Right on time, ISIS has released the Safeguards report for Iran — Implementation of the NPT Safeguards Agreement and relevant provisions of Security Council resolutions 1737 (2006), 1747 (2007) and 1803 (2008) in the Islamic Republic of Iran GOV/2008/15 (May 26, 2008).
And the big news, at least from the wonky angle, is a new centrifuge design, the IR-3:
3. On 10 April 2008, Iran informed the Agency about the planned installation of a new generation sub-critical centrifuge (IR-3) at the Pilot Fuel Enrichment Plant (PFEP). On 19 April 2008, the Agency confirmed that two IR-3 centrifuges had been installed at PFEP. In February 2008, Agency inspectors noted that Iran had also brought 20 IR-1 centrifuges into PFEP, which were run in a 20-machine cascade for a short time, after which they were removed.
Why is Iran building an IR-3 before mass producing the IR-2? As I understand it, the IR-2 is an interim design until Iran can design better scoops — tubes that withdraw the enriched uranium — that will allow the IR-3 take advantage of the faster speed of a carbon fiber rotor.
What The Hell Are Scoops?
Scoops are little bent tubes that withdraw the enriched uranium from the top of the centrifuge.
Figure 22.2.10. Centrifuge columns and
a scoop found in Iraq. A column with
feed, product, and waste connections is
shown at the top of the photograph, and
a scoop is attached to the column at the
bottom of the picture. The scoop extends
from the right end of the column and
touches the tape measure.
DOE published the best description and picture — well the only picture — of scoops I could find in Annex 3 of the Handbook for Notification of Exports to Iraq.
Scoops are small tubes designed to operate in a fixed position inside a spinning gas centrifuge rotor to extract UF6 gas and transfer it to the center post where it is removed from the centrifuge machine. A centrifuge scoop is similar in appearance to the Pitot tubes used on aircraft to measure airspeed. Figure 22.1.1 shows a cutaway view of a centrifuge machine with scoops located at the top and bottom inside the rotor. Each scoop is permanently mounted to the center column with ends very close to the wall of the rotor. Figure 22.2.10 shows a typical centrifuge column with a scoop attached to the end of the column.
The length of the scoops is slightly shorter than the radius of the centrifuge rotor. Scoop tube diameters can be up to 12 mm, although scoops with larger diameters have been produced. One end of the scoop is bent to face the gas flow within the rotor tube and the other end is attached to the fixed center column of the centrifuge. Each scoop is manufactured from UF6-resistant materials. Typical materials of construction for gas centrifuge scoops include copper, aluminum, and stainless steel.
These are pretty tough to manufacture so that they don’t disrupt the internal gas flow. Pakistan imported a bunch of preforms, which produced a very nice explanation of scoops by ISIS.
IR-3 and Scoops
I first heard that Iran was working on an IR-3 after Ahmadinejad’s office released pictures of him strolling through Natanz. Iran, it was quite obvious, had assembled a cascade of IR-2 centrifuges.
Experts thought Iran wasn’t ready to install cascades, based on reports that Iran was having trouble developing scoops to withdraw the product. Bill Broad even quoted one anonymous European expert talking about difficulties with the IR-2, although he didn’t specify the nature of the trouble:
A European centrifuge expert who closely follows the Iranian program, including the evaluations of international inspectors, said difficult work remained on the IR-2. “They obviously have months, if not a year, of test work to do before they can consider proceeding with mass production,” the expert said, who spoke on the condition of anonymity because of the issue’s sensitivity.
That was probably a reference to the scoops, which were noticeably absent on the table of components.
So, one quite plausible suggestion was that Iran was installing a less than optimal IR-2 design in cascades, while working to manufacture a better scoop for a third design (the IR-3) that would make fuller use of the tensile strength of carbon fiber.
After Iran reported it to the IAEA, word got around.
IR2 and IR3 Estimates
Working from these assumptions, our friend Scott Kemp sends along some calculations on the IR-2 and IR-3.
|Design||(is like)||Speed (m/s)||Super- critical?||SWU (kg/a)||Current number, location and status|
|IR-1||P-1||330||supercritical||2.5||> 3000 machines installed in cascades|
|IR-2||short P-2||450||subcritical||2.2||~10 machines in one cascade, plus some stand- alone machines|
|IR-3||short P-2||600||subcritical||4.0||~2 prototype stand-alone machines|
Note. IR-1 is SWU estimate is based on observed efficiency of 42%; IR-2 and IR-3 based on an estimated efficiency of 60%.
Scott notes, based on the report, that Iran “is getting better results with its IR-1 (P1) cascades. The early cascades were suffering from instabilities, requiring low feed rates, high mixing losses, and low effective SWU values.”
Presumably Iran’s choice of whether to mass produce the IR-2 or IR-3 will reflect their success in developing better scoops.