Friday, December 7, 2012

Yeast Washing Exposed


Yeast washing is a means of cleaning yeast to separate the viable yeast cells from unwanted partials.
To wash yeast, The yeast is harvested from a fermentation vessel and combined with water.  It is allowed to settle until the water has separated to the top and the yeast to the bottom.  The water and the top of the yeast is then poured into several jars.  The yeast and debris that has settled to the bottom of the first container is then discarded.

Normally I don't wash my yeast, I just pour the cake into three or four quart size mason jars, but I have been curious as to what benefit may be achieved from yeast washing.  My understanding is that the intent is to separates the hops and dead yeast cells from the live yeast cells.  In the last few days I have been running experiments on this technique to find out what it does, and if it is worth the extra time. 

Yeast washing is beneficial, but not for the reasons that I had anticipated.

After the yeast has settled into the container it divides into roughly three sections.  Common brewing wisdom indicates that the top portion is mostly water, the light colored middle section contains viable yeast, and the darker bottom contains dead yeast, hops, and other debris.  However, it seems that this in not the case.

The viability throughout the container is roughly the same.

In four test cases the viability was not statistically different in these layers.  Tests were run with three slurries with 10%, 50% and 90% viability.  In all tests the viability of the yeast in each section did not vary more than one standard deviation.

What was interesting was that the bacteriological content was much higher in the top portion of the yeast containers than in the lower parts.  There was about 100 times more bacteria per live yeast cell in the top "liquid" section.

Another strange finding was that the concentration of non-yeast debris followed the cell density.  While the cell concentration at the bottom of the container was twice what it was in the middle, the viability was the same, and the concentration of non-yeast material per yeast cell was virtual identical.  The hops and other partials did not separate from the yeast.  So when you are throwing out the junk, it takes just about as much viable yeast with it as it takes debris.

The above experiments were done in test tubes.  To make sure this was repeatable, I tried yeast washing with a full sized batch.  The results were the same.  The viability of the part that would have been thrown away was the same as that of the part that would have been kept.  The viable cell density was also very similar. 

For further information on washing a a full sized batch see the new blog post: Yeast Washing Revisited.

The tests were conducted with WLP566, WLP004, EC-1118 and S-04.
Conclusion
When washing yeast, discard the liquid when washing yeast to remove bacteria.  Keep the thick slurry and add clean water on top of it for storage and to wash out additional bacteria.

I would be happy to talk through the numbers with anyone interested.


15 comments:

  1. Replies
    1. Thanks! The science behind brewing is exciting.

      Delete
  2. So you're saying to pour the bottom 2 layers or just the bottom layer?

    Doesn't the amount of hop material in the trub depend on someone's brewing setup. Some person may get a lot of hops in there fermentation vessel due to poor racking from BK.

    Would you still expect the same amount of hops per cell in the slurry on the very bottom and the middle layer in this case?

    On a separate note, great work! I love the stuff you are doing and it is extremely helpful!

    ReplyDelete
    Replies
    1. Thank you for your positive feedback. Knowing that people appreciate what I do keeps me working on this.

      All of the thick material that settled to the bottom of the container in the test cases I did contained a large amount of viable yeast. So, Yes, I recommend using all of the thickly settled material.

      Yes, like you said, the amount of hop material in the trub will be different for everyone, and will be driven by their techniques. Dry hoping and using pellets hops in the boil would increase the amount of hop trub. Emptying the entire content of the brew kettle (including the settled debris) will also increase trub in the fermenter.

      Therefore, no, I would not expect the same amount of hop trub per cell if preventative measures were taken to prevent hops from entering the feremnter such as the use of leaf hops, a hop spider, hop bag or other methods.

      Although, in all of the slurries that I have examined the largest portion of the slurry is protein, and not hops.

      Delete
    2. Don't we want to leave those proteins behind and not bring them along with the yeast for future pitching? Or are we not concerned because they should settle and be left behind in the trub in future brews?

      Delete
    3. The trub that remains after fermentation is actually beneficial to yeast. It aids is cell wall development. (Fix 1999) In my experience there is more protein from the wort than added when by the repitched slurry.

      Delete
    4. Awesome info. Thanks alot for your time and effort.

      Delete
  3. what about using a yeast cake from a 1.090+ beer to make starter? I would use .5 liter of yeast cake slurry with 1 liter of 1.040 wort on a stir plate, and step up to 4 liter to pitch in another 1.090+ 10 gallon batch. I have read the yeast are tired and used up, but these are the same people that have said tossing the settled slurry of washed yeast is the best practice. what do you think? Are these yeast of a big beer done, or can they be raised up to healthy numbers again?

    ReplyDelete
    Replies
    1. The problem I see with using yeast from a big beer is that the alcohol content is high. See the "ABV effect on viability" post for details. It is also true that more mutation occurs with stress, but genetic mutation can take years of re-pitching to really develop to a point where it is a problem in yeast.

      In my experience there is no problem with raising yeast from a high gravity beer. It's similar to culturing yeast from a bottle. It's takes time, but quite possible.

      Delete
    2. This is a great site with wonderful insight. Your post on ABV effect on viability I think answered my question. I just opened a RIS I bottled six weeks ago with 4oz corn sugar in 4g batch. The ABV is 10.89% after sitting on the yeast for 28 days. There is absolutely no carbonation. Based on your chart there is probably no viable yeast left to carb. I should probably add some healthy yeast to my batch before I bottle this high of ABV. What do you think? And is there some formula I could use based on volume I'm going to bottle?

      Delete
    3. For bottle conditioning it doesn't take much yeast. Many professional brewers use 1 million cells per ml.

      Delete
  4. Just, awesome information. Do you generally get the yeast going again in some kind of starter? I generally still make a starter, because... well frankly dry yeast is cheap, dme is cheap, I know it works, and mostly because I spent $80 on a stir plate. By the way your recent posts have changed how I extract brew. I now own a pressure cooker, and I haven't used my wort chiller for an extract batch since reading your post. Keep up the good work.

    ReplyDelete
    Replies
    1. Thanks! Generally I'll pitch directly from the fridge and estimate 1 billion cells per ml if I'm skipping the viability count.

      Delete
  5. Hi Steven, thanks for the great report! I'm curious about your method -- you say "Tests were run with three slurries with 10%, 50% and 90% viability." From what you say, it sounds like viability is a something you controlled rather than being a dependent variable. If so, how did obtain those specific viability numbers? Otherwise, did the viability readings from your slurry samples just happen to be those numbers?

    Lastly, you mention 4 different strains of yeast at the end of your article (WLP566, WLP004, EC-1118 and S-04). Did you use three slurry samples of each type (12 total experiments) or just three total? How did the viability measurements fare for each?

    Thanks for any additional info you can provide!
    Tyler

    ReplyDelete