Modelling a strike-slip fault system affecting porous carbonates in Favignana Island (Sicily, southern Italy)

Investigating the deformation processes as well as the characteristics and distribution of their end-products is a crucial issue to improve geo-fluid exploitation in carbonate reservoirs (≈50% of natural geo-fluids). Indeed, besides the primary controls on the petrophysical properties of limestones, which are due to nature and organization/shape of the constituent elements (i.e. grains, pores, cement, clay minerals), both containment and migration of fluids in these rocks are influenced by fault zones and fractures. In this contribution we integrate quantitative structural analysis and numerical modelling approaches aiming at testing a new workflow useful to create a 3D discrete fracture network (DFN) model of a reservoir starting from outcrop data collected in Favignana Island (Sicily, southern Italy). The presence of several quarries in the Island provides 3D exposures of ≈25 m-thick Lower-Pleistocene high-porosity grainstones crosscut by two conjugate sets of strike-slip faults. This fault system, documented by Tondi et al. (2012), is comprised of three types of structure: single compactive shear bands (CSB); zones of bands (ZB); and, faults. CSBs are narrow tabular features with porosity less than the surrounding host rocks, and have thicknesses and displacements on the order of a few mm. The growth process for these structures involves localizing further deformation within zones of closely-spaced CSBs and, possibly, along continuous slip surfaces within fault rocks overprinting older ZBs. The transitions from one growth step to another are recorded by different values of the dimensional parameters (i.e. length, thickness and displacement) for the structures. These transitions are also reflected by the ratios and distributions of the dimensional parameters. The DFN model was built by means of the Fracture Modelling module of the commercial software package Move from Midland Valley©. The analysis of an aerial photo was performed firstly to delimit the outcrop perimeter and geometry, needed to construct the model, and then to identify the major faults. The intensity of CSBs and ZBs, was calculated by integrating the distribution laws of the dimensional parameters (Tondi et al., 2012) with the lineament analysis tool of Move. The variation in intensity of deformation was used to guide the modelling of deformed zones, both within the fault zones and in the host rock in order to obtain a DFN reflecting a deformation pattern similar to the natural one. The DFN was then used to model the effect of deformation on the permeability of the host rock. Here, unlike in tight rocks where deformation generally increases permeability, the CSBs and ZBs reduce permeability whilst slip surfaces within the fault cores enhance fault-parallel fluid flow. Thus, by varying the porosity values attributed to CSBs, ZBs and slip surfaces the resulting DFN model was imposed to have a reduced permeability in CSBs and ZBs (relative to the host rock and the slip surfaces). This semi-automated process of lineament analysis followed by the use of power law distributions to model sub-seismic scale CSBs and ZBs is proposed as a potential modelling route for reservoir scale assessment of structural deformation controls on permeability in porous carbonate reservoirs.