Iran to Mass-Produce Improved Centrifuges

Peter Crail

Iranian officials announced last month that Iran would begin mass-producing a second-generation centrifuge in the coming months, a step that could lead to an increase in the rate at which Iran enriches uranium.

Centrifuges spin at high speeds to increase the concentration of the uranium-235 isotope to levels of around 4 percent for nuclear fuel or to much higher levels for possible use in nuclear weapons. Iran’s second-generation machines are believed to be capable of enriching uranium about 2.5 times faster than the centrifuges Iran is currently operating. (See ACT, November 2007.)

In a CBS Evening News interview that aired April 13, Atomic Energy Organization of Iran (AEOI) chief Ali Akbar Salehi said that Iran is considering installing more-advanced centrifuges at its commercial-scale uranium-enrichment plant at Natanz. A February International Atomic Energy Agency (IAEA) report stated that the Natanz plant has about 8,600 centrifuges installed, although only about one-half are operating. The agency noted in a 2004 report that Iran intended to install more than 50,000 machines at Natanz. However, Salehi told CBS that Iran may expand that number to as many as 60,000 in order to provide fuel for its first nuclear reactor at Bushehr as well as any additional nuclear power plants that Iran may build.

Iran’s Russian-built Bushehr plant is expected to begin operations in August. Russia has already provided the first load of fuel for the reactor. Moreover, the proprietary specifications for that fuel are owned by the Russian state-owned nuclear conglomerate Rosatom. A Russian diplomat said in April 2009 that he doubted that Rosatom sold Iran the right to make this fuel. (See ACT, May 2009.)

Aside from the Bushehr plant, Salehi said the second nuclear power reactor Iran plans to construct, at Darkhovin, will be finished “probably 10 years from now.” Iran provided preliminary design information for the plant to the IAEA last September, and construction is not believed to have begun.

Despite Iran’s lack of near-term need for enriched uranium to fuel its power reactors, the 1970s-vintage centrifuges it has been using, called the P-1, are known to be inefficient. (See ACT, December 2009.) Pierre Goldschmidt, former IAEA deputy director-general for safeguards, said in an April 16 e-mail that, because 50,000 P-1 centrifuges would be insufficient for one annual reload of the Bushehr reactor, it would “make a lot of sense” to move to more-advanced centrifuge designs to produce fuel if Iran were to produce the fuel domestically rather than importing it.

He added that the possible expansion of the Natanz plant to house additional machines would not have an impact on the way the plant is safeguarded.

Iran acquired the P-1 designs in the 1980s from the nuclear smuggling network led by Pakistani nuclear official Abdul Qadeer Khan, who had stolen the design from the enrichment consortium Urenco while working in the Netherlands in the 1970s.

Iran’s second-generation centrifuges are based on the P-2 design, which Khan also sold Iran in the 1990s. Because of Iranian difficulties manufacturing key P-2 components using maraging steel, however, Tehran modified the design to use carbon fiber. According to an Institute for Science and International Security (ISIS) report on April 13, centrifuge rotors using carbon fiber spin faster than those using maraging steel. Iran is still dependent on foreign sources of carbon fiber, and the primary manufacturer in Japan tightly controls such goods, ISIS President David Albright said in an April 19 interview. Carbon fiber materials that may be of relevance to a centrifuge program are subject to international controls, including UN sanctions against Iran.

Iran’s second-generation centrifuges, which it began testing in January 2008, include various P-2-based models referred to as the IR-2, the IR-2m, and the IR-3. It is unclear which model or models Iran intends to mass-produce.

During April 9 festivities marking Iran’s fourth annual “National Nuclear Day,” Iranian officials unveiled what they called a third-generation centrifuge design capable of enriching uranium six times faster than the P-1. The ISIS report says the third-generation model is “theoretically capable” of achieving this output because the rotor assembly is twice as long as that of Iran’s IR-2 centrifuge.

Salehi told reporters that Iran would begin testing the third-generation machines with uranium hexafluoride, the feedstock for uranium-enrichment plants, in the coming months and that it may be a year before a cascade of linked centrifuges is tested.

In addition to working on more-advanced centrifuges to increase enrichment capacity, Iranian officials also claimed in April that Iran would begin constructing one or two additional enrichment facilities this year. (The Iranian calendar year began in March.)

Last November, Iran announced that it would build 10 additional enrichment plants and that construction on two of them would begin this year. (See ACT, December 2009.) Iran is continuing to expand its Natanz plant and is constructing a second, smaller facility called Fordow near the city of Qom.

During his April 13 interview, Salehi said that construction would begin at one or two of the sites selected for these facilities, depending on a decision by the Iranian president. He indicated that it takes about four years to construct such a facility.

Although the IAEA has stated that Iran is legally obligated to provide design information for such facilities once a decision has been made to construct them, Salehi reiterated Iran’s claim that it is not required to do so until six months prior to the introduction of nuclear material. Tehran’s interpretation is based on a pre-1992 version of a safeguards provision regarding the time frame for informing the agency of the construction of nuclear facilities. Iran agreed to the newer version in 2003, but declared in 2007 that it would revert to the older requirement.

Assessing Timelines to a Weapon

As Iran takes steps to improve its enrichment capabilities, U.S. officials in April provided assessments for when Iran may be able to use those capabilities to build a nuclear weapon.

Testifying before the Senate Armed Services Committee April 14, Defense Intelligence Agency Director Lt. Gen. Ronald Burgess said “the general consensus” is that “we’re talking one year” as the time period in which Iran could produce enough highly enriched uranium (HEU) for one nuclear weapon if it decided to do so. Burgess added the caveat that the estimate was made without knowing “the exact number of centrifuges that we actually have visibility into.” A 2007 National Intelligence Estimate (NIE) said the intelligence community did not know whether Iran had made the decision to produce nuclear weapons, and that judgment “still stands,” Burgess said.

An annual U.S. intelligence report submitted to Congress in March similarly said it was uncertain whether Iran would decide to build an nuclear weapon. Unlike the report from last year, however, the 2010 document did not include reference to the 2007 NIE judgment that Iran had halted work related to weaponization in 2003. (See ACT, December 2007.)

Burgess indicated that an updated NIE is being produced but that the intelligence community has not yet made a decision on when it will be finished or released.

Secretary of Defense Robert Gates recently questioned whether it would be possible to verify whether Iran decided to make a nuclear weapon once it had all of the capabilities. “If their policy is to go to the threshold but not assemble a nuclear weapon, how do you tell that they have not assembled?” he said in an interview with NBC’s Meet the Press broadcast April 11.

Iran’s declared nuclear facilities and its known stores of enriched uranium are currently under IAEA safeguards, but Iran’s uranium mines are not under such monitoring. It is unclear whether Iran maintains in secret the additional facilities required to convert this uranium into a form usable for weapons.

Beyond the time frame to produce HEU, a U.S. official stated in Senate testimony that it would take Iran three to five years to produce a nuclear weapon once it has sufficient HEU.

At the April 14 Armed Services Committee hearing, Vice Chairman of the Joint Chiefs of Staff Gen. James Cartwright, in response to a question about when Iran may have a nuclear weapon, said that “three to five years is a historical estimate of how long it takes a nation with a low-enriching capability to move, both through the high-enrichment protocols and then to the things that would put it together to make it a weapon.”

Paul Pillar, former national intelligence officer for the Near East and South Asia, said in an April 19 e-mail to Arms Control Today that he was puzzled by Cartwright’s statement. “I find it hard to understand the application of some general, rule-of-thumb timetable to this question because each proliferator is apt to be in a different situation regarding the relative progress made on the separate tasks of fissile material production and weapons design,” Pillar said.

He added that “the principal pacing element” determining when a country acquires a nuclear weapon is the production of fissile material.