Natural earthquakes and mining induced seismicity share some similarities which have led to the field of mining seismology relying heavily on the theoretical developments and advances made in crustal seismology. Source models developed for crustal events have generally performed satisfactorily for shear slip dominated events, which are often associated with the largest mining induced events (1). However, events in mines can also exhibit complex, non shear-slip signatures due to the presence of voids which allow volumetric changes to take place. The closest equivalent in crustal seismology would be seismicity due to the volume changes of magma chambers (2, 3).
The difference between the source signatures in crustal and mining induced seismicity can be attributed to two effects. The first is the influence of the mining environment on the recorded waveforms through the complex interaction of the waves with the mining excavations and surrounding disturbed rockmass. This wave-excavation interaction results in complex waveforms being measured by the in-mine seismic system – the complexity being a post source effect of the wave path. These effects are widely discussed in literature (4-10). The mining environment, however, also allows for more freedom of movement with complex sources ensuing (8, 11-16), and with mining stopes (large excavations for the extraction of ore) contributing to the source mechanism (17-19). While implosive components in moment tensors have been widely observed in case studies (e.g. 11, 20, 21-24), there are relatively few studies (e.g. 17) that attempt to dynamically model stress waves excited by the coseismic closure of a stope.
This paper presents a numerical investigation on the influence of the mining environment on seismic sources, with a focus on pillar failure mechanisms in tabular mining. We investigate the influence of the mining stope (underground excavation or void) on seismic inversions for the scalar moment, corner frequency, source radius, stress drop and moment tensor using synthetic events created within elastodynamic numerical modelling software, WAVE3D. The main objective is to determine whether the source parameters calculated from the recorded waveforms are due to a combination of the stope source and the pillar sources, rather than being related only to crushing of the pillar or shearing in the pillar footwall. The main finding is that the presence of stopes, and types of pillars, have a significant impact on the seismic moment and other source parameters. This is important since moment is viewed as a robust parameter on which seismic magnitude is often based, however this study indicates that moments calculated for pillar failure in a tabular stoping environments are less representative of the shearing or crushing source than originally thought.
