DARPA Awards $53 Million for Solar Power Research

mygadgetbox writes "Defense Advanced Research Projects Agency (DARPA) will be giving a consortium led by the University of Delaware nearly $53 million in funding to more than double the efficiency of terrestrial solar cells within the next 50 months. DARPA wants the consortium to develop and produce 1,000 Very High Efficiency Solar Cell (VHESC) prototypes that are affordable and that operate at efficiencies of at least 50 percent. The goal is to create solar cells that operate at about 54 percent efficiency in the laboratory and 50 percent in production."

Comments

TATFA (think about the article)

[PresidentEnder]

Most of what Parent says is true- for the applications in question (home use, vehicle use, consumer use in general) the cost of solar cells is the limiting factor. However, this is DARPA we're talking about here- as well as a bunch of commercial clients. The applications mentioned in the article (primarily the military) rely heavily upon efficiency and not so heavily upon cost efficiency. Yes, it would be nice if GI Joe's GPS solar cell only cost fifty cents, but if it weighs a hundred pounds? I'd rather have a five-hundred dollar solar cell producing enough in half a pound. Efficiency is important in some areas, cost effectiveness in others. Research into efficiency isn't a total waste.

Solar Power is not just Photovoltaics

[necro81]

The article discusses that the goal is to improve the efficiency of solar cells to 50%. As I mention earlier in this thread, silicon-based semiconductor photovoltaics top out at a theoretical efficiency of about 25-26%. Other semiconductor technologies top out somewhere around 35%. These are the two technologies people think about when they imagine solar cells. I think the outlook for discovering and commercializing a semiconductor-based solar cell that's 50% efficient in the next 50 months to be very poor. I won't get into the physics, but the theoretical limitations have to do with the fact that semiconductor photovoltaics make inefficient use of the solar spectrum: a red photon will produce as much electrical energy as a blue photon, even though the blue photon is more energetic.

But solar power is not limited merely to what one can do with photovoltaics. When people talk about the many terawatts of solar power that falls on the surface of the earth, most of that solar goes into two things: photochemistry (like in plants) or to heating the earth's surface. Plants make very efficient use of the solar power that falls on them, and a black, nonreflective object will convert the incident solar power to heat (or reradiated infrared light) with extremely high efficiency. If we could focus efforts to developing technologies that capture sunlight first into chemistry or raw heat and converting that to electricity, rather than the direct conversion to electricity that photovoltaics do, we may have a better chance of reaching the 50% goal.

For instance, there was (is?) a solar power project that in the California desert that was a solar-thermal generator. Hundreds of mirrors focused sunlight onto a tower, much like the Archimedes death ray (which has received some press in /. lately). Instead of trying to burn a ship, the focused sunlight heated sodium to about 1200 Celcius, which liquified it. That sodium was passed through a heat exchanger to boil water, which made steam, which turned a turbine, in a similar closed-cycle technology to a nuclear plant.

I'll admit this isn't much use in the battlefield, which is what DARPA is aiming for, but it is not out of the question to consider a smaller solar thermal unit for an encampment, which used a different medium than sodium.