CO2 Reactor Design For Large Tanks

[RickyV]

As some of you may know I am working on setting up this tank 3000 Liter High Tech Planted Tank I am considering different reactor designs and wanted to receive input on them and what would be the best for a 3000l tank.

Here is one design, it is essentially a griggs style reactor from what I understand. I would have two 120cm reactors, 10 cm in width, with a 8000 lph pump splitting its flow into both reactors. The CO2 would be injected directly into the pump to crush the bubbles into microbubbles. The outlet of the reactors would go into the main return pump to the display tank. Thoughts on this design? Surely this is big enough for a 3000 liter tank with 800 liter sump.

The second design is the Reactor

This is a horizontal reactor that works using a CO2 pocket with water flowing down absorbing the CO2. I really like the simplicity of this design but it has not been tested as much as the vertical style reactor like in the first design. I am still leaning towards this horizontal reactor but I wanted to make sure I am considering all the options.

 

[RickyV]

With how simple it is to build reactor I made a small prototype reactor that I will be testing on my 1200 L tank at home. The protoype uses 2 60 cm pvc pipe, with a 20 cm U PVC for a total length of 140 cm. The PVC is 5 cm in diameter (2 in). I am connecting it to a controllable 4000 lph pump. I am first going to test it feeding the CO2 directly into the pump to see if I can avoid drilling into the PVC. This is probably undersized for my tank, but it is much bigger than what I already had so I am okay with that. It will give me a good idea of how this reactor works. If micro bubbles leak out, its either because of of too high flow (which I can turn down), the method for injecting the CO2 into it, or not long enough (which can easily be fixed). I will update after seeing it work. I am also curious how this reactor would work if I stood it up vertically.

 

[xZaiox]

8000 lph is a very strong pump, and to be honest I would highly doubt that you'll be able to use a reactor with the pump feeding straight into the reactor. You'll probably need to make a bypass to slow the flow down, because reactors typically have a 'sweetspot' where they operate most efficiently.

As your design is right now, I think all that would happen is that your pump would push the co2 all the way through to the outlet and it will off-gas. I'm not the most mechanically minded, so I may be wrong, and I would wait for other people to comment. I'm currently using a self-made cerges reactor with a 20" filter housing and it's being run off a fluval 407, which is rated at 1450 LPH. Even at this level, there is some slight bubble leakage where I get a few micro bubbles leaving the reactor. I don't think the reactor could take much more flow before it stops working. I absolutely would not be able to run a 8000 LPH pump through it, that's for sure

Good luck with your project, what an amazing opportunity! I'd love to work with such a large tank. 😄

 

[RickyV]

8000 lph is a very strong pump, and to be honest I would highly doubt that you'll be able to use a reactor with the pump feeding straight into the reactor. You'll probably need to make a bypass to slow the flow down, because reactors typically have a 'sweetspot' where they operate most efficiently.

Yeah its probably too strong. I will likely get a pump where I can control the flow or use a bypass to figure out that sweetspot. I guess with reactors lower flow may be ideal to give CO2 more dwell time.

Good luck with your project, what an amazing opportunity! I'd love to work with such a large tank. 😄

Thanks I am really excited!

 

[Wookii]

Wil be interested to hear the results. How have you modified the lower pipe to maintain an area of gas above the water?

 

[RickyV]

How have you modified the lower pipe to maintain an area of gas above the water?

The two pipes are actually sitting next to each other on the same level. Its essentially still like the original reactor but I curved it to save space. The diagram I made may be confusing but its actually a top view/side view of what is inside.

 

[Wookii]

The two pipes are actually sitting next to each other on the same level. Its essentially still like the original reactor but I curved it to save space. The diagram I made may be confusing but its actually a top view/side view of what is inside.

Oh right I see that makes sense, I thought from the drawings it was vertical.

 

[hypnogogia]

@RickyV I suspect injecting CO2 directly into the pump defeats the whole idea of the horizontal reactor.

Also, looking at your plan, will the horizontal reactor be mounted so that one tube sits above the other, or will it be flat so that both tubes are at the same height?

 

[RickyV]

@RickyV I suspect injecting CO2 directly into the pump defeats the whole idea of the horizontal reactor.

Also, looking at your plan, will the horizontal reactor be mounted so that one tube sits above the other, or will it be flat so that both tubes are at the same height?

Yes that is probably true. I was just curious how different it would work if I did it this way before I drilled into it.
It will be flat with both tubes at the same height.

 

[RickyV]

So I have been running this reactor for a day now and here is the pH profile.

The pH drops to close to 7 by lights at full power and stays relatively stable throughout the day. The reactor does produce some micro bubbles, however not enough to bother me. I personally like having some bubbles to give the plants both dissolved CO2 and CO2 bubbles on their leaves. The bubbles are likely because I am injecting the CO2 directly into the pump creating micro bubbles in the reactor. Another cause is probably the flow of the pump, I am running the pump on the lowest setting but it is still likely too strong. The reactor is also undersized for my size tank. I am going to try adding a ball valve to reduce the flow to something very slow and report back what happens. If nothing changes I will inject the CO2 into the top of the PVC like in the original design. I may also add extra length to the reactor.

For reference this is how much CO2 I am injecting into the reactor.

 

[22802]

So I have been running this reactor for a day now and here is the pH profile.View attachment 206509
The pH drops to close to 7 by lights at full power and stays relatively stable throughout the day. The reactor does produce some micro bubbles, however not enough to bother me. I personally like having some bubbles to give the plants both dissolved CO2 and CO2 bubbles on their leaves. The bubbles are likely because I am injecting the CO2 directly into the pump creating micro bubbles in the reactor. Another cause is probably the flow of the pump, I am running the pump on the lowest setting but it is still likely too strong. The reactor is also undersized for my size tank. I am going to try adding a ball valve to reduce the flow to something very slow and report back what happens. If nothing changes I will inject the CO2 into the top of the PVC like in the original design by 22802. I may also add extra length to the reactor.

For reference this is how much CO2 I am injecting into the reactor.

When we apply the 17.7 ratio (our preliminary estimation for for size of the reactor for any given tank surface), your reactor should be adequate for about 750 liter tank size (depending on dimensions) - not the 1200 liter that you actually have in this experiment.
Indeed you don't hit the 1.5 pH drop that I used as a reference, but it is an interesting new data point that you do hit 1.0 (or a bit more as you don't know if the tank was fully degassed in the morning?).

When we build up more experience as a user group it may be helpful to collect data what the reactor dimensions should be for 1.0 pH drop, as that is more commonly used for the hobby than 1.5. If you calculate your tank surface area divided by the effective reactor surface area then we have our first data point for that

We calculated the dimensions for your 3800 liter project based on 1.5 pH drop, and you may have saved some effort having now an estimation what would be required for your target of 1.0 in that project, while being inherently safe and under no circumstance exceeding this level. I also agree with @hypnogogia that you want to avoid your pump making micro bubbles and that you prefer to inject CO2 straight into the reactor.

Note: the maximum effective surface interface in the reactor is probably less than tubewidth*length, as the bubbles will start escaping when the gas pocket touches the top of the water exit tube. I am sure you thought about it, you may cut an elbow piece so that CO2 starts to escape when the tube is exactly half filled and the surface area is maximised.

 

[RickyV]

When we apply the 17.7 ratio (our preliminary estimation for for size of the reactor for any given tank surface), your reactor should be adequate for about 750 liter tank size (depending on dimensions) - not the 1200 liter that you actually have in this experiment.
Indeed you don't hit the 1.5 pH drop that I used as a reference, but it is an interesting new data point that you do hit 1.0 (or a bit more as you don't know if the tank was fully degassed in the morning?).

When we build up more experience as a user group it may be helpful to collect data what the reactor dimensions should be for 1.0 pH drop, as that is more commonly used for the hobby than 1.5. If you calculate your tank surface area divided by the effective reactor surface area then we have our first data point for that

We calculated the dimensions for your 3800 liter project based on 1.5 pH drop, and you may have saved some effort having now an estimation what would be required for your target of 1.0 in that project, while being inherently safe and under no circumstance exceeding this level. I also agree with @hypnogogia that you want to avoid your pump making micro bubbles and that you prefer to inject CO2 straight into the reactor.

Note: the maximum effective surface interface in the reactor is probably less than tubewidth*length, as the bubbles will start escaping when the gas pocket touches the top of the water exit tube. I am sure you thought about it, you may cut an elbow piece so that CO2 starts to escape when the tube is exactly half filled and the surface area is maximised.

Yeah it is indeed undersized, but I am pretty happy with it. I forgot to mention that the reactor kept spitting out microbubbles all the way until the lights were out which is why I assume the pH increased so slowly after shutting the CO2 off. I will keep experimenting with it to figure out how to maximize its efficiency at dissolving CO2. I'll eventually get to injecting straight into the reactor, what would be the best way to implement that part? Drilling a hole and inserting an airline tubing barb and supergluing it in? Maybe I could do another hole at the end like you have mentioned, but connect some tubing to it and feed it back to the pump.

 

[22802]

If I were experimenting with this one I would do following:

• Reduce water flow to less than 500 litre per hour (my gut feel), just to make sure there is a gentle flow in the reactor, but nothing that creates noise or bubbles. I would use a bypass for this.
• I would modify an elbow piece at the water exit, so that the maximum CO2 build up is exactly half of the tube. With the dimensions that I see on your photo I suspect that CO2 will be escaping before the reactor surface is at its maximum (only a small CO2 gas pocket in perhaps the top 30% of the tube), and therefore using the reactor below its true potential.
• I would let the reactor purge its trapped air (either by purging valve or just let it spit air/CO2 bubbles for a couple of hours), and then give it strong injection with CO2 to test its maximum capacity. Once we see pure CO2 bubbles escaping, dial back CO2 injection until all CO2 gets absorbed and no more bubbles escaping. This will be a good indication of the maximum reactor power, assuming you compare pH with properly degassed water.

Drilling a hole and inserting an airline tubing barb and supergluing it in?

That's what I did, but don't trust me on my DIY skills. I used PVC glue.

Maybe I could do another hole at the end like you have mentioned, but connect some tubing to it and feed it back to the pump.

I don't remember mentioning this? I would definitely not recommend to make the reactor more complicated than necessary, a simple river with water will do the job.

 

[Hanuman]

what would be the best way to implement that part? Drilling a hole and inserting an airline tubing barb

That's what I did in my reactor. I drilled a hole that was slightly undersized and then screw the air hose barb. I then used epoxy all around (inside and outside) to make sure no CO2 or water would escape.

On this pic I still had not put the epoxy.

 

[RickyV]

So I've been thinking a little bit and I think I want to have some CO2 microbubbles in the 3000l tank. I'm thinking of injecting the CO2 directly into the pump to achieve this like in the prototype reactor. Though the reactor on the big tank will be so big I'm curious how much if any will leak out of the reactor.

Edit: I guess if I can probably control the amount of microbubbles using flow, more flow = more bubbles.