FLOW AND RADIONUCLIDE TRANSPORT IN THE SATURATED ZONE BENEATH YUCCA MOUNTAIN: PAST ACCOMPLISHMENTS AND NEXT STEPS

The U. S. government has decided that it is appropriate to proceed with the preparation of a License Application to dispose of high-level radioactive waste at Yucca Mountain. Hydrogeologic site characterization studies of the Saturated Zone (SZ) during the past twenty or more years have built confidence in our understanding of the hydrogeologic setting and fundamental processes controlling groundwater flow and radionuclide migration. This paper summarizes the past suite of characterization and modeling activities of numerous investigators and presents results of recent saturated zone radionuclide transport modeling. In addition, a discussion of future efforts that might be undertaken is presented. In the current Total-System Performance Assessment models for the Yucca Mountain Project, the system as a whole meets performance requirements even though uncertainty in the SZ barrier performance is relatively large. Hydrogeologic variability and conceptual model uncertainties are factors that preclude precise predictions of solute transport in any natural system, and the Yucca Mountain SZ is no exception. Therefore, future work should focus on building confidence in SZ conceptual and numerical models. Potential activities to build confidence in our understanding should, to a large extent, parallel those representing the current state of the art in the science of flow and contaminant transport in groundwater. In the author’s opinion, these include: field investigations of the influence of faults on large-scale groundwater flow and contaminant transport; transport field tests combined with laboratory studies aimed at discovering fundamental transport mechanisms; development of integrated, multiscale coupled models of the hydrologic system; application of stochastic hydrology theory to radionuclide transport in the Yucca Mountain SZ; and application of theories and methods for treating uncertainties in conceptual models.