The FPT0, FPT1 and FPT2 tests were respectively performed in the Phebus facility in 1993, 1996 and 2000. The objectives of these three tests were to study the bundle degradation and the fission products releases during a severe fuel damage scenario, in different conditions : oxidising in FPT0 and FPT1 tests, and steam-starved in the FPT2 tests. These two different conditions were obtained using different steam flow rates injected at the bottom of the twenty rods bundle.
From the beginning of the tests interpretations, different calculations using ICARE2 code were performed. The first ones used the V2 mod1 version, and during the years the successive new versions and so new models were used. This leads to not necessary use, for the three tests, the same modelling options or models to calculate the bundle degradation (cladding burst, oxidation run-way and materials relocations).
In this presentation it is shown that with the same hypothesis and models used for the three tests (with only two justified exceptions concerning the cladding dislocation criteria and the double face oxidation in the burst zone), and using the last code version (ICARE2 V3 mod1.4) with standard candling model, the test calculations indicate good results compared to the measurements (fuel temperatures, Hydrogen production, duration of the steam-starved phase for FPT2 and materials relocations). Even if the state of the calculated damaged zone in the bundle is not exactly the same than the state given by the post-mortem examinations, the material mass distributions are however rather well calculated in comparison to the experimental ones.
Moreover, the new recommended “best-fitted” correlation (using Cathcart and Prater correlations) combined to the introduction of the blanketing effect, leads to very satisfactory results for these three tests.
In the next future, standard candling model should have to be replaced for these three calculations, by the “MAGMA” modelling.