Joshua PollackNon-Technological Safeguards

Cross-posted from

I’ve just stumbled across a paper, almost a year old now, by one Peter Friend, the head of Safeguards and Security at Urenco. It concerns the safeguarding of a centrifuge enrichment facility.

Yes, this again.

It includes a short section on “new inspection techniques.” Here are the highlights:

There are many organisations – particularly in USA – currently aiming to develop new equipment and new techniques for safeguards verification purposes. But many of the developers (who might not have many contacts with IAEA or with operators experienced in safeguards implementation) seem to be too interested in the technology per se, and should give a lot more thought into the practicalities.


In Urenco’s view, the presence of a competent inspector on site provides more effective safeguards than the use of complex remote monitoring equipment.

(VERTIC’s Persbo has mentioned the idea in the past as well.)

Without regurgitating Friend’s entire list of concerns — see page 7 of his paper if you’re interested — it suffices to say that there are many complexities involved with designing and installing new monitoring technologies in centrifuge plants, especially if the plant is already standing.

One might add to this a certain lack of trust between the monitors and the monitored: just what is that gizmo doing, anyway? And those third parties meant to be assured by the monitoring may have concerns that the gizmos can be gamed, one way or another, if there’s no one around to keep an eye on them. So having a permanent on-site presence does seem preferable in many ways.

(To be sure, sorting out the modalities, including who would make a mutually acceptable on-siter, is not entirely simple. Also, I do think continuous flow monitoring would be an excellent idea. These are not mutually exclusive ideas, or shouldn’t be.)

But there’s another benefit to having a small team of on-site inspectors always present. They can really get to know the people.

It’s the People, Stupid

Without pretending to know more than I do, let’s just say that there can only be so many humans in a given country, such as Iran, with the requisite expertise in working with centrifuge enrichment technology. Getting to know those humans and what they are doing seems like the best possible monitoring technique.

Call it social verification, right?

Concerns about breakout potential are clearly mounting — even if one doesn’t indulge in worst-case thinking — and undeclared centrifuge facilities are notoriously difficult to detect. So if you are worried about both a breakout at a declared site and the possibility of an undeclared site somewhere else, how would you guard against them? There’s reason to be doubtful that even the Additional Protocol, by itself, would suffice to detect undeclared plants with confidence.

You’ll sometimes hear this same argument made in favor of a multinational fuel center; personally, I find it pretty compelling, at least compared to alternative strategies. But there’s a long way to go before any such proposal can be realized. The good news is, even if the multinationalization idea can’t be achieved, the idea of a full-time presence can be adapted to safeguarding a national facility.

The difficulty, of course, remains in getting the monitored side to agree.


  1. Andreas Persbo

    One option that is often overlooked, but should be mentioned, is the establishment of more regional IAEA safeguards offices. At present, I think there are two (Toronto and Tokyo). The establishment of an office in the Middle East could be an idea, given the present situation. This would reduce travel times and, obviously, travel costs.

  2. Rick Wallace (History)

    You and Peter both make excellent points about the advantages of inspector presence over technology. One case where technology might be preferable is when the host state will not tolerate continuous inspector presence (you’d be surprised at how many that is) – and renegotiating baseline agreements can be excruciatingly slow. Another may be if one does a cost-benefit analysis. High-tech might be expensive, but you’d be surprised at how much equipment you can buy for the cost of a team (or even single) inspector on site continuously (including travel, subsistence, etc.).

  3. robusto (History)

    I’ve just stumbled across a paper, almost a year old now, by one Peter Friend

    So the paper was an old Friend—

Pin It on Pinterest