Jeffrey LewisGive An Inch, They Swim All Over You

The United States and Europe may let Iran continue converting uranium. This is a really bad idea.

David Sanger at the New York Times reports that IAEA DG Mohamed ElBaradei and SECSTATE Rice discussed a new proposal that would allow Iran to convert uranium at Isfahan (Esfahan), shipping the UF6 to Russia for enrichment.

The new proposal, officials from both Europe and the United States said, is an effort to give Iran a face-saving way out of its tense standoff by arguing that it has retained what it contends is its right to enrich uranium as a signer of the international Nuclear Nonproliferation Treaty, but has simply chosen to do so at facilities in another country.

According to officials briefed on the discussion between Ms. Rice and Dr. ElBaradei on Tuesday at the State Department, the two talked about letting Iran take a financial stake in an enrichment facility in Russia.


The proposal on the uranium program, they said, would follow the same model: Russia and other nations would ensure that the uranium shipped to Iran would not be usable in a weapon. All of the nuclear waste would also have to be shipped out of the country.

Francois Murphy at Reuters and George Jahn at AP have more.

This is a bad deal—Iran’s conversion facility is a bottleneck in Tehran’s program. In the short-term, if Iran continues to improve various conversion related technologies (including the use of pulse columns for solvent extraction) then the international community is not delaying Iran’s possible bomb program.

The proposal is very similar to a recent South African plan to allowed Iran to continue to produce UF4 and UF6 at the Uranium Conversion Facility (UCF) in Isfahan, with the product being shipped to South Africa for storage during negotiations.

Mark Hibbs (“Iran’s critics reject South African proposal to let Iran operate UCF,” Nuclear Fuels 30:20, September 26, 2005, 8) reported that Western states opposed the South African plan because it allowed Iran to improve its conversion capabilites:

According to one Western official attending a meeting of the IAEA board last week, “Iran has technical problems to solve in three key areas of its nuclear program: making clean UF6, enriching uranium, and weaponization.”

If the South African proposal were to be accepted, he said, “Iran would be able to solve one of the three problems and take a step closer to their goal,” which he said was having the technical means to produce weapons-grade fissile material and nuclear weapons.

“We told South Africa this was a very bad idea,” said an official from another Western board delegation.

Those were good arguments then, they are good arguments now.

Over the long-term, the IAEA will be challenged to verify that Iran is not diverting significant amounts of UF6 for a clandestine centrifuge program. Iran has employed accounting practices for its nuclear material that would embarrass Ken Lay. As a colleague of mine recently said:

Since a nuclear fuel cycle is really just a chemical process with a really big chemistry kit, “stuff” gets lost. When these “process losses” happen, there’s usually a margin of error accounted for in the IAEA’s material balancing. They don’t expect countries to be perfect with accounting, just reasonable. Iran was able, according to the BOG reports, to declare process loss for material that had actually been diverted.

See where this one is heading?

Iran’s use of the process loss to hide material (depleted UO2, UF3O8, UF4) diverted to clandestine activities is fairly well documented in Implementation of the NPT Safeguards Agreement in the Islamic Republic of Iran, Report by the Director General, November 10, 2003, (GOV/2003/71).

Absent the imposition of some extraordinary accountancy mechanism, this is a really, really bad deal.

For more information on accounting for nuclear materials, see Chris Eldridge, Protection, Control, and Accounting of Nuclear Materials: International Challenges and National Programs—Workshop Summary.


  1. j (History)

    You are correct—this is a bad proposal. First, Bob Joseph gave in on the India deal and now he is letting an Iranian UF6 capability slide. Had this been a proposal by the Clinton or Kerry Administrations, Republicans and their house organs at AEI and Heritage would be screaming bloody treason.

  2. mark gubrud (History)

    1. “depleted UO2… diverted to clandestine testing”?

    2. Any idea how much UFx would need to be diverted toward how large a clandestine centrifuge plant in order to make a bomb’s worth of HEU? I doubt that this could be easily written off as a process loss.

    3. If Iran can have clandestine centrifuges why can’t it have clandestine U conversion facilities?

    4. Why is Iran’s right to develop an indigenous fuel capability only something Iran “contends” it has under the NPT?

    5. Which is a really worse idea: a) The US and EU “let” Iran conduct some limited and monitored amount of indigenous U mining, conversion, enrichment and fuel fabrication, under Iranian acceptance of enhanced IAEA safeguards and adherence to the NPT, or b) another or a wider war?

  3. Jeffrey Lewis (History)

    1. I should have said “experiments” or “activities.” I’ll update that.

    2. To be specific, my worry about diversion centers on the possible use of diverted UF6 for clandestine development of centrifuges.

    But, now that you mention it, Esfahan has a throughput of 200 metric tons of UF6 per year. If Iran can claim 1 percent process loss above whatever the real number is (I haven’t seen a number I fully comprehend yet), Iran should be able to divert 2 metric tons per year—Iran could accumulate enough feed stock to enrich enough uranium for a bomb over a period of several years.

    I would regard that as a relatively low risk, all things considered.

    3. Iran might have clandestine uranium conversion facilities, although the technical problems associated with this plant suggest Iran remains in the very early stages of UCF development.

    4. Don’t be horrid. That’s Sanger’s language, not mine.

    5. I don’t see the point of asking a rhetorical question here. I don’t, obviously, see these as the only two options. You’d be better served by making that argument explicitly. A third option, for example, is merely to let Iran build a bomb.

  4. mark gubrud (History)

    Okay, I don’t generally mean to be horrid, and appreciate your letting me know if I am. Point taken on Sanger’s language, and your point about the possible diversion of modest quantities of UF6 for a clandestine development program, assuming Iran needs to conduct its centrifuge development program clandestinely. My assumption is that Iran will insist on, or at least claim a very high price for renouncing, its legal right to the full fuel cycle, but probably can be persuaded to both limit it and place it under adequate safeguards for a more reasonable price.

    I did read your post as uncharacteristically hardline and don’t share your concern that entertaining proposals for “face-saving” and entangling (in arms control) arrangements (even as bizarre as this one) is such a bad idea, partly because I remain unconvinced that Iran will find its conversion problems so terribly difficult to solve in the slightly longer run nor that they need to test-run the large facility in order to be making some progress on solving them (assuming they have already identified what some of those problems are); and partly because I doubt that Iran will be very interested in this idea, or that if it were, the US would be.

  5. matt tompkins (History)

    Rice is denying it now anyway, so either Sanger’s sources were a little over-eager or the administration isn’t ready to back such a proposal publicly…

  6. Bill M (History)

    What are the implications of shipping large amounts of UF6 back and forth across the world? Gases are hard to ship in bulk and easy to lose control of. Is an accidental release more or less guaranteed?

  7. AHM (History)

    This kind of creative thinking (letting Iran create UF6, then enriching it in Russia) is, as you argue, suboptimal, but it is still the best alternative available that has any chance of agreement on Iran’s part. The best assurance against Iranian nuclear weapons development is to keep Iran within the NPT and get it to ratify the Additional Protocol, keeping IAEA inspectors swarming around the country. Keeping its program public and open also preserves a later military option: Isfahan is, as you argue, a bottleneck, and could be taken out if necessary.

    As for diversion: A risk exists, but so does a risk of being caught. The Europeans are really playing a delaying game here, hoping in the long term that a more moderate regime will come to power with more modest nuclear aims or that Iran will become sufficiently socialized as to stop seeking weapons.

  8. Cheryl Rofer (History)

    I think AHM has the only reasonable analysis.

    I agree with Jeffrey that from both technical and arms control viewpoints this proposal is severely flawed. But buying time might be worth it. It also gives something to the Russians, who would like to develop a legitimate nuclear-supplier industry, including enrichment and reprocessing.

    Shipping UF6 shouldn’t be too much of a problem. Container technology is developed, and pressurized UF6 is a solid at room temperature, which makes it easier to handle.

    Shipping those quantities bothered me for a while, but the Soviet Union shipped large amounts of ore concentrate and yellowcake large distances, so the volumes (and costs) don’t matter if your goal is other than economic.