UC Berkeley

One of the unresolved hurdles facing present and future nuclear energy systems is waste management. Recent efforts to identify promising fuel cycles have attempted to value waste management through metrics reflecting intrinsic characteristics of the wastes rather than the environmental impacts, risks, and challenges involved in their disposal. This dissertation applies three waste management models to a broad set of fuel cycle options to explore the relationship between fuel cycle characteristics and waste disposal. First, long-lived fission product inventory is used as a proxy for repository performance, avoiding the computational expense and complexity of a full performance assessment. Second, the attractiveness for diversion and recovery of fissile materials from waste streams for potential proliferation is evaluated using a figure of merit, which values the quality of the material and its retrievability. Finally, repository area and surface storage requirements are determined based on thermal constraints for waste disposal in three geologic environments.

These models are used to analyze fuel cycle options taken from the Fuel Cycle Evaluation and Screening (FCES) study. A Python package has been developed to characterize waste streams using FCES data including mass balances, discharged fuel composition(s), and details about fuel cycle technologies such as reactor type and reprocessing method. The package extends the FCES mass flow calculations to include waste package loading for spent fuels and waste loading fraction for high-level wastes. Benchmarking against FCES metric results demonstrates good agreement.

The long-lived fission product inventory is shown to be sensitive to reactor thermal efficiency — because reactors that are less efficient are required to produce more fissions — and to the extent of recycling, because the recycled fissile material generally has greater long-lived fission product yield than enriched uranium. Specific fission product isotopes demonstrate sensitivity to fuel isotope, neutron spectrum, and residence time. Fissile material in all waste streams is shown to become attractive for recovery from waste as the self-protecting dose rate decays, but the time before that occurs is longer for high-level wastes that concentrate highly radioactive nuclides. Because high-level wastes are more dilute in fissile material, more waste packages need to be intercepted to obtain usable quantities of fissile material. The area and storage time requirements are shown to be highly sensitive to thermal properties and constraints of repository design. Fuel cycles recycling long-term heat-generating transuranic isotopes in fast reactors perform well, whereas those that utilize limited- recycling of actinides perform poorly. Parametric analysis of waste package and waste form loading with respect to these metrics demonstrates opportunities for the integration of fuel cycle operations and waste management.