A Framework for Multiphase Pore-Scale Modeling Based on Micro-CT Imaging

Abstract

We demonstrate how to use pore-scale modeling combined with high-resolution imaging to make predictions of multiphase flow properties. Experiments were performed on two sandstone samples that were mixed-wet after contact with crude oil: Bentheimer and a reservoir rock. Flow experiments were combined with high-resolution X-ray imaging from which the pore space, fluid configurations and local contact angles can be measured. We first show that both lattice Boltzmann modeling and a pore network model can predict the fluid occupancy to within experimental and model uncertainty in Bentheimer using the measured contact angles. We then used the greater computational efficiency of the network model to simulate flow in a large network representing the reservoir sample. By calibrating the contact angle to match the observed pore-by-pore arrangement of fluid, the model was able to make predictions of relative permeability and capillary pressure that were within the bounds of experimental and model uncertainty. The results provide a framework for predictive image-based pore-scale modeling, where wet and dry images of rock samples are used to characterize both the pore structure and wettability.