Combined prediction model for mining subsidence in coal mining areas covered with thick alluvial soil layer

The entire overburden stratum above a coal mining area is considered to be composed of an alluvial soil layer and a bedrock layer. In underground mining, alluvial soil has a major effect on ground subsidence. This effect is especially significant and not negligible when the alluvial soil accounts for a large proportion in the entire overlying stratum of coal mining areas covered with thick alluvial soil (CMATASs). In this study, the applicability of the most popular probability integral method (PIM) used for CMATASs was analyzed, and it was found that the PIM is not suitable for ground subsidence prediction in CMATASs for two main reasons: (1) the subsidence basin range predicted by the PIM was smaller than the measured range; thus, the predicted basin converged more rapidly than the measured basin at the edges. (2) Poor fitting results were obtained at the subsidence basin edge. The mechanism of mining subsidence in CMATASs and the reasons why the PIM is not suitable for CMATASs were investigated in terms of the internal deformation and crack (void) distribution within the rock mass induced by underground coal mining. The results indicate that the alluvial soil is compacted and then subsides under vertical compression deformation, which increases ground subsidence. The bedrock is subjected to the weight/load of the thick alluvial soil layer. Because of the vertical compression deformation inside the rock, its internal voids (spaces) are compacted; thus, the replaced voids (spaces) are transferred to the ground surface, resulting in an increase in ground subsidence. However, the effect of the alluvial soil on ground subsidence cannot be detected by the PIM; moreover, the PIM is unable to represent the mechanism of the mining subsidence in CMATASs. Thus, the PIM cannot be used for predicting subsidence in CMATASs. To precisely calculate subsidence in CMATASs, we propose a combined prediction model (CPM) for mining subsidence in CMATASs based on soil mechanics and stochastic medium theory. The new CPM was applied to the mining process in the Huainan Coal Mining Area. The calculation results show that subsidence predicted by the CPM better fits the measured subsidence values, with a relative error of 4.9%, and that the fitting accuracy is improved by 18% compared to the relative error of the PIM (6.0%). Thus, the proposed CPM is more suitable for predicting ground subsidence caused by underground coal mining in CMATASs, and can be used to provide more accurate predictions for ground subsidence in similar coal mining areas.