Jeffrey LewisIran's UF6 Is "Crap"

Regular readers know that Iran’s uranium conversion facility (UCF) at Esfahan (Isfahan) stinks—a fact Paul first reported in the October 2004 edition of Arms Control Today and which forms the basis for IC judgements about Iran’s proximity to a nuclear weapons capability.

Nuclear Fuel and Reuters (hat tip SQ and GSN) have written four articles in the last month detailing problems with the UCF. Francois Murphy and Louis Charbonneau with Reuters have the best quote:

“The UF6 is crap,” said a Western, Vienna-based diplomat, who follows Iran’s nuclear case closely.

The three Nuclear Fuel stories (Platts asked me to remove the .zip file … I suspect my posting was fair use, but I was moved by Ms. Margaret Ryan’s usage of all caps to indicate that I did NOT have permission), all by Mark Hibbs, are less colorful, but provide loads more detail.

Hibbs reports that Iran is unable to remove metallic impurities from yellowcake (U3O8). Impurities will impede enrichment and may clog cascade piping or collect on surfaces, crashing unbalanced centrifuges.

Apparently the Iranians lacked the sweet skills to execute the Chinese blueprints—something the Chinese told the Iranians after cutting off assistance. Iran has subsequently taken an alternative, less sophisticated route to purifying U3O8:

In 1990, China had agreed to build UCF in Iran as a turnkey project. When in 1997 the U.S. had persuaded China not to build the plant in Iran, China nonetheless sold Iran blueprints for the conversion plant. Iran built the plant according to Chinese specifications between 1998 and 2003. Without additional hands-on assistance from Chinese engineers, however, Iran failed to master the U3O8 purification process.

In 2004, Iran responded to the difficulties by removing mixer-settlers called for by the Chinese process, and replacing the equipment with pulse columns.

The pulse columns are of Iranian design and were tested at the Tehran Nuclear Research Center (TCNC), the leading center of Iran’s uranium conversion research and development program during the 1990s and before UCF was built. During the R&D program for the pulse columns, Iran first test operated columns made of glass before it tested, and then began installing at UCF, columns made of metal. Vienna officials said Iranian experts told the IAEA that the Chinese purification process was “too complex” to master without outside assistance.

Hibbs also reported that Iran has suffered uranium process losses of about 40% and has not been able to demonstrate operation of the plant at even half of its declared throughput of 200 metric tons U as UF6.

One weird thing: One of Hibbs’ articles says the opposite: That Iran had trouble with pulse columns and switched to mixer-settlers. IAEA DG Mohamed ElBaradei told the Board that Iran moved “from mixer settlers to pulse columns” because the former “was technically and mechanically complex and more difficult …” But this is way outside my expertise, so further clarification is appreciated.

Update: The original post had a sweet illustration from the New York Times, but a reader noticed the graphic—however handsome—was wrong. Damn, that’s metaphoric.


  1. mark gubrud (History)

    One reason for starting work on a difficult engineering project is so that you can run into the problems you will run into and then solve them. The Iranians have about five years to work out their problems and get some kind of UF6 production that isn’t “crap” going before the Natanz facility will be ready for full-scale operation in time to get HEU production up and running on something like the schedule suggested by the NIE (first half of next decade) and by any commonsense reading of publicly-available information about Iran’s overt program.

    Then again, if one is a believer in covert Iranian nuke plants, one could as well discount all of this as disinfo. I don’t, but I also wouldn’t denigrate Iranian capabilities to the point of suggesting they’ll never be able to make the bomb at all (if they so choose). The available facts clearly support the consensus view that they will probably be able to do that by 2015 at the latest, and possibly quite a bit sooner if we keep pushing them into it.

  2. anon.

    ok, they have a gas problem. and left alone, their program will take longer. so this will give diplomacy more time, more toward the upper ceiling of the iiss estimate. and time is a valuable currency in international relations. so we are not in a hurry. and maybe both, the technical problems and diplomacy, will convince them to give it up.

    but can we count on their technical problems to convince them? should technical problems drive diplomay?

  3. Cheryl Rofer (History)

    No, we shouldn’t count on technical problems to convince them, but the problems do give us more time for diplomacy.

    A couple of clarifications: the impurities in the UF6 are not “metallic” in the sense of little particles of metal. They are just as gaseous as UF6. But they are lighter, and it is the lighter U-235 that is desired as the product. So they will come out with the product (in much larger amounts) and will have to be removed, unless they chemically react with the steel and unbalance the centrifuge rotors, which will ruin the centrifuges.

    I’m a bit surprised (see Mark Hibbs’s articles) that the pulsed columns, mixer-settlers, whatever, aren’t at the U3O8 production plant. Or maybe there are some there, and they’re not working very well.

  4. Tom (History)

    “their program will take longer. so this will give diplomacy more time”

    Just to play devil’s advocate for a moment…

    The converse is also true in the sense that the longer diplomacy takes the more time they have to work the kinks out in their programs.

    If diplomacy takes long enough the Iranians could simply present the world with a fait accompli program.

  5. ramin safi (History)

    this is a question: the 37 tons yellowcake was from a uo2 import by iran, right? if so, and the uf6 is in fact ‘crap’,then does that mean they have lost their chance at producing uf6, this time around or could they still remove the impurities?