MAIN LESSONS LEARNT FROM FISSION PRODUCT RELEASE ANALYSIS, FOR THE UNDERSTANDING OF FUKUSHIMA DAI-ICHI NPP STATUS

摘要

Over the months following the Fukushima Dai-ichi accident, the CEA (French Alternative Energies and Atomic Energy Commission) has gathered a technical crisis team with the purpose to analyse on-line the situation and to give a diagnosis concerning the status of the reactors and the spent fuel pools (SFP). Among all the information which were delivered, a specific attention was given to the Fission Product (FP) measurements as soon as they became available. Two main sources of FP data were analysed: the data delivered by TEPCO or other organisations and the FP measurements taken by the CEA through the Japanese radionuclide Particulate station (called JPP38) located at Takasaki, and part of the International Monitoring System of the Comprehensive Test Ban Treaty Organisation (CTBTO). The analysis was performed by taking into account the knowledge gathered by the CEA on the FP behaviour in severe accident conditions over the past decades, through the experimental programmes performed in France and abroad. In particular the main findings learnt from the VERCORS separate effect experiments and PHEBUS integral tests, classifying the FP release in four groups with decreasing volatility, has been largely considered. Based on these FP measurements, two major questions have been successively addressed: (1) where the release originated from the SFP, in particular the SFP of unit 1F-4, or from the reactor cores themselves? (2) How damaged were the cores? The methodology used to determine the origin of the FP release was based on the ratio of volatile FP of different half-lifes, which was compared to the calculated ratio from the cores and from the SFP. One generally used the ~(131)I/~(137)Cs ratio, but also the ~(132)Te/~(137)Cs or ~(136)Cs/~(137)Cs ratios when they were available, since these ratios are more reliable. The conclusion from this analysis is that the main source of the FP release came from the cores and not from the SFP. To assess the level of core degradation, a combined approach has been used: (1) quantitative analysis of the cumulative ~(133)Xe source term estimated from atmospheric transport modelling simulation based on several combinations of detection locations, thanks to the CTBTO network, (2) semi-quantitative analysis, based on the presence of FP of decreasing volatility within the measured samples, indicating an increasing degradation level, completed by the ratios of these FP over ~(137)Cs, which were compared to the calculated ratios in order to verify the consistency of the measurement. The conclusion from this analysis is that the three cores 1F-1, 1F-2 and 1F-3 were significantly melted. In addition, this analysis has highlighted the importance to get as soon as possible reliable FP measurements, in particular for low volatile FP which are generally more difficult to measure: indeed, during the early weeks of the Fukushima events, some confusing data were delivered, leading to the risk of doing a wrong diagnosis on the reactor or SFP status. Finally, associated with new and reliable FP instrumentations in the containment, this simple methodology of analysis could be a way to improve the severe accident management in such situations.