(A Swiss-made Bottom Bearing from Iraq similar to the Dutch model discussed below.)

Bet you’re all hot and bothered now. Along with “portal perimeter monitoring”, “bottom bearing preform” just sounds naughty.

Mark Hibbs is following an obscure Dutch export control case that has been virtually ignored by the US press. Henk Slebos was convicted of illegally exporting a variety of dual-use items to Pakistan, for use in their nuclear weapons program.

The charge on which he was acquitted, however, is the interesting one. Hibbs reports that, in 2001, Slebos sent 10,000 bottom bearing preforms (an important component of a centrifuge) to the Institute for Industrial Automation in Rawalpindi:

The Alkmaar court then and since did not divulge that Slebos had in 2001 also shipped thousands of other steel bearing balls to IIA that precisely matched the metallurgical and design specifications for the bottom bearing of the Urenco centrifuge known as CNOR. Intelligence sources told NuclearFuel that evidence that Slebos exported these goods to Pakistan was considerable.

Western officials said that Pakistan had given up on CNOR by around 1985 in favor of another Urenco-design machine, G-2. It was strongly believed that when Slebos exported these items to IIA, in 2001, they were destined for use in so-called P-1 centrifuges. These centrifuges, based on CNOR, were to be set up with Khan’s assistance in Iran, Libya, and perhaps North Korea, the sources said.

Slebos was acquitted because the Dutch government failed to promptly inform him that the sale was prohibited by a Dutch catch-all clause. Slebos does not, apparently, dispute the facts of the export.

The fact that the bearing preforms matched the CNOR centrifuge is interesting because it means they were for re-export, likely to Libya or Iran.

Hibbs previously reported (more) that the CNOR formed the design basis (along with the SNOR design) for the P-1 centrifuges at Natanz. Joby Warrick and Glenn Kessler reported that the design of the centrifuges had been substantially modified—so the technical details are murky.

I wonder … For how many preforms, at present, can we account?