13 March 2011

~ REAL INFORMATION FROM EXPERTS on meldown: Metafilter ~


A few answers to a few questions.

1) Can the core undergo nuclear detonation? No. The concentration of 235U and 239Pu is far too low, there are far too many other fissionables that would prevent the full prompt-critical reaction from being sustained, and there's no inertial confinement. The hardest part about building a nuclear weapon is holding it together while it's trying to explode.

2) Can the core undergo other sorts of explosions ? Very hard, but not impossible. SL-1 and Chernobyl Reactor 4 both had large steam explosions when the reaction rate jumped by a couple of orders of magnitude due to loss of control. SL-1 was an inadvertent control rod removal, Chernobyl #4 was a combination of positive void feedback, existing reactor poisons, and a bad design.

These sort of runaways are, as far as we can tell, impossible in a PWR or a BWR, which count on water as a moderator *and* coolant -- if you lose the coolant, you limit the reaction rate. As the coolant heats up, the reaction rate moderates.

BWRs and PWRs also depend on pressure to work (more pressure in a PWR, of course) -- thus, the pressure vessels are much stronger than the basic reactor vessels of graphite moderated reactors like Windscale #1 & #2, or the RBMKs at Chernobyl. You can exchange a moderate radioactive release to completly prevent this possibility -- vent steam, add water, keep the pressure low. This is basically what Fukushima #1 is doing.

3) Can we get a criticality accident? Harder, but not impossible, and actually more likely than an explosion, but you need severe core damage for it to happen. With the control rods in and coolant limited, you have a bunch of neutron absorbers in place, and with the coolant very hot, fewer prompt neutrons are being moderated into thermal neutrons, which are the ones more likely to cause a further fissioning.

But if the heat becomes extremely high, the core can melt. Liquids flow, and they'll flow down into the bottom. Intermixed with that will be bits of control rod, but it'll basically be random. Get enough fuel material together, and not enough control material, and you can reach criticality. This is often hard on people nearby. It's not an automatic thing, though -- Chernobyl #4 melted down after the explosion, but the "corium" that flowed out of the reactor and into the basement didn't go critical. This release of core material was very limited, easily contained, and while it made that building suck, it didn't do the massive damage.

The core material blown out of the top of the reactor by the steam explosion, followed by the core material carried up in the ashes of the burning graphite moderator, that's what caused the massive release of radioactive materials. If Chernobyl had just melted down and flowed into the basement, we'd have been a lot better off.

Note that the graphite fire that put this stuff up into a fly ash plume that could carry for miles won't happen here, because there's no graphite to burn.

4) Will any of these reactors run again? Depends. Fukushima 1 reactor #1, no -- even if it turns out there is very little core damage, they were going to decommission this reactor anyway. If F1#3 has a similar issue, they'll probably decommission that one as well. 4-6 are newer, and were shut down completely before the quake for maintenance, and should be fine - #6 is a new reactor, a BWR-5 rather than a BWR-4 of 2-5, or the BWR-3 that #1 is.

Shutting down #1 and #3 will cost Japan 1.24GW of power. Shutting them all down will cost them 4.69GW of power, which is a big chunk. Shutting down the four BWR-5s at Fukushima 2 would cost them another 4.4GW of power.

The big problem isn't the boric acid -- though it is an acid, and it can increase corrosion. The big problem is the chlorides in seawater. They'll need to be cleaned out before the reactor is safe to use again, and given the age of the Fukushima #1 reactors, it probably won't be worth doing so -- indeed, F1#1 is at end-of-life, anyway, and one report I saw said it was slated to be shut down permanently for decommissioning at the end of the month. There's no way they'll bother to repair it -- or even, if for some miracle, there is no core damage at all, bother to restart it.

Fun fact: The other three reactors at Chernobyl were kept running after the accident on #4. Why? They couldn't afford to lose the 3GW of power that shutting them down would have done. #2 ran until a turbine fire damaged it's power plant, and the other two were shut down at the end of the 1990s.

5) Why evacuations? Part of it, to be honest, is fear. A government think "Hmm, if I don't evacuate, and somebody gets contaminated, it will be all over the news forever." Governments also like doing something, and they can't do much at the plant proper, so an evacuation, which they can do, fits. The anti nuclear lobby has wedged it into any mind that any radiation is deadly (if so, we're all dead multiple ways after the various nuclear tests) and a government that does nothing when something bad happens is one that loses elections. See why the TSA is still around.

But there are failure modes that could result in at least short term severe releases, the biggest being a steam explosion compromising the core, and getting people away from a reactor that you don't think you have complete control over is not a bad idea. As I said before, I think the biggest problem in F1#1 is loss of sensors -- they're really not sure of the state of the reactor. If you don't know, assuming worst handleable case is a good idea.

6) Why am I not panicking? I don't automatically trust the news. It's been repeatedly filters, and when I have a Japanese Government Official Who Is Not A Nuclear Engineer talking to a Japanese report, WINANE, being translated by someone who either doesn't speak English or Japanese natively, and WINANE, then being reported by some US or UK news site by someone WINANE, then being liveblogged or tweeted by someone WINANE, and all of them get paid or renown for putting the most dramatic headlines possible on this, you should realize that all is not as it seems.

And, really, people. Worst case is a triple core compromise. Kindly go look up "atmospheric nuclear testing" for just how much radioactive material was pumped into the air. If everything goes wrong, all three reactors meltdown and escape containment it will be a mess -- but it won't even be a Chernobyl like mess, because you won't get the combination of steam explosion and graphite fire that you had there, and you have vastly better physical containment.

Double really, people. We *nuked* Hiroshima and Nagasaki. We deliberately constructed devices so that they would undergo an extended prompt critical reaction, and we dropped them on cities.

Are those cities empty wastelands, bereft of life, with nobody allowed to enter them?

And those were small weapons. Look at some of the beasts we cranked off in the Pacific, in the USSR, in Australia, in Africa, hell, in Nevada. Repeatedly.

This is a bad situation. This is not that. And not only is this not that, as we get more and more information, I can tell you this for certain.

In the end, you will be hard put to find the people who died from these reactors amongst the thousands, and now starting to look like tens-of-thousands, who died from the quake and the tsunami.

If you think the reactors are the worst thing that has happened here, you are failing badly at understanding risks.

If you want to know what scares me about Japan right now? It's the broken transport, which means more will die because they can't get supplies, and it's the statement from the JMA stating that there's a 70% of a Magnitude 7 aftershock in the next three days.
posted by eriko at 10:16 AM on March 13 [50 favorites]



once in 70-100 years is not extraordinary in Japan

Lets say every 100 years 'a black swan' happens - what's the 99 year plan to no longer be using said 100 year 'whoops, there's the black swan now you are hosed' tool?

The plant most damaged was supposed to be shut down within a couple of months - or so I've read.

And my memory on the plant model - 25 year original design life. Time will tell if some of the damage is because of the old plant beyond what it was designed for.
posted by rough ashlar 



There have been zero 9.0+ in Japan and only four greater than 9.0 ever. Bearing in mind that the scale is logarithmic and a 9.0 is a significantly different beast than the previous Japanese maximum of an 8.5. 

My mistake, I mis-remembered that there was a 9.0 in 19th century.

However, there was an 8.3 in 2006 and this design of plants is not designed to be near epicenter of such magnitude.

The tsunami is caused by the quake so a safe design has to last when both hit in quick succession - that's just common sense. 



The tsunami is caused by the quake so a safe design has to last when both hit in quick succession - that's just common sense.

Agreed, it seems obvious in retrospect but two things.

1) The newer generations of reactors don't seem to have suffered the same fate so again, a 40-year lag in lessons learned isn't something to be ignored. I could be wrong, as others have said getting solid information is difficult but it seems like a design flaw that has been long since corrected. Obviously anyone who knows more specifically and corrects or corroborates me would be appreciated.

2) I honestly have no idea what requirements or guidelines there are for protecting against tsunamis. I mean really before the Banda Aceh wave all the previous events in wikipedia are before 1908. It doesn't seem like something that had really been planned for in any industry. Obviously nuclear plants are and should be held to a higher standard but this is an event that erased entire towns, it's not something that was foreseen by anyone, even the otherwise extremely well prepared Japanese.


thebestsophist, I've also heard that NHK has been reporting that there's a 70% chance of a magnitude 7 quake in the next 3 days.


The tsunami is caused by the quake so a safe design has to last when both hit in quick succession - that's just common sense.

Not all quakes cause tsunamis. Not all tsunamis are caused by quakes (but most are.)

Now, all 9+ megathrust earthquakes do seem to cause tsunamis, but they are very rare -- so rare that it might be coincidence! An Mw 8.5 earthquake near Sumatra in the Indian ocean produced no tsunami. A 9.1 near the same spot killed or injured over a quarter million by tsunami.

To contain, indefinitely, a complete core meltdown. For that purpose, a large and thick concrete basin is cast under the pressure vessel (the second containment), which is filled with graphite, all inside the third containment. This is the so-called “core catcher”. If the core melts and the pressure vessel bursts (and eventually melts), it will catch the molten fuel and everything else. It is built in such a way that the nuclear fuel will be spread out, so it can cool down.

That would be news to me -- and this graphite wouldn't be exposed to air, like the moderator in Chernobyl #4 was. The Fuel Pool (which is the core catcher on these BWR) is stated to be reinforced concrete with a cooling system, not graphite. I'm looking for the full design documents on a GE BWR-3 with a Mark I containment.

This is a good drawing (with a couple of annotations) on the basic layout. Note the torus below. This is where vented steam goes to condense, and if there was a full core meltdown, this is where it would flow.

Remember -- there are a lot of reactor designs out there. The most common BWRs are GE BWR 1 through 6, and the ABWR. In design is the ESBWR. This is a BWR-3 with a Mark 1 containment, if it's not talking about an BWR-3 with an Mark 1 containment, it's not actually telling you anything about Fukushima 1 Reactor 1.

2 comments:

Anonymous said...

I think If a doped uranium is allowed to undergo fission in open in the same area.. It may be effective.!

Anonymous said...

You need to comment on the spent fuel.
Several core's worth of fissiles there,
no more borated water, and subject
to a 9.0 ---could be accidentally
fissile there, and zero containment.