Hot Flue Gas-Enhanced Coalbed Methane Recovery: Comparative Discrete Fracture Network and Equivalent Continuum Modeling

Abstract Hot flue gas injection provides a potential approach for enhancing methane displacement from low-permeability coal, but methane transport is strongly controlled by the heterogeneous pore-fracture structure of coal. How different representations of fractures affect the predicted displacement response remains insufficiently understood, particularly when discrete fracture network (DFN) and equivalent continuum medium (ECM) models are compared under identical initial and boundary conditions. To address this issue, CO2–H2O–coal experiments and CT characterization were combined to obtain fracture statistics, which were used to construct DFN and ECM multiphysics models coupling solid deformation, gas flow, heat transfer, and species transport. The models were validated against displacement experiments under isothermal injection pressures of 1 and 2 MPa, with deviations below 5%. The DFN model predicted more pronounced preferential transport and faster early-stage methane removal than the ECM model. At 393 K, the CH4 concentration at Monitoring Point B after 10 min was 0.49 kg/m3 for DFN and 0.87 kg/m3 for ECM. Increasing injection pressure and injected-gas temperature enhanced methane displacement, whereas increasing the initial CH4 pressure mainly increased the initial methane concentration without altering the stronger early DFN response. These results indicate that fracture representation substantially affects the predicted transient and spatial characteristics of methane displacement. DFN is more suitable for fracture-dominated and strongly heterogeneous transport.