| The methane diffusion coefficient in coal matrix is one of the basic parameters required in simulation of gas transport in coalbeds. Its value has been assumed to be constant throughout the production life of a coalbed methane reservoir. The adequacy of this assumption, however, needs to be examined. The objective of this study is to conduct an experimental investigation in order to measure the micropore diffusion coefficient and examine its dependency on reservoir pressure, desorption of methane, and water saturation. The techniques used in previous studies are either inadequate or incapable of serving for this purpose. This investigation introduces the transient interflow method to measure D. A modeling equation is derived based on Fick's Law and solved using the experimental results to obtain the value of D. The numerical solution evidently show that the diffusion coefficient is methane concentration, pressure and water saturation dependent. A continuous decrease in the value of D with decreasing methane concentration in coal is observed, and related to the changes of dominant flow mechanism in coal matrix and the shrinkage of coal matrix induced by desorption of methane. Results of the experiments conducted at different pressures indicate that the release of methane from coal at late time diffusion is much slower at a lower pressure. This may partly be attributable to the higher degree of coal matrix shrinkage at low pressures. The methane diffusion in coal matrix, therefore, should not be described by a unipore model or a single diffusion coefficient model. The experiment with a wet sample shows that the value of D is a strong function of water content in coal. The decrease in the value of D due to the presence of water in coal also changes with mean concentration of methane in coal. The diffusion coefficient measured for a water content 0.53% may be 30 times smaller than for the dry sample at late time diffusion. |
