I am giving a talk today at AAAS on the Bandar Abbas Uranium Production (BUP) Center, which was the subject of Jonathan Tirone’s story last week.

The talk is by invitation, but I simply plan to walk through the images to see if we can confirm the BUP is operating at full capacity and discuss the policy implications of this development.

Basic Info About the Bandar Abbas Uranium Prodcution Center (The BUP)

The first slide simply reproduces this table from the 2007 OECD Red Book that shows basic technical information about the BUP (as well as the Ardakan Uranium Production center).

- Center #1 Center #2
Name of production center Bandar Abbas Ardakan
Production center classification Existing planned
Start-up date 2006 2009
Source of ore:
• Deposit name Gachine Saghand
• Deposit type Surficial metasomatite
• Resources (tU) 100 900
• Grade (% U) 0.200 0.0553
Mining operation:
• Type (OP/UG/ISL) OP 10% OP, 90% UG
• Size (t ore/day) 55 500
• Average mining recovery (%) 85-90 85-90
Processing plant (acid/alkaline):
• Type (IX/SX/AL) AL AL
• Size (t ore/day); for ISL (L/day or L/h) 48 400
• Average process recovery (%) >70 >75
• Nominal production capacity (tU/year) 21 50

Source: Uranium 2007: Resources, Production and Demand, OECD, 2008.

As I understand this table, this simply means that, at Gchine, they pull out about 55 tons of ore a day, of which more than 85-90 percent contains uranium.

And since 85-90 percent of 55 is 48, the BUP mill processes 48 tons of ore per day. About .2 percent of the ore is uranium, of which more than 70 percent can be recovered. And, since everybody gets one day off a week (Friday in Iran, right?): 48 t ore x 0.2% uranium x 70% recovery x 312 days is 21 t of uranium yellowcake per year.

The interesting point, for an analyst, is that since such a tiny amount of uranium is extracted, we basically know that 48 t of leftover ore a day ought to go in the tailing pond. Which brings us to the pretty pictures from April and October 2009.

Can we estimate if the mill is operating at full capacity?

I think, from looking at how fast Iran has filled the first tailing pond, we — actually, you, dear readers — ought to be able to assess whether the mill is operating at full capacity.

Here is a brief list of images that show the empty, one-third full, two-thirds full and full (well mostly), along with links so you can do it yourself:

D# elapsed Date Fill m3
0 0 5 Mar 2008 Empty
262 262 25 Nov 2008 One-third 9,000
413 151 26 Apr 2009 Two-thirds 5,200
573 160 3 Oct 2009 Full 5,500

The first column just shows how many days have elapsed since the first photograph; the second column how many days since the previous photograph. The third column is the date of the image and my estimate of how full it is (using the Mark 1 Eyeball). The final column expresses the volume of tailings the BUP — processing 48 tons of ore per day, six days a week — ought to churn out over the period since the last image.

If we assume full production over 573 days (again, breaking on Fridays), the BUP should produce about 19,700 m3 of tailings.

(I used a dry density of tailings estimate of 1.2 t/m3, which is almost certainly too conservative.) That would suggest that the pond need only be about 3 m deep. That looks plausible to me, again using the Mark 1 Eyeball, which means I can’t exclude the possibility the the BUP is operating at full capacity. As we discussed, however, I am not the best imagery analyst. So, dear readers, this is where you come in.

Deepening Suspicions

The rapid development of Gchine is suspicious, since Iran spent considerably more time and effort touting the Saghand deposit — giving presentations to international conferences, tours to journalists and even setting up a webpage (since taken down).

Then IAEA DDG for Safeguards Pierre Goldschmidt — in a highly confidential report to the IAEA board, which of course was leaked — asked pointedly in June 2005 “why the work on the very promising Gchine project was suspended by the AEOI from 1994 to 2000 to focus on a much less promising ore deposit at Saghand.”

The implication, of course, is that work wasn’t suspended at all, but that the mine and mill were intended to be part of a covert program.

The company AEOI contracted to develop the BUP was Kimia Maadan, which seems to have been linked to other elements of the suspected clandestine program centered on “all the sketchy stuff” that Mohsen Fakhrizadeh was up to at Lavizan Shian.

The fact that the Gchine mine is now the main mine in Iran suggests Goldschmidt’s question is not one the Iranians want to answer.

Policy Crap, or A lame Zzzzzzzzz policy prescription

As one editor once called something I submitted. (Really!)

We know, thanks to Mark Hibbs (“All of Iran’s UF6 centrifuge feed now indigenously mined, milled,” Nuclear Fuel, December 15, 2008) that all of Iran’s yellowcake is now mined and milled right there in Iran.

The bad news is that, as Iran produces more yellowcake, we need to worry that some of it could end up in a covert facility, like Qom, for conversion and enrichment. The good news is that this should force a realistic assessment of the status of Iran’s program. This assessment should conclude that we should improve our ability to monitor the Iranian nuclear program, rather than attempting to place unverifiable limits on it.

I guess you could say it depends on whether you are a “The tailing pond is half-full or half-empty” sort of person.

For me, the place to start is with the Additional Protocol. Iran “declared” the Gchine mine as part of its May 2004 Additional Protocol declaration. The AP, to which Iran stopped adhering in 2005, would obligate Iran to provide basic information about the mine, as well as permit periodic access to IAEA inspectors.

The Additional Protocol is not, however, enough. Our ultimate goal needs to be a broader political settlement with Iran, in which the Islamic Republic has incentives to refrain from exercising its bomb option. An intrusive monitoring arrangement that emplaces a verifiable barrier between Iran’s civilian program and a bomb won’t guarantee non-weaponization, but it can sharpen the choices provided sufficient incentives (and disincentives) are also on the table.

In this context of a larger settlement with Iran, we might imagine a more venturesome endeavor to safeguard the mine. R. Scott Kemp has written a fascinating paper, On the Feasibility of Safeguarding Uranium Mines, that makes the case for perimeter monitoring of such facilities:

A different approach is to turn the entire mining site into a controlled area, including as many of the post-processing plants (milling, leaching, and purification) as possible. In this case, the unimpeded movement of uranium material inside the mining site is permitted. One needs only to verify that all the uranium exiting the site is accounted for.

With full-time inspectors, an effective material control area is easily imagined. A fence with real-time sensors to detect a breach could be placed around the majority of the mine’s operations. A limited number of portals could be equipped to monitor vehicles and personnel as they entered or exited the controlled area. The unnoticed diversion of one significant quantity, or 250 tons of unprocessed ore, from a well-controlled area is unlikely.