The Geysers—a geothermal power plant in northern California operated by Calpine—has been pumping out electricity harvested from steam heated deep within Earth since the 1920s. Since the 1990s Calpine has been pumping in treated wastewater from surrounding communities to replenish water lost during all those years of power production. And since 2000 the plant began producing more than six million megawatt-hours—enough to power 750,000 homes—annually, all while operating nearly 24 hours a day, seven days a week. "We run 98 percent availability," says Dennis Gilles, Calpine's senior vice president for geothermal operations. "That is one of the highest availabilities of any power source."

Tapping this geothermal resource is the subject of a new study prepared for the U.S. Department of Energy by the Massachusetts Institute of Technology (M.I.T.). The researchers estimate that more than 13 million exajoules (EJ), or 1.23 x 1022 British thermal units, lurk deep beneath U.S. soil and 200,000 EJ—or "2,000 times the annual consumption of primary energy in the United States in 2005"—is recoverable, without taking into account cost.

In some places, such as the Geysers, cost is minimal as nature does much of the work: fracturing the subsurface rock, filling it with fluid and heating it. In such locations, one merely has to build the conversion apparatus to turn the heat in steam into electricity, typically at a cost of $1,700 per kilowatt. But the M.I.T. team is proposing a much more ambitious plan that calls for constructing geothermal plants where none naturally exist, drilling down into heated bedrock, creating an open reservoir, and pumping water into it to be heated. "The biggest cost is basically in drilling the wells," says Brian Anderson, a chemical engineer at West Virginia University in Morgantown who helped develop the economic models for the report. "We're looking at a superlinear relationship between depth of completion and cost."

Of course, the know-how for such deep drilling already exists in the oil industry, which shells out roughly $9.4 million a pop for its 18,000-foot-plus deep wells, according to economist Michael Moore of the University of Calgary, another study co-author. In return, such a man-made geothermal power system would harvest 40 percent of the heat in that bedrock and then convert 15 percent of that heat into useable electricity. "You've got two closed loops there," explains Ronald DiPippo, a co-author and mechanical engineer at the University of Massachusetts, Dartmouth. "There is an environmental impact when you have to reject heat but that's a heck of a lot cleaner than burning coal or having to dispose of nuclear waste."

Based on costs alone (that is, how much of the geothermal resource could be tapped at prices as low as 6 per kilowatt-hour, the typical price of electricity from coal), more than 100 gigawatts of geothermal power could be developed for just $1 billion spread out over the next 40 years—the price tag of just one advanced coal-fired power plant and one third the cost of a new nuclear generator. "This is a very large resource that perhaps has been undervalued in terms of the impact it might have on supplying energy to the U.S.," says lead author Jefferson Tester, a chemical engineer at M.I.T. "Geothermal is the one sort of forgotten orphan here that could be much more compatible with our existing grid system than other renewables."

But geothermal is not entirely risk-free. A recent effort to build such a geothermal power plant in Basel, Switzerland, came to an abrupt halt when it triggered an earthquake measuring 3.4 on the Richter scale, too small to cause damage but large enough to be felt by humans. "We generate between 3,000 and 5,000 earthquakes a year," Calpine's Gilles says. "In a typical day, we experience on average 10. I can guarantee you won't feel any of them." It remains unclear what will happen with the Basel power plant, and the problem may result from its siting; Basel was leveled by an earthquake in 1356. "You are not going to want to put a geothermal facility like this where you have a danger of lubricating a big fault," Tester notes. "There would be no chance of doing this under the radar screen here as happened in Basel."

Geothermal energy can provide stable electricity that complements the intermittent power provided by other renewables, such as wind and solar. Creating such enhanced or engineered geothermal systems requires federal funding for research, the scientists say, something the Bush administration did not include in its fiscal year 2007 budget request. Congress allocated $24 million in 2006, and that funding level may be maintained this year, but it is still about half the amount that the report says is needed to develop geothermal into a resource that could power 25 million homes nationwide. "The surface has barely been scratched as far as harnessing the potential of this indigenous resource, not dependent on foreign fuels," Gilles adds. "We would like to see greater efforts exerted towards realizing its full potential."