Pore scale study of gas sorption and transport in shale nanopore systems

Xu, Rui, Ph. D.
Journal Title
Journal ISSN
Volume Title

Shale gas production accounts for about 70% of the total natural gas production in the US. Yet it remains a nontrivial task to characterize the petrophysical properties of shale core samples either by experimental analysis or numerical simulations. Shale matrix has low porosity and permeability resulting from nanometer-scale pore sizes. Surface properties of shale can be quite inhomogeneous arising from complex mineralogy and diagenesis. Heterogeneous morphology and topology of the pore structure poses significant challenges on understanding fluid distribution and flow capacity. Pore scale simulations provide insight into the fundamental mechanisms of thermodynamics and hydrodynamics in tight porous materials, and can supplement experimental characterization of shale petrophysical properties (e.g. absolute/relative permeability, capillary pressure curves). However, challenges exist in creating representative pore structures tailored for specific simulation tools, incorporating the appropriate and relevant physics for the problems to be simulated, and interpreting, calibrating, or validating the simulation results. In this work, we used two types of pore scale simulation tools, namely pore network modeling (PNM) and lattice Boltzmann method (LBM), to study gas adsorption/desorption and transport behavior in shale matrix. For the first part of the work, a dual-scale PNM was integrated with lattice density functional theory (LDFT) to study nitrogen adsorption/desorption in mesoporous materials with pore sizes smaller than 200 nm. Critical pore structure parameters (i.e. porosity, pore size distribution, and pore throat connectivity) were characterized by calibrating the simulated nitrogen sorption isotherms to experimental results, and were then used to construct PNMs to study supercritical methane transport. We found that the pore structure characterization results were nonunique and highly dependent on the assumed pore shape. Scanning electron microscope (SEM) images were used to further constrain the description of pore shapes. Advection and diffusion of methane at reservoir conditions were simulated and compared, and suggestions were made regarding the choice of representative pore shape in PNMs for single phase advection/diffusion calculations. We next used LBM to study two-phase thermodynamic and hydrodynamic problems in nanopore systems in shale. Both 2D and 3D LBM models were developed with consideration of mesoscale fluid-fluid and solid-fluid interactions to model gas adsorption in complex geometries, and phase separation occurs automatically without the need to track the interface. This overcomes the pore shape deficiency of PNMs in cases where nanoporous media reconstruction exists. LBM models were then calibrated to LDFT and validated against experimental adsorption data for both subcritical and supercritical gases for the first time. We studied and compared nitrogen sorption hysteresis in two model nanopore system reconstructions representing the interparticle and intraparticle pores in shale. As another example of many possible applications of our developed model, we studied water adsorption and condensation in a reconstructed clay pore structure based on SEM image analysis, and explored the effect of surface wettability on adsorbed/condensed water distribution and connectivity. Supercritical methane flow simulations with the existence of condensed water were conducted using a 3D hydrodynamic LBM model that considers nanoscale flow physics for high Knudsen number flow. The relative permeability of methane as a function of water saturation and surface wettability was calculated and compared to available experimental data measured on geosynthetic clay liners. We demonstrated the wide applicability of our model and suggested future applications