Wednesday, March 28, 2012

JNST: "A scenario of large amount of radioactive materials discharge to the air from the Unit 2 reactor in the Fukushima Daiichi NPP accident"

Journal of Nuclear Science and Technology has a paper by Fumiya Tanabe of Sociotechnical Systems Safety Research Institute published online on March 28, 2012.

Tanabe is the one who said last August that there was a second "meltdown" in Reactor 3 on March 20-21, in which the melted fuel dropped through the Reactor Pressure Vessel onto the floor of the Containment Vessel, releasing a large amount of radioactive materials that caused the spikes in radiation levels in wide areas in Tohoku and Kanto.

In November last year (11/19/2011), he also disclosed his analysis of the Reactor 2 Suppression Chamber, and concluded that it may have been damaged by the earthquake. This paper looks to be that analysis, now peer reviewed and published.

The paper was received on December 9, 2011, accepted final version for publication on January 24, 2012.

So it is possible to disclose the outline of the analysis before submitting the paper to a peer-reviewed magazine, and the magazine has no problem accepting and publishing the paper. (All those researchers in Japan and elsewhere in the world who withheld their data, analysis, research until their papers were published by peer-review magazines, what would you say now?)

From Taylor & Francis Online:

A scenario of large amount of radioactive materials discharge to the air from the Unit 2 reactor in the Fukushima Daiichi NPP accident

Fumiya Tanabe


Based on an analysis of the measured data with review of calculated results on the core melt accident, a scenario is investigated for large amount of radioactive materials discharge to the air from the Unit 2 reactor. The containment pressure suppression chamber (S/C) should have failed until the noon on 12 March 2011 only by seismic load due to the huge earthquake on 11 March or by combination of seismic deterioration and dynamic load due to steam flowing-in through safety relief valve. Opening of the two safety relief valves (SRVs) at 14 March 21:18 should have resulted in discharge of large amount of radioactive materials through the S/C breach with the measured air dose rate peak value of 3.130E-3Sv/h at 21:37 near the main gate of the site. The containment drywell (D/W) should have failed at 15 March 06:25, at the cable penetration seal due to high temperature caused by the fuel materials heating up on the floor of the D/W, which had flowed out from the reactor pressure vessel. Then large amount of radioactive materials should have been discharged through the D/W breach with the measured air dose rate peak value of 1.193E-2Sv/h at 15 March 9:00.

Is Tanabe saying that venting of Reactor 2 on the night of March 14, 2011, with the Suppression Chamber already broken due to the earthquake, caused a discharge of a large amount of radioactive materials through the Suppression Chamber?

What a tragi-comical sequence. Amateur hours.

If the Reactor 2 Suppression Chamber was already damaged by the earthquake, what about other Reactors?


Anonymous said...

The researchers are lost too

Anonymous said...

laprimavera "So it is possible to disclose the outline of the analysis before submitting the paper to a peer-reviewed magazine, and the magazine has no problem accepting and publishing the paper. (All those researchers in Japan and elsewhere in the world who withheld their data, analysis, research until their papers were published by peer-review magazines, what would you say now?)"

In my discipline, authors who submit papers to journals for publication, usually post full preprints of the manuscripts on the publicly accessible e-print archive before submitting the manuscripts to journals for peer review. I can't understand the excuses which the Japanese and other scientists are using to withhold data related to the Fukushima NPP disaster from the public.

BTW, laprimavera, send an e-mail to johnoah at gmail dot com if you want a full copy of the article by Tanabe.

JAnonymous said...

As I already said earlier, and I think this is exactly the proof I needed. An MSM paper and a peer-reviewed paper differ by several orders of magnitude.

This said, the real reason why scientists sit on their data waiting for peer reviews (which can take up to 2 years, how 'novel' is that when you publish 2 years late) is that they are SCARED that someone steals their idea and makes a better paper. No need to be scared if you are a good and confident researcher.

Actually, excellent scientists who do care, usually publish a 'technical report' as early as possible to warn the masses. About a disease outbreak, about flawed practices, and so on...

This makes the upcoming paper only better because of the time that people have to assimilate the information. And this inevitably hypes the paper, making non-scientific audiences want to know what's in the actual peer-reviewed paper.

Also this keeps reminding me about the farce of whaling for research, although this is getting OT.

Anonymous said...

It is common in Math and Physics to publish a preprint of your scientific work on before publication in refereed journals. Therefor the attitude to "hide" such important information seems strange to me.

elbows said...

Responding to this part of your blog post:

"Is Tanabe saying that venting of Reactor 2 on the night of March 14, 2011, with the Suppression Chamber already broken due to the earthquake, caused a discharge of a large amount of radioactive materials through the Suppression Chamber?

What a tragi-comical sequence. Amateur hours."

Venting details get a bit complex, especially at reactor 2 where it is unclear whether earlier vet-venting or the brief dry-venting actually happened. But a picture is emerging where the large release of radioactive substances at reactor 2 was caused because they failed to vent at the right time, they left it too late....

The SRV stuff they mention is to release pressure from the reactor vessel, not containment, They have to do this to get the factor pressure low enough to pump in water by fire truck.

But this release from the SRV goes to the suppression chamber. And this is where the big problem at reactor 2 starts. On March 12th they switched the source of water for the RCIC cooling system to use the water from the suppression chamber. But they failed to monitor the pressure and temperature of the suppression chamber at all until March 14th. So by the time the RCIC was failing on 14th and they needed to get ready to pump water, the suppression chamber was already in a state where it would not be able to properly deal with the steam released from the reactor via SRV.

To work around this problem, they needed to vent the suppression chamber. But these attempts failed. So they tried to dry-vent the dry-well instead, but this attempt either raffled totally or was only successful for a few minutes.

As a result of these failings, one or more of the following happened:

Reactor vessel only depressurised slowly, causing delays in water injection.

Suppression chamber got damaged because it couldn't handle the steam from the reactor SRV valves

Drywell failed whilst still under high pressure

The interim cabinet committee report from a few months ago is very good at explaining the failings that happened at all 3 reactors. Based on these details, here is what I believe they should have done to minimise radioactive release:

They should not have waited for the RCIC cooling system to fail. They should have switched to pumping water earlier rather than waiting till RCIC failed. Well before noon on the 14th they should have wet-vented via the suppression chamber, then depressurised the reactor via SRV, then pumped in water.

I can forgive them for problems getting the pumping & venting ready because the explosion at reactor 3 ruined a lot of their preparations. But it is inexcusable that they did not start monitoring the suppression chamber temperature & pressure until late on the 14th of March. If they had measured it sooner they would have realised that the RCIC would fail soon, and that suppression chamber was in a poor state to handle SRV. They would have had more time to successfully vent the suppression chamber (wet vent) and as a result we would probably not be looking at reactor 2 as the main culprit for serious contamination to Japanese land.

So to be clear, faults in Mark 1 containment mean that if the core melts, you have to vent, and the containment will still fail in some way anyway. This was unavoidable, it was the best that could be done for Fukushima. But at least wet-venting removes some of the worst radioactive substances, and the containment doesn't fail at a moment when pressure is very high. Failure to manage this at reactor 2 is what sets it apart from reactors 1 & 3.

Anonymous said...

The (planned, voluntary) venting would have brought with it responsibility for inevitable contamination. As things stand, it's an unfortunate accident and no-one is responsible.

elbows said...

@Anonymous That may have been a factor very early on, with reactor 1, but Im quite confident it didn't have much affect on later decisions. They tried to vent at reactor 2, including the much more polluting dry-vent, but they mostly failed for technical reasons. And they were not afraid to vent reactor 3 even when they knew the risk of the building exploding. Venting is the lesser of 2 evils.

If the rain had reached central Tokyo at the same time as the March 15th plume, rather than what actually happened with the rain not making it much past the mountains, things would have been much worse.

The political consequences of venting are not as bad as the consequences of venting. This is more true for wet venting than dry venting, but even dry venting is preferable to having delays pumping in water, dramatic containment failure, etc.

elbows said...

Oops that last paragraph was supposed to read 'The political consequences of venting are not as bad as the consequences of not venting'.

Anonymous said...

It should not be assumed that there was earthquake damage to the torus. A seismic analysis would be necessary. AFAIK, TEPCO has not done a formal seismic analysis on the torus, but I might have missed something in the deluge of information that has been posted (not all of it in English). However, the sloshing of water inside the torus due to the earthquake may have developed forces strong enough to cause damage. It is a question of resonance. The long duration of the earthquake is particularly troublesome in this regard.

elbows said...

If the SuppressionChamber had failed that early, then I don't know as the RCIC cooling system would have lasted as long as it did, since it was switched to use suppression chamber as a water source on March 12th.

Also they did not start measuring suppression chamber pressure until late on March 14th, so we don't have good data for the earlier period.

But we do have drywell pressure data from earlier, and interestingly when TEPCO and NISA ran analysis of the accident using models, they could not get the model output to match the actual measured pressure of the drywell during the early days of the accident, before reactor 2 had its real bad days on 14th and 15th. The model suggested that the drywell pressure should have increased more rapidly than was actually the case. They eventually got the model to match the measurements, by adding an assumption that some damage occurred to the drywell quite early on. This was the reason we had stories a long time about an assumed drywell leak of so many cm, but there isn't actually any real observations or data about this, its just an assumption to make their model fit the reality, so it seems plausible.

Even though they thought this early damage was in drywell rather than suppression chamber, I expect that early damage to the suppression chamber is still an alternative possibility, but there isn't much that tells us quite how likely this was. Whatever happened at reactor 2, we know that the situation didn't get real bad till the 14th when reactor water levels could no longer be maintained. A leak before then would not have resulted in notable contamination as the fuel had not yet melted. Looking at the timing of when reactor 2 RCIC failed, it is also possible that the explosion at reactor 3 caused a problem.

arevamirpal::laprimavera said...

Thank you elbow for sharing your thinking.

elbows said...

Thanks, and sorry that I am unable to explain things in a way that uses less words!

I'm sure there are sometimes some mistakes in my thinking, and I an not an expert, I just read a lot of government reports and background papers and over time things started to make more sense. But there is so little real data that many things will remain assumptions or speculation rather than facts for many years.

Anyway as I have mentioned several times how lucky Tokyo was not to get more contamination than it did, I should provide my source. It is the page that is labelled '2333' of this paper:

Anonymous said...

I was thinking along the lines of the wetwell cracking and losing water, thus allowing the PCV to vent directly to the atmosphere. Perhaps there is a simpler way. One of the ducts that go from the PCV to the torus may have cracked, inside the torus, but above the water level, providing a dry vent path.

elbows said...

Well there are lots of possibilities, lots of explanations that would result in significant release from reactor 2, the challenge is working out which ones actually happened, and that may yet take a while.

When looking at timeline of events at reactor 2 I realised that I seem to have been making a mistake when I say that suppression chamber pressure & temperatures were not taken till late on March 14th. Looks like they started taking them on the morning of the 14th. Regardless, they should have started taking them on March 12th.

Atomfritz said...

Thanks to all commenters for the input.

I am still somehow not convinced that the midnight venting at March 14/15 really failed.

1. The alleged rupture disk failure is, first, extremely improbable by itself at double the rupture disk working pressure, second there seems to be no actual confirmation/proof of the rupture disk not opening. It's just Tepco stating they believe so, without having proof by inspection or indicators.

2. The discussed alternative paths of reactor2's big release through parts of of the reactor building instead through the vent stack seem unplausible to me. Where are the highly radioactive condensates you'd expect then?
The few millisieverts in RB#2 to me just seem to contradict the assumption that a significant part of the core's radioactivity escaped through these rooms.
How could this be possible when the vent stack of unit 3/4 measures 10+ sieverts from condensates of #3's vent?
Just think of how much water runs down in a cold bathroom in winter if you take a hot steamy bath, even if you open the bathroom's small window.

If the radiation escaped via the electric ducts as the study says, wouldn't there have to be expected way more radioactivity than has been found near the electric installations in the RB?
On the other hand, the "exit duct" maybe was near the blowout panel and blowed the steam directly into the outside? (Has already a Quince taken measurements there?)

I am still clueless.

elbows said...

Do you mean the vent stack of 1/2? Thats where the 10+ Sv/hr is as far as I know. It could even have come from a possible brief reactor 2 vent for all we know.

However a brief dry vent at midnight does not explain the high levels of radiation on site & away from site during periods much later on the 15th than the supposed vent.

I know what you mean about radiation levels in the building, but the robot did pick up several hundred mSv/hr on the top floor either side of the top of reactor containment cavity. I don't know what levels are reasonable to expect, since the much higher value in the pipe can be influenced by the fact that its a pipe, a very narrow route which very nasty stuff could have flowed down at high pressure during important moment of reactor fuel melt. Also the robots radiation detection is probably not as good at getting readings directly next to the hotspot as a human with the 'detector on a pole' would be. The actual radiation levels right next to the failure points where the steam came from for months could be higher than the survey data we have seen so far.

And if we want to consider all possible areas of doubt, we don't know what role reactor 3 may have played in emissions on the 15th, if any, or even spent fuel pool at 4 for that matter (even though lack of obvious fuel damage in this pool has seriously decreased the emphasis on this possibility as time has gone on).

arevamirpal::laprimavera said...

On March 15, I think there was a white smoke rising from Reactor 3.

RSMC Beijing report says there was a second explosion of Reactor 3 on March 15, at 11AM.

Atomfritz said...

thank you for correcting me, I obviously remembered wrong. You are right, such a high radiation at the #1/2 vent stack's foot could indeed point to a dry vent event.

Really need more information about radiation condensation effects and how much residual radiation would remain.
You probably are right too with stating that Quince isn't that good in finding the actual hot spots like a pole-equipped human.

Don't even know for sure where Quince's radiation counter is located. Some documents state "in 1 m height", others state 1.20 meters.
So surface radiation could be much higher than one might think.

Then the actual radiation at floor level near the #2 concrete lid could be multi-sieverts, where Quince strangely didn't respond for 11 secs for two times. Which in turn could have been caused by a radiation-induced hangup and the following reset sequence, followed by quickly being pulled back from the danger zone?

And, the main release path even could be not the circular outer side of the concrete cap where Quince "hanged up", but at the line separating its two half-circle parts, where Quince possibly hasn't been at all.

Looking from this perspective, it even could be possible that there was first a short dry vent, following by drywell leaking and containment compromise due to extreme strain, caused by the expansion and deformation of the drywell cap being heated by the glowing RPV and losing its tight closing.

The chamber between the steel drywell cap and the radiation protection concrete lid then could have worked as a steam condenser and steam cooler, thus reducing the condensation contamination in the reactor hall itself. So it also would have dampened the violent release pressure somewhat.

At least this would explain the missing of apparent cracks in the concrete which would have been big enough to allow for such a large pressure release and be consistent with the relatively smooth, not-too-violent steam billowing out of #2's blowout panel.

But I am still speculating as I still don't understand why RB #2 is so "clean" compared with #1 and #3, in spite of its alleged much bigger release.

Reactor 3 steamed/dry vented for weeks. Containment lid definitely damaged. Containment also leaks into the building, for example at the hatches, even if the pressure didn't get as high as in #2. Sign of an in-containment pressure surge, possibly due some sort of explosion?
And this minor releases compared to #2??

SFP 4 partially ran dry.
Do the spent fuel assemblies really have to melt before releasing nasties? Doesn't it suffice to lose the Zirconium cladding barrier by transforming into a brittle and porous Zirconium oxide?
Radioactivity values at the plant's MPs dropped sharply immediately after water levels restored, with no official explanation given.
Really reason to exclude releases from SFP 4??

Above all, remember that all these release estimates are only more or less arbitrary estimates based on release model assumptions, not actual release measurements.

Anonymous said...

Atomfritz, it is possible that the fuel rods release gasses simply upon being heated up. No major deterioration of the cladding is needed, microscopic cracks and pores can be enough. Such cracks and pores are routinely found in even properly-cooled used fuel.

I am sure that watering SFP #4 contributed to the significant decreases in dose rate on-site, but I am unsure about the mechanism. Perhaps even just the reduction in gamma shine and stopping the water from bubbling would be enough to matter?

I am saying this because from the recent movies we can say for sure that the fuel in SFP #4 was not uncovered for any period of time - there is no warping, no discoloration, nothing.

Atomfritz said...

I am not so sure that we can deduct from the SFP videos that it never ran dry so that part of the fuel got exposed.
We saw only a handful of the more than 1300 used assemblies.

Could well have been a Potemkin village. What if we were shown only a section of older fuel assemblies?
Reportedly substantial part the fuel stored in the pools was older than 12 years and develop way less heat than "fresh used fuel".

A Zirconium-steam reaction could even have happened if only part of the fuel was exposed.

But, there could indeed have been other mechanisms that come into effect when water boils while the fuel is still fully covered. Concrete debris damaged fuel would then release contaminants to air via steam bubbles.

I don't think the gamma shine would have much an effect at the plant gate radiation detectors as it would have been directed into the air.

There were some radiation measurements from helicopters which indicated high gamma radiation above some reactor buildings, but I don't find the actual values they metered in which height.
But they were quite high, maybe from SFP gamma shine too and not only from the radiating steam plumes.

So many possibilities.

Post a Comment