Preliminary Evaluation of Potential Induced Seismicity Risk at a Nearshore Carbon Storage Candidate Site

Abstract

Seismicity or microseismicity events induced by geologic carbon storage (GCS) operations are a major concern for a carbon storage project. The level of induced seismicity risk can be evaluated prior to the site operation and well managed during the injection stage through carefully designed operational procedures and monitoring programs. In this study, we developed a coupled hydro-mechanical model to evaluate the potential induced seismicity risk for carbon storage in a deep saline aquifer at a nearshore candidate site located in a coastal industrial park in Taiwan. For considering the uncertainty of site geology and operations, various geologic and injection scenarios were simulated to obtain the possible ranges of pressure buildup and stress change at the nearest major fault for the site as well as the estimated impact area of the site. In this study, the numerical simulations were carried out by the GEM simulator. The GHG and rock mechanical modules were coupled in the GEM simulator to simulate the simultaneous fluid flow and rock deformation caused by CO2 injection. Several CO2 injection scenarios were simulated using the GEM simulator to obtain the pressure and CO2 plume distributions due to the injection activity. The calculated values of pressure and CO2 saturation were then used to estimate the possible impact locations in the reservoir and compared with the previous study results of injectivity risk. At this candidate site, the nearest major fault, about 17 km far from the injection well, was selected for the preliminary induced seismicity risk evaluation. One million metric tons of CO2 were injected annually by 1-km horizontal well for a period of 50 years with a post-injection observation period of 450 years. In addition to the pressure buildup, the indicators of the potential induced seismicity risk, including stress responses and rock deformation, were estimated using the coupled geomechanical model. The results show that even if the large injection and high response scenario is considered, the simulated CO2 plume is mostly centered around the injection well, gradually spreading upward and moving toward the north-western due to the geologic formation structure. As expected, the corresponding maximum pressure and stress changes occur at the injection well, and the pressure and the effective stress responses propagate further away from the injection well, compared with the CO2 plume. Our highest-risk scenario shows that the pressure increases by about 20 percent at the injection well area while the pressure increases only about 1 percent at the major fault. The values of maximum change in the maximum and minimum principal stresses are -3300 kPa and -1700 kPa, respectively, all occurred around the injection well. The maximum vertical displacement is around 1.2 cm, located close to the bottom of the caprock, and there is very tiny displacement occurred at the location of the fault. Therefore, we suggest that if the operations, such as injection pressure and injection rate, can be properly managed at the source injection well, the GCS induced pressure and stress changes at the fault will remain very low.