A convergence-guided molecular modeling framework for CO2/N2/CH4 competitive adsorption in coal during hot flue gas enhanced methane recovery
Abstract
Hot flue gas injection for enhanced coalbed methane recovery (ECBM) holds potential to enhance methane recovery and enable geological CO2 sequestration. Predictions of CO2/N2/CH4 competitive adsorption and transport in molecular simulations are sensitive to coal molecular model size, coal rank representation, and temperature and pressure conditions. Quantitative criteria for representative molecular models remain lacking. This work treats coal macromolecule number selection as a finite size convergence issue and establishes a convergence scoring framework accounting for density, pore structure, adsorption capacity, isosteric heat of adsorption, and competitive adsorption selectivity. Integrating PoreBlazer, grand canonical Monte Carlo (GCMC), and molecular dynamics (MD) simulations, this work investigates adsorption, diffusion, and intermolecular interactions of CO2/N2/CH4 in lignite, bituminous coal, and anthracite. Results show minimum representative model sizes of 50, 40, and 30 macromolecules for lignite, bituminous coal, and anthracite, respectively. Larger models reduce isolated free volume artifacts and improve statistical representation of adsorption sites and diffusion pathways. In the ternary pore gas mixture, pressure drives transition from high affinity site occupation to progressive pore filling. CO2 enriches at favorable sites to displace CH4, while N2 shows weaker interaction and higher mobility. The competitive adsorption is governed by synergistic site occupation and pore filling, and an increase in temperature shortens the residence times of molecules at the adsorption sites, thereby suppressing adsorption. Nonlinear mean squared displacement indicates transport via intermittent trapping in connected free volume rather than unrestricted diffusion. This work provides a quantitative convergence criterion to replace empirical model-size selection in coal molecular simulations.
