The Simplified Local-Density Peng-Robinson (SLD-PR) adsorption model was tested for its generalization capability. In particular, a coal-property, structure-based generalization was employed. Although this SLD-PR model development is similar to our previous efforts with the Simplified Local-Density Hard-Sphere (SLD-HS) model, an important motivation in carrying out this generalization was the readily available mixing rules of the PR equation of state. This considerably simplified the direct extension of this model to mixture adsorption behavior.

Before performing the SLD-PR model generalizations, we have assessed the correlative capability of the model. On average, the model can represent all adsorption data within their experimental uncertainties. Then, simple correlations were developed to generalize the SLD-PR parameters (i.e., matrix surface area, solid-solid interaction parameter, and the slit width) in terms of gas properties and accessible adsorbent characteristics (the proximate and ultimate analyses of the coals). For this analysis, we used our adsorption measurements for methane, nitrogen and CO2 on ten dry and wet coals, including five Argonne Premium coals.

Our results indicate that the generalized SLD-PR model is capable of predicting the adsorption of the coalbed methane gases (methane, nitrogen and CO2) and their mixtures, on average, within twice the expected experimental uncertainties.