The Twilight of the Bombs

by Richard Rhodes

  Because nuclear weapons are well protected, national-security bureaucracies have postulated that a terrorist group that wants to acquire a nuclear capability will be forced to build its own bomb or bombs. Enriching uranium or breeding plutonium and separating it from its intensely radioactive matrix of spent fuel are both well beyond the capacity of subnational entities. The notion that a government would risk its own security by giving up control of a nuclear weapon to terrorists is nonsensical despite the Bush administration’s use of the argument to justify invading Saddam Hussein’s Iraq. A nuclear attack on United States interests by a terrorist group using a donated bomb would certainly lead to a devastating nuclear counterattack on the country that supplied the weapon, provided the supplier could be determined—a near certainty with nuclear forensics and other means of investigation.

  A terrorist group in possession of a homemade or stolen nuclear device is assumed to be undeterrable, however, and probably would be. “The more than 10,000 warheads5 in the U.S. nuclear stockpile on Sept. 11, 2001,” Stephen Schwartz wrote to the point, “had no impact on al Qaeda’s calculations that day or in all the days that followed.” More than any other single threat, more even than fear of a large-scale anthrax attack, the fear after September 11 that the next such assault might be nuclear changed the views of government leaders everywhere. The theory and practice of nuclear deterrence, whatever its real merits, had made the long nuclear standoff between the United States and the Soviet Union nerve-wracking but tolerable. Suicidal terrorists armed with homemade nuclear weapons are a fresh nightmare.

  But could a terrorist group build a deliverable atomic bomb? The United States has quietly explored the question for years. The earliest and perhaps best-known investigation was conducted in the mid-1960s at what was then the Lawrence Radiation Laboratory in Livermore, California (Lawrence Livermore today). The project was called the Nth Country Experiment; its purpose was to explore the “technical problems facing a nation6 wishing to acquire a small stockpile of nuclear weapons”—in other words, to see if a nonnuclear state without access to secret technical information could successfully design and build a nuclear weapon. A state has more resources than a terrorist group, to be sure, but the experiment was set up to test whether a minimal team—in this case, two young American physicists, recent Ph.D.’s without nuclear-weapons-design knowledge—could invent, on paper, a workable atomic bomb. “The goal of the participants,”7 their instructions read, “should be to design an explosive with a militarily significant yield.”

  The two novices succeeded, after three man-years of work. Significantly, they decided at the outset of their project, after a little preliminary consideration, that designing a gun bomb would be too easy—would not adequately show off their abilities—and moved on to the successful design they presented to their superiors, which was an implosion device.

  In the late 1980s, several experienced Los Alamos weapons designers, including Carson Mark, Ted Taylor, and Jay Wechsler, published a report which attempted to answer the question, “Can terrorists build nuclear weapons?”8 Experienced professionals can find it difficult to forget what they know, and it would seem the Los Alamos designers were not able to return themselves to a state of technical naïveté. What they called a “crude design” turned out to be an implosion device.

  For a “crude design,” they wrote, “terrorists would need something like 5 or 6 kg of plutonium or 25 kg of very highly enriched uranium (and more for a gun-type device).” The Los Alamos designers thought it “exceedingly unlikely” that any single individual could acquire all the necessary knowledge and skills to make an atomic bomb by himself, so they assumed “that a team would have to be involved.… Their number could scarcely be fewer than three or four and might well have to be more.” They would need months; they would need a secure location; they might very well irradiate themselves fatally by accident or blow themselves up. But if a team of terrorists weathered all these challenges, the Los Alamos designers concluded, it should be able to produce a weapon with a minimum yield of several tons of TNT equivalent, and very possibly of one hundred tons or more. Indeed, they wrote, “devices employing metal in a crude design could certainly be constructed so as to have nominal yields in the 10 kiloton range—witness the devices used in 1945.” The Hiroshima bomb—a gun-type weapon—exploded with a yield estimated to have been equivalent to 13,500 tons of TNT, 13.5 kilotons.

  An Ohio State University political scientist named John Mueller has argued that a terrorist attack with a nuclear weapon is extremely unlikely.9 Mueller has made a career of healthy skepticism about nuclear strategy, countering the persistent American politics of nuclear threat inflation with sensible arguments and solid facts. In a 2009 research paper for the International Commission on Nuclear Non-proliferation and Disarmament, he examined in detail the question of what he called “the atomic terrorist.”10

  Mueller minimizes—in my opinion, correctly—the risk of terrorists’ stealing a nuclear weapon or acquiring one by donation, for the reasons I’ve already reviewed. He correctly eliminates plutonium as a likely terrorist explosive, although he does so for the wrong reasons. Plutonium is not, as Mueller and many others seem to think,11 more dangerous to work with than uranium—not significantly more radioactive or toxic—but it is far more susceptible to predetonation during assembly. In plain English, firing one piece of plutonium into another would result in a meltdown, not a substantial explosion. That was why Los Alamos under Robert Oppenheimer scrambled to invent the implosion method of assembly in the final year of the Pacific War: A sphere crushed to critical density by high explosives doesn’t have time to predetonate.

  Acquiring enough HEU would be difficult, Mueller argues. No market for it exists; it’s carefully guarded; inside collaborators would have to fear being killed upon delivery rather than paid off; governments would deploy extensive security forces and offer huge rewards when the theft or diversion was discovered; crossing international borders with stolen HEU would be challenging. All these arguments ring true.

  Less certain is Mueller’s bomb-construction scenario. He thinks the terrorist team would need engineers and specialists, a machine shop, and at least months “or even a year”12 to fashion its bomb, and perhaps it would, but none of those requirements seems extravagant. He quotes Stephen Younger, the Los Alamos physicist and weapons designer who served as director of the Defense Threat Reduction Agency during George W. Bush’s first term, emphasizing the “daunting problems associated13 with material purity, machining, and a host of other issues.” Younger concluded, “to think that a terrorist group, working in isolation with an unreliable supply of electricity and little access to tools and supplies, could fabricate a bomb or IND [improvised nuclear device] is far-fetched at best.” Younger seems to be describing Al Qaeda trying to build a bomb in the caves of Tora Bora—where else would the supply of electricity be unreliable and tools and supplies scarce?—and his condescension toward the bomb-making capabilities of terrorists may be complacent.

  I would certainly prefer to believe that a terrorist bomb is beyond the technical and organizational skills of terrorists. Two personal experiences with physicists deeply knowledgeable about nuclear-weapons design leave me skeptical of such a conclusion. In 1986, when I helped the Nobel Laureate physicist Luis W. Alvarez write his memoirs, Alvarez commented one day on the importance of keeping track of highly enriched uranium. Speaking of a quantity sufficient to form a supercritical mass, he told me, “You can make a fairly high-level nuclear explosion just by dropping one piece onto another by hand.” Alvarez had been intimately involved in the development of the first atomic bombs, and knew whereof he spoke. Some years later I gave a talk at the National Atomic Museum in Albuquerque during which I quoted Alvarez’s comment. An audience member from one of the nearby weapons labs, Los Alamos or Sandia, told me afterward with some enthusiasm, probably speaking out of turn, “Yes, and if you can accelerate it even a little you can get a much bigger explosion.”
r />   Equally unsettling was my encounter with Ted Taylor, the former Los Alamos weapons designer whom John McPhee14 credits in his book The Curve of Binding Energy with having pioneered the miniaturization of U.S. nuclear weapons. I interviewed Taylor after the 1993 truck-bomb attack on the World Trade Center, the precursor to 9/11. Speculating about nuclear terrorism, Taylor told me he’d wondered how many people it would take “to start World War III.” To build a nuclear weapon, he meant, and detonate it under conditions that would lead the nation that was targeted to believe some other nuclear power was behind the attack and to retaliate with a nuclear strike, beginning a catastrophic nuclear exchange. He said he’d concluded it would only take one person. I thought then and think now that one person would suffice only if that person were Ted Taylor. But even if Taylor, who died in 2004, was one of a kind, his informed speculation that a minimum number of conspirators with the right training and skills might succeed in building a homemade nuclear weapon is chilling.

  As long ago as 1977, the U.S. Office of Technology Assessment confirmed a small-group scenario:

  A small group of people,15 none of whom have ever had access to the classified literature, could possibly design and build a crude nuclear explosive device. They would not necessarily require a great deal of technological equipment or have to undertake any experiments. Only modest machine-shop facilities that could be contracted for without arousing suspicion would be required. The financial resources for the acquisition of necessary equipment on open markets need not exceed a fraction of a million dollars. The group would have to include, at a minimum, a person capable of researching and understanding the literature in several fields and a jack-of-all trades technician.

  Which is, of course, an approximate description of the subsequent South African program that led to the successful design and construction of a small arsenal of uranium gun bombs.

  All these expert conclusions assume the availability of sufficient HEU to make at least one bomb. Where would Taylor’s lone gunman find one hundred pounds of uranium metal? I discovered one answer to that question, and confirmation of Alvarez’s suicidal hand-assembly scheme, in a report on nuclear security by Matthew Bunn and Anthony Wier of the Project on Managing the Atom at Harvard’s Belfer Center. The two Harvard experts wrote:

  Setting off a nuclear explosion16 with HEU can be done rapidly enough that [U.S. Department of Energy] internal security regulations require that security for U.S. nuclear sites where enough material for a bomb is present be based on keeping terrorists out entirely, rather than catching them as they leave the site, to avoid “an unauthorized opportunity … to use available nuclear materials for onsite assembly of an improvised nuclear device”—that is, to prevent terrorists from being able to set off a nuclear explosion while they were still inside the facility where they stole the HEU.

  Since terrorists who broke into a U.S. (or some other) nuclear site where HEU was available in critical quantities would hardly have the time or tools needed to build a bomb on-site, the DOE requirement for perimeter defense confirms the Alvarez scenario. Bunn and Wier wrote further that the Senate Foreign Relations Committee at one point “asked the three U.S. nuclear17 weapons laboratories whether terrorists, if they had the nuclear material, could make a crude but workable nuclear bomb. The answer given was ‘yes.’” Within a few months “the laboratories had actually built a gun-type device, using only components that, except for the nuclear material itself, were off the shelf and commercially available without breaking any laws. The device was actually brought into a secure Senate hearing room to demonstrate the gravity of the threat.”

  If an improvised nuclear device is relatively easy to make once the fissile explosive has been obtained, then the fundamental requirement for security from nuclear terrorism must be rigorous control of plutonium and (especially) highly enriched uranium. As two security specialists, Ivo Daalder and Jan Lodal, concluded in 2008:

  U.S. nuclear policies remain18 stuck in the Cold War, even as the threats the United States faces have changed dramatically. Today, the gravest threat comes from the possibility of terrorists bent on delivering a devastating blow against the United States acquiring the capacity to do so with nuclear weapons. This threat is compounded by the dangers of nuclear proliferation, as more and more countries hedge against potentially negative developments in their regions by acquiring the wherewithal to build the bomb. Then there is the increasing global demand for nuclear energy, which will spread the infrastructure necessary to produce fissile nuclear materials still wider. The world, in short, is on the verge of entering an age of more nuclear weapon states, more nuclear materials, and more nuclear facilities that are poorly secured—making the job of the terrorists seeking the bomb easier and the odds that a nuclear weapon will be used greater.

  Some of this argument is threat inflation, to be sure, but enough of it is realistic to be disturbing. At the opening of the second decade of the twenty-first century, with both nations and subnational entities becoming capable of fielding nuclear arsenals, the world faces a stark choice: eliminate nuclear weapons and secure their fissile explosives or expect them to be used.

  RICHARD BUTLER has a name for that choice: he calls it the axiom of proliferation. “It was identified by19 the Canberra Commission,” he told a group of students in Monterey, California, one spring day in 2008, “and in some ways I think it was its highest achievement: As long as any state has nuclear weapons, others will seek to acquire them.”

  Australia had taken the lead among states in examining nuclear abolition. In 1995, between Butler’s work on the indefinite extension of the NPT and the CTBT, he had heard from Canberra about an additional but complementary assignment. The Australian prime minister, Paul Keating, had encountered a journal article that had stirred him to action. “The article was about20 how ridiculous and threatening nuclear weapons are,” Butler told me. “Keating read it and said, ‘This is an appalling situation. We should find some way to get rid of nuclear weapons.’” After discussing the question with his foreign minister, Gareth Evans, Keating decided to empower a wide-ranging commission. Evans asked Butler to undertake the work, and from that fortuity the Canberra Commission on the Elimination of Nuclear Weapons was created in November 1995 with Butler as its chairman and a charter “to propose practical steps toward a nuclear weapon free world including the related problem of maintaining stability and security during the transitional period and after this goal is achieved.”21

  Butler invited seventeen knowledgeable and experienced experts and former government officials to join the commission, including Lee Butler, the retired chief of the U.S. Strategic Air Command; Michael Carver, a British lord who had been chief of general staff and chief of defense staff as well as commander in chief of the Far East in the British Army; the Sri Lankan diplomat Jayantha Dhanapala; Rolf Ekéus; the Japanese nuclear engineer and former ambassador Ryukichi Imai; the former U.S. secretary of defense Robert McNamara; the Chinese ambassador for disarmament affairs Qian Jiadong; the former French prime minister Michel Rocard; Joseph Rotblat, the Manhattan Project physicist and Pugwash Conference president who won the 1995 Nobel Peace Prize; the Russian astrophysicist Roald Sagdeev, who had been Mikhail Gorbachev’s science adviser; and the only woman, Maj Britt Theorin, a former Swedish ambassador for disarmament and a member of the European parliament.

  They debated across three continents between November 1995 and August 1996, when their report was presented to the Australian prime minister and subsequently to the United Nations and the Conference on Disarmament. “They were an extremely interesting22 and colorful group,” Butler told me:

  Maj Britt Theorin was a classic abolitionist, literally wearing beads and sandals; she would pull out her magic wand and ask why couldn’t we just declare nuclear weapons illegal and bring the world together? She was a handful. At the other end of the spectrum, Michel Rocard came on board by arranging to issue a statement of reservations—France might want to retain a few nuclear weapons,
you know. In the middle of the spectrum, Rolf Ekéus and Joe Rotblat, among others, were immensely helpful. Lord Carver, like most military men, took the view that nuclear weapons were useless and stupid and that we ought to get rid of them—anyone who has had the button in his hand will tell you that it’s a horror. Bob McNamara was terrific. There were times when he was rivetingly bright. He left his ego off the table; there was none of that “I was American secretary of defense” nonsense. All were agreed that the elimination of nuclear weapons was a sound purpose, but it wasn’t a group that found ready agreement.

  In that, I suggested, they seemed to be representative of their countries and the agendas they brought to the table—a practice run, as it were, of what a negotiation to eliminate nuclear weapons might look like had they carried their governments’ portfolios in that regard. Butler agreed. They came to consensus in the end.

  The Canberra Commission report opened with a statement of plain truths about nuclear weapons that decades of nuclearist sophistry have obscured:

  The destructiveness of nuclear weapons is immense. Any use would be catastrophic.

  Nuclear weapons pose an intolerable threat to all humanity and its habitat, yet tens of thousands remain in arsenals built up at an extraordinary time of deep antagonism. That time has passed, yet assertions of their utility continue.