Physics‐Based Evaluation of the Maximum Magnitude of Potential Earthquakes Induced by the Hutubi (China) Underground Gas Storage

Abstract

The world's largest underground gas storage facility in Hutubi (HUGS), China, is a unique case where cyclic gas injection‐extraction induced both seismicity and ground deformation. To assess the potential for future induced seismicity, we develop a framework physically based on a well‐constrained hydro‐geomechanical model and on fully‐coupled poroelastic simulations. We first interpret the spatiotemporal distribution and focal mechanisms of induced earthquakes and take it as a key step and a premise to estimate the magnitude and location of the largest potential earthquake. The sharp increase in seismicity was controlled by poroelastic loading on secondary southwest‐dipping thrust faults with spatial scales too small to be resolved by 3D seismic surveys. Both operational and local geological factors affect the seismic productivity at the HUGS site, distinguishing it from most cases of seismicity induced by wastewater disposal and hydraulic fracturing. We then conduct slip tendency analyses for major faults imaged by the seismic data, including the largest reservoir‐bounding Hutubi fault hydraulically connected to injection wells. The reactivation potentials of these imaged faults are estimated to be extremely low. Accordingly, future seismicity would most likely occur on failure‐prone secondary faults in regions with positive stress perturbation due to poroelastic loading. The maximum magnitude likely depends on the spatial scales of the secondary faults. As the occurrence of detected earthquakes is spatially and temporally consistent with the simulated evolution of Coulomb stress perturbation, the location of the largest potential earthquake probably depends on the sizes of the poroelastic stressing regions.

Three Mw ≥ 4.7 Earthquakes Within the Changning (China) Shale Gas Field Ruptured Shallow Faults Intersecting With Hydraulic Fracturing Wells

“…Possible surface rupture of the 2018 event had devastating consequences including ground shaking that exceeded design capacities of numerous structures. And, given the direct association with large‐scale HF activities, we suggest that an assessment of time‐varying hazard with physics‐based modeling may be needed to improve the accuracy of hazard forecasts (e.g., Rubinstein, et al., 2021); geodetic observations can be useful for developing such models (e.g., Jiang et al., 2021).…”

Retracted: Control of Multiple Detachments on Structural Development in the Southern Junggar Fold‐and‐Thrust Belt, Northern Tian Shan: Implications for Seismic Hazard Assessment

“…Although it is hard to predict when this fault could link to the deep detachment (forming a mature ramp), its seismic potential cannot be precluded with ease for the following two reasons: a) it is close to the City of Urumqi, with a population of over 3.5 million; b) the southern Junggar Basin is underlain by over 10 km of Mesozoic-Cenozoic sediments, which likely amplify ground shaking during earthquakes (Field, 2000;Hubbard et al, 2014;Middleton et al, 2016). We believe that a better understanding of its geometry and kinematics would be helpful in further assessing the seismic hazards related to this structure (Jiang et al, 2020(Jiang et al, , 2021.…”

Poroelastic model in a vertically sealed gas storage: a case study from cyclic injection/production in a carbonate aquifer

“…In general, both CM and RS models have been applied to study induced seismicity in different contexts (Dempsey & Suckale 2017;Langenbruch et al 2018;Norbeck & Rubinstein, 2018;Shapiro, 2018;Candela et al 2019;Jiang et al 2021). In our case, the CM K /CM sub K and RS models can explain the observed microseismicity in the examined time period of 2.3 yr significantly better than the Poisson and simple CM (i.e.…”

“…Here the pore-pressure variations are orders of magnitudes smaller, but the poroelastic stresses can give an important contribution in triggering seismicity (e.g. Deng et al 2020;Jiang et al 2021). The occurrence of seismicity just across the F2 boundary could also be ascribed to a not complete impermeability of the fault or the presence of a thick fault zone.…”