Simulation of the LIVE-J Experiments using MELCOR: Findings and Current Limitations

Abstract

During the late in-vessel stage of a Severe Accident (SA) the debris is assumed to accumulate as molten pool in the lower plenum of the reactor and form various layers according to the material density. This is the assumption adopted in the In-Vessel Melt Retention strategies for example. The case of Fukushima Daichii however, recent analyses suggest a different accident progression, where at first a debris bed forms in the bottom of the vessel, which later starts melting from the lower melting point material which permeates through the high-melting point debris bed. This configuration was reproduced experimentally at the LIVE-3D facility at KIT, using ceramic beads and nitrate as simulant materials. Heat up due to radioactive decay is simulated using internal wire heaters, and two experiments are performed: the LIVE-J1 experiment, which focused on the melting of the nitrate into the ceramic debris bed, and LIVE-J2, focused on the heat transfer characteristics inside the debris bed. It is still unclear whether SA codes, such as MELCOR, are capable to simulate this scenario and reproduce correctly the heat-transfer on the wall. Hence, in this paper we validate the MELCOR COR model against the LIVE-J2 experiment. Results are then compared with the ones previously obtained from CFD simulations, showing that the MELCOR model predicts higher pool and lower vessel temperatures. Applying an HTC derived from CFD calculations gives an improved forecast of both.