Technology Review - 29 Feb 08

Prof. Robert B. Laughlin
Department of Physics
Stanford University, Stanford, CA 94305

http://www.technologyreview.com/biztech/20356
(Copied 17 Dec 09)


Solar Without the Panels

Utilities are using the sun's heat to boil water for steam turbines.

By Peter Fairley
Friday, February 29, 2008

Grid-feeding trough: In a solar-thermal trough plant, hundreds of mirror arrays like this one track the sun from east to west, concentrating sunlight onto the receiver pipe suspended above them and heating the oil within to nearly 400 C. The captured heat can be used to produce steam and generate electricity, or it can heat large tanks filled with molten salt to store solar energy for a rainy day. - Credit: Abengoa Solar

Investors and utilities intent on building solar power plants are increasingly turning to solar thermal power, a comparatively low-tech alternative to photovoltaic panels that convert sunlight directly into electricity. This month, in the latest in a string of recent deals, Spanish solar-plant developer Abengoa Solar and Phoenix-based utility Arizona Public Service announced a 280-megawatt solar thermal project in Arizona. By contrast, the world's largest installations of photovoltaics generate only 20 megawatts of power.

In a solar thermal plant, mirrors concentrate sunlight onto some type of fluid that is used, in turn, to boil water for a steam turbine. Over the past year, developers of solar thermal technology such as Abengoa, Ausra, and Solel Solar Systems have picked up tens of millions of dollars in financing and power contracts from major utilities such as Pacific Gas and Electric and Florida Power and Light. By 2013, projects in development in just the United States and Spain promise to add just under 6,000 megawatts of solar thermal power generation to the barely 100 megawatts installed worldwide last year, says Cambridge, MA, consultancy Emerging Energy Research.

The appeal of solar thermal power is twofold. It is relatively low cost at a large scale: an economic analysis released last month by Severin Borenstein, director of the University of California's Energy Institute, notes that solar thermal power will become cost competitive with other forms of power generation decades before photovoltaics will, even if greenhouse-gas emissions are not taxed aggressively.

Solar thermal developers also say that their power is more valuable than that provided by wind, currently the fastest-growing form of renewable energy. According to the U.S. Department of Energy, wind power costs about 8 cents per kilowatt, while solar thermal power costs 13 to 17 cents. But power from wind farms fluctuates with every gust and lull; solar thermal plants, on the other hand, capture solar energy as heat, which is much easier to store than electricity. Utilities can dispatch this stored solar energy when they need it--whether or not the sun happens to be shining. "That's going to be worth a lot of money," says Terry Murphy, president and chief executive officer of SolarReserve, a Santa Monica, CA, developer of solar thermal technology. "People are coming to realize that power shifting and 'dispatchability' are key to the utility's requirements to try to balance their system."

In fact, the capacity to store energy is critical to the economics of the solar thermal plant. Without storage, a solar thermal plant would need a turbine large enough to handle peak steam production, when the sun is brightest, but which would otherwise be underutilized. Stored heat means that a plant can use a smaller, cheaper steam turbine that can be kept running steadily for more hours of the day. While adding storage would substantially increase the cost of the energy produced by a photovoltaic array or wind farm, it actually reduces the cost per kilowatt of the energy produced by solar thermal plants.

The amount of storage included in a plant--expressed as the number of hours that it can keep the turbine running full tilt--will vary according to capital costs and the needs of a given utility. "There is an optimal point that could be three hours of storage or six hours of storage, where the cents per kilowatt- hour is the lowest," says Fred Morse, senior advisor for U.S. operations with Abengoa Solar. Morse says that the company's 280-megawatt plant in Arizona, set to begin operation by 2011, will have six hours of storage, while other recent projects promise seven to eight.

Economies of scale: Spanish solar-power-plant developer Abengoa Solar plans to build and begin operating this 280-megawatt solar thermal power plant in Gila Bend, AZ, by 2011. The plant's rows of mirrors, thermal storage tanks, and power-generating turbines will cover nearly three square miles. Phoenix-based utility Arizona Public Service will buy the power--enough to supply 70,000 Arizona homes. - Credit: Abengoa Solar

Morse says that while the design of solar thermal power stations is rapidly diversifying, most will use essentially the same system for storing energy: tanks full of a molten salt that remains liquid at temperatures exceeding 565 C. "It's basically two tanks with a lot of heat exchangers, pipes, and pumps," says Morse. For a sense of scale, consider that the 50-megawatt plants that Germany's Solar Millennium is building in Spain near Granada will employ 28,500 tons of molten salt in twin tanks standing 14 meters high and 38.5 meters in diameter.

While molten salt is the most popular storage option, developers are experimenting widely to find the best means of collecting heat in the first place, and integrating collection and storage. Abengoa's plant in Arizona (see below image) will use a "trough" design in which arrays of parabolic mirrors concentrate sunlight onto a glass tube carrying a commercial heat-transfer oil such as therminol. Some of the heated oil heats the molten salt in storage while the rest directly generates steam. Abengoa Solar's vice president for technology development, Hank Price, says that the plant's trough energy-collection design is the one most commonly used today, thanks largely to improvements in the glass tubes. Ceramic-metal absorption coatings have increased the amount of heat captured by the tubes to the point that plants using them produce 30 percent more power than the first-generation solar thermal demonstration projects of the early 1990s.

SolarReserve, in contrast, is developing systems that directly heat molten salt. Its designs call for so-called power towers in which arrays of mirrors focus sunlight onto elevated towers. The company, launched in January, is a joint venture between energy investment bank U.S. Renewables Group and aerospace firm Hamilton Sundstrand, whose subsidiary Rocketdyne built molten-salt heat receivers for a 10-megawatt power-tower demo plant that operated in the early 1990s.

SolarReserve's Murphy says that the power-tower system should be cheaper to build than trough-collection systems, since it doesn't require miles of glass tubing. More important, he says, it should produce higher-quality steam. That's because it will directly heat its molten salt to about 565 C, about 165 degrees hotter than the oils in a trough plant.

That increased thermodynamic efficiency will be key, says Murphy, when water shortages force thermal power plants in hot, dry deserts to abandon water-based cooling of their used steam. (Steam that's passed through the turbine must be cooled and condensed so that it can be reused.) Alternative cooling techniques are more energy intensive, cutting into a plant's overall efficiency. The hotter a plant runs, says Murphy, the lower the losses from alternative cooling schemes. "We're going to experience 3 to 4 percent loss," he says, "and [the trough plants] are going to be losing 7 to 8 percent."

Abengoa's Price agrees that power towers do, in theory, have thermodynamic advantages, which is why Abengoa has built its own 10-megawatt demo in Spain and is building a second at 20 megawatts. But Price questions whether investors will support the direct jump to 100-to-200-megawatt power-tower plants that SolarReserve envisions. "There's a lot of technical risk in doing that," he says. "We need to scale up in a way that's financeable."