> Take into account the existence of night, cloudy days, and the fact you're probably not at the equator: divide that by four. That's still somewhat optimistic

> Take into account the existence of night, cloudy days, and the fact you're probably not at the equator: divide that by four. That's still somewhat optimistic; about 10--20% of peak is more realistic, based on what we've seen with existing solar installations. It's closer to 20% in sunny places, and closer to 10% in places with less sun. However, that still leaves us with a respectable amount of electricity. Good for them. Of course, the panels themselves are not the whole cost. There's other electrical equipment needed to make that energy usable, for example, and someone has to install and maintain the solar panels. I think the next big source of money in this market, after they get the cost of manufacturing the panels down far enough, is streamlining installation and maintenance. Perhaps someone could make durable "solar shingles" that are relatively easy to put on a house, or some such thing. Longer-term, solar photovoltaics look like they could provide a pretty hefty amount of energy cheaply, if we solve a lot of engineering issues. Where they fall down is producing power reliably, close to where it's needed. One way of dealing with this is to have the solar panels in space, but obviously that's pretty tricky. Another way of dealing with it is to use something else to provide the base load of your electrical grid -- nuclear could handle this with cheap green aplomb -- and use the solar power for things which can easily soak up cheap intermittent power, such as charging the batteries of electric or plug-in hybrid cars. Theoretically, it could also be used to power interesting energy-hungry chemical processes, like ammonia synthesis or aluminum refining. Making this financially viable is another big engineering challenge, but definitely a worthwhile one if it works out.