Fuel, Vol.97, 569-584, 2012
Microfracturing of coal due to interaction with CO2 under unconfined conditions
Laboratory experiments conducted in the past century have shown that exposure of coal to CO2 under unconfined, hydrostatic conditions leads to reversible adsorption and swelling. However, several authors also report irreversible changes in sorption capacity, sample volume, equilibration time and brittle failure strength. Some relate these effects to the formation of microfractures, while others consider "structural rearrangements'' in the macromolecular structure of coal to be responsible. In this study, we investigate the magnitude of irreversible swelling effects and changes in equilibration time in high volatile bituminous coal (Brzeszcze, Seam 364, Poland), and attempt to explain the results in terms of the operative microphysical processes. We also assess the implications for Enhanced Coalbed Methane (ECBM) operations. Our approach involves detailed dilatometry experiments conducted on fresh, unconfined, mm-scale coal matrix cylinders at CO2 pressures up to 100 MPa, and at 40.0 degrees C. Exposure of our samples to CO2 produced reversible and irreversible strains resulting predominantly from competition between adsorption-induced swelling and elastic compression. During the first or second cycle of exposure, substantial hysteresis was observed in volumetric behaviour, notably at CO2 pressures above 35-40 MPa. After two or three upward and downward CO2 pressure cycles, the measured strain response became fully reversible. Equilibration with CO2 took about four times longer during the first CO2 pressurisation cycle than in subsequent CO2 pressurisation cycles. Microstructural analysis and comparison showed that microfractures formed in the coal during first exposure to CO2. From the microstructural and mechanical data, we infer that microfracturing was responsible for enhanced CO2 penetration into the present samples. This, in turn allowed more homogeneous access of CO2, and caused adsorption-induced swelling of matrix material not previously accessed by CO2. We further infer that the enhanced penetration, sorption and swelling, in turn, resulted in the observed hysteresis in dimensional response and in the decrease in equilibration time seen in subsequent exposure cycles. Since most microfractures developed parallel to the bedding, roughly following maceral-maceral and bedding/layer interfaces, and because the largest permanent strains and strain hysteresis were measured perpendicular to the bedding, we infer that the formation of microfractures was caused by heterogeneous swelling, in combination with differential accessibility of the coal microstructure. No evidence was found that CO2-induced plasticisation of the macromolecular structure of the coal matrix played any role in the behaviour observed. Simple mechanical considerations indicate that at in situ stresses corresponding to a depth of 1000-1500 m, i.e. effective stresses in the range 25-35 MPa, adsorption-induced microfractures are unlikely to form. This means that improved access of CO2 to coal matrix material for ECBM production can probably be achieved only by inducing damage into coal seams, either by injection of solvents/solutes, or by performing active mining of the coal and/or the over-or underlying strata. (C) 2012 Elsevier Ltd. All rights reserved.