Pump water with no electricity, no gasoline, just gravity!

Sound crazy or impossible? Don't worry, it does obey the laws of physics, but I'll try to explain the operation later. This instructable shows how to build a fairly simple water pump that needs no energy input other than water flowing from a higher point to a lower point. Most of the pump is constructed from PVC, with a couple of bronze pieces thrown in for flavor. I was able to source all of the parts from a local hardware store (Lowes) for a bit under $100.

To function, the pump does require a reasonable amount of water that will drop at least 3'-5'. The level that the pump can raise water to depends on the water's head (total drop the water will make).

This design was worked out by Clemson University.

If you like what I've done, please take the time to give it a rating, and I'd love to hear your input. Thanks!

Step 1: Bill o' Materials

Before you can really do much, you've got to go out and buy some stuff. One of those sad facts of many projects. But if you want to build this (and it's a lot of fun to see it work), print out this list and head to the plumbing dept of your hardware store.

Materials for the Pump
  • 1-1/4" valve
  • 1-1/4" tee (buy two of these)
  • 1-1/4" union
  • 1-1/4" brass swing check valve
  • 1-1/4" spring check valve
  • 3/4" tee
  • 3/4" valve
  • 3/4" union
  • 1-1/4" x 3/4" bushing
  • 1/4" pipe cock
  • 100 psi gauge
  • 3/4" x 6" nipple
  • 4" x 1-1/4" bushing
  • 4" coupling
  • 4" x 24" PR160 PVC pipe
  • 4" PVC glue cap
  • 3/4" x 1/4" bushing
  • Short (4') section of 1-1/4" PVC pipe
  • Old Bicycle Innertube
This parts list comes directly from the Clemson website. I recommend you look there for help in identifying what each of the pieces look like, if you're unsure. I'm also not convinced that the 100 PSI gauge, or all of the things that make it possible, are necessary. This will probably drop the price a good bit, and I haven't found a need for it on my pump. The associated pieces are: 100 PSI gauge, 3/4" Tee, 3/4" x 1/4" bushing, the 1/4" pipe cock. Four things not needed. But have them if you like.

Connections Note Read through the instructable and understand all the pipe-fitting connections that will happen before buying materials. The store may not have exactly what you're looking for, and you may have to improvise. I wound up getting some different parts because my local store didn't have the exact parts I was looking for. This usually appears in the form of not having a threaded fitting, but having a smooth pipe connection, or vice versa. Not a problem, you can figure it out.

Installation Materials
  • Long section of 1-1/4" PVC ("drive pipe", connects pump to water supply)
  • Garden Hose (male end threads into 3/4" union, supplies pumped water)
  • Bricks, blocks, rocks to prop up and anchor pump
  • Shower Drain assembly (must be able to attach to 1-1/4" pipe, for attaching pipe to water supply)
Build Materials and Tools
  • PVC Primer (I used Oatey Purple Primer)
  • PVC Cement (Oatey again, just what they had)
  • Teflon Thread Tape
  • Hacksaw
  • Measuring Tape
  • Clamps
  • Pocket Knife
  • Lab gloves (keeps the chemicals on the pipe and off your hands)
  • Bike Pump (to inflate the innertube)
<p>designing one...as a pro to type to help my home rural technicians with the best of it kind while using the available local resources</p>
<p>Nice invention you have there! :)</p><p>Question: If i were to use this to pump water vertically up to a tank located 10m above, would it be possible? If so, would it be better than to use the conventional pump? Thank you so much!</p>
<p>After some research and calculations and I'm not sure if they mention it on this but me and my buddy found out that for every 1ft of elevation on the head you should be ale to achieve 7ft of lift.</p>
<p>Hi kelvintoys,</p><p>As is often the case, the answer is &quot;it depends&quot;. This pump's ability to lift water depends on how far its water source falls in getting to the pump. If you have a decent drop, it's certainly possible. If it's flat water, the ram pump probably isn't your best choice. Please post more details of your water source if you have the chance! </p>
<p>While I didn't read every single comment, no one seems to touch on the subject of weather. If you live in an area where temperatures stay below freezing, or sub zero, this project will not work. All the PVC pipes with split open if water isn't constantly flowing.</p>
<p>can you share link where you buy spring check valve ?</p>
<p>Can I use a 300 gallon tote and run the feed out from the bottom to drive the pump? </p><p>Thank you!</p>
<p>what will you do when the tote goes empty?</p>
<p>Ram pumps have been around for over 300 years. Still a very valid design if you have a lot of water dropping some distance, and need to move a little water a further distance. The old cast iron pumps could run 50+ years without maintenance. https://en.wikipedia.org/wiki/Hydraulic_ram</p>
<p>Any ideas on how this might integrate into a gutter downspout? I'd like to fill a rain barrel that's a bit uphill from my house.</p>
<p>I really appreciate your effort in providing this important information , thank-you very much..</p>
Can you use other materials then PVC
<p>I have constructed one water ram but its not working. I have checked all possibilities of air leakages from pressure tank and their is no leakage. I have attached the image for your reference. Please help me out. My source tank is about 15 feet height and delivery pipe is above source tank height. Still it is not working.</p>
<p>what i need to know is how to get the water out of my well with out electricity.</p>
<p>Your best bet is a windmill pump.</p>
<p>I watched a video earlier tonight about an air lift pump that a couple had built .It was quite interesting to watch.They used an air pump to power their air lift pump.It only served to pump the water out of the well not into their house. https://youtu.be/TjYLPW93-EA</p>
<p>Not sure about your well (depth, safety of the water, or whatever) but take a look here at this link for lehmans.com : </p><p><a href="http://non-electric.lehmans.com/search#w=well%20water" rel="nofollow">http://non-electric.lehmans.com/search#w=well%20wa...</a><br><br>I would love to set something up like this for my well, unfortunately I have really BAD water unless it goes through my filtration/softening system. It's not safe (bacteria) to even use for washing dishes or in the toilet. (Airborne bacterias, nasty stuff!) <br>BUMMER! I still need electricity! </p>
Sadly this pump won't help with that, but you might look into pulser or geyser pumps. You'd need some pretty specific terrain though. Probably easier to go with a windmill / solar pumped hybrid system. Best of luck!
<p>hi my name is giovanni Agrippino and are interested in buying this pump,<br>how we can contact? my address is Giovanni Agrippino, 87065 Corigliano Calabro, via capri snc, Cs, Calabria, Italy - tel 0983856093 - cell. 3331648415 - clementino.max@gmail.com</p>
<p>Great site,thank you.I hope you can give me some advice with my project.We need to pump water about 200 feet length and height about 3 feet.We have a small slow moving creek with no drop but i dug a 7 foot well beside the creek and want to place the pump on the bottom.Besides the problem of the waste valve water would it work?I want to put a short piece of hose on the waste valve so it can pump it out of the well.Can you give me any hope?</p>
<p>So the short answer is that the hydraulic ram is not the pump you're <br>looking for. You really need the water to fall a decent distance to <br>drive the pump. But I can give you hope. For low lift requirements and <br>in the setting of a slow moving watersource, I would look into screw or <br>spiral type pumps. Specifically investigate the Archimede's Screw, but <br>the DIY ones made out of a hose wrapped around a core. It won't pump <br>fast, but it should be able to lift water to a container 3 feet or <br>higher. After that, you just need to run pipe. Let me know if you have <br>more questions.</p>
<p>If I build the ram and set it 15' down in my lake, will it pump water straight up an additional 5 ' to a holding tank?</p><p>Thanks :)</p>
<p>theorically..no..it wont.</p>
<p>Can I put an inverted u-tube at the top of the standpipe to keep debris out? If so can I restrict it from 4&quot; or 3&quot; to say 2&quot;?</p>
<p>While I'm not entirely sure I'm imagining the right thing, I would say that you likely could put an inverted u at the top of the standpipe. <strong>However</strong>. </p><p>I'd hesitate before restricting the diameter. As you decrease the diameter of the pipe, the cross-sectional area will decrease by the square of the change, &amp; the volume of water that can pass through at a given pressure will also drop. If you look up some of the calculators online, you'll see that the diameter of the standpipe is important. </p><p>Ultimately, I'm not sure the inverted u would filter debris - but if you leave a comment with a bit more description of what you're thinking of, I'd be happy to think about it.</p>
Thanx for your answer, habolooby. I live in an oak forest. If I install two 90's at the top with the opening facing down, I could avoid leaves and branches falling into the opening. 3 and 4 inch 90's are expensive. I thought I might accomplish the same thing by reducing the diameter of the pipe first.
<p>I don't know if this image will work, but this is what I imagine you're thinking of. Figure A is your plan. To filter incoming water, you'd make a U out of two 90's and a T joint, with the 90's pointing down into the water. I'd be wary of anything that restricts flow into the pump. Instead, you might consider figure B. Here you would cap a piece of culvert, large pipe, a barrel - anything big really, set it so it's ~ 3/4 in the water, give it a vent on top, and have the stand pipe draw water from inside the cylinder. Bottom of the barrel / pipe / whatever is open to the water. Sorry for the crap quality, it's what I could come up with in ~15 minutes between work things. Let me know if you have questions. </p>
<p>I was referring to the stand pipe used to compensate for friction when your supply line is more than 100 feet.<br> <br><br> <br>Thanx for the info though.<br> <br><br> <br>Magfree</p>
The stand pipe coming out of the water has a cap. I'm not sure if that is what you are talking about but the big pipe acts just like the pressure tank for a well pump.
Something to consider for the waste valve is a modified foot valve for a well. The valve can be disassembled and the valve turned to work in reverse. I prefer this method because you can add weights to the valve that control how much pressure is needed to close the valve. I feel like this type of valve will hold up longer than a swing gate valve as well.
<p>I'm not very familiar with well foot valves, but I have seen designs for ram pumps that had adjustable waste valves. Sounds like what you're describing, and it has that nice tune-ability feature. Hardware store parts used here for simplicity, but by all means modifications and improvements are encouraged. If you have any photos / drawings of your approach, please post them - or make an instructable! I'd be happy to post a link or something on here. </p>
They are available at hardware stores. Well here anyway. I can post a picture of the modified one when I can get down to it. I don't have it in the water now because of the cold weather.
<p>Hi there. If you are pumping up to a tank, can you use a ballcock at the tank so as not to waste water? I assume the pipe pressure would equalise and stop the pump when the tank ballcock was shut, but wondered if when the ballcock opened again and the output pipe pressure dropped the pump could restart itself? I would appreciate your thoughts.</p>
<p>I'd agree with hulk1371's response, and would add that if you were put a valve at the end, closing it would be similar to raising the outlet pipe to an elevation higher than the pump could lift water. The pump would pressurize the system a bit, and then cease to cycle. Opening the valve would relieve pressure on the outlet side, but it is unlikely that the pump would spontaneously restart. You'd probably have to go and give the waste valve a poke. Hulk1371's advice is spot on.</p>
You tank should not be sealed air tight. If it is a large tank it would take a long time to build that much pressure though. You need something to run off excess water when the rank is full. The best setup is a top fed tank with the valve a little off of the bottom. (you want somewhere for sediment to go) you also want a clean out on the very bottom for when the sediment builds up. At the top of the tank you want a pipe that has a check valve to allow water and air to escape without allowing everything else in. The overflow should run to somewhere that you don't mind water draining too. <br>
I do not want to take away from this excellent instructable but I think you under estimate the value of the drive pipe. This makes all the difference on the overall effectiveness of the ram pump.
<p>I'd agree completely. The drive pipe (and available elevation drop) are pretty much determine your energy source. Didn't mean to undersell it! </p>
<p>Thanks for your post.... I'm so grateful with your great idea.... I will try to produce one sample in order to help a particular community where we constructed a Masjid(place of worship) here in the Philippines.</p>
<p>Hi <a href="/member/habolooby/" rel="nofollow">habolooby</a> this is a great find. Thanks for posting. I have a fast flowing stream with about 1.5m head needing about 30m of drive pipe. The outlet pipe is about 60m long and will feed a dam maybe 1m above the high point of the drive pipe.</p><p>a) what sort of life expectancy do these pumps have (ignoring damage from river debris) and</p><p>b) how do you keep the water filtered at the intake - do you need to fit a filter for river applications? Assuming &quot;yes&quot;, does the pump self clean the intake filter with a backflush when the pump stokes, or do you have to vigilantly inspect and clean?</p><p>Thanks in advance for your reply.</p>
<p>Sounds like a good place for the pump, I hope it works out well! </p><p>As to your questions - I'm sorry to say I don't have a good answer for the life expectancy. As something with only two moving parts, I'd hope it lasts for a good while. I haven't had the one I made in continuous operation - I'm not around the farm much these days. Might post the question as a reply to some of the other commenters who built these. </p><p>I would filter the intake. A large screen filter creating a bubble around the intake would probably be best, so it doesn't decrease the intake flow rate. </p>
<p>Can the output pipe from the ram pump drop in elevation before it makes its rise? For instance can the output line drop down a canyon before climbing the other side?</p>
<p>Short answer is yes. </p><p>It may affect the start-up operation of the pump, but in steady state is should be fine. Water finds its own level, and all the pump cares about is what back pressure it has to work against. If if spring check valve doesn't have a little pressure behind it to accelerate it when it closes, I think the pressure at the swing check never gets low enough for it to open, and the pump gets stuck. You might have the outlet pipe make a short elevation run before diving through the canyon etc.</p><p>If you're doing a super long run in small pipe, you might lose some flow to friction in the pipe, but typically the water in the outlet pipe flows slowly enough for this to be negligible. </p>
Thanks! it works great!<br> <br> I was wondering if spring check valves with more or less psi requirements make a difference; one spring check valve i found had a small week spring in it and others required some force to open? The one i used which you may have used, was one i picked in-inbetween the spring strengths i found.&nbsp;<br> <br> <br>
<p>The strength of the spring will determine at what pressure difference the valve closes, but I don't think it has a major impact on function. I have seen other designs that have adjustable versions of the swing check valve. This lets you determine how fast the water must be flowing before the valve activates, and thus how much pressure you build up. Here this element can be slightly adjusted by rotating the pump along it's long axis, changing the force of gravity on the swing check valve. </p>
<p>Hi,</p><p>I'm planning to use it to pump water upto 1 KM at ground level from creek to prawn pond, will it success ? need calculation formulas ? pond is around 3 Acre , need to recirculate water. I may need multiple pumps can you guide me ? </p><p>parab.shailesh@gmail.com</p>
<p>Sadly, I don't have the time to provide and explain the formulas, but I'll do my best to help here. The main concern the ram pump has is resistance. This is usually answered by how high it needs to pump water, that is how much pressure it must overcome to get water across the spring check valve. If your pond is at the same elevation as your pump, you will need to consider the resistance your 1 km of pipe will produce. This is a whole separate set of calculations, but is easy enough. If the pond is higher than your pump, you'll have to take the elevation change into account. If the pond is lower... skip the pump and just run the pipe! Hope this helped, and some time spent searching the internet should answer your calculation questions. Good luck! </p>
<p>So i am planning on building a two tiered pond with water fall, and thinking about using this. would it work if instead of water pulling out of the check switch valve it poored back into the lower tier of the pond?</p>
<p>I'm not sure I entirely understand your plans, but here goes. The <br>essential function of this device is that if you have water falling some <br> distance, a small portion of that falling water can be pumped higher, <br>and the rest will continue on to a point lower than the source. However, <br> if you're thinking of the perpetual waterfall, I'm afraid that's still <br>just a good optical illusion :)</p>
<p>habolooby:</p><p>Sweet find and good job publishing for us, although I'm lost as to what the purpose of the 1 1/4&quot; union is. Can you explain?</p><p>Thanks,</p><p>Matt</p>