (Figure 3h) comprise the best reservoir quality bodies within a delta s y s t e m . T h e g e n e r a l u p w a r d - c o a r s e n i n g c h a r a c t e r of distributary mouth bars tends to produce sandstone bodies that have their greatest permeability at the top. Conversely, distributary channel sandstone bodies are usually upwardfining and have their greatest permeability at the base (Sneider et al., 1978). Preferred orientation of flow m a y be expected to follow paleochannel trends.
Distributary mouth bars typically contain a high p e r c e n t a g e of interstratified clay that r e d u c e s vertical permeability. Hartman and Paynter (1979) document an example of such behavior from a Gulf Coast deltaic reservoir u n d e r g o i n g n a t u r a l w a t e r d r i v e . A f t e r several y e a r s of production, the better quality channel sands watered out, whereas oil remained in the poorer quality delta fringe deposits. In this field, by-passed oil was accessed by recompleting wells only in the delta fringe interval.
Wave modification acts to winnow delta mouth bar sandstones thus increasing their reservoir quality. In addition, wave processes and longshore currents enhance overall reservoir potential in deltas by redistributing sand as beach and chenier deposits in the interdistributary areas. Tidal reworking can affect reservoir quality at the delta mouth either by acting to winnow fines from the sands or by reducing effective permeability in distributary channels by the introduction of increased a m o u n t s of interstratified clay.
268
PART 6—GEOLOGICAL METHODS
Lacustrine Deposits
Lakes occur in a wide variety of geological settings. They are often very important d u r i n g the early rifting phase of basin formation on continental crust. Major hydrocarbonbearing lake deposits are associated with very large and longlived Tertiary lakes such as those of the western United States, Indonesia, and China. These deposits are characterized by siliciclastic, carbonate, and organic-rich sediments deposited under generally low energy conditions, often by suspension deposition. Other processes include turbidity flows in the lake interior and wave and current reworking along the lake margin.
Lacustrine rocks are generally the source rocks for hydrocarbons found in alluvial fan, fluvial, eolian, and deltaic rocks rather than the reservoirs. However, sandstone bars, beaches, turbidites, and fan deltas associated with lake margins can be reservoirs sourced by open lake deposits. The core and log response characteristics of these deposits are similar to those described from analogous marine environments.
Shoreline Deposits
In shoreline systems adjacent to active deltas, the geometry and internal anatomy of sandstone bodies are controlled by an interplay of tidal and wave processes. Clastic, nondeltaic shorelines with a tidal range of 0-2 m (microtidal) tend to be wave-dominated. Resulting sand bodies are elongate barrier islands and strandplains. A tidal range of 2-4 m (mesotidal) tends to produce short ("drum stick") barrier islands with extensive tidal flats and ebb tidal deltas. A tidal range of 4-6 m (macrotidal) tends to produce estuarine linear tidal sand ridges that are perpendicular to shoreline with associated extensive tidal flats.
Barrier islands (Figure 3f) illustrate the spatial variability in facies that affect reservoir properties. Sands in the beach or foreshore are very well sorted, lack interstratified clay, and exhibit excellent reservoir properties where not cemented. Tidal inlet and flood tidal delta deposits comprise another important grouping of reservoir quality rocks, particularly because they are most often preserved in the rock record.
Wireline log shapes through barrier island sequences vary depending on exactly where a well intersects the barrier island complex. Gamma ray, SP, and resistivity logs through the barrier core have an upward-coarsening motif (Figure 3f). Logs through the back barrier and lower shoreface are typically highly serrate and often lack a well-defined upwardcoarsening motif. Logs through the barrier inlet may exhibit upward fining.
In general, barrier islands have the best reservoir quality rocks at the top of the sequence. Reservoir quality drops off as one moves either seaward down the foreshore and shoreface into m u d s of the marine shelf or landward into the lagoon. High reservoir quality is also developed within the tidal inlet sandstones. Two major trends in directional
permeability are suggested by (1) the shore-parallel nature of foreshore and shoreface s a n d s t o n e s and (2) shoreperpendicular tidal inlet and delta sandstones. In coastlines dominated by tidal processes, extensive interbedded mud and sand "flats" occur in the intertidal area of the coast and sand bars in estuarine channels in the subtidal area. The reservoir quality of tidal flat environments varies as a function of sand to m u d ratio of the deposits. Reservoir quality of estuarine channel deposits also varies as a function of sand to m u d ratio and degree of bioturbation.
Shallow Marine Clastic Deposits
The marine shelf is an environment affected by storm- and tidal-driven waves and currents and sometimes by oceanic currents. Although shelf sand ridges of either storm or tidal origin formed during transgression are the best known examples (Figure 3g), sand bodies associated with the marine shelf also include reworked delta front and barrier sands, amalgamated storm sheets, and oceanic current deposits (Barwis, 1989).
Most marine sand bodies are upward coarsening with the best reservoir quality rocks at the top of the body. Gamma ray, SP, and resistivity logs have a corresponding upwardcoarsening character. In the case of storm-deposited sheet sands either attached or detached from the shoreface, amalgamation of individual storm deposits at the top of the bodies produces the greatest permeability and porosity and the most laterally continuous units (Atkinson et al., 1986; Scheihing and Gaynor, 1988). In the case of tidal- and stormgenerated shelf sand ridges, best reservoir quality is also at the top in the form of several different types of large scale cross bedding.
Deep Water Marine Deposits
Reservoir quality sand bodies form on both the continental slope a n d at the base of the slope. Slope e n v i r o n m e n t s include sand bodies formed within submarine canyons and gullies cut into the slope and as spillover sheets (Slatt, 1986). Sands can also accumulate on tectonically formed small basins within the slope itself.
Submarine fans may form at the base of slopes that have a delta-like appearance in plan view (Figure 3i). Internal facies vary from channelized sand and gravel bodies to sheet-like, thin, graded beds deposited by turbidity flows in distal parts of the fan. Vertical sequences through channelized portions of the fan typically s h o w an u p w a r d - f i n i n g character accompanied by an upward-fining wireline log motif. Vertical sequences through more distal parts of the fan show an alternation between sandstone and mudstone beds, so that wireline logs are typically interdigitate and irregular. Reservoir quality varys accordingly. M a n y variations of morphologies and internal facies configurations occur in submarine fans as a function of sediment supply, sea level, type of continental margin, and local tectonic features.
Carbonate Reservoir Models: Facies, Diagenesis, and Flow Characterization
F Jerry Lucia
The University of Texas at Austin Bureau of Economic Geology Austin, Texas, U.S.A.
INTRODUCTION
T h e f l o w characteristics of c a r b o n a t e r e s e r v o i r s are controlled by a combination of depositional and diagenetic processes. Depositional processes control the initial pore size distribution and the geometry of the individual depositional facies. The diagenetic overprint modifies the pore size distribution and controls the productivity of depositional facies. In some cases, reservoir quality and flow characteristics are totally controlled by diagenesis, as in karsted reservoirs.
Carbonate reservoir descriptions are based on observations of depositional and diagenetic fabrics and pore space from core and cuttings samples. The descriptions are correlated with wireline log responses and incorporated into geological facies a n d / o r diagenetic models in order to map porosity, saturation, and permeability.
CARBONATE SEDIMENTS AND ENVIRONMENTS
The majority of carbonate sediments are p r o d u c e d in shallow, w a r m ocean w a t e r s by extraction of calcium carbonate from seawater by organisms to form their shells or skeletal material. The sediments are composed of a spectrum of sizes a n d pore geometries. The D u n h a m classification (Figure 1) describes depositional textures in a m a n n e r that can be related to pore geometries. The grain-supported textures tend to have larger pore sizes than do mudsupported textures. The textures have different geometries in different depositional environments.
There are five basic carbonate depositional environments. From shore to basin, they are peritidal (tidal flat), shallow shelf interior, shelf inargin complex, slope, and basin (Figure 2).
The peritidal depositional environment is complex (Figure 2). Sediments deposited between mean high and mean low tide are called intertidal sediments, sediments deposited above m e a n high tide are called supratidal sediments, and sediments deposited below m e a n low tide are called subtidal sediments. In arid and semi-arid climates, evaporite flats (sabkhas) are present from which gypsum and halite are deposited. EoHan sand d u n e s composed of siliciclastic or carbonate grains m a y form on the supratidal surface.
The shallow shelf interior environment (Figure 2) is dominated by low energy waters that allow lime mud to accumulate. Storms, however, churn the sediment into suspension, winnowing out the fine-sized material and concentrating the coarse material. Near shorelines, the shelf
e n v i r o n m e n t m a y be c o m p o s e d of offshore bars a n d spits oriented parallel to shoreline. Shorelines that face heavy wave action accumulate carbonate sand or gravel. Tidal currents are concentrated in channels between islands and produce tidal deltas on the lee side of the island.
The shelf margin complex (Figure 2) is characterized by the p r e s e n c e of c a r b o n a t e s a n d s a n d reefs. Reefs are commonly found at the shelf edge where their rigid framework can withstand strong wave action and they can take advantage of the nutrients upwelling from the deeper waters. Carbonate sands derived from a reef or from plants and animals inhabiting the shelf edge accumulate along a wide belt that follows the break between the shelf edge and the slope. Tidal and storm currents mold the sand belt into tidal channels and bars.
The slope (Figure 2) is dominated by sediment transport seaward from the shelf margin. Fine-grained sediment settles to the bottom forming thin-bedded mudstones, while slumps, debris flows, and turbidity currents form coarsegrained bodies of breccia, conglomerate, and carbonate sand. The resulting facies patterns depend u p o n the relief of the shelf margin and the nature of the shallow water portion of the margin (Mcllreath and James, 1984).
S e d i m e n t s of the basin e n v i r o n m e n t (Figure 2) are d o m i n a t e d by very fine grained skeletons of planktonic microorganisms, which when lithified, become chalks. The are found in deep basins as well as on drowned shelves.
DIAGENESIS
The r e s e r v o i r q u a l i t y of s e d i m e n t s is m o d i f i e d by diagenetic processes. Some diagenetic processes are directly related to the physical n a t u r e of the s e d i m e n t , a n d the resulting geometries can be directly related to sedimentation patterns. Other diagenetic processes involve the interaction of sediment with flowing interstitial water. Understanding hydrodynamics is very important in predicting the geometry of the diagenetic effects.
Calcium Carbonate Cementation and Compaction
Calcium carbonate cementation and compaction are diagenetic processes that are initiated soon after deposition and by which intergranular pore space is progressively reduced producing systematic changes in petrophysical properties (see the chapters on "Evaluating Diagenetically Complex Reservoirs" and "Reservoir Quality" in Part 6). Compaction is a physical/chemical process, while cementation requires fluid flow. Microporosity within grains
269
270
PART 6—GEOLOGICAL METHODS
DEPOSITIONAL TEXTURE RECOGNIZABLE
Original Components Not Bound Together During Deposition
Contains mud ( particles of clay and fine silt size )
Mud-supported
Grain-supported
Lacks mud and is
grain-supported
Less than L O percent grains
Original components were bound together during deposition... as shown by intergrown
skeletal matter, lamination contrary to gravity, or sediment-floored cavities that are roofed over by organic or questionably organic matter and are too large to be interstices.
DEPOSITIONAL TEXTURE NOT RECOGNIZABLE
Crystal line Carbonate
( Subdivide according to classifications designed to bear on physical texture or diagenesis.)
Figure 1. Dunham (1962) classification of carbonate rocks according to depositional texture. (From Swanson, 1981.)
or between lime mud particles may be retained even though intergranular pore space is lost.
Dolomitization
Dolomitization is a d i a g e n e t i c p r o c e s s t h a t c o n v e r t s limestones to dolostones through a microchemical process of calcium carbonate dissolution and dolomite precipitation. Dolomitization can change the rock fabric and the petrophysical properties significantly because the dolomite crystals are commonly larger than the replaced limestone particles. Dolomite cement systematically grows on the dolomite crystal faces, reducing reservoir quality. Dolomitization requires the addition of large quantities of m a g n e s i u m through fluid flow.
Evaporite Mineralization
The most common evaporite mineral found with carbonate rocks is anhydrite and its hydrous form, gypsum. Gypsum is the common form at shallow depths, but it converts to anhydrite at depth in response to higher temperatures. Bedded anhydrite is commonly found in tidal flat environments and is an effective reservoir seal. Diagenetic anhydrite is found in reservoir rocks as nodules and poikilotopic crystals, which have little effect on reservoir properties, and as pore-filling crystals, which reduce reservoir quality.
Dissolution and Associated Processes
Dissolution is the diagenetic process by which carbonate and evaporite minerals are dissolved and removed, thus creating and modifying pore space in reservoir rocks (see the chapter on "Reservoir Quality" in Part 6). The effect of this process on permeability depends upon the geometry and location of the resulting voids relative to the rock fabric. In some cases, dissolution is fabric selective and results in f o r m a t i o n of isolated v u g s . In o t h e r cases, d i s s o l u t i o n
enlarges fractures and interparticle pores resulting in large, connected vugs. If the v u g s are large enough, the roof m a y collapse, forming a floor breccia and fractured roof.
CARBONATE ROCK FABRIC AND PETROPHYSICAL RELATIONSHIPS
Petrophysical Classification of Carbonate Pore Space
The Archie (1952) classification was the first attempt at relating rock fabrics to petrophysical rock properties. Lucia (1983) improved on the Archie classification by defining the pore size distribution in relationship to the particle size and the spatial relationship of the pore space to the particles.
Interparticle porosity (Lucia, 1983) is d e f i n e d as that pore space located between grains (intergranular porosity) or between crystals (intercrystalline porosity), but which is not significantly larger than the grains or crystals. Vuggy porosity is defined as pore space larger than or within rock particles. Isolated vugs are called separate vugs, while vugs that form a connected pore system on a reservoir scale are called touching vugs. (For more information on porosity, see the chapter on "Porosity" in Part 5.)
Rock Fabric and Petrophysical Property Relationships
The rock fabric elements important to petrophysical properties are particle size, interparticle porosity, separate v u g porosity, a n d the presence or absence of touching vugs. These elements can be identified and quantified by microscopic examination of core material and related to petrophysical properties measured on the same material.
In nonvuggy carbonates, permeability and capillary properties are a function of interparticle porosity a n d particle size. The addition of separate v u g porosity to a carbonate reservoir rock increases storage capacity, but has little effect on permeability. Separate vug porosity, however, has a large effect on the Archie m factor used in water saturation
MESTOne
TOPOGBA^
SOLUTION COLLAPSE
BRECCIA
PALEOTOPOG RAPHY
SUBUNCONFORMITY TRUNCATION
P
о a
СЛ O«—».
O ГЪ
БГ
Figure 2. Carbonate depositional environments. (Diagram by R. G. Loucks and C. R. Handford, unpublished.)
KJ 4Vl I—\
272
PART 6—GEOLOGICAL METHODS
calculations from wireline logs and on the acoustic transit time. Touching vug pore systems dominate fluid flow and override the effects of interparticle porosity, particle size, and separate vug porosity.
Relationships Between Rock Fabric and Wireline Log Response
Since most fields have few cores available, wireline logs must be used to identify rock fabric elements. This requires the calibration of wireUne log responses with core data.
Particle size can be determined from gamma ray, "porosity", and resistivity logs. Grain-supported rocks commonly have lower gamma ray activity then do mudsupported rocks. Other fine-grained rocks, such as shaley and organic-rich carbonates, commonly have the highest gamma ray activity. However, the level of g a m m a ray activity in some carbonates (dolostones in particular) is not related to particle size b e c a u s e of the p r e s e n c e of a n o m a l o u s concentrations of uranium. Water saturation is a function of particle size and interparticle porosity, a n d crossplots of porosity, water saturation, and reservoir height can be used to determine particle size.
Interparticle porosity can be determined by subtracting separate vug porosity from total porosity. Total porosity can be calculated from porosity logs, while separate vug porosity can be estimated f r o m crossplots of acoustic transit time versus crossplot porosity. Touching vug pore systems can be identified using borehole televiewer and resistivity scanner logs. (For more details, see the chapter on "Borehole Imaging Devices" in Part 4.)
Lithology can have a major effect on porosity logs. G y p s u m is of particular concern in carbonate reservoirs because it contains large quantities of b o u n d water. The bound water is seen by neutron logs as porosity, resulting in erroneously high porosity indications.
CARBONATE RESERVOIR MODELS
The key to describing the geometry and petrophysical p r o p e r t i e s of c a r b o n a t e reservoirs lies in c o m b i n i n g sedimentary and diagenetic descriptions, geological models, rock fabric analysis, and petrophysical rock properties. While each hydrocarbon accumulation has a unique geometry, certain combinations of sedimentation, diagenesis, and pore fabric seem to result in a similar distribution of petrophysical rock properties. Four such reservoir models are (1) the upward-shoaling cementation and compaction model, (2) the subtidal-supratidal dolomitization and sulfate emplacement model, (3) the karst-collapse model, and (4) the geological reef model.
Upward-Shoaling Cementation and Compaction Reservoir Model
The upward-shoaling model is based on a depositional m o d e l of s e d i m e n t a g g r a d i n g to sea level. As the w a t e r shallows, the energy conditions increase, resulting in a vertically stacked sequence from mudstones and wackestones at the bottom to packestones and grainstones at the top (Figure 3).
Cementation and compaction occur with burial, reducing the reservoir quality of the mud-supported mudstones and wackestones more than the grain-supported packestones and grainstones. The result is a vertical s e q u e n c e of lower porosity and permeability mud-supported rocks at the base and higher porosity and permeability grain-supported rocks at the top. Consequently, the best quality flow unit occurs at the top of the sequence (see the chapter on "Flow Units for Reservoir Characterization" in Part 6). The permeable units are confined to the grainstone bars.
If the grainstone bars are exposed to meteoric diagenesis, significant separate vug porosity can develop, causing a loss of permeability while retaining porosity. In the dolomite environment, grainstone bars are commonly cemented with anhydrite, and intercrystalline porosity in mud-supported sediments forms the permeable facies.
Subtidal-Supratidal Dolomitization and Sulfate Emplacement Reservoir Model
The subtidal-supratidal model (Figure 3) is based on the transport of carbonate sediment onto the shore by storm and tidal currents resulting in the progradation of the tidal flat environment over the subtidal environment. Subtidal i n t e r v a l s are c o m m o n l y c o m p o s e d of m u d s t o n e s , wackestones, packstones, and grainstones in no predictable order. When present, grain-supported sediments may be concentrated in the upper part of the subtidal section in the form of offshore bars and high energy shoreface deposits. Intertidal and supratidal sediments are typically muddy except in association with high energy subtidal sediments. A typical vertical sequence would show intercalated mud- and grain-supported sediments in the subtidal interval overlain by algal mats in the intertidal interval, and mud-cracked and desiccated wackestones and mudstones in the supratidal interval.
The subtidal-supratidal sequence is commonly dolomitized and contains anhydrite and gypsum. In the subtidal interval, dolomitized grainstones retain their intergranular pore space, except where cemented by anhydrite, and form permeable units. Dolomitization of the subtidal mud-supported sediments converts the tight, mudsupported limestones to permeable units because of the larger dolomite crystals and intercrystalline pore space. This produces two types of flow units in the subtidal interval: a dolomud-supported flow unit and a dolograin-supported flow unit. Each will have a unique porosity-permeability transform.
Karst-Collapse Reservoir Model
The karst-collapse model describes a touching vug pore system that is formed by massive dissolution of carbonate resulting from meteoric groundwater flow and subsequent collapse and filling of caverns (Figure 4). This process is i n d e p e n d e n t of the s e d i m e n t ' s original e n v i r o n m e n t of deposition. Most karsted carbonates are thought to be related to major dissolution concentrated in the vadose and upper phreatic zones producing a horizontality to the caverns. Cavern geometry is also controlled by fracture orientation, often resulting in caverns with linear trends.
Carbonate Reservoir Models
UPWARD-SHOAUNG RESERVOIR MODELS
CEMENTATION AND COMPACTION
WATER DEPTH
PERMEABlUTY
POROSITY
• DOLOMHTZAVON AND SULFATE АЛAЛ,ЛAЛЛI
EMPLACEMENT
Й? k Z Z
1'
Il
TJ1
SUBTlDAL-SUPRATIDAL RESERVOIR MODELS
CEMENlATnNANO COMPACTION
-iWi-Mi^WiWi^^
• DOLOMtTJZATK>N AND SULFATE
EMPLACEMENT
П FLOW UMT A / JrtOW UNTT B
I FLOW UNTT A
^^
Gralnetone with Intergranular Pore Space
LEGEND
Gralnstone with Separate-VUg (MoIdic) Pore Space
Dolograinetone wtth Intergranular Pore Space Anhydrite Cemented Dologralnetone
— — Subaerlal Exposure Surface
^^ '
Wackestones and Mudstones
Dolowackeetonee and Dotomudstonee with IntercrystalNne Pore Space lntertldal and Supralldal Sediments
Shaley Limestone
Figure 3. Schematic diagrams of the upward-shoaling cementation and compaction reservoir model and the subtidal-supratidal dolomitization and sulfate emplacement reservoir model.
274
PART 6—GEOLOGICAL METHODS
Cave roof Cave f i l l Lower collapse zone Intact cave floor
Figure 4. Schematic diagram of the karst-collapse reservoir model showing three karst facies. (From Kerans, 1989.)
Three karst facies can be described that relate to touching vug pore geometries. The roof facies is characterized by dissolution-enhanced fractures formed by cavern collapse. The cave facies is characterized by infill sediment. The floor facies is characterized by collapse breccia from roof collapse and small caverns. Karsted reservoirs are generally highly compartmentalized and very complex.
Geological Reef Reservoir Model
The geological reef model is a composite of the u p w a r d shoaling subtidal-supratidal and karst-collapse reservoir models. The difference is that the facies tracts are compressed onto a carbonate shelf of limited aerial extent with high relief
above the seafloor and with steeply sloping sides. The interior shelf or lagoon facies (Figure 2) located landward of the shelf edge normally contains a high percentage of mud. Grainstones, packstones, and boundstones associated with the reef facies are typically found along the shelf edge.
Compaction and cementation typically destroy the p e r m e a b i l i t y of the lagoonal m u d s , leaving the graindominated sediments and boundstones of the reef edge as reservoir rocks. However, selective leaching, dolomitization, and karsting can significantly alter the permeability patterns, as discussed in previous sections. The reservoir flow units can be very complex due to the numerous possible combinations of depositional and diagenetic events.
Reservoir Quality
S. P. Grier
Core Laboratories Carrollton, Texas, U.S.A.
D. M. Marschall
Core Laboratories Houston, Texas, U.S.A.
INTRODUCTION
The quality of a reservoir is defined by its hydrocarbon storage capacity and deliverability. The hydrocarbon storage capacity is characterized by the effective porosity and the size of the reservoir, whereas the deliverability is a function of the permeability. Effective porosity is the volume percentage of interconnected pores in a rock. The remaining space in the rock is occupied by the framework or matrix of the rock and, if present, nonconnected pore space. C o m m o n porosity types in sandstone and carbonate rocks are listed in Table 1 (also see the chapter on "Porosity" in Part 5).
The p e r m e a b i l i t y of a rock is a m e a s u r e of the rock's ability to transmit fluid (see the chapter on "Permeability" in Part 5). Permeability, measured in darcies, is a function of the size, shape, and distribution of the pore channels in the rock, the type and n u m b e r of fluids present, the fluid flow rate, the length and cross-sectional area of the rock, and the pressure differential across the length of flow.
At least within clastic rocks, there is generally a direct relationship between porosity and permeability (see the chapter on "Geological Heterogeneities" in Part 6). The exact relationship varies with formation and rock type; however, increased porosity is typically accompanied by increased permeability.
CONTROLS ON RESERVOIR QUALITY
Environment of Deposition
The initial p o r e n e t w o r k of newly deposited sediments a n d the quality of shallow b u r i e d reservoirs are generally d e t e r m i n e d by the e n v i r o n m e n t of d e p o s i t i o n (see the chapter on "Lithofacies a n d E n v i r o n m e n t a l Analysis of Clastic Depositional Systems" in Part 6). This dictates the grain characteristics, which in turn control porosity and permeability. In clastic rocks, these characteristics include grain size and sorting, sphericity, angularity, packing, and the abundance of matrix materials. The best reservoir quality rocks are well-sorted, have well-rounded grains, and contain no matrix material.
Sedimentary structures affect initial reservoir quality by imparting a preferential flow pattern in the reservoir. Planar bedding, laminations, or other stratification features can c r e a t e s t r a t i f i e d p l a n a r f l o w , e s p e c i a l l y if p e r m e a b i l i t y barriers such as clay partings, finer-grained laminae, or graded beds are present. Slump structures may reduce permeability by creating a tortuous flow path, or may increase permeability (and porosity) by causing a looser grain packing and by producing small faults. Bioturbation typically decreases reservoir quality by mixing adjacent
sands and clays, introducing the clay into the interstices among the sand grains.
Diagenesis
During and following burial, diagenetic events will modify the original pore network of reservoir rocks (see the chapter on "Evaluating Diagenetically Complex Reservoirs" in Part 6). Four main diagenetic mechanisms affect reservoir quality: compaction, cementation, dissolution, and recrystallization These mechanisms are controlled by the detrital composition of the rock, burial depth, burial time, burial temperature, pore fluids, and pore fluid pressure.
Compaction
Compaction reduces the porosity a n d permeability of a rock by causing the following: (1) grain rotation and rearrangement into a tighter packing configuration, (2) plastic deformation of ductile grains that flow into adjacent pores and pore throats, (3) fracturing and crushing of brittle grains, a n d (4) pressure solution in the form of grain suturing and stylolitization (McBride, 1984). Rocks that contain mechanically labile grains, such as clay clasts, altered rock fragments, or delicate fossils, are likely to experience a reduction in porosity and permeability as the ductile grains plastically flow into adjacent pore spaces. Brittle grams will fracture, shatter, or in the case of s o m e fossils a n d p o r o u s grains, collapse. A rock that consists of a f r a m e w o r k of strong minerals, such as quartz, tends to undergo only minor porosity and permeability reduction during compaction due to grain rotation and rearrangement into a tighter packing configuration.
Cementation
Cementation, the filling of original pore space by cements, m a y occur early or late in the diagenetic history of a rock (Scholle and Schluger, 1979; McDonald and Surdham, 1984). Table 2 lists some c o m m o n cement types. Precipitation of authigenic minerals usually reduces reservoir quality; however, early formation of some authigenic minerals can preserve the original porosity by protecting the rock from later degradation by compaction or cementation (Wilson and Pittman, 1977).
Dissolution
D i s s o l u t i o n of less chemically stable m i n e r a l s in sandstones and carbonates can sometimes significantly increase both the rock porosity and the permeability (Schmidt and McDonald, 1980). Dissolution tends to be especially important in carbonates that are buried to shallow depths and sandstones that are deeply buried.
275
276
PART 6—GEOLOGICAL METHODS
Table 1. Porosity Types
Type
Sandstones Primary intergranular Dissolution or vug Micropores
Fracture Carbonates
Interparticle Intraparticle Intercrystal Moldic Fenestral Fracture Vug
Characteristics
Interstitial void space between framework grains Partial or complete dissolution of framework grains or cement Small pores mainly between detrital or authigenic clays; can also occur
within grains (e.g., microporous chert) Breakage due to earth stresses
Pores between particles or grains Pores within individual particles or grains Pores between crystals Pores formed by dissolution of an individual grain or crystal in the rock Primary pores larger than grain-supported interstices Formed by a planar break in the rock Large pores formed by indiscriminate dissolution of cements and grains
Recrystallization
Recrystallization of carbonates and the alteration of grains and cements to clays can have a significant impact on reservoir quality in sandstones and carbonates. Dolomitization of limestones or calcite cement in sandstones typically increases porosity and permeability. Similarly, clay r e p l a c e m e n t m a y increase overall porosity of the rock; however, the pores associated with clay minerals tend to be micropores that contain irreducible water. Also, delicate clay flakes may become mobile with flowing pore fluids and migrate to, and clog, pore throats.
Structural Deformation
Fracturing and brecciation associated with folds, faults, and diapirs generally increase the reservoir quality of wellindurated rocks (see the chapter on "Evaluating Fractured Reservoirs" in Part 6). Fracture porosity is typically low, usually providing only about 1 % porosity; however, fractures in large reservoirs may hold considerable reserves. Fracture p e r m e a b i l i t y m a y be as high as tens of darcies a n d is directional in nature. Conversely, fractures filled by mineralization or with gouge may produce a permeability barrier in the direction perpendicular to the fracture. Brecciation along fracture or fault zones may occur due to shearing or dissolution and collapse. Except where mineralization has occurred in the breccia, brecciation can increase both porosity and permeability considerably. Closely spaced sealing faults can significantly compartmentalize a reservoir.
Wettability
Wettability in an oil reservoir controls reservoir quality by affecting the amount of water production (see the chapter on "Wettability" in Part 5). When the reservoir rock is oil-wet, water is located in the central portion of the pores and will flow through the pore system with the oil. Conversely, in a water-wet reservoir, the water is restricted to the perimeter of the pores and will not flow through the pore system until m u c h of the oil has been r e m o v e d . In a d d i t i o n , the
irreducible water saturations of oil-wet reservoirs tend to be much lower than those of water-wet reservoirs.
Capillary Pressure
The capillary pressure of a reservoir affects the magnitude and distribution of water saturation and thus the hydrocarbon volume in a given reservoir area (Leverett, 1941). The capillary pressure is a function of the capillary radius, the interfacial tension, and the contact angle between the water and the solid (see the chapter on "Capillary Pressure" in Part 5). In a reservoir, zones with larger pores and pore throats have lower capillary pressure, lower irreducible water saturation, and higher hydrocarbon pore volume.
METHODS OF ASSESSING RESERVOIR QUALITY
Numerous methods exist for assessing reservoir quality, ranging in scale from the macroscopic to the microscopic (see the chapter on "Evaluating Diagenetically Complex Reservoirs" in Part 6).
Macroscopic Techniques
Modern three-dimensional seismic data (Brown, 1986) can sometimes assist in predicting reservoir quality away from well control. Careful processing of seismic data allows a conversion of the seismic reflection amplitudes to estimates of acoustic impedance. Because lithology, porosity, and fluid s a t u r a t i o n s affect the acoustic i m p e d a n c e of a rock, a relationship can then be established between the seismic estimates of impedance and the rock properties determined from the logs or in the laboratory. (For information on comparing seismic data to rock properties, see the chapter on "Seismic Inversion" in Part 7.)
Wireline logs can be classified into three different groups based on the information they provide: (1) lithology indicators—gamma ray, sonic, density, and neutron logs, (2) porosity logs—sonic, density, and neutron logs, and (3) fluid saturation logs—resistivity logs (Asquith and Gibson, 1982). (For more on the information that wireline logs can provide,
Table 2. Common Cements of Sandstones and Carbonates
Cement
Quartz Calcite
Dolomite Anhydrite Gypsum Feldspar Siderite Zeolites Kaolinite Illite Chlorite Smectite
Common Crystal Form
Syntaxial overgrowth, prismatic Fibrous, bladed, granular, blocky,
poikilotopic, syntaxial rim Rhombohedral, blocky, granular Blocky, bladed Blocky, bladed, prismatic Syntaxial overgrowth, prismatic Granular, blocky, bladed Platy, bladed, fibrous, prismatic, blocky Platy Fibrous Platy Crenulate
Reservoir Quality
277
see the chapter on "Standard Interpretation" in Part 4.) In addition to lithology, porosity, and fluid saturations,
permeability sometimes can be inferred from log responses or a combination of log responses. The spontaneous potential log is m o s t o f t e n u s e d as a q u a l i t a t i v e i n d i c a t o r of the permeability of a formation. (For more on wireline log response to reservoir properties, see the chapter on "QuickLook Lithology from Logs" in Part 4.)
Another macroscopic technique used to determine reservoir quality is drill stem testing (DST) or formation testing. A drill stem test is generally performed after the well has been conditioned by sealing the zone(s) of interest and allowing the production of fluids (see the chapter on "Drill Stem Testing" in Part 3). The fluids are tested for hydrocarbon content and the pressures and flow rates are measured. The permeability can be inferred from the pressures measured over time, and the productive capability of the f o r m a t i o n is d e t e r m i n e d f r o m the t y p e s of fluid produced and the flow rates.
Mesoscopic Techniques
Core analysis measurements performed on representative core samples can more accurately assess reservoir quality (Keelan, 1972) and heterogeneities. Core analysis porosities are typically d e t e r m i n e d using one of three techniques: s u m m a t i o n of f l u i d s , r e s a t u r a t i o n , a n d Boyle's Law. Permeability on core samples is determined using one of two methods: steady-state or unsteady-state. Air (gas) permeability measurements are typically measured using a steady-state technique. The unsteady-state technique monitors pressure changes, flow rates, and fluid changes as a function of time to determine permeability (Jones, 1982). The unsteady-state method should be used to determine the air permeability for samples of low permeability to obtain the most accurate values. Liquid permeability measurements can be determined by either the steady-state or the unsteady-state method (see the chapter on "Permeability" in Part 5).
Capillary pressure can also be measured in the laboratory on core samples (Wardlaw, 1976). Various techniques are used to determine fluid saturations in the sample at various pressures so that a saturation profile at different pressures is created, which characterizes the irreducible water saturation and hydrocarbon pore volume of the rock.
Figure 1. Binary petrographic image of sandstone. Darkareas are pores and light areas are grains or cement.
Microscopic Techniques
Microscopic techniques used to assess reservoir quality include thin section analysis, petrographic image analysis, scanning electron microscopy, and X-ray diffraction (see the chapter on "SEM, XRD, CL, and XF Methods" in Part 5). Through thin section analysis, the pore types and distribution, the extent of reservoir e n h a n c e m e n t or d e g r a d a t i o n by diagenesis, and the influence of depositional textures on reservoir quality can be determined (see the chapter on "Thin Section Analysis" in Part 5).
Another microscopic method of assessing reservoir quality is through the use of scanning electron microscopy (SEM) with energy-dispersive X-ray. The SEM allows examination of a reservoir rock at very high magnifications with an excellent depth of field so that the pore network and clay minerals within the pores can be viewed. Energy-dispersive X-ray analysis provides an elemental analysis of the grains, cements, and clays identified by the SEM and is used to aid in determining the mineralogy. Such analysis is extremely important in evaluating the potential for formation damage by introduction of potentially reactive stimulation fluids.
Petrographic image analysis (Gerard et al., in press) is a relatively new technique that provides porosity and permeability values and capillary pressure curves for sandstone samples that are not suitable for conventional core analysis, such as cuttings, percussion sidewall cores, and unconsolidated core samples. Image analysis measures key t w o - d i m e n s i o n a l geometrical characteristics of the pore network in thin section using a research-grade petrographic microscope coupled with an image analysis system. The system generates a binary image representing porosity and rock material from thin section views of undamaged portions of the s a m p l e (Figure 1). From this image, p o r e area, diameter, perimeter, length, width, and aspect ratio can be analyzed and related to the three-dimensional porosity, permeability, and capillary pressure values that have been measured on conventional core samples.
Geological TJTHLeteroeene• it•ies
"
Roger M. Slatt1
Department of Geology and Geological Engineering Colorado School ofMines Golden, Colorado, U.S.A.
William E. Galloway
Department of Geological Sciences The University of Texas at Austin Austin, Texas, U.S.A.
INTRODUCTION
The term reservoir heterogeneity is used here to describe the geological complexity of a reservoir and the relationship of that complexity to the flow of fluids through it (see Alpay, 1972, for a discussion of definitions).
Reservoirs are inherently heterogeneous assemblages of depositional facies and subfacies, each with characteristic and commonly differing sediment textures, stratification types, and bedding architectures. Variability is compounded by postdepositional alterations of the strata, such as t h r o u g h compaction, cementation, and tectonic deformation. Geological heterogeneities have been classified in a variety of ways according to their size or scale; the common categories, u s e d h e r e , a r e wellbore, interwell , a n d fieldwide scales of heterogeneity (Figure 1).
Heterogeneities at the wellbore scale affect matrix permeability, distribution of residual oil, directional flow of fluids, potential fluid-rock interactions, and formation damage. Heterogeneities at the interwell scale affect fluid flow patterns, drainage efficiency of the reservoir, and vertical a n d lateral s w e e p efficiency of s e c o n d a r y and tertiary recovery projects. Heterogeneities at the fieldwide scale determine the in-place hydrocarbon volume, areal distribution, and trend of hydrocarbon production.
WELLBORE SCALE HETEROGENEITIES
Elements of wellbore heterogeneities include the p o r e network (pores and pore throats), grain size and composition, grain packing, lamination and bedding styles, sedimentary structures, lithofacies, and vertical stratification sequences. These properties can be readily described in a numerical or quantitative fashion because of the usual availability of rock samples and well logs. Rock cores provide the best information on lithofacies and stratification sequences, plug or whole core porosity, permeability, and fluid saturation (if oil-based drilling m u d w a s used d u r i n g coring). The use of log shapes for facies recognition, as well as sidewall samples, micrologs, and dipmeter tools can also provide indirect information on lithofacies and stratification types. (For more on lithofacies, see the chapter on "Lithofacies and Environmental Analysis of Clastic Depositional Systems" in Part 6.) Pore networks, grain size characteristics, and mineralogy can be analyzed by routine thin section petrography as well as by X-ray diffraction, scanning electron microscopy, capillary pressure measurements, and petrographic image analysis (see the chapter on "Reservoir Quality" in Part 6). Analysis of all or most of these properties is essential for adequate reservoir description because these properties provide the database and thus the foundation for reservoir description at larger scales. (For information on these types of analyses, see Parts 4 and 5.)
In clastic rocks, there is usually a direct relationship between primary depositional lithofacies and reservoir properties and performance. For example, sandstones that become progressively thinner bedded and finer grained stratigraphically upward also become progressively less permeable u p w a r d (Figure 2) so that during waterflood, both gravity and liigher permeability toward the bottom will pull water down. In contrast, sandstones that become progressively thicker bedded and coarser grained upward also become more permeable u p w a r d (Figure 2) so that during waterflood, gravity still pulls the water down, but permeability pulls the water up, resulting in better vertical sweep (Lassiter et. al., 1986; van de Graaff and Ealey, 1989).
Figure 1. Levels of reservoir heterogeneity. (Modified from Weber, 1986.)
1Formerly of ARCO International Oil and Gas Company, Piano, Texas, U.S.A.
INTERWELL SCALE HETEROGENEITIES
Elements of interwell scale heterogeneity include lateral bedding geometries, styles, and continuity; systematic lateral and vertical textural patterns; and resultant variations in reservoir quality. This scale of heterogeneity is probably the
278
Geological Heterogeneities
279
FT
FN
40-,
PERMEABILITY
CSE LOW
HIGH
PERMEABILITY
FT
FN CSE LOW
40-,
HIGH
30-
30-
20-
20-
10-
10-
O-1
(b)
Figure 2. Typical vertical stratification and permeability profiles of (a) fining- or thinning-upward and (b) coarsening- or thickening-upward sequences. Fining and coarsening refer to average relative grain size of individual laminae and beds, and thinning and thickening refer to the relative thickness of individual laminae and beds.
most difficult to quantify because wellbore data of the type previously described must be extrapolated to the interwell region. In many instances, between well correlations are difficult because lithofacies may not be continuous at interwell spacings. Thus, interpretation must be guided by an u n d e r s t a n d i n g of depositional e n v i r o n m e n t s and facies, interpreted from core analysis and compared with modern environments or outcrop analogs where actual observations and measurements have been made.
Relatively few reliable quantitative studies of depositional environments and facies have been published, and those that have suggest considerable variability in interwell scale heterogeneities among different depositional systems, as well as within any one system. Excellent examples include Scheihing and Gaynor (1988) and Krause et al. (1987) for shelf sandstones; van de Graaff and Ealey (1989) for fluviodeltaic sequences (Figure 3); and Jordan and Pryor (1992) for fluvial sands.
In the absence of sufficient quantitative information on d i f f e r e n t d e p o s i t i o n a l s y s t e m s , statistical m e t h o d s of predicting interwell variability have proven useful, (see Part 6). For example, the literature contains m a n y examples of statistics applied to analysis of the lateral continuity and
100 ft -ь-зо m
300 m
S 1000 ft
PERMEABILITY- DARCYS
m >1
HH > 0.1 - < 1
Ш 1 > o.oi - < 0.1
Figure 3. Lateral and vertical bedding and permeability heterogeneity of a typical fluviodeltaic sequence. (From van de Graaff and Ealey, 1989.)
280
PART 6—GEOLOGICAL METHODS
O
200
400
600
800
1000
Figure 4. Synthetic reservoir cross section showing the vertical and lateral distribution of shales (black) within a sandstone (white) sequence. (Modified from Haldorsen and Lake, 1984.)
spatial distribution of shales, since fluid flow in a reservoir is particularly sensitive to shale distribution. One commonly cited example (Weber, 1982) predicts the anticipated lengths of shales as a function of depositional environments, so that if the depositional e n v i r o n m e n t of a reservoir sequence is known, measurements of shale thicknesses in wells can be used to generate (using a random number generator) a synthetic reservoir cross section (Figure 4). This approach is particularly fruitful w h e n the lateral dimensions of the shale are thought to be less than well spacings (uncorrectable or stochastic shales, as o p p o s e d to c o r r e c t a b l e or deterministic shales whose lateral dimensions are greater than well spacings) (Haldorsen and Lake, 1984). Statistical methods have also been used to evaluate lateral variations in reservoir properties of sandstones. For example, Stalkup (1986) f o u n d considerable lateral variability in outcrop m e a s u r e m e n t s of permeability of shallow m a r i n e a n d fluvial sandstones a n d suggested that permeability distribution should also be described stochastically rather than deterministically.
Standard seismic reflection methods generally cannot resolve reservoir heterogeneities at the interwell scale. Crosshole seismic tomography (Stewart, 1987) offers promise for high resolution reservoir description at the interwell and fieldwide scales, as well as for monitoring enhanced oil recovery projects.
FIELDWIDE SCALE HETEROGENEITIES
E l e m e n t s of f i e l d w i d e v a r i a b i l i t y i n c l u d e r e s e r v o i r thickness, facies geometries and continuity, and bulk reservoir properties. Like interwell heterogeneity, heterogeneities at this scale are difficult to assess because information derived at smaller scales must be scaled up and generalized. Depositional models, determined by geological description at the smaller scales, provide the main basis for interpreting fieldwide reservoir architecture. (For more on depositional models, see the chapters on "Lithofacies and Depositional Analysis of Clastic Depositional Systems" and "Carbonate Reservoir Models: Facies, Diagenesis, and Flow Characterization" in Part 6.) It is very important to describe the reservoir at this scale adequately because reservoirs, being complex depositional systems, are often compartmentalized (Figure 5), and separate compartments may not be in communication (see the chapter on "Evaluating Stratigraphically Complex Fields" in Part 6).
Compartmentalization gives rise to regional patterns of variability in reservoir characteristics and production performance that directly reflect variability in facies distributions or, more precisely, geological flow units (see the chapter on "Flow Units for Reservoir Characterization" in Part 6). Unfortunately, it is not always possible to uniquely define the depositional environment and facies distribution because the reservoir, and the geographic distribution of the database from which to formulate an interpretation, may be smaller than the entire depositional system from which the reservoir originates (see Tillman and Jordan, 1987, and Slatt et al., 1993, for examples).
At this scale, analysis of conventional two-dimensional seismic data, inverted two-dimensional seismic (Seislog), and vertical seismic profiling may be useful in delineating gross architectural elements. With recent improvements in acquisition, processing, and interpretation procedures, threedimensional seismic is more widely used for reservoir delineation. Brown (1986) provides several examples of the d e t e r m i n a t i o n a n d m a p p i n g of structure, fluid contacts, porosity, bed thickness, lateral bed continuity, and various other attributes of reservoirs using three-dimensional seismic methods. (For more on using seismic data to help delineate reservoir heterogeneity, see Part 7.)
Geological Heterogeneities
281
Figure 5. Architectural elements of a barrier island sand body. (From Galloway and Cheng, 1985.)
Flow Units for Reservoir Characterization
W. J. Ebanks J r . M. H. Scheihing C. D. Atkinson
ARCO Oil and Gas Company Piano, Texas, U.S.A.
INTRODUCTION
Petroleum geologists and hydrologists have long recognized the need to define quasi-geological/petrophysical units to formalize their descriptions of rock strata as storage containers and conduits for flow of fluids. Maxey (1964) even proposed the introduction of the term hydrostratigraphic unit into the Code of Stratigraphic Nomenclature to fulfill this need. Other terms that have been introduced include reservoir facies (Langston and Chin, 1968), reservoir unit (Pettijohn et al., 1973), flow unit ( H e a r n et al., 1984; Ebanks, 1987), a n d lithohydraulic unit (Krause et al., 1987).
This c h a p t e r outlines one a p p r o a c h to z o n a t i o n of a reservoir for modeling and prediction of performance—the flow unit concept. The subdivision of a reservoir into flow units provides a practical means for reservoir zonation that m a k e s use of both geological a n d p e t r o p h y s i c a l data representing heterogeneity observed at several scales (see the chapter on "Geological Heterogeneities" in Part 6).
DEFINITION AND CHARACTERISTICS OF FLOW UNITS
A flow unit is defined as a mappable portion of the total reservoir within which geological and petrophysical properties that affect the flow of fluids are consistent and predictably different from the properties of other reservoir rock volumes (modified from Ebanks, 1987). How units have the following characteristics in common:
1. A flow unit is a specific volume of a reservoir, which is composed of one or more reservoir quality lithologies and any nonreservoir quality rock types within that same volume, as well as the fluids they contain.
Flow units are internally consistent, but not necessarily homogeneous, in terms of either geological or petrophysical properties. They may contain more than one reservoir quality lithology and they may include nonreservoir features such as shales and cemented layers. Petrophysical properties may correspond to those of lithofacies defined geologically. However, petrophysical similarities among lithofacies may indicate that those lithofacies should be grouped into a single flow unit if they are contiguous. Alternatively, petrophysical dissimilarities within a geologically defined lithofacies may dictate that subdivision of a single lithofacies into several flow units is warranted. A flow unit zonation differs in principle from a lithofacies zonation in that it integrates geological, petrophysical, and production data with the end purpose of describing fluid flow pathways in the reservoir, not the distribution of depositionally distinctive lithologies.
2. A flow unit is correlative and mappable at the interwell scale.
Because flow units are deterministic elements in a reservoir description, they must be of a scale that is correlative and mappable relative to well spacing. Flow units should be continuous at interwell scales, but they need not extend across the entire reservoir. Because some parts of a reservoir can be efficiently drained only on closer well spacings, the definition of s o m e flow u n i t s m a y c h a n g e with infill drilling and changes in p r o d u c t i o n mechanism (such as initiation of waterflooding) during field life.
3. Aflow unit zonation is recognizable on wireline logs.
Mappability in the subsurface requires that flow units be recognizable on wireline logs. Row units recognizable only in core are useful only if all wells have been cored. Some method must be found to translate a flow unit zonation based on core to a zonation based on the wireline log suite available in a particular reservoir.
4. Aflow unit may be in communication with other flow units.
Flow units may be in communication with one another across their boundaries, both in terms of pressure and in the ability of fluids to move vertically and laterally, or they may be completely isolated from one another by permeability barriers.
The fundamental requirement of flow unit delineation is that reservoir volumes in which properties that affect fluid flow differ are consistently distinguished. Furthermore, the extent of these volumes should be definable in the subsurface relying predominantly on wireline logs at existing well spacings.
METHODS OF DEFINING FLOW UNITS
There is no universally appUcable set of rules by which to define flow units. Dividing a reservoir into flow units requires an integration of stratigraphic, sedimentological, structural, petrographic, petrophysical, and field performance data. The process is summarized as follows (Figure 1):
1. Identify the major lithofacies, vertical sequences, and depositional environments from available core. Relate lithofacies, at the whole-core scale, to their mineralogical, textural, and pore level properties and to permeability, porosity, fluid saturations, and capillarity
282
Flow Unitsfor Reservoir Characterization
283
CORE LITHOFACIES
PORE PETROPHYSICAL
TYPES
DATA
GAMMA RAY FLOW
LOG
UNITS
Pc
к! j\s
Pc
Pe
B•
Pe
Figure 1. Major types of geological and petrophysical data applied to flow unit zonation of a well. Here the flow units are delineated on the basis of permeability contrasts due to Iithofacies changes and to the presence of a laterally continuous barrier to vertical permeability (flow unit 4). According to the definition of flow unit, it is permissible to define a flow unit that exhibits only weak flow or no flow through it. This property of flow units makes it possible to use a single numbering system for identifying both obvious flow units and probable permeability barriers that can be mapped at the same scale as reservoir quality flow units.
as measured on core plugs. Establish consistent relationships between rock properties and petrophysical properties. 2. Determine what lithofacies, or associations of lithofacies, are probable flow units based on petrophysical properties, changes in texture, cementation, fracture density, differences in sedimentary structures or bedding styles, and/or separations by prominent shales or other features that may bear on fluid distribution and flow. 3. Calibrate wireline log response to major rock types in as much detail as possible and with appropriate depth shifting of core to logs, in order to detect changes quantitatively in flow unit quality and to correlate major flow units to uncored wells. If cores are not available, cuttings, sidewall cores, patterns of textural change inferred from log signatures, cementation or shales detected on logs, downhole images of the borehole wall, microscanner logs, or other such information must be used in place of core (see Part 4 on Wireline Methods). 4. Establish the three-dimensional distribution of flow units by correlation of calibrated wireline logs. Knowledge of environments of deposition of the reservoir sequence is important to interpreting the style of correlation to be used and the expected patterns of external and internal geometry of any flow unit (see the chapter on "Lithofacies and Environmental Analysis of Clastic Depositional Systems" in Part 6). During correlation, the flow unit zonation established in individual cored wells may change somewhat. Tying
correlation horizons around a loop is critical because individual correlation sections alone can be deceptive. 5. Test the validity of flow units established by consideration of production logs (see Part 9), flow tests of small intervals, oil and water geochemistry (see Part 5), repeat formation tester (RFT) surveys (see Part 4), injectivity logs, tracer surveys, and any available data on patterns of production through time. Modify the flow unit definitions as needed to accommodate the physical measurements of flow, if a rationale can be found for the differences.
The distribution of petrophysical properties such as porosity and permeability can be mapped within flow units using well control only or by applying geostatistical p r o c e d u r e s to create stochastic realizations of these distributions "conditioned" on the well data (see Part 6). Geostatistical techniques that have a strong stochastic component are consistent with, and complementary to, the flow unit concept, which is itself mostly deterministic.
EXAMPLES OF APPLICATION OF FLOW UNITS
Examples of field studies that apply the flow unit concept are listed in Table 1. A larger n u m b e r of s t u d i e s h a v e identified and mapped lithofacies control on geometry and petrophysical properties of reservoirs but have not applied a flow unit classification to these subdivisions. Some examples are listed in Table 2 and illustrated in Figure 2. These studies are good examples of the stratigraphic and sedimentological component of the process of flow unit subdivision.
284
PART 6—GEOLOGICAL METHODS
Table 1. Examples of Field Studies Using Flow Units
Field Siliciclastic Reservoirs
Elk City, Oklahoma
Hartzog Draw, Wyoming
Pembina, Alberta Balmoral, North Sea
Carbonate Reservoirs Fields along the Cabin Creek anticline, Williston basin Rainbow Lake, Alberta Jordan, Texas
Application
Mapping of "flow unit-like" subdivisions of clastic rocks on the basis of lithological, well test, and production data
Flow units applied to reservoir characterization of shelf sandstones for an EOR project
Reservoir characterization of shelf sandstones by "lithohydraulic units"
Flow unit model of turbidite sands integrating geological and petrophysical properties for reservoir simulation and management
Combined lithological and petrophysical parameters of Ordovician-Silurian carbonates to recognize different reservoir types
Described "reservoir facies" within reef complexes
Defined "flow units" of San Andres dolomites based on a number of geological and petrophysical characteristics
References Sneider et al., 1977 Hern et al., 1984 Krause et al., 1987 Slatt and Hopkins, 1990
Roehl, 1967
Langston and Chin, 1968 Major and Holtz1 1989
Table 2. Examples of Lithofacies Control of Geometry and Petrophysical Properties of Reservoirs
Field/Area/Formation Gulf Coast
Frio Formation, Texas El Dorado field, Kansas
Hartzog Draw field, Wyoming
Spraberry Trend, Midland basin, Texas
Application
Continuity and internal facies and petrophysical properties of sandstones
Facies architecture of strand plain reservoirs
Effects of facies on productivity of deltaic sandstone
Effects of facies on productivity of shelf sandstone
Facies architecture and petrophysical properties of submarine fan reservoirs
References Morton et al., 1983
Tyler and Ambrose, 1985 Tillman and Jordan, 1987
Tillman and Martinsen, 1987
Guevara, 1988; Tyler and Gholston, 1988
FEET METERS
80-
-20
40-
-10
Oj-O
CENTRAL BAR (FACIES A)
BAR MARGIN (FACIES B & C)
BIOTURBATED SILTSTONE 3 (FACIES E)
INTERBAR (FACIES D)
FEET 0 Y
!•!•(lit
+
METERS 0 200 400 600 800
FLOW UNIT 1 FLOW UNIT 2
FLOW UNIT 3 FLOW UNIT 4
FLOW UNIT 5 • • • • • • •
iiiiiii
(а)
(b)
LITHOFACIES
LAMINITE
I I LIGHT-COLORED LUTITE
LO-O- AMPHIPORA
g g g SKELETAL RUDFTE STROM ATOPOROID & CORAL RUDITE MASSIVE STROMATOPOROID MASSIVE STROMATOPOROID & CORAL
EZl ARENITE
ШШ MASSIVE CORAL
STROMATOPOROID & CORAL DETRITUS
f75v»3 DENDRITIC CORAL
И
CRINOID
HI
DARK-COLORED LUTITE
(c)
FEET
V
METERS 0
1000
I -T L
500
5000
-Tj
1500
(d)
O
RESERVOIR FACIES
bTl O
S
с
СЛ
Hi O
O
M Si Г-Ь 3r * .
S
Figure 2. Some examples of lithofacies and flow unit subdivisions of clastic and carbonate reservoirs, (a) Lithofacies and (b) flow unit subdivision of the Shannon Sandstone
body in the Hartzog Draw field, Powder River basin, Wyoming. (Modified from Hearn et al., 1984.) (c) Lithofacies and (d) reservoir facies (flow unit) subdivision of the Rainbow
Ю
Lake reef reservoir ("A" Pool), Alberta, Canada. (Modified from Langston and Chin, 1968.)
Oeno
Effective Pay _D_ etermie nateion
Gerard C. Gaynor
^ ^ -- - mReservoir GeousysstAems Inc.
Robert M. Sneider
Robert M. Sneider Exploration Inc. Houston, Texas, U.S.A.
INTRODUCTION
Pay determination is a key component in the calculation of the expected v o l u m e of recoverable hydrocarbons f r o m a field u n d e r a set of k n o w n or predicted economic conditions. It is of i m p o r t a n c e f r o m the field discovery, t h r o u g h initial appraisal and development phases, to final abandonment. The uncertainties associated with pay determination can be circumscribed only by a thorough integration and interpretation of geological and engineering data.
PAY AND NONPAY CONCEPTS
Pay is defined as that part of a reservoir unit from which hydrocarbons can be produced at economic rates given a specific production method. This c o n c e p t of p a y links t h e physical characteristics of the reservoir (rock properties, fluid saturations, and capillary behavior) to the economic aspects of production (completion method, recovery techniques, and volumetric estimates of reserves). Nonpay is defined as the part of a reservoir unit that will not produce hydrocarbons at economic rates and includes intrareservoir barriers.
A reservoir rock is any porous a n d permeable rock capable of potentially containing hydrocarbons in its pore system. This statement implies that not all reservoir rocks qualify as pay. In some reservoirs, there may be intermediate pay types or a continuum between pay and nonpay intervals. Tliis situation may include reservoir units that have differing fluid saturations or pore geometries, or that are present at different elevations above the hydrocarbon-water contact.
The production methodologies—primary, secondary, and
enhanced recovery—affect the definition of pay. For example, beds with limited lateral continuity may qualify as pay under primary production, but may not be waterfloodable at contemplated injector-producer well spacings, thus disqualifying them as pay under secondary production. Thus, there are two separate but related questions regarding pay determination: first, the delineation of reservoir quality rock, and second, the classification of that part of a reservoir quality interval as pay.
PAY DETERMINATION TECHNIQUES
The steps and information necessary for the effective determination of pay in a reservoir are outlined in Table 1. Data sources include the following:
1. Results of sedimentological studies, preferably using conventional cores (see the chapter on "Lithofacies and Environmental Analysis of Clastic Depositional Systems" in Part 6)
2. Core analysis (see Part 5) 3. Analysis of the capillary system of selected core
samples (see the chapter on "Capillary Pressure" in Part 5) 4. Tliin section petrography and electron microscopy for mineral composition and pore geometry (see the chapters on "Reservoir Quality" in Part 6 and "Thin Section Analysis" and "SEM, XRD, CL, and XF Methods" in Part 5) 5. Well logs, including measures of the lithology, porosity, and fluid saturation (see Part 4)
Table 1. Basic Steps in Pay Determination
Step 1. Geologically characterize reservoir
2. Determine reservoir properties
3. Delineate reservoir and nonreservoir rocks and characterize pore space geometry
4. Evaluate pay and nonpay
5. Confirm pay zones
Procedure
Core description, wireline log calibration, lithofacies determination, depositional environment analysis
Core analysis (porosity, permeability, fluid saturation), wireline log analysis (porosity, fluid saturation)
Porosity/permeability crossplots, thin section petrography, pore cast electron microscopy, mercury injection capillary analysis; apply cut-off criteria (Table 2)
Mercury injection capillary analysis, fluid saturation analysis; apply cut-off criteria (Table 3)
Measure well performance using spinner, temperature, flowmeter data and production results; observations noted during drilling, including shows
286
Effective Pay Determination
287
Table 2. Reservoir and Nonreservoir Rock Criteria Based on Mercury (Hg) Injection Capillary Data
Criterion
Initial displacement pressure3 (psi) Capillary pressure (psi)
(1% bulk volume Hg fluid saturation) Bulk volume Hg fluid saturation at 1000 psi Bulk volume Hg fluid saturation at 2000 psi Distribution of effective pore throat
radii at 2000 psi capillary pressure
Reservoir <100
<300 >3% »3%
>50% of radii > 0.05 pm
aThe initial displacement pressure is the lowest pressure required to force a nonwetting fluid into a pore space system. From Sneider (1987).
Nonreservoir >100
>500 <2% <3%
>50% of radii < 0.05 ц т
6. Production history and well test results, including spinner and temperature surveys (see Part 9)
7. Ancillary data, including m u d logs, drillers' logs, and show evaluations (see Part 3)
The simplest, yet most useful, method for combining this information is a composite log, which displays the different classes of data in a format in which each data set is readily correlated by depth. From a detailed reservoir profile log, pay zones can be identified and correlated to uncored wells using well log curves that are calibrated to core data. Examples of this type of p r o c e d u r e can be f o u n d in Connolly a n d Reed (1983), Harris (1975), Hearn et al., (1984), and Hietala and Connolly, (1984).
An important component of effective p a y determination is a systematic, sedimentologically based reservoir zonation. This p r o c e d u r e provides a direct m e t h o d of evaluating the validity a n d r e p r e s e n t a t i v e n e s s of core m e a s u r e m e n t s in relation to the actual distribution of porosity, permeability, and fluid saturations within the reservoir. Core description should be integrated with well logs for calibration and correlation to uncored wells. Discussion of calibration techniques can be found in Connolly and Reed (1983), Hietala and Connolly (1984), and Sneider and King (1984).
Well and production tests are often taken over too large an interval in the wellbore to be precise in distinguishing pay and nonpay, especially in heterogeneous reservoirs. Spinner and temperature surveys can be good indicators of the loci of p r o d u c t i o n w h e r e the b o r e h o l e p e n e t r a t e s the reservoir if production rates are high enough. Electric logs can delineate hydrocarbon saturated intervals, but are not an effective tool for pay determination until they are calibrated with production tests, core analyses, or results from analogous reservoirs. The effective d e t e r m i n a t i o n of p a y relies on a n a l y s e s f r o m the physical s a m p l i n g of r e s e r v o i r a n d nonreservoir rocks. The different classes of information regarding reservoir behavior and pay determination may be irreconcilable or open to misinterpretation in the absence of a thoroughly understood geological framework.
Of all the m e t h o d s available for the p r e d i c t i o n of the b e h a v i o r of the r o c k - f l u i d s y s t e m , capillary analysis is essential in determining pay because the displacement characteristics of hydrocarbons are dependent on pore throat geometries, fluid saturations, and the respective fluid properties of immiscible w e t t i n g a n d n o n w e t t i n g phases.
M e t h o d s of capillary p r e s s u r e analysis (such as m e r c u r y injection) and the interpretation of capillary behavior in reservoir rocks can be found in Wardlaw and Taylor (1976). Mercury injection capillary pressure curves can be readily transformed for predicting fluid behavior during production, locating transition zones, and estimating water cut during production.
The initial delineation of reservoir quality rocks can be obtained by crossplotting such quantities as porosity, permeability, and fluid saturation in which these attributes are identified by lithofacies, depositional environment, or any other valid geologically based description that zones the reservoir into genetically distinct units. Hydrocarbon fluid s a t u r a t i o n w i t h i n t h e r o c k p o r e s p a c e is not a f a c t o r in determining reservoir rock quality. A set of guidelines that identifies reservoir quality and nonreservoir rocks in most cases is shown in Table 2. These criteria have been derived f r o m m o n i t o r i n g the p r o d u c t i o n history of d i f f e r e n t rock types in varied geological settings in h u n d r e d s of wells. A relative ranking system of reservoir a n d nonreservoir rock types can be established using this table in cases where some, but not all, criteria are met.
The location of reservoir quality rocks and their relative rankings should be added to the reservoir profile log. After reservoir quality rock has been identified, the initial determination of pay within the reservoir can be m a d e on the basis of cut-off values shown in Table 3. These cut-off values s h o u l d b e a p p l i e d absolutely only in t h e c o n t e x t of a geologically based characterization of the reservoir and never presumptively.
Pay determined by the foregoing procedure should be confirmed u n d e r a specific production method. The final step should be to correlate well and production tests, injection or production profiles, and any show or other information obtained during drilling to locate producing zones accurately. These zones should be compared with the expected pay as determined in the outlined procedure and the reasons for similarities and differences fully investigated and explained. Pay criteria specific to the field situation can then be refined. Uncertainties (and there always will be these) regarding the reservoir can be circumscribed and addressed using other investigative tools, such as field-scale tracer studies or wettability studies of the separate phases. The procedure allows the differentiation of classes of p a y that m a y relate to p r i m a r y , s e c o n d a r y , or e n h a n c e d recovery p h a s e s of field development or equity allocations.
288
PART 6—GEOLOGICAL METHODS
Table 3. Pay Cutoffs Based on Mercury Injection Capillary Pressure Data
Classification Pay
Intermediate Nonpay
Nonwetting (Mercury) Phase Saturation
Bulk Volume(%)
Pore Volume (%)
>4 ~3—4
>40 >30-40
~2-3
>22-30
>1-2 <1
>10-22 <10
The determination of pay is an estimate and is only as good as the data and its interpretation. A clear and obvious implication is that reliance on a single data source for pay determination, such as electric logs or well test results, is neither appropriate nor advisable. As new information, such as a relatively long-term production history, becomes available, pay delineation should be reevaluated.
Conversion of Well Log Data to Subsurface Stratigraphic and Structural Information
Jeremy M. Boak
U.S. Department of Energy Yucca Mountain Project Office Las Vegas, Nevada, U.S.A.
INTRODUCTION
Conversion of log depths to positions with respect to the surface location of a well is the first step in arriving at a consistent representation of structural or stratigraphic data for a field in a three-dimensional grid. This is particularly important in deviated wells where measured depth and thickness can differ significantly from the vertical depth and stratigraphic thickness.
MEASURED DEPTH
Measured depth (MD), or drilled to depth (DD), is the depth m e a s u r e d (by driller or logger) to a n y f e a t u r e of a well, whether a casing point, a sidewall core, or a significant geological marker. To use this depth in interpreting geological structure, the depth must be corrected to subsea depth or ground elevation (or rig floor or kelly bushing elevation) using information commonly found on log headers. For example, drilled depth minus kelly bushing elevation equals subsea elevation.
TRUE VERTICAL DEPTH
The true vertical depth (TVD) of a point within a well is the depth to that point measured on a line connecting the point to the center of the earth. It m a y be d e r i v e d f r o m the measured d e p t h by correcting for the deviation of the well. Critically i m p o r t a n t to an u n d e r s t a n d i n g of the threedimensional path of a well are the following facts:
1. Wells are not straight; even nominally vertical holes commonly show substantial horizontal displacement even if the deviation is too small to produce a large change in true vertical d e p t h (Figure 1).
2. Wells are not commonly deviated from the surface, but rather are drilled approximately vertical to a kick-off point, where deviation is built up to a planned degree (see the chapter on "Wellbore Trajectory" in Part 3). The rest of the well can be drilled at a constant angle, or the well can be returned toward the vertical to penetrate the horizon of interest (Figure 2).
3. Deviation is rarely constant in a well, even when that is the objective of drilling.
As a consequence, location of the position of any point in a well must be calculated using data from well surveys and an additive formula.
For a simple case (Figure 3), in which the well course is approximated as a series of straight line segments parallel to the individual survey measurements, the formula is as follows:
T V D = Z ( M D , - - M D 1 4 ) X C O S CLj N S D = Z ( M D ( - - M D M ) X s i n a , c o s p,
E W D = X ( M D , - M D ( 4 ) x sin a , sin (3, where
NSD = north-south displacement EWD = east-west displacement
a = inclination angle, in degrees from the vertical, from the survey
(3 = compass bearing, in degrees clockwise from north, from the survey.
i = survey point n u m b e r (7 = O at surface) The intervals can be defined in several ways depending on the accuracy and simplicity of calculation required. The tangential or terminal angle method (Figure 4a) a s s u m e s a constant deviation for the entire interval from one survey point to the next. Thus, the measured depths (MDi-, M D 4 ) for each interval coincide with the depth at the survey points, and the angle used would be for the lower survey point. Although easy to calculate, this method is likely to be substantially in error and is generally not recommended (Craig and Randall, 1976; Inglis, 1987). It is mentioned here
Figure 1. Surface projection of well course for a nominally vertical well. (Data courtesy of Don Clarke, Dept. of Oil Properties, City of Long Beach, CA).
289
290
PART 6—GEOLOGICAL METHODS
Survey
Point)-1
Pi-!
«И
Course
True Vertical Depth
North-South Displacement
East-West Displacement
Point / ot
a = inclination angle P= hearing angle
Figure 3. Segment of a curved well path showing angular and dimensional relationships between the top and bottom of the interval.
Figure 2. Well course of two types of deviated wells. True stratigraphic thickness and true vertical thickness of a dipping stratigraphic unit are shown in relation to the measured interval in a well penetrating the unit.
for historical reasons, as it has been widely used. Altern a t i v e l y , t h e angle averaging method ( F i g u r e 4b) u s e s t h e a v e r a g e for the t w o s u r v e y p o i n t s at either e n d of the segment.
A b e t t e r a p p r o x i m a t i o n , t h e balanced tangential method (Figure 4c), is derived by placing the interval depths (MD,, M D j 4 ) half way between the individual survey points, thus assuming that the deviation is constant in an interval around the measured point.
These f o r m u l a s are d e r i v e d for calculation of well positions from data at specific survey points. Calculation of the location of a stratigraphic top or other d e p t h of interest on the well log m a y require interpolation of the inclination and bearing angles between survey points. The simplest approach is to interpolate linearly between the survey points above and below the point of interest:
am = OCm + [(a,- - a M ) x (MDw - MDl4) /(MDi-MDm)]
P,„-PM + KP,- P/-i>X (MDm - M D m ) /(MD1-MDm)]
where m = marker d e p t h of interest i = next survey point below the marker depth
More sophisticated approaches to well-depth correction are the r a d i u s of c u r v a t u r e m e t h o d (Figure 5a) a n d the m i n i m u m c u r v a t u r e m e t h o d ( F i g u r e 5b). T h e radius of curvature method approximates the well path as a circular arc in the vertical plane, which is then wrapped around a vertical cylinder. The equations in the method (from Craig and Randall, 1976) are as follows:
T V D = ( 1 8 0 / 7 1 ) Z ( M D ; - M D m ) x ( s i n cc,- - s i n
/(a— aM)
aM)
N S D = (180/л)2 Z ( M D , - - M D m ) X (cos a M - cos a,)
x (sin p , - sin P M ) / [ ( a — a M ) x ( p - p M ) ] E W D = (180/jc)2 X { ( M D , - M D m ) / [ ( a - a M )
x (p— p M ) ] } x (sin a — s i n a M ) x (cos P—cos P M )
H D = (NSDj, - NSDf)2 + ( E W D b - E W D f ) 2 ) 1 / 2
where HD = net horizontal displacement b = bottom of interval of interest t = top of interval of interest
Survey Point
Conversion of Well Log Data
291
Survey Point
Calculated Displacement
Approximate Path
Well Course
Calculated. Vertical Depth
Y Survey
Point I
з»
Figure 4. Linear approximations of a curved well course by the
(a) tangential method (Craig and Randall, 1976; Dailey, 1977), (b)
angle averaging method (Craig and Randall, 1976), and (c)
balanced tangential method (Craig and Randall, 1976).
(also as a percentage of measured depth) for a well segment of constant deviation. For most map and cross section scales, depth corrections for wells deviated less than 3° are almost undetectable. For wells deviated by a larger n u m b e r of degrees, large depth and horizontal displacement deviations could lead to substantial misplacements of a geological m a r k e r and the consequent distortion of structural and stratigraphic relationships.
The minimum curvature method approximates the well path as a single circular arc. The equations in this method (from Craig and Randall, 1976; Inglis, 1987) are as follows:
ф = cos"1 [cos CCm COS OCi + sin otj sin a ) 4 cos(p— P m ) ] TVD = 180/jt X[tan (ф/2) x ( M D I - M D M )
x (cos a M + cos a,-)] / ф NSD = 180/л X[tan (ф/2) x ( M D J - M D M )
x (sin a M cos P m + sin a i cos P,)] / ф EWD = 180/тг E [tan (ф/2) x ( M D r M D J
x (sin a M sin P m + sin a;- sin P() J / ф
These methods are especially useful when the deviation angle is built or decreased rapidly with respect to the survey interval.
MEASURED DEPTH VERSUS TRUE VERTICAL DEPTH
Table 1 shows values for the percentage difference between measured depth and true vertical depth for varying degrees of deviation, and for the horizontal displacement
TRUE STRATIGRAPHIC THICKNESS
True stratigraphic thickness (TST) is the thickness of a stratigraphic unit measured in the direction perpendicular to the b e d d i n g planes of the unit (Figure 2). It is a critical measure for understanding both the structural and stratigraphic development of a field. The true stratigraphic thickness is derived from the true vertical depths by the following equation:
TST = (TVDfc - TVDf) x (cos 60 - [(NSD,, - NSDf)2 + (EWDfc - EWDf)2]1/2 x sin 80
In this equation, 6' indicates the apparent dip of the bed in the direction of the horizontal displacement (Figure 6), which is written as
8' = tan-1 [tan 8 cos(P - e)]
where 8 = true dip P = bearing of horizontal displacement between well penetration of top and bottom of unit, or = tan"1 (EWDfc - EWDf) / (NSDfc - NSDf) e = bearing of dip vector
If the well is straight (no change in deviation) for the
292
PART 6—GEOLOGICAL METHODS
Figure 5. Circular approximations of a curved well course showing angles used for the approximations, (a) Radius of curvature method showing chords of horizontal and vertical circles. This method assumes a constant radius of curvature (constant increase or decrease in deviation between survey points), (b) Minimum curvature method showing chord of single circle and the angle ф, which describes the chord.
length of the interval of interest, this formula reduces to
TST = (MDb - MDf) x cos (a + 6')
where MD = measured depth
It is important to note the sign convention for the two angles a and 8'. The deviation is measured from the vertical and is positive, whereas the dip is measured from the horizontal and is positive if it is in the same direction as the deviation a n d negative if the d i p is opposite to the deviation.
A n assumption m a d e here is that the dip of the top and
Table 1. True Vertical Depth Correction and Horizontal Displacement as a Function of Well Deviation
Deviation (a°)
1 2 3 4 5 10 15 45
Correction Factor (% of MD)
TVD
HD
0.02 0.06 0.14 0.24 0.38 1.52 3.41 70.71
1.75 3.49 5.23 6.98 8.72 17.36 25.88 70.71
bottom surfaces is essentially the same. The more closely the wellbore direction approximates the dip direction, the more sensitive the thickness calculation will be to stratigraphic thickness changes (see Figure 7a). The assumption is also violated if the well traverses a zone of strong curvature in the rock such that dip changes rapidly (Figure 7b). Such changes can be corrected for if sufficient data are available, but are commonly too small to be of significance.
To calculate a TST requires survey information as well as some measure of the d i p of the beds. Dip can be derived from dipmeter logs (with some caution) (see the chapter on "Dipmeters" in Part 4) or from m a p s of geological structure. In some instances, TST maps reveal anomalies in well correlation, resulting in iterative refinement of structural and stratigraphic models. It must be recognized that, where folding occurs, stratigraphic thickness trends on m a p s of true stratigraphic thickness will be distorted by compression.
TRUE VERTICAL THICKNESS
True vertical thickness (TVT) is the thickness of a geological unit in a well measured in the vertical direction (Figure 2). It is a valuable measure for volumetric calculations because it is u n a f f e c t e d b y variations in the d i p of the unit a n d can be derived by subtracting computer-gridded structural horizons. In a deviated well with a nonhorizontal unit, the TVT is difficult to calculate because, as the well steps out horizontally, it no longer cuts the bottom of the unit vertically
Apparent Dip
Conversion of Well Log Data
293
True Dip
True Stratigraphic Thickness at
Point A
True Stratigraphic Thickness at
Point B
Figure 6. Relationship between true dip of a planar surface and apparent dip of that surface in a plane at an angle of s to the dip direction.
below the point where it penetrated the top of the unit (Figure 5). If the dip is in the same direction as the deviation, the unit will appear thicker than it actually is, whereas if the dip is in the opposite direction, the unit will be shortened. The TVT is calculated according to the following formula:
TVT = (TVDft - TVDf) - [(NSDb - NSDf)2 + (EWDft - EWDf)2]1/2 x tan 8'
True Stratigraphic
Thickness
Erroneous Stratigraphic
Thickness
Figure 7. Conditions under which true stratigraphic thickness calculated according to its equation will be in error, (a) Variation in stratigraphic thickness is large in relation to the downdip deviation of the well path, (b) Change in dip produces an error in estimation of the true stratigraphic thickness.
Subsurface Maps
K w- weissenburSer
Conoco, Inc. Ponca City, Oklahoma, U.S.A.
INTRODUCTION
Reservoir properties are mapped to promote optimal field development. Subsurface maps dictate well placement and enable engineers to calculate reserves and monitor trends in reservoir performance. Geologists play a key role in subsurface m a p p i n g by using interpretations of depositional environments and diagenetic events to project reservoir data away from relatively few well control points (see other chapters in Part 6). In this sense, subsurface m a p p i n g is in great contrast to geological m a p p i n g of the earth's surface. Whether using traditional concepts (Landes, 1951) or "high technology" computer contouring hardware/software systems Qones et al., 1986), mapping interwell areas places a premium on interpretation rather than straightforward plotting of precise data. " M a p p i n g " is here limited to projections in plan view.
MAPPING SURFACES
A n u m b e r of s u r f a c e s are typically m a p p e d d u r i n g reservoir development to show closure and other limits to reservoir production. M a p s of top of pay a n d bottom of pay can also be "subtracted" to determine pay thickness.
Structure
Structure m a p s s h o w lines of equal elevation or d e p t h for a selected m a r k e r h o r i z o n (Figure 1) (see the c h a p t e r on "Evaluating Structurally Complex Reservoirs" in Part 6). Mean sea level is a useful reference datum. Commonly contoured horizons are top of zone or top of net pay. Control points are provided by surveyed wells and can be supplemented by seismic interpretations, especially offshore. In liighly developed fields, typically onshore, sufficient well control might exist to allow geostatistical interpolation between control points (see Part 8).
Fault Planes
Faults are special surfaces whose traces will show on structure contour maps (Figures 1 and 2). Faults form bounding surfaces for some reservoirs, and sufficient well control might exist to contour map the fault surface itself. Projections of subsurface data into the plane of the fault are also useful "maps" for reservoir development, but are more appropriately described as cross sections. (For details of c o n s t r u c t i o n of fault p l a n e m a p s , see the c h a p t e r on "Conversion of Well Log Data to Subsurface Stratigraphic and Structural Information" in Part 6.)
Unconformities and Subcrops
Surfaces of unconformity can be especially useful marker horizons for structure contour mapping (Figure 2). In many
fields, u n c o n f o r m i t i e s are the location of sealing shales and/or source rocks above reservoir pay. Subcrop maps, traces of p r o d u c t i v e zones, barriers, or m a r k e r horizons mapped on the unconformity surface are invaluable for planning well placement and for reservoir development.
Pressure
M a p s of reservoir pressure are useful throughout reservoir life (Figure 3). Pressures should be converted to a common depth datum, such as mid-reservoir, prior to contouring. (For information on obtaining pressure data, see the chapters on "Production Testing" and "Pressure Transient Testing" in Part 9, "Wireline Formation Testing" in Part 4, and "Drill Stem Testing" in Part 3.)
MAPPING THICKNESSES
I n t e r p r e t a t i o n s of d e p o s i t i o n a l t r e n d s , p r e - a n d syndepositional structural development, and reservoir storage capacity are based in large part on thickness information. An accurate meaning of thickness is critical in these and other analyses (see the chapter on "Conversion of Well Log Data to Subsurface Stratigraphic and Structural Information" in Part 6).
Isopach
A c o n t o u r m a p of e q u a l v a l u e s of t r u e s t r a t i g r a p h i c thickness is an isopach map (Figure 4). Except for vertical wells in horizontal beds, corrections for wellbore deviation and formation dip are needed to make isopach maps.
Isochore
A contour m a p of equal values of true vertical thickness is an isochore map (Tucker, 1988). Note that in c o m m o n practice, isochore maps are informally referred to as "isopach" maps, a term that properly should be restricted to true stratigraphic thickness.
Isochron
A n isochron map is a contour m a p of equal values of seismic traveltime between selected events (Tucker, 1988). Isochron m a p s are the seismic analog of isochore m a p s and, as such, are intended to derive thickness information from seismic data. Isochroning between events above and below a pay horizon, for example, would estimate pay thickness. Renick and G u n n (1989) present a good case history of using isochron and time-structure maps to generate "isopach" and elevation-structure maps. Their isochron-isopach approach delineated reef trends for further development drilling and used well penetrations through a shallow horizon for depth control on a deeper horizon. Phipps (1989) documents the
294
Subsurface Maps
295
Figure 1. Structure map ot the top of the T5 marker, Frio Formation, Brazoria County, Texas. (After Bebout et al., 1978.)
pros a n d cons of using isochron thins and structural highs as exploration drilling criteria for dolomitized Devonian limestones.
MAPPING TO CALCULATE RESERVES
When few production performance data are available, typically early in the life of a reservoir, reserves can be calculated by a volumetric analysis (see the chapter on "Reserves Calculations" in Part 10). For an oil reservoir, the basic volumetric equation is as follows:
N = 7758 х Л х Н х ф х ( 1 - Sw)/Boi where
N = original oil in place, stock tank barrels (STB) 7758 = conversion factor (acre-feet to barrels)
A = area of reservoir (acres) H = height or thickness of pay zone (feet) ф = porosity (fraction of bulk volume) S w = connate water saturation (fraction of pore volume) Boi = oil formation volume factor (dimensionless,
reservoir bbls/stock tank bbls)
A similar basic equation applies to gas reservoirs.
Net Pay
The product A x H is the reservoir bulk volume, and the product AxHx ф is the reservoir pore volume. The general determination of bulk reservoir v o l u m e involves m a p p i n g reservoir area in plan view and m a p p i n g net pay in terms of true vertical thickness to provide a c o m m o n presentation of dipping beds or deviated wells. An isochore m a p of net pay
296
PART 6—GEOLOGICAL METHODS
211/23a 211/24a
MARCH 1974
5 km
BOTTOM WATER INJECTOR
, PERIPHERAL WATER INJECTOR
PATTERN WATER INJECTOR
CONTOUR INTERVAL 2750 к Pa
I
I PRESSURE
SINK
PATTERN FLOOD AREA
Figure 3. Map of pressure response to pattern flood, Judy Creek field, western Canada, 1974 and 1975. Contour interval is 2750 kPa. (After Jardine and Wilshart, 1987.)
Figure 2. Structure of the base of the Humber unconformity (top of the Brent Group), Dunlin field, U.K. Northern North Sea mapped with 1979 and 1989 vintage data. Contours are marked in ft subsea x 100; contour interval is 100 ft. (From Braithwaite et al., 1989.)
should be contoured using well control points and interpolated or extrapolated using available seismic and well test data and the geologist's interpretation of depositional and diagenetic history.
"Net" pay (see the chapter on "Effective Pay Determination" in Part 6) implies that some formation thickness has been excluded from consideration by either (1) occurring below an oil-water contact (or above a gas-water contact), or (2) having porosity a n d / o r permeability values below a "cutoff" limit for productivity. Not all net pay is necessarily productive at a given well spacing. Discontinuous productive horizons between wells might be described, for example, by the concept of net pay to net connected pay ratio (Poston, 1987).
Porosity
The porosity (ф) in a reservoir zone can be determined from log and/or core data (see the chapter on "Porosity" in Part 5). The data in an individual well within the net pay interval can be averaged arithmetically and posted on a map for contouring. The averages should be weighted by thickness.
Water Saturation
The water saturation (Sw) within the net pay interval is typically estimated from well logs. Water saturations can also be d e r i v e d f r o m capillary p r e s s u r e testing of cores to determine the relationship of water saturation versus height above the oil-water contact (see the chapter on "Capillary Pressure" in Part 5). Like porosity, the water saturation data in an individual well within the net pay interval can be averaged arithmetically and posted on a map for contouring (Figure 5). The averages should be weighted by porosity.
Oil Saturation
In an oil-water system, the water saturation and oil s a t u r a t i o n (S0) s u m to 1. T h e r e f o r e , once S w has been determined, oil saturation can be calculated and mapped as S=I-S..,.
MAPPING FOR RESERVOIR MANAGEMENT
A variety of maps are used to predict or monitor reservoir performance.
Permeability
Permeability (k) can also be mapped and contoured (see the chapters on "Permeability" and "Core-Log Transformations and Porosity-Permeability Relationships" in Part 5). As for saturation values, some care must be exercised in mapping permeability because values must be derived from indirect measurements. Typically, permeabilities are derived from wireline log porosities transformed on the basis of core p e r m e a b i l i t y v e r s u s p o r o s i t y cross plots. Permeabilities can be reported at ambient laboratory conditions of pressure or adjusted to reservoir conditions of confining pressure. Similarly, permeabilities can be absolute permeabilities to air (nitrogen) or liquid or effective permeabilities to oil in the presence of irreducible water. Permeability values in an individual well are thickness
StRACtIAfJ 03-л AND D3-D POOlS
SIRACHAN LEDUC REEF CROSS SECTION
Subsurface Maps
297
FAULT TRACES
OIL/WATER CONTACT
Figure 4. (a) Cross section and (b) net pay isopach map of the Strachan gas field, western Canada. Contour interval is 100 ft. (From Hriskevich et al., 1980.)
0
1
1 I
2 KILOMETERS
I
CONTOUR INTERVAL = 0.1 PORE VOLUME
Figure 5. Porosity-weighted average water saturation map for Layer 2 of a Middle Eastern carbonate reservoir.
weighted and typically averaged harmonically, arithmetically, or geometrically, depending on flow geometry. Alternatively, flow capacity (ZcH) values derived from pressure transient testing can be divided by net pay thickness (H) to yield a liquid permeability value for a well.
Porosity Thickness
Reservoir storage capacity or porosity thickness (фH) is the product of porosity and net pay (Figure 6).
Productivity Index
To avoid coning, sand production, pipe collapse, or other harmful effects, wells might not be produced at their maximum wide-open flow rates. Therefore, the ability of a well to produce is usually determined by a productivity index (PI) (Kimmel and Dalati, 1987). The PI is a measure of the stock tank barrels (STB) of oil p r o d u c e d per d a y per psi drawdown under steady-state or pseudosteady-state flow conditions (see the chapter on "Production Testing" in Part 9). Changes will show on periodic maps of PI during reservoir life indicating trends in reservoir depletion or formation damage.
Solution Gas to Oil Ratio
Engineers forecast ultimate recoverable reserves by applying material balance equations or decline curve analysis to production liistory records. For example, in a depletiontype reservoir, the solution gas to oil ratio is sometimes plotted versus cumulative oil production on semilog paper ( G a r b a n d Smith, 1987). If s u c h a c u r v e s h o w s a g o o d straight-line relationship, the curve can be used to predict the trend of a cumulative gas or cumulative oil plot to estimate ultimate recovery.
The s o l u t i o n gas to oil ratio (GOR) is the a m o u n t of dissolved gas that will evolve from the oil as the pressure is reduced to atmospheric from some higher pressure. GOR is usually expressed in units of SCF gas/STB oil. A barrel of oil and its solution gas at reservoir conditions of temperature and pressure will usually "shrink" as the fluid is produced and brought to stock tank conditions (normally reported at 60 T and 14.7 psia). As GOR changes during reservoir life, GORs for individual wells can be mapped periodically to identify areas of the reservoir receiving or not receiving pressure support and serving as indicators for reservoir management action.
298
PART 6—GEOLOGICAL METHODS
B zone absent
Zone B
10,000 feet
Figure 6. Porosity thickness (<j>W) maps for the B and C zones from the San Andres Formation reservoir, Jordan field, Ector and Crane Counties, Texas. Contours in PV fraction-feet. (After Major and Holtz, 1989.)
Water Cut
Water cut is the fraction of a liquid production stream that is water, where oil cut = 1 - water cut. Like GOR, water cut will c h a n g e d u r i n g the life of a reservoir, a n d periodic mapping can serve as a performance indicator for reservoir m a n a g e m e n t . A variety of p e r f o r m a n c e features can be indicated by water cut maps, including water coning, directional permeability or channeling, and formation damage.
Cumulative Production
Cumulative oil or gas production is a parameter useful for ultimate reserves forecasts. Cumulative production can also be mapped periodically as a performance indicator signaling areas of the reservoir that may be responding in a manner seemingly unrelated to initial potential. Figure 7 shows an
example of cumulative production that was concluded to be only poorly correlated to storage capacity (Figure 6) in individual and summed zones of a carbonate reservoir (Major and Holtz, 1989). In this case, porosity did not necessarily indicate effective porosity.
OTHER MAPS
A variety of other m a p s can come into play d u r i n g the development of a specific reservoir. Maps of facies, facies architecture, paleoenvironment, and isolithology might be particularly important in selecting stepout well locations and planning reservoir development strategy. Other reservoir properties such as temperature can have value for specific reservoir engineering applications, particularly where potentially temperature-sensitive chemical stimulation, production, or recovery technology might be involved.
>1000
750-1000 500-750 250-500 <250
Figure 7. Cumulative oil production map for the А, В, C, and D zones from the San Andres Formation reservoir, Jordan field, Ector and Crane counties, Texas. Contours in MSTB/year/acre. (After Major and Holtz, 1989.)
Subsurface Maps
299
Geological Cross Sections
JeremyM Boak
U.S. DepartmentofEnergy Yucca Mountain Project Office Las Vegas, Nevada, U.S.A.
INTRODUCTION
Geological cross sections are graphical representations of vertical slices through the earth used to clarify or interpret geological relationships with or without accompanying maps. As with other tools applied to petroleum development, cross sections are used to portray geological information in a visual form so that reservoir characteristics can be readily interpreted. For example, a t h o r o u g h u n d e r s t a n d i n g of regional structural and stratigraphic relationships may lead to better characterization of reservoir flow units (see the chapter on "Flow Units for Reservoir Characterization" in Part 6).
T h e r e are t w o m a j o r classes of cross sections u s e d in understanding petroleum reservoirs.
1. Structural cross sections, which show the present day geometry of an area
2. Stratigraphic cross sections, which s h o w prior geometric relationships by adjusting the elevation of geological units to some chosen geological horizon (Figure 1).
A third type of cross section called a balanced cross section is a combination of these two. Tliis type attempts to portray the form of geological units prior to some episode of deformation (see the chapter on "Evaluating Structurally Complex Reservoirs" in Part 6). It can provide important conclusions about present day geometry and past stratigraphic relationships.
STRATIGRAPHIC CROSS SECTIONS
S t r a t i g r a p h i c cross sections s h o w characteristics of correctable stratigraphic units, such as reservoir sandstones or sealing shales. They may also be vital in understanding the timing of deformation by showing the d r a p e of sediment over d e v e l o p i n g folds or the thickening of the section across growth faults. The following elements of cross section design a r e p r e s e n t e d a s if t h e y w e r e a s e q u e n c e . In p r a c t i c e , however, each choice affects and is affected by the others.
Choice of Datum
The datum is the level or reference horizon f r o m w h i c h elevations and depths are measured in the cross section. In a stratigraphic cross section, the geologist takes advantage of the principle of original horizontality to p r o d u c e an interpretation of w h a t the chosen slice of the earth might have looked like at some time in the past. By "hanging" all the available vertical information on a stratigraphic horizon or d a t u m that can be correlated along the full length of the cross section, the data are transformed to reflect a different horizontal plane, one that existed at an earlier time (see Figure
la). The assumption that this surface was horizontal when deposited assumes no original depositional slope.
The p u r p o s e of the cross section is to d e t e r m i n e which horizon can serve as the datum. Because it is shown as horizontal, the thickness variations of the units directly above and below the datum are most simply interpretable on the cross section. The cross section in Figure l b uses the horizon labeled F as a datum because this has been interpreted as the top of a chronostratigraphic sequence (Slatt et al., 1993).
An unconformity is commonly used as a datum. In many circumstances, unconformities represent relatively uniform and geologically important time horizons and are therefore useful features on which to hang cross sections. However, caution must be used since the sedimentary layers may reflect paleotopographic relief.
Orientation and Layout of the Cross Section
T h e orientation of a cross section m u s t be chosen to balance the need for a clear representation of the features of interest with the availability of appropriate information. In development geology, this information comes largely from well data (geophysical logs, mudlogs, and cores), but in some places, outcrops and seismic reflection data can be used to constrain interpretations (see Parts 3,4,5, and 7).
Stratigraphic sections should be oriented perpendicular to depositional strike (dip or transverse section) to show facies changes toward or away from the basin margin. Strike sections parallel to the basin margin should be drawn to show lateral variations of particular b e d s or sequences. In the tectonic context of a basin, these axes are also structural axes. Determining the orientation of a stratigraphic section is also complicated by the fact that stratigraphic trends may be at any angle to subsequent structural trends.
W h e n the main source of data is well logs, it is traditional to lay out cross sections to connect wells, which may result in a zigzag path in map view. The cross section is built simply by connecting selected horizons with straight lines and avoids the errors introduced by inexact projection of the data onto a single plane of section. This type of layout results in a distorted v i e w of structural f o r m s if one also constructs a structural cross section of the same wells, as a p p a r e n t dips will vary along such a section, making a smooth structure appear irregular in form. In horizons with rapidly varying thicknesses, this approach can also create apparent irregularities in thickness.
For the p u r p o s e of s t r a t i g r a p h i c c o r r e l a t i o n a n d interpretation, the precise rendering of structural form m a y be of lesser i m p o r t a n c e . For e x a m p l e , F i g u r e 2 s h o w s a stratigraphic cross section in which horizontal scale is entirely schematic because stratigraphic and well log variations across a n u m b e r of fault blocks are the main features of interest and the details of lateral variations are of lesser importance. The
300
Geological Cross Sections
301
Lambert Y Coordinate
400
0
Elevation (ft. from -400
datum)
-800 -1200
J
5000 ft.
S
A
Lambert Y Coordinate
FGHIJK
L M NO
- I 1 11 1
-2000 Subsea Depth (ft)
-4000 -
(b)
-6000
FO Fl f
y
y r
H r/
x r
G 2000 ft. _
No vertical exaggeration
0
G6 G5
III I
5000 ft.
N Af
1
Figure 1. (a) Stratigraphic and (b) structural cross sections of the Ranger Formation in the Long Beach unit of the Wilmington field, California. Sections are projected onto a north-south plane. (From Slatt et al., 1993.)
path of a cross section that has bends in it (Figure 1), whether to accommodate well location or for other reasons, should always be shown on an index map.
The preference for sections that connect well locations may be conditioned by the c o m p u t a t i o n a l b u r d e n of projecting well log data onto a single vertical plane. For stratigraphic cross sections, this approach is generally sufficiently exact even when wells are moderately deviated because the vertical scale is exaggerated and differences from the vertical are minimized (see Figure la versus lb). But the increasing i m p o r t a n c e of d i r e c t i o n a l d r i l l i n g m e a n s that this approximation is no longer sufficient.
In a substantially deviated well, it is important to correct for the d e v i a t i o n of the w e l l b o r e to give a p r o p e r representation of the stratigraphic thickness of units. In m a n y areas, this can be accomplished by using a true stratigraphic thickness (TST) log (see the chapter on "Conversion of Well Log Data to Subsurface Stratigraphic and Structural Information" in Part 6).
Selection of Data
Once the choices of d a t u m a n d orientation h a v e been made, the next step is to decide what data must be displayed.
If the object of the cross section is to show lateral and vertical details of the s t r a t i g r a p h y , log p r o p e r t i e s are of u t m o s t importance.
Typically the SP or gamma ray log and one resistivity log are displayed (Figure 2). Porosity logs may also be important, and if seismic data are part of the cross section, the sonic log is a critical tool to demonstrate the velocity structure, and consistency of conversion of time to depth.
Lines connecting correlative formation or zone tops between wells will s h o w the lateral variation in thickness of these units. If it is important for the display to s h o w exact correlations on logs, these lines should be drawn horizontally across the log d i s p l a y a n d a n g l e d b e t w e e n the e d g e s of adjacent well displays, such as shown in Figure 2. Straight lines connecting the centers of the well displays m a y be more appropriate to provide a better representation of the thickness variations of units between wells. If thickness variations or the geometry of units is p a r a m o u n t in importance, then the logs can be reduced in scale so as to form a background or overlay to the formation data. Alternatively, they can be omitted entirely, and well courses can be represented as line segments, as shown in Figure 1(b).
If lithological data f r o m core a n d / o r cuttings are available, these can be displayed in columnar form between or
302
PART 6—GEOLOGICAL METHODS
A-771 DU F1
JUNIPERO A-863 FAULT
TEMPLE AVENUE FAULT
LONG BEACH BELMONT UNIT FAULT A-1 FAULT
BELMONT
FAULT C-814IA
-4000 FT
3900 FT—*
Figure 2. Schematic stratigraphic cross section along part of the north flank of the Wilmington anticline in the Long Beach unit showing log displays. Distance scale is irregular to make the cross section more compact. The left track of each log is an SP or gamma ray trace and the right track is a resistivity trace. (From Slatt et al., 1993.)
alongside log tracks and hung on appropriate well log horizons. Other data that may form an important part of the cross section include hydrocarbon shows, productive horizons, and geochemical data (such as vitrinite reflectance). The same procedures can be applied to constructing outcrop cross sections.
Vertical and Horizontal Scale
To show significant details of stratigraphic variation, it is usually necessary to exaggerate the vertical scale with respect to the horizontal scale on a stratigraphic cross section. It is important to realize the effect that this distortion has on reservoir geometry and angular relationships of geological surfaces. The small angular differences between stratigraphic horizons that account for thickness variations are strongly exaggerated in such a section. The apparent dip of a bed in a vertically exaggerated cross section is related to true dip by the following equation (Langstaff and Morrill, 1981):
tan bE=V tan 5
where 8£ = apparent dip in exaggerated section 6 = true dip V = vertical exaggeration, or = Iv/Ih, the ratio of vertical scale (Iv) to horizontal scale (Ih)
As a result of this relationship, low dips are exaggerated and appear larger, whereas higher dips all appear close to vertical. The effect is illustrated in Table 1, where selected
values of true a n d a p p a r e n t d i p are s h o w n for vertical exaggerations of five and ten times. Note that two horizons differing in dip by only 3° appear to differ by 14° and 22°, respectively, for the two values of vertical exaggeration. Any attempt to render structural form on a stratigraphic cross section is schematic but should take into account this effect. It is also important to remember that the image one creates with a stratigraphic cross section is a distortion of reality.
Labeling
While a working stratigraphic cross section may serve well without a great deal of labeling, any section presented to a wider audience should have at least its location, orientation, scales, and well locations properly indicated.
STRUCTURAL CROSS SECTION
A structural cross section is made to show the shape of a geological structure so as to evaluate the relationship of fluid contacts and compartments to that structure. Such features as spill points, rollover on faults, and fault geometry give an indication of the likely limits of field production. The form of a structure also provides information about its history and t h u s possibly the history of reservoir formation and oil migration.
Choice of Datum
For a structure cross section, the datum is sea level, with data plotted above or below that point according to its present position (see Figure lb).
Geological Cross Sections
303
Table 1. True Dip Versus Apparent Dip for Common Vertical Exaggerations (Horizontal Scale/Vertical Scale)
Vertical Exaggeration Five times
True Dip
1° 2 5 10 45 75
Apparent Dip
5° 10 24 41 79 87
Ten times
1°
10°
2
19
5
41
10
60
45
84
75
88
Orientation and Layout of a Cross Section
Linear cross sections are preferably oriented perpendicular to the major structural trends (dip or transverse sections). Bends in the section can be introduced to accommodate variable structural trends or to show different features. In a straight section, m u c h of the data will usually be projected into the p l a n e of section. A c c o m p l i s h i n g this projection requires detailed k n o w l e d g e of the strike direction. If the structural trend is variable so that the cross section is not everywhere perpendicular to strike, data should be projected along strike onto the section. To fully represent the structure, several transverse sections may be linked by a longitudinal or strike section running parallel to the strike. Strike sections may also be important in showing the plunge of a structure, c u l m i n a t i o n s in a fold, or the i m p o r t a n c e of s e c o n d a r y structures (for example, normal faults across a fold axis).
Some structures plunge steeply (>30°), producing distortion of the geometry in a vertical cross section, so that it may be preferable to construct a profile section in which the plane of section is perpendicular to the plunge of the structure rather than being vertical. This section will be important for understanding geological history and of less importance for u n d e r s t a n d i n g the relationship of f l u i d s in the associated reservoirs. However, one type of nonvertical section m a y be crucial to understanding the filling of reservoirs. This is the fault plane section (Allan, 1989), which is constructed from well or seismic data to represent the surface of a fault with the trace of units that intersect the fault on either side.
Selection of Data
For u n d e r s t a n d i n g the geometry of structures (folds and faults), an undistorted view of the shapes of geological units is important. Logs can be reduced in size with only the major units represented (Figure la). Where well control is dense and computers are available, it may be best to construct structural cross sections by using gridded and contoured stratigraphic surfaces and drawing each horizon as one would a topographic profile.
If it is important to demonstrate the control of structure on
fluid contacts, it may be vital to show the primary log data from which these are interpreted (see the chapter on "Fluid Contacts" in Part 6). Other data, such as dips from a dipmeter log, can be schematically represented (see the chapter on "Dipmeters" in Part 4).
Vertical and Horizontal Scale
Structural cross sections should be constructed with no or very little vertical exaggeration so that true dips and geometry of an interval can be depicted. The a p p a r e n t dip equation given earlier indicates that not only are small dip variations increased, but at high dips, the differences are minimized (see Table 1). For example, a relatively shallow thrust fault at 45° and a steep normal fault at 75° would appear to have dips of 79° and 87°, respectively, in a cross section at five times vertical exaggeration and are thus virtually indistinguishable at first glance. In addition, distortion due to vertical exaggeration may introduce apparent thickness variations between limbs and axial regions of folds.
CROSS SECTIONS IN THREE DIMENSIONS
W h e n the full three-dimensional aspect of a field must be shown, a single cross section or even a suite of cross sections m a y not be sufficient. The display of n u m e r o u s wells in a t h r e e - d i m e n s i o n a l a r r a y c a n b e a c c o m p l i s h e d b y a fence diagram, in w h i c h the d a t u m horizon is represented by the plane of the m a p . Well plots are displayed vertically, with the d a t u m at the well location on the m a p plane (Figure 3). The wells are the "fence posts," and the lines connecting formation tops are the "rails" that give this diagram its name. Geological r e l a t i o n s h i p s can also b e p o r t r a y e d on a block diagram, in which the sides and top of a schematic block cut into the earth at the location of interest are s h o w n in a threedimensional representation (Figure 4).
COMPUTER GENERATION OF CROSS SECTIONS
Cross sections are now routinely and rapidly constructed by computers. Computers process input data to construct cross sections in two ways, each of which has a pitfall. The m o r e sophisticated m e t h o d involves the use of p r o g r a m algorithms to interpolate and extrapolate from the limited available data to a complete cross section, such as the connecting of stratigraphic tops w i t h straight lines. This process can go w r o n g w h e n a well does not have a full set of tops picked such that false correlations are created. Thus, one pitfall to be a w a r e of is the i n d i s c r i m i n a n t application of software to computer generation of cross sections.
The second way computers function to streamline the preparation of cross sections is to allow the geologist to enter and manipulate one's own interpretations. This capability is especially useful when dealing with discontinuities such as faults that are not handled well by mathematical algorithms. H o w e v e r , a pitfall here is the indiscriminant application of p e r s o n a l interpretation or opinion. If o n e ' s interpretation disagrees with that made by the computer, it is important to discover why the two interpretations differ.
304
PART 6—GEOLOGICAL METHODS
Figure 3. An example of a fence diagram
Figure 4. An example of a block diagram.
Fluid Contacts
Alton A. Brown
ARCO Oil and Gas Company Piano, Texas, U.S.A.
INTRODUCTION
Positions of initial fluid contacts are critical for field reserve estimates and for field development. Typically, the position of fluid contacts are first determined within control wells and then extrapolated to other parts of the field.
Definitions of fluid contacts are based on comparison to capillary pressure curves (Figure D (see the chapter on "Capillary Pressure" in Part 5). The free water surface is the highest elevation at which the pressure of the hydrocarbon p h a s e is the s a m e as that of water. The hydrocarbon-water (oil-water or gas-water) contact is the lowest elevation at which m o b i l e h y d r o c a r b o n s o c c u r . T h e transition zone is t h e elevation range in which water is coproduced with h y d r o c a r b o n s . T h e gas-oil contact is the elevation a b o v e which gas is the produced hydrocarbon phase.
METHODS FOR DETERMINING FLUID CONTACTS
Methods for determining initial fluid contacts are listed in Table 1 and are discussed by Bradley (1987). These include fluid sampling methods, saturation estimation f r o m wireline logs, estimation from conventional and sidewall cores, and p r e s s u r e m e t h o d s . Once initial fluid contact elevations in control wells are determined, the contacts in other parts of the reservoir can be estimated. Initial fluid contacts within most reservoirs having a high degree of continuity are almost horizontal, so the reservoir fluid contact elevations are those of the control wells.
Some reservoirs have irregular or tilted fluid contacts (Figure 2). Reasons for differences in contact elevation in different control wells must first be determined to properly
PRODUCED WATER FRACTION
0
0.5
1.0
Figure 1. Contact definitions and relationship of contacts in a pool (right) to reservoir capillary pressure and fluid production curves (left). The free water surface is the highest elevation with the same oil and water pressure (zero capillary pressure). The oil-water contact corresponds to the displacement pressure (DP) on the capillary pressure curve. The transition zone is the interval with coproduction of water and hydrocarbons. The fraction of co-produced water is shown by the dashed line on the left. The gas-oil contact is controlled by the volume of gas in the trap, not the capillary properties.
305
306
PART 6—GEOLOGICAL METHODS
Table 1. Method for Determining Fluid Contacts Within a Well
Method
Fluid sampling: production tests drill stem tests RFT tests
Description
Directly determines fluid contacts by measuring recovered fluids
Advantages
Direct measure of fluid contact
Saturation determination: well logs
Saturation determination: core analyses
Pressure profiles: RFT tests
Estimates fluid contacts from changes in fluid saturations or mobility with depth
Estimates fluid contacts from changes in fluid saturation with depth
Estimates free-water surface from inflections in pressure versus depth curve
Low cost Accurate in simple
lithologies Rapid High resolution
Saturation estimates for complex lithologies
Saturation can be related to petrophysical properties
Little affected by lithology or coning
Pressure profiles: reservoir tests production tests drill stem tests
Estimates free water surface from pressures and fluid density measurements
Makes use of widely available pressure data
Limitations
Rarely closely spaced, so contacts must be interpolated
Problems with filtrate recovery on DST and RFT
Coning, degassing, etc. may lead to anomalous recoveries
Saturation must be calibrated to production
Unreliable in complex lithologies or low resistivity sands
Saturation measurements may not be accurate
Usually not continuously cored, so saturation profile is not as complete
Imprecise; data usually require correction
Only useful for thick HC columns
Most reliable for gas contacts Requires many pressure
measurements for profile Requires accurate pressures
Imprecise; data usually require significant correction
Only useful for thick HC columns
Most reliable for gas contacts Requires pressure tests from
both fluid zones and assumed or measured fluid densities to estimate contact Requires accurate pressures
extrapolate nonhorizontal fluid contacts to untested parts of a reservoir. Excluding interpretation or mechanical problems, the most common reasons for tilted fluid contacts are the following:
1. H y d r o d y n a m i c gradients 2. Reservoir heterogeneity (see the chapter on "Geological
Heterogeneities" in Part 6) 3. Semipermeable barriers
Situations can usually be distinguished because they are associated with different geological settings and result in different fluid contact characteristics (Table 2).
Hydrodynamic Gradients
A c o m m o n t y p e of n o n h o r i z o n t a l o i l - w a t e r contact is tilting in response to hydrodynamics, the m o v e m e n t of water
in the reservoir interval. Hydrodynamic conditions that affect fluid contacts are usually associated with active meteoric aquifers at relatively shallow depths. Indications of active meteoric flow include low salinity water, high topographic relief, and proximity to recharge areas.
If purely h y d r o d y n a m i c in origin, the fluid contact tilt can be extrapolated across the field as a flat plane that intersects the contact elevation in a m i n i m u m of three control wells. Regional fluid pressure data can be used to extrapolate the fluid contacts f r o m the contacts measured in one or two wells. Only corrected shut-in pressures unaffected by nearby production should be used for this evaluation.
H y d r o d y n a m i c potential (h) is usually m e a s u r e d as the elevation to which water would rise in an open borehole, called the potentiometric elevation. It is calculated f r o m the reservoir pressure by the following relationship:
h = P/ (pw x C) + (Em - Er)
(i)
Fluid Contacts
307
where ph = density of the hydrocarbon
Because the density difference between gas and water is greater than that between oil and water, the dips of gas-oil and gas-water contacts are the same and always less than t h o s e of o i l - w a t e r contacts r e s p o n d i n g to the s a m e hydrodynamic gradient. The dip direction is the same for all fluid contacts.
Example: D e t e r m i n e the position a n d d i p s of the fluid contacts from data in Figure 3. Well B is a discovery well with reservoir gas, oil, a n d w a t e r densities of 0.15, 0.8, a n d 1.0 g / c m 3 , respectively. Gas and oil contacts are at elevations of -5020 and -5080 ft, respectively.
Solution: First d e t e r m i n e (in English units) the p o t e n tiometric gradients using sea level datum:
Zzд = 2369/0.433 - 5160 = 311 ft hB = 2324/0.433 - 5090 = 277 ft hc = 2400/0.433 - 5300 = 243 ft and
ZzA-/7B/2mi = 17 f t / m i /zB-/jc/2mi = 17ft/mi
The potentiometric gradient is approximately constant along the section at 17 ft/mi. Calculating the contact dip from Equation (2), w e have
oil-water tilt = 1.0/(1.0 - 0.8) x 17 = 85 f t / m i gas-oil tilt = 1.0/(1.0 - 0.15) x 17 = 20 f t / m i
Figure 2. Geometries of fluid contacts, (a) Horizontal contacts indicative of hydrostatic conditions in homogeneous reservoir rock, (b) Tilted, flat contacts resulting from hydrodynamic conditions, (c) Contact elevation is constant for each lithology type, but pool contact is irregular due to reservoir heterogeneity, (d) Irregular contacts due to semipermeable barrier in an otherwise homogeneous reservoir.
where P = corrected shut-in pressure C = pressure gradient constant (0.433 psi/ft or 0.1 k g / c m 2 / m)
p w = specific gravity of water Em = pressure measurement
Er = reference elevation (not subsurface depth)
Potentiometric elevations are mapped and contoured to determine the change in potentiometric elevation per unit distance, called the potentiometric gradient. The h y d r o d y n a m i c tilt of a fluid contact can be estimated f r o m the potentiometric gradient and fluid densities by the following relationship (Hubbert, 1953; Dahlberg, 1982):
Hydrocarbon tilt = p w / ( p w - ph)
x potentiometric gradient
(2)
The fluid contacts are graphically projected a w a y from well B at the calculated dips to determine the contact elevations along the cross section.
Reservoir Heterogeneity
Reservoir rocks may have substantially different pore structures in different parts of a field. These heterogeneities may result in significant variations in hydrocarbon-water contacts, especially in low permeability reservoirs (Figure 4). W h e r e all reservoir facies are very porous, heterogeneity of depositional environments does not significantly affect fluid contact elevation.
Fluid contact elevations in different control wells can be empirically related to lithofacies at the contact. Where critical lithofacies are not penetrated at the fluid contact, the contact elevation of the lithofacies can be predicted f r o m capillary pressure and relative permeability tests (see the chapter on "Relative Permeabilities" in Part 5). The greater the difference in capillary pressure and relative permeability behavior for different lithologies within a reservoir, the greater the potential for fluid contact differences caused by heterogeneity. Because surface tension between oil and gas is usually low in subsurface reservoirs (Katz et al., 1957), the effect of reservoir heterogeneity on oil-gas contacts is usually small.
Fluid contacts can be extrapolated f r o m control wells if d i s t r i b u t i o n of d i f f e r e n t reservoir rock t y p e s a n d their capillary properties can be mapped. In many cases, the
308
PART 6—GEOLOGICAL METHODS
Table 2. Nonhorizontal Fluid Contacts
Jm Hydrodynamic tilt
Cause
Moving water at the reservoir interval
Reservoir heterogeneity
Semipermeable barriers
Differences of HC saturation and HC permeability due to capillary effects on different pore systems
Semipermeable barriers compartmentalize an otherwise homogeneous reservoir
Recognition
Presence of hydrodynamic conditions Fluid contact dip angles and directions are
constant over the field despite facies changes. Oil-water contact has a steeper dip than the gas
contacts, but all contacts dip in the same direction Significant dip to gas contacts are rare
Presence of several reservoir lithologies with different capillary properties
Fluid contacts at facies changes Oil-water and gas-water contacts affected;
gas-oil contact not affected Hydrocarbon and water production may be
intercalated
Fluid contacts horizontal, yet at different elevations in different parts of the pool
All types of fluid contacts affected Barriers are documented in locations which
might cause observed compartmentalization Contact changes do not correspond to
regional hydrodynamics or changes in capillary properties
distribution of rock types within heterogeneous reservoirs is poorly characterized during initial development, so the largest uncertainty in mapping the fluid contact is the uncertainty in the distribution of the lithofacies. The position of the hydrocarbon-water contact may need to be confirmed by well penetration.
-4900 -
2 MILES
2 MILES
WeIIC
-ф-
-5000 -
-5100 -
Semipermeable Fluid Barriers
Semipermeable barriers can divide a reservoir into c o m p a r t m e n t s with d i f f e r e n t fluid contacts even if the capillary properties of the reservoir rock are the same on both sides of the barrier. Semipermeable barriers can include faults, mineralized fractures, or semipermeable beds. The resulting pool has horizontal fluid contacts, but the contacts occur at different elevations on different sides of the barrier (Figure 5). The elevation difference between fluid contacts is related to the displacement pressure of the semipermeable barrier (Watts, 1987). Whereas fault or mineralized fracture compartmentalization is not readily recognized without detailed mapping, semipermeable beds are commonly recognized and the reservoir is separated into different pools corresponding to the different fluid contacts. Once the position of the barrier is mapped and the elevations of the contact on either side are determined from control wells, the fluid contacts can be mapped as horizontal surfaces within each c o m p a r t m e n t of the pool. Limited communication across semipermeable barriers is possible during production from the pool.
Figure 3. Example of calculating hydrodynamic fluid contacts from pressure data. Pressure elevations are shown by arrows. Calculated fluid contacts are shown by thin lines.
ANOMALOUSLY THICK TRANSITION ZONES
The transition zones calculated for homogeneous reservoirs may be relatively thin, particularly in coarsegrained reservoirs. However, thick transition zones are observed in many fields due to reservoir heterogeneity. Interbedded lithologies with different capillary behaviors may result in a thick transition zone in which some lithologies produce hydrocarbons and others produce water. Rocks with
Fluid Contacts
309
Figure 4. Effect of reservoir heterogeneity on fluid contacts, (a) Capillary pressure curves for facies A and B within the reservoir. The dashed line corresponds to the saturation trend of the well in part (b). Sharp changes in saturation correspond to elevations of facies changes, (b) Oil-water contact corresponding to capillary pressure curves. The free water surface (fw) is the same for all facies, but the different displacement pressure results in different oil-water contact elevations (thick arrows). The transition zones will also have different thicknesses due to different relative permeability characteristics not shown here. The vertical line is the well position corresponding to the saturation profile shown in part (a).
Figure 5. Irregular contact caused by semipermeable barriers in a reservoir, (a) Capillary behavior of the reservoir and barriers A, B, and C. (b) Fluid contact elevations result from charging of the reservoir from the left. Each compartment of the reservoir has a different free water surface and oil-water contact. The displacement pressure of bed A causes the contact elevation difference between contacts 1 and 2. The displacement pressure of fault B results in the elevation difference between contacts 1 and 3. The displacement pressure of the mineralized fracture C results in the difference in elevation between contacts 3 and 4. The gas column is not thick enough to invade across the fault.
310
PART 6—GEOLOGICAL METHODS
complex pore networks (such as combined fracture and matrix porosity) may also have a thick transition zone, with different fluid types produced from different pore types. The cause of thick transition zones may be evaluated by combined core examination and capillary pressure tests. Some intervals within transition zones characterized by interbedded lithologies may be brought into low water production by selective perforation.
Upward movement of hydrocarbon-water contacts may
leave a zone of residual hydrocarbon saturation below the present transition zone. The hydrocarbons might be interpreted as part of a transition zone from well log or core analysis, leading to an e r r o n e o u s a p p r o x i m a t i o n of fluid contacts or rock p r o p e r t i e s . The p r e s e n c e of r e s i d u a l hydrocarbon saturation below an oil pool can sometimes be detected by the presence of two inflections in the plot of hydrocarbon saturation against depth, one at the present transition zone and another at the original transition zone.
Evaluating Stratigraphically Complex Fields
Diana Morton-Thompson1
Consultant
Kalamazoo, Michigan, U.S.A.
W. E. Galloway
Department of Geological Sciences The University of Texas at Austin Austin, Texas, U.S.A.
INTRODUCTION
Stratigraphically complex fields are those that exhibit a high d e g r e e of vertical and lateral heterogeneity that is directly controlled by the environment of deposition (Table 1). This heterogeneity occurs at all scales (see the chapter on "Geological Heterogeneities" in Part 6). It can result in a highly variable distribution of rock properties within one reservoir and the division of a single stratigraphic interval into multiple reservoirs that act as independent, selfcontained compartments.
It is important to identify stratigraphic complexities early in field development so that (1) operations can be tailored to o p t i m i z e recovery and (2) data necessary to p r o p e r l y characterize and manage the reservoir throughout its life can be obtained. It is also important to recognize stratigraphic complexities in mature fields because these fields offer the opportunity for significant improvements in production with relatively low risk.
IDENTIFICATION
Many criteria exist for recognizing stratigraphic complexities in new fields versus mature fields.
New Fields
• High degree of lithological variation seen in cuttings and core.
• Small scale depositional packages • Vertical and lateral changes in wireline log character • Questionable correlation of depositional packages
between wells • Presence of n u m e r o u s vertical permeability barriers at
a variety of scales • Complex porosity-permeability and capillary pressure
relationships (a broad scatter in the data indicates different lithofacies and pore types) • Anomalous changes in pressure gradient • Variable oil and water chemical composition
Mature Fields
• Multiple fluid contacts at different structural elevations • Irregular reservoir distribution (such as a dry hole
surrounded by producers) • Anomalous fluid relationships (such as a gas zone
underlying an oil zone or an oil zone d o w n d i p of water) • Irregular and/or highly directional waterflood response • Poor injection profiles indicating irregular and poor waterflood sweep efficiency • Selective changes in produced oil and water chemical composition • Erratic pressures • Highly variable cumulative productions • Anomalous recovery factors • Variable GOR (adjusted to standard operating conditions) • Difficulty in history matching reservoir simulation models with actual production
PROBLEMS
Problems associated with a stratigraphically complex field include difficulties in (1) drilling and completion and (2) reservoir characterization and management.
Drilling and Completion Problems
1. A variety of m u d weights and types are needed to prevent formation damage in different producing zones.
2. Various stimulation treatments must be designed to optimize production for different stratigraphic zones or areas of the field.
3. Multiple completions a n d / o r selective injection equipment are required.
Reservoir Characterization and Management Problems
1. Correlations that subdivide the reservoir into meaningful producing zones must be consistent.
2. Maps must be generated that portray complex threedimensional reservoir geometry.
3. The a m o u n t and type of pay in infill locations must be predicted.
4. Three-dimensional conductivity and "tracking" fluid flow m u s t be understood.
5. Production or injection volumes need to be allocated to different zones or patterns.
1Formerly with Chevron U.S.A. and ARCO Research. 311
312
PART 6—GEOLOGICAL METHODS
Table 1. Tentative, Relative Depositional Heterogeneity Ranking of Reservoir Facies Based Primarily on Texas Petroleum Fields3
Depositional Environment
Heterogeneity Rank
Siliciclastic Systems
Submarine slope/fan/canyon fill, silty to muddy 9.0
Braided streams
8.0
Fine-grained meander belt
8.0
Tidal flat and estuary
8.0
Back barrier apron and tidal delta
8.0
Fluvial or deltaic sandstone on carbonate shelf 8.0
Alluvial fan
7.5
Fluvial-dominated delta
7.5
Coarse-grained meander belt
7.0
Fan delta
7.0
Shelf sandstones
7.0
Submarine slope/fan/canyon fill, sandy
7.0
Lacustrine shore zone and delta
6.0
Barrier shoreface
6.0
Eolian
4.5
Wave-modified delta
4.5
Wave-dominated delta
3.0
Strandplain, sandy
2.5
Barrier core
2.0
Carbonate Systems
Evaporitic flats (sabkhas)
9.0
Restricted platform, shoaling cycles, dolomitized 8.5
Open shelves, extensive diagenesis
8.0
Shelf edge reefs, drapes
8.0
Shelf edge, slope or basinal
8.0
Karstic unconformity zones
8.0
Open shelves, platforms or ramps
7.0
Small atolls, pinnacle reefs
7.0
Oolitic bars and barriers
6.0
Large atolls, pinnacle reefs
5.0
aLarge numbers indicate highly complex heterogeneous reservoirs with the greatest potential for improved recovery of in-place hydrocarbons.
Modified from Finley et al. (1988).
EVALUATION
Recognition and three-dimensional characterization of reservoir heterogeneity in a stratigraphically complex field requires a well-by-well analysis of the following:
1. Lithologicalinformation 2. Ruid types and saturations 3. Production history
Because of the inherent geological variability and the diverse developmental histories of different reservoirs, no one approach to reservoir characterization is universally applicable. Rather, evaluation must be tailored to integrate and maximize the use of all available geological, geophysical, geochemical, and engineering data.
The following steps outline a generalized plan for evaluating a stratigraphically complex field.
1. Understand the basic regional and local geological setting. • Review the literature and note pertinent paleogeography, depositional, and structural or tectonic trends, formation tops, unconformities, and sequence boundaries. • Stratigraphically complex fields need to be evaluated in the context of the total geological picture. It is important to distinguish reservoir heterogeneity due to depositional processes from structural or diagenetic overprinting.
2. Describe lithofacies and interpreted depositional environment(s) (see the chapter on "Lithofacies and Environmental Analysis of Clastic Depositional Systems" in Part 6).
3. Determine the relationship between lithofacies/depositional environment and reservoir distribution and quality. • Calibrate lithofacies, depositional interpretations, and petrophysical analysis with wireline log response (see the chapters on "Reservoir Quality" and "Evaluating Diagenetically Complex Reservoirs" in Part 6). • Define net pay (see the chapter on "Effective Pay Determination" in Part 6). • Define flow units and reservoir types (see the chapter on "Flow Units for Reservoir Characterization" in Part 6). • Construct and correlate high resolution cross section grids that show lithofacies and depositional relationships and establish reservoir zones and boundaries (see the chapter on "Geological Cross Sections" in Part 6).
4. Determine reservoir distribution. • Construct isopach, isolith, lithofacies/depositional, and net pay maps and cross sections to quantify the three-dimensional reservoir geometry (see the chapter on "Subsurface Maps" in Part 6). • It is often easier to map depositional geometry first and then use that as a guide to map net pay and other reservoir parameters. Electrofacies maps and cross sections that show SP or gamma ray shape can be useful in defining general reservoir heterogeneity and elucidating trends (Figure 1).
5. Determine reservoir quality • Construct average porosity, porosity height, average permeability, permeability height, and continuity maps and cross sections to define trends in reservoir parameters that affect fluid flow.
6. Determine hydrocarbon occurrence • Evaluate original fluid contacts and hydrocarbon saturation (see the chapter on "Fluid Contacts" in Part 6). • Map original hydrocarbon pore volumes. • Map present hydrocarbon pore volumes.
7. Integrate reservoir distribution, quality, and hydrocarbon maps with production histories and geochemical data. • Use other information that can help clarify the effect of reservoir heterogeneity on patterns of fluid flow, such as maps of initial production, fluid ratio
(percent water cut and GOR), and flood front migration (using changes in oil and water geochemistry). • Production engineering practices, particularly perforation and stimulation histories, need to be considered so that geological influences on reservoir performance can be distinguished from technologically induced "noise." Proper integration of data in this step is critical to ensure an internally consistent, calibrated reservoir description that can be used in subsequent reservoir modeling (see Part 9).
8. Run reservoir simulation models to predict reservoir performance.
Evaluating Stratigraphically Complex Fields
313
• Determine what parts of the reservoir are poorly drained because of poor sweep and ineffective wellbore communication (see the chapters on "Reservoir Modeling for Simulation Purposes" and "A Practical Guide to Reservoir Simulation" in Part 10).
9. Use reservoir simulation model to manage field. • Perform workovers, realign injection support, infill drill and propose enhanced recovery programs to optimize production.
10. Refine reservoir description and calibrate reservoir simulation model. • Continue this process as new data become available and the objectives of reservoir management change to meet economic conditions.
Figure 1. Example of electrofacies map showing distribution of SP log patterns. (From Galloway and Cheng, 1985.)
Evaluating Diagenetically Complex Reservoirs
Michael D. Wilson
Consultant Golden, Colorado, U.S.A.
INTRODUCTION
Major variations in levels of reservoir quality a n d degrees of lateral a n d vertical continuity within oil a n d gas fields are controlled primarily by depositional factors (see other chapters in Part 6). However, major inhomogeneities may also be produced by diagenetic alterations. These inhomogeneities in rock properties may transect or reverse trends produced by depositional controls and can significantly influence reservoir properties, including initial fluid saturations, residual saturations, waterflood sweep efficiencies, preferred directions of flow, a n d reactions to injected fluids. Extreme permeability stratification or the development of permeability barriers by diagenetic alteration may lead to the need to drill additional infill wells or reposition the locations of such wells, selectively perforate and inject reservoir units, manage zones on an individual basis, and revise decisions regarding suitability for thermal recovery operations.
A diagenetically complex reservoir is a reservoir in which the major inhomogeneities affecting fluid distribution and/or productivity are controlled primarily by diagenetic events. Diagenetic inhomogeneities are zones of reduced or enhanced porosity and/or permeability that are generated by one or a combination of the processes of cementation, compaction, replacement, dissolution, and fracturing. For a reservoir to be considered complex, the diagenetic inhomogeneities must exhibit a complex distribution that is not directly correlated with or controlled by depositional factors.
DIAGENETIC EVENTS THAT CREATE RESERVOIR HETEROGENEITIES
Diagenetic alterations are defined here as all physical and chemical alterations that affect a sediment subsequent to deposition, including tectonically produced fractures and
faults. Diagenetic alterations that have been observed to generate reservoir heterogeneities having a major influence on reservoir rock properties are s h o w n in Table 1. In sandstone reservoirs, carbonate and anhydrite cementation, clay authigenesis, secondary porosity generation, and fracturing are the most commonly reported alterations. In carbonate reservoirs, the diagenetic components most often observed are gypsum/anhydrite cementation, dolomite replacement, secondary porosity generation, and stylolitization. A cross section f r o m Longman (1983) illustrates the complexity of porosity d e v e l o p m e n t in one of the few well-documented examples of a diagenetically complex reservoir (Figure 1).
Zones of cementation, dissolution, a n d replacement m a y exhibit a w i d e variety of distributions. To prepare an accurate reservoir description, the geologist needs to understand the diagenetic processes, and the controls on these processes, which generated the heterogeneities present in the reservoir under investigation.
RESERVOIR DESCRIPTION METHODOLOGY
A s c h e m a t i c d i a g r a m illustrating the basic stages of reservoir description in a diagenetically complex reservoir is s h o w n in Figure 2. Geological activities follow the stages of well data analysis and core analysis, which constitute basic input to the geological phase of the analysis.
Stage 1. Construction of a Regional Geological Framework
To assist in the u n d e r s t a n d i n g of the depositional a n d diagenetic events that have created and modified the reservoir rocks, analysis of the regional geological framework can be very helpful. Elements of the regional geology most useful for this include regional thickness and lithofacies patterns, the plate tectonic history of the area, the local
Table 1. Examples of Diagenetic Alterations that Produce Major Reservoir Heterogeneities
Diagenetic Component
Cementation Siderite Dolomite Anhydrite Authigenic clays
Stylolitization and associated cementation
Dissolution Carbonate Anhydrite
Dolomite replacement Faulting/fracturing
Field
C Unit, Kuparuk Field, Alaska Oregon basin Field, Wyoming Pierce and Black Hollow Fields, Colorado Hankensbuttel-Sud Field, Germany
Dukhan Field, Qatar
Spindle Field, Colorado Howard-Glasscock Field, Texas Crane Field, Montana Anschutz Ranch Field, Wyoming
Reference
Gaynorand Scheihing, 1988 Morgan et al., 1977 Levandowski et al., 1973 Gaida et al., 1987
Dunnington, 1967
Porter and Weimer, 1982 White, 1984 Longman et al., 1983 Lindquist, 1983
314
Bakken
42-3 •
Bakken
12-2 *
Bakken
24-2 *
Bakken
31-11 11X-12
# '
*
Evaluating Diagenetically Complex Reservoirs
315
Log Analysis, Well Test Analysis
4>
Core Analysis
4>
Geologic ., Activities
Figure 1. Cross section in the Crane Field, Richland County, Montana, showing extreme irregularities in the development of porous dolomite zones in the Red River Formation below the Canhydrite. (From Longman et al., 1983.)
structural history, the history of exploration and production in the area, the burial history including major erosional and/or nondepositional events, and in particular, the thermal and pressure history of the reservoir.
Stage 2. Construction of a Depositional Model
The first stage of reservoir description is the preparation of a depositional model that depicts the vertical and lateral variations in depositional environment as reflected in lithologies, textural properties, fossil content, and sedimentary structures (see the chapters on "Lithofacies and Environmental Analysis of Clastic Depositional Systems" and "Carbonate Reservoir Models: Facies, Diagenesis, and Flow Characterization" in Part 6). This is necessary to assess the degree of heterogeneity attributable to depositional factors and the relative importance of depositional controls on major diagenetic alterations.
Stage 3. Construction of a Diagenetic Model
In this stage, the geologist must prepare a model that depicts the major vertical and lateral variations in diagenetic alteration as reflected in the abundances of cementing agents a n d / o r extent of recrystallization, dissolution, replacement, and fracturing.
Step 1. Basic Rock Description
The geologist first collects data relating to the relative a b u n d a n c e of diagenetic c o m p o n e n t s recognizable at the magnification levels of a binocular microscope or handlens. This evaluation may be conducted simultaneously with the rock description step of depositional model construction. The most useful data are derived from slabbed full diameter cores, but recourse to cuttings and sidewall cores is necessary where such cores are not available. In addition to describing the slabbed core, horizontal core plugs used for core analysis measurements should also be described. Recent advances in the methodology of cuttings analysis involve visual aids such as cuttings comparators, low magnification (20x) photographs of cuttings samples, thin section p h o t o m i c r o g r a p h s , a n d scanning electron microscope (SEM) micrographs of rock chips and pore casts. These allow more extensive utilization of these s a m p l e s for rock characterization (Sneider et al., 1983).
Figure 2. Stages in the generation of an integrated geological reservoir model.
Step 2. Quantitative Analysis
Semiquantitative a n d / o r quantitative analyses of the abundance of diagenetic components are generally performed using the standard techniques of thin section petrography, Xray diffraction (XRD), and scanning electron microscopy (see the chapter on "SEM, XRD, CL, and XF Methods" in Part 5). The objective of these analyses is to establish the types, abundances, and distributions of the various f r a m e w o r k , pore, and pore-filling components as well as to determine the temporal sequence of diagenetic alterations (paragenesis). Table 2 lists the p r i m a r y t e c h n i q u e s a n d their areas of application as well as similar information for supplementary and alternative techniques where factors of speed, cost, or available m a n p o w e r or expertise prevent use of standard techniques.
Sample sites for petrographic analysis are best selected on the basis of low magnification rock descriptions generated in step 1 a n d t h r o u g h e x a m i n a t i o n of s e m i l o g p o r o s i t y permeability crossplots (Figure 3) with values keyed to major categories of size, sorting, matrix content, cement content, or pore type, depending on their relative importance in a particular reservoir. Samples should be selected to span a wide range of porosities and permeabilities for each major t y p e of reservoir rock (for example, s a n d s t o n e s that are dolomite cemented, anhydrite cemented, quartz-overgrowth cemented, or argillaceous).
316
PART 6—GEOLOGICAL METHODS
Table 2. Techniques Used in Assessment of Diagenetic Alterations
Technique Primary Techniques
Thin section petrography
X-ray diffraction Scanning electron microscopy
Application
Mineral types and abundances; pore types and abundances; diagenetic sequence; texture (size, sorting)
Mineral types and abundances Pore network characteristics; diagenetic sequence
Supplementary and Alternative Techniques Fourier transform infrared spectroscopy Cathodoluminescence microscopy
Isotope geochemistry Fluorescence microscopy
Image analysis
Fluid inclusions
Chemostratigraphy Fission track thermometry Radiometric dating
Mineral types and abundances Recognition of subtle or hidden diagenetic features;
diagenetic sequence Interpretation of paleosalinity and paleotemperatures Recognition of depositional and diagenetic components
and textures in dolomitized or recrystallized limestones; porosity estimation Pore network characterization; grain size and morphology; mineral types and abundances (in combination with SEM or microprobe analysis) Temperatures and pressures of formation of surrounding crystal; composition of fluids involved in diagenesis Correlation of apparently homogeneous sequences Thermal history analysis Age dating of diagenetic events
Use of p l u g e n d s f r o m h o m o g e n e o u s horizontal core analysis plugs for thin section, XRD, or SEM sample p r e p a r a t i o n allows for the d e v e l o p m e n t of quantitative relationships between data from these analyses and data from core analysis measurements. Plugs containing significant inhomogeneities, such as laminae of distinctly different grain size or degrees of cementation, should be avoided or else erroneous variance in the data set will tend to blur what otherwise might be easily recognizable clear-cut relationships.
Zones having obviously low or high permeability as a result of their grain size, sorting, clay content, cement content, or extent of dissolution or fracturing, but which were not represented in the original suite of core plugs, should be sampled to evaluate their influence on flow properties. It is also important to assess the degree of vertical permeability across small scale potential permeability barriers such as stylolites, cemented zones, and healed fractures. If full diameter core analysis was conducted, a representative portion of each segment can be selected for analysis, and the porosity-permeability measurements for the whole segment can be assumed to apply to the portion sampled. Alternatively, core analysis and rock descriptive work can be conducted on plugs cut in selected segments of the full diameter core.
Petrographic analysis should be designed to generate quantitative data. Decisions as to which components should be measured separately in this analysis are based on several criteria. Generally, components that are relatively abundant (>10%) and whose abundance varies significantly in the interval being investigated should be measured separately. Porosity-permeability crossplots on which samples are plotted according to lithological data obtained from binocular
microscope analysis may reveal relationships indicating the relative i m p o r t a n c e of certain c o m p o n e n t s (Figure 3). Components that are finely crystalline, particularly diagenetic clays, may exert a major effect on permeability, even where their abundances are relatively low (3-5%).
Quantitative point count analysis is preferable to semiquantitative visual estimation. If such analyses are not feasible, use should be m a d e of visual comparators. If microcrystalline components having sizes <10 to 20 |im are present in more than minor amounts, it is recommended that X-ray diffraction analysis be used to estimate their abundance.
Step 3. Lithology-Log Response Analysis
If c o n t i n u o u s core is available, analysis can proceed directly to the p r e p a r a t i o n a n d c o r r e l a t i o n of vertical component profiles (steps 4 and 5). However, in many reservoirs, particularly those with thick gross pay intervals, core or sidewall core coverage is limited to only a few wells or intervals. To complete the lithological analysis, the geologist must resort to cuttings descriptions or log analysis (see the chapter on "Quick-Look Lithology from Logs" in Part 4). Commonly, the degree of detail that can be extracted from cuttings samples is not adequate for the objectives of the study, so the geologist must attempt to correlate log response with lithological data from intervals containing full diameter or sidewall cores and then apply these correlations to uncored intervals.
Step 4. Diagenetic Profile Construction
Once the lithological data for all wells involved in the study have been gathered, vertical profiles of diagenetic
Evaluating Diagenetically Complex Reservoirs
317
Suturing and
/
Overgrowth - Cemented /
Ss.
>/i
0.01 _ ,
0 2 41 6
Core Analysis
• -MicrocrystaIline Dolomite •-Mudstone
+ - Med.-gr. Ss1Abundant Matrix Fine-Grained Sandstones
«-Anhydrite Cemented
©-Dolomite Cemented
• -Cemented by Suturing
^ - Q u a r t z Overgrowth Cemented
- - A r g iIl l a c e o Iu s I
I
8 10 12 14 16 18
Ambient Porosity - %
component abundance should be prepared. Only components having a significant effect on reservoir quality should be profiled. In some instances, components measured separately but having similar overall effects on reservoir properties can be combined to form one profile.
If the abundances of the major diagenetic components are based strictly on visual estimation using a binocular microscope, qualitative (very minor, minor, moderate, or abundant) or semiquantitative (0-5%, 5-10%, or 10-15%) bar profiles can be prepared displaying sedimentological and petrophysical data (Figure 4). Where quantitative petrographic data are collected, they can replace the original bar profile data.
Figure 3. Porosity-permeability semilog crossplot with samples coded according to grain size, clay content, and dominant agent of cementation.
<—
Figure 4. Coregraph displaying geological and petrophysical parameters of a well based on core description.
I
Ambient Permeability
Grain SizeA-^Sedimentary Envi- Ambient Lithology £ Structures ronment Porosity
Sonic Porosity
Anhydrite Dolomite Quartz Cement Cement Cement
Zonation
Detailed Simpli tied
318
PART 6—GEOLOGICAL METHODS
Table 3. Distributions and Origins of Zones of Diagenetically Controlled Reservoir Quality Destruction and Enhancement
Distribution Cementation at top and/or base of reservoir
sandstone Cementation or dissolution along or in close
proximity to fractures or faults or in zones in communication with such features Cementation in zones that represent concentrations of chemically unstable detrital particles Zones of cementation or dissolution follow lamina having specific textural properties
Dissolution occurs in close proximity to major or minor unconformities
Dissolution occurs at crest of anticline or at updip pinchout of a reservoir unit
Increased cementation occurs below oil-water or gas-water contacts
Zones of increased cementation are facies selective and tend to follow zones reflecting specific subenvironments
Increased cementation occurs at positions of paleowater tables
Increased calcite/aragonite cementation in shallow marine phreatic zone where steep slopes or high energy conditions occur
Origin
Pressure and/or temperature decreases in fluids compacted out of bordering shale beds
Fluids moving along avenues of pressure release precipitate material or react with adjacent rocks
Variations in sedimentological conditions produce local concentrations of grains that dissolve and reprecipitate as cementing material
Migrating fluids are preferentially concentrated in highly permeable units or, under evaporitic conditions, in relatively fine units where capillary properties favor their retention
Invasion of fresh meteoric waters favors extensive dissolution and/or mineral alterations (e.g., kaolinitization of precursor clays or micas)
CO2 and/or organic acids generated during thermal maturation of organics seek structural or stratigraphic highs and generate acidic conditions
Early stage emplacement of oil or gas retards diagenetic alterations in the zone saturated with hydrocarbons
More extensive rates of evaporation in sabkha environments or more extensive pressure solution in low energy argillaceous zones
Increased evaporative precipitation occurs at or just above positions of paleowater tables if solutions are supersaturated; dissolution occurs just below if solutions are undersaturated
Conditions favorable to precipitation of calcite and aragonite include areas where CO2 degassing occurs, certain bacteria are present, and photosynthesis and/or respiration of reef organisms increase the pH of pore waters to values greater than 9
Step 5. Continuity Analysis
The objective of this step is to p r e p a r e a m o d e l that describes the three-dimensional distribution of the major diagenetic components represented on the vertical well profiles. The accuracy of and confidence in these correlations are governed by such factors as (1) the u n i q u e n e s s of the profiles constructed, (2) the degree of heterogeneity within the reservoir (see the chapter on "Geological Heterogeneities" in Part 6), (3) the degree to which the sample coverage in each well represents the variations present within the reservoir, (4) h o w well the geologist u n d e r s t a n d s the origins of a n d factors controlling the distribution of the major diagenetic alterations, (5) the well spacing, and (6) the thicknesses of the diagenetic zones.
Although some diagenetic components may display abundance trends that follow or are only slightly discordant with depositional trends, others may occur at large angles to t h e latter. S o m e of the m o r e c o m m o n d i s t r i b u t i o n s of diagenetic components and the geological factors controlling these distributions are listed in Table 3.
The basic correlation techniques in common use include marker and sequence analysis and, where continuity is very l i m i t e d , slice t e c h n i q u e s (Cant, 1984). C o r r e l a t i o n of diagenetic zones is most accurate when the origins and timings of the diagenetic events creating the components of interest are well understood, the sample and well spacing are relatively small, the diagenetic zones are relatively thick, and the sequence of zones is unique.
The distribution of reservoir fluids or pressures at the time of discovery of the reservoir, or at subsequent intervals during field development, m a y indicate the presence of continuous permeability barriers and thus may help to confirm the extent of s o m e diagenetic zones. W h e n the basic correlation techniques prove unsatisfactory or inadequate due to a high d e g r e e of c o m p l e x i t y or l o w d e g r e e of c o n f i d e n c e , the geologist may need to resort to special engineering techniques such as pulse testing (Pierce, 1977) or tracer studies (Wagner, 1977) (Table 4), or to probabilistic modeling (Hewett and Behrens, 1988) (see Part 8).
Preparation of contour m a p s for pertinent diagenetic
Evaluating Diagenetically Complex Reservoirs
319
Table 4. Methods of Assessing the Continuity of Diagenetic Zones
Technique Diagenetic profile correlation
Advantages
Rapid once basic data are collected; relatively inexpensive
Log correlation
Rapid; inexpensive
Fluid distribution Pressure distribution Transient and pulse testing Tracer tests Three-dimensional seismic
Rapid; inexpensive
Rapid; inexpensive
Provide information on reservoir connectivity, storage, transmissibility, and hydraulic diffusivity
Provides information on volumetric sweep, directional flow, flow barrier distribution and fluid velocity
Provides essentially continuous profiling across area of investigation
Disadvantages
May not provide unique solution; limited to zones with full-diameter and sidewall core coverage; degree of confidence low if zones correlated are thin or well spacing is large
May not provide unique solution; results dependent on ability to correlate lithology to unique log response
Degree of confidence low if zones correlated are thin or well spacing is large
Requires modem log suite (GR, density, neutron, or sonic)
Generally useful only over limited intervals and only in a few reservoirs
Generally useful only over limited intervals and only in a few reservoirs
Expensive; commonly does not provide unique solutions
Expensive; generally limited to closely spaced wells; slow
Resolution limited; expensive; interpretation may be difficult; may not provide unique interpretation
components may assist in the evaluation of zone continuity (see the chapter on "Subsurface Maps" in Part 6). However, the merging of data from more than one diagenetic zone may lead to misinterpretations of the degree of continuity present. Contour maps should always be interpreted in combination with detailed correlation cross sections. Preparation of threedimensional "spectragrams" may be helpful in visual correlation studies.
Stage 4. Preparation of an Integrated Geological Model
Following generation of the separate depositional and diagenetic models, data from these models must be merged to form an integrated geological model. To do this, the geologist must examine the two independent models and extract those factors that constitute the major controls on porosity, permeability, and hydrocarbon saturation. It may be useful to perform statistical analysis on the data using techniques such as correlation coefficient calculations, univariate regression, and cluster analysis, or multivariate techniques such as discriminant analysis and stepwise regression (see Part 8).
In some fields, it may be necessary to prepare separate structural and formation fluid models in addition to depositional and diagenetic models (Figure 2). In most reservoirs, however, structural and formation fluid factors are
not the major controls on fluid flow behavior, and the heterogeneities observed can be incorporated into the diagenetic model.
Stage 5. Reservoir Zonation
In this stage, it is necessary for the geologist to subdivide each well (reservoir quality profile) into relatively homogeneous units or zones. Initially it is best to construct a relatively detailed zonation reflecting major variations in reservoir rock characteristics regardless of their vertical dimensions (Figure 4).
Zonation may be facilitated by establishing criteria by which major categories of reservoir rock types present can be d i s t i n g u i s h e d . Semilog crossplots of p o r o s i t y a n d permeability keyed to texture, matrix content, and diagenetic component content are very useful. Through examination of these plots, the geologist can quickly separate samples into natural groupings (Figure 3). Because permeability directly reflects fluid flow capacity, it is the major parameter used to designate reservoir rock categories. Because porosity is commonly a major control on permeability, it generally exhibits a positive correlation with that variable. Attempts to create hierarchies of permeability heterogeneity based strictly on depositional criteria (Lewis, 1988) should be avoided in reservoirs where diagenetic alterations are major controls on
320
PART 7—GEOPHYSICAL METHODS
permeability heterogeneity. If natural g r o u p i n g s are not present, it may be necessary to set arbitrary group boundaries, such as porosity or permeability cutoffs.
Estimation of the lateral distribution of zones is then guided by relationships developed in the integrated geological model and documented in the form of maps and cross sections.
Stage 6. Applications Studies
Depending on the objectives of the study being conducted, the geologist needs to generate additional maps, cross sections, fence diagrams, or simulation models. For the purposes of reserve estimation, net pay maps are needed (see the chapter on "Effective Pay Determination" in Part 6). If the major controls on net pay distribution are diagenetic in nature, contouring of net pay m a p s should be guided by t r e n d s p r e s e n t on c o n t o u r m a p s of m a j o r d i a g e n e t i c
components. Average porosity maps should also be "guided" in this same fashion. Where simulation of fluid flow behavior is the object of the study, the geologist must assist in the preparation of a three-dimensional simulation model.
Stage 7. Model Testing and Revision
Where economics dictate, it may be necessary to test the accuracy of the models developed. This can include testing by history matching of pressures, production rates, and GOR values for segments of the model or full scale testing of the complete model (Weber et al., 1978) (see the chapters on "Product Histories" in Part 9 and "Conducting a Reservoir Simulation Study: An Overview" in Part 10). Testing can also involve drilling additional wells, conducting special engineering tests (pulse or tracer), and collecting geological data on additional samples. Revisions may also be required as additional wells, particularly infill wells, are drilled in the field.
Evaluating Tight Gas Reservoirs
Thomas F. Moslow
University of Alberta Edmonton, Alberta, Canada
INTRODUCTION
Gas reservoirs with estimated in situ gas permeabilities of 0.1 md (millidarcy) or less are officially recognized by the U.S. Federal Energy Regulatory Commission (FERC) as "tight gas reservoirs." This absolute value for classification as a tight gas reservoir was critically important during the late 1970s and early 1980s to qualify for federally allowed enhanced prices of tight gas. Since that time, however, and for all practical purposes, a tight gas reservoir is generally recognized as any low permeability formation in which special well completion techniques are required to stimulate production (Table 1). The most commonly used recovery technique is hydraulic fracturing, without which many tight gas reservoirs would not be economical (see the chapter on "Stimulation" in Part 9). Thus, most low permeability gas reservoirs are considered "unconventional."
Low permeability gas-bearing formations occur in almost all gas-producing sedimentary basins worldwide. In North America, the vast majority of tight gas reservoirs can be grouped into two main geological categories: (1) Devonian shales from eastern United States and Canada, and (2) low permeability sandstones from throughout the United States and from the Western Canada Sedimentary basin (Spencer and Mast, 1986). It has been estimated that in the United States alone, tight sandstone formations are likely to have recoverable reserves ranging from 100 to 400 tcf, and Devonian shales have recoverable reserves of u p to 100 tcf (Office of Technology Assessment, 1985; cited in Spencer and Mast, 1986). The successful exploitation of tight gas resources in the future will depend in large part on advancements made in the proper geological evaluation of low permeability reservoirs.
Table 1. Problems and Approaches to Evaluating Tight Gas Reservoirs
Reservoir Formation Problem Reservoir performance is controlled by
sedimentary facies, lithology and/or geometry
Reservoir performance and formation permeabilities are negatively impacted by presence of detrital or authigenic clays (very common)
Reservoir performance is dictated by origin and distribution of natural fractures
Fracture mineralization impacts reservoir performance.
Porosity and permeability trends are controlled by clay diagenesis or secondary cements
Reservoir gas accumulations lack a floored gas-water contact
Overpressured formation or reservoir (occurs frequently due to common distribution of tight gas with basin center locations and excessive overburden)
Underpressured formation/ or reservoir due to stripping (erosion) of overburden
Approach
Determine sedimentary characteristics and origin of facies through core description and construct predictive models for lateral variability and heterogeneity of reservoir units
Identify and map reservoir facies with the least detrital clay content; avoid treatment of formation by acidization or injection of any fluids; use oil-based muds; enhance recovery through artificial fracturing of the formation (Moslow and Tillman, 1986)
Evaluate the relationship between fracture occurrence and lithology (Pitman and Sprunt, 1986)
Determine the nature, origin, and timing of mineralization through petrographical and stable isotope techniques (Pitman and Sprunt, 1986)
Determine the petrological and mineralogical history of reservoir facies; identify and map "diagenetic facies" relative to sedimentary facies
Map structural, isothermal, and pressure gradient contours that are likely coincident with boundaries of the gas envelope (Rose et al., 1986)
Requires appropriate exploration strategies or reservoir engineering approach to gas recovery
Requires appropriate exploration strategies or reservoir engineering approach to gas recovery
321
322
PART 7—GEOPHYSICAL METHODS
AMOCO Whiskey Buttes No. 16
Lincoln County, Wyoming Moxa Arch, Green River Basin
GH Perforated Interval • Cored Interval Ш Gas Show
11,0004 11,050'
Facies Foreshore/Upper Shoreface Delta Plain (Crevasse Splay & Marsh) / Abandoned Channel
Distributary Channel
C- ^CZf I
-V
Lower Shoreface to Inner Shelf
IP 1012 MCF
ФСЫ1
0FDC
-11,000'
-11,100'
Figure 1. A cored sequence of tight gas reservoir facies and correlations to eletric log responses of the Frontier Formation, Green River basin, Wyoming. Lithologies and sedimentary characteristics are summarized in this kind of description; facies and environments of deposition are shown on the right. (From Moslow and Tillman, 1986.)
TOOLS AND METHODS
The extremely low permeability of tight gas reservoirs severely restricts the ability of gas to migrate appreciable distances. Consequently, the most important geological characteristic of this type of reservoir is the n a t u r e a n d distribution of porosity and permeability (Table 1). The most common reason for the minimal permeabilities is the occlusion of interstitial pore throats by detrital or authigenic clays or cement (see the chapter on "Rock-Water Interactions: Formation Damage" in Part 5). Thus, a proper geological evaluation of tight gas reservoirs requires a multidisciplinary approach to assess the depositional and diagenetic controls on reservoir quality and heterogeneity (also see other chapters in Part 6).
Facies Determinations
Since most tight gas reservoirs in North America are of detrital origin (shale, siltstone, and sandstone), primary processes of deposition, inferred f r o m the examination of sedimentary characteristics in core, can have a strong impact on preserved porosity and permeability trends. An example of a s e d i m e n t o l o g i c a l d e s c r i p t i o n a n d e n v i r o n m e n t a l
interpretation of cored facies f r o m a tight gas reservoir is shown in Figure 1.
Core to Log Correlations
Documenting characteristic log signatures for reservoir facies can provide a valuable tool for constructing regional cross sections, determining facies relationships, and extrapolating reservoir geometries in areas of minimal or nonexistent core control (see the chapter on "Quick-Look Lithology from Logs" in Part 4). Commonly, the gamma ray log provides the most distinctive log signature for individual facies (Figure 1). For low permeability gas reservoirs, crossover of the compensated neutron-formation density logs is the most reliable well log for indicating gas-saturated and porous intervals and for determining which intervals in the reservoir should be perforated and/or stimulated by hydraulic fracturing.
Stratigraphic Cross Sections
Lateral variability in facies relationships, and thus reservoir continuity and heterogeneity, are best determined from the construction of stratigraphic cross sections (see the
chapter on "Geological Cross Sections" in Part 6). An example of a cross section through part of a tight gas reservoir is shown in Figure 2. Facies interpretations are based on core descriptions a n d extrapolation of log signatures for each cored facies to adjacent uncored wells. Distributary channel sandstones form the reservoirs, and bay, marsh, and crevasse splay mudstones form the seal. The lack of production in the two wells to the east is attributed to the pinching out of these mudstone facies and substantiates its importance as a stratigraphic seal. Note the laterally discontinuous nature of individual reservoir sandstone beds as depicted in the cross section.
Petrophysical Properties of Reservoir Facies
Average core analysis values for porosity, permeability, oil, gas, and water saturation should be determined for each facies recognized to identify those facies of greater and lesser reservoir quality (Figure 3a). In gas-bearing sandstones, very low values of porosity and permeability are acceptable and expected. While the average air permeability values rarely exceed 1.0 md (millidarcy) for tight gas reservoirs, a significant difference in permeability values often occurs between facies (Figure 3b).
Anomalously high values from core analysis measurements may also identify zones of fracture porosity and permeability in tight gas reservoirs (see the chapter on "Evaluating Fractured Reservoirs" in Part 6). However, one
EvaluatingStructurallyComplexReservoirs
323
must be careful in interpreting such results because erroneously high measurements can also be produced by b y p a s s i n g or artificial f r a c t u r i n g of core s a m p l e s d u r i n g analysis. Checks should be made to ensure that a sufficient number of samples have been analysed for each facies or unit and that permeability and porosity values correspond to observed lithologies in core.
Petrological and Mineralogical Assessment
A petrological thin section, SEM, and X-ray diffraction analysis of core samples from each sedimentary facies is highly recommended in any geological evaluation of tight gas reservoirs (see the chapters on "Thin Section Analysis" and "SEM, XRD, CL, and XF Methods" in Part 5). Analyses of several tight gas sandstones have attributed the low average permeabilities, and thus poor reservoir quality, to the presence of authigenic or detrital clays or cements (Masters, 1984; Spencer and Mast, 1986). Since the occurrence of these constituents can be quite variable within a depositional system and can be facies d e p e n d e n t , a b r o a d r a n g e of porosities, permeabilities, and gas saturation values often exists in any reservoir (Figure 4). Identifying and mapping those units of greatest reservoir potential are key to a successful evaluation.
The common association of clays with tight gas reservoirs makes them very sensitive to formation damage. Hydraulic fracturing is therefore the least destructive and most preferred well stimulation technique.
с
West
о
Whiskey Buttes No. 20
IP 650 MCFD
Whiskey Buttes No. 19
IP 4,000 MCFD
Whiskey Buttes No. 5
IP 3,687 MCFD
Whiskey Buttes No. 18
IP 775 MCFD
Oj NJ
O
Whiskey Buttes No. 10
H
к
С'
СП
East
О
Г-ч
O
О
Whiskey Buttes No. 7
С)
с!
г-1
CNL/FDC
gС П
о
СЛ
I Perforated Interval Cored Interval
Facies ZZZ] Distributary Channel
Abandoned Channel HH Foreshore and Shoreface IzZ3 Inner Shelf HS Delta Plain
MOXA AREA
4
UK= ,1
C H O U t t C TiON
Figure 2. The depositional dip-oriented cross section through the Frontier Formation, Moxa arch area, Wyoming, showing facies relationships and inferred geometries. (Modified from Moslow and Tillman, 1986.)
CHANNEL FACIE8
NEARSHORE MARINE FACIES ( D > l f Front)
DELTA PLAIN
EvaluatingStructurallyComplexReservoirs
325
CHANNEL FACIES
NEARSHORE MARINE FACIES ( D d U Fronl)
DELTA PLAIN
ACTIVE
11.5
CHANNEL
I1II 9.8 8.4 I
UPPER SHOREFACE
SHELF TRANSITION
Vt,.
Ш i
(a)
Kmd-
LOWER SHOREFACE
o.6;
CHANNEL LAO
i i I ABANDONED CHANNEL 0.3,
UPPER
A SHOflEFACC
FORESHORE
.07
SHELF TRANSITION
0.8
(b)
.33-
.1
Figure 3. Histograms showing (a) average porosity values and (b) average permeability values for cored tight gas reservoir facies. (From Moslow and Tillman, 1989.)
FACIES
% TOTAL IN SUBSURFACE
CHANNEL FACIES
• PARTIALLY ABANDONED
1
• ACTIVE CHANNEL
65
• Fine Grained Sandstone • Coarse Grained Sandstone • Conglomeratic Sandstone
SHALLOW MARINE
24
(SHELF-SHOREFACE-FORESHORE)
DELTA PLAIN
10
(BAY-SPLAY-MARSH)
% OF PERFORATED
INTERVAL
12 80 (30) (37) (13) 8
PETROGRAPHIC RESERVOIR QUALITY
VERY LOW I LOW I MOD. I HIGH
Figure 4. Correlation of sedimentary facies and lithologies to petrographic reservoir quality. Distribution of reservoir facies in the subsurface is compiled from observations of cores, well logs, and cross sections. (From Moslow and Tillman, 1986.)
Evaluating Fractured Reservoirs
Ronald A. Nelson
Amoco Production Company Houston, Texas, U.S.A.
INTRODUCTION
A fractured reservoir is one in which naturally occurring fractures either have or are predicted to have a significant effect on reservoir fluid flow in the form of (1) increased reservoir permeability, (2) increased porosity, a n d / o r (3) increased permeability a n i s o t r o p y . Four basic types of reservoir fractures can be defined (Figure 1):
Type 1—Provide the essential porosity and permeability to the reservoir
Type 2—Provide the essential permeability Type 3—Provide a permeability assist to an already
producible reservoir Type 4—Impart no positive reservoir quality but create
strong reservoir anisotropy and inhomogeneity (Nelson, 1985) Anticipated exploration and development problems associated with these four reservoir types are summarized in Table 1.
EVALUATION
When evaluating a fractured reservoir, the analyst must follow these steps (Nelson, 1985):
1. Determine the origin of the fracture system found or the type of fracture system that is being explored for based on geometric characteristics of the fractures, their
ALL
POROSITY
Figure 1. Crossplot showing the relative positions of fractured reservoir types 1 through 3 and normal matrix reservoirs (m) in the percentage of porosity and permeability space. Symbols: kf= fracture permeability, <j)f = fracture porosity, kr= matrix permeability, фг= matrix porosity.
distribution in three dimensions, and empirical models of fracture system genesis (Stearns and Friedman, 1972). 2. Determine the reservoir properties of both the rock matrix housing the fractures and the fracture system (van Golf-Racht, 1982). 3. Determine the degree of fluid flow interaction or crossflow in either a qualitative or quantitative manner. 4. Classify the fractured reservoir on the basis of what the fracture system provides to overall reservoir quality.
This evaluation should be initiated early in the life of a field a n d can be terminated at any time if preliminary results suggest that either the reservoir is uneconomic or that fractures appear to be a minor component of total reservoir quality.
DATA ACQUISITION
The data needed to evaluate a fractured reservoir are commonly obtained from detailed observation of reservoir cores (including their associated well logs) and relevant outcrops. Cores and core-calibrated well logs give the onedimensional depth and lithological variation in fracturing in one position in the reservoir. Outcrops give appropriate twodimensional variation of the reservoir in m a p view within individual layers. Together, a three-dimensional representation of the reservoir can be a p p r o x i m a t e d a n d correlated with production histories.
CORE AND WELL LOG ANALYSIS
The following procedures have proven useful in fracture analyses of core (after Nelson, 1985):
1. Lay out all of the core from a given well for observation. Fit the core together and note continuous segments (see the chapter on "Core Handling" in Part 3).
2. Do initial observations while the core is clean and in its whole-core state (unslabbed) (see the chapter on "Core Description" in Part 5). In this state, all fractures are present and their angular relationships retained. The core can be slabbed later and a finer internal description done at that time. Large diameter cores are generally preferred to small diameter cores. The larger diameter core depicts more regularly spaced high angle fractures and thus gives a better representation of fracture spacing and a better measurement of reservoir properties from core analyses.
3. Have the core analyses (porosity and permeability) data in hand while observing the core to determine the relative effect of the features observed (see the chapter on "An Overview of Routine Core Analysis" in Part 5). Obtain three-dimensional whole-core analyses, if
326
Evaluating Fractured Reservoirs
327
Table 1. Potential Production Problems Inherent to Each Fractured Reservoir Type
Type Type 1. Fractures provide essential porosity and permeability Type 2. Fractures provide essential permeability
Type 3. Fractures provide a permeability assist
Type 4. Fractures create only permeability anisotropy
Problems
Rapid decline rate Often limited reserves Rapid coning Fracture intensity and width must be very high
Poor fracture and matrix interaction can isolate matrix hydrocarbons
Rates often need to be controlled to not pull water preferentially along fractures and not through matrix
Permeability contrast between fractures and matrix reduced sweep efficiency in flooding
Development must be highly tailored to the reservoir Matrix block size should be small
Nonrecognition of the fracture system leads to poor estimations of drainage area and reserves per well and unusual response in secondary recovery
Apparently isolated reservoirs are often really interconnected
Fracture system must have good continuity or "connectedness"
Reservoir often highly compartmentalized Flow predictions from standard core or log analyses do not
match reservoir performance Wells may display a larger skin factor than anticipated Best to have fractures of one orientation than multiple
intersecting trends
possible, when the fractures or other anisotropic features are predicted to be important. 4. Create the fracture stratigraphy (involves steps 5 through 16). 5. Construct a core deformation strip log at a relevant scale that includes the positioning of major Hthology and formation breaks, lost core and noncored intervals, oil-water or gas-water contacts (if available), fractures, and stylolites. Hairline fractures in the core are often difficult to see. These can be enhanced by painting the core with a volatile liquid and observing the drying pattern. Overlooked fractures may "jump out" at the observer by persistence of wetting along hairline fractures reaching the core surface. Additional characterization of hairline or larger fractures internal to the core can be accomplished by core X-radiographs or core computer tomography scans.
6. Record the fracture distribution with depth (Figure 2). (Recording of this and all following data could be done in a tabular format compatible with computer entry to facilitate output compatible with other strip and well logs.)
7. Relate fracture distribution to rock type. 8. Record the dip of fractures either real or apparent. 9. Back u p core observations with appropriate logs from
the same zones as core from the well for effective extrapolation to uncored wells (see the chapter on "Formation Evaluation of Naturally Fractured Reservoirs" in Part 4). Record the strike of features if
the core is oriented core or if it is locally oriented either mechanically or by oriented logs such as the Borehole Televiewer, Formation MicroScanner, or high resolution dipmeter (Plumb and Luthi, 1986). (For more on these methods, see the chapters on "Core Orientation" in Part 3 and "Borehole Imaging Devices" and "Dipmeters" in Part 4.)
10. Look for intersection angles of fractures as expressed on the outside surface of the core or on the ends of the samples and record the true or apparent angles (Figure 2).
11. Determine which of the fractures in the core are natural or induced (Kulander and Dean, 1985).
12. Describe stylolite distribution (position, rock type, and postulated G1) (Nelson, 1985; van Golf-Racht, 1982).
13. Determine fracture plane morphology, paying particular attention to any partial mineralization along the fracture planes that might act as a natural proppant during depletion (Nelson, 1985). If present, determine its mineralogy and predicted relative compressive strength.
14. Measure the relative size or height of the fractures, paying particular attention to any rock features that tend to control the vertical extent of the fractures, such as lithology breaks, bedding planes, stylolites, or unconformities.
15. Observe the width and width variation of the fractures. Measurements of width could be made with a scale, micrometer, or calipers or by impregnation with epoxy
328
PART 7—GEOPHYSICAL METHODS
for documentation in reports, if needed. 25. Select samples for mechanical testing, if appropriate. 26. If needed, set u p a checklist of important parameters
and document as many parameters as needed on a footby-foot basis (on each whole core piece is ideal). Such a checklist should be compatible with direct computer input of the tabular data (see Figure 2). 27. Write down impressions and conclusions arrived at before leaving the core. Include thoughts on the relative importance of fractures to production and flow, the permeability of the fractures present, fracture porosity (qualitative), fracture fill and relative compressibility of the fractures, fracture origin and continuity in the reservoir, the percentage of real versus induced fractures, and the fractured reservoir type.
Figure 2. Schematic of typical features to measure and record in core analysis. (From van Golf-Racht, 1982; courtesy of Elsevier. Typical recording format from Reiss, 1980; courtesy of Gulf Publishing Co.)
or plastic for either thin section measurement of epoxy width or dissolution of matrix leaving the width at the depth approximated (Nelson, 1985). 16. Estimate or measure fracture spacing and its variability with depth (Narr and Lerche, 1984; Nolen-Hoeksema and Howard, 1987). 17. Determine principal stress directions and the origin and continuity of the fracture system(s) (Stearns and Friedman, 1972; Nelson, 1985). 18. Determine the relative timing of deformational events from cross-cutting relationships or paragenetic sequence (Lindquist, 1983). 19. Relate fracture distribution to rock properties and rock fabric (such as composition, porosity, preferred grain orientation, bedding, and cross-bedding). Structural and petrological descriptions go hand in hand (Nelson, 1985). 20. Select samples for additional petrophysical or petrological determinations (such as X-ray, thin sections, and permeability) (see the chapters on "Thin Section Analysis" and "SEM, XRD, CL, and XF Methods" in Part 5). 21. Estimate the permeability of properly oriented individual fractures from core analyses (see the chapter on "Permeability" in Part 5). 22. Qualitatively estimate fracture and matrix porosity interaction or at least determine if there is any evidence of impedance to cross-flow (Nelson, 1985) (see the chapter on "Porosity" in Part 5). 23. Determine the fractured reservoir type (Nelson, 1985). 24. Photograph important relationships shown in the core
OUTCROP ANALYSIS
When collecting fracture data from outcrops, one should look for rocks similar to the reservoir rocks of interest on similar outcropping structures. Look at m o r e of the rock section than just the prospective reservoir rock, including potential sealing beds, and pay attention to relative fracture intensity between layered units.
1. Select outcrop areas having representative fracture patterns for specific structural positions. Record strike and dip data, or at least strike data. Additional sites can be chosen for one unit at different positions within the structure.
2. For a so-called quantitative measurement station (having enough fracture measurements to be statistically meaningful), all fractures in the measurement area should be recorded and should number from about 100 to 150 fractures. The actual number should depend on the complexity of the fracture patterns present.
3. For a so-called qualitative measurement station (not having enough fracture measurements to be statistically meaningful), only general fracture trends are recorded, along with a judgment of relative abundance of the individual fracture orientations. These are often used when spot checking areas between statistical measurement stations. If no major change in orientation and intensity of fracture patterns is observed away from nearby statistical measurements stations, the use of qualitative stations gives valid intermediate data.
4. At the individual measurement stations, the analyst should record as much of the following information as possible (Nelson, 1985):
• A locality number corresponding to a location map • A brief station location and description • The formation(s) and rock type(s) measured at the
station • Strike and dip of the bed(s) • Structural position and attitude of the structure • Bed thickness(es) • Fracture spacing (number of fractures per foot of
outcrop) along lines that are parallel and
MATRIX PERMEABILITY (kr;
k r = 1000
100
10 MD
50% -- 91% - - 99% --
1000p
TJ E
9.-1% - - 50% - - 91% - -
1.0% --9.1% - - 50% --
0.1% --1.0% --9.1% -3 O
100 —
m <
UJ
QC Ш O.
LU
OC z> hO <
a:
1.0x10
Evaluating Fractured Reservoirs
329
FRACTURE WIDTH (e) IN C M
Figure 3. Total reservoir permeability due to fractures plotted as a function of fracture width, fracture spacing, and matrix permeability. (From Nelson, 1985; courtesy of Gulf Publishing Co.)
perpendicular to the major fracture trend (with strike of measurement line), or parallel to bedding strike and dip (see La Pointe and Hudson, 1985) • Comments • A postulation of the maximum compressive stress direction (O1) at fracturing • A postulation of the origin of fracture sets
5. Each individual fracture measurement at the station should record as much of the following data as possible (see Nelson, 1985):
• A sequential number for the fracture • Fracture strike • Fracture dip • Fracture morphology (gouge, open, mineralized,
slickensided, or vuggy) • Fracture length or predefined class length • Comments
6. At convenient times, the fracture data should be plotted in preliminary form on either rose diagrams or pole plots (к diagrams). Such preliminary plotting is necessary in the field to establish trends and application to simple geological fracture models. In this way, working interpretive models can be created and altered or updated while field data are still being gathered. The observer should always examine fracture patterns in light of their relationship to their localities and to local structural configuration.
7. The number and frequency or spacing of quantitative measurement stations are generally high in the early stages of study in a region and decrease in relationship
to qualitative stations throughout the study. 8. When dealing with outcrops containing a
predominance of either contractional fractures or fractures related to unconformity surfaces (Nelson, 1985), much of the previous quantitative orientation data will be ill-defined due to their isotropic or irregular distribution in orientation. In these outcrops, matrix block size (fracture spacing in three dimensions) are very important, as are lateral distribution and lithology. 9. Photograph all outcrops measured and take block samples (about 10" x 6" x 5") of all major units of interest for petrophysical and possibly mechanical testing.
CROSSPLOTS
In d o i n g core a n d o u t c r o p a n a l y s e s of f r a c t u r e s to determine reservoir properties and reservoir type, it is often difficult to judge the relative effect of the fracture system. Two crossplots can be used to alleviate this problem (Figures 3 and 4). These plots are the percentage of total reservoir permeability (Figure 3) and porosity (Figure 4) as a function of fracture width and fracture spacing for various orders of magnitude of matrix values. Assumptions can be made for w i d t h of the fractures at d e p t h . Matrix properties are determined from core analyses, thin sections, etc. and the relative c o n t r i b u t i o n of the f r a c t u r e system for v a r i o u s spacings can then be read off the graph. Ideally, ranges in values for width and spacing of fractures are used and a box or area created on the graph within which the actual value is
330
PART 7—GEOPHYSICAL METHODS
MATRIX POROSITY (фг)
b = 20
10
1%
33% - - 50% - - 91% --
10.0
ш O CC
LLI CL
Z 4.8% - - 9.1% - - 50% --
0.5% - - 1.0% - - 9.1% -0.05% - - 0 . 1 % - - 1.1% --
(/) O DC O Q.
Ш
GC 3 OH< DC
0.01 1.0x10
FRACTURE WIDTH (e) IN CM
Figure 4. Total reservoir pore volume due to fractures plotted as a function of fracture width, fracture spacing, and matrix porosity. (From Nelson, 1985; courtesy of Gulf Publishing Co.)
likely to occur. These figures assume one set of regularly spaced fractures, hydraulic apertures, and parallel plate laminar flow (Nelson, 1985).
CHECKLIST FOR SEQUENCE OF STUDY
The following checklist will aid in outlining the work flow of a fractured reservoir evaluation:
1. Document fracture presence • Logs (Aguilera, 1980) • Cores (Nelson, 1985) • Anomalous flow rates (Aguilera, 1980)
2. Determine if structure is present • Seismic, gravity, magnetics • Structuremaps • Dipmeters (Plumb and Luthi, 1986)
3. Determine lithological control of fracture distribution • Logs (Aguilera, 1980) • Cores and fracture stratigraphy (Nelson, 1985) • Logs and flow tests or DSTs (Aguilera, 1980, van Golf-Racht, 1982)
4. Document fracture systems geometry • Borehole Televiewer or Formation MicroScanner (Plumb and Luthi, 1986) • Cores (Nelson, 1985) • Predictions (including relevant outcrops) (Stearns and Friedman, 1972; Nelson, 1985)
5. Document fracture morphology • Cores (Nelson, 1985) • BoreholeTeleviewer • Predictions (including relevant outcrops) (Nelson, 1985, p. 56)
6. Determine fracture type (origin)
Application of observations to empirical models using data from steps 1 through 5 (Stearns and Friedman, 1972; Nelson, 1985) 7. Predict fracture distribution and extent Extrapolation using fracture type and observations 8. Estimate fracture spacing and spacing variability • Cores (Nelson, 1985; Narr and Lerche, 1984) • Borehole Televiewer (Plumb and Luthi, 1986) • Predictions (including relevant outcrops)
(Nelson, 1985; La Pointe and Hudson, 1985) 9. Estimate fracture width
• Laboratorydata • Flowtestdata • Thin sections • Impregnation and dissolution 10. Estimate reservoir properties at depth • фш, к (Nelson, 1985; van Golf-Racht, 1982) • ф,, kf (Nelson, 1985; van Golf-Racht, 1982) • Using data from steps 7 through 9 11. Estimate fracture and matrix interaction • фуг/ фот interaction • kf/km contrast 12. Correlate small scale petrophysical properties with large scale reservoir engineering tests (Aguilera, 1980; van Golf-Racht, 1982) 13. Determine fractured reservoir type Correlate matrix and fracture properties and their communication to determine relative contribution of the fracture system and potential recovery problem (Nelson, 1985) 14. Make conclusions relevant to the type of evaluation • Early exploration evaluation • Estimation of economic potential • Recovery planning and reservoir modeling
Evaluating Structurally Complex Reservoirs
J. R. Hossack
BP Exploration Middlesex, U.K.
D. B. McGuinness
BP Exploration Houston, Texas, U.S.A.
INTRODUCTION
Evaluation of a structurally complex reservoir requires integration of geological and geophysical data to generate maps and cross sections that show the attitude, geometry, and thickness of key reservoir beds; the locations of crestal highs and synclinal troughs; the position, dip, and character of faults; and the location and orientation of the cutoffs of key beds on both sides of the faults. Also, fault blocks must be accurately delineated since they can effectively compartmentalize a reservoir. Maps and sections must be integrated so that they agree with each other, and they should be tested for viability and admissibility by balancing or backstripping.
SECTION CONSTRUCTION
Well and Seismic Constraints
Cross sections and maps are usually constructed simultaneously and need to be continually checked against each other (see the chapters on "Geological Cross Sections" and "Subsurface Maps" in Part 6). The first well in a field provides depth, velocity, dipmeter, and hydrocarbon distribution information that can improve the accuracy of predrill seismic and geological interpretations. Time pick corrections and seismic reprocessing generate improved seismic sections. Ideally, seismic sections should be time migrated and displayed with no vertical exaggeration so that true scale cross sections can be constructed for section restoration and balancing (Dahlstrom, 1970) (see Part 7).
Production wells m a y lie off the lines of seismic and geological section so that data will have to be projected carefully onto the lines using seismic cross lines or downplunge projection (DePaor, 1988).
Orientation
A balanced section can only be constructed in the direction of the regional displacement direction of the faults or in the d i r e c t i o n of flexural slip on the fold l i m b s (Dahlstrom, 1970). The appropriate direction can be chosen by an analysis of regional structure m a p s using the b o w string rule, lateral ramps, or drawing the section normal to the trend of the regional compressional or extensional folds (Woodward et al., 1985).
Structural Style
A p p r o p r i a t e seismic lines, close to the line of section, define the structural style of the folds and faults, so this style should be incorporated directly into the section (Dahlstrom, 1970). Dip domain construction methods are popular guides to section drawing (Suppe, 1983; Groshong, 1989a) in both compressional and extensional systems.
Use of Dipmeter
Cross sections can be more highly constrained using statistical curvature analysis techniques (SCAT) on dipmeter d a t a (Bengtson, 1982) (Figure 1) (see the c h a p t e r on "Dipmeters" in Part 4). This method allows determination of the positions in a wellbore of important axial, crestal, and fault surfaces and their strike and dip (Figure 2). Hence, structures can be projected in section away from wellbores and used to sketch the structure in profile.
Dip Isogons
Dip isogons, or contours of equal dip in the plane of the section (Figure 3) (Ramsay, 1967), can be used to fill in the geological section. The isogons can be located from projected dipmeter data and projected or correlated between the wellbores following the rules described by Ramsay and Huber (1987). A series of dip segments along the various isogons helps the geologist sketch the fold profiles. Interpolation between the isogons can also be carried out using the dip domain methods previously described or by cubic spline interpolation (McCoss, 1987).
Relationship of Folding and Faulting
Modern theories of structural geology generally relate the formation of folds to a c c o m m o d a t i o n on irregular fault surfaces (Hamblin, 1965; Dahlstrom, 1970). Generally, the folds are more obvious on seismic sections than faults, but fortunately there are geometric rules that allow us to predict one shape from the other (Suppe, 1983; Verrall, 1982; Gibbs, 1983; Williams and Vann, 1987; Groshong, 1989a) in both extensional and compressional examples. An example of a cross section solution explaining the relationship between extensional rollover and listric faults is shown in Figure 4.
386
332
PART 7 — G E O P H Y S I C A L METHODS
APIot FEET
no «то a »o i»o о
D Plot
4000
• K:
5000
.CRESTAL
\ TV I* 6000 sS ^
7000
<C\
8000
я
9000
1
10,000 ,V
11,000,CRESTALX*
У PLANE/ *
12,000 у : - : -
13.000U.
TPIot
LPIot
DVA Plot
SW
UNCONFORMITY
NE NW
mNU
ANTICLINAL CURVATURE CR ESTAL
xK HP-LAAMNhE
SYNCMLINNAALL. . i T f<T P ^
CURVATURE
ZJT
.— J
IPR
_
MINOR SW DIPPING THRUST
sr.
I
УЛ ANTICLINAL
\П . CURVATURE
to 40 IIO (40 IOO
I rSE
ALL DEPTHS
:•<': л
••) I Л .OrJ1CRESTAL LINE A j Y ' ,
/V •
CRESTAL TLNE.JT
••/.•.'
DOUBLE PLUNGE PATTERN
NW PLUNGE PATTERN
.^/^RESTAL
PLANE
A N T I C L I N A L / } 4 ^ . AXIAL PLANE:
SE
CURV A T U R E ^ V * ; DIPS 8 0 ° NE
STRATIGRAPHIC 1 SCATTER IN POINT OF ROCKS SS
-J
I
9800" TO 13,000"
120*
• > "'it
•. •x V
-.V4T7-
•
I .y
SE PLUNGE PATTERN
Figure 1. SCAT plots used to define the complex structure seen in the discovery well of the Rail Road Gap oil field, California. The five plot types are (from left to right) azimuth versus depth (A plot), dip versus depth (D plot), dip versus depth in the direction of greatest curvature (T plot), dip versus depth in the direction of least curvature (L plot), and dip versus azimuth (DVA plot). (From Bengtsen, 1982.)
NW PLUNGE - 6* INCREASING TO
7000
8000
WATER
oil^water.
|8E PLUNGE - 7* INCREASING
—CL
TO 9*
PLUNGE REVERSAL AXIS
ACTUAL O/W -
CONTACT
UNFORSEEN T
^ TEAR
/
X-FAULT /
LONGITUDINAL SECTION
9000
*AXIAL * -RF»LANBI
INFLECTION PLANE
(J) DC UJ ILU
S O -J
TOP OF C CONTOUR MAP PREDICTED FROM DIPMETER DATA
1.0 KILOMETER
PREDICTED O/W
CONTACT
INFLECTlONXAe.-
PLANE
^ ~
S b ' JL
^
Г1
TD 12,731* V ' T - - -
TRANSVERSE SECTION
Figure 2. Predicted transverse and longitudinal cross sections and contour map derived from SCAT plots. Depths are subsea depths (From Bengtsen, 1982.)
Evaluating Structurally Complex Reservoirs
333
Regional
h = heave
Total distance to fault 3 4 56 7 8
(a)
distance below erence bed
STRUCTURAL SURFACES DEFINED BY SCAT DIPMETER ANALYSIS
UN - UNCONFORMITY APa - ANTICLINE AXIAL PLANE - - IPa - ANTICLINAL INFLECTION SURFACE
K - FAULTS MARKED BY DRAG PATTERNS
HC - HANGINGWALL RAMP CUT OFF
слпп„
ISOGONS # - THE DIP VALUE ALONG THE ISOGON
2000m
SCALE : V - H
Figure 3. Cross section through an asymmetric ramp anticline, Whitney Canyon field, Wyoming, with SCAT and isogon data superimposed. Unconformities, axial planes and inflection surfaces have been identified from the dipmeter data and projected away from the wellbore. Isogons are contours of equal dip (see Ramsay, 1967) and can constrain the shapes of folds in section. (From Lammerson, 1982.)
Balanced Cross Sections
Balanced cross sections are used to test the viability or admissibility of a cross section. The deformed cross section is redrawn on a template in the undeformed state so that the beds are unfolded and the offsets on the faults removed (Figure 5). Section balancing requires reference pin-lines and loose lines at o p p o s i t e e n d s of the section f r o m which measurements of bed lengths are made. Bed thicknesses and bed lengths are generally retained so that the deformed and undeformed cross sections have the same area. For an ideal restoration, there should be no gaps or overlaps between adjacent fault blocks (Dahlstrom, 1970; Woodward et al., 1985).
Balanced sections were first constructed for thrust belts, but Gibbs (1983), Groshong (1989a), and Rowan and Kligfield (1989) have successfully applied the method to extensional and salt-related structures. Extensional section balancing is more difficult than compressional balancing because of the
Figure 4. Modeling extensional fault shapes from the rollover geometry, (a) The Groshong (1989b) method uses oblique simple shear with a reference grid constructed with a spacing equal to the fault heave. Distance 2 from the rollover up to regional elevation of the same reference bed is transferred to 2'; likewise, 2' + 4 is transferred to 4' and so on to complete the fault trajectory. Interpolation between these points is carried out using a half grid spacing, (b) Fault trajectory reconstruction by the Groshong (1989b) method uses simultaneous modeling of three horizons. Dashed trajectories are individual solutions; solid lines are the preferred solution. (From Hossack, unpubl. data, 1988.)
bed thickness changes that occur across faults. The balancing template has to show these thickness changes accurately. Generally, computer-aided methods are essential because they can sequentially backstrip the section to remove tectonic as well as c o m p a c t i o n strains. E x a m p l e s of these are described by Rowan and Kligfield (1989), Worrall and Snelson (1989), and Shultz-Ela and Duncan (1990).
MAP CONSTRUCTION
Structure Contour Maps
The geometry of the field is defined by a series of structure contour m a p s of key reservoir horizons (Figure 6a). The maps, showing several levels through the prospect or reservoir, are generated from well elevations of reference beds or depth-converted seismic sections (see the chapter on "Subsurface Maps" in Part 6). Workstations for threedimensional seismic interpretation considerably aid the process because the shapes of the structure contours and the faults are readily observable on horizontal seiscrop sections generated by the workstation (Brown, 1986) (see Part 8). Contour maps can be quickly generated from stacked seiscrop sections.
Faults must be located in wellbores by omission (extension fault) or repetition (reverse fault) of stratigraphic section.
334
PART 6 — G E O L O G I C A L METHODS
ч?
-J
8
1
I
< 1 Jl H l d W
I
- I
I
2 Mm
m
г* *
Ш
И* а
«л о -
(IJ) HidX)
These are defined on the electric logs by repetition or o m i s s i o n of p a r t s of the SP a n d g a m m a ray s i g n a t u r e s compared to a reference well that is believed to show an unfaulted section. Fault map trends and dip direction can also be defined by SCAT dipmeter analysis or on the stacked three-dimensional seiscrop sections. Generally, fault cuts have to be correlated from well to well to define the dip and curvature of the fault. Once these are estimated, fault contour maps can be generated by contouring the subsurface elevations of the fault cuts or, more directly, on the seismic workstation by stacking the seiscrop sections (Brown, 1986). The faults will offset the reference beds, and the amount of offset in section and map view must be estimated. Once the separation is known, a separation surface can be projected along the fault retaining the same trend, but adjusted in value by an amount appropriate for the offset on the fault (see Figure 6a).
Bed contours and fault contours have to be combined in a series of overlays to generate the structure map. Initially, individual fault blocks bounded on all sides by faults have to be contoured separately (Dickinson, 1954; Brown, 1986). The intersections b e t w e e n the bed a n d the fault c o n t o u r s of equivalent elevation value have to be identified to define the line of intersection of the bed and the fault. These lines are the fault cutoffs of the beds. There are two on each fault, one in the hanging wall and the other in the footwall. For extensional faults, there is a gap between the cutoffs where the key reference bed is omitted, and the gap in map view defines the heave across the fault.
Evaluating Structurally Complex Reservoirs
335
Map Restorations
Structure maps, like cross sections, can be tested for viability by restoring them to the undeformed state (Barr, 1985). Provided there are no areas of steep dip, a simple cutand-paste restoration may highlight areas in the map where faults have been drawn inaccurately. Figure 6b shows a restoration where most of the individual fault blocks can be restored in m a p view to close u p the fault gaps. The areas of overlap require additional r e d r a f t i n g or the collection of additional data.
Fault Sealing Characteristics
The m a p patterns of oil-water and gas-oil contacts are important features that define field geometry. Common or separate oil-water contacts and gas-oil contacts in separate fault blocks will define sealing and nonsealing faults. The sealing characteristics of faults can also be gauged from the mud weights used during drilling or the production data from the well after completion. These data can be plotted on the maps and sections. A special type of section is the fault plane section of Allan (1989), d r a w n within the plane of a single fault showing the positions of key reservoir and seal beds on either side of the fault and their contacts against one another across the fault (Figure 7). This type of section allows the interpreter to perceive top seal and cross fault seal potential and spill points and to identify undrilled prospective fault blocks. Fault plane sections may have to be drawn in many different sections, particularly where faults cross-cut or splay off one another (Allan, 1989; Downey, 1984; Smith, 1966,1980).
336
PART 7—GEOPHYSICAL METHODS
Figure 6. (a) Structure map and (b) (next page) restored structure map showing fault gaps removed. Remaining gaps and overlaps in the restored faults represent geometric incompatibilities in the interpretation. (From Galloway et al., 1983.)
Evaluating Structurally Complex Reservoirs
337
O
1500 SOOOft
(b)
0
4 0 0 8OOm
Figure 6 (continued)
338
PART 7—GEOPHYSICAL METHODS
Figure 7. Fault plane section and structure map of a model field to show the effects of synclinal and cross fault spilling, (a) Simple anticlinal closure cut by an extensional fault with two stacked reservoirs on both the downthrown and upthrown sides. Positions of cross fault spill points and synclinal spill points shown, (b) Fault plane section illustrating the synclinal and cross fault spill points. Reservoir beds are shown hatchured, whereas seal horizons are shown white. Note the effect of thick seal trapping across the fault. (From Allan, 1989.)
Statistics Overview
Т. С. Coburn
Marathon Oil Company Littleton, Colorado, U.S.A.
Brian R. Shaw
Battelle, Pacific Northwest Laboratories Richland, Washington, U.S.A.
INTRODUCTION
The p u r p o s e of statistics is to project or infer, from limited s a m p l e s , the character of a p o p u l a t i o n . In m o s t cases, particularly in oil and gas investigations, geological information is not derived from carefully designed sample schemes but, by design, represents anomalies. What successful company would drill on a regional trend as opposed to the top of a structure, on a bright spot, or at the crest of a reef? Statistical procedures p r e s u m e that sufficient data are randomly sampled from a population and that the average sample value approximates the population average. This is only possible if both high and low values are sampled without bias and enough samples are taken to stabilize the calculations. While proper sampling techniques are essential to formal statistical inference, geological samples are much too difficult or costly to obtain and cannot be discarded. Therefore, the robust testing of hypotheses and calculation of confidence intervals for statistical projections must be viewed in the restrictive light of geological d a t a . N o n e t h e l e s s , quantitative description and relationship inferences can be m a d e with the underlying awareness of the constraint of data quality.
It is also important to remember the effect of resolution and precision in analyzing quantitative geological data. J. C. Davis put it eloquently in his introduction to his classic text (Davis, 1986):
If you pursue the following topics, you will become involved with mathematical methods that have a certain aura of exactitude, that express relationships with apparent precision, and that are implemented on devices which have a popular reputation of infallibility.
While mathematical and statistical methods generate quantitative answers, one must always remain aware of the large disparity between geological samples and populations. Even in a producing field with "ample" well control, the creation of a structure m a p from well control represents the projection of a few 8-in. boreholes to h u n d r e d s of acres of surface area.
Even given this extreme difficulty, geological and statistical p r o c e d u r e s s h a r e the c o m m o n p r i n c i p l e of parsimony: the simplest explanation is superior to a complex solution to a problem. Recognition of this relationship can f o r m a basis for p r o p e r selection a n d application of the multitude of statistics available to the scientist.
Prior to beginning any statistical investigation, be sure to review any one of a n u m b e r of overview texts in geological
applications in the oil and gas industry, including Davis (1986), Harbaugh et al. (1977), and Krumbein and Graybill (1965).
CENTRAL TENDENCY
The simplest and most commonly overlooked statistical procedure is to plot the data. (Atkinson, 1985). Often a simple crossplot reveals the essential characteristics of a data set and allows for interpretation as well as proper selection of additional methods. In most cases, plotting of data reveals the n a t u r e of the data set and outliers or a n o m a l o u s data points to review for accuracy or measurement error and can indicate the s p r e a d or variability of the data. Eliminating measurement error is not uncommon even in commercial d a t a sets. For e x a m p l e , in a d a t a set c o m p o s e d of well information, if the kelly bushing is not k n o w n or uniformly subtracted from all wells, the resulting m a p will develop a severe case of volcanoes!
There are three measures of characterizing a population by describing the average value, or its central tendency. The most familiar measure is the arithmetic mean, which is simply the s u m of the values divided by their number. The mode is the value that occurs with the greatest frequency, and the median is the value that has as m a n y values above it as below it (Figure 1). As an e x a m p l e of c o m p a r i n g s o m e of the statistics discussed in previous chapters, consider the following values of porosity (in percent) that h a v e been measured on ten different sandstone samples: 15.1,16.5,18.8, 19.0, 22.0, 23.0, 25.0, 24.9, 31.9, and 43.0. Of the measures of central tendency, the arithmetic mean is the s u m of all these numbers divided in this case by 10, or 239.2 + 1 0 = 23.93. The median is 22.5 (halfway between 22.0 and 23.0), the value below which half the porosity values fall. The mid-range value is 29.05. The mode is the most frequently occurring value. Of the measures of dispersion, the range is computed to be 27.9, the variance is 61.79, and the standard deviation (the square root of the variance) is 7.86.
Although the mean, median, and mode convey the same general notion of centrality, their values are often different, as just demonstrated, because they represent different functions of the same data. Statistically, each has its strengths a n d weaknesses. Although it is sensitive to extreme values, the arithmetic m e a n is most generally used, partially because of c o n v e n t i o n a n d partially b e c a u s e of its c o m p u t a t i o n a l versatility in other statistical calculations.
The differences a m o n g these measures are a function of
339
340
PART 7—GEOPHYSICAL METHODS
Data Value
Figure 1. An asymmetrical data set. Thethreemeasuresof central tendency are different.
Figure 2. A symmetrical data set. The three measures of central tendency are identical.
the frequency distribution of the samples. The frequency distribution is nothing more than a plot of the values versus the number of times the value occurs, and it is often depicted as a histogram. Most values cluster around some central value, and the f r e q u e n c y of occurrence declines t o w a r d extreme values. There are several shapes of frequency distributions that commonly occur in nature. Data sets that are symmetrical about a central value develop the familiar "bell-shaped" normal distribution (Figure 2). Data sets that have numerous small values and a few large values develop an asymmetrical curve shape. Comparison of histograms plays a vital role in the study of various geological properties. For example, construction of a histogram might be used to d e t e r m i n e if a p a r t i c u l a r oil field exhibits a m u l t i m o d a l porosity distribution, indicating the presence of multiple lithologies. Another situation might involve a comparison of the d i s t r i b u t i o n s of p e t r o l e u m field sizes discovered worldwide in foreland and rift basins.
The three measures of central tendency are identical in symmetrical data sets (Figure 2) and are very different in asymmetrical data sets (Figure 1). This difference is crucial in arriving at essential estimates. For example, what is the most likely value for reserves for the next well we drill? If, as in most producing basins, there are a few huge fields and many subcommercial small fields, the most likely discovery is not the m e a n but the m o d e . D e t e r m i n i n g the s h a p e of the frequency distribution is critical to understanding which statistic to use. (For an excellent d i s c u s s i o n of the characteristics of petroleum data population distributions, see Harbaugh et al., 1977.)
Different geological properties and phenomena exhibit rather diverse distributions. For example, porosity is generally believed to be normally distributed, while permeability often tends to be lognormally distributed (that is, the l o g a r i t h m of p e r m e a b i l i t y t e n d s to be n o r m a l l y d i s t r i b u t e d ) . K n o w l e d g e of the g e n e r a l f o r m of the distribution is important to the selection of summary statistics because it helps prevent incorrect interpretations of the data. As a case in point, use of the arithmetic mean to represent average permeability is generally inappropriate because of the lognormality and high skewness of that property. Thus, the
geometric mean, which identifies the median of a lognormal distribution, is better suited to this situation. In geology, not all quantities of interest approximate a normal distribution, a n d for that reason, u n i f o r m u s e of a particular statistic simply as a matter of convenience should be avoided. Table 1 lists formulas that are commonly used to derive effective permeability.
There are two basic types of measured data: discrete and continuous variables. Discrete variables are measurements that can only be represented by counted values. For example, the number of limestone beds in a formation or the number of producing wells in a field can only be whole numbers. Continuous variables can have any value within the scale of measurement. Gamma ray log values, the porosity or permeability of a rock, or the subsea elevation of a formation are e x a m p l e s of c o n t i n u o u s variables. They can h a v e fractional values and can even have values the same as a previous sample.
VARIABILITY
Another characteristic of a frequency distribution curve is that it indicates the spread or dispersion of values about the measure of central tendency. This is commonly called the variance and is referenced to the mean, which indicates an assumption of a normal distribution. The variance can be regarded as the average squared deviation of the sample population:
n
where X = sample value x = sample mean n = number of samples
T h e standard deviation is a l s o u s e d to d e s c r i b e t h e dispersion about the mean, and it is simply the square root of
Statistics Overview
341
Table 1. Commonly Used Formulas To Derive Effective Permeability
Name Arithmetic mean
Formula
-
1
N
H
z
t
i-=4i
1
1
Application
Average of uniform, horizontal, parallel layers in linear flow. /(, and Iij are the permeability and thickness of layer /'. Ht is the total thickness.
Harmonic mean Geometric mean Radial flow
l t k = H ( Nh , V 1 Ki=lK<
(N Л l/N
№ к = Vi'=i
1/2
^ (^max ' ^min )
Average of uniform, horizontal, serial layers in linear flow. Used for vertical permeability estimates in shale-free sands.
Approximate average of an ensemble of uncorrelated random permeabilities in globally linear flow, /c,- is the permeability of each element in the ensemble
Radial inflow (well) permeability in homogeneous, anisotropic media. Zcmax and Zfmin are the major and minor axes permeabilities.
Cross bedding
кr= cos 2 a + si•n 2 а
к0
Zc90
Permeability in a direction at an angle a to cross bedding. Zc0 and Zc90 are the permeabilities parallel and perpendicular to cross bedding.
After Weber (1986).
the variance. This statistic gives a measure of the variation in units of the variable instead of in squared units. For example, the variance of data measured in feet would be square feet or area. The s t a n d a r d d e v i a t i o n is the s q u a r e root of this number, expressed in feet, which makes more sense for data measured in length.
HYPOTHESIS TESTING
Inference in statistics is derived from sample data and projected to populations by comparing sample statistics to the underlying population frequency distribution derived statistic. Hypothesis testing thus involves the relationship of a sample statistic to the theoretical population value. The importance of determining the frequency distribution can easily be seen.
S t a t i s t i c a l i n f e r e n c e is g e n e r a l l y p r e s e n t e d a s a true hypothesis to which a probability of not being true is assigned. The form of a statistical hypothesis is given both as the null or assumed event, HQ, a n d as the alternate event, H a .
H 0 : Ц = M0 H a : iL Ф IIq
A confidence level, or probability of the decision being in error, must be specified to determine whether or not to accept the null hypothesis (in this case, the null hypothesis assumes that the sample mean is the same as the population mean, p), or reject it in favor of the alternate hypothesis (which assumes that the means are different). Typical acceptable confidence
level values for this decision have been derived from population frequency distribution equations and usually are 0.005,0.01,0.05, and 0.1.
Generally, the test for this type of hypothesis is the f-test. While there are m a n y versions of h y p o t h e s e s that can be t e s t e d u s i n g this statistic, t h e g e n e r a l f o r m is t h a t of comparing a hypothetical value obtained from two samples from the same population, with the value calculated given the sample variance:
f_
X1-X2
^slX~12 +sx22
where
x i and x2 = sample m e a n s
and s ^ = sample variances
The hypothetical value can be obtained from generally available tables of t values, listed u n d e r the null hypothesis probability that w a s selected. If the t statistic exceeds the table value, then the null hypothesis, HQ, is rejected in favor of the alternate. A n o t h e r w a y of p h r a s i n g this result is that the means are statistically different.
This type of hypothesis testing assumes that the variance of the population being tested is the same as the test statistic— in this case, a Student's t distribution—which is normally distributed a n d has a m e a n of 0 a n d a variance of 1. If the variable under consideration does not have a known
342
PART 7—GEOPHYSICAL METHODS
population distribution of this character, it can be transformed to a standard normal distribution, which has a mean of 0 and a variance of 1. This is easily done by subtracting the mean from each observation and dividing by the sample standard deviation:
7 _ Xj-X S
where Z1 = ith transformed variable s = sample standard deviation
Confidence intervals around the population estimate also reflect the significance of a calculated statistic. A confidence interval is the range of possible values that contains the true
value of the population estimate with some specified level of confidence or probability. For example, the confidence interval about the m e a n of a normal distribution can be represented by
X_ t*sj^=r<H<X+—tj*=s
у Tl
л1п
where t* = Student's ^-distribution value n - number of samples s = sample standard deviation (j. = true population mean
The probability used for defining the f-distribution statistic also defines the range containing the true population value of the mean (|i).
Correlation and
Regression Analysis
O
J
T c Cobum
Mam"l0n 011 Company
Littleton, Colorado, U.S.A.
CORRELATION ANALYSIS
Correlation analysis, and its cousin, regression analysis, are well-known statistical approaches used in the study of relationships among multiple physical properties.. The investigation of p e r m e a b i l i t y - p o r o s i t y relationships is a typical example of the use of correlation in geology.
T h e t e r m correlation m o s t o f t e n r e f e r s to t h e l i n e a r association between two quantities or variables, that is, the tendency for one variable, x, to increase or decrease as the other, y, increases or decreases, in a straight-line trend or relationship (Draper and Smith, 1966; Snedecor and Cochran, 1967). The correlation coefficient (also called the Pearson correlation coefficient), r, is a dimensionless numerical index of the strength of that relationship. The sample value of r, which can range from -1 to +1, is computed using the following formula:
The simplest type of regression analysis involves fitting a straight line between two variables (Figure 1). In this case, one of the quantities is called the independent or predictor variable (usually denoted x), while the other is called the dependent or predicted variable (usually d e n o t e d y). This approach is often referred to as simple linear regression, or yon-x regression. It leads to the development of an empirical straight-line relationship between the two variables and has the following form:
у = ax + b
where a = slope b = !/-intercept in an (x, y) coordinate system
One feature of this line is that it always passes through the centroid of the data ( x, V/ ) defined by the two arithmetic averages. The formulas for the parameters a and b are
where the summation is made over the n sample values available and where
Xj = ith observation of the first variable y, = corresponding /th observation of the second
variable x, у = corresponding sample (arithmetic) means
Many useful types of correlation other than simple linear correlation exist. Correlation based on the ranks (numerical ordering) of measurements of two variables is an important alternative for simple linear correlation when the number of samples is small a n d / o r the joint distribution of the two variables is not simple. Multiple correlation and partial correlation are useful when studying relationships involving more than two variables. Cross correlation and autocorrelation are i m p o r t a n t to the analysis of repeated patterns observed in time and space, such as depth-related data recorded from geological stratigraphic sequences. Autocorrelation can also be used to measure the degree of similarity among porosity values measured at different locations.
Whereas correlation describes the linear association among variables, regression involves the prediction of one quantity from the others. Regression analysis is that broad class of statistics and statistical methods that comprises line, curve, and surface fitting, as well as other kinds of prediction and modeling techniques.
Z(*«-*)(y«-y)
a = —
, and b = y-ax
1(¾-*)2 i
There are two primary statistical applications for the regression line. It can be used as a sample-based estimate of an underlying linear functional relationship among the quantities of interest, or it can be employed as a predictive tool. To use the regression line for prediction, which is the most common usage in geology, simply substitute various values of x into the equation у = ax + b anc^ s o I y e f ° r the
corresponding predicted values of *.
By substituting the original x-values into the equation, it is possible to derive a measure of "goodness of fit" for the line. For every original x-value, a predicted value of у can be computed that differs from the corresponding original yvalue by a certain amount. This difference, y-y, is called the residual e (in the у-direction), or deviation from the regression line. T w o c o n s e q u e n c e s of the least s q u a r e s approach for the determination of the parameters a and b are that the sum of these residuals is zero and the sum of their squared values is as small as possible.
The sum of squares of the residuals is the basis for two measures of goodness of fit: the error variance s2 and the multiple correlation coefficient (also called the coefficient of determination). These two measures are given by the
343
344
PART 7—GEOPHYSICAL METHODS
• .
Sw
•
•
I . i
the following:
• Fitting through the origin, or forcing the regression line through any fixed point
• Regression in reverse, which consists of rewriting the regression of y-on-x as x = (y - b)/a
• Modeling nonlinear relationships, such as уA= a + bx + cxrJ
Various other functions of the x variable can be included in the previous relationship, such as polynomials and logarithms. Again, the regression parameters are determined so as to minimize the corresponding error variance.
Figure 1. Linear regression of x-on-y. Note the negative slope corresponding to a negative correlation. The regression line is determined so as to minimize the sum of squared deviations:
following formulas:
Х(У«"У«\ n-2
MULTIPLE AND MULTIVARIATE REGRESSION
The most important extension of the two-variable case is to situations involving more than two variables. When there is still one dependent variable but many predictor variables, the fitting technique is called multiple linear regression. When there are also more than one dependent variable, the approach is called multivariate regression (see the chapter on "Multivariate Data Analysis" in Part 6). The methods of simple bivariate regression extend directly to these multivariate situations. A typical geological application of multiple regression is the prediction of fold thickness from various geometric attributes, as given by the following equation:
Thickness = a + b (attitude) + с (tightness) + d (asymmetry)
R = 1 - -_ г
Z f c -yf
where n is the number of (x, y) pairs. The error variance s2 measures the scatter in the data about
the regression line. The square of the sample correlation coefficient, R2, takes values from 0 to 1 and can be interpreted as the fraction of the variation in the data explained by the regression line. The line is said to be a good fit of the data if S2 is small and R2 is simultaneously large.
Some i m p o r t a n t e x t e n s i o n s of t w o - v a r i a b l e linear regression analysis of particular interest to geologists include
Polynomial regression, sometimes referred to as curve fitting, is a special type of multiple regression frequently used in the earth sciences to model nonlinear relationships. Sphne fitting is another type of nonlinear curve fitting. A close relative to polynomial curve fitting is surface fitting, which has one or more spatial components as predictor variables.
The fitting of surfaces by least squares is an important component in most automated contouring software packages and is commonly used in computer generation of geological maps. Trend surface analysis, another mapping technique, is also based on the principles of least-squares fitting. Finally, some of the more specialized geostatistical techniques, such as kriging, are likewise rooted in the basic principles of least squares and multiple regression.
Multivariate Data Analysis
K. H. Esbensen
Norwegian Computing Center Oslo, Norway
A. G. Journel
Applied Earth Sciences Department Stanford University Stanford, California, U.SA.
INTRODUCTION
Most geological phenomena are multivariate in nature; for example, a p o r o u s m e d i u m is characterized by a set of interdependent quantities or attributes such as grain size, porosity, permeability, and saturation. Although univariate statistical analysis can characterize the distribution of each attribute separately, an understanding of porous media calls for unraveling the interrelationships among their various attributes. Multivariate statistical analysis proposes to study the joint d i s t r i b u t i o n of all a t t r i b u t e s , in w h i c h the distribution of any single variable is analyzed as a function of the other attributes distributions.
Multivariate observations are best organized and manipulated as a matrix of sample values, of size (n x P), where n is the number of samples and P is the number of attributes or variables. For example, a (5 x 3) matrix might represent five core samples at different depths on which f r e q u e n c i e s of o c c u r r e n c e of t h r e e d i f f e r e n t fossils are recorded. The purposes of multivariate data analysis is to study the relationships among the P attributes, classify the n collected samples into homogeneous groups, and make inferences about the underlying populations from the sample.
As in most statistical endeavors, multivariate data analysis is most efficient if backed by sound prior knowledge (such as geological interpretation) of the underlying phenomenon and a clear idea of project goals. There are many multivariate techniques, and many ways to apply each technique and interpret the results. It is the responsibility of the data analyst (geologist) to formulate the problem context, choose the appropriate multivariate technique, and apply it correctly.
The typical objectives of multivariate data analysis can be divided broadly into three categories.
1. Data description or exploratory data analysis (EDA)—The basic tools of this objective include univariate statistics, such as the mean, variance, and quantiles applied to each variable separately, and the covariance or correlation matrix between any two of the P quantities. Some of the P quantities can be transformed (for example, by taking the logarithm) prior to establishing the correlation matrix. Because the matrix is symmetrical, there are P(P -1)/2 potentially different correlation values.
2. Data grouping (discrimination and clustering)— Discrimination or classification aim at optimally assigning multivariate data vectors (arrays) into a set of
previously defined classes or groups (Everitt, 1974). Clustering, however, aims at defining classes of multivariate similarity and regrouping the initial sample values into these classes. Discrimination is a supervised act of pattern recognition, whereas clustering is an unsupervised act of pattern cognition (Miller and Kahn, 1962).
Principal component analysis (PCA) allows analysis of the covariance (correlation) matrix with a minimum of statistical assumptions. PCA aims at reducing the dimensionality P of the multivariate data set available by defining a limited number (fewer than P) of linear combinations of these quantities, with each combination reflecting some of the data structures (relationships) implicit in the original covariance matrix. 3. Regression—Regression is the generic term for relating two sets of variables. The first set, usually denoted by y, constitutes the dependent variables(s). It is related linearly to the second set, denoted x, called the independent variable(s). (For details of multiple and multivariate regression analysis, see the chapter on "Correlation and Regression Analysis" in Part 6.)
A NOTE ABOUT OUTLIERS
Multivariate data analysis techniques, particularly those relying on some minimization of square deviations, are sensitive to outlying data (values much larger or smaller than the corresponding mean). The data analyst should conduct ample checks for such influence. Once the outliers are detected, they should either be eliminated (if they originated through measurement error) or they should be classified into a separate statistical population. One of the easiest yet most effective ways to detect outliers is to look at the data in some suitable projection section of the multivariate "mathematical space," possibly using some of the techniques described later in this chapter, then iterating the analysis with the outlierscreened subset of the data.
A NOTE ABOUT DATA REPRESENTATIVENESS
As mentioned in the chapter on "Statistics Overview" (in Part 6), the available sample data are an incomplete image of the underlying population. Statistical features and relationships seen in the data may not be representative of
345
346
PART 7—GEOPHYSICAL METHODS
the characteristics of the u n d e r l y i n g population if there are biases in the data. Sources of biases are multiple, from the most obvious measurement biases to imposed spatial clustering. Spatial clustering results from preferential location of data (drilling patterns and core plugs), a prevalent problem in exploration and development. Such preferential selection of data points can severely bias one's image of the reservoir, usually in a nonconservative way. Remedies include defining representative subsets of the data, weighting the data, and careful interpretation of the data analysis results.
PRINCIPAL COMPONENT ANALYSIS
The main p u r p o s e of principal component analysis (PCA) is to identify the most significant variables in a multivariate problem. This is accomplished by reducing the dimensionality of the problem from P, the n u m b e r of original quantities Xv . . . xp, to fewer combinations of those quantities, called principal component variables: yv ..., yP{), P0 < P.
Often, the first f e w (P0 < P) first principal components t/• m a y account for a large fraction of the total variability of the o r i g i n a l P v a r i a b l e s Xi. T h e d e r i v a t i o n of t h e p r i n c i p a l components does not require any prior assumption on the distribution of the original variables. However, one should usually take care to standardize the measurement units of the X-S if they correspond to different attribute values. Tliis can be d o n e by d i v i d i n g each s a m p l e value Xi by the s t a n d a r d deviation of all measurements of its type.
Interpretation
A simple examination of these first principal components through their loadings я7 reveals a great deal about the total variance. If, for example, the first principal component y, explains a large proportion of the total variability, its loading factors au relate to the most "influential" variables Xi. These variables in their turn often allow a physical interpretation of yv B e w a r e , t h o u g h , t h a t t h e r a n k i n g of t h e p r i n c i p a l components y; is based on the variance correlation measure of variability. Alternative m e a s u r e s of variability such as the m e a n absolute deviation, as well as measures of nonlinear correlation, would lead to different decompositions. Again, the principal c o m p o n e n t s y} are best i n t e r p r e t e d against a prior k n o w l e d g e (such as geological interpretation) of the underlying phenomenon.
Filtering
T h e h i g h e r c o m p o n e n t s , yjr j > P0, u s u a l l y c o n t a i n significant "noise," whereas the first (lower) components carry the "signal" i n f o r m a t i o n of the initial P variables Xi. These higher components can be discarded if they are of no interest. In that case, the multivariate analysis is pursued in the lesser dimensional space of dimension P0 (there are fewer y's to deal with than original x's). However, in some cases, the noise m a y be of interest, such as w h e n it represents anomalies in a geochemical data analysis.
The relationship (matrix) defining the principal components i/• can be inverted, yielding the following inverse
linear relationship:
P = bOi + Z bijYj' for a11 *= - ' P
M
This inverse relationship can be used in estimating ( i n t e r p o l a t i n g ) a v a r i a b l e Xi f r o m p r i o r e s t i m a t e s of t h e principal components yr The first principal components y;, j < P0, can be estimated by s o m e type of regression p r o c e d u r e (such as kriging), while the higher components y;, j > P0, corresponding to random noise, can be estimated by their respective means.
Plotting the PCA Results
Many plots and scattergrams can be produced using the results of PCA and can be interpreted with the help of prior knowledge about the underlying phenomenon. Such plots may reveal clusters (grouping) or trends in the physical space or the P0 dimensional space of the first principal components. Interpretation of these clusters and classification of the initial data set (of size n x P) into more homogeneous subsets in the multivariate and/or spatial sense may then be in order. Data analysis can then be p u r s u e d within each of these n e w subsets.
DISCRIMINANT ANALYSIS (CLASSIFICATION)
Discriminant analysis (DA) a t t e m p t s to d e t e r m i n e a n allocation rule to classify multivariate data vectors into a set of p r e d e f i n e d classes, w i t h a m i n i m u m p r o b a b i l i t y of misclassification (Davis, 1986). Consider a set of n samples with P quantities being measured on each. Suppose that the n samples are divided into m classes or groups. Discriminant analysis consists of two steps:
1. The determination of w h a t makes each g r o u p different from the others. The answer may be that not all m predefined groups are significantly different from each other.
2. The definition of an allocation rule, usually taking the form of a "score" equal to a particular linear combination of the values of the P quantities.
Using this allocation rule, additional (new) samples can be classified into the predefined groups, and the corresponding probability of misclassification can be estimated (Figure 1).
D i s c r i m i n a n t a n a l y s i s r e q u i r e s the d e f i n i t i o n of a "distance" between any two groups. A widely used measure is the Mahalanobis distance (see Davis, 1986, for further details).
CLUSTER ANALYSIS
The p u r p o s e of cluster analysis (CA) is to define classes of samples with multivariate similarity (Hartigan, 1975). No prior assumption is needed about either the n u m b e r of these classes or their structures. Cluster analysis requires and
Mullivmiate Data Analysis
347
1= 1
2345
67
89
Rgure 2. Dendrogram (by aggregation). Starting from л samples, combine the two most similar samples (here 2 and 3). Then, combine the two nearest groups by either joining two samples or aggregating a third sample to the previous group of two (1 is aggregated to 2 and 3). At the next step, 4 and 5 constitutes a new group, which is then aggregated to the former group (1, 2,3). The aggregation process stops when there is only one group left In the last step, group (1,2,3, 4,5) is aggregated to group (6, 7,8, 9).
There is a large (and growing) variety of types of cluster
X,
analysis techniques (Hartigan, 1975):
Figure 1. Plot of two-bivariate distributions, showing overlap between groups A and B along both variables X1 and X2. Groups can be distinguished by projecting members of the two groups onto the discriminant function line. (From Davis, 1986.)
(unfortunately) often depends heavily on a prior choice of a distance measure between any two samples, (xiU i = 1,..., P) and (xu,, i = 1, . . . , P). Examples of distances include those represented by the following equations:
Generalized Euclidean distance
1 /Jfc
dw J] ™i(x,l ~ xUl
, with к > 0,Wj ь О,
where Wi = weight indicating the relative importance of each variable X1
Ccnrehition type distance
• Hierarchical techniques provide nested grouping as characterized by a dendrogram (Figure 2).
• Partitioning techniques define a set of mutually exclusive classes.
• Clumping or mixture techniques allow for classes that can overlap.
• Multimode search techniques look for zones in the Pdimensional space having relatively high concentrations of samples.
• Multinormal-related techniques capitalize on a prior hypothesis that each group is P-variate normally distributed.
• Certain other techniques are specific to binary variables.
The problem of preferential sampling in high pay zones, which may lead to more samples having high porosity and saturation values, is particularly critical when performing cluster analysis. If spatial declustering is not done properly before CA, all results can be mere artifacts of that preferential sampling. A related problem is linked to sample locations Uj and Uy not being accounted for in the definition of, say, the Euclidean distance between two samples I and I'.
2w
d„.~ - 1
dw = YixU-xHlY /=1
O n e c a n a l s o d e f i n e a d i s t a n c e du. b e t w e e n a n y t w o variables Xi and Xi by setting the previous summations over all n samples. Such distances between variables lead to definition of classes of variables having similar sample values. Such classes (clusters) of variables can help defining subsets of the P variables for further studies, with reduced dimensionality.
with x{l = Xj(Uj) and Xjj, = Xj(Ur) being the two measurements on variable Xj taken at the two locations и, and ur
In conclusion, although cluster analysis aims at an unsupervised classification, it is best w h e n applied with some supervision and a prior idea of what natural or physical clusters could be. Cluster analysis can then prove to be a remarkable corroboratory tool, allowing prior speculations to be checked and quantified.
Monte Carlo and Stochastic Simulation
MethodS
A G Journel
°чтЬт*
Stanford, California, U.S.A.
MONTE CARLO METHOD
The Monte Carlo technique consists of generating many different joint outcomes of random processes (Figure lc) and then observing the behavior of response values that are functions of these o u t c o m e s . Such behavior can be characterized by probability density functions (pdf) of the response variables, as depicted on the right of Figure 1(c) (Journel, 1989).
For example, the input variables might be porosity (ф), oil saturation (S0), and a binary indicator (I) set equal to 1 or 0 depending on whether the sample location belongs to a certain pay formation. The unique response value is the volume of oil in place defined by a particular function of the various input variables, called a transfer function (TF). In this example, the transfer function is a summation representing the total volume V.
In the ideal case of exhaustive sampling, the values used as input are known, so the response value is unique and deterministically known. However, in practice, the previous three input quantities (ф, S0, Г) are poorly sampled at best, and the unsampled values must be interpolated, as on the left of Figure 1(b). In addition, the transfer function itself, as in the case involving flow simulators, is usually a rough approximation of the actual process taking place in the reservoir. Consequently, the resulting response values (there are usually many) are uncertain estimates of the actual values.
If a critical decision is to be m a d e on these estimated response values, such as pursuing the development of a given project, it is essential to evaluate the uncertainty attached to these estimates or, better still, to evaluate the probability distributions of the response values, as shown on the right of Figure 1(c).
A Monte Carlo approach to evaluation of these response distributions consists of the following steps:
1. Model any aspect of uncertainty about either the input
variables or the parameters of the transfer function by use of the concept of random variables. For example, the joint spatial distribution of the three variables porosity, oil saturation, and indicator of formation presence can be modeled by three, usually interdependent, random functions.
2. Draw joint realizations (outcomes) of these random
variables or functions. Each realization represents an alternative equiprobable input set to the transfer function. Many such realizations could be retained, as on the left of Figure Ic.
3. Transfer the input uncertainty through the transfer
function into sets of response values. The histogram of the response values provides a probabilistic assessment of the impact of input uncertainty on that response, as shown on the right of Figure 1(c).
Step 1 consists of the determination and probabilistic modeling of the most consequential aspects (for the transfer function being considered) of the input uncertainty. Returning to the example of evaluation of oil volume in place, i m p o r t a n t e l e m e n t s of i n p u t u n c e r t a i n t y i n c l u d e the respective spatial continuity in space of the variables porosity, oil saturation, and formation boundaries; the spatial correlation of porosity and saturation; and whether these two variables d e p e n d on the proximity of the f o r m a t i o n boundaries. Indeed, the greater the spatial continuity of any variable, the smaller (on average) the estimation error when interpolating an unsampled value from neighboring data.
Step 2 consists of drawing realizations (outcomes) from a multivariate set of r a n d o m functions. These realizations reflect the statistical properties of the random function models (such as histograms and correlograms) in addition to honoring the actual data values at their locations.
Step 3 involves the repetitive application of the transfer function on each of the realizations of the input variables. If that transfer function is a simple, well-defined mathematical function (such as the equation defining the oil volume in place), step 3 poses no particular problem. However, if the transfer function involves a complex flow simulator, repetitive runs of such a transfer function can be tedious if not prohibitive in computer time. Various approximations are then possible, such as streamlining the transfer function, modeling the response distribution(s) from only a limited number of response values, or bounding type approaches in which only some quantiles of the probability distribution of the response values are determined.
The concept of M o n t e Carlo a n a l y s i s is g e n e r a l l y straightforward and this approach is often used. However, depending on the complexity of the problem, particularly the transfer function being considered, this approach can be quite difficult and time consuming (see Box and Draper, 1987). The small book by Hammersley and Handscomb (1964) provides a general discussion of Monte Carlo analysis.
STOCHASTIC SIMULATION
Stochastic simulation is a tool that allows Monte Carlo analysis of spatially distributed input variables. It aims at providing joint outcomes of any set of dependent r a n d o m variables. These random variables can be
• Discrete (indicating the presence or absence of a character), such as facies type
• Continuous, such as porosity or permeability values • Random sets, such as ellipses with a given distribution
of size and aspect ratio, or shapes d r a w n at random from a frequency table of recorded shapes
Monte Carlo and Stochastic Simulation Methods
349
Input
Trudi
(a)
Interpolation
(b)
Equiprobable Inputs
(С)
Transfer Function
•C5>
-CS)-
CED
Output (K response values)
У к , k=1,...,K
exact values
K estimates
Apdf
• Ук к=1 К
distributions
Figure 1. The Monte Carlo approach for transferring input uncertainty into a distribution of response values, (a) Ideal case—The input space and transfer function are perfectly known, resulting in the exact response values, (b) Traditional approach—The input values are interpolated from sparse data and the actual transfer function is estimated, resulting in estimated response values with usually no assessment of their uncertainty, (c) Monte Carlo approach—Input uncertainty is modeled by a series of equiprobable input sets which, after processing, provide a probability distribution (pdf) for the response value(s).
The set of r a n d o m variables can be any combination of these types. Some random variables can be functions of the geographic coordinates, in which case they are called random functions.
The dependence characteristics of these random variables are usually limited to 2-by-2, or bivariate, dependence, as opposed to k-by-k dependence involving к variables at a time, with к > 2. Bivariate dependence characteristics can be of the simpler linear correlation type, as defined by the correlation coefficient between any two variables. Theycan also be of the more complete type involving the bivariate probability distribution function.
The joint outcomes must verify any number of, or all of, the following typical conditions:
1. All outcomes are equally probable, which does not imply that they are all similar. In particular, one realization can be quite similar to some and still different from others.
2. The histogram of the simulated values of any one attribute reasonably reproduces the probability density function of the corresponding random variable model.
3. The dependence characteristics between any two random variables are reproduced by the corresponding realizations. These two random variables can relate to the same attribute at two different locations in space or to two different attributes at either the same location or at two different locations.
4. All outcomes honor the sample data values.
In a spatial context where all the random variables relate to the same attribute at different locations, condition 4 amounts to honoring the sample values at their locations. This condition is also known as the exactitude condition, and the corresponding realizations are referred as being conditional to the data values.
There are as many algorithms for generating joint realizations of a large number of dependent random variables as there are different models for the joint distribution of these r a n d o m v a r i a b l e s , w i t h a n e q u a l l y large n u m b e r of implementation variants. With the advent of extremely fast computers with vast memory, the field is exploding with new algorithms being proposed regularly. The book by Ripley (1987) gives an excellent s u m m a r y and an attempt at classification of the algorithms, yet as of 1990, it can no longer be c o n s i d e r e d complete. A g o o d generic discussion of simulation topics is given in Hohn (1988).
Particular mention should be given to stochastic simulations based on self repetitive fractal models (Hewett, 1986). Such m o d e l s c o r r e s p o n d to p a t t e r n s of spatial variability that repeat themselves whatever the distance scale used.
The present ability to generate a large n u m b e r of very large stochastic simulations very quickly far outstrips the capability to look at the corresponding (stochastic) images and the capability to process them with realistic flow simulators. The bottleneck for systematic utilization of the Monte Carlo approach is no longer stochastic simulation but rather computer graphics and flow simulators that are presently much too slow.
350
PART 7—GEOPHYSICAL METHODS
Part 6 References Cited
Aguilera, R., 1980, Naturally fractured reservoirs: Tulsa, OK, PennWell Books, 703 p.
Ahlbrandt, T. S., and S. G. Fryberger, 1982, Introduction to eolian deposits, in P. A. Scholle and D. Spearing, eds., Sandstone Depositional Environments: AAPG Memoir 31, p.11-47.
Allan, V. S. 1989, Model for hydrocarbon migration and entrapment within faulted structures: AAPG Bulletin, v. 73, p. 803-811.
Alpay, O. A., 1972, A practical approach to defining reservoir heterogeneity: Journal of Petroleum Technology, 24, p. 841848.
Archie, G. E., 1952, Classification of carbonate reservoir rocks and petrophysical considerations: AAPG Bulletin, v. 36, p. 278-298.
Asquith, G., and C. Gibson, 1982, Basic well log analysis for geologists: AAPG Methods in Exploration Series, 216 p.
Atkinson, A. C., 1985, Plots, transformations, and regression: Oxford, U.K., Oxford Press, 282 p.
Atkinson, C. D., B. G. Goesten, A. Speksnijder, and W. van der Vlugt, 1986, Storm-generated sandstone in the Miocene Miri Formation, Seria Field, Brunei (N.W. Borneo), in R. J. Knight, and J. R. McLean, eds., Shelf Sands and Sandstones: Canadian Society of Petroleum Geologists Memoir 11, p , 213-240.
Barr, D., 1985,3D palinspastic restoration of normal faults in the Inner Moray Firth—implications for extensional basin development: Earth and Planetary Science Letters, v. 75, p. 191-203.
Barwis, J. H., 1989, The explorationist and shelf sand models—where do we go from here?: 7th Annual Research Conference Proceedings, Gulf Coast SEPM, p. 1-14.
Beaumont, E. A., and N. H Foster, 1987, Reservoirs II— Sandstones: AAPG Treatise of Petroleum Geology Reprint Series, n. 4,573 p.
Bebout, D. G., R. G. Loucks, and A. R. Gregory, 1978, Frio Sandstone reservoirs in the deep subsurface along the Texas Gulf coast—their potential for production of geopressured geothermal energy: The Univ. of Texas Bureau of Economic Geology Report of Investigations, n. 91,93 p.
Bengtson, C. A., 1982, Structural and stratigraphic uses of dip profiles, in M. T Halbouty, ed., Deliberate Search for the Subtle Trap: AAPG Memoir 32, p. 31-45.
Berg, R. R., 1986, Reservoir Sandstones: Englewood Cliffs, N.J., Prentice-Hall, 481 p.
Box, G. E. P., and N. R. Draper, 1987, Empirical model building and response surfaces: New York, John Wiley, 669 p.
Braithwaite, С. I. M., J. D. Marshall, and Т. C. Holland, 1989, Improving recovery from the Dunlin field, U.K. Northern North Sea, in Formation Evaluation and Reservoir Geology: 64th Annual Technical Conference of the Society of Petroleum Engineers, San Antonio, TX, Oct. 8-11, SPE 19878,18 p.
Brown, A. R., 1986 Interpretation of three-dimensional seismic data: AAPG Memoir 42,194 p.
Cant, D. J., 1984, Subsurface facies analysis, in R. G. Walker, ed., Facies Models: Geoscience Canada, Reprint Series 1, p. 297-319.
Connolly, E. T., and P. A. Reed, 1983, Full spectrum formation evaluation: Canadian Well Logging Society Journal, v. 12, p. 23-69.
Craig, J. T., Jr., and В. V. Randall, 1976, Directional survey calculation: Petroleum Engineer, March, p. 38-54.
Dahlberg, E. C., 1982, Applied Hydrodynamics in Petroleum Exploration: New Yorky Springer Verlag, 161 p.
Dahlstrom, C. D. A., 1970, Structural geology in the eastern margin of the Canadian Rocky Mountains: Canadian Society of Petroleum Geologists Bulletin, v. 18, p. 332-406.
Dailey, P., 1977, A guide to accurate wellbore survey calculations: Drilling-DCW, May, p. 52-59 and 118-119.
Davis, J. C., 1986, Statistics and data analysis in geology: New York, John Wiley, 646 p.
DePaor, D., 1988, Balanced section in thrust belts, Part 1— construction: AAPG Bulletin, v. 72, p. 73-90.
Dickinson, G., 1954, Subsurface interpretation of intersecting faults and their effect upon stratigraphic horizons. AAPG Bulletin, v. 38, n. 5, p. 854-877.
Downey, M. W., 1984, Evaluating seals for hydrocarbon accumulations: AAPG Bulletin, v. 68, p. 1752-1763.
Draper, N. R., and H. Smith, 1966, Applied regression analysis, 2nd ed.: New York, Jolin Wiley, 709 p.
Dunham, R. J., 1962, Classification of carbonate rocks according to depositional texture, in, Classifications of Carbonate Rocks—A Symposium: AAPG Memoir 1, p. 108-121.
Dunnington, H. V., 1967, Aspects of diagenesis and shape change in stylolitic limestone reservoirs: 7th World Petroleum Congress Proceedings, v. 2, p. 339-352.
Ebanks, W. ]., Jr., 1987, Flow unit concept—integrated approach to reservoir description for engineering projects, abst.: AAPG Bulletin, v. 71, n. 5, p. 551-552.
Everitt, B., 1974, Cluster analysis: London, Heinemann Educational Books Ltd., 122 p.
Finley, R.}., W. L. Fisher, S. J. Seni, S. C. Ruppel, W. G. White, W. B. Ayers, Jr., S. P. Dulton, M. L. W. Jackson, N. Banta, V. A. Kuusliraa, K S. McFall, M. Godec, and Т. V. Jennings, 1988, The University of Texas Bureau of Economic Geology Report of Investigations: Austin, TX, The University of Texas Bureau of Economic Geology, v. 179, 69 p.
Gaida, K. H., D. G. Kessel, H. Volz, and W. Zimmerle, 1987, Geologic parameters of reservoir sandstones as applied to enhanced oil recovery: SPE Formation Evaluation, v. 2, p. 89-96.
Galloway, W. E., and E. S. S. Cheng, 1985, Reservoir facies architecture in a microtidal barrier system—Frio Formation, Texas Gulf Coast: The University of Texas Bureau of Economic Geology, Report of Investigations, v. 144, 36 p.
Galloway, W. E., and D. K Hobday, 1983, Terrigenous Clastic Depositional Systems Applications to Petroleum, Coal, and Uranium Exploration: New York, Springer Verlag, 423 p.
Garb, F. A., and G. L. Smith, 1987, Estimation of oil and gas reserves, in H. B. Bradley, ed., Petroleum Engineering Handbook: Richardson, TX, Society of Petroleum Engineers, p. 4-1-40-38.
Gaynor, G. C., and M. H. Scheihing, 1988, Shelf depositional environments and reservoir characteristics of the Kuparuk River Formation (Lower Cretaceous), Kuparuk field, North Slope, Alaska, in A. J. Lomando, and P. M. Harris, eds.,
References Cited
351
Giant oil and gas fields—A core workshop: Society of Economic Paleontologists and Mineralogists Core Workshop 12, p. 333-389. Gerard, R. E., C. A. Philipson, F. M. Ballentine, and D. M. Marschall, in press, Petrographic image analysis—an alternate method for determining petrophysical properties, in I. Palaz, and S. Sengupta, eds., Automated Pattern Analysis in Petroleum Exploration: New York, SpringerVerlag.
Gibbs, A. D., 1983, Balanced cross section construction from seismic sections in areas of extensional tectonics: Journal of Structural Geology, v. 5, p. 153-160.
Groshong, R. H., 1989a, Half graben structures—balanced models of extensional fault bend folds: Geological Society of America Bulletin, v. 101, p. 96-105.
Groshong, R. H., 1989b, Structural style and balanced cross sections in extensional terranes: Houston Geological Society Short Course Notes, Feb. 24-25,128 p.
Guevara, E. H., 1988, Geological characterization of Permian submarine fan reservoirs of the Driver Waterflood Unit, Spraberry Trend, Midland Basin, Texas: The Univ. of Texas Bureau of Economic Geology Report of Investigations, n. 172,44 pp.
Haldorsen, H. H., and L. W. Lake, 1984, A new approach to shale management in field-scale models: Society of Petroleum Engineers Journal, Aug., p. 447-452.
Hamblin, W. K., 1965, Origin of "reverse drag" on the downthrown side of normal faults: Geological Society of America Bulletin, v. 76, p. 1145-1164.
Hammersley, J. M., and D. C. Handscomb, 1964, Monte Carlo methods: Chapman and Hall (reprint 1983), 178 p.
Harbaugh, J. W., J. H. Doveton, and J. C. Davis, 1977, Probability methods in oil exploration: New York, John Wiley, 269 p.
Harris, D. G., 1975, The roles of geology in reservoir simulation studies: Journal Petroleum of Technology, May, p. 625-632.
Hartigan, J. A., 1975, Clustering algorithms: New York, John Wiley, 351 p.
Hartman, J. A., and D. D. Paynter, 1979, Drainage anomalies in Gulf Coast Tertiary sandstones: Journal of Petroleum Technology, Oct., p. 1313-1322.
Hearn, C. L., W. J. Ebanks, Jr., R. S. Туе, and V. Ranganathan, 1984, Geological factors influencing reservoir performance of the Hartzog Draw field: Journal of Petroleum Technology, v. 36, Aug., p. 1335-1344.
Hewett, T. A., 1986, Fractal distribution of reservoir heterogeneity and their influence on fluid transport: Society of Petroleum Engineers, SPE Paper 15386.
Hietala, R. W., and E. T. Connolly, 1984, Integrated rock-log calibration in the Elmworth field, Alberta, Canada, Part II—well log analysis methods and techniques, in J. A. Masters, ed., Elmworth—Case Study of a Deep Basin Gas Field: AAPG Memoir 38, p. 215-242.
Hohn, M. E., 1988, Geostatistics and petroleum geology: New York, Van Nostrand Reinhold, 264 p.
Hriskevich, M. E., J. M. Faber, and J. R. Langton, 1980, Strachan and Ricinus West gas fields, Alberta, Canada, in M. T. Halbouty, ed., Giant Oil and Gas Fields of the Decade 1968-1978: AAPG Memoir 30, p. 315-327.
Hubbert, M. K., 1953, Entrapment of petroleum under hydrodynamic conditions: AAPG Bulletin, v. 37, p. 19542026.
Inglis, T. A., 1987, Directional Drilling, Petroleum Engineering and Development Studies, Volume 2: London, Graham and Trotman, chap. 9, p. 155-171.
Jardine, D., and J. W. Wilshart, 1987, Carbonate reservoir description, in Tillman, R. W., and K. J. Weber, eds. Reservoir Sedimentology: SEPM Special Publication 40, p. 129-152.
Jones, S. C., 1972, A rapid accurate unsteady-state Klinkenberg permeameter: Society of Petroleum Engineers Journal, v. 12, n. 5, p.383-397.
Jones, T. A., D. E. Hamilton, and C. R. Johnson, 1986, Contouring geologic surfaces with the computer: New York, Van Nostrand Reinhold, 320 p.
Jordan, D. W., and W. A. Pryor, 1992, Hierarcliical levels of heterogeneity in a Mississippi River meander belt and application to reservoir systems: AAPG Bulletin, v. 76, p. 1601-1624.
Journel, A. G., 1989, Fundamentals of geostatistics in five lessons: Washington D.C., American Geophysical Union, Short Course in Geology, v. 8,40 p.
Katz, et al., 1957, Handbook of Natural Gas Engineering: New York, McGraw-Hill, 802 p.
Keelan, D. K., 1972, A critical review of core analysis techniques: Journal of Canadian Petroleum Technology, v.2, p. 42-55.
Kerans, C., 1989, Karst-controlled reservoir heterogeneity and an example from the Ellenburger Group (Lower Ordovician) of west Texas: The Univ. of Texas Bureau of Economic Geology Report of Invetigations, n. 186,40 p.
Kimmel, J. D., and R. N. Dalati, 1987, Potential tests of oil wells, in H. B. Bradley, ed., Petroleum Engineering Handbook: Richardson, TX, Society of Petroleum Engineers, p. 32-1-32-16.
Krause, F. F., H. N. Collins, D. A. Nelson, S. D. Mochemer, and P. R. French, 1987, Multiscale anatomy of a reservoir— geological characterization of Pembina-Cardium pool, west-central Alberta, Canada: AAPG Bulletin, v. 71, p. 1233-2260.
Krumbein, W. C., and F. A. Graybill, 1965, An introduction to statistical models in geology: New York, McGraw-Hill, 475 p.
Kulander, B. R., and S. L. Dean, 1985, Hackle plume geometry and joint propagation dynamics, in O. Stephansson, ed., Fundamentals of Rock Joints: Proceedings of the International Symposium, Bjorkliden, Sept. 15-20, p. 85-94.
Lammerson, P. R., 1982, The Fossil basin and its relationship to the Absaroka thrust system, Wyoming and Utah, in R. B. Powers, ed., Geological Studies of the Cordflleran Tlirust Belt: Rocky Mountain Association of Geologists, p. 279340.
Landes, K. L., 1951, Subsurface maps and sections, in Petroleum Geology: New York, John Wiley, 660 p.
Langstaff, C. S., and Morrill, D., 1981, Geologic Cross Sections: Boston, MA, IHRDC, 108 p.
Langston, J. R., and Chin, G. E., 1968, Rainbow Member facies and related reservoir properties, Rainbow Lake, Alberta: AAPG Bulletin, v. 52, n. 10, p. 1925-1955.
La Pointe, P. R., and J. A. Hudson, 1985, Characterization and interpretation of rock mass joint patterns: Geological Society of America Special Paper 199,37 p.
Lassiter, T. K., J. R. Waggoner, and L. W. Lake, 1986, Reservoir heterogeneities and their influence on ultimate recovery, in L. W. Lake, and N. B. Carroll, Jr., eds., Reservoir Characterization: Orlando, FL, Academy Press, p. 545-560.
Levandowski, D. W., M. E. Kaley, S. R. Silverman, and R. G. Smalley, 1973, Cementation in Lyons Sandstone and its role in oil accumulation, Denver basin, Colorado: AAPG Bulletin, v. 57, p. 2217-2244.
Lindquist, S. J., 1983, Nugget Formation reservoir characteristics affecting production the Overthrust Belt of
352
PART 7—GEOPHYSICAL METHODS
southwestern Wyoming: Journal of Petroleum Technology, July, p. 1355-1365. Longman, M. W., T. G. Fertal, and J. S. Glennie, 1983, Origin and geometry of Red River Dolomite reservoirs, western Williston basin: AAPG Bulletin, v. 67, p. 744-771. Lucia, F. J., 1983, Petrophysical parameters estimated from visual descriptions of carbonate rocks—a field classification of carbonate pore space: Journal of Petroleum Technology, March, p. 629-637.
Lupe, R., and T. S. Ahlbrandt, 1979, Sediments of the ancient eolian environment—reservoir inhomogeneity, in E.D. McKee, ed., A Study of Global Sand Seas: U.S. Geological Survey Professional Paper 1052, p. 241-252.
Major, R. P., and Holtz, M. H., 1989, Effects of geologic heterogeneity on waterflood efficiency at Jordan field, University Lands, Ector and Crane counties, Texas: 64th Annual Technology Conference of the Society of Petroleum Engineers, San Antonio, TX, Oct., SPE 19874, p. 633-640.
Masters, J.A., 1984, Elmworth—Case Study of a Deep Basin Gas Field: AAPG Memoir 38,316 p.
Maxey, G. B., 1964, Hydrostratigraphic units: Journal of Hydrology, v. 2, p. 124-129.
McBride, E. F., 1984, Compaction in sandstones—influence on reservoir quality: AAPG Bulletin, v. 68, p. 505.
McCoss, A. M., 1987, Practical section drawing through folded layers using sequentially rotated cubic interpolators: Journal of Structural Geology, v. 9, p. 365370.
McDonald, D. A., and R. C. Surdam, eds., 1984, Clastic Diagenesis: AAPG Memoir 37,434 p.
Mcllreath, I. A., and N. P. James, 1984, Carbonate slopes, in R. G. Walker, ed., Facies Models: Geoscience Canada Reprint Series 1, p. 245-258.
Miall, A. D., 1984, Principles of Sedimentary Basin Analysis: New York, Springer-Verlag, 490 p.
Miller, R. L., and J. S. Kahn, 1962, Statistical analysis in the geological sciences: New York, John Wiley, 481 p.
Mitra, S., 1986, Duplex structures and imbricate thrust systems—geometry, structural position, and hydrocarbon potential: AAPG Bulletin, v. 70, p. 1087-1112.
Morgan, J. T., F. S. Cordiner, and A. R. Livingston, 1977, Tensleep reservoir study, Oregon Basin field, Wyoming— reservoir characteristics: Journal of Petroleum Technology, v. 29, p. 886-896.
Morton, R. A., Ewing, T. E., and Tyler, N., 1983, Continuity and internal properties of Gulf Coast sandstones and their implications for geopressured fluid production: The University of Texas Bureau of Economic Geology Report of Investigations, n 123,70 pp.
Moslow, T. F., and R. W. Tillman, 1986, Sedimentary facies and reservoir characteristics of Frontier Formation sandstones, southwestern Wyoming, in C. W. Spencer, and R. F. Mast, eds., Low Permeability Sandstone Reservoirs: Studies in Geology Series 24, p. 271-295.
Narr, W., and I. Lerche, 1984, A method for estimating subsurface fracture density in core: AAPG Bulletin, v. 68, p. 637-648.
Nelson, R. A., 1985, Geologic analysis of naturally fractured reservoirs—Contributions in Petroleum Geology and Engineering #1: Houston, TX, Gulf Publishing Company, 320 p.
Nolen-Hoeksema, R. C., and J. H. Howard, 1987, Estimating drilling direction for optimum production in a fractured reservoir: AAPG Bulletin, v. 71, p. 958-966.
Office of Technology Assessment, 1985, U.S. natural gas availability—gas supply through the year 2000: U.S.
Congress, Office of Technology Assessment, OTA-E-245, 252 p. Pettijohn, F. J., P. E. Potter, and R. Siever, 1973, Sand and Sandstone: New York, Springer-Verlag, 618 p. Phipps, G. G., 1989, Exploring for dolomitized Slave Point carbonates in northeastern British Columbia: Geophysics, v. 54, n. 7, p. 806-814. Pitman, J. K., and E. S. Sprunt, 1986, Origin and distribution of fractures in Lower Tertiary and Upper Cretaceous rocks, Piceance basin, Colorado, and their relation to the occurrence of hydrocarbons. AAPG Studies in Geology, n. 24, p. 221-234.
Plumb, R. A., and S. M. Luthi, 1986, Application of borehole images to geologic modeling of an eolian reservoir: 61st Annual Technical Conference of the Society of Petroleum Engineers, New Orleans, LA, Oct. 5-8, SPE 15487,11.p.
Porter, K. W., and R. J. Weimer, 1982, Diagenetic sequence related to structural history and petroleum accumulation: Spindle field, Colorado: AAPG Bulletin, v. 66, p. 2543-2560.
Poston, S. W., 1987, Development plan for oil and gas reservoirs, in H. B. Bradley, ed., Petroleum Engineering Handbook: Richardson, TX, Society of Petroleum Engineers, p. 36-1-36-11.
Qiu Yinan, 1984, Depositional model, heterogeneous characteristics, and waterflood performance of sandstone reservoirs in a lake basin case study of oilfields, eastern China: Proceedings of the I l t h World Petroleum Congress, v.3, p. 113-125.
Ramsay, J. G., 1967, Folding and fracturing of rocks: N e w York and London, McGraw-Hill, 568 p.
Ramsay, J. G., and M. I. Huber, 1987, The techniques of modern structural geology_Volume 2, in Folds and Fractures: Orlando, FL, Academic Press, 700 p.
Reading, H. G., ed., 1986, Sedimentary Environments and Facies, 2nd ed.: Boston, MA, Blackwell Scientific Publications, 615 p.
Reiss, L. H., 1980, The reservoir engineering aspects of fractured formations: Houston, TX, Gulf Publishing Company, 108 p.
Renick, H., Jr., and Gunn, R. D., 1989, Triangle Ranch Headquarters field development using shallow core holes and high-resolution seismic data: Geophysics, v. 54, n. 11, p. 1384-1396.
Rider, M. H., 1986, Geological interpretation of well logs: New York, John Wiley, 175 p.
Ripley, B. D., 1987, Stochastic Simulation: New York, John Wiley, 237 p.
Roehl, P. O., 1967, Stony Mountain (Ordovician) and Interlake (Silurian) facies analogs of recent low-energy marine and subaerial carbonates, Bahamas: AAPG Bulletin, v. 51, n. 10, p. 1979-2032.
Rose, P. R., J. R. Everett, and I. A. Merin, 1986, Potential basincentered gas accumulation in Cretaceous Trinidad Sandstone, Raton Basin, Colorado: AAPG Studies in Geology, n. 24, p. 111-128.
Rowan, M. G., and R. Kligfield, 1989, Cross section restoration and balancing as aid to seismic interpretation in extensional terranes: AAPG Bulletin, v. 73, p. 955-966.
Scheihing, M. H., and G. C. Gaynor, 1988, Reservoir heterogeneities of storm-generated shelf sandstone bodies—implications for reservoir management, abst.: AAPG Bulletin, v. 72, n. 2, p. 244.
Schmidt, V., and D. A. McDonald, 1980, Secondary Reservoir Porosity in the Course of Sandstone Diagenesis: AAPG Continuing Education Course Note Series No.12,125 p.
Scholle, P. A., and P. R. Schluger, eds., 1979, Aspects of Diagenesis: SEPM Special Publication, n. 26,443 p.
References Cited
353
Schultz-Ela7 D., and Duncan, K., 1990, Users manual and software for Restore, version 2.0: The Univ. of Texas Bureau of Economic Geology, 75 p.
Siemers, С. T., and Tillman, R. W., 1981, Recommendations for the proper handling of cores and sedimentological analysis of core sequences, in C. T. Siemers, R. W. Tillman, and C. R. Williamson, eds., Deep-Water Clastic Sediments—A Core Workshop: SEPM Core Workshop, n. 2, p. 20-44.
Slatt, R. M., 1986, Exploration models for submarine slope sandstones: Transactions of the 36th Annual Meeting of the Gulf Coast Association of Geological Societies, Continental Slope—Frontier of the 80's, p. 295-304.
Slatt, R. M., and G. L. Hopkins, 1990, Scaling of geological reservoir description to engineering needs: Journal of Petroleum Technology, Feb., p. 202-210.
Slatt, R. M., S. Phillips, J. M. Boak, and M. B. Lagoe, 1993, Scales of geological heterogeneity of a deep-water sand giant oil field, Long Beach unit, Wilmington field, California, in E. G. Rhodes, and T. F. Moslow, eds., Marine Clastic Reservoirs—Examples and Analogs: New York, Springer-Verlag.
Smith, D. A., 1966, Theoretical considerations of sealing and nonsealing faults: AAPG Bulletin, v. 50, p. 363-374.
Smith, D. A., 1980, Sealing and nonsealing faults in Louisiana Gulf Coast salt basin: AAPG Bulletin, v. 64, p. 145-172.
Snedden, J. W., 1987, Validity of the use of the spontaneous potential curve shape in the interpretation of sandstone depositional environments, in B. R. White, and R. Kier, eds., Transactions of the 34th annual meeting of the Gulf Coast Association of Geological Societies and 31st annual meeting of the Gulf Coast Section of SEPM, v. 34, p. 255263.
Snedecor, G. W., and W. G. Cochran, 1967, Statistical methods, 6th ed.: Ames, Iowa State Univ. Press, 593 p.
Sneider, R. M., 1987, Practical petrophysics for exploration and development: Houston, Robert M. Sneider Exploration, Inc., AAPG Course Notes, 300 p.
Sneider, R. M., and H. R. King, 1984, Integrated rock-log calibration in the Elmworth field, Alberta, Canada—Part I, Reservoir rock detection and characterization, in J. A. Masters, ed., Elmworth—Case Study of a Deep Basin Gas Field, AAPG Memoir 38, p. 205-214.
Sneider, R. M., F. H. Richardson, D. D. Paynter, R. R. Eddy, and I. A. Wyant, 1977, Predicting reservoir rock geometry and continuity in Pennsylvanian reservoirs, Elk City field, Oklahoma: Journal of Petroleum Technology, July, p. 851866.
Sneider, R. M., C. N. Tinker, and L. D. Meckel, 1978, Deltaic environment reservoir types and their characteristics: Journal of Petroleum Technology, Nov., p. 1538-1546.
Sneider, R. M., H. R. King, H. E. Hawkes, and Т. B. Davis, 1983, Methods for detection and characterization of reservoir rock, Deep Basin gas area, western Canada: Journal of Petroleum Technology, Sept., p. 1725-1734.
Spearing, D. R., 1974, Summary sheets of sedimentary deposits: Geological Society of American Publication MC-8.
Spencer, C. W., and R. F. Mast, 1986, Introduction, in C. W. Spencer, and R. W. Mast, eds., Low Permeability Sandstone Reservoirs: AAPG Studies in Geology Series, n. 24, p. iv-vi.
Stalkup, F. I., 1986, PermeabiHty variations observed at the faces of crossbedded sandstone outcrops, in L. W. Lake, and H. B. CarroU, Jr., eds., Reservoir Characterization: Orlando, FL, Academy Press, p. 141-180.
Stearns, D W., and M. Friedman, 1972, Reservoirs in fractured rock: AAPG Memoir 16, p. 82-100.
Stewart, R. R., 1987, Seismic tomography: SEG Continuing Education Program Notes, GENIX Tech. Ltd.
Suppe, J., 1983, Geometry and kinematics of fault-bend folding: American Journal of Science, v. 283, p. 684-721.
Swanson, R. G., 1981, Sample examination manual: AAPG Methods in Exploration Series, 65 p.
Tillman, R. W., and D. W. Jordan, 1987, Sedimentology and subsurface geology of deltaic forces, Admire GSO Sandstone, El Dorado field, Kansas, in R. W. TiUman, and K. J. Weber, eds., Reservoir Sedimentology: SEPM Special Publication 40, p. 221-292.
Tillman, R. W., and R. S. Martinsen, 1987, Sedimentologic model and production characteristics of Hartzog Draw field, Wyoming, a Shannon shelf-ridge sandstone, in R. W. Tillman and K. J. Weber, eds., Reservoir Sedimentology: SEPM Special Publication 40, p. 15-112.
Tucker, P. M., 1988, Seismic contouring—a unique skill: Geophysics, v. 53, n. 6, p. 741-749.
Tyler, N., and W. A. Ambrose, 1985, Facies architecture and production characteristics of strandplain reservoirs in the Frio Formation, Texas: The Univ. of Texas Bureau of Economic Geology Report of Investigations, n 146,42 p.
Tyler, N., and J. C. Gholston, 1988, Heterogeneous deep-sea fan reservoirs, Shakelford and Preston waterflood units, Spraberry Trend, West Texas: The Univ. of Texas Bureau of Economic Geology Report of Investigations, n. 171, 38 p.
van de Graaff, W. J. E., and P. S. Ealey 1989, Geological modeHng for simulation studies: AAPG Bulletin, v. 73, p. 1436-1444.
van Golf-Racht, T. D., 1982, Fundamentals of fractured reservoir engineering—Developments in Petroleum Science No. 12: Amsterdam, Elsevier Scientific Publishing Co., 710 p.
Verrall, P., 1982, Structural interpretation with appHcations to North Sea problems: Geological Society of London.Course Notes No 3, JAPEC (UK).
Walker, R. G., ed., 1984, Facies Models, 2nd ed.: Geoscience Canada Reprint Series 1,317 p.
Walker, R. G., and Cant, D. J., 1984, Sandy fluvial systems, in R. G. Walker, ed., Facies Models: Geosciences Canada Preprint Series 1, p. 71-84.
Wardlaw, N. C., 1976, Pore geometry of carbonate rocks as revealed by pore casts and capillary pressure: AAPG BuHetin, v. 60, p. 245-257.
Wardlaw, N. C., and R. P. Taylor, 1976, Mercury capillary pressure curves and the interpretation of pore structure and capillary behavior in reservoir rocks: Bulletin of Canadian Petroleum Geology, v. 24, p. 225-262.
Watts, N. L., 1987, Theoretical aspects of cap-rock and fault seals for single- and two-phase hydrocarbon columns: Marine and Petroleum Geology, v. 4, p. 274 - 307.
Weber, K. J., 1982, Influence of common sedimentary structures on fluid flow in reservoir models: Journal of Petroleum Technology, March, p. 665-772.
Weber, K. J., 1986, H o w heterogeneity affects oil recovery, in L. W. Lake, and H. B. J. CarroH, eds., Reservoir Characterization: Orlando, FL, Academy Press, p. 487-544.
Weber, K. J., 1987, Computation of initial well productivities in aeolian sandstone on the basis of a geological model, Leman Gas field, U.K., in R. W. Tillman, and K. J. Weber, eds., Reservoir Sedimentology: SEPM Special Publication 40, p. 335-354.
Weber, K. J., P. H. Klootwijk, J. Knoieczek, and W. R. van der Vlugt, 1978, Simulation of water injection in a barrier-bar-
354
PART 7—GEOPHYSICAL METHODS
type, oil-rim reservoir in Nigeria: Journal of Petroleum Technology, v. 30, p. 1555-1565. White, T. C., 1984, Dolomitization, sulfate solution, and porosity development: The San Andres Formation, Howard-Glassock field, Howard County, Texas, in AAPG Southwest Section Transactions: West Texas Geological Society Publication 84-78, p. 91-102. Williams, G., and I. Vann, 1987, The geometry of listric normal faults and deformation in their hanging walls: Journal of Structural Geology, v. 9, p. 789-795. Wilson, M. D., and E. D. Pittman, 1977, Authigenic clays in sandstones—recognition and influence on reservoir
properties and paleoenvironmental analysis: Journal of Sedimentary Petrology, v. 47, p. 3-31. Woodward, N. B., S. E. Boyer, and J. Suppe, 1985, An outline of balanced cross sections, 2nd ed.: University of Tennessee Department of Geological Sciences, Studies in Geology 11. Worrall D. M., and S. Snelson, 1989, Evolution of the northern Gulf of Mexico with emphasis on Cenozoic growth faulting and the role of salt, in A. W. Bally and A. R. Palmer, The Geology of North America—An Overview: Geological Society of America, v. A, p. 97-138.
Part 7
GEOPHYSICAL METHODS
edited by
Peter M. Duncan
Landmark Graphics Corporation Houston, Texas, U.S.A.
Contents
• Introduction • Seismic Data Acquisition on Land • Marine Seismic Data Acquisition • Basic Seismic Processing • SeismicMigration • Displaying Seismic Data • Seismic Interpretation • Mapping with Two-Dimensional Seismic Data • Three-Dimensional Seismic Method • Vertical and Lateral Seismic Resolution and
Attenuation • Synthetic Seismograms • Forward Modeling of Seismic Data • Seismic Inversion • Amplitude Versus Offset (AVO) Analysis • Checkshots and Vertical Seismic Profiles • Cross-Borehole Tomography in Development
Geology • Full Waveform Acoustic Logging • GravityMethod • BoreholeGravity • Magnetics • ElectricalMethods • ReferencesCited
Introduction
The inclusion of a geophysical section in this reference manual is a testament to the great leap forward that the geophysical discipline in general and the seismic method in particular have made in the last 15 years. To the geologist and engineer, geophysics has been the "black box" inaccessible to all but a few skilled practitioners. Despite impressive technology and computer power, the geophysicist was able to furnish only a fuzzy picture of the subsurface, highly subjective in its final interpretation and, generally, with neither the resolution nor repeatability to be of use in the development world.
The advent of three-dimensional (3-D) seismic surveys and computer-assisted interpretation have changed all that. The geophysicist can now contribute to the understanding of a reservoir on the scale necessary for development, both structurally and stratigraphically. Increasingly, a geophysicist is part of the team that concurrently uses all available data to obtain a better reservoir model and formulate a better development plan. Increasingly, geologists, reservoir engineers, and petrophysicists are being exposed to the practices of geophysics.
The purpose of Part 7 is to introduce the basic techniques and technology of geophysics, to facilitate communication among the disciplines, and to make geophysics more accessible to the practicing development geologist. As in
Peter M. Duncan
Landmark Graphics Corporation Houston, Texas, U.S.A.
other parts of the Manual, each chapter introduces concepts and terms, procedures and pitfalls only briefly and then points the interested reader in the direction of more detailed information.
This part begins with the seismic method, taking the reader from acquisition through processing and interpretation. Special extensions of the method such as 3-D, modeling, amplitude versus offset (AVO), vertical seismic profiles (VSP), tomography, and inversion are handled in chapters separate from the basics. To round out Part 7, brief discussions are included of the other geophysical techniques as they apply to the development problem, namely, gravity, magnetics, and electrical methods.
Acknowledgments
Part 7 of the Manual represents the efforts of many people. Stuart Buchan originally designed this part, but was unable to complete the editing job owing to other commitments. I would like to thank all the authors for their assistance in helping me complete the task Stuart began. We must all thank various companies who allowed for the time and data examples that the authors have used. I especially thank the two reviewers who greatly assisted in my efforts. Finally, a heartfelt thanks to the volume editors for their patience.
Seismic Data Acquisition on Land
Dale M. Short
Conoco Inc. Casper, Wyoming, U.S.A.
INTRODUCTION
There are f o u r basic c o m p o n e n t s of l a n d seismic acquisition:
1. Location—knowing where the Hne is to be shot 2. Source—some means of transmitting acoustic energy
into the subsurface 3. Receivers—a means of gathering the energy as it is
reflected by changes in rock properties in the subsurface 4. Recorder—a device for preserving the gathered information for processing
LOCATION
Once the location of the seismic program (a line or group of lines) is laid out on a program map, and prior to any field activities, permits must be obtained from governments and/or private landowners. Permits are commonly granted, and in the case of private ownership, fees are paid for trespass a n d / o r d a m a g e s . D u r i n g the p e r m i t t i n g p h a s e of t h e program, many potential problems with the location of the program must be diagnosed, such as impassable terrain, areas restricted to vehicular traffic (e.g., national parks and wildlife sanctuaries) or other obstructions (e.g., towns, buildings, and refineries). The actual s u r v e y i n g of the p r o g r a m is fairly straightforward. Standard survey techniques for determining x, y, and 2 coordinates generaHy provide adequate accuracy, given sufficient control. However, in remote or previously unsurveyed areas, satellite stations may be necessary to p r o v i d e a c c u r a t e p o i n t s of r e f e r e n c e . A b o v e all, reproducibility of the survey is necessary to be able to go back and drill at the shotpoint interpreted as "the place to drill."
SOURCE
Numerous energy sources exist, including various explosives (dynamite and primacord) (Figure 1), gas or air guns, weight drop mechanisms, vibrator systems (Figure 2), and even firearms, such as .50 caliber machine guns (fired single shot) and shotguns. Applications, advantages, and disadvantages of some of these sources are listed in Table 1.
Explosive sources are typically buried. Some examples include dynamite in "shot holes" ranging in depth from 10 to 300 ft (3 to 100 m) and primacord in trenches covered by a few inches of soil. However, in some cases, both of these sources can be exposed on the surface. The other sources are generally truck mounted, but specialty vehicles are often used as transport when required by surface or environmental conditions (e.g., wet ground or arctic tundra). Many areas are inaccessible, and require helicopter support. In such cases, all aspects of the operation must be portable enough to be loaded onto helicopters and then move by hand into position.
Source locations, or shotpoints, are sometimes single points, but are often positioned in arrays. A shotpoint array is specifically designed to impart maximum seismic energy into the ground and/or minimize seismic noise such as ground roll. Arrays may be as simple as lining up several vibrator trucks (Figure 3) or as complex as drilling and shooting sixpoint star patterns (Figure 4). Shotpoint intervals range from over 1000 ft (300 m) to less than 100 ft (30 m) and are often in intervals evenly divisible into 5280. However, modern seismic is typically shot with fairly short shotpoint intervals ranging from 55 to 440 ft (17 to 134 m). Shooting patterns can be off end or split spread (Figure 5). Split spread is probably the most c o m m o n and has an equal n u m b e r of geophone groups in front of and behind the shot.
Figure 1. Dynamite shot pattern being detonated in the desert. (Photo courtesy of Western Atlas International.)
Figure 2. Vibrator truck. (Photo courtesy of Western Atlas International.)
358
Seismic Data Acquisition on Land
359
Table 1. EnergySources
Source Explosives
Application General acquisition
Vibrators
General acquisition
Gas or air guns
Acquisition in rough terrain
Weight drop
General acquisition. Desert acquisition.
Advantages
Short, broad band pulse. Source can be placed below weathering layer.
Minimal cultural and environmental impact. Controlled frequency source. Narrow frequency band can increase signal to noise ratio. No shot holes required.
Cheap. Heli-portable. Can use multiple sources. No shot holes required.
Cheap. No shot holes required.
Disadvantages
Requires expensive shothole drilling. Frequency dependent on material in which explosion occurs. Dangerous to handle.
Shallow weathering can cause statics problems. Base-plate to ground coupling can be a problem. More complex recording and processing.
Limited energy penetration. Ground roll problems. Requires longer offsets. Shallow weathering can cause statics problems.
Cannot synchronize sources. Ground roll problems. Requires longer offsets. Shallow weathering can cause statics problems.
RECEIVER
A geophone is a mechanical device that transforms seismic energy into electrical voltage (Figure 6). Individual geophones are often wired together and configured in arrays along a cable. These arrays are designed much the same as source arrays and for the same basic reasons, that is, to maximize detection of reflected energy and to reduce the a m o u n t of noise. H o w e v e r , it is i m p o r t a n t that for any geophone array to work, the individual phones must be properly planted and not just thrown out on the ground, stuck into trees, hung in bushes, or set on rocks.
There are two basic types of cable systems: analog and telemetry. The analog systems have a pair of wires for each geophone group and several additional pairs of wires for rollalong. (Roll-along allows for shooting to continue while geophones are picked up behind the shot and moved into position in front of the shot.) For example, a 96-channel system may have 72 pairs of wires for the front part of the cable and the same for the back. Likewise a 240-channel s y s t e m m a y h a v e as m a n y as 144 p a i r s of wires. The advantage to this sort of hard-wired system is that it can be used in most any type of terrain. However, if these cables get too long, the signal may be attenuated by leakage or obliterated by 60-Hz noise. These problems can be overcome by telemetry systems (also known as distributed systems), which have an analog connection from the geophone group to a processor. The processor or station box amplifies the analog signal, filters, digitizes, and transmits the digital signal to the recording facility by wire, optical fiber, or radio. Hybrids of these two systems can be used to accommodate varying field conditions.
RECORDING
Modern recording systems can be as varied as the source and the detection systems. The recording system does many of the same things to the data that the processor or station box in the distributed systems does, such as amplifying, filtering, multiplexing, and digitizing, but in the end, the data are recorded (generally on magnetic tape). Parameters such as sample rate, record length, and recording filters can be controlled in the recording process. These instruments are often the limitation in the n u m b e r of channels that can be recorded. Typical modern units record u p to 120 channels, but these can be linked together to form master-slave units. Higher n u m b e r s of channels are the wave of the future in recording systems.
FINAL REMARKS
Modern land seismic acquisition has advanced from a shallow structural determination technique to a sophisticated subsurface measuring tool. Future developments, such as cables with increased numbers of channels, telemetry, and m o r e a d v a n c e d e l e c t r o n i c s h a v e s p a w n e d a r e a s of specialization beyond any one person's grasp. This necessitates good communication between the explorers and their partners in the acquisition and processing sides of the business. Knowing what specific information the explorer is looking for allows the acquisition and processing personnel to properly design a program that maximizes resources, prevents overkill, and accomplishes the task at hand.
360
PART 7 — G E O P H Y S I C A L METHODS
Figure 3. Vibrator truck in a simple "in line" array. (Photo courtesy of Western Atlas International.)
Figure 4. Six-point star shot pattern,
A
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX V
B XXXXXXXXXXXXXXXXXXXXXXXX V x x x x x x x x x x x x x x x x x x x x x x x x
V SHOT X GEOPHONE CENTERS
Figure 5. (a) Off end shooting, (b) Split spread shooting.
Figure 6. Geophone. (Photo courtesy of Western Atlas International.)
Marine Seismic Data Acquisition
James F. Desler
TGS-Calibre Geophysical Co. Houston, Texas, U.S.A.
INTRODUCTION
Marine seismic data acquisition techniques and capabilities have advanced at a rapid pace over the past few years. Today the typical marine seismic source is much more powerful, efficient, and effective due to improvements in towing technology, air gun array design, and the air guns themselves. Towing two parallel streamer cables in threedimensional surveys is now the rule rather than the exception, and recording between 240 and 480 channels of seismic data is common. The ability to determine the position of everything in the water on a shot-by-shot basis has also improved significantly.
SOURCE
Over the years, a large variety of different marine seismic energy sources have been used, but today two sources are used for almost all of the marine seismic shot. These are the air gun and the water gun, with the air gun being by far the most common.
In a water gun, a piston is driven through the water so fast that a vacuum bubble is produced. When this bubble collapses, acoustic energy is radiated. The pressure signature from a water gun has pressure variations (a precursor) before the main pressure pulse. This precursor is caused by the initial acceleration of the water, a n d special w a t e r g u n processing (signature deconvolution) must be performed. Geophysical contractors who use water guns have developed this special processing.
In an air gun, high p r e s s u r e air (typically 2000 psi) is stored in a chamber, and upon receipt of a firing command, this air is expelled into the water. This release of air generates a strong pressure pulse and forms an air bubble. This bubble oscillates, g e n e r a t i n g a s e q u e n c e of d e c a y i n g p r e s s u r e variations (bubble train) that follow the initial pulse. Air gun arrays a r e d e s i g n e d u s i n g air g u n s t h a t h o l d d i f f e r i n g amounts of air, so that the bubble oscillations are of different periods and therefore tend to cancel one another while the initial pulses reinforce. Thus, a signal approaching an impulse is generated because the primary objective of seismic method is to measure the acoustic impulse response of the earth.
Recently, sleeve guns, w h i c h are d e s i g n e d to be m o r e efficient and reliable than conventional air guns, have been introduced. A disadvantage of the sleeve g u n is that it is currently offered in only a few discrete sizes, which handicaps the air g u n array designer. Clustered guns are another recent advancement. When two or three air guns are fired in close vicinity of one another, it is possible to get a strong initial pulse and a weak bubble sequence because the composite bubble formed is not spherical and thus does not tend to support oscillations.
The output of an air gun array is typically illustrated by a normalized pressure time sequence called a signature. The signature of a 1000-in.3 air gun subarray is shown in Figure 1. One measure of the strength of the source is the peak-to-peak pressure, which is often quoted in pressure units of bars at 1 m (bar meters). Another measure of the performance of the array is the peak-to-bubble ratio, which is the peak-to-peak magnitude of the initial pulses divided by the magnitude of the residual bubble oscillations. These simplistic measures of performance can be used to compare different sources provided the signatures have been recorded using the same techniques, especially the same field filters (Johnston et al., 1988).
The strength of marine seismic sources has increased at a steady pace over the past 20 years, as shown in Figure 2. Stronger sources produce detectable signals from deeper reflectors. Very large air compressors are necessary to charge the air guns, which typically fire every 10 sec. In addition to increased strength, air gun arrays consisting of six, eight, or ten subarrays (strings) are deployed in various geometries to improve directivity and/or to reduce shot generated noise. The simple wide array is common and very effective (Lynn and Larner, 1989).
RECEIVER
The basic element used to detect the reflected seismic energy is the hydrophone, a piezoelectric device that creates an electrical signal in response to pressure changes. Multiple h y d r o p h o n e s are m o u n t e d in a streamer cable, which is a jacketed tube filled with a liquid less dense than water. Weights and liquid are combined so that the streamer cable is near n e u t r a l b u o y a n c y . Then, w i t h the h e l p of d e p t h controlling devices called birds, the cable can be positioned at a specified depth below the surface of the water. Acliieving neutral buoyancy is called balancing the cable.
Multiple hydrophones mounted throughout a certain length of the streamer cable are electrically connected in series and parallel so as to form what is called a detector array, a station, or a group. Typically, a group is 25 m in length and contains from 25 to 40 hydrophones. Group lengths range from 6.25 to 50 m. It is common to put two, three, or four groups in a cable section (75 to 100 m long), which is a unit with connectors at each end. When a section is damaged or fails for any reason, it can easily be replaced.
The length of a streamer cable ranges from 3000 to 6000 m. Shorter cables can be deployed in areas with obstructions to improve control and thereby avoid damage. The length of cable required for a job d e p e n d s on the objectives of the survey. In general, the deeper the objective, the longer the cable should be. A rule of thumb is that x, the length of the cable, should equal z, the depth of the objective.
361
362
PART 7—GEOPHYSICAL METHODS
Figure 1. Typical air gun array signature.
YEAR
Figure 2. Air gun array strength.
Some streamer cables contain electronics to convert the analog signal from the hydrophones to a digital form. These are called digital cables, while the traditional streamer cables are called analog cables. In an analog cable, there must be a pair of w i r e s for each g r o u p . As the n u m b e r of g r o u p s increases, the number and weight of the copper conductors increase, which necessitates that the diameter of cable gets larger and larger. There are analog cables with as many as 240 channels (Carlini and Mazzotti, 1989). Digital cables, however, can multiplex data onto one or a few conductors, so there is not a weight problem due to conductors. Some digital cables have as many as 1000 channels. Analog cables may be more susceptible to electrical leakage, but a well-maintained analog cable can gather data equal in quality to that gathered using a digital cable.
In 3-D seismic surveys, the seismic boat commonly tows two and even three parallel cables spaced laterally apart so that two or three seismic lines are collected per traverse of the vessel. In a two-cable application with a single source, two seismic lines spaced L meters apart in the subsurface will be acquired if the streamer cables are spaced 2L apart.
GHOSTING
The air-water interface reflects upcoming acoustic energy, a p h e n o m e n o n referred to as ghosting. D u e to ghosting, certain frequencies are destructively interfered with while other frequencies are boosted. The lowest frequency boosted by the maximum amount is calculated by dividing the velocity of sound in water (1500 m / s e c ) by four times the depth of the detector or the source. Therefore, a detector or source at a depth of 10 m causes the 37.5-Hz component of the seismic signal to be e m p h a s i z e d . O d d multiples of this frequency are also boosted. The frequencies that are completely eliminated (notched) by ghosting are the even m u l t i p l e s of the f r e q u e n c y just calculated. The notch frequencies caused by a detector or source at 10 m are 0, 75, 150, and so on. Thus, the depth of the source and the depth of the detector are important survey parameters.
NAVIGATION
It is essential that the precise location of the survey be known, and this is where navigation is important. The work horse of the industry, certainly in the Gulf of Mexico, is the shore-based navigation system Si/ledis. In the Gulf and in other areas, base stations have been placed on production platforms to extend the range of coverage. Aiiother system, Star-Fix, uses commercial satellites and a shore-based network of stations to determine the position of the satellites precisely. Given this information, the position of the vessel can be calculated. Another satellite system is the Global Positioning System (GPS), which is now being installed by the U.S. government. It is still unclear whether its capabilities will be intentionally degraded for commercial applications. GPS (or differential GPS) has the potential to make all other navigation systems obsolete.
The advent of 3-D seismic has increased the demand for accuracy in positioning. One needs to know within a few meters the position of the source and the position of each detector group for every seismic shot. Underwater, acoustical systems and/or laser measurement through the air are used to determine the positions of things in the water near the vessel. Extremely accurate magnetic compasses are attached along the streamer cable, and sometimes a navigation receiver is placed in the tail buoy at the end of the streamer cable. Accuracies of a few meters are desirable.
VESSEL
The ship that carries and tows the source and streamer cable and that houses the personnel needed to operate and repair everything must be fairly large and self contained. Currently, the ideal sized conventional seismic vessel is about 250 ft (75 m) long. Comfortable living conditions are required to attract and keep qualified personnel. The electronic and mechanical systems on a seismic boat are complex, and when something breaks far out at sea, the people onboard need to be able to fix it.
QUALITY CONTROL
It is essential to have stringent but fair quality control (QC) specifications on each and every aspect of the marine seismic data acquisition system. Air guns will break. The QC specifications must address how many air guns can fail before acquisition must be terminated for repair. Things will go wrong with the streamer cable. How many channels can go bad before the cable must be repaired? Sometimes there are other seismic crews in the area. How much interference noise (Lynn et al., 1987) can be tolerated?
The details of this very important subject are beyond the scope of tliis note, but in my opinion, it is almost as easy to keep things running at 90 to 100% capacity as it is to keep things running at 70 to 80%, so it makes sense to demand liigh performance.
Marine Seismic Data Acquisition
363
CONCLUSIONS
We can expect marine seismic data acquisition technology to continue to advance. The objectives remain the same—get deeper penetration and higher resolution in a cost effective manner. The performance objectives—deeper penetration and higher resolution—will be achieved by increasing the signal to noise ratio. This is accomplished by increasing the signal and/or decreasing the noise.
Seismic data processing is important (see the chapter on "Basic Seismic Processing" in Part 7), a n d m u c h of the increased signal to noise will be done on the computer, but only so much. When better data are acquired in the field, the final result will be better illumination of the exploration objectives.
Basic Seismic Processing
Peter M. Duncan
Landmark Graphics Corporation Houston, Texas, U.S.A.
INTRODUCTION
The seismic data written to tape in the dog house, whether on land or at sea, are not ideal for interpretation. To create an accurate picture of the subsurface, we must remove or at least minimize artifacts in these records related to the surface upon which the survey was performed, artifacts related to the instrumentation and procedure used, and noise in the data obscuring the subsurface image. Treatment of the data to achieve these ends is commonly referred to as seismic data processing. Through processing, the huge volumes of data taken in the field are reduced to simple images for display on paper or the work station screen. This simple image, while it contains less data about the subsurface, is readily accessible to the interpreter and has many of the artifacts and errors just listed removed. Figure 1 shows a single, unprocessed (raw) field record taken from a line. Figure 2 is the same line of data after processing to illustrate how the field records are turned into an interpretable image.
BASIC FUNCTIONS
The processing sequence designed to achieve the interpretable image will likely consist of several individual steps. The n u m b e r of steps, the o r d e r in w h i c h they are applied, and the parameters used for each program vary from area to area, from dataset to dataset, and from processor to processor. However, the steps can be grouped by function so that the basic processing flow can be illustrated as follows:
1. Database building—The myriad of numbers on field tape must each be uniquely related to shot and receiver positions on the surface of the earth, an elapsed time after the shot that originated the reflection or echo (traveltime), and a reflection p o i n t on the s u b s u r f a c e of the e a r t h at a n y traveltime. The p r o p e r assignment of these geometrical properties is fundamental to all that follows. As computers move into the field, more of this work will be done at the dog house.
2. E d i t i n g a n d f u n d a m e n t a l c o r r e c t i o n s — O b v i o u s experimental failures due to humans or machines are flagged for removal from the records. Differences in traveltime related to elevation and other surface conditions at the shot or receiver are removed, as are the timing peculiarities of the field apparatus. The weakening of the signal with distance from the source is also corrected by a simple multiplication of the signal by a geometrical spreading factor.
3. Signal to noise enhancement—Portions of the record showing low signal to noise ratio, usually determined visually but based on certain models of signal p r o p a g a t i o n in the earth, are removed by filtering the recording. Where organized (nonrandom) noise is recognized, one usually tries
to determine the origin of this noise to better predict how it will manifest in the signal and hence derive the most efficient filter to remove it. Removal of water bottom multiples is an example. Redundant samples of the same subsurface location that occur in a p r e d i c t a b l e f a s h i o n as a result of the multichannel recording technique are summed together to reduce r a n d o m noise in a process called stacking (Sheriff, 1984).
4. Enhancement of resolution in time—To the extent that the earth is a perfectly elastic medium, the reflection from any interface is instantaneous, that is, it has no width in time. Ideally, we should be able to determine the time of a reflection absolutely and achieve infinite resolution. Unfortunately, this is not possible. First, the signal sent into the earth is not infinitely short. Rather, it is a pulse with some finite width. If more than one interface is encountered within the width (in time) of the source pulse, the responses will interfere and the reflection received at the surface will be a complex sum of all the reflections created. One can think of the source pulse as a running sum over the ideal reflection sequence. Second, the hydrophone or geophone receiver and the seismic recording device each have a characteristic response time, that is, they take time to react to any signal such that a pulse is smeared or averaged over a time wider than the pulse itself. Reflections occurring at shorter intervals than this characteristic time will be summed together. Finally, the earth is not perfectly elastic so s m e a r i n g of the signal occurs t h r o u g h the n a t u r a l mechanism of transmission in the earth.
The mathematical process used to compute the result of such interactions is called convolution. Reversing the process is called deconvolution (Sheriff, 1984). If one k n o w s the response time of the instrument and receivers (hydrophones or geophones) used, one can calculate the summing function that has been applied to the signal and can remove it or deconvolve it from the seismic records. Similarly, the source pulse or wavelet and the nonelastic properties of the earth can be removed using the process of deconvolution in an attempt to eliminate all time-averaging effects and turn the seismogram into a series of narrow reflections with greater resolution in time.
5. E n h a n c e m e n t of r e s o l u t i o n in s p a c e — J u s t as the seismic source has width in time, which reduces temporal resolution, it also has width in space, which reduces spatial resolution. As the seismic wavefront travels outward from the source, it not only gets weaker (as a result of energy conservation), but also causes reflections from a larger and larger area. (Consider light from a flashlight or ripples on a pond.) All of these reflections are recorded at the receiver location as a single sum without regard to the origin of the reflection except for time of travel. The spatial width of the signal must be narrowed as was the time width. This spatial deconvolution is analogous to the process of triangulation to locate the source of an observed signal. Many observations of
364
Basic Seismic Processing
365
Figure 1. A single shot record as it is recorded in the field. The shot is at station 60. There were 120 geophones laid out in this "split" spread. Two seconds of data were recorded. (Courtesy Landmark/ITA.)
the same reflection from different points on the earth are required so that different traveltimes are available for a given reflection. Predictable patterns in arrival time allow for the determination of the location of the reflector. Signals from all but those reflectors directly beneath the surface position of a trace are removed from the trace. This effectively collapses the spatial spreading of the signal to a single downgoing ray. Spatial resolution approaches the trace interval. Seismologists call this p r o c e s s migration (see the c h a p t e r on "Seismic Migration" in Part 7).
6. A e s t h e t i c s — T h e u n d e r d e t e r m i n e d n a t u r e of the seismic interpretation problem means that interpretation remains a mostly subjective application of pattern recognition by highly experienced individuals. It is thus understandable that considerable time and effort is expended in any processing project on the final parameters of seismic display so as to satisfy the individual tastes of the interpreter. Such things as frequency content, gain, trace spacing, and type of display are all up for grabs (see the chapter on "Displaying Seismic Data" in Part 7).
366
PART 7—GEOPHYSICAL METHODS
Figure 2. A seismic section produced by processing six shots such as those in Figure 1. (Courtesy Landmark/ITA.)
TYPICAL PROCESSING STEPS
Given the broad categories of processing functions just described, this section briefly defines the common programs by their generic names in the order they would normally be applied. Some steps may be applied more than once at different times in the sequence, while others may be skipped for a particular dataset.
Demultiplex—The name given to sorting the traces from time ordered storage (all receiver stations at a given time) to receiver ordered format (all times for a given receiver) or trace sequential format. Many modern instruments do this in the field, but much data still comes in from the field multiplexed.
SEG A and SEG B formats are multiplexed, SEG Y is a trace sequential fonnat, and SEG D can be either way.
Edit—The process of flagging traces or pieces of traces to be ignored for one reason or another.
Geometry—The association by unique identifier of each recorded trace with shot and receiver locations.
Antialias filter—A low pass filter applied before resampling the data to a coarser time scale to prevent aliasing. Aliasing is a phenomenon in which high frequency data m a s q u e r a d e s as l o w f r e q u e n c y e n e r g y as a result of undersampling. To sample a signal properly, there must be at least t w o s a m p l e s w i t h i n the shortest period of interest. Antialias filters remove frequencies above the sampling limit (Nyquist frequency) of the new sampling time. The operation
Basic Seismic Processing
367
120
Figure 3. The shot record ot Figure 1 after the application of a gain recovery algorithm to replace the energy lost as the signal traverses the earth. (Courtesy Landmark/ITA.)
is performed before the sampling is reduced. Gain recovery—The correction for the loss in amplitude of
a signal as it travels through the earth and spreads its energy over a larger surface area. This involves multiplication of the signal by a number that increases with time. The exact time variant multiplier can be based on the theoretical concept of spherical s p r e a d i n g (related to the square of the distance traveled), can be based on measurements of amplitude decay with time made on the data itself, or can be entirely arbitrary. An example of the effect of gain recovery is given in Figure 3.
Deconvolution—The removal of the frequency-dependent response of the source and the instrument. The instrument response is normally known and can be removed exactly. The source shape is not usually known but can be measured directly (marine air gun signatures) or estimated from the signal itself under certain assumptions. Signature deconvolution, wavelet deconvolution, spiking deconvolution, gapped deconvolution, predictive deconvolution, maximum entropy deconvolution, and surface consistent deconvolution are various manifestations of the attempt to
368
PART 7—GEOPHYSICAL METHODS
Figure 4. The shot record after a statistical deconvolution process has been applied to "shorten" the wavelet and increase time resolution. (Courtesy Landmark/ITA.)
remove the source width from the observed reflections (Yilmaz, 1987). The resulting reflection sequence always has some s m o o t h i n g f u n c t i o n left, usually called the residual wavelet. A t t e m p t i n g to be too exact about deconvolution u s u a l l y r e s u l t s in a v e r y noisy section. The effect of deconvolution is seen in Figure 4.
Statics—The removal of traveltime artifacts relating to the placement of the source and receiver at or near the earth's surface. Differences in traveltime to the same reflector which
result from elevation differences and near-surface velocity changes at different source and receiver stations must be removed. The relative elevation of each shot and receiver location and the near surface velocity must be known to make these corrections. An elevation datum is chosen, and the distance above or below that datum is measured for each source and receiver. The difficulty is in knowing what velocity to use to convert this elevation difference to a time correction to be added to or subtracted from the entire trace
Basic Seismic Processing
369
(hence the term statics). Refraction statics, surface consistent statics, and residual statics are all techniques used to estimate and apply the appropriate velocity and time corrections (Figure 5).
Demultiple—Strong reflections can act as a secondary source of seismic energy that will interfere with the primary reflections and confuse the interpretation. Such secondary reflections are called multiples. The most c o m m o n are water bottom multiples, but interbed multiples also exist. The demultiple process attempts to remove these (Yilmaz, 1987).
f-k or apparent velocity filter—Acoustic signals that are not reflections from subsurface layers appear in shot records (Figure 1) as s t r a i g h t lines r a t h e r t h a n h y p e r b o l i c curves. These events have a constant "apparent velocity" as they travel along the receiver cable. This simple organization allows them to be isolated from the reflection signal a n d to be removed from the record. A c o m m o n w a y to d o this is with the FK (sometimes called pie slice) filter. Judicious selection of the r a n g e of a p p a r e n t velocities to be r e m o v e d can eliminate linear noise. Too wide a filter can remove too much information from the section and causes serious interpretation problems.
Normal moveout (NMO) correction—The reflection from a given horizon does not arrive at the same time at different receivers along the length of the seismic cable or spread (see the chapter on "Seismic Migration" in Part 7). However, if the velocity at which the sound traveled is known, the arrival time difference (moveout) at each station can be predicted. Conversely, knowing the arrival time difference, the velocity the sound traveled can be determined under certain model assumptions. Usually the velocity of the earth as a function of time is determined at a few locations over the survey. This model can then be used to calculate moveout as a function of time everywhere in the survey. The moveout is subtracted from each seismic record such that the reflections from a given horizon will appear flat. This facilitates identification of reflectors and stacking. Figure 6 demonstrates the NMO process.
Dip moveout (DMO) correction—NMO corrections are m a d e u n d e r the assumption of horizontal planar reflectors. If the reflector has appreciable dip, then the actual movement will be slightly different. The DMO correction is a method for estimating the effect of dip on moveout and removing it from the records as well.
Common midpoint (CMP) stack—This is the single most effective step for noise reduction in the processing flow. The shooting procedure results in many traces being acquired with the point midway between source and receiver (called the midpoint) being coincident on the earths surface. The only difference between the traces is the distance between source and receiver (offset). Once these traces have been NMO (and DMO) corrected, they are really redundant s a m p l e s of the s a m e reflection. A d d i n g them together increases the signal to random noise ratio by the square root of the n u m b e r of r e d u n d a n t samples. The process reduces the field data to a stacked section consisting of one trace for each midpoint location, assumed to have been recorded with a shot and receiver coincident at the midpoint location (see Figure 2).
Poststack filter—Usually a band pass filter, this process excludes frequencies above a certain value (high cut) and below a lower value (low cut) to retain that part of the signal with the highest signal to noise ratio. The values are usually set by trial and error a n d judged by a visual comparison of sections. The values may be different for different time gates of the section. Typically, the deeper reflections (later time) have less signal at high frequencies because these frequencies are absorbed or scattered more readily in the earth. Consequently, a lower value for the high cut frequency must be used as the bandpass is applied to later times on the trace.
Poststack mix—This is a simple procedure that averages together adjacent traces to enhance the signal to noise ratio. It causes a concurrent loss in horizontal resolution.
Migration, display and other advanced processing techniques are available and essential to the complete utilization of the seismic data (see other chapters in Part 7).
CONCLUSIONS
Seismic processing attempts to enhance the signal to noise ratio of the seismic section and remove the artifacts in the signal that were caused by the seismic method. The end result should be a more interpretable section. The process has s o m e very subjective elements. The selection of various parameters is done most often heuristically with more emphasis on satisfying the personal taste of the interpreter than on the rigorous physics of signal processing.
370
PART 7—GEOPHYSICAL METHODS
0.0 0. 1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 .0 1. 1 1.2 1.3 1 .-4 1.5 1 .6 1.7 1.8 1 .9
Я
мшм
0.0 0. 1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 . 0 1 . 1 1.2 1.3 1.4 1.5 1.6
1.8 1.9
-1980 -1287 -594
132
858
1551
(a)
Basic Seismic Processing
371
-1980 -1287 -594
132
0.0
858
1551
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1 ШШШШШШ^ ШШ , J I M ' / , J , , I ' » , / ; . ( i f . i l l « Ufi'f,',«j4»<5« I I » Л Ji « "V,.«i*V-5 <• ,W' «4
1.2
Ш 1.3 Ш Ш Ш !
1.4
1.5
1.6
1.7
1.8
1.9
(b)
Figure 6. (a) A gather ot processed traces with a common surface location. Shot-to-receiver offset is zero at the center of the gather and increases to about 2000 m on either end. The offset related curvature of the reflections is due to normal moveout. (b) Normal moveout correction (NMO) has been applied and the horizons are flat. The gather is now ready to be summed or stacked to produce one trace on Figure 2. (Courtesy Landmark/ITA.)
<—
Figure 5. (previous page) The application of statics corrects for differences in arrival time caused by elevation or weathering, (a) The valley in the data to the left of station 1500 represents an anomaly that persists throughout the time length of the record, (b) This "static" effect has been corrected. (Courtesy Landmark/ITA.)
SeismicMigration
"
KenLamer
DepartmentofGeophysics Colorado School of Mines Golden, Colorado, U.S.A.
David Hale
Advanced Geophysical Corp. Englewood, Colorada, U.S.A.
INTRODUCTION
Virtually all seismic data processing is aimed at imaging the e a r t h ' s s u b s u r f a c e , that is, o b t a i n i n g a picture of subsurface structure from the seismic waves recorded at the earth's surface. Deconvolution, for example, aims to sharpen reflections, and common midpoint (CMP) stacking exploits data redundancy to enhance signal-to-noise ratio while producing a seismic time section that simulates what would have been recorded in a zero-offset seismic survey, that is, one in which a single receiver, located at each seismic source
source-receiver
position, records data generated by the source at that position. Of the many processes applied to seismic data, seismic
migration is the one most directly associated with the notion of imaging. Until the migration step, seismic data are merely recorded traces of echoes, waves that have been reflected from anomalies in the subsurface. In its simplest form, then, seismic migration is the process that converts information as a function of recording time to features in subsurface depth. Rather than simply stretching the vertical axes of seismic sections from a time scale to a depth scale, migration aims to put features in their proper positions in space, laterally as well as vertically.
All the issues in seismic migration reviewed here are treated in the collection of reprints found in Gardner (1985).
CMP (x)—
boulder
depth ' (Z)
velocity = V
t modeling
4 migration
CMP (x)
diffraction
Figure 1. Schematic depth section (top) and zero-offset time section (bottom) for a single boulder at depth.
WHAT MIGRATION ACCOMPLISHES
The migration problem is illustrated in Figure 1. The u p p e r p a r t of the f i g u r e d e p i c t s a z e r o - o f f s e t s u r v e y conducted over a subsurface medium that is homogeneous (constant P-wave velocity) with the exception of an isolated boulder at some depth. Also shown are the straight ray paths traveled by seismic waves from each of five different source positions down to the boulder and back up to receivers located at the sources. Clearly, reflections from the boulder will be observed at all the surface locations, not just the one directly above it. Also, the reflection time clearly increases as the source-receiver pair is moved farther from the point directly above the boulder. The bottom part of Figure 1 shows schematically the seismic section that would be obtained for this survey. Reflections occur along a hyperbolic diffraction pattern with the apex at the same CMP location as that of the boulder.
The task of migration here is to convert or map reflections along the diffraction into a single point at the position of the boulder. The reverse process, by which the boulder gives rise to the observed diffraction pattern, is called modeling.
While the earth's subsurface is more complicated than that shown in Figure 1, the seismic data that would be obtained over the real earth can for all purposes be represented as a superposition of many diffraction curves generated by each of many boulder-like anomalies in the subsurface.
Figure 2 shows another depth section and associated seismic section for a subsurface consisting of a single dipping reflector. For a constant-velocity subsurface, the many weak diffractions from very closely spaced points along the reflector (of which five are s h o w n in the figure) give rise, through constructive and destructive interference, to a net
372
Seismic Migration
373
reflector
t modeling
* migration
Figure 3. Schematic depth section showing normal incidence ray paths for two-way travel between source-receiver positions and a dipping reflector.
HOW MIGRATION IS ACCOMPLISHED
One senses the massive scale of the computer intensive two-dimensional mapping involved in the transformation from unmigrated to migrated data in Figure 4. While these schematic sections depict what migration aims to accomplish, they say little about how it is done. All of the many methods of doing migration are founded on solutions to the scalar wave equation, a partial differential e q u a t i o n that m o d e l s h o w waves propagate in the earth. A simple form of the wave equation.is as follows:
Figure 2. Schematic depth section (top) and zero offset time section (bottom) for a dipping reflector at depth.
reflection along the straight-line envelope of the diffraction curves. Note that the reflection is displaced laterally from the true reflector position (the line connecting apexes of the d i f f r a c t i o n curves). It is this lateral m i s p o s i t i o n i n g of reflections from dipping reflectors that gave rise to the term migration for the process that corrects the positioning.
Figure 3 shows another perspective on this mispositioning. Reflections recorded at zero source-receiver offset follow ray paths that are perpendicular to the reflector. As a result, the reflection from the point on the reflector beneath point P, for example, would be recorded by the geophone at location G, to the right.
Figure 4 s h o w s the application of migration to CMPstacked field data. The superposition of diffraction curves evident in the u n m i g r a t e d data of Figure 4a gives rise to crossing reflections that can not plausibly be interpreted as structure. By correcting for lateral mispositioning of dipping reflectors and "collapsing" diffraction curves to zones defined by the diffraction apex, migration converts the recorded waves to a subsurface picture (Figure 4b) depicting both broadly and tightly folded anticlines and synclines.
dlP B1P _ 1 dzP dz2 + Bx1 ~ V2 dt2
w h e r e P(x,z,t) is the seismic a m p l i t u d e as a f u n c t i o n of reflection time t at any position (x,z) in the subsurface, and V is the seismic wave velocity in the subsurface, a function of both x and z. Disturbances initiated by a seismic energy source are assumed to propagate in accordance with solutions to the wave equation. Migration, then, involves a running of t h e w a v e e q u a t i o n backward in time, s t a r t i n g w i t h t h e measured waves at the earth's surface P(x,z = 0,t), in effect pushing the waves backward and downward to their reflecting locations.
All current computer-based approaches to migration involve this backward solution to the wave equation. The earliest form, Kirchhoffsummation migration, intuitively follows the situation depicted in Figure 1. In essence, recorded amplitudes on CMP traces are summed along the diffraction trajectories dictated by the assumed subsurface velocity distribution, and the s u m s are placed at the apexes of the curves, one curve for each sample point in the output migrated section.
Finite difference migration numerically integrates the wave equation by the method of finite differences to push seismic waves backward into the subsurface. A third category of migration approaches is f-k migration, which operates via Fourier transforms in the frequency wavenumber (f-k)
374
PART 7—GEOPHYSICAL METHODS
(b) Figure 4. (a) CMP stack of data from the Santa Barbara Channel, offshore California, (b) Result of migration.
domain. In general, Fourier transform methods provide elegant means of solving partial differential equations. When applied to migration, this elegance is often complemented by high computational efficiency.
Each of the different approaches has specific advantages such as computational efficiency, accuracy for imaging steep
reflectors, and accuracy in the presence of spatial variation of velocity. Likewise, each can produce undesirable processing artifacts related to some limitation in data quality such as poor signal to noise ratio, too coarse a spatial sampling interval, and missing data (e.g., due to seismic source misfires).
VELOCITY: THE KEY PARAMETER
Regardless of the migration approach implemented, the key parameter of the process is velocity. Since migration involves pushing waves back to their reflecting points, it is essential that the waves be pushed backward through the same medium through which they have propagated. Clearly, waves will not get back to the correct position at a given time if the velocity used in the migration process differs from the actual subsurface velocity. Unfortunately, subsurface velocity is seldom well known, particularly in geologically complex areas. Today's migration algorithms are highly accurate when supplied with the correct subsurface velocity. Because subsurface velocity can only be estimated, however, migration yields only an estimate of the true subsurface.
Where lateral variation of velocity is modest (as in many places in the Gulf of Mexico), migration methods in the class called time migration have performed adequately. Where lateral velocity variation is severe (as in many overthrust areas), more computationally intensive depth migration is required. Note that the terms depth and time migration do not relate to whether the migrated results are presented as a function of time or depth. Results of both migration categories are most often displayed in time (as in the examples shown here) because of added uncertainties in results converted to depth. While depth migration is capable of accurate subsurface imaging where velocity is complex, the required accurate estimation of velocity is difficult and time consuming.
Seismic Migration
375
POSTSTACK VERSUS PRESTACK MIGRATION
While migration algorithms are capable of accurately imaging reflections from steep interfaces, shortcomings in CMP stacking lead to destruction of such reflections before conventional poststack migration is applied. Two alternatives to poststack migration of CMP stacked data preserve reflections from steep interfaces. Migration can be applied to the unstacked data (so-called prestack migration) so that the data need not be reduced to an approximation to zero offset before migration. The improvement in imaging of steep reflectors by this approach, however, is bought at the price of a great increase in the amount of computation required for the migration.
A cost-effective and accurate alternative to full prestack migration is to apply poststack migration to data that have had the added step of dip moveout (DMO) applied after normal moveout (NMO) correction, but before the data are stacked. DMO, a form of partial prestack migration, completes the process that NMO only imperfectly accomplishes—it converts data recorded with separated sources and receivers to a close approximation to zero-offset data, preserving reflections from both gently dipping and steep reflectors. Figure 5 shows the improvement in imaging of the steep flank of a salt dome achieved by poststack migration when applied to DMO-processed data.
The additional accuracy of either DMO or prestack migration over that of conventional poststack migration demands special care in the field acquisition of seismic data. Too coarse a spatial sampling, that is, too large a geophone
Figure 5. (a) CMP stack showing reffections from a salt dome in the Gulf of Mexico, (b) Stack after DMO processing. Steep portions of diffractions and reflections are now preserved, (c) Migration of the DMO-processed data shows the steep flank of the salt dome to be a particularly strong reflector.
376
PART 7—GEOPHYSICAL METHODS
group interval, may preclude high resolution imaging of steep reflectors by any migration method.
The example in Figure 6 shows imaging of reflections from steep faults. While migration of CMP-stacked data (not shown here) shows the faulting, reflections from the faults themselves are absent. Details of the fault reflections seen on the DMO-processed result can be diagnostic of sealing along the faults.
The schematic diagrams shown here have been twodimensional (2-D) representations, and the illustrations have all involved 2-D migration of 2-D seismic data. Invariably, the earth's subsurface has three-dimensional (3-D) complexity. As a result, the mispositioning of recorded reflections extends in two lateral directions, and migration must be done as a 3-D process (see the chapter on "Three-Dimensional Seismic Method for Reservoir Development" in Part 7.). It suffices here to state that migration is fundamentally incomplete unless it is applied as a 3-D process to 3-D data.
Figure 6. (a) Unmigrated and (b) migrated stack of DMOprocessed data from the Gulf of Mexico.
Displaying Seismic Data
Kevin Reddy
Conoco Inc. Casper, Wyoming, U.S.A.
INTRODUCTION
Selecting the correct seismic display is vital to realizing the maximum potential for geological interpretation. The seismic trace contains much information that can be associated with changes in rock properties in the subsurface. Lithology, porosity, pore fluid, and a host of other variations are often detectable in the seismic response but can easily be masked by an improper display.
The time to think about display needs is at the outset of processing. The processor should be aware of the type of anomaly you are trying to discern, such as porosity, gas zone, or facies changes. This allows the correct selection of a data processing flow and display of the desired information.
POLARITY
In exploration seismology, reflections are recorded from velocity or density contrasts in the rock sequence. By SEG convention, an increase in velocity, such as from a slow velocity shale to a high velocity dolomite, is recorded and processed in such a way as to be displayed as a positive peak and is said to have a positive coefficient of reflection. If the rocks were reversed (that is, travel is from a high velocity rock to a slower velocity rock), the coefficient of reflection is said to be negative and is displayed as a trough. When this convention is followed, the display is called normal or positive polarity. Sometimes it is desirable to reverse the display convention. Then the display is said to be reverse or negative polarity.
In most structural plays, normal polarity would be the appropriate way to display data. However, when using lines recorded with different energy sources, it may be handy to have a normal and reverse section when you try to tie different data sets. Also, in stratigraphic work, it may be helpful to determine which polarity best fits your well ties, then work with that polarity. Occasionally, neither polarity seems to fit. When this happens, you may want to discuss the problem with the data processor who may be able to adjust the phase of the data to better fit the well or existing seismic data. This process is called phase shifting.
VARIABLE AREA WIGGLE TRACE
The variable area wiggle trace is the commonly accepted work horse in the petroleum exploration field. Each trace displays the data associated with its common depth point as a continuous trace oscillating on either side of a zero amplitude line (Figure la). Commonly, the area of the curve (trace or "wiggle") to the right of the zero line (positive or normal amplitude) is shaded black, while the negative side to the left is unfilled (Figure lb). The shading need not start at zero, but the area to be shaded (either positive or negative)
may be selected to enhance the visual display of the amplitudes desired. This shading of part of the area of the curve gives the display its name variable area wiggle trace (Figure lc). The term bias is used to describe the type and amount of shading desired. Zero (0) bias means all positive amplitudes are shaded. A 25% positive bias means the shading is offset by 25% of maximum amplitude from the zero line. A negative bias tends to shade in more of the negative values. Another factor is the choice of the trace deflection or excursion. A 2.5 excursion allows the maximum amplitude of the trace to cross over 2.5 trace spacings. This greatly affects the look of the data.
VARIABLE DENSITY AND SUPERIMPOSED DISPLAYS
A variable density display can be created by assigning different densities of shading to different amplitude values. Higher amplitudes are shaded darker, while lower amplitudes are less dark (Figure Id). It is also possible to superimpose the wiggle traces over the variable area or the variable density to create a new composite display.
COLOR AMPLITUDE DISPLAY
Color amplitude display is a variation on variable density in which color is substituted for shading. The use of many colors allows for increased amplitude resolution. Color presentation often more easily identifies geological changes in vertical and lateral stratigraphic sequences. A display of a variation in amplitude, phase, frequency, velocity, or any other trace-related attribute can be represented by different colors or shades of gray to black. These displays are very useful in stratigraphic studies and for detection of lateral variation in facies. The color band or shades of gray is your choice, limited only by machine capability.
SQUASH PLOTS
Squash plots are a very simple and sometimes effective method of enhancing the visibility of an event within a data set. These plots can be accomplished by several methods. The traces in the display can be moved closer together and allowed to overlap, or more commonly, adjacent traces are added together (trace mix). The number of traces summed and the mode of weighting should be determined from the structural and stratigraphic setting. Care must be taken, as a large number of trace sums in an area of steep dips may distort the picture and render undesirable results.
377
378
PART 7—GEOPHYSICAL METHODS
Figure 1. Various types of seismic displays: (a) wiggle trace, (b) variable area, (c) variable area wiggle trace, and (d) variable density. (Courtesy of Conoco Inc.)
AUTOMATIC GAIN CONTROL, TRACE BALANCE, AND FILTERING
At display time, scaling and filtering parameters must be selected. Energy decays very rapidly as it penetrates the earth and is reflected back to a receiver. If the energy reflected from the deeper horizons were displayed with the same amplitude scale as that from the near surface, peaks and troughs from deeper events would be invisible to the eye. Automatic gain control (AGC) is a method to normalize all amplitudes in a trace by looking at time w i n d o w s over that trace and adjusting the amplitude of events within the window relative to a chosen standard.
The choice of the design windows for the AGC affects the display. If the window is very short in time, the program tends to raise all amplitudes to the standard and very little differentiation between events is preserved. Generally, a window of about 250 msec is a good place to start. As the gate becomes longer, more of the original amplitude relationships are preserved. A single window trace equalization is sometimes desirable in processing data intended for relative amplitude studies, while a fast (short) AGC could be helpful in data having poor signal to noise ratio.
Another method of gaining the section is programmed gain control. This method applies a predetermined gain function to the trace and is most commonly encountered in field displays.
Since some form of gain is required to display the data at a usable scale, what are often called true amplitude sections are in reality relative true amplitude sections. The relative changes in amplitude are preserved by the method of gain used, and yet the entire data set is interpretable from the same display.
Filtering, that is, limiting the frequency content of the display, is an important aspect of the final display. A certain amount of filtering is done during the processing sequence to enhance the signal to noise ratio. At display time, one has a final opportunity to choose the frequency band that highlights the seismic target best. This highly subjective decision is usually made by testing several ranges and then choosing the one that works for that particular play. Usually the useful band width for the shallow section is much greater than for the deep section. Noise content may also vary with time and position along the line. A wide variety of constant or time/distance varying filter techniques is available.
Frequently, the noise degrading the signal may be very similar to the signal in dip, frequency, and other properties. At this point, you may have to choose how much signal you are willing to give up to reduce or eliminate the noise.
Seismic Interpretation
D c Nester
Orion Energy Inc. Houston, Texas, U.S.A.
Michael J. Padgett
Pennzoil Exploration & Production Co. Houston, Texas, U.S.A.
INTRODUCTION
____
Simply defined, seismic interpretation is the science (and art) of inferring the geology at some depth from the processed seismic record. While modern multichannel data have increased the quantity and quality of interpretable data, proper interpretation still requires that the interpreter draw upon his or her geological understanding to pick the most likely interpretation from the many "valid" interpretations that the data allow.
The seismic record contains two basic elements for the interpreter to study. The first is the time of arrival of any reflection (or refraction) from a geological surface. The actual depth to this surface is a function of the thickness and velocity of overlying rock layers. The second is the shape of the reflection, which includes how strong the signal is, what frequencies it contains, and how the frequencies are distributed over the pulse. This information can often be used to support conclusions about the lithology and fluid content of the seismic reflector being evaluated.
The interpretation process can be subdivided into three interrelated categories: structural, stratigraphic, and lithologic. Structural seismic interpretation is directed toward the creation of structural maps of the subsurface from the observed three-dimensional configuration of arrival times. Seismic sequence stratigraphic interpretation relates the pattern of reflections observed to a model of cyclic episodes of deposition. The aim is to develop a chronostratigraphic framework of cyclic, genetically related strata. Lithologic interpretation is aimed at determining changes in pore fluid, porosity, fracture intensity, lithology, and so on from seismic data. Direct hydrocarbon indicators (DHI, HCIs, bright spots, or dim-outs) are elements employed in this lithologic interpretation process.
This chapter discusses a basic three-step methodology, which when followed provides for a more complete and accurate geological interpretation from seismic data.
STEP ONE: INTERPRETATION PLAN
Accomplishing a successful interpretation requires that the interpreter first carefully consider the following questions.
What Are My Objectives?
An interpreter should clearly understand what conclusions are required from the data. Because so much information is available on the seismic, it is important to
focus maximum attention on extracting the data pertinent to completing the objective task. Does the objective require evaluating the entire dataset from first sample to last, one stratigraphic sequence, or just one specific amplitude anomaly? This dictates what combination of the three basic interpretation types should be used, when the interpretation should be completed, and what supporting databases are required.
What Are the Regional Tectonic, Structural, and Depositional Trends?
It is important for the interpreter to have a basic understanding of what tectonic influences and depositional systems occur within the area of the seismic survey to be investigated. Although this preconceived earth model may be vague and incomplete, particularly in frontier basins, it provides interpreters with insight and constraints as to what types of structures, faulting, and stratigraphic geometries may exist. The interpretation of fault styles, structural geometries, and facies patterns must be consistent with regional tectonic forces and basin infilling.
What Seismic Patterns Should I Be Looking For?
Perhaps the most common interpretational pitfall, and certainly one of the most dangerous, is the mapping of events, amplitude, or AVO changes without qualification as to what geological analog they represent. To prevent this mistake, it is critical that all types of available geological data be gathered and merged with the seismic data. Key to this merging are well-constructed synthetic seismograms, vertical seismic profiling (VSP) data, and/or seismic models (see the chapters on "Synthetic Seismograms," "Checkshots and Vertical Seismic Profiles," and "Forward Modeling of Seismic Data" in Part 7). This verifies the seismic signature of the target, the location of the mapping horizon, and the adequacy of the time-depth functions. Varying the synthetic seismogram or model parameters allows for the prediction of seismic responses for various lithologies and fluid types.
STEP TWO: BUILDING AND MERGING DATASETS
After developing an interpretation plan, the next step is to begin assembling the complete dataset. An inventory of available seismic data of all vintages is made, including p wave seismic, shear wave seismic, well data, velocity surveys, and VSP. The data are scanned for quality and
379
380
PART 7—GEOPHYSICAL METHODS
suitability. At this point, a determination can be made whether the available data can reasonably support the goals of the project.
Available well, core, test, paleontology, and outcrop data are gathered and organized for integration with the seismic data. Where available, gravity and magnetics data should be tied to the seismic data to identify the location of basement, salt bodies, igneous intrusives, and shale masses. Another type of data that sheds light on the geological conditions of a specific reservoir is pressure and production history data. These data can provide information on the presence and proximity of faulting and the size of fault blocks.
If digital data are available, a decision must be made whether to use a workstation or proceed with a "paper interpretation." Generally, a workstation offers the interpreter a valuable edge achieving a "correct" interpretation where detail is important (see the chapter in "Two-Dimensional Geophysical Workstation Interpretation: Generic Problems" in Part 8). For much regional work, paper is often still the most used medium due to the display limitations of the workstation screen. Seismic interpretations on paper, however, can always be digitized later for computer mapping or incorporation into a workstation project.
STEP THREE: INTERPRETATION
The process of interpreting seismic data eventually comes down to putting pencil to paper or cursor to screen. After building an exploration analog by integrating the available geological data, it is advisable to scan the dataset to observe the basin setting, major structural components, and major stratigraphic components, such as reefs, shelf breaks, and major sequence boundaries. While scanning, major faults can be picked as a guide to establishing the dominant structural style.
After scanning, detailed mapping begins by working outward from a point where geological information exists, preferably a well location with a synthetic seismogram. The horizons selected for mapping and observed fault cuts are correlated from the well to the seismic. The interpreter then begins to pick these same events away from the well on the seismic, being careful to tie at all other well locations.
Critical to the interpretation process is comparing how horizons and faults tie at line intersections. Significant effort is expended correcting misties of faults, horizons, and sequence boundaries at every line intersection. In this regard, closing the interpretation in loops around the seismic grid is a
particularly effective technique. On a workstation, a quick way to check for misties is a contour map. Misties will be evident by groups of unreasonable contours. In addition, workstations can be very helpful for working out the misties among varying vintages of two-dimensional data by applying time and phase shifts automatically (see the chapter on "Mapping with Two-Dimensional Seismic Data" in Part 7).
Tying all lines in both 2-D and 3-D data sets is the only way to reliably construct a three-dimensional model of the subsurface using two-dimensional images. Tying around data loops is also the best way to correlate from fault block to fault block. Otherwise, faults must be jumped using reflection character, sequence analysis, or additional well control.
After all lines are picked and tied, the results of the interpretation are then summarized and presented as maps. Basically, any observation that can be made using seismic data can be posted on a base map and mapped. Maps that are routinely made include
1. Time structure maps with faults 2. Depth structure maps 3. Seismic facies maps for reservoir, source, or seal
analysis 4. Seismic amplitude maps for DHI analysis 5. Thickness maps inferred from seismic tuning analysis 6. Fault plane maps 7. Fault plane maps with cross-fault sand juxtaposition for
seal analysis 8. Isocliron or isopach maps showing growth or thinning
in a stratigraphic interval 9. Seismic velocity maps for lithology determination or
depth conversion
In addition, many combinations of these maps can be made, such as seismic amplitude plotted on top of structure. The only limitations in constructing these maps are the imagination and skill of the interpreter (see the chapter on "Mapping" in Part 7).
The overall aim of seismic interpretation is to aid in constructing the most accurate earth model or reservoir description possible. This can best be accomplished when the seismic data are merged with petrophysical, geological, and engineering databases. While the process of interpreting seismic data is basically the same on paper or in a workstation environment, the workstation offers advantages in data management, manipulation, and display and it allows for a more convenient integration of other data types.
Mapping with Two-Dimensional Seismic Data
I. R. Gordon
I "'- l mocoj
11Сл Casper, Wyoming, U.S.A.
MAPPING TWO-WAY TIME
Most two-dimensional seismic reflection lines are presented in the format of horizontal distance versus twoway traveltime (time sections). Using interpreted time sections and a geographic base map, one can draft structure contour maps.
Preparation for Mapping
Before mapping, double-check that variations in the frequency, phase, and (especially) polarity of wavelets were accounted for during interpretation. Lines that were processed differently may appear to correlate well, even though they actually do not (Tucker and Yorston, 1973). Also, make sure that the interpreted events tie at the intersections between lines. You may construct a map perfectly, but if the lines are misinterpreted, your map will be misleading (see the chapter on "Seismic Interpretation" in Part 7).
The basemap should have clearly legible line and shotpoint numbers. The shotpoints should be printed at intervals that are smaller than the wavelength of the structures that will be mapped. The base map should be printed accurately, but sometimes mistakes occur. Remember that the accuracy of the base map partly determines the accuracy of your geological interpretation. If you question the accuracy of any part of the base map, consult with the person who drafted the map or check the field records yourself.
Recording Times of Events
Begin mapping by recording the two-way times of the event being mapped for every shotpoint along each line and for all Une intersections. The precision of time picks will vary with the scales of lines, but be as precise as possible. The numbers that are recorded are raw times. It is advisable that the times be recorded on a rough draft base map and in tabular form.
In some cases, the distances between the shotpoints printed on the lines are too large for resolving relatively small structures. In this case, it might be necessary to interpolate a ground position and then record a time. However, be aware that a relative position between two shotpoints on a Hne may not correspond to the same relative position between the same two shotpoints on the basemap. In short, the only ground positions that you can locate accurately on the base map are those that are printed on the base map. If highly accurate intermediate ground positions are important for your interpretation, refer to the person who drafted the map for help or check the field records yourself.
Misties
Usually, raw times do not match perfectly at the intersections between two lines. These differences in time are known as misties, and can range in magnitude from 1 msec to tenths of 1 sec. Misties often result between lines that are processed differently, particularly when different datums are chosen for the statics routine. Small misties may also result because normal migration routines cannot compensate perfectly for the geometry of dipping events. Commonly, misties result from "end-of-line" effect, where one line crosses the poorly migrated end of another line. Misties can also arise from errors in the interpretation or in the reading of two-way time.
Misties are corrected mathematicaUy. Differences of a few milliseconds are normally small enough that they can be averaged. However, when the differences approach the magnitude of the structures that you want to resolve, they must be treated by other methods. Averaging large misties ("splitting the difference") is ill-advised because doing so may introduce nonexistent structures to your map (Figure 1).
In general, adjusting for large misties in a group of lines starts by identifying a subgroup with very small misties. This subgroup is used as a base to which the times of events on all other Hnes are adjusted (Figure 2). Adjustments are made by shifting all the times of all the events on a given line by the same amount. Properly honoring the data requires that, for a given line, all the times for every event be adjusted by a constant value (Figure 2). For example, do not adjust one event up 5 msec and another 15 msec on the same line; this would introduce more time (i.e., depth) between the two events.
The end result after adjusting for misties will not be perfect. In cases where scores of lines are being used, some of the final misties may actually be quite large. However, the object is to minimize the error, which is an attainable goal.
Contouring
After the misties are adjusted, the revised two-way times can be plotted on the final base map. Finally, the points can be contoured at an interval that you deem appropriate.
OTHER TYPES OF MAPS
Velocity Gradient Maps
A velocity gradient map is constructed at an intermediate step between a time map and a depth map. During conversion from time to depth, a velocity gradient map compensates for lateral changes in velocity, which is preferable to using a single velocity function (Figure 3).
381
382
PART 7—GEOPHYSICAL METHODS
LINE Q
O
N
.1
.2
.3
.4
.5
Figure 1. False structure (dashed lines) created when misties are averaged. Dots show times of events on seismic lines A, X, and N where those lines intersect line Q. Solid lines show true attitude of beds. If dashed events were mapped, false structure would appear.
(b)
.2 о 528-84 5(P26) с A561-80 5(g)
X +24 +7 (-59)
X
4* 0CI0O <Ю4 (-24) X X
(-81)
(-21)
(-7) X X
(-49) (-4)
+59 +81 +49
X (+53)
X +21 +4 (-53)
X
9200 ft./8
Figure 2. (a) Crude base map illustrating seismic line intersections, (b) Table showing misties at seismic line intersections (times in milliseconds). Circled lines constitute a group having small misties. A group can be used as a base to which times on all other lines are adjusted. For example, times on line A561-80 could be shifted down about 52 msec.
Average Velocity A = 8286 f t . / s Average Velocity B = 9700 • f t . / s
Figure 3. Illustration showing effect of lateral differences in velocity on conversion of time and depth. If a single velocity function of 8300 ft/sec were used, errors of +6 ft at A and -440 ft at B would appear on the depth map.
Construction requires a base map and velocity data. The object is to contour the average velocity down to an event.
Velocity data generally come from five sources: vertical seismic profiles (VSPs), checkshot surveys, synthetic seismograms, stacking velocities, and well depth to time correlations. The latter is an easy and reliable method of determining velocities. Simply match the time of a horizon on a seismic line with the depth of that horizon in an adjacent well and you can calculate a velocity. This method may not work in highly deformed rocks, in which one is unsure exactly what the two-dimensional seismic line is imaging. However, depth to time correlations generally work well.
Vertical seismic profiles and checkshot surveys are also excellent sources because they show the actual traveltime of
sound through material over a known distance, thereby yielding true velocities. Good estimates of velocities are provided by synthetic seismograms. Synthetics are made from sonic logs and show the cumulative travel time through the rocks where a sonic log was run. Knowing the cumulative travel time for a given depth, one can calculate a velocity (distance divided by time).
Approximate velocities can be calculated using the stacking velocities that were picked during processing (Dix, 1955) (Figure 4). This is the poorest source of velocity information, but it may be the only source in areas where no wells have been drilled. Stacking velocities are usually printed at the top of each seismic line. Use the nearest shotpoint printed under the stacking velocities as the "ground position" for your calculated average velocities. Keep in mind that stacking velocities are not true velocities; they are just the velocities that the processor interpreted as the best at tuning events during processing. Occasionally, these velocities can vary from true velocities by more than 20%. However, they generally approximate the root mean square velocities from which average velocities can be calculated.
Depth Maps
Basically, depth maps are constructed by multiplying a one-way time map by a velocity gradient map. Hence, the times on a two-way map must be halved before this
Source
Receiver
(a)
Mapping with Two-Dimensional Seismic Data
383
V|b = Interval velocity of interval b Va ,Vb = Stacking velocities of intervals b and a ta.tb = Two-way travel times down to interface
a and b. (b)
Figure 4. (a) Illustration of ray paths, intervals, and interfaces used to help explain the Dix formula, (b) The Dix formula for calculating interval velocities, which assumes that interfaces are flat and smooth.
Figure 5. Velocity map with velocities marked at grid intersections.
calculation takes place. A basic map can be made by simply multiplying one-way time by velocity at every point where velocity data exist and then contouring the products. A much better result is obtained by gridding both the time and velocity maps (Figure 5) and then multiplying the time by the velocity at each grid point. Both grids can be constructed by interpolating values between points where data are available. After multiplying time by velocity at each grid point, contour the products. The gridding method is easily done with a computer and mapping software.
Time Interval Maps
Time interval (or isotime or isochron) maps are commonly used for interpreting changes in thickness between interpreted horizons (Figure 6). To map time intervals, calculate the difference in time (normally two-way time) between two events at each shotpoint and contour the resultant values.
Time Slice Maps
A time slice map shows geology in the horizontal plane at a given time (normally two-way time) (Figure 6). In essence, making a time slice map is analogous to making geological maps from cross sections. As a first step toward construction, imagine that all the data above your chosen time disappear
and that your base map lies directly on top of the chosen time. Next, locate each intersection between an interpreted event and the chosen time and plot that intersection at the ground position immediately above. Finally, link the points for each respective event in a manner that is consistent with your knowledge of the structure on the lines.
COMPUTER-AIDED MAPPING
Interpretations derived from seismic lines can be mapped efficiently with the help of a computer workstation, which performs repetitive calculations very quickly. First, the interpretations must be entered into a workstation, either by interactive (on-screen) interpretation or by digitizing interpretations that exist on printed lines. Mistie corrections are then performed by the computer. Two-way times and time intervals can then be posted on a base map. Contouring time and time interval maps can be done by the workstation, but the result usually requires some hand editing (see the chapter on "Overview of Development Geology Workstation" in Part 8).
Constructing depth maps is possible once hand-drawn time and velocity maps are digitized into the workstation. The computer grids and multiplies the time and velocity maps, and the resultant values can then be contoured.
384
PART 7—GEOPHYSICAL
METHODS
Figure 6 (a) Block diagram showing the time that is mapped for a time slice map. (b) Interval that is mapped on time interval map. (c) Time interval map.
Three-Dimensional Seismic Method
Oz Yilmaz1
Landmark EAME Services Group Weybridge, Surrey, U.K.
INTRODUCTION
Subsurface geological features of interest in hydrocarbon exploration are three-dimensional (3-D) in nature. Examples are salt diapirs, overthrust and folded belts, major unconformities, reefs, and deltaic sands. A two-dimensional (2-D) seismic section is a cross section of a 3-D seismic response. Despite the fact that a 2-D section contains signal from all directions, including out-of-plane of the profile, 2-D migration normally assumes that all of the signal comes from the plane of the profile itself. Although out-of-plane seismic signals (sideswipes) are often recognizable by the experienced seismic interpreter, the sideswipe signal causes 2-D migrated sections to mistie. These misties are due to inadequate imaging of the subsurface resulting from the use of 2-D rather than 3-D migration (French, 1974) (see the chapter on "Mapping with Two-Dimensional Seismic Lines" in Part 7).
processing, traces are collected as common-cell gathers (bins). These gathers are used in velocity analysis, and common-cell stacks are generated. Typical cell sizes are 25 by 25 m for land surveys and 12.5 by 37.5 m for marine surveys.
Conventional 3-D recording geometries often complicate the process of stacking the data in a common-cell gather. Cable feathering in marine 3-D surveys can result in traveltime deviations from a single hyperbolic moveout within a common-cell gather. For land 3-D surveys, azimuthdependent moveout within a common cell gather is an issue.
After stacking, the 3-D data volume is sometimes (but not always) migrated in two stages. First, a 2-D migration is applied along the in-line or cross-line direction. Then the data are sorted, and a second pass of 2-D migration is applied along the orthogonal direction. Before the second pass of migration, the data sometimes need to be trace interpolated along the cross-line direction to avoid spatial aliasing.
PERFORMING 3-D SURVEYS
A typical marine 3-D survey is carried out by shooting closely spaced parallel lines (line shooting). A typical land or shallow water 3-D survey is done by laying out a number of receiver lines parallel to one another and placing the shot points in the perpendicular direction (swath shooting). Other recording geometries have also been used in acquiring 3-D data. Shooting in circles has been done in the Gulf of Mexico to delineate salt domes. Shooting around a lake or a topographic high to achieve subsurface coverage under the surface obstacle has also been tried.
In marine 3-D surveys, the shooting direction (boat track) is considered to be the in-line direction, whereas in land 3-D surveys, the receiver cable is along the in-line direction. The direction that is perpendicular to the in-line direction in a 3-D survey is called the cross-line direction. In contrast to 2-D surveys in which line spacing can be as much as 1 km, the line spacing in 3-D surveys can be 50 m or less. This dense coverage requires an accurate knowledge of shot and receiver locations.
The size of the survey area is dictated by the areal extent of the subsurface target zone and the aperture size required for adequate imaging of that target zone (Yilmaz, 1987). This imaging requirement means that the areal extent of a 3-D survey almost always is larger than the areal extent of the objective. A few hundred thousand to a few hundred million traces normally are collected during a 3-D survey.
PROCESSING OF 3-D DATA
The basic principles of 2-D seismic data processing still apply to 3-D processing. In 2-D processing, traces are collected as common midpoint (CMP) gathers, while in 3-D
3-D VERSUS 2-D MIGRATION
Three-dimensional migration often produces surprisingly different sections from 2-D migrated sections (see the chapter on "Seismic Migration" in Part 7). The example in Figure 1 shows a no reflection zone on the 2-D migrated section, while the same zone contains a series of continuous reflections on the 3-D migrated section that are easily correlated with reflections outside that zone. When we do 2-D migration, we confine the movement of the energy into the plane of the line itself. So the energy contained in the unmigrated stacked section in Figure la is indeed the same as the energy contained in the 2-D migrated section in Figure lb, except that it has been moved somewhere else on the section. As a result of moving the energy during 3-D migration within the 3-D volume, some energy moved into the section (Figure lc) from others and some moved out of the section and migrated into the others.
From the field data example, we see that 3-D migration provides complete imaging of the 3-D subsurface geology. Specifically, 2-D migration cannot adequately image the subsurface and introduces misties between 2-D lines in the presence of dipping events. However, 3-D migration eliminates these misties by completing the imaging process.
The difference between 2-D and 3-D seismic methods is the way in which migration is performed. Just having a dense coverage on top of a target zone, for example, a 25-m in-line trace spacing and a 25-m cross-line trace spacing, will not necessarily provide adequate subsurface imaging unless migration is performed in a 3-D sense. But one does need the dense coverage to perform the 3-D migration accurately. Also, to be able to map small subsurface features, sufficiently close spatial sampling is needed.
1Formerly with Western-Atlas International.
385
386
PART 7—GEOPHYSICAL METHODS
3-D INTERPRETATION
The 3-D data volume is available to the interpreter as vertical sections in both the in-line and cross-line directions and as horizontal sections (time slices). The time slices allow the interpreter to generate contour maps for marker horizons (Figure 2). The interactive environment provides an effective and efficient means for interpretation of the sheer volume of 3-D migrated seismic data. Fault correlations, horizon tracking, horizon flattening, and some image processing techniques can be adapted to the interactive environment to help improve interpretation. Since the 3-D volume provides detailed constraints on interpretational decisions, drilling programs based on 3-D seismic usually have high success rates.
Figure 1. (a) A CMP-stacked section from a marine 3-D survey, (b) The corresponding 2-D migrated section, (c) The 3-D migrated section. (Data courtesy of Amoco Europe and West Africa, Inc.)
Three-Dimensional Seismic Method
Figure 2. (a) Selected time slices from a marine 3-D survey and (b) a time-structure map of a marker horizon derived from the 3-D volume of migrated data. (Data courtesy of Western Geophysical, Division of Western-Atlas International.)
Vertical and Lateral Seismic Resolution and Attenuation
R. E. Sheriff
University of Houston Geosciences Department Houston, Texas, U.S.A.
VERTICAL RESOLUTION
Resolution is the ability to distinguish between objects, that is, to see a second object in the presence of another. Concerning seismic data, vertical resolution relates to how far apart two interfaces must be to distinguish separate reflections from them or how thick a bed must be to allow distinguishable reflections from the bed's top and bottom. The length (in time) of seismic wavelets produces confusion because successive reflections overlap, so it is desirable that the source wavelet (Figure 1) be short. The wavelet should have a distinctive, sharp peak to be timed. Side lobes are undesirable because they can be mistaken for true reflections. They add to confusion because of interference. The wavelet that is naturally embedded in the data can be replaced in data processing by a different more desirable wavelet to facilitate interpretation, a procedure usually called wavelet processing. To achieve the desired characteristics, the embedded wavelet (Figure 2) needs a broad spectral range (1.5 octaves) including high frequency content, and the phase response should be zero phase. Data quality, the mode of display, and the interpreter's experience are also clearly factors in being able to resolve reflections.
Assuming that the three foregoing conditions have been met, let's examine a simple model of a very gentle wedge for which the physical properties above and below the wedge are the same (Figure 3). This might represent the pinchout of a sand body embedded in shale or any other combination of lithologies. When the wedge is thick, the reflections from the top and base of the wedge are distinct, and timing of their separation gives the wedge thickness (Figure 4). As side lobes begin to interfere, amplitude begins to change and minor timing errors result. As the wedge thickness approaches onefourth of a wavelength, constructive interference produces an amplitude buildup known as the tuning effect. The finite frequency of the wavelet will not let the peaks and troughs
representing the two reflections come closer, that is, the time separation becomes constant and it is no longer a measure of the wedge thickness. The tuning thickness is usually taken as the limit of resolution or limit of separability. Since wavelength equals velocity divided by frequency, this one-fourth wavelength represents about 25 ft for moderately shallow beds having a velocity of about 6000 ft/sec for a wavelet with 60-Hz dominant frequency, or about 250 ft for a deep bed with 15,000 ft/sec velocity and a 15-Hz wavelet. Resolution decreases rapidly with depth because velocity increases and frequency decreases.
HORIZONTAL RESOLUTION
The waves giving rise to a reflection event are reflected from a fairly large, roughly circular area of the reflecting interface known as the first Fresnel zone. Reflections from this zone arrive at a geophone so as to constructively interfere. The radius of this zone is often taken as the horizontal resolution for unmigrated seismic data. (A nomogram for Fresnel zone calculation is given by Sheriff, 1980.) For the 60Hz wavelet at 6000 ft/sec previously mentioned, this radius would be about 500 ft at a depth of 5000 ft. As with the wavelength, the Fresnel zone size also increases rapidly with depth.
Seismic migration effectively collapses the in-line aspect of the Fresnel zone, so this measure of resolution is not appropriate with the migrated seismic data that are usually used for interpretation (see the chapter on "Seismic Migration" in Part 7). In principle, after migration, horizontal resolution is reduced to trace spacing. However, migration
^DISTINCTIVE, SHARP POINT
*
4 f\
TO BE TIMED J MINIMUM SIDELOBES
V \/
^^
MINIMUM LENGTH Figure 1. Desired wavelet for positive reflection.
Figure 2. A1.5-octave zero-phase wavelet passing frequencies between 14 and 65 Hz that passes 100% between 22 and 55 Hz (top), and an embedded wavelet actually achieved (bottom).
388
BED THICKNESS IN WAVELENGTHS
Figure 3. Seismic response of a wedge pinchout.
smears out noise, including contributions from events to the side of the line, and this often becomes the limiting factor. Other factors contributing to limit the horizontal resolution include uncertainty in the velocity and the mathematical completeness of the migration algorithm. One of the most important limitations results from the spatial sampling because of the geophone group interval. At least two samples per wavelength are required by sampling theory to prevent aliasing. An aliased signal appears in the seismic record at lower frequency and/or dip than is real. This yields an equation relating the geophone group interval s to the frequency / and the maximum reflection dip d (in milliseconds per unit of distance) that can be properly imaged:
d = 1000/sf
For a 300-ft group interval and a 50-Hz frequency, dips greater than 0.07 msec/ft are aliased (or I l 0 for a velocity of 6000 ft/sec). Other considerations are also involved, but clearly many situations require shorter geophone group intervals than commonly used.
ATTENUATION
The amplitude of seismic waves decreases with seismic traveltime (or depth) because of a number of factors (Sheriff,
Vertical and Lateral Seismic Resolution and Attenuation
389
200%
2T
У
ш Q
I-
o.
S < O 100%
Ш
N
< OоSс
Z
о
о
A/4
A/2
A/4
A
BED THICKNESS IN WAVELENGTHS
Figure 4. Amplitude (solid line) and time thickness (dashed for observed, dotted for actual) for a 1.5-octave wavelet reflected from the wedge.
1975). Seismic processing endeavors to correct for these factors so that amplitude is proportional to reflectivity. However, enough information about all the amplitudechanging factors is never available, so relative amplitude preservation means that gain changes slowly in the same way for every trace. The objective is to preserve the lateral significance of amplitudes so that as one follows a reflection event, changes in amplitude will be proportional to changes in reflectivity. However, the amplitudes of different events may not be proportional to their respective reflectivities.
The earth filters out higher frequencies as a seismic wave travels. This causes a lowering of frequency content with time on seismic records, which in turn causes changes in wave shape. Thus, shallow reflections often have dominant frequencies around 60 Hz, whereas 15 to 20 Hz is usually the dominant frequency for deep reflections. Q-compensation is sometimes applied in processing to compensate partially for the loss of high frequencies with time.
Synthetic Seismograms
KanRekosb
David Hicks
Sierra Geophysics Inc. Kirkland, Washington , U.S.A.
INTRODUCTION
The more control the geoscientist has in mapping the subsurface, the greater the accuracy of the maps. Control can be increased by the correlation of seismic data with borehole data. The synthetic seismogram (often called simply the "synthetic") is the primary means of obtaining this correlation.
Velocity data from the sonic log (and the density log, if available) are used to create a synthetic seismic trace. This trace closely approximates a trace from a seismic line that passes close to the well in which the logs were acquired. The synthetic then correlates with both the seismic data and the well log from which it was generated.
CALCULATING A SYNTHETIC SEISMOGRAM
The calculation of a synthetic seismogram generally follows these steps:
1. The sonic and density curves are digitized at a sample interval of 0.5 to 1 ft. (If the density curve is not available, the sonic alone may be used.)
2. A computer program computes the acoustic impedance log from the sonic velocities and the density data. The data are often averaged or "blocked" to larger sample intervals to reduce computation time and to smooth them without abasing the log values.
3. The resulting acoustic impedance curve is then used by the program to compute reflection coefficients at each interface between contrasting velocities.
4. A wavelet is chosen that has a frequency response and band width similar to that of the nearby seismic data. The synthetic wavelet is convolved with the reflection series for the entire well survey and generates a synthetic seismic trace. A potential pitfall in synthetic
generation is using a wavelet of fixed frequency over the entire survey. Care should be taken to choose a wavelet whose frequency is similar to a key interval of the seismic data to which it will be compared.
The resulting trace is displayed at the same vertical scale as the seismic section for direct comparison. To improve the match with the seismic data, the synthetic seismic trace can be recomputed using different wavelets and filters. Figure 1 shows an example of a synthetic seismogram and associated well log data used in its generation. Different wavelets have been convolved to produce two additional synthetic seismogram displays.
The synthetic trace can now be compared to a trace from the seismic line. This is commonly done by laying the synthetic directly on top of the appropriate seismic trace and adjusting the synthetic vertically until the two coincide. Through a trial-and-error process, the interpreter determines at what point the synthetic trace "best fits" the seismic data.
In an ideal world, there is an obvious agreement between the seismic line trace and the synthetic seismogram. A formation top or other correlation marker on the well log can then be tied to the corresponding seismic horizon with relative ease. In the real world, however, the interpreter may be, and often is, confronted with a synthetic trace and seismic data trace that bear little resemblance to each other. Variations in the quality of the well log data can have a major impact on the final synthetic display. A sonic log that was generated from a borehole containing numerous washed out zones will contain areas of unrealistic velocities. Careful editing of the well log data can help to smooth spurious data readings and generate a more realistic synthetic. Care should be taken, however, whenever well log data are edited. For more details about synthetic seismograms, the interested reader can refer to Seismic Exploration Fundamentals (Coffen, 1978).
390
Synthetic Seismograms
391
Depth Time Density Sonic Velocity Acoustic
RC Raw Seq Seq
Ft Sec G/CC US/Ft Ft/S Impedanc Series Synthetic 1
2
Depth M
OСПO TCO-; CTOTCiqSrJ-O; g_ c—g
T—' c\i c \ i o J c \ i CNi г - о
TOOОo-
CoОOOMОoOOr-CОoOOMoOOOCOoOOO^ "oC4iot/-—:oC,pSooJ*T+o'-.
*
IfiifiPfiPifiif -305
MUUISlillll -610 • 1 1 1 IiilllIlli -914
fHsil f l^Pll -1219
-1524
IiifiiiItiM ЗЗДДОЯ
«3$$ -1829
щщтщпт
IlliiliiiiIiIiIlf
-2134 -2438
-2743
Well Tops
Sand 1
Sand 3 Reef Unconformity Gas/Water Top Delta BTM Delta Channel Top Salt
Figure 1. Synthetic seismogram and associated well log data.
Forward Modeling of Seismic Data
George Mellman Paul A. Kunzinger
Sierra Geophysics Inc. Kirkland, Washington, U.S.A.
INTRODUCTION
Forward numerical modeling of seismic data is the use of geological models of the earth to simulate seismic field experiments. Models can be one, two, or three dimensional and consist of depth horizons and associated P wave velocities, S wave velocities, and densities
The most common use of forward modeling is for verification of structural and stratigraphic interpretations. For example, synthetic seismic sections derived from forward modeling can be compared to stacked sections to verify the original interpretation. If needed, the original interpretation or model is altered and the process repeated until a desired correlation between the actual data and modeled results is observed (see the chapter on "Synthetic Seismograms" in Part 7).
An example of the use of modeling to aid in stratigraphic interpretation is the use of amplitudes of seismic data to infer the thickness of a thin sand layer. If layer thickness is less than a wavelength, variations and thickness appear as tuning effects, with systematic changes in amplitude and wave shape correlating directly with sand thickness. Producing forward models with varying sand thickness using well data to calibrate, and comparing these synthetics with observed seismic sections, provides a means of accurately determining sand thickness where well data are sparse or absent.
Another use of forward modeling is the determination of the seismic expression of expected features of the geology. Figures 1 and 2 illustrate the use of seismic modeling to determine if porosity in a carbonate reef interval is detectable on seismic data. Figure 1 is a strike cross section through a 3D geological model representing a carbonate reef play. Reef structures are located in the interval between the depths of 2500 and 3500 ft. The structure located on the left side of the cross section contains porosity, while the structure on the right is tight. Porosity is represented in the model by a lower interval velocity and lower density than the surrounding rock (see the chapter on "Porosity" in Part 5).
This study included the generation of two synthetic seismic sections using 3-D ray tracing techniques. One section represents a synthetic stacked seismic section (created using normal incidence ray tracing), while the second represents a synthetic migrated section (created using image ray tracing). The objective of modeling is to determine if there is a noticeable difference between the seismic signature of the tight carbonate and that of the porous carbonate. The synthetic migrated section is illustrated in Figure 2. The positive event at traces 20 through 26 and between 325 and 375 msec represents the top of the tight anticlinal structure. There is a distinct character change of this seismic event corresponding to a change in porosity in the geological model. The porosity is identified by the dim spot on traces 6
through 11 and the velocity pulldown of the underlying event. The results from this modeling indicate that seismic data could be used to identify porosity in this geological setting. A good description of modeling methods and a number of case study examples can be found in Fagin (1991).
METHODS OF FORWARD MODELING
A number of forward modeling methods are available, and the choice of method generally depends on a tradeoff between the accuracy necessary and the desired computing time. In general, the type of data to be modeled, the complexity of the model, and the aspects of the data that need to be accurately modeled dictate the method that should be used.
While field experiments always produce shot gathers, there are approximations available in numerical modeling that produce other data types directly. To simulate stacked data, for example, an exploding reflector model is often used. Virtual explosive sources are placed along all reflecting interfaces, with source strength proportional to the normal incidence reflection coefficient at that interface. This method produces accurate traveltimes and good zero offset amplitudes for geometric arrivals. Diffraction amplitudes are not accurate in this approximation, and multiples are not present. However, a complete stacked section can be generated in about the time it takes to generate a single shot gather. This efficiency makes the exploding reflector model the method of choice for most stacked section simulations. Another approximation, called image rays (Hubrel, 1977), makes it possible to simulate time-migrated data directly, again with approximations, but with significant time savings.
There are two classes of seismic modeling: ray tracing and wave equation methods. Implementation of both classes exists for one, two, and three dimensions; shot gather; common midpoint (CMP) gather; and stacked data simulation.
Ray theory uses the fact that energy in the form of rays travels along minimum time paths in the model. As in optics, rays bend when velocities change, obeying Snell's law, and are partially reflected when velocity or density discontinuities are encountered. Traveltimes of reflected arrivals correspond to the times of the minimum time paths, while amplitudes are a combination of geometric spreading and reflection coefficient. The reflection coefficient, based on the Zoeppritz equations for elastic media, depends on the velocities and densities on both sides of the interface, as well as on the incidence angle of the ray.
Ray methods usually give very accurate traveltimes and accurate amplitudes for geometric arrivals if the model is sufficiently smooth. These methods are efficient, and computing time is low to moderate. Diffractions and multiple
392
Forward Modeling of Seismic Data
393
DISTANCE: KILGFEET 0. 0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0
VELOCITY F/SEC 15000.
17000.
19000..
19894. 22000.
VflR1 Il
25000.
Figure 1. A strike cross section ot a carbonate reef play. The reef structure on the left contains porosity, while the reef structure on the right is tight.
CMP N 0 .
21
26
31
200
O LLJ 300 CO z:
400
Figure 2. Synthetic seismograms for the model in Figure 1. By synthetic modeling of a migrated section, the expected seismic signatures of reefs containing porosity and tight reefs have been obtained.
394
PART 7—GEOPHYSICAL METHODS
reflections can be added, though at high computing times and never with complete accuracy. Ray methods are the methods of choice for many structural verification problems and are quite useful for stratigraphic problems if the interfaces are reasonably smoothly varying.
Wave equation methods solve the propagation problem over the entire model, rather than performing local solutions as in ray methods. Two commonly used wave methods are the Kirkhhoff (Hilterman, 1970) method and the finite difference method (Kelley et al., 1976). Kirkhhoff methods use a local approximation to obtain a reflected wave field at an interface. These methods can provide accurate traveltimes for geometric and diffracted arrivals and accurate amplitudes for geometric arrivals, even in complex media. Diffraction amplitudes are usually not very accurate for these methods, and multiples are usually absent. Computing time is moderate.
Finite difference methods can provide very accurate results even in the most complex media. These methods provide
exact numerical solutions and can include all wave phenomena, such as diffractions, multiples, and ground roll. The only Umitations on finite difference methods are imposed by computing time, which is high in two dimensions and very high in three dimensions. This effectively limits grid size and hence the resolution obtainable. However, these are still the methods of choice for highly faulted complex models for which ampUtude accuracy is important.
Often, the largest limiting factor on the accuracy of a forward model is not the method chosen to solve the problem, but rather the inappropriate use of one- or two-dimensional models of the geology when two- or three-dimensional models are necessary. The controlling factor for the model dimension is not the dimension of the data but rather the complexity of the geology. The presence of appreciable dip requires the use of two- or three-dimensional models. The presence of out-of-plane structures or dip requires the use of three-dimensional modeling. Failure to use models of the correct dimension can give rise to appreciable traveltime and amplitude error, as well as missing or misinterpreted arrivals.
Seismic Inversion
Brian Russell
Hampson-Russell Software Services Ltd. Calgary, Alberta, Canada
INTRODUCTION
The object of seismic inversion is to convert the seismic interpreter's view of the earth reflections as a function of time to the geologist's view of the earth velocity as a function of depth. This is not as easy as it may seem. The reason for this is that seismic data measurements are taken at the earth's surface and involve sending a sound pulse through the earth and recording the echoes from each reflecting interface. The farther this sound pulse has to travel, the more the pulse is distorted and the less information it is able to carry back to the surface. Obviously, drilling a well and running a set of logging tools gives us much more information. However, the advantage of the seismic method is that coverage can be made over large areas of the earth's surface. This is especially true of the large three-dimensional surveys that are now routinely being acquired. For this reason, seismic inversion is an important processing tool.
SEISMIC INVERSION METHODS
As just mentioned, seismic data give us a "fuzzy" picture of the subsurface over a large area, whereas wells give us detailed geological information at a few points. This suggests a general approach to the seismic inversion method, as shown in Figure 1. On the left side of this flowchart, the input is our best reflectivity estimate from the seismic data. On the right side, we introduce geological constraints taken from a sonic log or seismic velocity information and produce a forward model. We then combine the information from the seismic data with controlling information from the seismic model and produce a final inversion result.
There are two main types of inversion currently being used. The first is band-limited inversion, which involves directly integrating the seismic trace. Since the seismic trace lacks a low frequency velocity trend because of the bandlimited wavelet, the inverted trace lacks the "excursions" seen on the original sonic log. We must therefore add in the low frequency component from the geological model. The result is a frequency band-limited version of the original sonic logs. This is shown in Figure 2 for a carbonate reef example. The heavy trace in the center is a sonic log from the producing well. The zone of interest is at 1150 msec.
The second type of inversion, which is more recent than the band-limited method, involves producing a "blocky" output rather than a band-limited output. There are several methods that produce this type of output, and they are
s o m e t i m e s referred to as sparse-spike or model-based methods.
These methods work by producing a forward model that best reproduces the seismic data when converted to synthetic form (that is, when the reflection coefficients are convolved with the wavelet). This method involves starting with a simple "guess" of this model and changing this guess iteratively until the error between the model and the observed seismic data is minimized (see the chapter on "Forward Modeling of Seismic Data" in Part 7). The results of doing such a model-based inversion are shown in Figure 3 for the same traces shown in Figure 2. Notice that the carbonate reef is visible, and looks like the blocked version of the log.
LOOl
Figure 1. A flowchart showing the general concept behind seismic inversion.
Figure 2. Band-limited inversion of several traces around a reef. The integrated sonic log from the reef well is shown as the dark trace in the central part of the figure.
395
396
PART 7—GEOPHYSICAL METHODS
Figure 3. Model-based inversion of several traces from around a reef. The integrated sonic log from the reef well is shown as the dark trace in the central part of the figure.
THREE-DIMENSIONAL CASE STUDY
As an example of inversion applied to real data, let's consider a 3-D case study that was done in the Taber area of southern Alberta, Canada, by Western Geophysical. Figure 4 shows a schematic interpretation of the geology of a river channel zone. The area of interest is the Glauconitic Formation of the lower Cretaceous upper Mannville Group, which is characterized by rapidly changing lithological facies. The reservoirs are sandstones with porosities of 15% surrounded by impermeable siltstone.
A 3-D survey measuring 2.2 by 1.7 km was acquired, as indicated by the rectangle in Figure 4. These data were then processed through to final migrated stack, and the result was inverted using the SLIM inversion method of Western Geophysical. This approach is similar to the model-based algorithm discussed earlier. The results of inverting one particular line from the 3-D volume are shown in Figure 5. Two producing wells and one dry hole are shown in this line.
Figure 4. A schematic diagram of a river channel zone from the Taber area of southern Alberta. The 3-D survey was done in the area outlined by the rectangle in the center of the map. (Courtesy of Western Geophysical.)
The producing interval is between 640 and 660 msec on the two producers, whereas the low velocity zone on the dry hole comes from a sand that is porous but not prospective. Figure 6 shows a series of time slices (that is, horizontal cuts through the 3-D data volume at constant time intervals) over the complete dataset (see the chapter on "Mapping with TwoDimensional Seismic Lines" in Part 7). Borehole B from Figure 5, which was the dry hole, is indicated by the faint white crosshairs at the centers of the graphs in Figure 6. The lightly shaded areas show areas of low velocity material at the zone of interest. Notice that this well is obviously mispositioned with respect to the low velocity reservoir material.
CONCLUSIONS
In this brief summary of inversion methods, we have considered two different approaches: band-limited and model-based. Band-limited inversion is a robust method that produces a smooth and continuous output. Model-based inversion produces a blocky output that contains more geological detail than band-limited inversion. However, both methods can be useful in the delineation of hydrocarbon reservoirs.
Seismic
Migration
397
Figure 5. The model-based inversion of the line through the river channel system of Figure 4. (Courtesy of Western Geophysical.)
Figure 6. Time slice displays through the 3-D survey of the river channel zone. The lightly shaded areas show low velocity material in which reservoir sands are present. (Courtesy of Western Geophysical.)
Amplitude Versus Offset (AVO) Analysis
Allan T. Long R. C. Anderson
Berrong Enterprises Ltd. Houston, Texas, U.S.A.
COMPONENTS FOR AVO ANALYSIS
The amplitude versus offset (AVO) phenomena on seismic data can provide substantial exploration and development information. Under good conditions the information extracted can be as detailed as an elastic layered model of the earth in the vicinity of the exploration or development target. To accomplish this goal, a systems approach to the problem is used. Major components of such a system analysis procedure are as follows:
1. AVO feasibility studies—Ascertain the general applicability of this technique to the problem at hand.
2. Proper AVO data processing—Emphasize maintaining and enhancing the proper amplitude relationships.
3. AVO scanning—Find anomalous regions of amplitude with offset on seismic data.
4. Detailed AVO modeling supported by well log data— Provide base information upon which to judge and interpret the AVO study using well log data in the vicinity of an exploration target.
5. Elastic model inversion at prospective locations—Identify potential reservoirs and describe them numerically prior to drilling.
Readers interested in additional information on AVO analysis should begin by consulting the work of Hilterman (1990), Neidell (1986), and Ostrander (1984).
1-
O<L>
-u
2
1 <' 1Oс
оtS: 1-,
0 •
/
/ у•
1-
У У У
O
(a)
-»5
1-
(b) 45
о
ч
(C)
45
O
(d) 45
— acoustic term — shear term
Angle of Incidence (degrees)
Figure 1. Possible combinations of the acoustic and shear terms.
AVO FEASIBILITY STUDIES
Whether or not there is an application for amplitude versus offset analysis in a given exploration setting is an often posed problem. The explorationist should ask several questions. Are the AVO phenomena that may be present large enough to overcome the various noises in the system? Are there sufficient differences between geological conditions and seismic expressions to distinguish between them?
A reflection coefficient can be thought of as consisting of the sum of two components—an acoustic term and a shear term. The acoustic term depends only on the impedance of the two layers and the angle of incidence. The shear term depends on Poisson's ratio in the two layers and also on the angle of incidence.
An AVO response not only occurs from an interface characterized by a contrast in Poisson's ratio, but also from interfaces that are entirely acoustic—that is, the measured response is more than a simple contrast in Poisson's ratio.
Consequently, four response cases are possible, as shown in Figure 1. Two of these cases—(a) and (c)—represent constructive summation of the two terms, yielding a large AVO response. The other two cases—(b) and (d)—are destructive, and the total AVO response is small. A third case consisting of little shear response and strong acoustic response may produce high AVO and may not be associated with a Poisson's ratio contrast.
In addition to the acoustic and shear components of the AVO response, there are the effects of AVO tuning from thin beds. In contrast to normal incidence tuning, AVO tuning varies with offset and gives rise to amplitude and waveform changes concurrently.
Feasibility modeling requires the following: (1) a depth-velocity model, (2) density and velocity trend curves, (3) porosity ranges, (4) fluid content variation, (5) Poisson's ratio trend curves, and (6) a geological description of potential reservoir and encompassing environments. Figures 2 and 3 show model examples consisting of two fluid types. Figure 2 is a brine sand model, and Figure 3 is a gas sand model. The sensitivity of the seismic expression can now be studied by changing model parameters. If the model studies indicate measurable AVO responses, then AVO processing and interpretation techniques can be undertaken at an early stage in the exploration effort.
AVO DATA PROCESSING
Seismic data processing for AVO requires that certain steps be applied to common depth point (CDP) gathers (the individual traces prior to stacking). Areas of importance include (1) generalized amplitude corrections, (2) signal to noise ratio improvement, (3) robust deconvolution, and (4)
398
Amplitude Versus Offset (AVO) Analysis
399
Figure 2. Brine sand model.
Figure 3. Gas sand model.
OFFSET
prestack migration prior to AVO analysis in structurally complex areas.
Generalized amplitude corrections are probably the most important aspect in extracting and restoring proper amplitude relationships in the data. Amplitude corrections must compensate for (1) irregular source strength, (2) source array effects, (3) inelastic attenuation, (4) transmission loss effects, (5) spherical divergence, (6) receiver array effects, (7) receiver sensitivity, and (8) receiver vertical directivity.
Surface-consistent amplitude corrections are also necessary when amplitude variations in the data may relate to source and receiver environment and not to the geology. These corrections are distributed among the surface terms of source, receiver, and offset, plus a single subsurface (CDP or geological) term.
Normal moveout (NMO) and residual statics corrections must be performed as accurately as possible, for any AVO analysis is degraded by inappropriate corrections. The recent use of interactive processing workstations is providing explorationists with the necessary control over the selection of correct stacking velocities (see Part 8).
Signal to noise ratio improvements must occur on individual traces in the CDP gathers. Robust deconvolution is important to preserve stability of the wavelet across all offsets. In the land data case, a surface-consistent deconvolution method is usually desirable. To further stabilize the wavelet, deconvolve the data to a desired target waveform with frequency cutoffs predetermined by knowledge of the input data's usable frequency range. Trace-dependent deconvolution, such as a spiking operator, can sometimes produce severe distortions in reflector waveform when the signal to noise ratio is low on some of the individual traces.
Multiple interference can render amplitude with offset analysis of the CDP gathers meaningless. Generally, multiples are more dominant in data from the marine environment, but can also be strong in land data. Multiples
OFF8ET (METER8)
FINAL CALIBRATED
GATHER
NOISE REDUCTION
1
NOISE REDUCTION
2
Figure 4. Calibrated NMO-corrected CDP gathers after relative amplitude processing and application of noise reduction methods designed for single traces.
are usually not visually evident in land data because of trace spacing and lower fold and will appear as strong random noise. But proper use of multiple removing programs, such as radon transform methods, can yield excellent results.
Figure 4 shows NMO-corrected CDP gathers that have been processed with the proper amplitude corrections and cascaded noise reduction methods. It is critical that noise be removed without smearing the signal or introducing artifacts, which often occurs with multichannel filtering techniques.
400
PART 7—GEOPHYSICAL METHODS
OFFSET (METERS)
OFF8ET (METERS)
Figure 5. Gas well AVO model, involving blocked well logs and calibrated (wavelet-processed) seismic data.
CALIBRATED GATHER
AVO MODEL
Figure 6. Comparison of calibrated seismic data and the gas well AVO model.
AVO SCANNING
The computer can efficiently measure the change in amplitude across the offsets of a CDP gather for each time sample and can produce a single number that is the best fit slope to that change in amplitude. This collection of numbers can be plotted like a seismic cross section to yield an amplitude versus offset gradient section. The gradient section can then be interpreted, and AVO anomalies can be correlated to the stack section, on which an interpreter has already identified key horizons and potential structural or stratigraphic trapping areas. The purpose of this step is to identify quickly regions of anomalous AVO behavior, which can then be studied further with detailed modeling and rock physics calibration.
AVO MODELING
By using seismic data and available well logs, an understanding of the origin and nature of AVO phenomena can be realized by detailed modeling. A time match between seismic and well data can be found, along with the seismic wavelet for modeling. Then, using estimates of Poisson's ratio (empirically derived), elastic model inversions are performed to estimate a more precise Poisson's ratio.
The result of this approach applied to a gas sand problem
is displayed in Figure 5. The AVO anomaly at 460 msec is the resultant response to this gas sand. An alternative display of calibrated CDP gathers and matched AVO model for the same gas sand problem is shown in the variable area plot in Figure 6.
AVO INVERSION
Amplitude versus offset anomalies at sites in the vicinity of previous detailed modeling can be inverted to obtain acoustic parameter information. Using the layered model from the previous step as the starting point for the inversion process, the final compressional velocity, density, and Poisson's ratio information can be found. This is then used to predict the probable geological circumstances at the inversion site.
A successful inversion of an AVO response provides the following:
1. Tlie origin of the anomaly, such as acoustic response, rigidity contrast, or AVO interference
2. The thickness of the pay zone 3. The value of Poisson's ratios near the anomaly 4. The lithology, porosity, and fluid content of a given
layer
Checkshots and Vertical Seismic Profiles
B. A. Hardage
Bureau of Economic Geology The University of Texas Austin, Texas, U.S.A.
CHECKSHOT SURVEYS Surface-recorded seismic data often comprise the largest
database that must be dealt with in reservoir development. However, seismic data have one shortcoming that can limit their usefulness—the reflection events used to map the seismic sequences and the seismic facies that describe the areal and vertical distributions of reservoir and sealing units are measured as functions of seismic traveltime, not as functions of depth. To understand reservoir performance, the boundaries of these units need to be mapped in depth. Thus, the concept of the velocity checkshot survey has been developed to establish time-depth calibration functions at control wells so that surface-recorded seismic images can be reliably converted to the depth images that are needed to do reservoir volumetric calculations.
CHECKSHOT SOURCE-RECEIVER GEOMETRY The purpose of a velocity survey is to produce a down-
going seismic wavelet at the surface near a well and then to measure the time required for that wavelet to travel to a
Figure 1. The source-receiver geometry commonly used in onshore checkshot surveys.
known depth where a seismic receiver is positioned in the well. This borehole receiver is locked successively at several different depth levels, and the vertical traveltime to each level is measured (Anstey and Geyer, 1987). Each measurement of the source-receiver traveltime is a checkshot, and the compilation of all of the traveltime measurements into a time-depth calibration function is referred to as a checkshot
survey.
The source-receiver geometry used in onshore velocity checkshots is shown in Figure 1. If possible, the energy source should be the same as that used to record the surface seismic data near the well. A buried explosive charge is shown in this diagram, but other common onshore energy sources include Vibroseis or air guns operated in a waterfilled pit near a well. Offshore, essentially all checkshot surveys involve air guns as the seismic energy source.
Ordinarily, the borehole receiver is first lowered to the deepest checkshot level, and the traveltime to this deepest receiver position is measured for one or more surface shots. The receiver is then moved upward a distance of 200, 500, or 1000 ft (61,152, or 305 m) to record the checkshot, or vertical traveltime, at successively shallower levels.
The time-depth calibration function and velocity analyses that can be calculated from checkshot measurements are more reliable if each source-receiver travel path is a vertical straight line rather than an oblique, refracted path (Goetz et al., 1979). Consequently, if a well is deviated, then the surface position of the source should be readjusted each time the downhole receiver is moved to a new depth level, as shown in Figure 2, so that the travel path always remains as vertical as possible.
Offshore, the vertical traveltime to a receiver is defined relative to sea level. Since the energy source is below sea level when it produces the down-going wavelet, an amount of time equal to the air gun depth divided by the sound velocity in water is added to the measured time to adjust it to a sea level origin.
Onshore, an arbitrary depth coordinate is chosen as the time datum. In Figure 1, the datum is above the shot depth, and in such a case, the vertical distance between the shot depth and the datum depth is divided by the velocity in that interval. That time adjustment is then added to the measured traveltime to each receiver. If the depth datum is below the shot depth, as in Figure 2, this adjustment time is subtracted from the measured traveltime.
When a checkshot survey well penetrates formations that exhibit complicated structural dips, it is advisable to position an energy source on both the updip and downdip sides of the well so that two different traveltime measurements are acquired at each receiver depth. One of the travel paths is usually a better approximation of a straight line than the
401
402
PART 7—GEOPHYSICAL METHODS
the travel path to each receiver is as nearly vertical as possible.
Figure 2. The source-receiver geometry commonly used to record checkshots in deviated wells.
other. For example, in Figure 3, source position A is preferred when the receiver is at depth Zv but source position B is the better choice for a receiver at depth Z2. Usually, the traveltimes measured for sources A and B are simply averaged at each receiver depth because the structural dips and formation velocities are rarely known with enough precision to predetermine which travel path is the better approximation of a straight line.
In surveys where the structure is simple horizontal layering but where significant lateral velocity variation occurs, it is also advisable to record traveltimes from shots on opposite sides of the well and average the times so that the checkshot values are not biased with a velocity that is unrepresentative of the prospect area.
When there is sufficient velocity and dip information and adequate presurvey preparation time to allow ray trace modeling of the source-receiver travel path, it is helpful to calculate and display the anticipated ray paths for several possible source and receiver locations to determine which source position produces the best approximation of a straight line travel path to each desired receiver location.
VERTICAL SEISMIC PROFILES
A vertical seismic profile (VSP) is recorded in essentially the same way as a checkshot survey (Balch and Lee, 1984; Hardage, 1985). The major difference between a VSP and a checkshot survey is that VSP data are recorded at much smaller spatial sampling intervals than checkshots. While a receiver may be moved a vertical distance of 200 to 1000 ft (61
to 305 m) between checkshot levels, it should be moved no more than 50 to 100 ft (15 to 30 m) when recording a VSP. Specifically, the vertical distance between successive VSP traces should not exceed one-half of ^min, where Xmin is the shortest wavelength contained in the recorded VSP wavefield.
When a seismic wave field is recorded with this small spatial sampling interval, several processing techniques can be used to separate the down-going and up-going wave fields. Once the up-going wave field is isolated from the more dominant down-going wave field, the up-going reflection events can be properly analyzed and interpreted and used to produce improved imagery of the subsurface.
VSP SOURCE-RECEIVER GEOMETRIES
Several types of VSPs can be recorded by altering the position of the energy source relative to the receiver. The term offset is used to describe the horizontal distance between the source and receiver. If the receiver is directly below the source, the recorded data are called a zero offset VSP. If there is a significant horizontal distance between the source and receiver, the recorded data are referred to as an offset VSP. Examples of offset and zero offset geometries are shown in Figure 4. A common misuse of the term offset is in describing the horizontal position of the energy source relative to the wellhead rather than the position of the source relative to the location of the subsurface receiver. For this reason, the geometry in Figure 4(d) is an offset VSP, not a zero offset VSP.
In a flat-layered earth, the reflection points associated with a zero offset VSP occur close to the vertical line passing through the source and receiver coordinates. Thus, the image made from these data will illuminate the subsurface in only a narrow vertical corridor passing through the receiver location. However, if there is structural dip, the reflection points
Zero-Offset VSP Geometry
Small Distance
K^-H
Well Source Head
Large Distance
Receiver Level N
Receiver Level N
Receiver Level 1
(a)
(b) Receiver Level 1
Offset VSP Geometry
Large Distance
Small Distance
Well Head
Source
Receiver Level N
Receiver Level
Receiver Level 1
(C)
Figure 4. Examples of the source-receiver positions involved in (a and b) zero offset and (c and d) offset VSP recording geometries.
Checkshots and Vertical Seismic Profiles
403
associated with a zero offset VSP can occur at significant horizontal distances from the vertical line passing through the source and receiver. When properly processed, such data can produce high resolution images extending from the receiver position to the farthest reflection point coordinate.
For offset VSPs, reflection points are always distributed over some horizontal distance, so offset VSP recording geomety is often used to produce seismic images that traverse portions of a reservoir near survey wells.
Cross-Borehole Tomography in Development Geology
J. H. Justice
Advanced Reservoir Technologies Dallas, Texas, U.S.A.
M. E. Mathisen J. R. Bulau A. A. Vassiliou
Mobil Research & Development Corporation Dallas, Texas, U.S.A.
INTRODUCTION
Cross-borehole tomography can be thought of as an extension of sonic logging to the reservoir cross section between two boreholes. The information obtained from crossborehole tomography, when properly interpreted in the context of all available information, can often be invaluable for preparing accurate geological cross sections for reservoir development and planning.
Important differences exist, however, between crossborehole seismic velocity images and sonic logs. Sonic logs usually measure velocity of sound in rocks at very high frequencies (5 to 40 kHz), whereas seismic tomography measures the velocity of sound at seismic frequencies, usually in the frequency range of 20 Hz to as high as several kilohertz. Sonic logs represent the measurement of velocities at the wellbore and are usually plotted as a continuous curve in depth. Because seismic tomography measures the twodimensional velocity field between the wellbores, it is usually represented by a color-coded map in which a color is assigned to the seismic velocity at each point. This map, or plot, is referred to as a tomogram. Other displays such as contour plots can also be used, but this is not common.
TOMOGRAPHIC DATA ACQUISITION
To acquire seismic tomography data, a source of seismic energy is lowered to the survey depth in the source borehole, and receivers are lowered to an appropriate depth in one or more boreholes that are used to record the seismic data (Figure 1). The source and receivers each occupy a number of stations, usually regularly spaced over a depth range that includes the zone of interest in the reservoir. The spacing of these stations and the vertical interval they cover (aperture) play a role in determining the final spatial resolution of the tomogram. With receivers operating in additional wells, data for several tomograms can be acquired simultaneously.
With source and receivers in place, the source is fired and a recording is made. In traveltime tomography, this recording is later used to compute the time required for the seismic energy to travel from source to receiver. Data are acquired for every combination of source and receiver stations. Ray paths connecting sources and receivers for a typical survey are shown in Figure 2. These traveltime data are then used to
infer the seismic velocity field between the wellbores. If attenuation measurements can be made on the data, then the seismic Q-factor can also be imaged in the reservoir. Since the number of source/receiver pairs is usually in the thousands, data acquisition can be a slow process. Wirelines with multiple receivers appropriately placed can reduce the survey time.
Because usable seismic frequencies are an important factor in resolution, sources with the widest possible bandwidth are desirable, and data sampling rates of the recording instruments must be correspondingly high (usually on the order of 10,000 samples per second for each channel). Usually multicomponent receivers, which record data using three orthogonally mounted geophones, are used, so that both shear and compressional waves can be identified. If it is possible to reconstruct both compressional and shear wave tomograms, then Vp/ Vs and Poisson's ratio can be computed in the interwell volume.
BOREHOLE SOURCE TECHNOLOGY
Seismic borehole source technology is receiving considerable attention, and a number of sources are in use or under development to meet the various requirements of power, bandwidth, repeatability, short cycle time (between shots), depth, and temperature, to name a few. It is likely that no single source can meet all requirements, and the source chosen is the one most suitable to the survey conditions (see the chapter on "Seismic Data Acquisition on Land" in Part 7). Examples of source technology under development or in use today include
• Sparkers (electrical) • Air gun (compressed air) • Piezoelectric (pulsed or swept frequency) • Hydraulic (vibratory) • Pneumatic (vibratory) • Chemical explosive • Magnetostrictive (pulsed or swept frequency)
Vibratory sources typically generate a swept frequency signal that sweeps across a preset frequency band in a fixed time period (usually on the order of a fraction of a second to several seconds). The swept frequency signal is then
404
Cross-Borehole Tomography in Development Geology
405
Figure 1. Schematic illustration of tomographic data acquisition.
correlated with a reference signal to produce a pulselike waveform that is used in processing the data. The advantage of a swept frequency signal is that a high level of energy can be transmitted while keeping the rate of energy input to the borehole within an acceptable range. Higher frequency sources are generally most useful over short interwell distances and/or well consolidated sediments.
DATA PROCESSING
Cross-borehole tomographic data could be processed using algorithms similar to those used in medical X-ray image reconstruction (CAT scan). However, this is not done because seismic energy does not propagate along a straight line between source and receiver as X-rays do. As a result, special algorithms have been developed for seismic tomography. These algorithms must simultaneously reconstruct both the seismic velocity field of the reservoir as well as the path of the energy through it. Algorithms based on ray tracing or some form of the wave equation are usually used. These algorithms solve a system of thousands of nonlinear equations to reconstruct the velocity field between the boreholes for compressional and/or shear waves.
SURVEY TYPES
Two major types of tomographic surveys are in use today. These consist of single tomograms for imaging reservoir characteristics, and time-lapse tomography for imaging time evolution processes in reservoirs such as those associated with enhanced oil recovery (EOR).
In time-lapse tomography, a baseline survey is taken, ideally before an EOR (or similar) project begins. The resulting tomograms may or may not be interpreted at this time. After an appropriate time interval, during which the EOR program is operating and inducing changes in the reservoir, a second survey is taken. The resulting tomograms will be of little additional value unless the changes in the reservoir have also induced changes in the seismic velocity field. Fortunately, there are some well-documented situations in which this does occur (see Nur, 1984; Justice et al., 1989, 1992, 1993). For example, heating of heavy oils by steam injection can produce large reductions in the seismic velocity field in the affected part of the reservoir.
With at least two time-lapse surveys in hand, the differences in the successive images (Figure 3) should relate to changes induced in the reservoir by the process, all other factors being constant.
Single tomographic images, however, are often difficult to interpret. Changes in the observed seismic velocity field across the reservoir may be due to the combined effects of lithology, pore fluids, and the physical state of the reservoir.
406
PART 7—GEOPHYSICAL METHODS
- -
^ "V:
V v • -;
.; . : . \
г ... л t ;s
t '
fe.
1000 FT
500
600
700
800 500
600
700
800 400
p 500
b.
jE 6oo
OL
LU Q 700
800
Vp
- 7333
- 718$
- 7000
- 6833
•6866
- 8500
- 6333
• 6166
- 6000
• 5833
- 5666
- 5500
Lx'-i-J
•
5333 5166
5000
100 200 300 400 500 600 700
CROSSHOLE DISTANCE
Figure 3. Time lapse tomograms acquired at 6 month intervals across a thermal EOR project. (After Justice et al., 1989.)
410 FT-
Figure 2. Ray path diagram documenting the distribution of ray paths across a thermal EOR project tomogram.
Often geological stratification is easily discernible (Figure 4) due to the contrast in seismic velocities characteristic of different types of rocks The results of faulting and other structural features in the reservoir can also sometimes be seen, when associated with a sufficient contrast in seismic velocities.
APPLICATIONS
Numerous applications have been and are being identified for cross-borehole tomography in development geology, including
• Identification of high porosity zones between wells (where porosity is clearly related to seismic velocity)
• Locating well sites for infill drilling • Monitoring enhanced oil recovery programs • Structural and stratigraphic mapping of the reservoir
and preparation of cross sections • Monitoring reservoir dynamics such as movement of
the gas cap or fluid contacts • Improved reservoir characterization and modeling
TOMOGRAPHY DATA ANALYSIS
Tomography interpretations need to document reservoir properties and/or the production process in a way that development geoscientists and engineers can steadily understand and use in subsequent production planning. Before interpretation, tomography data quality and potential interpretation pitfalls should be evaluated. Appropriate tomography display parameters also need to be selected. Interpretations, which can then be completed based on correlations with reservoir data and models, should be presented using integrated data displays.
Tomography data quality should be analyzed to identify possible data-dependent limitations on interpretation. Travel time error must be evaluated to estimate velocity uncertainty. Spatial resolution should also be estimated based on shot and receiver intervals. Ray path diagrams should be analyzed to identify zones that may have insufficient coverage.
Vp TOMOGRAM
Cross-Borehole Tomography in Development Geology
407
SHALE-CARBONATE SEQUENCE
Figure 4. Single image tomograms documenting stratigraphy and reservoir zones. Heterogeneity of the lower reservoir is documented by the contrast in log curves and the corresponding change in the tomographic velocity fields. (After Justice et al., 1990.)
Numerous interpretation pitfalls, which can make part or all of an interpretation uncertain, are also important to evaluate. Since velocity is affected by many factors, incomplete geological characterizations or log data increase the uncertainty of tomography interpretations. Sufficient well data are therefore necessary to support interpretation adequately. In reservoirs where EOR processes cause rapid changes, correlation of tomograms with older logs may not be valid. Efforts should be made to run logs at the same time the tomography data are acquired.
Tomography display parameters are one of the more important factors governing the ease of interpretation and acceptance of results. To facilitate correlations with crossborehole geology, tomogram velocity fields should be scaled to match major stratigraphic and reservoir units (Figure 4). The definition of smaller velocity fields may be appropriate to show details such as reservoir heterogeneity within selected major velocity fields. Stratigraphic units such as formation tops and sedimentary facies should also be delineated. In
addition to defining velocity fields that match geological units, the use of standard colors and/or graphic symbols for rock units is important. Cold colors (blues) should be used for high velocity fields with a transition to hot colors (reds) for low velocities. Using these guidelines for display should result in tomograms that closely resemble geological cross sections and are more readily understood and utilized by geologists, engineers, and management. Analysis of tomograms with reference to reservoir models and any available surface seismic should provide the basis for interpretations that can be summarized with integrated data displays, as illustrated in Figure 5. Most important are the well data to velocity field correlations and the log to tomography correlations documenting lithology, porosity, fluid saturation, and temperature. The resulting log and tomography display should provide the data needed to qualitatively document cross-borehole structure, especially dip and faults, reservoir heterogeneity and homogeneity, fluid contacts and any EOR flood fronts.
408
PART 7—GEOPHYSICAL
METHODS
Vp TOMOGRAM
HEAVY OIL SANDS
1000
eRleeSc' VELOCITY"
SANDSTONE
CLAYSTONE
SANDSTONE
AMALGAMATED BRAIDED CHANNELS MEANDERING CHANNEL
LEVEE/INTERCHANNEL
CHANNEL/LEVEE COMPLEX
VELOCITY
ELEC. RES.
- 8000 - 7000
9200 9000
Figure 5. Integrated data display documenting cross-borehole tomography interpretation across thermal EOR project. Log-defined stratigraphic units and fluid saturation zones correlate with tomographic velocity fields and provide the basis for interpretation of reservoir and fluid properties. (From Justice et al., 1990.)
Full Waveform Acoustic Logging
С. H. Cheng
Earth Resources Laboratory Cambridge, Massachusetts, U.S.A.
INTRODUCTION
Full waveform acoustic logs (FWAL) provide information about the seismic and lithologic properties of a formation. Conventional acoustic logs measure only the P (compressional) wave traveltime through the rock. Full waveform acoustic logs, however, provide a large amount of other information. This includes S (shear) wave traveltime; Stoneley wave traveltime; and P, S, and Stoneley wave amplitudes. P and S wave traveltimes and amplitudes provide information about the lithology and the fluid content of the formation. Stoneley wave velocity and amplitude have been shown to be useful in the detection of fractures and the estimation of relative formation permeability.
TOOLS
Full waveform acoustic logging tools are similar to conventional acoustic logging tools. The main differences are that (1) the source is usually of lower frequency, centered at 5 to 10 kHz rather than at 20 kHZ; (2) the source-receiver spacing is longer, on the order of 10 ft (3.05 m); and (3) the receiver is commonly an array consisting of four to twelve phones.
The source generates a pressure pulse. Seismic waves propagating down the borehole are recorded by the receivers. Instead of just the first arrival, the whole wavetrain traveling down the borehole is recorded and digitized, with the typical digitization interval being 5 (isec. The length of the time
window recorded varies from 2 to 10 msec, depending on the source-receiver separations (Minear, 1986).
WAVE TYPES
There are three wave arrivals that can easily be identified in a full waveform acoustic log (Cheng and Toksoz, 1981). In order of arrival time, they are the P wave packet, the S wave packet, and the Stoneley wave (Figure la). The first two are analogs of the compressional and shear head waves refracted along the borehole wall. In a "soft" formation where the shear wave velocity is lower than the acoustic velocity of the borehole fluid (such as in semiconsolidated sediments), there is no refracted S wave arrival (Figure lb). The Stoneley wave is a pressure pulse created by the wave guide effect of the borehole and is called the tube wave in checkshot surveys or vertical seismic profiling surveys. Because the wavelength of the pressure pulse is approximately the size of the borehole diameter, there is a significant borehole wave guide effect. Thus, the refracted P and S wave arrivals along the borehole or formation boundary are not pure P and S waves. However, the first arrivals from these packets still travel with the formation P and S wave velocities, respectively (Cheng and Toksoz, 1981; Paillet and White, 1982). The amplitudes of these waves are affected by a combination of the formation P and S wave attenuation and that of the borehole fluid. The Stoneley wave travels with a velocity that is slower than both the formation shear wave velocity and the acoustic velocity of the borehole fluid.
(a)
TIME (msec)
(b)
TIME (msec)
Figure 1. FWAL microseismograms recorded (a) at two source-receiver separations and (b) in a "soft" formation at two source-receiver separations.
409
410
PART 7—GEOPHYSICAL METHODS
O
1000
2000
3000
4000
5000
TIME (microseconds)
Figure 2. FWAL microseismograms across a fracture zone.
DATA PROCESSING
With some FWAL tools, the slowness (inverse velocity or time needed to travel a fixed distance) is obtained the same way as in conventional sonic tools by picking the P wave arrival using a threshold detection algorithm and measuring the moveout between two receivers (Willis and Toksoz, 1983). Because of the lower frequency content, this method is not as accurate as that used with conventional sonic tools. The newer generation of FWAL tools take advantage of the larger number of receivers. Several different array processing techniques are used, the most common being semblance stacking along different slownesses (Kimball and Marzetta, 1984; Hsu and Baggeroer, 1986; Lang et al., 1987). This method can also be used to obtain the slownesses of the later arrivals, namely, the S wave and the Stoneley wave.
In a "soft" formation where the S wave velocity is lower than the acoustic velocity of the borehole fluid, no refracted S wave can be detected. The S wave slowness can be estimated indirectly using the Stoneley wave slowness (Stevens and Day, 1986), but a number of factors including permeability can affect the Stoneley wave slowness (Williams et al., 1984; Burns et al., 1988). The only reliable method is to use the direct shear wave logging tool (Zemanek et al., 1984; Chen, 1988).
APPLICATIONS
The most direct use of FWAL is the measurement of formation shear wave velocity. Together with P wave velocity and density, one can obtain the shear modulus and compressibility of the formation, which are very important in engineering applications. P wave to S wave velocity ratio is a good indicator for lithology, and borehole S wave velocity information is necessary for tie-in with shear wave reflection
Д&Т ( ^ s e c / f t )
Figure 3. Plot of the difference between the measured slowness and the predicted elastic slowness (ДДТ) against the core measured permeability values for both the limestone-dolomite and the sand-shale examples. (After Burns et al., 1988.)
profiles, amplitude versus offset studies, and elastic wave equation migrations, among many other uses.
The FWAL is also commonly used to identify and characterize fractures. Fractures are easily identified by a significant attenuation in all the wave modes—P, S, and Stoneley. An example of data across a fracture zone is shown in Figure 2. Various models are available to estimate the permeability of the fracture from the Stoneley wave attenuation across a fracture (Paillet, 1983; Tang and Cheng, 1989) and reflection from a fracture (Hornby et al., 1989).
Similar to fractures, the FWAL can also be used to identify and characterize permeable zones (Williams et al., 1984; Bums et al., 1988). Stoneley wave velocity decreases and attenuation increases with formation permeability. These changes can be attributed to the interaction between the pore fluid and borehole fluid (Rosenbaum, 1974). A correlation between core measured permeability and change in Stoneley wave slowness for two different formations is shown in Figure 3.
The Gravity Method
T. R. LaFehr
LCT, Inc. Houston, Texas, U.S.A.
INTRODUCTION
The purpose of gravity surveys is to aid in the detection and delineation of subsurface geological features. Gravity meters measure the differences in gravitational attraction between points of measurement. Nongeological effects are removed from the measured values to produce gravity anomalies that are caused by rocks having anomalous density in comparison with the density of the surrounding rocks.
FIELD MEASUREMENTS
The unit of gravity measurement is the Gal, named after Galileo, and is equal to 1 cm/sec2. The earth's gravity field is typically about 980 Gal (Telford et al., 1976; LaFehr, 1980). Field instuments are capable of detecting differences in the earth's field between 0.000001 and 0.0001 Gal (or between 1 pGal and 0.1 mGal), depending on the instrument. Other factors, such as errors in positioning data or the movement of the meter, usually cause larger uncertainties in the observations than those inherent in the gravity meter itself. Measurements are made on the ground surface, on the water bottom, in ships on the sea surface, in aircraft, and in boreholes (LaFehr, 1983). Table 1 gives each method of acquisition, a generally accepted range of measurement accuracy and a list of factors that may affect accuracy, productivity, and cost.
Proper line orientation and/or station spacing may be important to achieve adequate anomaly identification. Stations (or lines) that are spaced greater than one-half the geological depths of interest result in partially or inadequately defined anomalies. Primary lines are ideally oriented in directions perpendicular to geological strike. Repeated stations are a necessity in land, borehole, and underwater work, while tie lines must be acquired in surface ship and airborne surveys. Regional control (beyond the area of immediate interest) is usually important to acquire,
although other surveys or government data may be available for this purpose.
DATA REDUCTION
Corrections are applied to the field observations for instrument drift, tidal effects of the varying position of the sun and moon, the rotation and oblateness of the earth, variations in station elevation, water depths, vehicular motion (for sea and airborne surveys), and terrain. The result is called the Bouguer anomaly, which may be plotted on profiles and contoured on maps (Erwin, 1977; LaFehr, 1991).
INTERPRETATION
The Bouguer anomaly is usually separated into components arising from the geological features of interest (residuals) and other components (often termed the regional field) (Nettleton, 1954). The residual Bouguer anomaly may then be interpreted in terms of the most probable geological causes. An example is shown in Figure 1. Regional fields may or may not be simple, depending upon their sources.
Two general approaches can be taken toward the interpretation of the residual anomaly: forward and inverse calculations. In the forward calculation, the geological source is assumed or approximated from well, seismic, or geological data and the resulting anomaly is compared with the residual Bouguer anomaly. Differences between the calculated and residual anomalies may suggest required changes in the geological model or in the anomaly separation. In the inverse calculation, reasonable limits are placed on the rock densities, depth, or dimensions of the proposed geological source and the residual anomaly is "inverted" to find a possible geological solution. Inversion is nonunique, however, and the solution cannot be considered final without corroborating information.
Table 1. Gravity Methods and Factors Affecting Production, Cost, and Accuracy
Method Land Precision land Helicopter with inertial surveying Borehole Underwater
Surface ship
Airborne
Accuracy 0.03-0.05 mGal
3-10 pGal 0.05-0.10 mGal
2-20 pGal 0.03-0.10 mGal
0.1-2.0 mGal
3-10 mGal
Factors
Surveying, terrain, station spacing, program size Same as conventional Mobilization, logistics, program size, station spacing, landing sites Reading interval, total depth, mobilization Positioning, sea conditions, station interval, program size,
mobilization Seismic production, mobilization, sea conditions, positioning,
program size Mobilization, program size, positioning, air conditions
411
412
PART 7 — G E O P H Y S I C A L METHODS
Figure 1. Example of observed, regional, and residual Bouguer anomalies.
Borehole Gravity
Alan T. Herring
Edcon Denver, Colorado, U.S.A.
BASIC CONCEPTS
The borehole gravity meter (BHGM) can be described simply as a deep-investigating density logging tool. Applications range beyond this simple description to include detection of oil- and gas-filled porosity and detection and definition of remote structures, such as salt domes, faults, and reefs.
One of the great advantages of the BHGM as a density logging tool is that it is practically unaffected by nearborehole influences, which are the scourge of nuclear tools: casing, poor cement bonding, rugosity, washouts, and fluid invasion. Another advantage is the fundamental simplicity of the relationships among gravity, mass, rock volume, and density. Complex geology can be easily modeled so that the response of a range of hypothetical models can be studied and understood before undertaking a survey.
What is actually measured is referred to as BHGM apparent density, which is a simple function of the measured vertical gradient of gravity. To obtain an apparent density measurement, gravity is measured at two depths. The accuracy of the computed density depends on the accuracy of both measured differences: gravity and depth. Operationally, BHGM surveys resemble vertical seismic profiling (VSP) surveys. The BHGM is stopped at each planned survey level, and a 5- to 10-min reading is taken. The blocky appearance of the log reflects the station interval (Figure 1).
The log is not continuous. BHGM measurements are taken at discrete depths usually at intervals of 10 to 50 ft, depending on the resolution required. While the BHGM has remarkable resolution in the measurement of density over intervals of 10 ft or more (less than 0.01 g/cm3), surveys requiring closer vertical resolution must sacrifice density resolution.
APPLICATIONS
The spectrum of BHGM applications is defined on one extreme by density logging and on the other by remote sensing of structure. The first extreme sometimes focuses strictly on formation and reservoir evaluation questions, while the other extends to basic exploration. Figure 1 is an example of both applications. In fact, the purpose of the survey was to detect carbonate porosity in a reef environment that was missed by the other logs. For this objective, the useful radius of investigation of the measurement is on the order of 50 ft. The sharp negative density anomaly observed between 6330 and 6370 ft suggests porosity obscured by nearborehole effects or poor volume sampling. However, the broad departure of the BHGM and gamma-gamma logs over the depth range of the logged section is typical of a structural effect, in this case the edge of the reef complex which is within a few hundred feet.
Density Logging
Borehole gravity density measurements are unhindered by casing, poor hole conditions, and all but the deepest fluid invasion. The BHGM measurement samples a large volume of rock, which provides a density-porosity value that is more representative of the formation. This is especially beneficial in carbonate and fractured reservoirs (Rasmussen, 1975). BHGM surveys have been used to find hydrocarbon-filled porosity missed by other logs in both open and cased holes. Gas-saturated sands are a particularly easy target (Gournay and Maute, 1982).
BHGM Density (g/cm3)
2.00
3.00
Gamma_Gamma _Den_s i t у
2~00 ~
2*50
3.00
Г г Ll
l E
I ;
г г г. J
H -J
5 п
1I
l
I
Ц I
fJ 1Г.
г
г
LI
J
г1 Vi
J
IS
Figure 1. An example of a BHGM log. The sharp difference in density between 6330 and 6370 ft is caused by porosity not detected by the gamma-gamma density log. The broader difference anomaly observed over the length of the logged interval is explained by the structural influence of the reef complex.
423
414
PART 7—GEOPHYSICAL METHODS
The wide radius of investigation has also been successfully used to determine gas-oil and oil-water contacts in reservoirs where other measurements have been ineffective (van Popta et al., 1990; Schultz, 1989). BHGM density measurements have been used to calculate hydrocarbon saturations: the larger the fluid density contrast, the larger the measured effect. Gas saturations are therefore the easiest to measure. The difference in densities measured by the gamma-gamma log and the BHGM can be used to calculate the difference in oil saturation between the invaded and undisturbed zones, which can in turn give an estimate of movable hydrocarbons.
Remote Sensing
A useful and practical rule of thumb for BHGM remote sensing applications is that a remote body with sufficient density contrast can be detected by the BHGM no farther from the wellbore than one or two times the height of the body. A salt dome with 15,000 ft of vertical relief would have a definitive signature a few miles away. A channel sand 20 ft thick would be detectable no more than 40 ft away. Local geology, and in particular the thickness of local density units, defines the effective radius of investigation of the BHGM.
Computer modeling of BHGM measurements can be used to develop relatively detailed salt dome flank or reef flank model interpretations. Modeling is particularly effective where seismic data can be integrated into the modeling process; a model is sought that is consistent with both data sets (Lines, 1988).
THEORY
Measurements of gravity differences repeatable to about 3 pGals have been achieved using a LaCoste and Romberg borehole gravity meter. The unit "Gal" is named after Galileo, and 1 Gal = 1 cm/sec2. The earth's gravity field varies from about 979 Gals at the equator to 983 Gals at the poles. Thus, 1 (iGal is about one part in IO9 of the earth's field. The BHGM is clearly a remarkably sensitive instrument.
The underlying assumption in computing apparent density is based on an earth model made up of a layer cake of infinite horizontal slabs. For such a model, the density of any slab is exactly given by the gravity gradient through that slab. The gradient measured at any point within the slab is constant, and the slabs above and below it have no effect on
Table 1. Maximum Operating Conditions for the LaCoste and Romberg Gravity Meter
Conditions
Temperature Pressure Deviation
Sonde Diameter
4-1/8" Sonde
5-1/4" Sonde
115°C 12,000 psi 14°
260°C 20,000 psi 14°
the gradient within it. This simple assumption serves effectively in a majority of cases. Modeling of more complex geometries is not difficult and is routinely used in computing structural corrections to apparent density.
The formula for apparent density is given approximately by the following equation (there are also small corrections for latitude and elevation):
p = 3.6824- 0.03913 Ag/Az
where p = density (in g / cm3)
Az = depth interval (in ft) Ag - change in gravity (in pGals)
The constant density term compensates for the earth's normal vertical gravity gradient.
OPERATING LIMITATIONS
The gravity meter itself is a delicate spring balance that measures changes in weight of a small proof mass. The meter must be leveled at each station, and it is accurately thermostatted at a temperature of about 126°C. The present LaCoste and Romberg meter cannot operate in a wellbore deviated from vertical by more than 14°. Table 1 shows a summary of maximum operating conditions for the meter.
For operating temperatures liigher than 126°C, the meter is operated inside a Dewar flask. Using the flask, operating temperatures up to 260°C are possible. Operations from floating platforms are also possible and have been carried out. An operating limitation is the platform motion noise transmitted down the cable to the BHGM. Because of this noise, readings shallower than about 4000 ft are usually not possible without some form of downhole tool clamp.
Magnetics
Michael S. Reford
Consulting Geophysicist Aylmer, Quebec, Canada
BASIC CONCEPTS
The earth's magnetic field has magnetized certain minerals in the upper part of the crust. Magnetite, an accessory mineral, is by far the most important. This magnetization has two parts: remanent magnetism, which was acquired when the rocks were formed, and induced magnetism, which is proportional to the present earth's field and the magnetic susceptibility of the rocks. The magnetization disappears at temperatures above the Curie point, so that only the topmost 15 to 25 miles of rock are involved. Most sedimentary rocks contain little or no magnetite, and the crystalline basement rocks are by far the most important contributors to local variations in the magnetic field. Aeromagnetic surveys have been enormously successful in mineral exploration as an aid to geological mapping since they help to outline different rock units and allow interpretation of structure. They have been widely applied in the reconnaissance of sedimentary basins for mapping the basement surface. More recently, detailed high resolution aeromagnetic surveys have been aimed at weak anomalies arising from the sedimentary section (Figure 1).
The main magnetic field of the earth is caused by electric currents in the core. The intensity of the field is measured in gammas or the equivalent SI unit of nanoteslas (nT), and it varies from about 60,000 gammas in a vertical direction at the poles to 30,000 gammas in a horizontal direction at the equator. The International Geomagnetic Reference Field (IGRF) is a mathematical approximation of this field, usually removed in processing survey data to make it easier to study the anomalies caused by crustal rocks. The IGRF is regularly updated since the field varies with time.
AEROMAGNETIC SURVEYS
The airborne magnetometer measures the total intensity of the earth's magnetic field. Modern instruments have a sensitivity of 0.01 gammas (or 0.01 nT). Gradients of the field are sometimes measured as well. Data are recorded by flying along a set of parallel lines that have perpendicular control lines to tie them together. Over smooth terrain, the magnetometer can be flown as low as 100 ft above the earth's surface, but heights of 500 to 1500 ft are more typical. The line spacing for complete detail should not be much greater than the depth of the anomaly sources that are the survey target. The data are usually processed to produce total intensity contour maps (with intervals down to 1 gamma and occasionally less) and digital data files. The processing often includes calculation of the vertical gradient or second vertical derivative of the total intensity, which sharpen and
emphasize local anomalies. In middle latitudes, reduction to the pole can make anomalies appear as if measured in a vertical instead of an inclined magnetic field. Filtering techniques are applied to separate anomalies of different wavelengths. Depths of anomaly sources can be calculated using a variety of model shapes, usually working with individual profiles. Finally, the interpreter determines the most probable geological anomaly sources in terms of their boundaries, size, shape, depth, and magnetization and assembles a coherent geological whole.
During the past 50 years, most of the earth's land surface has been surveyed by aeromagnetics with varying detail, and a high proportion of the results have been published. Airborne and marine surveys of the oceans are less complete, but have provided the basis for studies of oceanic crust and continental drift. A sampling of regional magnetic studies has been presented by Hinze (1985), and publication of the magnetic anomaly map of North America (Committee for the Magnetic Anomaly Map of North America, 1987) showed results on a continental scale. Paterson and Reeves (1985) have reviewed the state of the art in magnetics.
APPLICATIONS
Since sedimentary rocks have little magnetism, the classic use of magnetic surveys in petroleum exploration has been to map basement rocks, as well as any intrusives or volcanics within the sediments. Basement depths are determined and faults are interpreted from disruptions in the magnetic patterns. Such applications are most common in the early stages of exploring a basin, and thus, magnetic data have been little used in development programs. However, the high sensitivity of modern instruments has allowed detection and mapping of weak magnetic anomalies arising from within the sediments. Tlie sources of these intrasedimentary anomalies vary from one region to another and include salt structures, faulting, and diagenetic formation of magnetite from seeping hydrocarbons, as suggested by Donovan et al. (1979). It has been difficult to establish the sources with certainty, especially in the presence of magnetic anomalies introduced by human activities, such as the use of steel in oilfield development. Reynolds et al. (1990) have recently studied this problem over the Cement oil field.
The application of magnetics to development geology is unlikely to be important except in two special circumstances. The first situation is when the field is related to basement structure and both regional and local magnetic features are involved. The second case is when intrasedimentary magnetic anomalies have been detected and may provide a key to local boundaries.
415
S
LANCASTER SOUND
DEVON ISLAND
MAGNETIC PROFILE (Regional correction applied)
200 S
INTERPRETED BASEMENT CROSS SECTION
JONES
SOUND
E LLESM E R E
ISLAND
BJORNE PENINSULA
2
H
I
Dl
S
СЛ
S
Г4
m
а о осл
N
BAUMANN FIORD
- O SEA LEVEL
-
10,000'
- 20,000'
-
30,000'
-
40,000'
IO
20
30
40
50 M
IOO M
150 M
200 M
250 Miles
Figure 1. This long aeromagnetic profile from the Canadian Arctic islands illustrates the main use of aeromagnetics in petroleum exploration. Each small circle on the basement cross section shows a depth determined from the magnetic profile. At the south end, the magnetic anomalies are sharp over the shallow basement of Devon Island. Anomalies broaden over Jones Sound where the basement deepens and then become sharper as it rises again. The basin develops fully to the north, and minor sharp anomalies (shown on expanded scale) indicate dikes and sills within the sediments.
Electrical Methods
Arnold S. Orange
Arnold Orange Associates Austin, Texas, U.S.A.
INTRODUCTION
The electrical geophysical methods are used to determine the electrical resistivity of the earth's subsurface. Thus, electrical methods are employed for those applications in which a knowledge of resistivity or the resistivity distribution will solve or shed light on the problem at hand. The resolution, depth, and areal extent of investigation are functions of the particular electrical method employed. Once resistivity data have been acquired, the resistivity distribution of the subsurface can be interpreted in terms of soil characteristics and /or rock type and geological structure. Resistivity data are usually integrated with other geophysical results and with surface and subsurface geological data to arrive at an interpretation.
Electrical methods can be broadly classified into two groups: those using a controlled (human-generated) energy source and those using naturally occurring electrical or electromagnetic energy as a source. The controlled source methods are most commonly used for shallow investigations, from characterizing surficial materials to investigating resistivities down to depths as great as 1 to 2 km, although greater depths of investigation are possible with some techniques and under some conditions. The natural source methods are applicable from depths of tens of meters to great depths well beyond those of interest to hydrocarbon development.
Possible applications of electrical methods for the development geologist range from the investigation of soil contaminants and the monitoring of enhanced oil recovery (EOR) projects to reservoir delineation and the evaluation of geological stratigraphy and structure. The application of electrical methods has been primarily confined to the onshore environment. The offshore use of some techniques is possible, particularly for permafrost delineation and shallow marine geotechnical investigations.
ELECTRICAL PROPERTIES OF MATERIALS
The application, interpretation, and understanding of electrical methods requires a familiarity with the relationship between soil and rock characteristics and the resistivities obtained from electrical data. The resistivity of subsurface rock formations is one of the physical properties determined through the process of logging that is performed on most oil and gas wells, utilizing instrumentation inserted into the wellbore (see Part 4 of this Manual). The concept of formation resistivity plays an important part in log analysis. Although there is a correlation between rock resistivities measured by well logs and those measured by electrical methods, the log is used to investigate properties only in the
immediate vicinity of the wellbore while electrical methods yield information on bulk properties averaged over a considerable volume of material.
The resistivity of most soils and rocks (including virtually all of the rocks of interest to hydrocarbon exploration) at the frequencies utilized by electrical methods is controlled by the fluids contained within the rock (Parkhomenko, 1967) (see the chapter on "Determination of Water Resistivity" in Part 4). This is because the dry soil or rock matrix is a virtual insulator at DC and near DC frequencies. The pore fluid is in most cases water, with dissolved salts. The salinity is the primary factor in determining the resistivity of the pore fluid, with pore configuration also playing a part. Of lesser importance at oil reservoir depths is the temperature of the formation. Oil and/or gas, when present, occur over such limited formation thicknesses that their effects on bulk average resistivity is, in most cases, undetectable.
Faulting or fracturing of porous sedimentary formations in most instances has little effect on the bulk average resistivity since the additional fracture porosity changes the already high porosity by only a small percentage. However, in very tight rocks, such as igneous, metamorphic, and nonporous carbonate rocks, where intrinsic porosity is very low, the fluids in joints, cracks, and faulted zones may become the primary conducting paths (see the chapter on "Porosity" in Part 5).
In summary, the factors affecting in situ average resistivity are the total porosity, including fault and fracture porosity, and the resistivity of the fluids present within the rock. The average resistivity can be considered constant over the frequency range of interest to most of the methods under consideration here.
CONTROLLED SOURCE METHODS
Controlled source methods use generated currents or electromagnetic fields as energy sources. An advantage is the control over energy levels and the attendant positive effects on signal to noise ratio in areas of high cultural noise. A disadvantage of controlled source methods is that the complex nature of the source field geometry (the geometry of the electromagnetic field or currents induced with the earth by the transmitter) may present quantitative interpretation problems in areas of complex geology.
In the DC method, a current (usually a very low frequency square wave and not actually direct current) is injected into the earth through a pair of current electrodes, and the resulting potential field is mapped. Various geometries of current and potential electrodes have been employed, with the choice primarily based upon the depth and geometry of the survey target. The measured surface potential field is
417
418
PART 7—GEOPHYSICAL METHODS
interpreted in terms of the subsurface resistivity distribution through modeling and inversion techniques (Zody, 1989). Induced polarization (IP) and complex resistivity (CR) techniques are special cases of the DC method in which the induced potential field is measured and interpreted in terms of mineralogy and/or soil characteristics. IP and CR have been applied with some success to hydrocarbon exploration through the measurement of geochemical alteration halos that have been found to be related to reservoirs under some conditions.
In the electromagnetic (EM) method, an electromagnetic field is produced on or above the surface of the ground (Nekut and Spies, 1989). This primary EM field induces currents in subsurface conductors. The induced currents in turn reradiate secondary EM fields. These secondary fields can be detected on or above the surface as either a distortion in the primary field (frequency domain methods) or as they decay following the turning off of the primary field (time domain methods). Both loops and grounded wires are used to generate the source field. Resistivities are calculated from the observed electromagnetic field data using modeling and inversion techniques.
EM techniques have been adapted to a variety of surface and airborne configuration, with the airborne instruments generally limited in penetration to 100 to 200 m. Airborne electromagnetic surveys have proven very effective for mapping the shallow resistivity distribution, leading to costeffective surveys over large areas. Surface loop or grounded wire systems are applicable to depths well in excess of 1 km, although high power transmitters are required as depth increases. The resolution attainable is normally considered as a percentage of penetration depth, such that absolute resolution decreases with depth.
In the controlled source magnetotelluric (CSMT) method, a low frequency electromagnetic wave is generated, and the electrical and magnetic fields are measured at some distance from the transmitter. The wave impedance of the electromagnetic wave at the receiver is calculated from the electrical and magnetic field values as a function of frequency and then interpreted in terms of the subsurface resistivity distribution. Depths of penetration in excess of 1 to 2 km are attainable under suitable conditions.
Ground probing radar (GPR) is used for detailed investigations of the shallow subsurface. An extremely short pulse is generated and transmitted into the earth and reflections are received from interfaces between materials of differing resistivity and dielectrical constant. GPR instrumentation is sophisticated but highly portable. Depth of penetration is limited from less than 0.3 m in silty soils to over 100 m in permafrost, freshwater-saturated sand, and some very low porosity rocks. Successful applications include the measurement of ice thickness, the location of cracks in ice, permafrost studies, the detailed mapping of the bedrock surface, the examination of soil stratification, and the mapping of contaminant plumes in the shallow subsurface. An important application of GPR is locating buried pipes, tanks, and other objects that reflect the radar pulse.
NATURAL SOURCE METHODS
Natural source methods take advantage of naturally occurring electrical potentials and electromagnetic fields as energy sources. Advantages of natural source methods are that there is no dependence on an artificial energy source and that the natural electromagnetic field is well understood. The principal disadvantages are the unpredictability and lack of control over energy levels and the attendant effects of cultural noise on the signal to noise ratio.
The self-poteiitial (SP) method examines the slowly varying surface potential field caused by electrochemical and electrokinetic actions in near-surface materials (Sills, 1983). Potentials can form, for example, at interfaces between materials containing fluids with different ion contents, or they can be caused by moving groundwater or by differential oxidation of ore bodies. The method has been applied successfully in geothermal and mineral exploration and in the delineation of certain groundwater contaminants. Field procedures are straightforward, with the potential measured between carefully designed electrodes using what is essentially a highly sensitive DC voltmeter. The potential field is mapped along profiles or on a grid of measurement stations. Interpretation is generally qualitative, with SP anomalies interpreted in terms of the shape and depth of the causative body or fluid flow.
Magnetotellurics (MT) is an electrical method of geophysical exploration that makes use of naturally occurring electromagnetic energy propagating into the earth to determine the electrical resistivity of the subsurface (Orange, 1989; Vozoff, 1986). The low frequency electromagnetic field is measured, and the wave impedance is calculated and expressed in terms of the resistivity of the subsurface. The depth of investigation is a function of the frequency of the electromagnetic wave, taking advantage of the fundamental principle that the lower the frequency of a wave, the deeper the penetration into the crust. MT surveys generally involve applications that range in depth from a few hundred meters to 10 km or more.
The resistivity versus depth cross section developed from MT data can be interpreted in terms of rock type. Spatial variations in the resistivity-depth relationship observed at closely spaced locations on the surface can be interpreted in terms of subsurface geological structure. While MT cannot be used to detect oil directly, the identification of favorable rock types and the presence of geological structure capable of trapping hydrocarbons is critical to successful exploration. MT data are interpreted using forward and inverse modeling techniques. Resolution is considered low when compared with exploration or exploitation seismology, but may be adequate in certain instances to provide valuable information concerning reservoir geometry, rock characteristics, and a regional geological framework. For the larger deep reservoirs, MT may be considered as a possible candidate for EOR monitoring if model studies indicate that the resistivity changes over time associated with the operation are within the resolving power of the method.
APPLICATIONS FOR DEVELOPMENT GEOLOGY
The following is a brief summary of some of the many possible applications of electrical methods of interest to the development geologist:
• Evaluation of various characteristics of the shallow environment. Examples of such characteristics are the classification of unconsolidated materials based on their electrical properties, the identification of a lateral and/or vertical freshwater-saltwater boundary, the depth to bedrock, and the identification and mapping
Electrical Methods
419
of conductive groundwater contaminants. • Monitoring of reservoir stimulation and enhanced
recovery projects, where the stimulants and propants or flood materials can be expected to modify the resistivity of the formations. • Investigation of permafrost and ice characteristics in the Arctic. • Investigation of stratigraphy and structure, in particular as an adjunct to seismic data and in those areas where seismic data are poor or unreliable. • Seafloor geotechnical mapping, as an adjunct to high resolution seismic studies.
420
PART 7—GEOPHYSICAL METHODS
Part 7 References Cited
Anstey, N. A., and R. L. Geyer, 1987, Borehole velocity measurements and the synthetic seismogram: Boston, MA, IHRDC, 355 p.
Balch, A. H., and M. W. Lee, eds., 1984, Vertical seismic profiling—technique, applications, and case histories: Boston, MA, IHRDC, 488 p.
Burns, D. R., С. H. Cheng, D. P. Schmitt, and M. N. Toksoz, 1988, Permeability estimation from full waveform acoustic logging data: The Log Analyst, v. 29, p. 112-122.
Carlini, A., and A. Mazzotti, 1989, Optimized receiver array simulation based upon resolution constraints: Geophysical Prospecting, v. 37, p.607-621.
Chen, S. T., 1988, Shear-wave logging with dipole sources: Geophysics, v. 53, p. 659-667.
Cheng, C. H., and M. N. Toksoz, 1981, Elastic wave propagation in a fluid-filled borehole and synthetic acoustic logs: Geophysics, v. 46, p. 1042-1053.
Coffin, J. A., 1978, Seismic Exploration Fundamentals: Tulsa, OK, PPC Books, A division of Petroleum Publishing.
Committee for the Magnetic Anomaly Map of North America, 1987, Magnetic anomaly map of North America: Boulder, CO, Geological Society of America, 4 sheets, scale 1:5,000,000.
Dix, C. H., 1955, Seismic velocities from surface measurements: Geophysics, v. 20, p. 68-86.
Donovan, T. J., R. L. Forgey, and A. A. Roberts, 1979, Aeromagnetic detection of diagenetic magnetite over oil fields: AAPG Bulletin, v. 63, p. 245-248.
Erwin, C. P., 1977, Theory of the Bouguer anomaly: Geophysics, v. 42, p. 1468.
Fagin, S. W., 1991, Seismic Modeling of Geologic Structure: Tulsa, OK, Society of Exploration Geophysicists, Geophysical Development Series, v. 2.
French, W. S., 1974, Two-dimensional and three-dimensional migration of model-experiment reflection profiles: Geophysics, 39,265-277.
Gardner, G. H. F., ed., 1985, Migration of Seismic Data: Tulsa, OK, Society of Exploration Geophysicists Monograph Series, 462 p.
Goetz, J. F., L. Dupal, and J. Bowles, 1979, An investigation into discrepancies between sonic log and seismic checkshot velocities: Australian Exploration Association Journal, v. 19, pt. 1, p. 131-141.
Gournay, L. S. and R. E. Maute, 1982, Detection of bypassed gas using borehole gravimeter and pulsed neutron capture logs: The Log Analyst, v. 23, n. 3, p. 27-32.
Hardage, B. A., 1985, Vertical seismic profiling, Part A— principles, 2nd ed.: Oxford, U.K., Pergamon Press, 509 p.
Hilterman, F. J., 1970, Three-dimensional seismic modeling: Geophysics, v. 35, p. 1020-1037.
Hilterman, F. J., 1990, Is AVO the seismic signature of lithology? A case history of Ship Shoal-South Addition: Geophysics—The Leading Edge, June, p. 1522.
Hinze, W. J., ed., 1985, The utility of regional gravity and magnetic anomaly maps: Society of Exploration Geophysicists, 454 p.
Hornby, B. E., D. L. Johnson, K. W. Winkler, and R. A. Plumb, 1989, Fracture evaluation using reflected Stoneley wave arrivals: Geophysics, v. 54, p. 1274-2188.
Hsu, K., and A. B. Baggeroer, 1986, Application of the maximum likelihood method (MLM) for sonic velocity logging: Geophysics, v. 51, p. 780-787.
Hubral, P., 1977. Migration—some ray theoretical aspects: Geophysical Prospecting, v. 25, p. 739-745.
Johnston, R. C., D. H. Reed, and J. F. Desler, 1988, SEG standards for specifying marine seismic energy sources: Geophysics, v. 53, p. 566-575.
Justice, J. H., A. A. Vassiliou, S. Singh, J. D. Logel, P. A. Hansen, B. R. Hall, P. R. Hutt, and J. J. Solanki, 1989, Acoustic tomography for monitoring enhanced oil recovery: Leading Edge, v. 8, p. 12-19.
Justice, J. H., A. A. Vassiliou, M. E. Mathisen, S. Singh, P. S. Cunningham, and P. R. Hutt, 1992, Acoustic tomography in reservoir surveillance: Society of Exploration Geophysicists, Reservoir Geophysics, p. 321-334.
Justice, J. H., M. E. Mathisen, A. A. Vassiliou, I. Shiao, B. R. Alameddine, and N. J. Guinzy, 1993, Crosswell seismic tomography in improved oil recovery: First Break (in press).
Kelley, K. R, R. W. Ward, S. Treitel, and R. M. Alford, 1976, Synthetic seismograms—a finite difference approach: Geophysics, v. 41, p. 2-27.
Kimball, С. V., and T. L. Marzetta, 1984, Semblance processing of borehole acoustic array data: Geophysics, v. 49, p. 274281.
LaFehr, T. R., 1980, Gravity method: Geophysics, v. 45, p. 1634.
LaFehr, T. R., 1983, Rock density from borehole gravity surveys: Geophysics, v. 48, p. 341.
LaFehr, T. R., 1991, Standardization in gravity reduction: Geophysics, v. 56, p. 1170.
Lang, W. W., A. L. Kurkjian, J. H. McClellan, C. F. Morris, and T. W. Parks, 1987, Estimating slowness dispersion from arrays of sonic logging waveforms: Geophysics, v. 52, p. 530-544.
Lines, L. R., A. K. Schultz, and S. Treitel, 1988, Cooperative inversion of geophysical data: Geophysics, v. 53, n. 1, p. 820.
Lynn, W., and K. Larner, 1989, Effectiveness of wide marine seismic source arrays: Geophysical Prospecting, v. 37, p. 181-207.
Lynn, W., M. Doyle, K. Larner, and R. Marschall, 1987, Experimental investigation of interference from other seismic crews: Geophysics, v. 52, p. 1501-1524.
Minear, J. W., 1986, Full wave sonic logging—a brief perspective: Transactions of the Society of Professional Well Log Analysts 27th Annual Logging Symposium, Paper AAA.
Neidell, N. S., 1986, Amplitude variation with offset: Geophysics—The Leading Edge, March, p. 47-51.
Nekut, A. G., and B. R. Spies, 1989, Petroleum exploration using controlled source electromagnetic methods: Proceedings of the IEEE, v. 77, n. 2, p. 338-362.
Nettleton, L. L., 1954, Regionals, residuals, and structures: Geophysics, v. 19, p. 1.
Nur, A. M., Seismic monitoring of thermal enhanced oil processes: Society of Exploration Geophysicists Expanded Abstracts, 54th Annual Meeting, p. 337-340.
Orange, A. S., 1989, Magnetotelluric exploration of hydrocarbons: Proceedings of the IEEE, v. 77, n. 2, p. 287317.
Ostrander, W. J., 1984, Plane-wave reflection coefficients for gas sands at nonnormal angles of incidence: Geophysics, v. 49, p. 1637-1648.
Paillet, F. L., 1983, Acoustic characterization of fracture permeability at Chalk River, Ontario, Canada: Canadian Geotechnical Journal, v. 20, p. 468-476.
Paillet, F. L., and J. E. White, 1982, Acoustic modes of propogation in the borehole and their relationship to rock properties: Geophysics, v. 47, p. 1215-1228.
Parkhomenko, E. Iv 1967, Electrical properties of rocks: New York, Plenum Press, 314 p.
Paterson, N. R., and С. V. Reeves, 1985, Applications of gravity and magnetic surveys—The state of the art in 1985: Geophysics, v. 50, p. 2558-2594.
Rasmussen, N, F., 1975, The successful use of the borehole gravity meter in Northern Michigan: The Log Analyst, v. 16, n. 5, p. 3-10.
Reynolds, R. L., M. Webring, V. J. S. Grauch, and M. Tuttle, 1990, Magnetic forward models of Cement oil field, Oklahoma, based on rock magnetic, geochemical, and petrologic constraints: Geophysics, v 55, p. 344-353.
Rosenbaum, J. H., 1974, Synthetic microseismograms— logging in porous formations: Geophysics, v. 39, p. 14-32.
Schultz, A. K., 1989, Monitoring fluid movement with the Borehole gravity meter: Geophysics, v. 54, n. 10., p. 12671273.
Sill, W. R., 1983, Self-potential modeling from primary flows: Geophysics, v. 48, n. 1, p. 76-86.
Sheriff, R. E., 1975, Factors affecting seismic amplitude: Geophysical Prospects, v. 23, p. 125-138.
Sheriff, R. E., 1980, Nomogram for Fresnel-zone calculation: Geophysics, v. 45, p. 968-973.
Sheriff, R. E. 1984, Encyclopedic Dictionary of Exploration Geophysics: 2nd ed..: Tulsa, OK, Society of Exploration Geophysicists, 323 p
References Cited
421
Stevens, J. L., and S. M. Day, 1986, Shear velocity logging in slow formations using the Stoneley wave: Geophysics, v. 51, p. 137-147.
Tang, X. M., and С. H. Cheng, 1989, A dynamic model for fluid flow in open borehole fractures: Journal of Geophysical Research, v. 94, p. 7567-7576.
Telford, W. M., Geldart, L. P., Sheriff, R. E., and Keys, D. A., 1976, AppHed Geophysics: Cambridge, U.K., Cambridge Univ. Press, 860 p.
Tucker, P. M., and Yorston, H. ]., 1973, Pitfalls in Seismic Interpretation: Tulsa, OK, Society of Exploration Geophysicists Monograph Series n. 2,50 p.
van Popta, ]., et al., 1990, Use of Borehole gravimetry for reservoir characterization and fluid saturation monitoring: Society of Petroleum Engineers, SPE 20896.
Vozoff, K., ed., 1986, Magnetotelluric methods: Tulsa, OK, Society of Exploration Geophysicists, Geophysics Reprint Series, 763 p.
Williams, D. M., J. Zemanek, F. A. Angona, C. L. Dennis, and R. L. Caldwell, 1984, The long space acoustic logging tool: Transactions of the Society of Professional Well Log Analysts 25th Annual Logging Symposium, Paper T.
Willis, M. E., and M. N. Toksoz, 1983, Automatic P and S velocity determination from full waveform acoustic logs: Geophysics, v. 48. p. 1631-1644.
Yilmaz, O., 1987 Seismic Data Processing: Society of Exploration Geophysicists, Tulsa, OK, 525 p
Zemanek, J., F. A. Angona, D. M. Williams, and R. L. Caldwell, 1984, Continuous acoustic shear wave logging: Transactions of the Society of Professional Well Log Analysts 25th Annual Logging Symposium, Paper U.
Zody, A. A. R., 1989, A new method for the automatic interpretation of Schlumberger and Wenner sounding curves: Geophysics, v. 54, n. 2, p. 245-253.
Part 8
INTEGRATED COMPUTER METHODS
Contents
• Introduction • Introduction to Contouring Geological Data with a
Computer • Using and Improving Surface Models Built by
Computer • Log Analysis AppUcations • A Development Geology Workstation • Two-Dimensional Geophysical Workstation
Interpretation: Generic Problems and Solutions • ReferencesCited
edited by
Brian R. Shaw
Battelle Pacific Northwest Laboratory Richland, Washington, U.S.A.
Introduction
Brian R. Shaw
Battelle Pacific Northwest Laboratory Richland, Washington, U.S.A.
Almost every aspect of development geological activity has been brought into the computer age. Part 8 of the Manual is not an attempt to describe and catalog the myriad of computer products and equipment available for everyday use, but rather to introduce more immediate generalized areas of analysis in development geology. The chapters here do not describe computer-aided drafting techniques, such as the preparation of base maps, GIS systems, relational databases, or other computer-assisted drafting topics. Rather, they focus on the geological nature of reservoir descriptions. All of the other parts of the Manual have their computer adaptations and implementations so this part focuses on integrated areas of development work. In particular, three significant topics are covered: computer-assisted surface analysis (contouring), digital well log analysis, and workstation interpretation pitfalls and techniques.
The first topic examines the fundamentals of computer contouring. The production of a map is the first and foremost result of any reservoir evaluation. Without a map, the relationships among each and every well and their productive capacities will not be understood. The first chapter by Joel Gevirtz describes the underlying algorithms commonly used in computer-assisted contouring. It is critical in producing computer-assisted maps that a development scientist understand the assumptions and results of the program. As Joel clearly describes, different parameters lead to significantly different maps of the same data. The second chapter by David Hamilton describes the fine-tuning and improvement of computer-generated surfaces. One of the major advantages of using a computer to describe surfaces is that the updating and revision of reservoir maps can be done with ease and accuracy. These maps can also be made to represent the geological aspects of a surface. This addition to computer-assisted surface analysis is a critical step.
The topic of computer-assisted well log analysis can, and indeed has, taken up entire books. The introduction to digital evaluation of logs provided here is intended to expose the reader to the potential tools for effective log analysis. Charles
Cleneay focuses on the fundamental features common to log analysis packages and the integration of geological, geophysical, and engineering data to result in a complete log evaluation. Probably the most useful aspect of computerassisted analysis is the integration of cross-discipline information to create meaningful and accurate portrayals of reservoir behavior.
The workstation is the place where this integration takes place. Tom Anderson describes the workstation environment, pointing out the basic characteristics of an integrated system. While the topics of operating systems, database programs, and hardware could be addressed, each installation is different and rapidly changing. The need for proper integration of the database and attention to standards are central to proper system development. Joe Stevens and Carl Marrullier present an excellent introduction to the pitfalls and techniques necessary in interpreting geophysical data in a workstation environment. Other parts of this Manual focus on the interpretation itself, but the computer adds a dimension of flexibility impossible with paper prints. Unfortunately, it also adds potential problems. This last chapter takes the reservoir analyst through the proper stages of data preparation and assimilation. Even if the computer was not involved, this would present a desirable methodology, but with the potential for error in large databases, it is critical to assess technique as well as interpretation.
Acknowledgments
The authors have taken time from their busy schedules to produce these chapters. We would like to acknowledge Conoco, Western-Atlas, InterScience, Zycor, Landmark Graphics Corporation, and ВНР Petroleum New Ventures for allowing them to pursue this activity. Mr. Tarek Ghazi of Conoco and colleagues, industry associates, and fellow scientists reviewed and contributed to the preparation of this effort.
425
Introduction to
t
•
1
•
1
Contouring Geological
Data with a Computer
J L Gev,rtz
InterScience
^ НОИЭШ,
U.S.A.
INTRODUCTION
A basic tool for analysis and display of spatial geological data is the contour map. A contour map displays variation of a geological variable, such as thickness, depth, or porosity, over an area of interest with contour lines of equal value. Often, one or more contoured maps form the basis of detailed analysis of potential or actual reservoirs and are used to estimate volumes of fluids contained within pore spaces of the geological feature of interest. These studies require that maps be graphically presentable and that values underlying the graphic representation of the surface be reliable and reproducible. Errors inherent in mathematical estimation procedures must be understood so that reliability of values (volumes, percentages, and so on) obtained from these maps can be estimated.
Shapes of geological surfaces are complex and not readily approximated by simple mathematical functions because they result from a multitude of interacting processes that vary at different spatial scales. Ideally, spatial data should be examined with a spatial sample of regular geometric design. These designs can capture the range of variation exhibited by most spatial phenomena. However, such designs are, for all practical purposes, impossible for most geological work, although in some instances recent developments in satellite imagery allow their economic implementation. In most cases, subsurface geological features are sparsely sampled relative to their complexity and the samples are highly biased to geophysical and/or geological anomalies. Therefore, values of a variable across an area of interest must be estimated by interpolating from a sparse, irregular control point set.
Several control point patterns are commonly encountered in geological practice. These include random patterns or clusters (Figure 1). Geophysical data that contribute to a geological study are gathered in lines. Lines are a special case of clustered points. Each pattern has its own spatial characteristics and must be understood before a meaningful contoured representation can be constructed. Most geological data usually exhibit properties of both end-member patterns. Gridded patterns are rare in geological practice. Most commercial contouring packages compute statistics that when used with visual inspection of the pattern on a base map can greatly aid selection of an appropriate contouring method.
variation were prepared by hand. Hand-contoured maps represent a geologist's best approximation of the shape of a surface under investigation. Ideas based on the regional geological framework and the geologist's bias arising from prior experience are an inherent part of the hand-contoured map. Hand-contoured maps cannot be reproduced exactly, and values implied by the contours cannot be recovered.
In contrast, high speed computing facilities have engendered methods whereby an "objective" surface can be created by applying mathematical interpolation procedures to a control point set. These methods are free of any geological bias or interpretation introduced during map preparation because they produce a representation of a surface that is constructed by an "unbiased" and decidedly ungeological mathematical formulation from data measured at selected control points. Computer contoured maps can be reproduced easily by presenting the program with the same data and options as those used to create the original map. Values underlying the contoured representation can be obtained by the same interpolation procedure used to generate it.
Procedures usually used in hand contouring require that the geologist choose a contour interval that best displays ideas to be conveyed by the map. Computer-contouring methods, in contrast, require that the geologist select parameters that will ultimately determine the mathematical basis that computes and draws the finished map. Many sets of parameters can be used to produce a contoured representation of a surface sampled by a sparse set of control points. Maps will be similar in overall appearance, but will differ in specific areas because each set of parameters causes different mathematical procedures to be invoked. Each procedure produces a different map. (For example, compare figures 1 and 2 of Philip and Watson, 1983 and figure 11.07 of Clarke, 1990 with Figures 3, 4, 5, and 7 of this chapter). Suitable parameters for a particular mapping project are selected by carefully inspecting both density and distribution
COMPUTER CONTOURING VERSUS HAND CONTOURING
Contour maps that represent three-dimensional geological surfaces are prepared by time-honored procedures involving estimation methods. Prior to the advent of fast computers and computational algorithms, maps showing geological
(a)
(b)
Figure 1. (a) Random points, (b) Clustered points.
426
Contouring Geological Data with a Computer
427
Figure 2. Triangular mesh prepared from Davis (1973) data.
Figure 3. Contoured triangular mesh of Figure 2.
of control points from which the map will be made. Data values between control points are obtained by some
form of interpolation for both manual and computer contouring. For hand-contoured maps, interpolation required to estimate position and shape of individual contours is accomplished by eye or by simple averaging techniques. A triangular mesh or a rectangular grid provide a basis on which to interpolate data from control points. These frameworks rely on a complex mathematical interpolating function (bicubic splines, high order polynomials) to estimate contour positions between data points. This function is a polynomial that is "flexible" and can represent a wide variety of curve shapes. However, these functions have no direct geological significance. They have a continuous derivative everywhere within a triangle or a rectangle and therefore are at least once differentiable. This ensures that slope information implied by the control point set will be more faithfully rendered by the computational procedure.
It is important to recognize that all contouring methods, mathematical or otherwise, are interpolation methods and therefore include error in the resultant surface. This error is related to both the density and location of the measured control points used to construct the surface.
TRIANGULATION
Triangulation connects control points into a mesh of locally equiangular (Delaunay) triangles (Figure 2). Contour positions within the bounds of each triangle are estimated by interpolating from control point values that are triangle vertices. Each member of the triangular mesh is handled separately, and the surface is created by assembling triangles. Interpolation and contouring on a triangulated mesh requires few decisions from the geologist. Control point data are presented to the method along with a contour interval, and a
contoured representation of the required surface is produced. Figure 3 is a contour representation of control point data presented by Davis (1973) produced by interpolating on a triangular mesh (Figure 2).
Contours prepared on a triangular mesh will always strictly honor all data points used for the interpolation. Triangulated meshes are easily updated, therefore adding new control points and updating maps is simplified. However, contours prepared on this mesh often look "rough" and are less desirable in appearance. Some more sophisticated mapping packages provide smoothing procedures to render maps of more acceptable appearance. Figure 4 is a portion of a surface contoured using a triangulation scheme. Notice the irregular shape of the contours. The original surface that was sampled to produce this map is a fourth order polynomial. The shape of this surface is characterized by smooth contours. Irregularities seen in Figure 4 are artifacts of the interpolation procedure used to estimate the contours within individual triangles. However, the relative positions of the contours are a good approximation to those for the original surface.
Triangulation is not a method of choice when surfaces derived from several horizons for the same area of interest are desired. Operations between surfaces (such as subtraction of the lower from the higher) require that data exist at each control point for both horizons. Often, this is not the case with well data and requires an estimated point to be submitted where data are missing (Jones et al., 1986).
RECTANGULAR GRIPPING
Rectangular gridding, in contrast to triangulation, first uses data at measured control points to interpolate values to a set of grid nodes at a predefined spacing. These values are then used to estimate positions of contours crossing each grid
428
PART 8—INTEGRATED COMPUTER METHODS
Figure 5. Contours from a 13 x 13 grid using nearest neighbor search. (Data from Davis, 1973.)
Figure 4. Surface contoured on a triangular mesh. The original surface is a fourth-order polynomial.
rectangle. The complete surface is assembled from contiguous grid rectangles. For most geological applications, grid squares are used rather than the more general rectangle. Interpolation and contouring of an irregularly spaced control point set on a rectangular grid requires many decisions from the geologist.
To obtain a contoured surface representation by a rectangular gridding method, the geologist must decide on a grid spacing, a search criterion, and the method with which to interpolate control point data to grid nodes as well as a suitable contour interval. Most commercial mapping packages include a variety of options for spacing, search criterion, and method. These decisions require an understanding of the relationship between control point density/distribution and the texture of the surface.
Contours resulting from interpolating control point data to equally spaced grid nodes do not strictly honor control points from which they were made because contouring procedures consider estimated values at grid nodes rather than those for the original control points. This can be controlled to some extent by setting a smaller grid spacing. However, smaller grid spacings result in greater interpolation error in sparse data areas so that the map will be aliased by many small features that are artifacts of the gridding procedure itself and have no geological meaning. Setting a suitable grid size requires careful consideration of the distribution and spacing of the control points. Many commercial mapping packages provide average control point distance measures that can be of help when selecting the grid size and search criterion.
Because geological data are rarely presented on a uniform grid and are most often irregularly distributed across the map area, the number of control points used to estimate values at grid nodes is an important consideration. Several search procedures have been devised and are included in most mapping packages. These include nearest neighbor, circular, quadrant, and octant searching.
Nearest neighbor searching uses the nearest neighbors of a grid node to estimate nodal values. The number of neighbors to use can be decided arbitrarily or can be taken as nearest neighbors defined by a Delaunay triangulation of the control point set. The number of nearest neighbors determined from irregularly spaced control points can vary so that each grid node can be estimated from different numbers of control points depending upon their distribution across the map area. Figure 5 is a contour representation of the same data used in Figure 3 using nearest neighbor search and a 13 x 13 grid (Figure 6).
Circular, quadrant, and octant neighborhood searching procedures attempt to balance the number and distribution of control points used to estimate each grid node. Most mapping packages include procedures to estimate density and control point spacing, and these statistics should be examined carefully before deciding on search criteria for a particular project.
Creating a map grid also requires that an interpolation procedure for estimating values at grid nodes be selected. There are numerous methods that have been devised to accomplish this, including weighting methods, trend projection methods, and statistical methods. The more common gridding methods are well summarized by Davis (1973) and Clarke (1990).
Weighting methods assign weights to values at control points based on their distance from the grid node being estimated. There are various strategies for devising weighting
Contouring Geological Data with a Computer
429
Figure 6. A 13 x 13 grid showing the relationship between grid nodes and control points for the Davis (1973) data set.
schemes. The most commonly encountered scheme used for geological data is inverse distance weighting. For this scheme, values at control points are weighted by the inverse of the distance from the node. Variations of this scheme allow the control point values to be weighted by the inverse of distance raised to a selected power. Positive powers cause the influence of more distant points to make a smaller contribution to the value estimated at the grid node. Selection of the power to which the distance is to be raised depends upon the surface "roughness" and a feeling for the relationship between control points and surface shape. Several of these weighting schemes are reviewed by Clarke (1990) and Davis (1973).
Trend projection methods are an adaptation of a linear regression technique called trend surface analysis. This method has been devised because geological subsurface sampling rarely provides observations at the highest or lowest points on a surface, and it is sometimes desirable to allow the interpolation procedure to exceed the measured maximum and minimum. Trend projection methods use one of the search criteria previously described to select points that are taken in groups of three and fitted exactly to a plane using a least squares or bicubic spline methods. The grid node estimate is obtained by averaging projections of these planes. This method can be quite effective for smooth surfaces where regional dip orientation remains relatively constant over a large area of the map. This method can produce a surface that is more highly textured than the actual surface in highly deformed areas where the dip direction changes rapidly over small distances. Sampson (1978) reviews this method in detail.
Figure 7 is the same portion of the surface shown in Figure 4. The map in Figure 7 was produced by a gridding method with nearest neighbor search. Contours for this map are
Figure 7. A representation of the fourth-order polynomial of Figure 4 contoured on a grid prepared using a nearest neighbor search criterion.
smooth, and their shape closely approximates those of the original fourth-order polynomial surface from which control points were obtained. However, contours are not in the same geographic positions as in the original surface, and some control points are not strictly honored.
STATISTICAL METHODS
Several statistical methods can be used to construct a regular grid from control point data. Trend surface analysis is a regression method that fits a power series polynomial to control point data. This method has been used in geological practice to isolate regional trends from sparse control point sets. It is not designed to honor control points, but is used instead to separate regional variation from local variation (for example, separating structural variation from stratigraphic variation). It is often used hi exploration, but it has little use in detailed mapping of reservoirs.
Kriging is a statistical method devised by Krige (1951) and developed by Matheron (1971) to estimate gold reserves in ore bodies. It has found considerable application as a gridding method in the petroleum industry. The method is based on the theory of the regionalized variable first formulated by Matheron (1971) and popularized by Clark (1979) and Journel and Huijbregts (1978). Regionalized variable theory breaks spatial variation into three components. The drift is largescale variation that can be attributed to regional variation, a
430
PART 8—INTEGRATED COMPUTER METHODS
smaller scale random but spatially correlated part, and still smaller scale random noise. The method uses knowledge of spatial variance of the drift to derive a set of weights for control points that are unbiased statistical estimates. If all of the statistical assumptions are met, it can provide contours that are unbiased estimates. It also provides estimation variance at each grid node. Therefore, the method is statistically superior to the gridding methods discussed earlier. It also strictly honors control points.
Knowledge of the drift function is necessary to use the method to interpolate control point data onto grid nodes. This knowledge is embodied in a function, termed the semivariogram, that can be estimated for several orientations from geophysical data (Olea, 1975). If the semi-variogram cannot be obtained experimentally, it is assumed to be linear or exponential. This assumption can greatly reduce the confidence of estimate thereby defeating the power of the method. Although it is the most complex of the methods discussed here, it has considerable application in reservoir analysis (see the chapters on "Correlation and Regression Analysis," "Multivariate Data Analysis," and "Monte Carlo and Stochastic Simulation Methods" in Part 6 and "Reservoir Modeling for Simulation Purposes" and "Conducting a Reservoir Simulation Study: An Overview" in Part 10).
A WORD ABOUT DISCONTINUITIES
Most interpolation procedures, particularly those that involve some form of gridding, assume that the surface being estimated is spatially continuous. Discontinuities such as faults or stratigraphic pinchouts are not successfully modeled by these gridding methods. Most mapping packages allow the geologist to enter a fault trace that effectively divides the area into subareas that are gridded and contoured separately. Pinchouts and other stratigraphic discontinuities are handled by a fence that defines the zero contour. No contouring will appear on the wrong side of the zero line (Jones et al., 1986). Triangulation procedures can model faulted or other discontinuous terranes with somewhat more success without the inclusion of fault traces or zero contour lines.
It must be kept in mind that faults require geometry that is related to the type of fault and the properties of the deformed material. Mathematical interpolation does not consider these
factors when contouring faulted terranes. Fault mapping procedures that consider geometric implications of the type of faulting have not yet been developed; thus, interpolation must suffice until such procedures become available.
LEARNING COMPUTER CONTOURING
The previous sections have introduced numerous methods of computer contouring. Aside from the general considerations of data distribution and density that have been mentioned, little can be said to recommend the "best" contouring method for all problems. Much more can be said about each method, and discussions of the strengths and weaknesses of each can and do fill many volumes (see, for example, the periodical Mathematical Geology (Plenum Press) for debates on the relative merits of various Kriging procedures). The best approach to learning to apply the methods described here (and others) to practical problems (such as reservoir estimation) is to apply them to known data and compare the results.
Several approaches can be taken. Each will provide some insight into the workings of the various methods and will help in choosing which methods are advantageous in which projects. The first and simplest approach is to take a sample from a topographic map (preferably one with considerable relief) and apply various contouring methods to it. Compare computed results with the original by overlaying output and by subtracting grids estimated by two methods. How do the various methods compare? Which methods overestimate the surface? Which underestimate it?
A second approach is to compare surfaces produced by available methods on a surface that has been used as a standard for other work. A good one is that of a small drainage basin used to prepare Figures 3 and 5 (it can be found in Davis, 1973). Others can be obtained from CEED II, which is an evaluation of mapping packages (Geobyte, 1986).
Finally, although computer contouring is a large and growing discipline, it remains something of an art form in which experience is the best guide. It is not exhaustively covered in this brief introduction. It is best to use this chapter as a guide for further study and experimentation with a commercial mapping package. Until better methods are discovered, interpolation is the only route available for estimating the shape and variance of subsurface surfaces.
Using and Improving Surf^iCG Models Bllilt
by Computer
David E. Hamilton
Subsurface Computer Modeling, Inc. Austin, Texas, U.S.A.
TECHNIQUES FOR IMPROVING MAPS
The first surface model and map generated by computer are often unacceptable. Many ways exist to improve the model, some commonly used techniques are described below.
Do Not Display (Blank) Bad Portions of the Model
Most computer modeling algorithms do not extrapolate adequately. Many mapping systems automatically extrapolate to the map edge, and extrapolations are often not needed. Parameters usually exist for constraining extrapolations to near the data during model construction (Figure 1). If these d o n o t exist, t h e n a p o l y g o n can b e d r a w n around the data area and the model displayed only inside that polygon. Polygon blanking is also used to blank bad data areas for rush projects when no time exists for making corrections.
Apply Filters to the Model
A common problem with computer-generated surface models are surface structures that are not supported by data. That is, the s t r u c t u r e s are b y - p r o d u c t s of the surfacemodeling algorithm. Filters such as least squares, biharmonic, laplacian, and others can be applied to existing surface models to remove these unsupported structures. Data should be honored while filtering so that data values continue to fall on the correct side of contours.
Sometimes undesired contour wobbles are caused by data. For example, shot point values for a seismic line that parallels strike and whose shot point z values fluctuate about a contour value will cause contours to wobble through the data. This wobble is d u e to noise in the seismic data and is usually undesirable. Many filter programs allow surface
models other than the one being filtered to act as upper and lower constraints within which the resultant surface model must stay. By shifting the original surface model u p and d o w n b y small a m o u n t s ( m a g n i t u d e of the wobbles) a n d using these new surfaces to constrain filtering, a smoother model that still honors surface form can be created (Figure 2).
Rebuild Using Different Modeling Parameters
Most mapping systems have many parameter switches to adjust surface modeling algorithms. Only a few are frequently used to improve the model, but the ones used often make drastic improvements. Parameters found to be useful affect items such as a m o u n t of smoothing, n u m b e r of data points required, n u m b e r of sectors (compass directions from which data are used) that must have data, distance to look for data, type of algorithm used, type of filter used, size of grid increment, a n d w h e t h e r a coarse regional m o d e l is built and refined to the desired detail.
Rebuild Using Multiple Step Modeling Methods
Surface-modeling algorithms are designed to solve certain problems. When surface form varies significantly from this design, the results become unacceptable. With experience, most users learn which surface forms a particular algorithm is a p p r o p r i a t e for. A general rule of t h u m b is that least squares or weighted average algorithms work best when modeling gently undulating horizontal surfaces. Algoritlims that project slope are effective for modeling gently u n d u l a t i n g tilted surfaces. F e w if a n y a l g o r i t h m s effectively model complex surface forms or surfaces with sparse data.
A multi-step modeling process will sometimes achieve acceptable results in these situations. Three multi-step modeling methods are described here.
(a)
Figure 1. Contour maps of the same surface data, (a) Unconstrained extrapolation into nondata areas, (b) Contours constrained to areas near data.
Figure 2. Cross section showing the output from filtering being constrained between models built by shifting the initial surface model up and down slightly.
431
432
PART 8—INTEGRATED COMPUTER METHODS
У
(a)
(b)
Figure 3. Cross sections through the same data, (a) Extrapolated values for a weighted average algorithm tend to "come back" to the average of near data values, (b) Acceptable surface extrapolation achieved by creating a first-order trend, modeling residuals between data and trend, and adding the trend and residual surfaces.
(a)
(b)
Figure 4. Cross sections through the same data, (a) The surface model does not honor the data, (b) Surface is shifted to data by modeling the error between the data and the original surface and then adding the original and error models.
Regional Trend Assist
This technique is sometimes used when the surface is tilted, projection u p and d o w n dip is desired, and algorithms that project slope d o not produce acceptable results (Figure 3). T h e g e n e r a l s t e p s a r e (1) b u i l d a first- or s e c o n d - o r d e r t r e n d surface t h r o u g h the data, (2) back interpolate f r o m the trend surface a z value at each data location, (3) subtract the back interpolated values from the original data creating difference values, (4) build a surface m o d e l of the difference, a n d (5) a d d the difference surface to the trend surface.
Error Correction
This technique is used to correct a surface model that does not honor the data. It is also used to u p d a t e a surface model w h e n n e w data are a d d e d and w h e n it is undesirable to rebuild the model, only to adjust it to the n e w data. The general steps are (1) back interpolate f r o m the original surface m o d e l a z value at each data location, (2) d e t e r m i n e the error at those locations by subtracting the interpolated value from the data value, (3) m o d e l the error, a n d (4) a d d the error to the original model (Figure 4). Tliis procedure is commonly used to shift a surface model built from seismic data so that it passes through (honors) well data.
Directional Bias
Surface m o d e l s that stretch along axes of anticlines a n d s y n c l i n e s a r e easily p r o d u c e d if d i r e c t i o n a l b i a s capabilities a r e available in t h e s u r f a c e - m o d e l i n g a l g o r i t h m (Figure 5). If these capabilites are not available, this effect (single direction bias) can be built using a multi-step approach (Jones et al., 1986). The general steps are (1) rotate the data so the bias direction is n o r t h - s o u t h , (2) divide the у coordinate by the bias m a g n i t u d e , (3) build a grid, (4) convert the grid to data, (5) multiply the у coordinate b y the bias m a g n i t u d e , (6) rotate the grid data back to the original coordinate system, (7) m e r g e t h e original a n d n e w d a t a , a n d (8) b u i l d t h e final grid.
Interactively Edit the Surface Model
Most mapping systems allow the surface model to be edited. Edits can be applied to either a display or a surface
(a)
(b)
Figure 5. Contour maps of the same data, (a) Most algorithms weight data isotropically and creat circular surface forms, (b) Single direction bias forces elliptical weighting, allowing surface form to stretch in one direction.
model. Editing can b e d o n e in a variety of ways, including
1. Alter z values at individual grid n o d e s or triangle vertices in the model.
2. Blank an area, a d d interpretation, and rebuild that portion of the model.
3. Edit the picture contours and have the model rebuilt to match those edits.
4. Push or pull on the surface model and see the effects in the displayed contours.
Add Interpretive Control Points
Interpretive control points (dummy points) are added to the original data set in positions and with values that force the surface model to have a specific shape. Once d u m m y points are a d d e d , the surface model is reconstructed. Usually several iterations of d u m m y point addition a n d editing are requried to achieve a desired result. Old d u m m y point interpretations will often conflict with new data, and they will need to be edited or removed before an acceptable updated model can be built.
Digitize and Model Hand-Drawn Contours
Often too few data are available to build an acceptable
Using and Improving Surface Models Built by Computer
433
Figure 6. Cross section showing two conformable surfaces. Dashed line represents direct modeling of lower surface data. Solid lines represent direct modeling of upper surface data and conformable modeling of lower surface data.
Figure 7. Cross section showing four conformable surfaces. The second from the top is the control and is modeled using structure data. The other surfaces are built using the conformable method.
surface m o d e l or to s u p p o r t the detailed s h a p e of a geologist's i n t e r p r e t a t i o n . If this p r o b l e m exists o v e r m o s t of t h e m a p area, then editing the output model or using d u m m y points is not feasible. Instead, hand-drawn maps should be used. Contours from hand-drawn maps are digitized and used as input for the surface modeling algorithms. Some algorithms a r e specifically d e s i g n e d f o r digitized c o n t o u r s . If o n e of these is not available, there are usually specific parameter settings that make point-modeling algorithms effective for modeling digitized contours.
INTERSECTING SURFACE TECHNIQUES
Conformity
A technique used to model conformable surfaces is thickness addition or subtraction. It is used because directly g r i d d i n g each surface in a g r o u p of conformable surfaces m a y not produce the best results. Often variations in data distributions allow one surface to project past another or to h a v e significantly m o r e f o r m definition than others of that sequence (Figure 6). Tlte conformable technique builds a grid for the surface with the best data distribution (control surface) within a sequence of conformable surfaces a n d then a d d s or subtracts the adjacent interval's thickness to generate conformable surfaces above or below it. The newly constructed structural surface now becomes the surface to which thickness is added or subtracted to produce the next higher or lower surface. The process continues upward and d o w n w a r d until all surfaces within the sequence are constructed (Figure 7). This a p p r o a c h w o r k s well for complete data sets and vertical wells, although additional steps are required to handle deviated wells, partial penetrations, or missing data (Jones et al. 1986).
The conformable technique is often used to transfer the shape of an existing surface to a n e w surface while honoring the data of that n e w surface. C o m m o n applications include stream channels a n d the top of d r a p i n g rock units (Hamilton and Jones, 1992).
Figure 8. Cross sections showing that surfaces that intersect due to (a) baselap or (b) truncation will incorrectly cross one another.
Baselap and Truncation
For c o m p u t e r m a p p i n g , t h e t e r m baselap can b e d e f i n e d as the a b r u p t t e r m i n a t i o n of a h i g h e r surface (usually depositional) against a lower surface. A similar definition c o u l d b e u s e d f o r truncation—the a b r u p t t e r m i n a t i o n of a lower surface against a higher surface (usually an unconformity).
Most computer mapping systems build a surface model using only data for that surface. When one surface laps onto or truncates another, the initial surface models will almost always cross (Figure 8). This is expected and m u s t be corrected. The following discussion describes methods for handling baselap and truncation in a grid-based mapping system.
Baselap
Baselap can be achieved in several ways. To baselap one grid onto another for the p u r p o s e of cross section display a n d volumetrics work, the elevation z values at each grid node are compared and the m a x i m u m value is retained in a new grid. This makes the two grids coincident in the area where the u p p e r grid is missing d u e to baselap (Figure 9).
C o n t o u r m a p s of b a s e l a p p i n g surfaces s h o u l d h a v e n o contours in areas of baselap because the surface does not exist there. To baselap one grid onto another for map display, the elevation z values at each grid node are compared and the v a l u e of t h e b a s e l a p p i n g grid is set to m i s s i n g if l o w e r or k e p t
434
PART 8—INTEGRATED COMPUTER METHODS
Figure 9. Cross sections showing a baselapping surface (a) as coincident with the lower surface in areas of baselap (for cross section display) and (b) as missing in areas of baselap (for map display).
Figure 10. Cross section showing surfaces before baselap operations. The zero contour of the model built by subtracting the two surfaces defines the subcrop line.
(a)
(b)
Figure 12. The middle surface baselaps onto the lower surface and is truncated by the higher surface, (a) Cross section showing proper relationships, (b) Map showing surface contours and lines of baselap and truncation.
Figure 11. Map showing contours and subcrop lines.
if h i g h e r t h a n t h e o t h e r g r i d (Figure 9). T h e subcrop line is t h e line of contact b e t w e e n a b a s e l a p p i n g
surface and the surface u p o n which it baselaps (Figure 10). To g e n e r a t e this line, the elevation z v a l u e s of o n e grid are subtracted from the other. The input grids must cross one another (grids before baselap) or the intersection cannot be established. The resulting intersection grid will have positive a n d negative values, and its zero contour will be the line of baselap.
To display contours and the subcrop line, contour the baselapped (blanked) surface grid and then on the same map draw only the zero contour from the intersection grid (Figure 11).
Truncation
With only a few modifications, the approach used for baselap can be applied to truncation. For cross section and volumetrics work, the two grids are compared and the m i n i m u m is kept as the n e w truncated grid. For contour display, the g r i d s are c o m p a r e d a n d the v a l u e s of the t r u n c a t e d grid a r e set to m i s s i n g if h i g h e r or k e p t if l o w e r t h a n
the other grid. The intersection grid for subcrop display is built just as it was for baselap.
Combining Baselap, Truncation, and Conformity
In projects involving more than two surfaces, the techniques for baselap, truncation, and conformity are often used in combination (Figure 12). The following rules are used to order the techniques:
1. O n a h a n d - d r a w n cross section s h o w i n g all surface relationships, identify the unconformities.
2. Identify the sequences of conformable surfaces. 3. Select a control surface (the surface that builds the best
grid) for each sequence. 4. Build grids for all unconformities and control surfaces. 5. For each sequence, use the conformable technique to
build all of the noncontrol grids. 6. Starting with the lowermost surface, perform truncation
and baselap operations, working up from the bottom.
U n i q u e aspects of a project often require these procedures to be modified. However, the procedures provide a useful outline for getting started and guiding you through a project.
Using and Improving Surface Models Built by Computer
435
F 1
F 2
A'
\
R
\
R' -
Дб''
Figure 14. Surface models are constructed for the faults and for each surface on each side of each fault. Operations between surface models prevent them from projecting past one another.
Figure 13. Separate surface models are built for each fault block, (a, b, and c) The surface for each fault block is allowed to extend past faults defining the block edge, (d) When displayed, contours are constrained to inside the fault block polygon and all models are displayed on the same map. (After Jones et al., 1986.)
FAULTS
Few programs automatically identify faults based on input data; therefore, an interpretation of the presence and position of faults m u s t be m a d e prior to computer mapping. Once the fault interpretation has been made, several techniques exist for incorporating them into a surface model. Some commonly used methods are described here.
Fault Block
The fault block or polygon technique uses polygons to isolate fault blocks. Typically, several sides of the enclosing polygon are faults, while some are lines (zero throw faults) used to close the polygon. Digitized contours and data points for the surface inside the polygons are used to build that fault block's surface model. A separate model is built for each of the surface's fault blocks. To create a contour m a p , each of the fault block surface models must be contoured separately with the contours constrained to stay within the fault block's polygon. All of the contours are d r a w n on the s a m e m a p along with the fault traces (Figure 13). Similarly, volumes must be calculated on a block-by-block basis and summed for the unit.
Special care must be taken when constructing a fault block surface near sides of p o l y g o n s that d o not represent faults. Since these portions of the polygons have been d r a w n across
unfaulted surfaces, grid node values should change smoothly across these polygon lines.
Fault gaps imply nonvertical faults, and traces with gaps would be expected to shift position from one surface to the next. Since digitizing polygons requires significant effort, the s a m e set of polygons is often used for all surfaces (vertical faults). It is i m p o r t a n t to u n d e r s t a n d h o w a vertical assumption affects unit volumes.
Fault Plane
The fault plane technique is used to model surfaces cut by nonvertical faults when enough data are available to model fault faces a n d structural surfaces on either side of those faults. Separate models are built for each surface on each side of each fault and for each fault plane. Baselap ( m a x i m u m ) and truncation (minimum) operations are used to prevent surface models from projecting through faults that bound them and to merge faults that intersect one another properly ( F i g u r e 14). T h e f a u l t s a r e t r e a t e d as if t h e y w e r e u n c o n f o r m i t i e s a n d t h e s u r f a c e s b e t w e e n f a u l t s as if t h e y w e r e sequences, thus the previously described techniques apply. For m o r e than a f e w faults (three or four), the ordering of the operations becomes complex. Both normal and reverse faults can be modeled.
If t h i s t e c h n i q u e is u s e d , t h e n c r e a t i n g d i s p l a y s a n d volumes for a surface or zone requires careful manipulation of a large n u m b e r of s u r f a c e m o d e l s . This is b e c a u s e each surface is represented by a suite of surface models, one for each fault block. Also, much care is required to model surfaces cut by faults that fade out in the m a p area.
Fault Trace
The fault trace technique uses fault trace locations and a faulted surface modeling algorithm to build continuous faulted surface models. The algorithm prevents data on one side of a fault f r o m being u s e d w h e n assigning values to the surface m o d e l on the other side of the fault (line-of-sight method) (Figure 15). This technique is used by a large n u m b e r of m a p p i n g programs.
For normal faults, the traces usually enclose an area called
436
PART 8—INTEGRATED COMPUTER METHODS
Figure 15. Faults act as barriers beyond which data cannot be seen from the location for which a surface value is being calculated, (a) A grid node (indicated by +) to the west of fault A can only see data in the hatchured area, (b) A grid node farther to the south of fault A can see more data, thus the surface smoothly changes form around the fault ends. (After Jones et al., 1986.)
t h e fault gap. T h e g a p r e p r e s e n t s t h e a r e a w h e r e t h e s t r u c t u r e surface is missing. Nodes in this area are typically set to missing, although they are sometimes assigned values representative of the fault plane. Traces that d o not h a v e gaps imply vertical faults and therefore will not change position from surface to surface. Traces for nonvertical faults should shift from surface to surface, and those for significant throws will show a gap in map view.
Contouring, cross section, volumetrics, and other surface display and manipulation algorithms must be modified to use fault traces. When modified, these algorithms do not use surface model values from one side of a fault for contouring and v o l u m e calculations on the other side of the fault.
Vertical Separation
Surface-modeling algorithms that use a fault's vertical separation (see Tearpock and Bischke, 1991, for definition) can use d a t a f r o m o n e side of a fault w h e n b u i l d i n g a surface model on the other side of the s a m e fault. This is because data on the opposite side of a fault are adjusted by that fault's separation before being used to calculate the surface's value.
Vertical separation modeling is usually handled either by (1) building a separation m o d e l for each fault a n d adjusting all the data at once or (2) adjusting each data point for vertical separation at the time it is used for surface modeling. In the first approach, a vertical separation model is built for each fault. All of the separation m o d e l s for faults that affect a particular surface are added together. Data for that surface are shifted by the total separation at each location, moving them to their prefault position. A surface m o d e l is constructed using the adjusted data, and the total vertical separation model is then subtracted from the unfaulted surface model, creating the final faulted surface model (Figure 16).
The second approach is similar to the fault trace method in that it alters the u s e of a data point on the opposite side of a fault. However, instead of not using the point, it adjusts the point's z value by the vertical separation of the faults that lie between the point a n d the n o d e being calculated (Figure 17).
Figure 16. (a) Surface model built with no fault constraints, (b) Model of vertical separation, (c) Unfaulted structure model built after removing vertical separation from data, (d) Faulted structure model built by subtracting separation model from unfaulted structure model. (After Jones et al., 1986.)
\+20
V-60
-7418_ A - '—'
-7421\ 7 4 2 3
x
~~ " I
\ ~~ —4-
\
-7437
___ -4-7376
/ >Г -7421
(S)+10
\
FRULTS
+ DflTfl POINT
° FAULT VERTEX
• NODE BEING CflLCULflTED
Figure 17. The data value is adjusted by the separation of faults crossed by the line connecting the data point and the location for which an estimate is being made.
Using and Improving Surface Models Built by Computer
437
Figure 18. The cell is centered on the grid node and lies either inside or outside the polygon. The cell's area is multiplied by its z value (thickness) and that volume is added to volumes for all other cells inside the polygon.
Once constructed, cross sections, contour maps, volumes, and so on are commonly generated from these models using algorithms similar to those used for fault trace models. There are other methods for modeling with vertical separation, but r e g a r d l e s s of w h i c h m e t h o d is u s e d , t h e y all r e q u i r e significantly more information about faults than other faultmodeling methods. Often m u c h of this information is not available. W h e n this happens, most of these p r o g r a m s will "degenerate" to working as the fault trace method does.
Combined Methods
Some programs blend two or more techniques together to produce faulted surface models. For example, one program blends the fault plane method with vertical separation modeling to produce a fault model that carries surface form across faults and properly migrates fault position with depth (Banks, 1990). Another program blends the fault plane method with the fault trace method to produce faults that properly migrate with depth (Raven and Hooper, 1991).
VOLUMETRICS
Estimates of volumes are normally calculated between two structure surfaces, above a fluid contact, or sometimes below another fluid contact.
The Volumetrics Algorithm
Following is a brief description of the r a n g e of v o l u m e calculation techniques used. This discussion uses a gridbased system, although similar procedures could apply to triangulated or other systems. The discussion assumes that volumes are calculated within a bounding area (polygon).
Figure 19. The cell's corners are defined by grid nodes. Thetop is defined by two or more planes passing through the node z values and lie inside the polygon. The prism of volume under each plane is calculated and added to volumes for all other prisms inside the polygon.
Volume by Point Count
Grid n o d e s will fall either inside or outside the polygon of interest. Statistics can be generated for the nodes that fall inside the polygon. Those statistics and the following equation are used to determine volume:
N u m b e r of n o d e s x A v e r a g e value x x-grid increment x y-grid increment = Volume inside the polygon
This approach assumes that each grid cell extends from a base u p to a flat top, w h i c h is t h e g r i d n o d e ' s v a l u e , a n d that t h e n o d e is i n t h e c e n t e r of t h e cell. If t h e c e n t e r of t h e cell is i n s i d e t h e p o l y g o n , it is c o u n t e d ; if it is o u t s i d e , it is n o t (Figure 18).
Volume by Simple Plane Fits
Grid n o d e s occupy corners of grid cells, and a line is d r a w n diagonally across a cell dividing it into two triangles. The value at each corner of the triangle is k n o w n ; therefore, a flat plane can be fit through these points. The base above which volumes are to be calculated is also a plane, and its value is k n o w n . Since the d i m e n s i o n s of the sides of the triangular p r i s m are k n o w n f r o m the grid increments, all of the information needed to calculate the prism's v o l u m e is available. The v o l u m e of all triangle prisms totally inside the polygon are calculated and summed. Any prisms partially within the polygon are subdivided into smaller prisms with the values at the corners of the smaller p r i s m s linearly interpolated from the three original triangle corners. The volumes of these partial values are calculated, s u m m e d , and a d d e d to the v o l u m e of prisms totally within the polygon (Figure 19).
Volume by Mathematical Fit And Integration
This approach integrates under a smooth mathematical surface passed through the grid nodes in and around the cell
438
PART 8—INTEGRATED COMPUTER METHODS
Figure 21. Thickness is normally defined by grids representing the top and base of reservoir and the fluid contact(s).
Figure 20. A mathematical surface is fit to the grid cell. Calculus is used to integrate the volume under the curve, inside the grid cell, and inside the polygon. All cell volumes inside the polygon are added together.
for which volumes are being calculated. As with the simple plane fit described previously, node values are at cell corners and volumes are calculated and s u m m e d for each cell and partial cell within the polygon. A common approach is to use 16 grid n o d e values, 4 f r o m the cell of interest a n d 12 f r o m the adjacent cells. A third-order polynomial is fit to these 16 n o d e values. By applying calculus to this polynomial over the area of the grid cell, t h e exact v o l u m e u n d e r the s u r f a c e is d e t e r m i n e d ( F i g u r e 20). If o n l y a p o r t i o n of t h e cell is contained within the polygon, then the integration is p e r f o r m e d only within that part of the cell.
Discontinuities and Other Constraints
Faults define surface discontinuities that must be a c c o u n t e d for if v o l u m e s a r e to b e calculated accurately. A fault cuts a grid cell into t w o portions, each of which m u s t have volumes calculated seperately. There are many a p p r o a c h e s to d o i n g this, all of t h e m c o n c e r n e d w i t h projecting the surface u p to the fault in a reasonable manner. Each side of the fault is w o r k e d i n d e p e n d e n t l y unless vertical separation is available; then a more sophisticated surface estimating approach can be used.
Political b o u n d a r i e s , p o r o s i t y cutoffs, or other t y p e s of constraints must also be incorporated into the calculation. In some systems, these are incorporated into the model before going into the volumetrics program. In others, they are handeled automatically by the program.
Modeling and Volume Calculation in a Thickness Domain
A typical goal when preparing models for input to a volumetrics program is to create one model that represents the thickness of h y d r o c a r b o n s (Figures 21 a n d 22). This is done by solving the following equation:
HPT = GR x N:G x PRx ( I - S W )
Figure 22. The gross hydrocarbon rock thickness is progressively reduced by net to gross ratio, average porosity, and oil saturation, until only the thickness of pores filled with hydrocarbon remains.
where HPT = hydrocarbon pore thickness, which is a surface m o d e l representing the thickness of hydrocarbons. This is integrated over the area of interest to determine hydrocarbon pore volume (HPV). GR = gross hydrocarbon rock thickness, which is a surface model representing total thickness of the interval containing hydrocarbons. N:G = net to gross ratio, which is a model or constant representing the ratio of porous (pay) rock thickness to gross rock thickness. PR = average porosity, which is a model or constant representing the average porosity of the p o r o u s rock. SW = water saturation, which is a model or constant representing the average water saturation.
There are m a n y considerations w h e n building each of these surface models.
Building the GR Model
To build a surface model for the gross hydrocarbon rock thickness (GR), w e e n v e l o p e the v o l u m e of interest by building one surface model for its top and another for its base. These surface models should cross where the volume goes to zero.
The top of volume model is built by comparing all surface m o d e l s that define part of the top of v o l u m e envelope (top-E),
Using and Improving Surface Models Built by Computer
439
.S=NONPRY
Figure 24. Incomplete data for net and porosity due to partial penetrations, truncations, baselaps, and so on create problems when building models of these and related variables.
Figure 23. The envelope technique is used to define one grid for the top of reservoir and another for the base of the reservoir. These are subtracted to create the gross hydrocarbon rock thickness.
creating a single m o d e l that is the m i n i m u m of their values. Next, compare all surface models that define part of the base of v o l u m e envelope (base-E) and create a single model that is the m a x i m u m of their values (Figure 23).
Once constructed, the base of volume model is subtracted f r o m the top of v o l u m e m o d e l , creating the gross rock thickness model. This model is positive where thickness exists and either zero or negative w h e r e it does not. The negatives in this case are desired as they allow clear definition of the reservoir edge (zero line).
Other Variables for Volumetrics
Often porosity and water saturation (and sometimes net to gross ratio) are input as constants representing the average value over the area of interest. Otherwise the variables are entered as surface models. Creating these surface models is complex a n d can b e d o n e in m a n y w a y s . O n e of the m o s t common is to digitize a hand-drawn m a p and build a model. Another is to build a model from well data. During model construction and use, certain issues must be considered:
1. Incomplete information caused b y partial penetration, eroded top, faulting, and so on is often encountered. W h e n this happens, the value of net to gross ratio or any other parameter being modeled will only represent the portion of the rock unit present. Figure 24 demonstrates this problem for a partially penetrating well. The N:G for the middle well is 0.875 (or 3.5/4), while for the left and right wells, which fully penetrate the unit, the N:G is 0.55. Clearly the partial unit value d o e s n o t correctly r e p r e s e n t t h e " t r u e " u n i t value. If t h e missing portion was pay, then the N:G would be 0.95 [(3.5 + 6)/10]. If t h e m i s s i n g p o r t i o n w a s n o n p a y , t h e n
the N:G would be 0.35 (3.5/10). The true N:G lies somewhere between these two limits. Special techniques must be used to model incomplete data (Jones et al., 1986). 2. Modeling water saturation using standard algorithms and well data generally does not produce acceptable results. This is because the a m o u n t of w a t e r in the oil or gas portion of the reservoir is d e p e n d e n t u p o n porosity, p e r m e a b i l i t y , h e i g h t a b o v e fluid contact, a n d o t h e r factors. G e n e r a l l y several e n g i n e e r i n g f u n c t i o n s (Jcurves) that relate porosity, h e i g h t a b o v e fluid contact, and water saturation are used to convert structure top, s t r u c t u r e base, fluid contact, a n d p o r o s i t y m o d e l s to a water saturation model (Hamilton and Jones, 1992). The resulting model can be adjusted to honor the existing water saturation data at wells.
3. Net to gross and sometimes average porosity can change rapidly in the area extending from the point w h e r e t h e b a s e of reservoir e n c o u n t e r s t h e fluid contact to t h e reservoir e d g e ( w e d g e zone). If t h e vertical distribution of net rock is not h o m o g e n e o u s throughout the reservoir, then these variables may change significantly in the wedge zone relative to values where the reservoir is full thickness. Often these changes are ignored. Sometimes the reservoir is separated into subzones, with a full suite of volumetric models constructed for each subzone. Three-dimensional modeling of net and porosity is a m o r e precise solution.
4. If m o r e t h a n o n e of the f o u r m o d e l s i n p u t to t h e H P T equation contain negative values, then additional incorrect volumes could result. This is because these m o d e l s a r e m u l t i p l i e d together, a n d if t w o h a v e negative values at the same location, the resulting value will be positive, creating a volume where no volume should exist. A commonly used safety measure is to clip porosity, water saturation, and net-to-gross models to a m i n i m u m value of zero, eliminating the problem. Zeros in the model often produce a very jagged zero line contour. However, that is preferred rather than significant volume errors. There are techniques for correcting these jagged zero contour lines.
440
PART 8—INTEGRATED COMPUTER METHODS
Modeling and Volume Calculation in a Structure Domain
Some programs calculate volumes directly from structure surface models. This has several advantages, two being simpler input and the ability to define horizontal levels above and below which volumes should be calculated. If this type of program allows only one surface for the top a n d one for the base of volume, then the same techniques for enveloping the volume described previously must be used. How the porosity, N:G, and water saturation grids are used may vary from program to program and must be understood for one to have confidence in the results.
OTHER SURFACE MODELING TOOLS
In addition to algorithms for gridding, contouring, and cross section display, there are many other tools required for effective computer mapping. Whether a grid, triangulated mesh, or other surface model is used, most of these tools are n e e d e d . W h e r e a p p r o p r i a t e , the term surface model is used instead of grid to make the descriptions more general. Here is a list of some of those tools:
• Single data and surface operations—Add, subtract, multiply, and divide by a constant, trigonometric functions, blank max, blank min, clip max, clip min, normalize, dip and azimuth, statistics, rotate, and so on.
• Dual data and surface operations—Add, subtract,
multiply, divide, union, intersection, max, min, blank max, blank min, clip max, clip min, and so on. Polygon blanking—One or more polygons are used to blank or set to another value data or surface values inside or outside the polygons. Merge data—Combine two data files by concatenation, union, or intersection. Range subset data—Select a range of acceptable values for one or more data fields a n d build a subset of the original data set that fits within those ranges. Resample grid—Given an old grid and a new arrangement of grid nodes (due to changes in origin, x\j limits, or x-y grid increments), calculate z values for the new grid using surrounding node values from the old grid. Grid to data—Convert a grid to data. Back interpolate—Given x-y locations from a data file, calculate z values at those locations from a surface model. Either create a new field in that data file, replace all of an existing field, or replace missing values in an existing field. Filter surface model—Smooth a surface model while either honoring or not honoring a given set of data. Optionally use a variety of constraints to limit h o w the filter is applied or the area over which it is applied. Volumetrics—Calculate volumes below a surface model (isochore), above a plane, and within a polygon. Optionally calculate volumes between two or several surface models and within a polygon.
Log Analysis Applications
CharIes A Cleneay
Petrophysical Consultant Houston, Texas, U.S.A.
INTRODUCTION
Log analysis is undergoing major changes through the addition of n e w logging tools and improved computerized interpretation software. These features make accurate and detailed geological descriptions from wireline data feasible. Now, wireline data are being integrated with geological, geophysical, and engineering data through software to produce more accurate and comprehensive answers to geological questions. This chapter briefly summarizes computerized log analysis packages (LAPs) by reviewing basic features and emphasizing the fundamental ideas that make LAPs useful, flexible, and powerful tools for development geologists.
FEATURES
Log analysis packages usually store data in a depthoriented database. This directly associates a well's scientific data (including wireline traces, core data, tops, and test intervals) w i t h the specific d e p t h s of their occurrence. Depending on the LAP used, database depth values may change (1) by a constant increment (usually based on the smallest c o m m o n sampling increment) or (2) by varying increments (based on each trace's unique sampling nature).
To use this data successfully for display and calculations, the user needs to learn the p u r p o s e and m e t h o d of operation for each of the seven basic LAP features. They are
• Data input • Data editing • Environmental corrections • Data processing • Data display • Data output • Data management tools
Log analysis packages should allow the user to do any or all of these in the order d e e m e d appropriate b y the user.
DATA INPUT
Digital data a r e c r e a t e d b y w r i t i n g v a l u e s i n t o a file in numerical form. Wireline data are now routinely captured on a magnetic medium at predetermined sample increments. Increments vary from tool to tool, from service company to service company (even for comparable tools), and according to the depth recording system (English or metric) used by the service c o m p a n y . The quantity of data values recorded at a given depth increment can also vary. Most logging tools record only one value per increment for each specific trace (slow channel data). Others (for example, a full waveform acoustic tool) acquire multiple values at each depth
increment (fast channel data) in order to later replicate and use the entire acquired range of values.
Each wireline company has its own proprietary format for recording digital data. The two most common formats are LIS (the de facto standard) and BIT. Considerable efforts are being m a d e to standardize all of the various formats into a single industry-wide standard known as the API digital log interchange standard, or DLIS (Froman, 1989).
At the customer's request, digital wireline data can be transmitted f r o m the logging truck's c o m p u t e r directly to (1) the client, (2) other p a r t n e r s , a n d / o r (3) the logging company's main computer for immediate processing, retransmission, or rerecording for delivery to the client. When digital data are read into a computer by a LAP, they are automatically converted to the LAP's internal data format for storage and use.
Analog data include all data presented on p a p e r or films:
• Traceplots (commonly called logs) • Numerical listings • Text information
Traceplots can be converted to digital data through manual digitizing or optical scanning. Most LAPs include log digitizing software. Many commercial data firms also sell digitizing services and may even have digital libraries for sale. Numerical listings include core and geochemical data. Both are usually sampled somewhat sporadically and thus have varying sample increments. Text information includes descriptions, explanations, formation tops, and test results. This information is extremely useful when annotating graphics. Both numerical listings and text information are well suited for keyboard entry into the database.
DATA EDITING
Options for editing of the data include the following:
1. Merge various traces into a single trace. 2. Rescale hybrid scales and improperly digitized traces. 3. Smooth (filter) traces. 4. Assign cutoff values. Compare trace values to a single
limiting value. Beyond that value, assign the data either a single value or null values. 5. Discriminate against invalid data. Using this technique, the user specifies minimum and/or maximum value limits for a primary trace. Then, instead of m o d i f y i n g actual trace values, the trace is scanned, and at depths where data occur outside the limits, flags are set in a separate discriminator trace. When applying the discriminator trace during data displays or calculations, any depths containing flags will either be eliminated from the display or be assigned default calculation values.
441
442
PART 8—INTEGRATED COMPUTER METHODS
BASE TRACE
UNSHIFTED TRACE
-100
0
-80
20
2000
2050
Figure 1. Interactive depth shifting. Theusermarkscorrelating inflection points and shifts off-depth traces to base trace depths.
6. A p p l y depth corrections. These fall into two categories: • Depth shifting traces against each other. To do this, the user visually compares base and unshifted traces, marks corresponding data points (Figure 1), and then shifts the off-depth data to the base trace depths. • Correct for true vertical depth (TVD), true vertical thickness (TVT), a n d / o r true stratigraphic thickness (see the chapter on "Preprocessing of Logging Data" in Part 4).
7. Baseline the spontaneous potential (SP). Interactively flattening t h e SP to a s h a l e b a s e l i n e at a single v a l u e (Figure 2) allows the user to look at SP values quantitatively in order to calculate water resistivity (Rw) and estimate shale content.
8. Convert data scales (both ways): conductivity to resistivity, raw data to porosities, neutron porosities to a different matrix, metric to English depth units, percent to decimal, and so on.
9. Data normalization. This procedure assumes the values in an individual data trace are credible but require some modification. This involves modifying data values with an atypical distribution and/or range to a "normal" distribution and range (see Figure 3 and discussion of histograms). Proper normalization m u s t first account for borehole conditions during each run, and geological c h a n g e s t a k i n g p l a c e across a w i d e r geographic area. Normalization is accomplished by applying the equation:
у = ax +b
where у = corrected trace
RAW SP
BASELINED SP
Figure 2. Interactive spontaneous potential (SP) baseline flattening. The user selects points on the raw SP curve, which represent zero deflection (that is, baseline = 100% shale). By projecting the baseline between two consecutive points, SP deflections are calculated and redisplayed as a baselined (or "static") SP.
a = constant (distribution) multiplier value x = trace to be normalized b = constant (range) value (Figure 3) (Note that p r o p e r normalization of a trace affects all values f r o m all d e p t h s of that particular trace in the same way. Normalization values should be determined for each r u n of each tool, a n d applied only to that run.) 10. Rename, copy, a n d delete curves. 11. Test a n d set c u r v e values. If a trace v a l u e satisfies a criterion (such as <, =, or >), then modify the value as specified. 12. Interactively view a n d / o r edit curve values depth by depth. 13. Enter or m o d i f y well heading information. 14. Enter or modify run specific information (such as m u d properties, bottom hole temperature, true depth, and service company).
ENVIRONMENTAL CORRECTIONS
Each service company has numerous unique correction charts and algorithms for each wireline tool. By applying LAP-provided corrections for mud properties and other runspecific d a t a , t h e u s e r can quickly r e m o v e t h e effects of environmental conditions local to the tool (such as m u d weight, temperature, and borehole salinity) and work with the most technically correct wireline data (see the chapter on "Preprocessing of Logging Data" in Part 4).
Log Analysis Applications
443
DISTRIBUTION = (X„ _ X1 ) RANGE = X1 to X 2 1
X2
VALUE
Figure 3. A histogram display of a trace's value range. Frequency nodes in a trace's data values (X1 and x2) within a given formation are related to geology. Node values are usually consistent and mappable for that interval if observed in multiple wells in an area. If node values are atypical for a given well due to tool miscalibration, a correct distribution and range can be determined and the trace normalized.
DATA DISPLAY
Data display is the most frequently used LAP feature. Properly displayed data allows users to do the following:
• Accurately present both raw and interpreted data in a concise and meaningful format.
• Visually examine data from multiple wells using exactly the same scales. (Note that this usually requires consistent trace naming, matrix parameters, and decimal/percent conventions for all wells in the database.)
• Select consistent parameters for detailed log analysis calculations.
• Visually correlate all geological, core, wireline, geophysical, and engineering data both within each well and from well to well.
Visual examination of a specific formation across a b r o a d area does the following:
• Encourages relative and absolute value comparisons, thereby promoting generally improved log quality control.
• Allows determination of individual trace normalization values for each well.
• Highlights geological features and data trends through visual pattern recognition.
All of the t h r e e m o s t c o m m o n g r a p h i c s d i s p l a y s (traceplots, crossplots, and histograms) have optional interactive features that allow users to display and/or identify meaningful information quickly on the screen.
Traceplots
Traceplots (TPLTs) v i s u a l l y r e l a t e d a t a v a l u e s t o depth. W h e n p l a n n i n g a n y TPLT display, careful use of display variables (including scales, intervals, grids, track quantities a n d widths, n u m b e r of curves, line types a n d weights, colors, symbols, spacing, shading, and annotation) can be used to c o n v e y a n i m m e n s e a m o u n t of i n f o r m a t i o n w i t h o u t o v e r w h e l m i n g a n o b s e r v e r (Figure 4). S o m e of the m o r e powerful LAPs allow interactive TPLT display, correlation, a n d database storage of formation tops f r o m one or m o r e wells displayed simultaneously on the screen.
Crossplots
Crossplots (XPLTs) relate t w o or m o r e d i f f e r e n t trace v a l u e s to each other at the same depth, such as core porosity and bulk density. Each type of wireline tool m e a s u r e s a different rock property. By studying the same XPLT in many wells, distinctive data plot patterns related to these rock properties allow users to identify lithologies, porosities, parameters, and other geological and/or engineering relationships.
Several interactive graphic XPLT techniques and other features have been developed that make crossplots even more useful and powerful:
• Pickett plots, w h i c h are log-log plots of resistivity versus porosity (Figure 5) allow interactive parameter identification of m (the cementation exponent) and the p r o d u c t (я x R w ) (empirical c o n s t a n t x f o r m a t i o n w a t e r resistivity), as well as visually displaying water saturation (Sw).
• Polygon isolators drawn on the screen around patterns recognized b y the user (left side of Figure 6) identify the enclosed data in the database for future reference. A dual XPLT/TPLT screen display of the same data (Figure 6) allows XPLT pattern recognition and isolation, and TPLT depth identification (usually with tic marks or color shading at the corresponding TPLT depths). The reverse procedure (TPLT depth interval isolation and XPLT identification) is also useful. Storage and redisplay of the same polygon (using the same XPLT) on another well's data reinforces previously recognized patterns.
• Chart overlays (only available for certain data combinations), relating wireline data to known lithologies and total porosity.
• Statistical and user-drawn best fit lines and / or curves. • Color selection to enhance the display. • Drop options to remove low frequency data from cell
plots. • z plots, w h i c h s u p e r i m p o s e a third (z) trace v a l u e on t o p
of the x a n d у coordinates. (Its value can be indicated by a letter, number, or color.) • Three-dimensional plots, allowing rotation of the XPLT around the x, y, and z axes. This usually requires specialized high performance graphics hardware.
Histograms
Histograms (HISTs) plot a trace's d a t a v a l u e s a g a i n s t their frequency of occurrence (Figure 3), s h o w i n g the distribution
444
PART 8—INTEGRATED COMPUTER METHODS
4?
Sst,f-gr. mass, w/occ x-lam. Bur
Sltst. intbd w/ Clst1 x-lam. Bur
Cist, intbd w/ Sltst,x-lam, extr bur
Sltst. intbd w/ Cist, x-lam. Bur
Cist, intbd w/ Sltst, x-lam, Bur
Sst,f-gr, mass, w/occ x-lam. Bur
Sst. f-gr.x-bed
Cist, intbd wI Sltst.x-lam, Bur
Sst.f-gr,x-bed
Cist, intbd W/ Sltst.x-lam, Bur
Resistivity
Porosity
Figure 4. A traceplot displays trace values by their depth ot occurrence. Users should carefully plan details of the display to maximize visual impact, legibility, amount of information conveyed, and any logical relationships in the data. (Traceplot courtesy of Schlumberger. Faciolog is a trademark of Schlumberger.)
a X Rw (= .015
Log Analysis Applications
445
TRACE 3 ISOLATED DATA
DEPTHS
«
A log Resistivity / \ log Porosity
Д l o g Resistivity
n
.1
1
10
TRUE RESISTIVITY (ohhm)
Figure 5. A Pickett plot allow users to interactively draw a line intersecting water wet points (Sw = 100%). This line identifies the cementation exponent (m) and the product of a x Ryf. (empirical constant x formation water resistivity) and relates water saturation (Sw) to porosity and true resistivity.
TRACE 2 CROSSPLOT
TRACEPLOT
Figure 6. Dual plot contains crossplot (featuring data isolator polygon) and traceplot. User interactively draws polygon on the screen, which identifies the enclosed data in the database. Corresponding depths are immediately marked on the traceplot, in this case with tic marks.
of data across its range of values. A display of data f r o m t w o wells on the s a m e HIST (Figure 7) allows users to observe significant data n o d e shifts between the two. Exact values of shifts can be determined by interactively moving data from one well across the other until a visual "best fit" is achieved.
The ability to combine data from many wells into a single composite XPLT or HIST allows the user to see at a glance the entire r a n g e a n d d i s t r i b u t i o n of the d a t a for a n y trace. Annotations on TPLTs, XPLTs, and HISTs are extremely useful when preparing displays for presentation and reports.
Advanced Graphics
Most of today's specialized high technology tools h a v e d e v e l o p e d specialized graphic displays of both r a w and processed data. Understanding both the derivation and the presentation of the d a t a is essential to u n d e r s t a n d i n g a n d interpreting the information presented. Examples include borehole imaging and dipmeter data (see the chapters on "Borehole Imaging Devices" and "Dipmeters" in Part 4).
DATA PROCESSING AND OUTPUT
Common LAP processing methods include statistics, s i m p l e e q u a t i o n s , a n d m o d e l s . Statistics can v a r y f r o m s i m p l e to c o m p l e x , a n d t h e y v a r y f r o m L A P to LAP. Simple equations are usually predefined in both batch and interactive LAPs. They generally cover the most c o m m o n types of calculations. Some LAPs allow users to define and process their own one line equation.
Models a r e m o r e c o m p l e x t h a n either statistics or e q u a t i o n s because they employ logic, data tests, and iterative computations. Four levels of m o d e l processing complexity are available.
• Simple m o d e l s a s s u m e clean rocks of a single matrix type, which greatly simplifies calculations.
• Advanced models require users to identify and define matrix, shale, water, and other parameters from TPLTs, XPLTs, a n d / o r HISTs. They m u s t also select which of several predefined lithology models and which shale and water saturation equations will be used to give the most accurate answers. Note that software on logging truck computers usually have levels one and two dataprocessing capabilities. To benefit from this at the wellsite, users must already be acquainted with the service company software prior to arrival at the wellsite and must provide the logging engineer with accurate parameters.
• Expert level models use predefined tool response equations for each selected mineral to statistically estimate mineral, porosity, and fluid percentages in the formation. The user selects minerals presumed to be present and designates which (environmentally corrected) wireline data to be used. Results are checked by monitoring statistical variations and by reconstructing raw input data values based on the interpreted volumes. Reconstructed values are then compared with original values. A good visual overlay implies a reasonably accurate interpretation. Differences require addition or deletion of selected minerals a n d / o r modification of parameter values.
• User-defined modeling, the fourth level of data processing, requires k n o w l e d g e of c o m p u t e r programming. At this level, the user must define constants and variables; determine input and output traces; determine the exact logical order in which equations and tests are to be applied to derive the desired output; and then code, test, and debug the model. These can be simple to complex, depending on the user's skill levels and needs.
446
PART 8—INTEGRATED COMPUTER METHODS
SHIFT AMOUNT = - 2 . 5
BASE: WELLl DATA Q
S H I F T : WELL2 DATA
Ц
> u
ZUJ
D
Oиas1
CK
F
OVERLAP DATA
Q
Ift a
VALUE
Figure 7. A two-well histogram allows users to compare data interactively from one well to another by shifting the second well's data across the base well on the screen. A visual best fit is usually satisfactory for determining the amount of normalization required.
LAPs process depth by depth across specified intervals. These boundaries are typically selected by the user based on interpreted lithology changes selected from logs. Log analysis parameters vary widely from formation to formation (or even smaller intervals) due to changes in geological conditions at the time of deposition. Therefore, the selection of consistent formation tops and intervals across a wide geographic area is of great i m p o r t a n c e to the final i n t e r p r e t a t i o n results. Integration with a common geological database containing formation tops and intervals allows consistent use of the same intervals by geologists, geophysicists, and engineers.
Mainframe computers process data either interactively (while the user waits) or batch (at the computer's convenience). Many LAPs use both types.
Data processing results are stored either as a continuous trace or, in the case of a s u m m a t i o n model, as a z value (a single-valued d a t u m such as net feet of p a y or subsea depth) associated w i t h a well's latitude (x) a n d longitude (y). Values of 2 can be either m a p p e d or u s e d for f u r t h e r calculations. Thus, integration with a mapping package allows them to be plotted quickly and easily. Integration of z values with an engineering package allows quick calculation of reservoir volumetrics.
Data can be output from the LAP as computer files, printer listings, a n d / o r graphics plots. Files in various formats can be transferred to tapes, floppy disks, or other computers. Most LAPs allow printer listings of data, a n d some have m o r e flexible f o r m a t s t h a n others. G r a p h i c s plots t a k e t h e f o r m designated by the user, usually as TPLTs, XPLTs, or HISTs.
DATA MANAGEMENT TOOLS
To simplify use, each LAP vendor has created unique data management tools. These tools, whether implemented in command form or as a menu system, shield the user from the low level computer operating system that manipulates the data and files. In command form, the low level operating system routines (OPEN, READ, COPY, WRITE, and so on) can be condensed into a single LAP command (for example, DEPSHIFT) which is selected by the user. In menu form, the user is presented a series of choices to be m a d e interactively using a mouse or keyboard. The software then performs these low level system routines based on the user's selection.
Although usually transparent to the user, these tools affect software function at every step f r o m the point of data entry t h r o u g h analysis a n d presentation of the final p r o d u c t to integration with other software products. How well the data m a n a g e m e n t tools h a v e been designed d e t e r m i n e (1) h o w smoothly and intuitively the user learns and applies the LAP features a n d (2) h o w quickly, easily, a n d flexibly the software processes the data. In the final outcome, these tools determine how well the LAP is accepted by user communities throughout the industry.
A Development Geology Workstation
Tom C. Anderson
Conoco, Inc. Casper, Wyoming, U.S.A.
INTRODUCTION
While the basic m e t h o d s of d e v e l o p m e n t geology have changed little in recent years, the tools available to the geologist for applying those methods have evolved dramatically. This facet of our w o r k is changing so rapidly that this part of the M a n u a l is likely to become dated faster than any other. Nevertheless, it is worthwhile to summarize the current state of c o m p u t e r applications a n d identify the key components of a workstation.
Three main approaches have been used in providing computer tools for the geologist. The first was a centrally located mainframe attended by white-coated operators and fed by keypunched card decks, or later, remote video terminals. These have largely been superseded by one or both of the other methods, although m a i n f r a m e s still have their place in certain applications. The second method was fostered by the advent of mini-computers and could best be t e r m e d distributed computing, w h e r e the c o m p u t i n g p o w e r was placed in operating regional offices. These computers were often networked together within a company to provide for data sharing and some central databases. The third approach has been the explosion in personal computers, or PCs, over the past decade. Initially only offering what could be termed business applications, there are now numerous geoscience applications available, and a useable development geology workstation can be built solely around a PC.
Today, the geologist is likely to h a v e a hybrid of any or all of the above, often with everything f r o m PC to m a i n f r a m e networked together, and each computer platform filling its appropriate role in the total system. The mainframe is best at managing large databases and providing certain computerintensive applications such as reservoir simulation. Minic o m p u t e r s offer local peripheral sharing of plotters and digitizers, file serving to local area networks, and scientific applications such as map gridding and contouring. The PC may overlap significantly with the mini-computer in providing geoscience applications, but it is still the work horse in general purpose applications such as word processing and spreadsheet calculations. In addition, it has become the best platform for graphics-oriented programs.
All of these elements are illustrated in Figure 1, except for the remote mainframe, which may not be present at all. Great variability is seen among various installations; for example, laser printers can be connected to a dedicated print server on the network rather than directly to the PC, and digitizing can be done directly into the PC. In some cases, color graphics terminals directly connected to the minicomputer are still in vogue, but these are largely being supplanted by PCs running terminal emulation programs instead.
INTEGRATION
The time-honored sequence for computing is
Data i n p u t —> Processing —> O u t p u t
Unfortunately, d e v e l o p m e n t of specific p r o g r a m s , w h e t h e r by commercial software vendors or "in-house," has followed this model independently for each major task and ignored the potential for data sharing or interfacing with other programs. As a result, we have programs that cannot talk to one another, or if they can talk, one p r o g r a m will say "I'll talk to you as long as you speak my language, not yours." Thus, much time is spent transferring a n d / o r reformatting data among applications. Rather than having to be concerned with these roadblocks, a more desirable situation would be to have all the data a user wants directly available and usable by a n y application desired. A r e q u i r e m e n t of this ideal is universally accepted standards, which is discussed later.
THE DATABASE
Data to the development geologist means many things. First, it means well data, including, but not limited to, well locations, header information (such as operator, year, total depth, status, elevation, and API gravity), deviation surveys, f o r m a t i o n tops, fault cuts, results of drill stem tests and p r o d u c t i o n tests, core data, a n d a host of calculated values such as isochore, true vertical depth, true stratigraphic thickness, and so on. Equal in importance is log information, including curves or traces, logging parameters such as mud type and resistivity, and analysis parameters such as formation water resistivity or cementation and saturation e x p o n e n t s . In s o m e projects, s u r f a c e geology is of great i m p o r t a n c e , consisting of bed a t t i t u d e s a n d surface expression of contacts a n d faults. Seismic data can be critical, including time cross sections with interpreted horizons that need to be tied to the well control using an accurate interval velocity model. The geologist may also deal with periodic or cumulative production data by well, lease, or field. Also, the interpretation process generates new data elements, including zone average porosity, net pay, and hydrocarbon pore feet.
These data types are stored in some form of a database, which can range from a simple spreadsheet or applicationspecific custom files to powerful relational database management systems. The data are entered into the database by one of four methods: (1) direct keyboard entry of text or numerical values; (2) digitizing from a map, seismic section, or log print; (3) extracting or d o w n l o a d i n g f r o m another c o m p u t e r system followed by reformatting (if necessary) a n d direct entry as digital data; or (4) as derived or computed
447
448
PART 8—INTEGRATED COMPUTER METHODS
Mini-computer
Color Electrostatic Plotter
Dot Matrix Printer
Laser Printer
Figure 1. Principal components of a development geology workstation.
data stored back into the database by an application program. The more powerful databases provide utilities for search and retrieval, sorting, reporting and statistical analysis, and interfaces into applications.
USER INTERFACE
The geologist interacts with the computer and its programs through the user interface. This is the appearance or "look and feel" of the system to the user. Well-designed user interfaces a r e called user friendly a n d a r e s u c c e s s f u l in g u i d i n g the novice t h r o u g h a m a z e of choices to reach the final results. There are generally t w o classes of user interfaces: c o m m a n d d r i v e n a n d m e n u d r i v e n . A command-driven system p r e s e n t s an often cryptic prompt to the user and expects him or her to learn a set of c o m m a n d s to tell the computer p r o g r a m w h a t to do next. These are very flexible but more difficult for the beginner to learn, and they are largely being replaced by the second method.
A menu-driven system p r e s e n t s t h e u s e r w i t h a s e r i e s of menus containing choices that can be made at that step and often have "context-sensitive" help available to further define e a c h c h o i c e if n e e d e d . M e n u s y s t e m s a r e m u c h e a s i e r to learn, but they can become cumbersome as the user becomes m o r e proficient. S o m e t i m e s the best of b o t h s y s t e m s is available by the provision for native commands that can by-
pass wordy menus and add flexibility for the experienced user.
A n enhancement to the m e n u approach is the graphical user interface, which combines text menus with graphical objects or icons to represent choices. Inherent with this m e t h o d is consistency of design, so that the s a m e type of function, such as editing, is always presented in the same place and the same way in all applications using the interface. These applications are also designed to share data, both text and graphics, among themselves. Also featured are standard methods for moving about within the data using scroll bars a n d consistent keyboard c o m m a n d s . Because of the graphical approach, even text-oriented applications such word processing present a what-you-see-is-what-you-get (WYSIWYG) display, incorporating font selections, character sizes, and even integrated graphics such as symbols and pictures.
The previous discussion has essentially described the evolution of user interfaces f r o m primitive to m o d e r n , a n d it appears that all applications are moving toward incorporation of a standard graphical user interface in the future. Another d e s i r a b l e f e a t u r e u s u a l l y p r o v i d e d b y t h e s e s y s t e m s is multitasking, w h i c h is t h e ability to d o m o r e t h a n o n e t h i n g at o n e time, with each process resident in its own window that can be arranged on the screen like papers on a desktop. Coupled with standardized data sharing, these systems go a long way toward achieving the integration goals described previously.
A Development Geology Workstation
449
APPLICATIONS
What does the development geologist want to do with the computer? Certainly he or she has the same needs as any other worker for general purpose business applications (word processing, spreadsheets, and so on), and these are not discussed here. In addition to these, numerous geoscience applications have been developed to make the geologist's job easier, a n d these are summarized in Table 1.
The major task of the geologist has always been to m a k e maps, and this is where the first computer applications were developed and where the greatest progress has been made in refining techniques (see the chapter on "Contouring Geological Data with a Computer" in Part 8). Computerassisted map making can be merely posting values from the database on a basemap for hand contouring, or it can make use of one of the m a n y specialized algorithms to compute a grid and contour that grid automatically. An intermediate approach is to digitize hand-drawn contours and compute a grid that exactly models the geologist's interpretation. The gridding step is desirable because it allows mathematical operations between surfaces (such as computing an isopach from two structure grids), volumetric reserve calculations, a n d three-dimensional perspective views of the surfaces, which are practically impossible to do by hand.
Second to mapping, but closely tied to it, is log analysis (see the chapter on "Log Analysis Applications" in Part 8). The computer can help by plotting multiple runs, curve types, and text information onto a composite log, or it can compute water saturation curves from input curves using the Archie equation or more complex variants of it. Crossplots of any curve against any other curve (such as a Pickett plot) can be generated. These types of analyses are not restricted to a single well. With the proper application program, an entire field study can be processed, complete with field-wide crossplots by zone. Often the output data from the log analysis process is imported into the mapping package to be contoured.
Geophysical applications round out the big three (see Part 7). A basic tool in the kit is g e n e r a t i o n of s y n t h e t i c seismograms from input sonic and density log data (which should be integrated with the log analysis database). Interpretation workstations for two-dimensional and threedimensional seismic data make the geoscientisf s job easier by displaying the raw data in flexible views, assisting the picking of h o r i z o n s a n d storing the i n t e r p r e t a t i o n s in a c o m m o n database for mapping. Seismic modeling in one, two, or three d i m e n s i o n s can h e l p test h y p o t h e s e s of s t r u c t u r a l or stratigraphic interpretations. There are also applications for potential fields modeling, including gravity, magnetics, and electrical methods.
M a n y of the remaining items listed in Table 1 are simply programs to display specific geological data types in traditional forms expected by the geologist. A rapidly growing area is geological modeling, which includes basin and maturation modeling, plate tectonic reconstruction, and cross section reconstruction a n d balancing. Certainly the Hst will grow as new ways are found for the computer to assist the geologist.
Table 1. Selected Geoscience Computer Applications
Mapping Base maps Contours Geological Topographic
Log analysis Basin and maturation modeling Digitizing (logs, seismic, maps) Potential methods modeling Decline curve analysis Plate reconstruction Log plotting Cross section plotting Three-dimensional modeling Statistics plotting Synthetic seismograms Seismic two-dimensional modeling (stratigraphy) Geostatistics and fractals Utilities
Apparent dip to true dip Three-point problems TVD, TVT, and TST Coordinate conversions Map projections Well deviation plotting Stereonet plotting Ternary diagrams Rose diagrams Geochemical analysis and plotting Hydrology Stratigraphic column plotting Core description Strip logs Section balancing Expert systems Economic analysis
DIRECTION AND STANDARDS
In the future, we should expect the geologist's workstation to evolve t o w a r d greater integration of d a t a b a s e s a n d a p p l i c a t i o n s a n d t o w a r d g r e a t e r ease of u s e t h r o u g h incorporation of a graphical user interface. A d o p t i o n of standards by industry and software vendors will become essential. Table 2 Hsts a n u m b e r of possible standards in use at this time. Several of these are de facto standards that have gained acceptance f r o m w i d e s p r e a d use or the lack of an alternative. Others have been proposed as official standards by various organizations but may have not yet received acceptance. A recent d e v e l o p m e n t is the formation of the Petrotechnical Open Software Corporation (POSC), which has the goal of establishing a n d p r o m o t i n g petroleum industry standards for software. The ideal standard should be nonproprietary, free to aH users, widely accepted, and poHced by some professional organization. It is beyond the scope of this overview to discuss these standards in detail, but some definition of terms is desirable.
450
PART 8—INTEGRATED COMPUTER METHODS
The AAPG Computer Applications Committee has proposed the AAPG-B data exchange format for general purpose data transfers among computer systems, applications software, and companies (Waller et al., 1990). For log curves, the Sclilumberger LIS (log information standard) has become a de facto standard, and extensions to it have been proposed (Froman, 1989). Another log data format called LAS, for log ASCII standard, has been proposed by the Canadian Well Logging Society (Struyk et al., 1990), which may supplant LIS. The Society of Exploration Geophysicists oversees several standards for seismic data formats, the most common being SEGY for seismic trace data and SEGPl for location data. A de facto standard for offshore shotpoint location (also called navigation) data is the UKOOA format, from the United Kingdom Offshore Operators Association. A format for transferring wellsite data called WITS, for wellsite information transfer standard, has been proposed by the International Association of Drilling C o n t r a c t o r s (IADC) (Rose et al.,1989).
While database standards are still evolving, most users prefer a full function relational database management system (RDBMS). A standard query language, called SQL for Structured Query Language, is receiving acceptance from all quarters. Several commercial database products are available that support SQL. It thus becomes unimportant which product is used since applications can interact with the database via this standard interface. Direct retrievals from the database are available to users who learn SQL, but because many users do not wish to learn a command language, other products are available that build SQL statements from "fill-int h e - b l a n k s " f o r m s or e x a m p l e p r o m p t s . S o m e of these operate within a graphical user interface, letting users point a n d click their selections w i t h a m o u s e . M o s t of the commercial databases offer or plan to offer a distributed database method, in which the actual database spans numerous computers in a network. The user will be able to store and access local data locally, yet still access other needed data from halfway around the world and not be concerned with the difference.
The graphical user interface standard being adopted by most players is X-Windows, from MIT, with an associated window manager called MOTIF, proposed by the Open Software Foundation (OSF). PC users will recognize the
Table 2. Standards in Computer Industry
Application
File formats Well data Log curves Seismic traces Shotpoint locations Wellsite
Databases Model Query language
User interface Workstation
Operation system Central processing unit Network
Standard
AAPG-B LAS or LIS-II SEGY SEGP1 or UKOOA WITS
Relational (RDBMS) SQL X-Windows & MOTIF
UNIX RISC OSI/FDDI
strong resemblance it bears to the Macintosh user interface and Windows on IBM-compatible computers.
The trends in hardware platforms, operating systems, and network connections are toward standards as well. The UNIX operating system presents advantages in running on computers from many vendors, thus making software portable across platforms. The newer, high performance workstations use a central processing unit (CPU) that employs RISC, which stands for reduced instruction set computing. This makes the processor operate at much faster effective speeds. Finally, network connections will likely m o v e a w a y f r o m the d e facto Ethernet s t a n d a r d of t o d a y toward FDDI, for fiber optic distributed data interchange, using OSI (Open Systems Interconnect) protocols. The benefits will be a ten-fold improvement in network transport speed.
W h a t this alphabet s o u p of b u z z w o r d s m e a n s for the development geology workstation is that in the future, applications software will have the same "look and feel," will run on computers from many hardware vendors, will share data among themselves, and will be able to access a common database. The end users will not need to be concerned with data transfers and reformatting, the computer itself or its operating system, or where the data are located.
Two-Dimensional Geophysical Workstation Interpretation: Generic Problems and Solutions
Joe D. Stevens
CAEX Consultants Houston, Texas, U.S.A.
Carl A. Marrullier
CAM-Offshore Bellaire, Texas, U.S.A.
INTRODUCTION
This chapter describes generic problems and solutions in loading and interpreting two-dimensional seismic projects on interactive workstations. Two key assumptions about the workstation are made here: no hardware constraints exist for memory or storage and an efficient database manager exists for data manipulation. A database manager allows data from a variety of sources to be compiled and m a n a g e d as one coherent set; its structure defines user friendliness (see the chapter on "A Development Geology Workstation" in Part 8).
Any workstation interpretation project requires some data preparation. The intent of this discussion is not to walk a user through every necessary procedure but to describe why and h o w some of the basic problems might be solved. The details of preparation will vary according to h a r d w a r e and software, project type, a n d the type of data available for a given project; however, methodology will remain similar. With a little forethought, many (though not all) potential problems can be anticipated and thus minimized in the loading process. Ultimately, it is the interpreters' responsibility to ensure that data in the project conforms to a format and standard that satisfies company requirements. The speed and interactive capabilities of today's workstations allow raw paper and digital data to be input and then transformed into a final interpretation. The goal is to produce paper media exhibits that convey ideas and concepts successfully to management.
Preparation of a project for interpretation on a workstation differs between two-dimensional (2-D) and threedimensional (3-D) data sets. Three-dimensional data are generally more coherent because only a single high density data set is usually involved. Two-dimensional projects are rarely confined to one data set, are widely spaced, are randomly oriented (Valusek et al., 1990), and offer many challenges to the capabilities of an interactive geoscience system (Brown, 1990). Because of data variables such as vintage, datum, field parameters, processing, and so on, 2-D data require more careful preparation than 3-D data. This chapter focuses on the 2-D case.
In general, interpretation projects can be partitioned into t w o b r o a d p h a s e s : t h e preparation phase a n d t h e execution phase. Preparation involves interaction w i t h data processing and data loading. Execution deals with working screen displays (map and seismic window manipulation) as a technique during interpretation.
PREPARATION PHASE
For this discussion, the preparation phase deals with the three major considerations leading to interpretation: base map, seismic, and geological preparation. Each have several subdivisions, as follows:
1. B a s e m a p p r e p a r a t i o n • Projection system • Projectsize • Seismic data files • Welldatafiles • Cultural information files
2. Seismicpreparation • Datum reference plane • Polarity and phase convention • Traceamplitudescaling • Tracebalancing • Storageformat
3. Geological preparation • Well data, such as curves, tops, and lithology • Modeling • Velocityinformation
Base Map Preparation
Technical data are surveyed and locations positioned most commonly by latitude and longitude. Coordinate projection schemes attempt to redefine latitude and longitude positions, as t h e y w o u l d a p p e a r o n a s p h e r e , to s o m e x-y r e f e r e n c e system converted to a flat plane. The reason for this conversion is ease a n d speed of data m a n i p u l a t i o n and posting. Latitude a n d longitude data are in degrees, while xy systems are in decimals (Coffeen, 1990). A common projection system for positioning all project data (including proprietary seismic, contractor seismic, well, and cultural data) is critical for the relationship of multiple data sets and proper base m a p displays (Figure 1). Without a c o m m o n projection system, one cannot reliably integrate multiple x-y data sets for display. Lacking this relationship, z information cannot be related properly for interpretation. Files containing technical base map information include seismic survey data, well location data, and reference annotation. The relationship files reside separately from the data files, allowing for seismic lines and well data to be posted on a base map. Technical data files that contain the z information, such as seismic traces a n d well deviation data, are linked to the x-y location files.
Initial project size a n d a d d i t i o n of d a t a d u r i n g interpretation is a prime factor to consider in the preparation
451
452
PART 8—INTEGRATED COMPUTER METHODS
Figure 2. Seismic preparation.
Figure 1. Base map preparation: critical and noncritical.
p h a s e of a n y 2-D project. With s o m e software, project size is often awkward to modify. For example, detailing a prospect may demand more seismic data, and the additional profiles may not be available at project creation time. A 2-D project should be created of such a size that n e w data could easily b e a d d e d w h e n required (Figure 1).
Anticipating the availability of various seismic display types (such as stack, migration, inversion, or attribute) is facilitated by creating several data files, each containing a different process. This allows the user to select which data type to w o r k with. Addition of seismic data files is simple and can be done at any time during a project.
Files w i t h o p t i o n a l x-y i n f o r m a t i o n c o n t a i n c u l t u r a l o r specialty information. Cultural information, that is, lease or topographic data, is considered m a p enhancement not critical to a project. Cultural information allows technical information and subsurface interpretations to be related to natural or man-made surface features. These files are at times used to highlight certain information to augment m a p displays. Examples of customized files might include m a n made features (towns, roads, pipe lines, and leases), geomorphic data (waterways and topographic relief), highlighted wells (horizon penetrations, exploratory wells, and platform locations), or contour data (gravity, magnetics, and bathymetry).
Seismic Preparation
Displaying various vintages of data with the s a m e d a t u m , polarity, phase, and scaling is a convenience afforded by the workstation (Figure 2).
Correcting for d a t u m shifts b e t w e e n vintages of seismic data is a tedious chore on paper. In the electronic environment, these shifts are assigned to each seismic line and the data adjusted at the time of display. After loading relative to t h e x-y location, d a t a a r e a d j u s t e d to a c o m m o n r e f e r e n c e plane. Some data require a bulk time shift to tie other data sets. Bulk shifts are executed primarily in two ways: a correction time is added or subtracted to each profile and a correction velocity is applied to the profile. D u r i n g interpretation, corrections dealing with a horizon are linked to that horizon, thus allowing for time variant adjustments. These conveniences allow the explorationist to concentrate on interpretation and decision making and not on data management (Valusek and Chan, 1989).
Polarity is routinely adjusted to a common standard for a project at the time the data are loaded. Polarity and phase matching are done by cross-correlation. If, however, the data are not loaded with the same polarity convention, one can display lines in either or both polarities with ease. A common polarity convention among the data is useful when attributes, such as amplitude, are computed along an interpreted horizon.
Phase matching requires rotating the reflection data to m a t c h the w a v e s h a p e s of the data. P h a s e m a t c h i n g of different data vintages is not normally done during the loading process. The responsibility for polarity and phase likeness falls o n t h e interpreter. If t h e d a t a for a project a r e currently being processed, these parameters can be attended to during processing by the processing geophysicists with guidance from the interpreter.
On the workstation, amplitude scaling and trace balancing are done at loading or can be performed interactively. The p u r p o s e of scaling data is to match the m a x i m u m amplitudes of the various data sets. O n c e the data are all in c o m m o n character, interpretation and postprocessing enhancement are optimized. Multiple vintages of seismic data m a y require that each set be scaled separately. W h e n mixed vintages of data are used, scaling must be to the same amplitude reference (Howell and Pepper, 1988). Fixed point data should be scaled so that the greatest amplitudes (positive or negative) are contained within the hardware display limits. In scaling the data, an attempt should be made to accommodate approximately 95% of the amplitudes. Here, a m i n i m u m of data is sacrificed (clipped) d u e to h a r d w a r e display constraints. Statistical analysis of the data is p e r f o r m e d to determine values that lie in the extremes. The data are then scaled to exclude only the extreme values. Preparing the data in this manner allows the geophysicist to begin interpretation as soon as a project is loaded.
Scaling can be done interactively and modified during interpretation. Interactive scaling is time consuming and f r u s t r a t i n g if o n e is n o t well v e r s e d in the p r o c e d u r e . Scaling is most efficient w h e n performed in the loading process.
The electronic environment provides the interpreter with a g r e a t e r d e g r e e of detail t h a n d o e s p a p e r . This detail is quantified by the size "word" written to tape and used for transfer of data. Typically, processed data are written as 16or 32-bit w o r d s . This density of i n f o r m a t i o n translates to large v o l u m e s of i n f o r m a t i o n a n d requires a great deal of storage space. Retrieval and display time are slowed, even with the latest hardware. To minimize the display time and reduce storage space, data are reformatted during loading to
Two-Dimensional Geophysical Workstation Interpretation
453
Figure 4. Executing a project.
Figure 3. Geological preparation.
an 8-bit word size for structural and stratigraphic interpretations. Once interpretation is complete, the 32-bit data can be loaded for detailed computations and attribute analysis.
Geological Preparation
Well information is u s e d in a variety of ways. The most basic is i n c o r p o r a t i o n of geological tops into the seismic interpretation (Figure 3). These tops are input digitally or via keyboard entry. Once in the database, this information is easily manipulated and can be used for modeling.
Models generated from log data aid the seismic interpreter in relating seismic reflections to lithological information. Well tops associated with horizons on the seismic profiles fully integrate an exploration concept. Lithology types from the log model can be superimposed on the seismic traces for display or modification. By integrating a sonic log and d e n s i t y log (if available), o n e - d i m e n s i o n a l s e i s m o g r a m s that are created can help to identify seismic reflections representing geological interfaces. Log models can also be converted to 2-D synthetic seismic models by integrating velocity information and convolving the model with a seismic wavelet. Models can imitate seismic response to test geological concepts. Seismic resolution of geological features are determined, and simulated profiles are generated to tie wells.
Velocity information is important in relating geological data to seismic data. This information is derived by the interpreter f r o m a variety of sources, the most c o m m o n being check shot surveys obtained from wells, time-depth charts derived from stacked seismic data or seismic to well correlation, and velocity function curves. These data are input in digital form via digitizer pad or keyboard entry (Figure 3). When concatenated to a well, the wellbore, log curves, and geological tops can be displayed with the seismic. A three-dimensional velocity field can be created for conversion of time surfaces to d e p t h m a p s .
With a time-depth relationship established, the synthetic seismograms are then adjusted to tie the seismic to the well control. Velocity data are used to adjust synthetic curves, seismic and well log models, depth converted horizons, and interval maps.
EXECUTING A PROJECT
W h e n executing a project, options dealing with display of the seismic data and maps should be considered. The workstation allows for a variety of display presentations that enhance structural or stratigraphic features and attributes of seismic data. Display options span the spectrum in importance. The f o u r basic factors concerning the display of b o t h seismic profiles a n d m a p s are (1) the type of display, (2) vertical a n d horizontal scales, (3) annotation, a n d (4) color schemes (Figure 4).
There are a variety of display types (variable density, variable area, a n d any of the wiggle or wiggle plus displays) available to the user of a workstation. S o m e enhance desired features or individual attributes, while others suppress u n w a n t e d ones. It is arguable which display type, combined with a scale and color scheme, showcases a feature best. That a variety of display types are available is the important factor.
The workstation has the capability of displaying seismic d a t a w i t h a w i d e r a n g e of vertical a n d horizontal preset scales. These window scales can be coupled with zoom or magnification options that provide the interpreter with countless scale combinations to view data. N o longer is the interpreter confined to one or two fixed display scales or types because the data can be viewed in multiple ways. One can enlarge a limited time and shot point range within a w i n d o w to focus on a particular feature of interest. Regional interpretation can optimally use decimated data. Similar effects are achieved in m a p view: a m a p can be enlarged to show each trace or reduced to show regional trends.
Also, seismic annotation is tailored to show as much or as little information as desired. Options include the following: (1) shot point, C M P , or trace n u m b e r s ; (2) seismic line ties; (3) time a n n o t a t i o n a n d / o r timing lines; (4) line direction; (5) line n a m e or internal identification; (6) data file used; (7) well type or identifiers; a n d (8) horizon a n d fault intersections. M a p view annotation is also a feature that is customized to the situation. Options that can be modified on the base map are (1) line, shot point, C M P , or trace n u m b e r s ; (2) wells a n d well tops; a n d (3) cultural data. C u s t o m i z i n g o p t i o n s for an interpreted m a p are (1) sequence of a p p e a r a n c e a n d color of display items; (2) ribbon or contour display of a horizon; a n d (3) fault contours, heaves, or polygons. A similar array of options are available for perspective views.
Use of color further increases the dynamic range visible in
454
PART 8—INTEGRATED COMPUTER METHODS
the seismic display. The value of color graphic displays is increasingly recognized. Admittedly, colors may be overused by novices; however, subtle features can be brought into focus with p r o p e r use of color schemes. Color s c h e m e s can be specifically tailored to complement selected attributes and display types.
Some interpreters are inclined to use one display (type, scale, annotation, and color) for structural work and another scheme for stratigraphic work. Other explorationists use one scheme for all their interpretation.
CONCLUDING REMARKS
A flowchart s u m m a r i z i n g t h e s t e p s of project p r e p a r a t i o n and execution is shown in Figure 5. The topics touched on in this chapter could easily fill a volume. System software documentation should provide details describing "how-to" steps, but manuals generally do not address "when and why" situations. As most interpreters become familiar with a system, they learn h o w , w h e n , and w h y a procedure is invoked and follow the same path for every project. The workstation is a toolbox of m e t h o d s to speed and enhance the procedure. It is the interpreter's creativity, ingenuity, and imagination that generate meaningful interpretations. Explorationists use a n u m b e r of techniques simultaneously to test and revise working hypotheses (Valusek and Chan, 1989). One is more inclined to test m a n y ideas quickly since it will not so take long to try them (Coffeen, 1990). A clever user can find paths that achieve results that were not deliberately programmed. An example might be as simple as using faults as horizon surfaces or vice versa for unique displays. These innovative uses, once proven, are occasionally programmed as menu items in new software releases.
Effective use of any workstation for interpretation is m a d e easier by a good user interface. An interpretation can be achieved by the experienced user significantly faster on a workstation than on paper. Time saved by interactive interpretation allows more detailed analysis in the final p r o d u c t . It s e e m s axiomatic that a significant majority of geoscientists place a high priority on ease of use. This user friendliness includes language or terms that are obvious in meaning, straightforward paths to achieve goals, and manipulations of the data that are meaningful. The h a r d w a r e and software vendors have designed exemplary systems. Totally integrated interpretation systems are the stated goals of vendors that market such systems, although the industry has many improvements to make in achieving these goals.
Figure 5. General flow chart for project preparation.
Part8 References Cited
Banks, R. B., 1990, Modeling geological a n d geophysical surfaces: Geobyte, v. 5, n. 5, p. 20-23.
Brown, J. A., 1990, Workstation issues—3-D a n d 2-D integration: Reflections, Southeastern Geophysical Society.
Clark, I., 1979, Practical geostatistics: N e w York, Elsevier Applied Science Publishers, 129 p.
Clarke, K. C., 1990, Analytical computer cartography: Englewood Cliffs, NJ, Prentice-Hall, 290 p.
Coffeen, J. A., 1990, Seismic on screen—an introduction to interactive interpretation: Tulsa, OK, Pennwell Books.
Davis, J. C., 1973, Statistics a n d data analysis in geology: N e w York, John Wiley and Sons, 550 p.
Geobyte, 1986, CEEDII—mapping systems compared, evaluated: Geobyte, v. 1, n. 5, p. 25-40.
Froman, N. L., 1989, D L I S - A P I Digital Log Interchange Standard: The Log Analyst, v. 30, n. 5, p. 390-394.
Hamilton, D. E., and T. A. Jones, eds., 1992, Computer Modeling of Geologic Surfaces and Volumes: AAPG Computer Applications in Geology, n. 1,297 p.
Howell, G. S., a n d R. E. F. Pepper, 1988, M a p p i n g seismic amplitude and seismic attributes: Reflections, Southeastern Geophysical Society, April 1988.
Jones, T. A., D. E.Hamilton, and C. R. Johnson, 1986, Contouring geologic surfaces with the computer: New York, Van Nostrand Reinhold Company, 314 p.
Journel, A. G., a n d C. J. Huijbregts, 1978, Mining geostatistics: N e w York, Academic Press, 600 p.
Krige, D. G., 1951, A statistical approach to some mine valuation problems on the Witwatersrand: Journal of Chemical Metallurgy and Mineralogy Society of South Africa, v. 52, n. 6, p. 119-139.
Matheron, G., 1971, The theory of regionalized variables and its application: Paris, Les Caliiers du Centre de
References Cited
455
Morphologie Mathematique, Ecole Nationale Superieur des Mines, Booklet 5,211 p. Olea, R. A., 1975, Optimum mapping techniques using regionalized variable theory: Lawrence, KS, Kansas Geological Survey, Series on Spatial Analysis, n. 2,137 p. Philip, G. M., and D. F. Watson, 1982, A precise method for determining contoured surfaces: Australian Petroleum Exploration Society Journal, v. 22, p. 205-212. Raven, J., and N. Hooper, 1991, Computer modeling of multiple surfaces with faults (abstract): AAPG Bulletin, v. 75, p. 658. Rose, R. J., M. R. Taylor, a n d R. E. Jantzen, 1989, Information transfer standards for well-site data: Geobyte, v. 4, n. 2, p. 9-13. Sampson, R. J., 1978, Surface II graphics system (revision 1): Lawrence, KS, Kansas Geological Survey, Series on Spatial Analysis, n. 1,240 p. Struyk, C., R. Bishop, D. Fortune, E. Foster, D. Gordon, T. d'Haene, D. Joyce, S. Kenny, H. Kowalchuk, and M. Stadnyk, 1990, LAS—a floppy disk standard for log data: Geobyte, v. 5, n. 2, p. 23-29. Tearpock, D. J., and R. E. Bischke, 1991, Applied Subsurface Geological Mapping: Englewood Cliffs, NJ, Prentice Hall. Valusek, J. E., and A. W. Chan, 1989,2-D seismic workstations—tools or toys?: Asian Oil & Gas, Jan. 1989. Valusek, J. E., M. Padgett, J. Austin, and I. Page, 1990, Zoned autopicking of seismic boosts CAEX: Oil & Gas Journal, Feb. 26. Waller, H. O., D. Guinn, M. N e r k o m m e r , and B. Shaw, 1990, AAPG-B—committee offers revised exchange format for transferring geologic and petroleum data: Geobyte, v. 5, n. 2, p. 11-21.
Part 9
edited by
PRODUCTION E™NTGS-IITN-X ETTE-TR-TIi NTXGTS^ MЛ /EГ ГTГHГ ТOТDП ПSР
stePhen A Holditch
s- A• Holditch & Associates, /nc. College Station, Texas, U.S.A.
Contents
• Introduction • Production Histories • WellCompletions • Stimulation • ProductionTesting • Pressure Transient Testing • Surface Production Equipment • ArtificialLift • Production Logging • Production Problems • Workovers • ReferencesCited
Introduction
Stephen A. Holditch
S. A. Holditch & Associates, Inc. College Station, Texas, U.S.A.
The m a i n e m p h a s i s of p r o d u c t i o n e n g i n e e r i n g is to optimize recovery and profit from an individual well. A production engineer is concerned with well completions, surface production equipment, well surveillance, and w o r k o v e r s to increase flow rates a n d ultimate recovery f r o m oil and gas wells. The production engineer looks at problems on a well-by-well basis. Production engineers must be versatile and must be able to integrate results from geological, reservoir engineering, and petrophysical studies into a master plan that will optimize production and recovery from each wellbore in a field.
Part 9 begins with a chapter on production histories by Brent Hale. Often, production data can be analyzed both to understand the nature of the reservoir and to predict future production given different operating scenarios. The next two chapters discuss various aspects of well completions (Stephen A. H o l d i t c h ) a n d s t i m u l a t i o n (John G i d l e y ) . If t h e w e l l completion and stimulation treatment are properly planned a n d executed, then most of the problems associated with producing operations tend to diminish in intensity.
One important production engineering tool that can be used to obtain data for making important decisions is well test analysis. Well test analysis can be associated with either production testing or pressure transient testing. These subjects are covered in Part 9 by David E. Lancaster and W. John Lee, respectively.
Well completion must be accompanied by the appropriate design of surface facilities and artificial lift systems. It is one thing to get the oil and gas from the reservoir into the wellbore, but it is quite another to design the appropriate systems for lifting, separating, a n d metering the fluids so that they can be sold. T w o chapters in this part of the M a n u a l devoted to these problems are "Surface Production Equipment" by James Jennings, and "Artificial Lift" by D. D. Smallwood.
Even if the best well completion, artificial lift system, a n d surface facilities are designed and applied, production p r o b l e m s will n o r m a l l y occur d u r i n g the life of a well. Production logging methods and well test analysis methods often provide the data needed to diagnose production
problems. In this part, production logging methods are discussed b y James J. Smolen a n d p r o d u c t i o n problems by Bradley M. Robinson.
Once production problems are diagnosed, a workover must be designed to improve the well condition. The need for workovers and the type of workovers that are normally performed in the field are discussed by Alan F. Osborne.
Acknowledgments
The contributing authors in Part 9 responded quickly with high quality material. As Part Editor, I thank them for their efforts and results. Special thanks go to the outside reviewers: Ralph Veatch (Amoco Production Company, Tulsa, Oklahoma) and Mark Looney (Texaco USA, E&P Technology Division, Bellaire, Texas). The individual authors would also like to acknowledge their respective companies:
John L. Gidley Brent Hale Stephen A. Holditch James Jennings David E. Lancaster W. John Lee Alan F. Osborne Bradley M. Robinson D. D. Smallwood James J. Smolen
John L. Gidley & Associates, Inc., Houston, Texas
Northwest Pipeline, Salt Lake City, Utah
S. A. Holditch & Associates, Inc., College Station, Texas
Texas A&M University, College Station, Texas
S. A. Holditch & Associates, Inc., College Station, Texas
Texas A&M University, College Station, Texas
Shell Oil Company (retired), Houston, Texas
S. A. Holditch & Associates, Inc., College Station, Texas
Conoco Inc., Lafayette, Louisiana Consultant, Missouri City, Texas
We, the editor, outside reviewers, and authors, trust the information in this part of the M a n u a l will be helpful to the members of AAPG a n d all other readers.
526
Production Histories
Brent Hale
Northwest Pipeline Corporation Salt Lake City, Utah, U.S.A.
INTRODUCTION
Production histories of gas and oil wells can be analyzed to estimate reserves and production rates and to validate results of complex reservoir studies. Because accurate p r o d u c t i o n data are commonly available on most wells, production data analyses can be widely applied by the production engineer. Two main analysis methods are decline curve analysis and type curve analysis. These methods are simple and economic to a p p l y a n d can p r o v i d e valuable data on the quality of the reservoir. Decline curves are generally easy to use, but type curves usually provide better forecasts for complex reservoirs. If l a r g e r i s k s o r l a r g e c a p i t a l e x p e n d i t u r e s a r e n e e d e d to conduct a project, well testing and simulation can be used to s u p p l e m e n t a n d / o r u p g r a d e the q u a l i t y of f o r e c a s t s generated by production histories (see the chapters on "Production Testing" in Part 9 and "Reservoir Simulation" in Part 10).
This equation directly relates reserves to the rate change over the life of the well a n d to the constant percent decline rate for the well.
Harmonic and Hyperbolic Decline
H a r m o n i c a n d h y p e r b o l i c d e c l i n e are v a r i a t i o n s of constant percent decline. The decline rate changes as a well ages, as seen f r o m 0 to 6 years in Figure 1. A n a d d i t i o n a l c o n s t a n t , n, w h i c h d e s c r i b e s h o w t h e initial d e c l i n e rate, Di, c h a n g e s w i t h t i m e is n e e d e d to d e s c r i b e t h e flow b e h a v i o r d u r i n g years 0 t h r o u g h 6 in Figure 1. T h e rate e q u a t i o n is
q = Qiil +nDitr1'"
where n varies from 0 to 1 but is constant for a single decline curve. Harmonic decline is a special case in which n = 1. To compute reserves, the following equation is used:
DECLINE CURVE ANALYSIS
Decline curve analysis relates past p e r f o r m a n c e of gas a n d oil wells to future performance, but it does not anticipate changes in performance d u e to operating conditions or changes in reservoir behavior (for more on decline curves, see the chapter on "Reserves Estimation" in Part 10.)
Constant Percent Decline
When the production decline rate is a constant, constant percent decline can be used effectively to forecast future p e r f o r m a n c e . T h e log of flow r a t e is p l o t t e d v e r s u s time. In Figure 1, a straight line exists f r o m 6 years to 24 years. This constant decline can be extrapolated to determine future rates with the following equation:
q = qi(l + D)t
I„n\l A-n
„1-й
NP =
J^i
-q
T^n
The mathematics here are more complex than for constant percent decline, but are necessary to describe more complex production histories.
TYPE CURVE ANALYSIS
Type curves differ from decline curves in one important respect: type curves are theoretical forecasts of reservoir
CO O
Hyperbolic
1000-
where
<D
q = production rate at time t
qj = initial rate
<3
QZ
D = decline rate
t = time
Reserves, Np, or the volumes p r o d u c e d over the life of a project are
-300-
^l-OO-
N _ (?i-?a) p ln(l + D)
w h e r e q = a b a n d o n m e n t flow rate
O
6
12
18
24
Time
FigureL Semi-log decline history.
527
Production Histories
461
IO-2
IO-1
IO0
IO1
IO2
IO3
IO4
Dimensionless Time
IO-3
IO-2
IO"1
IO0
IO1
IO2
IO3
Year
Figure 2. Simple type curve.
Figure 3. Log-log production decline history.
performance and can be used to forecast reservoir behavior that has not yet been observed. Because there are many different types of reservoirs, a great variety of type curves have been generated. Type curves are generated using analytical or numerical solutions to Darcy's law. A simple type curve is s h o w n in Figure 2. The reciprocal dimensionless rate is plotted versus dimensionless time. The curve is suitable for wells produced at constant pressure where the rate declines as a function of time. (For m o r e information on dimensionless variables and type curves, see the chapter on "Pressure Transient Testing" in Part 9.)
Flow Regions
T h r e e flow r e g i o n s can exist o n a typical t y p e c u r v e , s u c h as that illustrated in Figure 2. These are the infinite acting region or unsteady-state region, the transition region, and the semi-steady-state region. Early in its producing life, a well acts a s if it is p r o d u c e d f r o m a n i n f i n i t e l y l a r g e r e s e r v o i r . F l o w rate d e c r e a s e s r a p i d l y a s a well p r o d u c e s b e c a u s e fluids must move increasingly longer distances to get to the well. T h e c o n s t a n t l y c h a n g i n g r a t e s a r e called unsteady-state rates. As boundaries are encountered, a transition occurs and rate changes begin to reflect depletion of gas or oil in a limited reservoir. Fully developed boundaries result in production at steady-state rates if p r o d u c e d f l u i d s a r e r e p l a c e d b y w a t e r i n f l u x . If t h e r e s e r v o i r p r e s s u r e d e c r e a s e s a s a w e l l is p r o d u c e d , t h e flow rate will also decrease. Semi-steady-state is defined as the production mechanism when the pressure everywhere in the reservoir is declining uniformly.
Dimensionless Variables
Most type curves are presented using dimensionless rate or dimensionless pressure versus dimensionless time. Both axes of the g r a p h s are presented as logarithms. Data are multiplied by a constant to generate dimensionless variables for convenience so that a single type curve can be used for different wells with different reservoir properties.
A typical analysis consists of three main phases. First, an appropriate type curve must be selected by answering the following questions. Are we analyzing a gas or oil reservoir?
Does it have high or low permeability? Is the reservoir naturally fractured? Is it necessary to consider complex rock a n d fluid p r o p e r t i e s ? Did the well p r o d u c e at a c o n s t a n t rate or a constant pressure?
Second, production data must be graphed according to the requirements of the type curve being used for the analysis. By g r a p h i n g the reciprocal rate versus time (Figure 3) for a well producing at constant pressure, one can then use the appropriate type curve to analyze the field data.
Finally, the data are matched to the type curve. When the m a t c h is complete, the axes of the type curve and the production curve are compared to determine the constant difference between them. With this relationship, the future performance can be forecast either graphically or by using computer models.
Flow Systems
Either linear flow or radial flow is commonly used to a n a l y z e p r o d u c t i o n d a t a a n d to f o r e c a s t r e s e r v e s . In linear flow s y s t e m s , fluid flows in p a r a l l e l p a t h s to a f r a c t u r e . In radial flow s y s t e m s , fluid flows in a p a t h that c o n v e r g e s to t h e wellbore, causing high pressure drops near the wellbore b e c a u s e of h i g h fluid velocity (see the c h a p t e r on "Fundamentals of Fluid Flow" in Part 10).
Linear flow is o b s e r v e d as a s t r a i g h t line (Figure 2) o n a log Дp or log 1 /cj v e r s u s log At g r a p h of field d a t a . Linear flow is common in low permeability reservoirs where pressure changes very slowly and fluids move slowly through the reservoir. Radial flow is expected in wells with high permeability or in old wells where the drainage area is large.
A l t h o u g h l o w p e r m e a b i l i t y is n e c e s s a r y for linear flow to occur over an extended time period, a hydraulic fracture must a l s o exist or t h e linear flow p a t t e r n s will q u i c k l y c h a n g e to r a d i a l flow p a t t e r n s . T h i s c h a n g e is w h e r e t h e s t r a i g h t line begins to curve to the right (Figure 2). After it begins, radial flow d o m i n a t e s p r o d u c t i o n until d e p l e t i o n occurs.
Other Factors
A n u m b e r of factors can influence p r o d u c t i o n data a n d cause it to deviate from ideal trends. The most common is
462
PART 9—PRODUCTION ENGINEERING METHODS
Table 1. Deviation of Actual Well Production from Expected Decline Rate
Deviation Production higher
than expected
Production lower than expected
Reasons
Poor initial test data Formation cleanup Inadequate initial formation
evaluation Mechanical problems Response to enhanced recovery
Poor initial test data Formation damage Mechanical problems Water influx Gas influx
well damage that results in higher than expected pressure drops near the wellbore and lower than expected production rates. Often, well damages change as a function of time d u e to factors such as scale deposition, salt or sulphur deposition, and the use of inhibitors in the tubing string. Another factor can be wellbore storage, which is the accumulation or d e p l e t i o n of gas in a w e l l b o r e as p r e s s u r e s increase or d e c r e a s e . This s t o r a g e of gas d i s t o r t s w e l l h e a d f l o w measurements. If the impact of storage a n d d a m a g e is not recognized and accounted for in the analysis, the resulting forecast can be incorrect.
WELL SURVEILLANCE
Production history analysis has significant economic and conservation benefits. Production problems need to be identified and corrected as soon as possible, usually with a
Initial Production
TIME (linear scale)
Figure 4. Operations impact on decline curve.
workover. To justify a workover economically, the engineer has to predict future performance under existing conditions and what would be expected after the workover. Incremental production or environmental protection must justify the cost of a n y w o r k o v e r project. M o n i t o r i n g completed projects is necessary to determine which workovers are successful and should be used in the future and which should not be applied to future wells.
By m o n i t o r i n g well p e r f o r m a n c e , o n e c a n o b s e r v e if a n d when the actual well production deviates from the expected decline rate. Table 1 illustrates possible reasons for deviation.
The information in Figure 4 illustrates how decline curves can be used to u n d e r s t a n d the production behavior of a well. In this figure, one can see that changes in operating conditions change the decline curve. Reserves and economics can be forecast by using exponential decline curve fits to the production data.
Well Completions
Sk'phenA-HoIditch
S. A. Holditch & Associates, Inc. College Station, Texas, U.S.A.
PURPOSE OF WELL COMPLETIONS
The p u r p o s e s of a well completion are to
1. C o n n e c t t h e reservoir to t h e s u r f a c e so that fluids can be produced from or injected into the reservoir
2. Provide a conduit for well stimulation treatments 3. Isolate the producing reservoir from other zones 4. Protect the integrity of the reservoir, especially in
unconsolidated formations 5. P r o v i d e a c o n d u i t to m e a s u r e t h e c h a n g e s in flow r a t e
and pressure needed to run a well test
The wellbore provides the only path from the surface to the reservoir. To a large extent, the successful production a n d depletion of a reservoir d e p e n d s u p o n the successful completion and workover operations applied to a well. In most companies, the production engineer designs and supervises all completion and workover operations.
TYPES OF COMPLETIONS
C o m p l e t i o n s can b e d i v i d e d into t h r e e categories: open hole completions, liner completions, a n d perforated casing completions. In most wells, conventional single perforated casing completions are used; however, multiple, alternate, or slim hole completions may be used under certain conditions. The choice of completion type should be closely coordinated with the d e v e l o p m e n t of the reservoir m a n a g e m e n t plan. For example, the size, weight, a n d g r a d e of the tubular g o o d s will be determined based u p o n the ultimate use of a wellbore. A n injection well may require stronger casing than a production well.
Open Hole Completions
Open hole completions were originally used in the early days of the petroleum industry w h e n most wells were drilled with cable tools. Normally, casing was run as the hole was drilled. When the formation was penetrated and oil and gas b e g a n to flow, drilling ceased a n d t h e w e l l w a s p r o d u c e d a s an open hole completion. As rotary rigs began to drill a majority of the wells, it w a s still c o m m o n to complete a well u s i n g a n o p e n h o l e c o m p l e t i o n . If t h e w e l l n e e d e d to b e stimulated, nitroglycerine was used to rubbleize the formation near the wellbore. Figure 1(a) illustrates a typical wellbore diagram for an open hole completion. Although not common in most areas, open hole completions are still used today in certain situations, such as the horizontal well completions in the Austin Chalk and in the Devonian shales in Appalachia. In a typical open hole completion, casing is set prior to drilling into the producing interval. A n o n d a m a g i n g fluid c a n t h e n b e u s e d to d r i l l i n t o t h e p a y section.
O n e important d i s a d v a n t a g e of a n openhole completion is that production casing must be set prior to drilling and l o g g i n g t h e reservoir. If for s o m e geological or e n g i n e e r i n g reason the target formation is not productive, then money has been spent to set casing in a well that may be plugged as a d r y hole. A n o t h e r d i s a d v a n t a g e is the lack of control that occurs w h e n an open hole completion is made. One cannot control t h e flow of fluids f r o m t h e reservoir into t h e w e l l b o r e nor the injection profile in an open hole completion. In a d d i t i o n , if t h e f o r m a t i o n is n o t c o m p e t e n t , s l o u g h i n g z o n e s can cave into t h e w e l l b o r e a n d restrict flow to t h e surface.
Liner Completions
Liner completions are similar to open hole completions in that casing must be set prior to drilling the producing interval. The a d v a n t a g e of a liner completion is that the d r i l l i n g fluid s y s t e m c a n b e c h a n g e d a n d a n o n d a m a g i n g fluid u s e d to drill t h e p a y i n t e r v a l . S e v e r a l t y p e s of liner completions are commonly employed in well completions. These include
• Slotted Uner • Screen and liner • Cemented Hner
The slotted liner completion is similar to an open hole completion and has all the major advantages and disadvantages discussed for open hole completions. The
(x
a
) '
„
.
Production
x( b7)
— Casing —
Liner ^ Hanger
Figure 1. Wellbore diagram of (a) an open hole completion and (b) a slotted liner completion.
463
462
PART 9—PRODUCTION ENGINEERING
METHODS
(b)
Production ~ Casing —
Liner Hanger
Production — Tubing
-4
Packer
Figure 2. Wellbore diagram of (a) a screen and liner completion and (b) a cement liner completion.
only difference is that a slotted liner is h u n g in the open hole interval to m i n i m i z e s l o u g h i n g of the f o r m a t i o n into the wellbore (Figure lb). A screen and liner completion is similar to the slotted liner completion in that a screen and liner is set in the open hole section of the wellbore. The difference is that gravel is sometimes placed behind the screen (Figure 2a). The advantages and disadvantages are the same as for open hole completions. The screen and liner completion is used primarily in unconsolidated formations to prevent the m o v e m e n t of f o r m a t i o n m a t e r i a l s into the w e l l b o r e , restricting t h e flow of reservoir fluids.
The cemented liner completion is used when intermediate casing is set in a well prior to reaching total depth (Figure 2b). Many times intermediate casing is used to isolate zones behind pipe such as low pressured intervals that tend to cause lost circulation problems or to isolate zones such as sloughing shales or salt layers. Intermediate casing is also set in transition zones between normally pressured intervals and geopressured intervals. After the casing is set, the weight or chemistry of the drilling fluid can be changed to continue drilling the well. The c e m e n t e d liner completion is advantageous because the particular intervals behind the liner can be selectively perforated. This selection will allow one to c o n t r o l b o t h t h e p r o d u c t i o n a n d injection of fluids in t h o s e intervals. The m a i n d i s a d v a n t a g e of a cemented liner is the difficulty encountered in obtaining a good primary cement job across t h e liner. If a g o o d c e m e n t job is o b t a i n e d , t h e n a cemented liner completion is very similar to a perforated casing completion.
Perforated Casing Completions
A perforated casing completion, illustrated in Figure 3a, is the most commonly used completion technique today. The m a i n a d v a n t a g e of this type of completion is that the well can be drilled and logged to total depth prior to running and cementing production casing. By obtaining cores a n d logs of
Figure 3. Wellbore diagram of (a) a perforated completion and (b) a single completion inside perforated casing.
the potential producing interval, one can estimate the economic value of that wellbore prior to committing f u n d s to complete the well. Another advantage of a perforated casing completion is that it is easier to obtain a good cement job, c o m p a r e d to a liner c o m p l e t i o n . If t h e p r i m a r y c e m e n t i n g job is properly performed, one can selectively produce from and inject into the reservoir and isolate intervals as required. The p r o d u c t i o n of a n u n c o n s o l i d a t e d f o r m a t i o n can be accomplished by performing an inside gravel pack. Finally, the perforated casing completion is quite adaptable to multiple completions and alternate completions.
Single Completions
Once the decision has been made to either complete a well using an open hole completion, liner completion or perforated casing completion, a decision must be made on how many different intervals will be produced. The most common method is the single completion in which only one interval is produced at a time (Figure 3b). A single completion is simple and results in fewer operating problems and less cost than multiple completions. Single completions are common on l a n d w h e r e t h e reservoirs a r e either s h a l l o w or v e r y d e e p . If the formation is very shallow, then drilling costs are minimal and single completions are usually best. In very deep wells, single completions are preferred because of the complexity and expense involved with a dual or triple completion in reservoirs deeper than 10,000 ft.
Multiple Completions
In certain cases, multiple completions may provide the best control of reservoir operations. Multiple completions include the tubing-casing dual, dual tubing strings, and the typical triple completion consisting of three tubing strings. Figures 4 and 5 illustrate these multiple completions.
The more complex the completion, the more trouble one can expect in both completion operations and in subsequent workover operations. Multiple completions should be considered only in special situations. These situations include
Single Tubing String •
Dual
Tubing
String
(b)
Well Completions
465
3 Tubing Strings
Dual • Packer
Triple Packer
E
NNNNN
Dual I l t l l Packer
• •••/ N N/N/N4S/NN4/N4\/N/чs •
is
Figure 4. Wellbore diagram of (a) a casing-tubing dual completion and (b) a completion with dual packers and dual tubing strings.
Single
^
Packer
areas where drilling costs are very high or where the area allocated for drilling wells is at a premium. Such areas include offshore areas, highly populated areas, and remote land locations.
The main advantage of multiple completions is that two or more reservoirs can be simultaneously produced from a single wellbore. If all g o e s p r o p e r l y , t h e e c o n o m i c s of u s i n g multiple completions can be attractive; however, operating and workover costs can be quite high. These operational factors must be considered when considering multiple completions.
Alternate Completions
Alternate completions are used to allow inexpensive recompletions. Normally, a single interval is completed and produced to abandonment and then, using wireline tools, the lower interval is isolated and the upper interval opened to production (Figure 6). Alternate completions eliminate the n e e d for costly w o r k o v e r s u s i n g rigs. It is possible to h a v e a dual well with two alternate completions; however, as the completion gets more complex, more problems will occur. Alternate completions are used in areas where workover costs by conventional methods can be quite high. The most c o m m o n u s e of alternate completions is offshore.
Slim Hole Completions
In certain reservoirs, the economic value of the reserves d o e s not justify the e x p e n d i t u r e s of n o r m a l c o m p l e t i o n m e t h o d s . A n e x a m p l e of such a reservoir w o u l d b e a series of thin, small lenticular dry gas intervals. Under certain conditions, the reservoir might not w a r r a n t the use of large production casing and tubing to produce the gas. For such reservoirs, slim hole completion techniques might be justified
Figure 5. Wellbore diagram of a conventional triple completion.
in which small pipe, such as 4-1/2-in. or 2-7/8-in. casing, is run and cemented in the wellbore. This small casing serves as the tubing and casing for the wellbore. This completion practice is used to minimize completion costs and should be used only for wells that are expected to have short producing lives and that do not require artificial lift.
Special Considerations for Producing and Injection Wells
The completion technique applied to a well depends on whether the well is a producing well or an injection well. For a producing well, the n u m b e r of different reservoirs that m a y eventually be produced is significant. A decision must be made on whether a single or multiple completion will be made. O n e should be a w a r e of stimulation treatments that may be required or the possible need for artificial lift e q u i p m e n t to p r o d u c e fluids f r o m t h e wellbore. I n f o r m a t i o n on Christmas trees, tubing spools, packers, and downhole equipment needed to produce the well should be carefully designed and included in the completion program.
If a well is to b e a n injection well, o n e s h o u l d b e a w a r e of the intervals that will require completion for the injection of fluids. T h e e n g i n e e r s h o u l d h a v e a n i d e a of t h e a n t i c i p a t e d injection rates a n d pressure that will influence the design of
462
PART 9—PRODUCTION ENGINEERING
METHODS
Circulation Sleeve
_Alternate Zone
Main Charge
Casei \ :—
Detonating
Primer / Charge
-Liner
^
—
Figure 7. Construction of a typical shaped charge.
Producing Zone
Figure 6. Wellbore diagram of a single completion with an alternative zone.
the completion program. Extra effort should be made to ensure that the casing, tubing, wellhead, and cement programs are carefully designed and that the mechanical condition of the well d u r i n g completion is excellent.
PERFORATING
The major p u r p o s e of perforations is to provide effective flow communication between the wellbore and the reservoir. To create a perforation, a perforating gun is used to "shoot" a hole through the casing and cement. The hole is called a perforation tunnel a n d p r o v i d e s a p a t h w a y for fluid flow f r o m the reservoir to the wellbore.
O v e r 90% of w e l l s a r e p e r f o r a t e d u s i n g a shaped charge perforating system. The shaped charge, or jet charge, came into use in the oil field shortly after World W a r II. The perforating shaped charge is a derivative of the charge used to construct a bazooka. The shaped charge has essentially replaced the bullet gun, which was the main perforating tool prior to World W a r II.
Figure 7 illustrates the c o m p o n e n t s of the s h a p e d charge. The m a i n characteristic of the s h a p e d charge is the coneshaped cavity. To fire the shaped charge, the primer cord ignites the primer charge, which in turn ignites the main explosive charge. A s the p o w d e r burns, it reaches the apex of the cone. The cone collapses and creates a jet stream that is capable of penetrating the steel casing, cement, and formation. Figure 8 illustrates the f o r m a t i o n of the jet stream f r o m a shaped charge.
Types of Guns
All perforating guns can be classified into three broad categories:
1. E x p e n d a b l e g u n 2. Semi-expendable gun 3. Retrievable, hollow carrier g u n
Expendable Gun
A n expendable gun is o n e in w h i c h all p a r t s of t h e g u n below the collar locator are either destroyed during the perforation explosion or left in the well as debris in the rathole. Expendable guns are used in shallow wells and usually cost less than other types of guns. Expendable g u n s allow the use of larger jet charges; however, the g u n s can cause casing damage in unsupported casing.
Semi-Expendable Gun
A semi-expendable gun h a s a m e t a l b a r that h o l d s t h e s h a p e d charges. After firing, some of the g u n disintegrates, part of the jet charges fall to the b o t t o m of the hole, a n d the metal bar is recovered. The semi-expendable gun costs slightly more that an expendable gun with comparable shot density. Like the expendable gun, the semi-expendable gun can use larger jet charges than the hollow carrier guns, a n d the possibility of casing damage exists.
Retrievable Hollow Carrier Gun
A retrievable hollow carrier gun is the most widely used gun because it
• Is liiglily reliable • Is mechanically rugged and strong • Can be used in high temperature and pressure
environments • Leaves no debris in the hole • Eliminates casing damage • Offers high charge performance
The hollow carrier gun contains all parts inside a steel tube that is resistant to wellbore fluids. When the gun is fired, most debris remains inside the tube and is retrieved at the surface.
Well Completions
467
Shaped Charge
Г CT
Pressure On Target ~ 5 x 106 psi
Velocity Of Jet
20,000 ft/sec
~
Jet
^sT /
-¾
Figure 8. Jet stream generated by shaped charge.
Gun Size and Method of Conveyance
Perforating guns also come in different sizes and can be run using a wireline or tubing.
Casing Gun
Typical casing guns range from 3-1/8 to 5 in. outside diameter. These large diameter guns have to be run inside c a s i n g w i t h t u b i n g o u t of t h e hole. C a s i n g g u n s are advantageous because maximum size jet charges can be used, resulting in maximum hole size and penetration.
Tubing-Conveyed Gun
A tubing conveyed g u n is a large diameter casing g u n (31 / 8 to 5 in.) that is r u n on t h e e n d of t u b i n g . This arrangement allows one to use a large diameter gun and to perforate the well with tubing in the hole and the wellhead installed. The gun can be fired by dropping a bar inside the tubing.
Thru-Tubing Gun
The guns that are run inside the tubing to perforate the casing are called thru-tubing guns. Typical thru-tubing guns range in diameter from 1-3/8 to 2-3/8 in. Because thru-tubing guns are much smaller than casing guns, smaller jet charges must be used; therefore, the resulting perforations are smaller in diameter and length, compared to a casing gun.
Perforating Methods
A well can be perforated three different ways:
• Overbalanced using a casing gun • Underbalanced using a tubing-conveyed gun • Underbalanced using a thru-tubing gun
W h e n t h e well is overbalanced, t h e p r e s s u r e in the w e l l b o r e is greater than the reservoir pressure. After firing the gun, the w e l l b o r e f l u i d s will f l o w into t h e p e r f o r a t i o n s . If m u d is u s e d , the solids in the mud will plug the perforations; therefore,
Figure 9. Two perforating methods: (a) Perforating overbalanced with a casing gun and (b) perforating underbalanced with a thrutubing gun.
462
PART 9—PRODUCTION ENGINEERING METHODS
clean wellbore fluids are preferred when perforating overbalanced. In contrast, when the pressure in the wellbore is less than the formation pressure, the situation is called underbalanced perforating. To perforate underbalanced, one should use either a tubing-conveyed gun or a thru-tubing gun. Also, the wellhead should be installed and a pressuretested lubricator should be on the wellhead.
Figure 9 illustrates the two methods: (1) overbalanced p e r f o r a t i n g with a casing g u n a n d (2) u n d e r b a l a n c e d p e r f o r a t i n g w i t h a t h r u - t u b i n g g u n . Figure 10 s h o w s qualitatively the difference in hole size and hole penetration for the two methods. The selection of the perforating method is influenced by the d e p t h of the zone, the temperature, the pressure, the need for well stimulation, and the need for sand control procedures.
Figure 10. Qualitative description of perforations using (a) a casing gun and (b) a thru-tubing gun.
Stimulation
John L. Gidley
John L. Gidley & Associates, Inc. Houston, Texas, U.S.A.
INTRODUCTION
The need for well stimulation arises f r o m either one of the following conditions:
1. F o r m a t i o n permeability is i n a d e q u a t e to allow the well to produce at rates high enough for the timely recovery of investment in drilling a n d completing the well.
2. The well has been completed in a formation having adequate permeability, but the formation near the wellbore has been damaged by the drilling or completion process.
The first case, low formation permeability, requires a reservoir stimulation technique, which can be either hydraulic fracturing or acid fracturing. The second case, formation damage, requires a damage removal technique, which is usually matrix acidizing but m a y occasionally involve hydraulic fracturing. Solvent or surfactant treatments are sometimes used for damage removal.
Figure 1 illustrates the division of well s t i m u l a t i o n problems on the basis of formation permeability. Formations h a v i n g an average effective permeability of 1 m d (millidarcy) or less generally require reservoir stimulation methods, while those h a v i n g effective permeabilities of 10 m d or greater generally require damage removal methods. Those formations in between these two permeabilities may require either type of treatment.
A f e w examples m a y serve to illustrate this m e t h o d of classifying well stimulation problems. The "tight gas sands" of the Rocky Mouxitain area h a v e effective permeabilities often in the microdarcy r a n g e (1 |id, or microdarcy, = 0.001 m d ) . If e c o n o m i c c o m p l e t i o n s a r e to b e d e v e l o p e d f r o m s u c h formations, hydraulic fracturing (a reservoir stimulation method) must be used to stimulate reservoir performance (see the chapter on "Evaluating Tight Gas Reservoirs" in Part 6).
In contrast, the younger permeable formations along the Gulf Coast, such as those of Miocene age, often have very high permeabilities—commonly, several hundred millidarcies. When such formations are unproductive upon initial completion, it often indicates severe formation damage. Although hydraulic fracturing methods have occasionally been employed to stimulate such wells, s a n d s t o n e a c i d i z i n g t e c h n i q u e s , w h i c h are c a p a b l e of removing near-wellbore damage, are normally the preferred stimulation method.
HYDRAULIC FRACTURING
The p u r p o s e of hydraulic fracturing is to change the flow pattern in the reservoir from one that converges radially on t h e w e l l b o r e (Figure 2) w i t h flow resistance c o n c e n t r a t e d n e a r t h e wellbore, to o n e w h e r e flow in t h e reservoir is linear to a h i g h l y c o n d u c t i v e f r a c t u r e t h a t c o n d u c t s fluid to t h e w e l l b o r e w i t h m i n i m a l flow resistance (Figure 3). T h e success of this o p e r a t i o n d e p e n d s u p o n the conductivity of the h y d r a u l i c fracture and how successfully this conductivity can be retained following the treatment. These, in turn, depend largely u p o n the design a n d execution of the f r a c t u r i n g treatment.
Fracturing Treatment
A hydraulic fracture is created in a subsurface formation w h e n a fluid is p u m p e d into t h a t f o r m a t i o n at a r a t e faster than the formation can accept the fluid through its matrix
Wellbore
i r r^.—
Hydraulic Fracturing
Matrix Acidizing
And
Surfactant Treatments
•
л
Acid Fracturing
Solvent Treatments
Minifracs
I
I
I
0.1
1
1 0
100
1000 10,000
Formation Permeability, md
I i
Reservoir Stimulation
Damage Removal
Treatments
Treatments
Figure 1. Spectrum of stimulation treatment applicability.
Figure 2. Flow to unfractured well converges at wellbore. 469
462
PART 9—PRODUCTION ENGINEERING
METHODS
Fracture
ч
// /
\ \ \
T=T==7
^=LJL
Wellbore
Figure 3. Flow to fractured well is linear to fracture, followed by flow that is down fracture to the wellbore.
permeability. For example, Figure 4 illustrates the pressure v e r s u s time b e h a v i o r at the surface of a well that is being fractured. As fluid is p u m p e d into the target formation at a rate sufficient to fracture the well, the pressure at the formation face builds up rapidly to the point that the formation fails and a fracture is developed.
Continued injection at high rate causes the fracture to extend from the wellbore, and this growth in fracture length is accompanied by an increase in fracture width at the wellbore a n d p r o p o r t i o n a t e l y t h r o u g h o u t t h e fracture. T h e fluid b e i n g pumped in the fracture at this time contains no proppant and is often referred to as the "pad." Its success in extending the f r a c t u r e d e p e n d s u p o n h o w m u c h of t h e fluid injected can b e retained within the fracture. Leak-off through the face of the f r a c t u r e r e d u c e s t h e a m o u n t of fluid r e t a i n e d in t h e f r a c t u r e . Only that fluid retained within the fracture is effective in extending the fracture. To reduce fluid leak-off, certain m a t e r i a l s a r e a d d e d to t h e f r a c t u r i n g fluid to f o r m a filter cake on the face of the fracture a n d to p l u g the porosity a n d permeability of the face of the fracture.
As fracture width grows, a point is soon reached where the fracture can accept a propping agent, and the proppant is a d d e d to t h e f r a c t u r i n g fluid (Figure 5). Typically, p r o p p a n t is a d d e d initially at a l o w c o n c e n t r a t i o n ( a b o u t 1 lb of p r o p p a n t per gallon of f r a c t u r i n g fluid), then the concentration is s u b s e q u e n t l y increased as it b e c o m e s apparent that the lower concentrations of p r o p p a n t are being accepted b y the fracture. The p u r p o s e of the p r o p p a n t is to hold the fracture open. Common proppant materials include sand, resin-coated sand, and ceramic beads.
W h e n the desired a m o u n t of p r o p p a n t has been p u m p e d into the fracture, p u m p i n g is ceased. The fracture begins to close while the system remains shut in, as s h o w n in Figure 6. After a few minutes or hours, depending on the conditions, t h e s u r f a c e v a l v e s a r e o p e n e d a n d fluid is p r o d u c e d at t h e surface from the fracture. This cleanup period is designed to r e m o v e t h e f r a c t u r i n g fluid f r o m t h e f o r m a t i o n a n d to initiate t h e p r o d u c t i o n of reservoir fluids. Reservoir " c l e a n u p " o c c u r s over a period of several days for relatively permeable wells to a period of w e e k s or m o n t h s for l o w permeability wells. Typically, d u r i n g the c l e a n u p period, only 30-70% of the f r a c t u r i n g fluid u s e d in t h e o p e r a t i o n is r e c o v e r e d . H o w e v e r , the success or failure of the fracture t r e a t m e n t d o e s not u s u a l l y correlate w i t h t h e p e r c e n t a g e of fluid r e c o v e r y d u r i n g the cleanup period.
LlJ <H <о
U-
Ш Z
§1
Cn H(Л < ш S
DSF- О=
LL
Breakdown Pressure
TIME
Figure 4. Formation fails and fracture is initiated.
Rather, the success or failure of the fracturing treatment depends more upon how successfully the proppant was placed in the fracture, the conductivity of the p r o p p a n t within the fracture after the treatment, and the potential of the well as an appropriate stimulation candidate, in other words, its remaining reservoir pressure, reserves, and other considerations. In some formations, where the hydrocarbon is located p r e d o m i n a n t l y in natural fractures, the success of the treatment d e p e n d s u p o n the orientation of the hydraulic f r a c t u r e relative to the p r e f e r r e d direction of the n a t u r a l fractures.
Fracture Orientation
A hydraulic fracture will propagate in a plane perpendicular to the least principal earth stress. For example, consider a well at 1000 ft in a tectonically relaxed area with a terrastatic gradient of 1.0 p s i / f t of depth. In such a shallow well, the overburden stress can be overcome more easily than either of t h e t w o p r i n c i p a l h o r i z o n t a l stresses. If a n a t t e m p t is m a d e to fracture such a well by p u m p i n g into it at high injection rates, the pressure adjacent to the exposed formation will increase rapidly. When that pressure becomes 1000 psi or slightly greater, there is sufficient pressure to lift the overburden. At this point, a horizontal fracture occurs (that is, a fracture perpendicular to the vertical wellbore or parallel to the surface of the earth).
Alternatively, consider a well that is 5000 ft deep in a formation h a v i n g an average terrastatic gradient of 1.0 p s i / f t and is also in a tectonically relaxed area. When an attempt is m a d e to fracture this well hydraulically, it is found that a fracture can be p r o p a g a t e d with a b o t t o m h o l e pressure of 3500 psi. Since this pressure is inadequate to lift the o v e r b u r d e n weight of the earth, it is inferred that the fracture is not horizontal b u t rather vertical (that is, in the plane of the vertical axis of the wellbore). Consequently, a good rule of t h u m b is that in wells deeper than about 2000 ft, in
Stimulation
471
Figure 6. Injection is terminated and fracture closes on the proppant.
Ol
ь< <о
LL
ш
§
1 г
сл I -
(/) < ш 5
SF
с
о
LL
TIME
Figure 5. Proppant is introduced to fracture.
tectonically relaxed areas, most hydraulic fractures will be vertical.
The a z i m u t h direction of the hydraulic fracture d e p e n d s u p o n the relative m a g n i t u d e of the two principal horizontal stresses. Contrary to certain folk lore, there is no effective method for redirecting the direction of a vertical fracture in a deep well (2000 ft or deeper). It depends entirely u p o n the relative m a g n i t u d e of the t w o principal horizontal stresses in the neighborhood of the wellbore. The hydraulic fracture will propagate perpendicular to the least principal stress. Fracture orientations tend to remain roughly the same over large geographical areas. East of the Rocky M o u n t a i n area, vertical fractures tend toward a northeasterly direction. This orientation is altered somewhat in the Gulf Coast area where fractures seem to parallel the coast line. In the Rocky Mountain area, fractures occur in unpredictable directions that are highly d e p e n d e n t u p o n the direction of local tectonic forces.
Fracturing Fluids and Proppant Placement
T h e effectiveness of a f r a c t u r i n g t r e a t m e n t is largely dependent upon how successfully the proppant has been placed within the productive zone and how much conductivity the p r o p p e d fracture has relative to that of the formation being stimulated.
Placement of the p r o p p a n t within the fracture is governed b y t h e p r o p e r t i e s of t h e f r a c t u r i n g fluid. C e r t a i n fluids totally suspend the proppant during its transit down the wellbore a n d into t h e f r a c t u r e . O t h e r fluids a l l o w t h e p r o p p a n t to d r o p through the fluid after the fluid and proppant enter the f r a c t u r e . T h e first c a t e g o r y of fluids is o f t e n r e f e r r e d to a s perfect support fluids, w h e r e a s t h e s e c o n d is d e s c r i b e d a s equilibrium bank fluids.
Perfect support fluids are created by adding to the f r a c t u r i n g fluids m a t e r i a l s t h a t a r e c h e m i c a l l y r e a c t i v e a n d
c a p a b l e of d e v e l o p i n g s t r u c t u r e w i t h i n t h e fluid. G u a r g u m , a natural product from the guar bean, builds viscosity in water. In addition, certain materials, typically organometallic c o m p o u n d s , react with solutions of g u a r g u m and build s t r u c t u r e b y "cross-linking" t h e g u a r p o l y m e r s . T h e s e fluids, when properly designed and formulated, permit the suspension of the p r o p p a n t within the fracture t h r o u g h o u t the fracture closure period. By this means, p r o p p a n t distribution throughout the fracture can be made to conform to that designed in the treatment.
In very hot wells, especially those treated at high pumping rates, cross-linked fluids will lose m o s t of their viscosity b y shear degradation during the pumping process and through e x p o s u r e to h i g h t e m p e r a t u r e s . If t h e s e fluids lose viscosity, they also lose their ability to suspend the proppant. This p r o b l e m has been corrected with the d e v e l o p m e n t of delayed, cross-linked gelled f r a c t u r i n g fluids. Such fluids first b e c a m e available d u r i n g the early 1980s. T h e y are capable of retaining their viscosity within the fracture even under the severe conditions described.
T h e second t y p e of f r a c t u r i n g fluid, e q u i l i b r i u m b a n k fluids, a l l o w t h e p r o p p a n t to fall t h r o u g h t h e f r a c t u r i n g fluid and settle in an equilibrium bank that builds away from the wellbore. Such fluids may provide a fracture completely filled with p r o p p a n t near the wellbore, b u t the height of the proppant bank decreases rapidly with distance away from the wellbore. T h e s e fluids h a v e a p p l i c a t i o n in d a m a g e r e m o v a l treatments where short, highly conductive fractures are required, but they have limited utility in creating long fractures with high conductivity.
Fracture Conductivity
A n i m p o r t a n t factor affecting the success or failure of a h y d r a u l i c f r a c t u r e t r e a t m e n t is t h e c o n d u c t i v i t y (in situ) of t h e fracture developed. The role of this variable can b e illustrated by a graph (Figure 7), first presented by McGuire and Sikora in 1960. This graph shows h o w the productivity index ratio (//J0 shown on the vertical axis) developed by the treatment relates to the fracture length and the permeability contrast b e t w e e n the p r o p p e d fracture a n d the permeability of the formation.
Suppose a well that is 6000 ft deep is fractured and the confining stress on the proppant under the most severe p r o d u c i n g c o n d i t i o n will b e 4000 psi. If a 2 0 / 4 0 m e s h O t t a w a s a n d is u s e d d u r i n g the t r e a t m e n t , the permeability of this s a n d will be a b o u t 100 d (darcys). S u p p o s e that the w i d t h of the p r o p p e d fracture created is only 0.1 in. after fracture closure and that a 0.1-md formation is being stimulated. Assume that the well is on 40-acre spacing; therefore, the scale
462
PART 9—PRODUCTION ENGINEERING
METHODS
»f/xe =
SyS^
0.5
Zy/^^
0.4
0.3
/х
0.2
^^ '
0.1
I
I
I
IO2
IO3
IO4
IO5
IO6
12 wkf /40" RELATIVE CONDUCTIVITY, — 1 ^ —
Figure 7. Increase in productivity from fracturing.
factors in Figure 7 are 1.0 for both the vertical a n d horizontal axes. Also a s s u m e that a p r o p p e d fracture length of 660 ft has been created. This m e a n s the fracture extends to the radius of drainage for this well.
The McGuire-Sikora graph indicates that this treatment will result in a relative c o n d u c t i v i t y of IO5, L/re = 1.0 ( w h e r e L is t h e l e n g t h of t h e f r a c t u r e a n d re t h e d r a i n a g e r a d i u s ) , a n d a s t i m u l a t i o n ratio of a b o u t 11.5. That is, the well after fracturing should have a stabilized production rate of about 11.5 times the production rate prior to stimulation. This prediction a s s u m e s that the value of skin prior to the fracture treatment was zero (that is, no stimulation and no damage).
If this s a m e t r e a t m e n t h a d b e e n a p p l i e d to a reservoir w i t h p e r m e a b i l i t y of 10 m d i n s t e a d of 0.1 m d , the relative conductivity w o u l d h a v e been r e d u c e d b y a factor of 100 (that is, to IO3). For this case, the stimulation ratio would have been slightly greater than 2 as compared to 11.5 for the 0.1-md reservoir. To improve stimulation in the high permeability reservoir, a proppant having a permeability substantially greater than that of the Ottawa sand in the previous example would need to be used. However, even the best available 20/40 mesh proppants have a permeability only about four times greater than sand at these low confining stresses. Thus, the ability to i m p r o v e the response is limited to the permeability of the p r o p p i n g agent that can be afforded for each treatment. The other choice is to increase the packed fracture width. Fractures four or five times wider are possible, but are often difficult to create. Hence, highly permeable formations are difficult to stimulate with the existing materials and techniques now available by hydraulic fracturing, unless the formation has been damaged during the drilling or completion operations.
Geological Considerations Important in Fracturing
All types of productive formations can be stimulated by hydraulic (proppant) fracturing, including sandstones, carbonates, cherts, and shales. The main requirements for success in fracturing a well are that the formation contain hydrocarbons and have some natural permeability (whether
through natural fractures or within the matrix) to allow oil and gas to reach the hydraulic fracture. However, perhaps the single most important parameter that should be known prior to considering a well as a candidate for fracturing is the formation permeability. High permeability makes a well difficult to stimulate by fracturing, while low permeability formations can easily be stimulated by hydraulic fracturing.
The location of c o n f i n i n g shales is also i m p o r t a n t in f r a c t u r i n g . If t h e p r o d u c t i v e f o r m a t i o n is effectively isolated between two massive shales that may serve as barriers to the vertical extension of the fracture, then the hydraulic fracture will be confined, the fracturing fluid and materials can be concentrated upon the productive formation, and the treatment will normally be successful. Thus, certain lithological sequences are more conducive to effective stimulation by fracturing than others.
ACIDIZING
Acidizing can be divided into three types of operations:
1. Acid fracturing, w h i c h i n v o l v e s injecting acid at rates above those that will be accepted by the matrix, fracturing the formation, and etching the face of the f r a c t u r e to d e v e l o p a p e r m e a b l e flow p a t h
2. Matrix acidizing, w h i c h i n v o l v e s injecting acid at l o w rates to permit the u n i f o r m penetration of the formation without fracturing it
3. Acid washing, w h i c h i n v o l v e s u s i n g acid to d i s s o l v e scales and precipitates within the wellbore by moving acid across the encrusted surfaces. The acid used in acid washing seldom enters the formation and for that reason is not discussed further.
Acid Fracturing
Acid fracturing is accomplished in a manner very similar to that described for hydraulic fracturing. The only significant difference is that it involves formations that have significant solubility to acid and no p r o p p a n t is used d u r i n g the treatment. Instead, as the fracture is created, acid etches the face of the fracture. The regions that are etched most deeply supply the fracture porosity and conductivity needed to i m p r o v e flow to t h e wellbore.
The acid used in acid fracturing is normally hydrochloric acid (HC1). Occasionally, organic acids such as formic or acetic acid may be used, either alone or in mixtures with hydrochloric acid. Organic acids are used principally in hot wells (>250 T ) to reduce the corrosiveness of the acid mixture.
Hydrochloric acid is commonly used in acid fracturing because of its low cost per unit of dissolving p o w e r (hydrogen ion concentration) and because it develops soluble salts (principally the chlorides of calcium and m a g n e s i u m ) w h e n reacted with limestone or dolomite.
Typically, 15% (by weight) HCl is used, and this is often referred to as "regular" acid. Half strength acid (7.5% HCl) is used in some applications in which dissolving power is not the main consideration. Double strength acid (28% HCl) has found application in certain areas, but the higher strength is
Stimulation
473
more corrosive and requires more effective inhibitors than 15% HCl.
Acid f r a c t u r i n g is c o n f i n e d to the s t i m u l a t i o n of carbonates. Acid fracturing is not useful for stimulating sandstone formations because the acids reactive with sandstones, such as hydrofluoric-hydrochloric acid (HF/HC1) do not provide the differential etching on sandstones in the manner that hydrochloric acid does on limestones or dolomites.
Acid fracturing has both advantages and disadvantages relative to proppant fracturing (which may also be used on carbonates). A m o n g the a d v a n t a g e s of acid fracturing are that the operation can be carried out in the field with little risk of mechanical failure. Fracture treatments carrying p r o p p i n g agents will sometimes screen out; such failures do not occur with acid fracturing. In addition, acid fracturing can often be carried out for less expense than proppant fracturing because of the use of less e q u i p m e n t (for example, no p r o p p a n t handling e q u i p m e n t or blender is required with acid fracturing). Finally, acid fracturing can be designed with somewhat less sophisticated tools than proppant fracturing and thus can often be accomplished with greater certainty.
The d i s a d v a n t a g e of acid f r a c t u r i n g is that fluid loss during the fracturing operation cannot be controlled as effectively in as in proppant fracturing since the acid reacts with the formation and often bypasses the material being used to reduce fluid loss. Normally, fracture half-lengths of only 50-200 ft can be achieved with acid. The net effect is that acid fracturing cannot be used to create long fractures, especially in reservoirs with t e m p e r a t u r e s above 200 °F. To create long fractures in low permeability carbonates, hydraulic fracture treatments carrying a propping agent should be used. In many cases, the choice between proppant fracturing and acid fracturing can be made only after a thorough evaluation of the potential and limitations of each treatment for the specific job intended.
Matrix Acidizing
Matrix acidizing is used to enlarge the pore spaces for fluid flow by dissolving the pore lining materials. Matrix acidizing of limestone formations involves simply enlarging the pores by d i s s o l v i n g p a r t of the m a t r i x . Matrix a c i d i z i n g of sandstone formations is almost always directed at dissolving
pore lining materials such as clays that may adversely affect the permeability of the system w h e n exposed to fluids from the drilling or completion operation. (For more on fluid-rock interactions, see the chapter on "Rock-Water Interactions: Formation Damage" in Part 5.)
Matrix acidizing is simply a damage removal operation. Even in limestones, the a m o u n t of acid required to provide reservoir stimulation by dissolving formation and enlarging the wellbore can readily be shown to be exorbitant. Thus, in limestones this technique is often used to overcome nearwellbore damage, such as that caused in the perforating process, or to bypass plugged porosity caused by drilling or completion operations. True matrix acidizing is used very little in carbonates since these materials are not as susceptible to formation damage as are sandstones.
Matrix acidizing is mainly used in sandstone formations. Both hydrochloric acid and mixtures of H F / H C l acid are used in matrix acidizing. Hydrochloric acid is often an effective stimulant in formations that contain some soluble carbonate. In this case, dissolving the carbonate develops an alternate path for fluid flow and provides a measure of stimulation.
More frequently, mixtures of H F / H C l acid are required to stimulate formations where the damage resides in the clay fraction. The corrective action is to dissolve the clays near the wellbore and remove them completely. By this means, the zone most susceptible to damage during drilling or completion operations is cleaned of the material.
The acid formulation usually employed contains 3% HF (hydrofluoric acid) and 12% HCl. At higher temperatures, mixtures of 1.5% H F and 6% HCl appear to be more effective than the 3% HF formulation.
Mixtures of H F / H C 1 acid require greater care in their application than does HCl alone. For example, HF reacts with the sodium in formation waters to produce insoluble fluorosilicates or with calcium to form CaF2, an insoluble precipitate. For this reason, a spacer fluid must be used ahead of an H F / H C 1 t r e a t m e n t to displace the connate water. HF/HC1 treatments, especially on oil-bearing formations, require an afterflush containing additives and solvents to clean u p the products of reaction and avoid emulsification of the crude with acid byproducts. Certain mutual solvents are useful as additives in the afterflush to effect cleanup.
Production Testing
David E. Lancaster
S. A. Holditch & Associates, Inc. College Station, Texas, U.S.A.
INTRODUCTION
Production tests are r u n to obtain an indication of well productivity. Some production tests are performed in open hole (such as drill stem tests) and can be used in making completion decisions. Others (such as single-point, multipoint, and swab tests) are performed after the well is completed a n d generally involve routine m e a s u r e m e n t s of oil, gas, and/or water production under normal producing conditions. The test results can be used to determine reservoir properties, to assess the degree of d a m a g e or stimulation, to identify production and reservoir problems, or to satisfy the reporting requirements of regulatory bodies. Production tests can also be performed when more conventional well tests (such as pressure drawdown and buildup tests) are impractical due to time constraints, well conditions, or extremely low well productivity.
DRILL STEM TESTS
Drill stem tests (DSTs) are used to obtain (1) samples of the reservoir fluid, (2) m e a s u r e m e n t s of static b o t t o m h o l e pressure, (3) an indication of well productivity, a n d (4) shortterm flow and pressure buildup tests from which permeability and the extent of d a m a g e or stimulation can be estimated (Lee, 1982). A DST is run in the open hole after drilling, and is often used in deciding whether to complete a particular zone. The total test duration is frequently a function of hole condition, a n d the tool assembly m u s t be retrieved from the open hole after the test is completed (see the chapter on "Drill Stem Testing" in Part 3).
To run a drill stem test, a special DST tool is attached to the drill pipe and run in the hole opposite the zone to be tested. A DST tool typically includes two or more clock-driven, b o u r d o n - t u b e recording p r e s s u r e gauges, a set of flow valves, and one or two packers. The tool isolates the formation from the m u d column in the annulus. When the tool is opened, reservoir fluid can flow into the drill p i p e (and possibly to the surface); pressure is recorded continuously during the test.
Most DSTs (Figure 1) consist of t w o flow p e r i o d s a n d t w o shut-in periods. The pressure gauges record the initial hydrostatic m u d pressure (pihm) while going into the hole. The initial flow period (pm to pm) is a short p r o d u c t i o n period, usually only 5 to 10 min. Pressure rises d u r i n g the flow p e r i o d as fluid collects in the drill s t e m a b o v e the p r e s s u r e gauges. The objective is to release the hydrostatic m u d pressure and draw down the formation pressure only slightly. The first shut-in period (pfn to pis) s h o u l d b e long e n o u g h to allow the reservoir pressure to return to its initial value. A shut-in time of 1 h o u r is usually preferred.
In t h e s e c o n d flow p e r i o d (pif2 to pff2), t h e objective is to c a p t u r e a large s a m p l e of f o r m a t i o n fluid a n d to r e d u c e the pressure as far into the reservoir as possible. This flow period
s h o u l d be at least 1 hour, a n d if reservoir fluid is p r o d u c e d to the surface, flow rates should be m e a s u r e d . The second shutin p e r i o d (p{{2 to /Jfsi) is l o n g e r t h a n t h e first a n d is u s e d to estimate formation properties in a manner similar to that for analyzing conventional b u i l d u p tests. The duration of the final shut-in period d e p e n d s on well behavior d u r i n g the flow test; it m a y range from one-half to twice the flow time. C o m p a r i s o n of the final or extrapolated reservoir pressure from this second shut-in period to that from the initial shut-in period may suggest depletion has occurred during the DST. If so, t h e z o n e b e i n g t e s t e d is a l i m i t e d , n o n c o m m e r c i a l reservoir. Following the second shut-in period, the final hydrostatic m u d pressure is measured (Pfhm) and the DST tool is pulled from the hole.
SINGLE-POINT TESTS
Single-point tests are usually simple productivity tests that typically involve a m e a s u r e m e n t (or estimate) of initial or a v e r a g e reservoir p r e s s u r e a n d a m e a s u r e m e n t of flow rate a n d flowing b o t t o m h o l e p r e s s u r e (which can b e e s t i m a t e d from flowing surface pressure) at stabilized producing conditions (Allen and Roberts, 1978). From these data, the productivity index, PI, can be calculated as follows:
PI = 1
q/iB
(for oil) =
(for gas)
-2
2
(1)
P - P Wf
P -P w f
where q = flow rate (STB/day or M S C F / d a y ) p = initial or current average reservoir pressure (psia)
pwf = flowing bottomhole p r e s s u r e (psia) /л = viscosity (cp) B - formation volume factor (rcf/MSCF)
T h e p r o d u c t i v i t y index can be a u s e f u l indicator of well productivity a n d wellbore condition d u r i n g the life of a well. PI will generally decrease over time due to declining reservoir pressure, changes in producing conditions, a n d / o r production problems.
Single-point tests can also be used to estimate formation permeability (Lee et al., 1984) with an iterative solution of the transient r a d i u s of drainage equation (Equation 2) a n d the pseudosteady-state flow equation (Equation 3), as follows:
1/2
kt
=
(2)
376ф/лсt j
k- 141.2qnB In Jd- 3 + S,
(3)
HP-Pwf)
T 4
474
Production Testing
475
BASE UNE
UJ tr D (Л
СП
IxJ CC
QL
t
TIME
Figure 1. Typical drill stem test pressure chart. (From Earlougher, 1977.)
where rd = transient r a d i u s of drainage (ft) к = permeability (md) t = flow time (hr), generally best estimated by dividing
cumulative fluid production by flow rate, q ф = porosity (fraction) Ct = total compressibility (psia-1)
B = formation volume factor (rb/STB for oil or r b / M S C F for gas)
rw = w e l l b o r e r a d i u s (ft) h = net pay (ft) s' = a p p a r e n t skin factor
To solve for permeability, an arbitrary value of permeability is a s s u m e d (0.1 m d is often a good first estimate), a n d Equation (2) is solved for rd,. Then, this v a l u e for rd is used in Equation (3) to solve for permeability. For each iteration after the first, use the permeability calculated f r o m Equation (3) in solving for rd f r o m Equation (2); this p r o c e d u r e usually converges in three to four iterations.
The need to estimate an apparent skin factor, which is usually not k n o w n , is the biggest limitation of this method. Pressure buildup tests run in other wells in the same reservoir often p r o v i d e a g o o d estimate of typical skin factors. L o w permeability wells are generally broken down and balled out after completion and prior to testing; in these wells, a skin factor of - 1 to - 2 is o f t e n a r e a s o n a b l e a s s u m p t i o n . If a well has been damaged by drilling fluids and the perforations h a v e not b e e n b r o k e n d o w n , a skin factor of +2 to +5 (or more) is a p p r o p r i a t e (see the chapter on " F u n d a m e n t a l s of Fluid Flow" in Part 10).
The single-point test method for estimating permeability is valid for c o n s t a n t flow rate p r o d u c t i o n , c o n s t a n t b o t t o m h o l e pressure production, or smoothly changing bottomhole p r e s s u r e s a n d flow rates. T h e m e t h o d is r e c o m m e n d e d f o r e s t i m a t i n g p e r m e a b i l i t y f r o m p r e f r a c t u r e flow test d a t a only; it does not work well with postfracture flow data. The method is particularly useful in low permeability reservoirs w h e r e operators d o not run b u i l d u p tests routinely because of the long test times required to overcome wellbore storage
300
CM _o
OT
Q-
100
5 CL CM CL
IO 100
/ I / Q
A
/
/
/ /
/
/ Q
и n
O O O OitJ U-
<fO
IOOO
3 000
Figure 2. Multi-point test used to estimate absolute open flow potential.
e f f e c t s a n d r e a c h r a d i a l flow (see t h e c h a p t e r o n " P r e s s u r e Transient Testing" in Part 9).
MULTI-POINT TESTS
Multi-point tests are typically used to establish gas well
deliverability and absolute open flow potential; these tests
may also be referred to as gas well deliverability tests,
b a c k p r e s s u r e tests, or flow-after-flow tests (Allen a n d Roberts,
1978). Multi-point tests typically require the m e a s u r e m e n t of
g a s flow r a t e s a n d s u r f a c e p r e s s u r e s at f o u r stabilized flow
conditions; surface shut-in pressure is also measured.
Generally, a n increasing flow r a t e s e q u e n c e is p r e f e r r e d to a
decreasing rate sequence.
T h e s u r f a c e s h u t - i n a n d flowing p r e s s u r e m e a s u r e m e n t s
are converted to b o t t o m h o l e conditions a n d a log-log plot of
-2
2
V -P wf v e r s u s flow rate, q, is g e n e r a t e d (Figure 2). T h e f o u r
points define a straight line with a slope that is generally
b e t w e e n 0.5 and 1.0. This straight line is extrapolated to
determine gas flow rate at a point where the flowing
bottomhole pressure is zero; this rate is referred to as the
a b s o l u t e o p e n flow (AOF) potential of t h e well.
Multi-point test data can also be used to estimate
p e r m e a b i l i t y u s i n g a v a r i a b l e r a t e flow test a n a l y s i s ( O d e h
and Jones, 1965). For gas wells, the data are plotted as
p2~^fn In
versus
-' cIn
i% j=о
;
l
o
g
(
f
„
-
f
;
)
where pw{n = flowing b o t t o m h o l e p r e s s u r e at t i m e tn (psia) qn = last of n d i f f e r e n t flow rates ( M S C F / d a y ) Mi = fy.i - Ii ( M S C F / d a y ) , qQ = 0 tn = total ( c u m u l a t i v e ) flowing t i m e f o r n c o n s t a n t r a t e flow p e r i o d s (Iir) tj = t i m e at w h i c h flow rate w a s c h a n g e d (hr)
462 PART 9—PRODUCTION ENGINEERING METHODS
The plot is made on Cartesian coordinate paper. A straight l i n e w i t h s l o p e , m', s h o u l d r e s u l t ; t h i s s l o p e is u s e d in Equation (4) to calculate permeability:
1637Tpz
Jt =
(4)
m'h
where z = real gas compressibility factor T = reservoir temperature (°R)
The apparent skin factor can also be determined from this plot. If the intercept of the plot (the value of Ap2/q w h e r e the time plotting function is zero) is denoted by b, then
(5)
Since multi-point test data are available for virtually all gas wells, this method of analysis often proves to be a useful w a y of estimating permeability and skin factor, especially w h e n drawdown or buildup tests are not available.
SWAB TESTS
Swabbing can b e d e f i n e d as p u l l i n g a f u l l - d i a m e t e r tool f r o m the wellbore; this pulling action is similar to that of a plunger in a syringe, and it initiates fluid flow into the wellbore (Whittaker, 1987). On occasion, oil or gas wells may not flow fluid to the surface on completion. W h e n this occurs, a s w a b b i n g unit is r u n to r e m o v e the hydrostatic column of fluid in the wellbore a n d allow the well to kick off a n d flow.
In some oil wells, the bottomhole pressure may be insufficient to lift fluid continuously. Because a surface flow rate cannot b e m a i n t a i n e d a n d m e a s u r e d , r o u t i n e flow a n d buildup tests cannot be used to evaluate reservoir properties a n d d e t e r m i n e w e l l p r o d u c t i v i t y . Ixi t h e s e i n s t a n c e s , a s w a b b i n g tool can be r u n at regular intervals to k e e p fluid flowing from the formation m o r e or less continuously.
By carefully recording the v o l u m e of fluid recovered f r o m each swabbing r u n as a function of time, one can determine the rate of fluid feed-in f r o m the f o r m a t i o n to the wellbore. These data can then be used to estimate well productivity or formation permeability using the methods outlined in the discussion of single-point tests. The information gained from a swabbing test can also be particularly useful in determining whether a pumping unit should be installed on a well and in determining the proper pump design.
Pressure Transient Testing
wj0hn
Texas A&M University Petroleum Engineering Department College Station, Texas, U.S.A.
PRESSURE DRAWDOWN AND BUILDUP TESTS
Information Derived from Tests
Pressure buildup and drawdown tests provide an o p p o r t u n i t y to obtain estimates of the following well a n d reservoir properties:
• Permeability to the produced phase (oil, gas, or water), which is an average value within the radius of investigation achieved in the test
• Skin factor, which is a quantitative measure of d a m a g e or stimulation in the well
• Current average pressure in the drainage area of the tested well
• Verification of flow barriers (such as faults) and estimates of distance to these barriers
Single-Rate Flow Tests
A single-rate flow test is run using the following procedure. Start with the reservoir at a uniform, stabilized pressure. Then place the pressure measuring device as near the perforations as possible. Flow the well at a strictly constant rate (control with a variable choke or other control device). An alternative method for gas wells is to measure surface pressures and calculate bottomhole pressures from surface measurements. The rate at which pressure changes with time reflects the formation properties. An example of this type of test is a single-point deliverability test in a gas well (see the chapter on "Production Testing" in Part 9).
Pressure Buildup Tests
In a pressure buildup test, one should stabilize the rate in the tested well for several days, that is, maintain the rate approximately constant. Then place a pressure measuring device as near the perforations as possible several hours before shut-in. Shut the well in and let the pressure build up. The rate at which pressure builds up with time reflects the formation properties. (For m o r e details of pressure b u i l d u p a n d flow tests, see M a t t h e w s a n d Russell, 1967.)
MULTI-RATE FLOW TESTS
These tests are run much like single-rate tests, except that the rate is changed by discrete amounts one or more times while the test continues. A n example of this type of test is a flow after flow deliverability test on a gas well, which is also called a four-point or back-pressure test.
When the Tests Are Applicable
Flow tests can be useful when the reservoir is at uniform pressure, such as when a new well is completed or when a well has been shut in for a lengthy period. Flow tests are appropriate when a well must continue to produce revenue even t h o u g h a test is n e e d e d . Analysis of f l o w tests is simplest when the rate is held strictly constant and in all cases requires known rates at all times during the test.
Buildup tests are appropriate at virtually any time in the life of a well because they simply require that the well be shut in. Buildup tests have the advantage that the rate (zero) is m u c h m o r e easily controlled than in a "constant rate" flow test. For this reason, buildup tests are overwhelmingly the preferred type of pressure transient test in practice.
How the Tests Are Analyzed
Wellbore effects d o m i n a t e early test data. The end of wellbore effects is f o u n d using log-log plots of test data, which are compared to preplotted type curves, as illustrated in Figure 1. The s h a p e s of test data plots are also used to identify the reservoir type, such as homogeneous acting, naturally fractured, layered, or hydraulically fractured. Derivative t y p e c u r v e s (basically the slope of a plot of p r e s s u r e v e r s u s the l o g a r i t h m of time) are particularly helpful for identifying reservoir type and wellbore effects, as shown in Figures 2(a) and (b).
Traditional analysis is focused on semi-logarithmic plots of test data, w i t h slopes of straight lines on these plots used to determine permeability. Figure 3 is a typical semi-log plot of f l o w test data, a n d Figure 4 is a typical semi-log plot of buildup test data. The "correct" semi-log straight line is indicated on these figures; the Hne is identified with the help of type curves (see the chapter on "Production Histories" in Part 9). On the buildup test plot, shut-in bottomhole pressure is plotted versus the logarithm of the ratio of p r o d u c i n g time, fp, plus shut-in time, At, to shut-in time. Tliis plot is called a Horner plot, n a m e d for t h e p e r s o n w h o p r o p o s e d it in the petroleum literature.
Simple equations allow us to estimate permeability and skin factor once the correct semilog straight line is identified and its slope, m, is estimated. These equations apply to both drawdown and buildup tests. The following equations are used for oil wells:
162.6cjB/J.
mh
477
462
PART 9—PRODUCTION ENGINEERING
METHODS
IO2
LLI
CT
Z>
V) LD
IO1
Ш
СГ
CL
CO
CO
Ш
_J
Z
O
tD
Z Ш
I0<
APPROXIMATE END OF WELLBORE EFFECTS
MATCHING
PARAMETER
|0SO
DAMAGED WELL
10го
IO16
щю IO8 IO6
IOsIo' IO 5 1 6.10"1 io-1' ,
6.10-2 10"г 5.IO"3
IO1
IO2
DIMENSIONLESS TIME GROUP
Figure 1. Type curve that identifies the end of wellbore effects.
and
s = 1.151 APlhr - l o g m
+ 3.23
where к = effective permeability to produced phase (md) cj = f l o w r a t e ( S T B / d a y ) B = formation volume factor (rb/STB) m = slope of semi-log straight line (psi/log10cycle) h = net pay thickness (ft) s = skin factor (dimensionless)
Aplhr = pressure c h a n g e in first h o u r of test (psi) ф = porosity (fraction) H = viscosity (cp) ct = total c o m p r e s s i b i l i t y of f o r m a t i o n a n d its fluids (psi"1) rw = w e l l b o r e r a d i u s or h o l e size (ft)
Similar equations are used for gas well test analysis. Extrapolation of pressure in a b u i l d u p test to Horner time
ratio of unity p r o v i d e s an estimate of original reservoir pressure (new well) or "false" pressure, which serves as the
basis for determining current drainage area pressure, P , for a well with some pressure depletion in its drainage area caused by e x t e n d e d p r o d u c t i o n of fluids. Figure 4 illustrates extrapolation of pressure to time ratio of unity.
Also, d i s t a n c e to b o u n d a r i e s of flow b a r r i e r s is f o u n d f r o m semilog plots by deviation from a previously established semilog straight line. In the simplest case, which is uncommon except for wells very close to boundaries, the late time slope doubles. Figure 5(a) shows the usual case, where the slope increases at late times but does not double. Figure 5(b) shows the less common case where the slope actually doubles at late times. The time at which the deviation occurs a n d the a m o u n t of deviation can be u s e d to estimate the d i s t a n c e f r o m t h e tested well to t h e flow barrier.
LONG-TERM PRODUCTION TESTS
The same information is available from long-term production tests as from short-term flow tests, including permeability, skin factor, and distance to boundaries. Hydrocarbons in place in the tested well's drainage area can frequently be estimated from these test data. Once boundaries have affected the test data, long-term production d a t a can b e e x t r a p o l a t e d to p r o v i d e a forecast of f u t u r e production to the economic limit and can thus provide a reserve estimate for the well.
How the Tests Are Run
These are not tests in a formal sense. The rate is simply monitored as a function of time while the well is produced at a n a p p r o x i m a t e l y c o n s t a n t b o t t o m h o l e p r e s s u r e (ВНР). If constant ВНР cannot be maintained, one should at least
IOO iW=: шcc
Z CL
Pressure Transient Testing
479
DAMAGED WELL WITH WELLBORE STORAGE
HOMOGENEOUS-ACTING RESERVOIR (RADIAL FLOW)
MINIMUM HETEROGENEOUS
BEHAVIOR
UPWARD TREND SINGLE OR MULTIPLE BOUNDARIES FELT. OPEN TO FLOW IN AT L E A S T
ONE DIRECTION,
DOWNWARD TREND CLOSED SYSTEM OR C O N S T A N T PRESSURE .BOUNDARY
10
100
1000
10,000
(a)
DIMENSIONLESS TIME GROUP
(b)
LOG OF ELAPSED TIME
Figure 2. (a) Derivative type curve used to match derivatives of test data, (b) Shapes of derivatives of test data for various reservoir conditions.
END OF WELLBORE EFFECTS
CORRECT STRAIGHT LINE
SLOPE = m
START OF BOUNDARY EFFECT5
END OF
WELLBORE
CL
EFFECTS
START OF BOUNDARY
QT
EFFECTS
CORRECT
CD
STRAIGHT LINE
0.1
10
100
1000
SLOPE = m
FL0WIN6 TIME, hours
Figure 3. Typical flow test data graph.
measure surface pressures continuously (from which one attempts to calculate the variable BHPs).
These "tests" are applicable at any time. The data are most r e a d i l y a n a l y z a b l e if t h e well is p r o d u c e d at a p p r o x i m a t e l y constant ВНР or when ВНР and flow rate are known continuously as functions of time.
How the Tests Are Analyzed
The simplest of these tests—those with production data from wells produced at constant ВНР or with smoothly changing rates and ВНР—can be analyzed with simple type curves, such as Fetkovich's, illustrated in Figure 6. Formation properties are obtained f r o m matches of actual field data to the type curve. Forecasts are made by projecting future performance along the type curve that matches postproduction performance.
A n essential requirement in this m e t h o d of analysis is that there is enough production history for boundary effects to have influenced production data. This same requirement also applies to conventional decline curves and decline curve analysis—if boundary effects have not been felt, the decline curve projection is totally meaningless and certainly incorrect. Figure 7 s h o w s a type curve match of past performance and indicates how production data can be extrapolated into the future.
IOOO
IOO LOG ( V A I
At
IO )
Figure 4. Typical buildup test graph (Horner plot).
More complex tests, with abrupt changes in rate and ВНР, are more readily analyzed with computer reservoir simulators. These simulators are used to history-match production data to obtain a reservoir description, which is then used to obtain a long-term production forecast and thus to estimate reserves. Figure 8 s h o w s an example of a history m a t c h of p r o d u c t i o n d a t a a n d a forecast of f u t u r e p e r f o r m a n c e of the well u s i n g the reservoir description obtained from the history match.
INTERFERENCE AND PULSE TESTS
Interference and pulse tests are run to obtain the following information:
• Whether tested well pairs are in pressure communication
• Directional permeabilities between tested well pairs • Average porosities in the areas influenced by the tests
462
PART 9—PRODUCTION ENGINEERING
METHODS
EFFECTS OF FLOW
(a)
BARRIER BEGIN
<n
WELLBORE
CL
EFFECTS
OL
END
X
Ш
6200
(b)
5800
5400
(0
V) 5 0 0 0 CL
QT
I
4600 _
CQ
4200
/ / / / /
/ / Cp
0 ¾ = SLOPE = 2m
Л
Cf/
3800 3400
J C T ^ * SLOPE = m CORRECT SEMILOG STRAIGHT
LINE
'-^r-) L0G / 1 D + A T
,
3000 1000
I
I
100
IO
l P t Al
Al
Figure 5. (a) Typical buildup curve shape with flow barrier, (b) Doubling of slope on Horner plot for well near barrier.
TRANSIENT DATA
(N0 BOUNDARY EFFECTS)-*
ACTUAL DATA POINTS
FITTING CURVE TO EXTRAPOLATE Л 0 FUTURE TIMES
^ o.oi
0,001
IO"2 IO-'
IO0
IO1
DIMENSIONLESS TIME
Figure 6. Fetkovich's type curve for analyzing long-term production data. (After Fetkovich, 1980.)
IO0
IO1
DIMENSIONLESS TIME
Figure 7. Actual production data matched on a Fetkovich type curve.
How the Tests Are Run
Interference tests are r u n by first shutting in the portion of the reservoir in the area affected by the test. Then one produces (or injects into) one well (called the active well) and measures the pressure response in the offset wells. Figure 9 shows a typical interference test pattern, and Figure 10 is a plot of a typical response in a n observation well.
Pulse tests are performed by first producing (or injecting into) the active well for a few hours. The active well is then shut-in, then returned to production, shut-in again, and so on in a regular, repeating pattern. The response in the offset wells is then measured while continuing to produce all wells in the field except those directly involved in the test. Tliis is possible because the "noise" caused by continued production of wells not directly involved in the test can be filtered out using the response caused by the repeated on-off pattern in the active well. Figure 11 s h o w s a typical response in a pulse test observation well.
Comparison of Pulse and Interference Tests
Interference tests are usually much more expensive than pulse tests because of the loss of r e v e n u e arising f r o m h a v i n g to shut-in a major portion or all of the tested reservoir to conduct the test. Also, ambiguity exists in interference test interpretation because it is uncertain that an observed response was actually caused by the active well. In a pulse test, if a r e p e a t e d s i g n a l is r e c e i v e d in a n o b s e r v a t i o n w e l l , there is little d o u b t that it w a s caused by the rate changes in the active well.
When the Tests Are Applicable
Interference and pulse tests are applicable whenever one needs to know whether two or more wells in a formation are in pressure communication. They are also used in improved recovery processes (water, gas, or other fluid injection projects) in w h i c h k n o w l e d g e of directional p e r m e a b i l i t y differences and reservoir description generally are critical to the success of the project.
VARIABLE PRESSURE INPUT INTO SIMULATOR
Pressure Transient Testing
481
O OBSERVED RATES SIMULATED RATES
012
345678
TIME, years
Figure 8. History match of production data and forecast of future performance.
Figure 9. Active and observation wells in an interference pulse test. (After Earlougher, 1977.)
L_ -I <t ш
TIME
ACTIVE WELL
OBSERVATION W(ELtL
ESTABLISHED^-^ TREND
^—(TIME LAG)
At
-
TIME
Figure 10. Schematic illustration of rate history and pressure response for an interference test. (After Earlougher, 1977.)
These tests work best in formations that have higher permeabilities, closer s p a c i n g , a n d s i n g l e - p h a s e flow. In g a s reservoirs, high formation pressure is more desirable than low pressure in achieving a successful test.
CL
CX
UJ
Ш £Г ZD O CO
CO LLI
СП
CL
MTTT
ЛЕс с > Г
е Г
е
г .
СП сл сл сл
m m m гп
— то ы
Figure 11. Schematic illustration of rate (pulse) history and pressure response for a pulse test. (After Earlougher, 1977.)
How the Tests Are Analyzed
Simple tests in homogeneous-acting, isotropic or anisotropic, infinite-acting (that is, no boundary effects during the test), single-layer reservoirs can be analyzed with readily available interpretation charts and type curves. Most actual tests require computer reservoir history-matching for analysis because of heterogeneities, boundaries, a n d layers.
Surface Production Equipment
James Jennings
Consultant Hearne, Texas ,U.S.A.
INTRODUCTION
There are three major c o m p o n e n t s of surface production equipment:
1. W e l l h e a d 2. Separators and heater treaters 3. Tank batteries and meter facilities
Production engineers often design all equipment on the lease. After the oil and gas leaves the lease, pipeline or facilities engineers take over.
For a gas well, the following equation is used:
d2R, cj = 390-
VGT
where q = gas flow rate (MSCF/day)
Ptf = flowing tubing head pressure (psia) d = choke size (in.) G = gas specific gravity T = wellhead t e m p e r a t u r e (°R)
WELLHEAD
T h e wellhead is t h e e q u i p m e n t at t h e s u r f a c e t h a t p r o v i d e s support for the tubulars inside the well, a pressure seal between the tubulars, and a m e a n s of controlling production f r o m the well. Typically, the wellhead consists of a casing head for each casing string, a tubing head, and a Christmas tree. For each string of p i p e in the well, casing, or tubing, s o m e m e a n s of s u p p o r t a n d p r e s s u r e sealing m u s t be provided. Tliis is the function of the casing a n d tubing heads. T h e Christmas tree p r o v i d e s t h e n e c e s s a r y v a l v i n g a n d c h o k e s to control the production f r o m a well capable of flowing. For a well that is being pumped, the Christmas tree is replaced by wellhead equipment that accommodates the pumping operation.
Figure 1 shows a typical wellhead for a flowing well. N o t i c e that a c h o k e is p r o v i d e d to c o n t r o l the rate of production from the well in addition to the tubing wing valve, w h i c h p r o v i d e s for a c o m p l e t e shut-off of the production. The choke can be either fixed or variable in size. The choke is nothing more than a small orifice, usually from 1 / 8 to 3 / 4 in. in d i a m e t e r , that restricts t h e flow rate.
O t h e r valves a r e p r e s e n t on t h e side of the w e l l h e a d . T h e s e a r e called casing valves a n d t h e y p r o v i d e access to t h e various annulii between casing strings and tubing. Normally, the wellhead is fitted with pressure gauges for monitoring pressure within the different annulii and in the tubing.
T h e flow rate f r o m either a n oil well or a g a s well c a n b e easily e s t i m a t e d f r o m t h e w e l l h e a d p r e s s u r e if t h e w e l l h e a d p r e s s u r e is a t least t w i c e t h e flowline p r e s s u r e . F o r a n oil well, the Gilbert equation is commonly used:
_ Pr t«fS< <7 =
600VK
SEPARATORS AND HEATER TREATERS
Once liquids are brought to the surface, the oil, gas, and w a t e r m u s t be s e p a r a t e d for ease of m e a s u r e m e n t a n d t r a n s p o r t a t i o n . A separator is a vessel u s e d to s e p a r a t e liquid
WING VALVE
MASTER VALVE
TUBINGHEAD ADAPTER c5I
TUBING HEAD -
CASING HEAD -
TUBING HANGER
CASING VALVE CASING HANGER
where q = g r o s s liquid flow r a t e ( b b l / d a y )
Ptf = flowing t u b i n g h e a d p r e s s u r e (psia) R = gas to liquid ratio (MSCF/bbl) S = choke size (1 / 6 4 in.)
Figure 1. Typical wellhead for a flowing well with a single-wing, single-completion threaded manifold.
482
Surface Production Equipment
483
Figure 3. Horizontal separator.
Inlet
Figure 2. Vertical separator.
from gas. In some cases, the liquid may be additionally separated into individual oil and water streams. A separator is c o m m o n l y given a n y of the following names:
• Gun barrel • Free water knock out • Knock out • Trap • Scrubber • Stage separator
A heater treater is s i m p l y a s e p a r a t o r t h a t is d e s i g n e d to separate primarily oil f r o m water. H e a t i n g of the mixture normally speeds up and improves the separation process.
Several physical processes are commonly used in the separation process:
• Gravity settling • Centrifugal force • Impingement • Electrostatic precipitation • Filtration • Heat
The design of a particular separator d e p e n d s on the n a t u r e of the f l o w s t r e a m to be s e p a r a t e d . For a gas well, the ^ З ^ Ш ^ Ч Ш а Ц ^ ^ ^ a j a t e ? . a „ s m a l i . a m a v m t Qf liquid f r o m t h e gas. In an oil well, the separation m a y involve a small a m o u n t of gas for the a m o u n t of liquid. In general, the well stream separator must separate the mostly liquid fluids from the mostly gas fluids. In addition, it must separate liquid h y d r o c a r b o n f r o m liquid w a t e r a n d r e m o v e m o s t of the entrained liquid mist from the gas.
To accomplish the separation, the separator is usually designed to control and dissipate the well stream flowing
Figure 4. Spherical separator.
energy. Once gas and liquid velocities are slow enough, gravity causes the liquid to settle and the gas to rise. The size of the vessel m u s t be such that a d e q u a t e time is allowed for this settling to occur b e f o r e the fluid leaves t h e s e p a r a t o r . If water is to be separated from oil, then the liquid residence t i m e d e p e n d s o n t h e v o l u m e of t h e fluid b e i n g h a n d l e d a n d the specific gravity of the t w o liquids. M a n y times, a mist extractor c o m p o s e d of v a n e s , m e s h p a d s , or a cyclonic passage is used to remove residual liquid droplets from the gas stream.
T h e r e a r e t h r e e t y p e s of s e p a r a t o r s : vertical ( F i g u r e 2), horizontal ( F i g u r e 3), a n d spherical ( F i g u r e 4). H o r i z o n t a l separators are found in both the single tube and double tube design. Advantages of the vertical separator include
• Good for predominantly liquid streams • Can handLe tjroducinestream surees without carryover • Occupies little space (small footprint) • Easily cleaned of sand a n d m u d
A d v a n t a g e s of the horizontal separator are
• Good for predominantly gas streams • Easy to fabricate, ship, and install • Low profile
462
PART 9—PRODUCTION ENGINEERING METHODS
Figure 5. Vertical heater treater.
And for the spherical separator, the advantages are
• Good for high pressure gas wells • Compact, small size
Figure 5 shows a typical vertical heater treater. Notice that a heater treater is simply a separator in which a firetube has been placed to heat the liquid mixture as it enters the vessel. This heating of the oil and water mixture reduces the viscosity and promotes the separation of the two phases. The fuel for the firetube is usually taken from the gas produced from the well.
V-8
Figure 6. Orifice meter for gas measurement.
TANK BATTERIES AND METERING
Tanks must be provided to hold both oil and water for shipping or disposal. Usually, at least two oil tanks are used, o n e for shipping and one for filling. The v o l u m e of oil being shipped is sometimes determined by simply measuring the height of the fluid in the tank, or "strapping" the tank. M a n y of the m o r e m o d e r n production facilities h a v e lease automatic custody transfer (LACT) units installed. These stations continuously measure the flow into the shipping point and periodically sample the product being shipped so that oil gravity, temperature, pressure, and water content are known. The metering in this case is done with a positive displacement meter.
The m e a s u r e m e n t of gas is usually d o n e with an orifice meter. Figure 6 shows such an installation. For a given meter installation, the gas flow rate depends on the pressure on both sides of the orifice a n d the t e m p e r a t u r e of the gas. These pressures and temperatures are normally recorded on a circular chart. This chart is later used to determine the total g a s flow o v e r a p a r t i c u l a r t i m e p e r i o d . M a n y i n s t a l l a t i o n s today have digital flow calculating and recording devices installed at the meter. These installations have proven to be a c c u r a t e , a n d t h e y p r o v i d e f o r t h e t e l e m e t e r i n g of t h e flow information from a remote installation.
Artificial Lift
D. D. Smallwood
Conoco Inc. Lafayette, Louisiana, U.S.A.
INTRODUCTION
Artificial lift r e f e r s t o t h e m e c h a n i c a l l i f t i n g of w e l l b o r e fluids to t h e surface. M e c h a n i c a l lifting of w e l l b o r e fluids is required w h e n reservoir pressure is insufficient to drive reservoir fluids to t h e surface. Artificial lift e q u i p m e n t also c a n b e u s e d to i n c r e a s e p r o d u c t i o n f r o m flowing w e l l s b y reducing the producing bottomhole pressure.
A n u m b e r of different types of artificial lift s y s t e m s are currently in use. The four primary artificial lift systems are
• Electric submersible p u m p (ESP) • Gaslift • Hydraulic p u m p (piston and jet p u m p ) • Beam pump
ELECTRIC SUBMERSIBLE PUMPS
A n electric s u b m e r s i b l e p u m p (ESP) consists of a centrifugal p u m p coupled to an electric motor. The p u m p a n d m o t o r combination is r u n in the well on the b o t t o m of the t u b i n g s t r i n g a n d is set b e l o w t h e o p e r a t i n g fluid level in t h e well (Figure 1).
The electric motor powers a centrifugal p u m p that forces fluid into the p u m p up through the tubing and out at the surface. The electric motor is powered by an electric cable s t r a p p e d to the side of the tubing string. Lift capacity for
each p u m p is adjusted by changing the n u m b e r of stages in the centrifugal p u m p a n d / o r by changing the h o r s e p o w e r of the electric motor.
Electric submersible p u m p s are beneficial in wells that m u s t Hft h i g h v o l u m e s of fluids f r o m less t h a n 10,000 ft d e e p . ESPs are often u s e d in the late stages of waterflooding w h e r e high water cuts require lifting large production volumes from each well.
GAS LIFT
G a s lift is t h e p r o c e s s of lifting fluids f r o m t h e w e l l b o r e using high pressure gas as the energy source. Downhole gas lift e q u i p m e n t consists of a series of gas lift valves spaced at predetermined depths in the tubing string. The tubing string is set in a packer above the casing perforations (Figure 2). Gas is normally injected d o w n the tubing/casing annulus and enters the tubing via the gas lift valves. A surface gas compressor is used to provide the high gas pressure required to open the gas lift valves.
G a s lift s y s t e m s m o v e t h e r e s e r v o i r fluids to t h e s u r f a c e b y r e d u c i n g t h e h y d r o s t a t i c p r e s s u r e of t h e fluid c o l u m n in the tubing below reservoir pressure. The injected gas e x p a n d s as it m o v e s u p w a r d in the tubing, p r o v i d i n g additional lift. Gas lift systems can be installed to operate continuously or intermittently. Gas lift is commonly used in
Figure 1. Electric submersible pump. (From Conoco Inc., 1990.)
Figure 2. Gas lift system. (From Conoco Inc., 1990.)
485
462
PART 9—PRODUCTION ENGINEERING
METHODS
CASING
PRODUCED FLUID & POWER OIL ~
PACKER
S • a
POWER OIL TUBING PUMP
activates t h e p i s t o n t y p e p u m p w h i c h lifts t h e p r o d u c e d fluids a n d h y d r a u l i c fluid u p t h e casing a n n u l u s t o w a r d t h e surface. The hydraulic fluid most commonly used is produced oil f r o m t h e well itself. W h e n t h e p r o d u c e d fluids a n d h y d r a u l i c fluid a r e p u m p e d to t h e surface, t h e oil is s e p a r a t e d a n d s o m e of t h e oil is r e u s e d a s t h e p o w e r fluid.
The hydraulic jet p u m p system uses a ventura pressure drop to commingle the power hydraulic fluid and the p r o d u c e d fluids. T h e jet p u m p s y s t e m d o e s n o t r e q u i r e t h e hydraulic fluid to be as clean as that for the piston type system. The jet system also allows for wider production rates than does the piston type pump system. Hydraulic pumps are used primarily in deep wells requiring lift volumes greater than those capable from beam pump systems.
ШШЯШШШШЯШ
PRODUCING
<
OIL
••И FORMATION
Figure 3. Hydraulic piston pump. (From Conoco Inc., 1990.)
offshore applications and in areas where an abundant supply of g a s exists. O f t e n t h e g a s p r o d u c e d f r o m t h e well is separated from the produced fluids and reinjected into the same well.
HYDRAULIC PUMPS
A hydraulic p u m p operates similarly to a gas lift system, w i t h h i g h p r e s s u r e p o w e r fluid u s e d a s t h e e n e r g y s o u r c e in place of high pressure gas. There are t w o different types of hydraulic pumps: piston and jet. A piston type p u m p assembly consists of a hydraulically operated m o t o r at o n e end and a plunger type p u m p at the other end (Figure 3). H i g h p r e s s u r e h y d r a u l i c fluid is p u m p e d d o w n t h e t u b i n g string and enters a reciprocating hydraulic motor. This motor
BEAM PUMPS
B e a m p u m p i n g s y s t e m s w e r e o n e of the first t y p e of artificial lift used in the oil field and are still the most widely u s e d m e a n s of artificial lift. A b e a m p u m p s y s t e m lifts fluid by reciprocating a rod string that activates a positive displacement p u m p . The positive displacement p u m p is seated in t h e t u b i n g string a n d set b e l o w t h e o p e r a t i n g fluid level in the well (Figure 4). A surface p u m p i n g unit provides the power to reciprocate the rod string.
The surface p u m p i n g unit is m a d e u p of t w o p r i m a r y components: a prime mover (motor) and a walking beam connected to a pivotal post. The walking beam operates like a seesaw on the pivotal post, providing a reciprocating motion to the rod string. O n each u p w a r d stroke of the rod string, a v o l u m e of p r o d u c e d w e l l b o r e fluids is lifted u p w a r d in t h e tubing string toward the surface. The capacity of the b e a m p u m p i n g system is set by the size of the d o w n h o l e p u m p , the stroke length of the rod string, a n d the speed at which the rod string is reciprocated. When p u m p capacity exceeds wellbore fluid entry, the surface pumping unit can be set up to run intermittently by shutting down the pumping unit for a set period of time.
T h e l i m i t a t i o n s of the b e a m p u m p i n g s y s t e m are a p p r o x i m a t e l y 150 bbl of fluid p e r d a y at 12,000 ft of d e p t h . L a r g e r fluid v o l u m e s can b e p r o d u c e d w i t h b e a m p u m p i n g systems at shallower depths. Beam pumping systems have been used in wells as deep as 15,000 ft.
Artificial Lift
487
Figure 4. Beam pumping system. (From Conoco Inc., 1990.)
Production Logging
James J. Smolen
Consultant Missouri City, Texas, U.S.A.
INTRODUCTION
After casing is set in place, wireline surveys are often run to evaluate the integrity of the completion. Such s u r v e y s include production logs and mechanical integrity i n s t r u m e n t s . Production logs a r e u s e d to e v a l u a t e f l u i d production and m o v e m e n t both inside and outside of the casing downhole. The production logging tools are small in diameter and are run through tubing for evaluation of the well as it is producing. Mechanical integrity instruments, which assess the condition of the casing or cement a r o u n d it, are generally larger in diameter. These surveys are run before the tubing is in place, or else it must be removed and the well shut-in.
PRODUCTION LOGS
The main applications of the production logs include
1. Locating sources d o w n h o l e of undesired fluid phase production such as water entries
2. Isolating mechanical problems such as leaking pipes, leaking packers, and fluid m o v e m e n t in cement channels behind pipe
3. Evaluating the effectiveness of well treatment or workover operations by comparing the before and after job surveys
4. Accumulating baseline well performance information for comparison with later monitor surveys
5. In EOR projects, maintaining injection efficiency by evaluating the injection profiles of individual wells in a field
temperature survey response to two gas entries into a well. Noise logs are also used to evaluate fluid movement
downhole. Unlike temperature surveys, noise logs are not r u n continuously across the interval of interest. Instead, a n u m b e r of stationary readings are taken at different depths d o w n h o l e . The m o v e m e n t of fluids, especially gasses, generates turbulence or noise, which gets louder as the flow rate or pressure drop increases. Figure 2 shows how a noise log can be effective at detecting movement downhole. In this schematic diagram, a source, sink, and restriction to flow are the noise sources. The frequency spectrum of the noise is also o b s e r v e d to f u r t h e r i m p r o v e the u n d e r s t a n d i n g of f l o w downhole.
Radioactive tracer s u r v e y s use a tool c o m p o s e d of an ejector capable of ejecting shots of radioactive tracer material into the flow stream, usually of an injection well. Such an instrument has either one or two gamma ray detectors spaced below the ejector. By various techniques, the operator chases the ejected radioactive material as it moves with the injected fluid. By noting the position, time, a n d size of the tracer signal, an accurate overview of the injection profile can be established. Special techniques are also available to detect
Production logs include (1) those designed to detect flow in and around pipes (temperature, noise, radioactive tracer, f l o w m e t e r , and fluid identification logs) and (2) those designed to evaluate flow quantitatively. Often combinations of these logs are required to be effective (Schlumberger, 1989; Atlas Wireline Services, 1986; Society of Petroleum Engineers, 1985).
Flow Detection In and Around Pipe
Temperature surveys are the most common surveys to locate fluid movement downhole. Small entries and even flow in channels behind pipe can be detected. Generally, if a well is not flowing, the t e m p e r a t u r e of the fluid in the borehole will eventually approach the formation temperature, called the geothermal gradient. W h e n a well is p r o d u c e d , formation fluids enter the borehole a n d m o v e uphole. Gasses typically cool when entering the borehole while liquids do not. In either case, their movement uphole is easily detected by d e v i a t i o n s of the b o r e h o l e t e m p e r a t u r e f r o m the geothermal gradient. Figure 1 illustrates a typical
Figure 1. Temperature survey showing two gas entries and the geothermal gradient.
488
Production Logging
489
i
D1 D
Flow profile
Time
E ft
d1iH
/
Time
N O I S E LEVEL Figure 2. Noise log responses to fluid movement downhole.
injected fluid channeling through the cement to undesirable zones. A schematic d i a g r a m of a tracer tool is s h o w n in Figure 3.
Quantitative Flow Evaluation
Q u a n t i t a t i v e evaluation of f l o w profiles in injection or producing wells is common. Injection wells are most often evaluated with radioactive tracer techniques, while producing wells, where multiphase flow may be encountered, are evaluated using flowmeters with fluid identification devices.
The most effective technique with radioactive tracers is the velocity shot technique, illustrated in Figure 3. The tool is stationary during such a test, and the gamma count rate is recorded at the surface. In Figure 3, tests were made above, between, and below the perforations, and the surface recordings are s h o w n to the right of the well sketch. The highest velocity and flow rate are recorded above the perforations, while zero flow is detected in the lowest interval. By measurement of the traveltime between detectors, At, and using the known spacing between detectors D1 and D2, the flow rates can be calculated and an injection profile constructed, as shown on the right of the figure.
In producing wells, spinner flowmeters are used to measure the bulk flow rate, even in multiphase flow conditions (Anderson et al., 1980). T w o such flowmeters are s h o w n in Figure 4. Thefull bore flowmeter in Figure 4(a) is r u n continously across the interval of interest, while the basket type flowmeter in Figure 4(b) uses stationary measurements. Although these devices can determine the bulk flow rate,
Time
Percent injected fluids
50
100%
Figure 3. Tracer velocity shot technique and injection profile.
fluid identification tools are required to evaluate the kinds of fluids present in the flow. These fluid identification instruments measure the pressure gradient, bulk density, or capacitance of the flowing mixture. The flowmeter and fluid identification devices are usually run as a combination on the same tool string. Results typical of such a tool string are shown in Figure 5. In this example, zone A produces water, while the zones above it are all gas producers. A plug set between zones A and B will be effective at eliminating the water production in this example.
MECHANICAL INTEGRITY LOGS
The well mechanical integrity survey logs include two groups. The first group, cement evaluation surveys, assesses the d e g r e e of c e m e n t fill a r o u n d the casing a n d can be effective at locating potential channels for fluid m o v e m e n t . The second group is the casing inspection surveys, in which acoustic, mechanical, and electromagnetic measurements are used to evaluate internal and external casing conditions.
490
PART 9—PRODUCTION ENGINEERING METHODS
Figure 4. Two types ot flowmeters, (a) Full bore flowmeter, (b) Diverting basket type flowmeter.
Cement Evaluation
Cement evaluations are primarily done with cement bond logs or pulse-echo cement evaluation tools. These are acoustic devices w h o s e m a i n objectives are the m e a s u r e m e n t of cement annular fill around the casing.
The cement bond log (CBL) measures the degree to which cement contacting the pipe on the outside attenuates an acoustic signal traveling along the pipe (Pardue et al., 1963; Fitzgerald et al., 1983; Western Atlas International, 1985). Figure 6 illustrates how the acoustic signal is affected by the presence of cement. The initial portion of the acoustic signal or signature indicates the amplitute of the signal traveling along the pipe. The amplitude curve records the amplitude of this initial portion or pipe signal. A low amplitude indicates good bond, while a very high signal amplitude shows free pipe. This amplitude measurement can be converted to percent annular fill of cement. The variable density log (VDL) at the far right of Figure 6 is a contour m a p of the received wavetrain signature as it changes with depth.
The pulse-echo cement bond log (CET) operates in an entirely different acoustic mode than does the CBL (Froelich et al., 1982). The pulse-echo tool is effective at measuring the compressive s t r e n g t h of cement b e h i n d pipe, as well as detecting the presence of liquid or gas behind pipe. The main presentation of the pulse-echo tool is the cement m a p s h o w n on the right of Figure 7. The dark areas correspond to cement, and the white areas indicate the lack of it. With such a cement map, likely channels can be readily detected.
DownhoMlelS P<wrroduction
5000
10.000
Figure 5. Flow profile in a producing well.
Production Logging
491
SIGNATURE
1000 0
VARIABLE DENSITY
1000 0
Figure 6. CBL logs and their response to various cement conditions.
Casing Inspection
The pulse-echo tool is also an effective acoustic casing inspection device, providing such information as internal diameter and wall thickness. Mechanical calipers, having 40 to 80 feelers, are excellent at inspecting the inside of the casing for small- or large-scale damage. Electromagnetic devices are useful in inspecting for pipe damage on either the inside or outside of the casing and can be useful in examining the outer concentric strings of pipe.
Minimum compressive strength (CSMN)
Figure 7. Pulse-echo cement bond log showing cement top and channel on the cement map presentations.
Production Problems
Bradley M. Robinson
S. A. Holditdi & Associates, Inc. College Station, Texas, U.SA.
INTRODUCTION
The p r o d u c t i o n of oil or gas f r o m u n d e r g r o u n d reservoirs involves chemical and mechanical processes that affect the wellbore. M a n y of these processes can eventually cause a problem with the well, resulting either in a decrease in p r o d u c t i o n or in failure of e q u i p m e n t installed d o w n h o l e or at the surface. Most of the serious p r o b l e m s can be avoided or delayed through preventive maintenance techniques or early recognition f r o m regular analysis of p r o d u c i n g rates, fluids, and the mechanical condition of the well. Such practices can prevent a costly workover that may be required to restore production from the well and m a y also prevent total loss of the wellbore.
WATER-GAS CONING
When water underlies an oil column or a gas cap exists above, the p h e n o m e n o n of coning can occur w h e n the oil is produced. Muskat (1949) noted that coning is most likely to occur w h e n a w e l l b o r e only p e n e t r a t e s a p o r t i o n of the formation or when the perforations are placed only in the oil zone (Figure 1). The horizontal and vertical pressure g r a d i e n t s created by the p r o d u c t i o n of oil c a u s e s a deformation of the oil-water (or oil-gas) contact. W h e n the vertical pressure gradient exceeds the hydrostatic gradient d o w n to the water zone, the water is pulled toward the well, resulting in a cone-shaped distortion of the oil-water contact. If a well is p r o d u c e d too hard, water b r e a k t h r o u g h can occur. The same basic principles apply to the d o w n w a r d coning of gas f r o m a gas cap.
To minimize the production of u n w a n t e d gas or water, flow rates from the well must be maintained so that the
pressure gradients around the wellbore will not cone in the unwanted fluids. Flow performance calculations or reservoir models can be used to help predict the optimum production rates. However, because of pressure depletion within the oil zone, a pressure differential will eventually develop between the water and/or gas zones, such that encroachment will eventually occur. After this point, the well may soon become uneconomical due to the production of these u n w a n t e d fluids.
Several authors have studied these types of reservoirs to determine which production methods result in the m a x i m u m a m o u n t of p r o d u c e d oil. Byrne and Morse (1973) evaluated various reservoir and completion scenarios and determined that increasing the well penetration and increasing the pressure drawdown at the producing well would result in an increase in the water-oil ratio. However, there w a s a corresponding increase in the oil production of the s a m e o r d e r of m a g n i t u d e . Their results s h o w e d no indication that ultimate recovery was adversely affected by the increased producing rates.
NEAR WELLBORE RESTRICTIONS
Restrictions can occur in the formation or within the wellbore that can cause a decrease in oil or gas production. These restrictions are a result of changes in the formation or fluid properties around the wellbore, chemical reactions within the formation or the wellbore, mechanical problems, or inadequate completion techniques.
Partial Penetration
If only a portion of the p r o d u c t i v e f o r m a t i o n h a s been drilled or perforated, this is referred to as partial penetration. Gas or oil that is flowing from the reservoir to this limited area will cause large pressure gradients near the wellbore. Fluid flow within the pores may also reach turbulent velocity, thus generating additional pressure drop and r e d u c i n g the productivity of the well. Studies by Harris (1966) and Odeh (1968) on partial penetration have concluded that the effects are similar to skin damage around the wellbore (that is, an altered permeability that is lower than the reservoir permeability). To overcome these effects, the well m u s t be d e e p e n e d to e x p o s e all of the productive interval or more perforations must be added to expose additional reservoir to flow. Figure 2 illustrates the effects of partial penetration on p r e s s u r e d r o p a r o u n d the wellbore.
Figure 1. Wave coning.
Scales and Precipitates
Plugged or inadequate perforations have the same effect as partial penetration; that is, reservoir fluids are restricted from flowing into the wellbore. Perforations can become
492
Production Problems
493
reduce this problem, a method must be implemented, such as hydraulic fracturing, to help decrease the pressure drop around the wellbore (see the chapter on "Stimulation" in Part 9). Eventually, pressure will deplete further out in the reservoir and the flow capacity of the primary fluid will be reduced.
Figure 2. Effect of partial penetration on flowing pressure.
plugged with scale or other solid particles that are precipitated from formation fluids (see the chapter on "Rock-Water Interactions: Formation Damage" in Part 5). As these f l u i d s are p r o d u c e d , e q u i l i b r i u m conditions of pressure and temperature become altered (especially near the wellbore) causing precipitation of these particles. Some precipitates can form when incompatible fluids come in contact downhole. Workover and completion fluids should be tested with formation fluids for compatibility. Precipitates can also occur in a formation after an acid treatment.
The most c o m m o n types of scale are calcium carbonate and calcium sulfate (gypsum). Sodium chloride (salt), sulfur, and other minerals can also be precipitated under certain conditions. The formation of these scales or solids can be m i n i m i z e d or p r e v e n t e d t h r o u g h use of chemical inhibitors, provided the produced fluids are analyzed a h e a d of time. If n o t , it m a y b e n e c e s s a r y to p e r f o r m a workover to clean out the well.
Paraffin or Asphaltenes
S o m e oils can p r e c i p i t a t e a p p r e c i a b l e a m o u n t s of paraffin or asphaltenes as the temperature and pressure are reduced in the near wellbore region. Most often these deposits form near the surface where the temperature is lowest. However, there h a v e been m a n y confirmed cases of p a r a f f i n b u i l d u p near the perforations. Removal of these deposits can be accomplished with heated oil or chemical solvents.
Decrease in Relative Permeability
As the well is produced, the lowest pressures are observed near the wellbore. In oil reservoirs, this pressure d r o p can result in the release of gas. Similarly, in a gas reservoir, the lower pressure can result in the formation of condensate fluids around the wellbore. The increase in s a t u r a t i o n of a second p h a s e of f l u i d d e c r e a s e s the permeability to flow of the p r i m a r y fluid. The ability of these different fluids to flow is defined by the relative permeability characteristics of the rock and the fluids. To
Emulsions
Emulsions can also form in or around the wellbore and may create a block that restricts or completely cuts off p r o d u c t i o n f r o m the r e s e r v o i r . A n emulsion is s i m p l y a m i x t u r e of one liquid (the d i s c o n t i n u o u s phase) that is dispersed within another liquid (the continuous phase) and is likely to be formed when oil and water are mixed. Stable emulsions are the most likely to block production and are also the most difficult to remove. For an emulsion to become stable, an emulsifying agent (fine particles and/or surfactants) must be present. Although surfactants can cause emulsions, other surfactants at the proper concentrations can break emulsions. Also, special deemulsifier agents are available to remove emulsion blocks.
Formation Collapse
In loosely consolidated or weakly cemented formations, the severe pressure drop that occurs around a wellbore can cause the formation to collapse. When this occurs, the pore structure is altered and the permeability is reduced, causing skin damage around the wellbore. To prevent this, a small hydraulic fracture treatment can be performed early in the life of the well. The fracture treatment will help minimize the pressure gradients and reduce the chance of formation collapse around the wellbore.
As the pressure is depleted throughout the reservoir, the overburden pressure caused by the overlying rock tends to compress the formation. This overburden pressure can also alter the pore structure and reduce the permeability. The d e g r e e of s e v e r i t y is a f u n c t i o n of the f o r m a t i o n compressibility. Soft formations, such as those found in the Gulf Coast, can experience reductions in permeability by a factor of 10 or greater. More c o m p e t e n t s a n d s t o n e s or carbonate reservoirs may only experience a small reduction in permeability. These effects must be taken into account when predicting long-term performance; otherwise, predictions will be too optimistic.
CORROSION
The potential for corrosion must be recognized and diagnosed early in the p r o d u c i n g life of a field. Otherwise, expensive workovers and repairs may be necessary. Corrosion within the wellbore can occur in various forms a n d is o f t e n a c c o m p a n i e d by the f o r m a t i o n of scale (discussed previously). Corrosion can be divided into three categories:
• Oxidation • Electrical corrosion • Chemical corrosion
494
PART 9—PRODUCTION ENGINEERING METHODS
0
Щ
Figure 3. Effects of corrosion on metal pipe.
Oxidation is essentially the exposure of metal to oxygen primarily from air, but it can also be obtained from oxygen in p r o d u c e d formation water. The result is the formation of rust on the tubulars in the wellbore. Prevention of rust can be accomplished through the use of special resistant alloys in the tubulars, chemical inhibitors, and cathodic protection.
Electrical corrosion is c a u s e d b y the n a t u r a l f l o w of a n electrical current from the metal pipe (anode) into the formation (electrolyte) or into another metal that is lower in the electromotive series (cathode). As current flows, the a n o d e will corrode. To prevent this type of corrosion, the expected anode can be insulated or coated. Cathodic protection can also be installed to slow d o w n the flow of current.
Chemical corrosion is c a u s e d b y s o l u b l e o r g a n i c a c i d s within the produced formation water or carbon dioxide within gas in the reservoir. These agents react with the metal and cause severe pitting (Figure 3). The incidence a n d severity of this type of corrosion is increased w i t h higher bottomhole temperatures.
Another form of corrosion is that f r o m h y d r o g e n sulfide (H2S), w h i c h is a c o m b i n a t i o n of electrical a n d chemical corrosion. The severity and potential d a n g e r of this type of corrosion makes it worth special mention (see the chapter on "Wellsite Safety" in Part 3). Hydrogen sulfide causes damage to tubulars by several different mechanisms. First, a sulfide scale forms and causes pitting as the steel acts as an anode and transmits current to the sulfide scale. As a result of f o r m a t i o n of the s u l f i d e scale, h y d r o g e n is liberated, which is absorbed by the steel. The hydrogen combines with the carbon compounds in the steel, reducing its strength and causing it to blister and crack. Since hydrogen sulfide is a poisonous gas, its production should be c o n s i d e r e d d a n g e r o u s b e c a u s e of the p o t e n t i a l for damage to surface and downhole equipment. Special corrosion inhibitors are commercially available to help protect the metal e q u i p m e n t f r o m these types of corrosion.
LIQUID LOADING
Liquid that enters the wellbore at the bottom of the hole can only f l o w to the s u r f a c e if the d i f f e r e n c e in p r e s s u r e between the bottom a n d the top of the well is greater than the h y d r o s t a t i c p r e s s u r e of the fluid c o l u m n , p l u s a n y friction that occurs as a result of flow u p the tubing. If the pressure is not sufficient, the fluid will remain static in the w e l l b o r e . T h i s is r e f e r r e d to as liquid loading. If g a s is primarily being produced, water flowing from the formation, or condensed within the tubing, can also a c c u m u l a t e in t h e w e l l b o r e if t h e g a s is n o t f l o w i n g at sufficient velocity to lift the water from the well. Accumulation of this static fluid column can impose an additional backpressure on the formation that can significantly r e d u c e the p r o d u c t i v i t y of the well or can actually "kill" the well so that it does not flow at all.
When liquid loading occurs, it m a y be necessary to change the wellbore configuration or to install artificial lift equipment. For example, the size of the tubing affects the velocity of the fluid flowing u p the well. If smaller tubing is used, the velocity of the fluid will be greater because of the smaller cross-sectional area. From a s t u d y of over 100 flowing gas wells, Turner et al. (1969) developed a useful correlation to d e t e r m i n e the m i n i m u m flow rate (and thus, velocity) required to lift liquids continuously from a gas well. Turner's correlation can be used to determine the size tubing required to remove water continuously from a well that is flowing at a specific rate.
In oil or gas wells that m a k e large v o l u m e s of water or condensate, it may be necessary to install artificial lift equipment to p u m p the liquids to the surface. Most c o m m o n m e t h o d s of artificial lift i n c l u d e rod p u m p s , submersible pumps, and gas lift valves (see the chapter on "Artificial Lift" in Part 9).
MECHANICAL FAILURES
Mechanical failures can occur in a well for a variety of reasons. They are usually related to one of the following: (1) the corrosion of d o w n h o l e equipment, (2) the collection of d e b r i s or scale in the wellbore, (3) the p r o d u c t i o n of formation sand or collapse of the formation, (4) insufficient cement protection, or (5) the u s e of e q u i p m e n t that is not designed to withstand the depth, temperature, or pressure of a well.
The type of e q u i p m e n t normally p r o n e to such failure is the tubing, casing, rod string, downhole pumps, packers, gas lift valves, and plugs in the wellbore (Figure 4). After long-term exposure to downhole conditions, breakage or w e a r can occur in a n y of this e q u i p m e n t . Extensive workovers (specifically, fishing operations) can be p r e v e n t e d by r e g u l a r m a i n t e n a n c e of t h e d o w n h o l e equipment (see the chapter on "Workovers" in Part 9). A proper maintenance schedule can be established for any field once one has properly analyzed the produced fluids a n d conditions of the reservoir.
Specific diagnostic techniques can be applied in an effort to determine where a failure has occurred. For example, a
PARTED RODS STUCK PUMP
CASING LEAK PLUG LEAKING
Figure 4. Examples of mechanical failure.
Production Problems
495
casing leak can be detected by running a packer in the well and pressure testing the casing at various intervals in the hole. Problems with rod pumping equipment can be d i a g n o s e d t h r o u g h the u s e of an i n s t r u m e n t called a dynamometer. This i n s t r u m e n t a n a l y z e s the strain o n the polish rod at the surface as the p u m p goes through an upstroke and downstroke cycle. Other diagnostic methods are available, and in some instances, it may be necessary to perform several tests before mechanical problems can be pinpointed.
Workovers
Alan F. Osborne
Shell Oil Company (Retired) Houston, Texas, U.S.A.
INTRODUCTION
A workover is any operation d o n e on, within, or through the wellbore after the initial completion. Although proper drilling, cementing, and completion practices minimize the need, virtually every well will need several workovers during its lifetime to satisfactorily fulfill its purpose.
To help understand w h y workovers are required, it is useful to visualize a wellbore. A typical wellbore schematic is s h o w n in Figure 1. The production casing is only 5 to 9 in. in diameter, and it has been run into a corkscrew-shaped hole that is not entirely circular and may be substantially deviated from vertical. An attempt has been made to cement the casing centrally, at least across the productive intervals, to obtain a hydraulic seal and casing support. Inside the production casing, even smaller equipment is run for completion and production or injection purposes. Figure 2 illustrates the tight
fit of these concentric strings. A l t h o u g h the diameters are small, the surface area of the equipment is significant and may be exposed to corrosive fluids and pressure for long periods of time.
Another visualization to keep in mind is the very small wellbore size relative to the reservoir area to be drained by that well. A minor restriction (such as scale, fines, paraffin, or pipe dope) at the wellbore has a significant effect on production rate and ultimate recovery.
REASONS FOR WORKOVERS
W o r k o v e r s m a y be r e q u i r e d for o n e or m o r e of the following reasons:
• Unsatisfactory production or injection rates • Supplemental recovery project requirements • Regulatory requirements • Competitive drainage • Reservoir data gathering • Lease requirements • Abandonments
Figure 1. Components of a well.
Tlie majority of workovers are d o n e because the well is not performing up to expectations (see the chapter on "Production Problems" in Part 9). Careful analysis will determine if the decline is abnormal or the result of normal r e s e r v o i r d e p l e t i o n . If d e e m e d a b n o r m a l , it m u s t b e determined if the problem is in the reservoir inflow system, the wellbore outflow system, or both. Inflow problems can be corrected with stimulation procedures such as acidizing, fracturing, scale, or paraffin treatments (see the chapter on "Stimulation" in Part 9) or by reperforating or additional perforating. Outflow problem resolution may require equipment changes, cleanouts, or chemical treatments.
Mechanical p r o b l e m s such as failures of the cement, tubulars, packers, wireline components, safety valves, or artificial lift equipment are also reflected in the well performance. Cement and casing problems are most often corrected by squeezing cement into the damaged areas (Figure 3). Very serious problems of this nature occasionally require a sidetracking operation. Depending on the initial wellbore configuration, some mechanical problems can be corrected by applying wireline methods. More often, however, the equipment must be pulled using a rig. This may b e a r o u t i n e o p e r a t i o n , b u t it m a y b e a m a j o r task if the equipment is stuck or has parted downhole. To recover parted or stuck downhole equipment, a fishing operation may be required (see the chapter on "Fishing" in Part 3).
Initiation of s u p p l e m e n t a l r e c o v e r y projects c a u s e s workovers because many wells must be converted to injectors, observation wells, or disposal wells. Also, all wells need to be completed in correlative zones, thus a significant
496
8" Hole
5-1/2" Casing
-2-7/8" Tubing
1" Rod with Slimhole Coupling
Workovers
497
TUBING
SQUEEZE PACKER
PROOUCING ZONE
а Г М - - T--V
HOLE IN — CASING
RETRIEVABLE - BRIDGE PLUG
L
Figure 2. Horizontal cross section of a well.
n u m b e r of wells need perforating a n d stimulation a n d / o r zones shutoff via squeeze cementing or plugs (cement or mechanical).
Workovers done for the reasons just noted are generally completed to yield a profit in a reasonable length of time. Those done for the following reasons may have little profit incentive, but are still necessary for prudent, legal operations:
• Regulatory requirements may initiate workovers to reduce GORs, isolate pressures, or install safety equipment.
• Competitive conditions occasionally generate workovers designed to protect correlative rights, which often affects the timing m o r e than the goal of doing that particular job.
• Data gathering, particularly of special reservoir information (such as pressures, fluid samples and zonal productivity/injectivity tests) required for reservoir management or in connection with pilot tests, may require a workover.
• The requirement to maintain production to hold a lease may encourage a workover to allow more time to determine proper lease disposition.
Figure 3. Casing repair via squeeze cementing.
• Workovers are normally required to temporarily abandon or to plug and abandon a well. Operations may include squeeze cementing, setting plugs (cement or mechanical), retrieving production equipment, and perhaps cutting and pulling casing. Workover costs vary depending on well conditions and regulatory requirements.
WORKOVER CONSIDERATIONS
Although the considerations necessary to design a workover are virtually the same as those noted for standard well completions (see the chapter on "Well Completions" in Part 9), some require special mention:
• Safety must receive special attention. The tendency to consider a workover as "routine" sometimes reduces the level of safety attentiveness.
• Well pressure m a y be a primary consideration. If the open zone has experienced significant depletion, the circulation of fluids becomes difficult or even impossible. The well killing process will introduce significant amounts of fluids into the reservoir with the
498
PART 9—PRODUCTION ENGINEERING METHODS
Figure 4. Recompletion.
potential for formation damage. If the wellbore is n o w open to high pressure, safety is of p a r a m o u n t importance. If existing well equipment has been in the well for a long time, it m a y not be capable of containing high shut-in pressure. • In thermal projects, the elevated temperatures create special needs for doing a workover safely. • The condition of the well equipment m a y be a major factor in what and how much is done in a particular workover. The toll taken by corrosion, erosion, and mechanical wear m a y be significant, particularly if the casing is involved. Extra funds and time may need to be allocated to cover potential squeeze cement jobs, fishing operations, and equipment replacement.
TYPES OF WORKOVERS
To some degree, workovers are classified on the basis of the reasons for them (discussed earlier). Another way to describe types of w o r k o v e r s is to divide them into (1) jobs performed primarily to influence the reservoir and (2) jobs applied to the wellbore (including the cement) and its associated equipment. Workovers done primarily to influence the reservoir can be subdivided into (1) w o r k d o n e for the zone already open and (2) work done to shut off the existing z o n e in f a v o r of o p e n i n g a n e w z o n e , t e r m e d a recompletion.
Operations typically done in an existing zone may include stimulation, reperforating, perforating additional intervals,
and plugging off u n w a n t e d perforations (because of high gas or water production).
Recompletion work varies depending on whether the new zone is above or below the currently open zone. If above, the lower zone will be abandoned via squeeze cementing (Figure 4), a cement plug, or a mechanical bridge plug, and the zone of interest will be perforated a n d stimulated. Appropriate outflow equipment will then be installed. If the n e w zone is below, the existing zone will probably be squeeze cemented (in the same manner as repairing casing, as shown in Figure 3). The cement in the wellbore is drilled out, the lower zone perforated and stimulated, and production equipment run. Some wells have had sufficient well work done upon initial completion to permit recompletion by simply using a wireline to run plugs and open sleeves.
Wellbore workovers can include casing or equipment repair, but may also simply involve cleaning out fill over the producing zone or circulating chemicals to remove scale or paraffin.
Many workovers involve the installation, maintenance, and repair of artificial lift equipment in the wellbore (see the chapter on "Artificial Lift" in Part 9). About 85% of the oil wells in the United States are on some form of artificial lift. Approximately 85% of those wells are on rod p u m p , 10% on gas lift, and 5% on submersible or hydraulic pumps. Rigs are required for most artificial lift workovers; however, many gas lift wells can be serviced with a wireline.
WORKOVER EQUIPMENT
W o r k o v e r s can be d o n e with conventional rigs (smaller but similar to drilling rigs) or nonconventional systems. Conventional rigs can be equipped to handle almost all types of work that m a y be required (see the chapter on "Land Rigs" in Part 3). Nonconventional systems allow specific types of work to be done without pulling the tubing, disassembling the Christmas tree, or even killing the well. This is accomplished by using lubricators and packoff equipment at the surface and by running the required equipment inside the p r o d u c t i o n tubing. C o m m o n types of n o n c o n v e n t i o n a l systems are wireline units, coiled tubing units, and snubbing units. Generally, these systems are designed to do a relatively narrow scope of work.
Wireline units use special equipment (Figure 5) on a solid wireline to gather data (pressure, temperature, and depth) and to set, manipulate, and/or retrieve tubing plugs, sliding sleeves, flow regulators, gas lift valves, safety valves, and wireline fishing tools. Wireline equipment can also be used to cut paraffin and remove wellbore fill. Prior planning, good operators, good equipment, and reasonable well conditions are prerequisites to the success of this method.
Coiled tubing units (Figure 6) find application in cleanout work, stimulation, plugback jobs, and unloading wells with nitrogen. Coiled tubing units can have 10,000 to 15,000 ft of pipe (usually 1-in. outside diameter) that can be reeled continuously into the wellbore. The pipe can be used to p u m p into the well a n d circulate fluids such as water, acid, or cement. The use of a dynadrill permits some drilling with coiled tubing. Limitations of coiled tubing include the reduced strength of the tubing and low p u m p i n g rates (high
I WIRELINE SOCKET STEM
Workovers
499
JARS
KNUCKLE JOINT
RUNNING OR PULLING TOOL
Figure 6. Coiled tubing unit.
Figure 5. Wireline equipment, (a) Lubricator, (b) Typical string.
friction pressure because of the diameter and length). Coiled tubing can be run into wells against pressures u p to 5000 psi.
Snubbing units (Figure 7) use hydraulic pressure and rams to introduce small, coupled tubing into the well, if necessary, against pressures above 5000 psi. The tubing is stronger and can be rotated so that tougher cleanout or fishing jobs can be accomplished. Snubbing units are used for the same type of jobs as are coiled tubing units, but their compactness is an advantage in offshore work. They are slow and expensive, but in the right applications, they are the more economical way to accomplish the task.
Figure 7. Snubbing unit.
500
PART 9—PRODUCTION ENGINEERING METHODS
Part 9 References Cited
Allen, T. O., and A. P. Roberts, 1978, Production Operations, Volume 1: Tulsa, OK, Oil and Gas Consultants International, 225 p.
Anderson, R. A., J. J. Smolen, L. Laverdiere, and J. A. Davis, 1980, A production logging tool with simultaneous measurements: Journal of Petroleum Technology, February, p. 191-198.
Atlas Wireline Services, 1986, Interpretative methods for production well logs, 3rd ed.: Houston, TX, Document No. 9441.
Bryne, W. B., Jr., and R. A. Morse, Water coning may not be harmful—Part I: The Oil and Gas Journal, Sept. 3, p. 66-70.
Conoco, Inc., 1990, Beam Pumping Engineering and Operating Manual, July.
Craft, B. C., and M. F. Hawkins, 1959, Applied Petroleum Reservoir Engineering: Englewood Cliffs, NJ, PrenticeHall.
Earlougher, R. C., Jr., 1977, Advances in Well Test Analysis: Dallas, TX, American Institute of Mining, Metallurgical and Petroleum Engineers, Society of Petroleum Engineer's Monograph 5, 264 p.
Fetkovich, M. J., 1980, Decline curve analysis using type curves: Journal of Petroleum Technology, June, p. 10651077.
Fitzgerald, D. D., B. F. McGhee, and J. A. McGuire, 1983, Guidelines for 90% accuracy in zone isolation decisions: Richardson, TX, Society of Petroleum Engineers, SPE 12141.
Froelich, B., A. Dumont, D. Pittman, and B. Seeman, 1982, Cement evaluation tool—a new approach to cement evaluation: Journal of Petroleum Technology, August.
Govier, G. W., 1975, Theory and Practice of Testing Gas Wells, 3rd ed.: Calgary, Alberta, Canada, Energy Resources and Conservation Board, Pub. ECRB-75-34, 504 p.
Hale, B. W., 1986, Analysis of tight gas well production histories in the Rocky Mountains: Richardson, TX, SPEPE, July, p. 310-22.
Harris, M. H., 1966, The effect of perforating on well productivity: Journal of Petroleum Technology, April, p. 518.
Lee, W. J., 1982, Well Testing: Dallas, TX, Society of Petroleum Engineers of AIME, 159 p.
Lee, W. J., T. B. Kuo, S. A. Holditch, and D. A. McVay, 1984, Estimating formation permeability from single-point flow data: Proceedings of the 1984 S P E / D O E / G R I Unconventional Gas Recovery Symposium, Pittsburgh, PA, p. 175-186.
Matthews, C. S., and D. G. Russell, 1967, Pressure buildup and flow tests in wells: Dallas, TX, Society of Petroleum Engineers Monograph Series No. 1,172 p.
McGuire, W. J., and Sikora, V. J., 1960, The effect of vertical fractures on well productivity: Transactions, AIME, v. 219, p. 401.
Muskat, M., 1949, Physical principles of oil production: New York, McGraw-Hill, chap. 5.
Odeh, A. S., 1968, Steady-state flow capacity of wells with limited entry to flow: Society of Petroleum Engineers Journal, March, p. 43.
Odeh, A. S., and L. G. Jones, 1965, Pressure d r a w d o w n analysis, variable-rate case, in Pressure Analysis Methods: Dallas, TX, American Institute of Mining, Metallurgical and Petroleum Engineers, Society of Petroleum Engineers Reprint Series No. 9,256 p.
Pardue, G. H., R. L. Morris, L. H. Gollwitzer, and J. H. Moran, 1963, Cement bond log—a study of cement and casing variables: Journal of Petroleum Technology, May.
Schlumberger, 1989, Cased hole log interpretation principles/applications: Houston, TX, Schlumberger Educational Services, Document No. SMP-7025.
Society of Petroleum Engineers, 1985, Production logging: Richardson, TX, SPE Reprint Series No. 19.
Turner, R. G., M. G. Hubbard, and A. E. Pukler, 1969, Analysis and prediction of m i n i m u m flow rate for the continuous removal of liquids from gas wells: Journal of Petroleum Technology, Nov., p. 1475-1482.
Western Atlas International, 1985, Acoustic Cement Bond Log and Prolog CBL: Houston, TX, n. 2206.
Whittaker, A. H., 1987, Mud logging, in H. B. Bradley, ed., Petroleum Engineering Handbook: Richardson, TX, Society of Petroleum Engineers.
Zaba, J., and W. T. Doherty, 1970, Practical Petroleum Engineers' Handbook, 5th ed.: Houston, TX, Gulf Publishing Company, chap. 6.
Part 10
edited by
RESERVOIR ENGINEERING METHODS
E-G- (skiP>Rhodes
"" ЕхEрxЫploТiTеeсcНh Inc. Houston, Texas, U.S.A.
Contents
• Introduction • Petroleum Reservoir Fluid Properties • Fundamentals of Fluid Flow • Reserves Estimation • Drive Mechanisms and Recovery • Waterflooding • Enhanced Oil Recovery • Reservoir Modeling for Simulation Purposes • Conducting a Reservoir Simulation Study:
An Overview • ReferencesCited
Introduction
E. G. (Skip) Rhodes
Landmark Graphics Corporation Houston, Texas, U.S.A.
As development geology continues to emerge as a legitimate career track within some companies and within the industry in general, cross-training between the geological and engineering disciplines may become more commonplace. Meanwhile, the geologist and the reservoir engineer must meet half way, quantifying the qualitative aspects of reservoir properties that affect well behavior and impact reserves estimates and that constrain reservoir management techniques.
This part of the Manual on reservoir engineering contains a thread of continuity f r o m specific to general. The eight chapters that follow provide an overview and introduction to the field of reservoir engineering rather than offering a handbook in applied methods. In some cases, complete t r e a t m e n t of a particular aspect of reservoir behavior has been simplified to the generalized case. Workers who are immersed in a specific type of reservoir are encouraged to use the References Cited list as a "road map" to additional sources.
In the first chapter on reservoir fluid properties, Curtis Whitson has presented reservoir fluids in the context of three p h a s e s : w a t e r , oil, a n d g a s . C o m m o n l y t e r m e d PVT ( p r e s s u r e - v o l u m e - t e m p e r a t u r e ) behavior, these aspects of reservoir fluids have a profound effect on the performance character of i n d i v i d u a l wells a n d entire fields. The PVT p a r a m e t e r s of a r e s e r v o i r m a y a p p e a r as a l a b y r i n t h of complex formulas and diagrams that occasionally condemn prospects strictly on engineering terms. Whitson has repackaged the essentials into a succinct reference guide.
The second chapter by Michael Golan is a companion article to Whitson's that explains the behavior and m e c h a n i s m s associated w i t h the m o v e m e n t of reservoir fluids into the wellbore. The details of fluid flow behavior are r e g a r d e d by s o m e as a m e r e extension of the s a m e labyrinth in which PVT resides. Golan has attempted to distill this relatively complex subject into a f e w pages of "need-to-know" principles and terminology. Flow potential is s h o w n to consist of a ratio of two independent factors— rock permeability and fluid viscosity. Both factors are a constant concern for the development geoscientist.
The third contribution, by Rick Sustakoski and Diana Morton-Thompson explains the complex interaction of fluid properties, drive mechanism, and reservoir geometries that enter into reserve calculations. The determination of reserves
is one of the most iterative processes between engineers and geologists. Sustakoski and Morton-Thompson review the formulas and reservoir factors to which both the engineer and the geologist must contribute their findings in order to determine accurate reserves.
P7Iuids enter the wellbore because they are driven by some m o d e of energy stored within the reservoir. This energy, referred to as the drive mechanism, is the topic of Stephen Sills' chapter on drive mechanisms and recovery. Although dependent on PVT properties and flow mechanisms, each major drive mechanism has a distinctive production pattern. Recognition and optimization of this behavioral pattern early in the history of the reservoir allows potential for greater recovery efficiency.
The next two chapters—waterflooding by Sam Sarem and enhanced oil recovery by Vernon Breit—explain the principles, behavior, and recovery expectations associated with secondary and tertiary reservoir management after the reservoir's natural primary energy has been depleted. The principles of PVT and fluid flow behavior emerge again in these two chapters, reminding the development geologist of the importance of physical and thermodynamic behavior of water, oil, and gas within reservoirs.
As the reservoir engineer's access to greater desktop computing power grows along with declining hardware prices, the engineer is requiring a higher degree of specificity from the geologist with regard to reservoir modeling input. The last two contributions explain the methods by which geological data are prepared for use in reservoir simulation. In the first of these chapters, Koen Weber provides a list of data types used in simulation. The companion chapter, by Scott L a u d e m a n , describes the preparation of fluid and matrix properties prior to simulation input. Laudeman concludes by demonstrating that the geosientist may provide the ultimate quality control on simulation results.
Acknowledgments
I wish to acknowledge and thank the authors of Part 10 and their employers for their patience and cooperation. Also, several external reviewers invested considerable effort in the r e f i n e m e n t of this part, a n d these e f f o r t s are m u c h appreciated.
574
Petroleum Reservoir Fluid Properties
Curtis H. Whitson
Pera a/s Trondheim, Norway
INTRODUCTION
Petroleum reservoirs may contain any of the three fluid phases—water (brine), oil, or gas. The initial distribution of phases depends on depth, temperature, pressure, composition, historical migration, type of geological trap, and reservoir heterogeneity (that is, varying rock properties). The forces that originally distribute the fluids are gravity, capillary, molecular diffusion, thermal convection, and pressure gradients. It is generally assumed that reservoir fluids are in a static state when discovered or, more correctly, that fluids are moving at a very slow rate relative to the time required to extract the fluids (10 to 50 years). Clearly the fluids may still be in a dynamic state in terms of geological time.
Because gravity is the dominant force in distributing fluids through geological time, hydrocarbons migrate upward and are trapped against impermeable cap rock. Gas overlies oil which overlies water. However, because the reservoir pores are usually saturated completely by water before hydrocarbon migration and because capillary forces acting to retain water in the smallest pores exceed gravity forces, an initial (connate) water saturation will always be found in hydrocarbon-bearing formations. The connate water saturation may vary from 5 to 50% with the hydrocarbons still having sufficient mobility to produce at economical rates.
This chapter reviews the physical and thermodynamic properties of gas, oil, and reservoir brine. As commonly done, the p h a s e and v o l u m e t r i c b e h a v i o r of p e t r o l e u m reservoir fluids is referred to as PVT (pressure-volumetemperature). Two important general references on PVT are Katz et al. (1959) and Society of Petroleum Engineers (1981).
PROPERTY DEFINITIONS
Some basic fluid property definitions are provided here:
Formation volume factor (FVF)—The ratio of a phase volume (water, oil, gas, or gas plus oil) at reservoir conditions, relative to the volume of a surface phase (water, oil, or gas) at standard conditions resulting when the reservoir material is brought to the surface. Denoted mathematically as Bw (bbl/STB), B0 (bbl/STB), Bg (fbVSCF), and B1 (bbl/STB).
Solution gas-oil ratio (GOR)—The amount of surface gas that can be dissolved in a stock tank oil when brought to a specific pressure and temperature. Denoted mathematically as Rs (SCF/STB).
Solution oil-gas ratio (OGR)—The amount of surface condensate that can be vaporized in a surface gas at a specific pressure and temperature; sometimes referred to as liquid content. Denoted mathematically as rs (STB/MMSCF).
Liquid specific gravity—The ratio of density of any liquid measured at standard conditions (usually 14.7 psia and 60 T) to the density of pure water at the same standard conditions. Denoted mathematically as y„ (where water = 1).
API specific gravity—Another common measure of oil specific gravity, defined by yAPI = (141.5/y0) -131.5, with units in "API.
Gas specific gravity—The ratio of density of any gas at standard conditions (14.7 psia and 60 T) to the density of air at standard conditions; based on the ideal gas law (pV = nRT), gas gravity is also equal to the gas molecular weight divided by air molecular weight (Mair = 28.97). Denoted mathematically as yg (where air = 1).
Bubblepoint pressure—At a given temperature, this condition occurs when an oil releases an infinitesimal bubble of gas from solution when pressure drops below the bubblepoint.
Retrograde dewpoint pressure—At a given temperature, this condition occurs when a gas condenses an infinitesimal drop of oil from solution when pressure drops below the dewpoint.
Saturation pressure—An oil at its bubblepoint pressure or a gas at its dewpoint pressure.
Critical point—The pressure and temperature of a reservoir fluid where the bubblepoint pressure curve meets the retrograde dewpoint pressure curve (see Figures 1 and 2), representing a unique state where all properties of the bubblepoint oil are identical to the dewpoint gas.
Composition or feed—Quantifies the amount of each component in a reservoir mixture, usually reported in mole fraction. Typical components in petroleum reservoir mixtures include the nonhydrocarbons N2, CO2, and H2S and the hydrocarbons C1, C2, C3, /C4, nC4, /C5, nC5, C6, and C7+ (C7+, or "heptanes-plus," includes many hundreds of heavier compounds, such as paraffins, napthenes, and aromatics). Asphaltenes are also found in reservoir oils.
Saturated condition—A condition where an oil and gas are in thermodynamic equilibrium, that is, the chemical force exerted by each component in the oil phase is equal to the chemical force exerted by the same component in the gas phase, thereby eliminating mass transfer of components from one phase to the other.
Undersaturated condition—A condition when an oil or a gas is in a single phase but not at its saturation point (bubblepoint or dewpoint), that is, the mixture is at a pressure greater than its saturation pressure.
575
Table 1. Composition (in mol %) of Several Reservoir Fluids
Component or Property
CO2 N2 C1 C2
C3 /C4 пСл
'C5 nC5
C6s
c7+ GOR (SCF/STB) OGR (STB/MMSCF)
TAPI
M7+
V
Dry Gas
0.10 2.07 86.12 5.91 3.58 1.72
—
0.50
—
—
—
OO
0
— — —
Wet Gas
1.41 0.25 92.46 3.18 1.01 0.28 0.24 0.13 0.08 0.14
0.82
69,000 15 65.0
132 0.750
Petroleum Reservoir Fluid Properties
505
Gas Condensate
2.37 0.31 73.19 7.80 3.55 0.71 1.45 0.64 0.68 1.09
8.21
5965 165 48.5 184 0.816
Volatile Oil
1.82 0.24 57.60 7.35 4.21 0.74 2.07 0.53 0.95 1.92
22.57
1465 680 36.7 240 0.864
Black Oil
0.02 0.34 34.62 4.11 1.01 0.76 0.49 0.43 0.21 1.16
56.40
320 3125
23.6 274
0.920
RESERVOIR WATER
The water found in petroleum reservoirs is usually a brine consisting mostly of s o d i u m chloride (NaCl) in quantities f r o m 10 to 350 p p t (%o); seawater has about 35 ppt. Other compounds (electrolytes) found in reservoir brines include calcium (Ca), magnesium (Mg), sulfate (SO4), bicarbonate (HCO3), iodide (I), and bromide (Br). Brine specific gravity increases with salinity in units of about 0.075 per 100 ppt.
At reservoir conditions, the brine that is sharing pore space w i t h h y d r o c a r b o n s a l w a y s contains a limited a m o u n t of solution gas (mainly methane), from about 10 SCF/STB at 1000 psia to about 35 SCF/STB at 10,000 psia for gas-water systems and slightly less for oil-water systems. Increasing salinity decreases gas in solution. Water compressibility ranges f r o m 2.5 to 5 x IO-6 psi-1, decreasing with increasing salinity. Water viscosity ranges from about 0.3 cp at high temperatures (>250 °F) to about 1 cp at ambient temperatures, increasing with increasing salinity. Finally, reservoir brines exhibit only slight shrinkage (<5%) when produced to the surface.
PETROLEUM RESERVOIR CLASSIFICATIONS
Petroleum reservoirs are usually classified into five fluid categories (Cronquist, 1979):
1. D r y g a s 2. W e t g a s 3. Gascondensate 4. Volatileoil 5. Black oil
The first three of these are gas reservoir fluid types, which are in a gaseous state at virgin reservoir conditions, meaning that the critical temperature of the reservoir fluid is less than the reservoir temperature. Dry gas and wet gas fluids consist mainly of light and intermediate hydrocarbons (N2, CO2, H2S,
and C1 to C2), in which no liquids will condense in the reservoir rock during pressure depletion. Wet gases produce high API condensate (distillate) at surface conditions in amounts usually less than about 5 STB/MMSCF. The OGR should remain constant throughout the depletion of a wet gas reservoir.
Gas condensates, in contrast, contain significant amounts of C5+ c o m p o n e n t s , a n d they exhibit the p h e n o m e n o n of retrograde condensation at reservoir conditions, in other words, as pressure decreases, increasing amounts of Uquid condenses in the reservoir (down to about 2000 psia). This results in a significant loss of in situ condensate reserves that m a y only be partially recovered by revaporization at lower pressures. Gas condensate reservoirs exhibit producing gas-oil ratios from 2500 to 50,000 SCF/STB (400 to 10 STB/MMSCF). Gas cycling projects designed to avoid liquid loss from retrograde condensation can usually be justified for fluids with liquid content higher than about 50 to 100 STB/MMSCF. Offshore, the m i n i m u m liquid content to justify cycling is about 100 STB/MMSCF.
Reservoir oils are classified as either black oil or volatile oil, the former being more commonly discovered in the first 50 years of the oil industry. Volatile oil reservoirs have become the "norm" in the past 20 years, mainly because discoveries are at greater depths with higher initial pressures. A clear demarkation between these two oil types is not easily made, although a gas-oil ratio of about 750 SCF/STB is probably a good indicator (black oils have lower GORs). Volatile oils may have GORs u p to 2500 SCF/STB and formation volume factors as large as three (meaning that the oil shrinks by a factor of three w h e n p r o d u c e d to the surface). Another characteristic of volatile oil reservoirs is that the reservoir gas that evolves and flows into the wellbore will contain significant quantities of Hquids that m a y eventually contribute the m a j o r i t y of s u r f a c e oil p r o d u c t i o n at late stages of depletion.
Table 1 gives s o m e typical reservoir fluid compositions and properties. Figure 1 shows a pressure-temperature
506
PART 10—RESERVOIR ENGINEERING METHODS
Reservoir temperature, ° F
Figure 1. Pressure-temperature phase diagram. Reservoir classification would be oil if reservoir temperature were less than 127 F and gas if reservoir temperature were greater than 127 F.
diagram for a specific reservoir fluid composition. Depending on reservoir temperature, this fluid would be defined as an oil or a gas. An oil exhibits a bubblepoint pressure at saturated conditions, while a gas condensate exhibits a dewpoint pressure at saturated conditions.
If a reservoir contains both a gas cap and an oil zone, then both fluids are normally at saturated conditions initially. Initial pressure equals the dewpoint of the gas cap fluid, a n d it equals the bubblepoint of the underlying oil (Figure 2). The repeat formation tester (RFT) has m a d e the determination of initial fluid contacts possible in reservoirs with reasonable permeability, that is, >1 md. A saturated gas cap in equilibrium with an underlying saturated oil, for example, will be seen as a sharp discontinuity in RFT pressures at the gas-oil contact.
In the past 20 years, deeper petroleum reservoirs have been d i s c o v e r e d a n d the t r a d i t i o n a l i n t e r p r e t a t i o n of a reservoir containing both gas and oil has changed. An alternative interpretation in some gas-oil reservoirs is that composition varies continuously with depth. Here the fluids at the shallowest elevations are gas condensates, while the fluids at greater depths are oils. Sometimes the initial reservoir pressure may be greater than the saturation pressure of all mixtures in the reservoir, implying that the reservoir is entirely undersaturated even though a gas is at the top and an oil is at the bottom of the reservoir. Reservoirs of this type would not show a sharp contrast in RFT pressures at the d e p t h w h e r e the fluid changes f r o m a near-critical gas to a near-critical oil. Instead they would show a continuously increasing pressure gradient (for example, from 0.2 to 0.3 psi/ft).
Figure 2. Pressure-temperature phase diagrams of gas cap and oil fluids in a reservoir that is initially at saturated conditions.
FLUID PROPERTY CORRELATIONS
Relatively accurate correlations are available for estimating the key fluid p r o p e r t i e s of reservoir s y s t e m s (Table 2). Standing (1977) and McCain (1990) give useful reviews of property correlations for oil and gas, and other correlations are available. Note, however, that for specific producing provinces (such as the Gulf Coast or the North Sea) more accurate correlations may exist.
Equations of state (EOS) a r e n o w c o m m o n l y u s e d to calculate p h a s e a n d v o l u m e t r i c b e h a v i o r of r e s e r v o i r mixtures. In particular, EOS are useful for predicting phase behavior of miscible and immiscible displacement processes resulting from the injection of gases such as carbon dioxide, nitrogen, and lean or enriched natural gas in oil and gas condensate reservoirs. EOS do not usually predict phase and volumetric behavior of reservoir mixtures accurately, thereby requiring a d j u s t m e n t of c o m p o n e n t properties to match experimental PVT data (Whitson and Brule, 1993).
LABORATORY PVT EXPERIMENTS
Experimental PVT measurements are usually obtained for (1) large oil and gas fields, (2) volatile oil and gas condensate reservoirs, and (3) reservoirs w h e r e gas injection is a potential EOR (enhanced oil recovery) m e t h o d . T w o types of fluid samples can be taken during production, or when a well is shut-in:
1. Bottomhole samples, preferred for oils 2. Separator samples, which must be recombined at the
producing GOR during sampling
Recombined separator samples are standard for gas condensate fluids, but they m a y also be used for oil reservoirs.
Table 2. PVTProperties
Gas
Oil and Water
Pseudocritical properties Z factor
Bubblepoint pressure Solution GOR Bubblepoint FVF Density
Isothermal compressibility Viscosity K values Interfacial tension Diffusion coefficients
Bottomhole sampling is preferred for oils if the reservoir is undersaturated (that is, the initial pressure is higher than the bubblepoint pressure).
Standard PVT experiments include compositional gas chromatography (GC) analysis through heptanes-plus (C7+), constant composition expansion, differential liberation expansion, constant volume depletion, and multistage surface separation. Other PVT measurements include true boiling point (TBP) distillation of the C7+ material and multicontact gas injection experiments. Table 3 summarizes these experiments, indicating when they are performed and on what types of reservoir fluids.
Compositional analyses are used to describe the reservoir fluid makeup on a component basis, including calculation of BTU (energy content) of gases, optimization of separator
Petroleum Reservoir Fluid Properties
507
Table 3. Summary of Laboratory Analyses Performed on Reservoir Oil and Gas Condensate Systems
Laboratory Analysis
Oils
Bottomhole sample
*
Recombined composition
+
C7+ TBP distillation
+
C7+ simulated distillation
+
Constant composition expansion
*
Multistage surface separation
*
Differential liberation
*
Constant volume depletion
+
Multicontact gas injection
+
Key: *, standard; +, can be performed; - , not performed.
Gas Condensates
+ * + + * + * +
conditions for liquid yield, and characterization of an EOS for compositional simulation. Differential liberation and constant volume depletion experiments are designed to provide quantitative information about the volumetric behavior of oil and gas condensate reservoirs during pressure depletion. The multistage separator test is used together with differential liberation and constant volume depletion data to calculate b l a c k oil p r o p e r t i e s Rs, B0, Bj,, a n d rs. M u l t i c o n t a c t g a s injection experiments provide important volumetric and compositional data that can be used to "tune" an equation of state (or alternative) model for simulation of gas injection processes.
Fundamentals of Fluid
T-i-i
A
*"
Michael GoIan
University of Trondheim The Norwegian Institute of Technology Trondheim, Norway
INTRODUCTION
Well flow calculations focus essentially on two aspects of fluid flow: pressure profile along the flow path a n d the rate versus pressure relationship at key points of interest (nodes), as illustrated in Figure 1. The main parameters of interest (all in units of psia) are
pR = reservoir pressure pwf = wellbore (bottomhole) flowing pressure Pwh ~ wellhead pressure psp = separator pressure pSJ- stock tank pressure
The corresponding rates are
q0 = oil production rate (STB/day) q& = gas production rate (SCF/day)
The p r e s s u r e difference (pR - p w f ) is called the reservoir drawdown. It is the primary force driving reservoir fluids into the wellbore. Generally, production rates increase with increasing drawdown. Flow into the wellbore induced by d r a w d o w n is called inflow.
The relationship between the production flow rate measured at the stock tank, q0, and the bottomhole flowing pressure, pwf, is called the inflow performance relationship (IPR). The IPR of a well can be determined directly by production test data, or it can be predicted from reservoir data. Whether presented graphically or expressed by a formula, the IPR is a statement of the production capacity and is widely used to design and analyze the production performance of wells
Good general references on flow in reservoirs a n d wells include Golan and Whitson (1991), Bradley (1987), and Craft et al. (1991).
EMPIRICAL IPR EQUATIONS
Several IPR formulas have been developed to represent the i n f l o w b e h a v i o r of v a r i o u s t y p e s of wells. M a t c h i n g a
WATER
Figure 1. Pressure conditions in a simple production system.
formula to multi-rate p r o d u c t i o n data (Figure 2) allows determination of the value of the characteristic constants in the equations, which in turn characterize the productivity of the well. The empirical formulas are the primary tools to quantify well productivity and to perform production calculations.
Productivity Index Equation for Undersaturated Oil
The production rate in undersaturated oil wells is linearly proportional to the drawdown, and the IPR is a straight line (Figure 2a). The equation is
cIo = KPr-PJ
The characteristic constant relating the oil rate to the d r a w d o w n is called the productivity index, Jr and is defined as
j=
PR
%
~Pwf
with units of S T B / d a y / p s i . The productivity index states the n u m b e r s of S T B / d a y p r o d u c e d for every psi of p r e s s u r e drawdown and thus reflects the productivity or deliverability of the well.
W h e n pwf equals atmospheric pressure, the rate is called absolute open flow (AOF) and is often designated as qmax. A O F is a useful indicator of well productivity.
Back Pressure Equations for Saturated Oil and Gas Wells
The equation for oil and gas wells is
Io = cW "Pwf2)"
It has two characteristic constants: the back pressure constant, c, a n d the back pressure exponent, n. The exponent n is a dimensionless n u m b e r between 0.5 and 1.0. It approaches 1.0 for low rate wells and 0.5 for very high rate wells. Values of n and с can be determined graphically f r o m a log-log plot of multiple rate test data in the f o r m of (pR2 - pwf2) v e r s u s q (Figure 2b). The data point can be fitted to a straight line whose slope is 1/??.
Quadratic Equation for Saturated Oil and Gas Wells
For saturated oil and gas wells, the equation is
W - P w V = Aq + Bq2
The characteristic constants A and B are the corresponding
508
Fundamentals of Fluid Flow
509
,Pn
Cartesian ]
Pwf
^max
I I IX
(a)
Pl plot
, I Log - Log
-
CM
5 CL
I
<N CLtr
JZ^
Slope
=
1
n
0.5 < n < 1.0
I., I L I I... I
(b)
q
Back pressure plot
2
£.
Figure 3. Two-phase relative permeability.
CM
Q? I CJ
CM o?
Cartesian | ^ 3 I j Slope = B
(C)
Quadratic plot
,Pr I Cartesian
Vb
f^f
\
(d)
P r > Pb > Pwf
"ease" with which the fluid flows through it. The viscosity is a property of the fluid and reflects the resistance of fluid to flow.
For multiple phase flow, the presence of a second phase in the p o r o u s media reduces the a p p a r e n t permeability of the first p h a s e (for example, the presence of gas reduces the apparent permeability of oil). The presence of each phase in the porous m e d i u m is quantified by the saturation, S, which is the ratio of fluid v o l u m e in a given porous rock to the pore volume of the rock:
Figure 2. Plots of multi-rate production data.
slope and the intercept of the straight line obtained f r o m a Cartesian plot of the multiple rate test data (Figure 2c) in the following form:
W 2 ~ Pwfi/cI versus (A+ Bq)
Extended Range Undersaturated Oil IPR
For wells producing below bubblepoint pressure, pb, while the reservoir pressure is above the bubblepoint (pwf < p b < pR), the IPR assumes the shape shown in Figure 2(d). It can be represented by the following equations:
S0 = V 0 / V p a n d Sg = Уp
where S0 = oil saturation (fraction) S k = gas saturation (fraction) V0 = volume of oil in a given pore volume Vg = volume of gas in a given pore volume Vp = pore volume
The apparent or effective oil permeability, keo, of one phase in the presence of a second p h a s e can be arranged as the product of two terms: absolute permeability, кя, and relative oil permeability, kTQ:
for Pwf>Pb' cIo = ^Pr-PJ
and
for pwf < pb, q0 = /(pR - pb) + (J/2pb)(pb* - pwf2)
EXTENSION OF DARCVS LAW
Darcy's law (see Part 5), which was originally developed for w a t e r flow, has been e x t e n d e d to describe flow of hydrocarbon reservoir fluids (compressible and multiple phases).
For single-phase oil flow, the proportional constant that relates flow rates to pressure differences in the original Darcy's law is broken down into two independent factors: rock permeability, k, and fluid viscosity, ji. For a linear flow system, this gives
q = (A/L)(k//л) Ap
The permeability is a p r o p e r t y of the rock that reflects the
The absolute permeability is a property of the rock and is essentially the permeability measured with single phase or at 100% single phase saturation. The relative permeability is a dimensionless quantity whose magnitude is between 1.0 and 0, d e p e n d i n g on the saturation (Figure 3). Relative permeabilities are measured in core (petrophysical) laboratories, and the results are reported versus the saturation. Some laboratories normalize relative permeability values with values different than the single phase or 100% saturation value, so some caution needs to be taken in interpreting and using reported data.
RADIAL FLOW
Darcy's law can be applied to an ideal well model producing a constant steady-state production rate. The model assumes cylindrical flow in the reservoir w h e r e flow across the formation is horizontal and fluid moves radially toward the wellbore. It also assumes constant pay zone thickness, constant isotropic permeability, and an ideal liquid (homogeneous incompressible liquid in which viscosity is
506
PART 10—RESERVOIR ENGINEERING METHODS
Slope • 141.2 Pressure distribution in infinite radial reservoir
/У/////Л Damaged Undamaged reservoir
УУ/у/УУ
In (r/rw)
No-Flow outer boundaries
Constant pressure boundaries
Figure 4. Pressure distribution in a radial reservoir.
pressure independent). For infinite size reservoirs, the result is
Figure 5. Skin effect.
or in field units,
P = Pwf +
infr/r,)
That is, for every radius r there is a corresponding pressure p that increases logarithmically with r (Figure 4).
Pressure distribution in radial bounded reservoirs is similar to the infinite case for the m o s t of the d r a i n a g e volume. It is different, however, near the boundaries, as shown in Figure 4.
Units of Darcy's Law Formulas
T h e practical field system is w i d e l y u s e d f o r p r a c t i c a l petroleum engineering calculations. It is a hybrid system that consists of various metric, English, and oil field units. It uses the millidarcy (md) for permeability. Other dimensions and units in this system are as follows:
q0 = stock tank oil rate (STB/day, or stock tank barrels per day)
qa = gas flow rate (SCF/day, or standard cubic feet per day)
Id0 = viscosity (cp, or centipoise) B0 = formation v o l u m e factor (RES bbl/STB, or
reservoir barrels to stock tank barrels) pR = reservoir pressure (psia) pwf = bottomhole pressure (psia)
к = permeability (md) h = pay zone thickness (ft) rw = wellbore radius (ft) re = radius of drainage (ft)
The formulas in the following section are in practical field units.
Restricted Entry into the Wellbore (Skin Effect)
The actual pressure distribution differs in most cases from the ideal pressure distribution derived for the ideal model. The following additional pressure drop at the wellbore results from near wellbore phenomena (Figure 5):
zVskin = Pwf(ideal) ~ Pwf(actual)
The most important phenomena are usually flow convergence d u e to limited penetration of the pay zone (partial penetration), impaired permeability adjacent to the wellbore (formation damage), and flow restrictions in the perforations.
It is convenience to express Apskin as a dimensionless quantity, s, called the skin factor, which is linearly proportional to Apskin:
141.2q0/J0B0 APskin =
kh
The skin factor can be determined from well test interpretation and can easily be converted to actual Apskin or Pwf (actual)'
THEORETICAL PRESSURE-RATE RELATIONSHIP (IPR)
Darcy's law applied to the ideal well model gives an IPR equation expressed in t e r m s of reservoir p a r a m e t e r s . For undersaturated oil (a single-phase homogeneous liquid) in radial reservoir with no-flow outer boundaries, the following equation apphes:
tJ O
fc/z(pR-Pwf) =141 .Iii0B0 [ln(re Zrw)-
0.75 + s]
Fundamentals of Fluid Flow
511
This equation is in harmony with the empirical IPR equation for single-phase oil, where
T
%
PR ~Pzof
kh W l 2 ц 0 В 0 [ l n ( r e Zrw)-0.75 + s]
Radius of Drainage
The r a d i u s of d r a i n a g e to be u s e d in the radial flow equation and the productivity index expression is
re = (А/к)05
where A is the well's drainage area in square feet and the radius is in feet. If the drainage area is given in acres, it has to be converted to square feet using the relationship 1 acre = 43,560 ft2.
Expansion of the Ideal Well Model
The theoretically derived IPR equations can be expanded to include the production of gas and saturated oil. For low pressure gas reservoirs (below 2000 psia), the following equation is used:
0.703kh(pj-p2wf) T^i Z [ l n ( r e / r w ) - 0 . 7 5 + s]
where cjg = gas flow rate (SCF/day) T = temperature (R° = F0 + 460) Z = gas compressibility factor /Ji, = gas viscosity (cp)
All the other parameters have the same units as the oil flow equations. There are rigorous methods used to express IPR of gas wells at reservoir pressures higher than 2000 psia. However, the low pressure equation is good for most well productivity calculations.
For saturated oil (simultaneous oil and gas flow), the equation used is
kh cIo = 1 4 1 . 2 [ l n ( r e / r w ) - 0 . 7 5 + s] 2/M°Bp°R
(P2R-PLF)
w h e r e t h e t e r m kro/Jd0B0 is e v a l u a t e d at a v e r a g e reservoir pressure, pR/ and saturation conditions.
There are two practical considerations to keep in mind when using these equations. First, depending on well spacing, the value of ln(re/rw) is in the range of 6.5 to 8. A v a l u e of 7 is a good a p p r o x i m a t i o n in m o s t p r o d u c t i v i t y calculations. Second, the skin factor, s, is positive when the entry to the wellbore is restricted by the skin effect. It is negative if the productivity is better than predicted b y the ideal radial model (for example, in stimulated or fractured wells).
IPR DERIVED FROM RESERVOIR SIMULATION
For the particular case of a solution gas drive reservoir, there is an IPR formula derived by fitting an equation to the results from computer simulations performed for a wide variety of wells with a w i d e range of reservoir parameters. The formula given in normalized form is
cI о =_ 1 - 0 . 2 P wf
cI max
PR
/ - 0 . 8 P Wf
PR
This formula is useful for low rate oil wells. It overestimates the productivity of high rate wells (those producing more than 2000 STB/day).
Example: A well in a solution gas drive reservoir has a reservoir pressure of 4000 psia. A single test point is Cj0 = 200 S T B / d a y with pwf = 3220 psia. Substituting the data in the IPR equation a n d solving for qmax gives
200
!-Oi^Voi cImax
I 4000
= 624 STB/day
322ov
I 4000 J
S u b s t i t u t i n g t h e c a l c u l a t e d Cjmax a n d t h e g i v e n pR in t h e equation calculates points on the IPR, as follows:
Pwf (PSia)
4000 3000 2000 1500 1000
q0 (STB/day)
0 250 437 508 562
EFFECT OF DEPLETION ON IPR FORMULAS
Depletion u s u a l l y results in deterioration of the IPR (Figure 6). These changes can be predicted quantitatively by calculating the changes of the characteristic constants: J, c, qax, A, and B. The exponent n does not change with depletion. If subscript p denotes the present depletion state and subscript / a future state, the present and future characteristic factors are related as follows:
For undersaturated oil reservoirs:
JF = JP
(VOB0)P (MOBOH
Low pressure gas reservoirs: cf = cp a n d nf = np
506
PART 10—RESERVOIR ENGINEERING METHODS
ЧгЧ Pwl
Under saturated oil
fWt
\ \
Q0
Saturated oil
a5I -
Sf i
In. I Illll Illll q0
Saturated oil
Figure 6. Depletion deterioration of IPR.
Saturated oil reservoirs:
WS2
S1 > S2
fWf
V% VX
q0
Under saturated oil
^^
S^S2
W Pwf
Saturated oil
Figure 7. Stimulation effect on IPR.
c_l
(Jcr0Zli0B0)f
(pR)f
Cr, ( W - M o ) p (PR)J
or
Assuming that n does not change with depletion, the new IPR equation at pR = 3520 psia is
q0 = 310[1 - (pwf/3620)2]0-8
fyo max ^f _ (kro/ILL0B0)f (<7o max )p (Icr0ZfI0B0)
(pR)f (pR) p
where nf = n =n.
Example: The present state IPR of an oil well in a solution gas drive reservoir is
Чо^Чо max = - (Pwf/PR)2]"
where ^omax = 480 STB/day pR = 4000 psia и = 0.8
Additional reservoir data are as follows:
PR (PSia)
4000 3520
/сГ0/р0Б0 (cp-')
0.223 0.147
EFFECT OF STIMULATION AND DAMAGE ON IPR EQUATIONS
Stimulating a well and reducing its skin factor improve the IPR of a well (Figure 7). Using the subscript 1 to denote a prestimulation or preservice state and the subscript 2 to denote a new post-treatment state, we see the characteristic parameters in IPR equation change with changing skin factor as follows:
Undersaturated oil reservoirs:
J2 _ [ l n ( r e / r w ) - 0 . 5 + s]] J1 [ln(re/rw)-0.5 + s]2
Saturated oil and gas reservoirs:
2/1-1 C2 Г [ln(re / r w ) - 0.5 + Sj1 C1 | [ l n ( r e / r w ) - 0 . 5 + s].
At a future stage w h e r e pR = 3520 psia, the m a x i m u m rate will be
(0.147)(3520) 0.8 = 310 STB/day
(0.223)(4000)
2и-1
(go max >2 = [ N r e / r W )-0-5 +sIi
where n, = iu = n.
Reserves Estimation
Rick J Sustakoskl
Chevron Overseas Petroleum Inc. San Ramon, California, U.S.A.
Diana Morton-Thompson
Consultant Kalamazoo, Michigan, U.S.A.
INTRODUCTION
________
To better understand reserves estimation, a few important t e r m s r e q u i r e definition. Original oil in place (OOIP) a n d original gas in place (OGIP) r e f e r to t h e t o t a l v o l u m e of hydrocarbon stored in a reservoir prior to production. Reserves or recoverable reserves are the v o l u m e of hydrocarbons that can be profitably extracted from a reservoir using existing technology. Resources are reserves p l u s all other hydrocarbons that may eventually become producible; this includes known oil and gas deposits present that cannot be technologically or economically recovered (OOIP and OGIP) as well as other undiscovered potential reserves.
Estimating hydrocarbon reserves is a complex process that involves integrating geological and engineering data. Depending on the a m o u n t and quality of data available, one or more of the following methods m a y be used to estimate reserves:
• Volumetric • Material balance • Production history • Analogy
These methods are summarized in Table 1.
VOLUMETRIC ESTIMATION
Volumetric estimates of OOIP and OGIP are based on a geological model that geometrically describes the volume of hydrocarbons in the reservoir. However, due mainly to gas evolving from the oil as pressure and temperature are decreased, oil at the surface occupies less space than it does in the subsurface. Conversely, gas at the surface occupies more space than it does in the subsurface because of expansion. This necessitates correcting subsurface volumes to standard units of volume measured at surface conditions.
One basic volumetric equation is
N = 7758Акф(1 - Sw)/Boi
where N = OOIP (STB)
7758 = conversion factor from acre-ft to bbl A = area of reservoir (acres) f r o m m a p data h = height or thickness of pay zone (ft) from log a n d / o r core data ф = porosity (decimal) from log and/or core data Sw = connate water saturation (decimal) from log and/or core data Boi = formation v o l u m e factor for oil at initial conditions
Table 1. Summary of Methods Used to Derive Hydrocarbon Reserves
Method Volumetric
Application
OOIP, OGIP, recoverable reserves. Use early in life of field.
Material balance
Production history Analogy
OOIP, OGIP (assumes adequate production history available), recoverable reserves (assumes OOIP and OGIP known). Use in a mature field with abundant geological, petrophysical, and engineering data.
Recoverable reserves. Use after a moderate amount of production data is available.
OOIP, OGIP, recoverable reserves, Use early in exploration and initial field development.
Accuracy Dependent on quality of reservoir description. Reserves estimates often high because this method does not consider problems of reservoir heterogeneity. Highly dependent on quality of reservoir description and amount of production data available. Reserve estimates variable.
Dependent on amount of production history available. Reserve estimates tend to be realistic. Highly dependent on similarity of reservoir characteristics. Reserve estimates are often very general.
513
506
PART 10—RESERVOIR ENGINEERING METHODS
Table 2. Estimation of Primary Recovery Factor
Drive Mechanism
Depletion Solution gas Expansion
Gas cap drive Water drive
Bottom Edge Gravity
Primary Recovery Factor (%)
18-25 2-5
20-40
20-40 35-60 50-70
(reservoir bbl/STB) from lab data; a quick estimate is Boi = 1.05 + (N x 0.05), w h e r e N is the n u m b e r of h u n d r e d s of ft3 of gas produced per bbl of oil [for example, in a well with a GOR of 1000, Boi = 1.05 + (10x0.05)]
Another basic volumetric equation is
G = 43560Л/г ¢(1 - S w ) / Bgi
where G = OGIP (SCF)
43560 = conversion factor from acre-ft to ft3 Bgi = formation volume factor for gas at initial conditions (RES ft3/SCF)
Recoverable reserves are a fraction of the OOIP or OGIP and are d e p e n d e n t on the efficiency of the reservoir drive mechanism. The basic equation used to calculate recoverable oil reserves is
Recoverable oil reserves (STB) = OOIP x RF
where RF = recovery factor, which equals RFp + RFs
The primary recovery factor, RFp, is estimated from the type of drive mechanism (Table 2). The secondary recovery factor, RFs, equals
RFs = E 0 x E A X EV
where E d = displacement efficiency E a = areal sweep efficiency Ev = vertical sweep efficiency
These efficiency terms are influenced by such factors as residual oil saturation, relative permeability, reservoir heterogeneity, and operational limitations that govern reservoir production and management. Thus, it is difficult to calculate the recovery factor directly using these terms, and other methods, such as decline curves, are often applied.
The basic equation to calculate recoverable gas reserves is
Recoverable gas reserves (SCF) = OGIP x RF
In this case, the recovery factor (RF) is typically higher than for oil reservoirs; it is often near unity for dry gas reservoirs.
MATERIAL BALANCE ESTIMATION FOR OIL
The material balance technique mathematically models the reservoir as a tank. This method uses limiting assumptions and attempts to equilibrate changes in reservoir volume as a result of production. Aquifer support and gas cap expansion can be accounted for by using this method.
One general equation is
Change in pore volume = Change in oil volume + change in free gas volume + change in water volume
where C h a n g e in pore v o l u m e = NBoi/(1 - SwJcfP C h a n g e in oil v o l u m e = NBoi -(N - Np)Boi Change in gas volume = (GBgi - GBg) + [NpRp (N - Np) - NRJBs due to gas produced, evolved, and encroached from a gas cap Change in water volume
= -NBoiSwi/(1 - SwJcwP - We + WpBw due to connate water volume change, encroached water, and produced water where Bg = formation v o l u m e factor of free gas Bgi = formation volume factor of free gas at initial conditions cf = formation (rock) compressibility (psi-1) cw = water compressibility (psi-1) N = OOIP(STB) Np = cumulative oil produced (STB); from production liistory data P = Change in reservoir pressure due to production, that is, initial pressure minus current pressure; taken from field pressure surveys Rp = cumulative gas-oil ratio, or total produced gas (in SCF)/total produced oil (in STB); from production history data Rsi = inital solution gas-oil ratio (SCF/STB) Swi = initial connate water saturation (decimal) Wc = cumulative a m o u n t of water encroachment; f r o m map and field data Wp = cumulative water produced; from production history data
Another general equation is
Np[Bt +(Rp-Rsi)BgJ-(We-WpBw)
N =
(П R N I MBtj (BT-BTI) + - ^ - (
B(TG>-
B
R
GI
)
ч , +
Bti(CwSwi+ J-^R-
Cf )P
where Bt = total (two-phase) formation volume factor Bti = total formation volume factor at initial conditions M = gas cap size expressed as a fraction of initial reservoir oil volume; from map data
This equation assumes thermodynamic equilibrium between oil and gas, a uniform pressure distribution, and a uniform saturation distribution in the reservoir. Additional equations
Reserves Estimation
515
can be derived from the general material balance equation for specific reservoir types.
A simplified equation can be used for a quick estimate of initial oil in place. This equation assumes a closed reservoir system (no active water drive), no initial gas cap, and initial reservoir pressure close to the bubblepoint:
N BobVt +(R-Rs)B N =—
Bob(Vt-Vi)
/5.61
where 5.61 = conversion factor from volume/volume to ft3/bbl Bob = formation volume factor for oil at the bubblepoint; determined for specific separator conditions R = gas-oil ratio, or GOR, equal to produced gas (in SCF)/ produced oil (in STB); from production history data Rs = solution gas-oil ratio (SCF/STB) or gas solubility in oil Vi = initial v o l u m e of oil plus liberated gas as a function of pressure measured at reservoir temperature V1 = volume of oil plus liberated gas as a function of pressure measured at reservoir temperature; determined u n d e r flash liberation conditions
This equation can also be used to predict Np (how much a reservoir can produce, or recoverable reserves) assuming N is d e t e r m i n e d b y a n i n d e p e n d e n t m e t h o d a n d R, the gas-oil ratio, can be controlled throughout the life of the field.
MATERIAL BALANCE ESTIMATION FOR GAS
The material balance technique for calculating gas reserves, like material balance for oil, attempts to mathematically equilibrate changes in reservoir volume as a result of production. The basic equation is
Weight (or SCF) of gas produced = weight (or SCF) of gas initially in the reservoir - weight (or SCF) of gas remaining in the reservoir
The equations used to calculate OGIP are
Gas reservoir with active water drive:
GpBg-(We-WpBw)
G —
Gas reservoir with no water drive (We = 0):
These equations can also be used to predict Gp (recoverable reserves) assuming G is determined by an independent method and the production conditions remain constant.
Reservoir Simulation
The material balance m e t h o d is actually a subset of the mathematical techniques that are available to modern petroleum engineers. Reservoir simulators use material balance as well as fluid flow equations to model the reservoir as a g r o u p of interconnected tanks. The advent of powerful computers has m a d e the use of numerical simulation quite common for estimating reserves and recovery as well as initial volume in place. Since reservoir simulation can account for performance history through history matching, this method i n c o r p o r a t e s facets of all the t e c h n i q u e s discussed. With sufficient data and p r u d e n t use of simulators, this m e t h o d provides the best recovery estimates for complex reservoirs.
PRODUCTION HISTORY ANALYSIS
Production history analysis is used to estimate economic ultimate recovery (or recoverable reserves) and the expected economic life of a reservoir. The rate of p r o d u c t i o n and cumulative production at any point in time can also be estimated. This method relies on historical production data to extrapolate f u t u r e p r o d u c t i o n performance. A variety of curves can be used (Figure 1), the most c o m m o n being a semilog plot of rate of production versus time (Figure 2). These data are easily obtained through operator records or state regulatory agencies.
Three mathematical models can be used to describe decline curve (usually rate versus time) behavior. They are
• Exponential decline • Hyperbolic decline • Harmonic decline
Exponential and hyperbolic decline are commonly used to describe reservoirs. Harmonic decline is an infrequently applied special case of hyperbolic decline.
The different types of decline behavior are not necessarily mutually exclusive. Often different decline curve characteristics are related to different stages of reservoir development, and the overall trends can be significantly affected by workovers or stimulation, infill drilling, a change in lift mechanics, or secondary or tertiary flood initiation (Figure 3).
F o r m u l a s u s e d to calculate the rate of p r o d u c t i o n , cumulative production, and economic life of a reservoir are given in Table 3.
GpBg+(WpBw) G =
Bn-Bl &
where G = OGIP (SCF)
Gp = cumulative gas produced (SCF)
ANALOGY METHOD
The analogy method for estimating reserves directly compares a newly discovered or poorly defined reservoir to a known reservoir thought to have similar geological or petrophysical properties (depth, lithology, porosity, and so on). While analogy is the least accurate of the m e t h o d s
506
PART 10—RESERVOIR ENGINEERING METHODS
100%
TIME
RATE VS. TIIVlE
wP
RATE VS. CUM. PRODUCTION
OIL-CUT VS. CUM. PRODUCTION
TOP OF SAND
o/w
CONTACT
NP
OIL WATER CONTACT VS. CUM. PRODUCTION
о
N
CUM. GAS VS. CUM. PRODUCTION
о
TIME
RESERVOIR PRESSURE VS. TIME
Figure 1. Production history curves. (From IHRDC, 1982.)
LOG q
ECONOMIC LIMIT
EXPONENTIAL HYPERBOLIC
TIME
t, ь
presented, it is often used early in the life of a reservoir to establish an order-of-magnitude recovery estimate. As the field matures and data become available to make volumetric OOIP or OGIP estimates, analogy is often used to establish a range of recovery factors to a p p l y to the in-place volumes. Evaluating recovery in this fashion is particularly useful when some performance history is available but a decline rate has yet to be established. Analogy should always be used in conjunction with other techniques to ensure that the results of the more computationally intensive methods make sense within the geological framework.
Figure 2. Semi-log plot of rate of production versus time. (From IHRDC, 1982.)
Reserves Estimation
517
CUMULATIVE OIL (IVlM BBLS)
Figure 3. Relationship of decline behavior to decline curve characteristics. (From Carr and Viret, 1986.)
Table 3. Decline Equations
Solving for
Rate of production Cumulative production Life of reservoir
where q, = Rate of production at time t (BOPD). Ql = Rate of initial production (BOPD).
Qec = Economic limit rate of production (BOPD). D = Decline rate (decimal).
Exponential
q,=
Np =
(q-q,)/D
t=(MD)\n(q/qec)
Dl = Initial decline rate (decimal). t =Time (years). n = Exponent usually between 0 and 0.7
Np = Cumulative production (STBO).
Hyperbolic
qt=q(1
+nDfrvn
Np = q ^ - r i ) D M - n - q r ) t=[(q/qec)n-MnDl
Drive Mechanisms and Recovery
Stephen R. Sills
ARCO Exploration and Production Technology Piano, Texas, U.S.A.
INTRODUCTION
The natural energy of a reservoir can be used to move oil and gas toward the wellbore. Used in such a fashion, these s o u r c e s of e n e r g y a r e c a l l e d drive mechanisms. E a r l y determination and characterization of the drive mechanism(s) present within a reservoir may allow a greater ultimate recovery of hydrocarbons. Drive mechanisms are determined by the analysis of historical p r o d u c t i o n data, primarily reservoir pressure data and fluid production ratios.
The three primary oil reservoir drive mechanisms are solution gas drive, gas cap drive, a n d water drive (Clark, 1969). Reservoir pressure trends and producing gas-oil ratio trends of these three drive mechanisms are s h o w n in Figures 1 and 2, respectively. A combination or mixed drive occurs w h e n two or m o r e of the p r i m a r y d r i v e m e c h a n i s m s are present in the same reservoir. A combination drive may also occur when one or more of the primary drive mechanisms are assisted by gravity drainage. T a b l e 1 s h o w s t h e e n e r g y s o u r c e s a n d ultimate recovery ranges of the major drive mechanisms.
SOLUTION GAS DRIVE
In a solution (or dissolved) gas drive reservoir, the oilbearing rock is completely surrounded by impermeable barriers. As the reservoir pressure drops during production, expansion of the oil and its dissolved gas provides most of the reservoir's drive energy (Figure 3). Additional energy is obtained f r o m the expansion of the rock and its associated water.
Depending on its discovery pressure, a solution gas drive reservoir can be initially either undersaturated or saturated (Odeh, 1986). In an undersaturated reservoir, the reservoir pressure is greater than the bubblepoint of the oil. N o free gas exists in the reservoir while the pressure remains above the bubblepoint. The reservoir drive energy is provided only by the limited expansion of the oil, rock, a n d water. In a saturated reservoir, the reservoir pressure is at the bubblepoint. As soon as oil is produced, the pressure drops a n d b u b b l e s of solution gas f o r m in the reservoir. This solution gas liberation causes the oil to shrink, but the oil shrinkage is more than offset by solution gas expansion, the p r i m a r y s o u r c e of r e s e r v o i r d r i v e e n e r g y b e l o w the bubblepoint.
Production Trends
Solution gas drive reservoirs show characteristic changes in reservoir pressure, producing gas-oil ratio, and oil and water production rates during the life of the reservoir. If the reservoir is initially undersaturated, the reservoir pressure falls quickly d u r i n g oil p r o d u c t i o n because of the small compressibilities of oil, water, and rock. Pressure d r o p s of several hundred pounds per square inch can easily occur over a matter of months. Because the only gas p r o d u c e d is that which evolves from the produced oil in the wellbore, the gas-oil ratio (GOR) remains constant until the reservoir reaches the bubblepoint.
Once reservoir pressure reaches the bubblepoint pressure or if t h e r e s e r v o i r w a s initially s a t u r a t e d , t h e r e s e r v o i r pressure declines less quickly due to the large compressibility
cс a .Я>
Water Drive
б
о
•
\ Gas Cap Drive
> о
(QCЛD) CC
Solution \ Gas Drive
100 Solution
Gas Drive
_Q
80
CO
</)
E
60
I
CC
O
O
o>
40
с
о "3OO
20
Gas Cap Drive
Water Drive
20
40
60
80
Oil Produced - % of OOIP
20
40
60
80
Oil Produced - % of OOIP
Figure 1. Reservoir pressure trends by drive mechanism.
Figure 2. Producing gas-oil ratio trends by drive mechanism.
518
Table 1. Ultimate Recovery Ranges by Drive Mechanism
Drive Mechanism Solution gas drive
Gas cap drive
Water drive Gravity drainage
Energy Source
Recovery (%OOIP)
Evolved solution gas expansion
5-30
Gas cap and evolved solution gas expansion
20-40
Aquifer expansion
35-75
Gravity
5-30 (incremental)
Drive Mechanisms and Recovery
519
of the gas bubbles forming in the reservoir. The producing GOR rises quickly as the bubbles link u p and begin to flow and can increase to as much as ten times the initial GOR. If reservoir pressure continues to fall, the producing GOR will eventually drop as the gas expands less and less as it flows u p the wellbore.
Oil production rates fall quickly once the producing GOR begins to rise. Wells must be placed on artificial lift early in their life (see the chapter on "Artificial Lift" in Part 9). Initially, little or no water is produced. As reservoir pressure d r o p s , a small a m o u n t of water m a y be p r o d u c e d as the interstitial water saturation expands and exceeds the critical value required for flow.
Recovery
Oil recovery from solution gas drive reservoirs is usually low, ranging from 5 to 30% of the original oil in place (OOIP) (see Table 1). Typically less than 5% of the OOIP is recovered above the bubblepoint. In general, the better solution gas drive recoveries are obtained in reservoirs with relatively low oil viscosities and fairly homogeneous rock properties. Recovery can sometimes be improved with completion strategies that conserve reservoir energy by minimizing the producing GOR.
GAS CAP DRIVE
In a gas cap d r i v e reservoir, the p r i m a r y source of reservoir energy is an initial gas cap, which expands as the reservoir pressure drops (Figure 4). Additional energy is provided by the expansion of solution gas released from the oil. Less significant drive contributions are provided by the expansion of the rock and its associated water.
Production Trends
Gas cap expansion causes reservoir pressure to fall more slowly in a gas cap drive reservoir than in one producing in a solution gas drive. The rate of pressure decline is closely tied to the relative size of the gas cap, w i t h larger gas caps resulting in a more gradual pressure decline as oil is produced.
Early in the life of a gas cap drive reservoir, the GOR rises slowly because the higher reservoir pressure keeps more gas in solution in the oil. Later, the GOR increases dramatically as
Figure 3. Solution gas drive reservoir.
the expanding gas cap reaches the highest wells on structure. The GOR continues to climb as the gas-oil contact moves farther down structure and gas cap gas production increases. Oil production rates fall less quickly than in a solution gas drive reservoir due to the slower decline in reservoir pressure. Artificial lift may not be required as early in the field's life since wells tend to have longer flowing lives. As in a solution gas drive, little or no water is produced.
Recovery
Oil recovery from gas cap drive reservoirs typically ranges from 20 to 40% of the original oil in place. The actual recovery obtained d e p e n d s on the size of the initial gas cap, the structural geometry of the reservoir, and the way the field is managed.
Gas cap drive recovery increases with the size of the initial gas cap if gas cap gas production can be minimized. This is done most easily in steeply dipping reservoirs or those with thick oil columns which allow the wells to be perforated as far as possible below the gas-oil contact. Recovery can also be improved by shutting-in wells when they begin to produce large amounts of gas cap gas. In addition, the produced gas may be returned to the gas cap using gas injection wells located high on structure.
506
PART 10—RESERVOIR ENGINEERING METHODS
Figure 4. Gas cap drive reservoir.
WATER DRIVE
In a water drive reservoir, the oil zone is in communication with an aquifer that provides the bulk of the reservoir's drive energy. As oil is produced, the water in the aquifer expands and moves into the reservoir, displacing oil. Depending on the aquifer's strength, additional energy may be provided by solution gas expansion. Much less significant contributions are p r o v i d e d by the expansion of the reservoir rock and its associated water.
The geometry of the aquifer determines whether it is a bottom water or an edge water drive (Figure 5). In a b o t t o m water drive, the aquifer is present below the entire reservoir and water influx moves vertically upward into the oil zone. In an edge water drive, the aquifer is located on the flanks of the reservoir and the water moves upward along the reservoir dip.
Production Trends
In a water drive, the reservoir pressure response to p r o d u c t i o n d e p e n d s on the size and permeability of the aquifer and the rate at which the reservoir is produced (Dake, 1978). If the reservoir is produced at a low rate, the aquifer is able to replace the fluid volumes produced and reservoir pressure remains fairly constant. At high production rates, the aquifer is unable to keep u p with withdrawals and reservoir pressure drops. If the rate is then reduced to a low level, reservoir pressure will rise. The m a g n i t u d e of "high" and "low" production rates for a particular water drive reservoir are determined by the size and permeability of its associated aquifer.
Edge Water Drive Figure 5. Edge water versus bottom water drive reservoirs.
Bottom Water Drive
Drive Mechanisms and Recovery
521
In a strong water drive reservoir, the producing GOR remains fairly constant, reflecting the stable reservoir p r e s s u r e . H o w e v e r , if the a q u i f e r is u n a b l e to m a i n t a i n reservoir pressure, the producing GOR will rise accordingly.
Oil rates remain high under strong water drive until water breaks through in the producing wells. Water production usually occurs early in the field life of d o w n structure wells, and the water-oil ratio (WOR) continues to increase with time as the oil-water contact moves upward. Gas lift may be required for high water cut wells to continue to flow.
Recovery
Oil recovery from water drive reservoirs typically ranges from 35 to 75% of the original oil in place. The actual recovery obtained d e p e n d s on the strength of the aquifer, the sweep efficiency of the encroaching water, and the w a y the field is managed.
Water drive recovery increases with the strength of the aquifer if water production can be minimized. As with gas cap drive reservoirs, this is done most easily in reservoir geometries that allow wells to be perforated a considerable distance from the fluid contact.
Water drive recovery also depends on the aquifer's sweep efficiency. Sweep efficiency is a measure of h o w effectively the encroaching water displaces oil. Higher sweep efficiencies and oil recoveries occur w h e n the viscosity of the oil is low compared to that of the water and oil flows m o r e easily than the encroaching water. Water drive reservoirs with high viscosity crudes have lower sweep efficiencies and oil recoveries b e c a u s e the w a t e r t e n d s to m o v e a h e a d of or "finger" through the oil, leaving behind unswept oil.
Water drive recovery can be improved by balancing production rates across the field so that the oil-water contact moves up as uniformly as possible. Since water drive is usually more efficient than solution gas drive, in some cases it is possible to increase recovery by producing the reservoir at a rate low enough that the aquifer is able to maintain a high reservoir pressure.
COMBINATION DRIVE
Most oil reservoirs produce u n d e r the influence of two or more reservoir drive mechanisms, referred to collectively as a combination drive. A common example is an oil reservoir with an initial gas cap and an active water drive (Figure 6).
Production Trends
The production trends of a combination drive reservoir reflect the characteristics of the d o m i n a n t drive mechanism. A reservoir with a small initial gas cap and a weak water drive will behave in a way similar to a solution gas drive reservoir, with rapidly decreasing reservoir pressure and rising GORs. Likewise, a reservoir with a large gas cap and a strong water drive may show very little decline in reservoir pressure while exhibiting steadily increasing GORs and WORs. Evaluation of these production trends is the primary method a reservoir engineer has for determining the drive mechanisms active in a reservoir.
Aquifer •:
Map View
Figure 6. Combination drive reservoir.
Recovery
The ultimate recovery obtained from a combination drive reservoir is a function of the drive mechanisms active in the reservoir. The recovery may be high or low depending on whether displacement or depletion drive mechanisms dominate. Water drive and gas cap expansion are both displacement type drive mechanisms and have relatively high recoveries. Solution gas drive is a depletion type drive and is relatively inefficient.
Recovery from a combination drive reservoir can often be i m p r o v e d by m i n i m i z i n g t h e effect of d e p l e t i o n d r i v e mechanisms by substituting or augmenting more efficient ones through production rate m a n a g e m e n t or fluid injection. To do this, the drive mechanisms active in a reservoir must be identified early in its life.
GRAVITY DRAINAGE
Gravity drainage, or gravity segregation, is the tendency of oil, gas, and water to segregate in a reservoir during production d u e to their differing densities (Figure 7). As a
506
PART 10—RESERVOIR ENGINEERING METHODS
secondary drive mechanism, gravity drainage occurs only in combination with one or more of the primary oil reservoir drive mechanisms.
Conditions conducive to gravity drainage include thick reservoirs with high vertical permeabilities or thin reservoirs with steep dips. In a solution gas drive reservoir perforated down dip, gravity drainage can cause released solution gas to migrate upward and oil to flow downward, conserving reservoir energy and increasing recovery to near that of a water drive.
The rate of oil gravity drainage in the reservoir is usually low compared to field production rates. Over time, however, gravity drainage can be extremely efficient and recoveries higher t h a n a n y of the p r i m a r y d r i v e m e c h a n i s m s are possible.
Hff
йШШ
v//////
- 4с
v////
' ; '
r>>>-
- г' '
El
I шшшЯщ
тшт
JH
Figure 7. Fluid segregation by gravity damaged' 1 / C
Waterflooding
A. M. Sam Sarem
Unocal Science and Technology Brea, California, U.S.A.
INTRODUCTION
Waterflooding is a process used to inject water into an oilbearing reservoir for pressure maintenance as well as for displacing and producing incremental oil after (or sometimes before) the economic production limit has been reached. Tliis is d o n e through the displacement of oil and free gas by water. In waterflooding, water is injected into one or more injection wells while the oil is produced from surrounding producing wells spaced according to the desired patterns. There are many different waterflood patterns used in the industry, the most c o m m o n of which are illustrated in Figure 1.
QUICK ESTIMATION OF WATERFLOOD RECOVERY
The overall recovery (Er) is a product of displacement efficiency (Ed), invasion or vertical sweep efficiency (EV) and the pattern or areal sweep efficiency (EP).
E J ^ D ^ EV ^ EP
where E r = overall recovery (fraction of initial oil in place recovered) E d = displacement efficiency or volume of oil displaced divided by total oil volume (fraction) Ev = vertical or invasion efficiency (fraction of vertical reservoir section contacted by injection fluid) Ep = pattern efficiency or pattern swept by total pattern area
The displacement efficiency is a function of residual oil saturation (Sor) of the swept region. The following equation gives the displacement efficiency as a function of Sor a n d the interstitial (irreducible) water saturation (Swi):
E0=a -sor-sj/a - s j
where Sor = saturation of residual oil (fraction) Swi = saturation of irreducible water (fraction)
The displacement efficiency d e p e n d s u p o n the ratio of the viscous to capillary forces or capillary number. In enhanced recovery processes, the interfacial tension between the oil and water is reduced to improve the capillary number (Willhite, 1986).
The vertical sweep efficiency is a function of the vertical heterogeneity (layering) and the mobility ratio (M). The mobility ratio d e f i n e d here is the ratio of the relative permeability to water at Sor (Icrw) to the relative permeability of the oil at Swi multiplied by the oil-water viscosity ratio
(jljfiw) as expressed in the following equation:
M = (IcrwZkro) x (/J0Zliw)
where M = mobility ratio, or the ratio of relative permeability to water at Sor divided by the relative permeability of oil at Swi multiplied by the oil-water viscosity ratio Zcrw = relative permeability of w a t e r kro = relative permeability of oil Hq = viscosity of oil /Jw = viscosity of water
The vertical heterogeneity is most commonly described by the Dykstra-Parsons permeability variation (Vdp), defined as the ratio of the standard deviation of the permeability of various layers (a) divided by the mean permeability ( k ) , as given in the following equation:
where V d p = Dykstra-Parsons permeability variation к = absolute permeability к = mean permeability or s u m of the permeabilities divided by the n u m b e r of permeabilities
In practice, the permeability variation is determined by arranging the permeabilities in descending order and determining the percent-greater-than values for each permeability. From a plot of к versus percent greater than on a log probability graph sheet (as shown in Figure 2), the values of к at 50% a n d к at 84.1% are read, a n d Vdp is determined as follows:
^DP = ^50 - ^84.1 )/^50
Dykstra and Parsons (1950) have published charts for determining the vertical sweep efficiency (V0) or conformance from the mobility ratio and the permeability variation, as shown for WOR = 5 and 0.1 in Figures 3 and 4, respectively.
The pattern efficiency is a function of the previously defined mobility ratio, the flood pattern, a n d given water cut. Figure 5 shows the pattern efficiency as a function of mobility ratio and water cut for a five-spot pattern. Craig (1971) shows similar plots for several other common patterns.
A w o r d of c a u t i o n is a p p r o p r i a t e at this point. The k n o w l e d g e of directional p e r m e a b i l i t y can be crucial in placement of pattern injectors. The injectors should always line u p in the direction of m a x i m u m permeability. As shown in Figure 6, infill drilling can rotate the flow pattern and in some cases it can reduce the areal sweep efficiency, as
523
506 PART 10—RESERVOIR ENGINEERING METHODS
1
A
»
V
»
\
/
TWO-SPOT
/
N
THREE - SPOT
A INJECTION WELL О PRODUCTION WELL
PATTERN BOUNDARY
O IO Д ^ I^
I
O IO
1
I
^ v
• I
O IO I
•Ч I X
A " I
vV
I
I O I I /
O •O
REGULAR FOUR-SPOT
О Д O / О ^Д O /
O I о ^д O / O ^д
О / о >Ko / O
\
/
/ \
/
N/ /
4 /-
о ^Д" о / о
о
/ \
/
/ N
> \
/ N
O 7 O ЧД" о / о чд
SKEWED FOUR-SPOT
д OД о д о
\
/^ \\
OV OX о>
•
\
S
/
< ° >; о л о
о /X \ \ \
/ /
O/X \ %N
• /
O/\ \>
•
4 /
S
•
д' O ^ O ^ O
FIVE- SPOT
д
д
о
д—д
/
\
/
/
\
/
о
4A-^
о
\
/
\
д—tL
о
чд—д
'
\
/
/
\
/
O ^
/
\
/
Д' O
\ \
чд—d
о
чд—д
SEVEN-SPOT
A OA
I
I
I
I
A- -Д- -Д-
l
I
I
I
Д OA
I
I
I
I
I
A- - A - - Д -
-Д
I I
I I
I
Д
Д
A
NORMAL NINE-SPOT
0 ДО
1
I
I
I
O- - -O O
I I 0 1
I I о I
I
I
0 O--O
1
I
I
I
ОА о
Д O-д O O
I I А о
INVERTED NINE - SPOT
O O O
/
O O
O O r / O Д
\
Ъ—O
O
O
о—о
д
O O
INVERTED SEVEN-SPOT
?—9—9—9—9
! j
• *I I*
I
* I
I
I
t *I *I
Iiiii
AA^ A
Iltil
I IIII
I
I
I
I
I
Illl
< j > - - - f - - o — <ь
! I: I IIII Illlt ДI lA l A l Д l Д
DIRECT LINE DRIVE
O-4-0-+-0-+-O-*-o
--A , I I I I
Д A-
O--f-O-T- o - r - o - f - o I
I I I
I
—A
A
A
A— —
STAGGERED LINE DRIVE
Figure 1. Flooding patterns. (From Craig, 1971.)
Waterflooding
525
0
0.2
0.4
0.6
0.8
1.0
WOR = 5 BBL/BBL
Figure 2. Permeability variation example problem. (From Craig, 1971.)
Figure 3. Vertical sweep efficiency (coverage) as a function of WOR, M, and permeability variation (Vdp), where WOR = 5. (From Dystra and Parson, 1950.)
exemplified in this figure for a five-spot and line-drive case. As shown, the sensitivity of the sweep efficiency to the ratio of maximum to minimum areal permeability depends heavily upon whether the maximum permeability is in the direction of injection to injection wells or injection to production wells.
Example of Estimation of Waterflooding Recovery:
Estimate the overall recovery of a five-spot waterflood for the following cases:
Case
l/n
M
0.7
3
0.7
1.5
0.5
1.5
0.7
3
0.5
3
Other properties are as follows:
Kw AT SOR = 0 . 2 km at SWI = 0 . 9 5 Ii0 = 10 cp at reservoir temperature /IW = 0.7 cp at reservoir temperature
SOR = 0 . 3 0 SWI = 0.28
Solution: E R = E ^ X EY x EP ED = ( I - S W I - S O R ) Z ( I - S W I ) = (1 - 0.28 - 0.3) / (1 - 0.28) = 0.583 M = (KJkJ x (k,/AO = (10)(0.2)/(0.7)(0.95) = 3.0
WOR
0.1 0.1 0.1
5 5
EV = 0.06 for M = 3, VDP = 0.7, WOR = 0.1 (from Figure 5) E P = 0 . 5 7 for M = 3 , 1 /M = 0 . 3 3 ,
WOR = 0.1(/W = 0.1/(1 + 0.1) = 0.09) (from Figure 6) ER = 0.583 x 0.06 x 0.57 = 0.02
The rest of the solution is given in the following table:
Ciie
0.7 0.7 0.5 0.7 0.5
M WOR K
3
0.1 0.583
1.5 0.1 0.583
1.5 0.1 0.583
3
5
0.583
3
5
0.583
K1
0.06
0.09 0.27 0.5 0.185
K
0.57 0.61
0.61 0.835
0.57
K
0.02 0.03
0.10 0.24
0.06
MATHEMATICAL MODELING OF WATERFLOODS
The f o r e g o i n g a n a l y s i s is o n e of m a n y s i m p l i f i e d approaches that can be taken as a first approximation for the prediction of waterflood recovery, as fully explained by Craig (1971) and Bradley (1987). For a more accurate recovery p r e d i c t i o n , m a t h e m a t i c a l m o d e l i n g of the reservoir is essential. However, the most important aspect of reservoir m o d e l i n g is the construction of the model, which requires detailed k n o w l e d g e of the reservoir characteristics. (For information on geostatistical methods used to determine the most probable realization of the reservoir structure, see the statistics chapters in Part 6.) Once the most probable reservoir realization is d e t e r m i n e d , the history matching of p r i m a r y production can be used to refine the model before it is used to predict the waterflood behavior. For this purpose, any of the many commercially available softwares can be used.
506 PART 10—RESERVOIR ENGINEERING METHODS
CL
Ш H CL Ш сn UJ 70 o< c
0.2
0.4
0.6
0.8
WOR = 0.1 BBL/BBL
0.2
0.4 0.6 0.81.0
2.0
4.0 6.08.010
RECIPROCAL OF MOBILITY RATIO, 1/M
1.0
Figure 5. Effect of mobility ratio on oil production for the five-
spot pattern. (From Craig, 1971.)
Figure 4. Vertical sweep efficiency (coverage) as a function of WOR, M, and permeability variation (Vdp), where WOR = 0.1. (From Dystra and Parson, 1950)
LINE DRIVE
5 SPOT
9-
-9
•
i
LINE DRIVE I—-O 1
•—-O-SIJ
Figure 6. Effect of directional permeability on sweep efficiencies for varied degrees of permeability anisotropes at a fluid mobility ratio of 1. (From Landrum and Crawford, 1960.)
Enhanced Oil Recovery
Vernon S. Breit
SIMTECH Consulting Services Inc. Golden, Colorado, U.S.A.
INTRODUCTION
Conventional recovery methods (primary and secondary) typically extract approximately one-third of the original oilin-place in a reservoir. Estimates (made in the late 1970s) of worldwide oil in-place range u p to 1.5 trillion barrels; using that figure, it is estimated that the oil remaining as a residual oil saturation after conventional recovery would be approximately 1.0 trillion barrels (National Petroleum Council, 1976). Several enhanced oil recovery (EOR) techniques—generally grouped together as tertiary production schemes—have targeted this huge unexploited reserve.
In the past, chemical, thermal, and miscible techniques have been used by the industry on a commercial scale. EOR t e c h n i q u e s r e q u i r e the injection of chemical c o m p o u n d s dissolved in water, the injection of steam, or the injection of a gas that is miscible with the oil in place. As a result, all current EOR techniques are much more expensive to implement than normal secondary water injection projects. Therefore, the a m o u n t of oil that can ultimately be recovered by existing EOR techniques is directly related to the price of crude oil.
All EOR projects begin w i t h an analysis of the n a t u r e , location, a n d causes of residual oil s a t u r a t i o n s (Sor) that remain after primary and/or secondary recovery operations. The m a i n factors that control the v a l u e of Sor are p o r e g e o m e t r y , rock w e t t a b i l i t y , a n d the p r o p e r t i e s of the displaced (oil) and displacing (injected) fluids. Fluid p r o p e r t i e s of particular interest are interfacial tension, viscosity, and density. In combination with the heterogeneity of the reservoir, these properties result in the overall recovery (Er) for any recovery scheme.
The overall recovery is the p r o d u c t of d i s p l a c e m e n t efficiency (Ed), a n d sweep efficiency (ES). The displacement efficiency is inversely proportional to the residual oil saturation, while the sweep efficiency is inversely proportional to the mobility ratio (M) between the injected fluids and the oil in place (see the chapter on "Waterflooding" in Part 10). M is usually stated in terms of the relative permeability of a fluid phase (kr) divided by the phase's viscosity (ju) relative to the same ratio for the other phase, such as for a waterflood:
M =
(krJnJ/(kr0//i0)
In simplest terms, the pursuit of incremental oil recovery for EOR reduces to selecting the process that minimizes M or Sor for a particular reservoir. The choice of process to be used is also dependent on other considerations such as depth, temperature, and a m o u n t of oil remaining in place prior to EOR. Unfortunately, the implementation of all current EOR processes involve factors that can both increase and decrease
M. For example, a miscible gas flood decreases M by increasing the permeability of oil relative to the injected gas through the elimination of interfacial tension, but at the same time, the low gas viscosity relative to oil results in an increase in M.
Much of the design phase of an EOR project is spent in the search for the combination of processes and injection schemes that will m a x i m i z e oil recovery relative to the cost of implementing the particular process. Specifically, the phases in designing an EOR project include the following:
1. Basic screening criteria • Oil viscosity • Remaining oil in place • Geology • Formation water salinity • Formation depth and temperature • Estimate of potential benefit • Cost
2. Labtests • Ruid properties • Rock properties • Core floods • Chemistry of injection material
3. Volumetric calculations to determine cost benefit ratio 4. Detailed technical studies
• Geological • Reservoir simulation • Economic projections 5. Field trials
It is possible to skip m a n y of these steps by using reservoir engineering analogy when neighboring reservoirs that have undergone EOR have properties similar to the reservoir in question. In addition, field trials are often not attempted due to the length of time they r e q u i r e a n d the difficulty in selecting a p o r t i o n of a r e s e r v o i r to s t u d y that is representative of the reservoir as a whole.
CHEMICAL FLOODING
The basic p u r p o s e s of chemical flooding are to a d d a material (chemical) to the water being injected into a reservoir to increase the oil recovery by (1) increasing the water viscosity (polymer floods), (2) decreasing the relative permeability to water (cross-linked p o l y m e r floods), or (3) increasing the relative permeability to oil and decreasing Sor by decreasing the interfacial tension between the oil and water phases (micellar and alkaline floods). The process is depicted schematically in Figure 1. Chemical additives to reduce interfacial tension are detergent type compounds such as petroleum sulfinates and are so expensive that chemical
527
506
PART 10—RESERVOIR ENGINEERING METHODS
Mobility ratio is improved, and the flow of liquids through more permeable channels is reduced by the polymer solution resulting in increased volumetric sweep.
Injection Fluids
Production Well
Iniection Well
(Single 5-Spot Pattern Shown)
Driving Fluid (Water)
Fresh Water Buffer to Protect Polymer
Polymer Solution
For Mobility Control
Alkaline Solution Forms Surfactants In Situ For Releasing
Oil
Additional Oil
Recovery (Oil Bank)
Preflush to Condition
Reservior
Figure 1. Schematic diagram of chemical flooding (alkaline). (Courtesy of U.S. Department of Energy, Bartlesville, Oklahoma.)
floods are often technical successes and economic failures. Successful design of chemical floods always revolves around minimizing the a m o u n t of chemicals needed to achieve the desired change in interfacial tension and/or mobility ratio (Shah and Schechterx 1971).
This minimization is achieved by preceding the chemical injection with a preflush to buffer the chemicals from reactions with the in situ water a n d following the chemical injection with the injection of a polymer solution to maintain a f a v o r a b l e mobility ratio for the flood. O n e of the m a j o r problems with the injection of surfactants and other chemicals into reservoir rock is that the chemicals are surface active. Thus, they have a great affinity for the minerals found in reservoirs, c a u s i n g a d s o r p t i o n of chemicals f r o m solution onto the rock in great quantities.
MISCIBLE GAS FLOODING
The concept behind miscible flooding, such as carbon dioxide floods, is that the best w a y to eliminate the interfacial tension between the in-place oil and injected fluids is to inject a solvent for that oil (a solvent being a material that is miscible in all proportions with the dissolved material). The residual oil saturation to displacement by a solvent would be 0.0. In practice, a solvent must be found that is miscible with the oil and costs no more than the oil. Finally, the volumetric sweep efficiency of the process m u s t be high e n o u g h to m a k e this scheme economical. Because the relative permeability effects have been removed by the miscible nature of the displacment,
the sweep efficiency is primarily a function of the viscosity of the solvent relative to the viscosity of the oil (Stalkup, 1983).
Current applications of miscible flooding have concentrated on carbon dioxide (CO2), hydrocarbon gas, and nitrogen injection processes. The gas solvents tend to be much less viscous than reservoir oils so that the sweep efficiencies are often very low for miscible gas floods. Design efforts center around finding methods to improve this volumetric s w e e p efficiency. In the case of carbon dioxide floods, the gas is injected into the reservoir in small slugs that are alternated with water slugs as a m e a n s of lowering the mobility of the injected fluid (Figure 2).
Secondary benefits of miscible gas injection include the effects of the solubility of the solvent in the oil phase. As the carbon dioxide, hydrocarbon gas, or nitrogen dissolve in the oil phase, the oil is swelled and its viscosity lowered. Both of these phenomena add to the mobility of the oil relative to the injected solvent. In practice, although miscible displacement implies no residual oil saturation in the area swept by the solvent, a small residual saturation is left due to economic considerations of p r o d u c i n g at high GORs and the p h a s e behavior of the system prior to the attainment of miscibility in the reservoir.
THERMAL RECOVERY
All thermal recovery processes involve the use of heat to accelerate the oil recovery process. The heat can be generated at the surface and injected into the reservoir, as in the case of
Enhanced Oil Recovery
529
Figure 2. Schematic diagram ot carbon dioxide flooding. The viscosity of oil is reduced, providing more efficient miscible displacement. (Courtesy of U.S. Department of Energy, Bartlesville, Oklahoma.)
steam injection (Figure 3), or generated in the reservoir by injecting a fluid such as air that is combustible with the inplace oil (Figure 4). Tlie choice of which technique to use to a d d thermal energy to the reservoir d e p e n d s on an analysis of the oil reservoir and the economics of generating the energy. However, a major goal of all thermal methods is to reduce the viscosity of the in-place oil. For most thermal processes, this is accomplished by heating a very heavy oil (API gravity < 20°), which dramatically reduces its viscosity. In the case of lighter oils, some components of the oil can be vaporized by the heating, thus increasing their mobility by displacing them to the production wells in the vapor phase.
In a thermal flood, much effort is devoted to treating the boiler water and the stack gases resulting f r o m the burning of
p r o d u c e d oil or gas to g e n e r a t e heat. Because of environmental considerations, this usually limits the technique to unpopulated areas.
Thermal recovery techniques have been the most successful of the techniques i m p l e m e n t e d to date. This is because the remaining oil saturation prior to a thermal flood is usually liigh so that the EOR target is large. The choice of hot water injection, continuous steam injection, cyclic steam injection, or in situ combustion is d e p e n d e n t on the depth of the reservoir, the thermal properties of the s u r r o u n d i n g f o r m a t i o n s , and the fluid p r o p e r t i e s of the in-place oil. Therefore, the design p h a s e of a t h e r m a l project typically concentrates on determining how fast energy can be transferred to the reservoir by the process being considered a n d the cost of producing that thermal energy.
506
PART 10—RESERVOIR ENGINEERING METHODS
Stack Gas Scrubber
Steam Generator
Production Fluids (Oil, Gas and Water) Separation and Storage Facilities
Injection Well
MS
Production Well
щш
Steam and Condensed Water
Hot Water
Oil and Water Zone Near Original
Reservoir Temperature
Figure 3. Schematic diagram ot steam flooding. In this method, heat reduces the viscosity of oil and increases its mobility. (Courtesy of U.S. Department of Energy, Bartlesville, Oklahoma.)
Injection Well
Air Compressor
Production Well Combustion Gases
Oil and Water
;' ; ' : ' - ^ - t ^ i >: >
1. Injected Air and Water Zone (Burned Out) 2. Air and Vaporized Water Zone 3. Burning Front and Combustion Zone (600° - 1200=F) 4. Steam or Vaporizing Zone (Approx. 400°F)
i : <: г ^ т д - о - ^
5. Condensing or Hot Water Zone
T jI
(50° - 200°F Above Initial Temperature) J
6. Oil Bank (Near Initial Temperature)
l 'I
7. Cold Combustion Gases
'I
l
Figure 4. S c h e m a t i c d i a g r a m of in situ c o m b u s t i o n . T h e mobility of oil is increased by reduced viscosity c a u s e d by heat a n d solution of combustion gases. (Courtesy of U.S. Department of Energy, Bartlesville, Oklahoma.)
Reservoir Modeling for Simulation Purposes
Koen Weber
Shell Internationale Petroleum Maatschappi/ B.V. The Hague, The Netherlands
INTRODUCTION
R u i d flow in a reservoir is controlled by bed continuity, the p r e s e n c e of b a f f l e s to f l o w , a n d the p e r m e a b i l i t y distribution (see the chapter on " F u n d a m e n t a l s of Fluid Flow" in Part 10). Reservoir heterogeneities influencing fluid flow range from large scale faults and discontinuities d o w n to thin shale intercalations, sedimentary structures, and even pore scale features (Figure 1).
Simulation studies (see the chapter on "Conducting a Reservoir Simulation Study: An O v e r v i e w " in Part 10) performed at early stages of field development are d o n e to estimate parameters such as optimal well spacing, while at later stages the objectives may be infill drilling or secondary recovery. The role of the geologist is to p r o v i d e reservoir models that give a sufficient description of those parameters that control the fluid flow relevant to the planned simulation
1. Sealing fault Semi- sealing fault Non - sealing fault
2. Boundaries genetic units
3. Permeability zonation within genetic units
4. Baffles within genetic units
5. Laminationj Cross - bedding
6. Microscopic heterogeneity Textural types, Mineralogy
7. Fracturing - Tight - Open
Figure 1. Classification of reservoir heterogeneity types.
study (Harris, 1975). The key to effective and economic field d e v e l o p m e n t p l a n n i n g lies in early recognition of those reservoir characteristics that control drive mechanisms, sweep efficiency, and consequently, well spacing requirements and possible need for pressure support. The significance of the v a r i o u s h e t e r o g e n e i t y types for oil recovery is summarized in Table 1.
The steps in the modeling process are as follows:
1. Determining the facies of the reservoir rock through data gathering from cores and logs
2. Rock typing for each environment of deposition to estimate permeability from log derived porosity and estimation of vertical permeability
3. Correlation of all wells to provide a f r a m e w o r k for the delineation of a simulation model
4. Determination of an optimal grid block pattern using the flow unit principle
5. Mapping the reservoir properties in each grid block layer
DATA GATHERING
The realism a n d reliability of the resulting m o d e l is a f u n c t i o n of the reservoir h e t e r o g e n e i t y b u t also of the planning that has gone into the data gathering. Defining suitable data gathering schemes for a specific field requires multidisciplinary cooperation and a sound understanding of the significance of the data (Hurst and Archer, 1986). Analog cases and sensitivity studies on prototype models can guide this process which is crucial to the reliability of the results. In Table 2, an overview is given of the value of the different types of data for the identification a n d quantification of reservoir heterogeneities. Only a part of the data are of geological nature. Reliable information on reservoir connectivity is often derived from production and pressure tests (see the chapters on "Production Testing" and "Pressure Transient Testing" in Part 9). In particular, it is very difficult to determine large scale vertical permeability from core and log data.
The detail to which a reservoir can be modeled is a function of b o t h the degree of heterogeneity a n d the data density. At an early stage of d e v e l o p m e n t , one m a y be restricted to seismic data, sparse well data, and production tests only. In these circumstances, we reap maximum benefit from knowledge of the typical flow characteristics associated with the prevailing facies and diagenetic overprint a n d of a data base to predict reservoir body continuity and architecture. Regional experience in analog fields can often be used to guide the modeling process (Slatt and Hopkins, 1990).
The data that form the basis for reservoir modeling ideally comprise regional data, seismic data, cores, logs, pressure
604
506
PART 10—RESERVOIR ENGINEERING METHODS
Table 1. Significance of Reservoir Heterogeneity Type for Oil Recovery
Reservoir Heterogeneity Type
Sealing fault Semi-sealing fault Nonsealing fault Boundaries as genetic units Permeability zonation
within genetic units Baffles
within genetic units Laminations,
cross-bedding Microscopic
heterogeneity Textural types Mineralogy Tight fracturing Open fracturing
Reservoir Continuity
• X • X
Symbols: • = strong effect; X = moderate effect. Source: Weber (1986).
SweeD Efficiency
Horizontal
Vertical
• • •
• •
• •
X
•
X
•
X
X
X
•
•
ROS in Swept Zones
Rock-Fluid Interactions
X
X
•
•
•
X
•
• • •
measurements, wireline formation tests, pulse tests, and wellplanned production tests (Table 2). Modern three-dimensional seismic data can be u s e d for a r a n g e of m o d e l i n g purposes from structural analysis to reservoir properties such as thickness, lithology, porosity, and pore fill (Ruijtenberg et al., 1990) (also see Part 7 on Geophysical Methods).
The effect of large scale features such as faults can be estimated on the basis of geological experience and modeling (Weber et al., 1978) or it must be evaluated by pressure measurements or fluid level differences.
ROCK TYPING AND PERMEABILITY ESTIMATION
The first priority in describing the reservoir rock is the determination of the environment of deposition and the range of lithofacies that occur within the reservoir (see the chapters on "Lithofacies a n d E n v i r o n m e n t a l A n a l y s i s of Clastic Depositional Systems" and "Carbonate Reservoir Models: Facies, Diagenesis, and Flow Characterization" in Part 6). Regional stratigraphic information, cores, and sidewall samples are used for this purpose. Of particular interest is the rock t y p i n g t h r o u g h a s t u d y of p o r o s i t y , p e r m e a b i l i t y , petrography, and capillary properties (Figure 2) (see Part 5 on Laboratory Methods). For simulation purposes, permeability is a major parameter, and estimating the permeability profile in noncored wells is of prime importance (Wolf and PelissierCombescure, 1982). The basis for these techniques is m u l t i v a r i a t e a n a l y s i s of the c o m b i n e d l o g g i n g d a t a . Discriminant analysis of log response using a core calibrated system usually leads to the best results. In general, one has to combine several rock types into an electrofacies class mainly because of the poor vertical resolution of the nuclear logs if r u n in s t a n d a r d fashion. If the p o r o s i t y a n d permeability relationships of combined rock classes differ little, this is an acceptable simplification (Figure 3).
WELL CORRELATION
The next step in reservoir modeling is correlating the reservoir intervals from well to well (Figure 4). This procedure is strongly dependent on facies and well spacing. A requirement is a s o u n d database of genetic unit geometry, for example, width to thickness ratio of a specific sand body type. W h e n no deterministic correlation can be m a d e of reservoir units, it may be necessary to use probabilistic modeling techniques, but in such cases only prototype simulations are usually carried out (Haldorsen et al., 1987).
The framework for constructing simulation models is controlled by facies distribution, major permeability contrasts, and impermeable layers (Weber and Van Geuns, 1990). A g a i n , m a x i m u m u s e s h o u l d be m a d e of r e s e r v o i r engineering data with emphasis on wireline formation pressures. Geological predictions of sealing surfaces are often unreliable because of the presence of small scale faults or local erosion.
GRID BLOCK PATTERNS
The reservoir has to be subdivided in such a way that all major baffles are represented and that the individual layers and bodies can be described using average petrophysical properties that lead to realistic flow simulations. For this p u r p o s e , the term flow unit h a s been i n t r o d u c e d , w h i c h is defined as a v o l u m e of rock characterized by a specific combination of geological and petrophysical properties that influence fluid flow (see the chapter on "Flow Units for Reservoir Characterization" in Part 6). Flow units can be composed of a combination of stratigraphic elements if they do not essentially differ in fluid flow criteria (Slatt and Hopkins, 1990).
The subdivision of the reservoir into flow units is typically a multidisciplinary activity because geological and reservoir
Table 2. Value of Data for Identification and Quantification of Reservoir Heterogeneities
Reservoir
Detailed
Heterogeneity
Seismic
Type
Sealing fault
•
Semi-sealing fault
•
Nonsealing fault
•
Boundaries as
genetic units
X
Permeability zonation
within genetic units
Baffles
within genetic units
Laminations,
cross-bedding
Microscopic
heterogeneity,
textural types,
mineralogy
Tight fracturing
Open fracturing
X
Reservoir Pressure Hor. Distrib. Vert. Distrib.
•
X
•
X
X
X
•
•
X X
•
X
Prod. Tests
X X X X
• •
Production Data/Tests
Pulse Tracer Prod.
Tests Tests Hist.
•
X
•
X
X
X
X
X
X
X
X
X
•
•
X
•
Symbols: • = great value; X = occasionally of value. Source: Weber (1986).
Prod. Logs
X X X
Well Logging
Stan- Spe-
ROS
dard
cial
Rock Samples Outcropor Cores SWS Analog
Cuttings Reservoir
XXX
X
X
X
X
X
•
X
X
•
X
•
•
•
•
X
•
X
X
X
X
X
X
X
X
X
X
506
PART 10—RESERVOIR ENGINEERING METHODS
CORE STUDIES
Ш *L
•iiy'-; Ф
м - ••'•'
Ш• Г
SEDIMENTOLOGY SHELF MUDS
0
TRANSGR^SANDS
UPPER
P
SHOREFACE
L. SHOREFACE TRANSITION ZONE
SHELF MUDS
ROCK QUALITY / FACIES ANALYSIS
Figure 2. Analysis of core data for facies identification and rock quality assessment.
LOG-FACIES CALIBRATION
WITH FACIES OR GENETIC UNITS / ELEMENTS
Ф I H
V:
Г
HHI
§ -fr J
O
Ф / Ф1
RESERVOIR SUBDIVISION
INITIAL SUBDIVISION
\J
UNIT 1
GENETIC SUBDIVISION
1.1 TRANGR. UNIT, LOW 0 / к
UPPER SHOREFACE 1.2 - HIGH 0 / к
4
LOWER SHOREFACE 1.3 - LOW/ MODERATE 0 / k
T
SEALING SHALE
Figure 3. Log-facies calibration and determination of facies-related rock characteristics.
engineering information has to be used together. Economic considerations with respect to n u m b e r s of grid blocks to be used also play a role.
Reservoir s i m u l a t i o n o p e r a t e s on the principle of simultaneously solving the flow equations between adjacent blocks of rock in response to offtake f r o m wells. The larger the n u m b e r of grid blocks u s e d , the closer the m o d e l resembles the geological prototype. However, even with the fastest computers, the simulation models are restricted to a m a x i m u m of some 60,000 grid blocks. For economic reasons, most simulations m a k e use of less than 10,000 grid blocks. Consequently, grid block sizes normally range from 100 to 300 m in diameter and from 5 to 15 m in thickness (see the chapter on "Conducting a Reservoir Simulation Study: An Overview" in Part 10).
Practical grid block sizes usually imply the amalgamation of a few geological features. Thus, an averaging procedure is required to obtain realistic overall flow characteristics for the blocks as a whole. The influence of d i s c o n t i n u o u s shale breaks on vertical permeability, for example, can be estimated
using a statistic approach (Begg et al., 1985). Averaging horizontal and vertical permeability over grid block size units is a difficult task. In practice, the effective horizontal permeability usually ranges from the arithmetic to the geometric average of the permeability profile of the block. The more continuous the sublayers of the flow unit, the closer the average lies to the arithmetic average. The more random the permeability, the closer it gets to the geometric average. Geostatistical methods have become popular to tackle these problems (Journel and Alabert, 1990). Vertical permeabilities are difficult to measure, and the values used are often based either on experience for a given facies or on vertical pulse tests or other pressure data evaluation.
MAPPING OF RESERVOIR PROPERTIES
The geological input to three-dimensional reservoir simulation m u s t consist of structural m a p s a n d p r o p e r t y maps. Typically, maps must be prepared for every simulation
Reservoir Modeling for Simulation Purposes
535
Figure 4. Correlation ot reservoir units and subdivision of reservoir in flow units.
Figure 5. Mapping of reservoir properties per grid block layer to provide input for the reservoir simulation.
m o d e l layer, s p e c i f y i n g the d i s t r i b u t i o n of n e t / g r o s s , isochores, porosity, horizontal and vertical permeability, capillary pressure curve characteristics, and water saturation (Figure 5). Also, the geologist and the reservoir engineer have to cooperate to define pseudo-relative permeability curves for different internal grid block heterogeneity types. The m a t c h i n g p h a s e of the simulation s t u d y requires similar cooperation to arrive at a final model with realistic properties.
MODELING CARBONATE RESERVOIRS
M o d e l i n g of c a r b o n a t e reservoirs is generally m o r e difficult than modeling clastic reservoirs. The reason is that
carbonate rocks usually undergo a much more complex pattern of diagenetic processes. As a result, the permeability distribution can be complex and poorly related to original facies distribution. A further complication can be the occurrence of open fractures.
Carbonate reservoirs without fractures can in principle be treated similarly to clastic reservoirs. Reservoir simulation of so-called d u a l porosity reservoirs is difficult because of the problem of quantifying the degree of capillary contact across fractures. Recently, methods have been proposed to tackle the problem of block-to-block interaction (Por et al., 1989). In all cases, carbonate reservoirs require a considerable amount of core studies and frequent use of the scanning electron microscope and its auxiliary equipment.
Conducting a Reservoir Simulation Study: An Overview
Scott K. Laudeman
ApTech Associates Pasadena, California, U.S.A.
INTRODUCTION
This chapter describes the steps necessary to conduct a detailed reservoir simulation study (also see the chapter on "Reservoir Modeling for Simulation Purposes" in Part 10). A simulation study requires description of the reservoir's rock and fluid properties, validation of completion a n d production history, and extensive history matching to validate and modify this input data. When liistory matching is complete, n u m e r o u s p r e d i c t i o n s of field and well p e r f o r m a n c e characteristics are calculated for various development scenarios.
The reservoir characterization required to define porosity and permeability for each grid block in a reservoir simulation model is more stringent and at the same time more loosely defined than reservoir characterization required in detailed development geology studies. A simulation engineer can spend countless hours defining the reservoir model. Nevertheless, despite the plausibility of the interpretation used to d e v e l o p this description, the ultimate test of the simulation model's validity is its reproduction of production data. The smaller relative uncertainty of production data with regard to input data dictates that simulation engineers may take many liberties in matching that data, while rationalizing their choices with a broad degree of latitude. Development geologists are the most likely source of quality control for the reservoir description developed and modified for a reservoir simulation study.
segments using a coarse grid. Small models can provide insight into the mechanics of production performance (such as viscous flow, gravity, or heterogeneity). For example, a fluid injection project can be studied using detailed areal and vertical models rather than a coarse grid three-dimensional model. Extrapolating results from these mechanistic models to field performance can be accomplished using their results to modify general recovery characteristics defined by coarse grid models.
PREPARING FLUID PROPERTIES
T w o types of fluid descriptions are used in reservoir simulation studies: black oil and compositional. The black oil description expresses the fluid's volumetric (formation volume factor and solution gas-oil ratio) and flow (viscosity)
MODEL SPECIFICATION
A simulation study begins by choosing a model type from a m o n g those s h o w n in Figure 1. A grid, which determines the resolution at which the complex reservoir flow equations are solved, is then selected. The model's size is defined by the n u m b e r of grid blocks resulting from the grid overlain on the field or field segment being studied. In general, the accuracy of results from simulation studies is greater for smaller grid block sizes. Smaller grid block sizes permit more detailed descriptions of reservoir heterogeneity and more accurate resolution of fluid fronts and phase behavior. An o p t i m u m grid size is often determined by running test cases at several different grid sizes. The largest grid size at which no appreciable change in the results occurs is selected for the study.
Practical limits on the size of reservoir simulation models are often imposed by computational expense or capabilities. These constraints m a y dictate that the size of the reservoir segment being simulated must be reduced or the grid block size increased. Simulating small characteristic segments of a field using a fine grid may be preferable to simulating larger
Figure 1. Typical models used in reservoir simulation, (a) Zerodimensional tank, (b) One-dimensional linear, (c) Onedimensional radial, (d) Two-dimensional cross-sectional, (e) Two-dimensional area, (f) Two-dimensional radial, (g) Threedimensional. (From Mattax and Dalton, 1990; Copyright © 1990 Society of Petroleum Engineers.)
536
Conductiong a Reservoir Simulation Study: An Overview
537
у 0.4
.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Oil S a t u r a t i o n , S0
1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Oil Saturation. S 0
Figure 2. (a) Pseudo-relative permeability and (b) capillary pressure curves calculated for two-layer thicknesses compared to laboratory measurements. (From Coats, 1967; after Matta, 1990; Copyright © 1967,1990 Society of Petroleum Engineers.)
characteristics as a function of pressure for three phases: oil, gas, and water. In complex fluid mixtures, the fluid's volatility often dictates that volumetric and flow characteristics are not only a function of pressure but also of composition. For these fluids, a compositional description uses an equation of state to describe the fluids' volumetric and flow characteristics. An equation of state describes a fluid in t e r m s of the f u n d a m e n t a l physical p r o p e r t i e s of its components: methane, ethane, heptanes-plus, and so on. These fundamental physical properties—critical pressure, critical temperature, critical volume, acentric factor, and interaction coefficients—are unique for each compositional fluid description derived for a simulation study.
Black oil fluid descriptions are used to describe most oil and gas fields. Primary depletion, waterflooding, and gas injection can all be simulated with black oil models. Volatile oil reservoirs or gas condensate reservoirs generally require compositional models. These models may exhibit such complexities as a fluid whose density is Unearly proportional to depth or whose phase switches repeatedly between oil and gas. Thermal models, used to simulate steam injection, may use either black oil or compositional fluid descriptions. Black oil thermal models describe fluid properties as a function of t e m p e r a t u r e as well as p r e s s u r e . The i m p o r t a n c e of oil volatilization in thermal recovery often dictates that compositional models are used to simulate thermal recovery processes.
PREPARING MULTIPHASE FLOW PROPERTIES
Fluid saturations and the p r o d u c e d fractions of oil, gas, and water are determined by capillary pressures and relative permeabilities specified as functions of w a t e r saturation.
Equilibrium (initial) fluid saturations are directly d e p e n d e n t on capillary p r e s s u r e , w h i c h is itself a f u n c t i o n of height above the fluid contacts, the fluid densities, porosities, permeabilities, and the surface chemistry of the fluids. Once production or injection commences, fluid movement is controlled by the relative permeability of each phase (except at very low velocities w h e r e the effects of capillary pressure are important). In a reservoir consisting of two fluid phases, oil-water, gas-oil, or gas-water capillary pressures and relative permeabilities must be specified. For a three-phase system, relative permeabilities and capillary pressures for two of the three possible systems are specified.
Capillary pressure m a y be expressed using the Leverett Jfunetion. This f u n c t i o n can b e u s e d to calculate capillary p r e s s u r e as a f u n c t i o n of each grid block's porosity and permeability. Sufficient data may also exist to correlate relative permeability curves' initial and residual saturations with porosity. Rather than assign individual capillary pressure and relative permeability curves for each grid block, average curves can be derived for several ranges of porosity and permeability values (also referred to as regions).
Relative permeability and capillary pressures characterizing grid blocks may differ considerably from laboratory measurements. Laboratory measurements m i n i m i z e the effects of g r a v i t y a n d h e t e r o g e n e i t y . An example of these pseudo-relative permeabilities and capillary pressures is s h o w n in Figure 2. The importance of gravity and heterogeneity effects is greater with larger layers. Pseudo-relative permeability and capillary pressure curves are often developed on small detailed studies for larger scale models. In some cases, pseudo-relative permeability and capillary pressure is developed during history matching. Pseudo-relative permeability and capillary pressure may be d e p e n d e n t on fluid a n d saturation history. Their ability to
506
PART 10—RESERVOIR ENGINEERING METHODS
I • I—I— —
'
, •;• «
Л
»•
W
•
• < -I
•
Л , .IA
•
LUG
\ — ^
'
—-Г-Ч
—•—
T
«
•
•
•
• •
•
-+Y
»
I »•
ГЛ
. H
A
V *! /
* ^M
-Л
•
>N * *
> T i•—•—
•
,
6
8
10
12
14
16
18
20
22
Core Porositv (%)
Figure 3. Model grid overlain on Khursaniyah field, Saudi Arabia. (From Boberg, 1974; Copyright © 1974 Society of Petroleum Engineers.)
account correctly for gravity and heterogeneity is limited. Where these effects are significant, a smaller grid size should be used.
PREPARING MATRIX PROPERTIES
Matrix properties describe reservoir characteristics specified at each point of the grid (matrix) overlying the reservoir model (Figure 3). Once a grid has been selected, the average depth, thickness, porosity, and permeability are calculated for each grid block. Digitized structure and isopach maps may be used with mapping and gridding software to calculate average depths and thicknesses. Mapping software is used to convert the digitized contour m a p s to an interpolated grid of values. The m a p p i n g grid should be several times finer than the reservoir simulation grid since the values in the mapping grid falling within grid block boundaries are used to calculate the depths and thicknesses for each grid block. If the m a p p i n g grids are fine e n o u g h , a simple averaging of the values within each grid block will suffice to calculate their values corresponding to the grid block centers.
Porosities are calculated for each grid block in a method similar to that described for reservoir depths and thicknesses. First, the porosities calculated at 0.5- to 1-ft intervals from well logs must be averaged for each simulator layer. These layer porosities are then contoured and gridded. More advanced
Figure 4. Core measurements from the Bradford Sandstone. (From Levorsen, 1967.)
techniques for calculating three-dimensional porosity distributions (such as stochastic and geostatistical methods) are topics of current research, but they are beyond the scope of this chapter. The porosity values for each grid block are calculated by averaging the porosity grid values that lie within the boundaries of each grid block.
Permeability is often correlated as a function of porosity (Figure 4). The porosities calculated at 0.5- to 1-ft intervals from well logs are used to calculate horizontal and vertical permeabilities for the same intervals. Permeability averaging techniques require special consideration. For porosity, a volumetric averaging technique produces the same volume of void space for the averaged block as exists in all the smaller blocks. For permeability, averaging techniques must ensure that the flow rates of the averaged blocks are the same as the combined flow rates of the smaller blocks. Arithmetic averaging is used to calculate the average horizontal permeability at the well for each simulator layer. Geometric averaging is used to calculate the average vertical permeability. These layer permeabilities are then contoured and gridded. Figure 5 shows that the choice of permeability averaging technique has a significant effect on the calculated average.
INITIALIZING THE MODEL
The first step in any m o d e l s t u d y is the calculation of pressures and fluid saturations before the onset of production. At this point in time, the model should be in static
Series Flow
Conductiong a Reservoir Simulation Study: An Overview
539
Parallel Flow
к = л/|Е11 /*,)
0.8
к = (L А,)/л
0.6
F inal Match
100
/
• /
/
10
Sl S
Parsi l l e l
0.4 First Trial
X oo 0 1
O Observed data
3
5
7
Time, daysx 10-3
Figure 6. Simulated water production for initial and modified reservoir description. (From Mattax and Dalton, 1990; Copyright © 1990 Society of Petroleum Engineers.)
— —
Ser ies
6 8 1 0 1 2 1 4 16 18 20
Number of Cells, n
Figure 5. Comparison of mean permeabilities calculated for series flow (geometric average) and parallel flow (arithmetic average). (From Lishman, 1970; Courtesy Canadian Institute of Mining, Metallurgy, and Petroleum.)
equilibrium. Distinct fluid contacts should appear at the appropriate depths, and the potential (the sum of pressure and the gravity head) should be equal everywhere in the model. A model that exhibits indistinct fluid contacts or uneven potentials (meaning that flow will occur in the next time step without production) has not been initialized properly.
HISTORY MATCHING
Reservoir simulation models are normally calibrated using production history. During this process, significant modifications to the reservoir description may be required to match historical performance. Figure 6 shows the initial trial and the final match achieved in a well model. History matching is often a trial-and-error process requiring 20 model runs or more before a satisfactory match between the predicted and observed performance is realized. In new fields, little data exists and reservoir performance predictions may be made without history matching, but this produces unreliable performance predictions. Any data at all, even well tests from discovery and delineation wells, should be used to calibrate the reservoir model. Because of the relatively minuscule volume of data sampled by well logs, it is simple to
construct a model that produces performance predictions that diverge dramatically from actual results. Any a m o u n t of data, unless it has been proven erroneous, is more credible than performance predictions from the most elaborately constructed simulator.
History matching is usually conducted by using the production data to specify the production rate for one phase, either oil or gas. With this phase specified, the model calculates the rates of the other phases. History matching attempts to match the rates of the unspecified phases and any pressure measurements.
Modifications in reservoir description needed to match history are rarely straightforward. Pressure decline is a function of initial fluid volume, the rate of fluid production, and fluid properties. The rate of fluid production is a function of the d r a i n a g e area p r e s s u r e , t h e p e r m e a b i l i t y of the formation and close to the well, relative permeabilities, and fluid properties. A discrepancy between observed and calculated pressures can be caused by an incorrect pore volume, incorrect interwell permeabilities that allow too little or too great fluid recharge after production, or incorrect relative p e r m e a b i l i t i e s r e s u l t i n g in the incorrect rate of production for the second producing phase. Failure to model larger scale features properly, such as interlayer communication, aquifer support, or faults, may also be responsible for discrepancies in pressure matches. Discrepancies in predicted rates can be equally as difficult to match. Wellbore conditions may also have a significant effect on history matching.
History matching continues to be an art. Even final matches are subject to uncertainty since the large number of parameters that can affect the results precludes identification of a single unique match. The participation of a geologist in the modifications to reservoir description that occur during liistory matching can greatly enhance the possibility that the final description will approximate the actual reservoir.
506
PART 10—RESERVOIR ENGINEERING METHODS
PREDICTING PERFORMANCE
Predictions almost always begin by calculating the projected p e r f o r m a n c e of the base case. The base case represents the mode of operations where everything remains exactly as it stands on the last day of the history match. In its strictest interpretation, the base case means no new wells, recompletions, workovers, or changes in a well's artificial lift status. In practice, the base case means these changes usually occur to the extent that they could have been foreseen (they are already in the budget or 5-year plan) or by extrapolating today's producing (or injecting) rules into the future (any well producing below a certain rate is converted to gas lift).
Each prediction case must specify each well's producing rules. The location of each well to be produced or injected during the prediction must be specified, including the time at which it is considered active. In the simplest case, wells are constrained to produce at a constant rate or at a constant bottomhole pressure. Economic limits can also be specified. Using simple rules, the simulator output must be run in
increments of 1 year or less so that the results can be scanned to determine whether changes should be made in a well's status, completion, or constraints to reflect the field's operating practices. This process is time consuming since the simulator must be constantly monitored and rerun.
Advanced simulators have tremendous flexibility with regard to producing rules. Operating constraints imposed by gas plants can be simulated by specifying a constant group rate. Water plant constraints can be simulated by specifying a maximum water rate. Gas li^ft injection or pumping can be simulated by calculating well rates using bottomhole pressures derived from wellbore gradient calculations and by allocating the total amount of gas lift gas or electricity to each well based on its productivity. Plug backs can be simulated by automatically recompleting producing grid blocks when saturations exceed a specified value in the initial grid block. Wells can be drilled to maintain a specified target rate by searching for high saturation grid blocks. These are just a few of the m a n y special features that can be used to simulate producing rules.
References Cited
541
Part 10 References Cited
Begg, S. H v D. M. Chang, and H. H. Haldorsen, 1985, A simple statistical method for calculating the effective vertical permeability of a reservoir containing discontinuous shales: Society of Petroleum Engineers Symposium on Reservoir Simulation, Dallas, TX, Feb. 10-13, SPE 14271.
Boberg, T. C. et al, 1974, Application of inverse simulation to a complex multireservoir system: Journal of Petroleum Technology, July, p. 801-808; Transactions, AIME, v. 257.
Bradley, H. B., ed., 1987, Petroleum Engineering Handbook: Richardson, TX, Society of Petroleum Engineers.
Carr, B. S., and J. Viret, 1986, PSD860030: Houston, TX, Chevron Exploration and Production Services, March.
Clark, N J., 1969, Elements of petroleum reservoirs: Dallas, TX, Society of Petroleum Engineers, AIME, p. 66-84.
Coats, K. H. et al., 1967, Simulation of three-dimensional, twophase flow in oil and gas reservoirs: Society of Petroleum Engineers Journal, Dec., p. 377-388; Transactions, AIME, v. 240.
Craft, В. C., M. Hawkins, and R. E. Terry, 1991, Applied Petroleum Reservoir Engineering, 2nd. ed.: Englewood Cliffs, NJ, Prentice Hall, p. 210-263.
Craig, F. F., 1971, Reservoir engineering aspects of waterflooding: Richardson, TX, Society of Petroleum Engineers Monograph Series, v. 3.
Cronquist, C., 1979, Evaluating and producing volatile oil reservoirs: World Oil, April, p. 159-166.
Dake, L. P., 1978, Fundamentals of Reservoir Engineering: The Netherlands, Elsevier Science Publishers, p. 79-102.
Dykstra, H., and R. L. Parsons, 1950, Secondary recovery of oil in U.S.: American Petroleum Institute Publication, p. 160174.
Golan, M., and C. H. Whitson, 1991, Well Performance, 2nd. ed.: Englewood Cliffs, NJ, Prentice Hall.
Haldorsen, H. H., P. J. Brand, and C J.Macdonald, 1987, Review of the stochastic nature of reservoirs: Proceedings of the Seminar on Mathematics of Oil Production, Cambridge, U.K.
Harris, D. G., 1975, The role of geology in reservoir simulation studies: Journal of Petroleum Technology, May, p. 625-632.
Hurst, A., and J. S. Archer, 1986, Sandstone reservoir description—an overview of the role of geology and mineralogy: Clay Minerals, v. 21, p. 791-809.
IHRDC, 1982, Production rate decline curves: PE107, Boston, MA, IHRDC.
Journel, A. G., and F. G. Alabert, 1990, New method for reservoir mapping: Journal of Petroleum Technology, Feb., p. 212-218.
Katz, D. L., et al., 1959, Handbook of Natural Gas Engineering: New York, McGraw-Hill.
Landrum, B. L., and P. B. Crawford, 1960, Petroleum Transactions, AIME, v. 219, p. 407.
Levorsen, A. I., 1967, Geology of petroleum, 2nd ed.: San Francisco, W. H. Freeman Publishing, p. 128-129.
Lishman, J. R., 1970, Core permeability anisotropy: Journal of Canadian Petroleum Technology, April-June, p. 79-84.
Mattax, C. C., and R. L. Dalton, 1990, Reservoir simulation: Richardson, TX, Society of Petroleum Engineers.
McCain, W. D., Jr., 1990, Petroleum Fluids, 2nd ed.: Tulsa, OK, Pennwell Books.
National Petroleum Council, 1976, An analysis of the potential for enhanced oil recovery from known fields in the United States—1976-2000: NPC, Dec.
Odeh, A. S., 1986, Reservoir fluid flow and natural drive mechanisms, in IHRDC Video Library for Exploration and Production Specialists, Manual for Module PE502: Boston, MA, IHRDC, p. 69-120.
Por, G. J., et al., 1989, A fractured reservoir simulator capable of modelling block/block interaction: SPE Annual Technical Conference and Exhibition, San Antonio, TX, Oct. 8-11, SPE 19807.
Ruijtenberg, P. A., R. Buchanan, and P. A. B. Marke, 1990, Three-dimensional data improve reservoir mapping: Journal of Petroleum Technology, Jan., p. 22-61.
Shah, D. O., and R. S. Schechter, 1971, Improved oil recovery by surfactant and polymer flooding: New York, Academic Press.
Slatt, R. M., and G L. Hopkins, 1990, Scaling geologic reservoir description to engineering needs: Journal of Petroleum Technology, Feb., p. 202-210.
Society of Petroleum Engineers, 1981, Phase behavior: Dallas, TX, SPE Reprint Series No. 15.
Stalkup, F. I., Jr., 1983, Miscible displacement: Richardson, TX, Society of Petroleum Engineers Monograph Series.
Standing, M. B., 1977, Volumetric and phase behavior of oil field hydrocarbon systems: Dallas, TX, Society of Petroleum Engineers, AIME.
Weber, K. J., G. Mandl, W. F. Pilaar, F. Lehner, and R. G. Precious, 1978, The role of faults in hydrocarbon migration and trapping in Nigerian growth fault structures: Offshore Technical Conference, Houston, OTC 3356.
Weber, K.}., 1986, H o w heterogeneity affects oil recovery, in Reservoir Characterization, L. W. Lake and H. B. Carrol, Jr, eds.: Orlando, FL, Academic Press, p. 487-544.
Weber, K. J., and L. C. Van Geuns, 1990, Framework for constructing clastic reservoir simulation models: Journal of Petroleum Technology, Oct., p. 1248-1297.
Whitson, C. H., and M. R. Brule, 1993, Phase Behavior: Society of Petroleum Engineers Monograph Series, in press.
Willhite, P. G, 1986, Waterflooding: Society of Petroleum Engineers Textbook Series No. 3, chap. 2.
Wolf, M., and J. Pelissier-Combescure, 1982, Faciologautomatic electrofacies determination: SPWLA Third Annual Logging Symposium Transactions, July.
Index
AAPG Computer Applications Committee, 450
Absolute open flow, 508 Accessories, on graphic log, 198-199 Acid fracturing, 472-473 Acid washing, 472 Acidizing, 472-473 Acquisition phase, 4 Acre, subdivision of, 7 Adhesive forces, 221 Aeromagnetic surveys, 415,416 After-tax net cash flow, 38,44-45 After-tax net operating income, 38 Age, of strata, 229-230 Agreements, oil and gas, 16-18 Air drilling, 78 Air gun, 359,361,404 Air respirators, 69 Alaskan native lands, 15 Algorithms, surface modeling, 431 Aliasing, 366
Allowable depletion, 43,45-46 Alluvial fan deposits, 265 Amortization, 46 Amott wettability method, 218-219,220 Amplitude, of wavelet, 389 Amplitude versus offset
analysis, 398-400 Analog data, 441 Analogy method, for estimating
reserves, 515-516 Angle averaging method, 290,291 Annular velocity, 96 Anomalies
in gravity, 411-412 in magnetism, 415-416 Antialias filter, 366 Anticline, arrow plot of, 161 API specific gravity, 504 Apparent velocity, 369 Archie equation, 172,183-184 Arrow plot, 160,161 Artificial lift, 485-486 Asphal tenes, 493 Asymmetrical data set, 340 Attenuation, of seismic data, 389 Automatic gain control, 378 Auxiliary wireline tools, 148-149 AVO analysis, 398-400 Back pressure equation, 508 Balanced cross section, 300,333,334 Balanced tangential method, 290,291 Band-limited inversion, 395 Barex, to preserve core, 129 Barges, 65
Barrier foil laminate, to preserve core, 129-130
Barrier island, 281 Base maps, 451-452 Baselap, 433-434
Beam pump, 486,487 Bedding, and permeability, 211 Bent housing motor, 72, 74 Bent sub, 72,74 Bias, 377
in estimating, 25 in risk decisions, 53 Bid strategy, 54 Bidding agreements, 17 Binary mixing diagram, 246 Biofacies, 230 Bit hydraulics, 96-97 Bivariate distribution, 347 Black oil, 505,536 Block diagram, 303,304,384 Blowout preventor, 64 Boot basket, 83,84
Borehole gravity meter (BHGM), 413-414 Borehole imaging devices, 163-166 Borehole source technology, 404-405 Borehole televiewer, 123-124,149,163-
165,192 Bottom hole assembly, 71-73 Bottom hole temperatures, 87 Bottom packer failure, 134 Bottom water drive, 520 Bouguer anomalies, 411-412 Boyle's law method, 206,209 Bradford Sandstone, 538 Braided river deposits, 265 Braidplain deposits, 265 Brine composition, 251 Brine sand model, 399 Brine-hydrocarbon system, 223-224 Bubblepoint pressure, 504,509 Buildup test curve, 479,480 Bulk density log, 214 Bulk properties, 242 Bulk volume, 207-208 Calcium carbonate cementation, 269-270 Calibration, rock-log, 200 Caliper logs, 176,192 Calipers, 148-149 Canadian Arctic Islands, 416 Capillary pressure, 221-225,537
and fluid contacts, 305,309 and reservoir quality, 276 Carbon dioxide flooding, 529 Carbonates
analysis of, 235-236 cements of, 277 classification of, 236,270 depositional environments of, 269,
271 diagenesis of, 269-270 facies of, 269,271 flow unit studies of, 284,285 heterogeneity of, 312 interpretation of, 189 porosity of, 204,206,276
reef play, 393 reservoir models of, 272-274,535 rock fabrics of, 270-272 Cased hole tools, 151-153 Cash flow see also Net cash flow maximum negative, 48 model of, 38-42 and present value, 36,40 worksheet of, 39 Casing inspection of, 491 repair of, 497 weight of, 95 Casing gun, 467,468 Casing valves, 482 Cathodoluminescence (CL), 237-240 Cement bond log, 490,491 Cement liner completion, 464 Cementation
calcium carbonate, 269-270 and reservoir quality, 275,277 in sandstone, 234 Central tendency, 339-340 Chance of success, 30-34 Checkshot surveys, 382,401 Chemical corrosion, 494 Chemical flooding, 527-528 Christmas tree, 482 Chromatography
flame ionization detection, 99,109 gas, 98,107-108,242-246 Circulation loss of, 87 system of, 62,64 time of, 96 Clastics analysis of, 263-268 depositional environments
of, 265-268 flow unit studies of, 284,285 heterogeneity of, 312 Clay minerals, 235 Clay swelling, 250 Clay types, 251 Cluster analysis, 346-347 Clustered gun, 361 Clustered points, 426 Coarsening-upward sequence, 279 Cohesive forces, 221 Coiled tubing units, 498,499 Coin toss, 31
Collapse, of formation, 493 Collars, weight of, 95 Color amplitude display, 377 Combination drive, 518,521 Combustion, in situ, 530 Command-driven system, 448 Commercial success, probability of, 32 Common midpoint gathers, 398-399
542
Common midpoint stack, 369,372,374,375,386
Compaction, and reservoir quality, 275 Compass bearings, 7 Compensated density logs, 193 Compensated neutron logs, porosity
from, 181 Compensated neutron
tools, 147,151,152 Completions
probability of, 31-32 types of, 463-465 Complex resistivity, 418 Composition, of fluids, 504 Compositional analysis, 233-234,504,536 Computer methods, 423-454 Computer-aided mapping, 383-384 Computers contouring by, 426-430 for making cross sections, 303 standards used in, 449-450 surface models built by, 431-440 in workstations, 447-450 Condensate analysis, 241-246 Conductivity, of fractures, 471-472 Confidence interval, 342 Confidence level, 341 Confining pressure
and permeability, 211,227 and porosity, 208 Conformity, modeling of, 433-434 Coning, 492 Constant percent decline, 460 Contact angle, 218,220, 221,224 Continuity analysis, 318-319 Continuous coring, 201 Continuous intervals, 124 Continuous variables, 340 Contour maps by computers, 426-430,449 by hand, 426-427,432-433 from 2-D seismic data, 381,383 of geological data, 426-430,449 of surface, 431,432 structure, 333-338 Contoured triangular mesh, 427 Contracts, oil and gas, 16 Controlled source methods, 417-418 Conventional coring, 115-118 Conveyance, of interests in land, 9 Core
alteration of, 127-128 analysis of, 237-240 depth control in, 214 description of, 198-200 of fractured reservoirs, 326-328 handling of, 125-126 orientation of, 122-124 preservation of, 128-130 routine analysis of, 201-203 spatial resolution of, 214 types of, 201 Core barrel liners, 116 Core barrels, 115,117,119,122,125 Core catchers, 115,117 Core plug, 215
Core type basket, 84 Core-based orientation, 123 Coregra ph, 317 Core-log correlation, in tight gas
reservoirs, 322 Core-log transformations, 214-215 Coring
conventional, 115-118 sidewall, 119-121 Correlation of fluid properties, 506 of reservoir units, 535 of strata, 229-230 of wells, 532 Correlation analysis, 343-344 Correlation coefficient, 343 Correlation devices, 144-145 Corrosion, 493-494 Cost depletion, 45 Costs
of ownership, 10-11 and revenue, 24 uncertainty in, 27-29 Crane field, 315 Critical flow velocity, 250,251 Critical point, 504
Cross bedding, Formation MicroScanner image of, 165
Cross sections from aeromagnetic profile, 416 of carbonate reef play, 393 computer generation of, 303 stratigraphic, 300-302 structural, 302-303,331-333 of surface data, 432,433
Cross-borehole tomography, 404-408 Cross-line direction, 385 Crossplotting
of fracture data, 326,329-330 of log data, 443,445 M-N and MID, 186-188 of neutron-density data, 178,183 of porosity/permeability, 215-216,
317,326,329-330 Cumulative production, on maps, 298,299 Curve fitting, 344 Cut fluorescence, 113 Cuttings, 104-105
evaluation of, 113-114 lithology on log, 101 Darcy (unit), 210 Darcy's law, 210-212,509-510 Data gathering, for reservoir
simulation, 531-532 Data grouping, 345 Data input, 441 Data management, 446 Data processing
of AVO data, 398-399 of FWAL data, 410 of seismic data, 364-369 of tomographic data, 405 of well log data, 445 Data representativeness, 345-346 Database, 364,447-448,450 Datasets, in seismic
interpretation, 379-380
Datum, 300 DC method, 417-418 DCFROR, see Discounted cash flow rate
of return Decision tree analysis, 33 Decline curve analysis, 460,462,517 Deconvolution, 364,367-368 Deep water marine clastic deposits, 268 Delta deposits, 267 Demultiple, 369 Demultiplex, 366 Dendrogram, 347 Density, of drilling fluid, 76 Density logs, 176-177,413-414
porosity from, 181 Density tools, 147 Dependent variable, 343 Depositional environment, 230-231
of carbonates, 269,271 of clastics, 264-268 and reservoir quality, 275-276 Depositional model, 315 Depreciation, 46 Depth maps, 382-383 Depth migration, 375 Depth section, 372,373 Depth shifting, 442 Derrick, 62
Desander/desilter, 104 Detection
of gas, 109-113 of pressure, 79-82 Detector array, 361 Development cash flow model of, 39 chance factors in, 32-33 economic parameters in, 48 geology workstation for, 447-450 income tax for, 45 risk in, 52-53 Deviated wells, 402 Diagenesis
assessment of, 316 of carbonates, 269-270 and fossils, 231 and reservoir heterogeneity, 314-320 and reservoir quality, 275,318 Diagenetic model, 315 Diagenetic profile construction, 316-317 Diagenetically complex reservoir, 314-320 Differential sticking, 83 Difficult lithologies, 186-189 Digital data, 441 Dip
computation of, 159-162 isogons, 331 true versus apparent, 293,303 Dip moveout, 369,375-376 Dipmeters, 123,149,158-162,192-193, 331,333 Directional drilling, 71-73 Directional permeability, 526 Discontinuities on maps, 430 surface, 438 Discount rates, 35-36
544
INDEX
Discounted cash flow, 24 Discounted cash flow rate of
return, 35-36,39, 47-48,49 Discounted payout, 48 Discounted profit to investment ratio, 49 Discrete variables, 340 Discriminant analysis, 346 Dispersed muds, 77 Displays
of log data, 443-445 of seismic data, 377-378 Disposable inner core barrels, 116 Disposition phase, 4 Dissolution, 270 and reservoir quality, 275 Distributed computing, 447 Distributions combining, 26 field size, 27 Diverting basket flowmeter, 490 Dix formula, 382-383 Doglegs, 87
Dolomitization, 270,272 Drawdown, 508 Drill cuttings, 104-105 Drill ships, 66, 67 Drill stem tests, 131-137,474,475 Drilled to depth, 289 Drilling
air, 78
in complex fields, 311 directional, 71-73 horizontal, 73-74 measurement while, 88-90 overpressure problems in, 79-82 problems of, 87-88 rate of, 80 Drilling fluid, 76-78
circulation of, 95-96 Drive mechanisms, 518-522 Drive recovery, 518-522 Dry core preservation, 128-130 Dry cut sample, 105 Dry gas, 505 Dual completion, 465 Dykstra-Parsons permeability
variation, 523 Dynamite shot patterns, 358,360 Dynamometer, 495 Economics
equations of, 24 parameters of, 47-51 of property acquisitions, 54-55 Edge water drive, 520 Editing
of seismic data, 364,366 of surface model, 432 of well log data, 441-442 Effective pay determination, 286-288 Effective permeability, 341 Effective porosity, 204 Electric submersible pump, 485 Electrical borehole scanning, 149,163-166 Electrical corrosion, 494 Electrical methods, 417-419 Electrical properties, of materials, 417
Electromagnetic method, 418 Emulsions, 493 Engineering methods
production, 457-499 reservoir, 501-540 Enhanced oil recovery (EOR), 527-530 Envelope technique, 439 Environmental corrections,
of logs, 168,442 Environments
carbonate, 269-274 clastic, 263-268 Eolian deposits, 266-267 EOR techniques, 527-530 Equations of state, 506 Equilibrium bank fluids, 471 Equipment
mudlogging, 98-99 rig safety, 68-69 surface production, 482-484 taxes on, 43 wireline, 499 workover, 498 Evaporites
interpretation of, 189 mineralization, 270, 272 Exactitude condition, 349 Execution phase, of interpretation, 451,453-454 Expected investment, 50 Expected net present value, 24,50-51 Expected present value, 55 Expected value, 30-34 Expendable gun, 466 Exploratory data analysis, 345 Explosives, 359 Facies
analysis of, 230-231,263-268 of carbonates, 269,271 identification of, 534 of sandstones, 266 and tight gas reservoirs, 322 Fan deltas, 265 Farm-ins, 16 Farm-outs, 16 Fault blocks, 435,436 Fault gaps, 336,436 Fault planes, 435,436 maps of, 294 sections of, 303,338 Fault separation, 436 Fault traces, 435-436 Faults
in reservoirs, 331,333 sealing character of, 335 surface models of, 435-436 Federal income tax schedule, 44 Federal lands leasing of, 14-15 ownership of, 10 Feed,504 Fence diagram, 303,304 Fetkovich type curve, 480 Field size distributions, 27 Fields new versus mature, 311 stratigraphically complex, 311-313
Fieldwide heterogeneity, 280 Filtering, 431
of data, 346 in seismic displays, 378 Filtrate invasion, 127 Fines mobilization, 250,251-252 Fining-upward sequence, 279 Finite difference methods, 373,394 First-appearance datum, 229-230 Fishing, 83-86 f-k migration, 369,373-374 Flame ionization detection
chromatography, 99,109 Flammable gases, hazards from, 69-70 Flat mill, 85 Floaters, 65 Flooding
by chemicals, 527-528 by miscible gas, 528-529 patterns of, 524 thermal, 529,530 by water, 523-526 Row
barrier to, 480 detection of, 488 of fluids, 508-512 to fractured well, 470 multi-phase, 537-538 quantitative evaluation of, 489 radial, 509-510 rates of, 132-133, 135,137,482 regions of, 460 systems of, 461 to unfractured well, 469 Flow profile, 489,490 Flow tests, 477, 479 Flow units, 282-285 Flowline, 112 Flowmeters, 489,490 Fluid
artificial lift of, 485-486 contents of, 128,505 density of, 224 drilling, 76-78 expansion/expulsion of, 127 fracturing, 471 hydraulic, 486 loss control, 76 production curves, 305 properties of, 504-507,536-537 recovery of, 132 sampling of, 156 saturation by, 201-203 segregation by gravity, 522 Fluid contacts, 305-310 geometries of, 307 Ruid flow, 508-512 Fluorescence, 113 Flushing, wellbore, 112 Fluvial deposits, 265-266 Ruviodeltaic sequence, 279 Foam drilling, 78 Folds, in reservoirs, 331,333 Formation
collapse of, 493 evaluation of, 88-90,151-153,
192-193
INDEX
545
fracture pressure of, 93 permeability of, 155-156 pressure of, 154-157 resistivity of, 145,183 Formation density logs, 176-177 Formation integrity test, 93 Formation MicroScanner, 149,163-166,193 Formation testers, 154-157,506 Formation volume factor, 504 Forward calculation, 411 Forward modeling,
of seismic data, 392-394 Fossils, 229-232 Four-arm dipmeter, 158 Fractures
conductivity of, 471-472 Formation MicroScanner image
of, 166 hydraulic, 469-472 natural, 192-193 orientation of, 470-471 pressure from, 93 in reservoirs, 326-330 zone of, 410 Framework grains, in sandstone, 234 Free surface, 221
Free water level, 222, 224,305-310 Freezing core, 130 Frequency distribution curves, 340 Fresnel zone, 388 Frio Formation, map of, 295 Full bore flowmeter, 489,490 Full waveform acoustic logging
(FWAL), 409-410 Full-closure coring system, 117,118 Future value, 35 FWAL microseismograms, 409-410 Gain recovery, 367 Gal (unit), 411,414 Gamma ray logs, 174-175
and sandstone correlation, 264 shale volume from, 180 Gamma ray tools, 144,145,151,152 Gas classification of, 505 detection of, 109-113 extraction of, 106-108 material balance estimation for, 515 specific gravity of, 504 types of, 112 Gas cap drive, 518-520 Gas chromatography, 98,107-108
of oil, 242-246 Gas condensate, 505,507 Gas flooding, 528-529 Gas gun, 359 Gas lift system, 485-486 Gas monitors, 69 Gas sand model, 399 Gas show, 109 Gas slippage, 211-212 Gas-oil contact, 305 Gas-oil ratio, 297,518 Geological age, 230-231 Geological cross sections, 300-304 Geological data, preparation of, 453
Geological and geophysical (G & C) costs, 43
Geological heterogeneities, 278-281, 307-309,531-532
Geological methods, 259-349 Geological reservoir model, 274,315, 319 Geological structure,
confidence in, 30-31 Geological success, probability of, 30 Geological time scale, of Alaska, 230 Geolograph, 91 Geophone, 359,360 Geophysical methods, 355-419,451-454 Geothermal gradient, 488 Ghosting, 362 Grain volume, 207,209 Graphic log, 198,199 Graphical crossplots, 187 Gravitational method, 206 Gravity
anomalies in, 411-412 in borehole, 413-414 Gravity drainage, 518,522 Gravity meter, 413-414 Gravity surveys, 411-412 Grid cells, 437,438,532-534,537-538 Grid nodes, 437 Grids, rectangular, 427-429 Gross hydrocarbon rock
thickness (GR), 438-439 Ground probing radar, 418 Guns
air, 359,361,404 gas, 359 perforating, 466-467 Hard rock core, handling of, 125 Harmonic decline, 460 Hazards, at wellsite, 69-70 Heater treater, 483-484 Heavy duty core barrels, 115-116 Helium porosity, 215, 216 Heterogeneity
fieldwide, 280 geological, 278-282 interwell scale, 278-280 of reservoir, 307-309,531-532 of stratigraphy, 311,312 wellbore scale, 278 Histograms, 443,445,446 Histories, of production, 460-462,481,
515-516,539 History matching, 539 Hoisting system, 62 Hole instability, 87 Horizontal resolution, 388 Horizontal separator, 483 Horizontal well, 73-74 Horner plot, 133,135,477,479,480 Hydraulic fracturing, 469-472 Hydraulic horsepower, 96-97 Hydraulic pump, 486 Hydraulics, 96
Hydrocarbon charge, confidence in, 31 Hydrocarbon evaluation, on log, 101 Hydrocarbon reserves,
estimation of, 513-517 Hydrocarbon-water contact, 305
Hydrodynamic gradients, and fluid contacts, 306-307,308
Hydrogen sulfide, 69-70 Hydrophone, 361 Hydrostatic pressure, 93-94 Hypothesis testing, 341-342 Image rays, 392 In situ combustion, 530 Income tax calculations, 45 Independent variable, 343 Indian lands, 15 Induced magnetism, 415 Induced polarization, 418 Induction logs, 185 Induction tools, 145 Inflow, 508
Inflow performance relationship (IPR), 508,511-512
Injection, in hydraulic fracturing, 470 Injection wells, 465-466 In-line direction, 385 Intangible drilling and development
costs (IDCs), 43,44 Integrated computer methods, 423-454 Interference test, 479-481 Interparticle porosity, 270 Interpretation
of seismic data, 379-380,386 of tomography, 406-408 on a workstation, 451-454 Intersecting surface techniques, 433-434 Interwell scale heterogeneity, 278-280 Inverse calculation, 411 Inversion of AVO data, 400 of seismic data, 395-397 Investment efficiency, 49-50 IPR equations, 508,511-512 Isochore map, 294 Isochron map, 294-295 Isopach map, 294,297 Jackups, 65 Jet bit, 72,73 Jet impact force, 97 Jet nozzle calculations, 96 Jet pump, 486
Joint operating agreements, 16 Junk basket, 83,84 Karst-collapse reservoir model, 272,274 Key seats, 83,87 Khursaniyah field, 538 Kicking off, 72,90,290 Kickoff point, 290 Kill weight of mud, 93 Kirchhoff method, 373,394 Kriging, 429-430 Laboratory methods, 195-252
in PVT, 506 Lacustrine deposits, 268 Lag time, 95-96,113 Land
maps of, 5 petroleum, 1-19 Land rigs, 62-64 Land seismic aquisition, 358-360 Landman, petroleum, 4 Last-appearance datum, 229-230
546
INDEX
Laterolog device, 145 Leak-off test, 93 Lease, clauses of, 13-14 Lease bonus, 43 Lease exchange agreements, 18 Lease operating expenses, 43-44 Leasehold costs, 43 Leasing
of federal lands, 14-15 of state lands, 14-15 Leverett J-function, 537 Limestones, see Carbonates Limit of resolution, 388 Linear completion, 463-464 Linear flow, 461 Linear matrix solutions, 187 Linear regression, 343-344 Liquid loading, 494 Liquid specific gravity, 504 Lithofacies, 263-268,282-284 Lithology
devices, 144-145 difficult, 186-189 and fossils, 231 on graphic log, 198-199 quick-look from logs, 174-179 Lithology-Iog response analysis, 316 Local area network, 447,448 Log analysis package (LAP), 441-446,449 Log probability paper, 25-26, 28 Logging data, preprocessing of, 167-169 Logging tools cased hole, 151-153 open hole, 144-149 operating limitations for, 150 responses in sedimentary
minerals, 190-191 Lognormality, 25-26,27 Logs, see Mudlogging; Well logs Long radius horizontal wells, 74 Lost circulation, 87 Lot and block descriptions, 7 Louisiana, overpressure
problems in, 80,81 Lubricator, 499
Macroscopic techniques, for reservoir quality, 276-277
Magnet, 83,84 Magnetics, 415-416 Magnetotellurics, 418 Maintenance phase, 4 Mapping surfaces, 294-295 Maps
computer generated, 426-430 computer-aided, 383-384,431-433 of depth, 382-383 of land, 5 preparation of base, 451-452 of reservoir properties, 534-535 restoration, 335 seismic interpretation, 380 of structure, 333-338 of subsurface, 294-299 time interval, 383,384 time slice, 383,384,387 of 2-dimensional seismic data, 381-384
velocity gradient, 381-382,383 Marine clastic deposits, 268 Marine seismic acquisition, 361-363 Master orientation line, 124 Material balance estimation
for gas, 515 for oil, 513,514-515 Math, wellsite, 93-97 Mathematical surface, 438 Mathematical variables tables of, 94,136 of volu metrics, 438 Matrix
acidizing of, 472,473 properties of, 538 in sandstone, 234-235 Maximum negative cash flow, 48 Mean, 25,27,29,339 Meandering channel sequence, 264 Meandering river deposits, 265-266 Measured depth, 289,291 Measurement while drilling, 88-90 Mechanical core orientation, 122-123,124 Mechanical failures, 494-495 Mechanical integrity logs, 488,489-491 Median, 25,27,339 Medium radius horizontal wells, 75 Menu-driven system, 448 Mercury injection system,
222-223,287-288 Mesoscopic techniques, for reservoir
quality, 276-277 Meters, on tanks, 484 Metes and bounds, 6-8 Microresistivity devices, 145 Microscopic techniques, for reservoir
quality, 276-277 MID crossplot, 186,188 Migration, of seismic
data, 365,372-376,388 Mill, 85 Millidarcy, 210 Minerals
identification of, 186,187,188, 190-191
as scale, 493 Minimum curvature method, 291,292 Miscible gas flooding, 528-529 Mist drilling, 78 Misties, 381,382 M-N crossplot, 186,187 Mobility ratio, 526 Modal analysis, 233-234 Mode, 25,27,339 Models
of AVO data, 400 of carbonate reservoirs, 535 cash flow, 38-42 computer algorithms, 431-440 geological reservoir, 315,319 of ideal well, 511 reef reservoir, 274 for reservoir simulation, 531-540 of seismic data, 392-394 of surface, 431-440 of waterflooding, 525-526 Money, time value of, 35-37
Monte Carlo method, 348,349 Moveout, 369,371 Mud pump output, 95 Mud system, 106-108 Mud tank level, and overpressure, 80 Mud weight, 92,93,112 Mudlogging
data sheet, 102 drill cuttings analysis, 104-105 equipment, 98-99 gas extraction/monitoring, 106-108 layoutof unit, 99 the mudlog, 101-103 personnel, 100 services, 99-100 standard form, 103 Muds, as drilling fluids, 77 Multi-phase flow properties, 537-538 Multiple completion, 464,465 Multiple linear regression, 216-217,343 Multiples, 369 Multi-point tests, 475-476 Multi-rate flow test, 477 Multi-rate production data, 509 Multivariate data analysis, 345-347 Multivariate regression, 344,345-347 Multiwell extension project, 41-42,46,48 Native-state analysis, 227 Naturally fractured reservoirs, 192-193 Navigation systems, 362 Nearest neighbor search
criterion, 428-429 Negative polarity, 377 Net cash flow, 38,49 Net pay, on maps, 295-296 Net present value, 35,40,47 Net revenue interest, 24 Network, local area, 447,448 Neutron-density logs, 177-179,
182-183,185 Neutron-porosity logs, 177 Noise log, 489 Nondispersed muds, 77 Nonflowing drill stem tests, 131 Nonhorizontal fluid contacts, 305-308 Nonpay, 286,288
Nonreservoirs, versus reservoirs, 287 Nonwetting, 221 Normal distribution, 340,342 Normal fault, arrow plot of, 161 Normal moveout, 369,371,399 Normal polarity, 377 Normalized rate of penetration, 92 Null hypothesis, 341 Odor, of hydrocarbons, 114 Offset VSP, 402-403 Offshore descriptions, 7-8 Offshore rigs, 65-66 Offshore safety, 68 Oil
analysis of, 241-246,507 classification of, 505-506 flow rate of, 462 material balance estimation
for, 513,514-515 mobility of, 529-530 Oil and gas contracts, 16
Oil and gas interests conveyance of, 9-12 ownership of, 9-12
Oil and gas lease, 9,13-15 Oil and gas taxes, 43-46 Oil saturation, on maps, 296 Oil-based muds, 77-78 Oilfield brines, 248 Oilfield water analysis, 247-248 Open hole completion, 463 Open hole tools, 144-149 Orientation, of core, 122-124 Orientation marks, 125 Orifice meter, 484 Original gas in place, 513 Original oil in place, 513 OSHA safety standards, 68 Outcrop analysis,
of fracture data, 328-329 Outer continental shelf, 15 Outliers, 345 Overbalanced perforating, 467 Overpressured reservoirs
causes of, 79-80 detection of, 80-82 Overshot, 85 Ownership, of property interests, 10-11 Oxidation, of pipe, 494 Packed bottom hole assembly, 73,74 Packer seats, 131 Packers, 132 Packing arrangements, for spheres, 207 Paleogeography, and fossils, 231 Paleomagnetic core orientation, 123 Paleontology, 229-232 Paraffin, 493 Partial penetration, 492,493 Pay determination, 286-288 Payout, 48 Pendulum bottom hole assembly, 71 Penetration, rate of, 91-92 Percentage depletion, 45 Percussion sidewall coring, 119 Perfect support fluids, 471 Perforated casing completion, 464 Perforating, 466-468 Perforation tunnel, 466 Performance prediction, 540 Permeability of cores, 203 directional, 526 due to fractures, 329 effective, 341 estimation of, 532 and flow, 283,539 of formation, 155-156 from FWAL, 410 on maps, 296-298 from multi-point tests, 475-476 and porosity, 215-217 pseudo-relative, 537 relative, 226-228,493 of reservoirs, 210-213 and rock-liquid reaction, 251 in tight gas reservoirs, 321-322,325 variation problems, 523,525 Perserva tion, of core, 128-130
Personal computers, 447 Personal safety equipment, 68-69 Personnel
mudlogging, 100 on a rig, 67 Petrographic analysis, 233-236,249 Petroleum landman, functions of, 4 Petrophysical analysis, and tight gas
reservoirs, 323 Petrotechnical Open Software
Corporation (POSC), 449 Phase diagram, pressure-
temperature, 506 Phase shifting, 377 Photoelectric effect, 145,148,
177,186,187 Pickett plots, 443,445 Pipe, corrosion of, 494 Pipe shear, 85,86 Piston pump, 486 Plastic, to preserve core, 129 Platforms, 65 Plug analysis, 201,202 Point bars, arrow plot of, 162 Polarity, 377
Polynomial regression, 344 Poor boy junk basket, 84 Pore geometry, 210-211 Pore types, 204-206 Pore volume, 206,209
and capillary pressure, 222-225 due to fractures, 330 Porosimeter, 207,209 Porosity, 172 of carbonates, 270 of cores, 203,214 estimation of, 180-183 on graphic log, 199,200 on maps, 296 and permeability, 211,215-217 and rate of penetration, 92 of reservoirs, 204-209 of sandstone, 235 and sedimentary texture, 204-205 in tight gas reservoirs, 325 tools of, 145-148 types of, 276 Porosity thickness, on maps, 297,298 Positive polarity, 377 Possum belly, 104,106 Poststaek migration, 369,375-376 Poststack mix, 369 Potentiometric elevation, 306-307 Practical field system, 510 Precipitates, 492-493 Predicted variable, 343 Predictions
accuracy of, 27 of performance, 540 Predictor variable, 343 Preparation phase, of interpretation, 451-453 Present value, 24,35-36 table of factors of, 37 Pressure bubblepoint, 504,509 capillary, 221-225
INDEX
547
confining, 208,211,227 contour map of, 157 detection of, 79-82 drawdown, 477 drill stem test, 475 by drive mechanisms, 518 of fluid flow, 493,508 in formation, 154-157 formation fracture, 93 hydrostatic, 93-94 increasing with depth, 80 loss at jet nozzle, 96-97 maps of, 294 in radial reservoir, 510 retrograde dewpoint, 504 static reservoir, 135 versus fluid flow rate, 510-511 Pressure buildup tests, 477,479,480 Pressure coring, 117 Pressure gauges, 132 Pressure transient testing, 477-481 Prestack migration, 375-376 Pretest formation tester, 155 Pretest pressure response, 156 Prices, uncertainty in, 27-29 Primary recovery factor, 513-517 Principal component analysis, 346 Probability
of commercial success, 32 of completion, 31-32 of failure, 32 of geological success, 30 and ranges, 25 Producing property, 24 Producing wells, 465-466 Production decline rate, 460,462,516,517 Production engineering methods, 457-499 Production histories, 460-462,481, 515-516,539 curves of, 516 Production logs, 488-491 Production problems, 492 Production taxes, 43 Production testing, 474-476,478-479,508 Production trends, by various drives, 518-521 Productivity, from fracturing, 472 Productivity index, 474,508 on maps, 297 Profit, 43,49 Profit or loss equation, 24 Programmed gain control, 378 Property descriptions of, 5-8 estimating value of, 54 evaluation of, 24 laws of, 9 ownership of, 10-11 Proppant placement, 471 Pseudo-relative permeability, 537 Pulse tests, 479-481 Pulsed neutron tools, 151-152,153 Pulse-echo cement bond log, 490,491 Pump output, 95 Purchase agreements, 18
548
INDEX
PVT properties, 504-507 Quadratic equation, 508-509 Quality control, in marine
seismic data, 363 Quality of reservoir, 275-277 Quantitative analysis, 315 Quick-look lithology, 174-179 Radial flow, 461,509-510 Radioactive tracer survey, 488-489 Radioactivity, of formation, 144 Radius of curvature method, 290, 292 Radius of drainage, 511 Radius of investigation, 133,136 Random functions, 349 Random points, 426 Ranger Formation, cross section of, 301 Rate history, 460-462,481 Rate of penetration, 91-92,101 Rate-related tests, 250-251 Ray path diagram, 406 Ray tracing, 392-393 Receivers, of seismic data, 359,361-362 Recompletion, 498 Recording system, 359 Recoverable reserves, 513 Recovery
by drives, 518-522 efficiency of, 223 thermal, 528-530 by waterflooding, 523-526 Recrystallization, and reservoir quality, 276 Rectangular gridding, 427-429 Rectangular survey system, 6 Reefs forward modeling of, 393 reservoir model of, 274 seismic inversion data of, 395,397 Reflection, of seismic data, 373 Reflection coefficient, 398 Regional geology framework, 314-315 Regression analysis, 343-344,345 Regulatory costs, 29 Relative amplitude preservation, 389 Relative permeability, 226-228,493,509 Remaining reserves, 54 Remanent magnetism, 415 Remote sensing, 414 Repeat formation tester, 506 Reserves
calculation of, 28,295-296 estimation of, 513-517 uncertainty in, 25-29 Reservoir capillary pressure of, 223-224 carbonate models of, 272-274 classification of, 505 conditions of, 479 definition of, 286 description of, 314-320 diagenetically complex, 314-320 flow units in, 282-285 fluids in, 504-507 and fossils, 231-232 heterogeneity of, 278-281,307-309,
311-320,531-532 management of, 296-298,311
naturally fractured, 192-193,326 overpressured, 79-82 performance of, 27 permeability of, 210-213 porosity of, 204-209 pressure of, 135 properties of, 534 quality of, 275-277,318,325 simulation of, 531-540 stratigraphically complex, 311-313 structurally complex, 331-338 subnormally pressured, 82 tight gas, 321-325 versus nonreservoir, 287 water in, 505
water saturation in, 183-185 zonation of, 319-320 Reservoir drawdown, 508 Reservoir engineering methods, 501-540 Reservoir rock, confidence in, 30 Residual fluid saturation, 202 Residual wavelet, 368 Resistivity true, 169 water, 170-173,183 Resistivity logs, 192 Resistivity ratio method, 172-173 Resistivity tools, 145,147 Resolution, of seismic data, 364-365,388-389 Resources, 513 Restored-state analysis, 227 Restrictions, near wellbore, 492-493 Retrievable hollow carrier gun, 466-467 Retrograde condensation, 505 Retrograde dewpoint pressure, 504 Revenue and costs, 24 uncertainty in, 25-29 Revenue interests, 10-11 Reverse polarity, 377 Rig personnel, 67 Rig safety equipment, 68 Rigs
land, 62-64 offshore, 65-66 Risk, reducing, 52-53 Risk aversion, 52-53 Risk-adjusted value, 52 River channel, seismic inversion data of, 396,397 Rock description, 263-265,315 Rock fabrics, of carbonates, 270-272 Rock typing, 532 Rock-log calibration, 200 Rock-water reactions, 249-252 Rodessa Formation lithofacies, 231 Rotary sidewall coring, 119-120 Rotating system, 62 Roundness, 200 Rubber sleeve core barrel, 117 "Rule of 72," 36 Safety
in drill stem testing, 131 wellsite, 68-70 Salinity rate of change, 251-252
and rock-water reactions, 249-250 Sample variance, 341 Samples
analysis of, 105 catching representative, 104 collection of, 233 of core, 201 examination area, 99 lag, 104 of oil/condensate, 241-242 packaging of, 105 paleomagnetic plug, 123 preparation of, 104-105,205,233 selection of, 226-227 Sampling techniques, 201 San Andreas Formation, 298-299 Sandstones see also Clastics analysis of, 234-235 cements of, 277 depositional environment
of, 265-268 interpretation of, 188 pore systems of, 204,205 porosity of, 276 Santa Barbara Channel, 374 Saran wrap, to preserve core, 129 Saturation and capillary pressure, 222-225 condition of, 504 fluid, 201-203 history of, 227-228 in oil well, 508 pressure of, 504 Scalar wave equation, 373 Scales, 492-493 Scanning electron microscopy (SEM), 237,238,239 SCAT plots, 332 Screen and liner completion, 464 Scribe knife, 123 Seal capacity, 224 Sealed trap, confidence in, 31 Sectionalized descriptions, 7 Sedimentary rocks, analysis of, 263-268 Sedimentary structures, 198-199 Seismic data attenuation of, 389 displaying, 377-378 forward modeling of, 392-394 interpretation of, 379-380 land acquisition of, 358-360 marine acquisition of, 361-363 migration of, 388 preparation of, 451-453 processing of, 364-369 resolution of, 388-389 three-dimensional, 385-387,396 two-dimensional, 381-384,385 Seismic detection, of overpressure, 80 Seismic energy sources,
358-359,361,404-405 Seismic inversion, 395-397 Seismic migration, 372-376 Seismic option agreements, 18 Seismic section, 366 Seismic surveys, tomographic, 405-406
INDEX
549
Seismograms, synthetic, 382,390-391 Self-potential method, 418 Semi-expendable gun, 466 Semi-log decline history, 460 Semipermeable barriers, and fluid
contacts, 308,309 Semi-steady rates, 461 Semisubmersibles, 66 Semivariogram, 430 Separator, 482-483 Services, mudlogging, 99-100 Shale resistivity, 81 Shale shaker, 104 Shale volume, 180,181 Shallow marine clastic deposits, 268 Shaped charge, 466,467 Ships, seismic, 362 Shoreline deposits, 268 Short radius horizontal wells, 75 Shot holes, 358
Shot patterns, dynamite, 358,360 Shot record, single, 365,367,368,370-371 Shotpoints, 358 Show evaluation, 109-114
form, 110-111 Sidetracking, 73 Sidetrak Coring System, 119,120 Sidewall core analysis, 201,202,203 Sidewall coring, 119-121 Siliciclastics, see Clastics; Sandstones Simandoux equation, 184 Simple linear regression, 343 Simulation
effect on IPR, 511,512 of reservoirs, 531-540 Single completion, 464,466 Single shot record, 365,367,368,370-371 Single-point test, 474-475 Single-rate flow tests, 477 Six-arm dipmeter, 159 Skin effect, 136,510 Sleeve gun, 361 Slim hole completion, 465 Slotted linear completion, 464 Slowness, of Stoneley wave, 410 Snubbing units, 499 Soft sediment, handling of, 125-126 Solution gas drive, 518-519 Solution gas-oil ratio, 297,504 Solution oil-gas ratio, 504 Solvent, 528
Sonic amplitude logs, 192 Sonic logs, 192
porosity from, 182 Sonic tools, 148 Sorting
classes of, 200 and permeability, 211 Source rock, and fossils, 232 Source-receiver geometry, 401-403 SP, see Spontaneous potential logs Sparse-spike method, 395 Spectral gamma ray, 144,145, 151,152,175 Spherical separator, 483-484 Sphericity, 200 Split spread, 365
Sponge-lined coring system, 116,117 Spontaneous potential baseline
flattening, 442 Spontaneous potential logs, 170-172,
175-176,193 shale volume from, 180 Spontaneous potential tools, 144 Squash plots, 377 Squeeze cementing, 497 Stacking, 364 Staining, of samples, 113 Standard deviation, 340-341,342 Standard interpretation, 180 Standard normal distribution, 342 Standards, in computer
industry, 449-450 State lands, ownership of, 10 Static SP, 170 Statics corrections, 368-369,370 Statistical methods, in gridding, 429-430 Statistics, in geological data, 339-349 Steady-state method, for
permeability, 212,213,226 Steady-state rates, 461 Steam flooding, 530 Stiff diagrams, 248 Stimulation, 469-473 Stochastic simulation, 348-349 Stoneley wave, 409 Straight hole, 71 Strategy, in bidding, 54 Stratigraphic cross sections, 300-302
in tight gas reservoirs, 322-323 Stratigraphic thickness, 290-293 Stratigraphically complex fields, 311-313 Streamer cable, 361 String, 499
Structural cross section, 302-303 Structural dip, 160 Structurally complex reservoirs, 331-338 Structure
maps of, 294,295,333-338 style of, 331 Stuck pipe, 85 Student's t test, 341-342 Stylolite, Formation MicroScanner image of, 166 Subcrop lines, 434 Submersible pump, 485 Submersibles, 65 Subnormally pressured reservoirs, 82 Subsurface data, from well logs, 289-293 Subsurface maps, 294-299 Subtidal-supratidal dolomitization reservoir model, 272,273 Success, probability of, 31-32 Summation of fluids method, 201-202, 208 Support fluids, 471 Surface models, by computers, 431-440 Surface production equipment, 482-484 Swabbing, 476 Swanson's Rule, 26,27 Sweep efficiency, 521,523,525,526 Syledis, 362 Symmetrical data set, 340 Synthetic seismograms, 382,390-391,393
Tadpole plot, 160,161 Tangential angle method, 289,291 Tank batteries, 484 Tank metering, 484 Tapered mill, 85 Tax Reform Act, 44 Taxes, 29,43-46 Telemetry system, 359 Temperature, and permeability, 227,252 Temperature survey, 488 Terminal angle method, 289 Test tools, 131-132 Testing
pressure transient, 477-481 production, 474-476,478-479 Texture analysis of, 234 maturity of, 200 and porosity, 204-205 Thermal recovery, 528-530 Thickness on cross section, 300-302 on graphic log, 198-199 hydrocarbon, 438-439 mapping of, 294-295 stratigraphic, 290-293 Thin bed resolution, 158-159 Thin section analysis, 233-236 Three-dimensional seismic data, 385-387,396 Thrust ramp structure, 334 Thru-tubing gun, 467 Tight gas reservoirs, 321-325 Tilted fluid contacts, 305-308 Time interval maps, 383,384 Time lapse tomograms, 406 Time migration, 375 Time slice displays, 397 Time slice maps, 383,384,387 Time value, of money, 35-37 Title curative, 12 Title opinions, 11-12 Tomogram, 404,406 Tomography, cross-borehole, 404-408 Tool string, 132,133 Tool table, 150 Tools
of FWA L, 409 open hole, 144-149 for surface modeling, 440 in tight gas reservoirs, 322-323 Total porosity, 204 Townships, numbering of, 6 Trace, 377 Traceplots, 443,444,445, see also Well logs Tracer velocity shot technique, 489 Transfer function, 348 Transition zone, 305 anomalously thick, 309-310 Travel paths, 402 Trend surface analysis, 429 Triangular mesh, 427 Triangulation, 427 Triple completion, 465 True hypothesis, 341 True resistivity, 169
550
INDEX
True stratigraphic thickness, 291-293 True vertical depth, 289-291 True vertical thickness, 292-293 Truncation, 433-434 Tube wave, 409
Tubing-conveyed gun, 467,468 Tuning effect, 388 Two-dimensional seismic
data, 381-384,385 Two-way traveltime, mapping of, 381 Type curve, 460-462,478, 479 Ultrasonic reflection, 149,163-165 Uncertainty
in economics, 25-29 in permeability, 213,217 Unconformity Formation MicroScanner image of, 166 maps of, 294 Underbalanced perforating, 467,468 Undersaturated condition, 504,508 Undiscounted profit to investment ratio, 49 Unsteady-state method, for permeability, 212,213,226 Unsteady-state rates, 461 Upward-shoaling reservoir model, 272,273 USBM wettability method, 219,220 User friendly, 448 User interface, 448 Variability, 340 Variable area wiggle trace, 377,378 Variable density display, 377,378 Variable density log, 490 Variable intensity logs, 192 Variables, dimensionless, 461 Variance, 340 Velocity, and seismic migration, 369,375 Velocity gradient maps, 381-382,383 Velocity shot technique, 489 Vertical exaggeration, 302,303 Vertical heater treater, 484 Vertical resolution, 388 Vertical seismic profile, 402-403 Vertical seismic profiles, 382 Vertical separation,
surface modeling of, 436 Vertical separator, 483 Vertical sweep efficiency, 525,526 Vessels, seismic, 362 Vibrator truck, 358,360 Vibrators, 359
Viscosity, of drilling fluid, 76 Volatile oil, 505 Volume
calculation of, 437-440 of shale, 180,181 of wellbore, 94-95
Volumetric estimation, of reserves, 513-514
Volumetrics algorithm, 437-439 VSP data, 402-403 Vuggy porosity, 270 Washover pipe, 85 Water
produced with oil, 247-248 reactions with rock, 249-252 in reservoirs, 505 Water cut, on maps, 298 Water drive, 518,520 Water gun, 361 Water pH, 252 Water resistivity, 170-173,183 Water saturation, 183-185, 201-203,225 on maps, 296,297 Water-based fluids, 76-77 Waterflooding, 523-526 Water-gas coning, 492 Water-oil ratio, 521 Wave coning, 492 Wave equation methods, 394 Wave types, in FWAL, 409 Wavelet processing, 388 Wax, to preserve core, 129 Wedge pinchout, 389 Weight
of collars or casing, 95 of mud, 93 Weight material, 76 Weighted least squares minimization, 187-188 Well completions of, 463-468 components of, 496 correlation of, 532 course of, 289-293 cross section of, 497 deviation of, 292 fluid contacts in, 306 model of, 511 producing, 465-466 stimulation of, 469-473 support agreements, 17 surveillance of, 462 trades of, 16 Well logs
analysis of, 441-446 BHGM, 413 and cross sections, 300-302 data conversion, 289-293 display of data, 443-445 facies calibration of, 534 of fractured reservoirs, 326-328 full waveform acoustic, 409-410 measurement while drilling, 89,90 and overpressure, 81 production, 488-491 rate of penetration, 91-92
shape of, 174-175 SP, 170-171 and synthetic seismograms, 391 typical examples of, 146,153 Well mechanical integrity logs, 488,489-491 Well planning, checklist for, 60-61 Wellbore directional, 71-73 flushing of, 112 horizontal, 73-74 restrictions near, 492-493 skin effect in, 510 storage capacity of, 135 straight, 71 trajectory of, 71 -75 volumes of, 94-95 workover on, 496-497 Wellbore scale heterogeneity, 278 Wellbore storage constant, 133-135 Wellhead, 482 Wellhead taxes, 43 Wellsite
core alteration at, 127-128 math at, 93-97 methods of, 57-137 safety at, 68-70 show evaluation at, 109-114 Wet core preservation, 130 Wet cut sample, 105 Wet gas, 505 Wettability, 218-220, 227-228 and reservoir quality, 276 Wetting, 221 Whipstock method, 72 Whole core analysis, 201 Wiggle trace, 377,378 Wilmington anticline,
cross section of, 302 Winner's curse, 54-55 Wireline core barrel, 117 Wireline formation tester, 154-157 Wireline logs
of rock fabric, 272 and sandstone correlation, 264 Wireline methods, 141-193,264 Wireline spear, 85, 86 Wireline survey, 488-491 Wireline units, 498 Withdrawal efficiency, 223 Workovers, 496-499 Workstation
for development geology, 447-450 interpretation projects in, 451-454 X-ray diffraction (XRD), 237,238,239 X-ray fluoroscopy (XF), 237,238,240 Zero-offset, 372,373 Zero-offset VSP, 402-403 Zero-phase wavelet, 388 Zonation, of reservoirs, 319-320