Osama bin Laden has not yet succeeded in launching a nuclear attack. But it isn't because he can't. With enriched uranium, a handful of military supplies available on the Internet, and a small team of terrorists, he could assemble a nuclear bomb in a matter of months. This is how it will happen.
- By Jeffrey G. Lewis, Peter D. ZimmermanPeter D. Zimmerman is professor of science and security in the Department of War Studies at King's College London. He was previously chief scientist of the U.S. Senate Foreign Relations Committee and chief scientist of the U.S. Arms Control and Disarmament Agency. Jeffrey G. Lewis is executive director of the Managing the Atom Project at the Belfer Center for Science and International Affairs, John F. Kennedy School of Government, Harvard University. He publishes the leading arms control blog, www.ArmsControlWonk.com.
Nearly 3,000 people were killed five years ago when terrorists plowed airplanes into New York’s World Trade Center, the Pentagon, and a field in Pennsylvania. The terrorist attack was undoubtedly a terrible tragedy. But it could have been much worse. Eight years earlier, aides to Osama bin Laden met with Salah Abdel al-Mobruk, a Sudanese military officer and former government minister who offered to sell weapons-grade uranium to the terrorists for $1.5 million. He proffered up a 3-foot-long cylinder. The al Qaeda representatives agreed to the purchase, because after all, as one of them later said, "It’s easy to kill more people with uranium."
The cylinder turned out to be a dud. But had it actually contained highly enriched uranium, and if bin Laden’s deputies had managed to use it to assemble, then transport and detonate a nuclear bomb, history would have looked very different. September 11 would be remembered as the day when hundreds of thousands of people were killed.
Osama bin Laden’s long-standing interest in developing nuclear weapons is deeply troubling, and the attempt to purchase uranium from the Sudanese was far from an isolated incident. Al Qaeda operatives have repeatedly tried to acquire nuclear materials over the years. In August 2001, a month before the September 11 attacks, bin Laden received two former Pakistani nuclear officials, asking them to help recruit other Pakistani scientists with expertise in building nuclear weapons. After the military effort to oust the Taliban from Afghanistan, U.S. forces found extensive documents, including crude bomb designs, at an al Qaeda safe house in Kabul. In 2003, bin Laden sought a fatwa from an extremist Saudi cleric permitting the use of weapons of mass destruction, calling their acquisition a "religious duty." As recently as September, al Qaeda put out a call urging nuclear scientists to join its war against the West. Bin Laden’s attempt to purchase highly enriched uranium in the past belies the conventional wisdom that terrorists want a lot of people watching, not a lot of people dead. Clearly, some terrorists do want a lot of people dead.
Could a nuclear attack by bin Laden, or any other terrorist, actually happen? Some say it would be impossible, mistakenly believing that terrorists do not have the motivation, or the ability, to assemble the highly sophisticated, modern tools necessary for the task. Most observers, however, agree that a small group could construct a lethal nuclear weapon since they are conceptually simple devices. After all, the technology involved in creating a nuclear weapon is more than 60 years old. In fact, it is perhaps easier to make a gun-assembled nuclear bomb than it is to develop biological or chemical weapons.
A NUKE’S WORTH
Would terrorists build a nuclear device? Presumably, some terrorist organizations want to kill as many people as possible at the lowest cost. Like any organization, sophisticated terrorist outfits are concerned with "cost effectiveness." It is a gruesome business, but very similar attacks may result in widely different casualties depending on the target. For example, the bombing of the Marriott Hotel in Jakarta in 2003 killed a relatively small number of people compared to the 2002 Bali bombings, despite the use of relatively similar devices. But, if one considers the bulk of terrorist attacks, the relationship of cost to casualties follows a simple curve, with the cost per casualty increasing as the size of the terror attack increases — from the relatively inexpensive Madrid bombing (which cost less than $10,000, or around $50 per murder) to the September 11 attacks (which cost $400,000-$500,000, or about $170 per murder).
Some might claim that thinking about terrorist attacks in terms of cost-versus-casualty ratios fails to capture the essentially political ends of a terrorist group. Cost data from previous attacks suggest that al Qaeda is sometimes willing to pay a significant premium to attack high-profile, heavily protected targets that may produce fewer casualties, but have greater political implications, such as a U.S. embassy or Naval vessel. For example, the October 2000 bombing of the U.S.S. Cole in Yemen may have cost $10,000, but with 17 casualties, it added up to a pricey $590 per murder. Yet terrorists do not have to pay a premium for a nuclear attack; on a per murder basis, nuclear weapons are both cheap and can be used against high-profile targets. And a nuclear attack induces great fear. Its specter has hung over the world since the United States dropped Little Boy on Hiroshima.
To put it in strictly commercial terms, terrorists would likely find a nuclear attack cost effective. The simple appeal of nuclear terrorism can be illustrated with a hypothetical situation. A failed nuclear detonation, one that produced only a few tens of tons in yield, could kill 10,000 people in just a few hours if the device exploded in a crowded financial center. Not only would 10,000 persons represent the upward limit of a conventional terrorist attack, but that figure would also exceed the combined casualties in all of al Qaeda’s attacks over the entire history of the organization.
And that’s a "worst-case" scenario for the terrorists. A "successful" nuclear detonation would kill 10 times as many people. If terrorists could construct a successful device that killed 100,000 people for a cost of $10 million dollars — about $100 per murder — it would be a bargain, considering that most of al Qaeda’s attacks have been mounted in the $100 to $300 per murder range. A nuclear terrorist attack that cost $5 million would result in a cost per murder comparable to the Madrid bombings. So, just how difficult an enterprise would this be? What would a terrorist group have to do to build a bomb that would kill 100,000 people for less than $10 million?
Without providing a blueprint for how to make and detonate a nuclear bomb, let’s break down the costs of what such a project might entail.
In our scenario, terrorists would construct the nuclear device within the United States. Smuggling a fully constructed bomb across the border would cost more in time, personnel, and planning. And U.S. Customs is on the alert for bomb-like devices. Operatives would still have to enter the United States surreptitiously for the assembly, final check-out, and delivery of the bomb. Obtaining the required special machinery for bomb-making might also be more difficult abroad than in the United States. It is certainly possible that a terrorist group might not want to risk detection within U.S. borders and would prefer to make the bomb overseas. But, for purposes of this hypothetical situation, we chose a scenario that would be less uncertain for the terrorists by eliminating the risks of moving the bomb across a border.
What kind of nuclear device might a terrorist organization consider? Some bombs can produce an extremely big bang, but are difficult to build; other devices produce a smaller explosion but are comparatively easy to construct. In order not to publish anything that would make the terrorists’ work easier, we have chosen a crude but well-known design concept, widely available on the Internet, that is similar to the device that the United States used when it bombed Hiroshima.
Our device consists of a gun that fires a highly enriched uranium "bullet" into a cylindrical target block, also made of highly enriched uranium. Terrorists could use a surplus light artillery gun barrel, something that’s easily available today on the global arms market or the Internet for much less than $10,000. The target is hollow with a hole to receive the bullet, and is simply bolted onto the muzzle of the gun. The explosion in Hiroshima produced a yield of approximately 12.5 kilotons, which killed about 100,000 people instantaneously.
How many people would it take to construct a crude nuclear device? In a 1977 government report on safeguards against nuclear proliferation, the U.S. Office of Technology Assessment estimated that a small group, including a "person capable of searching and understanding the technical literature in several fields and a jack-of-all-trades technician," could build a nuclear device for a sum that "need not exceed a fraction of a million dollars." Adjusted for inflation, that’s less than $3 million today.
The constraint we have placed on our would-be bomb-makers is a total of 19 persons — the same number of hijackers who orchestrated the September 11 attacks — working over the course of a year in the United States. We estimate that a three-person physics team, including a relatively senior physicist and two postdoctoral students, would be capable of rendering the design in three to six months. Their salaries during the course of a year would total approximately $200,000. In addition to the physics team, the project could comprise a few small engineering teams to address the following: casting the uranium for the device, constructing the proper gun, assembling the supercritical mass of uranium, overseeing the electronics, and finally, the actual detonation.
In many respects, the most difficult task for nuclear terrorists would be casting the uranium metal, which melts at high temperatures, into appropriate shapes. The metallurgy team would include at least one person with experience in advanced casting techniques. A vacuum furnace is probably required to reduce oxygen contamination and prevent the uranium from igniting. The team would likely need to practice using either natural uranium or some surrogate before casting the final core. The group could find the vacuum furnace to fit their specifications by searching on the Internet, and could probably purchase it for less than $50,000.
The actual pit — or core of the weapon containing the highly enriched uranium — could be fabricated quickly. When China built its first nuclear bomb in 1964, a single technician named Yuan Gongfu used a lathe to shape the highly enriched uranium in just one night. New or used lathes large enough to properly finish the roughly cast pit can be bought on the Internet, even on eBay, for $10,000. These instruments are probably as capable as the one Yuan used more than 40 years ago. Computer-controlled machine tools are not necessary. Our terrorist outfit could probably find all the standard machine shop equipment it would need in any university physics department. None of the tools requires special licenses to purchase.
The machining group would also have the task of designing and building the required structure for the device and assembling the whole. This requires at least two or three people able to carry out such common laboratory tasks as welding, brazing, and hard soldering. One member of the group should bring the skills of a draftsman, and preferably good abilities to use computers to design complex shapes.
To detonate a nuclear bomb, terrorists do not need to fashion the right type of gun. "Team Gun" would likely consist of three or four people, at least one of whom is familiar with the interior ballistics of guns in the appropriate size range. Their principal task would be to find a surplus artillery piece of the correct size and to build a projectile. Such recoilless rifles are widely available in the United States and Canada as military surplus, though they require a license to purchase. A hobbyist could easily refurbish a recoilless rifle for just a few thousand dollars. In all likelihood, the gun team would want to test-fire the gun with a dummy nuclear projectile to verify its actual speed. One or two shots should suffice. Barrel life, usually hundreds of shots without maintenance, would not pose a problem. It is unlikely to take more than six months for such a group to adapt and test a reliable gun.
Meanwhile, the electronics team would likely include one or two technicians and probably be headed by an engineer with at least a baccalaureate degree in electrical engineering or experimental physics. Its principal job would be to design the circuitry that arms and fires the device at the desired point, and which prevents it from detonating accidentally. In addition, the group would need to assemble or purchase, install, and calibrate neutron detectors to be used in testing the device.
The terror group will also require a large, remote area for their "mini-Los Alamos." Their biggest concern would be the noise caused by firing the gun during their tests. We would choose a 150-acre ranch in an isolated area — relatively small compared to the 500,000 acre ranch that Japanese doomsday cult Aum Shinrikyo purchased in Australia. We estimate that such a ranch might be purchased for $150,000 in remote areas of the United States, such as Texas or Wyoming, and require another $50,000 in temporary improvements to build the foundry, machine shop, electronics lab, and other equipment.
Once complete, the nuclear device itself is likely to be less than 9 feet long. Although it would not fit easily in a sedan, it could be transported in a van or small panel truck with, say, a couple drivers and a couple more people to keep an eye on the device. The plotters could target any number of major metropolitan areas and would be free to choose based entirely on their desire to travel unobtrusively and undetected, presumably across a large fraction of the United States. Nevertheless, the transportation phase of the operation would pose significant risks for the terrorists. For the first time, the device would be moved, most likely on public roads, with little security.
The plotters would want to travel on busy roads so as to lose themselves in the traffic stream. In the event that the government became aware of the plot, it would seek to install targeted roadblocks. These checkpoints would force the terrorists to back roads, where they may be likely to attract the attention of local police. The trip would consume no more than 40 driving hours, and could easily be completed in four to five days traveling between dawn and dusk. The cost to deliver the device using a typical rental truck from our hypothetical Wyoming ranch to, say, New York City or Washington, D.C., would be less than $3,000.
At this point, the project would have employed roughly 17 people full time for about one year. Purchasing the necessary items — from land to supplies, and surplus gun barrels to vacuum casting equipment — is a specialized business, particularly because the purchases could be traced by law enforcement. The plotters would want to avoid being personally involved in purchasing supplies, so there could be a need for one or two specialists who would be responsible for clandestine procurement. Even so, the entire active team would number no more than 19 people.
URANIUM, THE HOLY GRAIL
Altogether, a terrorist organization like al Qaeda could plausibly build and deliver a nuclear weapon for less than $2 million. That leaves substantial room in our budget of $10 million for the cost of buying highly enriched uranium.
Estimating the cost of acquiring fissile material is quite difficult. A terrorist group would be most likely to either purchase uranium on the black market, or attempt to steal the material from a poorly guarded facility. The "market" for fissile material is an unusual one. For the most part, the market is a mix of criminals and con artists on one side and police and informants on the other.
We know of only one particularly disturbing instance in which smugglers obtained a significant quantity of highly enriched uranium: a 1994 case in Prague where police arrested a smuggling ring involving Czech, Slovak, and Russian nationals. The ring had obtained 10 kilograms of uranium and told undercover officers they could come up with another 30 kilograms quickly. The Prague case reveals the trouble in estimating an accurate market price. Although the Russian suppliers were asking $800 per gram, the Czech and Slovak middlemen doubled the price to $1,600 a gram. At those prices, terrorists would need to spend several tens of millions of dollars to acquire enough fissile material for a bomb.
But if a terrorist group subcontracted the smuggling operation to others who would haggle and bargain more than the Prague police did, it could probably demand significantly lower prices. Last year, the International Atomic Energy Agency reported that it had documented 18 seizures of stolen plutonium or highly enriched uranium. Clearly, groups are attempting to acquire the material necessary to build a bomb. And for every attempted theft that we are aware of, how many go undetected?
Al Qaeda’s attempt to purchase highly enriched uranium from the Sudanese for $1.5 million is perhaps a more realistic estimate of the cost of acquiring fissile material, even though that particular cylinder turned out to be a fake. Without knowing the mass of the cylinder in Khartoum or whether bin Laden believed additional purchases would be necessary, a terrorist group might need to make two or three such purchases at a cost of $3-5 million, with $4 million dollars as an average.
With all other expenses running a little less than $2 million, two things become clear. First, the biggest cost — and, by extension, the most difficult part of the operation — would be buying the highly enriched uranium. Likewise, the cost of designing a relatively more reliable weapon is trivial. A job worth doing, it would seem, is worth doing well.
Second, given our initial budget of $10 million, a well-financed terrorist organization need not worry unduly about scams or buying bogus nuclear explosive material. Falling victim to a scam — such as the failed al Qaeda effort to purchase uranium in Sudan—would not threaten the cost effectiveness of developing a nuclear weapon. The risk of buying poor quality or fake uranium may simply be the cost of doing business.
NOT NOW, BUT WHEN?
Our scenario does not suggest that terrorists would find building a nuclear weapon either easy or inexpensive. The most important obstacle remains the difficulty in acquiring enough nuclear explosive material to build a bomb. Moreover, only a very small number of terrorist groups fit the profile we have outlined: interested in mass casualties, well financed, and organizationally sophisticated. We have identified only two groups in recent history with all three qualities: al Qaeda and Aum Shinrikyo. So far, al Qaeda has not developed the necessary technical expertise to obtain the right material. Bin Laden, perhaps, has yet to find his Robert Oppenheimer.
To prevent him from doing so, governments can continue to use intelligence to break up terrorist plots. They can find ways to restrict funding to terrorist groups. (After all, spending $5-10 million to kill 100,000 people is a bargain only if you have $5-10 million to spend in the first place.) They can continue to prosecute their war on terror. But such solutions are anodyne. No one really knows how much highly enriched uranium there is in the world, or how close the wrong groups are to getting the right amount. The frightening truth is that fissile material, including nuclear explosive material, is an item of commerce, and moves from place to place. One of the side effects of our globalized economy is that opportunities for direct theft and bribing of nuclear custodians abound.
And when rogue states like Iran and possibly North Korea continue to enrich uranium, ostensibly for energy purposes, it is even harder to control what happens to it. Although building a nuclear device remains an expensive, complex undertaking out of reach for most organizations, a well-financed group that seeks to kill very large numbers of people may well find it an irresistible option. A wealthy organization seeking to kill several hundred thousand people could hardly find a more economical method than the detonation of a small nuclear device. That is reason enough to consider the nuclear threat a serious one. Just because a nuclear terrorist attack hasn’t happened shouldn’t give us the false comfort of thinking it won’t.