Fault Roughness and Induced Seismicity: Hydro-Shearing of Laboratory Faults

Abstract We explore induced seismicity in smooth and rough laboratory faults, by investigating the dynamic interplay between pore pressure distribution and fault slip, focusing on the relationship between overpressure ratio (OPR) which scales the ratio between experimental and theoretical critical pore pressure required to induce slip across a fault, and its roughness. We find discrepancies between rocks with different permeabilities. High permeability rocks (e.g. Berea sandstone) allow for a more homogeneous fluid distribution across the fault surface, thus generally maintaining an OPR close to 1.0. In low permeability rocks (e.g. Timna granite), heterogeneous pore pressure distribution across the fault surface prevails, triggering higher OPR values. Considering Timna granite where both smooth and rough faults were tested, we find that the smooth faults exhibit an OPR ≫ 1.0 whereas the rough fault surfaces exhibit an OPR ~ 1.0. We attribute this difference to the pore pressure distribution across the fault plane, where in the smoother faults induced pore pressure tends to concentrate near the injection borehole and thus flows slower across the fault plane, whereas in the case of rough fault planes the asperities in the fault aperture allow for faster diffusion and more diverse flow paths, thus generating a more homogeneous pore pressure distribution across the fault plane resulting in an OPR ~ 1.0. We compare between injection-driven and displacement-driven hydro-shearing tests. In injection driven modes, fault activation is characterized by gradual slip whereas fault activation under displacement driven modes is characterized by rapid slip. Finally, under undrained conditions, we find that slip velocities are enhanced, possibly due to suction effect during suppressed dilation.