A Comparison of Wine Transfer Methods - Oxygen Uptake and SO2 Loss

How to rack wine

A Comparison of Wine Transfer Methods: Oxygen Uptake and SO2 Loss

by Jane Jackson, Copyright by The Beverage People 2022

There are many points in the winemaking process where minimizing oxygen exposure is completely under the control of the winemaker and can have a big impact of the resulting quality and longevity of the wine. One such instance is during racking. The importance of this step led us, at The Beverage People, to experiment with different methods of transferring wine and measuring the changes in SO2 and oxygen uptake with each method.

 Our Experiment in Racking

Racking is necessary at multiple points and, when poorly executed, can damage a wine that has otherwise been obsessed over to that point. Three common methods of racking include pouring from container to container (or splash racking), siphoning with a racking cane or auto siphon, and pumping. The use of a pump is overkill when making small amounts of wine, and siphoning is generally impossible when making large amounts of wine such as in a barrel or stainless tank.

At The Beverage People, we set out to perform an experiment with those three methods of racking to determine the differences in their impact on oxygen uptake and loss of free SO2. Our staff gathered for the experiment and we made a few hypotheses before beginning.

Our Hypotheses:

Armed with a Vinmetrica SC-300 and their add-on Dissolved Oxygen Meter, we prepared to measure the SO2 and the dissolved oxygen before and after the various racking methods.  The dissolved oxygen tells us how much SO2 is needed to counteract the oxygen uptake during the racking process. 

We used a Vinmetrica SC-300 to perform our lab tests.


We started with a 5 gallon keg of 2021 Viognier.  The wine had been kept under pressure in the keg in cool storage for months.   The day before we experimented with transfer methods, we used our Vinmetrica SC-300 to test the free SO2 in the wine.  White wines, lacking the preservative tannic qualities of red wines, require higher levels of SO2 additions for stability.  Consider the following reference table which reports the ideal free SO2 levels for both red and white wines at various pH levels.

pH 0.8 ppm 0.5 ppm
White Wine
Red Wine
11 ppm
7 ppm

 Molecular SO2 needed for Stability (ppm)

We were happy to see that this keg of wine was at 40 ppm of SO2- a good amount for the long term storage of the wine and also a decently protective amount for the wine as we experimented with the transfer methods.  Had the SO2 been lower than ideal, we would have made an addition, especially knowing we would be racking it the next day.  It is always better to check and adjust SO2 soon before racking so that the wine is properly protected during the disruptive procedure.  Since we had just tested SO2, we had a baseline against which to judge the racked samples after racking.  Next we had to calibrate the Vinmetrica D.O. (dissolved oxygen) Meter.  This Vinmetrica add-on can be used with the SC-200/300 Meters.  It is cumbersome to work with but, if done carefully and properly, can offer very useful information regarding oxygen exposure/uptake of the wine.  After the D.O. meter had been calibrated, we tested the D.O. in our control keg.  It was at 0.1717 ppm.

Next, we chose common tools available to home winemakers for racking and prepared our experiment as follows:

Method 1 - Splash Racking.  A stainless baster-type thief to execute the splash racking (we didn’t actually want to torment our delicious Viognier by splash racking the whole keg).  We used this to vigorously squeeze our sample from the thief into our sample beaker.

Method 2 - Siphoning.  An auto siphon and attached tubing to gravity rack from keg to sample beaker.

Method 3 - Pumping. The Vintage Shop Variable Speed Diaphragm Pump fitted with rigid racking tubes on both ends, operated at medium speed to pump from keg to sample beaker.

After each method of racking, we tested SO2 and D.O. in our samples.  The results below had a couple of surprises for us.  Remember, we started with our control sample at 40 ppm of SO2 and 0.1717 ppm of D.O.

 Method 1   Splash Racking   36 ppm   1.94 ppm 
 Method 2   Siphoning   38 ppm   0.47 ppm 
 Method 3   Pumping   38 ppm   0.39 ppm 

The splash racking was obviously the most disruptive to the wine, resulting in greatest loss of SO2 and greatest uptake of oxygen.  The surprise, to us, was that the pump was actually much more gentle than we hypothesized- about as gentle as the gravity-fed auto siphon.  This gave us great confidence in recommending this particular pump as a gentle, but efficient way to move larger volumes of wine with minimal loss of SO2 and no serious oxygen uptake.

 Lesson Learned

With the free SO2 and D.O. numbers gathered in our experiment, we can calculate the amount of SO2 that must to added to the wine to compensate for the agitation losses and oxygen uptake. 

The following is a key fact needed for the calculation: For every 1 ppm of oxygen uptake, an additional 4 ppm of sulfite is needed to bind it.

With this knowledge, the winemaker can properly sulfite ahead of time based on the chosen method of transfer so that no additional oxygen uptake is incurred. 

For instance, performing a gentle method of transfer such as with an Auto Siphon or Pump requires an addition of 2 ppm of sulfite to compensate for the agitation loss from racking plus another 2 ppm for the oxygen uptake for a total addition of 4 ppm of sulfite.

A more vigorous method of transfer, such as splash racking, would require 4 ppm for the agitation loss from racking plus another 8 ppm for the oxygen uptake for a total addition of 12 ppm!  Without these additions, a wine that was otherwise protected becomes vulnerable to oxidation and a reduction in quality.

This article is part of a 3 part series on oxygen management in wine.  Next, enjoy our discussion of Managing Oxygen During and After Wine Bottling.

Learn best-practices in limiting oxygen exposure in your wine bottles to extend the shelf life and improve wine quality.