Methodology for bounding calculations of nuclear criticality of fissile material accumulations external to a waste container at Yucca Mountain, Nevada

The possibility of nuclear criticality, however remote, in the vicinity of the proposed repository at Yucca Mountain, Nevada generates justified concerns and may impact the performance of the repository. A heuristic approach is presented here for determining the amount, spatial distribution and other characteristics of fissile material accumulation in the rock beneath a waste package that could contribute to such an event. This study is concerned primarily with waste packages containing special spent fuel from the Department of Energy and high-level nuclear waste glass. Mixing with less alkaline waters and the subsequent drop in pH is the mechanism that is most efficient for precipitating fissile material from the waste package internal leachate, in contrast to natural deposits in which redox changes are the main precipitation driver. External accumulation size is determined by (1) computing the chemical composition of the leachate leaving a package as its internal materials degrade (with the batch geochemical code EQ3/6), (2) determining precipitation of fissile material into mineral phases (using the 1D geochemical code PHREEQC) as the effluent mixes with percolation water, and (3) heuristically scaling results to a 3D volume and computing the criticality coefficient (using the code MCNP). Loci for accumulation are the multiple lithophysal cavities and the fracture system. A bounding conservative approach is used by necessity in Step 3. Nuclear criticality is sensitive to small variations in the distribution of fissile material and parameters of natural systems vary by orders of magnitude. Because the most likely combinations of parameters are not conducive to nuclear criticality, this study focuses on extreme values of parameter probabilistic distributions, such as limited flow into the package associated with a large percolation rate, combinations of material degradation rates favoring actinide release, and very high host-rock porosity values. By considering these combinations, most favorable to criticality but unlikely, it was concluded that external nuclear criticality is not a concern at the proposed repository.