"Hot Case," Feature Article, January 2004

FEATURE FOCUS: Nuclear Power

hot case

A national lab is revisiting the past to find new methods of putting uranium into reactors.

by Jeffrey Winters,
Associate Editor

Oak Ridge National Laboratory is one of the crown jewels of the national nuclear infrastructure. It is here, some 60 years ago, that the world's first continuously operated reactor went critical. Enriched uranium processed at the lab fueled an atomic bomb dropped on Japan. And the lab also was instrumental in developing portable nuclear reactors for the Army and radioisotope power for deep space probes.

Even so, things aren't all as up-to-date as one might imagine. Take the furnaces used for depositing chemical vapor coatings. Engineer Rick Lowden uses them for materials science research for nuclear power applications, work that has taken on an increased importance in recent years. The furnaces themselves don't betray that emphasis, though. "The furnace I use is as old as I am," Lowden said.

The old furnaces have been upgraded and are still serviceable, and Lowden's work is making strides. Two decades have passed since Oak Ridge worked on designing new nuclear fuel particles. If he and his team can figure out how to make flawless coatings, nuclear power may wind up being made safer and cheaper, finally living up, in part, to its green potential.

It's a lot to ask from a carbon coating just a few microns thick.

The yellow gel spheres or uranium at Oak Ridge National Laboratory are transformed into carbon-coated particle fuels.

Nuclear reactors come in four main varieties, grouped on the basis of moderator or coolant: light water, heavy water, liquid metal, and gas cooled. Of the four, light water reactors dominate the commercial market, and their design is probably the most familiar. Uranium oxide powder is pressed to form small pellets that are loaded into metal rods; the rods are submerged in a tank of water that is heated by the fission of atomic nuclei in the fuel. That heated water produces steam to drive a generator.

The fuel pellets in a water-cooled reactor are supported and protected by the metal rods, so they can be relatively primitive in design. But the metal rods are also a limitation on the range of designs. Metals, though tough, have relatively low melting points; this means that the rods must be carefully cooled or they will succumb to temperature stresses and fail. Working temperatures in light water reactors are generally under 600°F, or 316°C.

Nuclear fuels, on the other hand, are capable of much higher temperatures, and some reactor designers would like to take advantage of that fact to achieve higher efficiencies. In particular, reactors that are gas-cooled, rather than water-cooled, can reach temperatures of more than 2,000°F, since high-temperature gases are easier to contain than high-temperature liquids (which must remain under pressure to prevent boiling). The U.S. Department of Energy has commissioned research into advanced gas-cooled designs that would employ extreme heat to generate hydrogen for use in fuel cell-powered vehicles, and such designs are being considered as part of the Generation IV Nuclear Energy Systems Initiative.

The temperatures needed to make such designs practical all but rule out the use of metal rods in the reactor core. Instead, the fuel would be loaded into ceramic vessels that would be cooled by a constant flow of helium gas. The best configuration of the fuel element is still open to debate. One design would press fuel particles into pinky finger-size cylinders and load them into graphite blocks riddled with channels. Gas would pass through the channels to carry away the heat of radioactive decay. This "prismatic" design has been studied extensively and a test reactor of this design is currently operating in Japan.

Another design that has received much attention is known as the pebble bed reactor. In this scheme, coated fuel particles are formed into billiard ball-size spheres, which are stacked in the containment vessel. Gas flows through the gaps between the stacked balls to convey the heat to a heat exchanger. The pebble-bed design has been tested in Germany, and South Africa has an active program. Vice President Richard
Cheney's energy task force touted pebble-bed reactors, and a 10-megawatt reactor with this design started operation in China in 2001.

One feature of the pebble bed design is the movement of the fuel pebbles through the containment vessel. Spheres are continually being removed from the bottom of the stack and checked to see if they have enough fissionable material: Those containing sufficient fuel are returned to the top of the pile, while those made up of mostly spent fuel will be sent to waste deposits. The sealed pebbles are designed to contain the spent fuel indefinitely, which is intended to make waste disposal if not trouble-free, at least less fraught with hazard than at present.

NEW CONTAINMENT

When you take water out of the containment vessel, you are also removing one of the key moderators of the nuclear reaction. Moderators slow the neutrons that are produced by nuclear fission and that help sustain a chain reaction. (Unmoderated neutrons move too quickly to be easily captured by uranium nuclei, and too many wind up escaping the fuel assembly.) The lighter the element, the better the moderator, since neutrons are slowed by collisions with atomic nuclei, much the way billiard balls are, and neutrons that run into a heavy nucleus shed very little momentum. Hydrogen, the lightest element, works the best, and liquid water is perhaps the easiest way to deliver hydrogen in a dense form.

Gas-cooled reactors don't have the luxury of using the coolant as the moderator. The density of any gas is much too low to be useful. Instead, researchers have taken to encasing the fuel particles themselves with a moderator: carbon, which is both light and relatively impervious to high temperatures. Graphite can withstand temperatures of more than 5,000°F before it begins to sublime. (Graphite has no liquid phase.) In the prismatic design, the graphite block also serves to moderate neutrons.

Coated nuclear fuel particles are shown here after chemical vapor deposition.

Frank Homan is an engineer who was involved with the Oak Ridge effort to develop gas-cooled reactors in the 1970s and 1980s. High-temperature gas-cooled reactors were seen as being incredibly useful as possible sources of tritium, an isotope of hydrogen that is an essential ingredient in atomic weapons. They might also have pointed the way toward a new generation of nuclear power plants. "Light water reactors got their start in nuclear submarines," Homan said, "so the model was that a military application might be able to spin off something of use for the commercial sector."

Homan's group made considerable progress, but the need for tritium fell through the floor with the end of the Cold War. By the early 1990s, the Energy and Defense departments lost interest in the project, and the program was discontinued. The facilities and expertise built up at Oak Ridge were effectively mothballed.

GAS-COOLED RESURRECTION

Interest in gas-cooled reactors reappeared in the late 1990s. When an international consortium investigated building such reactors in the United States, federal energy officials realized that the country might fall behind Europe or Asia in developing the technology needed to make such reactors a commercial reality.

In 2001, Lowden and his colleagues were put on the task of resurrecting and improving methods to produce coated fuel particles for gas-cooled reactors. "That's when we were chosen to reestablish research capability in the coating of particle fuel," Lowden said. "Our job is to make the coatings as good as they're ever going to be."

Studies in reactors such as the Fort St. Vrain reactor in Colorado built in the 1970s discovered that out of every thousand fuel pellets, one would break or chip or be improperly coated. Considering that a single ball in a pebble-bed reactor contains 16,000 fuel particles, such a failure rate would be unacceptably high.

A researcher inserts a fuel stick containing non-nuclear coated particles into a hole in a hexagonal graphite fuel element block.

The fuel particles Lowden's team creates begin as a uranium-bearing liquid. That liquid is injected into another fluid to form droplets that eventually gel into small spheres. The spheres are then collected and dried to form uranium oxide kernels about a millimeter across.

The balls are placed into a chemical vapor deposition chamber to be coated with layers of carbon and silicon carbide and then compacted to make fuel particles. These particles, which look like nuclear caviar, can be pressed into various shapes, depending on the needs of the reactor design.

The goal of the work at Oak Ridge is to uncover and eliminate the sources of failure in the nuclear fuel-making process. Researchers there have built on discoveries by Chinese and Japanese researchers to reduce the pressure needed to compact the coated fuel particle into the finished pebble shape. And they've cut the number of steps in the entire process, which reduces the chances for mishaps.

The ceramic coating itself is a new composite, with sandwiched layers of graphite and silicon carbide. The hope is to make a shell that can seal in radioactive gases and solids produced by uranium fission. "We're putting a miniature containment vessel around each particle," Lowden said. The better the seal, the less likely it is that radioactive material will escape during an accident.

BREWING UP FUEL

Laying down the carbon coat requires vaporizing the material in high-tech furnaces. Even so, making nuclear fuel particles is as much an art as it is a science. Homan, who has consulted with Oak Ridge (serving as an institutional memory, of sorts), said that sometimes the quality of fuel is so dependent on the expertise of a couple of technicians that whole batches of fuel made while key personnel are on vacation often have to be rejected.

One of the important goals for the renewed Oak Ridge effort is to standardize the procedure so that near-perfect fuel can be made no matter who is at the controls. Rather than a craft like brewing, manufacturing fuel would become an industrial process.

The Oak Ridge work also involves developing non-carbon coatings. Carbon makes a great moderator, but for some fission applications, you need fast neutrons. Specifically, if you want to make new fuel inside a breeder reactor, fast neutrons are needed to turn thorium or low-grade uranium into more potent isotopes of uranium or even plutonium.

The Oak Ridge team has been investigating using non-carbon coatings that would wave by fast neutrons. Materials such as zirconium nitride have been looked at, with the goal of creating an easily dissolved ceramic that could be used in fast-breeder reactors. In such applications, the fuel pellets would be ground up and the casings removed; the remaining plutonium or uranium would then be used to make new pellets.

Breeding fuel in this way is seen as a way to stretch limited supplies of uranium into an enduring energy source. But in some respects, this is just another blast from the past. The U.S. fast-breeder reactor program was abandoned in the early 1980s. But as nuclear power gets a new look, almost everything old is now new again.

"When I started at Oak Ridge," Lowden said, "my job was figuring out what to do with these particle coaters we weren't using any more. Twenty years later, I'm putting them back."