I believe that given the extent and timing of the core damage (as manifest by the hydrogen explosions and sudden pressurization events) that some fraction of the core debris is already found on the drywell floor, especially for the higher-power units 2 and 3.

Unfortunately no-one knows the answer to this critical question with any significant degree of certainty. The main reason is lack of information about the status and activation of various systems in the reactor coolant system (RCS) and thereby the timing of events that lead to core uncovery.

It should also be noted that BWRs have their lower head full of penetrations for the control rods and the instrument tubes. The instrument tubes are not supported on the outside and they are susceptible to failure by melting of the welds that seal them in place, and there is no significant cooling from the outside to prevent such failures.

The presence of core material on the drywall floor has major significance on the long-term stabilisation of the reactor. The significance depends on how much debris is on the floor. Let’s suppose at one extreme there is a major fraction on the floor, and it is inside the pedestal area, where the walls of the reactor vessel are supported, where the debris bed would be quite deep actually, and then it is very likely it would not be covered by water. If the debris is not coolable, then it will attack the concrete, and depending on the kind of concrete, it would generate gases, and if there are unoxidised materials in the debris, hydrogen, which would vent to the outside along with radioactive aerosols found in the containment. I don’t know how much of a basemat the unit has, but am concerned about the long-term possibility of the debris melting through the bottom, and also melting through the pedestal walls that hold the reactor vessel.

If, on the other hand, all or the vast majority of core material is in the lower head of the reactor pressure vessel, then the question is, how can you cool it there, since you can’t fill up the drywell all the way to submerge the reactor vessel, since the suppression chamber, which drains the wetwell, is damaged at least in unit 2. In BWRs, even if the lower head could be submerged, there are some very big questions about whether outside cooling is sufficient to stabilise it. This method of cooling would be effective in keeping debris inside the lower head of PWRs, based on my experience with a concept of severe accident management we established for PWRs of this power [750 MWe]. However, this is not the case for BWRs. In addition to all of the lower head penetrations, which PWRs do not have, their flatter lower head shape and the metal skirt underneath the lower head near its attachment to the reactor vessel, both tend to prevent vapour from escaping. In such a case, there would be vapour blanketing the lower head from underneath, making it very difficult to cool.

Given the amount of radiation and hydrogen that the reactors have been putting out, I would expect core melt and degradation to a significant extent. Given this, and the fact that the TEPCO has put a lot of salt in it [by injecting seawater], I am puzzled about why the Japanese are saying that there is going to be a major fraction of the core still inside the core region and the reactor pressure vessel. I hope that would be the case, but even if it were, it would be hard to stabilise the melt inside the lower head now.

The presence of salt could complicate the process of stabilising of the reactors, but it could also be helpful. We cannot know the extent of each until we have more information about where the core debris is. Obviously, salt water injected into the core, combined with core water boiling, tends to increase core salt concentration. Accumulated salt can plug up coolant passages. But at high temperatures, the salt would melt, and circulate through the debris by natural convection. In that way, it would function as a coolant.

The principal issue is finding out where the core materials are. Only then can we reach the next step. I know it is extremely difficult to do so, but I don’t know if it is impossible. For the time being we need to be concerned about the open-ended water flow through the system, and the associated releases of radioactivity to the environment.



The optimistic point of view

Richard Lahey, the Edward E. Hood professor of engineering at the USA’s Rensselaer Polytechnnic Institute, was head of R&D for GE BWR thermal-hydraulic and safety technology. He spoke briefly to Will Dalrymple:
“I worked for GE in R&D on BWRs, including the Fukushima design type, and worked on the Fukushima plant in particular. Then in 1975 when I came to Rensselaer, we developed the APRIL simulation code to look at this kind of core melt and how it propagates in BWRs. What we found, which was since verified by experiment, is that the core preferentially melts through the control rod guide tube bottom, and comes out there rather than through the bulk head. That can be good news.
“The bad news is that it comes out early. At Three Mile Island 2 [during the famous March 1979 accident at the Pennsylvania, USA-based PWR], they were able to quench it before it came out. Although a core melt and relocation happens earlier in a BWR, it emerges from the lower head in a distributed way; the bottom head is perforated like a colander.
If corium is distributed on the floor evenly, that is not bad news, because it is easier to cool than a lump. Because they have been spraying water into the reactors, water should pool there to a depth of 500mm, and the water will sparge out any radioactive aerosols from the corium-concrete interaction.
But we do not know for sure that this is happening; there is no real instrumentation or visualisation. If core melt and relocation is happening, you really need to know, so you do not do something crazy like cut off the water or entomb the reactor with concrete.
If I were there, I would try to get a nuclear-hardened camera into the drywall and see the state of the lower head.