Fukushima: Unit 2 Discharge - Why Differs from Units 1 & 3?

[zapperzero]

From the 2nd report to the IAEA, http://www.meti.go.jp/english/earthquake/nuclear/iaea/pdf/20110911/chapter0-summary.pdf"

(Quoted text below is OCR extracted from an image of the page, there may be minor differences to the original)

" a policy of drywell venting was adopted
because the pressure on the S/C side was lower than the working pressure of the rupture
disk and the pressure on the drywell side was increasing, and an operation to open the
small valve of the drywell vent valve (AO valve) was performed once at 0:02 on the
15th; however, it was confirmed several minutes later that the small valve was closed.
After that, drywell pressure maintained a high level of values; large sounds of impact
occurred between around 6:00 and 6:10 of the lS, while S/C pressure indicated O MPa
abs. Lower drywell pressure was also confirmed at around 11:25 on that day."

My question is - does the abovementioned Drywell direct venting constitute a possible explanation for Unit #2's alleged large contribution to site releases?

This also seems to shed more light on why TEPCO seemed fairly sure something broke in the wetwell area? They were not able to vent from S\C?

The report raises interesting questions. Why would S/C pressure be lower than drywell? 0 MPa abs means the sensor was off or broken somehow. It measures using a water column? Maybe the water drained?

also

large sounds of impact occurred between around 6:00 and 6:10

sounds. plural.

To your question: yes, the drywell would have been "dirtier".

EDIT: I should probably revise my earlier proposed accident sequence.

 

[clancy688]

Via EX-SKF:

http://ex-skf.blogspot.com/2012/03/dry-vent-of-reactor-2-may-have-released.html

It shows the spatial dose of radiation at the front gate of Fukushima I Nuke Plant and the timeline of events at the plant in March. The largest spike seems to be around or after 12AM on March 15, and the label on the spike says "Dry vent at reactor no.2".

What's the difference between a "dry" and a "wet" went? Morever, according to the INPO report, the Unit 2 containment was never vented because the venting line-up wasn't completed (a rupture disk failed to break).

 

[zapperzero]

What's the difference between a "dry" and a "wet" went?

Wet venting is venting steam from the RPV through the water in the suppression chamber (aka wetwell) and then out of the suppression chamber into the stack. This scrubs most of the particulates and some of the gasses out of the vented steam, so the radioactive release is diminished.

Dry venting is when the steam goes directly into the outside air by some route or another. Obviously it is very far from being clean.

 

[SteveElbows]

It is good to see that this topic receives attention again, as I am still extremely dissatisfied that reactor 2's possible majority contribution to the total environmental contamination does not get much focus in the wider world. I've been fascinated by this subject for a long time now, but with little new info to help my understanding.

The ICANPS report was certainly much more readable than previous reports, made it easier to be confident about thing we already thought we knew, and added some detail that I don't think was discussed elsewhere.

It doesn't explain the venting contradictions, since it says in detail what we already heard from previous reports, that they had a range of nightmares when trying to vent, and failed. However I read with interest some of the comments on the ex-skf article, where someone has a theory that venting from the drywell may have occurred very briefly around midnight, not for long enough to reduce pressure in a noticeable way, but enough to allow substances to escape. To explore this theory further we would need to know why reports seem so certain that the rupture disk didn't fail.

The report isn't very helpful when it comes to why the measured pressure in the D/W and S/C diverged so much, but the report does emphasise the possibility that the S/C pressure readings developed a fault, especially as a reading of 0 as shown shortly after 6am on the 15th indicated less than atmospheric pressure, which makes no sense.

I think that the main thing from that report that was new to me was the explanation as to why the S/C was in a state that was inappropriate for receiving a load of stream from the reactor. They switched the RCIC to use the S/C water as a source early in the morning of March 12th, but they didn't pay any attention to S/C temperature & pressure until March 14th, nor did they attempt to complete an alternative to the RCIC until the RCIC was on its very last legs. Combine this with the multiple problems when trying to vent, fire truck running out of fuel without anyone noticing for a while, and some reactor pressure fluctuations which would have prevented water injection from working at various points, and we start to get a better idea of just why reactor 2s failure had such large implications for the environment. It will be interesting to perhaps learn one day as to how the fuel damage at this reactor compares to the others, but have a very long wait for that, if ever! Not sure how much we will really find out about the state of the drywall or S/C either.

 

[SteveElbows]

Dry venting is when the steam goes directly into the outside air by some route or another. Obviously it is very far from being clean.

I expect that the term dry venting is supposed to mean deliberate venting by one very specific route, as opposed to unplanned escapes from drywell containment due to some kind of containment failure. Most importantly this route still involves the stack, but as you mentioned does not include scrubbing via the S/C so it has larger environmental implications.

Sadly unlike the initial vents at reactors 1 & 3, we have no visual confirmation of anything leaving the stack during any of the reactor 2 venting attempts, mostly because the attempt at dry venting happened at night, and the old 'one published image an hour' camera was no use at night.

What I find especially annoying is that we don't even know exactly when steam was first noticed to be coming from the blowout panel hole in the side of reactor 2. We certainly know that plenty of steam etc came out of this hole later, as we have footage, but not knowing when it began is frustrating. Assuming it was from the reactor rather than the fuel pool, which is a reasonable bet given much later footage showing steam from the reactor arena and the robot-measured radiation readings showing high levels by one side of the floor above the reactor well, we may well imagine that this started on March 15th. But it would be good to know for sure, especially as events at reactor 4 building on the 15th also got in the way of really clear explanations of radiation data on that day.

 

[zapperzero]

I expect that the term dry venting is supposed to mean deliberate venting by one very specific route, as opposed to unplanned escapes from drywell containment due to some kind of containment failure.

I feel a bit guilty for being unclear - I meant to say that there is a number of routes by which one could achieve this kind of venting

One could route the steam through the SGTS or not; if the suppression pool water level is lower than that of the steam pipes for some reason, dry venting is also achieved, by default; there is also a choice of RPV valves that could be opened.

 

[Yamanote]

Probably the venting of unit 2 happened in the way Arnie Gundersen explained in one of his videos:

http://vimeo.com/36492960

 

[clancy688]

Nope, most definitely not.

That's only a "containment vent", not a reactor vent. Moreover, it doesn't vent high pressure. It only ensures, that, at high pressure, the pressure doesn't get any higher. And that only goes for the containment.
So when pressure inside the containment reaches a certain value (I think it was twice the normal pressure), the containment cap lifts and pressure escapes. The pressure falls slightly and then the cap closes again. So the pressure can't go over twice the normal pressure, but it also can't go below twice the normal pressure.
But that still doesn't vent the reactor pressure vessel. And the unfiltered release ("dry venting") can't happen that way. Because for reactor pressure to escape via a lifting containment cap, the gases have to travel through the wetwell (torus). Which they obviously didn't, that's why we have such a dirty release for Unit 2.
So there must've been an other path.

 

[zapperzero]

Because for reactor pressure to escape via a lifting containment cap, the gases have to travel through the wetwell (torus). Which they obviously didn't, that's why we have such a dirty release for Unit 2.
So there must've been an other path.

We don't know that. The torus may have been dry.

 

[SteveElbows]

Now you are confusing me a bit clancy688, either I misunderstood or else I don't think I agree with you about several things.

I thought that the terms wet and dry venting applies only to how the gases are released from drywell containment to the outside world, not how those gasses got into containment in the first place. If the gases were vented from the reactor vessel to the s/c, but then ended up in the drywell, and the drywell was then vented directly, I think that would still count as a dry vent.

A quote from the interim investigation report gives us some info about the venting of the reactor vessel into containment, with emphasis on how poorly the s/c seemed to handle this:

According to the reactor pressure gage, Unit 2 reactor pressure indicated 6.998 MPa gage at around 16:34 that day. It indicated still 6.075 MPa gage at around 18:03, more than one hour after they had started depressurizing.
They continued trying to open the SRV to depressurize the reactor. However, they had trouble in keeping the SRV open and the steam from the RPV barely condensed in the S/C because of high temperature and pressure in the S/C. Consequently, it took time to depressurize the RPV to the sufficient extent.
The reactor pressure was finally lowered to a level where water injection was possible at around 19:03 that day, when the reactor pressure gage indicated 0.630 MPa gage.

and

According to the reactor pressure gage, Unit 2 reactor pressure indicated higher than 1 MPa gage from around 20:54 until 21:18 that day (it indicated 1.463 MPa gage at around 21:18) and then it decreased due to depressurization. It again exceeded 1 MPa gage from around 22:50 until 23:40 that day (it indicated 3.150 MPa gage from around 23:20 until 23:25 that day) and then decreased again as a result of further depressurization. From around 00:16 until 01:11 on March 15, it again rose to over 1 MPa gage (it indicated 2.520 MPa at around 01:02 that day). At least during those periods of high values, Unit 2 reactor pressure seemed higher than the discharge pressure of the fire pumps and therefore it was highly likely that water had not been injected into the reactor.

and

From around 01:00 on March 15, Unit 2 reactor pressure indicated steadily staying above 0.600 but below 0.7 MPa gage and continuous water injection into the reactor became possible.

(from pages labelled as numbers 257,258 and 259 of http://icanps.go.jp/eng/120224Honbun04Eng.pdf )

Obviously what we can't tell from this is whether the bulk of the depressurisation of the reactor was achieved through the deliberate operation to depressurise the reactor vessel, as opposed to the fuel falling out of the bottom of the reactor vessel. Its been ages since I looked at the pressure charts for this period, I will have to refresh my knowledge on this front and comment again if anything seems relevant.

In any case I find the approach of looking at containment failure in terms of a very simple model of a cap rising and falling at certain pressures to be just a bit too flawed. We know that there are a range of potential failure points in containment, such as gaskets & flanges around not just the cap but also pipework points and equipment & personnel airlocks. And we know that heat contributes to degradation of such seals, so we aren't just dealing with pressure. Now it may turn out that the cap rising was a main culprit at one or all of the affected reactors, but its going to be a long journey to discover this. Certainly we saw some visual evidence of stuff escaping from locations around the top of containment at all three of the reactors. Reactor 3 offered the clearest view of this, could see it emerging from at least the area where the drywall concrete pit gates meet the floor area above the reactor. But we saw some signs at reactors 1 and 2 as well, along with the robot readings around that area of reactor 2 that I mentioned earlier. But this isn't proof that the cap lifted, cap or another part of containment could have degraded in a different manner.

 

[SteveElbows]

Does anybody know where a Japanese version of the following document might be found?

It is described on Tepco's English website as '- Fukushima Nuclear Accident Investigation Report (Interim Report Supplementary Volume) (PDF 212KB) '

http://www.tepco.co.jp/en/press/corp-com/release/betu11_e/images/111202e16.pdf

The reason I ask is that in the section dealing with reactor 2, on page 39 it says the following, which as you can see is suffering from poor wording just at the moment when it could reveal something about the status of the rupture disk:

At 00:02 on March 15, the AO valve (bypass valve) on the vent line from the D/W was opened, and it was thought that the vent line, with the exception of the rupture disk, was completed, however several minutes later it was discovered that the AO valve (bypass valve) in the vent line from the D/W was closed. As a result it was not possible to determine whether venting was successful (ruptured status of the rupture disk ruptured).

 

[Sorai]

Does anybody know where a Japanese version of the following document might be found?

It is described on Tepco's English website as '- Fukushima Nuclear Accident Investigation Report (Interim Report Supplementary Volume) (PDF 212KB) '

http://www.tepco.co.jp/en/press/corp-com/release/betu11_e/images/111202e16.pdf

This one: http://www.tepco.co.jp/cc/press/betu11_j/images/111202g.pdf

Seems to be the same info. They couldn't confirm whether or not the rupture disk was open:

3月15日0時02分頃、ドライウェルからのベントラインにあるAO弁(小弁)
の開操作を実施し、ラプチャーディスクを除くベントライン構成が完了した
と思われたが、数分後にはドライウェルからのベントラインにあるAO弁
(小弁)が閉状態であることを確認した。結果として、ベントの成否
(ラプチャーディスク開放の有無)は確認出来ていない。

 

[etudiant]

There has been speculation on EX-SKF that the rupture disk did rupture, but that debris from within the RPV clogged the piping for the vent. Given the violence of the events, this does seem plausible.

 

[SteveElbows]

This one: http://www.tepco.co.jp/cc/press/betu11_j/images/111202g.pdf

Seems to be the same info. They couldn't confirm whether or not the rupture disk was open:

Thanks very much, this is perfect, just what I needed. The Japanese version is much clearer than the English version. If I google translate the relevant sentence I get this:

As a result, the success or failure of the vent
(The presence or absence of open rupture disk) has not been confirmed.

Whereas the English version fails to make that point clearly at all, due to using rupture words three times.

So now we have heard them say that the rupture disk status is unknown, the theory on ex-skf that there may have been a vent for a couple of minutes, is not incompatible with official statements.

 

[SteveElbows]

I've just been saying elsewhere that it might be useful to look again at some graphs in this document:

http://www.kantei.go.jp/foreign/kan/topics/201106/pdf/attach_04_2.pdf

Specifically the CsI distribution graph for reactor 2 Tepco Case 2 on page 36, and Cs graph on page 37.

Note the presence of reactor building, FHB (fuel handling building I guess) and environment in these graphs, as well as the timings. Then compare to the other reactors.

 

[Yamanote]

It seems to me that successful venting through the designated path is almost impossible under accident conditions and without electricity, at least for the Mark I containment (don't know if it would work out better for other containment designs).

Is the "blow out panel" opened at unit 2 actually an improvisation or a feature?
Because units 1 and 3 explosions showed us that trying to contain gases and Hydrogen (which must be released anyway) within the secondary containment might not be the right approach in an emergency.

 

[zapperzero]

Is the "blow out panel" opened at unit 2 actually an improvisation or a feature?

Design feature. But it was popped from the outside, by hand, as it were, to avoid the accumulation of hydrogen. Seems to have worked, to a point.

Because units 1 and 3 explosions showed us that trying to contain gases and Hydrogen (which must be released anyway) within the secondary containment might not be the right approach in an emergency.

Nothing is supposed to be released into the reactor building. The very worst case scenario designed for is to vent some steam from the RPV through the S/C, then through the hardened vent (bypassing the undersized SGTS) and out the stack.

 

[rowmag]

Design feature. But it was popped from the outside, by hand, as it were, to avoid the accumulation of hydrogen. Seems to have worked, to a point.

Are you sure? My recollection is that they announced that they intended to open it, but then discovered that it had already popped off by itself. (As a result of the Unit 1 explosion?)

 

[zapperzero]

Are you sure? My recollection is that they announced that they intended to open it, but then discovered that it had already popped off by itself. (As a result of the Unit 1 explosion?)

No, I am not sure. This is my recollection however. I would be grateful if you can dig up a source, of course.

 

[Yamanote]

Nothing is supposed to be released into the reactor building. The very worst case scenario designed for is to vent some steam from the RPV through the S/C, then through the hardened vent (bypassing the undersized SGTS) and out the stack.

I agree that it was not supposed but happened for all three units. And like Steve Lochbaum said, even unit 4 paid sympathy to the others and exploded as well.

So in the future we have to suppose that unintended venting into reactor building will happen in such kind of accidents. Opening the blow out panel in advance might be a good idea and venting into another unit through shared equipment must be avoided.

 

[zapperzero]

shared equipment must be avoided.

You got that right, I think.

 

[rowmag]

No, I am not sure. This is my recollection however. I would be grateful if you can dig up a source, of course.

See the third slide (page-numbered "2") of this NISA presentation:

http://www.nisa.meti.go.jp/shingikai/800/28/006/6-3.pdf

"The reason there was no hydrogen explosion at Unit 2 is conjectured to be that by luck, the blow-out panel was opened as a result of the explosion at Unit 1, releasing accumulated hydrogen to the outside and thereby avoiding an explosion."

(The date on the title page of the above is 20 Jan. 2011, presumably a typo for 2012.)

 

[zapperzero]

See the third slide (page-numbered "2") of this NISA presentation:

http://www.nisa.meti.go.jp/shingikai/800/28/006/6-3.pdf

Thank you.

 

[SteveElbows]

Now that we can consider the possibility that reactor 2 could have dry vented for a few minutes, I would like to suggest that reactor 2 may be the source for the >10 Sv/hr contamination at the pipes at the bottom of the reactor 1 & 2 stack.

So I took another look at the gamma picture taken in August, on page 2 of this document:

http://www.tepco.co.jp/en/nu/fukushima-np/images/handouts_110802_01-e.pdf

I am interested in the gamma blobs on the right of the image. By looking at some other photos of the area, including the early high-ish res overhead ones from early on, I believe that this area of the photo is heading east. And we can just make out the large ducting pipe heading in that direction. But either side of that duct pipe are the smaller pipes, running parallel to the larger pipe, and their height above ground is near the bottom of the larger ducting. I believe the smaller pipe on this side is one from reactor 2, and the gamma blob fits with a point in this pipe.

You can see the pipe I mean in the last 3 photos on this page, it stands out fairly well because although it is small it looks pretty white in the photos. Download the zip of the images to see it even more clearly.

http://cryptome.org/eyeball/daiichi-npp/daiichi-photos.htm

 

[jim hardy]

Here's a somewhat scholarly report from last summer regarding release rates.

They suggest a ~30fold increase in I131 release rate "was probably caused by the damage of the suppression chamber of Unit 2 on march 15."

https://docs.google.com/file/d/0B-lM5qg9ztErMzY1NDAxNmUtNmFkNi00OWEwLWI0MzMtMGI5ZjFjZTk0NTBi/edit?hl=en_US&pli=1

i hope link works okay.
Please excuse if it's old hat, just i hadn't seen it before despite having tried to keep up.

 

[SteveElbows]

Thanks Jim. The documents detail is new to me but its conclusions are old news, having been used to come up with the official release estimates and that graph of release rates over time which we have sometimes referred to when talking about reactor 2. Incidentally that graph was updated slightly at some later point, refining in particular events around the 15th, so somewhere there is likely to be an update of that paper I would think, or a different but related study.

It was good to learn a bit more about the data they used. It matches pretty well with what we had gleaned from less wordy sources, that a lack of certain kinds of sampling at various points in time made some of the estimates rather uncertain. For example they mention that certain data was only available when the march 15th plume met wet weather in the north west , giving them a much better look at a narrow range of time than the 15th as a whole.

And for example their focus on the suppression chamber is very understandable given the explosion report, but since the report was written they have become less certain as to what exactly happened in the region of the suppression chamber, backing away from the explosion theory somewhat whilst not denying the possibility that the suppression chamber was damaged.

These are also some of the reasons why we have only recently had this discussion about the possibility that reactor 2 did actually manage to dry vent for a few minutes around midnight. We can find data that shows radiation levels rising at locations south of the plant in the hours after midnight, but they aren't of the type most used in the report, which gets its best glimpse of reality much later on the 15th when things have gone to the north west and fallen to the ground with the wet weather.

Given the stuff that fell to Earth later on the 15th, and the radiation levels on site during the daytime on the 15th, I don't think anybody is trying to suggest that the possible midnight dry vent was more significant than what came out of the reactor for hours during the day. But the vent is of interest because it has largely been discounted in the past, many of us thought that it was reasonably certain there had been no vent due to the wording used in various reports, lack of pressure drop, etc. But if it did happen then its of interest both in terms of demonstrating the differences in what came out of reactor 2 compared to the other reactors, and in possibly explaining the story of the contamination that went towards Tokyo as opposed to the stuff which affected the ground in the north west so much which is better understood. Mind you even though its better understood I am sure reactor 2 still manages not to get star billing when discussing the north west contamination in most reports, which often focus more on the weather and previous emissions from reactor 1 than the bad story of reactor 2 on the 15th.

When looking at what happened during the daytime of the 15th I no longer make the assumption that the stuff came mostly from the suppression chamber, since we don't actually know how much came out of the drywell as it depressurised that morning. Much later we saw signs of stuff emerging from refuelling floor level in the area of the drywell top, but that may have been responsible for relatively little emissions compared to, for example, the drywall depressurising via the suppression chamber failure point. I probably need to look at those graphs again that showed estimates for where substances may have gone, the one that mentions the FHB, but I don't expect to learn much and that stuff was based on modelling anyway. If we don't learn much more about reactor 2 from the next endoscope mission, then I am prepared for a very long wait before getting more detail that could shed light on nature of suppression chamber failure and its role in the emissions of the 15th.

Another reason why I try not to assume too much that the suppression chamber damage was the main release pathway, is that it was about the only form of containment damage that they have wanted to mention much, especially at the time, so it provided a very simple bit of narrative at that moment, 'here is the event that has taken things to another level'. Even later we only got vague comments that the possibility of containment damage at the other reactors could not be denied. We've seen them taking air samples from above likely failure points of refuelling floor level containment top areas, and sometimes steam emerging from such areas on videos, primarily at reactor 3 but occasional glimpses from others.

So for me right now, if I am looking for a shorthand way to describe why reactor 2s emissions were so bad, I'll look more broadly than the suppression chamber. The possible dry vent didn't change the pressure in a notable way, and so I can still speculate that reactor 2 was different from the others because one or more parts of containment failed and released substances whilst the containment was under rather high pressure, unrelieved by scrubbed venting. Whereas at the other 2 reactors, for all we know containment failures may not have had such an opportunity to release substances, due to venting or the damage to containment happening due to heat or explosions at a time when the pressure wasn't at such a peak. However as I write this I have forgotten some key data trends from other reactors so that last statement may be easy to disprove.

 

[jim hardy]

backing away from the explosion theory somewhat whilst not denying the possibility that the suppression chamber was damaged.

You clearly have a better grasp of how all the pieces of information fit together than i have.

I'm still at "The water hammer from safety valve blowdown into SC ripped out a pipe connection" stage of thinking.

Time will tell.

 

[SteveElbows]

You clearly have a better grasp of how all the pieces of information fit together than i have.

I'm still at "The water hammer from safety valve blowdown into SC ripped out a pipe connection" stage of thinking.

Time will tell.

I don't suggest that my thinking leads to any better answers than your stuff, or that it makes any of that irrelevant. I have a reasonable understanding of much of the released reports, and the ways that some of the factors may interplay with each other, But there is precious little data with which we can form solid conclusions.

There are a number of different ways that the suppression chamber may have failed. The committee report that was written in clearer and more critical language than other reports was very informative. It said that one of the main mistakes made at reactor 2 in the first days was that they switched the cooling system to use the suppression chamber water, and then failed to put a priority on getting pressure & temperature data from the suppression chamber for several days. They go into detail about how concerned the site manager was that the suppression chamber would not be able to handle SRV release from the reactor, a release which they really needed to do in order to get reactor pressure down to levels where water could be pumped from firetrucks. He thought that they needed to wet-vent first to give the suppression chamber a chance, but they couldn't wet vent and so resorted to dry vent attempts in a desperate bid to save containment. None of this happened in time so they had to depressurise the reactor anyway, and depressurisation was slow which indicated the suppression chamber was losing its ability to lend a useful hand. Later either its relationship with the drywall changed and then it failed, or at the very minimum its pressure gauge failed.

 

[jim hardy]

I don't suggest that my thinking leads to any better answers than your stuff,

It's not a competition so no sweat there.

There was understandable reluctance to vent their families live downwind.

depressurisation was slow which indicated the suppression chamber was losing its ability to lend a useful hand.

I wonder still, did it get too hot to quench more steam or was it being handed noncondensibles?

As i said in another forum, I'm waiing for the Nova show on it.

 

[SteveElbows]

I wonder still, did it get too hot to quench more steam or was it being handed noncondensibles?

Not sure, this evening was the first time I actually went looking for proper historical studies of such things. I hadn't realized that some problems of this nature were known as the 'Würgassen effect' because of what happened at that power plant decades ago.

I found reference to this stuff in this paper:

http://www.osti.gov/bridge/servlets/purl/5338909-FPWlHy/5338909.pdf

And a fair bit more more detail in understandable form in this document, from page 30 to 33:

http://wikdahl.se/Filer/Korr_3_07_Marviken_eng.pdf

I know this topic has been covered on the forum in the past but I don't know to what extent or whether that 2nd document I link to was referenced at the time.