American Nuclear Society

Home / Publications / Journals / Nuclear Technology / Volume 196 / Number 3

Core Melt Composition at Fukushima Daiichi: Results of Transient Simulations with ASTEC

H. Bonneville, L. Carenini, M. Barrachin

Nuclear Technology / Volume 196 / Number 3 / December 2016 / Pages 489-498

Technical Paper /

First Online Publication:October 18, 2016
Updated:December 7, 2016

The Accident Source Term Evaluation Code (ASTEC) is used to perform numerical simulations of the accidents at the Fukushima Daiichi nuclear power station in the frame of the Organisation for Economic Co-operation and Development/Nuclear Energy Agency Benchmark Study of the Accident at the Fukushima Daiichi Nuclear Power Station (BSAF) project. At present, simulations are available for Units 1, 2, and 3 of Fukushima Daiichi and for 6 days from the earthquake. A clear lesson from phase 1 of the project was that the uncertainties in the functioning of the safety systems and in accident progression are still large and there are many explanations for the measured thermohydraulic behavior. Rather than focusing on the thermohydraulic key parameters for which comparisons with measurements are available, this paper will address melt composition computation results that may provide insights relevant for the decommissioning process.

When molten corium relocates from the core down to the vessel lower head, the melt jets interact with water and may be totally or partially fragmented depending on the level of water. A U-Zr-O-Fe molten pool may form in the lower head, and because of chemical reactions, separation between nonmiscible metallic and oxide phases may occur. The models implemented in ASTEC enable the simulation of these phenomena. Up to five different axisymmetric corium layers in the vessel bottom head can be formed, which are, from bottom to top, a debris layer, a heavy metallic layer, an oxide layer, a light metallic layer, and another debris layer. An important process is the UO2 fuel reduction to metallic uranium by nonoxidized zirconium, which results in uranium transport to the dense metallic layer as demonstrated in the MAterial SCAling (MASCA) program.

Because of the large consensus on the accident progression of Fukushima Daiichi Unit 1, in this paper we present complex melt compositions before vessel failure for the current best-estimate cases for Unit 1. We do not present similar work performed for Units 2 and 3.

It should be underlined that in the case of vessel bottom failure, a part of this complex melt will relocate to the pedestal and molten core–concrete interaction will take place enhancing other complex physical phenomena with possible large consequences on the melt chemical composition and behavior.