[The problem with writing blogs on complicated subjects is that you always have to leave something out. I know, for example, that one can’t make an H-bomb, a fusion device, without using a garden variety fission weapon, an atomic bomb, to trigger the fusion reaction.[1]Fred kind of skipped over that point last time, and mostly assumed that if Israel had the technology to make A-bombs eventually it would learn how to make H-bombs as well. After all, we and the Russians did, so why couldn’t the Israelis?

Why, indeed? Here at Elemental Zoo Two we like facts, if we can get them, rather than assumptions. So I asked Fred if he could review what he had, go back down memory lane one more time, and give me some specific reasons why he thinks that, today, Israel can make fusion weapons. This is important, of course, because he originally estimated that Israel would have 200 of the things built and ready to go to war by 2030.

Fred said he would do that, but we shouldn’t expect him to talk like a physics instructor. He’s not one of those but he is an amateur historian, and can tell a good story that way.

That sounds fine to me.]

Thanks, G. You know, when we started this little project you didn’t tell me I might be writing a trilogy or more. But I agree on one point. Perhaps the best way to understand what Iran is doing is to look at what Israel has already done. Perhaps the Iranians are simply trying to emulate an obvious success story.

[Hey, I didn’t go that far. I didn’t say that the Israelis are destabilizing the Middle East with their obviously successful nuclear program!]

Anyway, let’s start todays little story with a discussion of fissile materials, i.e., the substances that, with a little help from humans, go boom!


As you know, in late 1960 U.S. intelligence confirmed that, in 1958-1959, Israel had started construction of a “sizeable reactor complex” that was “intended at least in part for the production of weapon-grade plutonium.[2] The construction was at Beersheba, Israel, and today the site as known as the Dimona reactor. Plutonium [or Pu-239][3] is, of course, one of the two substances identified at that time as possible fuel for atomic weapons. The other was U-235, an isotope of natural uranium [U-238].[4]

In the 1960s and 1970s Pu-239 was thought to be the easier fuel to make if smaller powers wanted to break into the nuclear club. [5] “All” they had to do was irradiate natural uranium; this would create a mixture of uranium, plutonium and other fission products; and thereafter the plutonium could be separated out “in a chemical separation plant.” Back in 1974 U.S. intelligence estimated that 16 countries [including Israel] could do that at some level.[6]


The other possibility was to take natural uranium, which ordinarily contains about .71 of a percent of U-235, and enrich that up to 90%; that could provide enough U-235 to make a bomb. But it wasn’t easy to make U-235 in 1974. “The method of enrichment commonly used [was] gaseous diffusion.” As practiced then the process required lots of electrical power, and “[had] not been practical on a small scale.” In 1974 gaseous diffusion facilities had been built only by the five nuclear powers.[7] That is, by the U.S., the U.S.S.R., the U.K., France and China.

[OK, I think I’m beginning to see why our people thought Israel was building plutonium-based bombs. If they had been refining U-235, even we might have noticed when the lights went dim in and around the country, and especially near Dimona. Apparently we didn’t see that, so plutonium was the answer.]

Something like that.[8] But even in 1974 our people thought enrichment technology was advancing, and ultimately U-235 could be refined more easily. The U.K., West Germany[9] and the Netherlands, as a consortium called Urenco, were experimenting with “the commercial use of gas centrifuges to enrich uranium for power reactor fuel.”[10] Urenco had two plants scheduled to open in 1976; and was negotiating 10 year contracts for enrichment services. Ultimately it intended to have “enough capacity to satisfy the fuel needs of 25 major reactors by 1985.”[11] By the way, Urenco is very much in the enrichment business today.[12]

[So what’s your point? Are you saying that the Israelis developed facilities similar to the one [or ones] planned by Urenco? Can the Israelis today produce enriched U-235, as well as Pu-239, to make nuclear weapons?]

I can’t really say based on the things I’ve read to date. But from all the talk about Iran’s centrifuges, it seems likely that country is pursuing the path Urenco followed many years ago. Why couldn’t the Israelis do the same thing? My guess is, probably they have. After all they are technically very competent and, unlike the Iranians, can operate free of any pesky outside inspections.


Now let’s turn briefly to the 1987 study done by the Institute for Defense Analyses. As we said last time, that document was declassified and released earlier this year. [13] Briefly, the study was very complimentary of Israel’s nuclear industry. The facilities at Dimona/Beersheba and other places, it said, were “the equivalent of … Los Alamos, Lawrence Livermore and Oak Ridge National Laboratories.”[14] In particular, Israel’s “SOREQ [today, the Soreq Nuclear Research Center”[15]] … [ran] the full nuclear gamut of activities from engineering, administration and non-destructive testing to electro-optics, pulsed power, process engineering and chemistry and nuclear research and safety.”[16] SOREQ had “the technological base required for nuclear weapons design and fabrication.”[17]

The IDA concluded SOREQ was “still hampered in being able to design and produce fusion weapons or other more complicated devices utilizing fusion and fission in the same configuration.”[18] I think that’s another way of saying Israel was having problems designing a deployable H-bomb. However,

It should be noted that the Israelis are developing the kind of codes which will enable them to make hydrogen bombs. That is, codes which detail fission and fusion processes on microscopic: and macroscopic leveI. However, it is doubtful they have the codes to completely design such devices as they involve more exotic radiation transport and are multidimensional. The Israelis do not yet have the capability to carry out these kinds of calculations.[19]

So what’s a country to do if it lacks the capability to make calculations of this type? Most likely it shops for a supercomputer or two, or three, or four.[20]


As you might guess, supercomputer exports were and are regulated in this country. I was going to say “tightly regulated,” but in view of what happened in 1996, perhaps that’s not so correct. Anyway, there’s an interesting report from the Congressional General Accountability Office [the GAO] that covers the subject from 1996 to 2006.[21]

The regulatory structure was as follows:

  • Only high performance computers were regulated. A machine was considered “high performance” if its capacity to process data exceeded a specified level. In 1996 that level was specified in MTOPS [millions of theoretical operations per second].
  • Not all potential exports were treated the same. The destination of the export was very important, and was evaluated from the standpoint of U.S. national security interests.
  • In 1996, for example, the executive branch organized countries into four computer tiers, with each tier above tier 1 representing a successively higher level of concern. Exports to higher tier countries required a government-approved license.
  • Very little or no hardware license requirements were placed on exports to Tier 1 or 2 countries. Tier 1 and 2 countries primarily included those in Africa, Asia, Central and Eastern Europe, Japan, Latin America, and Western Europe.
  • Exports of computers to tier-3 countries, such as China, India, Israel, Pakistan, and Russia, required a license.
  • Exports to tier-4 countries, such as Iran, North Korea, and Syria, essentially were prohibited.

[You know, I remember hearing about all this back in the day, but I never took the time to sort it out. This is really very helpful.]

Yes, the folks over at GAO often are that way.

[So what happened in 1996?]

Well Cray Research, an early manufacturer of supercomputers[22], wanted to sell a couple of its devices to Israel; so as required Cray applied for a license. The license was turned down by reviewers, reportedly because they were concerned the computers would be diverted to military purposes. “According to Cray, the negative vote by the Subgroup on Nuclear Export Controls (SNEC) was due to the fact that the machine ‘was going into a black hole because the site was classified as off-limits’ It might be used to design nuclear weapons or long-range missiles. The recipient [Israel] was ‘known but not trusted…’”[23]

[Well that sounds fairly definitive. How did the Israelis get around that? Or did they?]

Yes they did. Remember, a computer is “super” only if its capacity to process information exceeds a certain level. In 1995 the Cray machine operated at 2,625 MTOPS, and that was enough to require an export license. However, in 1996 someone persuaded the Clinton Administration to raise the baseline for defining a supercomputer to a minimum of 7000 MTOPS. So the Cray machine was demoted with the stroke of a pen; it was no longer “super;” a license was not required to export it; but its capacity to design nuclear weapons, etc. remained the same.

[This is a great country, isn’t it? Somebody did all of that just for Israel?]


Oh, no, I didn’t mean to imply that. Other countries took advantage once the gates were opened. Even the Russians bought the Cray machine.[24]

I’m just trying to make the point that, to the extent Israel lacked advanced computing power for its nuclear program, the problem was solved by 1996. That was 19 years ago. Since then the Israelis have had ample time to develop and refine their designs and deploy whatever number of thermonuclear warheads, etc. they might want. So I see no reason to change my initial estimate that they could have 200 such devices by 2030; indeed, they may already have them.

[Thanks, Fred. Once again you’ve amply justified the assumptions you made in your initial war scenario. Frankly, I don’t see why the Israelis think they need so much thermonuclear firepower. Surely not just to counter a few terrorists wearing dynamite jackets and carrying shoulder-fired missiles? Do they have greater ambitions? Or perhaps the Israelis are lonely, and just want to join the rest of us in the world of “nuclear giants and ethical infants.[25]

I don’t know.]

[1] Do you want an explanation of that, or do you just want to be depressed. If so, check out Wikipedia at nuclear weapon, at https://en.wikipedia.org/wiki/Nuclear_weapon

[2] See Post-Mortem in SNIE 100-8-60, Implications of the Acquisition by Israel of a Nuclear Weapons Capability, (Draft) at p. 1, par. 1, available at http://nsarchive.gwu.edu/nukevault/ebb510/docs/doc%2027A.pdf

[3] Today it seems to be used mostly by NASA. See the Wikipedia piece on Plutonium at https://en.wikipedia.org/wiki/Plutonium-238

[4] See the discussion at Nuclear Fission, a site hosted by Georgia State University, and available at http://hyperphysics.phy-astr.gsu.edu/hbase/NucEne/fission.html

[5] See Special National Intelligence Estimate, Prospects for Future Proliferation of Nuclear Weapons (Aug. 23, 1974) at p. 20, par. 37. The document was issued by the Director of National Intelligence, and is available at http://nsarchive.gwu.edu/NSAEBB/NSAEBB240/index.htm

[6] Id. at par. 1A4, 1A5.

[7] Id. at par. 1A6.

[8] See the discussion of gaseous diffusion at Global Security.org, available at http://www.globalsecurity.org/wmd/intro/u-gaseous.htm “Diffusion equipment tends to be rather large and consumes significant amounts of energy. The main components of a single gaseous-diffusion stage are (1) a large cylindrical vessel, called a diffuser or converter, that contains the barrier; (2) a compressor used to compress the gas to the pressures needed for flow through the barrier; (3) an electric motor to drive the compressor; (4) a heat exchanger to remove the heat of compression; and (5) piping and valves for stage and interstage connections and process control. The entire system must be essentially leak free, and the compressors require special seals to prevent both out-leakage of UF 6 and in-leakage of air. The chemical corrosiveness of UF 6 requires use of metals such as nickel or aluminum for surfaces exposed to the gas (e.g., piping and compressors). In addition to the stage equipment, auxiliary facilities for a gaseous-diffusion plant could include a large electrical power distribution system, cooling towers to dissipate the waste process heat, a fluorination facility, a steam plant, a barrier production plant, and a plant to produce dry air and nitrogen.”

[9] Remember, this was the Cold War. Germany was split in two, East and West.

[10] Id. at p.10, par. 7 & n.1 “The centrifuge process involves high speed spinning of uranium in gaseous form in cylindrical containers through many iterations, with the lighter isotope (U-235) gathering towards the center of the tube.”

[11] Id. at p. 10, par. 7.

[12] See Urenco, enriching the future, at http://www.urenco.com/

[13] See Townsley & Robinson, IDA Memorandum Report No. 317, Critical Technology Assessment in Israel and NATO Nations (April, 1987), available at http://fas.org/nuke/guide/snie4-1-74.pdf

[14] Id. at p. III-4

[15] Today the site address is http://www.soreq.gov.il/default_EN.asp   Good luck getting any information out of it.

[16] Id. at p. III-4.

[17] Id. at p. III-4.

[18] Id. at p. III-4. “As far as nuclear technology is concerned the Israelis are roughly where the US was in the fission weapon field in about 1955 to 1960.”

[19] Id. at p. III-4.

[20] For a really basic introduction – the kind I like – to supercomputers, check out the one at Wikipedia. It’s at https://en.wikipedia.org/wiki/Supercomputer#Applications

[21] See General Accountability Office, GAO-06-754R, Export Controls and High Performance Computers (2006), available at http://www.gao.gov/assets/100/94283.pdf

[22] Cray’s successor company is still with us. See Cray at http://www.cray.com/

[23] See The Risk Report, Israel Gets U.S. Supercomputer for Secret Military Site (May-June 1996), http://www.wisconsinproject.org/countries/israel/israelgets.htm “Cray’s application was voted down in October 1995 by a group of experts from five federal agencies because the computer was destined for a secret site where its use could not be verified. Officials were concerned that the machine could be used to design nuclear weapons or long-range missiles. ‘The policy was that if the export is going to a proliferant and you can’t safeguard it, and it is not part of a government-to-government program, you deny it,’ said a U.S. official familiar with the case.” See also Geek, U.S. eases supercomputer export controls (Jan. 3, 2002), at http://www.geek.com/chips/us-eases-supercomputer-export-controls-548688/

[24] See The New York Times, Gerth, Clinton Administration Tightens Rules on Supercomputer Sales to 13 Foreign Groups (July 1, 1997), available at http://www.nytimes.com/1997/07/01/world/clinton-administration-tightens-rules-supercomputer-sales-13-foreign-groups.html

[25] That’s from Omar Bradley. See Oxford Dictionary of Quotations (6th Edition) (Oxford, 2004) at p. 148, Omar Bradley, n. 23. “The world has achieved brilliance without wisdom, power without conscience. Ours is a world of nuclear giants and ethical infants.”