James ActonMore on the NORK declaration

I’m back from Wye now. The (fairly awful) photo is of Jeffrey, Wonk alum Jane Vaynman and me. It turns out that Jane and I have the same all-time favourite Foreign Affairs article: the one on monkey sex.

In my jet-lagged state I have been trying to catch up on North Korea news. According to Glenn Kessler in the the Post the declaration contained no surprises. Apparently it states the DPRK has extracted 37 kg of plutonium and contains “a brief reference to the U.S.-North Korean side deal” under which North Korea privately acknowledged US concerns about the alleged enrichment programme and proliferation to Syria just before submitting the declaration.

However, Monday’s Global Security Newswire has the following to say:

Pyongyang reportedly acknowledged in the declaration that it had collected roughly 30 kilograms of weapon-usable plutonium and used 2 kilograms in its October 2006 nuclear test blast, Kyodo News reported.

One nuclear weapon generally contains 4 to 8 kilograms of the material.

North Korea’s total plutonium holdings, including unextracted material in used fuel rods and material that remains in equipment at Yongbyon, is believed to total 44 kilograms, the source said (Kyodo News I, June 28).

Irritatingly the Kyodo story does not seem to be online (Ed. I put the full text in the comments) and I’m assuming Kessler has it right (not least because his article is more recent). However, if the Kyodo story is correct, it is pretty sensational. Not only would “roughly 30 kg” of extracted Pu be almost (but not quite) implausibly little, but a weapon containing 2 kg of the stuff would be very ambitious for a first shot.

I wonder if the declaration will be made public…

Anyone want to leak it to the Wonk?


  1. Glenn Kessler (History)

    The figure of 44 kilos unextracted is not inconsistent with what I have heard. For the newspaper, it was a bit complicated to explain that the North Koreans might have left a bunch of stuff in the pipes, so I focused the material that they could fashion into bombs—ie, the declaration that 37 kilos were extracted. But there is a second, higher figure, said to be in the mid-40s, for total plutonium. I have not heard they they claimed they only used 2 kilos for the test.

  2. Jeffrey Lewis (History)

    Kyodo: LEAD: N. Korea Says Used 2 Kg of Plutonium in 2006 Nuke Test: Source

    JPP20080628969032 Tokyo Kyodo World Service in English 1020 GMT 28 Jun 08

    [Computer selected and disseminated without OSC editorial intervention]

    Seoul, June 28 Kyodo — (EDS: ADDING DETAILS) North Korea’s declaration of its nuclear programs that it submitted this week states that it extracted a total of around 30 kilograms of plutonium and used 2 kg in the nuclear test it conducted in October 2006, a six-party talks source said Saturday.

    A nuclear weapon normally requires between 4 and 8 kg of plutonium. There is speculation that North Korea may have declared a smaller quantity to counter rumors that the 2006 nuclear test, which resulted in a relatively small explosion, was a failure by lending weight to its claims aboard of having the ability to develop small-scale nuclear weapons.

    According to the source, in addition to the extracted plutonium, there also appears to have been approximately 8 kg of unextracted plutonium that was generated in used fuel rods. Moreover, if plutonium that was left over in equipment of its nuclear facilities is included, the total amount of extracted plutonium rises to 44 kg.

    On Thursday, North Korea handed over the nuclear list to China, which hosts the six-way talks also involving South Korea, the United States, Japan and Russia, half a year after the original deadline under a six-nation agreement.

    In return, the United States as promised started the process of removing North Korea from the terror blacklist and lifted sanctions against it under the Trading with Enemy Act.

    North Korea is disabling its nuclear facilities in Yongbyon, about 90 kilometers north of its capital, under the six-way deal in return for energy aid and diplomatic benefits.

    Pyongyang on Friday demolished a cooling tower attached to its key Yongbyon reactor, in a symbolic demonstration underscoring progress in the disablement of the site.

    [Description of Source: Tokyo Kyodo World Service in English — English service of Japan’s largest domestic and international news agency, owned by nonprofit cooperative of 63 newspaper companies and NHK]

  3. Brad Lohaus

    Wonk — you seem concerned about 30 v. 37kg. What’s the big deal if you can get a bomb for 4-8kg. That’s a difference of one bomb.

  4. John Field (History)

    Two kilos.

    I suggest that means they got a compression of about 1.75 times alpha phase standard density. More if alloyed to delta phase. The shock data for Uranium says that will occur at an assembly speed of about 2800 m/s, probably a little more for alloyed plutonium. A uranium tamper would be easier than beryllium and seems likely if only to get as much as they can out of the little Pu they had.

    OK, so here is the big question:

    They told the Chinese they were shooting for 4 kton yield but they didn’t get near it. If so, they wanted assembly speeds up in the 4,000 m/s+ range. That’s fast.

    So, you choose :

    a) the bomb was big and heavy
    b) the bomb was sophisticated
    c) there is some kind of liar’s poker going on here

    Maybe we at the Wonks were too quick to dismiss their efforts?

  5. Stephen Young (History)

    I’m just assuming that Brad is joking here. When one bomb = one city, that makes it a big deal, especially when the owner of that one bomb has already sold missiles and, potentially, a plutonium-generating reactor to other regimes.

    And, of course, it also reminds us of the nonsensicalness of the thousands of warheads that remain in the arsenals of countries far more “sensible” than the Norks.

  6. Jeffrey Lewis (History)

    As I currently understand it, the declaration is something like 8 in the spent fuel, 6 in process loss and the rest as separated Pu.

    That’s plausible, which is not to say I believe it absent additional verification.

    Yes, I believe a missing bomb is a big deal.

  7. R. Kelly (History)

    “A nuclear weapon normally requires between 4 and 8 kg of plutonium.”

    I would have thought 1-4kg depending upon your objective. But, then again, who would beleive anything you read on the internet 🙂

  8. Brad Lohaus

    What matters is that NK has a bomb. The difference between 0 and 1 is a lot bigger than the difference between 1 and 2 if you take all international politics stuff into consideration, no?

    Now if what you are saying is that NK lied on their declaration and they are holding back so they can keep an “extra” bomb even if they give the others up at the end of the day, that’s a really big deal.

    But in the grand scheme of issues nuclear, presuming this is not deliberate cheating, how high on the list is it? Higher than NK having a bomb in the first place?

  9. Brad Lohaus

    Stephen — let me rephrase. From a US foreign policy perspective, if the US thinks NK has 5 bombs v. 4 bombs, how does that influence US foreign policy? Your argument would suggest that knowing NK has 5 bombs v. 4 bombs would significantly influence US policy.

  10. MTC (History)

    The Yomiuri Shimbun of July 1 (translated on July 3) spells out the fates of the 44 kilograms.


  11. Lao Tao Ren (History)


    I vote (b) the bomb was sophisticated.

    Very typical of Korean nature to try to leapfrog, and to be honest, (if (b) was the case), they got close. Damn close.

  12. Yale Simkin


    The whole point is to DISARM North Korea.

    That is why the quantities are of absolute importance.

    It does not matter very much if they have 5 or 4, it is whether they assuredly have ZERO rather than maybe one.

    There is infinite difference between 1 a-bomb and zero a-bombs.

  13. michael

    Moving on, am I really the first to note Jane’s new(?) haircut? Like it. Like it a lot.

  14. John Field (History)

    @Lao Tao Ren

    Yeah, I am suspicious it is (b) also. It paints a consistent picture, I think.

    If you wanted to do a diagnostic and you didn’t have hundreds of millions of dollars, what you are most likely interested in is total average compression. Presumably, you would choose to maximize the derivative of log(yield) with respect to compression evaluated around the pressure range where you expect to hit. This minimizes uncertainty in your post-shot estimation. Two kilos is going to be just about that point.

    Also, it uses less plutonium, halves the neutron background, greatly increases the premature buildup time, and might even piss off the rest of the world less. Presumably too, it also sends a backdoor signal to the intelligence people that the NORKs might know what they’re doing and not fooling around.

    I wish we knew how much the bomb weighed…

  15. Lao Tao Ren (History)


    If you assume that it is intended to sit atop a Nodong or a Taepodong-2, then it should be somewhere between 650 to 500kg assuming you are pushing for max. range rather than payload.

    There are other ways to get at the estimate, which is to look at the potential customer set and what are the limitations of their launchers.

    I did that exercise and came up with about the same mass.

    So say, 600kg +50/-100 as the range?

  16. Yale Simkin (History)

    I am mystified by that 2kg plutonium claim.

    That is very low.

    Making some assumptions about relector/tamper and a reasonable 2x compression, that would create only about 1 and a quarter critical masses. That means that if your implosion was even minimally out-of-spec, you would yield zilch.
    Why do that?
    If it was absolutely perfect, then you would get the apparent subkiloton yield ascribed to the NK test, but I feel that the test fizzled, not that it was designed for such low performance.

    The US weapons use low fissile mass and minimal explosives because they are boosted. The unboosted yield of US warheads is typically about 1/3 kiloton – sufficient to light the fusion boosting fuel and go multikiloton. That’s a major reason that they are pre-detonation-proof as they essentially pre-detonatate by design.

    Would a virgin builder try for such a design?

    I would assume that one’s first test would be quite conservative. With a near critical mass of plutonium, one would get a satisfying bang even if maximal, perfectly-timed, totally symetrical implosion was not achieved.


  17. Andrew Foland (History)

    The difference between the DPRK having 4 weapons or 5 may not be that big, but the difference between DPRK having 4 weapons, and DPRK having 4 while selling a fifth to Country X, is quite significant.

    Also, I believe 2 kg is WELL below what’s generally considered a “significant quantity” of Pu. Which will doubtless make somebody’s job much, much harder.

  18. Dave (History)

    I suspect that they did a series of sub-critical tests using progressively larger pits, using a similar set of implosion systems.

    If it was my program, I would have done a shot aimed a criticality of 0.9 or 0.95, which is enough that you can observe neutron multiplication with your instruments.

    Then, I would have done one with a criticality of 0.999, and the final shot would have had a design criticality of perhaps 1.001. This series of tests would provide much more infomation than a Trinity-style “let’s see how big this thing is” kind of display.

    The nice side effect of a series of tests like this is that only the last (biggest) one can be detected at any reasonable distance. My suspicion is that the NORKs conducted a very sophisticated test.

  19. Alex W. (History)

    One wonders what happened to the test directors for the NK “dud” test. One of the reasons you might want to be conservative in that sort of state is that the consequences for failure might be pretty high.

    David Holloway writes of how the USSR decided what to do in the event of their successful test in 1949:

    In deciding on who was to receive which award, Beria is said to have adopted a simple principle: those who were to be shot in case of failure were now to become Heroes of Socialist Labor; those who would have received maximum prison sentences were to be given the Order of Lenin, and so on down the list. This story may well be apocryphal, but it nevertheless conveys the feeling of those in the project that their fate hinged on the success of the test.

    (Quote from David Holloway, Stalin and the Bomb, p. 218.)

  20. Yale Simkin (History)

    What you describe is possible, but is it likely?

    One problem with sub-crit testing is the messiness. It is very chancy that you can contain the plutonium dust after the HE explosion (even US test with just a few kilograms of HE is risky), and doubly that you have to recover and clean the plutonium for re-use. Not only would it be chemically and physically contaminated, but it would contain some unwanted emitters.

    Further, considering that they would likely only get one chance at a super-critical test before the international community would react, does it make sense to waste it on a non-production, sub-scale core?

  21. Lao Tao Ren (History)

    Would be able to get around the sub-critical testing issues mentioned above by:

    a) having access to someone else’s sub-critical testing data and/or simulator

    b) testing with uranium and then the results massaged for Pu?

    c) Better yet, could it be done with depleted uranium and then massaged?

    If any of these are possible, it seems to be that the data could be obtained from somewhere or someone else, and then just tweaked a bit.

    Who says the NORKs have to do all the work “in house”?

    BTW, the Pakistan bomb design found on the Tinner computers was allegedly a uranium bomb, not Pu. If the NORKs used it, they would have to do some additional work.

Pin It on Pinterest