Jeffrey LewisTechnical Difficulties at Bushehr?

Middle East News Line (via GSN) cites Russian officials describing technical difficulties with Bushehr arising from changes to design specifications:

Russia’s effort to complete Iran’s first nuclear plant continues to be hampered by technical difficulties.

Russian officials said that despite significant progress over the last year, the Bushehr nuclear reactor plant has encountered technical difficulties. They said the key difficulty has been adapting the construction of the nuclear reactor core to Iranian specifications.

“The Iranians have changed their specifications several times and this has been a key challenge in completing the reactor on schedule,” an official said.

Arms Control Wonk reported on an unclassified Livermore briefing that last year that suggests Bushehr is not suitable for a nuclear weapons program.

One recipient of the briefing said that the design history of the reactor will complicate reprocessing of spent fuel to recover plutonium:

… the reactor, in large part because of its mixed design heritage is a piece of garbage. It is, of course, true that Pu will be produced as is the nature of fissioning uranium. That said, the reactor chemistry is going to produce really, really nasty fission products. It would not be overwhelmingly difficult to separate Pu from the spent fuel coming out of Bushehr but the bother to worth ratio is extremely high – you are going to get a lot of ‘bad’ Pu for very little “good”.

I identified the briefing, but it is FOUO.

The description of the briefing, emphasizing the German origins of the reactor, and the MENL story, emphasizing later changes by the Iranians, are slightly different explanations for the technical problems—but they may be related.

Also, Russian Minister of Atomic Energy Alexander Rumyantsev told reporters that the first shipment of fuel for Bushehr would occur “either at the end of the year 2005 or early 2006.” I have the full text of the interview, but its not very hot.


  1. Pavel Podvig (History)

    Do you have any details on why would the reactor “is going to produce really, really nasty fission products”? As I understand, it is a VVER-type (or very similar) reactor with standard uranium oxide fuel. Where would “nasty products” come from? Does the briefing say anything specific?

  2. Robert

    The unclassified Livermore briefing, you write, argues that Bushehr is not suitable for a nuclear weapons program. I’m trying to understand what exactly they mean.

    (1) Even if Bushehr yields spent fuel with high concentrations of plutonium-240 or plutonium-242, wouldn’t the “diluted” plutonium-239, after being separated from spent fuel through reprocessing, still be usable for a “first” less-than-optimal nuclear explosive?

    As the presence of undesired plutonium isotopes grows, the probability distribution of yields will change—with lower yields more likely, a larger failure-rate, and a lower expected/average yield.

    (2) But, still, wouldn’t a “diluted”-plutonium explosive be nevertheless “nuclear,” with an explosion that would give off at least three orders of magnitude more energy per pound than a high explosive?

    (3) The destructive-effect reduction of the “diluted” plutonium explosive would be smaller than the yield-reduction, right? (Area destroyed by blast overpressure decreases as the two-thirds power of the yield-reduction.) Wouldn’t radiation-reduction be even smaller than yield-reduction, right?

    The upshot is that, although “diluted” plutonium wouldn’t lead to an “optimal” nuclear explosive, it could still be used to make a “less-than-optimal” first nuclear explosive.

    Am I missing something here?

    BTW, I enjoy reading your weblog. My day isn’t right without several visits.

  3. Cheryl Rofer (History)

    All fission products are really, really nasty. I don’t know why VVER or any other reactor would be worse.

  4. Michael Roston (History)

    First, to whomever Robert is, if you’ve got tips on how to make a diluted plutonium weapon that works reliably, you should probably tender a contract at LANL and LLNL – sounds like you’re about to get real paid by solving all of our stockpile reliability problems.

    Second, in response to Pavel and Cheryl – the briefing I attended indicated that separating plutonium from Bushehr fuel would be so difficult and time-consuming that it would easily be detected. Additionally, as Jeffrey indicated was in this briefing, the fruit-to-chaff ratio would be so high in favor of the chaff that it just wouldn’t be worth it.

    Iran knows this fact, which is why they are pursuing uranium enrichment and heavy water production capabilities to substitute for what Bushehr can never be. Additionally, Bushehr then stands out as what it always was supposed to be – an excuse for conducting broader nuclear energy research in the country, which in turn gives it the Japan-like virtunal nuclear capability that it wants.

    Curiously, I wonder if these reports explain a lot of the dispute between Russia and Iran over who has to pay for reprocessing the fuel. The Iranians want the Russians to pay for it, and vice versa. Maybe the Russians don’t want to pay for it because it’s so difficult.

  5. John Field (History)

    I suppose that the problem caused by the many redesigns is that the thermal neutron spectrum now contains a large tail of largely cooled but not yet completely thermal neutrons which is then rapidly absorbed in resonance capture in produced Pu-239 to make Pu-240.Seems like this would lead to all sorts of other isotopes of plutonium eventually also.

  6. Pavel Podvig (History)

    No, the “fruit-to-chaff ratio” doesn’t seem to explain anything. I just don’t see what would would be the physics that would make reprocessing of Bushehr fuel any different from reprocessing of ordinary VVER fuel (which is quite difficult and time-consuming to begin with).

  7. AHM (History)

    The whole reactor-grade versus weapons-grade debate was resolved a while ago. reactor-grade materials can still be used to make a weapon; it’s suboptimal for various reasons, but feasible. See J. Carson Mark’s article in Science and Global Security, vol. 4, no.1 (1993), available on the S&GS website. I don’t see how the VVER Iran has would produce any worse plutonium than any other VVER.

  8. Jeffrey Lewis (History)

    I don’t think the Livermore briefing was related to the theoretical suitability of VVER reactors, but rather something arising from the design history of the reactor.

    John Field has suggested a possible explanation, which I’d like to see addressed directly.

  9. Pavel Podvig

    Even if we assume that the neutron spectrum in the reactor does not look like a regular VVER spectrum, the worst this can do is to slightly change the isotope composition of Pu or of fission products. I would seriously doubt that either of these would be significant enough to affect the chemistry of reprocessing, which the briefing seems to imply.

  10. mark gubrud (History)

    In every reactor, the neutrons come out hot and bounce around initially through the surrounding nuclear fuel before being thermalized largely in the moderator. So the fuel is always exposed to a large flux of hot and not-yet-”fully thermalized” neutrons. Being no expert on reactor design or the n capture spectra of Pu239, I can’t entirely discount John Field’s theory that the redesigns might lead to less than optimally rapid thermalization and excess production of 240 etc.. If this theory is correct, it suggests that the Bushehr plutonium would be worse than standard “reactor grade” plutonium, but probably still not something you would want to shrug off; at the very least it would make one hell of a dirty bomb, or a type of dirty neutron bomb.

    I don’t think Field’s theory would be related to the comment from the “recipient of the briefing”.

    Pu is not a fission product. The fission products are a spectrum of mid-weight isotopes that result from randomly splitting U235. They are going to be the same in any reactor, and exposure to an excess of above-thermal neutrons can hardly make them any nastier. They can be separated out chemically from the Pu-239; what can’t be separated chemically are the unwanted Pu isotopes.

    Iran does not want to use Bushehr to produce Pu for weapons. They can do that with the heavy water/natural uranium reactor they are building at Arak. But once Bushehr is operating, an air campaign that targets Arak and Natanz while sparing Bushehr to avoid “another Chernobyl” would leave Iran with a potential desperation source of plutonium that ought to have some deterrence value.

  11. John Field (History)

    Clearly, this quote is confusing. But, I still believe that the problem is related to unwanted Pu isotopes and not to other chemical elements.

    Note that the thermal absorption cross sections of Pu239 and Pu240 are similar, but the resonance integral(capture of suprathermal neutrons) of Pu240 is 40x larger.

    The implication is clear. Once Pu240 is in the fuel rods, suprathermal electrons can rapidly breed Pu241. A well thermalized reactor won’t do this nearly so much.

    The Pu241 lives only 14 years – inconveniently decaying in your warheads and then a bunch of things can happen involving Pu as well as Americium, but they are all way more of a problem than even Pu240.

    I believe that this is what they are talking about.

  12. Pavel Podvig (History)

    It’s an interesting hypothesis. Pu-241 (or, rather, Americium) is indeed a major headache. But as I understand there is enough Pu-241 in regular reactor-grade Pu to begin with. Any changes in neutron spectrum are unlikely to make noticeable difference. It would be interesting to do some calculations.

  13. John Field (History)

    OK, ok…over my morning coffee, I pasted together a program to demonstrate the effect.

    It is oversimple, but I think it makes the point. The cross sections come out of GE Nuclear’s Nuclides and Isotopes, 14th ed. I bred Pu239 until the fraction of Pu240+Pu241 reached 7% and varied the non-thermal neutron percentage.

    At 10% nonthermal neutrons, 1% of the product will be Pu241, which should be more than enough to cause trouble. You can peruse the table of results at, and the program is at the same address under /pufracs.cpp. It would be easy to modify as to introduce any other rates/reactions you like.

    It occurs to me that this hot electron limitation will NOT apply to the production of TRITIUM in a reactor of this sort.

    This may be a naive question, but it has always seemed to me that a nation interested in developing thermonuclear bombs would have a much easier time of it by using U235.

    The reasoning goes as follows : The hydrodynamics are so much easier that you can get to way, way over critical. This guarantees high efficiency. Which in turn guarantees that the boost will go off – you might even get away without a boost. We can be confident that it will get mighty toasty inside the holraum.

    Now, your high pressure equation of state data doesn’t need to be nearly as accurate for the design of the second squish to work properly. Right?

  14. Allen Thomson (History)

    > It occurs to me that this hot electron limitation will NOT apply to the production of TRITIUM in a reactor of this sort.

    Tritium production is something that one should keep in mind when contemplating potential proliferators. 3H is to nuclear weapons design as MSG is to Chinese restaurant cooking: it really brings out the flavor.

    Among other things, DT boosting takes a lot of the worry out of the unwanted Pu isotopes problems being discussed here. Note that India is alleged to have developed a process for obtaining 3H from heavy-water reactors; one might want to look for signs that Iran is interested in the process.

  15. mark gubrud (History)

    Again, I’m no expert, either, but I don’t think DT boosting takes the worry out of the unwanted Pu isotopes problem, which is that the high background of spontaneous neutrons from the 240+ isotopes leads to predetonation before full compression, resulting in a fizzle and no boost. Also, I don’t know why people keep referring to “hot electrons” when they mean warm neutrons. And I really don’t think we have a good explanation of why Bushehr redesigns would have led to too many warm neutrons. How much can the reactor core itself have been redesigned? More likely we are talking about the surrounding equipment, but even if the reactor core is redesigned, I’d like to hear the basis for supposing this may lead to excess warm neutrons. Also the main determinant of the Pu240+ content is the length of time the fuel has been in the reactor; a high-quality Pu production reactor is one in which the fuel is frequently reprocessed. If the Bushehr fuel has been used for commercial power production for many years and is therefore contaminated with a high level of 240 & 241, it can be reprocessed to clean it out and then start over in production mode (short cycles), assuming the Iranians develop a fuel reprocessing and refab capability to do this.

  16. Yale Simkin (History)

    ”..I don’t think DT boosting takes the worry out of the unwanted Pu isotopes problem, which is that the high background of spontaneous neutrons from the 240+ isotopes leads to predetonation before full compression, resulting in a fizzle and no boost.”

    Conditions allowing fusion boosting will occur when an implosion fission yield exceeds roughly 1/4 kiloton.

    Almost all US atomic weapons outside of sub and land-based ICBMs are variants of the Mk.61 warhead. This is a “Dial-A-Yield” design that can be set (depending on mod) to:
    0.3, 1.5, 10, 45, 60, 80, 170, and 340 kilotons.
    Notice the 1/3 kiloton minimum yield. This results if DT gas is not injected into the pit.

    In a fission bomb, the worst “predetonation” will occur when chain-reaction starts at the exact moment 1 critical mass forms, rather than the optimal assembly of multiple crit masses at maximum compression.

    This earliest possible predetonation is the same REGARDLESS of Pu isotope mix – its the same whether its super-grade of reactor-grade plutonium. This worst-case pre-det sets the lowest possible yield that a particular bomb design will produce.

    This is the lowest so-called “fizzle” yield. The minimum fizzle yield of the original 60 year-old design, the first crude a-bomb, the Trinity/Fat Man Gadget, was ~2/3 kiloton.

    Thus we see that even with the worst possible predetonation (which is independent of isotope blend) is more than enough to trigger fusion-boosting and therefore result in full-yield explosion.

    ANY reactor produced plutonium is capable of fueling completely reliable and nominal yield bombs.

    Even without boosting, a worst-case pretonation of of crude Trinity design is a fearsome weapon.

    2/3rd of a kiloton would kill almost everyone in a circle a mile across.