American Chemical Society
ef0c03501_si_001.pdf (732.59 kB)

Improved Methane Adsorption Model in Shale by Considering Variable Adsorbed Phase Density

Download (732.59 kB)
journal contribution
posted on 2021-01-14, 15:04 authored by Xingxing Kong, Hongjun Fan, Dianshi Xiao, Pengfei Mu, Shuangfang Lu, Shu Jiang, Guohui Chen
Numerous models have been used to describe the isotherm adsorption of supercritical methane in porous media. Many models assume that the adsorbed phase density does not change with pressure during the adsorption process. However, recent studies show that this assumption is unreasonable, and the resulting error is enormous. Therefore, we propose an improved isotherm adsorption model in shale that assumes that the adsorbed phase density keeps changing and that adsorbed phase volume remains constant during the adsorption process [the variable density adsorption (VD) model]. A logarithmic function is used to describe the change of the adsorbed phase density during the adsorption process. The product of the adsorbed phase density and volume is used to calculate the adsorption capacity. The fitting results for large amounts of methane adsorption data show that this assumption is reasonable. The fitting results are consistent with the molecular simulation, and it will be more convenient to obtain the truly adsorbed phase volume and density. The adsorbed phase volume and density obtained by the VD model show a good positive correlation with the total organic carbon, specific surface area, and micropore volume, which indicates the rationality of adsorption parameters fitted by the model. As a result of the correct calculation of the adsorption phase density, the gas in place (GIP) obtained by the VD model is lower than the supercritical Dubinin–Radushkevich model. The new model proposed this time provides a new tool for the study of shale methane isotherm adsorption and a new model for the calculation of GIP. Using this model, the adsorbed phase density and volume of methane can be obtained more conveniently and accurately. This will be a milestone in the VD model.