PV is used in both wholesale and retail power generation. The former involves large-scale arrays, where an electric utility places thousands of panels in a sunny place and ships the resulting electricity over high-capacity power lines to users. Worldwide, over a gigawatt of power was coming from such plants in early 2008, 80% of which were in Europe, where government policies are friendliest to renewable energy. But at the power-plant level, today's PV is far more expensive than conventional sources like coal. So while PV farms exist and new ones are being built, they only generate an adequate return for their owners when paired with massive subsidies and/or mandates requiring that utilities derive a portion of their power from renewable sources. Large-scale PV will not replace coal in the foreseeable future.
The real excitement is at the retail end of the market, where PV has a whole host of advantages. Unlike wind, coal, or natural gas, PV scales down. Three solar panels on a rooftop are just as efficient as a thousand in the desert. This kind of "distributed" power generation doesn't require any new real estate, since the buildings are already there. It doesn't require environmental impact assessments, since a roof is not an ecosystem. Distributed power makes the grid more robust, since it's much harder to disrupt a system that is fed by a million homes than one in which a single plant's power is transported over long-haul lines. Because daylight coincides with peak power demand, rooftop solar helps utilities avoid building new, expensive gas- and coal-fired plants. And-this is suddenly a big one-PV locks in a given electricity price for two or three decades, something no utility burning coal, gas, or oil can promise.
Because rooftop solar panels compete with the retail price (i.e., the cost that utilities charge their customers for electricity after shipping it long distances from large power plants) rather than the wholesale price of electricity, PV's cost disadvantage is modest in most markets. Meanwhile, utility power costs are rising along with fossil fuel prices, wages, and carbon taxes. These two trend lines-conventional power up, solar down-are intersecting in more and more places. In early 2008, when tax breaks and other incentives were included, solar was competitive in Japan, much of California, and several European countries, meaning that an installed solar system would pay for itself in 10 years or less. The next stage is "grid parity," where solar becomes cheap enough to compete head to head with the retail price of grid-delivered coal or natural gas-derived electricity. No new breakthroughs are needed; all it will take is another 5 or 10 years (depending on the market) of the same steady progress that's currently taking place. Even in cloudy Britain, solar should reach grid parity by 2020.