Jeffrey LewisNORK Enrichment Facility

Fuel Fabrication Facility. Empty Machine Shop from which lathes have been removed. Exhaust ducts are still in place. Photo credit: W. Keith Luse

Well, well, well.  It isn’t just a light water reactor that North Korea is constructing on the decaying Yongbyon site.

The North Koreans also showed Sig Hecker a 2,000 centrifuge enrichment facility in what was the empty Fuel Fabrication Plant at Yongbyon

Brazen, huh?

David Sanger has the story, although he is vague about whether he got it from Hecker or from Administration officials  (I have one colleague calling him Assistant Secretary Sanger, now):

The scientist, Siegfried S. Hecker, a Stanford professor who previously directed the Los Alamos National Laboratory, said in an interview that he had been “stunned” by the sophistication of the new plant, where he saw “hundreds and hundreds” of centrifuges that had just been installed in a recently gutted building that had housed an aging fuel fabrication center, and that were operated from what he called “an ultra-modern control room.” The North Koreans claimed 2,000 centrifuges were already installed and running, he said.

American officials know that the plant did not exist in April 2009, when the last Americans and international inspectors were thrown out of the country.

The location is fascinating.  We all thought it would be some secret tunnel under a mountain.  Instead, the North Koreans just installed them in plain sight and invited Hecker over for tea.

The choice of location conveniently dates the start of installation to sometime after North Korea tossed inspectors in April 2009.  That would be consistent with the timeline suggested by North Korean statements documented by Josh, in his post Parsing North Korean Enrichment.  Josh noted the statements on April 14, April 29, June 14 and September 4, culminating in the announcement that “experimental uranium enrichment has successfully been conducted to enter into completion phase.” Subtle, they ain’t.

I would guess North Korea started installing centrifuges in either June or September.  Assuming the North Koreans install one cascade a month, a perfectly plausible pace, that would easily result in about 2,000 centrifuges when Hecker visited.  (A nitpick: I am not sure 2,000 centrifuges installed over the course of a year qualifies as either vast or rapid — Sanger engages in some hyperbole to titillate a reader whose eyes might glaze over at “centrifuge” — but twelve cascades is, of course, be plenty big enough for a bomb program.  That’s the part of the problem, one doesn’t need a vast centrifuge facility to build a bomb.)

The choice of location also makes plain the North Korean’s point: That building was empty until the Six Party Process collapsed.  Draw your own conclusions, Yankee Doodle Dandy.

I despair of convincing whichever Kim we are dealing with this week to abandon North Korea’s enrichment capabilities, though I don’t like any of the other options any better.

This is yet another a perfect example of how AQ Khan helped distort our North Korea policy.  Khan asserted — inaccurately, it now appears — that North Korea already had such a facility in 2002.  Our political system responded to that possibility as though it were either true, or would be shortly, setting into the motion the events that resulted in the collapse of the Agreed Framework, the farce of the Six Party Talks and, now, North Korea’s construction of the very same enrichment facility whose hypothetical existence started this episode in the sorry saga of US-North Korean relations.

(As a side note, why do so few reporters understand that no one questions whether North Korea was seeking a uranium enrichment capability in 2002, only over the scale of the cheating?)

There are lots of questions about what sort of machines the North Koreans are spinning, how they are performing and so on.  That will provide plenty of wonkporn, though the Norks didn’t let Sig take any happy snaps. In the meantime, if you are looking for a nice, comprehensive account of North Korea’s enrichment efforts, I recommend David Albright and Paul Brannan’s, Taking Stock: North Korea’s Uranium Enrichment Program.  Obviously, they will want to update it, but I think it holds up very well.


  1. Cheryl Rofer (History)

    Here’s Sig’s report. PDF, but short.

    • Allen Thomson (History)

      Since I’m interested in intelligence matters, I hope we get some clarification on “this facility was undetected in the middle of the Yongbyon fuel fabrication site.” Did the US have no idea that the centrifuges were being installed there? Or were there indications that something was going but details are lacking? Or what?

      BTW, just for the sake of covering all possibilities, is there any argument to be made that this might be a Potemkin facility? I can’t think of anything that seems particularly plausible, but maybe someone who’s better at reading DPRK minds can come up with something.

    • joshua (History)

      I have a hard time thinking that the facility went undetected. They would have had to demolish the previous structure and build the new one. It has a new and different roof, in an eye-catching color, no less. Also, it’s a long, two-story building with a pitched roof. It resembles a “slice” or “strip” of centrifuge facilities elsewhere.

    • Allen Thomson (History)

      I’d be surprised if the building went undetected, but detection and identification aren’t the same thing. That other centrifuge buildings are long and narrow might not have been enough of a tell.

    • joshua (History)

      It’s not so much the narrowness as A) the length, B) the height, and C) the style of roof, all taken together. In any case, it’s hard to miss that they rebuilt the structure at the same time that they were talking about U enrichment.

  2. bradley laing (History)

    Maybe the reason why no one in intelligence mentioned a facility “in plain sight” is because someone got a telephone tap in place, and is afraid of letting anyone know.

    During WW II, a well written Chicago Tribune article about a Pacific War sea battle had the U.S. Navy fearing that the Japanese government would realize that their codes had been broken.

  3. bradley laing (History)

    Clarification: …would lead to the Japanese government realizing that Japanese codes had been broken.

  4. P.E.T. (History)

    That enhanced non-proliferation funding is really paying off – so far there’s no evidence that Mexico has acquired nuclear weapons, yet.

    • Anon (History)

      Setting an example for the rest of the world is really paying off also: go ahead make new modern warheads and don’t sign the CTBT or New START. So will we.

  5. markob (History)

    Great post, but do you really think that AQ Khan distorted US Nork policy or did the hawks actually wanted the policy to be distorted?

  6. Jalon (History)

    They probably revealed the site because they know we know they know we know…or something..

  7. Adalbert (History)

    North Koreans probably invited Hecker after recent publication UN report on the spread of Pyongyang’s weapons technology, and also public information about the arrest of Kim So-in and Kim Song-il.
    If the two nuclear scientists were working for foreign intelligence agencies invitation of Hecker would make sense.
    North Koreans want to ponder more concessions during the six-party talks and they need strong arguments, and since the U.S. and South Korea already know about this facility, the public disclosure of this fact only create more pressure on the superpowers during the talks.

  8. Daniel Pinkston (History)

    I’m not convinced Khan’s assertion is inaccurate. The DPRK has been interested in enrichment and working on it since the 1980s. Under their “military first” doctrine, I think it is naive to assume or believe there is no separate military facility somewhere. I won’t be surprised if they conduct a third nuclear test with a uranium bomb next year, which would pretty much confirm my suspicion. Oh, and it will occur under the “wise leadership of the genius General Kim Jong-un.”

  9. Greenish (History)

    Any word on whether Hecker noticed whether they were using Siemens PLCs? Have they plugged their USB ports with epoxy yet?

  10. Carey Sublette (History)

    Quote of the Day (from the NY Times):

    _South Korean Foreign Minister Kim Sung-hwan downplayed the newly operational facility, saying, “It’s nothing new.”_

    I suppose Kim’s right, as long as one overlooks the fact that is _completely_ new.

    • rwendland (History)

      Not new in the sense that, as Hecker noted, on 15 April 2009 the DPRK Foreign Ministry announced “we will proceed with our own LWR fuel cycle.”[*]

      Then on 4 Sept 2009 KCNA stated:

      “Experimental uranium enrichment has successfully been conducted to enter into completion phase”

      And on 3 November 3 2009:

      “Signal achievements have been made in the production of uranium ore thanks to the high zeal and strenuous labor drive of the officials and workers to satisfactorily provide nuclear fuel to the LWR power plant to be built by the DPRK itself in the future.”

      So to alert analysts this should be no surprise.

      Interesting though that the LWR is being built at Yongbyon, next to the 5MWe Magnox-type GCR. I suppose reutilising the existing spent fuel facility, and river water supply, saves some work. I suppose they will have to build a new, and larger, cooling tower – I wonder if the 5MWe GCR could utilise that as well should they ever want to restart it?

      [*] NB If anyone has a URL for the 15 April 2009 announcement, could they give it – I cannot track down that announcement replayed at KCNA.

    • joshua (History)

      The April 14 statement is here:

      For reference, all four statements from 2009 about an LWR or enrichment are linked here:

    • rwendland (History)

      Thanks Josh for that URL. I’d forgotten that article you’d written last year on the progressive KCNA LWR/enrichment statements – you were certainly on the ball there.

  11. rwendland (History)

    You just have to admire the Fox News love of the facts:

    “The Obama Administration and the North Korean HEU facility

    The confirmed report by a visiting American scientist that North Korea has constructed a facility at its Yongbyon nuclear complex to produce highly enriched uranium (HEU), apparently for use in a nuclear weapon …”

    by James Rosen | November 22, 2010

  12. Ben (History)

    The North Korean rationale for this setup is bizarre. Surely I can’t be the only one that thinks this. If they really want to increase their stockpile/bargaining position why not reactivate the 5MW reactor, or start a rebuild on a larger gas graphite reactor? Why go for the HEU route for heavier, less deliverable warheads? (greater concealment?). Why build a 25 MW LWR?, unless it’s a soviet style LWGR, it’s not going to be much use for Pu production. It’s a lot of effort to keep up the pretense of ‘peaceful activity’.

    Also, did they ever find the hidden reactor?

    • Carey Sublette (History)

      ” Why go for the HEU route for heavier, less deliverable warheads? (greater concealment?).”

      That would be my guess – look at the speculation right now about whether this cascade hall is the only one they built. This one is the showcase of their capability.

      A 20 MW reactor and reprocessing plant have much larger signatures.

    • Azr@el (History)

      Cost, even socialist states must deal with opportunity cost trade offs. Uranium 235 enrichment in a capitalist or socialist state is just cheaper in terms of resource allocation than plutonium 239 transmutation from scratch. And this takes into account the greater amount of U235 required for a device.

    • Mark Gubrud (History)

      Then again, it could be that this is really about electricity. Hecker seems to think so.

      1. If it doesn’t actually plan on collapsing, NK needs the juice.

      2. Nuclear is juche juice.

      3. They already have enough deterrence.

      4. A big U enrichment complex gives them the capacity to turn out HEU for more bombs, if they decide to.

      5. But it has a legitimate use, too. So, they may reckon, they can get a deal on their nuke weapons while keeping the bomb-in-a-jiffy capability. Their Iranian friends provide a nice example of this logic.

      6. Sure, they can also set up a covert plant with the same centrifuges. But I see no reason to assume they already are.

  13. Homer Williams (History)

    Please note that Siegfried S. Hecker published a trip report on Nov 20. Perhaps someone could read it and provide a detailed critique?

    Report Link:

  14. Ben (History)

    ‘A 20 MW reactor and reprocessing plant have much larger signatures.’

    I don’t disagree with you, but with North Korea having a reported 8-12 weapons already, aren’t we past the hide and seek stage?(there’s zero chance of a pre-emptive strike on a facility however large the signature). Also, concealment didn’t seem to bother them that much when they were building the now defunct 50MW and 200MW reactors.

    • Carey Sublette (History)

      A concealed nuclear weapon production option was not available when they started their program. It was only after gas centrifuges became a gray/black market commodity that this became a realistic choice.

      On the other hand – one additional possibility is that they are interesting in making two-stage pure fission weapons so that their 4 kt proven plutonium device could drive explosions one or two orders of magnitude larger. What if their second test included a radiation implosion experiment?

      8 to 12 100 kt bombs is a rather different prospect than 8 to 12 4 kt bombs.

    • Allen Thomson (History)

      > making two-stage pure fission weapons so that their 4 kt proven plutonium device could drive explosions one or two orders of magnitude larger.

      An interesting possibility that I’ve wondered about myself. Do you have an analysis of what a primary-driven uranium secondary (add DT for boosting if desired) might achieve? I assume that it gets way more compressed than if HE were used around an equal mass of uranium.

    • George William Herbert (History)

      Allen writes:

      An interesting possibility that I’ve wondered about myself. Do you have an analysis of what a primary-driven uranium secondary (add DT for boosting if desired) might achieve? I assume that it gets way more compressed than if HE were used around an equal mass of uranium.

      “A lot”.

      Depends on how much HEU you use. Orange Herald, anyone? Layer-cake (tamper-Li6D-HEU), 117 Kg HEU, 720 Kt. SOB? 60 Kg HEU, 500 Kt.

      There’s certainly more captureable and redirectable energy in the NK Pu device than was needed to implode either of those assemblies.

      The assemblies also don’t need as much enrichment; energies of compression are so much higher that lower enrichment will go substantially supercritical. Compare the critical masses of various enrichments of HEU from 20% up with the shock compression energy of U and reasonable volumes; lots of options, many of which are far fewer SWU than primaries need…

  15. Ben (History)

    Makes you think…I wonder if they had problems with reprocessing, it would explain the fizzles I guess.

  16. Allen Thomson (History)

    As I understand the Hecker report, the LWR is supposed to produce 20-30 MW electric power (from what generators?) with a base thermal power of around 100 MW (leaving 70-80 MW waste heat to be disposed of where?).

    Assuming that all gets up and running and is actually optimized for power production, how much plutonium would the reactor produce per year in actual operation?



    Experimental 25 to 30 MWe LWR construction

    At the 5 MWe reactor site we were taken to a construction site that had been identified previously from overhead imagery. The chief engineer of the 5 MWe reactor showed us the site and answered questions, but only when pressed.


    It is designed for a power level of 100 MW (thermal). He chose not to specify the electrical power, but said that the conversion efficiency is typically 30 percent. Therefore, I estimate the electrical power to be roughly 25 to 30 Mwe.

  17. bradley laing (History)

    Could there be a division within the government of North Korea, one that would lead to one faction doing something a second faction would not want to do?

    Maybe it “doesn’t make sense” because groups with overlapping areas of responsibility were fighting for resources.

    Th one-stage Mark-18 Nuclear Bomb was built, as a parallel program to the two stage bombs.

  18. GWR (History)

    David Rothkopf in FP topay: “These events may also ultimately be seen as wins for Kim on two other levels. First, the nuclear facility almost certainly required international collaboration. If it turns out that support came in part from, say, Pakistan, already suspected of helping the Koreans develop a nuclear ballistic missile capability, it would be deeply embarrassing and awkward for the United States.”

    I’m interested to here Jeffrey’s and others’ thoughts on whether international collaboration was required and whether this means that the AQK network might still be operational.

    • Jeffrey (History)

      As I told Joby Warrick, “It is naive to think that somehow these guys aren’t still doing business.”

      Recall that Pak-NORK cooperation continued AFTER AQ Khan “retired.” That was a major objection of mine to the WaPo article where Khan claimed North Korea had a full scale plant. He had been cut out of the loop by that time, and was guess-timating based on his past experience at KRL and in North Korea (which was interesting speculation, though not dispostive).

      So, given that Pakistan continues to operate a program based on illicit foreign help and continued to assist North Korea for some time, I would conclude that both Pakistan and North Korea continue to maintain active supply networks that can support their programs, as well as future ones (Burma and Algeria come to mind).

      What isn’t clear to me is whether, having received assistance from Pakistan, North Korea set up its own independent network, or if the two remain intertwined.

    • joshua (History)

      I think it was the North Koreans who provided missile technology to the Pakistanis, not the other way around. Still, you can’t rule out some variety of collaborative development.

  19. George William Herbert (History)

    Generic thoughts –

    One doesn’t leap straight to a 2,000 centrifuge public cascades set without other big facilities first, including a big centrifuge production line.

    It’s possible that this was the first installation of note, but that seems unlikely to me.

    It’s also possible that this was a recent develoment. If it relied a lot on foreign design aid or parts, Pakistan is (hopefully) off that market these days. They were in it previously. Iran is in it now as a consumer; they could also be a producer/exporter now, though it’s not clear that they are doing that. The specific centrifuge models aren’t apparently what Iran’s open facilities use. But Iran could have stuff under other mountains we haven’t seen inside.

    Could this have been purely a civil power reactor program move, doing what they claim it is without a military component? Remotely possible, but those are pretty advanced centrifuges to be starting out a commercial program with.

    Do they have legitimate electrical power problems which would justify a reactor? Probably.

    Given the shellings, one has to wonder what exactly they’re thinking or planning overall. If the shellings were intentional, and not an accident of some sort or misunderstanding of the SK exercise.

    All in all, not a hope-inducing week.

    • Jeffrey (History)

      “Installation of note” is an interesting phrase. What do you mean by that?

      I would expect there to be some other facility with small (1, 10, 20, etc.) test stands and a pilot cascade (164 machines) or two, but not more than that.

      The Iranians went from a pilot plant that did test stands, followed by a single cascade (164), then moved to mass installation at the adjacent Fuel Enrichment Plant.

      I’d be surprised if the North Korean model were different — ie a pilot facility someplace with no more than, say, 300, centrifuges followed by the facility at Yongbyon.

      Do you disagree?

    • Carey Sublette (History)

      I am certain there are people on this site that have been keeping close track of the chronology of Iran’s centrifuge developments. How does the rapid installation we see in the DPRK match up with Iran’s setting up cascade halls of similar size?

      How long did it take Iran to set up an operating hall of 2000 machines after they had set up the original pilot cascade? If installation and startup (as is claimed) proceeded here more rapidly than in Iran it might suggest prior experience.

  20. ikje (History)

    I have several remarks to this story.

    I also came along with the remarks made above that a hybrid weapon could be a target for the uranium enrichment plant.

    A Uranium only, bullet/barrel type of bomb would be much easier to construct and detonate.
    It would detonate with about 30 kt power, much more than their two Pu sizzles of 0.8 to 4 kt yield.
    It would prove that the NORKs have mastered full explosion capability.
    It would open up the possibility for a hybrid nuke with enhanced power in the 500kt range.

    The installation of 2000 centrifuges into the old fuel preparation building is an incredible achievement in terms of engineering.
    As I understand, the building was empty as of april 2009, looking according to the picture provided with this post, which would fit with satellite images.

    In order to install the ultracentrifuges, one will have to remove the floor, put a solid concrete foundation in it, let it harden, at least 54 days. Add tubing, electric wiring, cooling systems. And install ready build and dry tested centrifuges at a rate of 5 to 10 centrifuges a day. The installed centrifuges will have to be leak tested, run in, vacuum tested, sensors must be validated etc. This will take more than 20 days/centrifuge to accomplish. It can be done in parallel if sufficient test equipment is present, but this is highly unlikely.

    To my humble opinion, it is impossible to mount an empty hall in 18 months with 2000 centrifuges and get the circus running.

    Unless it is a circus.

    Maybe the centrifuges are just mock-ups. But even then: installing 2000 mock ups in a fully adequate modified old building is a gargantuan task to be accomplished in 18 months.

  21. ikje (History)

    The answer to this fantasy is embarrasingly simple:

    The NORKS already have enriched U235 to 90%+ and are weaponizing it. The next test explosion will be this weapon. But to explain their possession of U235, they need to show a working smoking gun at a place that is not related to their actual enrichment site.

    And remember this strange short notice about their mastering of nuclear fusion, earlier this year?
    Exciting times, sure.

  22. ikje (History)

    The NORKS think that Uranium enrichment will solve their problem with their stack of plutonium, which is obviously not suited for the big bang. This may be due to the isotopic composition of the Pu they have recovered after long labour, which has costed so many lives, to no avail. It appears that the Pu they have recovered from the YoungByong reactor contains too much Pu240 to be useful for a Pu only implosion type of device. Since Pu cannot be purified in a centrifuge, as does Uranium, they have decided to go for Uranium enrichment.

    In combination with their apparent experiments with LithiumDeuteride, it makes perfectly sense. We should not underestimate the NORKS.

  23. ikje (History)

    The implications are telling: the NORKS had the centrifuges ready and spinning for a long time before. If you want to be seemed to be able to install 2000 centrifuges in 18 months, you are actually telling the world that you had these centrifuges spinning all the time. And that you now are willing to harvest what they have been preparing. The next nuclear test of the NORKS will be a uranium device, and occur within 6 months. They are already preparing for it.

  24. Allen Thomson (History)

    > Maybe the centrifuges are just mock-ups.

    While I think the prudent assumption should be that the centrifuges are genuine, the available evidence and general logic are equally consistent with the mock-up (“Potemkin”) hypothesis. We have the report of knowledgeable observers who were allowed a brief look at the cascades, no physical access, weren’t allowed to take pictures, only got to ask a few questions of the DPRK local official, and didn’t have the opportunity to verify the answers to the questions. Lots of room for uncertainty there.

  25. Mark Gubrud (History)

    ikje’s speculations are interesting, but they are just that.

    I think the most one can conclude from the argument is that it is unlikely that the centrifuges and ancillary equipment were produced and assembled from scratch in a mere 16 months, and that they are all currently tested, run-in and operational.

    More likely this was already well in preparation before April 2009 and it is certainly possible that most of the centrifuges have so far only been installed and are not yet operational. I would discount mock-ups as unlikely.

    Hecker reported that the cascades were configured for LEU and that is NK’s story, so it is hard to see their strategy in showing this facility “to explain their possession of U235”.

    I agree it is likely their “fizzles” are due to Pu-240 but they ought to realize, and incidents like the Cheonan and the shelling this week suggest that they do, that their 0.8-4 kt bombs, on top of the rest of their weapons pile, are scary enough. SK and the US are sufficiently deterred from attacking or escalating with NK. What more would NK gain from having Mt-class nukes (which would be harder to deliver)?

    The greater threat to the DPRKs survival comes from within. What they really need is electricity.

    • Murray Anderson (History)

      Remind me why the U.S. has bombs bigger than 20 kilotons. Bigger is generally perceived to be better, for deterrence or anything else. In terms of numbers of bombs, more is also perceived as better, at least up to a point way beyond what North Korea has.
      Also, HEU can be sold at a high price. The commercial value, if it were available as a commodity from Urenco or Tenex, would be around $40,000 a kilogram. North Korea could charge several times that to interested parties. North Korea produces little of value, other than missiles and now perhaps HEU, and HEU is easier to deliver.

      Murray Anderson

  26. ikje (History)

    As far as I know, you can not transport a tested centrifuge, install it and run it. You will have to take it apart, ship the parts, reassemble it and start testing anew. In many cases you will have to replace parts due to damage by transport. In any case you will have to match the centrifuge to the control systems, which is not done so easily. As for HEU or LEU configuration, this depends on the lay-out of the tubing of the cascades. I guess that the NORKS did not allow to let Hecker see the lay out of their tubing. The last stages of enrichment need special precautions to prevent criticality incidents, but the other stages are not easily discriminated. Especially when there are no cooling tubes around the centrifuges visible, as Hecker noted in his report (contrast this to the images of Natanz).
    In my opinion, it is impossible to get 2000 centrifuges up and running in 18 months, even when starting with pretested and certified centrifuges, in a building that has to undergo a complete interior rebuild for fundations, tubing, electricity (do you have an idea how much electricty 2000 centrifuges need? They would have to build a complete new UHT supply for the facility), cooling, massive vacuum plant and the very complicated control cabling? Install and test it? No way. During start up, the centrifuges need to be heated in order to prevent sublimation of solid UF6 on the rotors, which would destroy them immediately.

    And I could go on like this. It is simply impossible to install 2000 centrifuges in 18 months and get them running.

    • Jeffrey (History)

      Dear Ikje:

      I am losing my patience with you. You make lots of assertions, but don’t really provide any evidence.

      If you want to continue to participate in this forum, I am going to require you to actually do some work, rather than just type whatever happens to be passing through your head.

      You assert: “It is simply impossible to install 2000 centrifuges in 18 months and get them running.”

      Between, 17 February 2007 and 19 August 2007, Iran installed 12 cascades in the A24 module and had them all enriching uranium (1,968). By 3 November 2007, they installed another 6 cascades that were also enriching uranium (another 984).

      There is an argument that installation work adds some lead time — in that case, on 12 December 2007, Iran had zero centrifuges under construction in the A26 module. Eighteen months later, by 31 May 2009, Iran had 12 cascades enriching (1,968 centrifuges) in A26 module.

      I am willing to listen to arguments about why the North Korea case is different, but I am not willing to simply let you assert one false thing after another.

  27. ikje (History)

    Dear Jeffrey,

    Sorry to hear that you are losing your patience. My remarks were well researched, but I assumed that everyone on this blog would be aware of the simple facts that can be verified by all.

    My data are based on the IAEA reports, which are very difficult to compare due to format changes during the reporting period. It is also good to remember that prior to 13-5-2007 no inspection of Natanz has taken place, AFAIK.

    My data:

    under construction vacuum test vacuum operation operating days centr const centr test centr vac op centr operating total centr/day

    17-2-2007 gov/2007/8 2 2 328 328 0 0 656
    13-5-2007 gov/2007/22 3 2 8 85 492 328 0 1312 2132 17
    19-8-2007 gov/2007/48 2 1 1 12 98 328 164 164 1968 2624 5
    3-11-2007 gov/2007/58 0 0 0 18 76 0 0 0 2952 2952 4
    12-12-2007 gov/2008/4 0 0 0 18 39 0 0 0 2952 2952 0
    7-5-2008 gov/2008/15 15 0 1 20 147 2460 0 164 3280 5904 20
    30-8-2008 gov/2008/33 12 0 1 23 115 1968 0 164 3772 5904 0
    7-11-2008 gov/2008/59 13 0 0 23 69 2132 0 0 3772 5904 0
    1-2-2009 gov/2009/8 3 0 9 24 86 492 0 1476 3936 5904 0
    31-5-2009 gov/2009/35 1 0 13 30 119 164 0 2132 4920 7216 11
    12-8-2009 gov/2009/55 14 0 8 28 73 2296 0 1312 4592 8200 13
    2-11-2009 gov/2009/74 1 17 12 24 82 164 2788 1968 3936 8856 8
    31-1-2010 gov/2010/10 3 0 2 23 90 492 0 328 3772 4592 -47
    24-5-2010 gov/2010/28 0 21 7 24 203 0 3444 1148 3936 8528 -2
    28-8-2010 gov/2010/46 0 25 6 23 96 0 4100 984 3772 8856 3
    5-11-2010 gov/2010/62 0 25 0 29 69 0 4100 0 4756 8856 0 5

    Which means an average of 5 centrifuges installed/day during 3 years. The data show surges of massive centrifuge installation alternated with long periods of no activity. Striking is gov/2010/10, which seems to imply the removal of almost half of the number of centrifuges? The reports are very skinny, difficult to get a consistent reporting picture. As for the tubing and other facilities, I am working on that.

    I hope this additonal information will be able to convince you that I am not merely speculating.

  28. ikje (History)

    Dear Jeffrey,

    As promised, my data on the support systems.
    A21 A22 A23 A24 A25 A26 A27 A28
    tubing installation running tubing installation running tubing installation running tubing installation running tubing installation running tubing installation running tubing installation running tubing installation running
    17-2-2007 gov/2007/8 final stage 2 2
    13-5-2007 gov/2007/22 final stage 5 8
    19-8-2007 gov/2007/48 final stage 4 12
    3-11-2007 gov/2007/58 completed 18
    12-12-2007 gov/2008/4 operational 18 started started started
    7-5-2008 gov/2008/15 operational 18 going on no report 16 2 going on going on
    30-8-2008 gov/2008/33 operational 18 going on no report 13 5 going on going on
    7-11-2008 gov/2008/59 operational 18 going on no report 13 5 going on going on
    1-2-2009 gov/2009/8 operational 18 going on no report 12 6 going on going on
    31-5-2009 gov/2009/35 operational 18 going on no report 6 12 going on 8
    12-8-2009 gov/2009/55 operational 18 going on no report 8 10 going on 14
    2-11-2009 gov/2009/74 operational 18 going on no report 12 6 going on 18
    31-1-2010 gov/2010/10 operational 1 17 going on no report 1 6 going on 14
    24-5-2010 gov/2010/28 operational 18 going on no report 12 6 going on 16
    28-8-2010 gov/2010/46 operational 1 17 going on no report 12 6 going on 18
    5-11-2010 gov/2010/62 operational going on no report going on
    going on no report going on

    The remarkable facts are that it seems to take at least 1.5 year to get the tubing ready in a building that was designed to harbour an uranium enrichment plant. Which means that all tubing, cabling for the main facilities was already in place.

  29. ikje (History)

    Dear Jeffrey,

    As for my claims that the NORK plant could not have been build, the following from gov/2009/74:

    10. The DIV included a detailed visual examination of all areas of the plant, the taking of
    photographs of cascade piping and other process equipment, the taking of environmental samples and
    a detailed assessment of the design, configuration and capacity of the various plant components and
    systems. Iran provided access to all areas of the facility. The Agency confirmed that the plant
    corresponded with the design information provided by Iran and that the facility was at an advanced
    stage of construction, although no centrifuges had been introduced into the facility. Centrifuge
    mounting pads, header and sub-header pipes, water piping, electrical cables and cabinets had been put
    in place but were not yet connected; the passivation tanks, chemical traps, cold traps and cool boxes
    were also in place but had not been connected. In addition, a utilities building containing electricity
    transformers and water chillers had also been erected.

    I think it describes pretty accurately what I claimed to have been neccesary for the YeounBeoung plant, and I remain assured that this could not have been achieved within 18 months.

    Please, illuminate me!

    • Jeffrey (History)


      Alright, I admit you are serious. I apologize for losing my patience. If you will accept my apology, I would appreciate it if you helped me work through this.

      So, you think the crucial factor is the lead time necessary for installation work, including equipment and subheader pipes, etc.? Am I correct?

      In that case, than I think we should focus on the A26 module, rather than the A24 module, since the IAEA reports offer more commentary on the state of installation work. (That’s my recollection, I might be wrong about that.)

      I see no reference to installation work on 19 August 2007; some work had begun by 3 November 2007.

      The Iranians had 2 cascades enriching by 7 May 2008 and 12 cascades enriching (and another 6 under vacuum) by 31 May 2009.

      7-9 months to complete installation work and install two cascades, followed by an installation of another ten cascades over the course of a year. That’s 19-21 months.

      A bit faster than the North Koreans, but not so different. I could imagine compressing installation times, especially when it comes to adding cascades. You use the average rate, but why not use the max Iranian installation rate? There are times when the Iranians seemed to be hurrying to make a political point, other times when they seemed to be taking their own sweet time.

      Anyway, here is the data as I collected it:

      Date of Visit to Natanz Day # from 17/02/07 TOTAL Enr TOTAL Vac A26 Enr A26 Vac A26 Constr
      28-Aug-10 1298 23 54 6 12 0
      24-May-10 1202 24 54 6 12 0
      31-Jan-10 1089 23 41 6 1 11
      2-Nov-09 1017 24 53 6 12 0
      12-Aug-09 935 28 50 10 8 0
      31-May-09 852 30 43 12 6 0
      1-Feb-09 733 24 33 6 9 3
      7-Nov-08 629 23 24 5 1 12
      30-Aug-08 560 23 24 5 1 12
      7-May-08 445 20 21 2 1 15
      12-Dec-07 298 18 18 0 0 0
      3-Nov-07 259 18 18 0 0 0
      19-Aug-07 183 12 12 0 0 0
      13-May-07 85 8 8 0 0 0
      17-Feb-07 0 0 0 0 0 0

      Date is the date the inspectors were at Natanz, Day number is the number of days elapsed since 17 February 2007, ENR is the number of cascades enriching; VAC is the number of cascades under vacuum. A26E, V and C are the number of cascades enriching (E), under vacuum (V) and under construction (C) in the A26 module.

  30. ikje (History)

    Dear Jeffrey,

    Thanks for your reply. Yes, I think the pipe and cabling installation takes far more time than the actual installation of the centrifuges.

    I would rather prefer to look at the data for A28, since this is more identically covered in the various reports than A26, of which no data about installation of tubing etc could be found by me. Installation in A28 started before 12/12/2007, first mentioned in gov/2008/4. In gov/2009/35 the first installed cascades are mentioned to be observed on 5/31/2009. This would imply an installation time for tubing cabling of at least 14 months, since on 2/1/2009 installation was still going on. On 11/2/2009, 18 cascades were installed, but not running. This gives an estimate of 7 to 8 months for about 2000 centrifuges. The basic infrastructure needs several months (curing time for concrete etc). Even when you can do some operations in parallel, it comes more to at least two years to get the thing installed.
    I think we see the same thing in FFEP, one year after the first inspection on 11/2/2009, still no installed centrifuges. And this plant had the EHT power supply, the cooling systems, the gas conditioning system and the vacuum facility already installed at that date. All provided the Iranians were continously working at max rate, but I see no reason why they should not?

  31. ikje (History)

    Dear Jeffrey,

    I am afraid my last post went into the spam-bin, so I will repost it here without the link. If it was a duplication, please feel free to correct, and I apologize.

    Dear Jeffrey,

    And what added up to my concerns was the report by ISIS, “Taking stock” by Albright and Brannan 10/8/2010. They were unable to find the required massive acquisitions of critical components for a 3000 centrifuge plant in the various reports. (Maybe we can find them in the Wikileak documents?). In this report is also a picture of how centrifuge parts are shipped, in this case to Libya: disassembled, parts individually packaged. I based my assumption on the need of in-situ assembly of any centrifuge that has been transported over any distance on this picture.

  32. Jeffrey (History)


    After the original 3,000 centrifuge model, I think the cascades were next installed in A26.


    “There has been no installation of centrifuges or centrifuge pipework outside the original 18-cascade area. Work to install feed and withdrawal infrastructure and auxiliary systems is continuing.”

    EXPLANATION: We know there is only preliminary work occurring outside A24.


    “There has been no installation of centrifuges outside the original 18-cascade area. Installation work, including equipment and sub-header pipes, is continuing for other cascade areas.”

    EXPLANATION: Now we know that there is installation work outside A24. But where?


    “Installation work has continued on four other units as well. [Note 2: Units A25, A26, A27 and A28.] On 7 May 2008, two 164-machine (IR-1) cascades of one of the four units [Note 3: Unit A26] were being fed with UF6, and another cascade of that same unit was in vacuum without UF6.”

    EXPLANATION: Thanks to the footnotes, we now know that the installation work was occurring in three modules: A25, A26, A27 and A28. But Iran installed the first two cascades outside of the A24 module in A26.

    I think this is right, no?

  33. ikje (History)


    I think you have the right citations. However, I disagree with your conclusions. As for Gov/2007/58: I was confused by this. It would imply that the Iranians installed 18 cascades in A24 and started running them without the necessary withdrawal/feed tubing? Does not make sense.


    “There has been no installation of centrifuges outside the original 18-cascade area. Installation work, including equipment and sub-header pipes, is continuing for other cascade areas.”

    Continuing means that it had been observed to be started. There is no report about the start of installation work but gov/2008/4, and this reports only the start of installation in A25, A27, A28.

    When comparing the duration of the installation work in A25, A27, A28 and FFEP with the not mentioned construction work in A26, I get the idea that A26 was already installed at the first inspection, but was simply not mentioned in the reports as such.
    As for the tubing, I just finished reading Glaser 2008.

  34. Mark Lincoln (History)

    “”Can loud debate and haughty speech stop sky-covering enemies’ rush? When a cavalry ravages, can we repulse it with talk? Can pinfeather thrust it?”
    – Gwanghae-gun Ilgi

    The Democratic People’s Republic of Korea is fatally weak and it’s rulers know it.

    “When whales war, shrimp are crushed.”
    – Old Korean Saying

    We have been watching a shock and awe show. A demonstration of technical prowess followed by a demonstration of military will and perhaps fuel-air artillery rounds.

    Bravado. loud debate and haughty speech, of the militaristic fashion.

  35. rwendland (History)


    Have you spotted this:

    Posted Date : 2010-12-03 (NK Brief No.10-12-3-2)

    The Korea Trade-Investment Promotion Agency (KOTRA) looked into the Chinese government’s import and export figures and determined that North Korean exports to China during the first eight months of the year were worth 650,000 USD, 20.6% more than during the same period last year …

    … The second- and third-largest imports were listed as “nuclear reactor, boiler, and machinery” (127,000 USD) …”

    Interesting? If this is correct, seems China is helping with the LWR. Maybe the 2012 target isn’t so unlikely.