Jeffrey LewisPimp My Cascade

Bill Broad and David Sanger report that Iran is set to make a major announcement.

Iran is expected to declare in coming days that it has made a huge leap toward industrial-scale production of enriched uranium — a defiant act that the country’s leaders will herald as a major technical stride and its neighbors will denounce as a looming threat.”

I was holding off blogging about this, waiting for the Robot Economist to make good on his threat to photoshop an image appropriate for the caption “MC Ahmadinejad seyz the rims keep spinnin’ even after the centrifuges stop.”

But Steve Clemons calls and I have a little bit of a scoop.

The announcement—which has been bouncing about in the press—is likely to be that the Iranians installed a pair of 164 centrifuge cascades at the underground Fuel Enrichment Plant near Natanz.

(Okay boys, start painting the giant targets on the roof).

Anyway, that isn’t the scoop. The scoop is that, in advance of the February 24th deadline set by UN Security Council Resolution 1737 for Iran to “suspend … all enrichment-related and reprocessing activities, including research and development …” the Iran Atomic Energy Organization will declare that Iran has mastered cascade operations, has enough enriched uranium for current purposes and that it will place it’s cascades in warm standby.

I’ll write more on warm standby tomorrow, but today I want to focus on Iran’s technical progress—essential to understanding what a “warm standby” suspension might mean.

The claim that Iran has mastered cascade operations is a bit of political theater, as the story Broad and Sanger and another by Peter Beaumont in The Observer make clear.

Broad and Sanger are general in describing technical troubles suffered by Iran, simply quoting IISS’s Mark Fitzpatrick on delays. (You can read Fitzpatrick’s assessment, “Assessing Iran’s Nuclear Programme” in Survival 48:3, Autumn 2006, full text.)

Broad and Sanger do, however, report one useful nugget of information: that IAEA inspectors confirmed that Tehran successfully manufactured “parts for about 3,000 centrifuges…”

Whether those parts are up to snuff is another question. A couple of months ago, I noted that Iran has components for 5,000 centrifuges—but that many of those components are of poor quality:

… Iran probably doesn’t have enough components or the ability to manufacture them. Albright and Hinderstein cite “senior diplomats in Vienna” as claiming that Iran has “components for up to 5,000 centrifuges” and “other senior diplomats” suggesting that some components are of poor quality and that Iran has components for an additional 1,000 to 2,000 centrifuges.

Hard to know which components are scarce, although Mark Hibbs has been all over this story about the export of ball bearing preforms to Pakistan and the question of whether Iran can indigenously manufacture it’s P-1 centrifuge.

[Yeah, okay, pretty much igore the next five paragraphs and read “Bellows, Bearings and Rotors” instead.]

Paul Beaumont emphasized the issue of the bearings in his article in The Observer:

Instead, say experts, the break-up of the nuclear smuggling organisation of the Pakistani scientist Abdul Qadheer Khan has massively set back an Iran heavily dependent on his network.

A key case in point is that Tehran originally procured the extremely high-quality bearings required for the centrifuges’ carbon-fibre ‘top rotors’ – spinning dishes within the machines – from foreign companies in Malaysia.

With that source closed down two years ago, Iran is making the bearings itself with only limited success. It is the repeated failure of these crucial bearings, say some sources, that has been one of the programme’s biggest setbacks.

[As an aside … Beaumont also mentions a second issue —Iran’s lack of access to maraging steel—which is not precisely relevant. Iran’s centrifuges are the relatively crude P1 design, which uses aluminum. Of course, a shortage of aluminum er, maraging steel means that Iran has to use P1 design.]

Anyway the bottom line, at least for today, is that until Iran demonstrates that it can manufacture large numbers of high-quality centrifuge components, the claim that they will install 3,000 machines seems far-fetched.


  1. abcd

    Well, over the weekend, Iran announced – through its Fars news agency – that Iranian scientists had found an herbal cure for AIDS. No lie. See:

  2. yale (History)

    They do not need 3K machines nor continuous process in linked cascades to have a bomb in a hurry.

    —- This is not so much a prediction, as a Don’t-Be-Surprised.

    Israel is never going to stand for an Iran with fissiles, and will attack if they see no other way.

    The Bushehr NPP is finishing up and a “binary” moment is imminent. It may be targeted in early March.

    Pictured below is some of the 80 tons of VVER-1000 fuel Russia is scheduled to deliver in March if sanctions don’t interfere.

    This fuel is already purified with none of the tricky contaminants that can trouble further enrichment.

    It is ready for direct conversion to Uranium hexafluoride (UF6) centrifuge feedstock with no intermediate steps.

    Even more importantly, VVER-1000 fuel is enriched 88% of the way to 90% HEU.

    Iran’s centrifuges may process about 2 SWUs per year. An HEU implosion bomb requires much less than 20 kilograms of HighlyEnriched Uranium (HEU).

    It take 4k SWU to make 20 kilograms of HEU from natural uranium.

    Diverting less than 2% of the VVER-1000 fuel to create 20 kilograms of 90% HEU, requires not 4k SWUs, but only 500 SWU (with 3.3% enriched “tails”) And as pointed out before, it is sparkling clean.

    An SINGLE 164 centrifuge cascade (in simple batching mode), using VVER-1000 fuel, could create a bomb’s worth of HEU in 18 months.

    6 such cascades makes a bomb in 3 months, and 3,000 centrifuges is 12 bombs in a year.

    If the Iranians chose instead to build a gun-type bomb, light weight and narrow, easily carried in their existing missile designs, then the numbers would be:1 cascade = 3 years6 cascades = 8 months3k P-1s = 4 warheads per year=============================

    The Iranian will alway be able to build and conceal at least small cascades of centrifuges (that cat is out of the bag) so stopping the delivery of pre-enriched fuel is a no-brainer.

    In addition to blasting the fuel cycle sites, Israel would also hit Bushehr. This would stop the Russian fuel import.

    Once delivered, the LEU would be virtually indestructable. The fuel rods may get mangled, but the U-oxide pellets would be highly survivable. Tens of tons would be easily recovered from any possible attack.


  3. CKR (History)

    About a year ago, it seems to me that Iran was claiming that it would have those 3000 installed by, er, next month.

  4. Eric Hundman (History)

    Good post Jeffrey. I agree with the general conclusion that technical issues make any installation of 3,000 ‘fuges unlikely in the near term.

    However, I’m confused as to why the installation of two 164-centrifuge cascades is bouncing around in the news now. Paul reported on it almost two months ago (; did the story just get buried?

  5. Robot Economist (History)

    abcd – lol. What will those Iranian superscientists come up with next?

    yale – I could have sworn WER-1000 was a pressurized light water reactor, which means it can operate using LEU as opposed to HEU. Why would the Russians supply them with HEU if they didn’t need it and why are they handing over fuel that is already weapons grade (minimum 85% U235)?

    Dr. J – Sorry about the delay, the combination of flu and Superbowl eat up more time than I anticipated.

    I know plenty has been said about the tangible aspects of Iranian technical expertise, but what about the operational side? If I remember my “Uranium Enrichment and Nuclear Weapons Proliferation” correctly, the Iranians will also need to master the flow rate and pressure vs. RPMs equation before their centrifuges can be expected to put out more than 2-5 SWUs a year.

  6. Robot Economist (History)

    Dr. J – The picture is totally up now btw.

  7. Lisa Simpson

    Answer for Eric: The difference is that the previous story reported the installation of centrifuges at the small, above ground Pilot Fuel Enrichment plant where test operations were to be carried out. This new story refers to the installation of cascades in the underground facility that is designed to hold more than 50,000 centrifuges for commercial operations. This does not change their capability any more than installing the same additional centrifuges in the Pilot plant would have, but it indicates a (false?) confidence in their ability to build up to commercial operations of the larger centrifuge plant.

  8. Lisa Simpson

    Regarding yale’s comment, and Robot’s follow-up question, I think that yale’s post is designed to be inflamatory, and therefore leaves an incorrect impression. Robot is right, the VVER fuel is LEU (4-5 % enriched). What I think yale means in using the 88% statistic is not that the fuel is 88% U-235, but that 88% of the effort to enrich natural uranium to weapons grade has already been done. This like to be close to technically correct (I haven’t done the math) because it takes so much more effort (SWU) to get the natural, 0.7%, uranium up to the 4-5% needed to fuel Bushehr vs. going from 4-5% to 90%. It certainly is a confusing way to express it, and I think that was on purpose given the tenor of the post. In my opinion, diversion of the Bushehr fuel is low on my list of concerns about Iran, and is a red herring for the policy debate.

  9. Mark Gubrud

    Yale is not saying that the Bushehr fuel is 88% U235, but that it has already absorbed 7/8 of the SWUs needed to go from natural to 90% HEU, plus it is clean.

    Jeffrey’s story is that Iran will offer “warm standby” as a minimum complaince with the UNSC suspension order. That will probably satisfy Russia and maybe the Europeans, probably not Bush and certainly not the Israelis.

    Delivery of the Bushehr fuel would then put Iran where it wants to be: within a relatively short shot of getting enough fissile material for a bomb or a few.

    If all this is correct, it does look like we are fast approaching a crisis point.