Interpretation of the results of the CORA-33 dry core boiling water reactor test

All boiling water reactor (BWR) degraded core experiments performed prior to CORA-33 were conducted under ‘wet’ core degradation conditions, in which water remains within the core and continuous steaming feeds metal-steam oxidation reactions on the in-core metallic surfaces. However, one dominant set of accident scenarios would occur with reduced metal oxidation under ‘dry’ core degradation conditions and, prior to CORA-33, this set had been neglected experimentally. The CORA-33 experiment was designed specifically to address this dominant set of BWR ‘dry’ core severe accident scenarios and to resolve partially phenomenological uncertainties concerning the behavior of relocating metallic melts that drain into the lower regions of a ‘dry’ BWR core (the ex-reactor experiments at Sandia National Laboratories will further address these uncertainties). CORA-33 was conducted on 1 October 1992, in the CORA test facility at Karlsruhe. A review of the CORA-33 data indicates that the objectives were achieved; i.e. core degradation occurred at a core heat-up rate (characterized by the absence of any temperature escalation caused by oxidation) and a test section axial temperature profile (at incipient structural melting) that are prototypic of full-core nuclear power plant simulations under ‘dry’ core conditions. Simulations of the CORA-33 test at Oak Ridge National Laboratory (ORNL) have required the modification of existing control blade-canister materials interaction models to include the eutectic melting of the stainless steel-zircaloy interaction products and the heat of mixing of stainless steel and zircaloy. The timing and location of canister failure and melt intrusion into the fuel assembly appear to be adequately simulated by the ORNL models. This paper will present the results of the post-test analyses carried out at ORNL based on the experimental data and the post-test examination of the test bundle at Karlsruhe. The implications of these results with respect to degraded core modelling and the associated safety issues are also discussed.