Endless Soda Water: Carbonation Unexpectedly Low

[Positronic_Matrix]

I have served carbonated water from my keezer on and off for a few years. A keg typically lasts a week and it’s then up to me to pull it out and refill the water. Because I can get busy, sometimes we can go a week or more without carbonated water to the dismay of my family. To rectify this issue, I purchased a KegLand continuous soda carbonation keg lid along with a new 9.5 L (2.5 gallon) corny keg. The continuous soda carbonation keg lid has a ball connector on the lid for a water input, that is turned on and off by a float. Thus, as carbonated water is poured, filtered tap water enters which is carbonated via a gas line with a carbonation stone. The idea with the 9.5 L corny keg, is that because it would be constantly refilled, I wouldn’t need a full keg and thus could put the half-sized keg up on the keezer hump, freeing room for a beer keg on the floor.

The issue I am having is that although the pressure on my CO₂ gas line has not changed, I received feedback that the water is not as carbonated as it was before. I am currently on travel but when I get home I am certain I’ll be able to solve the issue by experimenting with increased gas pressure.

My question for the community is, is there some physics going on with the keg having both a pressurized gas and water input? I had assumed that once the water had filled, the tank would look identical to a standard keg and the carbonation stone would deliver the same quantity of gas to the liquid. Is this not the case?

 

[sibelman]

Could time be a factor? As flat water is added, the overall carb level drops until enough CO2 dissolves to take it back up to the desired level.

 

[IslandLizard]

What is the CO2 pressure set at?
How long is the serving line? What is its internal diameter?

Does it pour with big bubbles, different from the previous system?
Is the first pour of the day OK, but successive pours start lacking carbonation?

 

[doug293cz]

Could time be a factor? As flat water is added, the overall carb level drops until enough CO2 dissolves to take it back up to the desired level.

This.

A carbonation stone is not magic. All is does is increase the surface to volume ratio (by orders of magnitude) which allows faster CO2 absorption than you can get from the headspace. But, there is still a maximum obtainable absorption rate, and if the water consumption and refill rate exceeds the CO2 absorption rate, the carb level will go down. It will recover with time, if consumption is paused.

A problem with carb stones used at the high pressures required for soda water carb levels, is that the CO2 flow rate will be much higher than the water can absorb the CO2, even with all of the tiny bubbles percolating up thru the water. The pressure then builds up in the headspace (due to the CO2 not all being absorbed on its way up the keg), the flow of CO2 stops, and now you are only absorbing CO2 from the headspace, just as if the carb stone wasn't there.

The optimal use of a carb stone is when the flow rate of CO2 is maintained at or below the rate at which CO2 can be absorbed from the bubbles percolating up thru the liquid. This effectively means starting at a low pressure, and slowly ramping up the pressure to maintain the required flow rate until the desired level of carbonation is achieved. The pressure in the system should then be at the "chart" value for the temp and desired carb level.

Brew on

 

[Positronic_Matrix]

The CO₂ pressure is set to 140 kPA (20 psi). The serving line is 2 m (6 feet) which was sized for beer at a lower pressure. Lines are standard 3/16” (4.8 mm) ID. When pouring from the 19 L (5 gal) keg previously, I’d get large bubbles as the gas would leave solution in the tubing. When pouring from the new 9.5 L (2.5 gal) keg, there are fewer large bubbles as if the pressure appears to be is lower.

The first pour is lower pressure than expected. Following pours are very similar, as expected when drawing 250 mL (2.5%) from a 9500 mL vessel.

 

[Positronic_Matrix]

A carbonation stone is not magic.

This is steady-state carbonation [edit: where the system was allowed to sit without use to achieve maximum carbonation]. I put the system together while my family was out of town (Thanksgiving) and found that [even] after a couple days ([again] without use) the soda still lacked full carbonation. As they were returning to town and I was leaving town (business), I ran out of time to debug the issue and asked them to give me an update. They confirmed pressure remained low (after another 24 hours [without use]) on [their] first use.

There are a few variables that changed. It’s a smaller keg, there’s a large float assembly that fills half the head space, the carbonation stone is different, and pressurized water is present.

 

[Positronic_Matrix]

If one has two pressurized vessels feeding a common volume isolated by check valves on each input, both inputs would flow at a rate proportional to the difference in pressure until equilibrium was met. If the input pressures were equal, the filled volume would contain exactly half volumes from each input. If one input was at a higher pressure, it would resist the input from the second.

Could it be that due to the pressurized water flowing into the tank that vessel creates a resistance to the flow of CO₂. Thus, the pressure reads the same, however the net volume of CO₂ delivered is actually less?

 

[doug293cz]

This is steady-state carbonation. I put the system together while my family was out of town (Thanksgiving) and found that after a couple days (without use) the soda still lacked full carbonation. As they were returning to town and I was leaving town (business), I ran out of time to debug the issue and asked them to give me an update. They confirmed pressure remained low (after another 24 hours) on first use.

There are a few variables that changed. It’s a smaller keg, there’s a large float assembly that fills half the head space, the carbonation stone is different, and pressurized water is present.

No, it is not steady state. For it to be steady state, you would have to have a constant rate of water consumption, matched by an equal constant rate of water replacement, and a constant rate of CO2 flow thru the stone, matched so that CO2 concentration in the water is constant. You have none of these conditions met in your system. You have a cyclic system.

Brew on

 

[doug293cz]

...

Could it be that due to the pressurized water flowing into the tank that vessel creates a resistance to the flow of CO₂. Thus, the pressure reads the same, however the net volume of CO₂ delivered is actually less?

Your float valve level control isolates the water in the tank from the water line pressure, except when water is actually flowing into the tank, so that should not be an issue. Just like the water in your toilet tank at home is not pressurized by the water line pressure.

Brew on

 

[IslandLizard]

The serving line is 2 m (6 feet) which was sized for beer at a lower pressure. Lines are standard 3/16” (4.8 mm) ID.

Much longer and thinner lines (such as 3mm or 4mm EVA Barrier) surely would help keep your water better carbonated while dispensing. Even more so when you're increasing CO2 pressure to improve carbonation levels.

I don't think that's the key issue though... I think it's what's inside the headspace.

In your old setup you may have noticed that carbonation actually improved somewhat over the first few days. As water was being dispensed, the headspace increased, leaving a larger amount of CO2 available. Besides, that water was fully carbonated already.

Here's my theory:
In your new continuous supply system, the composition of the gasses in the keg's (small) headspace is likely different.
I wonder how much O2 and N2 is brought in with the water. Those 2 gasses build up in that small headspace mixed in with residual (undissolved) CO2 from your carbonation stone. Aside from the partial pressure of CO2 being low, the stone can't release any more CO2 bubbles once the (headspace) pressure reaches 20 psi. And that may happen way before the water mixture gets fully carbonated.

 

[doug293cz]

I wonder how much O2 and N2 is brought in with the water. Those 2 gasses build up in that small headspace mixed in with residual (undissolved) CO2 from your carbonation stone.

This could be a small factor, if the headspace is purged when the keg is initially filled. In that case the O2 and N2 partial pressures in the headspace start out at ~0. So, initially any dissolved O2 or N2 in the incoming water will tend to diffuse into the headspace until equilibrium is reached between the concentration in the water and the headspace partial pressures of each gas.

This can be compensated for by increasing the CO2 pressure to keep the CO2 partial pressure at the value needed for the desired carbonation level. At some point this situation will become stable enough that you no longer have to fudge the CO2 pressure upwards. Unlike beer, O2 in the headspace will not degrade the water, so no need to try and periodically remove excess O2 (or N2 for that matter.)

Aside from the partial pressure of CO2 being low, the stone can't release any more CO2 bubbles once the (headspace) pressure reaches 20 psi. And that may happen way before the water mixture gets fully carbonated.

I think, as I tried to say above, that this is the primary issue.

Brew on

 

[Positronic_Matrix]

In your new continuous supply system, the composition of the gasses in the keg's (small) headspace is likely different.
I wonder how much O2 and N2 is brought in with the water. Those 2 gasses build up in that small headspace mixed in with residual (undissolved) CO2 from your carbonation stone. Aside from the partial pressure of CO2 being low, the stone can't release any more CO2 bubbles once the (headspace) pressure reaches 20 psi. And that may happen way before the water mixture gets fully carbonated.

I’m not sure if this is the cause of my current issue but you have good intuition. When I first set the system up, it would not reasonably carbonate, as if there were no carbonation stone in place. It hit me that the tank was air filled when I forced water and CO₂ into the existing atmosphere under pressure and that might be interfering with the carbonation. I purged the tank a few times and a couple hours later it was carbonated.

 

[Positronic_Matrix]

OK. I have a theory.

I believe where I’m getting hung up is assuming the pressure in the system is constant. In actuality, when gas flows into the keg there is a pressure differential or potential. In physics, a potential (e.g., voltage, pressure) is required to do work. If no potential exists, then no work can be done. Thus, a pressure differential (potential) must exist between the regulator and tank when refilling in order for the gas to be driven into the tank (work is performed).

Given that the gas regulator acts as a constant pressure source, the only variable in the system is the pressure at the keg input. If the keg pressure is increased, the pressure differential (or potential) is decreased, reducing the system’s ability to do work. This is a relation in physics which states that power flow (energy transfer per unit time) through a hose is given by the difference in potential times the flow rate.

If one were to conduct a thought experiment with two kegs, one with a low pressure and one with high pressure, the tank with the lowest pressure and thus highest differential will have the greatest potential for work. That work will be realized as transfer of more gas into the keg and into the water inside that keg. It takes energy to move CO₂ into solution and that energy comes from the pressure differential. Thus, although both tanks would reach pressure equilibrium, the tank with the lowest starting pressure would undergo the greatest work, the greatest energy (gas) transfer, and thus have the highest carbonation level.

So, returning to the endless soda water configuration, it has parallel gas and water inputs. Because the keg is being pressurized in parallel by a secondary source (water), this must increase the keg pressure, decreasing the pressure potential, and reducing the ability for the gas bottle to do work. The result is a decrease in carbonation level.

Thus, a two-input keg is harder to carbonate.

 

[doug293cz]

I’m not sure if this is the cause of my current issue but you have good intuition. When I first set the system up, it would not reasonably carbonate, as if there were no carbonation stone in place. It hit me that the tank was air filled when I forced water and CO₂ into the existing atmosphere under pressure and that might be interfering with the carbonation. I purged the tank a few times and a couple hours later it was carbonated.

The carbonation charts and calculators assume that the headspace is 100% CO2. If you start with CO2 at atmospheric pressure in the headspace the gauge pressure is 0 psi, but the absolute pressure, and CO2 partial pressure, is 14.7 psi. If the chart/calculator says you need 20 psi for your desired carb level, and you set your regulator to 20 psi and connect to the keg, then the absolute pressure in the keg (and CO2 partial pressure, is 14.7 + 20 = 34.7 psi. The equilibrium carbonation level depends only on the temperature and CO2 partial pressure in the headspace. So, in this case you need a CO2 partial pressure of 34.7 psi to get your desired carb level.

Now if your headspace starts out full of air at atmospheric pressure, and you connect CO2 at 20 psi, the total absolute pressure will be 34.7 psi, but the CO2 partial pressure will only be 20 psi, and you will be under-carbed.

If you start with a sealed keg full of air at atmospheric pressure, and you force water into the keg to 3/4 full, the absolute pressure in the keg will be 4 * 14.7 = 58.8 psi, and the gauge pressure will be 44.1 psi. If you connect CO2 at 20 psi, nothing happens if you have a check valve on the CO2 line, or if you don't have a check valve, air flows back into your regulator. In either case, no CO2 flows into the tank, you have 0 psi CO2 partial pressure, and you end up with 0 carbonation.

If one were to conduct a thought experiment with two kegs, one with a low pressure and one with high pressure, the tank with the lowest pressure and thus highest differential will have the greatest potential for work. That work will be realized as transfer of more gas into the keg and into the water inside that keg. It takes energy to move CO₂ into solution and that energy comes from the pressure differential. Thus, although both tanks would reach pressure equilibrium, the tank with the lowest starting pressure would undergo the greatest work, the greatest energy (gas) transfer, and thus have the highest carbonation level.

It doesn't work like this. As stated above, the equilibrium carbonation level depends only on the temperature and the CO2 partial pressure in the headspace. The starting level of carbonation in the liquid does not affect the final equilibrium, but will affect the rate at which CO2 is absorbed.

Brew on

 

[Bobby_M]

I have the same setup in my home kegerator. The reason the carbonation is lower is because the turnover of product is faster than the carbonation rate. More seltzer is being poured than the stone can keep up with so the full tank of water is never reaching equilibrium. If it were a 5 gallon keg, you'd get closer to equilibrium because the product turnover would be a smaller portion of the total product available. I see this playout real time if I'm having a party and 20 people are drinking seltzer all night. Even with a 5 gallon keg, the carbonation level drifts downward.

20psi is objectively low for seltzer in the first place. I run mine at 26psi and that's also on the lower end.

You'll also benefit from putting a check valve on the water input line to keep CO2 from continuing down the line. When I first installed the continuous lid, I was also carbonating my drinking water holding tank under the sink (20 feet of tubing away) by mistake.


Here's one way to buffer the usage.. The incoming water sits in a keg in advance of the carbonation to make sure it's cold. That means by the time it gets to the serving keg it will accept carbonation noticeably faster. That's assuming the incoming water into your current system is warm-ish.

Another way is to make sure the water spends some time in the cold space, such as wrapping a bunch of tubing around the keg or through a stainless coil and then have it run through a tee that has a diffusion stone mounted into it. This forces the incoming water to get carbonated before it even hits the tank.

I personally did something in between those two. When the incoming water enters the kegerator, I have 25 feet of 1/4" ID PE tubing rolled up and stashed where ever it can fit so it pre-chills somewhat. It helped enough. If it didn't, I'd add an inline carbonation tee as the second picture shows.

If I were you, I'd add some tubing length into the cold area and increase the CO2 to 25-28psi and see if that solves it for you.

 

[Positronic_Matrix]

If I were you, I'd add some tubing length into the cold area and increase the CO2 to 25-28psi and see if that solves it for you.

Thank you for the detailed post with the high quality pictures. I took your advice and increased the pressure to 28 PSI and that solved the problem. The family is very pleased with the quality of the carbonated water and I’m a bit embarrassed for overthinking the issue when a turn of a knob made all the difference.

I appreciate the advice regarding chilling the water prior to carbonation. Ground water is currently cool but I will implement a similar coil solution as you have before temperatures start to warm up.

Thank you!

 

[ODI3]

Why not put the flow lid in keg one?
The system would take a lot longer to stabilize, but at the end of the day, you would end up with 2 tanks carbonated, and new water added to keg one would have very little impact on tank two serving until you went through like 7+ gal of stored carbonated water.

Also this design would minimize carbonic bite.

To get set up faster, you may want to fill and shake tank 2 as it would take too long to carbonate tank 2 via tank 1, but going forward after stabilized, tank 2 would be filled from tank 1 carbonated water instead of noncarbonated water.

This assumes that you dont want noncarbonated water as well.

 

[Bobby_M]

Why not put the flow lid in keg one?
The system would take a lot longer to stabilize, but at the end of the day, you would end up with 2 tanks carbonated, and new water added to keg one would have very little impact on tank two serving until you went through like 7+ gal of stored carbonated water.

I think it would be possible that undercarbonated water could leave tank one and then it doesn't have access to CO2 in tank 2.

Also this design would minimize carbonic bite.

Carbonic bite is the whole point of carbonating water.

To get set up faster, you may want to fill and shake tank 2 as it would take too long to carbonate tank 2 via tank 1, but going forward after stabilized, tank 2 would be filled from tank 1 carbonated water instead of noncarbonated water.

But it's still possible to over-serve driving tank 2 carbonation down with no way to recover. The two tank solution is more about sticking to 2.5 gallon kegs but just moving to a single 5 gallon keg would be closer to the solution you're suggesting.

This assumes that you dont want noncarbonated water as well.