A Thousand-Year Plan for Nuclear Waste

Tiny companies may help the DOE solve the biggest problems

The Hanford Nuclear Reservation is home to North America's worst witch's brew of hazardous waste. The 560-square-mile site in eastern Washington, which produced plutonium for nuclear warheads from 1943 to 1989, houses 53 million gallons of plutonium, uranium, iodine, mercury, and countless other contaminants known to cause cancer, thyroid disease, and other illnesses. One million gallons have already leaked into the soil and are oozing toward the Columbia River, threatening a million residents in the vicinity. When it comes to environmental disaster areas, "Three Mile Island doesn't even begin to compare," says Dirk A. Dunning, a nuclear specialist for Oregon's Office of Energy.

Watched over by the U.S. Energy Dept., the custodian of America's weapons labs, Hanford has been the focus of countless remediation studies. This fall, the DOE will start constructing plants that will transform the worst of the radioactive waste into glass--a process called vitrification. For at least the next 1,000 years, that should prevent the most lethal elements from leaching into the environment.

The plan is far from perfect. Hanford expects to produce about 500,000 tons of radioactive glass that would have to be buried on-site or shipped to other locations such as the proposed Yucca Mountain repository in Nevada. The cleanup could whack taxpayers with a bill exceeding $40 billion. And the job is so massive that the DOE's Office of River Protection, which is managing the effort, has already warned that it won't make its 2028 deadline unless it finds new technologies to improve the process. "We need to get smarter," says Harry Boston, who heads the office.

A tiny group of scientists in San Diego, backed by some of the biggest guns on Wall Street, believe they have a cure for the Hanford headache. Archimedes Technology Group Inc. has invented a machine to filter out the most toxic radioactive elements--the ones that must be enveloped in glass. The least dangerous sludge that's left over might then be processed in a less costly manner. The contraption may thus shave years and billions of dollars off the current cleanup plan. Archimedes has raised $42 million in personal investments from the likes of George R. Roberts, partner in New York leveraged-buyout firm Kohlberg Kravis Roberts & Co. Tully M. Friedman, CEO of Friedman Fleischer & Lowe and an investor in such companies as Levi Strauss & Co. is also on board. "If we could solve this problem, that would be a great legacy," says Friedman.

Nuclear-waste remediation could be a great business, as well. Aside from the work at Hanford, major cleanup efforts are under way or planned at dozens of sites (table). Then there's the high-level waste from 103 nuclear-power plants in the U.S. that will someday have to be removed from the deep pools it's stored in and placed somewhere safe. Factor in radioactive sites in Europe, Asia, Australia, and the former Soviet Union, and you end up with a potential market in the hundreds of billions of dollars.

The DOE and its national labs have been working for decades to develop better ways to deal with nuclear waste, but now more and more private companies are getting involved. The Department expects to award nearly $50 million in grants in 2003 to companies with new ideas for dealing with the materials. "We're looking for radical alternatives," says Gerald Boyd, assistant manager for environmental matters at the DOE's office in Oak Ridge, Tenn.

Unlike most waste startups, Archimedes is using private funding. It's building a demonstration unit of its invention, called the Archimedes Plasma Mass Filter, in San Diego, and will begin testing it next year with nonradioactive chemicals that mimic the composition of Hanford's myriad oozes. If it works, they'll negotiate a Hanford contract with the DOE.

Nobody disputes that the Plasma Mass Filter is novel. Until now, scientists have focused mostly on chemical separation--to no avail. "The people who brought us the bomb brought us a chemical knot we can't untie," says John Gilleland, Archimedes' CEO. The company's solution ties together different techniques used in plasma physics and nuclear fusion research. The filter, about the size of a large sport-utility vehicle, will use heat and electromagnetic fields to draw the most radioactive elements out of the waste and shoot them into glassmaking machines. Then the radioactive elements are mixed with molten glass and poured into stainless-steel containers. The glass solidifies, trapping the radioactivity in logs that can be buried or stored in vaults. By separating out the most egregious components, the Archimedes device might cut the amount of high-level waste that must be turned into glass by 75%.

Outside the U.S., Russia might be a major customer. Its man-made Lake Karachai in Siberia served as a dump for Soviet weapons manufacturers and contains roughly as much radioactivity as was released at Chernobyl. That waste may be headed for the Arctic Ocean. "If it gets there, we'll have an international disaster," says Deborah Brosnan, president of Sustainable Ecosystems Institute in Portland, Ore.

Before rescuing the world, Archimedes must prove its invention works. While each component of the filter has been tested elsewhere--the plasma technology is used to make computer chips, for example--the parts have never been tried together. "Given the risks," says investor Friedman, "we had to go in believing we could lose all our money."

Lately, the DOE has been open to all suggestions. Earlier this year, at the Office of River Protection, Harry Boston met with Archimedes. The filter "is not ready for prime time, but we're still thinking about it," says Boston. He'd better be. In the past half-century, science has devised many methods to spew radioactive wastes, but precious few to effectively neutralize them.

By Arlene Weintraub in San Diego, with Laura Cohn in Washington, D.C.

    Before it's here, it's on the Bloomberg Terminal.