We use a two-dimensional plane strain dynamic rupture model with strongly rate-weakening friction and off-fault Drucker-Prager plasticity to model damage zones associated with buried second-order thrust faults observed in the SSC reservoir. The modeling of ruptures propagating as self-sustaining pulses is performed in the framework of continuum plasticity where the plasticity formulation includes both deviatoric and volumetric plastic strains. The material deforming inelastically due to stress perturbations generated by the propagating rupture is assumed to be the damage zone associated with the fault. Dilatant plastic strains are converted into a fracture population by assuming that the dilatant plastic strain is manifested in the form of fractures. The cumulative effect of multiple slip events is considered by superposition of the plastic strain field obtained from individual slip events. The relative number of various magnitude slip events is chosen so as to honor the Gutenberg-Richter law. Results show that the decay of fracture density (F) with distance (r) from the fault can be described by a power law F = F 0 r À n . The fault constant F 0 represents the fracture density at unit distance from the fault. The decay rate (n) in fracture density is approximately 0.85 close to the fault and increases to~1.4 at larger distances (>10 m). Modeled damage zones are approximately 60-100 m wide. These attributes are similar to those observed in the SSC reservoir using wellbore image logs and those reported in outcrop studies. Considering fault roughness affects local damage zone characteristics, these characteristics are similar to those modeled around planar faults at a scale (~10 m) that affects bulk fluid-flow properties.

We analyze fracture-density variations in subsurface fault-damage zones in two distinct geologic environments, adjacent to faults in the granitic SSC reservoir and adjacent to faults in arkosic sandstones near the San Andreas fault in central California. These damage zones are similar in terms of width, peak fracture or fault (FF) density, and the rate of FF density decay with distance from the main fault. Seismic images from the SSC reservoir exhibit a large basement master fault associated with 27 seismically resolvable second-order faults. A maximum of 5 to 6 FF/m (1.5 to 1.8 FF/ft) are observed in the 50 to 80 m (164 to 262 ft) wide damage zones associated with second-order faults that are identified in image logs from four wells. Damage zones associated with second-order faults immediately southwest of the San Andreas Fault are also interpreted using image logs from the San Andreas Fault Observatory at Depth (SAFOD) borehole. These damage zones are also 50-80 m wide (164 to 262 ft) with peak FF density of 2.5 to 6 FF/m (0.8 to 1.8 FF/ft). The FF density in damage zones observed in both the study areas is found to decay with distance according to a power law F = F 0 r −n. The fault constant F 0 is the FF density at unit distance from the fault, which is about 10-30 FF/m (3.1-9.1 FF/ft) in the SSC reservoir and 6-17 FF/m (1.8-5.2 FF/ft) in the arkose. The decay rate n ranges from 0.68 to 1.06 in the SSC reservoir, and from 0.4 to 0.75 in the arkosic section. This quantification of damage-zone attributes can facilitate the incorporation of the geometry and properties of damage zones in reservoir flow simulation models. INTRODUCTION AND MOTIVATION Field observations of relatively large-scale faults and damage zones frequently show that fault zones consist of a fault core

The URTeC Technical Program Committee accepted this presentation on the basis of information contained in an abstract submitted by the author(s). The contents of this paper have not been reviewed by URTeC and URTeC does not warrant the accuracy, reliability, or timeliness of any information herein. All information is the responsibility of, and, is subject to corrections by the author(s). Any person or entity that relies on any information obtained from this paper does so at their own risk. The information herein does not necessarily reflect any position of URTeC. Any reproduction, distribution, or storage of any part of this paper without the written consent of URTeC is prohibited.

We predict pressure and stress in the 3D Mad Dog field using the Full Effective Stress (FES) pressure-prediction workflow. The FES workflow incorporates the full stress tensor (e.g., lateral stress and deviatoric stresses) into pressure prediction: it uses a geomechanical model to predict mean total and shear stresses in the 3D field and a relationship between velocity and equivalent effective stress (instead of vertical effective stress) to account for both mean- and shear-induced pore pressure generation. In complex geologic settings, such as salt basins or thrust belts, compaction depends on non-vertical and differential stresses; in such settings, the FES method offers a significant improvement over the traditional approach, that is based on the vertical effective stress. We focus our study on the anticline below the Mad Dog salt at the original platform area. We quantify the mean and shear-induced overpressures and show that shear-induced pressures account for 80% of the total overpressure in front of the salt. We also show that shear-induced pressures are the source of more than 1.5ppg overpressure in the anticline below salt, where the mean-stress approach alone predicts underpressures (less than hydrostatic). Higher pressures and the decrease in lateral stress in the anticline area lead to a 1ppg drilling window (defined in this paper as the difference between the pore pressure and minimum principal stress at any given depth). This drilling window is shifted to higher overpressures by 0.4ppg compared to the VES prediction. We find that the stress ratio in the mudrocks decreases to ~55% of its uniaxial value. Furthermore, we show that the velocity-informed geomechanical model is able to predict the pore pressure regression observed at Mad Dog and the regional hydraulic connectivity in the area. The three-dimansional (3D) geomechanical model is built in Horizon (Elfen). The known pressure regression in the sands is modeled; mudrock pore pressures are initialized using the VES estimate. Modified Cam Clay is used to quantify mean- and shear-induced compaction. Overall, we demonstrate that incorporating the full stress tensor is important for pressure and stress prediction at Mad Dog, and that the FES method, by providing both pressure and stress, can help improve drilling-window estimates.