Nuclear Plant Safeguards Could Thwart Weapons Proliferation, Experts Say at AAAS Briefing


R. Scott Kemp

With more than 20 nations, including many in the war-torn region of the Middle East, announcing their intention to build large power reactors, the question arises: Will the spread of nuclear power also mean the spread of nuclear weapons? Two specialists offered a sobering assessment at a Capitol Hill briefing co-sponsored by AAAS.

Without strict and effective safeguards on reactor fuel and the plants that produce it, they said, there are no good ways to detect clandestine efforts to divert nuclear materials for weapons use.

R. Scott Kemp, a doctoral student at Princeton University, said the only assured way to keep an eye on spent reactor fuel—which can be reprocessed to separate out bomb-usable plutonium—is to have on-the-ground safeguard procedures and inspections.

Spy satellites and remote sensing instruments can pick up clues to the location of covert reprocessing facilities, Kemp said, but there are ways to avoid detection, including reducing the tell-tale emissions of the isotope krypton-85 from the facilities. Moreover, emissions from large, commercial reprocessing plants make it even more difficult to spot the small additional releases from a covert reprocessing operation.

"We really do have to hope that safeguards at reactors can detect the diversion of spent fuel," Kemp said. "Otherwise, the ability to detect reprocessing is limited." The safeguards used by the International Atomic Energy Agency (IAEA) include video surveillance cameras, tamper-indicating seals on key equipment, and regular on-site inspections.


Robert W. Selden

Robert W. Selden, a former leader of the applied theoretical physics division at Los Alamos National Laboratory, also urged the use of international safeguards to avoid weapons proliferation. He said that all plutonium isotopes, including those in spent fuel, can be used directly in nuclear explosives. He dismissed as fallacious the concept that plutonium from spent fuel could somehow be "denatured" so it cannot be used in a weapon.

Kemp and Selden spoke at a 12 May Hill luncheon briefing on nuclear proliferation co-sponsored by the AAAS Center for Science, Technology and Security Policy and the Nonproliferation Policy Education Center.

Kemp, who serves as an occasional consultant to the United States government on nuclear proliferation, is completing work on his doctorate at the Woodrow Wilson School of Public and International Affairs, at Princeton. In addition to spent fuel reprocessing facilities, Kemp said, the facilities where nuclear reactor fuel is fabricated in the first place also pose monitoring challenges.

Those facilities include arrays of centrifuges used to separate uranium-235, an isotope capable of sustaining a chain reaction, from its more ubiquitous cousin in nature, U-238. The centrifuge halls can be buried underground, as Iran has done, making it essentially impossible to detect or monitor them with cameras on spy satellites. Kemp said the energy consumption of the Iranian centrifuge facility is small, about on par with a food-processing plant or hospital of the same size. So an infrared scan of the area's energy footprint would not show anything particular unusual. There also are virtually no atmospheric emissions from centrifuge plants, Kemp said.

Centrifuge plants do contain a lot of electrical systems, Kemp said, and those can generate identifiable artifacts (transient electromagnetic waves) on power lines. Those artifacts can propagate over long distances, he said. In theory, if you had access to the power lines in a nation, you could try to look for such signals. However, the signals can be filtered out, he said, and power companies in many industrialized countries routinely ask their customers to avoid such electromagnetic "noise" so that it does not affect other customers. A rogue centrifuge operation might also simply generate its own power and avoid hooking into the power grid altogether.

A conversion facility that produces gaseous uranium hexafluoride—the feedstock for a centrifuge—could be "leakier" than the centrifuge plant itself, Kemp said. But emissions from such plants, in the form of uranyl fluoride particles, probably cannot be detected by sensors at significant distances, he said. Kemp published an analysis last year, which he updated using aerosol emissions from Canada's Port Hope conversion plant. He said modern, high-efficiency air filters could make the emissions even more difficult to detect.

The United States needs to move away from too much reliance on "national technical means"—satellites, sensors, listening devices—in its search for covert nuclear programs, Kemp said. "The only way to real assurance is to think about verification, at any level, by managed access," he said. "In other words, the right for an inspector to go anywhere, on demand, with little delay." That is really the only kind of inspection that worked in Iraq, he said, during United Nations monitoring and dismantling of Saddam Hussein's nuclear facilities.

Such an on-site strategy may require more transparency on both sides, Kemp said, with the United States more willing to be open about its own nuclear programs in exchange for better access to facilities elsewhere. And it may require more diplomacy from the outset, he said, rather than wasting time trying to detect or sabotage a nation's ability to produce nuclear materials.

Safeguards are not infallible, of course, and there have been calls for the IAEA to thoroughly assess the reliability of its safeguards regime against a determined proliferator. "I'm not aware that the IAEA has done that," Selden said. "It sounds like a good thing to do." The amount of nuclear bomb-making material has grown by a factor of 6 to 10 over the past 20 years, while the IAEA's safeguards budget has not kept pace and the number of IAEA inspections per facility has declined, according to the report last year of the congressionally mandated Commission on the Prevention of Weapons of Mass Destruction Proliferation and Terrorism.

Selden has been addressing the proliferation risks of plutonium from spent reactor fuel since 1976, when he and Carson Mark, another nuclear weapons specialist, gave a series of non-classified briefings showing that reactor-grade plutonium is highly useful for constructing nuclear explosives. They noted that the critical masses—the amount needed to make a nuclear explosion—are very similar for weapons-grade and reactor-grade plutonium. The so-called "bare spherical critical mass" for high-quality, weapons-grade plutonium is 11 kilograms (about 24 pounds), Selden said, and for reactor-grade plutonium it is 13 kilograms (about 28 pounds).

The bottom line, according to Selden, is that all plutonium isotopes are capable of carrying out an explosive chain-reaction in the presence of fast neutrons. He said there has been a continuing misunderstanding by some in the reactor community about the plutonium-240 isotope, which accumulates in reactor fuel over time. Plutonium-240 does not fission in a reactor. Hence, it reduces the reactivity of the plutonium in the reactor, However, plutonium-240 will fission in a nuclear explosive, Selden said.

Reactor-grade plutonium rich in Pu-240 complicates matters for a bomb-maker. The background neutrons in the plutonium, so the argument goes, could trigger a chain reaction too soon (a condition called pre-initiation), allowing the fissile material to expand before it can go supercritical enough to produce a massive nuclear explosion. The result has been called the "fizzle yield."

But Selden said that even a "low-tech" bomb using reactor-grade plutonium with Pu-240 can have a yield on the order of several kilotons. (A kiloton is equivalent to a thousand tons of TNT.) The assembly velocities—the speed at which the critical mass of plutonium is compressed to initiate a runaway chain-reaction—can be high enough to produce a substantial explosion. "A few kilotons is clearly a significant weapon," Selden said.

And, in fact, the explosion of the Trinity Device at Alamogordo, N. M., on 16 July 1945—the world's first atomic explosion—was just the sort of large, relatively unsophisticated device that would have worked with reactor-grade plutonium as well as the weapons-grade plutonium that was used, Selden said. The yield of such a reactor-grade device, though smaller than Trinity's 20-kiloton yield, would still have been more than 1 kiloton, he said. "This is a real device," Selden said. "The bomb would have worked."

There also has been debate over the years on whether reactor-grade plutonium could be readily "denatured" so that it cannot be used even in a low-yield nuclear bomb. Israeli researchers recently argued that plutonium created in large nuclear reactors could be "de-clawed" by adding americium-241, a form of the basic synthetic element used in smoke detectors and industrial gauges. But Selden is not persuaded, nor are many in the arms control community who have found the Israeli study flawed.

Selden notes that americium-241 is an isotope that is made from the decay of the plutonium-241 isotope, and is present in any plutonium after some time passes. It is a fissile isotope, with about the same critical mass properties (for fast neutrons) as uranium-235. "If you don't want it in the plutonium, it can be removed by a chemical separation process that is well known," Selden says. "So I would say that it should not be called a 'denaturant.'"

There also have been suggestions that reactor-grade plutonium could be diluted with a non-fissile material. For example, plutonium oxide can appear in reactor fuel over time. The oxygen serves to dilute the potency of the plutonium. But the material can still be used in a nuclear explosion, Selden said.

There also is talk of "poisoning" the plutonium with a material such as boron, an absorber of neutrons that is used in control rods to limit or shut down the chain reaction in a nuclear reactor. But Selden said boron is not an effective poison in the presence of fast neutrons of the sort that occur in a runaway chain reaction. A plutonium-boron combination (imagine boron replacing the oxygen in plutonium oxide) could have a critical mass, for weapons purposes, of several tens of kilograms—say 70 kilograms or about 150 pounds, depending on the density of the combination. That's a lot of material, but if a rogue nation or terrorist group were determined to enter the nuclear club under any circumstances, it might not be an obstacle.

If you have any type of plutonium in sufficient quantities, Selden said, "you can make a bomb."