Permeability Evolution of Intact and Fractured Coals Subject to Varying Pore Pressures: Laboratory Tests and Analytical Modeling.

Due to the application of hydraulic fracturing, a proper understanding of the permeability evolution of multiscale fractures in coal is essential for predicting coalbed methane production and formulating development strategies. In this paper, we adopted laboratory seepage experiments (conducted under controlled pore pressure conditions) and analytical modeling techniques (including permeability characterization models for different fracture types) to investigate the permeability evolution patterns of multiscale fractures in coal under varying pore pressures. Experimental results reveal that when the pore pressure increases from 0.5 to 3 MPa, the permeability of intact coal presents a trend of first decreasing and then increasing, whereas the permeability of self-propped coal samples exhibits a monotonic increasing trend. Additionally, the permeability of the propped fracture is significantly higher than that of intact coal and self-propped samples, showing an approximately linear increase with increasing pore pressure. Notably, under the same pore pressure conditions, a higher proppant concentration and larger proppant size correspond to greater propped fracture permeability. The experimental results also indicate that the permeability of fractures propped with larger proppants is more sensitive to pore pressure changes. At low proppant concentrations, propped fracture permeability increases significantly with increasing concentration; however, this enhancement diminishes as the concentration further increases. Analytical modeling results demonstrate that the proposed permeability model for coal can effectively characterize both intact coal and self-propped samples, exhibiting stronger applicability than existing models, particularly for self-propped samples. Moreover, the proposed permeability characterization model for a propped fracture shows excellent consistency with experimental data and outperforms existing models in capturing permeability variations under different pore pressures, proppant concentrations, and sizes.