Applications du code intégral ASTEC dans le réseau SARNET

Titre du congrès :ICONE - 16th International Conference On Nuclear Engineering Ville du congrès :Orlando Date du congrès :11/05/2008

The ASTEC integral code is being developed by IRSN (France) and GRS (Germany) for simulation of severe accidents (SA) in Light Water reactors (LWR): safety studies including the evaluation of source term, Probabilistic Safety Analysis level 2 (PSA2) and assessment of SA management actions. It plays a key-role in the SARNET Network of Excellence on R&D on severe accidents (2004-2008) of the 6th Framework Programme of the European Commission. A substantial effort is being made to disseminate ASTEC and to perform jointly-executed research activities with the ultimate objective of providing physical models for integration into ASTEC and making it the European reference integral code. Thirty partners are assessing the ASTEC V1 successive versions through validation against experiments and benchmarks on plant applications with integral and mechanistic codes. This paper presents an overview of the work done in 2006 with the version ASTEC V1.2rev1 released by IRSN and GRS in Dec.05. In particular, this version included improvements of the documentation, mainly users manuals and guidelines, and CEA model improvements for the corium behaviour in the vessel lower head. For ASTEC adaptation to BWR and CANDU, the needs concern mainly the in-vessel core degradation and the corresponding specifications are being written. As to ASTEC validation, applications were performed on the following physical phenomena and experiments: circuit thermalhydraulics (PACTEL T2.1 and ISP33, PMK2, LOFT-LP-FP2); core degradation (CORA-13 and W2, QUENCH-11, LOFT-LP-FP2, Phébus FPT4); fission products release and transport (COLIMA, STORM, Phébus FPT0-1-2); Molten-Corium-Concrete-Interaction or MCCI (ACE L4, BETA, OECD-CCI2); containment thermalhydraulics and aerosol behaviour (LACE LA4, ThAI, PACOS Px1.2); and iodine in a multi-compartment containment (ThAI). The results are in general good, often close to results of mechanistic codes. Some models reach the limits of present international knowledge, for instance MCCI and hydrogen production during the reflooding of a degraded core. As to ASTEC benchmarking, applications for diverse accident scenarios (LOCA, Loss of Steam Generator Feedwater and SBO) were performed for several reactor types: PWR 900, Konvoi 1300, Westinghouse 1000, VVER-1000 and VVER-440. The main trends of results are similar to results obtained with MELCOR or MAAP4 codes. Some quantitative differences are due to modelling differences, i.e. on core degradation. The preliminary comparison on fission products results will be extended in 2007-08. Good results were obtained in the comparison with mechanistic codes such as ATHLET-CD, RELAP5-3D or COCOSYS. Exploratory calculations on VVER-440 showed the ASTEC capabilities to evaluate the possibilities of In-Vessel Melt Retention. For CANDU reactors, physically reliable results were obtained on fission products transport and behaviour. The ASTEC V1 code evolution will now be limited to feedback from the IRSN current Probabilistic Safety Analysis level 2 on 1300 MWe reactors and from SARNET applications. From now on, IRSN and GRS are preparing the new series V2 of ASTEC versions, taking into account the code evolution needs as expressed by the SARNET users. The first V2.0 release is planned at the end of 2008: the version will be applicable to EPR and will include the advanced core degradation models of the ICARE2 mechanistic IRSN code. In 2008, the partners will switch to the assessment of the V1.3rev2 version that was delivered in Dec.2007 and present their results at the 3rd ASTEC Users' Club organised by IRSN in April 2008.