by Kellbot
Air Backflow Problems with T-Joints
6:47 pm in Seeking Advice, Water flow by Kellbot
A few days ago I noticed some problems with my setup: when the pump would turn on sometimes the airflow would go back down through the source tube and start bubbling up the reservoir rather than lifting water up the lift tube. I could fix it by releasing the air pressure in the lines and letting the water re-prime itself.
Over the weekend I didn’t have much time to work with it, but thankfully every time I peeked into my office things seemed to be moving smoothly.
Today when I sat down to do some work, I heard the pump click on and once again the bubbling sound of air coming into the reservoir. I was perplexed as to why it only seemed to fail when I was around.
I didn’t have time to debug / re-prime the system at that moment, so I just ignored it for a few minutes. Then to my great surprise, it suddenly fixed itself and started delivering water to my plants again.
My theory is that sometimes, after the pump shuts off, water runs back down the lift tube and sits stationary at a height above the t-joint. When the system first turns on, the pressure from this thin, tall water column is too much for the air pump so the air starts to backflow into the reservoir. In a few minutes, the water stuck in the lift tube flows back down below the t-joint, and which point air starts flowing up again. The system resets itself and everything works.
Since I use black hose for my lift tube I can’t quite see what’s going on, but I’m curious if anyone else has run into similar problems.
Hi there, I’ve experienced this problem before too. The solution is to move the location of where the air is going in to the air lift tube (air intake). Per my observation, the air intake should be placed higher than the maximum loop of the water intake. I know it’s a little hard to understand so I’m putting in a diagram.
Check it out here: http://our.windowfarms.org/files/2012/03/PLANS1.jpg
Check his diagram http://our.windowfarms.org/files/2012/03/PLANS1.jpg and make X as big as possible and that should solve your problem. Many of those pumps can only pump 2 and a half or 3 ft deep so if you make X to be 3 ft, the air cannot go round the bend and it must go up the tube to your plants instead.
Thanks, I figured the height (or lack thereof in my case) was probably a contributing factor.
I’m trying to keep the reservoirs low for purely aesthetic reasons, so I’ll do some experiments with longer coils of tubing for the water intake to see if that can compensate for a lower X. I have a feeling that ultimately you really need the help from gravity, but maybe I can fudge things a little.
Hi,
I had this problem as well. However, you helped me solve the problem in short time.
Now I am facing another problem. From the t-joint onwards the water is shot up the air lift tube. What should slowly raise as a water bubble starts making sounds as you’re sipping the rest of a drink with a straw. The result is also similar. Less water is coming up. (The airpump is set to minimum air flow).
My theory is that the airflow is too fast. Though, a bigger sized airtube could fix the problem. Does anyone have any experience?
Regards
Georg
Hi, Georg, your t-joint might be a little high (submergence too low) or as you suggest, your airspeed might be too much. Tony has a video where he looked inside an air pump and he thinks that your can throttle the air (after the pump) without damaging the pump. So, I suggest that you put a little control valve just after the pump and dial it down and see if that increases the water flow. It probably will. If you are worried that throttling will damage your pump, you can use a Y or T joint just after the pump. One part of the Y goes to your windowfarm and the other part bleeds off excess air. Slowly open up the valve and the bubbling will slow down and hopefully you will start pumping lots of water with less air. You might only need a valve to bleed to the atmosphere but you can have a valve to both parts of the Y if you like.
Brian
I have also had the problem with air going back into the reservoir. I want to put a timer on it on the power but don’t think I can at the moment because its not reliable. I using normal clear airline. from the bottom of the reservoir to the floor is just under a foot. It doesn’t seem to make much difference how high i make the air intake. However the line running from the bottom of the reservoir to the t-join it not strait and is semi-windy, is it important to have this strait with a smooth curve at the floor. Do I need to make my reservoir higher to add pressure. Can any one offer me any advice,
many thanks
Hi Josh,
My water supply lines are coiled up, about 5′ long, and my reservoir bottles (1 liter soda bottles) are only about 6″ above the t-joint. The biggest factor I’ve found is air pressure. Not enough air gives you bubble-back, too much gives you lines that are loud and gurgling (like sucking up the last of a drink through a straw).
My general strategy is to make a very small change, wait 2-3 minutes, and then change more if it’s not fixed.
I found the t-joint system really difficult to debug without a good air valve (the one that came with my pump was junk), so if you don’t have one you may want to get one.
I’m working on a post that details more of the metrics of my set-up (heights, hose diameter and length, etc), hopefully that will help some.
cheers for the advice, I will give it a little tinker and see what happens. It would be really cool to see how exactly someone else has done it,
Josh – I posted a diagram that shows my setup a little better as part of this blog post: http://home.kellbot.com/2012/05/05/building-a-better-reservoir/
If you scroll down a bit, it also has links to each of the different parts I used (exact product links to the stores I got them from) so you can compare the specs to what you’re using.
Prior to setting up the 3rd column I was using a Whisper 10 air pump, but found it wasn’t enough to drive all three columns at once. The 40 is a bit of overkill, but I plan on adding a 4th column eventually.
I am having this exact problem with my kickstarter window farm. Initially, after the pumped stopped… the water would travel back down. Once the pumped re-started, I then had a bubble-machine.
I clamped the bottom valve in the reservoir and elevated the pump (it is now sitting about 4 feet up. And, by having the pump set fairly high… I can get it to work where it will pump properly each time.
Problem is… it is pumping a lot of water. If I dial down the pump, then I get the bubble machine. Thus, I was wondering what to do next.
Would cutting the air-tubing (pump to reservoir) in half so that the tubing goes vertical (Right now it has a steady incline up to the pump).
And I read things about air-control valves. How do they work? What size would I need to get?
Thanks… still all very new to this =p
I get the very same air back flow issue you speak of. I have been able to resolve it by maintaining water levels and adjusting the t-joint to the right level in relation to the reservoir.