Here, we report the results of dilatometry experiments conducted on unconfined, cylindrical coal matrix samples (¡«4 mm long and 4 mm in diameter) of high volatile bituminous coal, where we aim to measure the effective volumetric effect of CO2 and to separate this into a component caused by adsorption-induced swelling and a component caused by elastic compression. The experiments were performed using a high pressure eddy current dilatometer that was used to measure one-dimensional sample expansion or contraction (resolution <50 nm). The tests were conducted at a constant temperature of 40 ¡ãC, and CO2 pressures up to 100 MPa. Our results show that the matrix samples reveal anisotropic expansion over the full range of CO2 pressures used. Expansion perpendicular to the bedding was about 1.4 times the average expansion measured in the bedding plane. Net volumetric strains, which were computed from the net linear strain in all directions measured, reveal that the response of coal is characterised by an expansion-dominated stage below 10-20 MPa of CO2 pressure and a contraction-dominated stage at higher CO2 pressures. Our data demonstrate direct competition between adsorption-induced swelling and elastic compression in the coal matrix. We propose a model for coal swelling, which expresses the net volumetric strain as the sum of the adsorption-induced swelling strain and the elastic compression with the adsorption-induced swelling being taken as linearly related to adsorbed CO2 concentration. A comparison of experimentally determined adsorption-induced swelling strain with the adsorbed concentration of CO2 (data ) confirms the assumed linear dependence. We go on to compare our experimentally determined adsorption-induced swelling strains to those calculated from an adsorbed concentration model. Good agreement was found over the full range of CO2 pressures up to 100 MPa. This shows that combining this thermodynamically based model for adsorbed concentration with the elastic compression of our samples, obtained from their bulk modulus, provides a good description of the measured volumetric behaviour of our samples, and suggests that the physical basis for the model is also valid.
The implications of our results for ECBM operations are that compliant coals (low K), which exhibit little adsorption-induced swelling (hence low dependence C), will show relatively small reductions or even increases in permeability due to competition between swelling and compression when CO2 pressure increases during ECBM operations. These coals will tend to be more suitable for ECBM operations. Coals exhibiting high stiffness (K) and high adsorption capacity are less suitable for ECBM.