Jeffrey LewisIran Panic Induced By Lousy Reporting

Another IAEA DG Report on Iran. Cue the panic.

Iran Has More Enriched Uranium Than Thought!

Iran understates uranium stocks to IAEA!

Iran holds enough uranium for bomb!

Only the exclamation points are mine. These stories are scandalous. Daniel Dombey’s father is a physicist for god’s sake!

(Joby Warrick, on the other hand, took the relatively balanced view, with Iran Easing Aspects Of Nuclear Program in the Washington Post, February 20, 2009, A14. I wonder whether his editors will give him a hard time for not writing the same thing everyone else has written.)

Obviously, Iran would have to further enrich its stockpile of low-enriched uranium to make a bomb, which would be detected (at least at Natanz). We’ve been over this before.

The bigger misconception is that Iran has somehow misled the IAEA about its stockpile of low enriched uranium. The IAEA report does not suggest that Iran failed to declare material. Here is the actual paragraph from the report:

3. The Agency has finalized its assessment of the results of the physical inventory verification (PIV) carried out at FEP on 24–26 November 2008, and has concluded that the physical inventory as declared by Iran was consistent with the results of the PIV, within the measurement uncertainties normally associated with enrichment plants of a similar throughput. The Agency has verified that, as of 17 November 2008, 9956 kg of UF6 had been fed into the cascades since February 2007, and a total of 839 kg of low enriched UF6 had been produced. The results also showed that the enrichment level of this low enriched UF6 product verified by the Agency was 3.49% U-235. Iran has estimated that, between 18 November 2008 and 31 January 2009, it produced an additional 171 kg of low enriched UF6. The nuclear material at FEP (including the feed, product and tails), as well as all installed cascades, remain under Agency containment and surveillance. [Emphasis mine.]

Important phrases:

within the measurement uncertainties normally associated with enrichment plants of a similar throughput.

remain under Agency containment and surveillance.

Nothing is f*cked here, Dude.

Within the Measurement Uncertainties

In order to know that the IAEA discovered more LEU than Iran “declared,” one needs to look at the previous IAEA DG Report. The DG stated “As of 7 November 2008, the total amount of UF6 fed into the cascades since the beginning of operations in February 2007 was 9750 kg, and based on the operator’s daily accounting records, Iran had produced approximately 630 kg of low enriched UF6.”

That’s the discrepancy: 839 kg minus 630 kg.

Fascinating. Anybody think that all these reporters carefully compared the two reports? Obviously, not.

This is spin — though not for a bad cause. More on that in a moment. Someone decided to explain they found a bunch more LEU in order to put a little pressure on the Iranians.

[Correction: My cynicism got the better of me here. I hope to write a fuller correction soon.]

Now observe the phrase: “based on the operator’s daily accounting records.” A colleague and I have been talking about this since the last report — how accurate are the operating records? We noticed, for example, that the Iranian operating records claimed that Iran was enriching to 4.9 percent, while environmental samples indicated the level was only up to 4.0 percent. Now, based on the PIV, the number turns out to be more like 3.49 percent. I guess we are the only people who have been reading the footnotes in the damned IAEA reports.

(By the way, it is worth noting that there is almost exactly the same amount of U235 in 839 kg of 3.49 percent LEU and 630 4.9 percent LEU.)

The IAEA makes clear that the discrepancy between the stockpile estimated from daily accounting records and the physical inventory is “within the measurement uncertainties normally associated with enrichment plants of a similar throughput.” As a diplomat tells Heinrich, “faulty estimates that can arise from complexities in the phased enrichment process, not to any maneuver to divert LEU out of sight.”

This is going to kind of frighten you, but large scale industrial processes are not measured in bomb units, even though that would be awesome. This same reality, in a slightly different context, gives rise to the wonderfully titled Material Unaccounted For (MUF).

Think about this way: A 164-centrifuge cascade consumes about 70 grams an hour. If you have 18 cascades, you are consuming 30 kilograms of UF6 a day. (.07 kg x 24 × 18 = 30) Let’s say that you are wrong, by one percent a day, regarding how much of the 30 kg ends up in the product as opposed to the tails. You are underestimating your production by 300 grams day — within two years, you will underestimate your production by more than 200 kg. (This is merely expository. Note that the Natanz PIV covered February 2007-November 2008, during which the number of operating centrifuges grew from 0 to 24 with uneven operating patterns.)

Remain Under Agency Containment and Surveillance

This isn’t nearly as terrifying as it sounds because accountancy is just one safeguards tool.

Containment and surveillance is another. Again, as the IAEA points out, “The nuclear material at FEP (including the feed, product and tails), as well as all installed cascades, remain under Agency containment and surveillance.”

In other words, even if you aren’t sure how much he weighs, Elvis is still in the building.

So, why the fuss? Because the IAEA only does a PIV every year or so at Natanz. Containment and surveillance are wonderful, but it is nice to occasionally check the material itself.

As a “diplomat” tells Mark Heinrich:

But the diplomats said the verified LEU figure was based on an inventory check that inspectors perform only once a year.

In theory, this means there is a risk that any smuggling of enriched uranium out of Natanz for use at a secret site might not be noticed for some time.

U.N. inspectors are discussing with Iran how to improve its operating records to prevent any repeat of such large differences in accounting in future, the diplomats said.

I actually agree that we need to improve our monitoring of Iran’s activities. So, I think it is helpful to point out how inadequate current arrangements are even if the Iranians are complying with their obligations.

But no need to stir up panic.


  1. Steven Dolley (History)

    No need to panic, agreed, but the IAEA report still raises issues to be addressed. It shows that MC&A at centrifuge enrichment plants can potentially be wrong by large margins. That will likely require attention at some point.

  2. Yale Simkin (History)

    Breakout risk:

    Based on the Nov 18, 2008 to Jan 31, 2009 output of the 3,936 centrifuges:

    Current Production Rate:
    6.76 SWU/day

    2.31 kg/day of 3.49% LEU (hex)

    0.63 SWU/year/centrifuge (quite low, but good enough)

    Total LEU (hex) as of Feb 20, 2009:

    1010 kg (as of 1/31/09) + 173 kg (2/20/09)
    = 1,183 kg LEU (hex)

    Total accumulated SWU = 3,460 SWU

    To enrich 16.5 kg 90% metallic HEU (15 kg for core with 10% reusable waste)
    from 1,183 kg LEU (hex) = 650 SWU (w/1.66% tails)

    Days from Breakout to Core =
    650 swu / 6.76 SWU/day = 96 days

    If Iran broke out today, they would have a metal core in the last week of May, 2009

    1,467 more centrifuges were under vacuum as of 1/31/09.
    If these are now running hex, then the breakout time is 70 days, or last week of April.

    Before the end of 2008 (exactly as I predicted 2 years ago – and the NIE estimated 2015) a binary moment occured.
    Before then, Iran COULD NOT BUILD A BOMB , unless they diverted the Russian fuel. Now the question is WILL IRAN BUILD A BOMB?
    Before the end of 2008 the LAWS of PHYSICS prevented an Iranian bomb. Now it is only the terribly flawed blunt tool of international diplomacy, carrot and stick, that stands in the way of hell.

    Remember this LEU is produced for a “PEACEFUL” nuclear reactor. No such thing.

    I am disgusted.

  3. MWG

    Here is a recent review of safeguards approaches for centrifuge plants. At the end it identifies weaknesses of current approaches and potential measures to address them.

  4. Mark Konrad (History)

    The demonisation of Iran and the manufacturing of hysteria has been an ongoing project in the Western government and media entities, as you know Mr Lewis. I actually suspect that the Iranians knew their stockpile was a bit larger than the IAEA last reported. They wanted the inspectors to make that discovery. Now the Iranians can say “See, we’re not up to anything nefarious. If we were in weapons building mode we would have never underestimated our own supply by that margin. We would have been carefully monitoring our production and discreetly hoarding our excess product in a safe place far away from inspectors. But we haven’t done that. We’re all friends here !”

  5. mike

    there also appears to be a difference in dates, 7 Nov vs 17 Nov

  6. Ael (History)

    What is the current status of technical assistance that the IAEA is supposed to render to Iran?
    I (vaguely) recall that it had been suspended, or something. The current report does not appear to mention what assistance the IAEA is providing to (or withholding from) Iran.

  7. AWR (History)

    Great analysis. Reminds me of the button on the Hitchhiler’s Guide to the Galaxy: “Don’t Panic!”. The biggest problem is not what the Iranians are doing, it’s what they are NOT doing, raising these kinds of suspicions. And even though the “red line” of HEU production has apparently not been crossed, reasonable suspicions still remain. Maybe more dialogue can cast this doubt away? Only Iran and maybe the IAEA can know for sure.

  8. Jeffrey Lewis (History)


    Look at that! Wonder if that is a typo …

    Good catch! 10 days would actually be enough to account for the difference in feed amounts.

    Hmmm. This needs to be looked at carefully.

    You people are just a joy to have as readers. A real joy.

  9. Yale Simkin (History)

    Jeffery wrote:
    10 days would actually be enough to account for the difference in feed amounts.

    I’m not sure if that is the issue.
    There may very well have been different audit dates for the feed inventory. The problem is the output.

    The Iranian paper estimate was for 630kg of LEU through 1/31/09.

    The measured quantity was 839 kg, or 209 kg more LEU

    They are producing 2.3 kg/day of LEU from the cascades.

    10 days is 23 kilograms of LEU not 209 kilograms.

    I don’t disagree with the view that the difference is just a need for improved Iranian estimation skill.

    I am not sure where the comparison U235 content of 630kg of 4.9% vs. 839 kg of 3.49% LEU comes from.

    The only reference I can find in the older report to enrichment levels is the boilerplate statement that the LEU is <5% enriched, the same text as the current report.

    The real problem is the +SQ worth of contained U235.

  10. Purple Clarinet

    What about the recent reports obviously fed by ´anonymous IAEA diplomats´ that Iran has used up its stockpile of yellowcake acquired from South Africa during the 70’s? Interestingly Iranian officials came out quickly to deny this story and maintained that Iran has indeed enough mining/refining capacity to satisfy any potential needs of their nuclear industry.

    Are there any reliable estimates about Iran’s UO2 stockpile existing beyond the relatively well known ´South African deal´ of the Shah Era?

    Assuming that Iran has already exhausted the fairly large amount of UO2 imported from SA: Could this be an indicator for a substantially larger aggregate production of UF4 and UF6 than declared to the IAEA?

    Admittedly I’m somewhat surprised that there is that much of obfuscation and confusion about a basic fact like Iran’s actual stockpile of UO2.

  11. ataune (History)

    The truth of the matter is that the world is changing and media outlet such as FT, NYT or WaPo carry less and less influence in shaping the opinion of the elite and the public in the world.

    Most of the western world enlightened elite knows that strenghtening the NPT regime and extending its arms to everywhere in the Middle-East is in the interests of the world in general and countries like Iran in particular.

    For a long time now, the before mentioned media, and many others, have promoted an agenda of undermining the NPT regime by spreading innuendos, half-truth, semi-truth or plain lies.

  12. Bill Arnold (History)

    To my paranoid (and amateur) mind, Iran has appeared to be attempting to increase the uncertainty in the LEU production estimates for the last year or two. Once the uncertainty is large enough, they are a defacto nuclear-armed state, whether or not they go the whole way.

  13. pkr (History)

    Thanks for that post! Gonna show it to my editors when they next time question me why I haven’t got the story (as last time when the NYT already blew out that Iran has enough uranium for a bomb only to be proven wrong by careful analysis carried out by our friends at ISIS).

    I really have a hard time understanding why most U.S. media are so obsessed with the breakout capability. Even a senior U.S. diplomat certainly not suspect of playing down Iran’s nuclear ambitions last year called it politically irrelevant if Iran has reached this threshold or not. A state that is able to manufacture thousands of centifuges and run them close to their nominal capacity for a prolongued period of time will reach this point sooner or later. That’s why DG ElBaradei repeatedly said that mastery of the fuel cycle makes you a virtual nuclear weapons power – and Iran obviously ha achieved this status.

    It is a political decision of the Iranians if they proceed to building weapons or not. So neither the U.S. nor others in the international community have much of a choice left: the only way to prevent Iran from going nuclear is to try to influence the decisionmaking. The means for that clearly is diplomacy, like it or not. And I have not yet seen any suggestion for a new approach that would undoubtedly promise to work better than the current double track strategy. The question rather is which new incentives could convince Iran to give in and which new threats should be made to bolster the effort (France, the U.K. and Germany in the P5+1 format are ready to tighten sanctions in areas that really would harm Iran’s economy).

    My only objection is that insisting on that the Iranians have to stop their enrichment activities before we sit down with them at the negotitimng table will not help to break the stalemate. This approach was designed at a time when Iran was running one cascade with 164 centrifuges. In the current stage of Iran’s program the best result to be achieved is a maximum of transparence through the implementation of the Additional Protocol.

  14. mark hibbs (History)

    We did a piece on the LEU estimation issue. The Iranian discrepancy was at the top of the article.

    But the explanation why and the context of the matter was also in there. We were told on good authority:

    1.) the problem came up because of an error in Iran’s calculation of the distribution of tails and product in the plant’s cumulative output.

    2.) It did not suggest there was any uncertainty or error in estimations of the actual separating power of the centrifuges at FEP, as represented by the data provided by Iran during the DIV. In fact, and most or all of the mass media accounts of the affair missed this, the PIV showed that Iran had very accurately predicted how much feedstock had been processed at the plant.

    3.) Because the PIV in November was the first one carried out after Iran had generated a large amount of enriched product, it was the first time that a discrepancy in this magnitude arose.

    4.) The IAEA isn’t overly concerned about the Iranian miscalculation, but has told Iran that the IAEA would have more confidence in Iran’s estimates once Iran had rectified the errors in its formula.

    Jeffrey is correct that there has been a lot of misleading and tabloid-type reporting of this issue, with reporters itching to write that “Iran’s got a bomb’s worth.”

    To be fair, some of the press reports the last couple of days, for example the article in the NYT on Friday, did say that the discrepancy was not necessarily a big deal.

    My instincts tell me that between now and the next PIV, when the dust settles on this the Iranians will follow the IAEA’s advice and refine their math. Iran already has a pile of contentious problems with the IAEA and it doesn’t need to add to these by stubbornly refusing to correct something which the PIV clearly showed they got wrong.

  15. isb

    Just wanted to point out that the ‘bad reporting’ is actually based on an ISIS analysis of the IAEA report:

    “This total includes an additional 209 kilograms of low enriched material over what would have been expected based on the November 2008 IAEA report. The increase is the result of the IAEA conducting a so-called physical inventory verification (PIV) of the actual enriched uranium stock at the Natanz enrichment plant and discovering a higher amount than expected. The LEU estimate in the previous IAEA report was based on an Iranian calculation of the LEU product, which turned out to be low by approximately 30 percent.”

  16. hass (History)

    ISIS is complicit in the scaremongering. How many times has Albright been quoted calculating how “close to The Bomb” Iran is based solely on a theoretical calculation of uranium atoms it has? Based on that way of calculating things, both Niger and Malawi are “virtual” nuclear super-powers who “could” one day decide to “break out” and enrich their raw uranium at “secret” facilities whose existence cannot be disproved. Booga Booga!

  17. kme

    There is an incentive for states under IAEA monitoring to be biased towards under-estimating output rather than over-estimating. If they under-estimate, then the audit will find additional LEU – no big deal. If they over-estimate, then the audit will find missing LEU – and that is a big deal indeed.

    So with honest intentions, given the choice, one will always err on the side of underestimating.

  18. John Bragg (History)


    When Niger or Malawi get caught redhanded with a secret nuclear weapons facility at whatever the local translation at Natanz is, then we’ll worry about their raw uranium stockpile

  19. hass (History)

    John Brag- you mean the “secret” nuclear facility in iran for enriching uranium that the iranians had announced on national radio ten years earlier and which the IAEA said has no relationship to a weapons program? Anyway We can’t wait for the secret Malawian nuclear sites whose existence can’t be disproved to be discovered (cue Rice) because it would be too late by then.