Japanese Quake and tsunami

I have been busy working 13 hour shifts 6 day/week at my local nuclear plant since 3/5/11 and have not had much time to spend on-line. I just scanned most of this thread & noted that cthippo is doing a pretty good job of sharing good pertinent info.

Frequently we try to push the use of technology to make our lives better. Given a choice, our society is not going to give up cheap electricity. This event is very unfortunate but it may help us to learn and get better at one aspect of technology.

To be honest, we should keep things in perspective. This was a terrible natural disaster with much human consequence. So far the nuclear contribution to death and injury has actually been small. If the emergency workers get an even break from this point forward, the nuclear contribution to the overall impact should remain small.

I have spent most of my career working at a US nuclear plant that has the same basic design as Units 2 -5 of the facility. As a safety system engineer, my career has been all about protecting the health and safety of the public. Of course I have been going over what limited real information is available about the damaged facility. I have tried to understand the events, and I will share a preliminary summary of what seems the most likely series of events combined with my opinions with respect to some what-if stuff.

One one hand the event showed us is how well the GE guys designed the plant to start with. At the same time it also showed a huge flaw in the basic regulatory process that drove the original design. In recent years, the US NRC has been pushed some minor changes that helped address the flaw, but in my opinion, we still have a good ways to go. The flaw was the arrogant approach where the plant safety features focused only on fully automatic protection against a set of postulated design basis events. These fully automatic provisions in fact can turn out to be you worst enemy when you are hit with something you did not plan for. This could be a natural event that exceeds what you expected, it could be a combination of events/failures that stack up, or it could be an attack by nuts that high-jack airliners.

On the good side, the entire facility did great at surviving an earthquake much more severe than it was designed for. The three operating plants automatically shut down, no major structural or piping system failures were noted and the emergency diesels automatically started and provided the power needed to operate the emergency cooling systems. Later, after the explosions, the containment structure has still be doing the job of minimizing releases.

On the bad side, when water washed over the site (reportedly at 30' as compared to the 20' that they were designed for), the diesels were lost and the facility was not prepared to deal with the extended loss of power to the crucial big pumps.

Initially the Unit 1 isolation condenser (this uses reactor steam to boil water on the back side of a heat exchanger) provided passive reactor cooling. This works fine until you run out of water on the back side (AC power is used to replenish this water). Unit 2 & Unit 3 used battery controlled steam driven turbines to drive pumps to push water into the the vessels (at around 1000 psig) for reactor cooling. Unit 1 may have also used one of these systems after they lost their isolation condenser. However, the steam that boils off from the reactor is discharged through relief valves to the ~150,000 gallon suppression pool that is part of the Containment structure (steel lined concrete structure designed for 50 - 60 psig). This works fine until either the batteries run out or the suppression pool (also used as the water source for those turbine driven pumps) starts to boil.

Much of what went wrong was the lack of provisions and plans that would let the operators do what they knew they needed to do. When the reactors automatically shut down, the reactor pressure vessel has relief valves that discharge steam to prevent an over pressure condition. This leaves the pressure sitting around 1000 psig. When they lost the ability to put high pressure water into the vessel, they have some ability to use air operators on the relief valves to depressurize the vessel to allow use of lower pressure pumps for makeup. However, there is a reluctance to take this action, and most of the low pressure pumps available for make up are fed from those diesels that failed. As a result, they let the reactors stay at high pressure and just boil off water until the top of the fuel was no longer submerged. They may have tried depresurizing, but if so, the action was too late and/or the low pressure injection flow was inadequate. This led to a fuel overheat problem where a high temperature reaction produces hydrogen in the vessel. This was not an immediate explosion threat since there is no oxygen available at this location.

The steam discharging to the suppression pool carried the hydrogen to the containment. Again there is normally no oxygen at this location (it is kept inerted with nitrogen). However since there was no suppression pool cooling (it relies on the diesel for power) the pool approached boiling and there was a rapid increase in containment pressure. They opened some containment vents to prevent gross containment failure. However, they may have waited too long and the pressure probably caused a vent line to rupture or otherwise leak into the reactor building. This leak allowed hydrogen to accumulate at the top of the buildings.

A big quantity of hydrogen accumulated in Unit 1 and blew off the thin siding at the top of the building. This did not cause much damage below, an Unit 1 got to a relatively safe condition (reactor depressurized, fire water system pumping sea water to the vessel and containment vented) shortly thereafter.

On Unit 3, more hydrogen appears to have accumulated before it detonated. This caused a larger explosion. The explosion and falling debris from this explosion wiped out some portable generators that they were setting up along with causing damage to the other units. This explosion made the site unsafe for many of the actions that were needed to get everything under control.

All of the Containment structures still seem to be more or less intact and sea water is being pumped into the depressurized reactors. There is not much potential for the fuel in the reactors to "melt down" and get out of containment. The spent fuel stored in the fuel pools is probably a bigger concern, but I am hoping that cooling water (externally applied if needed) will be able to get the pools safe. Unless there is a big turn for the worse, this event will probably not release enough radioactive material to result in serious environmental damage or widespread significant health problems. I expect the damage to be more along the lines of the impact of the early A Bomb testing in the US.

Had this plant had better provisions and procedures for getting a safe as possible as fast as possible they would have been much better off. This would be using manual valves and manually controlled non-AC pumps for reducing reactor pressure and providing reactor makeup. It would be procedures and flow paths to manually vent containment early to avoid ever pressurizing it in the first place. In the US, we have been making some improvements in these areas since 911, but I do not think that the reactors in Japan were required to make similar changes.

Again, automatic systems are needed for certain roles and these systems did what they were designed to do. What they did not have is the simpler manual provisions and systems that are more capable of letting properly trained staff do what they need to do when things go really wrong. The thought process of over-reliance on automatic and fail-safe stuff is dangerous given that man’s ability to predict the future is so limited.

 

P Flados,

Thank you for bringing clarity and sensibiity to this thread.

-Tom

 

Thanks for filling in the stuff I was missing. Always nice to get a professional's view on it.

What do you know about the status of units 5 and 6? Is there a possibility they could be brought back on-line?

What is the current industry thinking on the consequences of a fuel melt accident based on the Chernobyl expeirence? They used to say it would burn it's way to the center of the earth, but there it just diluted in sand until it cooled into a lump. Has this changed the planning models?

 

Q1:

Units 4, 5 and 6 were all shut down and probably not fully assembled (vessel head off for refueling) when this started. Unit 4 got in trouble just after Unit 3 blew up right next to it. Unit 5 and 6 have been under control pretty much all along. The big challenge for them is first political. Will the public / government be OK with a restart effort? The site and/or plants would probably need changes to ensure no loss of Diesel power for an event that might be even slightly worse than the one they had. If you address these issues, then you will have to clean up the site enough to provide normal access for the rather large work force needed to finish the work that was on-going at Unit 5 and Unit 6.

If they are smart, they will do all of the above, but in a slow and deliberate fashion. In the US we are thinking about spending 11 - 14 billion dollars each for new reactors in the same size class of Unit 6. Japan does not have a lot of great alternatives to nuclear, and discarding two undamaged units would not be good.

Q2: The "melt down" sequence is actually not impossible and it is potentially significant. Lets discuss a no cooling water at all sequence. Spent fuel generates decay heat. With no cooling water, fuel would melt and collect into a large mass of very dense liquid metals. It would then melt through the bottom of the reactor (around 8" of carbon steel). If it spreads out a lot, the individual clumps would probably have enough heat transfer away such that the temperatures would be low enough to not attack the surface they land on. On the other hand, a bunch may re-form into a large mass just below the reactor. The bottom center of the mass of fuel has the least heat transfer away so it gets the hottest. The temperature will just keep rising until the material under the mass either melts or vaporizes. If you are lucky, the material would just continue to go straight down. Some fuel would probably be distributed along the way and some heavy material in the ground may dilute the material in the mass. If could eventually just stop when the heat transfer from the mass is enough to to keep the surrounding rock intact. If the mass were big enough and the geology right, the molten fuel could indeed eventually penetrate the crust where the heavy metals would eventually join the much much much larger lump of molten radioactive heavy metals (my opinion) that are at the center.

The real concern would be having the molten mass hit an aquifer or some other structure that would spread material on the way down.

A second bad sequence is to have the molten fuel mass melt through the bottom of a reactor while it is still at high pressure. A "high pressure melt ejection" would be very messy and we are not real sure what it would look like. If a bunch of water is just below the reactor (likely for most cases), something like a "steam explosion" with entrained fuel material could occur.

If you can just keep the fuel submerged in water (and preferable at low pressure), the possibility of any of the above goes away.

 

they should have backup gens on trailors off-site.

Close, but not too close. In the 5-20? mile range. That way if an extra big wave, or electo-magnitic pulse weapon, or riot, or terrorist knocks out the on site gens they would have a backup in place.

This isn't some extra major expense for some one in a billion event.

Lot of things could've knocked out the on site backups, like bad old fuel, or an incorrectly installed bearing that was about to burn out the last time tested but barely made it.

Plus, big mobile gens are something have retain a lot of value on the 2nd hand market, so you are really only renting. Or maybe just signing a contract to have 'first dibs' on such equip from a local tool rental place.

I can't believe in Japan, even after the quake, it took so long to bring in extra gens.

I would have guessed they would have big American military type helicopters landing new gens within an hour or so.

 

They did bring in portable generators. They had problems getting there due to damaged / debris filled roads and there were also debris problems at the site. If they had some at the site before hand, they probably would have washed away when the main diesels were lost in the 30' wave. Once they got them there, they had to set them up and connect them. The most important loads are probably the 4160 VAC, 3 phase pump motors that are in the 300 - 1000 hp range each with probably five or six needed to do different things all at the same time. Not some extension line thing that you just plug in.

Shortly after they got at least some of the portable stuff hooked up, Unit 3 made an absolute mess of everything. Trying to guess at what I read, this explosion probably killed some emergency workers out in the yard that were working on just this kind of stuff.

Hindsight will tell you that a backup power facility a little further from shore and located on a high hill would have been great. This kind of stuff is not cheap and it is not something that you can just borrow. You have to buy them, set them up, possibly build a hardened building for them, run cables to the plant, add connections at the distribution centers, periodically test the stuff, maintain the stuff and keep it all ready to use at a moments notice. The expense and effort continues for decades and normally it is all just to be ready for an event that usually never happens.

Same kind of hindsight would have helped us not lose a couple of space shuttles.

 

P Flados- Are the new reactor types in development/use less vulnerable or reliant on these ancillary systems such as the external pumps?

I have been told that "new" reactors have different core designs in which convection circulates heat away from the core in some passive manner.

It seems unlikely given how much heat must be removed but this was the opinion given.

 

The newest design GE nuclear plant (ESBWR) is described in documents you can find on the NRC web site. It makes extensive use of passive cooling techniques. It is also bigger than most of the competition.

The Westinghouse AP-1000 design also has a passive cooling system.

 

Is the pebble bed design ready for commercialization? I seem to recall that at least had it to prototype stage out at Idaho.

 

P Flados,

Again, thank you for your insight here.

Indulge me if you would: Why are you having to work such horrendous over-time? Is this not counter-productive to the effort? I would parallel it to having air traffic controllers work over-time. Is it not dumb? It's 78 hours a week for over two weeks! PM me if you like, I realize you may not have the luxury to speak freely.

-Tom

 

WTF - Hippo is interposing indications of sense into his Googling? I was just about to suggest building sloppy and moving housing projects full of progressives near US plants...

 

You get less radiation uptake living near a nuclear plant than you do downwind from a coal plant.

 

We'll switch to coal later.

 

This has rapidly fallen into one of those ridiculous thread like the Global warming thread.

No one knows what the hell they are talking about.

Its a Japanese reactor in Japan and no one has been to see it.

Not only has Japan kept its reactor information close to its chest the numbers of people missing are only those that have been reported missing.

Its beginning to be apparant that Japan is not the fabulous futuristic country I have been led to believe

 

I heard that