Making A Better Pulser Pump
Some engineering flights of fancy
This video caught my fancy so here are my thoughts on improving the design. There seems to be a lot of things which can be done which should make big improvements but consider everything in this post speculative spitballing. Anyone who wants to improve on this mechanism is free to try my ideas.
Technically it’s a bit wrong to say this mechanism has ‘no moving parts’. It does have moving parts, they’re just air bubbles which are being captured on the fly and hence aren’t subject to wear. The problem is that air bubbles don’t like behaving.
Starting with where the water comes in:
The mechanism in the above video is cheating a bit because the pump getting the water into the top is aerating it. A proper mechanism should have a way of getting air into the water when it’s coming in slowly and steadily. In particular it should have a mechanism for being able to recover if the mechanism as a whole ever gets overflowed so it isn’t stuck with no bubbles in it forever. The simplest mechanism for this is to have a section of the pipe going down which has holes in the sides. As long as water is flowing fast it will pull air bubbles in through the holes. If it gets backlogged water will escape through the holes and can be directed to the exit, making room for air to be let in. The ideal size and spacing of the holes is unclear. If the mechanism were big enough it would probably improve things a lot to split across multiple pipes which have air intake holes to pull more bubbles in. It might also be a good idea to make a whirlpool and stick a pipe in the middle to help the air go down but that gets complicated.
Once bubbles are captured the downward pipe should be split into a bundle of straws to keep the bubbles from coalescing and forcing their way upwards. The ideal diameter of the straws is probably somewhat dependent on their length but should be small enough that surface tension makes water form plugs. The length of the downward pipe in the above model seems to be way too long. It appears to be that this is being done to make the pulsing effect happen but there’s a better way of doing that which I’ll get to.
The intake for the air bubbles should come from the bottom of the chamber where the pumping upwards happens. That should lead upwards to a manifold which is a short pipe with a horizontal cap at the top with holes in it, all kept under water. Air will then build up in the pipe and result in a steady stream of bubbles coming out of the holes. The size and depth of the holes as well as the material they’re made out of and the width of the pipe relative to the rate of air coming in all affect the nucleation of bubbles. What should happen is that bubbles of a reasonably consistent size come up at a reasonably consistent rate in a nice steady stream instead of the chaos you see above. There’s probably a range of possible sizes and rates of bubbles which are possible and that needs to be studied.
Instead of a single pipe going upwards there should be a bundle of straws. The bottoms of the straws should splay out and have tapered inlets with a one to one correlation with the holes in the manifold so the bubbles from that hole go directly into that straw and push the water upwards. The ideal number and diameter of the straws is very dependent on how far the water is being pumped, how quickly the air is coming in, and what they’re made out of. They should be thin enough that surface tension causes water in them to form a plug and makes bubbles force the water upwards. The idea is to make the water flow up slowly and steadily, with the upwards force of the bubbles just barely able to force it to the height it’s being pumped to, without wasting any energy on the momentum from those pulses. Maybe this shift in emphasis makes the whole thing technically a different mechanism.
At the top the straws should flare away from each other so the water going out of one straw doesn’t fall into its neighbors.
Hopefully these changes can improve the efficiency of the system from awful to merely bad. You’d still only use it when you care less about efficiency than low maintenance or quiet or specifically want aeration. Using all those straws will reduce how well it works on water containing particulates.
No mechanism to entrain air bubbles is needed in the trompe. What actually happens in a 4 inch or 6 inch diameter pulser pump is that the water has a half meter or more of free fall. What does it do in a half meter? It accelerates. But water is incompressible, if water is moving faster in the pipe than the water that is going to the pipe, there is extra space, and that space is taken up by air. Eventually you can get a more or less steady state, where the water finds its speed and air bubbles are leaving at the bottom at about the same rate as they are being replaced by air coming in at the top. The straws idea might work. I have suggested it too, the issue is, how do you direct the right amount of air to each straw? In real life, when I had 2 airlift pumps working from the one trompe (with the airlift part and the trompe part separated), I had to have a tiny bead in the line before each one. (Otherwise, the tendency was for most of the air to go to whichever airlift pump was slightly lower). I use tiny airlift pumps connected to an aquarium air pump to water my garden. I usually have about 10 connected to the pump. You need a little adjustable air tap before each airlift pump, otherwise, all the air will go to the airlift pump with the lowest submergence, you tweak them all at the start, to get everything going, and then all should be good. (Taps are like variable resistors in an electrical circuit). Another thing to note is you do NOT want a whirlpool going into the trompe. Any time I had one, there was less air going down. What I did with my last trompe the 6 inch diameter one, was I got about a 6 inch long piece of pipe, cut it in half and screwed them together so they are like 2 C's back to back, then I put them over top of the pipe, and cut 4 notches in them so that the sat at the right heigh on the pipe. Then I put a plastic can over top. I had a hole in the top of the can to let in air. No more whirlpool with that. I think I had a 1/2 inch pipe going down to the entrance to the 6 inch pipe, but I don't remember. The 6 inch trompe was too big for the stream. It needed 500 or 600 liters per minute, but the stream usually topped out at about 350 liters per minute. This worked well for a 4 inch trompe. My stream had 45 to 55 cm of head. Thanks for being interested. People need to make trompes with 10 cm, 15 cm, and 20 cm and 4 inch 6 inch and 8 in diameter pipes. Trompes with half meter head, 3/4 meter, and 1 meter head, and going down 1, 1.5, 2 and 2.5 meters below the exit water level. That will finally tell people how efficient they are. I tried to guide students doing experiments with small models. They tend to be willful. One left out the taps, because he didn't understand the "why" and he got really low efficiency as a result. It was very annoying because the results were posted. My results for compressing air, with a half meter head and going to 2.5 meters down, was around and just over the 30% efficiency rate, while pumping water to 3.5 or 4 meters was about 6% efficient. And the thing that matters is cost, not efficiency. Think about it, the stream I used could have a half meter head pulser pump every 50 meters, so in 200 meters of stream length, you can have 5 pulser pumps pumping 4 liters of water per minute to 3.5 meters. That's 20 liters per minute gone from your 350 liter per minute stream flow in a very short distance. (or a whole lot more if you are pumping to 2 meters high). Much airlift pump calculations and experiments have been done, but very little work has been done on low pressure trompes. That is the part that needs experiments. It beats me why no plastic pipe company has done this work. They could sell a lot of pipes and a lot of fittings, if people started using pulser pumps. My greenhouses and some of my garden planters use circulating water, and this is supplied by a little aquarium air pump. But if I lived near a small river or a stream, I could easily supply way more than enough air with a pulser pump. A one meter deep trompe would be enough to supply my type of airlift pumps in a garden. I hope that helps, Brian
Love the systematic approach here. The insight about using a bundle of straws to prevent coalescence is elegent, and the self-recovery mechanism with holes is really clever for handling overflow states. I've ben tinkering with low-energy water systems at a small scale and the tension between efficiency and maintenace-free operation is exactly the tradeoff nobody talks about. That manifold design for controlled bubble nucleation could probably apply way beyond just pumps.