Investigation of steam explosion duration in stratified configuration

Highlights: • Analysis with MC3D code provided insight into stratified steam explosion phenomenon. • Explosion duration may indicate nature of explosion development. • Weak explosions reasonably well predicted with assumed premixed layer. • Explosion duration significantly underestimated for strong explosions. • In strong explosions probably important layer mixing during explosion. - Abstract: A steam explosion is an energetic fuel-coolant interaction process, which may occur during a severe reactor accident when the molten core comes into contact with the coolant water. In the stratified melt-coolant configurations, where the corium melt is released in a shallow water pool or in an initially dry cavity later flooded with coolant water, the melt-water interaction may occur on the interface between the bottom melt layer and the upper water layer. It was long believed that stratified melt-water conditions cannot generate strong explosive interactions, because it was assumed that there is no significant premixing of melt and coolant water prior to the shock wave propagation. However, the recently performed experiments in the PULiMS and SES (KTH, Sweden) facilities with corium simulant materials revealed that strong steam explosions may spontaneously develop also in a stratified melt-water configuration, where a formation of a considerable melt-water premixed layer was observed. Understanding the premixture formation mechanisms is crucial for predicting the stratified steam explosion energetics. One of the external indicators of the way the melt and water are mixed during the stratified steam explosion process is the explosion duration. In the article, the duration of the stratified steam explosion is studied by the analysis of experimental observations in the PULiMS and SES tests using simulation results obtained with the MC3D code (IRSN, France). The explosion characteristics are analysed by varying the available melt mass in the premixed layer, the triggering position, the melt droplet diameter, the void fraction in the premixed layer, the void fraction in the water above the premixed layer and the size of the fine fragments. The analysis showed that longer explosion duration for the experiments with strong explosions cannot be predicted solely by varying the premixed layer conditions. A significant increase of the explosion duration could be obtained only by increasing the size of the fine fragments, which depends on the nature of the explosion. Thus the observed discrepancies between the experimental results and simulations can be explained by the fact that in strong stratified steam explosions there is also significant melt-water layer mixing during the explosion propagation.