The calculations of pore structure parameters from gas adsorption experiments of shales: Which models are better?

Abstract

Currently, low-pressure gas (carbon dioxide and nitrogen) adsorption experiments have been widely applied to describe pore structure features of shales. However, there are no standard models in calculations of pore structure parameters from low-pressure gas adsorption experiments, which limits comparison between various studies. To evaluate the applicability of calculating methods, 9 Chang 7 shale samples were performed gas adsorption experiments and pore structure parameters are obtained using Monte Carlo simulation (MC), Density Functional Theory (DFT) and Barrett-Joyner-Halenda (BJH) methods. The results show that MC and DFT can accurately characterize micropore structures, but MC model provides a wider pore size range than DFT. Therefore, we recommend that MC method should be used to get micropore parameters from CO2 adsorption data. Under the influence of tensile strength effect (TSE), a fake peak located at ~3.8 nm in pore size distribution (PSD) lines obtained from N2 desorption curves with BJH model, and an overestimation for pore structure parameters in meso-to macropore dimension based on this method was found. This phenomenon implied that the parameters from N2 desorption using BJH model are not applicable for the calculation of parameters for pore structure. What's more, compared with DFT method, BJH model based on adsorption curve ignores the effect of adsorption potential on location of pore condensation transition, which may lead to an underestimation of the shale pore structure parameters in narrow mesopore range (<10 nm). Therefore, DFT method was recommended to characterize the meso- and macroporous pore structure of shales based on N2 adsorption data.