Matrix shri nkage and swelling can cause profound changes in porosity and permeability of coalbed methane reservoirs during depletion or when under injection processes, with associated implications for primary or enhanced methane recovery. Two models that are used to describe these effects are discussed. The first was developed by Advanced Resources International (ARI) and published in 1990 by Sawyer, et al. The second model was published by Palmer and Mansoori in 1996. This paper shows that the two provide equivalent results for most applications. However, their differences in formulation cause each to have relative advantages and disadvantages under certain circumstances. Specifically, the former appears superior for undersaturated CBM reservoirs while the latter would be better for a case where matrix swelling is strongly disproportional to gas concentration. Since its presentation in 1996, the Palmer and Mansoori model has justifiably received much critical praise. However, the model developed by ARI for the COMET reservoir simulation program, in use since 1990, has significant advantages in certain settings. A review of data published by Levine in 1996 reveals that carbon dioxide causes a greater degree of coal matrix swelling compared to methane, even when measured on a unit of concentration basis. This effect is described in this paper as differential swelling. Differential swelling may have important consequences for enhanced coalbed methane and carbon sequestration projects. To handle the effects of differential swelling, an extension to the matrix shrinkage and swelling model used by the COMET simulator is presented and shown to replicate the data of Levine. Preliminary field results from a carbon dioxide injection project are also presented in support o f the extended model.
展开▼
