(140g) Molecular Simulation of Methane Adsorption Behavior in Nanopores for Shale Gas Development: With Comparison between Graphite and Kerogen Models

Organic rich shale contains kerogen nanopores with characteristic lengths typically about tens of nm to about one hundred nm. For successful shale gas production, the amount of resources need to be evaluated at first. Most of shale gas is adsorbed in kerogen nanopores. In addition to the formation pressure and temperature, the amounts of adsorbed gas are also influenced by different sizes of nanopores. Molecular simulation is a powerful technique, and gets increasingly important in this type of investigation. However, a few issues need to be carefully discussed in order to better address the methane adsorption behavior for engineering applications.

In this paper, we have focused on three points about the computational techniques. First, we discussed the limitation on the use of graphite as a kerogen model. For that, we have compared the results for gas density inside the nanopores with graphite and full-atom kerogen slab models, respectively. The pressure and temperature are fixed at 10 MPa and 350 K, 400 K, while the pore size varies from 0.2 to 50 nm. Second, we discussed the limitation on the implementation of a long-range correction scheme due to the use of a cutoff for dispersion interactions. This scheme can be derived analytically, and appropriately applied for a homogeneous system, while saving the computational cost. However, there is no guarantee for this scheme to be applied in an inhomogeneous system, such as methane in kerogen nanopore system. For this point, we have selected different cut-off values to simulate methane adsorption in kerogen nanopore system (at 10 MPa and 350 K) with dispersion correction and without dispersion correction, respectively. In addition, particle mesh Ewald (PME) is also used to calculate the dispersion interactions as a benchmark for this system. The above two points were addressed by molecular dynamics (MD) simulations, where an anisotropic scaling without shear deformation was set to barostat for pressure control of the system. At last, we have also compared the results from MD and Grand Canonical Monte Carlo (GCMC) simulations.

In kerogen model, the simulation results show that the methane density in the nanopores with pore sizes larger than ~ 6 nm is almost identical to that in bulk phase at the corresponding pressure and temperature. When the pore size is smaller than ~ 6 nm, the density increases with decreasing pore size. However, it seems only to present pore size effect at ~ 2 nm in graphite model. The above simulations have been performed with PME scheme for the dispersion interactions. This indicates that the graphite model has been too simplistic to represent the real kerogen constituents. The reason for this deviation between two models is presumably because there are micropores in kerogen slab (including the interfacial region), where the methane density can be influenced each other in the micropore and mesopore regions. Furthermore, the simulation results with dispersion correction and without dispersion correction are compared. It is shown that the computational details, such as the cut-off values, can play a significant role in the calculation of methane density in the mesopores. In addition, the result (with PME scheme) is close to the simulation result without dispersion correction when we select larger cut-off (> 1.5 nm). Currently, we have been employing GCMC simulations to compare the results obtained from MD simulations. The purpose is to examine the “pore pressure” of methane as used in MD simulations.

The knowledge that we obtained in this study will help us select an appropriate scheme for investigation of methane adsorption in kerogen nanopores. The results will contribute to the quantitative evaluation of methane adsorption under a certain temperature and pressure for certain kerogen constituents, and enable us to connect the nano-scale simulation and macroscopic experimental results.