Mitigating risks of injection-induced seismicity using cyclic fluid injection: new evidence and insights from the laboratory

Anthropogenic earthquakes associated with high-pressure fluid injection into underground have been reported worldwide over the past decades. Fluid injection can reduce the effective normal stress on a critically stressed fault and potentially promote seismogenic fault slip. Cyclic fluid injection has recently been employed as an alternative fluid injection protocol to mitigate seismic risks. However, the mechanism of seismic risk mitigation by cyclic fluid injection remains unclear. We performed injection-driven shear tests on sawcut fractures in Bukit Timah granite, and compared the injection-induced fracture instabilities under monotonic and cyclic fluid injection conditions. The pressurization rate (i.e. 0.01 MPa/s) or volumetric rate (i.e. 0.2 mL/min) was kept constant in the monotonic fluid injection, and the cyclic fluid injection alternated between high and low injection pressures (i.e. 95%, 90%, or 85% of the injection pressure at fracture failure by monotonic fluid injection and 0 MPa) or volumetric rates (i.e. 0.2 and 0 mL/min). We particularly compared the injection pressures at fracture failure and associated seismic moments under different conditions. Our results show that the injection pressure at fracture failure in monotonic fluid injection is higher than that predicted by the Mohr-Coulomb criterion combined with the effective stress law. The deviation of the injection pressure at fracture failure from the predicted value is reduced by decreasing the peak fluid pressure during cyclic fluid injection. Our data also demonstrate that cyclic fluid injection can effectively reduce the injection-induced seismic moment compared to monotonic fluid injection.