A Monte Carlo paradigm for capillarity in porous media

Abstract

Wet porous media are ubiquitous in nature as soils, rocks, plants, and bones, and in engineering settings such as oil production, ground stability, filtration and composites. Their physical and chemical behavior is governed by the distribution of liquid and interfaces between phases. Characterization of the interfacial distribution is mostly based on macroscopic experiments, aided by empirical formulae. We present an alternative computational paradigm utilizing a Monte Carlo algorithm to simulate interfaces in complex realistic pore geometries. The method agrees with analytical solutions available only for idealized pore geometries, and is in quantitative agreement with Micro X‐ray Computed Tomography (microXCT), capillary pressure, and interfacial area measurements for natural soils. We demonstrate that this methodology predicts macroscopic properties such as the capillary pressure and air‐liquid interface area versus liquid saturation based only on the pore size information from microXCT images and interfacial interaction energies. The generality of this method should allow simulation of capillarity in many porous materials.