PREMIX experiments PM12-PM18 to investigate the mixing of a hot melt with water

Abstract

In the case of a hypothetical core melt-down accident in a pressurised light water reactor, hot melt may be relocated and mixed with water present in the lower coolant plenum. The amount of masses involved in the mixing process and the intensity of thermal interaction determine the extent of a possible steam explosion. Such processes have been investigated, on a technical scale, at the Forschungszentrum Karlsruhe in so-called PREMIX experiments. Melt was released from above into a water pool using alumina instead of corium. Eighteen tests, PM01 to PM18, have been performed from 1994 to 1999. PM12 to PM18 were a task of the EU-MFCI-project within the Fourth Framework Programme. For that purpose, geometrical key data of the experimental facility as well as starting conditions of some of these tests were chosen to meet as close as possible those of the FARO/FAT tests performed at JRC Ispra with molten corium and water. This was in order to facilitate comparison of results. Three tests (PM12, 13, 14) were started with almost identical parameters. The general course of events turned out to be very similar. Deviations nonetheless found in the results can generally be attributed to uncertainties in the procedure of preparing and controlling the melt supply. By this, reproducibility of the PREMIX experiments has been proven. The conditions of two others tests (PM16, 17) were set to meet those of two FARO tests, L-28 and L-31. These were the melt mass, water temperature and depth, system pressure, nozzle diameter, and duration of melt release. In most cases, under saturated water conditions, melt penetration, premixing, and steam production occurred in such a way that the bulk of water was prevented from close contact with the melt. Steam and water were in equal shares, around 50%, in the interaction zone. In case of large initial subcooling of the water, the share of the steam and, with it, the average distance between melt and water in the interaction zone were much smaller. Nonetheless, a steam explosion did not occur. The influence of the various parameters on the results is discussed. Special interest is devoted to the evaluation of the jet break-up length. The report gives a documentation of all relevant data. The data should be used for the validation of multiphase computer programmes which are presently developed in reactor safety research. (orig.)