Jeffrey LewisHow Much Pu in a North Korean Bomb?

How many nuclear bombs does North Korea have? That depends on their plutonium (Pu) stockpile and design skill.

In April 2004, David Sanger claimed AQ Khan told Pakistani interrogators that he saw three plutonium bombs in North Korea.

Today, buried deep in a related story, he casually drops that maybe the story was bullshit:

… elements of [Mr. Khan’s statements to Pakistani officials] have leaked out, including Mr. Khan’s assertion – doubted by several specialists in the American intelligence community – that the North Koreans once showed him what they said were three fully assembled nuclear weapons.

Khan’s purported story contradicts defector testimony that suggests North Korea has a single, 1,000 kg plutonium device, using 4 kg of Pu (more):

The highest-ranking defector after Hwang Chang-yo’p. A deputy to the Supreme People’s Assembly [SPA] and a researcher (doctorate degree holder) at the “Maritime Industry Research Institute” under the “Second Economic Committee.” A graduate of Kyo’ngbuk Middle School.

“North Korea built a one-tonne nuclear weapon with 4 kg of plutonium. However, there is no confidence in the performance of the manufactured nuclear weapon.”

Initially, I thought that 4 kg was too low.

Assuming that North Korea wanted a 20 kt yield—the yield of fat man bomb used to destroy Nagasaki— the North Koreans would need to be relatively skilled to use just 4 kg of Pu. Tom Cochrane and Chris Paine express the relationship between Pu mass and yield at varying levels of skill in The Amount of Plutonium and Highly-Enriched Uranium Needed for Pure Fission Nuclear Weapons:

Of course, the problem with this statement is that Pyongyang has “no confidence” in the weapon. So, maybe North Korea built a 4 kg design that doesn’t work.

The IC doesn’t really know how to assess the possibility that adversaries will attempt things that don’t work, as suggested by North Korea’s Taepodong 1 test. Of course, even a nuclear weapon with a small chance of working provides some measure of deterrence.

Still, building a lone, low-confidence fission device with 4 kg of Pu makes sense only if that was all the Pu available to North Korea at the time.

That’s plausible, though on the low end of the North Korean Pu estimates. (The Chinese also had real trouble reprocessing).

The CIA estimates that North Korea “has one or possibly two weapons using plutonium produced prior to 1992.” (David Albright estimated that was 7-11 kg of Pu).

But Charles Kartman—outgoing Executive Director of the Korean Peninsula Energy Development Organization—told AP that North Korea might have been severely constrained in the amount of fissile material:

“There is a maximum amount of plutonium that could have been reprocessed, and if that is true, then depending on the state of North Korean technology, it would have been sufficient for one, or at most, two (weapons),” Kartman said.

Now when you get to the number two, you are really applying the worst case scenario. Everything has to run right,” Kartman said. “You’re not going to get too many responsible scientists going along with the number two” from that time period in the mid-1990s.

So, the 4 kg estimate is plausible—largely because of the caveat that Pyongyang has no confidence in the weapon.

For information on the mass of North Korean bombs, see ACW on 30 April 05 and 10 May 2005.


  1. Stephen Schwartz (History)

    For people averse to logarithmic scales, the Bulletin of the Atomic Scientists recently published a handy table using the same data showing approximate amounts of HEU and weapon grade plutonium required for pure fission weapons of selected yields.