A Hydrologic-Geophysical Method for Characterizing Flow and Transport Processes within the Vadose Zone
Research Objectives Many DOE facilities inadvertently have released toxic or radioactive wastes into the unsaturated, or vadose zone, that comprises the near-subsurface beneath their sites. Contamination sources include leaky underground storage tanks and landfills, as well as cribs, ponds and trenches where hazardous wastes were essentially dumped on the ground and allowed to infiltrate (GAO, 1998). Such waste disposal methods were the result of not fully appreciating the potential public
... tential public health hazards posed by the contaminants, as well as a lack of awareness of the complexity of vadose zone flow and transport processes. This complexity and our continued lack of ability to account for important flow and transport processes in predictive modeling is underscored by the many documented cases of contaminants being found where they are not expected and/or not predicted. For example, at Yucca Mountain in Nevada, chlorine 36 from airborne nuclear detonations at the Nevada Test Site during the 1950's was found at much greater depths than thought possible. Inaccurate estimates of vadose zone flow and contaminant transport rates most likely result from simplifying assumptions required by the large scale and complexity of real world problems as well as the inability to incorporate pertinent physics governing flow and transport within the vadose zone into predictive models. These assumptions range from the treatment of heterogeneity where clastic sedimentary formations typically have complex hierarchical assemblages of bedforms. Here the structure and hierarchy of the bedforms are the end result of complex geologic and hydrologic processes. For example differences in sediment supply and stream flow can produce lateral and vertical variations in the bedforms at a variety of scales. It is very difficult to measure of the spatial variability of these systems due to the large number of samples required to statistically describe the full nature of the heterogeneity. Air permeameter techniques show promising results in rectifying this problem (cf. Davis et al, 1994) , but such measurements require extensively exposed outcrops. An obvious example of the need to adequately account for heterogeneities in the vadose zone is underscored by the common observation of preferential flow where small perturbations tend to grow with time eventually producing finger like flow paths producing fast transport pathways. A second problem associated with characterizing/modeling heterogeneous deposits is measuring hydraulic properties and up-scaling these properties to formation scale. Often field samples are collected from borehole cores or from excavations at a site. Hydraulic properties are then determined through laboratory procedures where significant precision and accuracy errors are commonly introduced. This error arises not only from the experimental method, but also from using repacked samples where little, if any of the original pore structure remains intact. The measurement error from hand samples may greatly exceed the spatial variability present in geologic formations. Additional problems in predicting flow and transport within the vadose zone is 'heterogeneity' resulting from direct human intrusion. For example, a report released by DOE on the Hanford site (DOE,1998) has identified man made structures such as unsealed wells penetrating contaminated zones, and funnel flow around tanks and waste cribs, as possible mechanisms for increased rates of contaminant migration. How these man made structures couple with the naturally occurring hydrologic conditions to influence transport is not well understood. Because of this, the estimated rates of contaminant migration within the vadose zone may often be in error. A better understanding of flow and transport modes within the vadose zone beneath DOE sites, including the influence of heterogeneity and man made structures, could lead to better conceptual models and therefore to better estimates of contaminant transport properties and rates. Incorporating this knowledge into predictive and performance assessment models would provide a sound basis for characterizing and/or removing some of the unknowns plaguing these programs. To address some of these issues, we performed a series of intermediate-scale field experiments to test assumptions concerning flow and transport mechanisms within vadose zone materials comprised of clastic sedimentary deposits similar to those underlying Hanford and other DOE sites in the western United States. The experiments involved geophysical imaging of water followed by a solute tracer introduced under controlled conditions at the ground surface. In addition we developed a series of numerical geophysical/hydrologic property estimation tools that allow for better estimates of spatial heterogeneity of hydrologic properties within the vadose zone. The benefits of the work to the DOE include: • Better conceptual models regarding flow and transport mechanisms active in natural heterogeneous vadose zone deposits and such deposits altered by human intrusion.