Evaluating Alternative Groundwater Flow Models with Geochemical Mixing Targets

Andrew V. Wolfsberg1, Zora V. Dash1, Edward M. Kwicklis1, Gregory J. Ruskauff2, Alaa H. Aly2

1 Los Alamos National Laboratory, awolf@lanl.gov, kwicklis@lanl.gov, zvd@lanl.gov, Los Alamos NM, USA
2 INTERA Inc. Greg.Ruskauff@nv.doe.gov, Alaa.Aly@nv.doe.gov, Las Vegas, NV and Denver, CO, USA

ABSTRACT

The Nevada Test Site Underground Testing Area (NTS-UGTA) Program is currently conducting corrective action investigations to predict the transport of radioactively-contaminated groundwater. The investigations involve the construction and calibration of multiple groundwater flow models for future use in contaminant migration simulations. Different geologic models address conceptual uncertainty in the stratigraphy and fault structures within the domain and different recharge/ boundary flow conditions address water balance uncertainty. For each combination of geologic and water balance conceptual models, groundwater flow models are calibrated to observed heads and estimated boundary fluxes. Then, reverse-transport modeling is used to predict locations and quantities of upstream groundwater sources mixing at specific downstream wells. The predictions are then compared with independently developed geochemical mixing model results that identify mixing ratios of source waters at downstream wells based on groundwater chemical compositions. Differences between simulated mixing ratios with the reverse transport model and geochemical targets are used in quantitative assessment of the various models. A statistical cluster analysis identifies groups of models with similar geochemical residual characteristics and is incorporated in the assessment, ranking, and discrimination of the alternative calibrated flow models for future use in contaminant transport predictions.