Talk:Nuclear meltdown

From Wikipedia, the free encyclopedia

A nuclear meltdown is also colloquially known as the China syndrome after the 1979 film The China Syndrome, which told the unlikely story of molten reactor material burrowing through the earth from California to China.

Note that (1) the movie did not portray that happening, it merely had a character half-jokingly suggest the possibility; (2) it's not unlikely, but impossible for a number of reasons - including the earth's center of gravity not being located anyhwere near China. Mkweise 20:25 Mar 9, 2003 (UTC)

Minor clarification: The term "china syndrom" was in common use among professionals long before the movie. The movie was considered to be a Hollwood "lets push an agenda" joke by the professionals of the nuclear industry. It liberally mixed technology/terminology/theory of several different nuclear tecnologies in a manner that can be loosely described as "typical Hollywood".67.174.53.196 03:30, 24 July 2006 (UTC)

However, I understood there to be a definite danger of the mass of reactor material melting its way down to the water table - in fact, I thought this was what a meltdown was. Is this really a danger? --Andrew 06:55, Apr 26, 2004 (UTC)

Obviously, the risk to the water table depends on how deep it is in the area around the reactor, as well as how far melted reactor fuel would go. I found a reference on the Web to data from the NRC's Reactor Safety Study (WASH-1400), that indicates that maximum penetration into the earth beneath a melting reactor would be on the order of 10-50 feet, depending on what the reactor was built on top of.

Interestingly, the report makes explicit reference to the meltdown material reaching the water table, and states that the boiling of groundwater would draw more heat from the melt, reducing the depth of penetration. "Since the ground underneath containment is well below the level of the water table, conduction heat transfer at the surface of the melt should be augmented by steam generation and convection."

I can't find any place where you can download the whole report as a PDF, but you can read excepts here: [1]


Para 3 refers to a chemical explosion when molten core and water mix - I believe the usual concern is a _steam_ explosion/Vapour explosion/fci, which is a physical phenomenon (essentially due to very rapid boiling) not chemical. In WASH-1400 this was postulated to lead to rupture of the pressure-vessel and 'alpha-mode' containment failure.

The last paragraph mentions that there have been several catastropic nuclear meltdowns on US nuclear submarines, but this is false. I will assume the article is conjecturing on the reactors of the lost submarines USS Thresher and USS Scorpion. All the radiation surveys and samples have shown that there is only a very small increase in radiation levels around the reactor compartment of the lost ships. While fuel may have melted (and this is debatable), it certainly was not a catastropic meltdown. Additionally, the words catastropic meltdown are probably too strong when discussing the various Russian submarine nuclear accidents (for example, I can find no event where lives were lost when fuel melted on a Russian submarine).--Burzum 05:15, 22 Oct 2004 (UTC)

Contents

[edit] Error on Fast breeder reactors

The article currently states "Fast breeder reactors are more susceptible to meltdown than other reactor types, due to the larger quantity of fissile material and the higher neutron flux inside the reactor core, which makes it more difficult to control the reaction. In addition, the liquid sodium coolant is highly corrosive and very difficult to manage."

This is outright wrong. Firstly, sodium is highly reactive and flamabble, but not corrosive to steel, zirconium, nor the fuel forms suggested for such a reactor (metallic U-Pu-Zr alloys or some ceramic fuel forms). In particular, sodium's extremely low reactivity with steel is generally considered a large advantage in favour of water cooled reactors. The use of liquid metal rather than pressurised water also means the reactor doesn't need to be pressurised, dramatically lowering the risk of a loss of coolant accident, and also allows it to rely on natural convection for cooling giving a high degree of passive safety. Furthermore, due to the neutronic properties of sodium sodium cooled reactors can be built with highly negative temperature coefficients, causing the chain reaction to cease well bellow temperatures that would damage the reactor.

Secondly, the larger quantity of fissile material is largely irrelevant, most reactors contain enough of it to sustain their reaction for years, and typically the limit of the reactor's energy production is down to other factors. The larger enrichment level might be an issue as it means the fuel has a higher power density, but this has less of an impact than the fuel form. Metallic fuels operate at much lower temperature than ceramic ones due to the greatly increased heat conductivity, as an example. Regardless, it is the power density and enrichment level that is relevant, not teh total quantity of fissile material.

Thirdly, whereas sodium is one of the best developed coolants for fast breeder reactors, it is far from the only one possible. Lead, Lead_Bismuth alloys, Molten salts and Helium have all been suggested (Interestingly, lead is far more corrosive than sodium, yet meltdown is less likely due to the higher boiling temperature ).

Finally, the risk of a nuclear meltdown is much more a matter of design and safety regulations than it is one of reactor type. As an example, the CANDU and RBMK reactors share some features, but because the RBMK has several serious design flaws it is generally very unstable and dangerous, meanwhile the CANDU reactors are among the safest nuclear powerplants ever built. Similarly, liquid metal cooled fast reactors are probably the least likely to undergo a loss of coolant accident ( and consequentially a meltdown ), whereas it is difficult to ensure sufficient cooling under low pressure conditions for gas cooled fast breeders. 213.55.27.154 19:46, 2 March 2007 (UTC)

[edit] Chernobyl Accident

Chernobyl wasnt a nuclear meltdown, and accidnet which was essentially impossible due to its core design. (Graphite tubes seperated all the fuel elements, meaning there was insufficent radioactive slag for a chain reaction). I have edited the topic to reflect this.

Also, while there was one partial meltdown of a reactor in a russian sub (K-219), there has not been a complete one, that I have ever noted. Once again I have edited the topic to reflect this fact.

(Comments by User:210.50.61.200)

Could you please provide a reference for these claims? In particular the claim about the Chernobyl accident could really use some support. Certainly the core material did melt, and was found in the basement, so it's not clear in what sense this should be removed from the list of meltdowns. Certainly this contradicts our article on the subject:
"[T]he fourth reactor of the Chernobyl power plant [...] suffered a catastrophic nuclear meltdown that resulted in a series of explosions and fire."
and, later,
"[T]he fuel rods began to melt [...]".
If you have references to cite, that would be great, as we could fix both articles. Until then, I'll leave this article in agreement with Chernobyl accident. --Andrew 09:24, Dec 7, 2004 (UTC)
I have done much research into the subject over the years and from my readings I have understood that the chernobyl accident happened as follows: A section of the core went critical, which caused the tubes to expand and crack. The tubes were made out of zirconium, which when exposed to oxygen forms Hydrogen. This hydrogen then exploded, rubutring all containment. Surely, fuel materials did melt, but this was a byproduct of the accident, and due to the core design, finding critical mass would have been difficult.
http://www.world-nuclear.org/info/chernobyl/inf07.htm
http://www.bellona.no/imaker?id=12663&sub=1
http://www.chernobyl.co.uk/
Pretty much, its a common mistake that Chernobyl was a meltdown, but it was really one very large steam explosion
I've read all those articles, and some of the others linked from Chernobyl accident. None of them say the core didn't melt, but none of them say it did. The following sources do claim a meltdown occurred, although I have no way to judge their reliability (although the second is a PhD thesis at MIT and the third is the Encyclopedia Britannica):
The following source does not claim it was a meltdown, and do list several other meltdowns, suggesting rather strongly that they don't think there was a meltdown there:
Perhaps more convincingly, the following site, which appears authoritative, claims there was a meltdown.
Specfically, they say
"At temperatures of over 2000°C, the fuel rods melted. The graphite covering of the reactor then ignited. In the ensuing inferno, the radioactive fission products released during the core meltdown were sucked up into the atmosphere."
They also give specific references, which are unfortunately in what appears to be German.
I think I would take it as fairly well shown that the core did in fact melt. Perhaps there is some confusion with respect to definitions - the definition we use here is simply "a meltdown is an accident in which the core melts, partially or completely". This may not be quite a standard definition; perhaps it is necessary, in the usual usage, that the molten mass be critical. If so, perhaps we could find a reference for that definition and clarify this article. --Andrew 04:38, Dec 9, 2004 (UTC)
Incidentally, the following site:
describes scientists chipping a piece off an "elephant's foot" of molten and refrozen fuel in the basement.
"[...]they found the elephant's foot was made of uranium and zirconium from the reactor fuel rods and silicon from sand packed around the reactor vessel. As the reactor core burned at thousands of degrees, molten fuel had apparently eaten through the concrete floor and oozed into the warren of rooms below, where it cooled and hardened. The uranium in this "Chernobyl lava," it turned out, was too dilute to threaten a new nuclear reaction."
and
"In May 1988, they drilled through concrete walls into the reactor pit itself–and found it empty. All of the fuel, it appeared, had either been blown out in the explosion or had oozed into the lower rooms as a dilute lava."
However, it does point out that the lava was too dilute to be critical (this was one of the reasons people ran into a radioactive building to look around). So while the core melted, it seems like no liquid critical mass was formed. But has such an event ever occurred? --Andrew 04:53, Dec 9, 2004 (UTC)
Also incidentally, zirconium cannot produce hydrogen on exposure to oxygen, as it is an element. I think what you mean is that it might produce hydrogen upon exposure to steam, which is entirely plausible, and one of the suggested explanations for the second explosion. Graphite simply burns on exposure to oxygen, and this certainly occurred, so the core was definitely very hot. --Andrew 04:59, Dec 9, 2004 (UTC)
I did mean zirconium + steam = hydrogen, thanks for correcting me. While the core did melt, it was not the actual cause of the accident, and it could be argued that it didn't really add to the problems of accident. I recall reading that the primaryreason that the core did melt was that when the core was covered with lead, sand, boron and other things it caused the heat to be traped inside the core, causing it to raise to the level required for Uranium to melt. Unfortunately I can not find a source for this information. I do agree with the above statement about differnce in definitions. I believe that a 'meltdown' is when the core melting is the entire accident, primarily due to a Loss of Coolant. Therefore, I class Three Mile Island as a meltdown accidnet, while I do not believe Windscale is one, as the fuel only melted due to afire starting in the graphite moderator - despite the fact both cores did melt.
I think of it like this: If your car was stolen, you would say it was stolen. If your car was stolen, and the theif crashed the car, you would say thatt the car was stolen. Its just like this. Chernobyl was caused by a explosion and fire. In the process of the fire, the core did indeed melt...but still, Chernobyl was an explosion.
The core melted. Therefore it was a meltdown. You can't melt a nuclear pile with fire, so let's not pretend it was some kind of sub-critical meltdown. It wasn't. The core went supercritical when the rods were plunged because their tips weren't graphite as they should have been. Not that that design is at all safe in the first place. The meltdown occurred after the explosion, and it's the explosion that caused the most trouble, but it was a meltdown also. This semantic debate has no merit. The meltdown DID add to the problems of the accident; that's why the core is still there, encased in concrete. Arguably this is the only real long-term problem, as the radioactive material that was ejected by the steam rupture is long since gone. Except perhaps for some strontium. But at any rate it's spread too thin to matter. --68.253.253.62 08:07, 6 September 2005 (UTC)
No. In terms of terminology you always call an accident according to the worst effect of that excident. If a fuel tank explodes, you don't call it a fuel leak. Even though massive amounts of fuel will leak into the ground, It's still an explosion.
Same with Chernobyl. Here we had a reactor which went promt supercritical and had a reactivity excursion. This is a huge difference to only supercritical by the way. You can read why in the section about nuclear physics on the wikipedia. Since the reactor went promt supercritical parts of it nearly instantly vaporized as well as several tons of water which nearly instantly vaporized. This caused the reactor to explode. BTW, already before the reactor exploded, the reactor again became sub-critical due to doppler widening. And after the explosion, the core was most definitely VERY subcritical.
The core meltdown did not add in any meaningfull way to the severity of the accident. This wase entirely governed by the explosion. Hance Chernobyl was an explosion. Dio1982 10:08, 26 March 2007 (UTC)

[edit] Windscale

In a similar vein, I recollect that Windscale was in fact a graphite fire caused by the build-up of Wigner energy in the graphite lattice. It was not related directly to decay heating levels. It was essentially chemical, but with an 'enhanced heat of combustion' Linuxlad 09:44, 25 Mar 2005 (UTC)

In the Windscale fire, the core certainly got too hot and melted; that's what a meltdown is. But you're right that some of the heat came from Wigner energy; once the pile got hot enough, it began to burn in air as well, releasing more heat. The reactor also remained critical, releasing yet more heat. So the incident was not a loss of coolant incident; in fact the coolant was air, of which there was too much. Perhaps they couldn't turn off the air for fear of the reactor overheating due to nuclear reactions; it's hard to say. But in any case, all this is irrelevant. All a nuclear meltdown is is an incident where the core melts.
I think the reason people are uncomfortable with this definition is that we have an idea that a nuclear meltdown is somehow the worst possible kind of disaster that can happen in a nuclear power plant, with severe long-term results. That's just not the case. What happened at Chernobyl and Windscale was worse than a simple meltdown; what happened at Goania was also worse (much worse than Windscale or Three Mile Island) but it had nothing at all to due with nuclear chain reactions, let alone a meltdown. A nuclear meltdown is a specific kind of accident, with the possibility to be very bad (by melting down to the water table and causing a steam explosion, say). There's a whole spectrum of other possible accidents that have the possibility to be very bad also. Some of them have a meltdown as a probable side effect. --Andrew 19:52, Mar 25, 2005 (UTC)


I'm not too sure you've got all your detail right there (but then the event happened when I was ten, I recollect). But I don't think the pile was critical or that nuclear heating played a significant role. The Wigner energy built up to very high levels because the graphite was normally too cool for the graphite defects to anneal out. The article here is written from the aspect of loss of adequate cooling for the nuclear-decay heat level - this wasn't really that - it was enhanced chemistry.

Of course, the melting is pretty irrelevant and a result of the article having started out with a PWR bias.

I am by no means an expert, but I have done some reading on this particular issue; perhaps we should discuss it at Talk:Windscale fire? Anything we decide there can of course be used to improve this article.
I agree that the melting hardly mattered in the Windscale incident. In general, a meltdown may be only an insignificant part of a real accident, especially if th e reactor is designed to contain it (as the Chernobyl reactor was). The other things that go wrong about the same time are often much worse. But for some reason, perhaps by the China syndrome and the publicity around Three Mile Island, "nuclear meltdown" has a magical weight in people's minds that is not matched by "steam explosion" or "fire", even though those have been the real problems in the Big Accidents. --Andrew 22:17, Mar 25, 2005 (UTC)

[edit] numerous clarifications and corrections

I was in grad school when TMI-2 occurred and on the job when Chernobyl-4 burned. I'm expert on BWRs and PWRs but not the older British reactors. I've tried to clarify the difference between reactor vessel and containment structure. Also, TMI was far more of a meltdown than was originally thought - once they removed the melted fuel, they found that the melt had cracked the RV's stainless steel liner but not attacked the base metal (this was unexpected and, so far, unexplained, but meltdowns don't penetrate RVs - Chernobyl's concrete RV failed due to the explosion, not the melt). The RBMKs have since been modified and de-rated (and, hopefully very soon, closed). Simesa 01:40, 26 Jun 2005 (UTC)

[edit] WASH-1400

There are still a few LWR-isms in it from a UK perspective, but probably near enough for most readers... Note again that 'meltdown to China' is NOT the worst accident, radiologically. 'Alpha-mode' (fci-induced) or HPME-induced containment failures would be much worse.

NB the melt just wandering down to the water table would be unlikely to be that dangerous. A fairly coherent in-core melt release together with efficient generation of a missile (usually from the top head) is necessary to fail a modern 'large,dry' containment (IIRC).

Later - I have added a stub for WASH-1400, Rasmussen's original 'Reactor Safety Study' for USNRC.

[edit] Technical

I added the technical template because the article (especially the Causes section) seems to assume a high degree of familiarity with the material. Perhaps the context template would be more appropriate. Either way, it calls attention to a problem. I am fairly familiar with the basics, but I find myself stopping after every sentence and trying to backtrack and see how it fits in with what's already been said. Example: "Borated water is injected by the emergency systems and thus in the large-break accidents, control rod insertion is not needed to stop the fission reaction." What is "control rod insertion"? It is discussed in the Nuclear reactor article, but if this article is about the failure of the system, it could at least overview the measures that have to fail. Even if it's ABCs to nuclear reactor buffs, there ought to be some explanation for the unwashed masses. On the other hand, I can be lazy in my research, but I don't think excessive clarity can ever be a drawback. icydid 02:05, August 23, 2005 (UTC)

[edit] Burzum's Changes

I've significantly updated the leading paragraphs and the causes section. Additionally I am going to have to create a couple of new articles on reactor containment (which is a superset of the containment structure article), loss of pressure control accident, uncontrolled power excursion or reactivity addition accident, and reactivity_(nuclear). Being a person who previously worked as a reactor operator, I may not always realize when certain parts are too technical, so I apologize beforehand. If any of the correct articles exist (or if you want to create them before I do) feel free to do so.--Burzum 07:10, 9 July 2006 (UTC)

[edit] Header

I have removed some information that was added to the header about TMI and 'fizzles.' The term fizzle is used for the failure of a nuclear bomb, not any condition of a nuclear reactor. The correct term for an extreme uncontrolled reactivity excursion incident is prompt critical (which is discussed further in the article). Additionally, the information that TMI was only decommissioned due to a coolant leak and that the reactor vessel did not suffer damage is almost certainly false. There are several publicly available images (including this one) that show extensive damage to the core. The claim that the reactor vessel did not fail does not mean that it did not suffer extensive damage and could be reused (such as temperature gradient stresses and pressure stresses). With the fuel assemblies looking like a pile of rubble there was never any doubt that the reactor vessel interior supports for the fuel assemblies were trashed and that irreparable damage to the vessel occurred. Cheers.--Burzum 22:25, 29 November 2006 (UTC)

After the corium was removed, it was found that the stainless steel liner of the reactor vessel had cracked but that the base metal was undeformed (although this was unexpected). Large portions of TMI are extensively and irremediably contaminated due to infiltration into the concrete (which is why some or all plants now seal their concrete, as I saw at a plant I worked at). I didn't add the part about fizzles, just tried to work with it. I'll settle for saying that the TMI RV didn't fail. Simesa 00:31, 30 November 2006 (UTC)
I don't think that the statement about TMI's RV is applicable because the introductory sentence says "nuclear meltdowns are typically characterized by severe damage to the nuclear reactor in which it occurs." TMI certainly isn't an exception since the fuel elements were destroyed, the inner RV fuel element support skirt was destroyed, heat stress damage to the RV occurred, the primary coolant system was contaminated beyond any reasonable design factor, the interior of the containment structure was severely contaminated, the U-tubes in the steam generators were cracked, etc. To say that the reactor did not suffer severe damage just because the reactor vessel didn't crack in half is false. If this doesn't convince you, then realize that the standard nuclear power plant definition of core damage is damage to the core that reduces the full power rating of a reactor or releases fission products to the environment. TMI was reduced to 0% power and suffered extreme core damage. This discussion about the RV doesn't belong in the header (which is an introduction to nuclear meltdowns). Cheers.--Burzum 11:55, 30 November 2006 (UTC)
All of the above is true. However, I'm thinking from a layman reader's point of view, and for many US readers (particularly those about to live near new plants) the first thing that will pop into their heads is "Did this happen at TMI?" (I'll also agree that it was expected that the corium melt through - last I read, why it didn't was still unknown. TMI certainly had no LVHDC.)
Comments from others?
Simesa 16:56, 30 November 2006 (UTC)
I am eventually going to need to revert back so that it is obvious that a nuclear meltdown causes significant damage to the nuclear reactor, not just the reactor vessel. I still hold that your TMI addition is extraneous and should be removed, but I would like to get some consensus here to avert an edit war. I also still hold that the TMI reactor vessel was severely damaged and unusable. As a former reactor operator I can see no way that the significant rearrangement in the core would not have affected brittle fracture pressure limit (BFPL) curves, heatup or cooldown rate limits, or maximum pressure limits. The metal of any reactor vessel is carefully heat treated for a desired strength. Melting fuel elements might not crack it but they will certainly affect this heat treatment. Cheers.--Burzum 21:14, 30 November 2006 (UTC)


I'd still like a statement in the introduction saying that a complete nuclear reactor malfunction not equivalent to a Hiroshima, but to a Chernobyl. That's not clear to people coming to this page, it wasn't clear to me until a discussion with Cadmium. When people think nuclear they think "mushroom cloud" in my back yard, and I used to think that too, and it's not clear from this article at first glance. Sillybilly 01:06, 1 December 2006 (UTC)
I have adjusted the header due to your concerns. Cheers.--Burzum 03:23, 1 December 2006 (UTC)

[edit] Nuclear accident versus Reactor accident

I see we've now agreed to go with "reactor accident" after all, so there's no need to belabor the point, but it'd be good for Wikipedia to explain the reasons behind this.

The term "nuclear accident", as can be seen by the somewhat-relevant-but-largely-irrelevant nuclear accident article, is a poorly constructed term with regard to semantics/meaning. Clearly, everything made up of matter (at our level of abstraction) has a nucleus...so the term is immediately suspect and non-descript, with vague implications of high radiation issues. It's basically non-scientific media shorthand for "somethin' nuke-u-lar" that's gone wrong. You won't find it being used (intelligently) in the nuclear profession...it's just too meaningless. The most appropriate shorthand, in the case of nuclear reactors, is the term "reactor accident," with the phrase "nuclear reactor accident" being most complete. "Reactor accident" is also the phrase used by the U.S. Navy's nuclear propulsion program.

As an aside, there's no real way of solving the semantic problems with the term "nuclear accident." Even a cursory review of the "nuclear accident" article will bear this out. It's a term that is best parsed by either expanding it to include other descriptors (e.g., "nuclear weapons accident") or simply eliminating the nearly useless term "nuclear" altogether (e.g., "reactor accident").

If it's of any assistance information-wise, I've served as the Engineer Officer (Chief Engineer) of a nuclear powered attack submarine during Admiral Rickover's era. That's not a badge I wear these days – nor does it matter with respect to the objective or subjective quality of my edits – but it may help with regard to communications here. --24.28.6.209 02:06, 13 November 2007 (UTC)

I'll have to disagree with you on cutting out of the link to nuclear accident. While it is true that industry and the military use more specialized terms, many watchdog groups still use the term 'nuclear accident' and I would argue that this term is fairly well known among the public. In fact if we are discussing strict naming we have to note that the term 'nuclear meltdown,' which is the title of this article, is used nowhere in industry or the military. But since Wikipedia is an encyclopedia for the general public we should have some discussion of these sometimes ambiguous names, like nuclear meltdown and nuclear accident. Cheers.--Burzum 03:54, 13 November 2007 (UTC)
It's easy to confuse something that is "often used" with "well known." There is no knowing what "nuclear accident" means, as it is a poorly constructed term. Using it often doesn't make it any more clear with frequency of usage. Again, we need look no further than Wikipedia's own nuclear accident article, which is something of an accident itself, as it is a hodge-podge of unrelated and ill-formed ideas.
You're quite right about "nuclear meltdown" itself being a media term, rather than being one that's well informed (or informing). But the fact that "watchdog" groups use these terms doesn't make them any more literate.
My position is that we need to deal with these poor language constructs without propogating them any more than is absolutely necessary. The ultimate goal is to inform. The good news is that this is not very difficult once we examine how well we're actually communicating. --24.28.6.209 10:57, 13 November 2007 (UTC)

[edit] Meltdown "theories"

For the benefit of User:Power2084 who refuses to participate in a discussion:

A molten core mass may well melt or burn anything below it. However, as it does so, the mass becomes diluted and eventually non-critical. For the mass to melt down much farther than the containment structure, it would need to remain in a critical state, and to do so, it needs to remain in a compact form, kinda like it is in the reactor but without anything to hold it together in such a form. This is exceedingly unlikely; I'm not a scientist but I'm pretty sure the Second law of thermodynamics backs me up here. Saying this could happen "in theory" grossly overstates the case. We could just as well say that the sun, in theory, might not rise tomorrow.

The "China Syndrome" is a proverbial term; no one seriously believes a meltdown will burn through all the way to China, or even through a significant portion of the Earth's crust. A number of editors have called you on this, and the burden is now on you to find a legitimate scientific source for your claims. Simishag (talk) 01:20, 6 March 2008 (UTC)

We should note that at Three Mile Island, the core did not breach the pressure vessel, and at Chernobyl the core only made it to the basement. It is not credible that the core could melt through several miles of the earth's crust. Paul Studier (talk) 02:53, 6 March 2008 (UTC)