Ex-vessel core melt phenomena have been studied to ascertain the feasibility of mitigation by water flooding or other means. The EC 5th Framework Programme ECOSTAR (Steinwarz et al., 2001) is concentrating on three important areas in relation to: melt release from the RPV, ex-vessel corium transport and long-term corium cooling. This programme builds on earlier projects CSC (Cognet et al., 1999), COMAS (Steinwarz et al., 1999) and CIT (Adroguer et al., 1999) to enhance the understanding of complex ex-vessel core melt behaviour, especially dispersion processes and jet formation, and their consequences.

To date, melt dispersion experiments using water/nitrogen fluids have indicated that lateral failures of the lower head lead to less melt dispersal out of the reactor cavity than do failures at the central part of the lower head. The new programme will examine the impact of fluid density on this conclusion. The erosion of different concretes with jets of iron melt and also oxide jets has also been studied. These experiments show that a metallic jet eroded the base-mat more deeply but that the oxide jet eroded a greater amount of the concrete. Melt dispersion experiments have been carried out in the DISCO facility at FZK, Karlsruhe and jet erosion is being studied in the KAJET facility, also at FZK.

Ex-vessel transport has been studied in the COMAS facility at the CARLA plant of Siempelkamp. This focuses on the spreading and distribution of the melt under molten core coolant interaction (MCCI) conditions. In a representative test, approximately 350 kg of oxidic melt are spread over a flat surface of siliceous concrete.

Reactor material experiments are in progress in regard to long-term stabilisation of the melt. Experiments with simulants have shown that phase segregation may exist within oxidic corium. Simulant experiments have been performed in the VULCANO facility at CEA using ZrO2(Al2O3. Experiments have also been conducted in the ISABEL facility to determine the plane front solidification limits. Experiments are conducted to examine the efficiency of both top and bottom flooding as a means of cooling. Several series of experiments on dryout and quenching with different particulate beds have been conducted at KTK. A large-scale top flooding of a melt pool experiment has been carried out at Siempelkamp. Bottom reflooding is being examined at FZK in 1D and 3D.

The melt coolability and concrete interaction (MCCI) (NEA Annual Report, 2002) project at Argonne is managed by the USNRC and aims to provide experimental data on the spreading of molten debris over the base of the containment and the effectiveness of water cooling from the top. It also aims to provide information on the 2D interaction of the molten corium with the concrete structure of the containment, including the kinetics.