First, we need a membrane material which is highly permeable with respect to water and water vapor, but as impermeable as possible with respect to all other components of air. This is not an impossible requirement, as such materials do already exist.

Alternate two layers of this membrane with two layers of spacer, and roll it up into a cylinder, such that we have a very large surface area in a small volume of space. Mechanically, this will be very similar to the core of a reverse osmosis water purifier.

Pass air from inside the building across one side of the membrane (at atmospheric pressure) while maintaining a partial vacuum on the other side of the membrane.

Water vapor will diffuse across the membrane from the side with the higher partial pressure to the side with the lower partial pressure -- in this case, from the air into the partial vacuum.

The partial vacuum is maintained by a compressor, which continuously transfers water vapor into a condenser. Heat is removed from the condenser by passing the dehumidified air around it. Condensate is continuously removed from the condenser by means of a condensate pump, and transferred either into a container or down a drain.

After the dehumidified air has passed through the condenser, it's returned to the inside of the building.

Ideally, there would be no air inside the condenser to interfere with condensation of water vapor, but in practice, a small vacuum pump will be need to discard the small amount of noncondensable gasses which will inevitably diffuse through the membrane. Most likely, the vacuum pump would be operated intermittently, activated either by an oxygen sensor, or by pressure and temperature sensors. If oxygen is present in the steam, or if the temperature is below water's boiling point for that pressure, then the vacuum pump must be activated.

To provide even greater comfort for the occupants of the house, the condenser could be cooled using air from outside the building, much like an air conditioner; in this case, the condensate could be sprayed onto the outside of the condenser, so that it's evaporation would help cool the condenser.

As a further improvement, we can improve both efficiency and elegance by replacing the condenser with a second membrane air handler, and blow outside air across this second membrane.

As long as the pressure of the pure steam inside the second membrane is higher than the partial pressure of water vapor outside the membrane, water molecules will diffuse through the membrane from inside to outside... even though the steam's total pressure will be much less than atmospheric pressure. In a small way, this is similar to evaporation, but without a phase change.
-- goldbb, Jun 21 2010

Another use of the type of membrane used. http://www.freepate...ne.com/4725359.html
[goldbb, Jun 21 2010]

As an additional consideration, if we use either of the two versions where humidity is moved into the air outside of the building, then we can combine the dehumidifier with an energy efficient evaporative cooler.

And if we're using the version with the atmosphere-cooled condenser, then some or all of the condensate could be used to supply that evaporative cooler... resulting in a system that can probably be considered a vapor compression refrigeration cycle.
-- goldbb, Jun 21 2010


I was thinking that you could dehumidify the air with your vacuum by lowering its pressure. And you could, but that would be a lot more work. I like this. Here is my question: permeable to water but not air I buy, but water vapor is not much like water in that it is gassy, not droplety. Do such membranes treat both states of water equally?

Linked is an oxygen concentrator: similar approach of changing constituents of air. These work because of an absorptive mineral specific for nitrogen.
-- bungston, Jun 21 2010


bungston, how were you thinking of using a vacuum to remove humidity?

As for a membrane which passes water vapor but not air, if you read the link I posted, you'll see that it's a patent for a device which uses precisely such a membrane in order to accomplish it's task. In fact, that patent is the inspiration for this halfbaked idea -- there is not a huge difference, except for what I'm using it for.

I don't see your oxygen concentrator link -- but don't most oxygen concentrators use pressure swing absorption? Hmm... on the other hand, the third link that came up when I searched for "oxygen and nitrogen molecules" was a page describing why oxygen escapes from car tires faster than nitrogen -- O2 molecules are smaller.

That would suggest that the same principle as I use for humidity removal *could* be used to efficiently concentrate oxygen, if we had two molecular sieves that each allowed O2 through but debarred N2. A compressor would move O2 from inside one sieve to inside the other.
-- goldbb, Jun 21 2010



random, halfbakery