[First Hit]

AAPG Annual Convention and Exhibition

Datapages, Inc.Print this page

Fluid Injection and Earthquake Size in Faulted Reservoirs


The phenomenon of induced seismicity due to fluid injection has been recognized and understood for many decades, which for instance resulted in cancellation of at least two geothermal projects in the past. However, unprecedented increase in seismicity in the U.S. mid-continent associated with oil and gas activities calls for a better understanding of the physical processes that controls seismicity and mitigates the associated risk. In this research, we use a numerical model to investigate the physical processes associated with injection-induced seismicity. We couple a three dimensional (3D) geomechanical model to a 3D reservoir simulator to model the nucleation, progressive growth, and spreading of an induced slip event on a fault and then determine the resulting Previous HitmagnitudeNext Hit of the seismic events. A 3D finite difference reservoir simulator solves for pressure diffusion through the formation. The fault is discretized into boundary-element patches with initial frictional strengths given by the Coulomb criterion; a rate and state friction relation is then implemented to account for unstable and stable rupture growth. In the present model, the fault responds to fluid pressure but does not define a conductive pathway. The results are then used to predict the magnitudes of the induced seismic events. We performed the following simulations: (1) sliding of a circular crack under a uniform shear stress drop (with comparison to analytical solution), (2) injection into a reservoir at the rate of 10,000 bbl/day where fault is located 10 m away from the injector, (3) injection into a reservoir with same rate as previous case, except that the fault is located at 100 m from the injector, and (4) injection into a reservoir with the fault 10 m away from the injector but at the rate of 5,000 bbl/day. The simulations generated earthquakes up to Previous HitmagnitudeNext Hit 2.5 and other observations that can be made are: (1) faults closer to the injector can generate higher-Previous HitmagnitudeTop earthquakes for a given injection rate, (2) higher injection rates can produce larger earthquakes for a given injector distance. However, including faults as a conductive conduits may alter the results and requires further testing.