Mitigation of injection-induced seismicity on undrained faults in granite using cyclic fluid injection: A laboratory study

Cyclic fluid injection has been demonstrated as a plausibly effective and controllable strategy to mitigate the seismic risks during hydraulic stimulation. The mechanism involved remains largely unconstrained, and our ability to control the activation of critically stressed, locally undrained faults is still limited. Injection-induced activation of these faults can be one of the most threatening scenarios as they likely perturb the stability of nearby faults beyond the stimulation volume. Here, we perform a series of laboratory fluid injection tests on critically stressed, locally undrained faults in low-permeability granite to offer insights into cyclic fluid injection as a possible solution for seismic risk mitigation. Our results show that cyclic fluid injection promotes fluid pressure diffusion on the faults, but a reduction in seismic moment release depends on several cycle-related factors, such as the critical injection pressure and injection frequency. Particularly, cyclic fluid injection could be inefficient for fluid pressure diffusion if the critical injection pressure is very close to the predicted pressure at fault failure, or over-reduced to cause excess fluid injection and long-term frictional healing. A proper design of injection parameters is thus essential to balance the energy budget between the seismic energy and hydraulic energy. Our study reveals that the effectiveness of cyclic fluid injection is also dependent on fault drainage conditions, stimulation requirements, as well as dynamic responses of faulted reservoirs, which could guide the future development of cyclic fluid injection.