Best way to measure total emitted weight of fermentation CO2?

Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum

Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

NeverDie

Well-Known Member
Joined
Jan 3, 2019
Messages
1,432
Reaction score
408
Location
Austin
The goal: indirectly figure out the amount of alcohol produced based on the amount of CO2 emitted.

From what I've read, it's nearly a 1:1 ratio, by weight, of emitted CO2 and generated alcohol. So, if 100 grams of CO2 are emitted, roughly 100 grams of alcohol would have been produced. If you know your must volume, then you could derive the alcohol percentage without having to open up your fermentation vessel and use a thief with a hygrometer.

So, that's the idea and the goal. It's probably easier to measure the total volume of CO2 emitted and then derive from that the weight of the CO2. So, what's the best sensor for that? Not sure how accurate it would be, but a digital bubble counter could be used to estimate it:

With a humidity and temperature sensor, you could probably also deduct the water vapor from the total gas emitted.

Is there a better way to automate the measurement? Or am I looking at this wrong, and maybe there's already some completely different way to automatically gauge the alcohol content in your wine or other brew?

Another way to measure the CO2 weight would be to put a scale under the fermenter and measure how much the weight goes down. I think that might be easiest, but for an automated system, you wouldn't be able to re-zero between the measurements, and so I don't know how accurate the measurements would be over time, short of automating a robotic way to lift the fermentation vessle up, re-zero the scale, and then place it down again.

Anyone tried either of these ways?
 
Last edited:
What you are describing sounds very similar to the Plaato airlock. I have never used it, though it is priced the same as a Tilt which I have several of. Either method will probably serve your purpose. If you are looking to build one yourself, I would look at Plaato's hardware and see if you can determine what they are using.

Cheers,
Joe
 
You could use an inverted bucket / container over water and put your blowoff tube into the inverted container.

This would yield a volume measurement that you could transform into mass with the ideal gas law.
 
I just found out about the Plaato airlock this week. Looks kinda cool but clearly you'd need a tame-running batch as blow-off hoses aren't happening...

Cheers!
 
I've been working on getting the Plaato airlock to work for some time and am still struggling. The science is there, I've just run into problems with potentially not perfectly sealed fermentors, or over active fermentations filling the airlock with yeast foam.

But your original post has me thinking/wondering if you could in fact just measure the weight of your fermentor continuously and empirically determine a relationship between weight change and gravity drop. That's essentially what Plaato does, except they're measuring bubble volume going by the sensor instead of reading a scale. But you start your fermentation entering the volume and OG of your wort. They have developed their algorithm then based on measured bubble flow rate to gravity.

So if you had the time and patience to develop your own data, you could weigh your fermentor at start and subtract the empty weight if your fermentor (which is the same as gravity/density x volume), then collect data of weight change from the start to gravity samples you take along the way. I'm guessing you probably need a pretty sensitive scale.
 
If it works as advertised, the Tilt solution sounds like a winner to me because the R&D would be already finished.

Alternatively, thinking a bit more about load cell drift, I bet it would be possible to derive an error correction formula to compensate for the bulk of the drift because (I'm guessing) it's likely to drift in a similar manner from one fermentation to the next rather than just randomly. Also, temperature of the load cell is known to affect measurements, so that could be included as a term in the error correction formula. In addition, with both the load cell and the gas emission approach one would need to deduct the volume of the sludge in order to derive ABV, but that seems well studied and it appears there are already formulas for that. If nothing else, a load cell based approach could act as a sanity check on the Tilt measurements, and if the discrepancy was too large the system could throw a red flag to summon human intervention rather than blindly plow ahead.

The absolute cheapest solution (in terms of material cost) though would seem to be counting bubbles digitally, so I'm curious what the shortcomings of that might be. If you already have a bubble airlock, then attaching a photointerrupter connected to an arduino (as in the youtube I linked to in the OP) looks both easy and cheap.

--------------------------------

Anyhow, the absolute easiest way to do this repeatably for a fixed volume of must is a simple before and after measurement. Weigh it before fermentation starts and then count the total bubbles emitted and weigh it afterward. Then take a hydrometer measurement for alcohol percentage and divide that by 1. the difference in weight and 2. by the number of bubbles. This, then, should reveal simple how much the alcohol percentage changes per gram of weight difference and per bubble. Am I right?

Of course, one could do a theoretical model that would be more general, but I think this empirical approach would likely be the easiest, which anyone could do.
 
Could one use an unloaded load cell as a reference to null out some of the error factors that the loaded cells see?

Cheers!

Good idea! I don't know the answer to that, but I bet you could use a load cell with a known weight dummy load that's near the target "must" weight to provide a correction offset.
 
I would expect CO2 isn’t the only thing you are losing due to partial pressure differentials - you will get a measurement, but have no idea how much is due to CO2 versus other mass loss (e.g., water vapor)
 
I would expect CO2 isn’t the only thing you are losing due to partial pressure differentials - you will get a measurement, but have no idea how much is due to CO2 versus other mass loss (e.g., water vapor)

Good point. In that case, the measurement might yield only an upper-bound on the alcohol content.

Is there an easy way to calculate/estimate how much loss that would be? Maybe a chemist or physicist would know that kind of thing.

I suppose one could chill the vapors through some kind of distillers coil before they enter the bubbler-airlock. That way they would drip back into the bottle and still be weighed.

Meh, if that proves to be necessary, then it's probably more hassle and more expense than its worth.

Alternatively, one could measure the temperature and humidity of the gases as they are emerge from the bubbler. That, combined with the volume of the gases emitted would allow at least the amount of water vapor to be calculated.

One thing that gives me hope is that a guy on one of the other threads says he uses a C02 mass flow meter to drive his calculations, and he says his results are always within a point or two of what the hydrometer says.
 
Last edited:
I concur that the Tilt works really well. I have a couple in case I'm fermenting multiple batches at once- can't imagine fermenting now without one. I mentioned above I'm evaluating the Plaato but haven't been able to get it to work well.

Apparently my fermentor which appears to be sealed and I still get airlock bubbles from, has some kind of partial leak that causes the calibration to be off. Other fermentors I have are even less airtight, or the configuration of their lid doesn't allow an airlock to fit together with my BrewJacket temperature system. The other limitation that's plagued me is that you have to have enough head space to completely contain your krausen. I've had a couple runs that were rendered useless on Plaato measuring gravity because I got krausen pushed up into it.

If you're trying to come up with some weight monitoring system just to see if it's theoretically possible, I think that's a cool idea. If you want to spend $130 on a Tilt, I concur that they work awesome.
 
Just want to listen in on this one - I hope to be looking at a stainless conical in a year or two, and really like the idea of something like this tied into temp control...

the beermkr group seem to be promising something along those lines. Alcohol content would be great, but if it can report and react to fermentation progress, that would be something else.
 
I tried playing around with counting bubbles several years ago and abandoned it because I wasn't able to correlate the data. The biggest problem I found was that during peak fermentation the airlock is going so fast that you get a more continuous stream for a short period rather than equal sized bubbles. So when I graphed bubbles per minute the rate actually went down during peak fermentation even though that's when the majority of CO2 was being released.

My airlock/sensor setup was nearly identical to the video linked in the OP. That video is only showing the "nice" part of fermentation where everything behaves as one would want.
 
Last edited:
The Plaato is close. Apparently it's my fermentor that's slightly leaky and not giving me good correlation to gravity. But it's spot on with its "Bubbles Per Minute" measurement throughout fast fermentation. It's much more sophisticated than the way I used to just manually count "glugs" in my airlock.

When fermentation is just starting, it starts climbing around 3 BPM, then gets going a bit at 50 - 60 BPM. But at peak it will be ripping along around 200 BPM. So it's metering/measuring in some fashion that's able to tell the difference between one glug of CO2 and a large volume of CO2.
 
Just want to listen in on this one - I hope to be looking at a stainless conical in a year or two, and really like the idea of something like this tied into temp control...

the beermkr group seem to be promising something along those lines. Alcohol content would be great, but if it can report and react to fermentation progress, that would be something else.

Sound interesting. In what manner would you like to see it react to the fermentation process?

------------------------------

Regarding counting of bubbles, I wonder if something made to "accurately" do that would perform any better than just the airlock bubbler, which I agree seems rather non-uniform in the kind of bubbles it lets through. Maybe something like this?
https://www.amazon.com/dp/B00IJLI7Y8/?tag=skimlinks_replacement-20

which is still fairy inexpensive and could perhaps serve as an airlock too.

-------------------------------------

Are you finding the Tilt to be as accurate as your hydrometer?

-----------------------------------------

Also, anyone try this thing? https://www.morebeer.com/products/brew-perfect-classic.html
It appears that the early software may have been a disaster, but now they claim to have updated software that works a lot better.
 
Last edited:
Sound interesting. In what manner would you like to see it react to the fermentation process?
In addition to what is currently available for temp control, I am envisioning something that measures CO2 to determine what stage of fermentation you are at. And then use that info to control temperature (be it glycol pump or a fridge) for the style/yeast/fermentation stage to keep things in the appropriate temp range. I'm a total newb here, but it seems ideal to time some temperature changes around primary/secondary fermentation and cold crash. This seems like a great way to automate that for when I'm not home, and to not stress the yeast and avoid off-flavors.
 
A lot of questions. Rather than do the quote thing, I'll just number them here:

1. What would I react to? For me, after a couple brews with the Tilt, I was adjusting temperature based on what I saw. As an example, I would watch the rate of change of SG. If it was going "too fast", I would drop the temperature a bit, "too slow" and I would nudge it up. I'll admit I have no rules or data to differentiate "too fast" vs "too slow", but if I automated/processed the Tilt output, I would monitor that slope and set up temperature adjustment rules based on the slope. Then after about 2-3 days, I'll see that SG can sometimes slow down. I'll bump up the temperature again until I start to see a good rate of SG dropping. At that point, I'm not concerned about the ester profile from the yeast, but I want to avoid a stalled fermentation.

2. Tilt accuracy? I hate hydrometers. Rather than rant on about why, let's just leave it at that. So before the Tilt everything I used was with my refractometer. I now use a digital refractometer, and use the refractometer gravity correction on Brewer's Friend after I've started fermenting since I know the alcohol skews the results. I've found my digital refractometer and the Tilt to be within 1 SG of each other, which is fully within their accuracy. I do find that my FG reading on the Tilt can tend to be more (2-3 SG points) off. So I just use the Tilt to follow the shape/progress of fermentation. When I see that it's done, I take my FG reading with my refractometer when kegging.

3. Tried BrewPerfect? I evaluated the new one (not the Classic), which supposedly had improvements over the Classic. I had some issues with it, and I found the App software to be a bit buggy. It's WiFi instead of Bluetooth, so you don't need to be in range to get a reading (or buy a cheap used iPhone to be a permanent BlueTooth to Internet logger like a lot of us do), and they are coming out with ability to do temperature control remotely via WiFi which is pretty cool (but haven't tried it).

Here's the links to my review of the Tilt and the BrewPerfect.

BrewPerfect Review
Tilt Review
 
Perhaps automated bubble counting is still useful as a way to track when fermentation has completed? I mean, if there have been zero bubbles over 3 days, can you safely call the fermentation done (or stuck)?

It's just that it's so cheap to do, it would be easy to automate the bubble counting on even small experimental batches.
 
Perhaps automated bubble counting is still useful as a way to track when fermentation has completed? I mean, if there have been zero bubbles over 3 days, can you safely call the fermentation done (or stuck)?

It's just that it's so cheap to do, it would be easy to automate the bubble counting on even small experimental batches.
It can be useful to show how the end of fermentation is slowing down, which you can use as a trigger to raise temps for a diacetyl rest, start spunding, dry hop addition, etc.
 
how about a spunding valve or switch actually that tied into an air source that powered a haskal pump and actually collected the CO2 and weighed it?
@MalFet talked many moons ago about using a mass flow meter prior to his spunding valve to measure volume and density of CO2 exiting the spunding valve. Granted he didn't save the CO2, but the measurement method seemed pretty cool.

Om the other hand I'm still waiting for @BrunDog to release v1.1 with tilt support so I can automate my digital spunding valve...
 
Anyone tried to use a hall effect flow meter to measure CO2? Can have it inline before or after an airlock...
 
Flow meters have minimum rate requirements that would be unlikely to be met metering CO2 flow from home brew scale fermentors...

Cheers!
 
Yeah, that does seem to be an issue. Mine is rated from 0.6L/min on up. Should be fine at the start of fermentation, but towards the end it would not be enough. One option would be to get some sort of pressure dump valve. Builds up pressure to a certain point and then dumps it before closing and building up again. Hmm, will see what I can figure out. Being in lockdown, I have the time ;)
 
So my flow meter seems to work to some extent. I ended up having the bubbles go up through a tube of water before going through the flow meter, this means that the bubbles take some water with them, which helps spin the wheel in the flow meter at an even lower flow. The design also makes it so there is a bit of build-up of gas before it releases and goes through the meter. It seems to read at even very low flow.

See figure below from an IPA I brewed. The green line represents the cumulative litres of gas produced. Ignore the other lines. There is a big gap because the power went out and my arduino wasn't saving the figure to memory, so it started at zero again. I took me a while to figure out the memory thing, implement it and get it going again. I doubt the actual number of litres are accurate and I can't yet convert to gravity or alcohol (would need many many more examples to work with). But at the very least, it gives a measure of the progress of fermentation.


flow_meter.png
 
start.jpg
slowdown.jpg
whole.jpg
whole.jpg
iSpindel works really well for me.
I can see the drop of gravity usually before 12 hours after pitch, when usually no bubble shows at that time. The slow down of gravity change speed is so obvious that it can be recognized in around 4 hours.
 
So my flow meter seems to work to some extent. I ended up having the bubbles go up through a tube of water before going through the flow meter, this means that the bubbles take some water with them, which helps spin the wheel in the flow meter at an even lower flow. The design also makes it so there is a bit of build-up of gas before it releases and goes through the meter. It seems to read at even very low flow.

See figure below from an IPA I brewed. The green line represents the cumulative litres of gas produced. Ignore the other lines. There is a big gap because the power went out and my arduino wasn't saving the figure to memory, so it started at zero again. I took me a while to figure out the memory thing, implement it and get it going again. I doubt the actual number of litres are accurate and I can't yet convert to gravity or alcohol (would need many many more examples to work with). But at the very least, it gives a measure of the progress of fermentation.


flow_meter.png


I have a couple of these mass flowmeters: Economical Gas Mass Flow Meters for Clean Gases with Optional Display

I just got BruControl set up and it was the first thing i hooked up, I've had these flowmeters for probably 10 years and had never used them. I had bought them for this specific purpose. I had 500cc/min model connected. It seemed to work well but I need more experience with it to understand the precision better. I'll probably just buy a Tilt in the end but I've had these flowmeters for so long wanting to try this that i had to get it set up.
 
once again i'm reminded of my 110lb scale that stays on, and is acurate to the 10th ounce...... just leave your fermenter on it.....i think even my 10 gallon batches are only around 80 or so pounds....$27 off amazon...
 
I got one for $45 off ebay, so less economical, but still not as much as a Tilt. I also can't use my Tilt. Despite other people's luck using these in conicals, I cannot get a signal out of a sanke keg once I install the spear.

I just got my new fermentation chamber control panel finished up to work with BruControl, so I will be testing this out on future ferments.
 
Resurrecting this thread. Anyone was successful getting some data of fermentation jug weight loss ?

I am not looking for automation, just estimate based on weight rather than measuring gravity with refractomer or hydrometer. The scale does not need to be extremely accurate, as weight difference between glucose and ethanol is quite significant. Bathroom scale should work.

Btw. the actual mass of ethanol to released CO2 would be ratio of their molar masses. 46.07g/mol (ethanol) / 44.01g/mol (CO2) = 1.0468.
 
Resurrecting this thread. Anyone was successful getting some data of fermentation jug weight loss ?

I am not looking for automation, just estimate based on weight rather than measuring gravity with refractomer or hydrometer. The scale does not need to be extremely accurate, as weight difference between glucose and ethanol is quite significant. Bathroom scale should work.

Btw. the actual mass of ethanol to released CO2 would be ratio of their molar masses. 46.07g/mol (ethanol) / 44.01g/mol (CO2) = 1.0468.
Based on CO2 production calculations here, 20 L of 1.050 wort fermented to 80% apparent attenuation will produce about 878 g of CO2. If fermenting at 65°F, the beer will retain about 0.88 volumes of carbonation or 20 L * 1.977 g/L * 0.88 = 35 g of CO2, so 878 - 35 = 843 g of CO2 would leave the beer. If the fermenter has 5 L of headspace, this will contain about 5 L * 1.977 g/L = 10 g of CO2, so 843 - 10 = 833 g of CO2 would leave the fermenter. 833 g = 1.83 - 1.84 lb.

Brew on :mug:
 
Back
Top