Severe accident progression in the BWR lower plenum and the modes of vessel failure

• The melt attack on the structure in the BWR plenum is considered.
• It is found that the welds holding the IGTs and CRGTs would be vulnerable to failure before other components.
• It is projected that the bottom part of the IGTs will fall off and create a succession of 7 cm holes in the vessel.
• The loading on the BWR containment could be much less than that projected currently.
• This projected vessel failure scenario needs confirmation through experiments.

Most of our knowledge base on the severe accident progression in the lower plenum of LWRs is based on the data obtained from the TMI-2 accident. It should be recognized that the lower plenum of a BWR is very different from that of a PWR. Unlike the PWR, the BWR plenum is full of control rod guide tubes (CRGTs) with their axial structural variations. These CRGTs are arranged in a cellular fashion with each CRGT supporting 4 rod bundles. There are also a large number of instrument guide tubes (IGTs), each generally placed in the middle of 4CRGTs. Both the CRGTs and IGTs traverse the thick vessel bottom wall and are welded to their extensions which come to bottom of the core. The core-melt progression in the lower plenum is controlled by the structures present and they, in turn, influence the timings and the modes of vessel failure for a BWR.The uranium oxide–zirconium oxide core melt formed in the 4 fuel bundles is directed by the structure below toward the water regions in-between the 4 CRGTs. The FCI will take place in those water regions and some particulate debris will be created, although there is insufficient water for quenching the melt. A FCI may occur inside a CRGT if and when the melt enters the CRGT at its top opening or the melt in the water region between the four CRGTs breaches the wall of the CRGT.The important issue is whether the welding holding the IGT inside the vessel will fail and the bottom part of the IGT falls out creating a hole in the vessel with release of water and melt/particulate debris from the vessel to the dry well of the BWR containment. Similarly, the failure of CRGT could have water and melt/particulate debris coming out of the vessel. These modes of vessel failure appear to be credible and they could occur before any large-scale melting and melt pool convection takes place. These modes of vessel failure and the melt release to the containment will have very different consequences than those generated by the other modes of vessel failure.Such BWR plenum melt progression scenarios have been considered in this paper. Some results of analyses performed at KTH have been described. We believe that the issues raised are important enough to consider a set of experiments for verification and validation of the melt progression in a BWR plenum. Such experiments are proposed.