A methodology for the evaluation of fuel rod failures under accident conditions

Abstract

Recent studies on the long-term behaviour of high-burnup spent fuel have shown that, under normal conditions of storage, challenges to cladding integrity from various postulated damage mechanisms, such as delayed hydride cracking, stress-corrosion cracking and long-term creep, would not lead to any significant safety concerns during dry storage, and regulatory rules have subsequently been established to ensure that a compatible level of safety is maintained. However, similar regulatory rules have not yet been developed to address failures of fuel rod cladding that could potentially lead to reconfigured fuel geometry under hypothetical transport accidents. At issue is the effect on cladding ductility of potential changes in zirconium hydride morphology during dry storage. Recent studies have shown that above a certain level of cladding hoop stress, the decaying temperature history during dry storage can cause the hydrogen in solid solution to precipitate in the form of radial hydrides, which, depending on their relative concentration, can induce brittle failures in the cladding. From a US regulatory perspective such cladding failures, if they were to cause fuel reconfiguration, could invalidate the cask's criticality and shielding licensing analyses, which are based on coherent geometry. This paper describes a methodology for high-burnup spent fuel to determine the frequency of cladding failure and failure modes under drop accidents, considering end-of-storage spent fuel conditions. The degree to which spent fuel reconfiguration could occur during handling or transport accidents would depend to a large extent on the number of fuel rod failures and the type and geometry of the failure modes. Such information can only be developed analytically, as there are no direct experimental data that can provide guidance on the level of damage that can be expected. To this end, this paper focuses on the development of a methodology for modelling and analysis that deals with this general problem on a generic basis. First, consideration is given to defining accident loading that is equivalent to the bounding hypothetical transport accident of a 9 m drop onto an essentially unyielding surface. Second, an analytically robust material constitutive model, an essential element in a successful structural analysis, is required. A model of material behaviour, with embedded failure criteria, for cladding containing various concentrations of circumferentially and radially oriented hydrides has been developed and implemented in a finite-element code. The hydride precipitation model, which describes the hydride structure of the cladding at the end of dry storage, and the hydride-dependent properties of high-burnup fuel cladding form the main input to the constitutive model. The third element in the overall process is to utilise this material model and its host finite-element code in the structural analysis of a transport cask subjected to bounding accident loading to calculate fuel rod failures and failure mode configurations. This requires detailed modelling of the transport cask and its internal structure, which includes the canister, basket, fuel assembly grids and fuel rods. The overall methodology is described.