IAEA FUMAC BENCHMARK ON THE HALDEN, STUDISVIK AND QUENCH-L1 LOCA TESTS

The International Atomic Energy Agency (IAEA) sponsored the Coordinated Research Project (CRP) on Fuel Modeling under Accident Conditions (FUMAC) to coordinate and support research on nuclear fuel modelling under accident conditions in member countries. The focus of the FUMAC CRP (2015- 2018) has been on loss-of-coolant accidents (LOCA). Various institutions performed fuel performance simulations of selected experiments using different fuel performance codes (e.g., FRAPCONFRAPTRAN, TRANSURANUS, ALCYONE, DIONISIO, SOCRAT, FTPAC, BISON, RAPTA) and system codes (e.g SOCRATE, ATHLET). One of the results of the FUMAC CRP is a comprehensive code-to-code benchmark of selected results, and a comparison of simulations with experimental data as well. This paper represents an overview of the current state-of-the-art of nuclear fuel simulation capabilities for LOCAs and paves the way to further analyses and future developments. More precisely, we discuss the results of the simulation of a subset of the experiments considered in the FUMAC CRP, i.e., (i) the Halden LOCA tests (IFA-650.9/10/11, but only IFA-650.10 is in detail presented in this paper), (ii) the Studsvik LOCA test NRC-192, and (iii) rod 4 of the KIT QUENCH-L1 bundle test. These experiments, briefly presented in the paper, cover a wide range of conditions relevant for LOCA scenarios from different sources. The presented benchmark results are considered in more detail at the end of the LOCA transient (e.g., time of failure, cladding outer diameter, cladding oxidation thickness…). The experimental data are always included in the comparisons, when available. The results are also critically discussed, with the aim of identifying modelling developments required for the improvement of LOCA analyses. Finally, the outcome is complemented with an uncertainty and sensitivity analysis in a separate paper in this conference.