North Korea Test not really a success

Many scientists now believe that the infamous Kim Jong-il fart may not be such a success as claimed. It is more likely a weaker version of the ad hoc test conducted by Pakistan as a response to India's test in 1998.

There are a few possible reasons for coming to such deduction

• DPRK does not have a lot of purified Plutonium or Uranium, to conserve nuke material it had to conduct a mini-test
  
• The technology may be immature -- e.g. the fuel (purity) grade is barely good for an explosion, but far from the optimum energy of a successful test (the consequence of this case would be dreadful for environmentalist, as a lot of unreacted nuke materials might have been left in the tunnel)
  
• The detonation process, though quite simple, still requires certain engineering precision, which the outdated NK technology may not have mastered
• The failure to seek attention might have prompted KJI to force a pre-mature test -- as happened for Taepodong-2 in July

So far this is consistent with technical data

• "Gary Gibson, senior seismologist at Australia's Seismology Research Center, said a 4.2 magnitude quake would be the result of a one kiloton explosion"
• This is small even if one allows for the margin of error, because "The nuclear weapon the United States exploded over Hiroshima in 1945 produced a 12.5-kiloton yield.
• A successful test would be one like what India did, with an average of 12 kilton/test. "In 1998, India carried out five underground nuclear tests at Pokharan in the western desert state of Rajasthan and declared itself a nuclear weapons state. The total yield of the first round of blasts measured near 60 kilotons...."Our biggest one was in the vicinity of 45 kilotons. That was thermo-nuclear," said S.K. Malhotra, head of the public awareness division of the Department of Atomic Energy.."
• Nuclear analyst Andrew Davies, from the Australian Strategic Policy Institute, said if the North Korean test yield was only a kiloton, Pyongyang may be disappointed..."A kiloton is a very low yield and would tend to suggest, I would have thought, that the device was not all they hoped it would be," Davies told Reuters. "If a nuclear, plutonium bomb fizzles, you can still get one or two kilotons quite easily. You still get a significant energy release. But an efficient device will give you more like 20 (kilotons)."
• Even the less successful (conducted without proper preparation) test by Pakistan, was believed to have generated at least 9-12 kilotons in strength. "Pakistan said two nuclear tests had a total yield of between 34-48 kilotons, while three others were sub-kiloton. It said a sixth test yielded 10-15 kilotons. The Southern Arizona Seismic Observatory said the two major tests yielded 9-12 kilotons, while the sixth yielded only 4-6 kilotons."

Jane's defence has a more concise summary that supported this analysis

• "Although details are tentative, initial and unconfirmed South Korean reports indicate that the test was a fission device with a yield of .55 kT. By comparison the nuclear bomb that was dropped on Hiroshima yielded approximately 12.5 kT. The figure of .55 kT, however, seems too low given the 4.2 register on the Richter scale. This could suggest - depending upon the geological make-up of the test site - a yield of 2-12 kT. If, however, the lower yield is correct, it would suggest that the test had been a "pre- or post-detonation" event (ie a failure), as it had been anticipated that North Korea's first nuclear test would have a significantly higher yield."

Though quite unlikely for the test yesterday, it is indeed possible to fake a nuke test with conventional chemical reaction, as verified by Lawrence Livermore National Lab with a 1.3 kiloton simulation in the desert of Nevada (though their objective was to find a way to detect real nuke tests with a simulation). The report also mentioned that if the explosion is more than 400m below surface, we may not see a collapsed crater -- as this might be the case for the NK test yesterday.

UPDATE (Oct10): Armscontrol Wonk is more adamant that it was a clear failure. I think he is right. More discussion on the likelihood of a failed test/hoax here.