Behavior of high burnup fuels under RIA and LOCA conditions

Full text: The paper describes and discusses results from an experimental program performed at Japan Atomic Energy Research Institute (JAERI) for high burnup fuel behavior during a reactivity-initiated accident (RIA) and a loss-of-coolant accident (LOCA). The program is comprised of RIA-simulating experiments in the Nuclear Safety Research Reactor (NSRR), LOCA-simulating tests in Reactor Fuel Examination Facility (RFEF), and cladding mechanical tests. To provide a data-base for regulatory criteria, the behavior of LWR fuels during an RIA is being studied in the NSRR. Numerous experiments have been performed by using pulse-irradiation capability of the reactor, in order to evaluate the thresholds, modes, and consequences of fuel rod failure in terms of fuel enthalpy, fuel burnup, coolant conditions, and fuel design. As for PWR fuels, we have performed twenty-nine tests so far, and the burnup ranges 38 to 79 MWd/kg. The tested PWR fuels include different design, 14x14 and 17x17, and variety of cladding, Zircaloy-4, low tin Zircaloy-4, ZIRLO, MDA and NDA, and different grain sizes of fuel pellets. Through the experiments, we observed cladding failure and subsequent fuel fragmentation and mechanical energy generation. The results from recent NSRR experiments reflect the better performance of the new cladding materials in terms of corrosion, the thinner oxides and accordingly lower hydrogen content generated during irradiation in the PWR. It can be concluded that the improved corrosion resistance gives a larger safety margin against the PCMI (Pellet/Cladding Mechanical Interaction) failure. Differently from U.S. and many European countries, the Japanese LOCA criteria are defined not by ring compression tests with post-quench cladding samples but by integral LOCA tests. In the LOCA-simulating test, cladding samples experience the whole LOCA sequence, ballooning and rupture, oxidation at high temperature, and quench. The Japanese criteria are based on a threshold of conditions resulting in rod fracture into multiple pieces. The test is performed with fully controlled axial constraint at quench. In terms of high burnup fuel behavior in LOCA, we examined an effect of initial hydrogen concentration with un-irradiated, artificially-hydrided cladding and irradiated samples. A recent LOCA test indicates that failure boundary is not reduced significantly by PWR irradiation in the examined burnup level. Hence, in the burnup level of the present study, differences were not significant between irradiated and unirradiated specimens in terms of threshold of fracture during quenching, although the fracture threshold is reduced as initial hydrogen concentration increases. (Author)