Ensuring the corium confinement within the containment is part of the design requirements for the European Pressurized Reactor (EPR). A specific strategy has then been imagined to collect and cool down the corium: it relies on corium retention in the reactor pit up to complete discharge from the vessel, then on spreading in a 170 m2 large core catcher, followed by flooding by water flowing upon its surface and within a cooling device located under the basemat.
As the heat is generated inside the corium as a volumetric power source, the heat flux to be evacuated per basemat surface unit grows with the corium height. Consequently, a research program, aimed at assessing the possibility to widely spread the corium, and correspondingly trying to forecast the maximal possible height of melt that needs to be anticipated, has been lead at the “Institut de Radioprotection et de Sûreté Nucléaire” during the last decade. It involved the development of a simulation software, namely the CROCO code, which modelling was consolidated and qualified by various experimental programs.
These latter schematically fall into two categories : experiments using simulant fluids, at low - CORINE1 - or high temperature - KATS - and experiments using prototypical melts – e.g. VULCANO2. The CORINE tests were equipped with a very accurate instrumentation, and provided an in-depth characterization of the corium flow, including the knowledge of the local temperature within the melt. The tests program included spreading with crust growth at the contact with the basemat, possibly with gas sparging to simulate the effects of corium/concrete interaction during the spreading. They were complemented by the KATS tests, which simulant material was heated at sufficiently high temperature to observe the effects of heat radiation to the atmosphere and interaction with the substrate. Finally, the VULCANO program, using prototypical corium melts, gave insight into specific material effects. It contributed to demonstrate, in particular, that the solidifying corium behaves like a slurry, and can be described as an equivalent homogeneous mixture with an effective viscosity. The continuously deforming skin behaviour of the upper crust, when the difference between melt solidus and liquidus temperatures is large, was also evidenced during these tests.
Despite extensive technological developments, scaling ratios between experiments and reactor case remained, until the end of the experimental programs, very large (from 300 kg of corium for largest experimental facilities to 300 tons for the « severe accident dry scenario » used for the EPR design). Consequently, it appeared that the development of software using the Computational Fluid Dynamics (CFD) approach, able to compute temperature and momentum exchanges at the flow boundaries instead of relying on experimentally fitted correlations, would allow more reliable extrapolation to the reactor: this was the basic choice for the CROCO software. This computer code describes all the aspects of the considered phenomenology, namely the solidifying corium flow, exchanging heat by convection with the basemat and by radiation with the atmosphere and interacting with the substrate. The liquid and solid phases are described by and equivalent slurry, obeying Newtonian or non-Newtonian constitutive laws; the gaseous phase, resulting from the concrete decomposition, is treated independently.
Applications have been performed to assess the capabilities of the EPR core catcher; they evidenced that a complete corium spreading onto the core catcher is obtained for a high enough discharge mass flow rate from the reactor.