Aquaponics - Any volume in, any volume out cluster siphon progress

This has been a real eye opener.

Perhaps my eyeopeningest "thing" of my 120 things in 20 years yet. At least within aquaponics. I have to say moving enough electrons to make a light glow was pretty cool.

Anyway, I felt reasonably comfortable with my understanding of the bell siphon. I thought I understood how it worked, what caused it to trigger, and what caused it to stop.

It turns out, I have no idea.

This thing's a total mystery.

It wouldn't be the first time something made perfect sense, but went un-noticed by me.

What I thought I'd do is be very clever and make a cluster of siphons.

There is this phenomenon where, when a bell siphon triggers the height, or head, of water can actually rise above the height of the surrounding depth.

I figured this would allow me to make a device that had a cluster of  bell siphons, that would be force triggered by an uber siphon.


Picture this... get a bell siphon, and fill it with smaller bell siphons.

The original standpipe would be the lowest. The bell surrounding this would also encompass all the other bells for each additional siphon in the cluster

I tried a device with 4 standpipes. This took just under an hour to build and most of that was looking for the parts.

3 of the standpipes had a mini bell, the 4th had a maxi bell that covered all the bells. The 4th standpipe was also slightly lower so that it would trigger first. This 4th standpipe was the one that didnt have a private bell on it. (unseen in the photo behind the glass bells)

The plan was, the 4th standpipe would fill first. This would drain the uber bell of air and actually raise the water level above the other standpipes. This would in turn, of course, decisively trigger them. (this worked)

What this would mean is that a tiny flow (enough to trigger only one mini-siphon) would instantly trigger all 4 siphons. It would also mean that the drain would be so fast that the siphon would shut down in a very decisive manner, because it would be effectively pulling four times as much water as it might be if only one siphon were involved. To aid the shut down there is an air breather pipe (seen sticking out on the left) with the other end inside the big cover all bell. The big cover all bell is not shown because it blocks everything else, but sits so the bottom rests at the level the small breather pipe on the left meets the brown tape.

Most importantly, if there was a flow anywhere greater than enough to trigger 1 siphon, and less than the amount 4 siphons could cope with, this system would start and stop very decisively.

In a few trials, the system worked perfectly. I could trigger the system so that it started and stopped repeatedly with a flow rate of ...

2.5 seconds to fill a 500ml jug (my laundry tap on full)

And the same siphon also triggered at...

39 or 29 (it's hard to read my note)  seconds to fill a 500 ml jug (my laundry tap on a trickle so that the stream was breaking up before it hit the jug

So it worked. It worked over a massively variable flow rate,  and I'd made some kind of revolutionary new device.

But strange stuff happened all the time when I tried to reproduce the results.

I couldn't do it.

Sometimes it would work, others not.

Keep in mind, this thing was made of blutac, wire, and brown packing tape, so there is a fair chance I'll be able to refine it, but in short, I'm left feeling that I have much less of an understanding of the dynamics existing within a bell siphon than I thought I had before I started. The siphons didn't even always trigger in the same order.

I hate it when that happens.

I love it when that happens.

The true value of the scientific method is that it can show you that you know less that you thought, or you are just plain wrong, just as easily as it can show you how fabulous you are. An independent judge of fabulousness!

Valuable, but never comfortable :)

We live and learn.

My next step, is of course, to figure out what's going on, and make this device work.

When it worked it was a pretty impressive thing to watch, and much too valuable to give up on.

Wheels within wheels.

Or at least bubbles within bubbles.

Way cool.



  1. I'd say, when you first tested it, you stumbled onto the optimum setup and it just worked, but when you went to reproduce it, something has moved slightly, the setup is slightly different.... enough to stop it from working. Perhaps the tolerance between working perfectly and not working at all is very narrow?

  2. It could well be, but I think there is a lot of room for normal refinements. Things like cutting the siphon tubes straight at the top can have huge impact. Also my seals were made of blutac and brown packing tape. That tape isn't even water proof :) so air is getting into what should be a siphon. There are many things wrong with it. Any many things to attempt before decide its a bust :) A few other people are now making them and trying to reproduce the results. The design should make the secondary siphons trip very easily.

  3. A tiny bit of air in will kill a siphon, because air is so much more fluid (less viscous) than water. I'd guess, as you infer about the tape not being waterproof, that you had a decent seal the first time, and not subsequently, so if you use real materials, it won't turn out to be as tricky as the faulty seals would suggest. Good on ya, and thanks for sharing it.

  4. I'm not really sure what happened in the end with this one.

    I suspect that the issue was with the venting of air through the mini siphons to the maxi siphon.

    I have a feeling that it all boils down to me gaining a seal the first few times I ran it, and then just enough times afterward, to keep me interested, by chaos inside the junction between the join of the siphons blocking the pipe.

    Where there are several pipes joining, it is easy to get unpredictable water flows. I think I hit on a water flow that kept air from entering through the other outlets. I tried submerging the siphon, but I get the feeling that just added yet another layer of complexity to the situation.

    I get distracted by new ideas a bit, and got a request for a from splitter or sequencer, so moved on from this idea to that.

    The flow splitter works perfectly, so I would encourage you to ignore this and move on to the flow spitter like I did :)

  5. "from splitter" should read "flow splitter"

    and "Anonymous said..." should read "120thingsin20years said...", but I'm using someone else's computer :)

    I'll prove it's me by not deleting this tomorrow. But don't believe it's me the next time, because now the nutters will enter the stage.

  6. This is what "appears" to be a venturi device used to prime a larger siphon. It's not a bad idea, either. You would need a trap of some sort downstream though I think in order to provide resistance to air entering the next siphon.

    1. Yeah that could work. I tend to try to avoid anything other than a free breathing exit though to try to avoid complications.

      Also I have conventional siphons that have run perfectly without adjustment for years now so I'm less fussed about trying to get this to work.

      My main system I run as constant flood to grow salad greens so that doesn't even require a siphon.

      I still think there is a place for this device, but it's perhaps more suited to a solar pump or something with more flow variations than my current system.


  7. This comment has been removed by the author.


Popular Posts