Investigation of conceptual and numerical approaches for evaluating moisture, gas, chemical, and heat transport in fractured unsaturated rock

We investigate the utility and appropriateness of various conceptual and numerical approaches for modeling several flow and transport processes in the unsaturated zone (UZ) at Yucca Mountain, NV, using a one-dimensional (1-D) column extracted from a three-dimensional (3-D) UZ site-scale model. Simulations of steady-state and transient moisture flow, transient gas flow, tracer transport, and thermal loading scenarios are performed, using a variety of numerical approaches to treat fracture–matrix (F–M) interactions, including an effective continuum model (ECM) and several varieties of dual-continua models. For the dual-continua models, we investigate the effect of varying the number of matrix gridblocks per fracture gridblock, the formulation applied for calculating F–M interface area, and whether or not global matrix-to-matrix flow occurs (dual-permeability vs. dual-porosity models). The key findings of the work based on a 1-D column are as follows. For steady-state moisture flow, most of the numerical and conceptual models provide similar results for saturation and fracture flow profiles. The ECM adequately models the steady-state processes because the system is not too far away from F–M equilibrium. For both transient moisture flow and transient transport in a steady flow field, the ECM is not adequate in general. Within the dual-continua models, the formulation for F–M interface area can have a major effect on the hydrodynamic response to an infiltration pulse and tracer arrival at various horizons, with fracture responses becoming earlier as F–M interface area decreases. The number of matrix blocks also has a significant effect on response time, with the more accurate multi-matrix-gridblock models yielding slower fracture response times. For transient gas flow arising from barometric pressure variations, the ECM adequately models the process, because F–M gas flow occurs rapidly compared to the time scale of the barometric pressure variations. For thermal loading, preliminary studies indicate that the ECM does not capture all the significant physical processes, because rapid fluid and heat flow can occur in the fractures before the matrix has a chance to equilibrate.