Fukushima Japan Earthquake: nuclear plants Fukushima part 2

[LabratSR]

Some tidbits from the NRC venting doc. Enclosure 4

"[EPRI’s] findings demonstrate that substantial decontamination factors for radioactive releases can be achieved by a comprehensive strategy that includes installed equipment, operator actions and capabilities that are largely consistent with the diverse and flexible coping strategy (FLEX)."

"The EPRI report concluded that “no single strategy is optimal in retaining radioactive fission products in the containment system.” The NRC staff agrees with this conclusion. Uncertainties surrounding severe accidents resulting from accident progression, status of plant systems and components, and operator response make it highly unlikely that accidents can be modeled and procedures developed to account for all potential scenarios."

"Core debris cooling is an important element of a robust strategy for mitigating releases. If debris cooling is not provided through water injection or spray into the drywell, containment failure or bypass is likely. Without core debris cooling, the containment can be challenged in several ways. Molten debris can come into direct contact with the containment wall, melting the liner and providing a release path to the environment. Elevated drywell temperatures in the containment atmosphere can cause seals and other containment penetrations to fail, leading to containment bypass. Finally, core–concrete interactions can generate large quantities of noncondensable gases that increase containment pressure and also can accelerate concrete erosion that could challenge containment integrity over time."

"The analysis also confirmed that Mark I drywell wall breach would largely negate any additional benefit of a hardened vent and external filter, if installed, in reducing releases or in preserving secondary containment (reactor building) accessibility and subsequent usefulness of equipment installed there for stabilizing plant conditions and avoiding or minimizing additional releases."


None the less, the recommendation appears to be for Severe Accident Filtered vents.

 

[etudiant]

From Nikkom:
This is the filter which TEPCO intalls right now at Kashiwazaki-Kariwa's vent lines.
Claimed to be capable of capturing 99,9% of contaminants.
Doesn't look too complicated or huge, right? Seems very small indeed, especially compared to the huge venting gravel beds used by reactors in Sweden and presumably Finland.
Presumably this unit is not meant to filter the massive steam plume created by a vented reactor that has just been shut down. Does anyone have any background for this filter and how/when it is expected to be used?
Also, the Kashiwazaki site has six or seven reactors, is there just one vent line or are there several lines and filters?

 

[zapperzero]

Also, the Kashiwazaki site has six or seven reactors, is there just one vent line or are there several lines and filters?

Isn't this (part of) the so-called SGTS?

 

[LabratSR]

I found this on the TEPCO site. The second page shows the vent system.

http://www.tepco.co.jp/en/nu/kk-np/safety/images/130804_01.pdf

 

[rmattila]

Seems very small indeed, especially compared to the huge venting gravel beds used by reactors in Sweden and presumably Finland.

The large gravel bed dry filter is only installed in the poorly situated and now shutdown Barsebäck NPP. Other Nordic BWRs use a compact wet scrubber filtered venting with jet nozzles injecting in NaOH water. The wet scrubber takes 99.9 % of Cs and around 99 % of elemental iodine, but unlike the large dry filter it's not very good at filtering organic iodine (perhaps 70 %) or noble gases (all pass through).

 

[etudiant]

The large gravel bed dry filter is only installed in the poorly situated and now shutdown Barsebäck NPP. Other Nordic BWRs use a compact wet scrubber filtered venting with jet nozzles injecting in NaOH water. The wet scrubber takes 99.9 % of Cs and around 99 % of elemental iodine, but unlike the large dry filter it's not very good at filtering organic iodine (perhaps 70 %) or noble gases (all pass through).

Thank you for the additional clarification.
This makes the industry's reluctance to install or retrofit such scrubbers more puzzling to me.
It seems like a fairly inexpensive retrofit/upgrade well worth the hassle in regulatory grief avoided.
I still do not understand how this can be effective while venting a megawatt power steam plume.
Is the vent path changed to include this unit if/when the cooling water runs out?

 

[rmattila]

Thank you for the additional clarification.
This makes the industry's reluctance to install or retrofit such scrubbers more puzzling to me.
It seems like a fairly inexpensive retrofit/upgrade well worth the hassle in regulatory grief avoided.
I still do not understand how this can be effective while venting a megawatt power steam plume.
Is the vent path changed to include this unit if/when the cooling water runs out?

See http://tvo.fi/uploads/julkaisut/tiedostot/ydinvoimalayks_ol1_OL2_ENG.pdf page 12 for a picture of the setup. The line connecting the scrubbers to the containment drywell (one of the two) has two manual valves that are kept open, and a rupture disk that will break automatically at certain containment pressure. Alternatively, for instance if the rupture disk fails to break, one of the two remaining lines can be used to bypass it by opening the manual valves in those lines. Usually the wet well line should be preferred, as it provides the additional scrubbing capacity of the condensation pool. Manual drywell path is needed if the containment is full of water and venting from wet well thus not possible.

EDIT: the picture on page 12 (page #10) is simplified and contains only one line from the drywell to the scrubber. Actually there are two as I tried to explain above.

 

[Bandit127]

BBC Video Inside Unit 4

On the page there is a video from above the SFP in Unit 4. Posted yesterday.
(I hope they haven't put restrictions on viewing it internationally).

http://www.bbc.co.uk/news/world-asia-24847381

 

[nikkkom]

but unlike the large dry filter it's not very good at filtering organic iodine (perhaps 70 %) or noble gases (all pass through).

I suspect no practical filter can trap noble gases.

 

[Hiddencamper]

Thank you for the additional clarification.
This makes the industry's reluctance to install or retrofit such scrubbers more puzzling to me.
It seems like a fairly inexpensive retrofit/upgrade well worth the hassle in regulatory grief avoided.
I still do not understand how this can be effective while venting a megawatt power steam plume.
Is the vent path changed to include this unit if/when the cooling water runs out?

As someone who works in a design department for a nuclear power plant, this type of modification is drastically more complex than it looks on the surface.

For one, you are extending containment to a location outside of the plant. You also have to add new penetrations to the containment which have a design to fail the containment in a controlled fashion. Fun fact, the primary containment is one of the only pressure vessels in all of the ASME code which is allowed to have no overpressure protection, due to the fact that it is contrary to nuclear safety for design basis accidents. There is no regulatory guidance or analysis which even supports doing something like this in the US, and if any plant did go out of their way to install it, it is very likely that it would cost easily 15-20 million dollars, and would require a rework if/when the NRC finally decides to put together regulatory guidance which explains what they think containment venting should look like.

Some design considerations that would have to be looked at (if I was preparing this engineering change authorization). Soil below where the vent unit is going to go will need geological reviews. A seismically capable pad needs to be built. The entire pipe route from containment to the vent unit and back to the elevated release point (which extends outside of secondary containment) needs seismic and weather/severe accident proof enclosures around it, and every piece of that needs calculations to determine the maximum theoretical force it can withstand to prove that it can exceed severe accident scenarios. All my leak rates for my primary and secondary containment need to be recalculated, and leak rate testing needs to be reperformed (which is challenging on the containment). The penetration work on the containment cannot be performed online, and would likely require an extended outage (not to mention that new containment penetrations have a potential risk of going bad...see crystal river 3). A lot of this work will have to go out to large external engineering firms who have the experience doing a lot of this analysis.

For me to replace a single section of pipe, or replace a single indicator in the control room, it takes about 70-80 pages of paperwork total (forms, drawing updates, authorizations, reviews, licensing analysis, testing requirements, parts list, programs impacts, procedure/training impact reviews, new vendor manuals, update forms for the master equipment list and the design basis database). For something like this, a new filter, it would likely be several thousand pages, cost over 10 million dollars in just engineering services, and take about 2 years to complete. At that point it doesn't matter if it is a "small" or "Simple" filter, the overhead cost in making changes to ASME pressure boundaries and extending containment outside of the plant is relatively astronomical. Plus when all is said and done, I need the NRC to agree to a license amendment and safety review, as this change absolutely is more than a minimal increase in the consequences of an accident (see 10CFR50.59). It takes about 1 year for the NRC to review these things, and they charge about 272 dollars per hour right now.

That's my view of it based on my experience as a design engineer at a nuclear power plant.

 

[nikkkom]

All this paperwork was added because practice showed that without it, nuclear industry does not execute due diligence and does not make their plants safe enough.

As to this particular filter, I don't see why it is not installed on the existing vent line, why a new vent line needs to be added (as opposed to "existing vent line is cut, and a detour through the filter is inserted into the cut"). This way, all changes are way outside containment.

 

[gmax137]

All this paperwork was added because practice showed that without it, nuclear industry does not execute due diligence and does not make their plants safe enough...

well that's debatable; but I think hiddencamper's exposition was really responding to etudiant's "fairly inexpensive" comment. The point is, even simple mods are hugely expensive in ways that are hard to imagine if you haven't worked in that world.

 

[Hiddencamper]

All this paperwork was added because practice showed that without it, nuclear industry does not execute due diligence and does not make their plants safe enough.

As to this particular filter, I don't see why it is not installed on the existing vent line, why a new vent line needs to be added (as opposed to "existing vent line is cut, and a detour through the filter is inserted into the cut"). This way, all changes are way outside containment.

There are many plants that don't do due diligence even with the paperwork. (Browns ferry...)

The idea for all the paperwork is due to configuration management. For every plant that is "built", there are 3 plants. The regulatory required plant, the designed plant (also known as the "paper plant"), and the physical plant itself.

When a plant is licensed, the regulator is actually saying that the paper plant is good enough to meet or exceed the regulatory required plant. The plant is then built per the paper plant, and QA/QC/testing performed to ensure the physical plant was built in conformance with the paper plant.

As far as the regulator is concerned, the paper plant is what you are licensed to, so anytime you want to make a change to the plant, you first have to update your paper plant. Then you go and do the licensing side to either demonstrate that you are still bounded by the approved regulatory plant design model, OR you get permission to deviate from that (license amendment). Then, finally, when all that is done, you are allowed to update the physical plant to bring it into conformance with the paper plant.

It really has nothing to do with diligence. The reason for all the paper work, is that is how you prove that your plant design can actually meet regulatory objectives.

As for the filter, remember that for design basis accidents the filter is not required at all. It only helps with beyond design basis accidents. For beyond design basis accidents, remember that in order to get into a beyond design basis accident, something extraordinary had to occur to make your ECCS fail. Putting a filter inside the plant just makes it vulnerable to the same common event which caused you to lose your ECCS in the first place (not to mention that Mark I/II plants have no place to put something inside their secondary containment, they are space limited and seismic/structural limited).

So just throwing a filter into the plant may look good to the public, but in reality it doesn't mean the filter will work for those beyond design basis events which you really need it for. Having the valves and equipment inside the plant means that you now have to send people into very high rad (lethal?) fields to open and close those valves (remember no guarantee of electricity). So really, you want this equipment to be outside the plant, in an enclosure that is hardened far beyond that of the plant itself, such that the events which would cause you to lose your plant ECCS (plane crashes, large tornados, extermely heat/frost, extremely flooding, tsunamis) don't cause you to lose the filter as well, and give you the ability to control the release rate from outside the high rad field. You would also need to make sure that the system doesn't breach containment during design basis accidents like LOCA, where you DO have your ECCS and you absolutely want to keep ALL the material inside containment.

 

[etudiant]

Is this not fiddling while Rome burns?
We have an illustration that the emissions from an uncontained failure post shutdown can be dramatically exacerbated by unfiltered venting. Instead of a local accident, we have hundreds of square kilometers made uninhabitable, at a cost in the multiple billions, possibly hundreds of billions.
Yet we are haggling over 10s or at most 100s of millions. Is there no International Atomic Energy Agency that can draw rational conclusions and set minimal world wide standards for nuclear installations?

 

[nikkkom]

It really has nothing to do with diligence. The reason for all the paper work, is that is how you prove that your plant design can actually meet regulatory objectives.

And you need to prove that because it is a *nuclear* plant, not a gas-fired one. Gas-fired plants have quite more relaxed rules.
Nuclear plants are more strictly regulated because accidents can be much worse than on a natural gas plant, and general public (via elected government, parliament and laws enacted by them) does not trust private owners to be careful enough without oversight.
That was my point.

So just throwing a filter into the plant may look good to the public, but in reality it doesn't mean the filter will work for those beyond design basis events which you really need it for.

Yes, it does not mean that it will work. It means that it MAY work. Say, 90% chance that it will. I bet Fukushima operators would *much* appreciate that instead of what they had!

So really, you want this equipment to be outside the plant, in an enclosure that is hardened far beyond that of the plant itself, such that the events which would cause you to lose your plant ECCS (plane crashes, large tornados, extermely heat/frost, extremely flooding, tsunamis) don't cause you to lose the filter as well

This filter is a passive device. In Fukushima, to survive tsunami it would need to only be secured to the ground strongly enough to not float away.

 

[nikkkom]

etudiant got it exactly right.

 

[etudiant]

Just as an aside on the topic of filtering, I would have thought that the nuclear industry would have been the first to ardently embrace the necessity of filters.
The 1957 Windscale incident in the UK was only a disaster rather than a catastrophe because the belatedly installed stack filter effectively limited the emissions from the burning reactor. It is noteworthy that it was not agreed within the industry that the filters were needed, but a senior scientist had the political clout to compel their installation after the plant was already under construction.
A tall stack as a pollution solution is bad policy if there is enough pollution around to keep the emission plume lethal even at 20-50 km. Scrubber and filter options have to be integral to the vent design and for an industry whose survival is at risk, this seems an unwise retrofit to refuse imho.

 

[LabratSR]

This filter is a passive device. In Fukushima, to survive tsunami it would need to only be secured to the ground strongly enough to not float away.

It seems that they should be planning for ALL scenarios, not just for what happened at Fukushima.

 

[mheslep]

Nuclear plants are more strictly regulated because accidents can be much worse than on a natural gas plant, ...

Are they really worse?

 

[etudiant]

It seems that they should be planning for ALL scenarios, not just for what happened at Fukushima.

No argument there, agree entirely.
But the idea of an unfiltered vent stack does give me pause.
I have spent time near Sudbury, Ontario, once the home of a world class nickel mine. The roasting plant for the ore has a 1250 foot stack, to keep the SO2 emissions dilute enough.
The effect is for 30 miles around, the rock is mostly bare, devoid of life courtesy of acid rain.
I don't think an unfiltered stack possibly emitting long life radionuclides in quantity is good engineering practice.

 

[Most Curious]

Would filtered venting made a lot of difference at Fukushima ? When the hydrogen explosions occurred the nasty stuff bypassed the venting system anyway.

 

[etudiant]

Would filtered venting made a lot of difference at Fukushima ? When the hydrogen explosions occurred the nasty stuff bypassed the venting system anyway.

The venting system was not brought into play effectively, afaik, partly because the valves were inoperable without power. There were burst disks, but how well they performed is uncertain, some apparently did not.
The system was unable to relieve pressure because there was no safe vent option that was available. Partly because the vent that was built in was not designed to deal with emissions from a melting reactor, venting was not ever an attractive option for the operators.


If anyone has an English language timeline of the operator choices and options during the period between the quake and the explosions, it would be a real service to post it as a sticky.
The Melcor codes show that the reactors melt down within a half hour of losing cooling, so the ability to depressurize and inject water was understood to be crucial, yet it did not happen.

 

[westfield]

As someone who works in a design department for a nuclear power plant, this type of modification is drastically more complex than it looks on the surface.

For one, you are extending containment to a location outside of the plant. You also have to add new penetrations to the containment which have a design to fail the containment in a controlled fashion. <SNIP>

Of course there is some engineering to be done but don't the existing "Hardened Vent" lines on these plants already do exactly what you say would need to be done.

Are you saying these additional filter proposals are not going to be retrofitted to these existing 'Hardened Vent" lines or are you saying most plants of this type do not have existing "Hardened Vent" retrofits?

 

[Hiddencamper]

Of course there is some engineering to be done but don't the existing "Hardened Vent" lines on these plants already do exactly what you say would need to be done.

Are you saying these additional filter proposals are not going to be retrofitted to these existing 'Hardened Vent" lines or are you saying most plants of this type do not have existing "Hardened Vent" retrofits?

I can only directly speak for what the US is doing as I've been following this pretty closely. In the US, all BWR Mark I plants have some form of hardened vent, although it is not really standardized and many of these vents don't meet all the new requirements. Mark II containment plants do not have a hardened vent. Mark IIIs are exempt as they are similar to a PWR containment and also contain many venting paths (probably too many).

The current US regulations do not require hardened vents to have filters, unless the utility is crediting filters to meet specific decontamination goals during a core damaging event. For example, Columbia generating station, due to the unique design of their Mark II containment, they are highly susceptible to wetwell bypass during an unmitigated core melt, and drywell vent filters may be required for them.

Hardened vents simply refer to a vent line which is capable of opening under post accident conditions and releasing to an elevated release point. Filters may or may not be a part of this. Currently passive filters are not required in US plants. The US industry is attempting to get permission to credit their FLEX strategies for wet scrubbing. The industry argues that FLEX equipment and strategies can achieve >1000 decontamination factor (remove > 99.9%), under more scenarios than a permanently installed filter, and would have a higher conditional success probability as the portable equipment would not be at the site at the time of the accident, meaning it is free from the common mode failure which caused the ECCS to fail in the first place. FLEX equipment also manages the core-damaging event, while a filter can only deal with releases caused by such an event. Those opposed believe that the complex strategies for wet scrubbing during a severe accident are challenging and would take resources away from managing the core damaging event. In any case, managing the core damaging event directly reduces the amount of radioactive material being released in the first place, which everyone agrees on.

 

[Hiddencamper]

The venting system was not brought into play effectively, afaik, partly because the valves were inoperable without power. There were burst disks, but how well they performed is uncertain, some apparently did not.
The system was unable to relieve pressure because there was no safe vent option that was available. Partly because the vent that was built in was not designed to deal with emissions from a melting reactor, venting was not ever an attractive option for the operators.


If anyone has an English language timeline of the operator choices and options during the period between the quake and the explosions, it would be a real service to post it as a sticky.
The Melcor codes show that the reactors melt down within a half hour of losing cooling, so the ability to depressurize and inject water was understood to be crucial, yet it did not happen.

http://www.cas.go.jp/jp/seisaku/icanps//eng/03Attachment2.pdf

There are some timelines on here. They don't talk about operator choices though. I have seen what you are looking for, but I'm having trouble finding it.

What I do know is that containment venting was challenging. On at least one unit, the rupture disc did not break. It appears the cause was the overpressure of the containment created enough leakage that pressure would not exceed the rupture disc capacity.

Also, just so we are clear, based on your comment about Melcor codes and the like. I think the timeline is a bit longer than that, as they were already 1 hour post scram (lower decay heat, and no shrink effect). If a plant was scramming on a loss of feedwater, I agree 30 minutes and you are at TAF (top of active fuel), this is primarily due to the shrink effect that occurs on a reactor scram, which causes water level to drop 30+ inches. I've had to deal with loss of feed scenarios a lot in the plant simulator, you go from +35 inches to -45 inches in the first two minutes on a loss of feed (top of fuel in my plant is around -160").

But, when you consider they were just under an hour after the scram, and water levels should have been restored to nominal ranges, there is substantially more time (up to 2 hours) prior to reaching TAF. Lower decay heat combined with greater actual inventory (no swell effect) give more time. Additionally if pressure is reduced even a couple hundred pounds, that adds another 15-20 minutes.

In either case, I get the impression you might be mixing up reactor depressurization with containment venting. It is crucial to activate the ADS (automatic depressurization system) to blowdown the reactor pressure vessel shortly after the TAF is uncovered. First off, the swell effect from ADS will cool the fuel for another 20 minutes without feed, second off this allows you to reduce RPV pressure below the shutoff head for portable pumps. This is the single best way to ensure adequate core cooling when you lose all your ECCS. But containment venting is not required for this strategy. Containment venting would only be required if you failed to inject with the portable pump for several hours. Containment venting is not required in the first few hours post accident as the suppression pool would handle the decay heat loads from the reactor for quite a while.

Alternatively, if the reactor was considered to be a lost cause, containment venting would be used to help support flooding the containment to the BAF (bottom of active fuel), and allowing conductive cooling through the metal skirt of the reactor to the water inventory.

 

[nikkkom]

It seems that they should be planning for ALL scenarios, not just for what happened at Fukushima.

you know, by now I'll settle for nuclear industry planning *at least* for the scenarios which already happened at TMI/Chernobyl/Fuku.

Because *it does not do even that* - a number of things in Fukushima, such as lack of radiometers with adequate range and lack of autonomous emergency lighting, should not be happening because Chernobyl should have taught these lessons already.

 

[rmattila]

http://pbadupws.nrc.gov/docs/ML1214/ML12143A336.pdf

Some specifics of the Swedish and Swiss filtered vents as seen by the NRC.

http://pbadupws.nrc.gov/docs/ML1217/ML12178A670.pdf

Report from the fact-finding trip.

 

[rmattila]

Would filtered venting made a lot of difference at Fukushima ? When the hydrogen explosions occurred the nasty stuff bypassed the venting system anyway.

The filtered venting system is designed to conduct the hydrogen from the containment to the stack in a controlled manner. At least in the Finnish and Swedish system this happens completely passively by means of a rupture disk, unlike the Japanese/American hardened vents, which require active opening of the valves in the venting line.

 

[rmattila]

Those opposed believe that the complex strategies for wet scrubbing during a severe accident are challenging and would take resources away from managing the core damaging event. In any case, managing the core damaging event directly reduces the amount of radioactive material being released in the first place, which everyone agrees on.

This argument is very difficult for me to follow, as the very idea behind the design in the 80's was to eliminate the need for unreliable decision-making in the case of a severe accident by enabling a completely passive initiation of venting. If your vent lines don't have filters, you probably want to keep the valves in the lines closed, and must count on the personnel to be able to open them at the right time.

 

[LabratSR]

This is about blocking the pipes and vents below the plant, not the famed ice wall.


TEPCO to start water-freezing work at Fukushima plant in Dec.


http://the-japan-news.com/news/article/0000785092



Link to an earlier Enformable article about it

http://enformable.com/2013/10/tepco-plans-new-freeze-mission-underground-tunnels-fukushima-daiichi/