Size and spatial fractal distributions of coal fracture networks under different mining-induced stress conditions

Coal permeability is a key issue in CO2 injection and enhanced coalbed methane (CBM) recovery, and it is determined by the fracture network, which is strongly influenced by mining-induced stress evolutions. For a comprehensive understanding of the size and spatial distribution characteristics of coal fracture networks under different mining-induced stress conditions, a series of laboratory experiments, computed tomography (CT) scans and image analyses focusing on 3D coal fracture systems have been conducted considering the stress conditions induced by three typical mining layouts, i.e., top-coal caving mining (TCM), non-pillar mining (NM) and protective coal-seam mining (PCM). The size and spatial distributions of mining-induced coal micmfractures and the anisotropic tortuosity characteristics of coal fracture networks have been quantitatively determined based on fractal theory. The results show that the size distributions of micmfracture geometries, the fractal dimensions of fracture size distribution and the tortuosity of the mining-induced coal fracture networks vary according to the mining-induced stress conditions, resulting in differences in the seepage capacity of coal masses. The coal samples subjected to PCM conditions have the highest percentage of large microfractures, and those exposed to NM conditions have the lowest percentage. The fractal dimensions (D-a and D-d) of the micmfracture size distributions of the typical coal specimens decrease in the order of PCM, NM and TCM conditions, and the D-a and D-d of coal specimens without pre-mining unloading-expansion simulation (PUES) are much lower than those of specimens with PUES. The tortuosity fractal dimensions (delta) of the mining-induced fracture network are spatially anisotropic. The ranges of delta for the coal masses exposed to PCM, NM and TCM conditions are 1.0-1.2, 1.2-1.3 and 1.25-1.4, respectively. Using fractal theory and the measured fracture network data, the anisotmpic spatial distribution of coal permeability can be theoretically estimated.