How to make water out of air?

I would like to know how to make a machine that is solar powered and turns air into water.
There are devices that create 30 Litres of water in 24 hours. Even though the device that I would like to make is powered by solar it should still provide water during the day provided that there is suffecient humidity in the air.
Please help!

Louvan (author) 1 year ago
Thanks atmoswater for the info.

The reason that I am so interested in changing air into water is because of the quality of the drinking water is terrible. A person can actually taste that there is something suspect with the water we receive from the muncipality.
So by using a dehumidifier, filtering the water through multiple stages i.e ultraviolet light, filter straws ect would this water be good enough to drink?

Hi Louis,

Yes, by proper use of water filters and UV sterilizer the system will be capable of producing water meeting WHO drinking water quality guidelines. For long-term use of the water it is best to have a water sample tested upon first installation and annually thereafter. Regular cleaning of the system (at least every 30 days) should be done. Filters and the UV lamp have to be replaced according to the manufacturers' instructions.


Louvan (author)  atmoswater1 year ago
Hi Roland,

Thanks for the info!
So the ideal situation would be to power a dehumidifier via a solar battery bank setup (using Lead Crystal Batteries) to ensure device runs 24/7. A switching mechanism would be used to switch between batteries. Is there a water filtration system that has a built in UV light water filter available commercially?

Thanks once again
Hi Louis,

An example of the type of filter + UV system that would be suitable is shown at


atmoswater1 year ago
Here is some information to better understand the water-from-air resource in Johannesburg. The first image shows the resource during Sept 25, 2012 (AM). The second shows the resource during Sept 25 (PM).

The water vapour density (grams per cubic metre) is the water vapour resource. It is fairly constant during any given day (depending on characteristics of the air mass over your region) even though the temperature and relative humidity vary markedly. The dew-point tracks the water vapour density and represents the water vapour content of the air.

On Sept 25, the water vapour density never rose above 7.6 grams per cubic metre. A mechanical dehumidifier, with the defrost control maintaining the evaporator wet coil temperature at about 5 °C, will have leaving air with a temperature of 5 °C and 100% relative humidity. This combination of temperature and relative humidity corresponds to a water vapour density of 6.8 grams per cubic metre (pyschrometric calculation). Water production is only possible when the water vapour density of the entering air is greater than 6.8 grams per cubic metre. My chart shows the maximum possible mechanical dehumidification efficiency (details at my blog post, So Sept 25 was not a good day for collecting water from the air in Johannesburg. But, better days are coming with the change of seasons! The third image is shows the water-from-air resource on a monthly basis. During Nov-Mar water-from-air production will be "fair" with the water vapour density ranging from 9.2 to 10.4 grams per cubic metre.
Johannesburg WFAR hourly am.JPGJohannesburg WFAR hourly pm.JPGJohannesburg WFAR monthly.JPG
Kiteman1 year ago
If you're asking us for a [design of] machine to extract water from the air, we need to know where you are - local levels of humidity etc.

Do you get regular fog?
Louvan (author)  Kiteman1 year ago
Hi Kiteman,

I am based in Johannesburg, South Africa. The humidity is 71 %.
I don't want to burn hydrogen, I am simply looking for a D.I.Y unit that sucks in air, condenses it and then makes it usable for drinking water. I know that the water would have to be filtered but there are drinking straws that filter water. The unit would have to be powered by a solar panel, perhaps 75 watts or so. I understand that most of the water would be captured during the evening and early morning but there should be enough humidity during the day to capture water using solar power ONLY!

Thank you so much for the interest.
Kiteman Louvan1 year ago
I came across a thing on the island of Tenerife - fog farms.

They consist of a fine-mesh* fence placed across the path of humid, fog-generating winds. Water condensed out on the mesh, and trickled down to water crops. No power required.

If you placed one on a hill-side, and fitted a guttering or trough beneath the fence, water would run down it to a vessel or reservoir.

The more fence you install, the more water you collect.

If you were determined to add power to it, you might think about a solar-powered peltier unit, cooling metal fins so that they collect even more condensation?

*The mesh looked like several layers of fishing net, piled loosely
Louvan (author)  Kiteman1 year ago
Thank you so much for all the info, it certainly helps.

I wonder what is the most effecient peltier? The reason why a peltier device would be suitable is due to the space requirements. If say a 200 watt solar panel powers two petiers and two 12V fans I wonder if this would create enough water to drip from the peltier modules into a holding tank?
Any ideas?

May you all be blessed!
Kiteman Louvan1 year ago
Sorry, reached the limits of my knowledge here, but I think you have enough information to do some useful research.
rickharris1 year ago
In the garden dig a shallow hole

In the middle put a cup

Over the hole place a plastic sheet

In the middle of the sheet place a small stone so that the sheet forms a slight depression over the cup

Leave over night and in the morning th cup will have water in it.

Why do you want to know this. if you have the time and energy in most places you can dig a well faster then "making water from air"

PS you don't make it from air you extract it from the air.

I guess there are systems around that do this called dehumidifiers

Moisture condenses on cold surfaces

Think about how to make the surface cold and your problem is solved.
Louvan (author)  rickharris1 year ago

Thanks for the advice.
I do not want to dig holes. Perhaps a peltier cooler would work?
if there is moisture in the air anything that is cool will cause it to condense

So you a peltier cooler will work. - Most air moisture is concentrated at night because the air temp drops. Even in deserts they get fog and dew.

You will need power perhaps a solar charged battery system.
I noticed that you wanted to make this thing solar powered.

Another trick that might work is using a desiccant
of the kind that can be regenerated by heat. 

Then you use a solar concentrator to sort of "boil off" the water trapped in the desiccant.  I am guessing you could do this at a temperature in the range of maybe 100 to 300C.   

Then you feed the water vapor to a condenser that is being cooled by ambient temperature air.  So the condenser temperature is a little hotter than ambient temperature, but I'm hoping that would work just because the relative humidity surrounding the heated desiccant is greater than 100%, i.e. it's hot and steamy.

Anyway, I think a setup like that might make more efficient use of your sunlight. 

More about desiccants:
A commonly used trick for extracting water from the air is to put that air in contact with a cold surface it can condense on.   The temperature of this surface has to be lower than the dew point for water,

Moreover, as the water condenses it gives up heat to the surface it is condensing on, at a ratio of  2257 joules of heat per gram of water condensed,  and that number is just the so called enthalpy of vaporization of water.

That heat has to be removed from the condensing surface somehow. (If it is not, then the surface will heat up to a temperature above the dew point.)

So how much heat must be moved, in order to condense 30 kg of water? 

Well, the answer to that question is:

Q = (30 kg)*(2257 J/g) = 67 710 000 J =  67.71 MJ of heat to be moved

That's kind of a lot of heat.

For the sake of comparison, you might ask how much heat your little 75W thermoelectric module, with a coefficient of performance (COP) of  maybe 0.6, how much heat could it move in a day?  Answer:

(75 J/s)*(24*60*60s)*(0.6) = 3 888 000 J = 3.8 MJ

And that's enough to condense about 1.7 kg of water in a day, which is a little smaller than your goal of 30 kg per day.

Sort of the most obvious improvement, I think, if you want to use an electrically powered refrigerator ( a heat pump) to do this, is to use a decently efficient refrigerator.  If I can believe this statement:

"Thermoelectric junctions are generally only around 5–10% as efficient as the ideal refrigerator (Carnot cycle), compared with 40–60% achieved by conventional compression cycle systems (reverse Rankine systems using compression/expansion)."

from the Wikipedia article on Thermoelectric cooling,

Then you can improve your heat pumping by a factor of about 6, just by using a refrigerator with a compressor.
framistan1 year ago
I dont think you have to dig a hole for RICKHARRIS's idea to work. Water evaporates from the soil. You just need an enclosed space over the ground, with a plastic sheet covering it. Water will condensate on the underside of the plastic and drip into the cup in the middle. No power required, except maybe sunshine.
mrmerino1 year ago
Or like, literally anything flammable?
mrmerino1 year ago
Burn Hydrogen?