Solar powered Sterling engine project, from Sandia

"The Stirling engine makes solar power so much more efficiently than photovoltaic solar cells can," said Robert Liden, chief administrative officer at Stirling Energy Systems Inc. (Phoenix). "That's because the Stirling solar dish directly converts solar heat into mechanical energy, which turns an ac electrical generator." The bottom line, he said, "is that large farms of Stirling solar dishes — say, 20,000-dish farms — could deliver cheap solar electricity that rivals what we pay for electricity today."

Under a multiyear Energy Department contract that started in 2004, Stirling Energy Systems will supply Sandia National Laboratories with solar dishes for integration into full-fledged power-generation substations capable of direct connections to the existing U.S. power grid. Right now about 20 EEs, including more than a dozen from Stirling Energy Systems, are working full time at Sandia to create the electrical-control systems to manage these sunshine stations.

By the end of 2005, they plan to have six dishes connected into a miniature power station capable of supplying enough 480-volt three-phase electricity to power about 40 homes (150 kW). The next step, in 2006, is a 40-dish power plant that will transform the combined output of the farm from 480 to 13,000 V, for distribution of industrial-level power to an existing substation. From 2007 to 2010, the program proposes mass-producing dishes to create a 20,000-dish farm supplying 230,000 V of long-haul power from its own substation directly connected to the grid.

If the project succeeds, the DOE predicts that by 2011, Stirling solar-dish farms could be delivering electricity to the grid at costs comparable to traditional electricity sources, thereby reducing the U.S. need for foreign sources of fossil fuels.

Power today costs from about 3 cents to 12 cents per kilowatt-hour, depending upon the customer's location and the time of day. The average is 6.6 cents/kW-hr for the industrial sector in 2004, according to DOE. In contrast, the Stirling solar-powered substations operate only during peak hours (daytime) but at potentially the same or less than the peak rates paid today — or "about 6.5 cents per kilowatt-hour during peak periods," said Liden of Stirling Energy Systems.

(there's another link at the bottom of this story, which is the full article. For some reason, the real link isn't working with the DU posting software)
http://www.gizmodo.com/gadgets/gadgets/solar-powered-stirling-generators-026232.php

.

I really have to get around to building the stirling powered roasting spit. (and the low temprature differential one to run off the heat of my monitor...)

They seem to be regular carnot type (heat) engines. I heard somewhere that they operated on spacecraft. Anyone know about that.

To produce the amount of power claimed here would require some huge dishes. In North America, the energy from full sunlight at noon is ~1.3 kilowatts per square metre, meaning that a 100% efficient conversion would require that each of the six dishes would have to have a surface area of 19 square metres- that would mean a diameter of 2.5 metres each. Since the best thermal power generators are about 40% efficient that takes the diameter up to 6.2 metres. I sincerely doubt that any material would absorb anywhere near 100% of the solar energy and convert it to heat, any plan that involves dishes less than 10m in diameter would be suspect. The bottom line is that 1.3kw/m2 is the absolute upper limit for any solar power project.

this does mean that the system is unworkable at all, only that there is - as is often the case in solar energy claims - some exaggeration about costs.

The economics of the system on the climate where it is used and the temperature available heat reservoirs (which will impact, of course the Carnot efficiency) and the costs for the construction of the apparatus. Even at 25% efficiency, however, there are still around 300 Watts per square meter. The roof of a 250 square meter (around 2200 square foot) ranch house would theoretically (depending on geometry) be able to put out around 80 kilowatts of energy at mid-day. This would be significant output.

Certainly though, the system is not workable everywhere.

But your point is well taken. I like how you think. You do the math.

I followed a few links on this product and it does look like a promising way to recover waste heat. According to the numbers there, at higher temperatures the engine approachs 50% efficiency which is better than internal combustion or turbines. The downside mentioned that it is sluggish to respond but for power generation that's a non-issue.

The other thing mentioned was that although they can run on just about any temperature differential, on anything less than 260 degrees Celsius the engine would be too large (in relation to the power delivered) to be practical. So for solar you're talking about a pretty big installation for a small return. OTOH this would be great for any industry that uses lots of energy and throws away a lot of heat- canneries and steel mills come to mind.

temperatures of up to 550C.

http://reslab.com.au/resfiles/hightemp/text.html

This solar plant in California operates at 350 C inlet temperature:

http://www.solarmillennium.de/pdf/TechDescripParabTrough.pdf

The California plant has a rated capacity of 105MWe, which is comparable to 1/5 the output of the Glen Canyon Dam when the thing has water in it, and 1/10 the output of a nuclear power plant running flat out.

I don't the plant's cost, but once it is amortized, I expect that it will run very cheaply indeed.

a big-screen TV. It is one of my prized posessions, and it will pass on to my descendents, if I believe they are worthy. With this lense, I can take a square meter of incoming sunlight, and focus it down to about 1/2 a square inch. I'm estimating, since any ruler I placed on the focal point would either burst into flame or melt. And I could't look at it anyway, since I'd damage my retinas.

With this lense, I have melted (among other things) glass, concrete, aluminum, copper and steel.

So, it ought to be possible to get temperatures even higher than 550C, although I don't know if it's practical on an industrial scale.

It's damnably fun, though. I recommend it.

than 550C, depending on the steel type:


http://hcrosscompany.com/metals/stainless.htm

The melting point of copper is 1084C.

The melting point of aluminum is 660C.

http://www.webelements.com/

From this same source we see that the heat capacity of copper is 24.4 J/(K-mol). Thus if you had one mole of copper (63.5 grams) theoretically at midday (assuming constancy of Cp, and P = 1300W/m^2) you'd be able to heat the copper over 50 degrees per second.

Impressive. I'll bet it's enormous fun.

As for the industrial applicability, I note that the plant described above is operating. 105 MW is a small, but respectable power plant.