The analysis and interpretation of MCCI-OECD and VULCANO experiments with real material in the case of a homogeneous pool configuration, performed during the last seven years, have brought valuable data concerning the 2D ablation features and indirectly the structure of the pool/concrete interface versus the concrete type.
This analysis of available real material MCCI tests concerning the heat transfer distribution within the pool suggests that no stable crust exists at the pool/concrete interfaces and that the heat is convected by gas bubbling up to a conductive boundary layer; at the opposite of the previous modelling approach proposed in last MCCI seminar in September 2007, the temperature at the interface between convective and conductive zones, the so-called solidification temperature, is not imposed. The heat convection transfer within the homogeneous pool is supposed to be isotropic and according to the analysis of existing 2D MCCI experiments, the 2D ablation is determined by the profile of the thermal resistance along the pool interfaces. The influence of concrete on the 2D ablation is explained mainly by the increase of the thermal resistance at the bottom corium/concrete interface in case of a siliceous one compared to the case of a limestone-sand concrete. This permits to build an improved modelling approach applicable to the reactor case for the homogeneous pool configuration or for the oxidic layer in case of a stratified pool.
Parametric calculations using the MEDICIS module of ASTEC code with the proposed modelling approach have been performed on a generic PWR reactor design using conservative boundary conditions. The present reactor study is focussed on the case of siliceous concrete. Indeed this concrete type might promote the pool stratification with metal below and lead then to a faster ablation kinetics than other concretes with higher gas content as it was shown already with the previous approach.
The new approach enhances the lateral ablation compared to the axial one in case of siliceous concrete in a more pronounced way than the previous approach did. The reason is that the lateral heat transfer is proportional to the difference between the bulk temperature and the ablation temperature and is no more limited by the solidification temperature but only by the thermal resistance of the pool/concrete interface which is assumed to be lower for the lateral pool interface than for the bottom one.
The results of presented calculations point out that the prevailing lateral ablation in the oxidic layer or homogeneous pool could counterbalance the influence of metal layer in case of stable stratification promoting at the opposite the axial ablation. It is shown more precisely that if using a realistic stratification criterion with a superficial gas velocity threshold for pool stratification onset consistent with simulant BALISE data and suppressing stratification for a thin metal layer, the axial melt-through time becomes comparable or even longer in case of a pool configuration evolution scenario with siliceous concrete than in case of a fixed homogeneous pool with limestone-sand concrete.
The new modelling approach gives similar results in case of a limestone-sand concrete compared to the previous one since the isotropic heat flux distribution in the homogeneous pool is maintained. Moreover it is shown that, if using a realistic stratification criterion as mentioned above, the pool stratification with metal below is very unlikely due to the high gas content in case of a limestone-sand concrete.
Finally the demonstration that, even in most severe conditions, no early basemat melt-through (before around 4 days or less) is possible excepted in case of a thin basemat appears to be within easy reach. To complete this demonstration it remains to exclude definitely the possibility of a sustained pool stratification with metal below promoting a fast axial ablation. This goal will be met first by confirming with a refined analysis of existing and future tests the prevailing lateral ablation in case of any concrete with a low gas content promoting pool stratification and second by modelling properly the heat transfer distribution in case of an intermediate pool configuration which appears to be more likely and long lasting than a stratified pool one.
Reaching this goal will not prove that concrete basemat melt-through can be avoided in dry conditions except in case of a very low corium inventory; therefore the issue of stopping MCCI by a cooling process, the so-called mitigation issue, has still to be solved.