Keg purging with active fermentation

[joshbuhl]

One caveat: this underestimates the O2 concentration because not all the CO2 will be evolved, a fair bit will remain in solution in the beer….I‘ll work than in next.

Okay, that was easy since I already had my own carbonation spreadsheet and I just had to copy in the formula for g/l CO2 at T&P. Here‘s an updated screenshot which considers the CO2 dissolved in the beer to not have been evolved:

This makes an about factor 2 difference in the remaining O2 concentration in this example.

 

[joshbuhl]

Yes, I know exactly what you mean. I get that way sometimes as well.

It's also satisfying that someone fully capable went thru the analysis in detail, and didn't find any blunders.

Brew on

So, plugging in your original example numbers with my corrections for yeast mass synthesis and DCO2 in beer, I now get about 223ppb O2 (instead of only 5ppb) remaining in the 20l keg with 5l headspace in FV (which also need to be purged. total volume to purge 25l) fermenting 20l of 1.050 Wort down to 1.010:

Cheers

 

[mac_1103]

So, plugging in your original example numbers with my corrections for yeast mass synthesis and DCO2 in beer, I now get about 223ppb O2 (instead of only 5ppb) remaining in the 20l keg with 5l headspace in FV (which also need to be purged. total volume to purge 25l) fermenting 20l of 1.050 Wort down to 1.010

...but...

Actual measurements have been made, but not published publicly, that show that the actual concentration of residual O2 is much lower than the upper limit calculated by my analysis.

The math is impressive, but clearly something is being missed?

 

[joshbuhl]

...but...


The math is impressive, but clearly something is being missed?

Good question, and well watched out/combined!

This calculation is absolute worst case. As @doug293cz originally reasoned and worked thru in his original analysis there should be some „push“ effect, i.e. not instantaneous complete mixing, but rather the evolving CO2 will be pushing out the gas ahead of it. There‘s more air next to the spunding valve on the keg being purged and none right above the fermenting beer. Evolving CO2 will be pushing out almost exclusively air at the beginning. I can only guess, though, that this is the (main?) reason for this discrepency.

Anybody notice anything else? Comments & criticisms from everybody welcome!

 

[doug293cz]

I tried. It won‘t let me choose the excel file to upload/attach, at least not on my iPad…anybody know why?

A simple way to upload a file type that is not supported for attachments on HBT is to put the file(s) in a .zip file. I use this trick for spreadsheets frequently.

Brew on

 

[joshbuhl]

Could you shar the actual excel spreadsheet? I'm lazy

A simple way to upload a file type that is not supported for attachments on HBT is to put the file(s) in a .zip file. I use this trick for spreadsheets frequently.

Brew on

Here you go. Thanks for the zip-tip @doug293cz. Nuthin‘ to say about the new corrections? Figured you‘d dig those…

 

[doug293cz]

Nuthin‘ to say about the new corrections?

I need to be in a certain mind set, and simultaneously have time available, before I can crawl thru a bunch of math that I don't use everyday. That hasn't happened yet.

Brew on

 

[AlexKay]

I know I've said this before, and I may have in fact said it before in this thread.

ppb in these calculations refers to partial pressures, or number of molecules: at 1 ppb O2, one out of every billion gas molecules in the container is dioxygen.

ppb as it is usually used in beer packaging refers to the proportion of oxygen dissolved in the beer, by weight. So at 1 ppb O2, there is 1 microgram of dissolved oxygen per kilogram of beer.

There's roughly a factor of 1000 -- working in your (and your beer's) favor -- separating these two meanings.

Say you purge a 5-gallon keg to the point where it has 1 ppm (that's "m" as in million) O2. That is to say you have a volume of 20 liters and an O2 partial pressure of 1x10^(-6) atm, and the ideal gas law tells you that this means there is 25 micrograms of O2 in that keg.

Now fill that keg with beer, somehow without displacing any of the oxygen in it. 19 liters of beer weighs (at least) 19 kg, and if all 25 micrograms of O2 are dissolved in 19 kg of beer, you get a dissolved oxygen concentration of 1.3 ppb (that's "b" as in billion.)

Now, if a professional brewery works very hard and spends a great deal of money on the very best equipment available, they might get 10 ppb dissolved oxygen in a can or bottle of beer. So far as I know, 1 ppb is not possible with today's equipment.

All of which is to say that once you're pretty sure you've purged your keg to part-per-million O2 partial pressure, any further improvement is not meaningful. You're virtually 100% certain to have other sources of O2 to worry about that are much larger.

Returning a moment to the sentence "somehow without displacing any of the oxygen in it" ... in fact, when you transfer beer into the keg, you will be pushing at least some of the gas in the keg out. If you can manage to mostly fill the keg, you could be looking at another 10-fold improvement. If you make efforts to completely fill (or overfill) your keg, you could do even better.

 

[joshbuhl]

ppb in these calculations refers to partial pressures, or number of molecules: at 1 ppb O2, one out of every billion gas molecules in the container is dioxygen.

ppb as it is usually used in beer packaging refers to the proportion of oxygen dissolved in the beer, by weight. So at 1 ppb O2, there is 1 microgram of dissolved oxygen per kilogram of beer.


All of which is to say that once you're pretty sure you've purged your keg to part-per-million O2 partial pressure, any further improvement is not meaningful. You're virtually 100% certain to have other sources of O2 to worry about that are much larger.

Yes, exactly. This is all just for the fun of getting the calculations as correct as possible.

As far as practical applications, there‘s a simple way to get *much* better results: despite these amazing results of purging an air-filled keg, you still fill it with sanitizer first, push that out and then let the rest of the fermentation CO2 run through. Then you‘re way below 1ppb by orders of magnitude, and you don‘t have to go crazy about getting the last few ml of headspace in the keg filled with sanitizer.

Doing this, you could probably daisy chain and purge multiple 20l kegs with one fermentation, which doesn‘t work if they‘re all air-filled.

 

[mashdar]

At some point all your hoses and connections become the weak point. One should consider diffusion rate into tubing if going to that extreme.

 

[mac_1103]

At some point all your hoses and connections become the weak point. One should consider diffusion rate into tubing if going to that extreme.

EVABarrier is good enough for months in the kegerator so I figure it should be good enough for a couple of weeks in the fermenter.

 

[Bassman2003]

I little active yeast would be the best insurance. Just transfer with like a gravity point remaining. Sort of like a mild bottle conditioning that will consume and extra O2 that makes it through.

 

[joshbuhl]

I little active yeast would be the best insurance. Just transfer with like a gravity point remaining. Sort of like a mild bottle conditioning that will consume and extra O2 that makes it through.

Yeast oxygen scavenging in the bottle is better than nothing, but unfortunately DO has done it's damage and is already bound up within 20 minutes (which is why you have to measure DO immediately fter packaging and can't take a bottle to a friedly brewery with a DO meter) and I haven't yeat read anything that yeast is quicker in scavenging.

 

[Bassman2003]

That is why I said insurance. To eat up the little bit that comes in from transfers, seals etc...

 

[mashdar]

EVABarrier is good enough for months in the kegerator so I figure it should be good enough for a couple of weeks in the fermenter.

Oh, yeah, I just mean you hit a point where theory and practice diverge. Luckily it appears to be well below an acceptable level.

Here's another one: Every time you attach a quick disconnect, there may be ~1ml of air above the gasket between the mating QD parts. That 1ml of 21% O2 into a 21L keg would yield 10ppm O2. Yet another massive source of O2 compared to the ppb discussed above. Luckily, you'll displace most of that with beer, and the remainder will give <10ppb in the beer because the density is so different.

(I've yet to find a non-crazy way to purge the QD interface...)

 

[DuncB]

If the keg has some pressure greater than the tube and QD connect and then vent the tube would that help?
Or have far end of tube QD open as you connect to a positive pressure fermenter or keg.

 

[Bassman2003]

If the keg has some pressure greater than the tube and QD connect and then vent the tube would that help?
Or have far end of tube QD open as you connect to a positive pressure fermenter or keg.

Sadly, no. It seems gas laws and movement is kind of a black box. Oxygen and CO2 and other gasses can coexist and move freely outside of pressure.

Now physical movement of the space can deter *some* gas from coming in but it is a challenge no doubt. (think disturbing a mud bottom in a pond and trying to direct the sediment in a certain direction). If one has the choice of lifting a PRV or having 6 feet of EVA barrier connected to a gas post with positive gas coming out and bleeding off at the end of the tubing, the tubing will be a better approach as it is farther away from the source.

 

[tracer bullet]

I finally started using the fermentation gasses to purge my kegs, after following this thread. This will be my 3rd brew doing so, the last 2 seem to have worked wonderfully.

@Bobby_M (BrewHardware) made the tubing. He welded some 1/2" SS tubing to Female Flare adapters in such a way that the nut still moves, to let me align things before I tighten them to the quick-disconnects. I later made the cut to one of the tubes to add a push-connect tube fitting (from McMaster) to help the assembly process, but in retrospect I really didn't need it. It still works but if anyone pursues something similar, FYI that this was not needed. Also, I originally had a different setup in mind with a host of adapters but he was kind enough to email me to ask what I was up to and offered a much better solution.

If it isn't obvious, it goes: Fermenter to keg, keg to Ball jar suck-back preventer, and then finally to some star san / iodophor combo.

(The fan is an aquarium cooler I use to do my chilling, but that's a different topic).

 

[Bassman2003]

Sweet. Yes Bobby is a big help.

 

[hooper]

I do similar, but I go fermenter to black (output) thinking CO2 is heavier than air.

 

[doug293cz]

I do similar, but I go fermenter to black (output) thinking CO2 is heavier than air.

Feeding the CO2 from the fermenter into the liquid out post (and "filling" with CO2 from the bottom of the keg) provides better removal of O2 from the keg. You have concurrent pushing of O2 up and out along with O2 mixing back into the lower part of the keg by diffusion. Feeding CO2 in thru the gas post will cause faster mixing of CO2 and O2 as the CO2 falls to the bottom of the keg before completely mixing with the gas already in the keg.

Brew on

 

[Bassman2003]

I do similar, but I go fermenter to black (output) thinking CO2 is heavier than air.

Yes, @tracer bullet, I did not notice before but it is better to run the gas into the liquid post and out of the gas post.

 

[tracer bullet]

Yes, @tracer bullet, I did not notice before but it is better to run the gas into the liquid post and out of the gas post.

I wondered about that, but then there's the whole "there's no such thing as a CXO2 blanket, it all mixes" which I believe in, and so I leaned towards it not mattering. Especially when we are talking a trickle of incoming CO2 taking several full days, at least. I feel like any existing O2 and incoming CO2 are going to constantly mix and it wont' really matter. I could of course be wrong.

 

[doug293cz]

I wondered about that, but then there's the whole "there's no such thing as a CXO2 blanket, it all mixes" which I believe in, and so I leaned towards it not mattering. Especially when we are talking a trickle of incoming CO2 taking several full days, at least. I feel like any existing O2 and incoming CO2 are going to constantly mix and it wont' really matter. I could of course be wrong.

Yes a persistent CO2 blanket is a myth, due to how gases spontaneously homogenize over time (a matter of minutes.)

It's the time factor that makes a difference, If you go back to post #3, I show that the rate of flow upwards thru the keg being purged is about equal to the speed of O2 diffusing into the CO2 at the bottom of the keg. This means that mixing will NOT be complete, and you will have at least a partial CO2 blanket while CO2 is being actively pushed into the keg. Once CO2 is no longer entering the keg, the gas inside will homogenize in about a half hour or so.

Because the dynamic flow effects are too complicated to deal with, when I did the calculations, I did it for the worst case of complete mixing of the CO2 with the existing gas in the keg. And the results showed that even with complete mixing, fermentation gas purging is highly effective. The fact that you don't have complete mixing if feeding CO2 into the bottom of the keg means that the purging is even more effective than the calculation predicts. If you feed the CO2 to the top of the keg, you will get more mixing, and the results will be acceptable, but not as good as they could be.

Brew on

 

[tracer bullet]

@doug293cz that makes sense, the time factor, and that reversing my setup could be better. I was thinking things were quicker, but perhaps not.

I'll go switch it.

 

[doug293cz]

@doug293cz that makes sense, the time factor, and that reversing my setup could be better. I was thinking things were quicker, but perhaps not.

I'll go switch it.

I wouldn't switch mid fermentation, but rather wait for my next batch.

Brew on

 

[Bassman2003]

I also think there is just a physical attribute as well. Bringing in the CO2 from the bottom means any force is pushing on all of the contents upward with the only escape being the small opening at the top. If the incoming gas is brough in from the top, any force will be pushing downward into the keg and the only escape will be through a tube at the bottom. So there is a chance some gas might not get pushed out because it stayed generally up the corner etc... New gas might go directly downward at straight out where we want the gases to mix and eventually clear the entire space and be replaced with only CO2 over time.