Methodology for tomographic imaging ahead of mining using the shearer as a seismic source

Abstract

Poor rock conditions in a coal longwall panel can result in roof collapse when a problematic zone is mined, significantly interrupting mine production. The ability to image rock conditions — stress and degree of fracturing — ahead of the face gives the miners the ability to respond proactively to such problems. This method uses the energy from mining machinery, in this case a coal shearer, to produce an image of the rock velocity ahead of the mining face without interrupting mining. Data from an experiment illustrates the concept. Geophones installed in gate-road roofs record the noise generated by the shearer after it has traversed the panel ahead of the mining face. A generalized crosscorrelation of the signals from pairs of sensors determines relative arrival times from the continuous seismic noise produced by the shearer. These relative times can then be inverted for a velocity structure. The crosscorrelations, performed in the frequency domain, are weighted by a confidence value derived from the spectral coherence between the traces. This produces stable crosscorrelation lags in the presence of noise. The errors in the time-domain data are propagated through to the relative traveltimes and then to the final tomographic velocity image, yielding an estimate of the uncertainty in velocity at each point. This velocity image can then be used to infer information about the stress and fracture state of the rock, providing advance warning of potentially hazardous zones.