The system of making and distributing electricity in the U.S. is a century old and miserably outmoded. Electricity is sent to factories and homes from large, central power stations, often built far from big cities because of the pollution factor. The old power lines leak like sieves; between 5 and 7 percent of U.S. electricity is lost through the nation’s 200,000 miles of high-voltage wires. But erecting new transmission lines is political drudgery, requiring cooperation across multiple local, state, and federal jurisdictions, any one of which can stall a project for years. In addition, the utilities have delayed expanding and upgrading existing power plants because doing so would require them to install state-of-the-art pollution controls that they contend are too expensive. Consequently, the U.S. grid has stagnated, with the capacity for generating power growing four times faster than the capacity to transmit it. Bill Richardson, the governor of New Mexico and a former energy secretary, calls the grid “third world.”

The decrepit system is a serious impediment to renewable-energy projects on a grand scale. To move wind and solar power to consumers from the breezy Great Plains and the sunbaked deserts of the Southwest, the U.S. needs about 20,000 miles of new transmission lines. It also needs a massive upgrade of the analog grid that directs the energy from place to place, with new computers, sensors, and communication gear to manage the network.

Unlike nuclear reactors and most fossil-fuel-burning plants, windmills and solar cells produce electricity only when the wind blows or the sun shines. An automated grid is crucial for managing the constantly fluctuating power from renewable sources.

The world as envisioned by Google includes a vast computer network that monitors and controls the nation’s electricity grid and sets prices for power based on real-time supply and demand. For example, the system could, on a particularly hot afternoon, send a signal to millions of utility customers warning that power prices are soaring. The information could be fed directly into an energy-management system linked wirelessly to people’s air conditioners and appliances, and their Jacuzzis, garden lights, and electric cars. After being programmed, the system would automatically shut down designated devices if prices hit preset levels, just as program trading automatically buys and sells stocks. For those without automatic systems, it would take just a few keystrokes from a computer at the office to power down selected machines at home and avoid being walloped by the price spike.

The grid itself would work in similar ways. If it faced shortages, it could send out a signal offering to buy back power stored in people’s electric car batteries for a healthy premium above what the same electrons cost just 15 hours earlier. Those interested would click accept on their computer screens. The network would locate their vehicles and automatically activate decharging. Eventually, demand and prices would drop, triggering dishwashers and clothes dryers to switch on. Electric cars would resume charging.

By Schmidt’s reckoning, the smart grid that Google wants to build would fix a massive market failure. Electric utilities, old and “structurally slow,” he says, haven’t invested in new technologies, even “when there’s a significant business opportunity before them.” In most states, they have an incentive simply to produce and sell as much power as possible. As monopolies that get paid for moving big volume, the utilities have no reason to make electricity use smarter. This, Teetzel says, has made power companies “traditionally risk averse.” He adds, “That’s not in our DNA.”

Google’s green-energy czar, Bill Weihl, is a former professor at the Massachusetts Institute of Technology who made his name as a world-class computer scientist in the 1980s and 1990s, then worked for Digital Equipment Corp. and Compaq, where he led research on distributed and parallel computing that produced 19 patents. Now he’s fully immersed in energy. He drives a Prius and has solar panels on his San Francisco rooftop, which he expects will pay off in energy savings after 20 years.

When Weihl came to Google in 2006, he was given a specific assignment by Brin and Page: to green Google’s sprawling data centers and “do it in a way that makes the rest of the world do it too,” Weihl explains.

Energy has preoccupied Google for years. The servers used for Web searches consume vast amounts of expensive energy, and running and cooling those servers has become a significant cost for the company. By altering the voltages and power supplies inside its servers, Google found it could reduce its overall energy consumption by as much as 50 percent below what most other companies use to run their systems.

The servers themselves are designed by the same group that builds and manages Google’s data centers, so energy considerations are integrated into every part of what Weihl calls the “total cost of ownership.” That means Google’s computer architects are supposed to think as much about heating and air-conditioning bills as its building architects do. “In many organizations, the guys in IT and the guys in facilities have never met,” Weihl says.

But even at Google, green energy doesn’t immediately pay. Renewable power, at roughly 8 cents per kilowatt-hour for wind and several cents more for solar, costs up to twice as much as electricity generated from coal. So Google is looking for ways to make renewable energy cheaper. It has invested more than $100 million so far in companies doing similar work, including a $15 million equity bet on Makani Power, which aims to produce utility-scale electricity from high-altitude kites, and $10 million investments in eSolar and BrightSource Energy, a pair of solar-thermal companies that use exquisitely calibrated mirrors to concentrate sunlight on a central water tower that acts as a boiler.

“Right now,” Schmidt said recently, “our primary mission is one of information.” If the company can use its core information business to promote more efficient energy use, he suggested, then “we’re clearly going to do that.”

Google’s green apostle after Schmidt is Dan Reicher, who directs energy and climate-change initiatives for Google­ .org, the company’s philanthropic arm. When we meet for lunch at Google’s cafeteria in its San Francisco office building—mahimahi burgers and po­lenta—Reicher, 52, is enthusiastic about his new role. An avid whitewater kayaker and conservationist, he spent a decade in the trenches as an environmental lawyer and prosecutor, and then eight years in senior posts at the Department of Energy under Bill Clinton, ultimately serving as assistant secretary in charge of energy efficiency and renewables. In 2007, Larry Brilliant, director of Google.org, lured him west from a private equity firm that Reicher had co-founded to invest in clean-energy projects. Google, Reicher tells me, has “all the tools under one roof” for an energy breakthrough: engineering, money, vision, and a lot of nerve.

But for Reicher, working with Google is a bit like going back to college. At the yogurt bar, surrounded by Googlers half his age, Reicher in his brown tweed looks like the wizened professor lunching with blue-jeaned students in the dormitory dining hall. When Google launched its initiative to develop low-cost green power—awkwardly named RE<C, for “renewable energy cheaper than coal”—some of the young engineers wanted to set a goal of producing 1.21 gigawatts of green electricity, the precise amount needed to power up Doc Brown’s flux capacitor of the time machine in the geek classic Back to the Future. (They even proposed linking their press release to a clip from the film.) “The adults prevailed,” Reicher says, relieved, and Google set the bar at an even 1 gigawatt, enough electricity to power a city the size of San Francisco.

Reicher says his mandate from Schmidt and the founders is to look for bold ideas. “For big impact, we’ll take some big risks,” he says. That may include investing in energy startups entering the dreaded “valley of death,” where so many go under for lack of funds. Google is also looking at filling some of the void in project financing created by the tanking economy. Reicher says, “You need look no further than the fact that AIG, Wachovia, and Lehman Brothers”—at one time, three of the biggest backers of green-energy projects—“are all gone” from the scene.