The modeling of fault activation, slip, and induced seismicity for geological CO2 storage at a pilot-scale site in the Janggi Basin, South Korea

We investigate the effects of possible fault activation and slip during geological CO2 storage operations numerically. To this end we develop a numerical simulator by coupling a multiphase flow simulator with an existing geomechanics simulator in order to model fault slip and seismic events induced by reservoir pressurization. After validation and verification of the developed coupled simulator, we apply it to a pilot-scale site of geological CO2 storage located in the Janggi Basin, South Korea. We design a fault structure with a fault core and surrounded damage zones, assuming the fault to be activated along the fault core only. Zero-thickness interface elements are employed in the numerical model to represent the fault core. We perform numerical tests with different injection scenarios to evaluate the impact of pore pressure on fault activation and its post-failure behavior. We model occurrence and the magnitude of seismic events, identifying sudden changes of the moment magnitude at the onset of fault activation. Specifically, we take CO2 injection with 100 , 200, 400 ton/day for 20 years, from which the cumulative moment magnitudes (Mw’s) of 1.46, 2.36 and 3.11 are estimated, respectively. However, the maximum instantaneous changes of Mw occur at the onset of fault activation for all the injection cases (i.e., approximately 0.1, 0.36, 0.75, respectively). As the injection rate increases, fault activation occurs early and the magnitude of induced seismicity becomes large, but substantial reduction of the injection rate can weaken following induced seismicity. Interestingly, the location of maximum surface lift does not correspond to the injection well due to heterogeneity of geology and geomechanics. We thus expect relatively low risk of induced seismicity overall within the range of injection rates considered in this study, but monitoring surface uplift and seismic signals is still necessary to prevent any unexpected events due to uncertainty of parameters.