Laboratory experiments and dual‐domain modeling of infiltration dynamics in partially saturated fractured porous media

Abstract

Abstract Many studies have shown that fracture networks can provide rapid pathways for water infiltration and time‐dependent recharge in the vadose zone of consolidated fractured rock formations. To investigate the control of fracture networks and fracture–matrix interaction on preferential flow and infiltration dynamics, we constructed a simple network system consisting of m × 2 sandstone blocks placed in between two glass plates. Water was injected from a point source directly into the fracture system with a constant flow rate of 1.5 g min−1. We used a dual‐porosity non‐equilibrium model to model the discharge dynamics and the internal fracture–matrix mass exchange and observed strong deviations from the laboratory results when the original parameterization was used. This specifically concerns the matrix–fracture volume ratio κ and the fracture flow velocity v that require modifications in terms of their definition and underlying process description. Although the original model assumes a perfectly coupled fracture and matrix domain, in the experiments the discrete nature of the fracture network led to a much stronger dominance of the rapid flow domain and hence to a reduction of κ. The newly introduced parameter κ* includes additional effects and processes related to the time‐dependent evolution and smaller size of the fracture–matrix interface. Furthermore, experiments of varying total vertical system size reveal convergence toward a unique parameter set and the existence of a representative elementary volume (REV) for the chosen setup. Though it performs less well for very small systems below REV scale, the unique parameter set describes discharge dynamics in sufficiently large systems with high accuracy.