Development and implementation of a dual-porosity pore network structure using X-ray computed tomography for pore network modeling purposes

Abstract

3-D pore network modeling based on high-resolution X-ray computed tomography (XCT) is a useful tool for simulating pore-scale processes and phenomena within porous media in fields such as chemical and petroleum engineering and groundwater hydrology. XCT images provide the opportunity to capture the true topology of the porous system, retaining important characteristics such as pore geometry, location, and connectivity. However, a major limitation of XCT is its inability to resolve features smaller than the image resolution such as intraparticle porosity and void-space within secondary phases such as clay and micrite, here called microporosity. Identifying this microporosity is important for modeling fluid flow through rocks lacking macropore connectivity or to better understand transport processes in systems where there is an apparent interaction between the bulk fluid and stagnant void space within the microporous phase. This work attempts to address the impact of incorporating microporosity into a physically representative macropore network on pore network models of single-phase permeability and quasi-static drainage. XCT images were collected for three geologic systems (a Castlegate Sandstone, an Indiana Limestone, and a Winterset Limestone) with very different pore structures, and then augmented with conventional core analyses to generate geologically based dual-porosity pore networks. A layered-sand system was imaged and analyzed as a semi-control for validation of our approach. The dual-porosity networks were developed using statistically generated pore networks for the microporous phases, which were integrated with the macropore networks generated on the resolvable void space. The results suggest that incorporating microporosity into a pore network model is essential for simulating permeability and drainage in systems of low macropore connectivity and high microporosity content such as the Indiana and Winterset limestones, but only aids in drainage simulation for systems of high macropore connectivity and low microporosity content such as the Castlegate Sandstone. The layered sand analysis suggests that a statistically based dual-porosity pore network could help predict both permeability and drainage. This research was only a first step towards developing an approach for the implementation of a dual-porosity pore network, and it highlights issues obstructing this process and the areas to which future research should be focused.