Contaminant Travel Time Estimates through a Thick Unsaturated Zone at Rainier Mesa and Shoshone Mountain, Nevada

While significant differences in the scale and persistence of preferential flow paths may exist between soil and rock, empirical evidence suggests that contaminant travel times are similar in diverse unsaturated porous media. In this investigation, residual contamination from underground nuclear testing within an approximately 1 km thick unsaturated zone at Rainier Mesa and Shoshone Mountain (RM/SM) motivates our testing of preferential flow theory in variably-saturated rock. We have developed a conceptual model for unsaturated-zone flow through RM/SM in which flow occurs solely as or in combinations of (i) preferential flow in fractures and faults, (ii) preferential matrix flow as fingers or funneling, and (iii) non-preferential matrix flow. Our conceptual model guides the application of a simple, empirically-based approach to estimate radionuclide travel times through the subsurface to the regional water table. Potential contamination flow paths include percolation through the weapons detonation chambers, flow through the access tunnels, and leaching of infiltration basins that accommodate drainage from some of the tunnels. Unlike the majority of preferential flow models, the transport travel-time model we employ depends solely on whether water supply is continuous or intermittent with time and does not distinguish between soil and rock. For the arid to semi-arid climate at RM/SM, a hypothetical continuous water supply results in simulated travel times that are orders of magnitude faster than those of intermittent water supply. The potential transformation of intermittent fluxes into continuous fluxes via focusing with depth of convergent preferential paths such as faults, fractures, fingers, and funnel flow would decrease estimated contaminant travel times at RM/SM. Friable volcanic tuffs, with no obvious preferential flow paths, could still support matrix-dominated preferential flow mechanisms such as funnel and finger flow resulting in solute travel times shorter than estimates based on non-preferential matrix flow.