August 9, 2014

On August 6th, Tepco announced that the fuel core of unit #3 at F. Daiichi may have melted about five hours earlier than had been previously estimated. In addition, the company said the added time-frame suggests that most of the corium (a mixture of melted fuel, control rods, structural metals and instrumentation) ate its way through the bottom of the reactor pressure vessel (RPV) and fell to the concrete base-mat beneath. It is estimated that the corium then penetrated 0.68 meters into the ~2 meter thick base-mat. Tepco cautions that their “analysis entails some degree of uncertainty”. Their degree of uncertainty might be considerable, if not astronomical.

On July 26, 2013, Tepco’s president Naomi Hirose vowed to improve Tepco’s public disclosure policy, saying “even if the evaluations do not show enough evidence, we will swiftly and honestly mention risks and worst-case scenarios without fearing the impact.” (Fukushima Commentary, August 24, 2013) His statement was in response to incessant allegations of non-transparency by the Japanese Press and parallel pressure from the Nuclear Regulation Authority. In effect, Tepco said that if everyone wants worst-case scenarios, that’s what they will get. The recent announcement on unit #3’s core damage may well be a computer-generated worst-case assumption. Although the results were based on sophisticated computer simulations, the degree of confidence in the conclusion is only as good as the data fed into the program. Much of the data fed into the simulation seems decidedly assumptive, thus we might put the conclusion into serious question.

There are several reasons why. The first, and perhaps most important, reason is that there is no way to look at the actual condition of the RPV’s bottom head and inspect the condition of the base-mat beneath, at this point in time. It is hoped that the use of Muon tomography will give the world a decent idea of where the solidified corium is located. But, this is currently in the developmental stage. It could happen as soon as next year. Regardless, assumption-based conclusions on the location of unit #3’s corium, at this point, cannot be given any degree of confidence; even those using the best computer simulations money can buy. The proof will be in the pudding, and Muon tomography should show us how well everything turns out.

Next, the data used to run the computer simulation contains numerous technical assumptions. One assumption is that prior estimates of the amount of water injected into the RPV by the High Pressure Coolant Injection system (HPCI) were too high. By re-crunching the numbers relative to HPCI, Tepco now estimates that less water was being fed into the core and allowed the degradation of the Zirconium tubes (cladding) holding the uranium pellets much earlier than prior timelines have shown. The heat generated by this was thus assumed to be much greater than earlier estimated, causing the fuel to melt about five hours prior to that previously proposed. But, did this actually happen?

Tepco has posted that even though HPCI operated, perhaps at a lower-than-previously-assumed output, until 2:42 am on March 13, “We assumed that no cooling water was injected into the reactor after 20:00 (8pm) on March 12.” This, in itself, is highly speculative and smacks of a rather questionable worst-case speculation. Even if HPCI was working at a reduced capacity, some water must have been getting inside the unit #3 RPV until HPCI stopped! If HPCI were merely “spinning its wheels”, it would have over-heated the mechanism and caused it to fail many hours before it actually did. Of all the assumptive data fed to the computer, this may well be the most speculative. Tepco’s chart on water level in the unit #3 RPV began to drop at about 8pm on March 12th, but it took more than four hours for that level to reach the top of the fuel core (TAF). If there was absolutely no water being injected during this period, the level decrease would have been much faster; not nearly so gradual. This implies that some water was being injected by HPCI – maybe not as much as HPCI is designed to supply, but certainly not nothing.

Yet another assumption in the computer data-feed concerns the fact that the water level in the RPV, as measured by the precious little instrumentation that was working at the time, stopped decreasing about the time HPCI failed. Tepco assumes that the water-level instrumentation must have been damaged and was giving false readings. They based this assumption on the fact that calibrations run months after-the-fact showed the instrument readings to be in error. However, the calibrations occurred after the hydrogen explosion with unit #3 at 11am on March 14. The shock of the detonation may well have caused instrumentation damage. The instrumentation itself is designed to operate effectively at all pressure and temperature conditions up to, and including, the point where the safety relief valves (SRV) open. The SRVs first opened at 9:36am on March 13. Although there seems to be no technical reason why Tepco is assuming the instrumentation must have been giving false readings, the assumption is made nonetheless.

Further, the measured water levels after TAF was exposed, never completely uncovered the entire fuel core. Even the computer simulation itself estimated that the core was never completely uncovered until the moment that the SRV’s were opened, depressurizing the system, and allowing low pressure fire pumps to inject water and quench the RPV’s internals. Without full core “uncovery” of the fuel core for many minutes, a full meltdown of the fuel is quite unlikely.

In addition, Tepco estimates that when about a fourth of the core was still water-covered (5:10am), the temperature above the core reached more than 2150oF (1200oC). This was not a temperature reached in air, but must have occurred in a steam-shrouded environment. In an enclosed steam environment, pressure rises in parallel with temperature increases. The saturated steam pressure would have been well in excess of the set-points for the SRVs to lift and exhaust into the suppression pool (Torus). This would have meant that the SRVs should have lifted on or before 5:10am – not 9:36am! Plus, we’re not looking at one or two valves that might have been stuck shut or malfunctioned. We’re talking about at least six of them. It is important to point out that SRVs work against spring pressure. They are fully “passive” in that no power source is needed for their operation. To assume all of these passive devices “stuck shut” and allowed pressure to skyrocket well above the design set-points borders on the absurd.

The above should be sufficient to make any rational adult skeptical of Tepco’s claim of a melt-through with unit #3 at Fukushima Daiichi. This does not mean that Tepco’s new estimate of unit #3’s core damage should be entirely rejected. For example, it is possible that HPCI injections were less than previously assumed. However, to say that there was no water going into the core at the point Tepco assumed is absolute speculation. In fact, without the conjecture that HPCI did nothing while it was unquestionably operating, the conclusion collapses.

It should be noted that nuclear energy worst-case scenarios rarely turn out to be correct. In fact, they are almost entirely wide of the mark. Cleverly-created worst-case scenarios, relative to nukes, have historically been the domain of hardened nuclear critics. To date, none of their tacit “guarantees” have come to fruition. Tepco’s recent worst-case scenario with respect to Fukushima unit #3 may be no better than the fantastical fabrications spawned by unprincipled prophets of nuclear energy doom.

References –

1 –

2 –