I think I've found Maxwell's Demon, however the demon is quite large...

[broli]

It might be that air, combined with heat sinks of large surfaces  would work better.
Also, it is probably not necessary to make the tanks have a large volume, the only thing that really affects the process is how high you make them, and how well shielded it is from the outside temperature.

Naboo

Luckily high performance heat shielding are easily available, namely radiant barriers. Two layers of a radiant barrier and some additional black pvc sheet can give a near 100% shielding rate. This project is quite easy to build. But I still don't fully understand it.

[Nabo00o]

Thank you for your response Broli, but I think I should set a fine balance between economy and efficiency  Of course it doesn't have to be that expensive, and also, the less surface to cover the cheaper.

The principal is based on something we find just as much in nature as in human-made environments.

Have you ever been in a lake and dipped your feet's in the water? Well then you may have noticed that the water gets colder the further down you get, in addition it is almost always occurring in really sharp layers of temperature, although it isn't completely necessary in this invention.
The same thing happens in valleys where there are no big openings and when there is no wind.
As a gas, air has a very low mass-density and is also easy to compress, both with temperature and with force. Since it is easily compressible, a small change in temperature will quickly change the density, and in a fluid, anything with a lower mass density will always float when affected by gravity.

This is what I use in the tanks, and by insulating them from the outside temperature, they can create their own "equilibrium" by allowing all that 'sinks to sink', and all that 'floats to float'. The result is that the total heat of the medium is split up into two potentials but without any input of energy, just like Maxwell postulated, and he couldn't be disproved either.


The next step is to allow the ambient temperature outside to be absorbed into the bottom, since it is hotter, it "wants" to travel there. When enough heat has been absorbed and the bottom temperature is the same as outside, then the top will be hotter because gravity will separate the potentials.

The last step is to make this higher temperature enter the cold part of the next tank. Since this fluid is much hotter than the bottom of the next tank, it will transfer its heat there if allowed ( a heat pipe will make this process very fast and efficient).

By connecting several tanks in this manner a stable equilibrium will eventually be reached where the bottom of the first tank will hold the outside temperature, while the top of the last tank will hold some kind of value higher than the beginning. What this value could be depends on a lot of variables, though it would be interesting to map the most important ones 

Naboo

[broli]

Below are the ingredients:

1) large diameter pvc pipes + caps
2) heat exchanger coils
3) radiant barrier
4) plastic insulation
5) optional: black white plastic sheeting used in grow rooms.
6) lots of duct tape
7) a thermometer with a long probe that you can stick in the last tube

The expensive part will probably be the heat exchanger coils unless you can make them yourself from copper tubes. One roll of any insulation can insulate MANY tubes.

[Nabo00o]

Great set of pictures there, but you forgot the heat pipe, and I have a feeling that it will be the most expensive part.  Btw, the coils could be replaced by anything which conducts heat well, we got lots of metal scrap lying around here, some stainless too...

Also, in what kind of business do you think they sell large pvc pipes? Probably not in your local plumbing store  Hmm.....

[broli]

120mm diamater and maybe 2m length for each pipe should be enough no?