Stepping up a yeast starter - why?

[shawnbou]

Can anyone help me understand exactly why it's recommended to step up a yeast starter for bigger beers?

What is the difference between, say, pitching yeast into a 2-liter starter and then stepping up to a 4-liter starter ... as opposed to just pitching the yeast directly into a 4-liter starter?

I've always understood that yeast multiply until they can't anymore, limited by the volume of wort and the available fermentable sugar. So why would I not just go straight to the last step to get to my intended pitching rate?

Thanks.

 

[MachineShopBrewing]

Yeast multiply based on a growth curve that factors in many things. One of the biggest factors is the pitching rate of cells per ml. You would see quite a different growth rate in a 2000ml stepped up to a 4000ml vs just a 4000ml.

Example (using yeastcalc.com)

100 billion cells starting

4000ml starter stir plated = 458 billion cells with a growth factor of 3.72

Stepped from 2000ml to 4000ml = 892 billion cells with a growth factor of 2.43 for the initial step and 1.68 for the second step.

Pick up a copy of the book "Yeast." All of the questions regarding yeast propagation and fermentation were tackled in there.

 

[WoodlandBrew]

In my experience there is no magical growth curve. The limiting factors are those that the OP mentioned. As long as your inoculation rate is reasonably high (more than 10 million per ml) it will take off and finish quick and attenuate well. If you gravity is reasonable (less than 1.050) the alcohol will not inhibit the yeast. And the final cell density is less than 300 million per ml fermentation will follow the Balling observation.
(for still starters)

There is a series of four posts on my blog right now with experimental results.

For the theory read here:
http://woodlandbrew.blogspot.com/2013/02/cell-growth-as-function-of.html

PS, the yeast calc magic inoculation rate is 65 million cells per ml. Use that and you will maximize your cell growth per volume.

 

[shawnbou]

So if I'm following, it's a function of the cells pitched per milliliter of wort, right?

If the viable cell count is very low - due to expired yeast, for example - there just aren't enough of them to succeed against a full-size starter, and that's where a smaller first step might be useful?

Similarly, if the OG is sufficiently high to need a really big starter, even a fresh supply of healthy yeast may need step-ups because the denominator of the function above (i.e., volume of wort) is so much greater?

If I'm capturing this correctly, then that's exactly what I needed to know. Thanks and cheers!

EDIT: I also do have the book Yeast on my list to buy and read already, but I haven't got it yet. Thanks for the tip.

 

[Demus]

You've got the gist. Basically starters have diminishing returns as you make them bigger. How you get to the recommended/desired cell count doesn't really matter, but on a big beer you may find it takes 4+ liters which may exceed your flask size or be otherwise inconvenient to make. 2 smaller starters generally produce more than one big one...

 

[Kaiser]

The lower growth at lower pitching rates is something I only observed in still starters but not in stirred starters. That meas that for stirred starters stepping up from 2L into 4L should grow as much as starting with 2+4 = 6 L. Check out the braukaiser model in the Brewer's Friend yeast growth calculator. That's for stirred starters and is based on my experiments.

If you let the yeast rest between the steps you may even be better off with one large step compared to 2 smaller steps since the resting costs the yeast energy which it could otherwise spend on growing.

There is still use in stepping up. But that's mostly for those who grow yeast from slants. If you start with a pack or vial of yeast there should be no need in stepping up unless you have equipment limitations. E.g. a 2L flask is the biggest flask you have.

Kai

 

[MachineShopBrewing]

There is still use in stepping up. But that's mostly for those who grow yeast from slants. If you start with a pack or vial of yeast there should be no need in stepping up unless you have equipment limitations. E.g. a 2L flask is the biggest flask you have.

To the OP, I would follow this advice. I have never had to step up a starter myself as I have a 5L flask. I have only used that flask a couple of times anyway for lagers and big Belgian beers. If you get your yeast fresh and have a stir plate, you won't really find a need to step up starters unless you are growing it up from slants or plates as Kai says.

 

[WoodlandBrew]

Agreed. I guess i didn't say it as well. My point in my first reply was that inoculation rate really only seems to come into play at extremes.

 

[shawnbou]

Fantastic information. This helps a lot. Thank you all!

 

[Demus]

So are you guys saying yeast calc and Mr. Malty are wrong? They've been pretty reliable and the mainstream accepted pitching information for some time. They both show pretty clearly that a 4 liter starter (with stir plate) yields about 458 billion cells (from 97 billion), and two 2 liter starters (stepped) yield 650 billion. I just want to be clear, you are saying this is wrong? What do you base this on?

 

[WoodlandBrew]

So are you guys saying yeast calc and Mr. Malty are wrong? They've been pretty reliable and the mainstream accepted pitching information for some time. They both show pretty clearly that a 4 liter starter (with stir plate) yields about 458 billion cells (from 97 billion), and two 2 liter starters (stepped) yield 650 billion. I just want to be clear, you are saying this is wrong? What do you base this on?

I would say Mr. Malty and Yeast calc are often wrong. Yeast calc is simply a derivation from Mr. Malty. (I emailed the developer of Yeast Calc and he offered that information) My conclusions are based on cell counts I have done and talking with others that have done cell counts.

Inoculation rate doesn't play as large a factor as the calculators indicate. The resolution of the calculators is also deceiving. Either path (one 4L or two 2L) with a 10°P wort will yield about 400 billion cells grown in a still starter or about 600 billion with a stirred starter. If you start with less cells it will take much longer, but the growth will be about the same.

On my blog I have a series of 3 posts showing data from 42 side by side starters. I haven't finished writing the concluding post, but this post with the raw data may be of some interest:
http://woodlandbrew.blogspot.com/2013/02/side-by-side-starters-2-of-4.html

 

[Kaiser]

So are you guys saying yeast calc and Mr. Malty are wrong?

At this point I’m only saying that they don’t reflect my observations from targeted experiments and practical yeast propagation.

They've been pretty reliable and the mainstream accepted pitching information for some time. They both show pretty clearly that a 4 liter starter (with stir plate) yields about 458 billion cells (from 97 billion), and two 2 liter starters (stepped) yield 650 billion. I just want to be clear, you are saying this is wrong? What do you base this on?

If you say that they have been working for brewers because they improved brewer’s beers I would say that they do that very well. Simply making a starter, regardless of actual growth model, will do that.

But how many brewers have actually counted the cells they got after propagation and correlated that to Mr Malty? (Yeast calc uses Jamil’s model and is not based on original research). The idea of these calculators is that you don’t have to count cells but I think brewers put too much trust into that result. Check out this link: http://braukaiser.com/blog/blog/2012/11/03/estimating-yeast-growth/ which goes into a comparison with Jamil’s calculator and why I think it may be flawed.

Based on what I’m seeing in his calculator and based on what I observe I think Jamil did not use a series of stirred starters to come up with the growth curve for stirred starters. I contacted him about this and he didn’t answer that question.

I recently finished building 2 double PWM controlled stir plates to speed up the experiments since I think it is important to use yeast growth observations from stirred starters. My results from still starters match Jamil’s calculator within an acceptable error band.

Kai

 

[Demus]

Very interesting, thanks. How about the recommended pitch count, are they fairly accurate? Obviously as an average home brewer I'm never going know my actual cell count, but the closer the estimate is to reality the closer I'll be to optimal. ...

 

[Kaiser]

Very interesting, thanks. How about the recommended pitch count, are they fairly accurate? Obviously as an average home brewer I'm never going know my actual cell count, but the closer the estimate is to reality the closer I'll be to optimal. ...

I trust the recommended pitching rates although they should not be seen as a dogma. Brewers are welcome to experiment.

Kai

 

[Demus]

Agreed. They can certainly be altered and even should be for certain styles...

 

[WoodlandBrew]

Agreed. The standard pitch rate that directly relates number of cells to weight of sugar makes good sense. It's a starting point, and a little adjustment may be needed from there for the style of beer, or your processes and taste.

 

[diS]

If you let the yeast rest between the steps you may even be better off with one large step compared to 2 smaller steps since the resting costs the yeast energy which it could otherwise spend on growing.

My though was that letting yeast to rest you enable it to build their glycogen reserves, so they can spend that reserves in next propagation step. In other words, energy that is spent in resting, yeast will use to multiply/growth when inoculated in suspension with enough sugars/nutrients.
That said, resting will impact on health and (probably) produce more healthy cells.

Am I misreading something?

 

[WoodlandBrew]

I don't think you are misreading anything. Yes, glycogen reserves are built at the end of a fermentation cycle. They are most useful at the start of fermentation when the cell membrane is not very permeable while the cell is still building its sterol content.

The analogy of glycogen in yeast to fat on people works for the most part. If a person has an average amount of fat on them and then they start running every day, he is going to burn off that fat. Even though this person is also eating everyday the fat energy reserve is depleted. Even after most of the fat is gone this person can keep running everyday as long as they are also eating. If this person lives in sunny Sothern California they can keep this up indefinitely. If this person live in the North East then winter comes along, and he isn't running everyday, and the fat is gained back.

 

[Kaiser]

My though was that letting yeast to rest you enable it to build their glycogen reserves, so they can spend that reserves in next propagation step. In other words, energy that is spent in resting, yeast will use to multiply/growth when inoculated in suspension with enough sugars/nutrients.
That said, resting will impact on health and (probably) produce more healthy cells.

Am I misreading something?

Yeast needs energy for two things: cell division and maintenance. The energy that goes to cell division will result in growth that we can observe by counting cells. Energy spent on maintenance is the energy yeast needs to stay alive. The longer yeast sits dormant, the more energy it is going to spend on staying alive. This energy largely comes from reserves built up when nutrients were present. If yeast has been dormant for a while its reserves will be depleted and the first thing it does when there are more nutrients available is to replenish these reserves before it can start budding. I.e. less energy is available for budding and you should see less growth.

So far this is only a theory based on a few observations and my understanding of yeast metabolism. I’m still planning to run a good series of experiments to test this theory.

Kai

 

[diS]

This leads me to 2 questions:
How much it takes to build all their reserves (if they are crucial ones besides glycogen) and how much it takes to spend them to the point where they were at the end of previous starter/step.
I am no a microbiologist, just trying to think logically..Inoculating in time between these two points should produce more growth (in well aerated starter with enough nutrients). Also, there is question about population age and how much healthy-to-reproduce cells there are, but if we assume it is 1st generation form slant/plate then it wold be benefit to rest starter before next step/fermentation.

I think that Woodland did a test with fresh/crashed yeast and crashed performed better. White/Zainasheff also states that storing the yeast for additional 8-12 hrs allows them to build glycogen reserves.

 

[Kaiser]

I don't have answers to these questions either, but that's stuff I'm looking into.

Kai

 

[WoodlandBrew]

I don't have any numbers or practical application of these principles. If you haven't read Fix (Principles of Brewing Science) I highly recommend it. Also search for EMP pathway and cellular scar theory.

 

[diS]

I didn't read it. Are there others valuable information's that are not in "Yeast"?
Looking forward for future researches on this topic, keep the good work guys!

 

[WoodlandBrew]

Thanks! Unfortunately, I haven't read "Yeast" so I don't know what the overlap is to Fix's book. He spends about 100 pages on yeast. My impression of the book is that Fix lays it down from a scientific perspective but doesn't really get into application.

 

[Demus]

Yeast goes the other route. Some scientific background but geared more toward practical application for brewing. ...