Fearing the collapse of an already faltering industry, Congress in 1982 drew up the Nuclear Waste Policy Act, ordering the Energy Department to open a nuclear waste storage facility within 16 years. But Congress did nothing in that time except select Yucca Mountain as the only location worth studying, delaying until 2002 to give the site final approval and allow the Energy Department to at long last apply for a license. As of December 2006, $13.5 billion from utility bills and taxes has gone into researching the site, and utilities have accumulated some 55,000 metric tons of waste. By 2010, the waste will exceed Yucca Mountain's limit of 70,000 metric tons. (Both Sproat and Energy Secretary Samuel Bodman recently advised lifting that cap instead of finding a second site.)

Doubts about Yucca Mountain's geologic suitability have piled up as well. Six hundred earthquakes have rumbled under Yucca Mountain in the last 20 years, one as great as magnitude 5.6. A panel of scientists put the chances of "igneous disruption" in the ridgeline's ancient field of volcanoes at one in 6,250 over the next 10,000 years — which seems low until you consider that, in most of the United States, the probability of a volcano erupting is zero.

Even the site's chief meteorological selling point — the dryness of the Nevada desert — may no longer play in its favor. For one thing, climates can change: In the winter of 2004 to 2005, enough rain fell in Death Valley, 20 miles to the West, to revive seedbeds that had lain dormant for a century. For another, the absence of water may not be as significant as the presence of air. "Yucca Mountain is an oxidizing environment," says Allison Macfarlane, an associate professor at George Mason University and editor of a book on Yucca Mountain, "and spent nuclear fuel is not stable in the presence of water and oxygen."

The Energy Department has tried to engineer away Yucca Mountain's geological deficiencies, promising to contain the waste in decay-resistant alloy canisters. It's also suggested using robots to install titanium drip shields in 100 years. (The robots have yet to be invented.) But Macfarlane argues that it's far better to encase spent fuel in rock where oxygen can't get to it. The U.S. already stores defense-related radioactive waste one-half-mile deep in 250 million-year-old salt formations at the Waste Isolation Pilot Project, near Carlsbad, N.M. Sweden and Finland both have plans to contain spent fuel rods in copper and bury them in airless crystalline bedrock, 500 meters beneath the surface.

"We have a really big country," says Macfarlane. "We're really lucky. There are lots of places that would be good geologically, and that wouldn't require transporting waste over long distances. We ought to be thinking about sharing the burden among different states, maybe among states that have nuclear power plants."