Don’t Underestimate the Mullahs

The New York Times
August 23, 2005

Valletta, Malta

Earlier this month Bush administration officials leaked to the press what they said was a new official estimate of when Iran might be able to build a nuclear weapon. Speaking anonymously, they told reporters that American intelligence agencies now believe it would take at least 6 and maybe as many as 10 years before that fateful day arrives.

Whew! Instead of worrying over the previous estimate of only five years, we can relax. And if this administration can’t figure out how to stop the Iranian bomb, there will be plenty of time for someone else to do it. Right?

Actually, no. We should be alarmed rather than comforted by this latest prediction. Consider this: American intelligence agencies completely missed Saddam Hussein’s giant machines for processing uranium to weapons grade before the Persian Gulf war in 1991. Then, overreacting to that mistake, these agencies wrongly reported that there were weapons of mass destruction in Iraq before the 2003 invasion. Now, they appear to be overreacting to their last overreaction by underestimating the threat from Iran.

What must Iran do to make a bomb? This month it started an essential part of the process. It resumed the conversion of about 37 tons of natural uranium into the gaseous form that can be fed into centrifuges. Those machines, by spinning the gas at high speed, enrich its potency – either to a low level for fueling a reactor, or to a high level for fueling a bomb. These 37 tons, which should be ready for enrichment in a month or so, would be sufficient for six to nine weapons.

Why does the administration think it will take up to 10 years to process this material? The intelligence estimate is secret, but foreign and American officials involved in monitoring Iran’s efforts tell me that Washington assumes Iran’s centrifuges are of poor quality and that Iranian scientists may have trouble connecting them into what is called a cascade, in which the uranium must flow from one machine to the next.

This prediction, however, discounts an overwhelming amount of countervailing evidence. First, an official at the International Atomic Energy Agency, which monitors Iran’s nuclear progress in detail, told me that his agency is confident that the Iranians can produce high-quality centrifuges. Officials at the agency also know that Iran has built a string of workshops as part of a plan to produce some 50,000 centrifuges, with an assumed production rate of many thousand per year. It also has thousands of components for the centrifuges on hand, some it made itself and others imported, likely from Pakistan.

It is unreasonable to assume that Iran could not, after deciding to begin a concerted effort, assemble a 2,000-machine cascade in a year. In 2002, Iranian scientists enriched a small amount of uranium in an experimental cascade at the Kalaye Electric Company, a secret operation in Tehran that the International Atomic Energy Agency didn’t discover until 2003.

After a year’s operation of such a cascade, Iran would have one bomb’s worth of highly enriched uranium, and could have built and started running 2,000 more centrifuges. Continuing at this pace would yield three bombs’ worth of enriched material in three years, and about six bombs’ worth in four. This is the sort of calculation that experts at American government laboratories have been doing for a long time, and one such scientist told me he was stunned by the administration’s 6-to-10-year estimate.

And then there is the problem of what we don’t know. Inspectors from the atomic energy agency frequently complain that Iran has never explained how far it got in its efforts to build a more advanced model of centrifuge that could save lots of production time. Iran got the blueprints for this machine around 1995 from the notorious Pakistani scientist Abdul Qadeer Khan, and imported hard-to-find components like specialized magnets.

This raises the possibility that Iran may have centrifuges or laboratories we still don’t know about, a risk that seems quite high given that for almost two decades Iran managed to hide work on uranium enrichment that international inspectors found out about only after visiting the site at Natanz in 2003.

And the concealment continues: last year Iran razed a building at one suspected nuclear site and scraped away the underlying soil to prevent analysis (hiding evidence in this way was a favorite ploy of Saddam Hussein’s in the 1990’s). Iran is also barring inspectors from following up their work at another site, the Parchin military complex near Tehran, which many suspect is being used for work on the non-nuclear parts of a nuclear weapon.

This latter activity – the making of bomb parts other than the uranium or plutonium metal that explodes – is easily hidden because it would most likely occur in parallel at laboratories not involved in creating the nuclear fuel. And it seems very possible that Iran received a complete bomb design, plus blueprints showing how to manufacture it, from Mr. Khan. In the 1990’s he sold both Iran and Libya packages of centrifuge technology; we know that in the case of Libya he threw in the bomb design for good measure. Why would he not have given the plans to his other good customer, Iran, as well?

Americans should resist the latest intelligence-agency lullaby. Given the dismal performance of our spies and analysts in recent years, why should we think they have suddenly wised up? Iran is determined to get the bomb – all the agencies agree on that – and dealing with that threat is not a job that can be left for the next administration.

Gary Milhollin is the director of the Wisconsin Project on Nuclear Arms Control in Washington and the publisher of