BEHAVIOUR OF CANDU® FUEL AFTER A LOSS OF COOLANT ACCIDENT IN AN IRRADIATED FUEL BAY

Abstract

After CANDU fuel is discharged from the reactor, it is submersed in large pools of water in Irradiated Fuel Bays (IFBs) to remove their decay heat. It has been postulated that a large-scale drainage scenario i.e. Loss of Coolant Accident (LOCA) in an IFB could lead to a release of radioactivity if the fuel reaches temperatures sufficient for a runaway oxidation of the Zircaloy-4 sheathing. The near miss at the Fukushima Diachii Spent Fuel Pools (SFPs) has prompted international efforts to better define the safety margins after irradiated fuel has been exposed to air. The purpose of this work is to support the development of an IFB severe accident code, specifically for CANDU type nuclear generating stations. A literature review and stress analysis of a fuel element were performed to determine the risk of overstraining and ballooning due to heat-up in a depressurized environment. It was concluded that overstrains were possible for elements with extremely high fission gas release, but only after embrittlement of the sheath due to oxidation. In the development of a severe accident code, defects arising from overstrain may be correlated to temperature to simplify the analysis. If a mass failure of fuel sheaths is predicted, the effect of ballooning is expected to become significant. In the concomitant analysis, the heat transfer behaviour of a 2-D bundle cross-section was simulated using COMSOL Multiphysics®. Several sensitivity studies were completed at steady-state to explore the effects of ballooning, fuel-sheath gap size and fill gas composition on the maximum temperatures in the fuel. It was shown the inclusion of conduction in the element’s interior enhanced heat transfer and resulted in slightly lower maximum temperatures. Transient analyses were also completed to determine when oxidation (1) becomes a significant effect, and (2) is estimated to transition to linear reaction kinetics after which it may become self-sustaining. The transition was predicted to occur within 2.6 hours for a 4 kW bundle, which could be considered as an upper bound for the heat generation produced by an exposed IFB bundle. Therefore, due to the time required for heat up and the delay in the initiation of breakaway, it was indicated that sufficient margin exists for operators to intervene before a passively cooled, isolated bundle undergoes a breakaway reaction. The cumulative effect of neighbouring bundles has the potential to adversely impact the heat dissipation process. Scaling up to a full-scale CANDU IFB or even IFB rack is difficult due to the complex bundle geometry and open rack design. In the last part of this work, a process to predict the steady-state temperature and mass fluxes of air as it passes through a rack of spent fuel using analytical models and computational fluid dynamics (CFDs) techniques is presented. The scenario acts as a lower bound estimate for the temperatures expected during a complete loss of coolant in a fuel bay by examining the 3-D heat-up of a stand-alone quarter rack without flow resistance of the bundles. The correct incorporation of flow resistance, in a detailed convection model that can be validated by experiment, is a necessary step before conclusions could be made about the safety of IFBs. However, the analysis as summarized in Section 8, using a CFD model for a 0.35 MW fuel rack, indicates that the maximum temperature of the air within the rack was about 575 K and located at the centre of the outlet. This result is encouraging to support the safety of IFBs, as the temperature is well below that required for a breakaway reaction.