Experimental Investigation on Effects of Water Injection on Rock Frictional Sliding and Its Implications for the Mechanism of Induced Earthquake

This study conducted water-induced fault slip experiments on saw-cut granite, sandstone, and limestone samples. Experimental results demonstrated that injecting 15 MPa pressurized water into the vicinity of a high-permeability sandstone fault could decrease the effective normal stress and induce fault slip but not significantly affect the stress of granite and limestone faults due to low permeability. When the pressurized water was injected into the fault plane, 1 MPa pressurized water could not significantly affect fault stress; however, the 15 MPa pressurized water caused a significant reduction in frictional strength and induced fault sliding. The actual pore pressure differed from the injection pressure and showed significant differences in three faults, resulting in the apparent difference in stress drop, slip duration, displacement, and sliding rate. Three faults showed velocity-strengthening properties at room temperature. The fault slip caused by 15 MPa pressurized water injection was a direct response of fault strength to the reduction in effective normal stress. The limestone fault was characterized by velocity-weakening behavior at 100 degrees C, and the sliding rate of the fault induced by the 15 MPa pressurized water injection was faster than that at room temperature. The experiment results suggest that high-pressure injection can dominate over velocity-dependent effects, inducing fault-unstable slips in velocity-strengthening faults, but is more likely to induce medium-strong earthquakes on the velocity-weakening fault.

Automated summary

This study conducted experiments on water-induced fault slip in different types of rocks. The results showed that injecting pressurized water into a high-permeability sandstone fault could induce fault slip, while it had less effect on granite and limestone faults due to their low permeability. The actual pore pressure differed from the injection pressure, resulting in differences in stress drop, slip duration, displacement, and sliding rate. The experiments also found that the behavior of faults differed at different temperatures.