This article discusses the innovative programs of Ausra, a solar power company based in Palo Alto, California, to convert sunshine into electricity. The team is trying to use the best tools available to design a renewable energy system that can be put together by largely unskilled labor and do it cheaply enough to be profitable. The paper also highlights that instead of using one parabolic surface, Ausra divides its mirrored reflectors into strips, each of which concentrates light into a set of pipes mounted 40 feet overhead. A single square mile of mirror field like this one near Bakersfield can generate as much as 80 MW. The Ausra manufacturing plant in Las Vegas is a garden-variety factory, using robotic welders and hard-hatted workers to build the trusses that support the mirrors. According to researchers, a look at a solar radiation map of the United States shows that finding sunshine ought not to be a problem. While the eastern half of the country has too many partly cloudy days to make much use of solar thermal power, the Southwest is one of the best locations in the world for it.

## Article

It was probably an impossible task: infusing a sense of accomplishment and excitement into the start of an automated sequence of tasks. And when Harry Reid, Senator from Nevada, pushed a button, there was no fanfare, no chords from Also Sprach Zarathustra, nothing to announce that this was a Big Event. Instead, a dozen or so robotic arms, partly obscured from view and placed a safe distance from the assembled dignitaries, began positioning a large mirror and welding together a steel framework.

Within eight minutes, the sparks stopped. It was time for pictures, followed by a reception.

To a certain extent, the official opening ceremony at the Las Vegas manufacturing facility for Ausra, a solar power company based in Palo Alto, Calif., was no big deal. But if the team of scientists, engineers, and venture capitalists behind Ausra are right, the mirror built that day in late June would be the first of countless thousands to be produced by the factory, all of which would be used to help turn desert sunshine into electricity.

The vision moving Ausra forward is also groundbreaking in another way. Solar-generated power has, up to now, been largely the product of high-tech, solid-state physics-the photovoltaic cell, which converts light directly into electricity. Ausra's product couldn't be any more different. It is galvanized steel and silvered glass and steam pipes. The system uses garden variety steam turbines. "This is as pure a mechanical engineering challenge as you'll find," said David DeGraaff, vice president of product engineering.

Indeed, what DeGraaff, who is a mechanical engineer, and others at Ausra have to do is counterintuitively difficult: Use the best tools available to design a renewable energy system that can be put together by largely unskilled labor. And do it cheaply enough to be profitable.

If they get all that right, and pick up a little regulatory luck, the company has a chance to change the way we make electricity in the United States.

THE SOLAR POWER TECHNOLOGY THAT PEOPLE MOST OFTEN ENCOUNTER IS THE PHOTOVOLTAIC CELL. Solar cells generate electricity in a fairly direct way: Photons of light create a voltage within a thin layer of semiconducting material. The voltage isn't particularly large, and the power generated in a square foot of typical photovoltaic material exposed to direct, noonday sunlight is less than 20 watts.

Photovoltaics have other problems, too. They can be difficult to manufacture and are often subject to raw material bottlenecks. But for all their faults, PVs have some outstanding qualities, from quiet operation and good reliability to having the ability to supply power on many different scales.

Solar cells have become so familiar-powering everything from pocket calculators to the International Space Station-that it's often easy to think that using PVs is the only way to get work from sunlight. There is, however, more to solar power than photovoltaics. Solar heat, for instance, has long been used in passive applications, such as heating water for domestic use.

Concentrate enough solar heat and you can do more than simply warm water. If sunlight is concentrated by a factor of dozens or even hundreds, materials at the focus can be heated to hundreds or even thousands of degrees, opening up all sorts of potential activities. Researchers in Switzerland, for instance, have used a giant mirror assembly to drive oxygen from zinc oxide to create pure metal, which could be used in making fuel. Others have suggested that sufficiently concentrated sunlight could crack water to create hydrogen. (The ideas are described in "Packaging Sunlight," published in the March 2004 Power & Energy, available at Mechanical Engineering Online, www.memagazine.org, under "Special Editions.")

Concentrated solar heat can also be used, directly or indirectly, to create steam to run through a turbine to generate electricity. Such concentrated solar power plants date back to a 70-horsepower pumping station established in Egypt by American engineer Frank Shuman in 1913.

With the growth of oil and natural gas as fuels in the early 20th century, however, solar thermal power became the province of dabblers and dreamers. That changed in the 1970s, when oil shocks and the early environmental movement spurred development of alternatives of fossil fuels. A 400 kW solar thermal proof-of-concept facility was built by Georgia Tech in 1977, and Solar 1, a 10 MW plant in Southern California consisting of a field of 1,800 mirrors concentrating sunlight on a point at the top of a tower, was running by 1981. Together with a few arrays of parabolic troughs, Solar 1 (and its successor, Solar 2) became emblematic of the attempt to tackle the Carter-era energy crisis by striking out in bold, new directions.

But after those first large-scale efforts in the Mojave Desert, solar thermal power faded from view. Part of that decline is due to a change in the energy climate: The cost of oil and other fuels began a decade-long drop beginning in the mid-1980s, which undercut much of the urgency to develop alternatives. Also, the Reagan administration was notably hostile to solar power in general, ripping down the solar panels that had been installed on the White House roof in addition to cutting the budget of federal research into alternative energy.

Some of the blame, however, rests on the performance of that first generation of solar power plants. The central tower-style facilities were bulky: Solar 1 and Solar 2, for instance, required 130 acres of mirror field to produce 10 MW. Each of the mirrors had to be independently steered to track the sun across the sky-no easy task. And the target tower was big and expensive; the central receiver for Solar 1 sat atop a 250-foot structure.

In any event, by the late 1990s, facilities such as Solar 1 were seen less as the wave of the future than as curiosities of a funky, discredited past. With oil selling for $10 a barrel and natural gas running around$2 per thousand cubic feet, investing in another giant, hugely expensive solar generating station would be the height of folly.

It would take another set of crises to make solar thermal seem again like something worth investing in.

LAS VEGAS IS A CITY BUILT ON SHEER DESIRE. The architecture along the famed strip of casinos is a brick and mortar testament to the idea that nothing is impossible if you want it bad enough and have the money to pay for it. Driving at a crawl along the boulevard, you'll see a scale-model Eiffel Tower, a faux Manhattan skyline, replica Venetian canals, and an Egyptian-style pyramid clad in black glass.

Even the throngs of tourists wandering through the 110-degree heat seem wildly improbable. How can people be so willing to fly or drive to the hottest corner of the continent in the middle of summer?

Drive south, away from the strip, and the fantasy falls away fast. Within a few miles, Las Vegas becomes just another Sunbelt metroplex, with big homes on small lots stretching away from the city. Some observers have claimed that these houses are every bit as much founded on fantasy-the belief in limitless availability of water in the midst of the desert, and in the unquestioned durability of the drive to gamble-as the casinos to the north. Indeed, at the outer edge of development are stands of half-built subdivisions and others that are little more than roads scratched into the dirt, victims of the real estate crash.

Then, everything just falls away and there is little more than highway and desert stretching out as far as the eye can see. Abandoned cars litter the side of the road and a 1,000-foot-high dust devil swirls menacingly near. And just at the limit of perception is a shimmering blue-ish patch at the far end of the valley, a patch that could easily be dismissed as a mirage except for one odd property: Its sides are straight and they meet at right angles.

Closer, much closer, the mirage resolves into a facility. More than 700 troughs made of parabolically curved mirrors stand supported by sturdy steel frames above a graded and raked dirt field. Each trough points toward the sun and running along the focus of the parabolic troughs are long pipes brilliantly lit by concentrated sunlight. If you stand near the facility on a scorchingly hot afternoon, the only sound is the low industrial hiss of hot fluid moving through the pipes and the slight rustle of a breeze.

The facility is Nevada Solar One, a $266 million power plant that's been running since 2007. Situated on 400 acres of desert flats in the El Dorado Valley north of Searchlight (the hometown of Sen. Reid), the plant can generate nearly 75 MW of power in full sun. Sunlight, concentrated some 70 times at the focal point, raises the mineral oil in the tubes to more than 700°F. The heat in the mineral oil then generates steam that powers a turbine. While Acciona, the Spanish company that owns the site, touts Nevada Solar One as the world s third-largest solar thermal plant, it isn't entirely clear how much power the facility's mirrors supply. A gas-fired heater contributes at least some of the energy needed to generate steam, and it isn't certain that the plant could operate without a contribution from fossil fuels. While Nevada Solar One is up and running, other solar thermal concepts are close to starting up. In the Mojave Desert, an array of large mirrored dishes is being erected to harness solar energy. Each dish, part of a project called SES Solar One, will focus light onto the heater head of a reciprocating Stirling cycle engine designed to generate up to 25 kW. When the first phase is complete, the facility will have some 20,000 of these dishes planted in the desert, generating as much as 500 MW of electricity. Oakland-based Bright Source Energy is revisiting solar tower technology. The company plans to build a field of mirrors to track the sun and concentrate sunlight on an elevated water tank. The 1,000-degree steam that is generated will run a turbine. The company, which has received funding from Google, set up a test Facility in the Negev Desert of Israel in June. IN THE EARLY 1990S, DAVID MILLS, AN AUSTRALIAN PHYSICIST, WAS INVESTIGATING DIFFERENT DESIGNS FOR CONCENTRATING SUNLIGHT FOR THERMAL POWER PLANTS. "The configuration I was working on had a lot of mirror area for the light collected," Mills said. The cost of a solar thermal plant is in many ways tied to the area of the mirrors: The more glass you need, the more expensive it is. Mills realized that there was more than one way to concentrate light onto a pipe. Most parabolic troughs use a continuous parabolic surface to bring sunlight to a sharp line focus relatively close to the mirror, which means the reflector has to be sharply curved. But if instead you used what is called a Fresnel mirror-essentially, an array of reflectors, each with a slightly different partial parabolic curve you could get a somewhat softer focus at significantly lower expense. What's more, if you place the point of focus away from the reflectors, Mills thought, the individual mirrors could have relatively shallow curves and be easier to handle. Placing the absorbing tubes some 40 feet up means they don't need to move in concert with the mirrors. That eliminates the need for moving joints in the pipe, which simplifies the operation and cuts the risk of leaks. The outlines of Mills's concept soon were in place. Large fields of Fresnel mirrors would reflect sunlight onto pipes containing water, heating it to make steam. The steam would be sent to a turbine to generate electricity, and the condensed water would be recycled back through the system. In many ways, Mills's basic idea was quite conservative: Except for the source of heat, an operation diagram of one of his solar thermal facilities would look remarkably like a conventional steam power plant. By the first part of this decade, Mills and his small company, Solar Heat and Power, had a demonstration plant up and running in New South Wales. In spite of a successful experiment, Mills found little support at home for further development of his concept. "The government supported coal and nuclear power," Mills said. "Programs that could get solar up to scale just weren't present. In fact, programs that had existed weren't operating any more." To tap into the American market, which was larger and more open to supporting solar power, Mills established Ausra and moved the company to California in 2006. Tapping into the pool of venture capital available in Silicon Valley, Ausra raised some$100 million. Some of that money was used to construct the factory in Las Vegas.

But to compete against conventional' generating stations, Ausra needed one more idea. One of the biggest stumbling blocks to adoption of renewably generated electricity has always been its intermittent nature, the old what do-you-do-when-the-wind-doesn't-blow question. It's a problem that's been handled many ways, from pumping water uphill in hydroelectric dams to storing electricity in giant batteries. The solution Ausra's engineers settled on was more direct-store some of the steam for later use.

According to John O'Donnell, until recently Ausra's executive vice president, "A Thermos stores as much energy as the battery in a laptop. But one of them costs five dollars and the other costs close to 10 times that."

Mills said that a solar thermal plant could save as much as two-thirds of its generated steam in storage tanks. That steam would be used during the evening and early morning hours to generate electricity. Instead of generating peak power during the afternoon and nothing after the sun goes down a plant would produce a smaller, but steadier level all day long.

That's the long-term idea, at least. Right now, Ausra is in the business of putting together bits of metal to do a specific job. One of the biggest differences between Ausra and other solar thermal start-ups is its emphasis on manufacturing its own components. To be sure, some elements will be purchased off the shelf-the turbines and generator sets, for instance-but DeGraaff and O'Donnell have emphasized that they want to have enormous control over the quality of the light-gathering material, and the only way to do that is to make it themselves.

"The other companies in this field don't publicize manufacturing because they don't really do any manufacturing," DeGraaff said. "We're operating our own manufacturing so that we can make a tightly integrated circuit-design, manufacture, construct, design, manufacture, construct. That's pretty new."

The 130,OOO-square-foot plant in Las Vegas is part of that plan. At full capacity, the line that was started in June could annually produce solar reflectors that could generate approximately 150 MW of electricity. The production lines are highly automated, with teams of robots welding together tubing and positioning the steel-backed mirrors. Most of the workers are new to this sort of work. One volunteered that he had been working at a dog food factory before signing on at Ausra.

The mirrors assembled there would be trucked to a field location. Unlike the central tower plants where each mirror has its own motors to move it in two axes, one motor moves a row of 10 mirrors at once, enabling them to track the sun. Forty feet above those mirrors, receiver pipes will be installed and connected to a steam turbine and generating set. A sunny field could be converted into a solar power plant in a matter of weeks.

But as futuristic as the idea of solar energy may seem, there's something reassuringly old-fashioned about the factory itself. Steel and glass are stacked along the walls. Forklifts hum over the poured concrete floor. Unlike so many places that make the bits and pieces of the "future"-biotech materials and silicon wafers and computer algorithms-everything was so solid and humanscale that a naive observer could understand how all the pieces were fitting together.

NOT FAR FROM LAS VEGAS IS HOOVER DAM. The iconic hydroelectric power station was considered a modern marvel when it was completed in 1936. The dam soars more than 70 stories above the floor of Black Canyon and impounds as much as 46 billion cubic yards of water in Lake Mead. The water-powered turbines in the base of the dam can generate as much as 2 gigawatts of power, though its average output is far less.

The power from Hoover Dam was instrumental in the development of Southern California after World War II and of Las Vegas in recent decades. But it is also a great point of comparison between the engineering feats of the past and the scale of solar development that Ausra and others have proposed in coming decades.

Although the actual structure of Hoover Dam has a relatively small footprint-660 feet wide at its base, and a quarter-mile long-the lake behind it covers nearly 250 square miles of desert. The surface area of the reservoir doesn't directly correlate with generating capacity, but if the same 250 square miles of desert were to be covered with a solar thermal generating facility of the sort Ausra is building, the output would be close to 20 GW. That's 10 times what Hoover Dam supplies.

In David Mills's vision, large chunks of the American Southwest would be devoted to solar thermal power plants. Since its water system is a closed loop, the Ausra design doesn't require a constant supply of water for cooling; such facilities could be set up virtually anywhere that has ample sunshine and a relatively flat surface.

A look at a solar radiation map of the United States shows that finding sunshine ought not to be a problem. While the eastern half of the country has too many partly cloudy days to make much use of solar thermal power, the Southwest is one of the best locations in the world for it.

"But most countries have enough solar resource to run their economies," Mills said. "That isn't the issue. But the question was whether you could implement a system at reasonable cost that could replace coal power-the capital equipment as well as the fuel."

Mills and his colleague Robert Morgan have recently published papers that have attempted to address this question. The big stumbling block for solar power has always been its diurnal nature: great on a sunny summer afternoon, missing in the dead of a winter night. But using electricity load figures from California and Texas, Mills and Morgan found that the peak loads corresponded with peak solar production. Combined with a system that could store solar energy for about 16 hours, such as Mills's steam tanks, solar thermal power could, in theory, run the U.S. electrical system most hours of the year.

"Because it tracks the day-night changes in electrical load," Mills said, "not only could solar thermal replace baseload generation, but it also could replace the much more expensive peaking component."

However, finding actual sites to put the power plants on won't be simple. For one thing, the desert Southwest is rugged. Most of the land slopes too steeply to accommodate a large field of coordinated mirrors. And of the remaining flat area, much of it is locked up in wildlife preserves or paved under the sprawl of Phoenix and Las Vegas. According to a recent study by the National Renewable Energy Laboratory for the Western Governors Association, only 12 percent of the land area in California, Nevada, Arizona, and New Mexico was suitable for solar thermal development.

That might be enough, though. That 12 percent works out to some 60,000 square miles, and even if not all of it was as intensely irradiated as Las Vegas, there are somewhere between 7,000 and 10,000 gigawatts of peak power available, or 2,000 to 4,000 gigawatts of base load, in just those four states.

"The bottom line is there is as much solar resource as you'd need," said Mills, "and the geographic requirements are actually very small."

The desert Southwest isn't the only part of the country interested in solar thermal power. O'Donnell said that Ausra has been approached by groups in Florida about the possibility of Ausra's building a plant there. Florida may be the Sunshine State, but it actually isn't as well suited to solar thermal as the Southwest. Partial cloud cover, common in Florida, robs solar thermal of a lot of its power. To make up for that, a solar thermal plant in Florida would have to be substantially larger than one in Arizona to produce the equivalent amount of electricity, meaning that the cost to grid would be higher. In some respects, it might be more efficient for a place like Florida to support a high-voltage dc transmission line and bring in solar power generated in Arizona.

"There's an economic development argument against that," O'Donnell said. Studies indicate that a solar plant would generate enough jobs in-state to make it worth the extra cost of electricity. "Even if the electricity is 40 percent more expensive, there's a motivation to build a plant in Florida."

For now, the company is in the process of building a 177 MW plant on the Carrizo Plain, near the coast of central California, and at a site near Bakersfield.

AFTER THE ASSEMBLED PRESS AND DIGNITARIES CLEAR OUT OF AUSRA'S NEW PLANT, THE FACTORY FLOORBEGINS TO RESEMBLE A WORKPLACE AGAIN. Workers in hard hats move pallets of material in advance of a day when the important metric is how many mirrors were made, not how many interviews were given.

Up close, the mirrors are impressive, but they also seem incredibly head-slappingly intuitive. From the side, the mirror glass looks pretty much like what you'd find if you pried apart your medicine cabinet. The corrugated steel plate the glass is glued to could be from the roof of a warehouse, the square cross-sectioned steel tubes from the rafters of a big box store. The materials are common, but when they are put together, the panels don't seem cheap.

"In some ways, it's like building boat trailers," DeGraaff said.

And that's the essential point: The panels are the product of a lot of high-end physics and engineering, with designs calibrated using finite element analysis and computational fluid dynamics. But the end result is a product that can be manufactured by the tens of thousands with non-skilled labor and robots, and assembled on site like so many giant Lego pieces.

To DeGraaff, who spent so much time refining the design and optimizing the manufacturing process, there's obvious pride in having simplified the idea of a field of giant solar thermal mirrors so much that it seems almost simple. But, he says, the real pride for him is elsewhere, larger.

Getting it right, DeGraaff said, involved "solving problems right in the core of mechanical engineering." In other words, mechanical engineers get to save the world .