Fukushima Japan Earthquake: nuclear plants Fukushima part 2

[Rive]

Could they have altered the result in any way or not?

By my understanding (and the posts from the last few weeks) to make a significant difference would have require so many violation of actual OpInst rules and so many knowledge about plant status and future events, that it's really a kind of 'what would have Bruce Willis done?' question.

everybody seems to agree that failure to recognize possibility of bigger-than-designed-for tidal waves and harden the plant against them is the culprit.

Agree. The underestimated height of possible tsunami and the presence of U1 with its far outdated design basis has already set up the situation, and any (realistic) course of decisions would have had only a very limited impact on the results.

 

[jim hardy]

Like Titanic- watertight doors weren't high enough.

 

[etudiant]

I recognize the plant was doomed by the tsunami, but was wondering whether there was a way to limit the resultant mess. Could doing absolutely nothing have been worse?

Astronuc has speculated that the sea water injections intended to cool the cores may in fact have damaged the fuel, which worsened the outcome.
There have been suggestions that most of the water injected flowed directly directly into the turbine building, with only very little actually reached the inside of the reactors, adding to the hydrogen build up rather than providing effective cooling.

 

[mheslep]

The Japanese containment vent design utilizes a rupture disc that's set at around twice containment maximum design pressure. So there was no way to vent early. Additionally for at least one of the units, the rupture disc did not rupture when it was supposed

No *automatic* rupture, yes, but could the disc have been manually blown?

 

[jim hardy]

Could doing absolutely nothing have been worse?

i can't opine, too interested ..

Does this say anything ?

http://www.nrc.gov/docs/ML1012/ML101270372.pdf

 

[mheslep]

By my understanding (and the posts from the last few weeks) to make a significant difference would have require so many violation of actual OpInst rules and so many knowledge about plant status and future events, that it's really a kind of 'what would have Bruce Willis done?' question. ...

I grant that discussion of alternative action is often ill informed (and is in my case). But I disagree that discussions of alternatives are necessarily indulgence in fantasy. The better comparison to a decisive personality, if there must be be one, is Captain Sullenberger and the landing in the Hudson. The Hudson River is not another numbered FAA runway. If Sullenberger had instead stubbornly headed back to the (out of glide range) runway with no engine power because of FAA regulations I'd have little time for those saying he was only following the rules, that nothing else could be done, or that pilot training and aviation operation are adequate, and that bird strikes and Tsunamis just happen.

 

[Hiddencamper]

I grant that discussion of alternative action is often ill informed (and is in my case). But I disagree that discussions of alternatives are necessarily indulgence in fantasy. The better comparison to a decisive personality, if there must be be one, is Captain Sullenberger and the landing in the Hudson. The Hudson River is not another numbered FAA runway. If Sullenberger had instead stubbornly headed back to the (out of glide range) runway with no engine power because of FAA regulations I'd have little time for those saying he was only following the rules, that nothing else could be done, or that pilot training and aviation operation are adequate, and that bird strikes and Tsunamis just happen.

Some things to think about:

When an accident happens, you are forced to respond to the conditions you are presented with.

To quote the BWR Emergency Operating Procedures:

"The EOPs impose various limits within which continued safe operation of the
plant is ensured and beyond which certain actions may be required. While conservative,
these limits have been derived using best-estimate engineering analyses rather than
licensing models. Consequently, these limits are generally not as conservative as the
limits specified in the Technical Specifications and conformance with these
guidelines does not necessarily ensure strict conformance with Technical Specifications
or other licensing bases. This does not imply, however, that operation beyond Technical
Specification limits is recommended. Rather, such operation may be required, and is now
permitted, to mitigate certain degraded conditions."

In other words, you should not just violate your license or regulations simply because you are in an emergency, but you may be forced to operate there. This is essentially what happened with the flight that landed in the Hudson.

Also remember that both the FAA and the US NRC allow violation of license/regulations as necessary to safeguard the plant and public. Japan had no such provision to the best of my knowledge.

http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/part050-0054.html

"(x) A licensee may take reasonable action that departs from a license condition or a technical specification (contained in a license issued under this part) in an emergency when this action is immediately needed to protect the public health and safety and no action consistent with license conditions and technical specifications that can provide adequate or equivalent protection is immediately apparent.

(y) Licensee action permitted by paragraph (x) of this section shall be approved, as a minimum, by a licensed senior operator, or, at a nuclear power reactor facility for which the certifications required under § 50.82(a)(1) have been submitted, by either a licensed senior operator or a certified fuel handler, prior to taking the action."

When transients happen at a nuclear plant, if anything didn't happen exactly per procedure or per plant design, the operators on shift will all go before a body consisting of regulatory personnel, senators/congressmen, company officials, to answer questions and be held accountable for actions they took. If you followed all the words on your EOPs as they were designed to be followed the likelihood of being prosecuted for an accident is slim to none, UNLESS you had some definitive knowledge that following the EOPs would result in a further accident.

In other words, trying to suggest the operators intentionally violate their license, government regulations, and the EOPs is insane and somewhat dangerous. The EOPs, tech specs, and other accident management documents were made with large amounts of engineering, to minimize the potential for fission product barrier loss, extend the coping time for events, and anytime you knowingly violate those you are making an on the spot decision with far less info about why those documents were made that way in the first place during a high stress situation. It carries a very high potential for a human performance failure. Something to think about

Also, to this day, I can't see any reasonable action that would have ensured safety unit 1.

Side note:

Do we need to have a talk about BWR EOPs??

 

[Hiddencamper]

No *automatic* rupture, yes, but could the disc have been manually blown?

There's no manual blow for these rupture discs. Japan's policy was to vent as late as physically possible and that is why these rupture disc point is set as high as it is. They used a Sandia national labs study that the containment should retain integrity at twice the design pressure as the basis for their decision.

 

[mheslep]

There's no manual blow for these rupture discs. Japan's policy was to vent as late as physically possible and that is why these rupture disc point is set as high as it is. They used a Sandia national labs study that the containment should retain integrity at twice the design pressure as the basis for their decision.

By manual blow I meant either demolition explosives, or hydraulics, or cutting torches, or whatever external device might be required to force open the disc. Yes the *policy* was undoubtedly double pressure. I'm exploring the applicability of the policy under these LOCA circumstances, and particularly would venting have lessened the possibility of the later H2 explosion.

 

[etudiant]

There's no manual blow for these rupture discs. Japan's policy was to vent as late as physically possible and that is why these rupture disc point is set as high as it is. They used a Sandia national labs study that the containment should retain integrity at twice the design pressure as the basis for their decision.

Very interesting detail and illuminating as to Japanese policy. Sort of an all or nothing approach.
Against that, Jim Hardy's reference ( http://www.nrc.gov/docs/ML1012/ML101270372.pdf ) concludes with this punch line:

'if the operators would be able to open both pressurizer relief valves after the core heatup
starts, this would have positive effect on further progression of the severe accident.'

To me, that suggests the Japanese policy may possibly have helped make matters worse than they could have been.Separately, I can only say 'Amen' to hiddencampers sharp reminder that manuals and procedures exist for good reason. They exist to guide operators in real life.

Fortunately we here at PF have the opportunity to speculate more freely. Also, in the case of Fukushima, the results were so poor that we must reexamine whether there was any possibility of some action that would have lesser consequences.

So the issue facing the operators, paraphrased, is how to set the reactors to melt down as gracefully as possible, knowing there is only a few hours of battery power to monitor the process.

 

[Hiddencamper]

Very interesting detail and illuminating as to Japanese policy. Sort of an all or nothing approach.
Against that, Jim Hardy's reference ( http://www.nrc.gov/docs/ML1012/ML101270372.pdf ) concludes with this punch line:

'if the operators would be able to open both pressurizer relief valves after the core heatup
starts, this would have positive effect on further progression of the severe accident.'

To me, that suggests the Japanese policy may possibly have helped make matters worse than they could have been.Separately, I can only say 'Amen' to hiddencampers sharp reminder that manuals and procedures exist for good reason. They exist to guide operators in real life.

Fortunately we here at PF have the opportunity to speculate more freely. Also, in the case of Fukushima, the results were so poor that we must reexamine whether there was any possibility of some action that would have lesser consequences.

So the issue facing the operators, paraphrased, is how to set the reactors to melt down as gracefully as possible, knowing there is only a few hours of battery power to monitor the process.

Jim's link is for pwr plants. It involves not depressurization during station blackout which can result in not having sufficient inventory for longer coping times, vs depressirizing and wasting inventory early but having the accumulators available.

Bwrs have a very different response. One major difference is that bwrs have an absolutely massive amount of steam relief capacity compared to PWRs, allowing a rapid emergency depressurization which also provides steam cooling to the core.

For a BWR, the safest place to be is with a depressurized flooded core. The challenge is even if you performed an emergency blowdown the moment the earthquake was over, on the loss of DC power the relief valves would have shut and the core would have repressurized. Additionally blowing down the core early removes IC capability, so there was no way to really say those actions could have been taken.

In order for things to be "graceful" the operators needed to A: recognize they had no valid level indication and make a transition to the core flooding EOP, B: quickly got batteries from cars to open up relief valves using car batteries, and C: lined up a portable or fire pump for injection. Even if the fire pump could not be lined up, just having the core repressurized will minimize the potential for containment failure.

The other issue is the diagnosis for entering the core flooding contingency is that you need to observe reference leg boiling. The operators should not simply enter because of a momentary loss of indication. However in this case the reference legs boiled by the time operators got indications back, so they never saw the transition.

This is ultimately one of the things that made daiichi and Daini different. With no DC power at daiichi the operators couldn't even make decisions in the EOPs, and could not take the proper or best steps to protect the core until it was too late.

Station blackout analysis for bwrs assumes you stay hot and pressurized, unlike a pwr.

 

[etudiant]

Thank you very much, hiddencamper, for this more detailed explanation. It certainly clarifies the gap between the BWR and PWR emergency procedures.
Two points that leave me still confused.
First, you note the operators need to go to core flooding EOP, but later say ...just having the core repressurized will minimize the potential for containment failure' .
Is that a typo and should be 'depressurized'?
Second is the discussion about the 'reference leg boiling'. Afaik, this is a topic that has not been explored in this thread. It seems a crucial indication, yet your comment suggests it must be observed, because there is no telltale to inform the operators of a crucial change in the reactor status. Is that correct?
Also, is there an overview discussion somewhere that would help inform the forum members such as myself to whom this aspect of reactor management is an unknown?

The final note, that 'BWR station blackout analysis assumes the reactor stays hot and pressurized' seems diametrically opposite to the core flooding EOP.
How do the operators decide which procedure is the proper one to follow?

 

[Hiddencamper]

Thank you very much, hiddencamper, for this more detailed explanation. It certainly clarifies the gap between the BWR and PWR emergency procedures.
Two points that leave me still confused.
First, you note the operators need to go to core flooding EOP, but later say ...just having the core repressurized will minimize the potential for containment failure' .
Is that a typo and should be 'depressurized'?
Second is the discussion about the 'reference leg boiling'. Afaik, this is a topic that has not been explored in this thread. It seems a crucial indication, yet your comment suggests it must be observed, because there is no telltale to inform the operators of a crucial change in the reactor status. Is that correct?
Also, is there an overview discussion somewhere that would help inform the forum members such as myself to whom this aspect of reactor management is an unknown?

The final note, that 'BWR station blackout analysis assumes the reactor stays hot and pressurized' seems diametrically opposite to the core flooding EOP.
How do the operators decide which procedure is the proper one to follow?

It was a typo. I did mean depressurized.

As for reference leg boiling, I read at one point that at unit 1, they got water level indication back at one point and it was high (top of the narrorw range indication), and believed it was real, as they also thought the IC was operating. This was incorrect, as reference leg boiling was in progress at this time causing an erroneously high level indication.

So, in the Reactor Pressure Vessel control EOP, the RC/L (Reactor Control/Level) branch, there is an override that says "If Level Indication is suspect, then exit this leg and enter the core flooding contingency". It also has a reference to the RPV saturation temperature table, which plots RPV pressure, drywell temperature, and saturation point on a plot. Generally, the EOP basis say that you should not say your level indications are suspect simply because you exceeded saturation temperature, however you do need to monitor for boiling. It's typically not appropriate to transition to core flooding simply because you have a momentary loss of level indication for other reasons, especially if you qualitatively knew what level was or what it was doing prior to losing the indication. So with no indications to go off of, and very few "data points", its hard to say the operators SHOULD have transferred to core flooding or not.

As for the core flooding contingency, the whole purpose of it, is that you lose the ability to determine if level is above top of fuel using your level indications, so instead you depressurize the core and flood it until you either achieve the minimum steam cooling pressure (for scram failure cases), or until water is overflowing out the steam lines (typically your suppression pool level will steady out). Then you can assure that you have adequate cooling because you have level at or above the steamlines, which is above the top of the fuel. You are also completely depressurized, so RPV temps will be <= 300 degF and lowering.

For BWR station blackout, that's a short term event (4-8 hours) with a specific progression, which assumes you have one of IC/RCIC/HPCI/HPCS in service. Reference leg boiling is not supposed to happen in the short duration of an SBO, as you are going to get some form of AC power back to at least 1 decay heat removal train within the coping time.

 

[etudiant]

Thank you, hiddencamper, very helpful and informative.
Is it not surprising that something as critical as the reference leg boiling should be so relatively poorly indicated?
Presumably this reflects the difficulty of determining the water level and state inside the reactor, (iirc the Three Mile Island accident also arose because the operators had no clear measure of the reactor water level). Is there any hope of a better sensor to resolve this uncertainty?

 

[Hiddencamper]

Thank you, hiddencamper, very helpful and informative.
Is it not surprising that something as critical as the reference leg boiling should be so relatively poorly indicated?
Presumably this reflects the difficulty of determining the water level and state inside the reactor, (iirc the Three Mile Island accident also arose because the operators had no clear measure of the reactor water level). Is there any hope of a better sensor to resolve this uncertainty?

It's kind of similar to TMI, but not the same. The issue PWRs had, was that there was no direct in vessel water level measurement, only pressurizer level, which can be inaccurate during a loss of subcooling margin with an open relief valve. For BWRs, all your level instruments are directly in vessel so you always know what core water level is (provided there's no reference leg boiling). In both designs, these indications are all vulnerable to reference leg boiling. You can confirm there's no boiling using a combination of qualitative factors and measurements of drywell/containment temperature and RPV pressure.

Another thing you can do, is install a gamma monitoring system which would detect if radiation levels rise due to a loss of water level, however that would be reliant on some form of electrical power as well. So it wouldn't have really helped in the Fukushima Daiichi case. Their in-drywell rad monitors would have been able to identify something wrong if they had power, without the need for a special monitor specifically for level.

The water level, RPV pressure, and drywell/containment temperature indications are all backed up by uninterruptable power supplies off of the station vital DC batteries. On a loss of all AC and DC power (something that wasn't analyzed), you lose all indications, so it doesn't really help. ALL DC power is never assumed to be lost under design basis events.

After 9/11 the US BWRs got portable "measurement devices" as part of the b5b program. You hook alligator clips up to the leads coming from your transmitters and it uses a built in battery to get a voltage measurement, which you can use a table to correlate to actual water level/ temperature/ etc. This would allow you to have indications even if you lost ALL AC and DC power. Japan didn't implement the b5b program or anything similar to it, so they didn't have this equipment available, which is why they had to go to great lengths to scavenge car batteries just to see what was going on.

 

[jim hardy]

Is there any hope of a better sensor to resolve this uncertainty?

i don't know that this principle is better but it's diverse...

there was research in 1980's for a level instrument based on a long tube with thermocouples, some heated and some unheated,
inside the vessel alongside the core. When level drops the heated ones get hotter than the unheated ones because steam doesn't conduct away the heat so well.
So by the differences in temperature one can infer level.

I don't recall whether a practical version ever made it to production.

 

[Hiddencamper]

For reference, this is the RPV control EOP for a generic BWR. This is the "less pretty" version of EOPs. Many plants use KLR services to generate their EOPs and they are much better structured and easier to look at. I'm honestly surprised at how hard it is to find legible copies of BWR EOPs on google.

Anyways, you can see the override down the RCL/1 RPV Level branch that says, IF RPV Level CANNOT be determined Enter EOP 206 (in this case, the RPV flooding EOP).

I know a lot of people are asking what could have been done. This is an example of what BWR operators have to work with. Sadly I cannot find the containment EOPs or some of the contingencies. 2 contingencies are built into this EOP (steam cooling and alternate level control)

http://www.nrc.gov/docs/ML1100/ML110060122.pdf

Edit: Looks like its hope creek. Here are some other EOPs:
RPV Flooding Contingency http://www.nrc.gov/docs/ML1100/ML110060135.pdf
Emergency Blowdown Contingency http://www.nrc.gov/docs/ML1100/ML110060125.pdf
RPV Flooding during ATWS (scram failure) contingency http://www.nrc.gov/docs/ML1100/ML110060126.pdf
Primary Containment EOP http://www.nrc.gov/docs/ML1100/ML110060123.pdf

Still can't find secondary containment or ATWS EOPs

 

[etudiant]

Thank you very much, hiddencamper, this post has made understanding the problem facing the Fukushima operators much easier.

A first reaction is: 'Oh my, these look like system engineering flowcharts, where a lot of the action implications can only be understood by people very well versed in the system.'
Navigating these procedure charts with partial or non functioning instrumentation seems unlikely to produce a good outcome.

It is not obvious either that increased computer systems support would be beneficial. Airbus pioneered the civil use of fly by wire, where all flight controls are operated by the computer based on pilot input, modulated by system overrides when these inputs are beyond the airplane operational limits. In practice, this sometimes creates conflict of the 'what's it doing now' variety, bad in an airplane, worse in a nuclear plant.
In both cases, a reset switch would be nice to have, but seems beyond our current capabilities.

 

[gmax137]

... I don't recall whether a practical version ever made it to production.

Yes, Jim! The Combustion Engineering supplied HJTC (heated junction thermocouple) system is used to monitor liquid level in the reactor vessel upper head. It does not, however, extend down into the vessel downcomer.

 

[jim hardy]

It does not, however, extend down into the vessel downcomer.

That'd be quite a mechanical feat , to put something long and slender(translate flexible) in the downcomer.
I can only imagine how turbulent is the flow there.

The whole core barrel flexes ever so slightly, and that can be detected in excore neutron detector signals by DSP ( which never ceased to amaze me) .

Thanks gmax !

old jim

 

[jim hardy]

Has anyone ever seen a better photograph of this little square ?http://cryptome.org/eyeball/daiichi-npp3/daiichi-photos3.htm 9th one down

at 10X

just curious if i missed something.

old jim

 

[Tom.G]

That's about all I could get out of it. I don't know how much detail will be left when this gets posted, so here is a brief description:

The bright spot seems the center of a larger structure about six times its diameter. To the upper left of the bright spot, there are two or three strong radial elements radiating thru the dark area, and the outer part of the structure has much faint radial detail showing as fainter dark lines

.

 

[Rive]

just curious if i missed something.

old jim

Brings back memories about the big pixel huntings in the first days...

Cryptome with those compressed and resized pictures never was a good site to start with. I'll look up something , but at first sight that spot looks like a simple beam crossing from the upper structure.

Ps.: As it seems I can only link images... Here it is: http://keptarhely.eu/view.php?file=20160817v00k1x3i5.png
http://keptarhely.eu/view.php?file=20160817v00k1x3i5.png

 

[jim hardy]

Brings back memories about the big pixel huntings in the first days...

i was glued to the computer for months.

Your link has somewhat better resolution.

remember how a long lens compresses distance..

as i said just idle curiosity now, one of those loose ends for me. Headbolt tensioner is above and slightly right as best i recall.

old jim

 

[Sotan]

A new topic appeared these days in the reports posted daily by Tepco: the drain sump pit located at the bases of Units 1/2 smoke stack.

http://www.tepco.co.jp/nu/fukushima-np/handouts/2016/images2/handouts_160825_02-j.pdf
(in Japanese)
In this report of Aug 26 they explain that the area around the basis of the smoke stack is still very radioactive - as confirmed by measurements taken from a distance at the end of 2015. The drain sump pit at the basis of the smoke stack needs some investigation to learn more about the water level and radioactivity. They have been preparing for this investigation using mockups, and now they are planning to go for it. (It is not a lot of water, if I understand correctly the sump pit is only about a cubic metre in volume - but probably it is very radioactive and poses a risk if it starts leaking.)
Basically, with a cutting tool manipulated with a crane arm from behind shielding panels, they will cut an opening in the concrete wall/roof of the drain sump pit, check the water inside and start operations for storing that water in a safer place.

http://www.tepco.co.jp/nu/fukushima-np/handouts/2016/images2/handouts_160826_03-j.pdf
(in Japanese)
Next-day report on the start of the operations.
The two photos are available in unusually good resolution here:
http://photo.tepco.co.jp/date/2016/201608-j/160826-01j.html

 

[mheslep]

still very radioactive

I wonder what qualifies as very radioactive in the context of units 1 or 2 and Tesco analysis.

 

[Bandit127]

I don't know what quantifies as "very radioactive" but it surveyed at 10 Sv/h after the accident so it probably still qualifies.

 

[Rive]

Also, those mobile rad-shields on the photos are surely not for just show...

 

[Sotan]

Hm... I checked again and to be honest I must say I inserted the word "very" without it being needed. So easy to get carried away. My apologies, the text doesn't really say "very". A more careful translation would be "the air radiation dose in the area is still high and hinders investigations". As for the radioactivity of the water - if there is any water in the pit - they don't really know anything, but they seem to be making careful preparations for its transfer if they find it there.
Page numbered 3 in the first link gives a few readings of the atmospheric radiation dose in that area at the basis of the smoke stack, most recently in October 2015; highest is 12 mSv/h, measured at 1.2m height.
As Bandit127 mentioned, in other spots, especially in the immediate vicinity of some pipes designated as "SGTS", they measured more than 10 Sv/h in August 2011 and still 2 Sv/h in October 2015. (http://www.tepco.co.jp/nu/fukushima-np/handouts/2015/images/handouts_151026_04-j.pdf - page 16)
This time, the workers operating the remote controlled machines will be using a room located at the basis of the stack, in which the radioactivity is much lower (0.04 mSv/h).

 

[Sotan]

Fukushima nuclear plant prepares for typhoon (NHK news)
http://www3.nhk.or.jp/nhkworld/en/news/20160829_22/