Hydraulic fracturing not only enhances the flow conductivity of coal seams but also transforms the CH4 adsorbed in coal seams into free CH4 and then outputs it via the desorption-diffusion process, thus improving the output quantity and efficiency of CH4. In this paper, taking the coal from Changping mine in Shanxi as the research object, the 3D macromolecule model of coal was established by using analytical test experiments, and the molecular dynamics simulation method of desorption and diffusion was applied to study the characteristics of methane desorption rate and diffusion behaviors under different driving water pressures through the parameters of gas-water concentration distribution, desorption and diffusion rates, and the interaction energies between the coal and the gas-liquid. The results show that the water drive can increase the desorption rate of methane. With the increase of methane adsorption pressure, the desorption effect of water drive becomes stronger and then weaker. The desorption effect of water drive was best when the methane adsorption pressure was 5 MPa and the pressure difference was 2 MPa. Generally, the water drive promoted the methane diffusion, but if the water drive time was too long, it formed a water plug, and the water drive methane inhibited the diffusion. In shallow reservoirs, water drive can obviously promote the desorption-diffusion of methane. The binding energy between CH4 molecules is mainly composed of the repulsive force, while that between water molecules is mainly the adsorption force. In the binding energy between CH4 molecules and coal structures, the van der Waals (VDW) force accounts for 99% of the effect; in the binding energy between water molecules and coal structures, the electrostatic force contributes 84% to ∼88% thereto. The VDW force is significantly influenced by the pressure, while the electrostatic force is slightly affected. As the water displacement differential pressure is increased, the binding energy between CH4 molecules and coal structures decreases.
Read full abstract