Mon May 15th, 2006 at 06:54:18 AM EST
September 11, 2001. In the wake of the terror attacks on the Pentagon and World Trade Centre, senior Bush administration officials became convinced that the U.S. must militarily overthrow the regime of Saddam Hussein.
On October 18, the CIA published a Senior Executive Intelligence Brief, Iraq: Nuclear-Related Procurement Efforts (SSCI, p. 36). Unlike formal assessment papers, SEIBs are not disseminated among the IC. Instead, they are distributed to `senior executives' such as the president and vice-president. SEIBs are usually drafted in response to specific policymaker questions and are narrow in scope. From July, 2001, until July, 2002, the CIA published nine SEIBs that discussed the tubes, none of which provided any information that supported the assessment they were intended for a nuclear program beyond a description of Joe's research.
In November, intelligence surfaced that Iraq had placed another order of aluminum tubes to the same specifications as those it had sought in July. The quantity required had increased from 60,000 to 100,000 tubes (Butler, p. 131). It is unknown whether the November order was seized like the previous tube shipment or if it managed to reach its destination.
At this time, senior policymakers appear to have become interested in Iraq's 1980s gas centrifuge program and, in particular, the IC's past assessments of it. Another SEIB, What We Knew About Iraq's Centrifuge-Based Uranium Enrichment Program Before and After the Gulf War, was published on November 24. The document noted there were "divergent views" on the tubes' likely end-use (RS, p. 196). Then, less than a week later, the tubes were being looked at again, this time by the Defense Intelligence Agency (DIA).
The Military Intelligence Digest (MID) is published by the DIA a little like a newspaper. It's distributed five times a week throughout the IC and contains a variety of intelligence products on diverse topics. On November 30, 2001, the MID included a supplement, Iraq: Procuring Possible Nuclear- Related Gas Centrifuge Equipment. The supplement was the first assessment of the tubes since DOE's eight-page analysis in August. It argued that a conventional end-use for the tubes was "possible" but unlikely. The tubes' use as centrifuge rotors was judged more plausible because their specifications were consistent with "earlier Iraqi gas centrifuge rotor designs." (SSCI, p. 91).
In DOE's August assessment, DOE had criticised Joe for assuming the Iraqis would build a 1950s Zippe centrifuge, when they had successfully built a more advanced Zippe/Urenco centrifuge with larger rotors made from maraging steel or carbon fiber. Besides, the tubes would not have worked in a 1950s Zippe centrifuge without extensive modification. The tubes' walls were too thick. However, the Military Intelligence Digest pointed out that the Iraqis had worked on an earlier gas centrifuge design in the 1980s: a Beams centrifuge. The tubes' walls might be too thick for any Zippe designs but, according to the MID, they were consistent with a Beams centrifuge rotor. Perhaps the Iraqis were planning a Beams centrifuge cascade?
|Table 6 - Specifications of the Iraqi Tubes, Beams Design, and Iraq's 1980s Beams Design|
|Iraqi Tubes||Beams||Iraqi Beams (1980s)|
(Source: SSCI, p. 109)
The DOE objected to the DIA's analysis on a number of grounds. Technically, yes, the tubes could be used in a Beams-type gas centrifuge without significant modification. However, the MID had ignored an important point: the Beams centrifuge had never actually worked (RS, p. 71). Not even Jesse Beams himself had managed to develop a working Beams centrifuge. Although the Iraqis had pursued a Beams centrifuge in the late-1980s, they could never get it to work either. The Iraqi Beams centrifuge had never operated faster than 25,000 rpm - far too slow to separate uranium isotopes (Albright, 1997). When Iraq procured the Zippe/Urenco designs in 1989, their Beams centrifuge project was quickly abandoned.
On December 17, two weeks after the publication of the DIA's MID supplement, DOE published Iraq: Seeking Additional Aluminum Tubes. The design the tubes most resembled was Zippe's 3" rotor. If the tubes were used in a Zippe centrifuge without thinning the walls, modifications to other parts of the centrifuge system would require "significant additional research and development." (RS, p. 209).
The DOE paper also detailed the technical difficulties Iraq would encounter in any centrifuge design using rotors with as narrow an internal diameter as that of the tubes. The paper argued that the "[Beams] centrifuge and the Zippe centrifuge have extremely low stage separation efficiencies that would lead to a very large number of centrifuge stages with a corresponding increase in cascade piping and complexity." (SSCI, p. 113). Iraq would need a cascade of between 12,000 and 16,000 centrifuges to produce enough HEU for a nuclear weapon in a year. A cascade that size would take years and years to build if it could be done at all. (In fact, DOE assessed that even a small cascade based on the tubes "capable of enriching gram quantities" of HEU would take Iraq close to a decade.) The only entities that had ever operated more than 10,000 centrifuges had been Russia and Urenco. A cascade that size required "significant operational experience" (SSCI, p. 113). DOE assessed it was doubtful that anyone could deploy a centrifuge cascade based on the tubes, let alone the Iraqis.
The December paper forcefully reasserted DOE's position that the tubes were unlikely to be intended for a gas centrifuge program. However, there appears to have been one aspect of the MID supplement that the DOE paper did not address. The November supplement had included a text-box that contained an analysis of the tubes by the National Ground Intelligence Center (NGIC).
The NGIC are a part of the U.S. Army and are considered the national experts on conventional weapons systems (RS, p. 55). The NGIC text-box in the MID supplement was titled, Conventional Military Uses Unlikely for Aluminum Tubes. It said, "Although 7075-T6 aluminum could be an acceptable metal for small rocket motor bodies, the 3.3-mm wall thickness and overall weight would make these particular tubes poor choices for rocket motor bodies. The thickness is roughly twice that of known small rocket motor bodies, and ...the 0.1 mm metal thickness tolerance along the 900 mm length is excessive for both rocket motor bodies and rocket launch tubes." (SSCI, p. 92)
NGIC's assessment was incorrect and, frankly, suspicious. DOE had published the dimensions of Iraq's Nasser-81 rocket in August. These were an exact match for the Iraqi tubes, including the wall thickness. It seems very strange that the U.S. rocket experts were unaware that the Nasser-81 - and the Italian Medusa rocket upon which the Nasser-81 is based - both had wall thicknesses of 3.3mm. It is also very strange that despite DOE's learning these facts months earlier, there is no record that DOE ever corrected NGIC's analysis or otherwise pointed it out.
It was true, however, that the "0.1mm metal thickness tolerance" was excessive for rockets (see Table 4). But Department of Defense rocket engineers pointed out to the Senate Intelligence Committee that "excessive" only meant "unnecessary" (SSCI, p. 102). The tight tolerances did not preclude the tubes' use in rockets but were in fact "perfectly usable" as such. Tightened tolerances only indicated the inexperience of Iraqi rocket engineers. (SSCI, p. 102) The Iraq Survey Group later learned that this had been precisely the case. Iraq's rocket engineers over-specified the motor casings' tolerances because they believed it would improve the accuracy of the Nasser-81. (ISG, p. 26)
Adding to the mystery is the MID supplement's level of classification. The supplement is classified "SCI" (RS, p. 206), which stands for `sensitive compartmented information'. SCI is one of the highest levels of classification within the Intelligence Community. SCI material can only be stored in specialized highly secure facilities and is only ever distributed on a need-to-know basis (Director of Central Intelligence, Directive 1/21, 1994). It is very strange that a Military Intelligence Digest supplement, which had already been disseminated widely, has been designated SCI.
By the end of 2001, only one assessment had claimed that the tubes could be used in a gas centrifuge without significant modification: the November MID supplement. The assessment was not only wrong, but was shown to be wrong within two weeks of its publication. Despite this, the MID supplement, and in particular its mysterious NGIC text-box, would play a significant role over the next year as the Bush administration made its case for war.