Japan Earthquake: nuclear plants


by gmax137
Tags: earthquake, japan, nuclear
rmattila
rmattila is offline
#10801
Aug5-11, 04:51 AM
P: 242
Quote Quote by MadderDoc View Post
OK, that's also what I figured. That would mean non-compressible gases, including any hydrogen from metal/water interactions and any hydrogen and oxygen produced from radiolysis from the RPV would be transferred to and accumulate in the S/C.
In addition to that, if there's fluctuation in the drywell pressure due to e.g. containment spraying or intermittent venting of steam in the drywell, the drywell pressure may occasionally fall below the S/C pressure, and this will let gases from the S/C flow back to the drywell through vacuum breaker check valves in the piping connecting S/C to the drywell.
SteveElbows
SteveElbows is offline
#10802
Aug5-11, 05:08 AM
P: 630
Quote Quote by MadderDoc View Post
I suggest the basis for the data estimate for that period may be in error. If there had in fact been any striking change or elevation in emissions during March 30th-31st, it would be expected to have shown up in the measurements from the site monitoring posts, but there is nothing there to be seen:
Yes that certainly seems possible.

They used dust sampling from locations well off-site in order to come up with the release estimates. On the 15th when the highest magnitude release is thought to have happened, they could not do dust sampling due to rain.

The figures used are shown in a table on page 4 of this document:

http://www.nsc.go.jp/anzen/shidai/ge...1/siryo4-2.pdf
SteveElbows
SteveElbows is offline
#10803
Aug5-11, 05:19 AM
P: 630
Quote Quote by MJRacer View Post
Is there a reliable estimate of what percentage of Cs-137 inventory has been released so far at Fukushima?
I don't know how reliable they are, but the accident analysis documents that are public tended to estimate very percentage releases, often in the 0.6%-1% range, but I believe one of their scenarios lead to a broader range of something like 1%-6%. Its been a while since I looked at these documents, but I imagine this was for reactor 2 since their other estimates also have more stuff released from this reactor.
SteveElbows
SteveElbows is offline
#10804
Aug5-11, 05:20 AM
P: 630
Continuing the radioactive stack news, they have measured 3.6 Sv/h at the 'stack drain pipe', photos included in this document:

http://www.tepco.co.jp/en/nu/fukushi...10805_01-e.pdf
NUCENG
NUCENG is offline
#10805
Aug5-11, 03:58 PM
Sci Advisor
P: 916
Quote Quote by westfield View Post
I don't think its that black and white - I'm also speaking as a layman so someone please correct me if I'm mistaken, early depressurization to me would mean the first line of defence on a loss of power is gone, namely HPCI and RCIC both of which require high pressure steam to work as I understand it. Obvioulsy the operators would have wanted to try and keep the ability to use those systems so initially depressurization would be the last thing they want to do. The only other systems they had to control heat in the reactors were the LP ones which all require electricity (again, as I understand it).

Of course Unit#1 with its Isolation Condenser instead of RCIC is a different case. That should have worked fine without power, something else went wrong there, perhaps as simple as running out of water.

I guess my point is that the operators are trying to remove the heat from the core to prevent a meltdown, if the heat is under control there will be no pressure issues and therefore no need to vent AT ALL - From what I understand of the systems early depressurization and venting prematurely will not help the heat problem and will remove several critical systems from the picture.

From the sparse reports I've read the operators at Fukushima 1 could not keep the IC and RCIC systems running for some reason and this is one of the most important questions in my mind. If they had functioning IC and RCIC then we might not even be here on this forum now.

Perhaps someone here in the business could clear these aspects of ECCS up for us?
Is it a bad thing to lose those steam driven HP systems and try "fight" the fight with a depressureized reactor?
Let me try. I will address the IC first as it is a very simple system. I will post a follow up on the more complicated possibilities with RCIC later.

The IC basically is a sytem to take reactor steam and condense it in a heat exchanger then route the condensate back to the vessel, removing heat in the process. The system runs on natural circulation. The steam rises to the condenser and is condensed. The condensate is cooler than the water being heated in the vessel so it flows back into the vessel. That is the theory. In operation all that is necessary is to open valves to allow the condensate to flow back into the vessel. The standpipe of condensate is kept filled by steam which is continuously available to the condenser. The condenser is basically a water tank that boils off and is vented to atmosphere. To keep it running all that is needed is to continue to add water to the tank. Since the tank is vented this can be done by a portable pump or fire truck.

Failure modes are azlso relatively straight forward. If the valve can't be opened the system won't work. At unit 1 the system was started, but apparently was stopped over concern about exceeding a design limit on cooldown rate. Later they tried to restart the IC, but it is not clear whether it worked. Power to the valve may have failed. The valve itself may have failed or the high temperatures in containment could have caused boiling in the condensate standpipe. This would have broken the driving force for natural circulation. Other possibilities are that the tank was damaged and leaked or boiled dry removing the coolant from the heat exchanger.

I have looked at the data dump from TEPCO from the first hour after the erathquake. It is clear that the IC was initiated and stopped after about 15 minutes. Following the tsunami there was no active instrumentation readings released so it is not clear what prevented reinitiation. The concern about cooldown rate was probably a mistake since the vessel was probably already on the way to core damage due to the extended SBO. I do think the mode of failure will be easy to identify when conditions permit examining the piping and valves.

Hope this helps. I will try to post on RCIC later tonight.
nikkkom
nikkkom is offline
#10806
Aug5-11, 04:39 PM
P: 549
Quote Quote by NUCENG View Post
Let me try. I will address the IC first as it is a very simple system. I will post a follow up on the more complicated possibilities with RCIC later.

The IC basically is a sytem to take reactor steam and condense it in a heat exchanger then route the condensate back to the vessel, removing heat in the process. The system runs on natural circulation. The steam rises to the condenser and is condensed. The condensate is cooler than the water being heated in the vessel so it flows back into the vessel. That is the theory. In operation all that is necessary is to open valves to allow the condensate to flow back into the vessel. The standpipe of condensate is kept filled by steam which is continuously available to the condenser. The condenser is basically a water tank that boils off and is vented to atmosphere. To keep it running all that is needed is to continue to add water to the tank. Since the tank is vented this can be done by a portable pump or fire truck.

Failure modes are azlso relatively straight forward. If the valve can't be opened the system won't work. At unit 1 the system was started, but apparently was stopped over concern about exceeding a design limit on cooldown rate. Later they tried to restart the IC, but it is not clear whether it worked. Power to the valve may have failed. The valve itself may have failed or the high temperatures in containment could have caused boiling in the condensate standpipe. This would have broken the driving force for natural circulation. Other possibilities are that the tank was damaged and leaked or boiled dry removing the coolant from the heat exchanger.

I have looked at the data dump from TEPCO from the first hour after the erathquake. It is clear that the IC was initiated and stopped after about 15 minutes. Following the tsunami there was no active instrumentation readings released so it is not clear what prevented reinitiation. The concern about cooldown rate was probably a mistake since the vessel was probably already on the way to core damage due to the extended SBO. I do think the mode of failure will be easy to identify when conditions permit examining the piping and valves.

Hope this helps. I will try to post on RCIC later tonight.
Excellent post. I also track (wait for) information about IC of unit 1. They stopped IC prematurely, and didn't restart it later. So far it's not known why. There may be a valid reason (such as: IC was damaged (several possible failure modes) and simply wasn't working at all) or it may have been an error, and if they had not stopped it, unit 1 could have additional ~8 hours of life. Unfortunately, in this case ultimately it didn't matter, these additional 8 hours would only delay the meltdown. It looks like we don't have a "unit 1 meltdown could have been prevented" situation here.

But in general, IC seems like an excellent meltdown prevention mechanism - provided there is a way to passively replenish or condense water which boils off.

I'm curious why some newer designs replaced IC with more complicated cooling systems such as RCIC. I just don't see why designers replaced a simple passive system with just two valves by a system with more valves, some pumps, etc. They should have _augmented_ it, not _replace_.
Dmytry
Dmytry is offline
#10807
Aug5-11, 05:57 PM
P: 505
Quote Quote by SteveElbows View Post
I don't know how reliable they are, but the accident analysis documents that are public tended to estimate very percentage releases, often in the 0.6%-1% range, but I believe one of their scenarios lead to a broader range of something like 1%-6%. Its been a while since I looked at these documents, but I imagine this was for reactor 2 since their other estimates also have more stuff released from this reactor.
The Chernobyl released something like 40% of the Cs-137 inventory IIRC, and had IIRC comparable inventory to Fukushima reactors. ZAMG estimated first 4 days release of 50% of the Chernobyl's Cs-137, based on CTBT sensors picking up the stuff that was blown into ocean. So, >7% give or take, assuming nothing was released from spent fuel pools. Very inaccurate of course. You can look up the inventory for those reactors and compare to ZAMG source term estimate directly. The source term estimates for Fukushima based on measurements are all very inaccurate because the wind was blowing to the ocean.

The percentage release predictions are a very sensitive subject because a high estimate is expensive for the plant operators (forces them to implement filtered emergency venting, which costs money). So as per usual they seem to just make some sort of lower bound calculation that is quite low indeed.
If the lower bound is too high, you must implement safety features, of course. Via common fallacy of confusing if x then y with if not x then not y, though, when the lower bound is not too high, no filtered emergency venting.

For the source term estimates by Japanese researchers based on couple Japanese land measurement stations , the wind was blowing to the east almost all of the time, i.e. most of the fallout never reached the sensors, i.e. whatever you base on those sensors puts a lower bound on the release. I think it's the same thing as the 55% core damage estimate (100% actual) - someone plugs numbers into software he doesn't understand, presents the lower bounds as the estimates, etc.
westfield
westfield is offline
#10808
Aug5-11, 11:48 PM
P: 143
Quote Quote by MadderDoc View Post
Ultimately HPCI and RCIC would fail due to loss of DC power, but until then the operator would of course want to use them to be able to inject water. I am not suggesting the RPV vessel might have been depressurized to atmospheric, only down to a level that would still keep those systems operable, i.e. to a pressure of about 1 MPa. It would seem to me to have provided a much better starting point, a colder and more water-filled reactor pressure vessel, once HPCI and RCIC failed. And in the meantime it would have protected the dry-well and associated systems from heat damage.

.
Ok, I misunderstood and was thinking we were talking total depressurization .


Quote Quote by MadderDoc View Post
The only respectable heat sink would seem to be the large body of water in the suppression pool. Once that had been filled, i.e. heated up, there would be no way to contain the heat produced by the core. From then on it would be either vent voluntarily, or build up excessive pressure and wait for something to give in.
Indeed, I gather that without some sort of Residual Heat Removal System in operation RCIC will, as you point out, eventually be useless once the S\C gets so hot it cannot function as a heatsink any longer.

There must be a compelling reason to have RCIC instead of IC's because it seems a big compromise to have RCIC that relies on several other systems to remain useful versus IC's which seem so simple and "stand alone". Not that having an IC system helped Unit #1's predicament.

I would be interested to know when the operators at Unit 1 saw the cooling rate was too fast why did they not just use one IC instead of either using both or none?
The cooling rate was too fast with two IC's running but the heating rate was too fast when they disabled the IC's. Surely a happy medium might have been acheived with just one of the IC's running? So many questions.
joewein
joewein is offline
#10809
Aug6-11, 02:47 AM
joewein's Avatar
P: 192
Quote Quote by zapperzero View Post
Goes to show how much trust we can put in TEPCO's cutely-colored site contamination maps.
I hope that worker got out of there in time
Talking about "cutely-colored site contamination maps", this seems to be the most recent one available still:
http://www.tepco.co.jp/en/nu/fukushi...20110802-e.pdf

Tepco published it on Aug 2, 2011.

At the joint U1/2 stack near unit 1 the yellow label says "70~100" (mSv/h). On a large bubble off to the left it says "U1/2 SGTS >10,000".

I think the pictures were taken towards the west, as you see the slope behind. Also, there's a small damaged two storey structure behind the stack. This is visible inside the stack frame on the west on aerial shots from the south on Cryptome.

The vertical brown-stained pipe is probably connected to one of the two smaller pipes that run along the fat pipe from the Y-section leading to units 1 and 2. There's one at each side of the fat pipe.

Ian Goddard speculates on his site that the brownish colour is not rust but cesium. However, for that the pipe would have to be leaky, for you to see cesium condensate on the outside, not just the inside.
joewein
joewein is offline
#10810
Aug6-11, 03:13 AM
joewein's Avatar
P: 192
Quote Quote by westfield View Post
There must be a compelling reason to have RCIC instead of IC's because it seems a big compromise to have RCIC that relies on several other systems to remain useful versus IC's which seem so simple and "stand alone". Not that having an IC system helped Unit #1's predicament.
While the IC condenses steam and returns it back into the RPV as liquid water, the RCIC pumps water into the RPV using its steam turbine, while its steam gets condensed in the wet well.

Unlike the IC, the amount of liquid water the RCIC can feed back into the RPV is not limited to the exact amount of high pressure steam it receives from the core, as it pumps from the wet well.

Where this becomes significant is when pressure in the RPV rises such that steam has to be vented from there into the wet well. With the IC alone that steam could not be replaced under station blackout conditions. Consequently, the water level in the core would have to drop. With the RCIC water in the RPV can be topped up after venting as long as the wet well has not exhausted its heat sink capacity. I think this explains why melt down occurred much later in units 2 and 3 than 1.
MadderDoc
MadderDoc is offline
#10811
Aug6-11, 04:30 AM
MadderDoc's Avatar
P: 698
Quote Quote by westfield View Post
<..>
I would be interested to know when the operators at Unit 1 saw the cooling rate was too fast why did they not just use one IC instead of either using both or none?
I reckon you'll be interested in this Tepco document, on the logs and testimony of operator response during the first days after the earthquake.

From the available evidence and testimony, the operators do in fact appear to have opted for the use of just one of the IC systems (the 'A' system) for the control of reactor pressure, judging it to be sufficient to keep the vessel at 6-7MPa, while they initially relied on the HPCI system for the control of the reactor water level.
tsutsuji
tsutsuji is offline
#10812
Aug6-11, 04:53 AM
PF Gold
P: 1,220
Units 1 and 2 injection rates are unstable again:
[unit 1] Water injection was arranged at approx. 3.9 ㎥/h at 9:02 am on August 5
since we observed reduction of the water injection into the reactor.
[...]
[unit 2] Water injection was arranged at approx. 3.9 ㎥/h at 5:50 pm on August 4
since we observed reduction of the water injection into the reactor.
http://www.tepco.co.jp/en/press/corp...1080501-e.html
Various troubles at the water treatment facility:
At 5:32 am on August 4, we stopped Water Treatment Facility to improve the flow rate. After the work, we activated the facility at 3:30 pm and restarted operation of the water treatment system at 4:13 pm. When we adjusted the flow rate of the system at 6:55 pm, a pump of the decontamination instruments was stopped and the whole water treatment system was shut down. We confirmed soundness of the pump and reactivated the system at 8:30 pm, and operation of the water treatment system was restarted at 8:50 pm.

-At 2:12 am on August 5, a process alarm was activated and the water treatment system was shut down. At 4:03 am, the system was reactivated and the operation was restated at 4:21 am.

-Around 7 pm on 8:04, we discovered water leakage from a flange of transfer hose of filtered water which is used to clean up salt in a vessel of the cesium adsorption instruments in the site bunker building.

http://www.tepco.co.jp/en/press/corp...1080501-e.html
http://www.tepco.co.jp/en/nu/fukushi...10805_02-e.pdf map showing the location of the leak at the site bunker building.

http://sankei.jp.msn.com/affairs/new...2450013-n1.htm A 700 l leak was found at the water treatment facility at 7 PM on 4 August. The water leaked from a pipe connection. It did not flow outside the building. When they are removed, spent adsorption towers are cleaned with freshwater to remove salt, which is a source of corrosion. It is water from this process that leaked.

http://www.yomiuri.co.jp/science/new...htm?from=main1 Until that leak occurred, Tepco had never measured the radiation from this water. The loose safety management is brought into sharp relief again. The leak is from the hose that takes the water back to the system after washing the towers. With 6,270,000 Bq/cm³ of CS-137, it is about the same radiation level as that of the water in units 3 and 4 turbine buildings basements.

http://www3.nhk.or.jp/news/genpatsu-..._shiunten.html The 700 l leak rang an alarm, after which the facility was stopped for more than 2 hours. Tepco has decided to delay the test run of SARRY initially planned to be performed for 2 days starting from 6 August. It is delayed to the middle decade of August or later. The reason is that this test requires to shut down the whole facility for 2 days, but Tepco cannot afford to do this as the water level in a waste treatment facility basement reached 30 cm below the maximum on 5 August.

http://www.47news.jp/CN/201108/CN2011080601000367.html At 12:50 PM on 4 August, there was a short blackout at the earthquake-isolated building. The electricity was restored within the first minute using a backup power source. It was found that a 2.5 m deep underground cable had been harmed during an excavation work performed as part of a ground water survey preparing the construction of the ground water shielding wall. The cable was changed and the electricity from that power line was restored within 3 and a half hours after the blackout. In the future, Tepco will make sure that underground cable maps are checked carefully enough before starting excavations.

SFP1 was being thirsty:
From 3:20 pm to 5:51 pm on August 5, we injected water to Unit 1 by using Fuel Pool Cooling and Filtering System.
http://www.tepco.co.jp/en/press/corp...1080602-e.html
MadderDoc
MadderDoc is offline
#10813
Aug6-11, 05:28 AM
MadderDoc's Avatar
P: 698
Quote Quote by joewein View Post
The vertical brown-stained pipe is probably connected to one of the two smaller pipes that run along the fat pipe from the Y-section leading to units 1 and 2. There's one at each side of the fat pipe.
The brown-stained section of the pipe is the last final bit of pipe before entry into the stack, this last bit is shared by the exhausts from unit 1 and 2 EGTS (Emergency Gas Treatment Systems).

Right above the stained part you see the forking out of one pipe up and towards south, that pipe is connected via a 90 deg bend to the smaller pipe along the fat pipe, coming from unit 2. The other fork proceeds vertically for a bit, then makes an upwards bend towards the north, and then a 90 deg bend to become aligned and connected with the smaller pipe coming from unit 1. I think you can make out the arrangement from the attached photo that is shot from the south-west.

Ian Goddard speculates on his site that the brownish colour is not rust but cesium. However, for that the pipe would have to be leaky, for you to see cesium condensate on the outside, not just the inside.
Cesium compounds are generally colorless, so taking the color brown as an indicator color for the presence of cesium appears like madness. Quite on the contrary. if a cesium mineral is found to be brownish a geologist will reasonably suspect the color is due to the presence of impurities, e.g. Fe3+ impurities aka 'rust'.
Attached Thumbnails
unit1-2_egts-exhaust.jpg  
Dmytry
Dmytry is offline
#10814
Aug6-11, 05:39 AM
P: 505
The stain is indicative of some sort of leak, which evaporated, depositing the dissolved material. You can guess it would include radioactive isotopes. Also the pipe bends inside meaning some dust could have deposited there. The piping would be very radioactive even if only a small fraction of the vented material had deposited.
AtomicWombat
AtomicWombat is offline
#10815
Aug6-11, 07:36 AM
P: 150
Exactly where does the pipe come from. It's hard to follow in the Cryptome pictures. It is not the large emergency relief duct from the airspace in the reactor building. It appears to be an emergency steam relief pipe from the reactor circulation, perhaps from the condensers.

It may have flooded as a result of too much water added to the RPV. It is possible that the bottom of the stack is filled with water. I notice the most severe corrosion is at the join between the pipe and the stack. There is thick brown layer of rust on the shield below it.
Attached Thumbnails
10svha.jpg  
joewein
joewein is offline
#10816
Aug6-11, 08:29 AM
joewein's Avatar
P: 192
Quote Quote by AtomicWombat View Post
Exactly where does the pipe come from. It's hard to follow in the Cryptome pictures. It is not the large emergency relief duct from the airspace in the reactor building. It appears to be an emergency steam relief pipe from the reactor circulation, perhaps from the condensers.
It comes from a filter room on the second floor, where also extremely high values were measured.

Attached are cropped images from a Cryptome set.
Attached Thumbnails
aerial-2011-3-30-0-50-45.png   japan-earthquake-2011-3-30-0-50-12.png   aerial-2011-3-30-1-10-7.jpg  
robinson
robinson is offline
#10817
Aug6-11, 10:52 AM
P: 201
So after something goes very wrong you vent the reactor directly into the atmosphere? What the .. ?
jim hardy
jim hardy is offline
#10818
Aug6-11, 12:39 PM
Sci Advisor
jim hardy's Avatar
P: 3,139
Quote Quote by robinson View Post
So after something goes very wrong you vent the reactor directly into the atmosphere? What the .. ?
better a controlled vent through filters than waiting for it to vent itself.
Think of the engine coolant reservoir in your automobile - keeps ethylene glycol off the pavement, within its design limits.

recall the operators were VERY hesitant to use it - the top brass had to directly order it.

old jim


Register to reply

Related Discussions
8.9 earthquake in Japan: tsunami warnings Current Events 671
New Nuclear Plants Nuclear Engineering 9
Gen IV Nuclear Plants Nuclear Engineering 10
New Nuclear Plants Nuclear Engineering 14
Astronomer Predicts Major Earthquake for Japan General Discussion 65