Dynamic criticality analyses in nuclear waste repository design

Suggestions that self-perpetuating nuclear criticality accidents (autocatalytic criticalities) may result from retiring weapons-derivative plutonium (w-Pu) or high fissile-content wastes at a high-level nuclear waste repository have not been satisfactorily scrutinized. Additionally, it has not been demonstrated how considerations for autocatalytic criticalities might be incorporated into repository maturation simulations known as performance assessments. Salient issues for the consideration of autocatalytic criticalities in performance assessments include the identification of critical configurations and the characterization of contributing physical phenomena.;After adopting a representative repository plan, this dissertation describes and utilizes quantitative, stylized evaluations for the evolution of incipiently critical configurations following facility closure. Assessments are limited to manifestations arising from the dilapidation of 2 candidate w-Pu waste forms. The evaluations consider configurations both internal and external to the affected waste packages. Quantitative measures for manifestation characteristics are obtained by implementing transient, deterministic models for the contributing mechanical, thermal, chemical and nuclear phenomena. Stochastic assessments determine both the likely and less-certain contributions from these phenomena to package degradation, material migration and sustained nuclear reaction.;Configurational descriptions enable dynamic simulations of the accidents. Deterministic models for the coupled thermal, hydrodynamic and nuclear dynamics affecting various stages of external accident progression are presented. The conceptual foundations for the models incorporate comprehensive consideration of the physical processes potentially germane to external criticalities. The dynamic characteristics of two autocatalytic criticalities, from a hypothetical and a plausible external actinide configuration, differ under quantitative evaluation. Estimates for the radiological and mechanical consequences of the accidents quantitatively establish the origins of the worst-possible repository performance detriments.;This work demonstrates that quantitative methods and models for autocatalytic criticality assessment are compatible and exercisable with existing performance assessment treatments. Additionally, stylized evaluations indicate that the candidate w-Pu repository wastes do not present the qualitatively asserted criticality hazards. Finally, the worst-imaginable autocatalytic criticalities are shown to have tolerable physical consequences from the standpoint of intended repository function.