hindsight- how to avoid repeat of Fukushima 4

[davefitz]

In anticipation of the need to incorporate "lessons not quite learned" from the last 3 meltdowns, here is my list economically and technically achievable means to avoid repeats:

a)add a hardwired, remote , backup operating room 0.7 km away from the reactor.
b)spent fuel storage pool located at least 0.5 km away from reactor
c)backup diesel genset at ALL coastal plants to have inlet air snorkel be at least 10 M above sea level. Ditto exhaust discharge nozzle.
d)use geopolymer concrete for containment dome- at least for the foundation- tolerates much higher temperatures than ordinary portland cement concrete.
e) (??) develop and implement a stirling engine based backup circ pump that is driven by primary coolant discharge , air cooled at cold end
f)(??) onsite mandatory storage of sufficient boric acid for loss of coolnat events- perhaps stored in a pressurized local accumulator that can be immeidiately discharged to the core , based on 2 independent keyed switches plus control system confirmation loss of coolant for ( XXX minutes).

 

[BJC]

"In case the electrical grid breaks down, develop technology to the point where a backup failsafe command can be sent piggybacked via a very remote laser to a sensor that would remotely activate the Boron release to the core. The OP's item "f" is assumed in place."
Poison pumps are powered from the emergency power system. For that reactor they are piston pumps. The boron solution is a solid at room temperatures. The injection lines have to be heat traced. There is a squib valve between the pumps and the reactor.

 

[burnt2x]

How about if they re-configure the shutdown command (or is it an emergency procedure) to allow for the immediate run-up/ re-synchronization of even the smallest unit (I guess in Fukushima Dai-ichi it was Unit No. 1 - 460MW) to prevent a total blackout at the plant site? If Fukushima Dai-ichi site has six units, perhaps the total in-house loads could load that unit by some %! (That is assuming the auxiliaries power centers were not damaged during the tsunami). If it took them 1 hour running on emergency diesels, they should have ample time to cut-back in one nuke generator, line up the emergency cooling systems to take power from that nuke unit before the emergency diesel got swamped over.

 

[trottiey]

Sure, bringing one reactor back up to cool the others might work, with some cleverness to override xenon poisoning and reject the excess heat. But now you're playing double-or-nothing. You've just turned a nuclear reactor back on right after it went through a record-breaking earthquake, and you haven't had time to give it a full inspection for damage. What if the coolant pumps are toast? Do you want a risk of loss of coolant on three shutdown cores, or a risk of one on-power core out of control?

You may be interested in thread466-294381 where several of the same back-up power issues were discussed.

 

[cloa]

Any Daiini is not that far away and its fully shutdown so they could bring it only a few percent online and supply plenty of power.

 

[burnt2x]

Sure, bringing one reactor back up to cool the others might work, with some cleverness to override xenon poisoning and reject the excess heat. But now you're playing double-or-nothing. You've just turned a nuclear reactor back on right after it went through a record-breaking earthquake, and you haven't had time to give it a full inspection for damage. What if the coolant pumps are toast? Do you want a risk of loss of coolant on three shutdown cores, or a risk of one on-power core out of control?

Yep, I guess xenon135 and samarium149 poisons the fission reaction after each nuclear trip but I guess you always have a poison override window right after a trip. Given the facts in Fukushima Daiichi, it seems the turbines and the steam generator/ nuclear reactor rode-thru the quake with out damages. I was basing my suggestion on the fact that for any external faults to the generators, it is possible to immediately run-up and bring the unit back to the system.

The good thing with a well-designed unit control is that you cannot line up the unit for connection back to the grid if the lockouts are not reset; and you cannot reset if there are problems detected by the relays' sensors. So, in effect, if a trip occurs and the unit elements get damaged, you can't run it back, answering your fear of putting back a unit which experienced a very strong quake.

 

[rcchap]

There are a lot of interesting ideas stated above. I certainly don't want to discourage discussion, but unfortunately most of the suggestions above won't work.

I have worked in the nuclear industry for 13 years, so here are the facts and my 2 cents:

- Fukushima was hit with a beyond design basis event. The information I have is that they lost offsite power as a result of the tsunami, and then went into a station blackout (SBO) condition as a result of tsunami damage to the fuel supply to the diesel generators.

- Most US plants are designed to accommodate about 4-8 hrs in a SBO. During this time steam driven pumps were used at Fukushima. Power is needed for instrumentation and controls, which comes from batteries. The batteries eventually need power to be charged, so when they go dead, you have nothing. At Fukushima, this happened after about 8 hours.

- As a result of this event, the NRC will likely ask most plants in the country to evaluate their capability to withstand an extended SBO condition, and plants on the coast lines will need to consider their capability to withstand a tsunami (or a seiche for an inland lake or reservoir). I also imagine that the beyond design basis procedures will be upgraded with possibly a focus on safely venting off hydrogen gas with no power without causing an explosion.

So for the ideas proposed above by davefitz:

a - No US plant has an offsite control room. That would be an incredibly bad idea, and would not have helped at Fukushima since they had no power. All control rooms are right in the middle of the plant. The control room must be an incredibly robust room with independent ventilation systems and tight security. The offiste facilities are typically just emergency response centers (desks and telephones).

b - The SFP will always be close to the reactor because the fuel must stay underwater (for shielding and cooling). Storage facilities that are farther from the plant are dry fuel storage, which does work, but not for fuel assemblies that have recently been used. And you still need a pool for a core offload.

c - Elevated air inlet and exhaust air for diesels is definitely a possibility. But the external fuel supply is what caused the problem, and that needs to be looked at.

d - No need to change concrete type. Normal concrete works just fine. Tearing down and rebuildling every containment in the US won't happen, and the NRC would never allow the use of an unproven and untested type of material, and it would also be cost prohibitive.

e - Many plants have steam driven pumps that use steam produced from the reactor to operate. They still need power for instruments and control valves.

f - Plants already have this. Boration is not the issue. It is long term removal of decay heat and loss of coolant inventory due to boiling.

And for the other ideas:

- Starting up a reactor just after a beyond design basis event would never happen. This would be incredibly reckless and foolish. You never start your reactor back up until you fully understand why it shut down and have performed inspections and surveillances on plant equipment as required.

- Changing coolant is not an option and is not needed. You would need an entirely different type of plant. The type of coolant is not the issue.

- An additional huge tank of water is not needed since all plants will have an ocean, lake, reservoir, or river to get water from.

Here are some better ideas:

- Tsunami proof emergency power and upgraded beyond design basis event procedures. Ensuring power is available addresses both core cooling and SFP cooling.

- Passively designed reactors that can cool using natural circulation and require minimal emergency power. Some new reactor designs are this way (AP1000).

- Improved hydrogen venting capability to prevent explosions.

 

[Zogzog]

"No US plant has an offsite control room"

No commercial plant

 

[rcchap]

Correct Zogzog.

All of my statements above apply only to the light water civilian nuclear power plants in the US, and are not meant to be applied to any other type of reactor.

 

[davefitz]

rcchap:

Good points all, and thanks for the education. Yet I still take issue with 2 items, going forward.

a) Today's availability of a DCS system makes it a nearly trivial task to allow a plant to have a remote control room for emergency use. That technology has been available since about 1985 ; being developed after the bulk of the US plants were built . Certainly the main control room needs to be close-coupled to the operating equipment for all the practical reasons you discussed, but after 3 emergencies, one can no longer pretend they do not happen. The most likely future emergencies at overseas plants would be war related ( Pak vs India, anyone vs Iran, asian conflicts are feasible) , but other causes are plausible, depending on your imagination- GPS spoofing or solar flares upsetting the synch of the grid, smart transmission system programming faults, and natural catastrophes.

b) the F-D-I 1-6 instant crisis shows an equal risk being associated with loss of coolant for the SFP- in hindsight, if each pool had been supplied with 2 redundant 12" NPS pipes routed from the pool to a termination perhaps 0.5 km away, then emergency crews could safely concentrate on supplying a coolant fluid loop to the pool when local devices have failed and local activity cannot be conducted .

Although the news reporting normally focuses on local damage caused by the event, there is long term damage that occurs far from the site that effects other countries, so perhaps there also needs to be an international agreement along the lines that war time targeting of such sites which leads to a level 5+ event represents a "crime against humainty " or some other such prohibition.

 

[cloa]

http://www.nytimes.com/2011/03/22/world/asia/22nuclear.html?pagewanted=2&_r=1&ref=world

Considerable reform of the regulation of the Japanese nuclear industry is needed.

 

[rcchap]

davefitz:

We can agree to disagree I guess. I know that there are advanced digital technologies out there, but they are rarely used in the nuclear industry, and would definitely not be used for reactor control. We keep it simple and analog. A networked control room will never happen because some hacker would find a way to cause problems. The current cyber security requirements already in place are very restrictive, and they are getting more restrictive every year, and these requirement would preclude a remote control room using a network.

There are digital computer systems used at plants to monitor parameters only ... they have no control functions.

There are also alternate control panels at the plant for specific equipment that are used in the event that the control room is uninhabitable due to a fire. These panels are inside the plant are used to control safe shutdown.

Keep in mind that if the plant is in an SBO condition, it doesn't matter where the control room is ... it is useless without power to the plant.

As for the SFP, I agree that upgrades should be done to be able to provide water to the pool during an extended SBO condition. The cascading series of events that unfolded at Fukushima where a core melt event turned into damage to the SFP and loss of SFP inventory clearly identifies an opportunity for improvement. Many plants have evaluated redundant methods that can be used to provide water to the SFP (including transferring water from various holdup tanks and using other cooling water systems or even fire protection). However, all of these methods require power. I imagine that this will be looked at very closely in the coming months.

On another subject, I have seen a report where Isolation Condensers were considered for installation at Fukushima, but it was decided not to install them. These would allow removal of core decay heat with natural circulation. I am sure the anti-nukes will get a lot of mileage out of that. Some US BWRs have these.

 

[VEBill]

With perfect hindsight...

It seems incredible that the 'spent' (a bit of a misnomer) fuel rods are removed from the highly-sealed reactor, withdrawn from the ultra-secure, ultra-strong primary containment structure, and then....they are dropped into what is basically a 3rd floor swimming pool under a tin roof.

Corrective action to better protect the spent fuel rods (and their water) when they're outside the reactor should be relatively inexpensive.

 

[burnt2x]

I know that there are advanced digital technologies out there, but they are rarely used in the nuclear industry, and would definitely not be used for reactor control.

I found this in the internet!

 

[VEBill]

Link?

 

[burnt2x]

Uhuh, sorry. Missed to add the link.
Google
Or Here

 

[rcchap]

What you have found is a simulation program. This seems to be a simple one. Most (if not all) US reactors will have a full scale simulator room that mimics the actual control room and is controlled by computers that are programmed to mimic plant operation. These are used for operator training and NRC license qualification.

So, of course, digital technologies are used for simulators and plant process monitoring systems and to perform calculations. My statement is still true. This type of digital system would not be used to control any actual plant equipment, at least on the existing US nuclear fleet.

The newer generation of plants will probably have more digital controls, but to what extent I do not know.

 

[rmw]

Not that I trust a thing that I see or read in the media, there was a scene shown of a control room on some TV channel a day or two ago that supposedly was the Fukujima plant in happier days. At least all the operators in the CR were all oriental.

I noticed that all or at least most all the stuff on the walls and operating consoles was analog.

Made a gauge tapper like me feel right at home.

rmw

 

[unclesyd]

Opening the door and trying to read a 28 channel temperature recorder. I didn't think there were any left except me.

Just heard on the news that there is serious possibility of compromise of the containment vessel in Reactor #3. Based on a previous report this could almost take out Japan as safe place to live.

 

[owg]

rmw - I think there are quite a few "gauge tappers" on this site.

HAZOP at

http://www.curryhydrocarbons.ca

 

[CodeRef]

rcchap is right on with the comments....for my $0.02...the ability to have natural circulation capability is key just as it is in Naval Reactors and stick with the PWR vs BWR..and my final comment...to pay for this retrofit, utilize the money set aside for the failed Yucca repository. Looks like Code Shops will be busy doing retrofits until new plant designs incorporate and cost out all that results from the above comments. Let alone the NRC mandates.

FAQ731-376 A question properly stated is a question half solved.