Monday, March 14, 2011

Where's Blinky?

The ongoing emergency at Japan's Fukushima Daiichi power plant raise some interesting questions about the current state of nuclear power generation.Specifically; why is there no "go-to-hell" plan in place for these facilities?

From what I can figure out (being, as a nuclear engineer, a hell of a former paratroop sergeant...) the issue with most of these pile-meltdown incidents is cooling.

Unlike a coal or gas plant, or a hydroelectric dam, a nuke plant can't be "turned off" by sluicing the fires or opening the spillway. Even with the control rods fully inserted a typical 460 Megawatt (MW) reactor like the Fukushima plant retains about 3% of its operating heat immediately after shutdown. Three percent of 460MW is about 14MW at SCRAM, or about 14 million Joules (1 Watt = 1 Joule/second). 4200 Joules will heat 1 kilogram of water by 1 degree C. So 14 million joules would heat 1 kg of water to ~3300 degrees C in a second.

So the problem is that even after you flick the "off" switch the damn thing has to be cooled, and cooled a LOT.

But almost all of the cooling schemes rely on some sort of power; steam generators, electric pumps...and as the Sendai earthquake showed us, in power stations built near large seismogenic faults, power is almost always lost.

You'd think that there would be a "last-ditch" cooling plan that did not depend on external power, but so far as I can tell there was no such plan in place for the Fukushima plant, and that such plans are nearly unthought of. But it seems to me, a natural pessimist, unthinkable to not-think of such a contingency. What would throw a nuclear plant off-line if not a major accident, and what would be more likely - where in the case of a fire, a storm, an earthquake, an internal control element failure - than losing electrical power?

Count me as someone who doesn't reflexively fear nuclear power. Nuclear seems like something that should be considered in our technologic society as an option, with positives and negatives much like other power production schemes.

But the active-cooling requirement seems like a real potential deal-killer. It seems to me that if you can't figure out a way to flood coolant through the core - without power, without pumps - then you have a significant roadblock to keeping these things safe. And safe with nukes is a problem, too. When the floodwaters from the failed hydropower dam recede, or the fire from the gas-power plant is extinguished, people can move back in and start to rebuild. But a core-containment failure has the potential to poison the surrounding area for years, even generations.Nuclear engineers are among the smartest, best trained people in the entire engineering profession. There has got to be an answer to this cooling problem; why haven't they found it?


  1. The pebble bed reactor would seem like a possible solution, as would some of the more conventional gas cooled designs like the standard British reactors, but YMMV since I first heard about pebble reactor's in a John Ringo Sci-Fi book.

  2. One of my roomies when I was living off base while stationed at Camp Pendleton was a Georgia-Tech-trained nuclear engineer at the San Onofre Nuclear Generating Station . . . and he was full of stories.

    So, a bit of second-hand information and no physics to back it up, but I do have a question. What are your views Chief of the core actually burning through the containment vessel?

  3. Also, check out Bob Cringely's post about Toshiba's 4S (Super Safe Small and Simple) reactors:

  4. seydlitz: I honestly have no idea. The brochure at the door says no can do. But the brochure also said that the Titanic was unsinkable, the Tacoma-Narrows Bridge was designed to code, and that New Coke tasted better than the old stuff. I think the problem with assuming that containment is fail-safe is that you're by definition dealing with extreme conditions. Can it? It's not supposed to. Have we actually smoke-tested it? Of course not. I'm a soils engineer; I'll believe a design works after it's still functioning after 30 years.

    EGrise: looks promising, other than the whole "sodium-cooled" thing. If I remember my HS chemistry sodium reacts explosively with water. So assume a cracked containment vessel and a tsunami - sounds like a little ol' dirty bomb to me. These reactors also seem to have had some teething troubles - corrosion?

  5. When the bright boys start saying that the situation is already scarier than Three Mile Island and could get worse... Well that gets my attention.

    Nobody REALLY knows what's going on in those reactors right now (partly because the control room was the first thing to become irradiated) but I'm seeing a pattern.

    1. Officious officials state that things are slowly coming under control
    2. Something happens (like the roof blowing off reactor #3 or the discovery that the spent fuel rods might not be getting any coolant).
    3. Officials publicly state that the recent event is was not foreseen and could have very negative consequences but they have faith that things will work out.
    4. Go to #1

    When viewed from a distance, this increasingly sounds a bit like a considerably more polite slow-motion Chernobyl.

    Latest update:

    Re EGrise: I haven't looked at your article but the Soviets played around quite a bit with sodium cooled reactors for their navy and eventually came to the conclusion that ultra-hot (both temp and radiation) high-pressure sodium is one of the most corrosive substances you could invent and that it was a really bad idea to use for a reactor coolant.

  6. Sodium and sea water? Sheesh, yes, I can see how that could be a Bad Thing. Good catch guys!

  7. seydlitz: In re: your earlier question, it appears that the latest explosion at the Daiichi plant has damaged part of the containment system. What happens there now if the core melts I don't know but I suspect is probably worse than before the damage.

    Again, I want to emphasize that this is NOT to argue against nuclear power generation. There are arguments on both sides based on the normal operating risks and benefits - to argue about whether nuclear power is a good idea because an M9 EQ can cause problems is like arguing against air travel because a Class V tornado can rip your airliner apart.

    But my question remains; the critical flaw in reactor design appears to be the need to physically pump coolant, be it water, sodium, whatever, in and out after SCRAM. That's a design flaw, like an automobile steering system that veers sharply to the right if the steering wheel is released rather than tending to track straight ahead or an electrical appliance that has bare wires sticking out of it.

    Engineers LIKE simple, elegant solutions. The fact that nobody seems to have figured out a solution to this suggests there ISN'T a solution, which in turn suggests that most nuclear plants have a single, serious loss-of-coolant safety flaw. And I'm not sure how comfortable I feel about that...

  8. Canadian Candu reactors use un-enriched uranium (so the fuel is much bulkier) Also their moderator is heavy water which can be used as emergency coolant. (ie. the entire reactor is immersed in a huge tank of water).

  9. My wife is a bona-fide nuclear engineer who should have her PhD in nuclear engineering in about a year. It's been interesting watching her pull her hair out the last few days listening to the media reporting.

    These are pretty old reactors - between 30-40 years. The designs are probably more like 50 years old. Designs today are a lot better, and have a lot more passive safety systems that don't require generators, pumps, etc. The problem is, of course, that building a power reactor is a huge capital investment so reactors tend to stay in service longer than they probably should.

    The containment for these reactors is pretty good. So far they've survived a 9.1 quake, tsunami as well as huge hydrogen explosions.

    Reading the news tonight, it looks like units 1, 2, and 3 are finally in cold shutdown. Unit 4 is on fire, possibly from the spent fuel pools. That's potentially pretty bad.

    Anyway, going forward Japan doesn't have many power options. They don't have a lot of resources, nor a lot of extra land area, and they are an energy-intensive society. Personally, I don't see many alternatives for them besides nuclear.

    Here in the US we've always got coal.

  10. I'm not about to argue with a bona fida nuclear expert but it doesn't sound like we're out of the woods yet, Andy.

    Check out this:

  11. Looks like I misspoke. The cold shutdowns were for the OTHER Fukishima plant, not Daiichi.


    No we're not out of the woods by a long shot.

  12. One news piece I read commented that the projected final shutdown date for the Daiichi plant was...March 2011.

    Ironic and not in a good way.

    I agree with you, Andy; given the downsides of all the other forms of power generation, and the reality that we're not going back to wickiups and whale oil, nuclear has to be an option.

    But the cooling issue just seems like something that HAS to get solved or the nuclear safety issue is just no going to go away, like the bad-weather control problems that doomed lighter-than-air flight. There just has to be a foolproof way to cool a reactor - all reactors, every reactor - after SCRAM. I have to think it's possible; the fluid mechanics can't be that difficult. But I think it would mean that siting becomes more difficult, and more critical. In essence, every reactor becomes the 21st Century equivalent of a mill wheel; water has to be able to gravity-flow through the core containment plant by simply cranking open two valves.

    This doesn't seem like rocket science to me, there must be a reason it's not a standard feature of every swinging nuke plant everywhere. Andy, can your wife give us a layperson's explanation?

  13. As a general comment, there are some interesting pilot projects that are working on harnessing wave and current energy (and Japan has nothing if it doesn't have waves and currents, tsunamis excepted). Wind, too. Solar...there's a lot of renewables out there to be captured. But the current technology isn't mature enough.

    So in my unprofessional opinion nuclear has to be part of the short- and medium-term power picture for Japan. They got hammered in the Seventies by petroleum prices, fought, and lost, WW2 in large part because they were so dependent on importing energy raw materials.

    IMO the same has to be said to technic civilization as a whole; if we want to continue, in the long run we really need to figure out how to get around the "fossil" (and nuclear IS a form of "fossil" power, just using a different mineral that coal) forms of power generation. The long-term downsides (carbon release, rad waste) are just too great. But in the shorter term they are here, and practical. But I think we need to be more critical of the industries' pursuit of short term profits over long term safety. That's not a "problem", in the sense that any power company executive who suggested "Let's spend $XX billion today to solve the rad waste problem that's gonna surface 10,000 years from now!" would get sacked, and rightfully so - his company's mission is to make money today, not worry about problems 100 or 1,000 years down the road. But someone should, and, in that sense, that's the "best" argument for an entity outside the profit-and-loss sphere (like a government) doing things like this; they SHOULD be looking down the road as far as possible, and deciding whether that money would be well spent, or not, to the best they can.

  14. My Japanese language skills are sadly eroded as it has been 60 years since I was stationed at Camp Fuji near the mountain of the same name. But as I recall, Daiichi, or more better Dai-Ichi, means number one. In Japanese culture it had several meanings. As it does in American English, it could mean the best (which would be ironic) or it could mean first in chronological sequence.

  15. 50 years ago not 60 - it appears I need I need remedial math as well as foreign languages

  16. Chief,

    Here's some good stuff my wife found:

    Layman explanation of what happened.

    Explanation of decay heat.

    Modern reactor designs incorporate more passive safety features. The Wiki on that is pretty good.

  17. Andy - the links are somewhat dated. They do not appear to mention that the containment systems have been breached on at least one (or more?) of the reactors due to hydrogen explosions. The hydrogen buildup that caused the explosion was possibly caused by the chemical reaction when the zircalloy cladding degenerated and started pulling oxygen out of the H2O used for cooling.

  18. Andy: Yep, I get what happened, and it still begs the question of passive cooling. The FD plant was an older design that required active measures to cool the reactor core. When you're reduced to fire hoses, you spell that eff you see kay ee dee.

    I note that most of the passive systems mentioned in the Wiki are also noted as having one of two flaws; either they produce fissile product materials (and present a risk of both accident and proliferation) or they require fairly sophisticated handling. Better than the current BWR designs, IMO, but still with significant drawbacks. Not deal-killers, but still a concern.

    I think the real issue with true passive water-cooling is that it forces you to site the plant where you'd site a hydropower dam, and at that point you raise the question "why not go with the dam, then?", which I'm not sure that a private nuclear operator wants to answer.

    Anyway, the news continues bad from FD. One thing I want to mention; the guys in the plant who are staying at their posts to try and control this emergency. Given the dose they're getting, these guys are killing themselves - and, being nuclear engineers must KNOW they're killing themselves - to try and save their country, and their neighbors.

    That's pretty fucking balls, and I thought someone should give them a shout-out. Banzai, you guys. Death is lighter than a feather; duty, heavier than a mountain.

  19. Hey Mike,

    Thanks for giving a scenario for the hydrogen explosions. I was wondering how that H2 got seperated from the O in the water. What the heck is zircalloy? A zinc alloy? Zirconium? Does anyone else have a take on that?

    FDC Chief,
    I agree on the heroics of the guys fighting that infernal blaze.

    Personally, I am a bit afraid as I live in Taiwan and we have four reactors (and a fifth pending) on a small island on the ring of fire. The closest one to me is about 100 klicks away right on the ocean. Gulp.

    James Caba

  20. Sorry Mike!

    Just read your link so please disregard zircalloy query.


  21. Here's a handy link for updates to the status of each reactor:

    JAIF Reactor Status.


    Like almost anything else, there's no way to ensure that nuclear power is risk-free. Whether or not a nuclear plant of a specific design in a specific location is worth the risks is ultimately a judgment call.

  22. Not good:

    "They’re not sure yet why this is happening. There’s concern that the spent fuel pools may have been damaged, and that there may have been some spillage. On top of that, the pumping systems that circulate water to keep the pools from getting hot are not operating. So they’re using fire hoses to cool the rods. They contemplated dropping water from a helicopter, but the radiation level was too high for them to do this."

    And can I restate for the record that fucking Rush Limbaugh is the world's most fucking pig-ignorant, irresponsible, bone-stupid north end of a south-bound horse with irritable bowel syndrome?

    "Rush Limbaugh is calling this "The Media Meltdown."

    Tell that to the engineers committing slow suicide by radiation poisoning trying to fight these fires, you fucking fat fuckwit. They should drop YOUR enormous jello-ass into the #4 plant spent-fuel pool, douchbag. It wouldn't put the fire out but it'd rid the world of some truly toxic waste.

  23. "Like almost anything else, there's no way to ensure that nuclear power is risk-free."

    Nonsense; "risk-free" is not the issue - LOWEST risk is the issue. Commonsense suggests that a gravity-flow cooling system is as near to foolproof as possible; everything else involces a "judgement call".

    The problem I have is that much of the nuclear development industry wants the rest of the world to accept a higher risk to ensure them higher returns. With certain things; automobile design, sleeping with attractive but crazy people, there's a judgement call.

    But when the downside is potentially that large, there's no judgement call. Safety is paramount. Whatever is safest is key. The only judgement call is "what is safest".

    You wouldn't let you kids juggle live grenades, regardless of whether the safety wire was in or not, right? That wouldn't be a judgement call; live grenades = not a toy.

    Same thing here. Why make a judgement call for higher hazard when the potential for disaster - regardless of the actual risk is small - is of such a large and long-term magnitude?

  24. Chief,

    Sure, there's also a judgment call in building densely populated areas along coasts with a long history of earthquakes and tsunamis (which is, after all, a Japanese word.)

    How many people died from tsunamis in the last decade? How many people died from nuclear accidents of any kind over the last century? The current focus on nuclear safety risk seems kind of bizarre given that entire communities were wiped off the earth only a few days ago. The problems that this disaster dealt to one nuclear plant are certainly very serious, but let's keep a bit of perspective when looking at the risk posed by nuclear power especially when compared to other risks.

    Also, let's compare apples to apples. The reactors in the Fukishima plant are vintage 1960's technology. They wouldn't get build today for the same reasons that Ford doesn't still build the 1966 version of the Mustang. The state of the art has progressed considerably.

    Even so, relatively minor changes would have saved this facility. Everything was fine until the tsunami came and took out the diesel generators. Had they been located elsewhere we'd probably be praising the Japanese nuclear industry as an example of how to do things right.

  25. Big "accidents" are never caused by one thing.

    There is usually a cascading series of events, each one of which could have averted the tragedy. At some point, however, you come up snake eyes on every roll and a lot of people get hurt.

    This is why you try and build things to fail "safe". You also need to pay attention to the near misses and the "there but for the grace of god..." events.

    The problem that you can't simply turn off a reactor is a grave one. The danger that a burning nuclear reactor can make large chunks of the surrounding landscape un-inhabitable for decades makes a grave problem a critical one.

    You can rebuild right after a tsunami. 500 sq kilometers around Chernobyl is still inhabitable after 25 years. Fukushima shows us that there is a non-trivial chance that another Chernobyl type disaster could happen.

    That is a serious downside which has to be factored into any cost-benefit-risk analysis.

    (Note that I am not saying that Fukushima will be as bad as Chernobyl, but that we are getting uncomfortably close to that possibility)

  26. "The current focus on nuclear safety risk seems kind of bizarre given that entire communities were wiped off the earth only a few days ago.

    Andy; go back to your Bible. Or Marlowe and Mary Shelley. Or just screen one of the versions of "King Kong". This is entirely predictable and, instead of bizarre, would be peculiar if the focus were the other way around.

    Humans have always feared Nature and Nature's works; why do you think we worshiped the gods of the sea and the volcano, the storm and the dark for so long? We've always expected that every so often the god's would become angry and destroy us. SSDD; we just sacrifice something (a Congressman would be my suggestion), rebuild, move on.

    But the "man loses control of his false command of Nature" is the modern version of this; the updated theme of eating of the Tree of Knowledge and suffering for it.

    The nuclear industry - for understandable reasons - helps contribute to this story by its assurances that nuclear power is inherently "safe". It's not - hell, it's a controlled thermonuclear reaction, ferchrissakes - but it can be made safer, safe enough to consider as a power source. But the real "judgement call" here is usually never made, as the industry has to pretend that the danger is nonexistent in order to get its permits.

    And so when the "incident" happens - and it will; earthquake, fire, storm, human error - the monster escapes; Kong on the Empire State Building, Godzilla destroying Tokyo. The scientist's reassurances are seen as self-deception at best and lies at worst. And the entire episode takes on the shape of the modern Prometheus, the fire loose, the people terrified, and the Titan (or the nuclear engineer) helpless to contain the danger.

    Bizarre? No; the whole story falls right into the classic theme. People use themes to make sense of their world, and right now the "Mad Scientist Loses Control Of The Monster" is following right behind "The Gods Are Angry". Not bizarre; classic.

  27. Chief,

    Compared to most other sources of energy, nuclear power is comparable or better safety-wise. You're right about human behavior though - I imagine that same people that went down to the beach to see the tsunami hit the west coast are the same people madly buying up iodine pills for the pending nuclear Armageddon that's going to blow from across the Pacific.

  28. Andy-

    Reminds me of the noted meteorologist who noted that tornadoes caused more death and injury in TX and OK than anywhere else. His hypothesis, as a native Texan? Texans and Okies tend to go outside and watch rather than take cover. Since my neighbors in TX regularly went outside to watch for tornadoes when the sirens went off, I tended to agree.

  29. "Compared to most other sources of energy, nuclear power is comparable or better safety-wise."

    Yes, and no. 99.99% of the time a nuclear plant is no more hazardous to your short-term health than the coal or gas plant down the street and several orders of magnitude less dangerous than the public highway running past the generating plants.

    But that 0.01% is a real ass-tickler. When the nuke plant fails catastrophically (and every human-engineered device eventually fails, just like everything natural eventually decays or dies) it poisons the area around it for years, possibly centuries.

    So it's a hell of a big fail.

    So the question would be; is it better to take the 99.99% of the beneficial time in return for the tiny risk of massive danger from failure?

    I think that the combination of potential disaster and the long-term issue of disposal of the spent fuel makes nuclear no more than one of the transient technologies between the present fossil generation techniques and the future dominated by renewable generation methods.