Friday, March 1, 2013

Side by Side Starters 4 of 4

There are three big take aways from this collection of data.
  1. For reasonable inoculation rates and gravities cell growth is not a function of inoculation rate or volume.  Cell growth is simply a direct function of sugar, and may be more accurately predicted by observing sugar consummation.
  2. Yeast taken from storage at 40°F (5°C) can out preform yeast taken from an active starter likely due to the glycogen content of the yeast.
  3. Attenuation is a function of both inoculation rate and gravity.  Higher inoculation rates and lower gravities lead to higher attenuation.


Cell growth observed over this wide range of inoculation rates and initial gravities showed cell growth proportional to the amount of initial sugar present.  In the case of the active culture of S-04 7.2 billion cells were grown per litter per initial degree Plato.  Considering that the average attenuation was 62% this indicates that 11.6 billion cells were grown per liter degree Plato of consumed sugar.  This tracks very well with the daily observations.  In a similar fashion the inactive starters grew 10.4 billion cells were grown per litter per initial degree Plato with an average attenuation of 68% making 15.3 billion cells grown per litter degree Plato of sugar consumed.  Both of these starters closely followed the equations derived from Ballings observation for starter cell growth.  Based on these observations the equation can be adjusted to account for the higher cell growth that is likely linked to proper aeration at the onset of inoculation.

Cells Grown = 14 * Volume of wort (Litters) * [Initial Gravity of wort (°P) - Final  Gravity of wort (°P)]


The yeast taken from refrigeration outperformed the yeast taken from a starter in terms of total attenuation and in cell growth.  When fermentation is allowed to run to completion yeast cells will build up a glycogen reserve.  It is possible that this extra glycogen allowed the refrigerated yeast to grow more cells given the same amount of sugar.  Further testing would be required to confirm this including measurement of the glycogen levels.


The percentage of sugar consumed seems to be a function of both the initial gravity and the inoculation rate.  A linear fit for these two parameters shows an excellent r squared fit of 0.9139 and 0.9795.  Although this shows wonderful correlation I am hesitant to say there is causation.  There are a other considerable factors, such as temperature, that are not accounted for here.  The equation also breaks down quickly at inoculation rates beyond the data set used.  It does seem reasonable at inoculation rates and gravities typically used for fermentation of beer.

The following equation can be derived from the linear fits:

A = Real Attenuation (as a decimal.  ie 0.71 = 71%)
G = initial gravity (in °P)
I = Inoculation rate (in million per ml)

A = 9.54E-4(G)-2.44E-4(G)(I)+4.23E-3I)+5.19E-1

    Shaped markers are measured data points. 
    The number of cells at zero initial gravity is the inoculation rate. 
    The "Active" chart represents data taken from an active starter.
    The "Inactive" chart represents cells taken from refrigeration.

    Real Attenuation in %sugar by weight.
    Each column is the number of cells used for inoculation. 
    Divide by 10ml for the inoculation rate.


  1. Thanks for the great work. I wonder if refrigerated yeast would still outperform yeast from a starter when pitched into partially fermented wort, such as when trying to re-start a stuck fermentation. Do you have any thoughts on this?

    1. That's a good question. From the daily data both started at about the same pace, but the refriderated yeast just kept going.

  2. Steven, thank you for the axis labels and the figure descriptions. Really appreciated. And thank you for your experiment as well. I am really fascinated by the fact that the refrigerated yeasts seem to grow to a higher concentration than non-refrigerated ones. Very well done. I as well don't think there is a correlation between the attenuation and the original gravity. I can imagine the amount of un-fermentable sugars (in %) in all the different original gravities would stay the same. By increasing the original gravity, the percentage of un-fermentable sugars will still be the same.

    On the other hand, don't you overpitch dramatically in all your experiments? A standard pitching rate for 10 mL of 15°P would be (1E6 cells mL-1 °P-1 x 15°P x 10 mL = ) 150E6 cells, right? And you used 290E6 cells 10 mL-1 as the lowest cell concentration. Are there any reasons for overpitching?
    Thanks again, cheers Sam

    1. Thanks for the suggestion on the chart labels.

      You are right that the percentage of each type of sugar are the same regardless of gravity, but there is more to attenuation than the types of sugar. In this case there was a pretty wide spread of attenuation.

      Yes, for an ale your pitch rate is correct. These were pitched to be more like a typical starter where the brewer is looking to double or triple the cell count. In a typical ale fermentation the cell count grows ten fold but takes about a week to get there.

    2. Thanks for your reply. Sure attenuation depends on the sugar composition. I was just pointing out that this aspect might be left aside in your experiment here since you used the same wort for every tube.
      Cheers, Sam

  3. So Woodlands research is showing that viability doesn't decrease very much over a months time at fridge temps. He also showed that stored yeast can out perform starter yeast on a small scale because the stored yeast has built up glycogen reserves. So, if I have a jar of yeast that has been sitting for about 4 weeks, would it be better to directly pitch at an assumed 1B cells per ml or do a small starter to activate the dormant yeast and then pitch.

    1. The yeast used in the test was in the refrigerator for about a month, so I would think you would be okay pitching it cold.

  4. Steve,

    Nice experiment and thanks for sharing your data.

    But I'm actually seeing the opposite when it comes to fresh and aged yeast. Fresh yeast (a day or two old) growths more cells per gram of extract than older yeast (5 days to a week old). The difference is quite noticeable. None of this has been published yet since I want to run a few more experiments in order to test my theory on yeast reserves. But I seem to be coming to a different conclusion than you.

    All my experiments are done with stirred starters. Maybe that makes a difference.


    1. Thanks Kai,

      That's very interesting, and makes more sense as it follows the conventional brewing wisdom. Even though this was 42 different fermentations they were all inoculated with the same yeast. I would like to run this again with a different strain under similar conditions. The "active" yeast was started on a stir plate and I am suspecting there is something wrong with it. Although the growth is nearly two times what I would get in a manually agitated starter, everything I grow using the plate has a fairly low viability.


  5. Thanks. Sorry for not directing the comment directly to you. That was actually a repost from HBT. I seem to have trouble submitting comments from my iPad mini.

  6. Steve,

    Are you growing the yeast on the stir plate in wort? When I propagate yeast on a stir plate I see MB based viability in the high 90's.


    1. Yes, that's what I would expect as well. Would a Zn plated screw (that I use for a stir bar) cause problems with yeast growth?