Jeffrey LewisPakistani Design In Switzerland

Greetings from Beijing

We now have David Albright’s essay on the advanced nuclear weapons design, as well as stories in the New York Times and Washington Post. (David Sanger, Nuclear Ring Reportedly Had Advanced Weapon, June 15, 2008; David Sanger and Bill Broad, Officials Fear Bomb Design Went to Others, June 16, 2008; Joby Warrick, Smugglers Had Design For Advanced Warhead, June 15, 2008.) There is a lot to sort through.

I have only a few comments to add to James’s excellent post.

This is the second design peddled by the AQ Khan network — and it seems to be considerably more advanced that the Chinese design found in Libya. (If you are interested in the Chinese design, which had a yield of 10 kilotons and mass of 500 kilograms, check out my post: More on Libya’s Bomb Design, October 8, 2005.)

According to Sanger, the design “is half the size and twice the power of the Chinese weapon…” and, writing with Bill Broad, and happens to “bear a strong resemblance to weapons tested by Pakistan” in May 1998.

The statement about 1/2 the size and 2x the yield is probably not true — though it may not matter very much. What does seem likely is that the device is small enough for the Nodong family, which includes Pakistan’s Ghuari and Iran’s Shahab.

1/2 the Size

In the second story, Sanger and Bill Broad describe the device as “miniaturized — using about half the uranium fuel of the older design to produce a greater explosive force.”

Warrick describes the Chinese design as “bulky and difficult to deliver”; Sanger and Broad, “unwieldy but effective.” That language suggests the issue is not mass, but diameter.

Note that half the diameter, half the mass and half the fissile material are all very different things. For example, a bomb with 1/2 the diameter would have 1/8 the mass.

I find the fissile material statement plausible — Hibbs and Albright claim that Pakistan used about 15 kilograms of HEU in its design. Cutting that number in half is quite plausible — Paine and Cochran calculated that a state with medium technical skill could squeeze 10-20 kt out of 7-9 kilograms of HEU. So, that’s in the neighborhood.

But reducing mass — and more important diameter — is more about reducing the amount of conventional high explosive (HE) that surrounds the fissile material. (See my discussion on this topic in the context of North Korea.) A physics package of this sort is, by mass, probably 90-95 percent explosives.

A desire to reduce the amount of explosives would explain the emphasis on advanced electronics in the Sanger and Broad story — less jitter, less explosive:

Clearly, someone had tried to modernize it, to improve the electronics,” one said. “There were handwritten references to the electronics, and the question is, who was working on this?”

If the device is, say 60 centimeters instead of 90 cm, than it might weigh as little as 200 kilograms instead of nearly 600.

You may remember the issue of electronics from James Risen’s State of War — the allegation that the US paid a Russian to give the Iranians a faulty firing set design to retard their program. Same thing, as far as I can tell from the news reports.

Oh and 60 centimeters isn’t an arbitrary number. More on that is a moment.

2x the Yield

Alright, this one I have more trouble with.

Pakistan’s nuclear tests on May 28 and May 30 had yields of 9-12 kt and 4-6 kt, based on the seismic signals. (The announced yields were 40-45 kt in five devices and 15-18 kt in the sixth device.)

In other words, I don’t understand how a Pakistani device that was tested can have twice the yield of the 4th Chinese test, because those two events had about the same bang.

Indeed, Warrick — in contrast to Sanger — implies that the yield is the same:

The lethality of such a bomb would be little enhanced, but its smaller size might allow for delivery by ballistic missile.

“To many of these countries, it’s all about size and weight,” Albright said in an interview. “They need to be able to fit the device on the missiles they have.”

I lean toward thinking the yield of the device is about the same, though I am open to be persuaded otherwise.

Implications for Iran

First, let me begin by noting that the Chinese design was deliverable by a missile — it was tested on a DF-2. But that’s just being pedantic — the DF-2 had significantly more throw-weight than the Nodong/Ghuari/Shahab. The 1966 Chinese design, as I have said before, wouldn’t be much good for Iran.

Some of you may recall the question that David Albright raised about the so-called Laptop of Death — that the RV was designed to accommodate a warhead of only 60 centimeters in diameter:

Another important question that is sidestepped by the misleading use of warhead in the article is whether Iran can build the relatively small nuclear warhead able to fit into the triconic re-entry vehicle apparent in photos of a 2004 flight test. Based on publicly available photos of the 2004 test launch, the nuclear warhead would require a diameter of about 600 millimeters. Achieving such a diameter would be challenging for Iran. For example, the diameter of the warhead in the design provided to Libya (and perhaps to Iran) by A.Q. Khan was about 900 millimeters. A legitimate question is whether Iran could successfully build such a small nuclear warhead without outside help.

Obviously, then, the 1998 Pakistani design may be helpful for a state like Iran.

On the other hand, it isn’t clear to me that simply receiving a design, without any domestic testing experience, is enough to have confidence. As North Korea discovered in trying to jump to a weaponizable device, there are apparently still a few tricks of the trade.

Still, interesting stuff.


  1. hass (History)

    Lets not let all the technical speculation hide the fact that there’s no evidence that Iran has this information — or wants this info — or needs it. As you yourself cited in a previous post, “Bronze medal” nuclear weapons design information is already far more accessible (in the public domain, in fact) and more useful for a country that wants to build nukes. If Iran wanted nuclear design information, all it needed to do was to visit your local friendly declassified government reading office.

  2. Lao Tao Ren (History)


    Welcome to the capital.

    The only thing I have to add to your excellent post is that there is one possibly compelling reason for DPRK to go straight to a weaponized design: that was a customer requirement.

    Hazarding a guess, the fizzle meant, no payday.

    And the world became a safer place as a result of it.

    The customer would not only have to be prepared to pay a lot, it would have to be paid in something that is safe from US embargoes against DPRK which included virtually locking them out of the international banking system. Typically that means being paid in gold.

    The customer is also likely to be a premier electronics powerhouse, though not necessarily in the electronics needed for nukes, which are not state-of-the-art in the customary commercial sense, but optimized for reliability, rad-hard, resistance to corrosion, speed and consistency, and a few other characteristics you just don’t need for your MacBook….

    Electronics engineers with this expertise are kind of scarce, but not impossible to find in CIS, South Africa, and a few other places.

    But again, this is idle speculation.

    While you are in Beijing, go for the Beijing Duck, the Mongolian BBQ, and the Olympics Mascots —- great gifts.

  3. China Hand (History)

    I find the timing of this revelation interesting. Musharraf seems to be hanging by a thread, and it looks like the civilian government is ready to rehabilitate Khan. Not allowing foreign questioning of Khan is becoming an assertion of Pakistani national pride and sovereignty. Maybe Albright’s essay with its explosive allegations of warhead related shenanigans is an effort to put Khan on top of the US-Pakistan agenda before he slips out of the grasp of the international community for good. Since US-Pakistan relations are entering the deep freeze thanks to the divergence of interests vis a vis the Taliban in Pakistan’s border regions, I think the prospects for the IAEA or anybody talking to Khan are slim.

  4. erichwwk (History)

    Nice Posts, both this and the one by James, asking all the right questions.

    So. Has David actually SEEN the weapons design? Enough to know whether it is one stage or boosted? Enough to get his own clues as to origin? I somehow find David Sanger’s “resemblance to Pakistani gadgets” not satisfying. Is all we have to go on re origin yield and physical size?

    Exactly where do you think anyone would be going with a ME nuclear weapons program? A non- missile delivery does get around the CEP and return address issues, but raises an additional one of wind force damage being dissipated by ground detonation. Or is this seen as an intermediate step, to a full blown ME Arab deterrent to Israel? How does our support of Saudi nuclear industry figure in all this?

    Also just because ones supplier has a product, does not, by itself identify the (any) end user- suppliers are often ahead of the consumer in product line.

    So that too makes me think the elephant in the room here is the timing.

    As David points out in his essay, the Libya design was moved to DC for safe keeping ( and as evidence to help with determining source of future design?) That the Swiss design was destroyed tells me something else was at play. Was the design perhaps of US origin? A US design with planted flaws? Why else would it be destroyed?

    And more importantly, why now? Is something afoot with Saudi Arabia or Israel?

  5. SQ

    This is an outstanding post, the sort of thing we all tune in for.


    It seems to be a HEU design that we are talking about, so it wouldn’t help the NKs that much, unless they’re much further along the uranium path than anyone imagines.

    China Hand:

    That’s too convoluted for me. The timing seems to be driven by revelations connected to the Tinner trial in Switzerland.

  6. Anonymous

    A related issue to all of this — Albright demands that the US gain access to Khan for questioning.

    But the real question folks should be asking is: why would talking to Khan help at all? From his media bluster, it is quite apparent he hates Westerners and denies everything. How would a U.S. officer question Khan in a way that would produce anything other than lies?

    I can’t think of beginning to do this without (1) taking Khan out of country and/or (2) employing more coercive interrogation means — and I don’t think either of those would work in any case, for many reasons. Nor is either of them feasible from a Pakistani standpoint.

    So let’s just drop the issue of gaining access to Khan already.

  7. Lao Tao Ren (History)

    Since we are splitting hairs, I was wondering if you can tell us what is the diameter specification for the physics package that goes into the four (4) Minuteman Nosecone Assemblies sent to Taiwan for reverse engineering?

    I seem to recall that 600mm is roughly the right diameter for the physics package.

  8. blowback (History)
  9. John McGlynn (History)

    What about Gordon Prather’s June 21 column*? Prather, who seems to have gobs of nuclear street cred, charges David Albright with being a know-nothing bonehead, at least in regard to Khan/Pakistan nuke technology. What say the wonks?

    *Another Tenet Sting Failure?
    June 21