Modeling Dynamic Gas Desorption in Coal Reservoir Rehabilitation: Molecular Simulation and Neural Network Approach

A thorough understanding of the control mechanisms of coal reservoir modification on methane adsorption and desorption is essential as this is a key technique for increasing the effectiveness of gas extraction. In this study, molecular dynamics simulations and neural networks were used to evaluate the effects of several coal reservoir alteration factors on gas desorption, from both microscopic and macroscopic perspectives. The findings demonstrated a direct correlation between coal pore size and the amount of methane adsorbed, as well as an inverse relationship between coal pore size and methane adsorption capacity and energy. The different methane-repelling properties of CO2, N2, and H2O, which are frequently used in coal reservoir reforming, are primarily due to the different diffusion capabilities of these three gases. The best reservoir reforming effect can be obtained by setting the pressure ratio of CO2 to N2 to 3.4:6.6. The thickness, depth, gas content, height, advance speed, rate of extraction, and daily production of coal are all closely interrelated, enabling a more accurate assessment of gas gushing.

Automated summary

This study evaluated the effects of coal reservoir modification on gas desorption using molecular dynamics simulations and neural networks. The results showed a correlation between coal pore size and methane adsorption and an inverse relationship between coal pore size and methane adsorption capacity and energy. The different methane-repelling properties of CO2, N2, and H2O are due to their different diffusion capabilities. The best reservoir reforming effect can be achieved by setting the pressure ratio of CO2 to N2 to 3.4:6.6.