Analyse critique des corrélations donnant la cinétique d’oxydation de zircaloy par la vapeur d’eau. Identification de points faibles et propositions d’amélioration dans les codes de calcul « Accidents Graves ».
Recent code-to-code comparisons between calculations of the same Severe Accident (SA) plant sequences carried out in the COLOSS project showed large differences between codes (ICARE/CATHARE, SCDAP/R5, MAAP4, MELCOR, ASTEC) regarding zircaloy (Zry) oxidation effects: different T behaviour (T escalation, Zry melting and UO2 dissolution, melt relocation) and different H2 production (timing, total H2). Part of these differences were suspected to be due to 1) uncertainties on correlations at High Temperature (HT) for T>1800K where there is a limited amount of data, 2) steam supply limitation at HT in the boundary layer of the cladding and 3) final cladding oxidation after consumption of ß-Zr. In the continuation of the investigations carried out in the COLOSS project on the Zry oxidation modelling in SA codes, a further activity was done by NSI-KI, FZK and IRSN to propose improvements in « existing parabolic-correlation models » used in SA codes.
Improvements were reported in a set of 3 Nuclear Engineering and Design papers. This presentation summarizes the first paper (Part 1) dealing with progress in the treatment of zirconium oxidation kinetics in different severe accident (SA) codes and convergence towards an agreed database required for the reliable verification of sophisticated LWR core degradation models.
Focused on the comprehensive experimental studies the available information on the high-temperature oxidation kinetics of Zry is evaluated. Important discrepancies between results for the high temperature range are interpreted in terms of different experimental and evaluation procedures. The critical assessment identifies the following items, which require separate consideration in the simplified convention of using fixed reaction rate correlations for the high - and low temperature ranges: the co-existence of two oxide phase sub-layers gives rise to a transitional kinetic response in an intermediate temperature range. Towards higher temperatures the validity of the correlations approach (assuming reaction rate control as in a semi-infinite solid-state diffusion system) is restricted further, on the one hand by kinetic control within the gas phase (initial oxidation range), on the other hand by metal matrix consumption (final oxidation range). The analytical treatment of the two last-mentioned concomitant phenomena is given, allowing reduction of apparent discrepancies between experimental results. A reasonable base of combined data is thus identified, for a moment with certain uncertainty, which is not suitable for direct application.
In Part II, the kinetic base is verified and refined by statistical evaluation of complementary experimental data. Improved correlations are proposed in the transition and in the high temperature zone allowing implementation in SA codes.
In Part III, validation calculations with the ICARE2 code are presented showing improvements on the prediction of Zry oxidation effects.
(2) FZK, Karlsruhe
(3) NSI-KI, Moscow