Iran: When Should We Panic?

If you are anything like me, you probably aren’t too thrilled by the idea of Iran with nuclear weapons. There’s something about the Middle East, revisionist ideology, sectarian divisions, and extreme rhetoric that does not mix well with fissile material. And overall, the NPT is, as Martha Stewart might say, a good thing. In terms of the “Sagan-Waltz debate”:, mark me down as a Sagan guy.

(Not _that_ Sagan. “This Sagan”:

But then, being like me, you probably aren’t too thrilled by the idea of headlong, precipitous action, either. There’s something about the Middle East, revisionist ideology, sectarian divisions, and extreme rhetoric that does not mix well with desperate measures, either military or diplomatic. So our first inclination is to counsel patience, not panic.


But let’s be fair: isn’t there some point when panic should kick in?

The subject arises by way of Jeff’s “jeremiad against the panic-mongers”: the other day, which he “softened a little”: afterwards.

A body could be forgiven for reading “FT’s lede”: (or the “Guardian’s headline”: and concluding that Iran had 1 SQ (i.e., one bomb’s worth) of HEU on hand. I don’t mean to “pick on the Brits”: too much: the “LA Times”:,0,3140113.story was just as bad, and the “NY Times”: wasn’t that much better.

Just One Letter Away From “Freakout”

“According to ISIS”:, what the Iranians actually have achieved is “breakout capability,” meaning the ability to take a stockpile of LEU and rapidly enrich it to 1 SQ (or more) of HEU.

Not everyone agrees. Mohamed ElBaradei, for example, says Iran is “still about a year away from this point”: The differences seem to have to do with just how much (or how little) HEU really constitutes a “significant quantity,” and how much U-235 would be lost during further enrichment. (Ivan Oelrich and Ivanka Barzashka have “a helpful explanation”: of the latter issue.)

Regardless, it is clear that this threshold will be crossed sooner or later. It does not matter whether every informed person agrees exactly which day or hour it occurred.

It is also clear that this threshold is primarily of a psychological or symbolic nature. It lacks practical significance, by itself. Nor does it translate unambiguously to a particular political outcome. (Would it frighten the neighbors into a more cooperative stance? Galvanize the world into concerted action? My guess is neither, but your guess is as good as mine.)

First, 1 SQ would be one heck of a thing to exit the NPT over. If the Iranians tested their first and only nuclear device to demonstrate that they had it, they would promptly stop having it. So 2 SQ would be the realistic threshold of concern, and even that seems a bit low. The North Korean precedent is instructive: they didn’t proclaim themselves to be nuclear-armed, or prove that point, until they had enough plutonium on hand for maybe half a dozen devices.

Second, breakout would not go unnoticed, for the reasons “spelled out here”: The IAEA inspects roughly once a month, not announcing the day in advance. To deliver a true _fait accompli,_ the Iranians would have to act within that window. (There are cameras in the cascade hall, but it’s not clear that they are monitored remotely.) This requires not just a big stockpile of LEU, but having enough SWU on hand, in the form of centrifuges, to get the job done. Perhaps this would be better described as “sneak-out.”

Third, just getting ahold of 1 SQ (or 2 SQ) of HEU is not the same as having a working weapon. Breakout is a risky undertaking, a way of gambling with the future of one’s country. It does not automatically produce an amulet against attack. In fact, it might invite an attack. But this doesn’t mean that it will never be attempted.

Update: Andreas Persbo provides “some insightful elaboration on these issues”:

OK. So When Should We Panic?

For the time being, let’s set aside Iran’s plutonium pathway, which is not progressing as rapidly as the uranium side.

There are two main uranium-enrichment scenarios to be concerned about. The first is called batch recycle: the Iranians break the seals and feed the LEU through the cascades at Natanz again until they’ve got the desired level of enrichment.

The second scenario, which is favored by ISIS, is diversion: the Iranians break the seals and cart off the LEU in trucks (presumably at night, unseen by spying eyes) to a second centrifuge plant, one not known to the IAEA. When the inspectors show up at Natanz, the Iranians would delay them until it was too late.

The scenarios differ at the margins. A centrifuge (of a given type) is a centrifuge, regardless of location. In the diversion case, we just aren’t aware of the second plant until it is too late. Also, it could be specially configured for more slightly more efficient LEU-to-HEU enrichment.

For simplicity (well, relatively), let’s consider just batch recycle with the IR-1, the centrifuge type that’s being installed and operated at the Fuel Enrichment Plant (FEP) at Natanz. We know they’re there, right? Two newer types are being tested in small numbers at the adjacent Pilot Fuel Enrichment Plant, but it’s not clear that there is a large-scale production capability yet.

So here, drawn from various IAEA reports, is the pace at which the Iranians have added centrifuge cascades at FEP:


What we can gather from this chart is a sense of the minimum rate of centrifuge production. No one can install more than they have. So the solid black line is a floor of about 18 cascades’ worth of machines per year, or one unit. (There are 164 machines in a cascade, so this means close to 3,000 machines/year.) The dashed extension of the line shows the implication of that floor: unless centrifuge production has (for some unknown reason) slowed considerably, a backlog of machines is building up, awaiting installation.

_Update: if the above chart and explanation elude you, “additional explanation”: is now available._

[No, it does not annoy me that Jeff had “a similar idea”: at the same time. Well, not too much. What really annoys me is that his graphic is nicer-looking and more informative. Grr.]

Some Assembly Required

So, how many centrifuges would it take to deliver a _fait accompli?_

This is going to disappoint you. The answer isn’t so clear. It’s subject to a number of uncertainties. So here are several answers.

It seems there are already a few different views on this question:

* “Houston Wood, Scott Kemp, and Alexander Glaser”: in _Physics Today._

* “Gregory Jones”: of RAND.

* The “Wisconsin Project”:

* “Oelrich and Barzashka”: at FAS.

* And, although it’s not fully explicit about its assumptions, the “Annual Threat Assessment of the IC”:, which contains two such estimates:

We judge Iran probably would be technically capable of producing enough highly enriched uranium (HEU) for a weapon sometime during the 2010-2015 time frame. INR judges Iran is unlikely to achieve this capability before 2013 because of foreseeable technical and programmatic problems.

Clearly, six is not enough, so I’ve done my own rough-and-ready estimate, working with the “FAS SWU calculator”: It’s ever so handy.

What makes this so tricky is that there are six important variables:

  1. The rate at which Iran adds cascades.
  2. The separative power of those cascades.
  3. The desired enrichment level.
  4. The amount of HEU needed per bomb.
  5. The number of bombs that are worth breaking out for.
  6. The time window that is worth the risk.

To make this manageable, let’s just say 90% enriched HEU in three weeks’ time. That leaves us with amount of HEU per bomb, separative power, number of bombs, and rate of addition. We can bound these variables, though.

The usual range for HEU-per-bomb estimates is 15 to 20 kg. (The IAEA’s official SQ is 25, but that’s generally considered too high.)

Separative power estimates for the IR-1 range between 2 and 3 kg SWU/yr. “Jeff estimates 2.3”:

Based on reasonable assumptions and the North Korean case, number-of-bombs should range between 2 and 6 or so, but since this is the Bomb we’re talking about, let’s be gloomy and think in terms of 1 to 3.

Then there’s rate-of-addition. Above, we came up with a floor of 18 cascades per year, so let’s say 18 to 36 cascades (1 to 2 units).

All of which gives us:

kg SWU/yr 3 3 2.3 2.3 2 2
Cascades/yr +36 +18 +36 +18 +36 +18
1 x 15 kg 2009 2009 2010 2010 2010 2010
1 x 20 kg 2010 2010 2010 2010 2010 2011
2 x 15 kg 2010 2011 2011 2012 2012 2012
2 x 20 kg 2011 2012 2011 2013 2012 2014
3 x 15 kg 2011 2012 2012 2014 2012 2015
3 x 20 kg 2012 2014 2013 2016 2013 2017

So there you have it: breakout, broken out. You’re free to ignore any portions of the table that you find unrealistic.

One final note. His graphic design skills may be sweet, but I do have one thing Jeff does not: the “musical bonus”:

If you got this far, you’ve earned it.

Leave a Reply

Your email address will not be published. Required fields are marked *