The Other Solar

Anybody who has burned leaves with a magnifying glass understands the idea behind concentrating solar thermal power. But the technology is more than just child’s play. For many companies, including the chemical firms that sell the heat-transfer fluids that are the lifeblood of the concentrating solar plant, it’s becoming a serious business.

Concentrating solar power (CSP) couldn’t be more different from its better known cousin, photovoltaic power, which takes advantage of the photoelectric effect. When photons hit cells made of polycrystalline silicon or other materials, they set electrons into motion, electrifying a circuit.

CSP works under a much more primitive principle: using the sun’s rays to heat something up. Under the dominant form of CSP technology today, parabolic mirrors focus sunlight on a glass tube. Heat-transfer fluids running through a steel pipe inside the glass tube warm up to nearly 400 °C and heat water to produce steam for a conventional turbine generator.

The two technologies are used in different ways. The emphasis of photovoltaic power is generating electricity at the end-use location. CSP, especially parabolic trough technology, is meant to be operated at an industrial scale.

The heat-transfer fluids of choice for parabolic troughs are blends of diphenyl oxide and biphenyl supplied by Dow Chemical and Solutia. These molecules have a long history. Dow has been making diphenyl oxide commercially since 1924. Biphenyl was chlorinated to make polychlorinated biphenyls (PCBs), coolant and insulating liquids that were banned about 30 years ago.

Because they are stable molecules that remain liquid under low pressures from ambient temperatures up to about 400 °C, diphenyl oxide and biphenyl lend themselves to CSP, according to Ravi Prakash, market development manager for specialty fluids at Solutia. “There are few other fluids out there that can do the same things,” he says.

After a 20-year hiatus, CSP is coming back into vogue, with new plants starting up in the southwestern U.S., Spain, and North Africa. But the temperature limits of diphenyl oxide-biphenyl blends are a barrier to making CSP more efficient, and scientists are looking for alternatives that can perform at higher temperatures.

At a December alternative-energy conference organized by the Chemical Development & Marketing Association and held at the University of Pennsylvania, Gilbert E. Cohen, president of the Morrisville, N.C.-based consultancy Eliasol Energy, profiled CSP’s history. In 1912, the U.S. inventor Frank Shuman built the world’s first parabolic trough plant to power an irrigation system in Egypt. But the technology didn’t catch on. Although Shuman’s concentrator design was not much different from today’s, “the apparition of ‘quasi-inexhaustible’ fossil fuels came out to be a more attractive and cheaper fuel supply,” Cohen told conference attendees.

In the wake of the late-1970s energy crisis, the Israeli firm Luz International collaborated with the U.S. Department of Energy to build nine parabolic trough plants in California’s Mojave Desert from 1985 through 1991. Still in operation today, they have a combined capacity of 354 MW. The first of these plants is based on a mineral oil heat-transfer fluid. The rest use diphenyl oxide and biphenyl supplied by Solutia’s predecessor, Monsanto.

It wasn’t until 2007 that another large-scale CSP plant was completed in the U.S. Spain’s Acciona Energy built the 64-MW Nevada Solar One parabolic trough complex at a cost of $266 million. Spain itself has been another hotbed for CSP development. The most ambitious project in the country was initiated by the German firm Solar Millennium and consists of three plants, named Andasol 1 to 3, each with a capacity of 50 MW. The first one opened in 2008. Unlike the Mojave Desert units, which use natural gas to supplement solar production, the Andasol plants include tanks of molten nitrate salts to store energy so they can run at night.

These plants might be just the first few flakes of a coming blizzard. In California alone, the Federal Bureau of Land Management has received right-of-way requests for 34 CSP plants, largely parabolic trough setups with more than 50 MW of capacity apiece. In Spain, more than 30 trough plants are planned. According to a report compiled by the market research firms CSP Today and Altran, about 680 MW of CSP capacity is operating today around the world; capacity for another 2 GW is under construction.

For Dow and Solutia, business in diphenyl oxide-biphenyl blends for CSP has been brisk. Dow fluids are used in six Spanish projects with a total of 300 MW of capacity, including 2,000 metric tons of fluid that Dow delivered to each of two of the Andasol plants. It also delivered 1,500 metric tons to fill Nevada Solar One.

Dow won another eight contracts for 2010, according to Christoph Lang, a technical service and development specialist for the firm’s heat-transfer fluids business. Among the various industrial sectors that the business serves, CSP has the greatest growth potential, he notes.

Solutia fluids are used in a 50-MW plant in Spain and in natural gas/solar hybrid plants in Algeria, Egypt, and Morocco with capacities of 170, 150, and 470 MW, respectively. “This is an exciting market,” says Richard Altice, vice president of commercial services at Solutia’s technical specialties business. “And our belief is that it will have more staying power than the initial installations in the 1980s. Climate change and energy security are different types of drivers than we saw then.”

Diphenyl oxide-biphenyl blends may be a good fit with CSP, but developers of the technology hope someday to have fluids that offer even better performance. Greg Glatzmaier is a senior engineer overseeing heat-transfer-fluid and thermal-storage research at the National Renewable Energy Laboratory (NREL) in Golden, Colo. He points out that at temperatures above 390 °C, biphenyl polymerizes, whereas diphenyl oxide undergoes a reaction that forms a bond between two phenyl groups of the same molecule.

Both reactions give off hydrogen, which dissipates through the metal pipe and into the glass tube, disturbing a vacuum that prevents heat loss by conduction. Because of this problem, the operating temperature of CSP plants that use diphenyl oxide-biphenyl blends must be kept below 400 °C.

The temperature limitation restricts the efficiency of converting heat into electricity at the plant’s steam turbine to about 38%, Glatzmaier says. Unencumbered by diphenyl oxide and biphenyl, CSP plants could in principle run at up to 500 °C. “If you can go higher, that conversion temperature goes up,” he notes. In addition, hotter heat-transfer fluids would raise the temperature of the molten salt storage medium, increasing the amount of energy the salts can store.

Better heat-transfer fluids would improve the overall economics of CSP plants. Right now, parabolic trough CSP incurs a cost of about 18 cents per KWh. NREL’s goal for the industry is to reduce that to 12 cents, bringing CSP closer to the 7 to 9 cents per KWh that natural gas power costs to generate.

Earlier in the decade, scientists investigated ionic liquids as alternative heat-transfer fluids, but they also ran into thermal stability barriers at higher temperatures, Glatzmaier says. Another alternative is molten salts. They can run at temperatures as high as 600 °C, but their freezing point of 120 °C or more poses a problem in plants that have miles of pipe, Glatzmaier points out.

Scientists can reduce the freezing point of salts by mixing different ones together. For example, Robert W. Bradshaw at Sandia National Laboratories has used calcium and lithium nitrate to decrease the melting point of sodium and potassium nitrate mixtures.

Last April, DOE awarded Sunnyvale, Calif.-based Symyx Technologies $1.5 million for a three-year project to come up with CSP heat-transfer fluids with an operating temperature range of 80 to 500 °C. However, at the end of the year, the company elected not to continue the project.

Symyx used high-throughput experimentation to test 5,000 different salt combinations, says Justin Raade, who was the principal investigator for the program. Although he won’t say how close to the target melting point the research group came, he notes, “We did make progress.”

Dow has been working to find heat-transfer fluids that can tolerate higher heat. Lang won’t reveal more details, although he does note that the company markets Syltherm 800, a polydimethylsiloxane fluid made by Dow Corning that doesn’t produce hydrogen at high temperatures. It also has a lower freezing point than diphenyl oxide-biphenyl blends’ 12 °C, which might make it suitable for colder climates.

However, Syltherm 800 doesn’t have the heat capacity or the density of the organic fluids. And it has yet to be used in a CSP installation.

Manufacturers of diphenyl oxide-biphenyl blends aren’t standing still, either. According to Solutia’s Altice, research shows that impurities exacerbate the instability of the mixture at high temperatures. Improving product purity could improve the fluid’s already good performance. “The aromatic ring structure is one of the most thermally stable structures that are out there,” he says.

Dow’s Lang agrees, noting that CSP plants will continue to use diphenyl oxide-biphenyl for the foreseeable future. “After more than 80 years, there is still no other organic fluid that we have found with a higher thermal stability,” he says. “It is really suitable for this solar application.”