Solar Technology

The January 2, 2008 Wall Street Journal has an article which, unfortunately, isn't visible except to subscribers, about United Technologies' Hamilton Sundstrand division. They are building a type of solar power plant that is a bit different. Part of it is unsurprising--the use of an array of mirrors to concentrate heat on a boiler filled with salt. This is similar to the Solar One project built near Barstow back several decades ago, which failed to meet its design goals because of the amount of power the plant consumed.

One of the great difficulties with solar power is that it is highly synchronized with sunlight. It doesn't help at night--or even much past 5:00 PM. Peak demand for power in areas where solar power makes sense is often in late afternoon and early evening, because of air conditioning demand, and use of dishwashers, washing machines, and other domestic electrical appliances.

I am reminded of one of the 1960s attempts to produce and sell "appropriate technology" solar ovens in India. It didn't work, because solar ovens only really worked in the middle of the day. (A friend of mine and I, along with our fathers, built a solar oven as a demonstration project in 6th grade. It worked great at cooking hot dogs and hamburgers--until about 2:30 PM, at which point it became quite disappointing.) As the song goes, "Only mad dogs and Englishmen go out in the noonday sun." Indian women, not being in either category, weren't too keen on using such an oven, and in addition, the main meal of the day in rural India wasn't lunch, but dinner.

What makes this new design by Hamilton Sundstrand interesting is that it is using the power not necessarily to directly produce electric power (which would limit it to morning to late afternoon), but to heat up the salt until molten, then pump the molten salt into insulated containers. The molten salt will stay hot for a very long time--long enough to boil water to turn generators during the night. This might have some merit, in large industrial applications. I don't see that it has much potential on a small scale, such as domestic power production. And I do not find it implausible that the same problem that made Solar One not work--that there's enormous energy inputs in making the components--might be hidden by the government subsidy involved.

The other problem of solar power being sun synchronized is not just day vs. night, but winter vs. summer. The farther you get from the tropics, the more severe the problem becomes. Not only do days get much shorter in winter, the farther north you go, but the sun angle becomes a problem. Here at 44 degrees north latitude, the maximum altitude of the sun above the horizon is 69.5 degrees--and that's at the summer solstice (when the sun is directly above the Tropic of Cancer). At the winter solstice, the sun is directly above the Tropic of Capricorn, or 47 degrees farther south--and then the highest position the sun can get for us is 22.5 degrees above the horizon.

You can store heat in molten salt over night, or even for a couple of days (which might solve the problem of cloudy days). But I shudder to think of the amount of heat that you would need to store to get you through the winter. This kind of system, however, does make sense for peak load power production.

UPDATE: A couple of readers pointed out that if you can solve the energy storage problem, it would be beneficial not only for solar power, but any energy production method that is not completely at your control, such as wind or hydropower. The difficulty is that batteries are not perfectly efficient at accepting a charge, they don't hold a charge indefinitely, and they don't last forever.

Another strategy that I have seen proposed is to pump water to an higher elevation. If the water is stored in a covered container, there shouldn't be any evaporative losses, and because the stored energy is gravity pulling the water to a lower level, the energy so stored will last for decades, if need be. But pumping water against gravity isn't terribly efficient, and turning it back into electricity later isn't terribly efficient. There is also a big capital investment in creating such a water storage mechanism, especially if it is covered. Of course, if you don't have access to natural elevation changes, building an artificial slope is extremely expensive.

One reader points out that perhaps too much energy is being spent on the electric car problem when the real problem that needs to be solved is: why are people living so far from work? If everyone lived a couple of miles from work (or telecommuted most days, like I do), it wouldn't so much matter what fuel we used for our cars. Of course, that would require social changes probably more dramatic than the technical challenges of electric cars.

UPDATE 2: Another reader points to this article on Wikipedia that claims that pumped water storage schemes can return 70%-85% of the original input energy, depending on evaporation of the body of water.