Nuclear Weapons

North Korea’s Bomb: A technical assessment [edited 16 October]

10.13.06 | 6 min read | Text by Ivan Oelrich

Last Sunday, North Korea apparently tested a nuclear explosive. The “apparently” is needed because the explosion was so small—by nuclear standards—that some have speculated that it may have been a large conventional explosion. What is the technical significance of the test, what does it mean, and what should we do now?

There is no question that the political and security implications of the test are huge and almost entirely negative. The technical implications are more mixed; the technical significance of the test is somewhat less than meets the eye.

There was early confusion about how large the explosion actually was, with U.S., French, and South Korean seismologists reporting a yield equivalent to about 500 tons of high explosive, that is half a kiloton, while the Russians reported that the yield was in the range of 10 to 15 kilotons, or twenty to thirty times larger. From the beginning, the source of this huge discrepancy was difficult to understand. Soon, the Russian seismic data were released and it became clear that even their own data did not support the Russian claim. Most reports as of yesterday had settled on the lower yield figure of about half a kiloton.

Reports appearing in the press suggest that the low yield indicates failure of the test. The simplest Manhattan Project style weapons will have yields in the ten to twenty kiloton range. Most analysts, me included, have assumed that that was the goal of the N. Koreans. (Sophisticated nuclear powers have built weapons with much smaller yields, with some bombs and nuclear artillery shells, for example, having yields of a fraction of a kiloton.) It follows that the low yield was a mistake, a major test failure. But it is certainly possible that they we aiming for a much smaller yield. For example, they may be working backwards, figuring how much bomb they can fit onto their missiles, and getting the best yield within that weight and volume constraint. So perhaps they were aiming for one or a few kilotons Even if this second approach is the one they were following, it seems unlikely, to me at least, they they were aiming at half a kiloton. So most likely the test fell short of their aims but we cannot know how much because even with perfect yield data, we still won’t know what their aims were.

So how has the situation changed with this test? From a technical perspective, less that we might first think. The outside world knew that the North Koreans had plutonium available from fuel rods that had been removed from the reactor at Yongbyon. We knew that at least some of the plutonium had been separated out of the fuel rods and, since separation is a fairly straightforward process, it was a fair assumption that most or all of the plutonium had been separated. So we knew about their plutonium supply (and the test tells us nothing more about that except that now they have a little less), but another key question remained: Could they fashion the plutonium into a bomb? We did not know, although the U.S. intelligence community concluded as far back as in the early 1990s that North Korea probably had built a couple of nuclear weapons.

There are two basic routes to a nuclear weapon, either using plutonium or enriched uranium. Enriching bomb-grade uranium is more difficult than producing plutonium in a simple nuclear reactor (although this is becoming less true with the continuing developments of gas centrifuges). Once a bomb-builder has the uranium, however, the construction of the simplest uranium bomb, a so-called “gun-assembled” bomb, is relatively easy. Plutonium is the opposite: producing the material is the easier step (especially if one does not worry about radioactive environmental contamination) but plutonium demands a more sophisticated and challenging bomb design, an “implosion” bomb. The Iranians seem to be on the uranium route (like Pakistan); the North Koreans have followed the plutonium path (like India).

Before the test, we did not know whether the North Koreans could build an implosion bomb or not. Had the test been successful, we would now know that they could, although we would still not know how close they were to a useable weapon; their test device might have weighed tons and been a once off, rigged up, laboratory experiment. But the test was not successful, so we still don’t know whether the North Koreans can build a workable implosion bomb. Presumably the North Koreans learned something from the test so the probability of the next test being successful is somewhat higher than the probability that the first test would have been successful. This is not much of difference, leaving us in pretty much the same position we were in before the test. So the political implications of the test are huge but the technical implications are quite limited.

Why might the test have failed? An implosion bomb uses conventional high explosives to compress plutonium until it becomes “critical,” that is, it will sustain a run-away chain reaction. The pressure from the conventional explosives has to be carefully controlled, for example, it must be symmetric or else it is like squeezing a ball of putty in your hand: pressure on one side doesn’t compress the plutonium, it just squirts it out the other side. The most likely reason for the failure is some problem with the compression and there is any number of reasons why the compression might not be adequate. Assuming the test was carefully instrumented (and given North Korean technology, this is not certain), the North Koreans should be able to narrow down the cause, which will give them a much improved chance for success with their next test.

We might be able to learn something ourselves about the test if radioactive debris escaped from the test site. It is not easy to completely contain an underground nuclear explosion. Russian tests often leaked. The US was much better at containing tests but even US tests leaked in a couple of cases. Of course, it is much easier to contain a half kiloton test than a ten kiloton test but some radioactive material might have leaked out. Detecting that would, first, confirm that the test was, in fact, nuclear and analyzing it might provide some limited information about the design of the weapon and the source and purity of the plutonium.

What does this mean about possible responses? First of all, there is something to be accomplished by responding. When the North Koreans broke out of the safeguards on their reactor, making several bombs’ worth of plutonium available, it was a disaster for the control of their weapon program. The outside world could keep an eye on the reactor and account for the materials there but once the material left the site, trying to track it was hopeless; the volume of the plutonium is small, any of thousands of buildings could house it. But there was one remaining important hurdle that would generate a clear signal if jumped by the North Koreans: a test. Had the test been successful, then the cat really would be out of the bag. There would be no way to track some uncertain quantities of plutonium and the North Koreas would have a design that, in principle, could be replicated to produce more bombs in any of hundreds of nondescript light industrial facilities. (I say “in principle” because a more sophisticated nuclear power would require more than one test, but the North Koreans may have substantially lower reliability and confidence requirements.)

But the test was not successful. The North Koreans no doubt learned a great deal from their test but they have not proven to themselves, or the world, that they have a design that works. If the first test were successful, the marginal value of subsequent tests would have been relatively much smaller but with an unsuccessful test, the value of the next test will be as great or greater than the last test. We should not give up and say this is now a lost cause. There is much to be gained by using threats of sanctions and other tools to stop follow-on tests.

I shall write again soon on the political and security implications.

[The fourth paragraph above was edited on 16 October. Some well-informed readers pointed out that the original assumption about what the N. Korean goal was could be wrong.]