Horizontal CO2 Reactor - Yugang 鱼缸 Reactor

[Yugang]

I am not 100% sure what is going on, but can share my observations and hope it adds a piece to the puzzle.

I am assuming here that during the day you have no pumps switching on an off, that may affect CO2 outgassing in tank or sump?
Are you using a precision regulator, or 'reactor overflow mode'? (Obviously not a pH controller, otherwise you would definitely not expect this pattern)

As you are aware, the CO2 ppm in a non planted tank (so we exclude CO2 consumption or generation by plants or life stock) is a balancing game between injection (reactor) and outgassing at the surface. Now if we bring in the plants, their consumption at lights on will lower CO2 ppm and therefore increase pH (a bit). Generally we assume that plant CO2 consumption is much lower than CO2 outgassing, and therefore the pH increase when lights on may be visible but is probably limited. This is different when the balancing game involves a relatively big plant consumption, as compared to CO2 loss through surface outgassing. In my tank, I did numerical modelling on that a couple of years ago, plant consumption was less than 5-10% of surface outgassing and the impact of photosynthesis was less than 0.1 pH. In your tank, with probably much more plant consumption this may be different and this may be the reason that you see so much pH drift when lights go on. You could reduce this, have more stability, by increasing your surface agitation (outgassing) while injecting more to achieve the same pH drop (as compared to fully outgassed) at lights on. You then created a situation where plant consumption is a lesser proportion of the total injected CO2. Does this all make sense?

Note on the above: the fact that your tank seems not fully outgassed in the morning (only 0.4 pH drop equivalent outgassing during night), may be another indication of relatively low outgassing / surface agitation.

With regards to absence of algae I can add that I have thrown very bad CO2 experiments at my tank (testing various injection methods), but to my surprise never really got any BBA algae outbreak from that. CO2 experiments do kill plants, but my hypothesis is that algae are mostly caused by unhealthy plants and their organics, and that a well maintained tank has a lot of tolerance for CO2 variations until BBA breaks out. I may be wrong, but these were my personal observations and interpretation.

I hope this is helpful, although not a full explanation/solution. I would increase surface agitation, or simply use a pH controller. When using a pH controller, the inherent safety feature of the reactor will mitigate the risks often associated with that.

Note: it may be useful to try one day not switching lights on, and compare the profile to a normal day with photosynthesis

 

[Unexpected]

I am assuming here that during the day you have no pumps switching on an off, that may affect CO2 outgassing in tank or sump?
Are you using a precision regulator, or 'reactor overflow mode'? (Obviously not a pH controller, otherwise you would definitely not expect this pattern)

No pumps are turning on only the lights. The reactor (if you mean by over flow mode) will only allow a 1.5pH drop then will purge. My regulator is not precision, but it's a GLA brand and I would assume this brand is semi stable. The flow through the counter is visually consistent (although not much merit here) throughout the day. The regulator is a potential weak point for sure as I found another leak and have since fixed it, again. I don't run the pH drop with the controller because overflow mode works better and my sand seems to raise KH just a smidge through the week. I'm hesitant to ever let the controller decide as the pH range is unstable.

Does this all make sense?

Yes it does.

Note on the above: the fact that your tank seems not fully outgassed in the morning (only 0.4 pH drop equivalent outgassing during night), may be another indication of relatively low outgassing / surface agitation.

This seems weird as my tank has several points of over flowing water into chambers in addition to the surface agitation I create with the return nozzles.

Actions I will take today. I will recheck for leaks, reduce on CO2 start time and increase injection rate as I'm mostly positive over flow mode calculated into length and diameter are near perfect with this reactor.


My thinking, and please correct me because surly I'm wrong. Do you think this may have something to do with the over under design? Meaning, and I'm just thinking here, The top reactor fills first and the second reactor would only fill with purge from the top reactor. Maybe the injection rate is too low and the lower tube depletes throughout the day (for reason I can't think of because this would defy physics) and this would be the pH increase we see with the trend upwards?

 

[Unexpected]

Here’s the surface agitation in the tank.

 

[Yugang]

No pumps are turning on only the lights. The reactor (if you mean by over flow mode) will only allow a 1.5pH drop then will purge. My regulator is not precision, but it's a GLA brand and I would assume this brand is semi stable. The flow through the counter is visually consistent (although not much merit here) throughout the day. The regulator is a potential weak point for sure as I found another leak and have since fixed it, again. I don't run the pH drop with the controller because overflow mode works better and my sand seems to raise KH just a smidge through the week. I'm hesitant to ever let the controller decide as the pH range is unstable.

Yes it does.

This seems weird as my tank has several points of over flowing water into chambers in addition to the surface agitation I create with the return nozzles.

Actions I will take today. I will recheck for leaks, reduce on CO2 start time and increase injection rate as I'm mostly positive over flow mode calculated into length and diameter are near perfect with this reactor.


My thinking, and please correct me because surly I'm wrong. Do you think this may have something to do with the over under design? Meaning, and I'm just thinking here, The top reactor fills first and the second reactor would only fill with purge from the top reactor. Maybe the injection rate is too low and the lower tube depletes throughout the day (for reason I can't think of because this would defy physics) and this would be the pH increase we see with the trend upwards?

We have an interesting puzzle here, and your surface agitations looks good to me

First, to make sure the reactor works correctly in overflow mode I believe it makes sense to do some tests with a high enough CO2 flow rate so that you see some small bubbles every now and then being blown in the tank. You are then absolutely sure that it is not your regulator drifting, and that both reactor tubes are used at the full capacity. This will give some CO2 losses, but you can later dial back injection and find the 'sweet spot' where losses are minimised but reactor optimised at its maximum capacity. For my reactor in my tank I now know that this optimum is about 7 bubbles per second, and I hear a slight trickle from small 1-2 mm bubbles escaping once every minute or so.

Furthermore, and my observation may not be correct from the picture that you posted, it seems that the pH has not yet stabilised at the time that your light kicks in. The picture may be a bit distorted, but it looks to me that your pH is still on its way down, and then suddenly bends up. Not really the pattern that I expect, and for me not clear what it means.

Then, I believe the logical test to do is run one day a pH profile when you do not turn on your light in the tank. This should give a nice smooth pattern of pH approaching -1.5 and then stay there all day. If next day you turn on lights as normal, you know what you're plants are doing with their CO2 consumption and hopefully it then all becomes clear.

I must admit that I only have experience with pH profiles (within a day, as well as day-to-day) in my own tank, and do not know what heavy planted tanks with lots of light can do. For a hight tech tank I have modest plant mass, modest lighting with some pearling but not impressive. Perhaps I underestimate how much CO2 the real champs like yourself can use during photoperiod, and maybe some other forum members can comment on their pH profiles.

I hope this makes sense

 

[Unexpected]

We have an interesting puzzle here, and your surface agitations looks good to me

First, to make sure the reactor works correctly in overflow mode I believe it makes sense to do some tests with a high enough CO2 flow rate so that you see some small bubbles every now and then being blown in the tank. You are then absolutely sure that it is not your regulator drifting, and that both reactor tubes are used at the full capacity. This will give some CO2 losses, but you can later dial back injection and find the 'sweet spot' where losses are minimised but reactor optimised at its maximum capacity. For my reactor in my tank I now know that this optimum is about 7 bubbles per second, and I hear a slight trickle from small 1-2 mm bubbles escaping once every minute or so.

Furthermore, and my observation may not be correct from the picture that you posted, it seems that the pH has not yet stabilised at the time that your light kicks in. The picture may be a bit distorted, but it looks to me that your pH is still on its way down, and then suddenly bends up. Not really the pattern that I expect, and for me not clear what it means.

Then, I believe the logical test to do is run one day a pH profile when you do not turn on your light in the tank. This should give a nice smooth pattern of pH approaching -1.5 and then stay there all day. If next day you turn on lights as normal, you know what you're plants are doing with their CO2 consumption and hopefully it then all becomes clear.

I must admit that I only have experience with pH profiles (within a day, as well as day-to-day) in my own tank, and do not know what heavy planted tanks with lots of light can do. For a hight tech tank I have modest plant mass, modest lighting with some pearling but not impressive. Perhaps I underestimate how much CO2 the real champs like yourself can use during photoperiod, and maybe some other forum members can comment on their pH profiles.

I hope this makes sense

I upped the CO2 last night and only had a .5 variance after the weird spike down then up at lights on. It's this regulator. Unfortunately, regulators are expensive here for a top end one. I guess I will need to do some research and spend more money . I upped the CO2 again and shortened the on time to lights on.

 

[Unexpected]

Huge difference with the increased injection rate.

 

[BigWave]

Unfortunately, regulators are expensive here for a top end one

@Unexpected Hit up @Alanle . He builds custom regulators and mails them all over the world. He's works with whatever parameters you give him to come up with a good working build. Im sure you guys can work something out. I have one of his regulators and its been going strong for some 7 yrs now.

 

[Unexpected]

@Unexpected Hit up @Alanle . He builds custom regulators and mails them all over the world. He's works with whatever parameters you give him to come up with a good working build. Im sure you guys can work something out. I have one of his regulators and its been going strong for some 7 yrs now.

Just bought one from him. Finally going to have a flow meter I’ve always wanted too.

 

[Aquascape_SG]

@Unexpected Hit up @Alanle . He builds custom regulators and mails them all over the world. He's works with whatever parameters you give him to come up with a good working build. Im sure you guys can work something out. I have one of his regulators and its been going strong for some 7 yrs now.

A quality co2 regulator is the foundation for a healthy planted tank.

 

[CFassett]

I want to thank @Unexpected for introducing me to this new forum, but most importantly for him being the first ever to build the Horizontal CO2 Reactor. I feel honored that "Yugang Reactor" as he calls it (Yugang 鱼缸 means fish tank) is introduced in @Unexpected journal, but it would be a pity if that journal gets cluttered with too much reactor talk. So let me open a dedicated thread for discussions, questions and help on this Horizontal Reactor.

I have been experimenting for years with CO2, and at some time I took videos of my (modified) Aquamedic reactor to watch the bubbles in slow motion. To my surprise I found that the combined surface area of the bubbles was not very high for achieving a 1.5 pH drop in my 250 liter tank, and that CO2 absorbs so fast in water that the lifetime of bubbles is quite short. So .... we don't need to juggle bubbles at all, we can just create a very simple absorption interface between flowing water and a pocket of CO2 above it. The Horizontal reactor is surprisingly simple, and has some benefits that we won't find in diffusers or conventional Cerges / Griggs bubble reactors:

• Simple rules to find the correct dimensions for any tank, a small nano tank or a huge tank the size of a swimming pool.
• Very easy and cheap to build with plumbing materials. No fragile parts, low risk of leakage or malfunction.
• 100% CO2 absorption efficiency
• No need to experiment with vortices, venturis, diffusers, needles wheels, impellers / rotors, multi stage reactors - it is just a pipe with a gentle flow of water.
• No noise
• No mist in the tank
• No maintenance, and stable performance over time
• Virtually no reduction of flow from pump
• A purging valve is optional, as the reactor will purge itself from excessive trapped air.
• The reactor can be configured so that in the event of a failure, CO2 injection will not exceed a limit and fish cannot be gassed. Inherently safe.
• The reactor can be configured so that we do not need a precision regulator, because the reactor controls the CO2 injection rate

View attachment 2030

Or a multi stage version ...
View attachment 2031

A detailed thread can be found on UKAPS, linked with the kind permission from Scapecrunch forum moderators #1

I am happy to help new users to verify the calculations on reactor dimensions for their tank, please send me a PM.

Super interesting, I have a few thoughts to add:

For context, I've been extremely frustrated by the sheer mass of CO₂ reactors, especially for small tanks, so I've been looking for ways to shrink them. This certainly shrinks them a bit, but also has the benefit of orienting them in a way that utilizes space that is often available. I still would like to try and shrink them further.

I'm considering using something like gutter downspouts to create a rectangular cross-section for increased/more consistent surface area. I realize that a nice effect of having the cylinder is that the more you inject, the greater the surface area becomes, so it sort of increases in efficiency as more efficiency is demanded, but that same feature adds significantly to the vertical space requirement. You should be able to use something that's 3" wide and 1" tall (inside measurements) and get the same benefit as a 3" ID pipe, with 1/3 of the space utilized. Reality may reveal some additional parameters: e.g. what is the minimum vertical space before efficiency is impacted? With the downspouts, many PVC options are available, and you can find PVC caps that fit and can be welded into place. It wouldn't be difficult to drill some holes and add bulkheads.

A cool thing about using a horizontal design is that the friction head of the water's surface against the bottom of the pipe will cause the water at the base to move more slowly than the water at the surface. This difference of speed will cause small pressure differentials and induce chaotic flow, allowing for vertical mixing of the water column, ensuring that all of the water is brought into contact with the CO₂, a distinct advantage over conventional "Cerges" style reactors, which have a ton of "wasted" vertical space, all for the purpose of preventing bubble egress.

When using something with a rectangular cross-section, we also have the option of adding some form of corrugation to the bottom surface fairly easily. I realize that this may induce some splashing and noise, but there may also be a way to create significant deformation of the water surface with an acceptably low increase in noise. This should result in a significantly increased surface area, as well as even more friction head, resulting in more turbulent flow (and therefore, vertical mixing of the water column).

Oh, and just a side note, I don't know about Hong Kong, but it's a heck of a lot cheaper for me to get some clear PVC pipe (I get Formufit furniture-grade) than it is to get acrylic tubes, and it fits standard PVC sizing.

 

[Art]

I like the innovating discussion here!

 

[toofewfish]

I found this idea of the horizontal reactor very logical once I started reading and thinking about it. I have 150 gallon tank with a approx. 80 gallon sump and had a hard time reaching the pH drop I was trying get to of 1.2/1.3 with my existing in-line vertical reactor. The reactor just didn't have the capacity to get where I needed. From reading Yugang's post, It just didn't have the interface area needed to absorb enough CO2. The more CO2 I pushed at it, it just created a larger pocket of CO2 at the top of the reactor. Also with the volume of water I have, it took over an 1 1/2 hours to drop to it's lowest point. After reading the posts from Yugang this all made sense to me, so I endeavored to give a horizontal reactor a try.

I took inventory of what PVC pipe I had around and I had a fair amount of 2 inch PVC pipe so I set about using Yogang's calculations to determine the length of reactor I would need. I did a final adjustment to compensate for the volume of CO2 before it would start spilling out of the outlet end as opposed to full diameter of the pipe. The calculated length worked well with the 5 ft lengths of 2" PVC that I had.


I didn't have a space constraint because my Sump system is in the basement below the living room where the tank is at. But I did build the reactor as a parallel system out of 2 5' lengths (120"). I 3D printed my brackets to space the reactor out around some other plumbing. I also went to a separate pump for the reactor and return the water from the reactor so that it exhausts out in front of my return pump to the tank. That way I can easily control the water flow thru the reactor.



Once I finished up the build, it was time to start some trials. I started adjusting my CO2 until I just started seeing a few small bubbles passing thru the outlet tubing. At this level of CO2 i was getting a 1.6 pH drop (which is more than I needed). I use a pH controller and a flow meter so it was quite easy to tune in my pH drop to the 1.3 I was looking for. With the Ki flow control, I can set the CO2 flow so that the controller is only there for any possible drops below my target of 6.0. I've been running for about 2 months now and everything is steady as a rock. Another great thing with this system is that I can get my full pH drop in the morning in less than 1/2 hour (that's for approx. 230 gallons) And yet another big bonus is that my CO2 consumption went down to about 60% of what I was using with the vertical reactor!! $$$

 

[Unexpected]

I found this idea of the horizontal reactor very logical once I started reading and thinking about it. I have 150 gallon tank with a approx. 80 gallon sump and had a hard time reaching the pH drop I was trying get to of 1.2/1.3 with my existing in-line vertical reactor. The reactor just didn't have the capacity to get where I needed. From reading Yugang's post, It just didn't have the interface area needed to absorb enough CO2. The more CO2 I pushed at it, it just created a larger pocket of CO2 at the top of the reactor. Also with the volume of water I have, it took over an 1 1/2 hours to drop to it's lowest point. After reading the posts from Yugang this all made sense to me, so I endeavored to give a horizontal reactor a try.

I took inventory of what PVC pipe I had around and I had a fair amount of 2 inch PVC pipe so I set about using Yogang's calculations to determine the length of reactor I would need. I did a final adjustment to compensate for the volume of CO2 before it would start spilling out of the outlet end as opposed to full diameter of the pipe. The calculated length worked well with the 5 ft lengths of 2" PVC that I had.

View attachment 2605
I didn't have a space constraint because my Sump system is in the basement below the living room where the tank is at. But I did build the reactor as a parallel system out of 2 5' lengths (120"). I 3D printed my brackets to space the reactor out around some other plumbing. I also went to a separate pump for the reactor and return the water from the reactor so that it exhausts out in front of my return pump to the tank. That way I can easily control the water flow thru the reactor.

View attachment 2588

Once I finished up the build, it was time to start some trials. I started adjusting my CO2 until I just started seeing a few small bubbles passing thru the outlet tubing. At this level of CO2 i was getting a 1.6 pH drop (which is more than I needed). I use a pH controller and a flow meter so it was quite easy to tune in my pH drop to the 1.3 I was looking for. With the Ki flow control, I can set the CO2 flow so that the controller is only there for any possible drops below my target of 6.0. I've been running for about 2 months now and everything is steady as a rock. Another great thing with this system is that I can get my full pH drop in the morning in less than 1/2 hour (that's for approx. 230 gallons) And yet another big bonus is that my CO2 consumption went down to about 60% of what I was using with the vertical reactor!! $$$

View attachment 2599View attachment 2606

That is so awesome! Thanks for sharing. I’m positive this is the future for big tank CO2 injection.

 

[Aquascape_SG]

Hi @toofewfish ,

First of all thanks for sharing your past experience with co2 dosing and your current journey into horizontal reactor ( aka Yugang Reactor ). You have certainly done your home work on building the reactor. Like you, I simply love DIY. My friends say I love my hardware stuff more than my fish or plants in my tank and I'm starting to believe that they maybe right after all...

The reactor just didn't have the capacity to get where I needed. From reading Yugang's post, It just didn't have the interface area needed to absorb enough CO2. The more CO2 I pushed at it, it just created a larger pocket of CO2 at the top of the reactor.

My tank is a 200 gal planted tank ( I have a weak spot for large tanks ). It's about 2 years old and from day one I have been struggling with co2 dosing. I have tried various methods of co2 dosing but none seems to be working due to the large water volume and the inbuilt attribute of degassing when running a sump. Now I'm running a vertical reactor ( from AliExpress). The co2 level from the reactor is a hit or a miss. I would constantly be hearing that dreaded gurgling sound due to a big gas pocket buildup inside the reactor. I was tearing my hair out on how to stop the gas buildup. I thought I was the one making the mistakes somewhere. I tried increasing and decreasing the co2 output, tuned the dedicate DC pump output for the reactor. It had a sea saw effect on the co2 levels in my tank. The uneven co2 levels led to many frustrating battles with BBA and I'm still fighting it till this very day. It was only after reading @Yugang"s post that I realized that I wasn't making any mistakes but it was how vertical reactor worked and it had a limited capacity. I was so relived after knowing this.

But I did build the reactor as a parallel system out of 2 5' lengths (120").

Am I understating that your new horizontal reactor 120 inch long? I live in the planet where everything here is in centimeters and meters...

3D printed my brackets to space the reactor

Are the DIY brackets strong enough to bear the weight of the tubes filled with water? Do they sag due to the added weight?

a flow meter so it was quite easy to tune in

A flow meter should be made mandatory for all co2 injected run large tanks. It's a very useful gadget and we can actually compare the output of co2 put with others and for our own record keeping. I hate bubble counters and taking slow-mo videos to count the BPS ( especially for large tanks like ours where co2 output is insane). I got mine thanks to @GreggZ and his most valuable guidance.

great thing with this system is that I can get my full pH drop in the morning in less than 1/2 hour (that's for approx. 230 gallons) And yet another big bonus is that my CO2 consumption went down to about 60% of what I was using with the vertical reactor!! $$$

Sounds sweet to my ears. Here in my country co2 refilling is very affordable but chucking 4 large co2 cylinders in and out of the car to be refilled isn't sweet at all. And let's not even talk about asking time out from the wife to run the errand.

oh btw, you have a lovely tank. Strange that I see many tanks here in this forum with rainbow fish ( which btw is my personal favorite too).
Maybe we should change the name of this forum to RainbowCrunch? @Art ( just kidding)

Cheers

 

[Art]

Another great thing with this system is that I can get my full pH drop in the morning in less than 1/2 hour (that's for approx. 230 gallons) And yet another big bonus is that my CO2 consumption went down to about 60% of what I was using with the vertical reactor!! $$$

This is really impressive and good to hear. I've never had a large tank where reaching saturation points with CO2 would be an issue. Good to have a tested option for others.

 

[Art]

A flow meter should be made mandatory for all co2 injected run large tanks. It's a very useful gadget and we can actually compare the output of co2 put with others and for our own record keeping. I hate bubble counters and taking slow-mo videos to count the BPS ( especially for large tanks like ours where co2 output is insane). I got mine thanks to @GreggZ and his most valuable guidance.

This is true for smaller tanks too. I'm thinking about it.

 

[Art]

Maybe we should change the name of this forum to RainbowCrunch? @Art ( just kidding)

I'm sure a few members would be delighted with that change!

 

[Yugang]

I’m positive this is the future for big tank CO2 injection.

Big tanks are really challenging with vertical reactors that work with CO2 bubbles, sure. But there are various reasons why also smaller tanks would be better off with a horizontal than a vertical bubble reactor.

If we would have a pipe with certain dimensions (diameter and length), used as a vertical bubble reactor, I calculated that a pipe with same dimensions used as a horizontal reactor will likely have a (much) higher capacity to inject CO2. The calculations are quite straight forward, we just calculate how many bubbles we would need in the pipe to provide the same water/CO2 absorption interface as compared to the length*diameter of same pipe when applied as a horizontal reactor. It appears very unlikely that we can keep enough bubbles in the pipe to match, or even come close, the surface area of a horizontal reactor.

Unfortunately my calculations were in one of the 600, mostly technical, posts that were deleted on another forum and I do not have a backup. If there is an interest, I can redo the calculations and post later in this thread.

I believe that also for smaller tanks horizontal reactors will do a better job than bubble reactors, but it may take a bit more time as the hobbyists with smaller tanks in many cases are satisfied with their current solution and are less inclined to look for (new) alternatives.

 

[toofewfish]

Am I understating that your new horizontal reactor 120 inch long? I live in the planet where everything here is in centimeters and meters...

Yes, it is that long. If I would have had to purchase PVC pipe, I would have gone to a larger diameter thus a shorter length. But as I said, I had the 2 inch pipe and space was much of a concern. At first, I built it as a "let's see" type of project, but the results were great.

Are the DIY brackets strong enough to bear the weight of the tubes filled with water? Do they sag due to the added weight?

I printed the 1st set of brackets out of PLA. They lasted about a month and then they started to get stress cracks and broke. I beefed up the design added larger fillets, more structure and then printed them out of ABS material. Time will tell if these hold up.

oh btw, you have a lovely tank. Strange that I see many tanks here in this forum with rainbow fish ( which btw is my personal favorite too).
Maybe we should change the name of this forum to RainbowCrunch? @Art ( just kidding)

Thanks, it's been a big project, but I'm really happy with the results. And of course you have to have Rainbows!

A flow meter should be made mandatory for all co2 injected run large tanks. It's a very useful gadget and we can actually compare the output of co2 put with others and for our own record keeping. I hate bubble counters and taking slow-mo videos to count the BPS ( especially for large tanks like ours where co2 output is insane). I got mine thanks to @GreggZ and his most valuable guidance.

I read Greggz's posts on the other forum and other people posts about flow meters as I was planning this system. I knew that was the answer for me with this large tank and put the system together from the start with the Ki flow meter. It makes adjustments so easy and controllable.

 

[Koan]

Birth registration of my newly built reactor

View attachment 2388

Some features may not be strictly necessary, but I included them for further testing:

• Acrylic pipe, transparent. Acrylic is not easy to get in HK, and the sizes are not compatible with standard PVC plumbing, so I ended up with a rather small diameter pipe and an overall design that is not very compact. I was not sure how to glue acrylic to PVC, so decided to use teflon tape for the fit. I paid about 7 USD for the acrylic.
• The CO2 comes in through a small tube, that sticks in the water. This allows me to use it as a bubble counter, A 30 sec video, and I count the bubbles while in slow motion to have an accurate measure for the injection.
• A gas purge valve. I will not often use it, as this reactor will be able to purge itself as described earlier.
• A bypass, so that the flow in the reactor can be slowed down while still using the maximum flow from my FX4 pump.
• Inlet and outlet design with 'elbows' as in below picture. The inlet will lead the water down, to reduce water splashing or making noise at high flows. The outlet will let CO2 purge when the gas pocket reaches 50% of the tube (inherent safety feature, limiting reactor max capacity).

View attachment 2390

I tested if the bypass is necessary, and tried what happened when all the flow from my FX4 (rated 2650 liter/hr) was lead through the upper reactor pipe. The reactor worked, but made some noise and the splashing of the water flow created a few bubbles that ended up in my tank. I now use the reactor with the bypass nearly fully open, and a gentle slow flow through the reactor.

I am currently using the reactor in 'overflow mode', that is that I inject CO2 at a rate that the reactor fills for 50% and starts to purge small bubbles from the exit. With this I know that I always use the same reactor capacity (not dependent on stability of my regulator), and have a stable injection rate that is fixed by the reactor geometry. I may use a pH controller in the future, in that case the overflow will give me a safety when something goes wrong, but that will have to wait until I've decided which controller to buy.
.

Yugang can you share the name of that gray PVC fitting you are using for inflow and outflow caps?

I am not familiar with that fitting, very slick! Clearly (sic) beats the h*ll out of drilling holes in end caps!