Towards an European consensus on possible causes of MCCI ablation anisotropy in an oxidic pool

•Major phenomena determining the 2D ablation during MCCI are identified.•The overall heat transfer coefficient to pool interfaces deduced from MCCI experiments is around a few 100W/m2/K.•The individual heat convection mechanisms do not account for the main trends of 2D ablation deduced from MCCI tests.•2D ablation is very dependent on the composition of concrete components and aggregates.•Pool/interface structures (crust, slag layer, gas film) determine the heat transfer towards interfaces. One of the most disputed issues raised by molten corium concrete interaction (MCCI) is how the 2D cavity ablation in an oxidic pool evolves: why is the ablation anisotropic with siliceous concretes and isotropic with carbonaceous concretes. The work performed in the frame of the SARNET2 WP6 group during the last 4years has enabled significant progress on this topic. This paper summarizes this progress using the analysis of recent 2D real material experiments in an oxidic pool and from analytical simulant experiments on 2D heat convection in a bubbling pool, including calculations and recalculations with MCCI codes available in Europe. Firstly, the effective heat transfer coefficients from the bulk pool to the bottom and lateral pool interfaces deduced from MCCI experiments lead to a range of a few 100W/m2/K. By contrast, a detailed review of possible 2D convection mechanisms shows that the individual heat convection mechanisms (without taking a crust into account) such as gas bubbling convection and solutal convection overestimates the overall heat transfer coefficient, and does not account for the main trends of 2D ablation deduced from MCCI tests, which are very dependent on the composition of concrete components and aggregates. This fact, in turn, points to the effect of more complex pool/concrete interface structures. On the basis of a thorough interpretation of the experimental database and of a detailed comparison of MCCI code predictions, a set of the most realistic and consistent assumptions are identified and major remaining uncertainties are listed.