A Gigantic Nuclear Furnace

During a discussion of my most recent post, a friend (I can't remember which one to credit, I'm sorry brilliant friends) suggested that solar power might be a reasonable alternative to biodiesel for running the Nigerian telecom stations. I agreed – it almost seemed too logical. I should have thought of it myself! Please note, MTN: my friends are brilliant!

A few years ago it seemed that solar power had gone to purgatory along with the electric car. Since the Pope abolished purgatory in December of 2005, however, it’s back! So what now?

Today, 95% of the world’s solar power is generated by flat plate silicon photovoltaics. These are what you occasionally see on tree-hugger’s roofs and your calculator. They just straight-up absorb sunlight over the whole surface of the cell and convert it to electricity. It’s all we've really had for quite awhile, since incentives for improving on them didn't really exist until that whole global warming thing came up. For a few reasons, they’re really not the best.

First, producing energy this way is expensive - estimated to cost between $0.22 and $0.44 per kilowatt hour (kWh). Natural gas, by contrast, costs $0.042 per kilowatt hour and subsidy-free wind is $0.05 per kWh. This is the wholesale price, though, and it’s not really a good comparison. Since solar energy is often made and used in the same spot, its cost ought to be compared to retail electricity prices. In 2005, the average retail price for electricity was $0.095 per kWh, for commercial it was $0.087 and industrial uses cost $0.057. Obviously not bridging the gap very much, there. It is especially interesting to note how inexpensive electricity currently is for industry, where the larger-scale solar set-ups I’ll discuss below might make the most sense.

What’s up with the cost? Flat photovoltaics are very material - intensive. Silicon is the active ingredient in the solar generation chemistry equation. And silicon is expensive. Computer manufacturers court it for semiconductor manufacturing and are willing to buy it lobster and steak (er, pay more for it) since they use a lot less, in a marginal sense. Since the entire surface of flat solar panels is designed to absorb and convert solar energy, a lot of silicon is needed. There is a lot of raw silicon-making material out there. But silicon refining capacity is short and it’s going to take awhile to catch up. Another issue with traditional solar cells is their efficiency - they convert on average only about 15% of the sun energy they absorb into electricity.

As my work in government has taught me, you can't come up with much savings by cutting out things that don't cost that much. So reducing or eliminating silicon from photovoltaic production has to be the way to go, though weight also must be given to the enemies you will make among silicon lobbyists. Is it worth it?

Gads of companies with visions of various currency symbols dancing in their heads must think so. They have scads of engineers challenging the silicon-heavy status quo. The Energy Foundation, a nonprofit grant-making and research organization, estimates that the U.S. solar energy sector alone has the potential to grow by $6.0 billion per year if the cost can be brought in line with other energy sources.

Let's check out one option - concentrator photovoltaics. They need just a little button of silicon because they are so brilliant. According to Parc Research, they can match the energy output of flats with only 1/1000th the amount of silicon. Mirrored lenses focus sunlight onto the little silicon spot, concentrating the energy intensity. A whole bunch of them are put together to make a functional receptor, like this one.


Rotation-ability is important so the lenses align with the sun correctly all day long.

Isn't that amazing? If you would like to steep your brain in yet more wonderment, just think: concentrator photovoltaics are approaching 30 to 40% efficiency. That's about double current levels. If each cell costs lots less due to reduced silicon use, yet can produce twice as much electricity from the same amount of sunshine...are we approaching some sort of market competitiveness? Parc puts the cost at half that of flat cell generation - so we're down to $0.11 to $0.22 per kWh: nearly competitive with residential costs, though the logistics don’t really work for most residential settings. Still, I do have a great idea for that old satellite dish on my uncle’s acreage…

Obviously, the cost is really not competitive with commercial or industrial rates in the U.S. Is it competitive with rates in Europe, Asia, Africa, Antarctica, South America or Australia? Um.

Let’s just look at Nigeria for now. If you read “Bio-wondering” you know that a lot of electricity there is produced privately. A December 2002 World Bank report put the average cost of private production at $0.06 per kWh and public production at $0.159. Nigeria’s government might explore this technology for advantages in cost, continuity of service and provision to remote areas. Private entities at the high end of provision cost ($0.208 per kWh) might also check this out. Of course, the technology costs will probably be different in Nigeria than they are in the U.S., but how they would be different is difficult to say.

A case study: a project being built in Australia has utilized a $1 million government grant to install concentrator photovoltaic “dishes” in an area where 2,500 people are spread across 100,000 square kilometers. Similar to Nigeria, power in this area is currently generated with diesel fuel. The government estimates a savings of 160,000 liters, (about 42,000 gallons) of diesel per year. But electricity in Australia is cheap (less that $0.04 per kWh). They have a ton of “clean” coal and an enormously flexible energy market, where electricity is bought and sold by the minute depending on who’s producing it most cheaply. However, supplying remote areas is clearly an obstacle for the Australians as well. And a full million dollars was required to make solar energy happen for them. That’s an almost unaffordable leap for private industry to make, especially when compared to burning biodiesel with not so much as a modification to the same old diesel generators.

The new solar technology looks promising – concentrator photovoltaics aren’t all that’s out there – but the costs still appear prohibitive, even in the places it could be most beneficial. It’s unfortunate, but the cost of fossil fuel consumption is still too low, even with a global warming premium fast approaching, to make direct capture of the sun’s energy economical. But this is only the beginning. If I didn’t believe in engineers’ ability to go beyond what anyone thought was possible in the 1980s, I wouldn’t have married one.