Multiscale pore network modeling and flow property analysis for tight sandstone: a case study

Abstract

Abstract Digital rock characterization enables high-fidelity quantification of core samples, facilitating computational studies of physical properties at the microscopic scale. Multiscale tomographic imaging resolves microstructural features from sub-nanometer to millimeter dimensions. However, single-resolution volumes preclude capturing cross-scale morphological attributes due to the inverse relationship between the field of view and resolution. Constructing multiscale, multiresolution, multiphase digital rock model is therefore imperative for reconciling this paradox. We performed multiscale scanning imaging on tight sandstone samples. Based on pore network model integration algorithms, we constructed dual-scale pore network model (PNM) and fracture-pore hybrid network model to analyze their flow characteristics. Results showed that the absolute permeability of the dual-scale PNM exhibited a distinct linear increase with the number of extra cross-scale throats and throat factor, but the rate of increase became smaller when the throat factor exceeded 0.6. For dual-scale pore network with cross-scale throat and throat factor of 1 and 0.7, the predicted porosity matched experimental results well. For the fracture-pore hybrid network model, the relationship between absolute permeability and cross-scale throat properties is similar to the dual-scale PNM. When fluid flow was parallel to the fracture orientation, permeability increased markedly with fracture aperture as a power-law function. However, the dip angle did not induce obvious permeability variation trends across different flow directions.