Transport and mixing of combustion-assisting agent in coal rock reservoirs for in situ methane explosive fracturing

Targeting the high water consumption and easy fracture closure of conventional hydraulic fracturing for ultralow-permeability coal seams, in situ methane explosive fracturing requires no external water, with detonation-generated complex fractures featuring high, durable flow conductivity, making it an emerging solution to boost low-permeability coal rock gas recovery. The core lies in achieving efficient mixing of a combustion-assisting agent (CAA) with methane inside the reservoir. To reveal this mixing process, a three-dimensional wellbore-reservoir flow numerical model integrating the Realizable k–ε turbulence model and porous media model was established. The migration patterns and mixing characteristics of the CAA under open-hole and perforated completion conditions were comparatively analyzed. The results show that perforated completion can significantly enhance both the maximum injection distance and the sweep efficiency of the system by constructing a “perforation-dominated radial expansion” composite diffusion pattern. The volume proportion of the explosive zone reaches 24.50%, representing a 41.60% increase (relative) compared to open-hole completion (17.30%), while the extension distance of the explosive zone increases from 1.80 to 2.30 m. Parameter sensitivity analysis reveals that reservoir permeability, which governs the matrix seepage capacity, is the most fundamental parameter determining injection effectiveness. Increasing permeability from 0.01 to 0.2 mD enhances CAA sweep efficiency by 35.90% (absolute). Perforation depth and delivery pressure are the most sensitive engineering parameters for regulating the extent of the explosive zone, whereas the influence of perforation diameter (10–18 mm) is negligible. Based on the Peclet number, completion structure factor, and permeability factor, a quantitative evaluation model for CAA mixing efficiency is established. Validated against ten operating conditions, the model yields prediction errors below 2%. This study provides a theoretical basis for coal rock gas reservoir explosive fracturing.