Calculating Liquid Yeast Viability Changes Over Time
Hi blog readers, I’m Braden Masur and I started working at Escarpment Labs in August of 2020 as a Co-Op student from the University of Guelph. During my time here so far, I was given the opportunity to evaluate shelf life of Escarpment Labs products. The current best before dates were originally set in 2017 based on some preliminary testing and observations performed on older systems. Since then, we have made huge strides in improving our processes (both manufacturing and packaging) and decided to re-evaluate the impact these changes have had on the shelf life of our products.
In order to serve brewers best, we maintain inventory of many of our yeast strains. While we prioritize freshness, and the vast majority of yeast goes out the door within a week of harvest), we do hold inventory of popular stains in order to absorb fluctuations in demand and ensure brewers get the yeast they need when they need it. Domesticated yeasts, such as clean ale and lager strains, have an internal shelf life of 30 days. Typically, these strains do not enjoy being stored for long periods of time as they want to get brewing! We have classified these strains internally as a Risk Level 1, as they are POF- and STA1- with a low risk of contaminating other products (see other blog posts for more information). Level 2 (POF+) and Level 3 (STA1+) yeasts tend to be more similar to wild yeasts, as they are better at adapting to their surroundings and are fine being stored for longer periods of time. Kveiks, as we all know, are just crazy strains and can survive anything and also have a longer shelf life!
For our homebrew pitches, the shelf life is extended for all products to 4 months, giving our homebrew retail partners more time to sell the yeast. We always recommend homebrewers make a starter if the yeast is within 1 month of expiration for best results. However, we’ve spot-checked some kveiks and found them over 90% viable after 6 months! We wanted to revisit yeast storage viability over time, as this data is very useful to brewers, who sometimes can’t use the yeast they receive on the initial timeline they had in mind. This knowledge will help us help our customers better - whether it’s a home brewer with old-ish yeast or a professional brewery, knowing how our yeast changes over time will enable us to provide more guidance into pitching rates, re-pitching practices and determine when it's time to get a fresh batch of yeast.
To begin this project, I searched through our internal database and read up on our current quality assurance protocols and the current standards that our yeast must meet before we sell it to our customers (95%+ viability, 3 rounds of QC). I then designed an experiment to analyze four popular yeast strains based on current shelf life and current availability. The selected strains were: Vermont Ale, Cali Ale, Foggy London Ale and Laerdal Kveik.
The experiment was performed using two batches of yeast for each strain in order to increase confidence in the results – or unfortunately in some cases; decrease confidence (LOL). Nine 10mL samples were taken from each batch, sealed in sterile tubes, and stored in a fridge at ~3°C for the duration of the experiment. This was to ensure that the same sample was not used for testing and that there would always be an accurate, undisturbed sample to test for the duration of the project (2 months). Once the samples were collected, weekly cell counts were performed with a microbiological differential stain, Erythrosin B, in order to determine the viability.
Test fermentations (small scale batches of beer) were also performed biweekly in order to ensure that the yeast was still performing to achieve the correct level of attenuation. This experiment was conducted for 8 weeks for each yeast sample starting on the day it was harvested from the tanks. Therefore, the time points for cell counts and test fermentations are different from one another, but all were tested for the same duration under the same conditions.
Sample enumeration was performed following our internal quality assurance protocol whereby a 1:200 dilution was performed and then counted on a hemocytometer to determine the cell viability and cell density. Live cells appear clear and dead cells appear red (dead = red is a handy reminder). Sample viability was calculated by dividing the live cells by the total cells. The cell density was calculated by the number of live cells/mL. Using the cell density, a pitching rate is calculated for the test fermentations. I also normalized the pitching rate to account for any dead cells over time. Note also that the test fermentation wort for this trial was highly fermentable (low in residual dextrins and starches).
Without further ado, here is the data: Vermont Ale samples showed mixed results with one batch sample dropping below 95% viability between 45 and 59 days. Overall, Vermont performed quite well in our trial!
Foggy London Ale also had mixed results with viability dropping faster than Vermont., but still around 90% after 60 days of storage. The apparent attenuation for both batches was consistent throughout the experiment.
Cali Ale samples were above 95% viability through 8 weeks of testing and fermentation cleared QC with consistently high apparent attenuation (~90%) - indicating that this yeast performs quite well at 60 days old!
For Laerdal Kveik, one of the batches decreased in viability faster than the other, but in a similar manner to Foggy London. As many brewers know, Kveik does really well with lower than normal pitch rates, so even a 4 to 6 month old pitch of Kveik can often give good results if the yeast gets lots of oxygen and nutrients.
Across the whole experiment, the attenuation of the stored yeasts in the test fermentations was very consistent.
We’re going to avoid making any extrapolations of the data for timelines beyond 60 days, but hopefully this project shows that lovingly-propagated yeast can survive extremely well with limited drop in viability over 60 days for many strains. If the loss of viability is linear, you can expect to see older pitches of Escarpment yeast (strain dependent) at surprisingly high viabilities even as the expiry date approaches. We will follow up and check some ancient homebrew pouches to provide some more data.
I want to say a big thank you to Luisa Muhleisen, Richard Preiss and Alex Mitro for providing guidance while planning the experiment and interpret data. Overall, this was a really valuable experience. I learned a lot about planning and executing a study and interpreting data. I was able to get a handle on the organization that is required to plan an experiment and meet deadlines, with the exception of a few missed timepoints. I also developed my ability to interpret raw data and draw conclusions from the results. These are essential skills that I have used when designing my next project which involves molecular techniques to edit a gene in some yeast! So, stay tuned for that!