TPO Measurement: the Importance of Methods Validation

Total Package Oxygen (TPO) can be such a confusing topic that it’s worth an ongoing discussion. Since real-life examples are often the easiest to grasp, today we’re going to look at TPO from the perspective of a brewer who is actively trying to troubleshoot his new seamer. Even if you aren’t canning, you should be able to apply this information to problems with a bottle filler.

In this case, the Brewer has purchased a new seamer that is specifically promoted as keeping TPO low, and yet his TPO measurements after seamer installation are nowhere near as low as the seamer company says they should be. When the brewer and I talked, I was immediately struck by what the seamer people had provided as the target TPO: 20 ppb, which is super low by any standard. So our next step was to try and understand the origin of that number.

The seamer manufacturer really wasn’t able to tell our brewer why his values were not meeting expectations for the seamer. After some initial probing, it turned out that the seamer people were sincere in their beliefs about the abilities of their seamer, but they didn’t have a lot of experience with beer, and had based their numbers on data from only one other customer. Knowing this, we then formulated some questions for that customer.

We’re still waiting to hear back, but here’s what we’ve asked:

  1. What was the reading from the package when measured on their dissolved oxygen analyzer?
  2. Was the package shaken or unshaken?
  3. If the package was shaken, did they calculate the TPO from the dO2?

Here’s what the answers to each of these questions should tell us:

  • If the answer to the first question is in the range of 7 to 10 ppb, then it really is in the realm of possibility that their TPO is 20 ppb, as long at the instrument was validated (more on that in another post), the package was shaken, and they did a correct TPO calculation.
  • If the answer to the first question is 20 ppb, then we need to know if the package was shaken. If it was shaken then it was at equilibrium, but did they calculate the TPO?
  • If the package was unshaken, then the value they were calling “TPO” was actually just the dO2. This represents only the oxygen concentration of the beer coming into the filler and the pickup in the filling process. To get the TPO, they still needed to measure the dO2 on shaken packages and calculate TPO based on the dO2, package temperature and the liquid and headspace volumes.

It will be interesting to see what happens, but my guess is that the numbers provided to the seamer people – and thus to our brewer — were not TPO. They may be shaken or unshaken DO2 – both are possible.

My final thought is that when someone tells you a dissolved gas value upon which you must rely, ask how it was derived. The reliability of measurement is only as good as the validity of the method by which it was obtained.

Do I Need to Let the CO2 in My Cans “Rest”?

I was helping a brewer evaluate the dissolved oxygen pickup from his filler before his can ends were applied. He was wondering if there was some sort of apparatus he should be using to cover the headspace of the can. I recommended that he just wait until the can end was applied and measure the unshaken dissolved oxygen though the closure. Then he asked the question that is the topic of this post: did he need to let the CO2 in the can “rest” prior to measuring his dO2? He had apparently been told that in order to get an accurate dO2 measurement, the can should sit open to the atmosphere for 20 minutes in order to let the COdo “something” – the intent was not really clear.

I had never heard of this, but my instinct told me it was a myth, and that it could in fact lead a brewer to underestimate the dissolved oxygen in his can of beer. When a fully carbonated package sits open to the atmosphere, the partial pressure of the CO2 is significantly higher than the oxygen content of the outside environment. As the beer sits the CO2 will rise out of the can and carry oxygen with it.

I wanted to be sure my thinking on this was clear, so when I got back to my lab I sacrificed a recently packaged beer in an aluminum bottle to be sure. First I removed the 28 mm closure and then I immersed a 12 mm probe. Since I was using an optical probe, I didn’t need beer flow to do my monitoring. Upon immersion the probe O2 reading quickly equilibrated to about around 20 ppb and stayed at this value for about ten minutes. As the CO2 began to escape from the can, the probe value steadily dropped to around 10 ppb, then slowly rose over the next two days to about 50 ppb. This was only one experiment, but the data was logical, and I’ve never had another reason to test this notion.

My final thought is that I believe the “resting” theory is myth, but if anyone has a different explanation, I’d love to hear it.

%d bloggers like this: