Revisiting Nitao's Analytical Model with Laboratory Experiments of Partially Saturated Fracture-Matrix Infiltration

We investigate infiltration into a single fracture embedded in an initially unsaturated sandstone with homogeneous matrix properties (doi: 10.1029/2023WR036323).

To outline the control of dual-porosity mechanisms, a classical analytical framework developed by Nitao (doi: 10.1029/91WR01369) was applied to model the observed infiltration behavior. Our study considered flow dynamics in terms of penetration depth, dominating flow regime (matrix- or fracture-dominated) related to applied flow rates, wetting front propagation in both domains, and the interference of matrix imbibition with the lateral boundary of the system. The employed model accounts for the matrix imbibition effect on fracture flow propagation.

Most interesting, matrix imbibition affected the observed discontinuous, partially saturated fracture flow (a combination of slugs and films) to behave, on average, like plug flow. Within the range of applied flow rates above a critical threshold, we found the model's plug flow assumption is not a relevant precondition for its applicability. Corresponding to the matrix imbibition state, fluid propagation in the fracture exhibits three characteristic scaling regimes (FP1-3). Only two scaling regimes are established for flow rates below a critical threshold, hence required to recover bulk infiltration for the chosen geometry. Furthermore, wetting fronts switch from fracture- to matrix-dominated at moderate to high flow rates, indicating a flow-rate-dependent limitation of fracture-dominated infiltration depth (source-responsive). While the scaling regimes agree with experiments for applied flow rates above the critical threshold, the model underestimates the initial penetration depth below. Here, we observe the direct onset of flow regime FP2 and the delayed transition into FP3.