Experimental Study and Modeling for CO2 Diffusion in Coals with Different Particle Sizes: Based on Gas Absorption (Imbibition) and Pore Structure

The aim of this work is to understand details about the relationship between gas absorption (imbibition), coal pore structure, and gas diffusion behavior in coals with different particle sizes. Four Chinese coals with nine different external surface areas (including particle and cylindrical shapes) were selected. The particle method was used to obtain the CO2 sorption kinetics at 313.2 K and similar to 0.35 MPa. The light microscopy, helium expansion, N-2 and CO2 low-pressure adsorption, CO2 high-pressure adsorption, and mercury porosimetry were used to characterize the pore structures of the different particle sizes. The unipore (UM), modified unipore (mUM), bidisperse (BM), and Fickian diffusion-relaxation (FDR) models and a newly developed bidisperse-relaxation (BR) equation were selected to predict the diffusion coefficients D/r(2) s (min(-1)) of the different particle sizes. The numbers of adjustable parameters, boundary conditions, pore structure assumptions, R-2 values, and the comparison results between experimental data and fitting results were used to assess the applicability of the different models. The results showed the following: (i) it is impossible to obtain the pure matrix diffusion information by the particle method; (ii) the parallel occurring gas absorption (imbibition) for gas diffusion might have become the sorption rate-limiting step when the particle size was sufficiently small; (iii) for a given coal, the pore structures of different particle sizes should be considered to be the same; (iv) the diffusion path length, rather than the pore structure, plays the key role in determining gas diffusion in different particle sizes; and (v) the BR is usable to predict the gas diffusion information for all different particle sizes when taking into account both the parallel occurring of gas absorption (imbibition) and the differences in the pore structures of various coals; however, the BR is not an appropriate model because of its larger numbers of adjustable parameters and inaccurate boundary condition.