Losin’ in Lausanne (5)

Omri Ceren emailed two reports from Lausanne overnight. Here is the first:

The P5+1 meeting with Iran got out a little while ago. Total running time: 1 hour, 14 minutes.

Functionally nothing happened this weekend from a news perspective – very little came out of talks, foreign ministers were still en route, etc. – which gave people plenty of time to muse over the Iran talks debate in the broadest terms. Short version: a 1 year breakout time is too short, and what the P5+1 are agreeing to won’t even be a 1 year breakout time.

1 year breakout not enough – Last week I sent around an email on this, which had the Hayden/Heinonen/Takeyh overview on why a one-year breakout time is too short to prevent an Iranian dash across the finish line. . The theoretical scenarios saying that the US could act in a year don’t work when you introduce real-world constraints: how long it takes the IAEA to detect cheating, how the IAEA referral process works, how the U.S. intelligence community would go about confirming a breakout, how the diplomatic debate would play out, etc. And even if the U.S. could mobilize in a year, there wouldn’t be any options available short of military action, because sanctions take much more than a year to being to work. So: detection time plus action time equals more than one year

It won’t even be 1 year – Then over the weekend Olli Heinonen – former IAEA deputy director-general – published a primer on how breakout calculations work (posted here). You should read the whole thing – it’s tightly written, and as good an explanation of this issue as I’ve read. He expands on the original argument that 1 year breakout time is inadequate, and gets into the nitty-gritty of IAEA detection and action. But he also adds the second part of the argument, about how the breakout time won’t even be 1 year. Given all the concessions that are likely involving centrifuges and R&D, it’s no longer tenable to suggest that it would take the Iranians that long to break out or sneak out. The way he describes the sneakout scenario is particularly specific.

Here is the second of Omri’s two messages, referring to the proceedings today:

It’s probably a make or break day in Lausanne. Technically the deadline is tomorrow night, but the parties should know by today whether they’ll get to an agreement in time. The full P5+1 met last night – Lavrov and Hammond didn’t get in until fairly late – and there’s a full meeting going on right now between the P5+1 and Iran. If things proceed the way they’re supposed to, they’ll get out of that meeting, then there will be a series of bilateral meetings, then they’ll be able to announce they’ve made it by tonight – in which case everyone will head off to Gevena for the announcement either this evening or tomorrow morning (and incidentally, look out for an email in a few hours on “why Geneva,” because that actually has the potential to become a Thing).

The NYT has the scoop driving the morning here. The Iranians appear to have pulled a very heavy-handed, last minute bait-and-switch. They bargained up their centrifuge numbers for months by saying that they’d ship out whatever material they enriched, the assumption being that who cares how much uranium they enrich to 3.5% as long as they don’t have it physically available to enrich further. Now that they’ve ratcheted up the number of centrifuges to over 6,000, they’re saying they won’t ship out the material.

In addition to being disastrous from a policy perspective, because it makes securing a 1 year breakout time almost impossible, it’s going to be a political problem: it looks like the administration got outplayed by Persian negotiators… again.

The NYT says the administration is doing damage control by floating that the Iranians will “dilute” their stockpile. Usually that implies they’ll commit to downblending their enriched stock – e.g. taking 20% enriched uranium and make it into 5% – but that doesn’t seem right in this context. The Iranians shouldn’t be enriching above 3.5% under a final deal, and what would downblending look like in this scenario anyway? The Iranians would enrich uranium, then dilute it, then enrich it, then dilute it? It’s the equivalent of digging a hole and filling it back up. How would they justify running the centrifuges?

Instead, “diluting” might mean that the Iranians will commit to turning the new uranium into uranium oxide, a form in which it can’t be enriched further. But that’s a chemical process that takes at most a couple of weeks to reverse, which means the Iranians would be making multiple nuke bombs’ worth of enriched uranium and putting it on the shelf. The debate over reversal has the potential to get mind-numbingly technical (right after the JPOA was announced, supporters of the deal tried to argue that Iran can’t reverse the oxidation process because of a hurdle involving piping technology; I ran the argument by one of the IAEA guys, who responded by literally laughing out loud and saying “how do they know what the Iranians have?”) But the debate isn’t really a close one. Here’s one assessment by Mark Hibbs, who is one of the top supporters of an agreement with Iran:

“Yadlin, cited as having told the Institute for National Security Studies that in Iran the reconversion of U3O8 to produce bomb fuel could be ‘completed in less than a week,’ walked back this estimate in a subsequent radio interview to ‘between one and two weeks.’ That’s more realistic. Experience from the uranium conversion industry and R&D sector outside Iran would suggest that Iran might be able to convert about 100 kg of U3O8 to UF6 in about two weeks–provided, however, that the work was carried out in a small facility using a dry process without purification, whereby perhaps three batches would be consecutively processed.”

Either way – downblending or oxidation – makes it difficult to see how the parties could stretch Iran’s breakout time to a year. Combine this new position with the Iranians’ demand to keep their fortified underground enrichment bunker at Fordow open, and the administration’s sales pitch on the Hill becomes that much harder.