The CH4 Production and CO2 Sequestration During CO2-ECBM Considering Fracture Networks: Multi-Physics Field Coupled Numerical Investigation and Field Application

Summary Accurate characterization of fluid flow, and competitive adsorption behaviors of CO2 and CH4 in coal reservoirs rich in complexity fractures is critical for optimizing CO2 enhanced coalbed methane (CO2-ECBM) recovery and assessing CO2 sequestration. A fully-coupled adsorption-temperature-hydraulic-mechanic (ATHM) model is established, incorporating mechanical deformation due to effective stress, temperature-induced swelling, and adsorption-induced swelling during CO2 injection. This leads to long-term permeability evolution caused by creep-desorption-compression. The numerical solution is obtained using the MRST. A transient source function-based embedded discrete fracture model (tSEDFM) is proposed to capture the transient flow within fractures. The ATHM-tSEDFM model's accuracy is validated against experimental and field data. After CO2 injection, the CH4 desorbs and preferentially migrates along the fracture network. Under multi-field coupling, permeability initially decreases and subsequently rebounds. When CO2 breaks through to the production wells after approximately 220 days, CH4 daily production declines sharply, and the CO2 sequestration capacity is significantly reduced. After continuous CO2 injection over 2000 days under varying fracture number, peak CH4 production generally decreases as fracture number increases. Both insufficient and excessive fractures constrain CO2 sequestration efficient. A higher CO2 injection rate enhances initial CH4 production and prolongs the stable period. However, once CO2 breakthrough occurs, CH4 production drops rapidly. The total amount of sequestered CO2 generally shows a positive correlation with the its injection rate. For fractured coals, implementing a CO2 injection-shut-in cycling strategy can alleviate the negative impact of the fracture network on CO2 sequestration. However, excessively long shut-in periods hinder CH4 migration, while excessively injection-shut-in transitions reduce CO2 displacement efficiency. A hybrid framework combining Bayesian optimization (BO) with the proposed simulator was established and applied to the 9-well group in the Qinshui Basin, with an injection rate of 480 m3/d and an injection-shut-in period of 110 days. The well group achieved a CH4 production of 66.3×104 m3 and CO2 sequestration of 10.72×104 m3. The proposed fully-coupled BO-tSEDFM-ATHM framework can better characterize and optimize the CO2-ECBM and CO2 sequestration in fractured coals, laying a solid foundation for optimizing the well pattern and sequestration strategies.