Accurate prediction of indicators for engineering failures in complex mining environments

Abstract

Due to seismic load, complex geological and geo-mechanical conditions of rock mass in Steeply Inclined and Extremely Thick Coal Seam (SIETCS) mines, the prediction and prevention of dynamic disasters remain quite challenging. Traditionally, physical models and numerical simulations are widely practiced for disaster prediction at shallow depths which becomes difficult in SIETCS mines. The drilling tests and above models are generally time-consuming and uneconomic providing miniature coverage. Whilst, rockburst driven disasters require detailed regional geological understanding and mechanical behavior of rock mass. For this purpose, a non-invasive and cost-effective geophysical workflow covering regional-scale geology is introduced for accurate identification and prediction of rockburst in a SIETCS mine of northwestern China. Passive seismological in-situ observations were made for a period of twenty-four months to compute the seismic source parameters, map fracturing process, identify regional lineaments and calculate important rockburst prediction indicators. The spatiotemporal distribution of induced-microseismicity, three-states stress theory, and mapped fractures helped in the identification of three rockbursts, several high-energy tremors, and the prediction of two new rockburst-prone regions. The mapped fractures through the novel ‘Seismo-Frac model’ showed that for the simultaneous excavation in B3 + 6 and B2 + 1 coal seams subjected to repetitive episodes of seismic deformation, the shear-displacement accumulation, prying effects of rock pillar and roof coupled with dynamic stress are the primary mode of deformation. The mapped fractures and lineaments are unevenly distributed in the minefield with a maximum density of the fractures concentrated in the B3 + 6 coal seam and rock pillar. Furthermore, through conditional probability analysis, the prediction efficiency of source parameter averages (E and N), seismological parameters (b-value and Δb), and maximum potential magnitude (Mm) was compared. A sharp-increase in source parameter averages while a sharp-decrease in the seismological parameters proved to be effective rockburst prediction indicators. This study also introduces decreasing occurrence frequency of fractures and lineaments as the new prediction indicators. This method, while a panacea for imaging the entire fracturing phenomenon, provides insights with widespread implications for academic researchers and industry practitioners to mitigate dynamic disasters in global underground engineering excavation.