It is of paramount importance to gain a better understanding of induced seismicity taking place in deep underground for sustainable energy developments. Notwithstanding the effort made in the numerical simulation of induced seismicity, there is still a large gap between analysis results and field observations. The present study aims at simulating spatially and temporally distributed fault-slip events whilst considering metre-scale stress heterogeneity. The result indicates that the heterogeneous fault confining stress is crucial in the occurrence of fault-slip. The analysis of b-values computed from multiple seismic events simulated on the fault plane demonstrated that the b-value decreases with the reduction of the effective normal stress, showing consistency with the characteristic of induced seismicity in the field. This implies the possibility of applying the simulation method developed in this study to the risk evaluation for seismic hazards through the b-value analysis based on the advanced fault-slip modelling approach.
Due to the depletion of shallow ore deposits, mining depths have been increasing around the world, leading to ore extraction under unfavorable geological and stress conditions, thereby causing serious problems that need to be addressed for sustainable developments at great depths. Mining-induced seismicity is well recognized as one of such problems, and seismic waves released could inflict devastating damage to underground facilities and openings in an extensive area in an underground mine when entailing rockbursts. Furthermore, seismic events with large magnitudes can cause noticeable ground vibrations, which may raise public concern depending on the intensity. For these reasons, it is of importance to develop methods for estimating the risk and mitigating the damage.
For decades, numerical modelling techniques have been intensively employed to analyze the stress regime of underground geological structures and estimate the potential of seismicity, including the identification of highly stressed region and source parameters of seismic events. However, there is a great gap between numerical modelling results and those of in-situ seismic monitoring in that the number of seismic events simulated with numerical models is significantly less compared to that detected with seismic monitoring systems (Sainoki and Mitri, 2014). Furthermore, for numerical modelling, it is quite challenging to estimate the risk of seismic events taking place in geological structures located away from active mining areas, where mining-induced stress change is negligible compared to that in the vicinity of underground openings, such as drifts and stopes.
