Coal Mine Subsidence Prediction Using A Boundary-Element Program

Historically, the surface subsidence over underground coal mines has been predicted using empirical profile or influence functions which have little or no connection to the actual mechanics of the subsidence. Without a mechanistic basis, establishing the exact site-specific parameters to use in these empirical methods has been problematic. A practical subsidence predictive method based on mechanics has the appealing capability of allowing the determination of site-specific parameters from fundamental properties of the overburden with minimal field calibration work. This paper presents several case studies where a mechanics-based, boundary-element program is used to back-calculate the surface subsidence associated with various panels at several Northern Appalachian coal mines. The program used in this case study is called LAMODEL, and it incorporates a frictionless, laminated overburden into a general purpose displacement-discontinuity code primarily designed for calculating the stresses and displacements in coal mines or other thin scam or vein type deposits. In this paper, the program is used to calculate both the underground convergence and the resulting surface subsidence at five longwall panels and a room-and-pillar section. The fitted subsidence from the model is compared with the field measurements and analyzed. The results from the case studies in this paper demonstrate that the laminated model with calibrated properties can easily provide moderately accurate subsidence predictions and is fairly flexible for fitting measured subsidence. However, additional subsidence predictive case studies arc recommended in order to ultimately evaluate the potential of the laminated overburden model for practical subsidence prediction