Topological origin of flow distributions in disordered porous media

Abstract

We investigate steady Stokes flow through porous media composed of two-dimensional disordered arrays of circular obstacles. We develop a theory for the statistics of flow rates based on a pore-network model that captures local flow correlations. We show that the flow rate distribution across the ensemble of pore bodies follows a Gamma distribution, and that the flow rate distribution through pore throats is fully determined in terms of it. Furthermore, this Gamma distribution can be directly linked to simple geometrical properties such as the coefficient of variation of pore throat widths, rendering the model parameterisable from minimal medium information. The resulting predictions agree closely with computational fluid dynamics simulations and show markedly better agreement than prior mean-field models that neglect local flow-rate correlations, clarifying how local splitting and merging shape flow in disordered porous networks.