A novel approach to model hydraulic and electrical conductivity in fractal porous media

Abstract:

Accurate prediction of conductivity in partially-saturated porous media has broad applications in various phenomena in porous media, and has been studied intensively since the 1940s by petroleum, chemical and civil engineers, and hydrologists. Many of the models developed in the past are based on the bundle of capillary tubes. In addition, pore network models have also been developed for simulating multiphase fluid flow in porous media and computing the conductivity in unsaturated porous media. In this study, we propose a novel approach using concepts from the effective-medium approximation (EMA) and percolation theory to model hydraulic and electrical conductivity in fractal porous media whose pore-size distributions exhibit power-law scaling. In our approach, the EMA, originally developed for predicting electrical conductivity of composite materials, is used to predict the effective conductivity, from complete saturation to some intermediate water content that represents a crossover point. Below the crossover water content, but still above a critical saturation (percolation threshold), a universal scaling predicted by percolation theory, a power law that expresses the dependence of the conductivity on the water content (less a critical water saturation) with an exponent of 2, is invoked to describe the effective conductivity. In order to evaluate the accuracy of the approach, experimental data were used from the literature. The predicted hydraulic conductivities for most cases are in excellent agreement with the data. In a few cases the theory underestimates the hydraulic conductivities, which correspond to porous media with very broad pore-size distribution in which the largest pore radius is more than 7 orders of magnitude greater than the smallest one. The approach is also used to predict the saturation dependence of the electrical conductivity for experiments in which capillary pressure data are available. The results indicate that the universal scaling of the electrical conductivity is valid from the percolation threshold all the way up to the complete saturation point. Our results confirm those reported previously by Ewing and Hunt (2006) who argued that the electrical conductivity should follow universal scaling over the entire range of saturation.