Compositional Mechanical And Transport Properties Of Carbonate Fault Rocks And The Seismic Cycle In Limestone Terrains A Case Study Of Surface Exposures On The Longmenshan Fault Sichuan China PDF Download

This body of work seeks to understand the development of fractures and normal faults in outcrop exposures of Cretaceous carbonate rocks throughout Texas and to assess their impact on subsurface hydrocarbon flow through three different projects. The first is a field-based investigation into the distribution of fractures in Cretaceous carbonate rocks and their role in normal fault formation. The second two efforts combine field investigations with fluid flow simulations. One of these combined studies assesses the impact of fractures and faults on secondary hydrocarbon migration in the Anacacho Limestone; the other assesses the impact of a particular fracture pattern on the Austin Chalk. The field-based investigation into fractures and fault formation focuses on the stratigraphic partitioning of fractures that formed through different mechanisms and how those fractures contributed to the formation of normal faults in sequences of carbonate formations. We find that the primary failure modes in these sequences of carbonates include opening-mode brittle failure, evidenced through the presence of joints and veins, and closing-mode ductile failure through pressure solution, evidenced by discrete solution surfaces (solution seams). The composition and texture of the examined carbonate rocks influences their primary mode of deformation. Matrix supported carbonate rocks failed primarily through pressure solution and now contain solution seams, a flaser rock fabric, and moderately dipping (~40 to 50°) fractures formed through networks of pressure solution seams. While in crystalline dolostones and also grain supported lithologies, joints and veins are most prevalent, indicating they are prone to opening mode brittle failure. The stratigraphic partitioning of failure modes influences the geometries and architectures of normal faults, since they appear to form through the shearing and coalescence of originally bed-confined pressure solution seams, joints, and veins. Normal fault segments exhibit a high angle (~80 to 90°) to bedding when crossing beds characterized by brittle failure (which contain predominantly joints and veins). Yet while crossing beds characterized by closing mode failure (which contain predominantly solution seams) normal fault segments exhibit moderate angles to bedding (~40 to 50°). The development of these faults localizes shearing-related fracture opening in the form of pull-aparts in brittle beds of these carbonate formations. This localized fracture opening may impact subsurface fluid flow. Insights from that field based research effort helped shape our approach when undertaking a multidisciplinary investigation of asphalt distribution in the Anacacho Limestone. In this investigation, field relationships between fractures, faults, and asphalt presence are evaluated at an open pit asphaltic limestone mine near Uvalde, TX. Based upon their distributions and geometries, we infer that relatively larger normal faults provided vertical flow paths through the Anacacho Limestone while strata-bound fractures enhanced the horizontal permeability of the formation. Directional variograms calculated from 75 subsurface measurements surrounding the mine indicate that the asphalt concentration is anisotropically correlated and that the maximum correlation length points in a similar orientation as the ~52° northeast mean strike of the fractures and faults. A globally-positioned laser rangefinder is used to measure faults and stratigraphic contacts within the mine. That data is then combined with lithologic descriptions from surrounding subsurface wells to construct a digital three dimensional model (3D) of the Anacacho Limestone. When the model is populated with asphalt concentration estimated with an ordinary block kriging algorithm, we find that the two largest normal fault zones qualitatively align within the central axis of an isosurface enclosing high values, thereby providing a positive correlation between the location and orientation of the normal faults and the asphalt concentration. The three dimensional model also provides a framework to numerically simulate secondary hydrocarbon migration. The simulation parameters are adjusted within physically realistic ranges to produce an oil saturation field in agreement with asphalt concentration estimates. Our results indicate that oil entered the Anacacho Limestone through normal faults, that oil flow was impacted by regional aquifer flow, and that fractures increased the horizontal permeability of the formation by an order of magnitude along their strike direction. Following that investigation, we develop a methodology to incorporate field-observed fracture networks into discrete fracture model fluid flow simulations. Here, we explore two methods to extract the 3D positions of natural fractures from a Light Detection and Ranging (LiDAR) survey collected at a roadcut through the heavily jointed Cretaceous Austin Chalk: (1) a manual method using the UC Davis Keck Center for Active Visualization in the Earth Sciences and (2), a semi-automated method based upon Gaussian and mean curvature surface classification. Each extraction method captures the characteristic frequency and orientation of the primary fracture sets identified in the field, although they have varying abilities to extract the secondary fracture sets. After making assumptions regarding fracture length and apertures, the extracted fractures served as a basis to construct a discrete fracture network (DFN) that agrees with field observations and a priori knowledge of fracture network systems. Using this DFN, we performed flow simulations for two hypothetical scenarios, with and without secondary fracture sets. The results of these two scenarios indicate that, for this particular fracture network, secondary fracture sets have little impact (~10% change) on the breakthrough time of water injected into an oil filled reservoir. Our work provides a prototype workflow that links outcrop fracture observations to 3D DFN model flow simulations using LiDAR data, possibly providing an improvement over traditional field-based DFN constructions. Moreover, the fracture extraction techniques may prove applicable to other LiDAR based outcrop studies.