Characterising the Dependence of Coal Permeability to Methane Adsorption, Pore Pressure and Stress: Laboratory Testing of Walloon Coals from the Surat Basin

摘要

This paper discusses a set of laboratory experiments on coal core samples, which seek to understand the changes in coal fracture permeability with varying confining stress, pore pressure and methane adsorption. Currently, no measurements exist to support QGC'sSurat Basin Coal Seam Gas (CSG) development, and therefore themagnitude of change in permeability relative to initial conditions is unknown. The set of experiments described here use a flow apparatus to measure the permeability of coal core samples as a single variable is changed. The ability to control the laboratory test to separate out the different variablesis an improvement on inferring coal permeability variations from well test data, which isaffected by multiple unknown variables at the same time. Four separate experiments have been carried out. The first investigated the impact of higher burial pressures on permeability, a four-fold reduction in permeability was measured with confining pressure increase from 150-750psi. The second replicated the reduction in pore pressure during dewatering the intention was to measure ‘stress-dependent permeability’, a halving of permeability was observed with pore pressure reduction from 220-30psi. The final two experiments used methane gas to investigate the effect of coal shrinkage with changing pore pressure. Under constant stress conditionspermeability halved and under constant volume conditions permeability saw a significant fifteen-fold reduction. Focus on the resulting magnitude of permeability change is more important at the current stage of this study than the absolute permeability change. The magnitude of permeability change can be used assensitivity in numerical simulation studies toassess the impact of pressure dependency of coal permeability and therefore better characterise coal reservoir behaviour and, consequently, improve the accuracy of any forecast derived from the reservoir model.