How High Can a Fishtank Air Pump Pump Water?




About: I am a stone mason. My hobby is making new solar cooking and gardening stuff. I have used solar heat to cook soil for a couple of years. In mother earth news in January, i read that their compost expert does...

I have used an air pump for a fish tank to pump water fairly high but I am curious about how much higher you can pump.
If you have a suitable site and you are curious too, maybe you can answer the question.

Teacher Notes

Teachers! Did you use this instructable in your classroom?
Add a Teacher Note to share how you incorporated it into your lesson.

1 Person Made This Project!


  • Made with Math Contest

    Made with Math Contest
  • Cardboard Speed Challenge

    Cardboard Speed Challenge
  • Multi-Discipline Contest

    Multi-Discipline Contest

18 Discussions


4 months ago

Well I study window farmers , they had an abundance of info on airlift water . Rather high too


4 years ago on Introduction

Hi what are the challenge rules? I have gotten my air lift to pump 2 feet with about 8 inches of submergence but at a much higher volume of water, enough to power my aquaponics setup. I am using a 5w 9.0L/m air pump. This air pump I invented (re-invented) is called a pneumatic ejector pump. Pneumatic ejectors which are more efficient than geyser pumps are limited only by the pressure the air pump can provide. Here is a video of the results


7 years ago on Introduction

I responded to you awhile back and had another thought. By the way sorry I saw the comment about this being a doing not thinking challenge although I have not tried this for what ever reason.

Have you or anyone ever tried cascading and using the same pump with a splitter to power the subsequent airlift section. It seems to me that using the numbers you show in your video 22inches of submergence with 13 feet is about 15% of the head total submergence.

I also saw the comment from Shascho re the Geyser pump but haven't found any real good schematics or video that really show how they go together.

Any way just a thought - has anyone improved on you numbers that you achieved?

1 reply

Someone got 16 ft from 2 ft submergence and when I moved I was able to rig up something that went to 18 ft. So 18 ft is the official "world record" at the moment.
I posted it on youtube at
Note that new tubing is either too "clean" or has some sort of water repellent coating. This causes problems for airlift until the tubing gets dirty. (Liquid holdup is super good in new tubing but that is a bad thing because you can end up with the water "stuck" in the tube and blocking the system and refusing to move up or leak down. It is a stupid problem but if anyone tries the experiment, be aware of it! Maybe leave your new tubing full of water for a few days before trying the experiment and that should remove the coating (or whatever causes the problem). The first I heard of the problem was from windowfarmers and I didn't believe them until it happened to me. (Because I have done airlift hundreds of times with no issue. (but always with old tubing)).
Cascading works, but it is hard to accurately divide the air in half or in 3rds.
I showed it in windofarms and other people have done it too. You need adjustable valves to get it just right. It is more fiddly than you would expect.


7 years ago on Introduction

Great challenge.
Have you considered the Geyser pump?
It's the basis of an entire industry.
It allows the air to accumulate at the bottom of the pipe
and by additive increase of its force,
accelerate a smaller 'plug' faster and higher.

I've never seen one in operation,
but here is a good explanation.

Keep up the good work, partners!

1 reply

Reply 7 years ago on Introduction

HEY , this isn't a thinking challenge, it is a doing challenge!
Thanks to Itsandbits1 for actually taking up the challenge. He is currently the WORLD CHAMPION in this event! (Though I would like video confirmation as proof). (He pumped to 16 ft high). 20 ft or 25 or 35 ft might be attainable!
Wiithin the limits outlined above in the video, feel free to take it up and take the crown away from Itsandbits1 If you do it by 31st and post your video before then, you get to be WORLD CHAMPION for 2011 and nobody can take it away from you!
Now back to your reply, I am afraid accelerating water faster and faster is not good. (Turbulence increases exponentially with the acceleration so you waste too much energy).
And, when the plug is going too fast it degenerates into "bubble flow" and this flow type transfers far less fluid.

The experiment has a couple of reason: First of all this might be a very useful effect (if it pumps a little bit higher) but nobody seems to know the limit. I am really surprised that progress is so slow.
The second reason is social engineering. Many people love science but seem to get drawn to stuff they cannot test (like relativity or that guy with the cat). This is an attempt to draw people to stuff that they CAN test and where they CAN make a useful contribution and CAN collaborate. What stops people doing this? For less than $50 and a couple of hours of your time, you might help design a new type of pump that can save peoples lives. (Because it is so cheap)


8 years ago on Introduction

maybe try a check valve at the bottom to stop the air and water from going out the bottom but it lets water in.

1 reply

Reply 8 years ago on Introduction

A check valve is not needed if your non return "U" is deep enough. Check valves are prone to failure and they also slow down flow a little bit. Over a day this slow down can really add up. Another big flaw with check valves is that if they are good they can stop back flow all together. This matters because there must be some back flow in this system. A spontaneous (very rare though) situation can occasionally arise where the combined length of plugs is greater than the air pressure. A good check valve makes this permanent! and your pump will stop. With the U, the last plug just drains back into the U and your pump keeps going.
Feel free to try it though.


8 years ago on Introduction

the part of the plug flow that you may be missing in your explanation is the airlift effect. If you take your 1 meter high water column and inject air plugs in equal amounts all the way up the column, you still have the same amount of water but the air is seperating it and you can end up with a lot higher column just because of thisbut the fact that there is a pressure differential between the air you put in and the ambient air will cause the higher pressure air in the system to keep pushing the plug of water higher and higher to get out and is supported on the bottom by the higher pressure air that keeps entering the system from below

3 replies

Thank you. I had enough pressure to push air 1 meter deep under water and that's it. The pump is unable to push the air deeper. So we are talking really low air pressure!
Engineering books (in their sections on airlift pumps) and most of the stuff online does not acknowledge that extremely low pressure air can pump water very high. (I got to 13 ft on about a meter pressure but only because I had ran out of tubes and ladder to go higher). It had no problem whatsoever pumping to 13 ft.
Sure I did not measure efficiency but just getting that high was surprising.
My points are 1 This is a very cheap pump, 2 this might work directly from a solar panel if dc pumps of the right voltage are available 3 It just pumps air and nothing but tubes need go down the well so it should be very robust and 4 Surely it is worth testing just in case it could be of use in poor villages? If you are pumping over 40 ft high, other pumps need quite a lot of hardware down the well, this just needs tubes! Thanks Brian

I managed to get 16ft with 24inch submergence using the small plastic drip irrigation hose. I actually had to go sideways a couple of places which made it more difficult because the water created air locks in the flat spots that had to be broken to keep the flow going but it managed with 35 feet of hose at very minimal flow and pressure.

Excellent, thanks. Could you do a video too? and put it as a response to mine? I am also gaiatechnician on youtube. I think irrigation hose is probably too small diameter. I found 1/4 or 5/16 worked pretty ok. I think 5/16 was better. Even if it is minimal flow, there may be niche applications for this. Mine went practically straight up so there were no flat spots. Thank you very much for your effort.
I will note it at my video. Brian


8 years ago on Introduction

A quick back-of-the-envelope calculation would consist of attaching a pressure gauge to the output of the pump and blocking the output off somehow.

Then (In Standard International Units) from the maximum pressure that the pump can exert,

p = rho g h

rho = density of water = 1000 kg/m^3
g = acceleration due to gravity = 9.8 m/s^2
p = pressure in Pascals
h = height of water column in metres.

The height will be a little less than this, due to pipe friction, but assuming that you're after only a little trickle at the end of the pipe, it should be pretty close.

5 replies

These pumps only exert enough pressure to push air about a meter below the surface of a container of water. (Depends on the pump). Using airlift and plug flow, mine pumped water to 13 ft high and I am pretty sure the limit is way more than that.

Theoretically, using formula given by ancienthart, that I agree on ;-)
it means that having water at ground level (p=10e5 Pascal), you can lift water with a pump that creates depression at a maximum of 10meters.

A good estimation of your pump pressure can be calculated following this.
Bernouilli's principle tells you that :
1/2mV^2 + mgz + P = Constant
which means fluid speed, elevation and pressure are kept along a circuit (forgetting about friction losses, compressible fluids hypothesis, ...)
Fluid Flow (mV) can be estimated counting the time required to fill a known container. (for ex. 4liter bottle is being filled in 30sec : mV=4/30=0.13 kg/sec)
V can be determined as you know m fluid mass (which is easy for water : 1L=1kg)
You can then determine your Constant in your Bernouilli formula.
Given the relevant hypothesis that you are not moving up or down, z in that Equation becomes 0, and therefore the constant gives you the P, which is the pressure at the exit of your pump, when speed is estimated to be zero.
I am using Bernouilli when Pressure is 0 and speed maximum, and Pressure maximum with speed V=0. Overall Energy (the Constant) is being kept.
Although I am not sure the pressure at pump exit will be the same as the pump entry, this can give you an idea.

Let me know if you attempted it...

I pumped water to 13 ft (3.9 meters) with 98 millibar air pressure (or a bit less) in the video. Why are we discussing Bernouilli's principle?
I just found this description of 2 phase fluid flow online and it has some math lower on the page. It also has some little videos demonstrating the different flow types which are definitely worth checking out.
The pumping was by plug flow. I am pretty sure it will pump over 10 meters if the plugs maintain themselves well as they go up the pipe.
However, I cannot do this because I do not have a 10 meter height difference on my property.
The maximum length of all the plugs in the pipe was 22 inches. (friction losses and pressure loss with turbulence would be critical to know how high you could go at the max). As a plug exits the pipe, the reduction in pressure means that another plug is formed pretty much immediately at the t- joint at the bottom.

Introducing Bernouilli's principle was to estimate pump output pressure...

This 10 meter limit exists in the case your pump is higher than the water you want to pump. Theory tells you it is impossible to ''suck'' water from an altitude over 10m. If your pump is at water level, then your pump is pushing the water up, which is a different story.

The thing that I do not understand is that your 98 millibar pump should be able to ''push up'' water up to +-1 meter (using formula given by ancienheart). When you have this water column of 1 meter, the pressure it gives you at the bottom will be equivalent to the pressure of the pump (98 millibar in your case), and therefore, pump pressure would need to be higher to continue to push water. I do believe you but then I am lost with what I understood from the theory... I am probably missing something...

Ok, I will try to explain what happens. My air pump can only provide I meter of pressure while the t joint is situated at 2 ft (about 60 cm.) As soon as air is turned on, it reaches the T and pushes the 60 cm of water above the T up and the water below the T down. So you have roughly 60 cm of water going at whatever speed the other 40 cm of air pressure can make it go at until it exits the top of the pipe.
Only when it exits the pipe will there be more water added at the bottom and we get a flow upwards that is at some equilibrium speed.
I have a bunch of videos on my site to try to show and explain airlift. There is also windowfarms. They use this type of airlift extensively and have done for about a year.
The big thing for me is I have no idea where the limits are with this type of "mycro" system and I cannot test to find out!
Maybe low pressure airlift with a meter of pressure and 1/4 inch pipes can lift to 20 ft, maybe 30 ft, or maybe it can lift to 60 ft! Until people try it we will never know. Even if it can only lift to 20 ft, it opens up a whole host of applications that have never even been tried before!