Planted in the New Mexico desert near Albuquerque, the six solar dish engines of the Solar Thermal Test Facility at Sandia National Laboratories look a bit like giant, highly reflective satellite dishes. Each one is a mosaic of 82 mirrors that fit together to form a 38-ft-wide parabola. The mirrors’ precise curvature focuses light onto a 7-in. area. At its most intense spot, the heat is equivalent to a blistering 13,000 suns, producing a flux 13 times greater than the space shuttle experiences during re-entry. “That’ll melt almost anything known to man,” says Sandia engineer Chuck Andraka. “It’s incredibly hot.”

The heat is used to run a Stirling engine, an elegant 192-year-old technology that creates mechanical energy from an external heat source, as opposed to the internal fuel combustion that powers most auto­mobile engines. Hydrogen gas in a Stirling engine’s four 95 cc cylinders expands and contracts as it is heated and cooled, driving pistons to turn a small electric generator. The configuration of the dish and engine represent the fruit of more than a decade of steady improvements, developed in collaboration with Arizona-based Stirling Energy Systems.

On a crisp morning this past January, Andraka and his colleagues fired up Dish No. 3. The temperature was around freezing, and the sky was 8 percent brighter than average—the contrast between the cold air and the hot sun helps the engine run more efficiently. When power began to flow from the 25-kilowatt system, it did so with the highest conversion efficiency ever recorded in a commercial solar device: 31.25 percent of the energy shining onto the giant dish flowed into the grid.

To Bruce Osborn, president and CEO of Stirling Energy, this merely confirmed something that he already knew: The system, which his company calls the SunCatcher, was ready to exit the laboratory. “The rocket science is already done,” he says. The challenge remaining is to turn the prototypes into a low-cost, mass-producible design—“just a question of good, old-fashioned engineering,” according to Osborn. To that end, Stirling Energy signed the two largest solar energy contracts in history with two Southern California utilities, promising to build up to 70,000 SunCatchers and provide power for a million homes. Construction starts next year.

Big promises from solar power companies are nothing new. “It is stern work to thrust your hand into the sun and pull out a spark of immortal flame to warm the hearts of men,” an AT&T publicity film crowed after the invention of the silicon photovoltaic (PV) cell in 1954. “Yet in this modern age, men have at last harnessed the sun.”

Well, sort of. The Bell Solar Battery, as it was called, had some successes—powering the first communications satellite, in 1962, for instance—but hopes of cheap, plentiful energy have remained elusive.

PV cells and concentrating solar thermal (CST), the two basic methods for harnessing the sun’s power, have made great strides since those early days. But inflation in the cost of raw materials, such as silicon, combined with decades of cheap fossil fuels has kept overall solar energy consumption in the U.S. at 0.08 percent. And a series of new technologies that looked promising in the lab have proved impractical on the open market, leaving many observers to conclude that the age of solar energy will always remain just around the corner.

Meanwhile, though, almost under the radar, a few solar technologies have reached maturity. A type of silicon-free solar panel, half as expensive as silicon cells, has rapidly turned Arizona-based First Solar into the biggest solar-panel maker in the country. And along with Stirling Energy’s SunCatcher, new CST designs promise to provide a steady flow of solar electricity—even at night.

Big power utilities love CST for two reasons, says Reese Tisdale, a senior analyst at Emerging Energy Research, based in Cambridge, Mass. “It’s large-scale and it’s [usually] steam-powered, so it’s not so different from the gas- and coal-fired plants they’re familiar with.” The idea is not new—in fact, nine CST plants with a combined capacity of 354 megawatts have been operating in the Mojave Desert since their construction between 1984 and 1991, powering the homes of 500,000 Californians and proving the design’s reliability. (An average coal plant produces about 670 Mw.) The plants use a “parabolic trough” design, with more than 900,000 mirrors, shaped like a skateboarder’s half-pipe in vast arrays over 1500 acres of desert. The mirrors adjust to track the sun across the sky, reflecting and concentrating its rays onto liquid-filled pipes. The hot liquid, in this case oil, then boils water, which produces steam to spin a turbine.

Progress on CST plants ground to a halt after natural gas prices plummeted in the 1990s. It wasn’t until last year that the next major plant in the United States opened: a 64-Mw parabolic trough system in Boulder City, Nev., called Nevada Solar One, built by the Spanish company Acciona. Now there are 13 other plants, totaling 5100 Mw, in advanced planning stages in ­Flor­ida, Arizona and California; most will use parabolic troughs. Stirling Energy pursued a different kind of system, one that offers more flexibility and better efficiency.

Bruce Osborn started his research career at Ford Motor Co., and the key advantage of his solar dish is one his former employers would understand. “Henry Ford used to say you can have your car in any color as long as it’s black,” Osborn says, “and that’s our approach, too.” The planned 900-Mw Stirling Solar Two plant near San Diego will eventually have as many as 36,000 identical dishes, and the 82 mirror panels that make up each dish come in only two shapes. That design choice causes a slight decrease in power output, in exchange for the advantages of low-cost mass production.

Modularity has other benefits, too. Since each 25-kw SunCatcher has its own Stirling engine producing electricity, there’s no single point of failure. “If something goes wrong with one dish, it doesn’t matter,” Osborn says. In contrast, the thousands of mirrors in a parabolic trough plant all feed a central turbine, so when the turbine is down for maintenance, power production stops. The SunCatcher design also shortens the wait for power during construction: Electricity will flow once the first 40 are built—a “solar group” that can churn out 1 Mw.

The breakthrough efficiency of the dish results from focusing the sun’s rays on a single spot instead of on a long pipe, which allows temperatures to reach 1450 F, compared to 750 F for parabolic troughs. In addition, the Stirling engine has a relatively flat effi­ciency curve: It produces close to maximum output even when the sun is obscured or low in the sky. So while the record 1-hour effi­ciency achieved earlier this year was 31.25 percent, the SunCatcher’s full-year, sunrise-to-­sunset efficiency is still a respectable 24 to 25 percent, roughly double that of parabolic trough systems.

Another twist on CST designs confronts the challenge that dogs every solar power scheme: “When the sun sets, that’s it for the day,” as Tisdale puts it. “But in Arizona in midsummer, it’s hot as hades, so people have their a/c cranked until 9 or 10 in the evening.” A hot liquid can be stored more efficiently than electricity; the analogy used by one industry executive is that a $5 thermos can hold as much energy in the form of heat as a $150 laptop battery can store electrochemically. Two 50-Mw plants that should begin operations by the end of this year in Spain will operate on this principle, using what amounts to a giant thermos filled with molten salt.

In the U.S., a thermal storage facility is scheduled for completion in Gila Bend, Ariz., in 2011. The 280-Mw Solana plant, being built by Spanish company Abengoa Solar, will use a parabolic trough design, but will incorporate a thermal storage tank that can keep the plant running for 6 hours with no sun. “We could design a plant that runs 24 hours a day,” says Fred Morse, an adviser for Abengoa who was formerly the Department of Energy’s solar czar, “but that would make no economic sense.” Instead, the plant is designed to cover Arizona’s peak energy-use periods, when power is most expensive.

The enormous scale of the Abengoa and Stirling Energy plants provides an answer to skeptics who doubt whether a few rooftop panels here and there can ever play a meaningful role in the world’s energy portfolio. But size also creates its own set of problems. For one thing, the power has to be transmitted to where it’s needed, and the empty deserts best suited for sprawling CST plants tend to be in the middle of nowhere. The site of Stirling Energy’s future plant for the San Diego market currently has enough transmission capacity for 300 Mw, or 12,000 dishes. The remaining 24,000 dishes will be built only if San Diego Gas & Electric is able to complete a proposed 150-mile transmission line between the plant and the city.

Water use is another issue. CST plants with steam turbines can require hundreds of millions of gallons of water to cool their con­densers—a challenge in regions where water is already at a premium. In this respect, Stirling Energy’s hydrogen­-based system has a significant advantage, since it only uses water to rinse the mirrors every few weeks. Osborn estimates that the San Diego plant, when producing power for 500,000 households, would use the same amount of water as 33 average homes.

Utility-scale solar power also requires enormous capital, which keeps it out of reach of people in the developing world, where such solutions are desperately needed. That’s a challenge RawSolar, an MIT spinoff, is trying to meet with a dish that is just 12 ft. wide, and simple and cheap enough to make for stand-alone operation. The nonprofit Solar Turbine Group, another MIT spinoff, built an even more bare-bones mini-CST system in Lesotho last summer, using spare car parts for the heat engine.

The most natural fit for small-scale solar, though, is the good old photovoltaic cell. It takes in sunlight and spits out electricity with no moving parts, requires no water and can be situated wherever electricity is needed, to avoid transmission losses. PV panels can generate useful amounts of electricity even in the weaker sunlight of northern states where big CST plants aren’t practical. Also, they’re ideal for homeowners, since they are simple to install and maintain—in fact, integrated building materials like PV roof tiles will make new homes even easier to connect.

In July, Southern California Edison installed the first of what will be 250 Mw worth of PV panels located on commercial rooftops throughout the utility’s territory, where power is most in demand. But instead of silicon, the panels were made of a thin film of cadmium telluride, or “cad-tel” for short. Thin-film PV has been touted for years as a cheaper replacement for traditional silicon cells, but past designs have had trouble scaling up to mass production. Cad-tel technology has “completely changed what people thought could be done with thin films,” says Larry Kazmerski, director of the National Center for Photovoltaics at the National Renewable Energy Laboratory in Colorado.

First Solar, the company that made the panels, estimates its manu­facturing cost to be $1.14 per watt and falling, about half the cost of comparable silicon panels. As a result, Kazmerski says, “There’s a big turn happening.” First Solar quadrupled its manufacturing capacity from 2006 to 2007, to 396 Mw, and it expects to exceed 1000 Mw next year. Two years after its initial public offering, the company’s market value is over $20 billion—double that of General Motors.

Cad-tel isn’t the only promising thin-film technology on the market. Newer panels developed using a copper indium gallium selenide (CIGS) semiconductor have efficiency ratings almost 30 percent higher than First Solar’s cad-tel PVs. The advances have sparked a flurry of startup companies. Venture capitalists are pouring in 20 to 100 times more money than government research funds are, Kazmerski says, creating what some are calling a dot.sun phenomenon.

California-based Nanosolar is among the companies racing to commercialize CIGS technology. But like First Solar, most of its sales have gone to European countries such as Germany and Spain, where long-established policies provide a stable, guaranteed price for solar power production. Here in the U.S., uncertainty looms about a 30 percent investment tax credit that is set to expire at the end of the year. For billion-dollar projects such as Abengoa’s Solana plant, extension of the tax credit is make-or-break: These projects simply won’t happen without an extension of at least eight years.

Ultimately, solar power will have to justify (and pay for) itself—and the market may be moving in that direction. The DOE predicts that solar electricity will be cheaper than the average grid price by 2015. What’s more, prices for natural gas have doubled in the past five years, coal has nearly tripled, and new nuclear plants won’t come on line for at least seven more years. Locking in a long-term contract with a solar plant whose fuel will never run out, on the other hand, is the very definition of energy security. “One thing we know about the sun,” Morse says, “is that the price never goes up.”