It is occasionally observed that a large seismic event takes place in a region near a fault but away from an acting mining area, despite the negligible mining-induced stress change. The objective of this study is to propose a methodology to quantify the volume of highly stressed zones generated in a fracture network within a fault damage zone.
To achieve this, a large-scale discrete fracture network including millions of fractures was generated in the domain of 600 m × 300 m × 300 m. Simultaneously, 12 numerical models with structured or unstructured meshes are constructed on a similar scale whilst varying the zone edge length. Subsequently, equivalent full anisotropic compliance tensors are computed for each zone using the crack tensor theory. Stress analyses were then performed whilst applying reginal stresses σR to the model boundaries.
The analysis results indicated that the volume of zones with σmax > σR increases with a decrease in the zone edge length. The analyses further demonstrated that the volume of extremely stressed zones (σmax > 4σR) continues to increase and does not converge even with the model containing more than 15 million zones. However, it was revealed that by assigning a specific range, e.g. 1.3σR < σmax < 4σR, the volume of zones with the maximum stress in the range converges when reducing the zone volume to 3.0 m3, irrespective of the mesh discretization scheme.
Assuming that extremely stressed zones undergo failure before mining activity, quantifying the volume of the moderately stressed zones is more crucial for assessing the risk of severe seismic events. The methodology proposed enables us to quantify the volume of highly stressed zones in an extensive region with a given fracture network. This study would contribute to analysing the risk of large seismic events caused by inherent stress anomalies.
As mining depths have been increasing since the last century due to the depletion of shallow ore deposits, more challenging in-situ ground and stress conditions have been experienced in many deep underground mines. Such considerably adverse conditions are known to increase the risk of rockburst that involves the instantaneous release of strain energy as well as the ejection of rocks with high velocity from the surface of underground openings (Ortlepp, 2000). As such, rockburst can cause devastating damage to underground facilities and mining equipment, being a life-threatening phenomenon. It is hence indispensable to develop a methodology that enables a more accurate estimation of rockburst potential and its prediction.