Total Package Oxygen 101 – What You Need to Know to Calculate TPO

The following post is the second in a series on Total Package Oxygen (TPO).

We’ll cover two topics in this blog post: 1) The parameters you need to do a TPO calculation and 2) Best practices for getting packages to equilibrium.

First, TPO calculation parameters. It’s a paradox, but the actual calculation for TPO is both complex and simple. The complexity part arises from bits of Henry’s Law, Boyle’s Law, physical properties of water vapor, and atmospheric barometric pressure. But these complexities are incorporated in the calculations, so the simple part is that in order to do our calculations, all we really need are these easily obtained variables:

Dissolved oxygen content
The package must be shaken, so the gases in the headspace and liquid are at equilibrium.

Liquid volume
Accuracy of liquid volume is not very critical, so using the average package fill will give a statistically valid TPO.

Headspace volume
The more accurately we can determine the headspace volume the more accurate the measurement. Knowing it to within 1 mL is best.

Package temperature
There are a lot of variables that rely on the package temperature built into the TPO calculation, so it is best to measure to within 1 oC.

So now we know our important package parameters. But how do we know we’re measuring our packages at equilibrium, so our calculations will be correct when we plug them into our formula?

We shake (equilibrate) packages in order to mix dissolved gases in the liquid with air trapped the headspace after the closure has been applied to the package. The question is: how long to shake the package?

People have tried all sorts of things through the years to determine the proper amount of shaking time. In the end it turned out that one of the best methods was to take old cans of beer with less than 2 ppb of dO2 and inject known volumes of air into the cans. By doing that and then shaking for various amounts of time, it could be determined how long it takes to be able to account for 100% of the oxygen when measuring TPO using the dissolved oxygen concentration and the other parameters outlined above.

Using a platform shaker at a minimum of 180 revolutions per minute, room temperature packages take 3 minutes to equilibrate. Packages at 5 oC, on the other hand, do best with 5 minutes of shaking. The basic rule of thumb is that the colder the package, the longer the shaking. When shaking cold packages, it is also important to remember that the temperature of the package is constantly trending toward room temperature, so the best practice is to shake continuously until you are ready pierce the package.

One short note on bottle conditioned beer: I’ve seen yeast so active upon packaging that they can consume statistically significant amounts of oxygen in the time it takes to shake a bottle, so I’ll have some hints about how to deal with that next time.

My final thought is to always shake thoroughly and consistently. Using some sort of a mechanical shaker will decrease operator error and insure against inconsistent results.



The Units of Gas Phase Measurements Demystified!

I don’t think there is anything more confusing than the units used when measuring gases dissolved in liquid compared to the units used when measuring in the gas phase.  Here is one more post on CO2 before I completely exhaust the topic – pun intended. Let’s see if I can make it clear.

Dissolved gas is always expressed in terms of weight per volume. When someone says that they have 10 ppb or 0.01 ppm of dissolved gas in their beer, what they really mean is that they have 10 micro grams per liter or 0.01 milligrams per liter dO2 content.  Again: this is gas expressed as weight of gas per liter.

Gas phase measurements, on the other hand, are always expressed as a volume of gas per volume of total gas. This can be in units of percent, bar (volume barometric,) mbar, atmosphere, or — the two most confusing units of all – ppm and “%CO2 purity.” Lets talk about those two:

  • Gas phase ppm is a comparison of the gas being measured to all of the gasses in the sample. If you have 100% O2, then you will have 1,000,000 ppm O2. Since it is cumbersome to talk about 1% or more as ten thousand parts per million, we use units of percent. But when we get to trace levels below 0.1%, then we start throwing around units of ppm.
  • Percent CO2 purity is all of the gas being measured that is only CO2. Here’s a table to help you sort it out.
Unit of Measurement O2 Content


O2 Content


O2 Content


O2 Content


O2 Content


Percent 100 1.0 0.1 0.01 0.001
bar 1.013 0.010 0.001 0.0001 0.00001
mbar 1013 10.13 1.01 0.10 0.01
ppm 1,000,000 10,000 1000 100 10
Atmosphere (ATM) 1.0 0.01 0.001 0.0001 0.00001
% CO2 Purity 0 95.0 99.5 99.95 99.995

Did you follow? Just in case I lost you, %CO2 purity assumes that all of the impurity in CO2 is air.  Since nitrogen comprises 4/5th of air, you have to take the oxygen content, add back in the nitrogen, and subtract all the air from the 100%.

My final thought is to add just a tad more data to compare ppm dissolved to ppm gas phase. Here’s a quick test: if you measure CO2 gas with an instrument reading, in the liquid phase, 0.001 ppm or 1 ppb dO2 at 20 deg C, what would be the equivalent gas phase reading?  At 20 deg C, the O2 content in the gas would be about 22 ppm. In terms of %CO2 purity, that would be about 99.99%.

O2 Impurity in Carbon Dioxide: How Much is Too Much?

I recently saw a post on a brewing forum where someone was wondering about air contaminated carbon dioxide and its impact on dissolved oxygen levels in carbonated beer. Air in CO2 really can raise the dO2 levels in your beer, so I thought it would be worth discussion.

Back sometime in the late 1980s (I think) someone passed along to me a table showing the exact amount of dO2 that would be picked up in beer if specific amounts of CO2 were injected into a process pipe or finished beer vessel. I’ve been hoping to find a copy of that article ever since, so maybe someone out there can help steer me to it. But in the meantime, with a tip of the hat to the original author(s), here is a copy of the table:

Co2 Injected O2 Impurity


O2 Impurity


O2 Impurity


0.5 V/V 7 ppb 35 ppb 142 ppb
1.0 V/V 14 ppb 71 ppb 284 ppb
2.0 V/V 28 ppb 142 ppb 567 ppb
Dissolved oxygen added to the beer during injection

So knowing all of this, what’s the best way to determine whether a CO2 supply is contaminated with air? There are two approaches. First is to simply measure your CO2 source in the gas phase using a low-level oxygen sensor that is accurate to at least 0.001%. The other is to measure the dO2 in your beer before and after CO2 injection. If you are measuring in the beer, use a measurement point that is furthest from the injection point so that the gas will have a chance to dissolve into the beer as much as possible before you measure. If you carbonate in a tank, just measure in the tank.

My final thought is that you may be doing everything right in the rest of your process, but if the CO2 you’re using to trim your carbonation is loaded with air, your beer may pick up a significant amount of oxygen.

Have you seen the article with the CO2 table? If you can point me to the journal and/or author, I’d be happy to post a reference.  Please leave the information as a comment below.

%d bloggers like this: