DIY Centrifugal Water Pump

Introduction: DIY Centrifugal Water Pump

While trying to put together an Aquaponics garden as a project for my family, I decided I wanted to make everything as cheap as possible.  One of the most expensive parts of the project is a pump to circulate the water from the fish tanks back to the grow beds.  After looking around the house and thinking on it a few days, I gathered the materials and put this together in about 3 hrs total.  My initial test run yielded approximately 180 gph rate.  Pretty impressive I think for something I basically just threw together.

Sorry I don't have pictures to cover every step, but I didn't decide to do the i'ble till after the project was complete.

Step 1: Required Tools and Materials.

For this instructable I used:
-1 12oz aluminum soda can
-1 smokeless tobacco container
- Stiff copper wire, wire coat hanger, etc (Will be pump drive shaft)
- Super glue  (I know not best choice but all I had at the time)

Sturdy scissors
Pocket Knife

Step 2: The Impeller

Cutting open an aluminum can is dangerous and results in VERY sharp edges. Care should be taken to prevent being cut by the aluminum and by whatever cutting utensil you use.

First I cut the top and bottom off the soda can then split it length wise so I would have a rectangle piece of aluminum sheet to work with.

I put the aluminum on the top of the tobacco container and used the back of my knife to imprint the edge of the lid into the aluminum which gave me the perfect line to follow when cutting out the base for the impeller.  You will gain the most benefit from having this piece as perfectly round as possible.

Once the base was cut out, I folded it in half in perpendicular directions so I could find the exact center.  I then carefully used my knife to poke small hole through the center.

With the remaining aluminum sheet, using the scissors and ruler, I measured and cut 4 strips that were 50mm long by 25mm wide to make the impeller blades. 

To form the blades, I measured and marked lines 2mm from each end to mark where the mounting 'feet' of the blades would be.  Next I folded each blade in half across the length leaving me with double thickness blade sections approximately 25mm in length.  Then I folded each of the feet to the outside of each folded sheet to form the 'feet'.

I mounted my blades using super glue and 'eye-balling' the angles.  I put the inboard edge of the blades off center causing the effect of a factory made impeller that has spiral blades.  Mounting the blades perpendicular will work, however a loss in efficiency of the pump will be noticed based on my research.

Allow the super glue to dry completely and take care not to glue your fingers to your impeller like I did.

Step 3: Pump Housing

Next I bore a hole through the center of the pump housing for the pump shaft to pass through.  I did this using the shaft itself so I would have a tight seal.  I was able to twist and push on mine to get it to push through fairly easily.  Care should be taken not to stab yourself with the shaft material when it passes through the plastic.  Also pushing too hard can split your pump housing.  If this happens you will need to start over because re-sealing the split will be more pain than it's worth.

For the discharge port I used my pocket knife to cut two vertical slits in the side of the housing.  Stopping a couple millimeters from the base of the housing.  
********BE CAREFUL*********
The plastic is thicker the closer you get to the base so cutting out the section with my pocket knife proved to be problematic, so I used my Dremmel with a cutting blade attached to make a clean cut and remove the plastic from the discharge port.

Step 4: Pump Cover

*******COMMON SENSE AND SAFETY**********
Using the Dremmel and the cutting blade, I cut an intake port into the metal lid of the tobacco container.  My port ended up being about  15mm square.  If the port is too small, it will cause cavitation and output will be weak.  Too large and you may not have any output.  Your port should be slightly larger than the opening between your inboard impeller blades.

Step 5: Assembly

To mount the impeller to its drive shaft, I used super glue and a small piece of plastic I had laying around.  Make sure you allow ample time for the super glue to try while ensuring that the impeller stays perpendicular to the drive shaft.  When I make my next pump, I will likely use 5-min epoxy for these connections as I anticipate the super glue will fail when I start using the pump for an extended period.

Once the impeller is securely mounted, insert the impeller and shaft into the pump housing and seat the impeller as far in as possible.  You may find, as I did, that your blades are too tall for the container at this point.  To solve this, I simply bent the top edges of the blades in the direction I intended my pump to turn.  For me this was counter clockwise.  It may also be possible to simply trim the top edges of the blades for fit, but you must ensure to trim equal amounts from each blade to maintain balance as much as possible.  Excess vibration caused by imbalance will cause your pump to leak and fail much sooner. 

Lastly you will install the pump cover.  Initially I just snapped the lid on the container and left it at that, but during initial tests, the pressure inside the pump proved to be too much for the lid to handle causing it to pop off.  To solve this I just put a fine line of super glue around the inner edge of the lid and re-installed then allowed enough time for the glue to dry and set.  After this adjustment, I was able to run the pump at full speed with a Dremmel and the pump cover did not pop off again.

Step 6: Summary

After working out the initial bug (pump cover popping off),  I was able to determine this project to be a success.  I was able to pump out a two quart container in about ten seconds.  With these numbers I was able to estimate that my homemade pump has a rate of around three gallons per minute or 180 gallons per hour.  This rate was achieved using a multi-speed Dremmel set on speed number two.   

Once I get it set up on my Aquaponic garden, I will take a video of it in action and add it here.

I hope you've enjoyed this instructable and when I make my next model I will start taking pictures sooner to cover the entire process.

Thanks for checking this out!

Further testing has indicated that the design I created does not care the direction of rotation.  In reality, performance seems to improve ever so slightly when I rotate the pump in the opposite direction I intended. (clockwise)  More updates to come.

Well once I finally got the pump mounted to it's drive motor and added the suction and discharge lines, I attempted another test before mounting it in my Aquaponic system.  Unfortunately, about 5 seconds into this test, one of my impeller blades cracked and lost it's stability causing it to jam into the suction opening.  I'll need to redesign the impeller (out of plastic/rubber this time) and post another update with the changes made.

Be the First to Share


    • Remote Control Contest

      Remote Control Contest
    • Digital Fabrication Student Design Challenge

      Digital Fabrication Student Design Challenge
    • Stone Concrete Cement Contest

      Stone Concrete Cement Contest



    Question 4 years ago

    i'm currently making a paper for centrifugal water pump like yours can you give me some advice of what type of performance testing can i perform????? thankss xxxx


    Answer 4 years ago


    Thank you for your question. Personally, when designing a centrifugal pump, I focus on three main factors. 1) Does the pump self-prime? What I mean by this, is when the pump is not already submerged in the water, does it generate enough suction when empty to pull the water from the source and begin pumping? If not, how can I maintain the prime on the pump so that it is self starting? 2) How much head can this pump produce? and 3) What volume can this pump produce relative to the current head?

    I hope this helps!


    Reply 9 years ago on Introduction

    Originally I used small fan motor, but found early on that the torque load was too much and causing it to over heat. As a result, I switched over to an old 12v cordless drill motor I had laying around. This worked great and the pump ran for almost a week before the impeller was totally destroyed.


    9 years ago on Step 2

    Shouldn't be the blades bent at a particular angle according to the direction we want the water to flow in?


    Reply 9 years ago on Step 2

    I actually found that bending the blades OPPOSITE the direction of rotation to be the least problematic. This allowed a small amount of water to pass between the blade and the pump face. Bending them towards the direction of rotation provided only slightly better performance, but as the blades began to weaken or if the pump was run at too high of an RPM the blades would flex ( trying to become straight again) and catch on the square hole I cut in the pump face. Perhaps this would not have been an issue had I made the suction port round. Unfortunately, other things have taken priority (life got in the way) and I haven't been able to continue testing of this design. If you conduct further tests, I would love to know your findings as well as what modifications you make to create a more reliable pump!


    Reply 9 years ago on Step 2

    I am new in this field..and I have been wondering that what if the impeller blades are kept straight?....will the pump still work?....I mean it does work due to the centrifugal force which pushes the water outward so the bending of the blades shouldn't matter right?...I have a bit of confusion so can you help me out here...thanks


    Reply 9 years ago on Step 2

    Yes, straight blades will work equally well. Just remember that the tighter you keep the tolerances the better the pump will perform. Any gaps between the impeller and the housing are areas for the pumped fluid to bypass the impeller effectively reducing pressure and flow rate.


    10 years ago on Introduction

    25 years ago I made a centrifugal pump like this, using phenolic plywood as body and a little piece of brass as turbine. It was motorized by the electric drill, and has an awesome power. I used it to filter the water of the canvas pool, successfully.