Jeffrey LewisNORK HEU 3 1/2 Years Old?

Greetings from Milan.

For reasons that I will decline to speculate on, Steve Hadley decided to take an internal debate within the intelligence community and make it very public:

This is especially true because some in the intelligence community have increasing concerns that North Korea has an ongoing covert uranium enrichment program.

Notice the passages I italicized: some in the intelligence community and ongoing.

Now, Glenn Kessler has the rest of the story — DIA (aka “some in the intelligence community”) claims that one of the particles found on smelted aluminum tubes and reactor documents was enriched in mid-2004, after North Korea imported the material from Pakistan (therefore, it is “ongoing” rather than cross-contamination):

David Albright, a former U.N. weapons inspector and president of the Institute for Science and International Security, said he had been briefed recently on the findings by government officials. He said “very few particles” had actually been discovered on the documents and the tubes, and that the DIA was basing its analysis on a single particle that, through age-dating techniques, was believed to be about 3 1/2 years old.

The dating could be significant because Pakistan has acknowledged providing North Korea with a sample centrifuge kit for uranium enrichment in the early 1990s. Many analysts have speculated that the tubes and the paper had been contaminated with enriched uranium from the Pakistani equipment. The DIA argued that a particle just 3 1/2 years old could only have been processed in North Korea.

The Energy Department disputed that, saying that the evidence did not exclude the possibility that the traces came from the Pakistani equipment. DOE analysts described the single particle cited by the DIA as an “outlier” from the other particles that were found, Albright said.

Albright said it was “irresponsible and inflammatory” for Hadley to highlight the concerns of just a segment of the intelligence community. “It fans the flames of controversy and hands Obama a hot potato.”

This is the classic pattern for leaks: Time and again, those who lose debates on the inside, appeal to the press. You just don’t usually see the National Security Advisor carrying the water for them.

James is working on a longer post, but in the meantime, the key question is “How big was that frickin’ HEU particle?”

The accuracy of age determination is a function of the number of atoms. In this case, we have a single particle, it is most definitely the size that matters.

James will address this question in some technical detail, but it is rather difficult to date recently enriched uranium with the kind of specificity talked about in the article (“three and one-half years”). Scott Kemp pointed me toward a discussion of this topic in the most recent report by the International Panel on Fissile Materials:

We conclude that it is critical to obtain at least one uranium particle with a diameter of 3 micrometers, and preferably larger, or five to ten smaller suspect particles that contain an equivalent amount of atoms. It would be extremely difficult or even impossible, however, to determine the age of small particles only a few years old. As the time-span between production and detection increases to two decades and more, age-dating becomes easier and more accurate.

“Extremely difficult or even impossible.”

That might explain why the Department of Energy is ot nearly as excited as our friends over at DIA (aka “Do It Again”). Then again, maybe they have some techniques that we don’t know about.

Also, the discussion doesn’t seem to indicate the statistical confidence in the finding. What size standard deviation are we talking about? Implicitly in the choice of “three and one-half years” is a standard deviation of six months or better (otherwise, it is a case of superfluous precision). That seems suspect to me.

As I say, James will delve into this. In case you can’t wait, here is some light bed-time reading:

— M. Wallenius, A. Morgenstern, C. Apostolidis, K. Mayer, Determination of the age of highly enriched uranium, Anal. Bioanal. Chem. 374, (2002), 379-384.

— A. Morgenstern, C. Apostolidis, K. Mayer, Age Determination of Highly Enriched Uranium: Separation and Analysis of Pa 231 , Anal. Chem. 72 (2002) 5513-5516

— Cong Tam Nguyen, Age-dating of highly enriched Uranium by gamma-spectrometry, Nucl. Instrum. Meth. B 229 (2005) 103-110

— G. Hall Uranium age determination by measuring the 230 Th/234U ratio Journal of Radioanalytical and Nuclear Chemistry, 264:2 (May 2005).


  1. Rwendland (History)

    Is there a firm claim anywhere that a particle is HEU? The WP story only seems to imply it is enriched.

    I recall that one of the suspect HEU traces in Iran was plausibly explained to be from spent HEU fuel from the TNRC research reactor, supplied in 1967 by the U.S.

    As Yongbyon has an 8MWt IRT type research reactor using enriched fuel, is this similarly a possible explanation? (IRT-2000 uses 10% enriched fuel.) The paper records are I believe from Yongbyon, and I’ve never seen a claim that centrifuge research is located at Yongbyon, so enriched fuel for the IRT does sound like a possible source that should be considered.

  2. A. Glaser (History)

    The text in the IPFM report is based on an article forthcoming in the Journal of Radioanalytical and Nuclear Chemistry: A. Glaser and S. Buerger, “Verification of a Fissile Material Cutoff Treaty: The Case of Enrichment Facilities and the Role of Ultra-trace Level Isotope Ratio Analysis,” 2009. In the article we conclude that:

    (a) As noted in the quote from the IPFM report, it is impossible to determine (or estimate) the age of a micron-sized HEU particle based on the Th-230/U-234 isotope-ratio if the particle is micron-sized, which is what we expect to find on a swipe sample, and less than 20-25 years old. In other words, an HEU sample has to be quite “large” in order to be able to specify an age of less than 5 five years using the Th-230/U-234-method.

    (b) We do also list potential other factors that might indirectly point to the age of the particle, in particular, a fluorine signature. From the article: “Uranium hexafluoride released from the equipment in an enrichment plant quickly reacts with atmospheric moisture to form hydrolyzed UF6 (e.g. UO2F2 and UO2F2-(H2O)x(HF)y). Over time, further reactions occur, in which the particle gradually loses its fluorine-content to form other uranium compounds. The dynamics of this process as a function of ambient conditions (e.g. moisture and temperature) has yet to be fully characterized in the open literature. Understanding this process would be highly relevant in the present context, because the absence of fluorine in an HEU particle could be a strong indicator for a particle from historic production.”

    Still, the claim remains rather dubious.

  3. Jeffrey Lewis (History)


    Thanks, that’s very helpful. Please let me know when your article is out.


    The second Glenn Kessler story — dated June 21, 2008 — indicated that highly enriched uranium particles were found on the documents.

    The United States in recent weeks has obtained new intelligence — fresh traces of highly enriched uranium discovered among 18,000 pages of North Korean documents — that are raising new questions about whether Pyongyang pursued an alternative route to producing a nuclear weapon, according to sources familiar with the intelligence findings.

    The uranium found on the smelted tubes is described as “enriched uranium” and the uranium in the most recent story is similarly described as “enriched.”

  4. Rwendland (History)

    Jeffrey, sorry I missed the HEU mention in the June 21, 2008 WP story.

    I googled to try to find out exactly what fuel enrichment the Yongbyon IRT research reactor uses, and it seems to use HEU fuel, probably 80% HEU, possibly with some 36% HEU fuel elements. See below for the best sources I found.

    It seems perfectly plausible that Yongbyon would keep the paper records of both on-site reactors in the same office. So it is plausible that the operators of the IRT were regularly coming into the office that had the 5MWe Magnox operating records. So would not a plausible explanation be that the pool-type IRT operators brought HEU micro-particles into the office, which got onto the 5MWe reactor records?

    Why has this plausible possible innocent explanation not been offered to the press? It seems more likely an explanation than off-site enrichment tests being the source.

    This seems somewhat reminiscent of one of the Iranian HEU traces story in the press. It was known from the outset the trace was on spent fuel containers at the Karaj Waste Storage Facility, and that HEU fuel from the research reactor had gone through this facility, so was a plausible source. The info about the HEU trace leaked to the press, but not the circumstances and a plausible innocent explanation. Could the same be happening here?

    The results of my googling the Yongbyon IRT research reactor are:

    ISIS says North Korea has 42 kg of “Russian-origin HEU provided for IRT reactor at Yongbyon.”

    GlobalSecurity and FAS say the research reactor was an IRT-2M model which was power upgraded to 8MWt in the 1970s by using 80% HEU fuel.

    A technical paper about the Czech LVR-15 10MWt research reactor states that it used IRT-2M fuel which is available in interchangeable 36% HEU and 80% HEU variants, both with have 3-tube and 4-tube versions. All 4 variants are configured into the reactor, which seems a fun mix-and-match play kit!

  5. Josh


    That’s a fascinating insight. The lack of greater curiosity about this issue until recently has been a puzzlement; a great deal would seem to hinge on the origins of the uranium traces and how they are interpreted.

    Most of all, it will help to determine whether the State Dept’s “plutonium first” strategy in NK/six-party talks will be deemed viable. Hopefully, well-connected journalists will take a continuing interest in this matter, while being mindful of the potential for manipulation.

  6. Rwendland (History)


    I noticed that an even better source on the IRT HEU fuel is Sig Hecker’s Feb 2008 Yongbyon visit report to the Senate Foreign Relations Committee, where he confirms it uses both 80% and 36% HEU fuel elements, reporting:

    Yongbyon officials … said the key to the IRT-2000 reactor is the fuel. They have not been able to get delivery of new fuel (Director Ri had previously told me all fuel was supplied by the Soviet Union, and that they had not received any new fuel since the dissolution of the Soviet Union). The most recent fuel used in the reactor was 36 and 80% enriched in U-235. I told them that it would not be possible to get new HEU fuel because of proliferation concerns. They indicated that it would be possible to convert the core back to low-enriched uranium, which is what the original fuel was when the reactor was delivered by the Soviet Union. They also stated that the reactor could be operated for several more decades with rather minor enhancements.

    I should have said that this of course would not explain 3.5 year old HEU. But this context should have been explained in the earlier June 2008 WP story, which in precis said “The aluminium tube HEU particles might plausibly have come from Pakistan, but the HEU particles on the Yongbyon paper docs cannot be explained away so easily, and raise enrichment worry.” Ignoring that Hecker’s visit report to Senate just 3 months earlier contained a plausible innocent explanation. A very poor show on briefing journalists, but was Condoleezza Rice also poorly briefed for her Heritage Foundation talk that the WP article was reporting on?

    Surely I cannot be the first person to identify in public the IRT HEU plausible explanation? There must have been many who realised. Surely someone has raised this earlier.

    Does anyone know if pool research reactors using old HEU fuel commonly give of HEU microparticles?