Yeast Overpitch/Underpitch Experiment From a Microbiologist

Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum

Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Really great thread. Thanks phattysbox!

Question: I'm just beginning to wash/harvest yeast. My question is, how do I know how much yeast I'm starting with when I reuse it? Or should I just follow the instructions here and make a 1L starter and multiply the yeast as you've laid out? Maybe said another way, is there a "terminal amount" of yeast that you'll get from a 1L starter and that's the basis for knowing how much you have/need to grow? I hope all this makes some sense! Thanks
 
I think another important variable here is yeast selection. 1056 is not expressive at all and I might see better differences with a characterful english ale yeast.

I'd suggest WLP007 - I made a fantastic tasting blonde ale with this yeast last spring, and then tried to brew it in the fall again but I underpitched and stressed out the yeast and the beer was completely different (and mostly undrinkable by my standards). I know a lot of people use this yeast as their 'house yeast', but I find it to be somewhat fussy about pitch rates and fermentation temps in the different beers I've used it in.
 
william_shakes_beer;3599217} OK said:
I think you're confusing a starter versus the target beer.

What I said above applies to a starter on a stir plate. The constant motion of the liquid provides a huge surface area for O2 exchange (and CO2 for that matter).

In wort, the surface area is much harder to manipulate since the volume is bigger and we dont have 5 gallon stir plates... ;)

Dumping wort back and forth from bucket to bucket multiple times is the best way to aerate your wort. However, you can only get about 6-8 ppm this way which is really the minimum of what yeast need. Not to mention your back breaking and contamination possibilities. Its not the best way to go.

Pure O2 is the only way to get oxygen in wort above 10ppm.
 
I'd suggest WLP007 - I made a fantastic tasting blonde ale with this yeast last spring, and then tried to brew it in the fall again but I underpitched and stressed out the yeast and the beer was completely different (and mostly undrinkable by my standards). I know a lot of people use this yeast as their 'house yeast', but I find it to be somewhat fussy about pitch rates and fermentation temps in the different beers I've used it in.

Good suggestion and I've been thinking about using this yeast.
 
Really great thread. Thanks phattysbox!

Question: I'm just beginning to wash/harvest yeast. My question is, how do I know how much yeast I'm starting with when I reuse it? Or should I just follow the instructions here and make a 1L starter and multiply the yeast as you've laid out? Maybe said another way, is there a "terminal amount" of yeast that you'll get from a 1L starter and that's the basis for knowing how much you have/need to grow? I hope all this makes some sense! Thanks

Great questions.

The only way to know how much yeast you have is by counting. You can do it with a microscope, but this is impossible for most homebrewers. Another way to do it is by diluting the slurry and plating the yeast. When you have a certain number of colonies at a specific dilution, you can calculate the number of cells in the original sample.

The equation:

number of cells/ml = colonies counted * dilution factor * 1ml/volume plated(mls)

For example lets say I have a 200 ml slurry of washed yeast cake with an unknown number of cells. First make a serial ten fold dilution.

1) Take 1 ml from the slurry and mix with 9 mls of sterile H20. This is a 1:10 dilution.

2) Take 1 ml from the 1:10 dilution and mix with another 9 mls of H20. The yeast is now diluted 100 fold or 1:100.

3) Do this for five dilutions, or 1:100,000

4) Using sterile technique at home, plate 0.5 mls (a few drops) onto a yeast plate (MYPG is best)

5) Grow at room temp for 3-4 days.

6 Count colonies. Lets say I get 139.


cells/ml = 139 * 100,000 * 1/0.5 mls

= 27,800,000 cells per ml


Since you have a 100 ml slurry (that you measured) you have

27,800,000 * 100 mls = 2,780,000,000 billion yeast cells in your slurry

Hope this helps. To count this way, you'll have to set up basic microbiology stuff. Cheaper than buying a scope.

J
 
The only way to know how much yeast you have is by counting. You can do it with a microscope, but this is impossible for most homebrewers. Another way to do it is by diluting the slurry and plating the yeast. When you have a certain number of colonies at a specific dilution, you can calculate the number of cells in the original sample.


Is there an easier way to ballpark the yeast quantity just so you don't over/under pitch from harvested yeast? I am just about to get into harvesting my own yeast (as soon as my next beer finishes) and would like to start out slow so I don't get scared away from biology (this happens more often than not with me, hence why I chose a dead science to study like electrical engineering :D).
 
Here's what I've done, which has worked well for 15-20 batches. It may not be optimum, but it certainly gets me in the ballpark.
1) Read the sticky on washing yeast
2) Check out Mrmalty.com. Check the options for calculating pitching rates from slurries
3) After the boil I cool and transfer to a 6 gal better bottle to settle. The great majority of the hot and cold break settles out in about 45 minutes. When I wash the yeast, there is very little debris with the yeast.
4) The sticky on washing yeast says to use the middle section as the slurry stratifies. I believe that is based on a slurry with a fair amount of trub. Since I have little non-yeast trub, I have found that I usually have the most yeast in the bottom (it may stratify into only 2 layers instead of 3). Bottom line - you may need to adjust based on your process.
5) I always use a starter to make sure the yeast are healthy and active.

I hope this helps:mug::mug:
 
Is there an easier way to ballpark the yeast quantity just so you don't over/under pitch from harvested yeast? I am just about to get into harvesting my own yeast (as soon as my next beer finishes) and would like to start out slow so I don't get scared away from biology (this happens more often than not with me, hence why I chose a dead science to study like electrical engineering :D).

There is.

However, it is woefully inaccurate and minor errors can ballon to large ones.

The key is to dilute the yeast to the point that you can just barely see them in solution. This corresponds to about 1 million cells per ml.

This can be tricky however, and will need some calibration with your eyes. You will have to dilute the yeast as I described above to the point you can see it in solution. The work backwards.

For example, if I diluted in 10 mls a yeast sample all the way out to 1:10,000 and can't see cells by my eye, I would work backwards making more dilutions until I can. For example, 1:9000, 1:8000, etc...

So lets say that cutoff is 1:5000. So in an original 20 ml slurry, for example, you would have:

1 million cells/ml * 5000 * 20 mls

= 100 billion cells
 
Please pardon my ignorance in this post as I might be totally off; but if the yeast are capable of running out of space and food, wouldn't their growth curve be logistic? And if there is some sort of population maximum for a given amount of volume/sugar available, couldn't you use that as a starting place?

If you could assume the yeast reach (or get close to) the population maximum, and your cold crash gets all of the yeast out of suspension, then you know how much yeast you are washing. Then you could split it between your jars, and you would know (roughly) how much yeast per storage container you have. You would have to step them up anyways, and in doing so would be subjecting them to another logistic curve, giving you another chance to rein in your assumptions.

Are these accurate assumptions, or am I just totally off my rocker?
 
Please pardon my ignorance in this post as I might be totally off; but if the yeast are capable of running out of space and food, wouldn't their growth curve be logistic? And if there is some sort of population maximum for a given amount of volume/sugar available, couldn't you use that as a starting place?

If you could assume the yeast reach (or get close to) the population maximum, and your cold crash gets all of the yeast out of suspension, then you know how much yeast you are washing. Then you could split it between your jars, and you would know (roughly) how much yeast per storage container you have. You would have to step them up anyways, and in doing so would be subjecting them to another logistic curve, giving you another chance to rein in your assumptions.

Are these accurate assumptions, or am I just totally off my rocker?

You are absolutely correct and no you're not off your rocker...:p

However, that involves determining all of the variables of a logistic equation for a particular yeast strain.

For example, one would need to know the growth kinetics at exponential phase (i.e. the slope of curve which is different for every strain), the starting number of yeast cells, the exact concentrations of sugars, and exactly when Pop_max is reached since there will be some cells that are dying and some dividing (establishing an equilibrium).

Not to mention changes in starter format, such as differences in extract used. Stirred versus shaken, etc... etc...

I mean dont get me wrong, it can be done. But mathematically modeling yeast couts from a starter seems like a lot of work to me.
 
Sounds like fun! Well, I forgot about how many variables there were, but at least my thought process wasn't off!

So, on the flip side of your first suggestion regarding the dilution the the yeast till you barely see them, could you condense the yeast instead? For example, put the washed yeast into the mason jars and then after a week or so measure the volume of yeast at the bottom of the container? I would assume there would be a bit of fluctuation in the density, but not too terribly much. I guess I would just need to figure out that density instead...

At this rate, it might be easier to just slant the yeast...
 
Is there an easier way to ballpark the yeast quantity just so you don't over/under pitch from harvested yeast? I am just about to get into harvesting my own yeast (as soon as my next beer finishes) and would like to start out slow so I don't get scared away from biology (this happens more often than not with me, hence why I chose a dead science to study like electrical engineering ).

How about this? When I buy a vial of liquid yeast, can I mark a line where the yeast comes up to, then refill with recaptured yeast up to that line and assume it's the same number of yeast as the original?
 
DCBrewer said:
How about this? When I buy a vial of liquid yeast, can I mark a line where the yeast comes up to, then refill with recaptured yeast up to that line and assume it's the same number of yeast as the original?

No. White Labs compacts the yeast to a cell density far greater than home brewers can achieve.
 
1) 1 liter starter - 1.040 + 1 wyeast smack pack

A wyeast smack pack is 125 ml. Does this mean I use 875 ml of 1.040 wort and combine with the 125 ml smack pack to achieve 1000 ml? Or do I add the 125 ml smack pack on top of the 1000 ml 1.040 wort? Is there any knowledge as to the gravity of the solution in a wyeast smack pack?

3) Dilute 1:10 into a new starter and grow for 24 hours. That means 100 mls into another 1 liter starter. This starter will go from 18 billion to another 100 billion.

Same question here... use 900 ml of fresh 1.040 wort with 100 ml of starter #1, correct?

8) After the third starter finishes, you should have close to 400 billion. You can then combine all the starters and pitch into beer.

I'm assuming you would cold crash and decant after adding #3 to the already-decanted #1/#2 combined solution. Right?

What are your words on bringing a starter up to an appropriate pitching temperature? Do you worry about the yeast cells building up glycogen as the cold crashed starter warms up to the appropriate pitching temperature? If I wanted to bring a 41 degree starter up to 68 degrees, sitting in a 70 degree room it is going to take several hours. I understand the yeast will be dormant throughout most of that but as it gets closer to 68, I imagine the yeast will begin to wake up. Since it's hard to say exactly how long a particular brewing session will take, or how long it will take for the starter to hit its ideal temperature, there is the chance it could be sitting at 68 for several hours before being pitched.

This method seems to be a bit different than what has been advised in the past. I like this method better as you don't need to build a giagantic 2 liter (or bigger) starter to reach the yeast levels needed.
 
...
2) grow for 24, no more. This is key. After 24 hours the yeast will still start to go dormant, building up glycogen reserves. Starters should be "split" when the yeast cells are actively replicating. You should go from 100 billion cells to about 180 billion
3) Dilute 1:10 into a new starter and grow for 24 hours. That means 100 mls into another 1 liter starter. This starter will go from 18 billion to another 100 billion.
4) Cold crash the first starter, decant wort.

At this point you have close to 300 billion, but you need more.

6) Do a third starter from the second as in step number 3, grow for 24 hours.
7) Combine starter #2 into #1, cold crash and decant.
8) After the third starter finishes, you should have close to 400 billion. You can then combine all the starters and pitch into beer.

Sorry for making this question a two-parter, but I'm really curious about a few more things.

I am a bit confused by your math. You state that after growing #1, you have 180 billion, and that growing #2 you go from 18 to 100 billion. Then you say we have around 300 billion cells.

But... 180-18+100 = 262, not 300...

Even so, if we are at 300 after starter #2 and we are at 400 after starter #3, that means that starter #3 grew 12% more yeast than either starter #1 or starter #2. Is this correct?

#1: Start with 100 billion, end with 180 billion. 180 - 100 = 80 billion new cells
#2: Start with 18 billion, end with 100 billion. 100 - 18 = 82 billion new cells
#3: Start with 10 billion, end with 100 billion. 100 - 10 = 90 billion new cells

I hope I'm not being too difficult here :) Just interested and curious.
 
Very curious. I plugged the numbers into YeastCalc and came up with the same final cell count; 400 billion cells, but the final counts for each starter were slightly different; this is assuming 100% viability for the Wyeast smack pack.

1000ml starter + 100 billion cells = 229 billion cells; with stir plate
divide final cell count 229 by 1000 = 23; put this number up in initial cell count.
1000ml starter + 23 billion cells = 103 billion cells; with stir plate
divide final cell count 103 by 1000 = 10; put this number up in initial cell count.
1000ml starter + 10 billion cells = 65 billion cells; with stir plate

229 + 103 + 65 = 397 billion
 
1000ml starter + 100 billion cells = 229 billion cells; with stir plate
divide final cell count 229 by 1000 = 23; put this number up in initial cell count.
1000ml starter + 23 billion cells = 103 billion cells; with stir plate
divide final cell count 103 by 1000 = 10; put this number up in initial cell count.
1000ml starter + 10 billion cells = 65 billion cells; with stir plate

229 + 103 + 65 = 397 billion

Nope nope nope nope

Final cell count:

(229-23) + (103-10) + 65 = 364
 
Why is 1:10 dilution more important than density?
1st step 1000 ml for 100B cells @ 1.040.
2nd step 1000 ml for 23B cells @ 1.040

Why not 1000 ml for 111B cells (~50 %) to keep
The same cell/sugar density?

This should give 250B cells ( linearly).
 
A wyeast smack pack is 125 ml. Does this mean I use 875 ml of 1.040 wort and combine with the 125 ml smack pack to achieve 1000 ml? Or do I add the 125 ml smack pack on top of the 1000 ml 1.040 wort? Is there any knowledge as to the gravity of the solution in a wyeast smack pack?



Same question here... use 900 ml of fresh 1.040 wort with 100 ml of starter #1, correct?



I'm assuming you would cold crash and decant after adding #3 to the already-decanted #1/#2 combined solution. Right?

What are your words on bringing a starter up to an appropriate pitching temperature? Do you worry about the yeast cells building up glycogen as the cold crashed starter warms up to the appropriate pitching temperature? If I wanted to bring a 41 degree starter up to 68 degrees, sitting in a 70 degree room it is going to take several hours. I understand the yeast will be dormant throughout most of that but as it gets closer to 68, I imagine the yeast will begin to wake up. Since it's hard to say exactly how long a particular brewing session will take, or how long it will take for the starter to hit its ideal temperature, there is the chance it could be sitting at 68 for several hours before being pitched.

This method seems to be a bit different than what has been advised in the past. I like this method better as you don't need to build a giagantic 2 liter (or bigger) starter to reach the yeast levels needed.

1) You could do either - just as long as when you go to the next step you take 1/10th of the starter volume. Sorry I didn't make that clearer before.

2) The gravity of a wyeast pack would be very low since it is pure concentrated yeast. I never break the wort "smack" pack as its not necessary.

3) Yes - you are right 900 mls + 100 mls. My mistake.

4) Yes, you can combine all the starters once decanted.

5) You know - this is an interesting point you bring up. Temp swings while making starters should be avoided. And while, yes the yeast will start to wake up and turn to their glycogen reserves, it really the introduction of sugars that gets mobilizing their starch reserves. I don't think this is a big issue, but maybe someone else has a different take on this.
 
Sorry for making this question a two-parter, but I'm really curious about a few more things.

I am a bit confused by your math. You state that after growing #1, you have 180 billion, and that growing #2 you go from 18 to 100 billion. Then you say we have around 300 billion cells.

But... 180-18+100 = 262, not 300...

Even so, if we are at 300 after starter #2 and we are at 400 after starter #3, that means that starter #3 grew 12% more yeast than either starter #1 or starter #2. Is this correct?

#1: Start with 100 billion, end with 180 billion. 180 - 100 = 80 billion new cells
#2: Start with 18 billion, end with 100 billion. 100 - 18 = 82 billion new cells
#3: Start with 10 billion, end with 100 billion. 100 - 10 = 90 billion new cells

I hope I'm not being too difficult here :) Just interested and curious.

As for the math - this was just an example and complete estimate. Could be less, or it could be more depending on how you make your starter. Keep in mind that even if you follow something like Mr. Malty exactly, you might not get the EXACT number. You might be close, and you may be over or under.

As for the growth numbers, I was just relating to what routinely get in the lab when I make my starters.

The growth of the starters, like the target beer, is subject to pitching rate. If there are more yeast present, as in starter #1, then there will be less growth. However, with starter #3, you are "pitching" less yeast into that starter and the yield will be higher.
 
Why is 1:10 dilution more important than density?
1st step 1000 ml for 100B cells @ 1.040.
2nd step 1000 ml for 23B cells @ 1.040

Why not 1000 ml for 111B cells (~50 %) to keep
The same cell/sugar density?

This should give 250B cells ( linearly).

The most important thing to do when stepping starters is not to add too much yeast to the next starter. This will inhibit growth and the yeast will just ferment the starter.

Keeping the density the same to the next starter will add more yeast to that starter. Or at least that's what I think you're getting at. Not sure what you mean by cell/sugar density.

Its not so much dilution but using total numbers. If you have a starter that has 200 billion total cells, the *ideal* step is to take 20 billion cells from that starter to the new one. This will ensure the most yeast. Also, yeast growth is logarithmic, not linear.

Hope this helps
 
phattysbox said:
The most important thing to do when stepping starters is not to add too much yeast to the next starter. This will inhibit growth and the yeast will just ferment the starter.

Keeping the density the same to the next starter will add more yeast to that starter. Or at least that's what I think you're getting at. Not sure what you mean by cell/sugar density.

Its not so much dilution but using total numbers. If you have a starter that has 200 billion total cells, the *ideal* step is to take 20 billion cells from that starter to the new one. This will ensure the most yeast. Also, yeast growth is logarithmic, not linear.

Hope this helps

Why is 1/10th the ideal? Each step thus has less yeast cells per ml. Isn't there an ideal # of yeast cells to ml of wort?

Just asking in hopes of obtaining more yeast with only 2 steps.
 
Why is 1/10th the ideal? Each step thus has less yeast cells per ml. Isn't there an ideal # of yeast cells to ml of wort?

Just asking in hopes of obtaining more yeast with only 2 steps.


1:10 dilution is what I find to be the best dilution to obtain the maximal yield in yeast growth. This number can vary from person to person. Some say 1:5 others 1:20. The 1:10 dilution gives you the ideal cells/ml once a starter has reached confluency. I will bet serious money that this is how WYeast and white labs prep there yeast.

Unfortunately, adding more yeast, as in a 1:5 dilution, will not grow more yeast. At least this is what I see when I make my starters. Yeast will only grow (replicate) to the point where there is still sugar and nutrients in the medium. Add more yeast (1:5 dilution) just depletes those reserves faster.

If you want to make the starter with less steps, the only way to do it, efficiently, is to make a bigger sized starter with more yeast. 2 smack packs in a 3 liter starter for example.
 
1 million per ml per °Plato. Maybe less for ales, more for lagers, but that is the number I shoot for.


Unless you mean for your target beer - I'll have to disagree. While this is perfect target for fermenting wort, growing yeast cells needs a much higher pitching rate. So we don't get confused, I'll call the starter pitch rate the "inoculation rate"

For the starter, you want a inoculation rate in the ball park of 5-15 million cells per ml / degree Plato.

For example, in a 1.5 liter starter of 1.036 gravity with a Wyeast pack of 100 billion cells gives you an inoculation rate of:

total cells = inoc rate * mls wort * plato

100 billion = (X) * 1500 mls * 9 Plato

(X) = 7.4 million cells per ml/plato


This will give you a decent yield in the starter.
 
Back
Top