The World in 2050_ Four Forces Shaping Civilization's Northern Future - Laurence C. Smith [38]
Theoretically,156 PV cells can convert sunlight to electricity with efficiencies as high as 31%, but most are considerably lower, around 10%-20%. If that sounds pathetic to you, then consider that the efficiency of plant photosynthesis, after three billion years of evolution, is just 1%. Nonetheless, a typical silicon-based solar photovoltaic panel, with 10% efficiency and a manufacturing cost of around three hundred dollars per square meter, produces electricity that costs around thirty-five cents per kilowatt-hour. That’s seven to seventeen times greater than coal-fired electricity. So sunlight, despite being far and away the world’s biggest energy source, is also the most expensive.
Finding a cheaper way to hijack sunlight is thus the single greatest barrier to the widespread use of solar power. Most photovoltaic panels are made of sliced wafers of extremely pure silicon that are highly polished, fitted with electrical contacts, sealed into a module, and encased in transparent glass. They are heavy, cumbersome, and expensive to make, and become even more costly when the price of silicon goes up. As ardent renewable-energy enthusiast Chris Goodall points out, installing large solar panels on the roof of his Oxford home costs about £12,000, yet the total market value of the electricity they produce after four years is just £300. While it makes sense for governments to subsidize such investments initially, eventually the technology must become competitive with fossil fuels in order to take hold.
That means the cost of PVs must fall to about one-fifth of what they are today, a huge challenge. It’s a materials-science problem and there is much exciting research under way, particularly in the area of “thin-film” photovoltaics that abandon heavy silicon panels in favor of exotic coatings of semiconductors like cadmium telluride, or even carbon nanotubes.157 The conversion efficiencies of these materials would probably be lower than that of traditional silicon PV cells (8%-12%), but if they could be manufactured cheaply—even printed as shrink-wrap for buildings, for example—the cost of PV electricity would tumble and we could start enshrouding the planet in electricity-making paints and films.
At the moment, photovoltaic paint lies in the sweat-soaked dreams of nanotech graduate students. A safer bet for 2050 lies in the expansion of so-called concentrated solar thermal power, or CSP, technology. Like wind power it has been around for years, and is already providing economically viable electricity from a handful of pilot installations. Unlike photovoltaics, CSP does not attempt to convert sunlight into electrons directly. Instead, in much the way that kids fry ants with a magnifying glass, CSP relies on mirrors or lenses to focus the Sun’s rays, heating a fluid like water, mineral oil, or molten salt inside a metal tube or tank. The fluid boils or expands, forcing a mechanical turbine or Stirling engine to move, making electricity. Sound familiar? It’s just plain old-fashioned electricity generation158 driven by a new source. And because CSP plants work best on hot, sunny days—a time when millions of air conditioners drive up the price of electricity—their product commands top dollar. Unlike photovoltaics, CSP requires no silicon wafers, cadmium telluride, or other fancy semiconductors, just a great many polished mirrors, the motorized steel racks to mount them on, and a traditional power plant.
To make the most sense, CSP plants should be located in deserts. Current operations include several in Spain and the U.S. states of California, Nevada, and Arizona. Seventy miles southwest of Phoenix a billion-dollar project is under way to spread mirrors across