Research on the conductivity and modeling of artificial fractures in coalbed methane reservoirs

This study addresses a critical gap in the literature by conducting the first systematic, multi-scale experimental investigation that explicitly distinguishes the conductivity contributions of main, branch, and micro-fractures in deep coalbed methane reservoirs. Experiments were performed under realistic reservoir conditions (65°C, 2% KCl brine) with proppant concentrations (2–10 kg/m2), three size ratios (30/50:40/70:70/140 = 1:2:7, 1:1:1, and 1:4:5), and closure pressures (10–30 MPa) over a 140-h period using authentic Ordos Basin coal cores. Three key innovations are presented: (1) quantitative comparison between coal-core and steel-plate tests reveals that conventional steel-plate tests overestimate conductivity by 7.8% versus coal; (2) main fractures dominate, but branch and micro-fractures provide sustained supplementary flow with similar relative decay rates (∼26–29%); (3) a novel dynamic model integrating fracture morphology, proppant size, closure stress, and time-dependent logarithmic decay is developed. Calibrated exclusively with coal core data, the model achieves an average prediction error of less than 1.5%—an unprecedented level of accuracy for coal-specific fracture conductivity modeling. This work provides a reliable engineering tool for life-cycle conductivity forecasting and optimal fracturing design in deep coal reservoirs, filling the gap between laboratory physics and field-scale fracture management.