| description abstract | The drift-shadow effect describes capillary diversion of water flow around a drift or cavity in porous or fractured rock, resulting in lower water flux directly beneath the cavity. This paper presents computational simulations of drift-shadow experiments using dual-permeability models, similar to the models used for performance assessment analyses of flow and seepage in unsaturated fractured tuff at Yucca Mountain. Comparisons were made between the simulations and experimental data from small-scale drift-shadow tests. Results showed that the dual-permeability models captured the salient trends and behavior observed in the experiments, but constitutive relations (e.g., fracture capillary-pressure curves) can significantly affect the simulated results. Lower water flux beneath the drift was observed in both the simulations and tests, and fingerlike flow patterns were seen to exist with lower simulated capillary pressures. The dual-permeability models used in this analysis were capable of simulating these processes. However, features such as irregularities along the top of the drift (e.g., from roof collapse) and heterogeneities in the fracture network may reduce the impact of capillary diversion and drift shadow. An evaluation of different meshes showed that at the grid refinement used, a comparison between orthogonal and unstructured meshes did not result in large differences. | |
