Description:

_Journal of Structural Geology 32 (2010) 1212–1230_ _Contents lists available at ScienceDirect_ _Journal of Structural Geology_ _journal homepage: www.elsevier.com locate jsg _How fracture systems affect permeability development in shallow-water carbonate rocks: An example from the Gargano Peninsula, Italy_ _B. Larsen a,*, I. Grunnaleite b, A. Gudmundsson c,1_ _a Department of Earth Science, The University of Bergen, Allegaten 41, N-5007 Bergen, Norway_ _b International Research Institute of Stavanger, Bergen, Norway_ _c Geoscience Centre, University of Go¨ttingen, Germany_ _article info_ _Article history: Received 10 December 2007 Received in revised form 6 May 2009 Accepted 8 May 2009 Available online 21 May 2009_ _Keywords: Fractures Stress fields Permeability Carbonate rocks Reservoirs Numerical modelling_ _abstract_ _Fracture networks control the permeability of many reservoirs. Since the fracture patterns of petroleum reservoirs in situ are difficult to study in detail, field analogues are very important for understanding their fracture-related permeability. Here we present the results of a study of the fracture system of carbonate rocks of Lower Cretaceous age in a quarry associated with the damage zone and fault core of a major fault zone on the Gargano Peninsula in South Italy. We measured the attitude of 1541 fractures and faults along several vertical and horizontal scan lines. There are two main fracture sets: one strikes between E–W and ESE–WNW, the other NNE–SSW. A total of 675 fracture-spacing measurements indicate log-normal spacing distributions, with an arithmetic mean fracture spacing of 0.29 m and a median of 0.15 m. The data, plotted on a log-log plot, suggest three main spacing subpopulations, each of which follows approximately a power law with different fractal dimensions. Subpopulation 1, where the spacing ranges from 1 to 10 cm and the straight-line slope D (вЂ?fractal dimension’) is 0.20, represents fractures confined to laminated carbonate mudstones (multilayers) that form the microbial mat deposits of a peritidal cycle. Subpopulation 2, where the spacing ranges from 11 to 55 cm and D is 0.77, represents fractures confined to thicker layers, forming a part of a peritidal cycle, the contacts of which are marked by stylolites. Subpopulation 3, where the spacing ranges from 56 to 243 cm and D is 2.81, represents fractures that dissect comparatively thick units of an entire peritidal cycle. For the spacing, the minimum coefficient of variation, Cv, defined as standard deviation divided by the mean, is 1.00 (essentially randomly spaced fractures) while its maximum Cv is 1.62, suggesting that some fractures form clusters, some clusters being denser than others. The clusters, composed of fractures with varying attitudes and therefore commonly intersecting, are likely to contribute significantly to the overall permeability of the carbonate rock. Fracture-aperture (opening) data (N Вј 324) also show a log-normal size distribution, with a mean opening of 1.01 cm and median of 0.29 cm. Log-log plots indicate that a part of this data groups into two subpopulations, I and II, each of which follows approximately a power law. The straight-line slope D (вЂ?the fractal dimension’) of subpopulation I is 0.46, whereas that of subpopulation II is 1.49. We present boundary-element models showing that laminated carbonate mudstones and their contacts modify the local stress fields so as to encourage fracture offset and, commonly, arrest. Our results also show that when a fluid-driven subpopulation 2 fracture approaches subpopulation 1 fractures, the induced tensile stresses may result in the opening up of many of the subpopulation 1 fractures directly above the tip of the subpopulation 2 fractures. If, in addition, the contacts between the multilayers are weak, they also tend to open up, thus generating a large interconnected cluster of vertical fractures and horizontal contacts. The results suggest that the tensile stresses induced by a comparatively large fluid-driven subpopulation 2 fracture may contribute to the formation of an interconnected cluster of subpopulation 1 fractures and associated contacts, thereby significantly increasing the permeability of the carbonate rock._ _Г“ 2009 Elsevier Ltd. All rights reserved._ _* Corresponding author. E-mail addresses: belinda.larsen@geo.uib.no (B. Larsen), ivar.grunnaleite@iris.no (I. Grunnaleite), a.gudmundsson@gl.rhul.ac.uk (A. Gudmundsson). 1 Present address: Department of Earth Sciences, Queen’s Building, Royal Holloway University of London, Egham TW20 0EX, UK._ _0191-8141 $ – see front matter Г“ 2009 Elsevier Ltd. All rights reserved. doi:10.1016 j.jsg.2009.05.009_ _1. Introduction_ _The hydromechanical properties of carbonate rocks are very variable. This is partly because carbonate rocks have different depositional patterns and, in many cases, heterogeneous fabrics._ _B. Larsen et al. Journal of Structural Geology 32 (2010) 1212–1230_ _1213_ _However, the variability in properties is partly due to the rocks having undergone extensive alteration after deposition which may have changed their original matrix porosity and permeability. In a subsurface reservoir with a relatively low porosity-related permeability, fractures (joints and faults) are commonly the main paths for fluid transport. Detailed studies of fracture patterns in outcropping surface analogues of such reservoirs are then needed to improve our understanding of fracture-controlled permeability in subsurface carbonate reservoirs._ _Several factors control the fracture distribution in a carbonate reservoir. These factors include rock lithology and diagenesis (e.g., stylolites), mechanical properties and layering, and the current and earlier local stress fields. Stylolites are seams of clay residue that form as a result of compaction through the mechanism of pressure solution. Pressure solution, in turn, is a diffusion-related, ductile deformation mechanism where material in areas of high stress, such as surfaces orientated perpendicular to the maximum principal compressive stress, is dissolved and transported to areas of low stress, such as surfaces orientated perpendicular to the minimum principal compressive stress (Van der Pluijm and Marshak, 2004). Stylolites are most commonly found in limestones where as much as 40% of the original sequence thickness may have been dissolved through stylolitisation. In rocks with considerable clay content, the stylolites are generally wavy, whereas where the clay content is less than about 10% the stylolites become irregular sutures that appear tooth-like in cross-section (Van der Pluijm and Marshak, 2004). While many authors distinguish between stylolites and dissolution seams, the former being a serrate (tooth-like, saw-like) surface and the latter a non-serrate surface (Karcz and Scholz, 2003; Stow, 2005), here we do not make that distinction and use the term вЂ?stylolite’ for both._ _Fractures in layered rocks are commonly stratabound, meaning that they are confined to individual layers so that the fracture height (dip dimension) is less than or equal to the layer thickness (e.g., Bai et al., 2000). Here we refer to such a layer as a вЂ?mechanical layer’. Mechanical layering has great influence on a variety of geological processes, in particular fracture propagation and arrest (Gudmundsson and Brenner, 2004; Gudmundsson, 2006). Stress rotation, for example, is common at the contacts between mechanical layers, resulting in fracture offset or arrest (Gudmundsson, 2006). Similarly, fracture spacing is often proportional to layer thickness (Price, 1966; Huang and Angelier, 1989; Narr and Suppe, 1991; Bai et al., 2000), although there are many known examples where fracture spacing in sedimentary rocks greatly differs from layer thickness (e.g., Wennberg et al., 2006; Odonne et al., 2007). By contrast, non-stratabound fractures, by definition, propagate through more than one layer and their spacing is thus not proportional to layer thickness (Odling et al., 1999)._ _Fracture spacing affects the mechanical properties of rocks and their permeability. The details of the processes and factors that control fracture spacing are, however, still poorly understood. There are many proposed explanations for fracture spacing (Hobbs, 1967; Gross, 1993). For example, Wu and Pollard (1995) studied the stress distribution between two adjacent fractures in a 3-layer model and varied the spacing-to-layer thickness ratio (S Tf). They found that at a critical S Tf value of 0.976 the stress perpendicular to the fractures changed from tensile to compressive between adjacent fractures, thereby suppressing the formation of further tensile fractures between the two. For S Tf values lower than critical, only flaws perturbing the local stress field (or hydrofracturing) may cause the initiation of new fractures. However, Li and Yang (2007), using a similar 3-layer model, concluded that tensile stresses remain at the layer boundaries when the S Tf ratio decreases, and these stresses can initiate fractures even though the S Tf ratio decreases below a вЂ?critical’ value._ _Fracture spacing thus relates to fracture initiation, propagation and arrest, and all these factors affect the permeability of a fractured reservoir. These factors also depend on local stresses, which, in turn, are a function of the mechanical properties of the rock layers, in particular their Young’s moduli and Poisson’s ratios. Mechanical properties and local stresses thus control how easily fractures link to form continuous networks, which to a large extent determine the reservoir permeability. The reservoir percolation threshold is reached when a set of fractures forms an interconnected network. Whether or not this threshold is reached depends on the proportion of arrested, offset, and stratabound fractures as well as fracture attitude._ Ключевые слова: e, r, o