Jane VaynmanHow long would it take Japan?

I’ve usually taken it as common wisdom that Japan, if it suddenly wanted to, could build a nuclear bomb in about 6 months. Articles and experts have quoted this figure.

However, there are reports quoting the Japanese newspaper Sankei of an internal Japanese report that puts the estimate closer to 3-5 years. The Japanese government denies the existance of this study:

The Japanese daily Sankei reported that experts at several government organizations concluded it would take at least three to five years to make a prototype weapon.

The experts also estimated that the project would cost about $1.68 billion to $2.52 billion and require the efforts of several hundred engineers, according to Sankei.

The experts did not say whether Japan should develop nuclear arms, the newspaper reported, only what such a project would require. The newspaper published a summary of the document, dated Sept 20 and titled “On the Possibility of Developing Nuclear Weapons Domestically.”

[Do any Japanese readers of ACW, or those with FBIS access, want to take a look for the original report and send us links or a translation if FBIS has it?]

I don’t know, this all sounds a bit odd to me. I am not sure where the 6 months claim originaly came from, and it does seem a bit fast now that I think about it. Maybe it comes from trying to put some numbers to as ‘little as a year’s time’ or maybe from a Septermber 2005 Forbes article which, as far as I can tell, first referenced a Japanese official saying that, it would take 183 days. (Exactly. On the dot. not 182 or 184…. what?!)

But 3-5 years? Why would it take a country with plenty of fissile material, reprocessing techonology, an advanced nuclear science establishment and a healthy economy, 5 years to build a prototype nuclear device? Some are worried that Iran could do it in 5, and they have nowhere close to the capablities and materials access of Japan. With a full industrial commitment, the US and Russia both did it in about 4 years starting from scratch. South Africa began working on a bomb design in 1971, with more intense work starting in 1974 and were ready for a “cold test” (without HEU) in 1977.

Two billion dollars is a lot of money, it is probably not something that would be a limiting factor if Japan wanted a bomb. Also, Japan gets about 35% of its electricity from 55 reactors. Would finding several hundred engineers be much of an effort? A few points of comparison: construction of the Rokkasho-mura reprocessing plant cost about 20 billion and employed 5,000 people.

ITAR-TASS says said that the difficulty in mobilizing the man power would be a moral issue rather than a techinical one: [I was emailed this article, so still looking for a link.]

The authors of the report said Japan possesses enough plutonium to manufacture thousands of warheads but such a project would have to mobilize several hundreds of engineers.

The latter will not be an easy task since the overwhelming majority of Japanese nuclear physicists are against turning their country into a nuclear power, an informed expert told Tass.

—Moscow ITAR-TASS in English 0148 GMT 25 Dec 06

Sure the physicists do not want to make weapons now, but that position is consistant with general public opinion in Japan. If public views were to shift – which would be a major change, but according to some experts, may be possible – I can imagine that the opinions of the scientific community would be just as likely to shift along with other Japanese citizens.


  1. Allen Thomson (History)

    Depends on what you mean by a “bomb”. Japan could do a Little Boy in not many months; a clunky Fat Man in a bit longer; light-weight, small-diameter, boosted fission devices longer still, about as long as a basic two-stage device. Bleeding-edge two-stage weapons suitable for MIRVed ballistic missiles would take a fair number of years, probably testing, etc.

  2. Haninah

    Two questions:

    1) Is it possible that by “prototype” they don’t mean Hiroshima-type, but a prototype of a modern (second- or third-generation) warhead? I mean, I can imagine that Japan could throw together a gun-type nuke in no time flat – presumably if they launched a nuclear program, they’d be more ambitious then that, and 3-5 years to go from zero to third-generation (something no country has ever done, as far as I know at least) sounds about ambitious enough for Japan.

    2) Not to be nitpicky or anything, but isn’t it a bit untrue that Russia built its first nuke from scratch? I mean, Klaus Fuchs ‘n’ all?

  3. John Field (History)

    Why would they want to do this?

    Maybe it would take so long because they don’t want to screw up the way the US and Russia did. Maybe they would want to build weapons that are really safe and that you can actually take apart and not want to contaminate miles around the factories and poison the workers. And, maybe they would design their primaries so as not to fool around in endless debates about LEP and performance margins 30 years from now.

    And, maybe after doing everything responsibly, they would come to the conclusion that it isn’t really such a great idea after all. In sum, just maybe they aren’t a bunch of idiots like us.

  4. Mark Gubrud (History)

    To state the obvious, the length of time it would take Japan to produce one or a few nuclear weapons, or to begin producing many, would strongly depend on the sense of urgency and hence the level of popular support and personal commitment of the scientists and engineers, resources made available by the government, and relaxation of safety and environmental protection requirements. So, depending on the situation in which the decision was taken to make nuclear weapons, the time to the first bomb might range from a few months to a few years.

  5. jane (History)

    Haninah – you are right, the Russians building the bomb from scratch was an oversimplification. The first Russian bomb was essentially a copy of the American design. Only a few people, including Kurchatov at the head of the project, knew that they were working from intelligence. David Holloway’s book, Stalin and the Bomb is one of the best sources on this history.

  6. MTC (History)

    Going from the information in the original articles on the front page and page 5 of the Sankei Shimbun of December 25, all of which is hearsay:

    1) An internal government report focuses on the possibility of Japan developing a miniaturized nuclear warhead, i.e., one that can be mounted on a missile.

    2) The analysis finds that the easiest route is to separate U235 from U238 using centrifuges. However, the report sees no way of appropriating the funds or acquiring the physical plant for such a program.

    3) The analysis allegedly finds that Japan’s current plutonium stockpile, reprocessed from the spent fuel of light Pu 240 in it for conversion into effective warheads. It also finds that the Toshiba laser process for separating Pu 239 from Pu 240 still unproven.

    The conclusion is that Japan would either have to build a graphite moderated nuclear reactor for Pu 239 production or embard on a crash program in uranium enrichment.

    4) The analysis finds that Japanese engineers and scientists have the knowhow to develop the implosion lens for a plutonium weapon.

    However, while it not impossible to test the design and the manufacturing of the implosion lens with achieving criticality, it would still be expensive and take a lot of time before a body of sub-critical tests resulted in a credible Japanese nuclear deterrent.

    5) In essence, the analysis provided strong technical reasons for continuing the current political policy of a non-nuclear Japan relying upon the protection of an extended U.S. nuclear umbrella.

  7. Robot Economist (History)

    I think the 5 year timeline for Iran that is frequently bandied about is time to a nuclear explosion test. It would probably take them much longer to get a nuclear down to a managable size that could be mounted on the tip of a Shahab II or III – which is something most analysts in the intelligence community overlook.

    The same could be said about the Japanese. Sure, they have some of the infrastructure and brainpower needed to support a nuclear weapons industrial complex, but they don’t have the experience or knowledge required to kick out a stream of small, reliable warheads.

    I haven’t read the article yet, but there is also the issue of putting a delivery system together as well. The Japanese have F-16s, which are capable of carrying nuclear-tipped cruise missiles, but they don’t have mid-air refueling assets.

  8. Strabo The Lesser

    It took the US four years in the 1940’s using slide rules and mechanical computers. Given material, a Japanese project would be much faster due to more advanced knowledge and better engineering.

Pin It on Pinterest