Modeling relative permeability of water in soil: Application of effective-medium approximation and percolation theory

Accurate prediction of the relative permeability to water under partially saturated condition has broad applications and has been studied intensively since the 1940s by petroleum, chemical, and civil engineers, as well as hydrologists and soil scientists. Many models have been developed for this purpose, ranging from those that represent the pore space as a bundle of capillary tubes, to those that utilize complex networks of interconnected pore bodies and pore throats with various cross-section shapes. In this paper, we propose an approach based on the effective-medium approximation (EMA) and percolation theory in order to predict the water relative permeability. The approach is general and applicable to any type of porous media. We use the method to compute the water relative permeability in porous media whose pore-size distribution follows a power law. The EMA is invoked to predict the relative permeability from the fully saturated pore space to some intermediate water saturation that represents a crossover from the EMA to what we refer to as the "critical region.'' In the critical region below the crossover water saturation S-wx, but still above the critical water saturation S-wc (the residual saturation or the percolation threshold of the water phase), the universal power law predicted by percolation theory is used to compute the relative permeability. To evaluate the accuracy of the approach, data for 21 sets of undisturbed laboratory samples were selected from the UNSODA database. For 14 cases, the predicted relative permeabilities are in good agreement with the data. For the remaining seven samples, however, the theory underestimates the relative permeabilities. Some plausible sources of the discrepancy are discussed.