Accurate seismic event locations are essential for analysis of mining-induced seismicity. Complexities in the seismic velocity structure around coal mines, particularly changes in velocity caused by the progression of mining, can cause systematic event mislocations. Synthetic event locations simulating the presence of a mining-induced low-velocity zone (LVZ) were compared with locations of seismicity on the second panel in a multi-panel longwall sequence for a deep, western U.S. coal mine. The locations of seismicity migrate away from the workings in a similar manner to the synthetic locations, suggesting an LVZ is influencing the locations. A parametric study using NonLinLoc was conducted to estimate the LVZ around the mine workings. The LVZ was applied in a time-varying manner to evaluate the resulting effects on the event locations. The best model according to travel-time root-mean-square errors and mining information is a model with a 30% velocity decrease extending 200 m into the roof and a 20% decrease extending 50 m into the floor. Relocating the seismic events with this model produces event locations that are more consistent with mining activities.
Dynamic rock mass failures, also known as rockbursts or bumps, pose a significant hazard to workers in underground coal mining. Monitoring mining-induced seismicity (MIS) in coal mines may provide useful information for understanding the performance of longwall mine designs, characterizing dynamic rock mass failures, and identifying the roles of geologic structures in the rock mass response. Gaining a better understanding of these factors can aid in the determination of strategies to mitigate the effects of dynamic failures. An important requirement for interpreting MIS in the analysis of mine performance is seismic event locations that are accurate relative to mine workings and geologic structures. Locating MIS associated with underground coal mining is relatively difficult compared to locating seismicity in other environments because of the complexities in the seismic velocity structure associated with shallow depositional environments. The high extraction ratio of longwall mining can also significantly modify the velocity structure as a result of subsequent caving.
Even with dense arrays of sensors, accurately locating coal MIS is difficult without an accurate model to characterize the velocity at which seismic energy travels through the rock mass. The shallow nature of coal MIS (depths generally < 1 km) poses challenges because the first-arriving rays of seismic energy to seismic sensors travel primarily in shallow velocity structure, which is more likely to be highly heterogeneous and anisotropic. The seismic velocity of a rock mass is dependent upon its elastic properties, competence, and stress state (Shearer, 2011). These factors can vary greatly as a function of location, especially in a coal mining environment comprised of sedimentary layers with very distinct properties. The shallow event locations can also be more sensitive to the effects of topography because variations in topography may become large relative to the ray paths. Many location algorithms do not explicitly account for topography, and may determine that the fastest ray path to a station is through a volume of air (e.g., a canyon) where minimal seismic energy actually traverses.