Stopping Nuclear Terrorism Is a Game of Odds, Not Certainty
Editor’s note: An Associated Press investigation into weapons traffickers attempting to sell nuclear material to terrorists prompted us to republish a WIRED feature from 2003 about the difficulty of preventing nuclear terrorism.
I’m standing near a row of deserted loading docks in Billerica, Massachusetts, and George Kinsella hands me a vial of cesium 137. “This,” he says, “is the kind of radioactive material you might see in a dirty bomb.”
As radioactive substances go, cesium 137 leads a fairly innocuous existence as a component of industrial instruments such as moisture gauges. Mishandled, though, it can cause severe burns or genetic defects, as it did at Chernobyl. I hand the vial back, fighting the urge to wash my hands, and Kinsella places it inside the trunk of a Mercedes sedan.
Then he shows me a black canister the size of a soup can: Wrapped in a shielding layer of tungsten, it contains cobalt 57. He climbs into a cargo container on the back of a flatbed truck and puts the canister down near the center.
The whole exchange looks like the kind of transaction that keeps Tom Ridge awake at night. As it happens, the loading docks belong to American Science and Engineering, the company where Kinsella works as principal software engineer, and he’s preparing to demonstrate its MobileSearch X-ray and radiation sensor technology. For the past decade, the 44-year-old firm has developed X-ray scanners that help customs officials detect contraband in the war on drugs; now it’s one of a handful of companies racing to manufacture devices that detect nuclear and radiological weapons.
Kinsella and AS&E chief technology officer Joseph Callerame usher me into an RV-like vehicle parked alongside the Mercedes and the truck. We climb into a small, air-conditioned space where two swivel chairs sit in front of a console outfitted with four monitors and a bank of flashing lights. It looks like the control room of a small TV station. Kinsella takes a seat, and Callerame and I stand behind him.
“Should we start the scan?” Kinsella asks. He makes a few quick keystrokes, and a low rumbling sound begins. A robotic arm mounted on top of our vehicle hangs near the far side of the Mercedes, then begins to slowly creep alongside it. “We’re almost starting to cross the cargo now,” Kinsella says. “Here we go.” A ghostly outline of the trunk scrolls into view on two of the screens. One image is an ordinary X ray, a blurred jumble of superimposed shapes with no apparent depth. The second image, using more advanced backscatter X-ray technology, has an unearthly quality, as though the side of the car was ripped off and a grainy black-and-white image was taken of the contents.
I’m so transfixed by the shot of the Mercedes, I don’t notice the rainbow stripe that appears beneath the outline of the trunk, shifting from green to bright red as the scan continues. “It’s already detected the radiation,” Callerame says. “Green indicates something’s there, red is more serious. Even this very small source reveals a potential danger. And you can see that the source is localized to the back of the vehicle.” As Callerame talks, the scanner moves past the Mercedes on to the flatbed. Suddenly there’s a steady beeping in the room. “When the threat is higher, that alarm goes off.”
Once we’re outside again, I glance back at the MobileSearch truck, glaring white in the early afternoon sun. From 20 feet away, it looks like the kind of trailer you might see at a construction site. But it’s also an early glimpse of a technology that could be seriously effective at reducing a certain kind of threat. Which would be a good thing, because as dangers go, this one is about as gruesome as it gets.
Worst Case Scenario
Imagine a terrorist driving east toward Washington, DC, a few minutes outside the Beltway on River Road. In the back of his van, he has 100 curies worth of cesium — about a thousand times what Kinsella handed me — along with a traditional explosive nearly as powerful as what Timothy McVeigh detonated in Oklahoma City. He’s 14 miles from the White House. If he makes it there, he could perpetrate an assault on the US more disruptive than the terror of September 11. A strategically placed explosion might kill hundreds and require thousands to be treated for radiation exposure. The cleanup would take months. If he were carrying a traditional nuke—the smallest of which is about the size of a large refrigerator—he could well execute the single most devastating strike in human history.
An atomic blast near the Capitol would vaporize everything in downtown DC. A zone framed by Georgetown and southeast Washington would suffer casualty rates of 98 percent. Three or four miles away—out to upper-northwest Washington, out to Reagan Airport across the Potomac—the casualty rate would run 50 percent, with most buildings damaged beyond recognition. The fallout would leave a trail of radiation sickness and, eventually, birth defects.
A dirty bomb set off on the Beltway, or on the outskirts of any city, would be a disaster. But depending on the traffic and the road, it’s likely that casualties from the explosion would be minimal and no one would die from radiation poisoning. The cost in life, disruption, and dollars might be the equivalent of a minor earthquake or a bad flood. A catastrophe, to be sure, yet nothing compared with the chaos that would ensue if such a device exploded in a crowded urban center.
In terms of an actual multi-kiloton nuke, the difference between a detonation on the Beltway and a detonation in front of the White House would be several orders of magnitude. In the Beltway blast, thousands would die and millions might suffer from the fallout effects. Park the bomb by the White House, however, and a million people could be killed in seconds. And much of the US government would be taken out in the process.
Right now, the government is focusing its resources on preventing that terrorist from getting his hands on a weapon of this scale. But how realistic is that? There are 30,000 nuclear weapons in the world, plus countless supplies of radioactive material that could be made into a dirty bomb, which spreads its poison with a conventional explosion. We live in the age of the “super-empowered angry young man,” in Thomas Friedman’s phrase, and any disgruntled group on the planet is a potential radiological event waiting to happen. No matter how hard you try to assuage or attack such groups, the odds are against you. So it’s prudent not only to keep potential bomb detonators away from bombs but also to keep bombs away from large groups of potential victims.
We need to make sure that van stays 14 miles away from the Capitol. And the way to do this isn’t with Star Wars technology—death rays beaming down from space. We need an older technology, as old as cities themselves. We’re talking about a wall.
Atomic Walls Against the Wicked
Walls have protected cities as long as there have been cities to protect. To guard against invasion, Nebuchadnezzar built a network of brick walls so that “the evil and the wicked might not oppress Babylon.” As weapons evolved, walls had to change with them. By the 16th century, artillery had advanced to the point that Vienna razed the scattered developments outside city walls so potential invaders would have nothing to hide behind. The battlements had become a broad detection zone. The width of the space created was 400 meters—the range of cannons at the time. The outline of that zone is still visible today in the avenues and parks that make up Vienna’s fabled Ringstrasse.
Airplanes and missiles rendered the city wall symbolic; these days, the important defensive barriers aren’t physical fortifications. But the advent of small nuclear weapons and dirty bombs—deliverable not by missiles and planes but by trucks and vans—suggest a new kind of urban perimeter defense, an atomic wall. Set up not as an actual barrier but as a vast array of sensors, such a technology would exploit the fact that any radiological or nuclear weapon leaves a footprint. For example, a ring of radiation detection devices deployed along the Beltway could scan every road, alley, and rail line that brings people within 14 miles of the White House. If nuclear material crossed the line, sensors would alert emergency response teams, which would intercept the vehicle before it entered the city. As in Vienna, the wall would be less a barricade that couldn’t be crossed than a zone under constant surveillance.
The basic technology behind such a system already exists, and while senior law enforcement and intelligence officials wouldn’t discuss the subject on the record, conversations with both government and private-sector experts indicate a high level of interest in such a system. This past August, the Department of Defense announced that bioweapons sensors would be deployed in certain cities. Three months before that, the House began discussing the Anti-Nuclear Terrorism Prevention Act of 2002, which authorizes $250 million for installing scanning devices in New York City ports and tollbooths. Really, that’s just a first step. “The idea of a truck containing a nuclear device in the center of our city is terrifying, but not impossible,”says Senator Charles Schumer, the New York Democrat who proposed the bill. “It doesn’t matter how good our airport security is if all it takes to bring a nuclear device right into midtown is putting it on a ship or bringing it in on a truck.”
When it comes to Manhattan, Schumer has it relatively easy: He has only four tunnels and 11 bridges to worry about. What about the cities that don’t happen to be located on islands? Any terrorist group well organized enough to plot a hijacking could find dozens of back roads into Los Angeles. How much would it cost to scan all the entrances to a city of that size?
“To track vehicles traveling along a highway, what you’d need is a sensor array, and probably a secondary array—along with some video technology to keep track of who you were scanning,” says Jim Winso, vice president of the San Diego-based SAIC (Science Applications International Corporation). The company is developing a radiation detection array—SAIC calls it a nuclear portal—that could cost less than $100,000 when purchased in bulk. The device would scan for suspicious cargo without slowing down traffic; once law enforcement identified an undocumented radiation source, they’d shut off the flow of vehicles at that particular access point, which would require roadblocks that could be set up in a matter of seconds. Once the vehicles in question had been contained, they would be examined with a more elaborate device—something like AS&E’s $2 million MobileSearch system, which combines gamma ray detectors with advanced X-ray technologies.
So when the terrorist heading toward the center of DC drives up the Beltway entrance ramp, his van would travel under a passive scanning device mounted on an overpass. If you remember high school physics, you know that radioactive materials emit invisible particles, including alpha and beta particles, which can be easily shielded against traditional scanners; most also emit gamma rays, which penetrate most materials and are far more difficult to conceal. In the split second he was under the scanner, 500 gamma rays might collide with it. Someone monitoring incoming data would notice the spike in radiation, and a video camera—the kind already used to catch traffic violators—would record his vehicle and license plate.
Within seconds, an emergency management team puts up a roadblock, slowing traffic to a standstill. A rapid-response team locates the vehicle and either searches it by hand or brings in a mobile X-ray unit to survey the contents. Of course, if the signature had suggested a less lethal material, the authorities could have opted to discreetly pull over the vehicle, as though they were nabbing someone for speeding.
An array for each of the 50 Beltway on-ramps, plus 400 more to cover the roads running under the Beltway, would cost $50 million. Twenty-five MobileSearch trucks that could be moved to a scanning zone in a matter of minutes would cost another $50 million. Throw in a final $50 million to build temporary barricades and reconfigure roadways for the system. Then assume that everything ends up costing four times as much: The final tab for a metropolitan atomic wall would be $600 million.
That’s a lot. But keep in mind that the Bush administration asked Congress for $8 billion for research into Star Wars-style technology, out of a proposed total military budget of $379 billion. In other words, for the portion of the 2003 budget allocated for missile shields, you could build atomic walls around the dozen biggest cities in the US.
A Network of Radiation Microsensors
“Every year there are about 300 cases of radiological materials that are either lost, stolen, or abandoned,” says Ralph James.
In his office at Long Island’s Brookhaven National Laboratory, where he serves as associate laboratory director for energy, environment, and national security, James is telling me how an atomic wall might work in practice. There’s a measured, Mr. Rogers quality to his speech that, along with his defense industry euphemisms, creates a false sense of normality. He mentions “consequence management” a few times before I realize he’s describing what needs to be done after a nuclear bomb goes off.
To James, the key technology behind an atomic wall is the ability to differentiate between types of radioactive materials. If a source is sending out enough gamma rays, an ordinary handheld Geiger counter will pick that up. But you won’t be able to identify the substance without a more sensitive detector. “You can think of gamma ray energies as frequencies,” James tells me, as we sit in his office decorated with plaques and honorary degrees accumulated over 30 years. “Just as you can tune your radio to go from one frequency to another, you can tune the sensors to different gamma ray energies.”
The gamma rays emitted by the terrorist driving toward Washington, for example, would register energies of 662 kiloelectronvolts. That energy profile would create a clear picture: cesium 137, and probably a lot of it. Definitely not the kind of material you’d normally see packed into the back of a van.
In general, detecting radiation is easy. The hard part is separating it from all the other radiation out there. “We have very sensitive detectors available—we can fly airplanes over the ground and pick up small increases in natural radiation from uranium ore deposits,” says William Miller, professor of nuclear engineering at the University of Missouri. “But there’s considerable variability in natural radiation levels. The presence of radon gas in homes and basements was discovered because a nuclear power plant employee kept setting off alarms at work.” Potassium decays in every human body, and anyone being treated with nuclear medicine would trigger a crude Geiger counter. Without more advanced sensors, an atomic wall would be a nightmare of nonstop false positives. You’d know there was something spitting out gamma rays—just not what it was.
“It’s like good and bad cholesterol,” says SAIC’s Winso. “You have ‘good’ radiation, and and you have undocumented radiation.” James and Winso believe that, in addition to screening out background noise, the system would need to keep track of known radiation sources—hospitals, for example—in an evolving database.
That same database would monitor information transmitted by mobile sensors. “We’d need a type of sensor network within the city—to disperse radiation detectors into a continuous monitoring set of stations,” James explains. “We can make these things very low-cost. They’re not going to be very smart sensors, but they could be no bigger than a wristwatch, distributed to police officers, firefighters, postal workers—enough people dispersed around the city so that it would be difficult to move radiological material around it. All these low-cost sensors could be connected to a network—if you saw a lot of them going off, that’s when you’d need to respond.”
On its own, each of these sensors would not be very sophisticated. “If we’re looking at a radioactive source that’s in the building across the street,” James says, gesturing at a one-story lab about 40 feet away, “and I have a detector that’s the size of a pinhead, it’s not going to have very many gamma rays impinging on it in the period of, say, 100 seconds.”
James believes you can get around those limitations by thinking of the sensors as small pieces, loosely joined—an atomic wall that functions as a web. “As you get farther away from a source, the sensors have to get bigger, because that radiation is spreading everywhere. So you want to be close. Now you reach a point where you don’t win by making these sensors bigger and more expensive by tiling them together. You win by having smaller, less-expensive detectors that are connected to a network.” Any given wristwatch sensor might fail to detect a radioactive source, but distribute enough of them in a detection zone, and you’ll start seeing a pattern of gamma rays trailing across the city. James calls it a “radioactive plume.”
Think back to the terrorist driving toward Washington. In the five minutes before he approaches the Beltway on-ramp, he drives by two police cars, one postal vehicle, and one fire truck, all of which are outfitted with Ralph James’ wireless microsensors. Gamma rays set off two of the sensors, and the signals produce a discernible plume headed along River Road. Even this small amount would likely register as cesium 137, though not whether it was enough for a bomb. But emergency-response teams would know to watch out for something emitting gamma radiation heading east on River Road, even before the terrorist hit the Beltway.
The Likelihood of Subverting the Network
An atomic wall naturally brings to mind the Star Wars missile shield proposed during the Reagan administration. But the widely held objection to the Strategic Defense Initiative—that it would threaten the deadlock of mutually assured destruction and thereby make one side more likely to pull the trigger—of course does not apply. As September 11 made all too clear, mutually assured destruction wouldn’t be much of a deterrent to our man with the cesium 137. Game theory has always had trouble accounting for players with no rational self-interest, and nuclear deterrence is no exception.
In fact, the irrelevance of MAD creates an opportunity. There’s a potential cost and inconvenience to building perimeter defenses, but there’s no longer any new danger posed by creating them. It really comes down to how much risk you think there is. There’s quite a bit of inconvenience at airports already; if urban residents truly felt that radiation warfare was a legitimate threat in their hometowns, it’s not hard to imagine them putting up with longer lines at the city limits. And cost? Spread over its 20-year lifetime, even the most advanced system would be a rounding error in the military budget.
Then there’s the more pressing issue: How easy would it be to subvert the network? After the scanning demo in Massachusetts, I sit down in a conference room with Callerame, and he walks me through the physics of concealment. High atomic-weight materials like lead can block gamma radiation, but the large quantities of lead that would be needed would show up on other scanning devices. Callerame’s solution is to combine radiation sensors with advanced X-ray technologies, like the backscatter system that produced the startling image of the Mercedes. “I still think you’re going to have to X-ray these things,” Callerame says. “If you run only a radiation detector and somebody shields their source well enough, you may not pick it up. On the other hand, if you’re simultaneously doing X-ray imaging, you’ll see this big blob in the middle of the cargo, which would be a dead giveaway of something being clandestinely brought in.” He shows me printouts of scans done at a demo in Washington, where they concealed the radioactive material in a container of lead the size of a bowling ball. In the image, the lead container pops out immediately, a bright-white circular shape in the middle of translucent grays. “Now, I should mention, even though we wrapped the cesium in this lead casing, we still managed to pick up the gamma radiation. It’s just easier when you do the two in combination.”
Experts agree that a mixed-sensor approach is the way to go. “Any remedy that a terrorist could employ [to conceal radioactive material] would automatically make him more vulnerable to detection,” says Winso. “You can use large quantities of lead to shield gamma sources, but that creates an awful lot of weight.” Also, the mixed-sensor approach would help alleviate civil liberty and health concerns. Passive detection systems like SAIC’s nuclear portal or Ralph James’ wristwatch sensors are noninvasive—in both medical and privacy terms. You’re not probing someone’s car, the way you would with an X-ray scan, you’re just listening for gamma rays coming out of the car. Other than radioactive materials, you wouldn’t be able to tell anything about the contents of a vehicle from a passive scan.
While the experts I spoke with seemed convinced that a dirty bomb attack was more likely than a nuclear weapon, there was also a consensus that nukes would be harder—but not impossible—to detect. Unlike a dirty bomb, which disperses radiation that’s already there, a nuke creates the majority of its radiation in the fission process. “This kind of system would have great utility for detecting radiological material that could be used in dirty bombs,” explains Philip Anderson, director of the Homeland Security Initiative at the Center for Strategic and International Studies. But if a nuke reaches US soil, it’s already too late. “With nuclear weapons, we have to go to the point of origin.”
Because nuclear weapons are larger, we could offset this by routing bigger vehicles through fewer portals, which would be outfitted with both X-ray and gamma ray scanners. Officials could distribute the cheaper microsensors along the city’s periphery, detecting suspicious radiation stored in smaller vehicles, while channeling all the trucks large enough to carry a real nuke past the more expensive probes. This would add some friction to the flow of commerce into a metropolis, but those roads are already teeming with tollbooths and weigh stations. Sure, building an atomic wall might involve rerouting some of the traffic coming into cities, but we’ve pulled off larger feats of urban engineering. And in those cases, the cost of doing nothing was far less.
No strategy to reduce the risk of nuclear terror is bulletproof. Even if scanning technology improves, it’s always possible that a van containing deadly material might slip through. Prevention is a game of odds, not certainty. But if you think the technology can reduce the chance of countless urban casualties, then at a certain price, it’s worth doing.
Before I leave the Brookhaven campus, James pulls me over to his bookshelves, near the patent grants displayed on the wall. He grabs a small jewel box off a shelf and opens it to reveal a gray block the size of a sugar cube. “This is the microsensor I was telling you about,” he says proudly, as though we’re looking at photos of his children. “It’s made of cadmium zinc telluride, which can detect gamma rays at room temperature. This is missing the electrodes and supporting circuitry you’d need for wireless transmission, but you can see how small it is.”
James hands me the sensor, and I imagine millions of them scanning trillions of invisible particles like an oversize urban immune system. I ask him if he thinks such a system will ever be built. He pauses for a second. “Let me just say that sensors have already been deployed in New York and Washington. I don’t want to make a comment beyond that regarding the specific locations to give information to terrorists.” His smile seems to say, We’re working on it.
To be sure, building an atomic wall wouldn’t be reason to stop doing all the other things we already do to keep the world free from radiological terror: weapons inspections, military strikes, peace marches. But like the city walls of old, an atomic wall of networked sensors might have the single most profound impact on our perception of safety, particularly for those of us living within the obvious targets. It would give the current system of policing radioactive materials a new kind of redundancy, one reassuringly close to home. You’d pass by those sensors at the exit ramp or the tollbooth, and you’d know that something, somewhere, was counting the gamma rays.