electricity from heat

[Tomo]

I was just wondering if anyone can think of a reason why heat can't be turned directly into electricity. I know that there are devices (thermoelectric converters) that turn temperature differentials into heat but I was thinking it would be cool if there was a device that absorbed the heat (energy) from the surrounding air and turned the heat energy into electricity. Obviously it would be great if such a device was possible because if you put the device in a hot room (in summer), the device would both cool the room and generate electricity.

[Qfwfq]

Hello Tomo, welcome to Hypography.

You might find it interesting to look up the 2nd principle of thermodynamics and entropy, if you're not familiar with them already. Basically it's because thermal energy is a chaotic agitation. At a uniform temperature there's no chance of extracting non-chaotic energy. Although there's such a thing as negatemperature, with which it's possible to see veeeeeery slight entropy violations, nobody has shown a useful way of doing such a thing and I doubt it will happen.

[CraigD]

Originally Posted by Tomo

… I was thinking it would be cool if there was a device that absorbed the heat (energy) from the surrounding air and turned the heat energy into electricity. …

Such a thing is, in principle, possible, without violating any thermodynamic laws.

The best known though experimental proposal is Maxwell’s demon. This 140-year-old scheme is intentionally very vague – Maxwell appears to have meant it to provoke thought, not as a serious engineering proposal.

After several erroneous explanations of the impossibility of Maxwell’s demon had been put forth and commonly accepted, Szilard proposed an actual, in principle possible, implementation of a Maxwell’s demon, commonly called “Szilard’s engine”. Although some nice print illustrations of this engine exist, the best I could find after a brief search is http://home.att.net/~numericana/answ...on.htm#szilard.

What Szilard’s engine does is convert random molecular heat into orderly mechanical work (energy), but at the same time, create random information. Once whatever memory is used to store the information necessary for its operation has been filled up, the mechanical work necessary to reset (“erase”) it is exactly equal to the work it performed, if the engine is 100% efficient.

It’s hard to imaging any practical use for an actual, physically implemented Szilard engine – some very exotic sort of super-fast, brief duration microprocessor cooler, perhaps – but what it illustrates about the relationship between binary information, temperature, and physical work, is profound. Some credit Szilard’s 1929 paper as being a major influence on Shannon’s 1948 “A Mathematical Theory of Communication”, arguably one of the most important publications of the 20th century, and widely considered to be the foundation of modern information theory.

[Abstruce]

Originally Posted by Tomo

I was just wondering if anyone can think of a reason why heat can't be turned directly into electricity. I know that there are devices (thermoelectric converters) that turn temperature differentials into heat but I was thinking it would be cool if there was a device that absorbed the heat (energy) from the surrounding air and turned the heat energy into electricity. Obviously it would be great if such a device was possible because if you put the device in a hot room (in summer), the device would both cool the room and generate electricity.

The most efficient way to convert heat to energy is to use a refrigerant like R-134. This gas will absorb the heat energy so you can extract the energy at a concentrated point by were you can drive a steam turbine of some type.

Because most refrigerants boil at hundreds of degrees below 0 you could assemble a collector to absorb heat from the atmosphere even in the winter in Alaska without the sun and turn this into useful energy.

[Qfwfq]

Maxwell's demon illustrates the fact that the 2nd of thermodynamics is a statistical principle. There is also an old anecdote about irreversibility. During a historic discussion of the topic, one physicist said: "Actually, there's no such thing as irreversibility, in principle. After the two gasses have mixed, all we'd have to do would be to reverse the motion of each molecule exactly enough and the whole sequence of collisions would backtrack to the separate gasses." Another one said: "All right, very fine, now YOU reverse the motions exactly enough!"

Originally Posted by Abstruce

Because most refrigerants boil at hundreds of degrees below 0 you could assemble a collector to absorb heat from the atmosphere even in the winter in Alaska without the sun and turn this into useful energy.

But not more than the energy necessary to re-liquify the refrigerant.

[Abstruce]

Originally Posted by Qfwfq

Maxwell's demon illustrates the fact that the 2nd of thermodynamics is a statistical principle. There is also an old anecdote about irreversibility. During a historic discussion of the topic, one physicist said: "Actually, there's no such thing as irreversibility, in principle. After the two gasses have mixed, all we'd have to do would be to reverse the motion of each molecule exactly enough and the whole sequence of collisions would backtrack to the separate gasses." Another one said: "All right, very fine, now YOU reverse the motions exactly enough!"

But not more than the energy necessary to re-liquify the refrigerant.

There are experiments that have proven that enough energy can be extracted to provide an over unity of the processes. Yet all energy is accounted for in the form of thermal energy and there must be an energy source.

[Kayra]

Any chance this is what he means?

http://www.powerchips.gi/faq.shtml

[Abstruce]

Originally Posted by Kayra

Any chance this is what he means?

http://www.powerchips.gi/faq.shtml

You can patent anything even things that do not work.

What you have is theory without being applied.

So yes it works in theory, it has not been developed into a working prototype for production.

There are other versions of this theory that have been produced and soon will hit the mass market.

[CraigD]

Originally Posted by Kayra

Any chance this is what he means?

http://www.powerchips.gi/faq.shtml

That’s not the impression I got.

Powerchips, are just ordinary heat engines – claiming high (80%) efficiency, but not challenging any established Physics. Tomo seems aware of the fundamentals of heat engines, and is proposing something that can get work from heat with no temperature difference.

Unless he rejoins this thread, I see no way to be sure.

[Tomo]

Thanks for the responses. I have a few queries and things to clear up.

I had a similar idea to Abstruce's about using refrigerants. The thing I don't get is how do you make "a collector to absorb heat from the atmosphere", wouldn't the collector just re-emit the heat into the atmosphere.
I was thinking that one could put some water in a sealed container and then pump out enough air so that the boiling point of the water is around room temperature (or similarly use a refrigerant which boils at room temp. if such a refrigerant exists). Then what would happen (maybe) is that the water at the bottom of the container would evaporate, rising up the container to turn a turbine, condense at the top of the container and fall back down to complete the cycle. I think that the water would condense at the top of the container even if the top and bottom of the container are at the same temp. The reason is that as the water vapour rises, it gains potential energy and thus loses kinetic energy (i.e. the temperature of the water vapour drops) allowing the water vapour to condense and fall back down.
Another thought that I had along the same lines is that if the container widens towards the top then the temperature of the water vapour would drop as it expands, making it easier to condensate.

I had a look at the website that Kayra referred to. A powerchip is not what I was thinking of as it needs a temperature gradient to work (similar to thermoelectric converters).
What I am ultimately asking is if electrical energy can be produced from the heat in a building, thus cooling the building and generating electricity. If, however, there is a natural temperature gradient in a building then it would be interesting to find out if powerchips and thermoelectric converters could be used to generate electricity. The only natural temperature gradient I can think of is the ceiling being hotter than the floor (because hot air rises and the roof of the building gets heated up by the sun). However, I imagine that this temperature difference is too small to be any use to a thermoelectric converter or a powerchip.
In addition, I don't think that thermoelectric converters would solve my intial query as I don't think that they actually cool down the surroundings. They generate electricity from the transfer of heat from a hot source to a cold source.

Just an additional question, is there any reason why a photovoltaic cell can't be tweaked to absorb Infrared Radiation (heat) instead of visible light, thus solving my initial query.