Jeffrey LewisDid Pakistan Test a North Korean Bomb?

Golly, I had no idea that Qatar Airlines has an entire premium Terminal for First and Business class patrons. Wow.

The other day a reader, Tom, asked about rumors that one of Pakistan’s nuclear tests in 1998 involved North Korean supplied plutonium:

Weren’t there credible allegations that at least one of Pakistan’s May 1998 nuclear tests involved a Plutonium based weapon which used (allegedly) North Korean supplied Plutonium?

There doesn’t seem to have been much revisiting of the issue lately that I can find.

Does anyone have anything that can elaborate on or discredit those allegations?

As far as I can tell, like a lot of things in life, it is complicated. I don’t buy it, though obviously it is worth verifying in the Six Party process or maybe wringing out of AQ Khan.

Pakistan claims to have tested six nuclear devices in 1998 — five on May 28 and one on May 30, all using highly enriched uranium.

About a year later — in January 1999, Dana Priest published an article in the Washington Post about a dispute between Los Alamos and Lawrence Livermore National Laboratories. (“U.S. Labs at Odds on Whether Pakistani Blast Used Plutonium,” January 17, 1999, A2, full text).

Priest claimed that:

  • A US aircraft — presumably a WC-135 — collected an air sample that contained plutonium shortly after Pakistan’s May 30 test.
  • Los Alamos National Laboratory conducted the preliminary analysis and concluded that Pakistan had tested a device involving plutonium. Given that Pakistan did not have a plutonium production capability in May 1998, North Korea was an obvious suspect.
  • Analysts at Lawrence Livermore National Laboratory disagreed with the Los Alamos conclusion, “alleging that Los Alamos contaminated and then lost the air sample from the
    Pakistan blast.” One official confessed to Priest (sorry, couldn’t help it) “there is some disagreement here, and experts at the labs need to sort it out.”
  • A second sample existed. Officials disputed whether it was “identical” — whatever that means — but the bottom line, according to Priest, was “scientists believe it will be possible to positively determine whether the initial analysis was faulty.”

This all leaked because it was included in a briefing materials prepared for President Clinton’s December 1998 meeting with then-Pakistani Prime Minister Nawaz Sharif in Washington.

About a year after that — presumably enough time for the second sample to be examined — Mark Hibbs (subscription only; full text) reported that the source of the plutonium in the air sample was from one of the Indian tests.

But on Feb. 3 U.S. officials close to the matter confirmed instead information from other sources suggesting that, when analysts at the Los Alamos National Laboratory (LANL) and the National Nonproliferation Center of the Central Intelligence Agency (CIA) obtained the raw data collected in Pakistan, a battle broke out among experts on how to evaluate it. Since then, sources said, the report that Pakistan used plutonium in its devices has been discredited.

Sources said it is currently believed that the plutonium found in the environmental samples in Pakistan was instead of Indian origin, and that comparative isotopic analysis suggests the plutonium was vented to the atmosphere by the explosions at Pokaran carried out two weeks before. One official said that meteorological data corroborated the hypothesis that small amounts of plutonium which were dispersed by an Indian blast out of the test shaft were transported by air currents to the area surrounding the Pakistan test site, which is located about 500 miles northwest of Pokaran.

(Vented Indian Plutonium Deemed Source of Reports Pakistan Tested Pu Weapons, Nuclear Fuel, February 7, 2000.)

I figured that pretty much settled the matter. Apparently, folks at LANL didn’t agree. So, when A.Q. Khan copped to assisting North Korea in 2004, someone called the New York Times.

David Sanger and Bill Broad reported in February 2004 “the old argument has been reignited in the United States’ national laboratories” (” Pakistan May Have Aided North Korea A-Test,” February 27, 2004).

In a clash between old rivals, the Lawrence Livermore Laboratory raised questions, claiming Los Alamos had erred, experts familiar with the dispute said. The problem was inadvertent contamination of the sample by American researchers, Livermore experts said. Eventually, a consensus emerged that the plutonium did come from Pakistan.

In April 2004, Sanger would repeat the claim
in another story about Khan (“Pakistani Says He Saw North Korean Nuclear Devices,” April 13, 2004).

The problem, of course, is that the claim given to Hibbs was not contamination by American researchers (though Priest raised that possibility) but debris that vented from the Indian test.

As far as I can tell we have two assertions that are at odds — Hibbs reports a consensus that the plutonium was from India; Sanger and Broad report a consensus that the plutonium was from Pakistan.

I suppose we don’t know who is right — Los Alamos or Livermore, Sanger and Broad or Hibbs.

But perhaps the North Korean declaration will shed some light on the issue. And, of course, if any of my readers with appropriate zip codes feel chatty, you know how to reach me.


  1. Lao Tao Ren (History)

    I have some issues as to how the question is framed.

    The larger question beyond the immediate one of whether the Pakistani test was of North Korean (or other) origin is, what degree of co-operation, movement of personnel, materials, and knowhow is occurring between states with illicit and NPT sanctioned nuclear weapons programs.

    Priors: Is it or is it NOT a Pu bomb?

    If yes, then:

    Lets not specify the source of the Pakistan Pu bomb, but let us begin by assuming that a Pu bomb was detonated.

    One approach is to look at hypothesis that gives a range of possibilities:

    – Pakistan did it on their own – an entire, ready to blow bomb sent to and used by Pakistan
    o fat man sized or bigger laboratory device
    o weaponized device that is much smaller – ready to use components were sent from elsewhere, but assembled in Pakistan
    o what components? Pit, Explosive lens assembly, Trigger, Odds and ends – No physical components, but personnel and technology
    o People with particular skills, Test data, Designs
    o Machinery to make critical pieces – Materials
    o Critical materials (i.e. partially enriched Pu)

    The point of this exercise is, I think we are making a bad decision to go down the path beginning with pointing fingers at North Korea as the suspect. By doing so, analytically, the other possibilities, that many other places and people could have contributed a piece here or there, has been foreclosed.

    Many of the pieces listed above could have been shipped / transferred and yet attract little notice. Who is going to notice a floppy disk of bomb design software and test data?

    When the possibilities are laid out in this fashion, it is clear that there are many more suspects beside North Korea. When the Libyan program was dismantled, we find, after the fact, that pieces came from not just Pakistan, but from Malaysia, etc. —- all over the place to attract as little attention as possible.

    On a broader policy level, the question is whether documented North Korean transfers would cause a material change in the “disablement” program as it is structured.

    At what level would there be an international consensus that North Korea has not come “clean enough”?

    Or, another way to parse the issue is: what are the kinds of information that would pose the least obstacles to obtain and provide a maximum of information?

    Here is the wish list:

    – how about soil samples from the actual test site (which would pretty conclusively tell us if there is Pu residuals from the test). If it vented into the air in minute quantities, it would have vented onto the ground.

    – How about comparing test site soil samples from the DPRK and Pakistan tests? Why is it that I have a feeling that this has already been done? And not done by one, but at least 2 or 3 countries. Grin.

    – get NK to disclose their bomb designs and as much evidence as possible that the disclosed design(s) were bona fide ones built by them (and not a spoof to throw off the investigators). That would mean things like detailed blueprints, looking at jigs, inspection of actual machinery and tools used in the manufacturing process (to show it was actually used and how it was used), and details of many other components that are hard to get / make and how they got a hold of them. That would tell you plenty about how much material OUGHT to have been consumed.

    – Want to really dream in Technicolor? Get the Pakistanis to disclose their bomb designs…..

    – Work backwards from the Pu claimed to have been produced by NK, and a pretty interesting picture should emerge.

    Why is it I am almost certain this exercise will not be done? GRIN

  2. kerbihan

    Three possibilities:
    Hypothesis 1: Indian Pu.
    Hypothesis 2: NORK Pu.
    Hypothesis 3: Pak Pu.

    Is it credible that the Pu came from the Khushab reactor? My recollection is that it was already operational, but I suppose that reprocessing would have been detected by US intel.

    Also, some Pakistani experts have referred to a possible “composite” HEU/Pu core for the 6th test.

    And yes, the First class part of the Doha terminal has a jacuzzi that you can use before or after your massage.

  3. Anon.

    Being a complete layman in the subject, I suspect this hasn’t been mentioned simply because it’s trivially impossible, but I’ll ask anyway so I could know. Could the Pu have been produced in a detonation of a relatively low enriched uranium device? One that had a significant U238 content that would have (partially) converted to Pu during the chain reaction. I’ve read it (somewhere unreliable; then again, what unclassified sources aren’t?) that c. 20% U235 is enough for weaponization, so U238 could be present in large quantities. Of course I’d expect it to show up in air samples as well, so lack of any may rule out this explanation even if it’s otherwise possible.

  4. Steven Dolley (History)

    Interesting question, but when you get down to around 20% U-235 enrichment the amount of uranium required for a critical mass is pretty enormous. Also, I don’t know why Pakistan would be interested in testing an LEU bomb, even if one were feasible, as one of its very first series of tests. Even if theoretically possible, such a bomb would be undeliverable.

  5. mike

    LANL making an error? Who would think such a thing? Those morons have been bilking the hard earning American taxpayers long enough with their pedestrian version of physics. It is about time light gets shown on that group and their charade ends. Good-bye St. Pete and the welfare they are accustom to – hello having to do real science and competing like real capitalist pig Americans.

  6. Andy (History)

    The 19-day timespan between the first Indian and last Pakistani/NK tests puts a few dents in this theory I think. Some key questions:

    1. Was there a NK weapon “waiting” in Pakistan? Probably not.

    2. Was there an agreement in place to allow NK to test, along with all the logistical and other details prearranged, before the Indian tests? This seems unlikely to me.

    3. Assuming this was an ad hoc opportunity for North Korea, are 19 days enough to negotiate the details with Pakistan, get the weapon there and assembled, etc., all in secrecy? Possible, but I think a sudden flight from North Korea to Pakistan or, less likely, a NK ship movement would probably have been detected along with other intelligence that might give additional weight to this possibility. I’m sure that intelligence collection systems were intensely focused on Pakistan after the Indian tests and the opportunity to detect such a shipment was therefore pretty good. That the reporting Jeffrey cites above lacks any comment at all on such matters indicates to me there is no evidence of any shipment and that the analysis and conclusion was made solely on the sample analysis.

    4. The purported North Korean device was the one detonated on 30 May with a yield of 12kt. Eight years later, in 2006, NK conducted a test on its soil which was a fizzle. It seems unlikely to me that North Korea could transport, assemble and test a device in Pakistan in 19 days, yet fail miserably at testing a device on its own soil with ample time and resources to prepare. Perhaps the 2006 test was a more complicated design, as Jeffrey has discussed here before, but that raises other questions.

    So in the final analysis, I think it’s possible, but pretty unlikely the 30 May test was North Korean.

  7. Lao Tao Ren (History)


    The flaw in the logic is assuming that it happened in 19 days.

    I mean, you can make the world in 6 days, but probably not a Pu bomb in 19.

    The scenarios I describe would have a much longer than 19 day timeline.

    Try years. Decades.

    The problem with nuclear technology is it is a relatively slow changing, low obsolescence rate technology that allow patient proliferators to patiently assemble it, bit by bit, piece by piece, over time.

  8. M Ahmed (History)

    Pakistan began work on the theoretical design of the bomb, as far back as 1972, when the “Theoretical Physics Group” was formed in the Pakistan Atomic Energy Commission. The PAEC by December 1973 had opted for the implosion design. At this stage, the PAEC plan was to build a plutonium device, but in the wake of India’s 1974 test, it became apparent that a uranium enrichment project needed to be added to the original plutonium program. But the same group formed in 1972 working on the plutonium device also continued to work on designs based on HEU, in PAEC. Also, by 1981, the New Labs pilot-reprocessing plant had been completed by PAEC and by 1985, work had commenced on the 50 MW production reactor at Khushab. Therefore, designs on based on HEU and Pu-239 would logically have been developed in parallel, since the Khushab reactor was expected to be complete by the mid-1990s.

    On March 11, 1983, PAEC carried out the first successful cold test of a working nuclear device. A second cold test followed shortly and till 1992, PAEC conducted 24 cold tests in which 4-5 different weapon designs were tested, and improvements made. By 1987, the design had been miniaturized to the extent that it could be easily carried by fighter-aircraft of PAF. And the design tested in the last cold test of 1992 was the one which was tested on May 30, 1998.

    So it did take decades of bomb design work, R & D and cold testing before the fifferent bomb designs were successfully tested on May 28 and 30, 1998. The design tested on May 30 was said to have been a very compact one that could easily be carried by missiles.


  9. Lao Tao Ren (History)


    Thank you for confirming it took decades.

    The question then become… how many fingers were in the pies during all those decades.

    Me think I see not just European, Asian, and…

    Just remember the American fingers in the French pie… and the many, many fingers in the Israeli one…

    The NPT is discriminatory in another fashion (than what the Indians say). When the proliferation trail goes back to… the place with the biggest cache of weapons and their allies or favored friends… the trail goes cold real fast.

    But when it goes to non-aligned or, heaven forbid, enemies….

    Like North Korea, lots of questions get asked until the trail leads where it is not suppose to.

    I for one, would like to see how much traffic there was to and from North Korea from (deleted list of places).

    Someone need to point out that it weakens an international regime when the hegemon is referred to as “just like the Chairman of the Central Committee”.

  10. Allen Thomson (History)

    > On March 11, 1983, PAEC carried out the first successful cold test of a working nuclear device.

    Is it known what kinds of diagnostic instrumentation the Indians (and others) have to, er, diagnose such tests? The US seems to like flash x-ray radiography, but IIRC back in the early days cruder techniques were used.

  11. Tom (History)

    Egad… my very own thread.

    Thanks Jeffrey.

    I do have one comment/question about the possible contamination of the May 30 tests with Indian Pu though…

    There’s a paper by Z. Mian and A. H. Nayyar hosted by Princeton as part of their Program on Science & Global Security called “An Initial Analysis of Kr Production and Dispersion from Reprocessing in India and Pakistan”.

    The paper in question has a fairly detailed explanation of the behaviour and detectability ranges of krypton as part of Pu production in both countries which suggests that while krypton from Pakistani production is detectable in India the reverse is not true due to the prevailing wind conditions and gaussian plume behaviour.

    So… my question would be how much difference is there between airborne sampling for krypton and Pu?

    Admittedly with out knowing the exact circumstances of the sampling of the May 30 test it would be hard to say if Los Alamos or Livermore is right but it would at least give a basis for judging the claims would it not?