How close is it?
The policy options available to the international community in its efforts to prevent Iran from obtaining a nuclear-weapons capability depend to a great extent on how close Iran is to reaching that goal. As Iran continued to forge ahead with uranium-enrichment operations, senior officials from the United States, United Kingdom, France, Germany, Russia and China gathered in New York in late September 2007 to discuss a possible new UN Security Council resolution. However, it seemed unlikely that there would be a consensus on imposing additional sanctions as long as Iran appeared to be cooperating with the International Atomic Energy Agency (IAEA) in answering questions about its programme.
Iran’s insistence that its nuclear programme is purely for civil energy purposes is viewed with scepticism by major powers. The Security Council has placed a series of sanctions on Tehran for ignoring its demand that the uranium-enrichment programme be suspended. In a speech to the UN General Assembly on 25 September, President Mahmoud Ahmadinejad said Iran would ignore the Security Council’s ‘political’ resolutions and would regard the issue as a ‘technical’ issue to be discussed with the IAEA.
The IAEA has also demanded the programme’s suspension, but as of late September, Iran had begun installing the last pair of 164-machine centrifuge cascades planned for now in the underground fuel-enrichment plant at Natanz, south of Tehran. Soon, all 18 cascades, or 2,952 centrifuges, will be operating as one module. Western intelligence estimates may need to be adjusted in accordance with Iran’s progress.
A key question is whether technological difficulties and manufacturing bottlenecks pose significant constraints. If the cascades were working round-the-clock at their intended efficiency, and if Iran decided to use them to produce highly enriched uranium (HEU), the nearly 3,000 centrifuges could produce one bomb’s worth (20–25 kg) of HEU in 9–11 months. For now, however, the number of centrifuges is less important than their performance. The key variable in determining how soon Iran could have a bomb’s worth of HEU is how long it will take to make the cascades work smoothly. At the extreme, Iran may be able to overcome glitches in one year, and thus be able to produce a bomb sometime in 2009, if it decided to go for broke and to expel IAEA inspectors. A longer timeframe is more likely: the US Central Intelligence Agency has consistently posited a range from the beginning to the middle of the next decade. In the most plausible break-out scenario, Iran would produce a stockpile of low-enriched uranium under IAEA verification, and only then expel the inspectors and enrich this stockpile to weapons grade in 5–8 weeks.
This would not by itself mean that Iran could produce a weapon. All the timeline estimates assume, without much supportive data, that Iran would also be able to acquire the technology to make a deliverable weapon out of the fissile material. The only reported evidence that Iran has worked on weaponisation came from a computer hard drive turned over by an Iranian defector in 2004 showing various stages of missile design plans. There is also a 15-page document in Iran’s possession – acquired unsolicited from the A.Q. Khan black-market network, Iran says – which describes how to cast uranium metal into hemispherical forms.
As of the end of August, 12 cascades at Natanz were being fed uranium hexafluoride (UF6). They were operating simultaneously, in parallel, in accordance with the normal configuration for power- reactor fuel. IAEA inspectors believe the centrifuges were spinning at or near the supersonic speed necessary for enrichment. The inspectors do not have visual access to the individual machines or to the control room, however, so they cannot say this with certainty. The IAEA monitors the UF6 input and can take swipe samples of the enrichment level. The cold traps where the output is collected are under IAEA seal, but its inspectors will not be able to measure this output until Iran withdraws the product, which might not happen until winter.
Technical difficulties
According to the IAEA’s 30 August report, 690kg of UF6 had been used in the cascades since February 2007. This is well below the expected usage rate for a set of cascades of this size, leading analysts to conclude that Iran is experiencing technical difficulties in running them, or has chosen for other reasons deliberately to slow the pace of work. Although the centrifuges are spinning constantly, they must have been operating much of the time since February with only inert gas. Otherwise the amount of UF6 used would have been much higher. Being able to operate the cascades with UF6 24 hours a day, 7 days a week for weeks on end is a key technical hurdle. Iran apparently has not yet reached this level of mastery of the enrichment process. There is evidence, however, that the cascade performance has improved. Each time the IAEA inspectors have visited Natanz – they are allowed to make some unannounced inspections – the cascades were being fed UF6. It remains to be seen whether the quantity of UF6 being fed into the machines at present is sufficient for production, or if this is still part of a process of testing the equipment.
This leads to the assessment reached by David Albright and Jacqueline Shire of the Washington-based Institute for Science and International Security on 30 August that ‘Iran likely has managed to learn how to operate individual centrifuges and cascades adequately. However, it still may be struggling to operate a large number of cascades at the same time in parallel.’
Iran chose to install the 18 cascades underground at Natanz before demonstrating that even one cascade in the above-ground pilot plant was operating smoothly. The normal procedure would have been to operate small cascades of 10 and 19 machines continuously and to push them to breaking point to assess stress levels before operating larger cascades. This appeared to be Iran’s earlier intention. But when the enrichment operations resumed in 2006, normal start-up procedures were omitted, and the full 3,000-machine module is now in effect a pilot plant. The reasons for taking this approach appeared to be political – to put in place as many centrifuges as quickly as possible in order to establish new facts on the ground that would put Iran in a better bargaining position. An alternative explanation cannot be ruled out: that Iran might have learned more about centrifuge operations at an unreported facility somewhere else. However, no evidence has surfaced pointing to a parallel, covert enrichment facility.
In the initial stages, Iran was experiencing centrifuge failure rates reportedly reaching 40%. Today, based on discussions with plant operators, IAEA inspectors believe the failure rate is far below that. They do not know for certain, however, because they do not have access that allows them to monitor individual machines.
The UF6 being fed into the cascades was domestically produced at the uranium-conversion facility at Esfahan, where 330 tonnes had been produced as of 14 August. The domestic UF6 is contaminated to some extent with other heavy metals present in the uranium ore but, according to officials who have seen the sampling data, this does not present a problem for the time being. The contamination, which is not as significant as had been reported earlier, will degrade centrifuge operations over time, reducing their lifespan to a third of the normal ten years. However, if commercial production is not Iran’s intention, this would not matter.
For now, Iran appears content to stop at the 18-cascade module. According to inspectors, additional bays in the cavernous underground facility have not been prepared for the installation of any more cascades. This is consistent with Iran’s public pronouncements. Although Iran still adheres to the goal of a 54,000-centrifuge plant, spokesmen consistently speak only of 3,000 as the interim goal. This was originally to have been met in December 2006, but implementation was repeatedly delayed. The pace of installation fluctuated, and installation of the final two cascades did not begin until early autumn.
Production limit in question
The upper limit of Iran’s centrifuge production is of critical importance to Western policy decisions. IAEA Director General Mohamed ElBaradei argues that it is already too late to stop Iran from acquiring uranium-enrichment technology. In his view, the international community should fall back to a second-best position of trying to keep Iran’s programme from growing larger. He contends it would be better to strike a deal to keep Natanz limited to 3,000 centrifuges than to see Iran expand the facility by several times its current size. However, if Iran could not realistically make more than 3,000, such a deal would be an empty bargain. Conversely, if Iran had the capacity to make many more than the number now installed at Natanz, then an air strike that destroyed the Natanz facility would not cripple Iran’s enrichment programme for very long.
The pace of installation led ElBaradei to say in May 2007 that if it was maintained, Iran would be on track to have 8,000 centrifuges running by December. He gave no basis for that figure, leading analysts to assume that it was a simple extrapolation. However, the estimate appears to have been based on an understanding of how many centrifuges Iran could assemble with the components it had to hand. In 2005, when the IAEA had access to Iran’s declared workshops, it conducted an inventory of centrifuge parts and counted approximately one million components. According to an informed official, Iran then had enough parts for approximately 10,000 machines.
The degree to which Iran is able to manufacture parts indigenously is central to an assessment of its centrifuge- production limits. Of the sets of components it inventoried, the IAEA did not know what proportion were imported or domestically produced. The agency, which no longer has access to the components or to the workshops, is not able to verify Iran’s claims that it is able to manufacture all the 92 components that go into a P-1 centrifuge (a first-generation Pakistan-designed gas centrifuge). Western intelligence agencies had earlier assessed that limitations in some key raw materials and the difficulty of producing certain sophisticated components presented bottlenecks in Iran’s centrifuge programme. Iran may not be able to produce the maraging steel necessary for producing certain centrifuge parts. IAEA officials now believe, however, that Iran has managed to import enough maraging steel (including from the UK in the 1990s) for its centrifuge purposes. Similarly, although Iran may not be able to produce its own ring magnets, it was able to import a sizeable quantity through an Asian intermediary before export controls were tightened. Whether Iran can also produce the delicate bottom bearings on which centrifuges spin is also unknown.
In 1995, Iran received design plans from the Khan network for the second-generation P-2 centrifuges. Khan’s chief foreign associate, B.S.A. Tahir, told investigators that the network also delivered three complete P-2s to Iran as samples, although this has not been verified. The IAEA did not believe Iran’s statements that it did not pursue any work on the P-2 design from 1995 to 2002 – and Ahmadinejad’s boast in April 2004 about research on advanced centrifuges belies this official claim. It would not be surprising to learn that Iran has a pilot P-2 plant – Iran has announced that it will ignore the routine safeguards obligation to make early declarations of nuclear facilities. Some degree of clarification about the P-2s may emerge if Iran provides answers about both the P-1 and P-2 development programmes in accordance with the Iran–IAEA work plan announced on 21 August. Iran and the IAEA have begun discussions about the agency’s questions, and the initial set of answers is due by 15 October, with follow-up clarifications scheduled for November. However, it is doubtful whether Iran will reveal anything about the current status of the programme: it can be expected to argue that the work plan is limited to questions dating from 2003, when the IAEA first enumerated Iran’s safeguards failures.
The IAEA’s questions did not stop as of that date, however. Indeed, the more the inspectors learned from their investigations of the Khan network in other countries, the more questions they had about Iran’s nuclear imports. Moreover, if the agency is ever able to close the books on Iran’s past activities and to verify any future deal that would set limits on Iran’s programme, it would need to obtain a clear baseline picture of Iran’s programme that runs up to the present. As the international community attempts to assess the present and future capacity of Iran’s nuclear programme with a view to determining future policies towards Tehran, there are still more questions than answers.
