Damage and fracture field evolution in the overburden of an extremely thick coal seam under multi-layered fully mechanized top-coal caving mining

Objective Multi-layered fully mechanized top-coal caving mining represents a primary method for exploiting extremely thick coal seams. Quantitatively characterizing the damage patterns and fracture field evolution characteristics of the overburden in stopes subjected to high-intensity layered mining is critical to the safe production of coal mines.

Methods The extremely thick B1 coal seam (average thickness: 53.59 m) in the No. 2 mine of the Dajing mining area in the Zhundong coalfield was selected as the engineering background. By integrating physical simulation using similar materials and numerical simulation, this study revealed the geometric and fractal characteristics of damage to the overburden during the multi-layered fully mechanized top-coal caving mining of the B1 coal seam. Furthermore, this study predicted and verified the height of the hydraulically conductive fracture zone in the overburden during the mining of various coal layers.

Results and Conclusions The simulation results indicate that during the mining of coal seam B1, high-intensity mining in a super-large space induced significant breaking damage to the overburden. The controlling effect of key strata on fracture development gradually weakened with the mining of the upper coal layer, and fracture evolution differed significantly during the mining of varying layers. The fractal characteristics of fractures in the overburden were determined using the fractal theory. Specifically, fracture propagation in the overburden predominantly occurred during the mining of the upper and middle coal layers, while the fracture field tended to stabilize during the mining of the lower coal layer. The fractal dimension of fractures in the overburden increased, decreased, and stabilized sequentially. The fracture field in the upper coal layer exhibited a staged development trend, while those in the middle and lower coal layers showed nonlinearly decelerating growth trends. A backpropagation (BP) neural network-based multi-parameter prediction model was established to predict the height of the hydraulically conductive fracture zone. Consequently, the predicted height of this zone was 176.3 m, 294.5 m, and 340 m under the mining of the three coal layers, highly consistent with physical simulation results. The results of this study provide an effective reference for subsequent mining face design of the B1 coal seam, as well as the safety management and control of relevant stopes.