PHOTO: Lawrence Livermore National Laboratory

9 April 2008—Part of the International Atomic Energy Agency’s job of ensuring nuclear safety is to make certain that civilian reactors are not diverting any nuclear material to make weapons. A few kilograms of plutonium is enough to make a nuclear weapon, and in the United States alone civilian reactors generate hundreds of bombs’ worth of plutonium every year. The IAEA could get much-needed monitoring help from a new type of detector that researchers at the Lawrence Livermore National Laboratory (LLNL), in California, and Sandia National Laboratories, in New Mexico, recently tested. By detecting particles known as antineutrinos that fly out of the reactor, the device measures the reactor’s power and how much uranium and plutonium are present in the core.

The IAEA’s inspectors could verify if this information matches up with what is expected for a reactor of its size. During their operation, nuclear reactors consume uranium and create plutonium, so knowing the power level, for instance, would tell the inspector how quickly plutonium is being built up in the reactor core. “If you operate at twice the power, you would build up plutonium at a significantly faster rate than at standard power,” says Adam Bernstein, an LLNL physicist who is leading the work.

The researchers have been testing a prototype of their detector at the San Onofre Nuclear Generating Station, in San Clemente, Calif., since 2003. The 2- by 3-meter, 1-metric-ton detector sits in an underground room about 25 meters from the reactor. In an upcoming Journal of Applied Physics paper, the researchers show that their device can detect a reactor shutdown within 5 hours, potentially blocking efforts to turn off a reactor and access the fuel for weapons. The detector can measure power levels with a 3.5 percent level of precision, good enough for the IAEA to consider its use.

These initial results have piqued the IAEA’s interest. According to Julian Whichello, a safeguards technology specialist at the IAEA, the agency had not taken an active interest in antineutrino technology for years, because earlier detectors were too large and had not been tested in a real setting. The American group has done the first practical demonstration, and its detector is promising, because it is not much bigger than other systems the IAEA currently deploys at reactors, Whichello says.

Antineutrinos are subatomic particles with little mass and no charge. Nuclear reactors are awash in them—about 10²¹ antineutrinos flow out of a reactor every second. These come from both the reactor’s uranium and plutonium. The uranium fuel in nuclear reactors contains two isotopes: uranium-235 and uranium-238. Of these, uranium-235 is the main source of antineutrinos. When it absorbs a neutron, its nucleus undergoes fission, forming daughter nuclei that produce antineutrinos. On the other hand, when uranium-238 captures neutrons, it sometimes forms uranium-239, which decays to form plutonium-239, also a source of antineutrinos.

Antineutrinos are an ideal nuclear-fission signature for detection. As opposed to other nuclear signatures, such as gamma rays or neutrons, antineutrinos do not react with matter easily, so they penetrate the reactor’s radiation shielding and travel for long distances. “You can install detection equipment quite a distance from the core,” says Andrew Monteith, a physicist at the IAEA, “which means you don’t have to interfere with the operation of the reactor [and you] don’t have to wire anything into the reactor building.”

But the particles are also harder to detect. The LLNL/Sandia detector contains stainless-steel cells filled with a liquid scintillator—a proton-rich material such as mineral oil. An antineutrino combines with a proton to create a neutron and a positron. These two particles, it turns out, are easier to detect, because they create light flashes in the scintillator. “What we look for is two flashes of light consistent with the signature of both the positron and neutron appearing at very close time coincidence in the detector,” Bernstein says.

In addition to antineutrinos, each fission reaction releases a quantifiable amount of energy. So by measuring the number of antineutrinos in a given time, the researchers can calculate the reactor power. They can also calculate the amount of uranium and plutonium in the core using the antineutrino rate. The fission of uranium-235 and that of plutonium-239 generate a different number of antineutrinos. As uranium in the reactor is gradually used up and plutonium is created, the antineutrino rate decreases.

The IAEA already uses a suite of technologies to monitor the nuclear material going in and out of commercial reactors. These include surveillance techniques such as satellites and cameras, as well as tags unique to each canister of nuclear material.

For smaller research reactors, the agency also keeps track of the reactor power. Small reactors are especially important to monitor, says the IAEA’s Whichello. They are simpler than power reactors, so it would be easier to secretly remove fuel for use as a weapon.

But today’s power-monitoring technology is complex. It involves installing devices that measure the coolant’s temperature and rate of flow at critical points in the reactor’s cooling system. The devices give inspectors an indirect indication of the reactor’s power. By contrast, antineutrinos give a direct measurement of the power and fuel content in the core. “You’re monitoring directly what’s going on in the core of the reactor rather than degrees of separation away from the reactor,” Whichello says.

Antineutrino detectors might also be easier and cheaper to implement. Coolant-measurement systems have to be designed differently for each nuclear reactor. They are also intrusive and difficult to install. “You have to position rather delicate probes on the cooling line in the facility and then lead wires and cables to a data-collection box,” Whichello says. An antineutrino detector, on the other hand, needs no connection to the reactor and could be discreetly situated up to 100 meters away from it.

Still, Monteith says it will take about a decade before we see the new detection technology at commercial reactors. Three other groups—in Brazil, France, and Japan—are planning to test their antineutrino detector prototypes in the next three to five years. These groups, along with the American researchers, still need to better their designs to meet the IAEA’s needs, Monteith says. They will, for instance, need to shrink the detector’s size. More important, they will need to switch to plastic scintillator materials, because the liquid scintillators used now are flammable. “Reactor operators won’t want flammable liquids on-site,” he says.