Yesterday’s news that Bush wrote to Kim, reminded me of some back-of-the-envelope calculations I did a couple of months ago on verifying North Korean plutonium production. After all, when (I’ll be optimistic and refrain from saying ‘if’) North Korea submits its declaration, the fun of verifying it will start. One thing the US/IAEA (delete as applicable) will want to make sure is that North Korea has not hidden a stockpile of plutonium. To do this, inspectors will try to estimate the amount of plutonium produced by North Korea and compare it to current plutonium holdings (minus the amount used in the test). The two figures won’t be exactly equal for reasons discussed below but if they’re the same to within, say, 1 kg (less than the amount required to build a bomb) then it’ll be smiles all round. If there’s a difference of 10 kg then things could get awkward.

So, how accurately can North plutonium be verified? I think there are four major sources of error (note that the amount of plutonium currently in weapons or storage can be measured very accurately—although whether North Korea permits it is another story):

1. The uncertainty in the amount of plutonium produced in the 40 MW reactor at Yongbyon. Based on satellite imagery alone, Albright and Brannan put this at 60 ±10 kg (with slight rounding). I’ll be sceptical and assume that North Korean operating records don’t help reduce this error but, in actual fact, they may do.

2. The uncertainty in the quantity of plutonium lost during reprocessing. Most reprocessing plants, even—how do I put this politely?—basic ones, are pretty efficient. In the absence of any more concrete information I think 95 ± 2 % is a fair guess for the extraction efficiency of North Korean plutonium.

3. The uncertainty in the amount of plutonium used in the test. Let’s say that the North Koreans used 6 ± 1.5 kg in their underground firework display last October.

4. The uncertainty in the quantity of plutonium that is/will be in the spent fuel pond at Yongbyon. This is rather hard to measure accurately. I’ll use Albright and Brannan’s number of 11.5 ± 1.5 kg but, again, North Korean records may help reduce this error.

Anyway, adding these errors give a total of error of about 10 kg—remember this number essentially represents the upper limit on the size of a secret North Korean plutonium stockpile. The bad news is that this isn’t good enough to account for one bomb’s worth of plutonium. The good news is that it’s almost entirely due to one source—the uncertainty in the amount of plutonium produced at Yongbyon—and nuclear forensics can help to reduce it considerably.

Specifically, if North Korea permits inspectors to take samples from the reactor’s graphite moderator for analysis then the amount of plutonium produced in the reactor can (hopefully) be estimated to within about 2.5% using a technique called GIRM (Graphite Isotopic Ratio Method) that was primarily developed to account for plutonium production is the former Soviet Union. This, in turn, would allow the US to put an upper limit of about 3 kg on the size of any North Korean plutonium stockpile—probably small enough for comfort.

All of this makes me optimistic that trying to verify the plutonium part of the North Korean declaration won’t derail the process. We just need to hope that (i) the reactor is not being disabled using a material that would interfere with nuclear forensics and (ii) that the North Koreans are ultimately willing to give inspectors rather a lot of access.