# Micro-scale solar thermal power

Are there any practical methods of converting solar thermal energy into electrical power on very small scales, eg: using a 300mm diameter paraboloid reflector with temperatures of approximately 800-1600 °C?

While there are many large scale concentrated solar power plants which use conventional steam turbine engines, I am wondering what other methods there are to produce electrical power from solar thermal energy at high temperatures but using a small solar collector, eg: thermoelectric modules, etc.

• As an physics engineer, my intuition is that this does not work efficiently. Thermal efficiency requires high temperature differentials, which requires lots of insulation. Insulation scales poorly to smaller dimensions. I.e. if you scale down a 1000 ton tank of molten salt down to one ton, the amount of insulation needed does not scale down by a factor of 1000 either, as its thickness doesn't scale - just the surface, which decreases by a mere factor of 100. – MSalters Aug 14 '13 at 7:31
• There are practical methods but the amount of electrical power that can be produced by a small-scale solar device is very small, though still useful. We have solar powered watches after all. – RedGrittyBrick Aug 14 '13 at 9:06
• I have a solar powered Seiko, still ticks even in winter indoors. – Sunnyskyguy EE75 May 11 '17 at 23:03

No matter WHAT you use to generate your electricity, you are still limited by your collector area, your latitude, and the fact that Earth rotates with an approximate 24-hour period.

Assuming you're in North America, your fundamental limits are:

1. $1.3 kW/m^2$ at high noon on a severe clear day in Tucson AZ.
2. $6 \text{ to } 8 kWh/m^2/\text{day}$.

Any competent solar energy text will give you these numbers. Also note that these are DC-to-light numbers: the energy is spread out across the electromagnetic spectrum, not concentrated in nicely usable wavelengths.

Applying those numbers to your proposed collector, you have:

$$300 mm \text{ diameter} = 150 mm \text{ radius}\\ \pi \cdot 0.15m \cdot 0.15m = 0.070 m^2$$

That means your limits are $92 W$ at high noon on a severe clear day in Tucson, and 424 to 565 Watt-hours/day.

I haven't kept up with photovoltaic conversion efficiencies. Last I heard, 16% was STILL about all you could get with inexpensive materials and processing. I keep hearing someone is working on 30% efficient cells, and will have product Real Soon Now...

Here's an article talking about a solar-heated Stirling engine system. They cite 31% efficiency, compared with 16% for photovoltaics. They're putting some of their power into an az-el sun tracking system, and they're in Arizona.

• This is good information, but it doesn't answer the question, which is expressly about thermal solar power or other methods. How much power is available wasn't the question. – Phil Frost Aug 14 '13 at 12:09
• The point I was trying to make is that a 300 mm diameter collector is not going to gather much power, no matter what you use to convert from sunlight to juice. – John R. Strohm Aug 14 '13 at 12:52
• @PhilFrost: Thank you for the reformatting. Can you provide a pointer to a guide on how I can do that formatting myself? – John R. Strohm Aug 14 '13 at 12:53
• Just search for $\LaTeX$ guides. – Phil Frost Aug 14 '13 at 14:15
• I agree with Phil. Useful information John, you got my vote- but I'm still left wondering what other solar - electricity conversion methods there are besides PV, steam and thermoelectric modules? – MachuPichu Aug 15 '13 at 0:54

Peltiers are a great way to harvest thermal energy. However, 800C is a bit much for most peltiers. The temperature is determined by the amount of heat flowing into the peltier and the peltiers thermal resistance (and whatever thermal sink you have). So it might be you could spread out the heat and use more low thermal resistance peltiers. There are some newer companies that do make peltiers that can handle that kind of temperature.

The second method is using a stirling engine, but you'd probably have to build one, there are a few that you might be able to rig up to a solar conentrator.

The last option is high temperature photovoltaics which is probably outside the range of hobbyists.

Power efficiency is too low to get much useful power /\$ on thermal piles. But heating water is very efficient if well insulated.

I think you have enough temperature difference there to run a thermoelectric cooler backwards, too.

you could heat up a thermal battery (an instulated water tank for example) and use the thermal differential between the tank and the outside to move a Stirling engine, I'm not an expert but i think it might be an efficient way to go about doing it.

• Thanks for reviving this 4 yr old question. If conversion losses are too high from 1kW/m² and efficiency is only 15% then you are down to 15mW/cm² in bright sun. – Sunnyskyguy EE75 May 11 '17 at 23:11