Jeffrey LewisQuestions about NORK Centrifuge Program

A former Department of State told Paul Kerr that “North Korea has probably imported enough components for 3,000-5,000 centrifuges and may have acquired enough for 6,000-7,000.”

Whether those components have been assembeled into centrifuges or whether North Korea can make UF6 feedstock is another question.

Malaysian components produced for
Libya’s nuclear weapons program.

After talking to a bunch of US officials, Paul Kerr concludes we don’t know jack about North Korea’s ability to convert or enrich uranium:

Publicly available intelligence assessments regarding a possible North Korean enrichment facility are inconclusive. For example, the CIA reported in November 2002 that North Korea was “constructing a centrifuge facility” capable of producing enough fissile material for “two or more nuclear weapons per year” as soon as “mid-decade.”

But subsequent agency reports to Congress covering North Korea’s nuclear programs in 2002 became increasingly vague, saying only that North Korea had the “goal” of constructing such a facility. A similar 2003 assessment said nothing about the program.


Providing yet another view, a knowledgeable former congressional staff member told Arms Control Today Sept. 27 that the Bush administration has never presented any “credible evidence” to relevant congressional staff that North Korea has ever sought to advance its enrichment efforts beyond a research and development program.

The question of whether North Korea has a facility capable of producing uranium hexafluoride could also prove difficult to resolve. The public evidence that Pyongyang possesses such a facility is thin, and the former State Department official described the administration’s intelligence on the matter as “pretty sketchy.” Knowledgeable current and former U.S. officials have articulated differing assessments on the matter both in published accounts and interviews with Arms Control Today.

More on North Korea’s phantom UF6 production facility tomorrow.

The uncertainty surrounding North Korea’s nuclear programs is a problem—North Korea will have to declare its nuclear facilities during Six Party Talks.

What will happen when North Korea doesn’t declare a gas centrifuge facility using all those components?


  1. Chris Clary

    Jeff, can we make requests? I’m hoping you can clear up the DPRK HEU program, b/c I’m all confused. Two things. First, from the Libyan and Iranian programs, it appears that rotors and ring magnets are still bottlenecks in centrifuge production. I’d be interested to know how North Korea could overcome this bottleneck, when Libya and Iran haven’t had much luck (apparently) even with Khan’s help. Note how much larger the estimated DPRK program is then the estimate of functioning centrifuges last provided by Albright and Hinderstein for Iran. (Though, if DPRK had a figure equivalent to Iran, those would not exactly be happy days either.) Second, I’m eagerly awaiting your DPRK UCF post. Two data points, I’d like your help on: Sanger and Broad, NYT, 3/31/05: “It is well known that the North routinely makes a precursor known as uranium tetraflouride at a plant near Yongbyon. Federal experts said converting that to the final product was relatively simple.” Compare to Demick, LA Times, 3/24/05: “North Korea has large reserves of natural uranium, but it is unclear whether it has the technology required to produce the gaseous uranium hexaflouride.” and Kessler & Linzer, 2/3/05: “Experts said it would be surprising if North Korea had built a conversion facility.”

  2. Michael Roston (History)

    Somewhere back in the day, from a source whose name I won’t repeat, I heard an analysis that energy-poor NK doesn’t have enough fuel to power those centrifuges to make enough uranium to build a nuclear deterrent. Worth thinking about.

  3. Allen Thomson (History)

    It would be interesting to get some specifics on the difficulties of UF6 production to the quality needed in centrifuges or diffusion plants.

    Looking at, the process appears to involve serious (HF and F2)chemistry, but no more than many others.