Oxygen uptake with vacuum racking

[Stressbaby]

I've been reading and learning more detail regarding SO2, trying to take a more thoughtful approach to SO2 management. One thing I've learned is that "gentle racking" introduces oxygen at 1-3mg/L. Splash racking introduces oxygen at 3-7mg/L (wine is considered saturated at around 7mg/L). So in addition to achieving the minimum protective levels based on pH and alcohol, and adjustments for binding, my SO2 additions also need to account for oxygen introduced at racking, generally with a ratio of 2:1 (SO2 addition : expected O2 uptake).

I use the All in One wine pump (@vacuumpumpman). My question is, what should be the expected rate of oxygen uptake with vacuum racking? I tend to think the rate is reduced, but not sure. Reduced how much? Any experimental or theoretical basis for an answer?

 

[mainshipfred]

I posted this on another thread. I was showing a commercial winemaker my set up he commented on the AIO and racking in the early stages of the wine. He advised me not to use the AIO early on because wine need O2 at this stage and the AIO by nature of vaccuuming out the air to transfer does not allow enough air (O2) to be introduced. Now this is his theory but it does reinforce yours. It's just a theory but it does make sense.

 

[cmason1957]

Not trying to be argumentative, but why would racking using vacuum introduce any more oxygen than racking using a siphon and gravity? Or to the second one introduce much less. Seems it ought to be about the same.

 

[Stressbaby]

Not trying to be argumentative, but why would racking using vacuum introduce any more oxygen than racking using a siphon and gravity? Or to the second one introduce much less. Seems it ought to be about the same.

Actually, it possibly could go either way. If the Ai1 pulls 75% of the air out (making the number up, not sure), then only 25% of the oxygen available for uptake is actually available.

On the other hand, my Ai1 actually does the "falling film" version; so that would seem to work in the opposite direction.

 

[mainshipfred]

Again it's just a theory but when you pull the vaccuum in the receiving vessel there is no longer any, or a reduced amount of, air in the vessel to mix with the wine. The only air in the transfer vessel would be the dissolved O2 already in the wine. Kind of makes sense to me but no proof.

 

[sour_grapes]

Only theory, but....

I do not believe that the pressure in the receiving vessel is reduced to any substantial degree.

You are starting with an empty receiving carboy at atmospheric pressure (i.e., filled 100% with air). You pump a little bit of the air out, and the wine in the filled carboy starts moving into the receiving carboy. Consider how much (or how little) pressure differential there must be between the carboys to make the wine move. The minimum would be something like 6" water column, or 0.2 psi (which is a reduction of approximately 1.5% of an atmosphere). That is the minimum, but say it is, I dunno, 10 times more. Okay, so you may possibly reduce the pressure in the receiving carboy by 15%. At this pressure differential, the wine would be squirting in there pretty good, as if you were dropping it from a height of 5 feet.

In fact, this gives a crude way to estimate the pressure differential. You take a look at the wine coming into the receiving carboy and splashing on the side of your vessel. Try to estimate from how high you would have to drop some water for it to be moving that fast. Take that number and divide by 34 feet. That is percentage reduction of the pressure in your receiving carboy.

So, in summary, I agree with Craig, and do not think there would be much difference compared to normal racking with a siphon.

 

[stickman]

The amount of air in the receiving vessel is proportional to the pressure, Steve may be able to comment here, but I would expect the pressure during racking to be only a few psi lower than what's required to get the wine to the receiving vessel elevation. I would estimate 15% to 20% reduction of oxygen available to contact the wine, so in a 5gal carboy you go from about 4 liters of oxygen to 3.4. The surface area for contact and wine temperature are other big factors; more surface area and lower temperatures means higher oxygen pickup. The other wild card is the CO2 content of the wine. Fresh wine is off gassing significant CO2 that that oxygen pickup is probably negligible, so a wine that has been degassed and is 6 months old is much more sensitive to oxygen pickup. The ratio for SO2 to O2 is 4 to 1, so 4 ppm SO2 is needed for every 1 ppm of oxygen pickup.

Maybe some of the members that have the DO option for the Vinmetrica could run a few samples during racking to provide some real data for home winemaking techniques.

 

[mainshipfred]

The amount of air in the receiving vessel is proportional to the pressure, Steve may be able to comment here, but I would expect the pressure during racking to be only a few psi lower than what's required to get the wine to the receiving vessel elevation. I would estimate 15% to 20% reduction of oxygen available to contact the wine, so in a 5gal carboy you go from about 4 liters of oxygen to 3.4. The surface area for contact and wine temperature are other big factors; more surface area and lower temperatures means higher oxygen pickup. The other wild card is the CO2 content of the wine. Fresh wine is off gassing significant CO2 that that oxygen pickup is probably negligible, so a wine that has been degassed and is 6 months old is much more sensitive to oxygen pickup. The ratio for SO2 to O2 is 4 to 1, so 4 ppm SO2 is needed for every 1 ppm of oxygen pickup.

Maybe some of the members that have the DO option for the Vinmetrica could run a few samples during racking to provide some real data for home winemaking techniques.

I was hoping you and Paul would chime in. I am doing some lab work today and have the DO analyzer. So what do you suppose I shoould do? Do I take an initial reading and then one after an AIO racking? My siphon broke some time ago and I never replaced it since I have the AIO so I can't compare that. Secondly do I use the splash rack cane or a cane to the bottom of the receiving carboy?

 

[Stressbaby]

Thanks to everyone for the thoughtful replies. Makes sense.

The ratio for SO2 to O2 is 4 to 1, so 4 ppm SO2 is needed for every 1 ppm of oxygen pickup.

This is an area of potential confusion for me. It's been a long time since AP Chem. The SO2:O2 ratio is 4:1 by weight, 2:1 molar, is that right? I'm wondering then if here in this link some of the calculations are wrong. For example, in 13.4 he has 3mg/L O2 uptake due to racking, but accounts for this racking O2 uptake with only 6mg/L SO2 when really it should be 12mg/L, and the total should then be 56mg/L

((4 * 3mg/L racking O2) + 16mg/L target SO2 for wine at pH 3.1) * 2 assuming 50% binding

 

[stickman]

I would think you should do what you normally do. If you are ready to rack some wine, just follow your typical procedure and test DO before and after, also record the wine type, age, temperature, and any other relevant data. It will take some time to collect enough data before a meaningful comparison can be made.

 

[Stressbaby]

I was hoping you and Paul would chime in. I am doing some lab work today and have the DO analyzer. So what do you suppose I shoould do? Do I take an initial reading and then one after an AIO racking? My siphon broke some time ago and I never replaced it since I have the AIO so I can't compare that. Secondly do I use the splash rack cane or a cane to the bottom of the receiving carboy?

My understanding is that the reaction takes several days.

Example:
You should be able to take an SO2 measurement before racking; rack the wine; take an SO2 measurement several days later, and based on the drop, calculate the rate of oxygen uptake with Ai1 racking. Lets say you start with 40mg/L. You rack, then wait a week, then you remeasure the SO2. You're going to have a tiny drop just from time (waiting a week) and oxygen absorption from the ullage, but shouldn't be more than 1-2mg/L total. If you find a week later the level is 29mg/L, probably 10mg/L of that drop was due to racking. Then assuming 4:1 ratio by weight of SO2:O2, we'd estimate 2.5mg/L of oxygen absorption from the Ai1 racking process. I'd test this but all I have is titrettes (Ripper method) and I don't think it is accurate enough to be helpful.

 

[Rice_Guy]

Only theory, but....

I do not believe that the pressure in the receiving vessel is reduced to any substantial degree.

You are starting with an empty receiving carboy at atmospheric pressure (i.e., filled 100% with air). You pump a little bit of the air out, and the wine in the filled carboy starts moving into the receiving carboy. Consider how much (or how little) pressure differential there must be between the carboys to make the wine move. The minimum would be something like 6" water column, or 0.2 psi (which is a reduction of approximately 1.5% of an atmosphere). That is the minimum, but say it is, I dunno, 10 times more. Okay, so you may possibly reduce the pressure in the receiving carboy by 15%. At this pressure differential, the wine would be squirting in there pretty good, as if you were dropping it from a height of 5 feet.

In fact, this gives a crude way to estimate the pressure differential. You take a look at the wine coming into the receiving carboy and splashing on the side of your vessel. Try to estimate from how high you would have to drop some water for it to be moving that fast. Take that number and divide by 34 feet. That is percentage reduction of the pressure in your receiving carboy.

So, in summary, I agree with Craig, and do not think there would be much difference compared to normal racking with a siphon.

Your logic is good,

* Steve's Ai1 can do a little over minus 29 inches mercury (or a little less than 1.0 atmosphere). I would never (again) rack with that much force since it looks kinda looked like an aerosol can spraying with a splash racking setup. You can assume 100% saturation at the vessel's pressure. With a submerged racking cane in the receiver you would be safe from the splashing effect, you can transfer 5 gallons in a few minutes. Gentle is good and I do not open the flow up all the way. My target is to transfer in roughly 10 to 15 minutes so I reduced tubing size to 1/4 inch rigid poly. At the start I may pull 18 inches of water (wine) column at the end I am probably 6 inches of water column as noted above. Yes I expect I have significant oxygen in the system, if I was going on memory I might be -2.5 psig (.15 atmosphere) controlled by pulsing the pump on/off.
* Steve's pump is good enough that is can implode a plastic carboy. With a glass carboy breakage will be a numbers game with bigger carboys having a higher risk. I would not run the pump till it is close to one atmosphere vacuum!
* The next post mentions CO2, at reduced partial pressure the CO2 is coming out of solution. CO2 is heavier than the oxygen and the nitrogen so it will displace essentially all the gas in the carboy, ie you would be using an inert environment so don't worry about it.
* End conclusions, grandpa did not have as many gauges as I have and could make drinkable wine. I routinely try to minimize oxygen with whites/ fruits but it might not matter for 1 year shelf life.
It would be interesting to have a DO probe and split a juice bucket into 1) splash racked w all transfers 2) racking with a submerged delivery cane 3) racking at 1/4 atmosphere in the receiver 4) racking with Steve's vacuum plug and 6 inches head space 5) ?? 2 more 1 gallon tests ??

 

[stickman]

@Stressbaby the article you indicated by Ben Rotter uses a factor of 2 for the SO2 binding with oxygen, but as usual with winemaking there is other literature with conflicting data. Pambianchi uses a factor of 2.5, Todd Steiner of OSU Eonology uses a factor of 4.

 

[Stressbaby]

@Stressbaby the article you indicated by Ben Rotter uses a factor of 2 for the SO2 binding with oxygen, but as usual with winemaking there is other literature with conflicting data. Pambianchi uses a factor of 2.5, Todd Steiner of OSU Eonology uses a factor of 4.

Yes, but he mixes it up, doesn't he? He says 2 molar, then in the examples he uses two by weight. Molecular weight of SO2 is 64, O2 is 32 so that off by a factor of 2.

 

[stickman]

I believe he indicates that in practice the theoretical reaction doesn't go to completion sighting several competing reactions, so he uses the lower figure of 2 (at least that's what I got out of the article, I could be wrong).

Clark Smith, Sulfur Dioxide Basics Revisited March 2012, indicates the following:

"In anticipating residual molecular SO2, the winemaker is obliged to estimate the decline in free SO2 brought about by the reaction with oxygen picked up during bottling. Since oxygen weighs 16 daltons, and SO2 64 daltons, there is a 1:4 stoichiometry to this reaction. For every 1 ppm of dissolved oxygen (DO) picked up during bottling, 4 ppm of SO2 will be reacted, usually within a few months. This should be subtracted from the anticipated FSO2, along with an additional 6 ppm if a cork is used (they dispense compressed air), before calculating the residual molecular SO2 in the bottle, which can vary widely based on pH."

I like how he indicates adding an additional 6ppm SO2 just for using a real cork.

 

[mainshipfred]

Way over my head but here are the results of the DO tests.

First wine 2018 Petite Sirah in 25 liter barrel MLF complete around 11-15-18 barreled 11-30. SO2 was at 13 ppm and brought it up to 45 ppm prior to racking. Racked from barrel to glass with 3/8 tubing.
Initial DO .332% after racking with the splashing cane 3.45%

Second wine 2018 Touriga in 7 gallon glass, MLF completed the same time. SO2 was at 19 ppm and brought it up to 40 prior to racking. Racked from glass to Barrel with 1/2" tubing.
Initial DO 3.31% after racking 7.82%

Both wines were at 57*.

 

[stickman]

Those numbers are low, maybe that's a good thing. I'm assuming these wines were not previously degassed. It looks like you're getting about .3 to .4ppm pick up. More data is needed to get a handle on how it may vary, but it's a start; @mainshipfred your effort is greatly appreciated.

Some data from a 2014 OSU study indicates the following:

 

[Stressbaby]

If my math is right that is 0.26 and 0.45mg/L which is well under what I see online. Good stuff, thanks.

 

[Ajmassa]

Interesting numbers BEFORE the test. The barreled wine had lower DO than wine in glass. Was there more rackings done to the Touriga than the PS?

 

[mainshipfred]

There was no attempt to degas since this stuff is going to sit at least til fall. It is strange the barrel had a lower DO and it was racked one more time to go to the barrel after MLF. The OSU study does not have a stat on vaccuum racking. I wonder if that makes a difference.