Extracting Ambient Thermal Energy Without Gradients

[Markanthony]

It seems possible to extract energy from the ambient air without temperature gradients (or at least with very small local gradients that will always be available). Being curious about this, I checked out some other threads and did not really find the specific question. (I did find some interesting posts about Maxwell's Demon, etc.)

Anyway, this is the concept:

The ambient thermal energy created by the Sun, the Earth's core, gravity, etc., are in abundance and there are decent ways of extracting it. According to the posts, however, it appears that thermodynamic theory suggest that extraction of thermal energy requires a temperature gradient. This seems counterintuitive. It seems that the molecules are flying around waiting to be put to work and constantly being re-energized by the aforementioned external sources. Statistically, it makes sense that one would not get too far trying to do anything useful since they are likely to exert the same pressure in all directions on any object. However, couldn't one imagine a very thin pizoelectric device sticking out of a box and being randomly strained in various random directions by the air molecules? Wouldn't this solve the statistical issue and thereby "get around" the thermodynamic theory?

 

[PeterDonis]

It seems that the molecules are flying around waiting to be put to work and constantly being re-energized by the aforementioned external sources.

Yes, and what does the "being re-energized" by the external sources do? It heats up the air, thus creating a temperature gradient. If you have a bunch of air inside a box, isolated from all external heat sources, and it's all at the same temperature, there's no way to extract any energy from it.

couldn't one imagine a very thin pizoelectric device sticking out of a box and being randomly strained in various random directions by the air molecules?

How would you get useful work out of such a device? Remember Newton's third law: if the air molecules are pushing on the device, the device is pushing back on the air molecules. So there will be no net energy transfer.

 

[Alec Dacyczyn]

...couldn't one imagine a very thin pizoelectric device sticking out of a box and being randomly strained in various random directions by the air molecules? Wouldn't this solve the statistical issue and thereby "get around" the thermodynamic theory?

How would you get useful work out of such a device? Remember Newton's third law: if the air molecules are pushing on the device, the device is pushing back on the air molecules. So there will be no net energy transfer.

I think he's suggesting making the sticking-out-device small enough that it would be buffeted by only one air molecule at a time. Of course, the stick would also have its own thermal vibration. Sometimes the encounters would transfer energy to the stick; and sometimes the stick would whack a molecule and transfer energy away. But if the devices could convert the kinetic energy of the impact into electrical energy and conduct it outside the system then the device would "whack the molecules back" with less energy. Over time the total energy in the box (ie: the temperature) would decrease. I've had a similar brain-wave with a box lined with billions of independent microscopic piezoelectric tiles with tiny rectifiers and a common DC power bus. If you could build such small vibration energy-harvesting devices then you could probably make them work without the gas molecules, just using their own thermal vibration.

The standard rebuttal seems to be, "But that would reduce entropy. And that's not allowed." I find the usual arguments against this sort of thing unsatisfying. I've also read Feynman explanation with the hypothetical molecular ratchet mechanism and his assertion that the ratchet could 'slip' just as easily as work correctly and allow the energy accumulation to reverse. I'm unconvinced. I feel like I'm hearing someone tell me that photovoltaic cells can't work because the electron would inevitably 'slip' back down the bandgap range and re-emit the photon.

I've never been able to shake my itching suspicion that thermoelectrically without temperature deferential is somehow possible.Maxwell's Demon rides again!

 

[mfb]

The random fluctuations would be of the order of 40 millielectronvolts of energy (the same energy scale as the molecules at room temperature), and in random directions with the voltage. There is nothing that could rectify electrical fluctuations that small, in the same way there is nothing that can "order" the flight direction of molecules.

Photovoltaic cells work with visible light emitted by the much hotter sun - about 1 to 3 electronvolts per photon. That is sufficient to extract the energy in a useful way. They do not work with thermal radiation for the same reason as above.

 

[Dale]

With that we will close the thread. We do not discuss perpetual motion machines on PF.