by Kevin Wells
How to set up a t-valve airlift.
1:52 am in How-Tos, made from scratch (without a kit), Materials and Resources, posts with pitcures!, Version 3.0 Modular Airlift Columns by Kevin Wells
First, some background. This is my first window farm. I have no prior experience in hydroponics, but have grown many aquatic plants. I started making a DIY window farm following the directions of a single-column, 5-bottle hanging V3 window farm. I found that the bicycle needle airlift method just was not as reliable as I had hoped. It would sometimes work, and other times, I would find it not working at all. I researched the site and found that others had set up a t-valve airlift, and it seemed like the way to go. The other guides did not seem to include all the information I needed to get it set up for myself, so I decided to try it anyway. Below, I’ll tell you what I used. I will also say that if this method seems ridiculously simple and it looks like it’s so short that I might be missing something, it’s because it is ridiculously simple and I’m not missing anything… I think.
Parts needed (in addition to the other parts used for the V3 hanging window farm):
Standard aquarium airline. I went with black silicone, because it looks nice and will stop algae from growing in the airline. I bought 25 feet, because it was cheap and I will probably use more when I add columns later.
T-valve. I purchased a metal t-valve from PetSmart. They have plastic ones for even cheaper.
Silicone glue. You want to make sure you get silicone glue that is 100% silicone. I got mine at a hardware store, but they also often carry this at pet stores/fish stores (for aquarium repair). The 100% silicone will ensure that there are no additives that could leak into your water and plants.
Steps taken to add the t-valve airlift to my V3 hanging window farm:
- To make the sport cap airline connector, first insert 1″ of airline into the sport cap of the water reservoir.
- Inside the sport cap, use the silicone glue to glue the airline in place. Make sure to form a complete seal. It must cure for at least 3 hours before you can get it wet. I recommend you let it cure for 24 hours before doing anything else with it.
- Measure/cut 1.5 feet of airline from the cap, and connect this to one of the two straight ends of the t-valve.
- Connect your airline from your air pump to the perpendicular end of the t-valve.
- Using your remaining airline, connect one end to the remaining straight end of the t-valve.
- Run this airline to the top of your window farm and into the top bottle. Secure using zip ties or what ever you prefer.
- ???
- PROFIT
Look at my awesome diagrams:
Thanks for the credit. Have you measured the inside diameter of your airline? (I use drill bits to measure the insides) . Also, a word of warning, I bought blue silicone airline and I found it almost unusable. I think it was because it is made of “silicone rubber” and water “sticks” to it. I think 3/16 or 1/4 inch are the best sizes for the internal diameter of your tubing.
I bet your airline is not silicone rubber! Thanks
Brian.
Brian, I have not measured it, but I do know that it is smaller than 3/16″ ID. It is standard diameter airline tubing for aquariums. It is black silicone airline, and I would not say it is unusable at all. It does not slowly drip, however; the water that is pushed out seems to come out more like small, sporadic pours. I am trying to find a good balance between noisiness and consistency of the water output. The higher I turn up the air pump, the more consistent the water drips are, but the output is extremely noisy. The lower I turn down the air pump, the more sporadic the water drips are, but the output is almost silent.
I think I will stick with the more silent, sporadic pours and use a timer to control the air pump. I hope that this will get the amount of water flowing over my baskets that will work best without having a ton of sputtering noise.
I also forgot to mention that I do have a check valve, with the output facing down, inside the sport cap of the reservoir. This may help prevent any water from being pushed back down, but I mainly used the check valve so the silicone was easier to apply to seal the cap.
Nice collaboration, guys!!
Hi, Kevin, it might have been that my silicone tube was brand new and there was something in it from the manufacturing process. An oil coating perhaps? Maybe if I had ran water through it for a while it would work fine.
The really scientific thing to do is compare 5 or 6 different tube diameters at all the different heights that we need for a week or 2. But who has the money or time to do that? There is another thing, the science of 2 phase flow. It is full of equations that would make your head hurt and non conforming data in narrow pipes too. (Which means it isn’t completed science!). From my reading of it over the years, and experiments too, I think that the most efficient flow is just a bit faster than “the slowest that works”. I still think your tubing is too narrow (you should size your tube to the output of the pump). I did experiments years ago (1990) and the best airspeed through the pipes was between 0.7 and 1.5 meters per second. (For 12 mm (half inch) and 19 mm pipes). That was for my “pulser pump”. I guess best airspeed will be a bit slower but similar for smaller tubes. In general I think smaller tubes would give more perfect “plug flow” IF you didn’t get a massive increase in turbulent reaction with the SIDES of the tubes as the diameter decreases. But you do get this increased turbulence because area of tube is related to the square of the radius (pi R Squared) and tube wall is directly related to the radius. (2 pi R.)
My tests last spring showed that 3/16 was pretty good, windowfarms settled on 1/4 inch. (Their pumps were probably a bit stronger than mine)
Your tubing is about 3 mm, so the “turbulence ratio” (if I can make up a term) is about 1/3. For quarter inch (6 mm ) it is about 6/36 so 1/6. So they would be getting about half the turbulence that you are getting! Turbulence is one of the causes of the noise and gurgle. In my case, way back then, half inch pipe “looked” so much better!
It squirted water out in quick powerful pulses with lots of noise.
BUT when I actually measured the output, the quiet 3/4 inch pipe was outperforming it by a wide margin!
I don’t mean to hog the conversation but I just thought, if you wanted to try a larger diameter airlift tube to go to the top, you only need to replace the tubing above the t-joint. All the rest can stay as is and your narrow airline delivering water to the t-joint actually helps even the flow of water into the airlift tube. Try it and see, it is just a minor change but it will probably give a big improvement. You might get the tubing at a hardware store or at a beer and wine maker store. I bring a drill bit of the right size to help me choose. You probably need 3/16 or 1/4 inch tubing.
Sometimes I have left a short piece of thin tube on the end of the t joint and press fitted the wider tubing onto it.
Thanks
Brian
Does anyone have any advice for someone about to attempt the construction of this model?
Joel,
All of the work I’ve done on this is listed here. Did you have a specific question about my setup?
No matter what I do it appears that the air wants to bubble back into the water reservoir. I’ve tried everything, different size hoses, nothing seems to keep it from flowing the wrong way. I put a check valve between the reservoir and the T-joint and that helps, but now the water just flows too slowly. Any other thoughts?
@BK Do you have a variable speed air pump? You might try decreasing the air flow. Also try raising the location of your t-valve, so there is less much water flowing upwards at any given time.
Thanks for replying so quickly Kevin. I do…well…sort of. I have a control valve on the output of my pump. It works great for adjusting the flow and it actually has two outputs (I use the second output to pump into an airstone in a second gallon of water to release the chlorine in our water over the week).
I think I may be too critical on myself, so I want to explain what I’m seeing right now. I have columns of water about 2-3mm in length/height, they are spaced about 2-4cm apart from each other, and each little column takes about 8-10 seconds to reach the top of my 5 column (4 plants, 1 reservoir) setup. There is actually quite a bit of water flow through the whole setup, it drips constantly. Does that sound about right?
That sounds like it is performing exactly as it is supposed to. The water is supposed to drip constantly. I think most people use a timer on their air pump to have 15-minute intervals, so water is not dripping nonstop. I think you just need to find out what works best for you by trial and error.
Thanks for the encouragement Kevin! I think this sort of “drips per minute” recommendation should be added to the instructions to help people understand when they have succeeded. I was trying too hard!
Video: http://www.youtube.com/watch?v=LDdaB93p3n0
As for the timer, why do we need to turn it off at all? Why not just let it run non-stop 24×7?
Honestly, I’m in the same trial-and-error stage as you. I don’t know what the drips per minute are supposed to be, and I only believe it needs to be turned off to prevent the roots from getting *too* wet. I think it’s because they need to get some exposure to air, but I’m not really sure. Maybe someone else who is more familiar with the process can answer your question. Like I mentioned in my post, this is my first time delving into hydroponics, so I’m here to learn too.
As frustrating as the airlift system has been, I’m having a blast.
I’ll post the drip/second question top level.
Oh yeah, and thanks for this initial post. Really kicked me over the edge in getting things working. I was pretty close to abandoning the needle valve method!
Improvements to your design: check valve from reservoir to T-valve (to assure bubbles ONLY go up), check valve from air pump to T-valve (to make sure the air pump doesn’t get borked).
I will definitely put this on the list of changes/improvements that need to be made to the directions, guys. Good work. I’m following! -Britta
Hi, BK, If you can place your pump higher than the reservoir level, no check valve is needed because the nutrient water cannot flow uphill. (I accept that many people cannot do this but in industry it is part of failure proofing a design). If you put the pump higher than the reservoir, it saves the cost of the check valve. Also, Check valves can fail and let water back into your pump but gravity will not fail!
Also, if you T-Joint is wide enough inside (about 4 mm) water can flow down through it and this prevents airlocking and means the the second check valve is not needed either. Kevin has his T-Joint and the pipe arrangement around them pretty well perfect. I am pretty sure that he will pump more water (and quieter) if he uses 3/16 inch tubing or 1/4 inch tubing for the reservoir up to the top of the airlift especially if he can find a T-Joint that is 3/16 or 1/4 internal diameter.
Thanks Brian
Brian, here is what I did to make mine perform better in the last few days:
I shortened the airline above the t-valve by about 6 inches. That’s it! This raised the t-valve to be about level with the bottom of the reservoir bottle. Now that there is less water trying to push up above where the air flows in, the water drips out way less erratically. I can now run my pump at the lowest setting and have water drip out regularly, with very little, if any, noise at all.
Hi, Kevin, having to shorten the airline (reducing the submergence) and putting your pump at its lowest possible speed is consistent with the tubing being a bit small for your pump. Basically your tube is at its maximum carrying capacity.
Hi.
So I built it and it works sometimes. But most of the time bubbles just come up from the tube into the bottom of the reservoir. I know I’m on the verge of a functional design since it does sometimes work – any advice to make it work all of them?
Once again, the problem is that my pump is just pushing bubbles up through into the bottom reservoir.
Nevermind I think I got it! I looked at the diagram and realized I had the tubing plugged into the wrong parts of the t-valve.
This design is sweet thank you guys.
Joel: Do you have a picture you could show us? Is there a loop at the bottom, below the t-valve? There must be a loop to prevent the bubbles from pushing down instead of up. You could try increasing the size of this loop, if possible. You could also try raising the height of your t-valve.
Increasing the loop size below the t-valve seems to be helping. I’m not sure I understand why that worked, but it seems to be the answer.
Though I will say that the flow is significantly less now. That said, it consistently works… I think.
Here’s my entirely non-scientific explanation: It works because the air will flow along the path of least water pressure. Increasing the loop size below the t-valve increases the amount of water that the air would have to push through to go backwards through the reservoir. If this amount of pressure is greater than the pressure of the water already in the airlift above the t-valve, the air will only flow upwards.
Word that makes a lot of sense actually.
I also had the problem with bubbles going back into the resevoir. What I did was attach a check valve on the tubing that leaves the bottle. If you try to blow air into a check valve, it only works one way, so I figured the same principle would apply here. Btw, good job Kevin. Motiviated me to get off my butt and start this windowfarm myself.
We had a lot of trouble with the bicycle needle airlift (although it’s working well now), but this seems really simple so I’ll be trying it for our next column. Thanks a million!
You’re welcome! Just as an update, I’ve found that the height of the t-valve is very important in relation to the flow of water. If the t-valve is too low, too much water will flow out of the airlift (and the air may possibly flow backwards into the reservoir). If the t-valve is too high (above the reservoir), water will not flow at all. Since posting this, I’ve tweaked the height of my t-valve to perfection, and get fairly consistent slow drips from my airlift with my pump on the lowest settings. If you have problems with your water flow or with air flowing into your reservoir, make sure to try adjusting the height of your t-valve! It will make a huge difference!
I’ve had similar luck with this design. Works fantastic and flows very very well!! My post with picture is here: http://our.windowfarms.org/2011/12/18/my-airlift-system/
Because it is discussed above, I thought I would add my experiences.
I recommend both a quality check valve and a pump that is higher then the reservoir. The check valves that came with my windowfarm kit are GREAT (hopefully the brand is listed somewhere)! However the cheap ‘spring’ based check valves I bought at a store DID let water flow back into the air pump tubes when their was enough pressure still in the airlift system above the T-valve (though the water backflow was really small and slow).
Without check valves I DID almost experience backflow of water into my pump even though it was higher then the reservoir (actually, I caught the water just as it was entering the pump, luckily). So I guess if you have ‘water and air pockets’ above your t-valve and if there is some odd air pressure in your tubing to your pump, you might run into water issues flowing back into your pump. My system seems to behave very odd with issues of this sort, but I just wanted to list this as a caution to everyone to keep an eye on things before you assume your pump is safe from water
Thanks for all the great info above, it really helped me out.
how much flow can I expact from an T-valve airlift?
Hey guys, thanks so much for the diagrams and instructions, super helpful! I’ve written up a little safety/failsafe guide at http://our.windowfarms.org/2012/01/01/ideas-for-safety-and-failsafes-holiday-mode/, hope this helps to prevent minor and major accidents.
Matthew, I had exactly the same issue with spring based check valves, the water flows back into the pump. Thankfully it survived a couple of drips! Hung the pump about 20cm above the water level now, as Brian said: Check valves fail, gravity won’t. I’ve experiences smaller pockets of water being pushed up higher than the water level towards the air pump by the air pressure within the tube, so make sure there’s a bit of “safety height” in your pump setup.
Kevin, thanks again for your diagrams – but I think its dangerous to only show part of the setup and leave the pump out. It took me a while (and lots of comment reading) to figure out that the pump isn’t in your pictures, and the red drawn line goes up towards a higher pump again (off picture). Given how popular this method has become, I guess you get some people giving it a shot based on the original post only – and might end up electrocuting themselves with a air pump sitting in a puddle of water, right?
I’ve built my one-column MAMA-style using t-valve. I bought a pair of check valves too, but according to my excessive tests, they’re not needed. I found these rules apply (I haven’t tested all of them ofc):
- Plastic 3/16″ t-valve works pretty well. Connecting a 1/4″ tube is easy: put an inch of 3/16″ tube to the t-valve, and pull the 1/4″ tube over it.
- 3/16″ ID tube is good enough for almost everything. In the end I’ve replaced airlift tube with a 1/4″ ID one, which produces more balanced drip flow, but no drip coffee maker noise.
- Silicone is not perfect for sports cup. Let it cure for a day for better results. Glue a shorter tube to the cup (I used a two inch 1/4″ ID tube for that), and attach a 3/16″ one to it with plumber’s tape. This makes the sports cap easily replaceable in case of failure.
- for me, a 1/4′ loop was not enough. I ended up using my first airlift tube (4′). Maybe a shorter one would be enough, though.
- when you flush water, be sure tubes don’t have water bubbles, or airlift won’t work. Run air pump while you wad loop tube with you fingers. Never ever let water bubbles go to the air tube. These bubbles can be pushed up to the air pump (see next item).
- make a loop of the air line, and hang it above reservoir water level. Once the pump stops, water flows into the air tube up to the reservoir’s water level. The loop comes useful against waterdrops *outside* of the air tube too.
- never put extension cords to the ground. Put away all electric stuff from the ground too. Water conducts electricity just enough.
I believe t-valve is the easiest method for pumping water, requiring less materials and manual skill.
Wow!! I saw the physics of this almost immediately. What a great design. The location of the T-valve relative to the top of the water reservoir is probably the key. Too low and the water pressure puts too big a spurt into the post-T-valve tube and the air pump will lift it too slowly. Too high and the water pressure puts spurts smaller than the air pump can easily handle thus wasting effort (and maybe making lots of extra noise?). I would guess that variations occur as the reservoir level rises and falls with the rate of water return. This design actually resolves several problems I was having in a design for big tanks.
Also, regarding the need for rinse vs soak: roots need to get oxygen in more volume than nutrients and moisture so they need lots of access to air.
This is the greatest site I have encountered in several years of work and research. Thank you Britta for the TED talk!!!
Make that “oxygen and nitrogen in more volume than nutrients”.
thank you guys for the t-valve air lift design!!! i was going crazy fiddling with the inflation needle. i know it works for some, but i just couldn’t get it to. the t-valve was the perfect solution. i now have 4 columns running smoothly and consistently, and just added my seedlings to the system. i am really happy right now
You are amazing
! Finally a fully working system. The airneedle system was leaking all the time, but this directly worked. Now I used 6mm OD airline, but I prefer the 9mm OD airlines. And with those it is not working yet. But I already experienced that once there has been some water in an airline for a few diays within the sunlight it starts working better… I hered because of the start of algae growing in it… We’ll see if it works 
.
Thanks!!!
Thanks so much for these directions! I have had so much trouble getting my airlift system to work! I will be getting the extra supplies tonight and giving it a shot. Seems like this way definitely works though!
Brian, I read your comment about larger diameter tubing, and I have tried it. I used a larger diameter when I tried getting the air needle method to work. The water does not have enough surface tension to maintain a bubble in larger tubing. The force of gravity will overcome the lift of the air, and the air bubble will push past the water bubble, and the water will sink back down. This happened with anything that was 1/4″ ID or larger.
My kit arrived last month, and I had not been able to get it to work. The lift system was bubbling air back into the reservoir even at the lowest setting, and obviously was not effective in getting water to circulate. After reading many of your notes (thank you, thank you!) I put some packing tape over the end of the t-valve and secured it with a small vinyl hair-band (It’s ok to laugh – it worked!) and made a tiny hole it in for water to enter. Now the system is working just fine!!! I too was considering sending the whole thing back! Perhaps the next iteration of the kit could include a small plug in the free end of the t-valve that has a tiny hole in it.
Lori