Air Powered Water Pump Prototype

An automatic water provision system is one of the foundations of a Global Village infrastructure. Yesterday we got one step closer to this with a successful demonstration of an air-powered water pump that Guy suggested. We will install this pump in our well, which we dug last year and which still has a manual pump handle. This will add an automatic sub-surface water source to our existing rainwater collection system, which we pressurize in a water pressure tank for distribution to the solar shower, sink, and dish washing station by the garden. Here are the results:

The good part is that the materials for the pump cost $30 (a bit lower than $600 off-shelf versions), not counting the air compressor or automatic timing. The next step is to add an automatic timer, connected to a solenoid, to turn the air pressure from the compressor on and off at the required intervals. We estimate that the duty cycle may be two seconds on followed by 6 seconds off. The video shows about 2 and 2 second cycles, respectively. We would appreciate any help on electronic timer and solenoids of choice.

Build one for yourself in a couple of hours. Some instructions are found on the OSE wiki.


  1. Inga

    We got water – we got life!! Well done you fabulous guys!

  2. mimarob

    One application for solenoids of this kind is in the special effects buissiness (smoke machines).

    The original machines are usually fitted with a small compressor that they switch on and off electrically, but for larger sets they like to have a big compressor away from the stage (for noise reduction) and then a solenoid valve.

    I know a guy in the field, I’ll ask when I see him!

    If you want a quick-and-dirty solution maybe try one of those small 12-V compressors used for filling car tires and then a relay with a timer circuit (2×555?) to turn it on and off? Might give you a to long duty-cycle and bad performance though.

    Another Idea I had was to hack a cheap air blow gun, perhaps just clamp it in a vice and have a strong servo move the lever. Biggest problem might be to adapt the air output back to an incoming connector.

  3. Rasmus

    Great job – reminds me of Engadget where they do a “teardown” of any new electronic gadget that comes to market.

    I wanted to suggest an extension. Rather than rely purely on electric power or diesel engine for compression, how about bypassing these precious energy forms altogether ? Keep in mind that using WIND to make compressed air is highly energy-efficient. So perhaps it is possible to design this in a way that allows a wind-driven air compressor to be plugged onto the pump.

    (side note: this principle of wind-to-compressed-air is under evaluation for larger-scale wind power: )

  4. Rasmus

    Could you guys comment on the energetic aspects of things ? Energy-wise, how does it compare to a purely electric pump ?

    Incidentally, this concept may have some advantages for the pumping of seawater. Since there are no metal parts that are exposed to the corrosive effects of salty ocean water, such a pump could be relatively maintenance-free (topic: “seawater agriculture”).

  5. mimarob


    Here is some stuff that can be bought from Farnell, there are different models that will operate at either 24,110 or 220 V

  6. mimarob

    Here is the same link in english this time, price seems to be about $65

  7. Jeremy

    Interesting, thanks mimarob.

  8. Mark

    What are the operating parameters of the air-powered pump? I have a well that is 300′ deep with a water table at 260′ and a yield of 2gal/min. Can this pump handle that depth?


  9. mimarob

    Hi Mark!

    Going back to my physics classes ages ago I remember that 10 meters of water would create a pressure of about 1 atmosphere.
    This is also the absolute limit on which you can suck water up.

    Since we are talking about 80 meters the pump needs to get to 8 atmospheres or above. I think that is the same as 8 Bar or 116 psi.

    Probably you need more in the real world but I think that is about what a normal compressor can deliver.

    On the power needs, I did some slobby calculations and ended up with about 100 Watts as an answer, meaning the compressor would be silent most of the time 🙂

    Now before destroying an expensive drill hole with non-approved equipment, a bench test would be nice.

    My suggestion would be to simply take it to a steep cliff somewhere, hang it by its hoses into a barrel of water and test the h-ll out of it!

    You could even have a third hose sending the water back again for a longer duration test.

  10. Jeremy

    Great calculations mimarob! We haven’t gotten the solenoid timer straightened out yet, so we don’t know how it’s going to work in the real world without any tests. I’m working with Ralf in Austria on using an arduino for the timer. We need to see if it will be cheaper than the $50-60 timers.

  11. Jeremy

    My conversation with Ralf is on the forum here:

  12. mimarob

    Hello again!

    If all you really need to do is to turn a solenoid on and off at certain adjustable intervals, maybe you can do without a microprocessor.

    I put a dirt cheap suggestion up on the wiki!

    I used a relay since I don’t know much about triacs!

  13. Jeremy

    Cool, thanks mimarob! It looks great! I’ll check it out and see if I can source the components.

  14. mimarob


    I think radioshack has most of this stuff.

    If you let Ralf do the triac part all is fine. Otherwise if you go with my simple transistor, make sure it can take the current of the relay coil.

    Also make sure the relay is speced for your voltage and don’t get yourself electricuted!!!

  15. Hydraulic Water Pump

    […] Air Powered Water Pump Prototype | Open Source Ecology […]

  16. ralf

    Jeremy, if you fear the Triac circuit you can still use the Arduino and take mimarobs relais circuit (you don’t need the transistor, just the relais and the diode connected to one of the Arduino output pins, the Arduino can drive 40mA on its outputs). The Triac is probably more lifetime-design, though (the relais, especially with inductive load will burn its contacts after some years).

  17. Jeremy

    Hi Ralf, I’ll need to study the triac and relay some more. We’d probably like to use the triac.

    I’d like to avoid the cost of the arduino for this component but we do need to use it for the CEB 2:

  18. Jeremy

    Earthling sent me a comment on the forums, what do you guys think?

  19. mimarob

    Sounds like good suggestions, but since your going with Ralfs suggestion to use the triac, you don’t need the mosfet to replace the transistor 🙂

  20. mimarob

    I edited the wiki according to Earthlings suggestion about putting the relay at the collector side. I don’t think it will work otherwise.

  21. Jeremy

    Cool! Check out Earthling’s design here as well: What do you think mimarob?

  22. mimarob

    No experience with SSR so I’ll just trust Earthling on that part.

    The expresspcb is a great service, but I think you’ll be just as happy with a simple stripboard design, especially since modifications are likely (Noone has actually built this construction yet).

    Above all, if you get it to work, get that into the documentation as-is!!

    If you fail I can rebuild the timer up to the led-part, in a few hours and verify the design.

  23. Solar Village 2010 | Open Source Ecology

    […] features are CEB floor, masonry stove, living roof, rainwater catchment, greenhouse in front, open source air-powered water pump, our own lumber, CEB water cistern, and keyhole growing space right in front. This means that in […]