Solar is more efficient than coal

Metaefficient reports that a new world record has been set for solar-to-grid efficiency. Some new kind of photovoltaics? No. Instead they used an old-school stirling engine.

If you're not familiar with this type of engine, I can tell you they're cool, and you should read a little about them. One of the reasons that I want to buy a metal lathe is so I can build my own sterling engines...

The operation of photovoltaics may be a mystery to you, as they are to me. Still, you might be wondering how a mechanical device proves more efficient than electrons dancing around at the molecular scale (or whatever they do).

The trick is conversion losses. While PVs have reached light-to-DC conversion efficiencies exceeding 40%, the overall system's efficiency is killed by the final stage: inverting the DC to the AC, as required for our power distribution grid. As is true in all engineering contexts, when considering green technologies, you must analyze the whole system. Remember to always look at efficiency: the amount to get out as a fraction of what you put in.

That's where stirling engines come in. A parabolic mirror collimates the incoming light onto the hot-side of the sterling engine; the cold-side is at ambient temperature. Since the stirling engine produces mechanical rotation, the DC/AC inversion is unnecessary. Skipping this step, it turns an AC generator, yielding an overall efficiency of 31.25%. By my estimates, this makes solar more efficient than coal.

I'll remind you that 31.25% efficiency means we let 68.75% of the total energy fizzle away as heat. For comparison:

1. A traditional coal power plant has efficiencies in the mid-30%, but newer coal plants can get efficiencies above 60% (energy stored in coal in, electrical energy out) [renewable energy access].
2. Grid transmission and distribution losses of 6--8% are considered normal (electrical energy in, electrical energy out) [renewable energy access].
3. Photosynthesis has an efficiency of 90% (usable light in, ADP and NADPH out). However, only 43% of the total solar incident radiation can be used, and so the overall efficiency is about 39% (sunlight in, ADP and NADPH out) [wikipedia].
   
4. Aerobic repiration has an efficiency of 40.4%, anaerobic is 29.1% (sugar in, ATP out) [wikipedia].
5. A computer's power supply is typically about 70--75% efficient, though higher efficiencies are available (electrical energy in, electrical energy out) [wikipedia].
6. An incandescent light bulb is about 2% efficient, and a consumer-grade compact fluorescent bulb is about 7--8% efficient. Present-day LED lamps are about 30% efficient (electrical energy in, light out) [wikipedia].
7. Driving a car takes 1,860 [kilo]calories per passenger-mile; walking takes 100 [kilo]calories per passenger-mile; cycling a mile takes 35 [kilo]calories per passenger-mile (energy in, change in position of a mass out) [world watch]. If we ignore everything except the effect--moving one passenger one mile--we can estimate the relative efficiencies: a car is less than 1/50 as efficient, and walking is about 1/3 as efficient as riding a bike.
   

So why does coal seem more efficient than solar? Because the prior statistic doesn't analyze the whole system. Coal is simply a vector for energy. The efficiency of a coal power plant is measured in terms of (energy stored in coal in, electrical energy out). To compare it to solar power, we must consider the process by which sunlight was converted into coal (via trees and a few millenia of geothermal activity), not to mention the energy expended in mining and transporting coal. We already know that photosynthesis has an efficiency of 39%, so a (generous) upper-limit on the solar-to-coal conversion is 39%.

Factor this in, and you will find that, solar is more energy efficient than coal. It just looks different because we have a lot of coal laying around.

Thanks metaefficient; you are my new favorite blog.

And thanks to the commentor, for pointing out my spelling mistakes.