Within the SARNET network of excellence in the 6th Framework Programme of the European Commission, the severe accident integral code ASTEC, jointly developed by IRSN (France) and GRS (Germany), has been validated against international experiments to evaluate the suitability and capability of new or improved models implemented in successive code versions up to V1.3rev2, delivered in December 2007. This paper focuses on the code applications concerning circuit thermal-hydraulics and core degradation to integral and separate-effect experiments: for the CESAR thermal-hydraulic module, BETHSY 9.1 b, PACTEL ISP 33 and T2.1, PMK2-SBLOCA, LOFT-LP-FP-2; for the DIVA core degradation module, CORA-13 and -W2, QUENCH-11 and -13, LOFT-LP-FP-2, Phébus FPT-4, FARO L14 and L28, LIVE-L1, OLHF-1, FOREVER EC2. Besides, the TMI-2 accident has been analyzed using the CESAR and DIVA modules in a coupling mode. The emphasis was put on the following new or improved models: i.e. in CESAR, reflooding of an intact core, condensation in the pressurizer, sub-critical break flow correlation, and new pressurizer spray model; in DIVA, corium behaviour in the lower head and lower head mechanical failure. For thermal-hydraulics in the circuits, good results have been obtained with ASTEC on the three integral experiments that cover various thermal-hydraulic flow regimes: LOFT-LP-FP-2 in Western PWR geometry and the two PACTEL experiments in VVER-440 geometry. These good results have been confirmed by the validation done on several BETHSY integral tests. For core degradation, the ASTEC results are good for early-phase models of core heat-up, oxidation and hydrogen production (before any quenching phase) on different CORA and QUENCH experiments and on LOFT-LP-FP-2. For the in-vessel late-phase, the results can be considered as good regarding debris bed melting (Phébus FPT-4), corium fragmentation at slump into vessel lower plenum (FARO), molten pool behaviour in lower plenum (LIVE-L1), and vessel lower head mechanics (OLHF-1 and FOREVER EC2). Furthermore, the first two phases of the TMI-2 accident before core reflooding are very well calculated by ASTEC. The main remaining modelling weaknesses concern the reflooding of a degraded core and the corresponding hydrogen production. The implementation of detailed magma 2D relocation models in the new series of ASTEC V2 versions (the first one being released mid-2009) will allow a more realistic simulation of late phase phenomena up to the failure of the lower head.
Read full abstract