Replies

SUMMARY

This paper has briefly described the approach taken by the BWRSAT Program
at Oak Ridge National Laboratory towards understanding the probable
sequence of events for an unmitigated BWR severe accident. There are
many associated uncertainties, and experimental verification of the
approach is certainly desirable.

For an unmitigated BWR severe accident involving the progressive relocation
of material from the core region into the lower plenum of the reactor
vessel, the control rod guide tube structure and the large amount of
water in the lower plenum would be expected to provide for distribution
and quenching of the relocating debris. Since the earliest relocation of
materials from the core region would consist of metals from the control
blades, channel boxes, and cladding, the lower portion of the bottom head
debris bed should be metals-rich. The subsequent collapse of fuel pellet
stacks into the lower plenum would provide an underwater decay heat
source and provide for continuous boiloff of the surrounding water.
After bottom head dryout, the debris bed temperature would begin to
increase.

The cluster of control rod guide tubes in the lower plenum would be
heated by the surrounding debris bed and would be weakened at high temperatures
to the point of failure. Loss of control rod guide tube
strength would cause collapse of the remaining standing outer regions of
Che core that: are supported by the guide tubes. This coLlapse would form
the upper portion of the bottom head debris bed while the stainless steel
mass of the control rod guide tubes would be subsumed into the surrounding
debris bed as they melt. Thus, there is expected to be a large
amount of stainless steel included in BWR bottom head debris.

As the bottom head debris reaches high temperature, failure of the bottom
head pressure boundary would occur at some point. Penetration failures
can occur by weakening of the stub tube welds supporting the control rod
drive mechanism assemblies or by failure of the instrument tube welds at
the reactor vessel wall. However, failure of a stub tube weld would only
cause a small downward motion of the associated control rod drive mechanism
assembly, and therefore, although gas blowdown would be initiated by
such a failure, gross release of debris from the vessel would not.

For the instrument tube, although there is nothing to prevent its complete
detachment from the vessel given weld failure at the vessel wall,
it seems probable that an earlier failure would be by opening of the tube
in the middle (hottest) point of the bottom head debris bed with subsequent
spillover of molten material into the tube with passage through the
vessel wall, causing heatup and creep-rupture of the tube just outside
the wall. Instrument tube failures in this manner would provide pathways
for release of molten debris from the vessel.

The individual components of the debris bed would be expected to leave
the vessel in the order in which they reach their melting points and
transform to the liquid state. Solid metallic material surrounding the
lower portion of the original instrument guide tube locations would be
ablated into the molten material flowing from the reactor vessel via
these pathways.

Gross failure of the portion of the reactor vessel bottom head underneath
the vessel support skirt would be expected to occur long after the penetration
failures discussed above. The reactor vessel bottom head wall is
thick, and there is relatively little wall stress after the vessel is
depressurized. BWR severe accident sequence calculations with the BWRSAR
code predict failure of the bottom head wall only after the majority of
the metallic debris has left the vessel.