AbstractPore pressure exists ubiquitously in the Earth’s subsurface and very often exhibits acyclic loading on pre-existing faults due to seasonal and tidal changes, as well as theimpoundment and discharge of surface reservoirs. The effect of oscillating pore pressure oninduced seismicity is not fully understood. This effect exhibits a dynamic variation ineffective stresses in space and time. The redistribution of pore pressure as a result of fluidflow and pressure oscillations can cause spatial and temporal changes in the shear strength offault zones, which may result in delayed and protracted slips on pre-existing fractures.
This research uses an experimental approach to investigate the effects of oscillating pore pressure on induced seismicity. With the aid of geophysical techniques, the spatial and temporal distribution of seismic events was reconstructed and analysed. Triaxial experiments were conducted on two types of sandstone, one with low permeability (Fontainebleau sandstone) and the other with high permeability (Darley Dale sandstone). Cyclic pore pressures were applied to the naturally-fractured samples to activate and reactivate the existing faults. The results indicate that the mechanical properties of the sample and the heterogeneity of the fault zone can influence the seismic response. Initial seismicity was induced by applying pore pressures that exceeded the previous maximum attained during the experiment. The reactivation of faults and foreshock sequences was found in the Fontainebleau sandstone experiment, a finding which indicates that oscillating pore pressure can induce seismicity for a longer period of time than a single-step increase in pore pressure. The corresponding strain change due to cyclic pore pressure changes suggests that progressive shearing occurred during the pore pressure cycles. This shearing progressively damaged the existing fault through the wearing of asperities, which in turn reduced the friction coefficient and, hence, reduced the shear strength of the fault. This ‘slow’ seismic mechanism contributed to the prolonged period of seismicity. This study also applied a material forecast model for the estimation of time-to-failure or peak seismicity in reservoir-induced seismicity, which may provide some general guidelines for short-term field case estimations.
|Date of Award||2010|
|Supervisor||Philip Benson (Supervisor)|