Hybrid venturi/vortex aeration system for RDWC

Been messing about with a hybrid venturi/vortex aeration system for RDWC, for a number of reasons

Keep in mind this is just a ‘proof of concept’ and not something I would put into use as you see it.

It consists of the tub container (rez) with a 4000 l/ph 30w submersible pump at the bottom. It pumps up through the venturi fitting at the top side of the bucket which is offset so that it creates rotation. The bucket has a 1.5 cm hole in the centre bottom that drains back into the reservoir via a ‘vortex’ or whirlpool.
The clear tube you can see running to the venturi just lets you elevate the air intake so if the venturi becomes immersed it will still suck air in.
As the water passes through the venturi it pulls air in from the tube and feeds the aerated stream in, driving a whirlpool in the process, if you tune the outlet and/or the drain hole then it is possible to get it so that the vortex collapses just prior to going down the drain hole. This is best for aeration.

Some benefits are:

  • Easy to scale up and can be added into the reticulation loop of an RDWC setup, for small setups air pumps are all great, but when it’s dozens etc it becomes a PITA. Other than the pump there is nothing here to go wrong.

  • No air stones required. So no more cleaning, replacing air stones or buying expensive scintered silicon dioxide stones so you can get more than a few grows before they dissolve.

  • Less noise. It requires a some tuning but this system is virtually silent.

  • In theory both veturi and vortex aeration are more efficient than air stones, although I don’t have a dissolved oxygen meter to test, the combination of verturi and vortex should be more efficient again.


Interesting idea, i like it. The only concern i’d have would be how do the roots handle the circular motion and the water force created? I can imagine roots twisted into a spiral and being constantly pruned by the water pressure? Root pruning isn’t so much a issue (thinking fabric/airpots) but the spiraled roots may? Like i said, interesting, and i hope to watch further development.


Yep I get that, tbh mate I had not even thought of having the roots directly in the ‘vortex’, though it’s amazing what they will tolerate it’s probably worth trying! :smile:

So the idea is that the roots don’t sit in the bucket part you can see directly, instead you have another circulating pump between the aerated rez and the buckets. All this arrangement does is oxygen saturate the water in your rez tank.
The circulation pump feeds into a manifold type distribution point to aerated water at a normalised rate to each bucket in the DWC setup. . I’ll work it out eventually so that the bucket overflow gravity feeds back to the rez, but I haven’t got to that point yet. The idea is to fully oxygen saturate the nutrient solution and then distribute it to the buckets, to properly evaluate it I really do need to invest in a dissolved oxygen metre.

Depending on how saturated you can get the water I had considered adding another venturi aerator in this ‘circuit’ as well. I’ll do up a schematic of a ‘scaled up’ version so it makes a bit more sense but think of what you see as a ‘mini’ version:)


Cool beans. I like venturi injection concept, in part, for the benefits you’ve noted.

Here is an attempt at using venturi injection along with a needlewheel pump.

Data collected in some trials has shown relatively fast DO saturation (and possibly) exceeding 100% DO for the temperature / atmospheric pressure.

How about starting a thread on your work / discoveries?


Hey that is very cool. Is this a setup being used? If so does it work well? I was just going to try with a venturi, but I have done a fair bit of bio brewing for my garden and I have used a vortex aerator and found it to be really effective at aerating.It’s the simpest thing in the world really and I have so had enough of airstones and air pumps! I am thinking I could use one of those needle wheel pumps. They are new to me, thanks!_


It certainly seems to do well on the DO, at least what I’ve measured so far. I have yet to evaluate the DO using the venturi + needlewheel during a grow cycle to validate the applied performance to the DO but I’m fairly confident that it’ll do well comparatively.

I think I read about the vortex technique being using in some scientific journals where the main end-application is for the aeration of large bodies of water. From what I recall, it has been shown to be an efficient technique.

The majority of the DO experiments that I’ve seen in the various forums, using the various techniques such as waterfall, air stone, etc, have run the experiments stand-alone. But, they have not measured the DO while the plants are in the system. The rate at which DO is depleted and the rate that the DO is replenished in actual application, I think, could be illustrative.


Yep vortex aeration seems what they use in water treatment plants and large fish farms (along with large venturi aerators) where sparging with air stones is not practical. The studies I have read it is just as or more efficient in comparison to any other method and I like the simplicity. And what’s not to like about a stable whirlpool?
For Mk. 2 I was thinking I make a taller and narrower vertical tube so the vortex is both higher and faster, this obviously increases turbulence and water/air contact time so it ‘should’ be much more efficient. For this experiment I just wanted to see if it was straight forward to get it circulating with a stable vortex. Turns out to be easier than expected. I have a 2 bucket system I have to setup for a friend on the cheap. I am thinking he might be the perfect crash test dummy :wink:

For even distribution of water/air to the buckets I have previously set up a system as your diagram indicates, however the I fed the circulation pump output into a central 6 inch ‘manifold’ pipe ( just regular 6 foot long PVC drain pipe) that was capped on each end and had an 1 inch outlet plumbed in for each bucket. The pump outlet connects to a inlet fitting on one end of the capped pipe. Then a 1 inch tube connects the manifold to the bottom of the bucket, so you get the nice upwelling effect I see in your video. The manifold works exceptionally well at balancing pressure so the buckets all fill and drain at the same rate.
I put an overflow at the top of each bucket that is plumbed so that it gravity feeds back into the rez. This setup still uses an air pump but I think If I used one of these skimmer pumps I could potentially get rid if the air pump altogether.
I would be very interested to see how fast the oxygen is consumed by plants, I can’t see it being any more rapid than fish would deplete it I would have thought?


Looking at those pin wheel pumps, it occurs to me the pinwheel idea could just as easily be passive and do the same job of breaking up the bubbles. As in the water upstream of the pump could simply be forced through a pinwheel impeller. Then the pumping could be done with a more efficient impeller design, you would be able to push more water around and therefore suck in more air. I just disassembled one of my submersible pumps to have a look and it woudl be reasonably easy to modify it to perform is a similar way. Am thinking I might have to fire up the 3d printer. :slight_smile:


Your experiment deserves it’s own topic…

Yes, the higher the velocity past the venturi, the higher the vacuum potentially generated.

I’d agree, the smaller the bubble size, the better.

I don’t think a passive technique of breaking up the bubbles would be able to create as “small” of a bubble size, though.

For some aquarium pumps, you can purchase the different impellers (including pinwheel) and swap them out. FWIW.


I just use 2 power heads in a 20 gallon res. No need for airstones and I don’t use the air feed. The circulation is all that is needed.


For reference, here are a couple of plots from some tests I did last year on DO (no plants) under different operating conditions:

For these, there were four modes of operation:

  1. pump speed set to low with the venturi off.
  2. pump speed set to high with the venturi off.
  3. pump speed set to high with the venturi injecting air.
  4. pump speed set to low with the venturi injecting air.

The initial “high” DO is from the municipal source supply, which, as can be seen, eventually out-gases. Just having the solution moving and fluming does not necessarily saturate the DO by itself. Oh, take the plots with a grain of salt since getting accurate/repeatable readings of DO takes a bit of detail. During some of this collection I was still getting acquainted with what to do/not do. Just thought you might be interested in seeing those.

The data collection and testing that I was working on (with help from @Larry3215) was put on pause “temporarily” as I had more pressing items to work on. But, now that I see you’re looking at venturi injection, my interest has been peaked again. I’ll probably look to get back to experimenting with this in the near future and to compare notes.

Side note, I was also doing some experimentation with the addition of surfactants to the solution along with the aeration. That has some potentially interesting results, as well.


That is so freaking cool. So in your plot ‘air’ indicates the active venturi? Interesting you got slightly divergent results at the end of the plot with the pump on low, I wonder what that’s all about? It’s obvious this is fully saturating the water, and given it’s going to hold the same amount of oxygen regardless of the aeration method it’s hard to see how this won’t function every bit as well as any air pump. It’s not like there is much of a lag in oxygen uptake even. It woudl be really interesting to see this next to a airpump aerator.

Having now spent a good few hours looking at venturi design and pinwheel pumps, I am thinking I might have a crack at making some in Inventor and 3d printing them out. I have a 10000Lph pump with busted impeller that would make a great bubble thrasher.

II found a reasonably good technical paper on the “analysis on the effect of venturi tube structural parameters on fluid flow” which runs a number or test on the elements that influence both velocity of fluid flow and probably more importantly vacuum.


It’s fairly maths heavy if you want to really get inside the actual fluid dynamics of it all, but you don’t have to in order to glean a lot of useful information. I sure wish I had paid more attention to fluid dynamics when I did engineering!


I was staring at all of my different airstones and diffusers the other day and I thought – I really should do a comparison of these to see which one works the best at aerating the reservoir. Unfortunately I don’t have a DO meter right now so I started looking around to find one.

Any suggestions, @Northern_Loki? Cheaper is better but I would take accuracy over cost.

I also have my venturi from my E&F setup so I could do a good side by side by side comparison. I’d love to see time to total saturation for each of these so we could truly compare. I know the venturi worked well for me last time, but there’s more data to consider here and I’d love to see it all. :slight_smile:


A friend has used a similar but more basic system for years. A water pump circulating the res. The intake line for the water pump sits on the bottom and has a small tube branching off that goes above the surface to get air.

I wouldn’t say it makes a vortex, but a ton of air gets in there very simply and cheaply.

Interesting design!


Um. Wow. Just to take the effort to do this. Now just do a run in each of these scenarios and see how it affects yield. Haha. Just kidding…good work.


Yes, correct. The venturi, in this case, is nothing more than some polypropylene tubing jammed into the pump feed line with a value to meter the air inlet. I need to construct something better than just a tube.

Part of what I was trying to evaluate was whether we could super-saturate the DO. From the trials, I did see positive signs of super-saturation on the order of several percent. But this needs to be re-evaluated carefully as I also saw some conflicting/confounding measurements (as can bee seen in the graph trends).
Yes, it is possible to achieve >100% saturation using active methods. Which techniques to use, and within our grasp, to achieve that is the question. The extreme end is nano-bubbles that collapse in solution under pressure achieving a DO that is much much greater than 100%.

I agree, I think other mechanisms will easily compete with airstones. And, could end-up being less costly along with other benefits.

So, for those graphs, they have a fairly long time scale. When the venturi is allowing gas into the system, I was seeing about 30 minutes to fully saturate 40 gallons of solution. It is fast relative to fluming (air off). We need some other experiments to see if there is actually any merit here versus vortex, airstones/air pump, waterfall, etc, etc…

Very interesting. Looking forward to seeing how this works.

That is awesome, I need to study this. Thank you.


:fist: Data geeks unite!

So for DO probes, there are two primary types. Polarmetric and luminescent(LDO). The preferable type is luminescent. Polarmetric is touchy, sensitive to solution velocity, and drifts. But, it also the least expensive.

What I have is a Hatch LDO probe along with a Hach meter. Their stuff is very expensive although, by being patient, these can be had via Ebay. I ebay’ed a used LDO probe and simply purchased a new LDO “cap”. From what I recall, I was able to get the probe for ~$100, the replacement cap ~$100, and the meter for ~$200. So, yeah, even used it is pricey. If you are willing to spend some sweat equity, industrial meters/probes can sometimes be had on the cheap.

A polarmetric Milwaukee DO probe + meter are available for under $200. There are downsides to the polarmetric method but may be perfectly suitable depending on how it’s going to be used.

For on-line process, the newest generation of LDO require no calibration ($$$), Hach LDO® Model 2. But, money to burn?

For what I’m doing, the probe needs to stay stable for long periods of time while submersed. LDO is better for that. And, I can say that DO probes in general are a bit tricky to use/understand in practice.

That would be excellent. If we can get folk with various set-ups measuring the DO %, rates, plants in-situ / ex-situ, we’ll have data that well exceeds what is currently available on the interwebs.

Understanding, measuring this is fairly difficult and expensive. It requires a certain amount of data geeky-ness to get sensible results. Right now, viewing the past experiments available from other sources, I feel their results may be a bit “misleading” as they avoid looking at things like saturation rate, etc. A comparison using an improved experimental methodology would be valuable.


If someone wants to let me borrow a meter I would have zero problems getting some data for the interwebs…and considering I have several different setups and I would like to know how they run vs each other…as I have a superponics cloner which is a dwc with top feed…I’m building a new cloner which will be done in 2 days when all the parts get here which it will be low pressure aero and fogponics too…so I can of course do one or the other and both…my veg setup is current culture rdwc but it’s upgraded and im running a alita60 with 2 large air stones per bucket for a 4 site 8gal…and my bloom setup is 4 site rdwc 27gal totes 3in pvc waterfalls I have x2 350gph pumps and my return pump is a 700gph 3in pvc I also have a drain kit I built which leaves 200ml in ea tote only…same setup for veg for the drain kit which they help with circulation to…so if anyone has a d.o meter that would let me borrow I wouldn’t mind getting data for everyone


Excellent idea!

I have been trying to keep my rez cool, so I am using an ice chest and adding frozen water bottles. But, with the ice chest lid closed up tight, the rez wasnt getting enough fresh air, and the water got nasty. Ive been playing with different ways to allow more air in while keeping the light out and keeping things cool, but none of them worked perfectly.

I have a 300 GPH aquarium pump fluming the rez. After reading this thread, I just stuck a section of air line tubing in the intake of the water pump, and ran it out side the ice chest through a small hole. Now it sucks in fresh air through that small tubing and the impeller in the pump breaks it up into small bubbles every time the pump runs. Extra freash air with no light contamination and I can keep the ice chest sealed tightly for less heat. The bubbles are not as small as you would get with a needle wheel impeller, but good enough for what Im doing. About the same size you get from a regular air stone.


Add a pvc drain pipe that drain from the top. Like a over flow on an ebb n flo table. Then insert venturi valve at the bottom. It creates a wicked venturi effect.