"The ideal starter" - Transcript of an article on braumagazin.de

[monkeymath]

In the thread Maturation rest for lagers we sort of went off-topic and started discussing yeast starters, and I backed my position - which is not to cold crash your yeast starter - by an appeal to authority, claiming this was supported by advice from Ulrich Preise, head of the yeast bank of Weihenstephan. Since the article I derive this from is entirely in German, and it might be of general interest, I thought I'd share a sort of transcript/summary here. To those of you who do speak German: please do not point out the parts where deviate from the original. These are deliberate lies which tie into a much bigger evil master plan.
I hope it can be of use to some of you.

The article itself is not written by Ulrich Preise. A couple of homebrewers from Munich had the opportunity to visit the yeast bank and ask Ulrich numerous questions. The article arose from these questions, but probably the author - Moritz Gretzschel, who is himself a wealth of knowledge on homebrewing - filled in some parts on fundamentals of yeast physiology etc. where he found it fit. This is just a guess on my part. Either way, I am certain that Ulrich Preise double-checked all the claims made in that article and that no major misinformation is contained therein.

The article is openly available online, so I hope it is fine to try and make its content available to a broader audience. I possess no rights to the article.

This is not a reproduction of the article, just an attempt to make the information contained therein available to a broader audience. I omitted some parts on closely related topics such as hygiene and yeast storage.

So let's jump right into it.

Fundamentals
There are two main aspects to consider for a starter. Yeast quality and yeast quantity. Let's talk about yeast quality first.
To find the optimal procedure for yeast propagation, we need to first take a look at the different stages of yeast reproduction. In terms of the number of active yeast cells, we can divide the fermentation into four stages:

1. The lag phase or latency phase. Broadly speaking, the yeast is adapting to the wort. It analyzes the nutrients available and prepares its metabolism through the activation of enzymes. Countless processes happen within the yeast cell, but they do not yet impact the wort. The brewer waits for fermentation to start.
2. The log phase. The yeast is fully active, there is still an abundance of nutrients and oxygen. The yeast reproduces at a constant rate over several generations. Fermentation starts.
3. The stationary phase. The wort is depleted of oxygen, nutrients are diminishing, the yeast cells feel it's getting crowded. We find an equilibrium of yeast reproduction and yeast death. The brewer observes a continued, but slowed down, alcoholic [= anaerobic] fermentation and a decrease in extract content.
4. The lethal phase. The wort is depleted of nutrients, the yeast starves to death or dies of the products of its own metabolism. Fermentation comes to a halt.

For us, the process of fermentation is more readily visible and measurable through the decrease is extract. First the waiting (hardly any decline), then the starting fermentation (showing the steepest decline) during "Hochkraeusen", then an increasingly gentle reduction which ultimately/asymptotically leads to a constant extract level.

In terms of yeast age, on the other hand, we may distinguish the following four stages:

• "children", that have just been created, but are still unable to reproduce,
• "young adults" in the prime of their youth, reproducing like bunnies,
• "adults" whose reproduction rate is slowly declining due to "bud scars" [no idea if that's the right word] which inactivate their surface,
• "seniors" which have a reduced metabolism and will die soon.

This implies that our quest for optimal yeast quality - both for pitching into a larger starter or into the wort - leads us to the following two goals:

• in terms of yeast age, we want to have as many "children" and "young adults" as possible (and consequently few seniors)
• in terms of yeast activity, we want the starter to be at the transition from log phase to constant phase (for maximum yeast quantity at maximum activity)

This is shown in the following chart (I include only a link and not the image itself) click here to see the image.

If we were to retrieve our yeast pitch from currently fermenting wort, it'd be advisable to do so when fermentation has just started rolling at full power (beginning Hochkraeusen), typically after 48 to 72 hours [your mileage may vary (greatly)]. For bottom fermenting beers, take some fermenting wort with yeast in suspension; for top fermenting beers the yeast swimming atop.

From that it is also apparent that, also when making a starter, one should not wait for the yeast to settle out and should not, as is often recommended, decant the supernatant: yeast only fully drops out after fermentation has finished. The material at the bottom contains a high percentage of middle-aged and old yeast cells, near the end of their activity. Instead, keep your starter on the stir plate and pitch it entirely into the next stage once it has reached the end of log phase (or the beginning of stationary phase). That way, the yeast quality will be at its peak.

To aerate or not to aerate?
With the method outlined above, continuous aeration is not necessary, in particular since that always harbours a risk of infection. If you own a stir plate, it is sufficient to run it at a high speed at the beginning, so that the vortex is audibly pulling in air. Afterwards [no mention of the precise timing is made, but I assume it should be after some 2-4 hours], the speed can be reduced, so that a vortex of only 2-3cm depth remains, which signals healthy mixing.
With an unaerated starter, it is not quite as crucial to pitch at exactly the right time as it is when aerating your starter. Moreover, should you wish not to pitch the entire starter - e.g. due to the color of the starter wort - , then given enough time you can wait for the yeast to drop out, in spite of the above.

Pitching rate
For bottom fermenting yeast, aim for 1-1.2*10^6 cells/ml per degree plato. For top fermenting yeast, aim for 0.5-1*10^6 cells/ml per degree plato.

Propagation steps
As homebrewers, we cannot truly work in a sterile environment, so we should ensure a quick take-off at every step of the propagation to leave no chance to any unwanted microbes. So let's apply the same pitching rate(s) as above to the starter wort used in the propagation and shoot for 10*10^6 yeast cells per ml. Under the right conditions (stir plate, 20 degrees celsius, sufficient nutrients) yeast should reach a concentration of 100*10^6 cells/ml within 24 hours, so this naturally suggests to always pitch at a ratio of 1:10 between each step (e.g. let a starter of 100ml wort work for 24 hours, then pitch into 1l of wort). Again, it is recommended not to decant the individual starters, but to pitch them entirely into the next stage.
To play it safe, homebrewers might be better off at a propagation ratio of 1:8. When working with older yeast or bottle dregs, you should start with much smaller steps of 1:2 or 1:3. Only the final step before pitching should always adhere to the ratio of 1:10.

Propagation wort
The ideal medium for yeast growth is regular wort. Of course, not everybody wants to produce a separate batch a couple of days prior to brewing just for propagation (although a simple no-sparge, 1 hour at 68 C, batch would be entirely sufficient). Left over (and then immediately frozen) wort from the last brew day is a great option: just dilute to the right strength and boil to sterilize. While boiling, you may pitch some yeast nutrient or even old yeast.
Otherwise, you may also use malt extract to create a wort of 6-8 degree plato, ideally with some yeast nutrient. Those that cool over night ["no chill"] might also perform the last (large!) propagation step with diluted and boiled first runnings while the wort is slowly cooling.

 

[mjdonnelly68]

Very informative. Dankes fur schreiben!

 

[VikeMan]

Thank You for translating/summarizing and posting this. It must have been quite time consuming. I'm initially struck by a few things here...

1) There's a "lethal" phase which is obviously overstated. If all or even a great number of yeast starve to death or die (relatively quickly) of the products of their own metabolism, crashed starters would not result in an increase in viable cells as compared to the original count. But we know they do. Taken literally, nobody would be able to make beer. And even the fresh pack of yeast we get from manufacturers would be filled with dead yeast. Now if the point is that a crashed started will have more dead cells than an active starter, I'll buy that. But what are the numbers? And how many viable cells are there in each scenario? Which leads to point 2...

2) His argument is based on an analogy between the life cycles of microbes and humans, rather than actual counts of anything, or sensory analysis or chemical analysis of results, or anything demonstrable. Kind of like many forum discussions! A major flaw, I think, in his analogy, is that humans don't split into two identical (but cell wall thinned) humans that are perfectly capable of reconstituting their cell wall reserves when provided the right conditions, like say in a batch of beer wort. Budding scars are not a permanent condition, leading inexorably to death.

3) In the stationary phase, he says "The wort is depleted of oxygen." It's true enough that the O2 has been depleted. But that happened back in the lag phase, most likely in the first hour. Not that this really argues one way or the other for starter strategy. Just thought it odd that Ulrich misunderstands what's happening with the O2.

 

[monkeymath]

1) There's a "lethal" phase which is obviously overstated. If all or even a great number of yeast starve to death or die (relatively quickly) of the products of their own metabolism, crashed starters would not result in an increase in viable cells as compared to the original count. But we know they do. Taken literally, nobody would be able to make beer. And even the fresh pack of yeast we get from manufacturers would be filled with dead yeast.

Maybe the word lethal seems a bit strong here. I think you have an idea of the supposed timeframe at which these processes are happening which is rather different from mine. The article doesn't specify it exactly, but I think this yeast death is more of a slow, steady decline.
I think we all agree that yeast in fully fermented beer slowly dies, don't we? And yes, the packs of yeast contain dead cells and that number increases as time goes by. That's why we keep the pack in the refridgerator: to slow down the inevitable. It's why you can't harvest yeast and pitch it five years later without any detrimental effects: yes, there will probably still be a few living cells, but this is far from an ideal pitch.

2) His argument is based on an analogy between the life cycles of microbes and humans, rather than actual counts of anything, or sensory analysis or chemical analysis of results, or anything demonstrable. Kind of like many forum discussions! A major flaw, I think, in his analogy, is that humans don't split into two identical (but cell wall thinned) humans that are perfectly capable of reconstituting their cell wall reserves when provided the right conditions, like say in a batch of beer wort. Budding scars are not a permanent condition, leading inexorably to death.

I think Ulrich was deliberately speaking in a simplified manner for a broad audience. I have no doubt he has the scientific background to back his statements up, even if the analogies are - by the very nature of analogies - simplifications of complex topics and may be lacking in parts. Even if the "argument" that you are alluding to may be, within the article, based on an imprecision, I strongly doubt that he himself arrived at this conclusion by means of an analogy.
As to why he does not deliver any "hard proofs", I can only speculate, but my guess would be that most, if not all, scientific research takes place in a controlled laboratory setup, and the exact numbers might not be transferable to a homebrew setup (even if such a thing as a "universal homebrew setup" existed). But unlike exact numbers, the guiding principles should be the same.
I also do not fully understand your objection about the difference in reproductive mechanism between humans and yeast cells. The yeast cells do reproduce more slowly as they age and they do eventually die - while their offspring may of course be very healthy at the same point in time. Kind of like humans, I think... ?

3) In the stationary phase, he says "The wort is depleted of oxygen." It's true enough that the O2 has been depleted. But that happened back in the lag phase, most likely in the first hour. Not that this really argues one way or the other for starter strategy. Just thought it odd that Ulrich misunderstands what's happening with the O2.

He actually states explicitly that at the beginning of the log phase, there is still an abundance of oxygen, which is depleted in its duration. He also puts the reproduction of yeast - which is closely linked to the availability of oxygen - in this timeframe, whereas the lag phase is described more as a preparatory step. Care to share some resources that use a contradictory notion of log and lag phase?

 

[VikeMan]

Maybe the word lethal seems a bit strong here. I think you have an idea of the supposed timeframe at which these processes are happening which is rather different from mine. The article doesn't specify it exactly, but I think this yeast death is more of a slow, steady decline.
I think we all agree that yeast in fully fermented beer slowly dies, don't we? And yes, the packs of yeast contain dead cells and that number increases as time goes by. That's why we keep the pack in the refridgerator: to slow down the inevitable. It's why you can't harvest yeast and pitch it five years later without any detrimental effects: yes, there will probably still be a few living cells, but this is far from an ideal pitch.

I agree with everything above. But I don't think it supports an argument of active pitch vs cold crashed. The timescales are not similar.

He actually states explicitly that at the beginning of the log phase, there is still an abundance of oxygen, which is depleted in its duration. He also puts the reproduction of yeast - which is closely linked to the availability of oxygen - in this timeframe, whereas the lag phase is described more as a preparatory step. Care to share some resources that use a contradictory notion of log and lag phase?

Oxygen is fully taken up in the lag phase. That's when it's used to make sterols to fortify cell walls. This has to happen before the exponential growth of cells, which happens in the log phase, and in fact defines the beginning of log phase. Here's what Chris White (White Labs) has to say about the phases: Fermentation Timeline - Brew Your Own

Dr. White doesn't explicitly state that the O2 has all been used up in lag, but what would yeast be doing (reproductively) with O2 other than building cell walls (which happens in lag)?

But here's some actual data regarding O2 takeup time. I actually watched it happening "live" (on the internet): Wort Study # 1 - Low Oxygen Brewing

 

[monkeymath]

I'm really glad to have sparked this conversation, thanks for offering a counterpoint @VikeMan !
I hope I do not come off as defensive - I'm merely attempting to clarify a couple of things that may have undergone a transformation
(1) the scientific head in Ulrich's head
(2) the analogies and simplified explanations that he presented to the group of homebrewers
(3) the explanations that were written down in the article
(4) the translation I provided

Admittedly, there were a few parts in the article that I could not wholly make sense of and which I either omitted (if they did not seem too relevant) or made an effort to repeat coherently and simultaneously close to the original.

Oxygen is fully taken up in the lag phase. That's when it's used to make sterols to fortify cell walls. This has to happen before the exponential growth of cells, which happens in the log phase, and in fact defines the beginning of log phase. Here's what Chris White (White Labs) has to say about the phases: Fermentation Timeline - Brew Your Own

Interesting to see them use the terms so differently. However, I'm quite certain this is merely a matter of definition or naming and where to subdivide the "stages" of fermentation. I do not think they actually disagree on the processes that happen, the order in which they happen or the effects they have.

1. the yeast analyzes the environment and available nutrients and starts to prepare its metabolism
2. the yeast consumes minerals, vitamins and oxygen, and goes on a yoga retreat to restore its youth
3. the yeast undergoes a stage of massive reproduction using the material it gathered in (2) and simple sugars
4. the yeast begins anaerobic fermentation

At the end of (2), the wort is basically depleted of oxygen, but still full of sugar. At the end of (3), the simple sugars are (mostly) gone and the yeast will have to work harder to break the more complex sugars down (using the mechanisms and enzymes developed in (2)).

Chris describes (1) and (2) collectively as the "lag phase", given that the brewer cannot see any activity. He refers to (3) and also (4) as the "exponential growth" phase, although he actually clearly states first that "The exponential phase occurs because yeast rapidly consume sugar. Wort sugar is consumed by yeast in a certain pattern. Glucose is used first, then fructose and sucrose. These are simple sugars and can be quickly shuttled into metabolism.", so I would think that we should already expect a slow down during this one phase, as the yeast has to turn from glucose to fructose to sucrose to maltose to, finally, maltotriose, which "is a tricky sugar for yeast to digest". It seems that no significant reduction of extract takes place during what he calls the "stationary phase".

Ulrich, on the other hand, defines the "lag phase" to be only step (1), which does not impact the wort. (2), (3) and partially also (4) are then part of the "log phase", where the yeast uses the oxygen and some of the nutrients in the wort to reproduce rapidly and happily. The tail end of (4) - when nutrients are becoming scarce - is then what Ulrich calls the "stationary phase".

Do you think that's a fair assessment of their difference in terminology?

EDIT: Quite interesting that link to ******************** - a complete uptake of oxygen within 1 hour after pitching yeast does not seem to conform to either Ulrich's or Chris' description, I think? And then it took another 4 hours for fermentation to start ... but what did the yeast do during that time, then?

 

[VikeMan]

Do you think that's a fair assessment of their difference in terminology?

Yeah, I dunno. I might buy that explanation, if the O2 wasn't being depleted by the end of one hour after pitching. For his definitions to fit, log phase (log stands for logarithmic growth, i.e. exponential growth of yeast population) would have to be starting before an hour has passed. It doesn't.

ETA: Didn't mean to be so short. IOW, yeast don't begin multiplying that soon, and nobody in the world that I have seen (with the possible exception of Ulrich) defines the start of log before the yeast have begun to divide. Given everything, I'm more inclined to think that for whatever reason, Ulrich either doesn't know how fast the O2 is taken up, or maybe had a brain fart when describing the process to the article writers, or maybe they misunderstood him.

 

[3 Dawg Night]

*applause.gif*

 

[RPh_Guy]

continuous aeration is not necessary

I'm not quite sure what you (or rather your translation) means by this statement. Oxygen is obviously necessary for yeast growth.

 

[TheMadKing]

From that it is also apparent that, also when making a starter, one should not wait for the yeast to settle out and should not, as is often recommended, decant the supernatant: yeast only fully drops out after fermentation has finished. The material at the bottom contains a high percentage of middle-aged and old yeast cells, near the end of their activity. Instead, keep your starter on the stir plate and pitch it entirely into the next stage once it has reached the end of log phase (or the beginning of stationary phase). That way, the yeast quality will be at its peak.

This is a false statement and my only real problem with this article. Otherwise thank you for the translation, it is much appreciated.

Yeast can flocculate for a variety of reasons. One of these reasons is because the temperature suddenly drops causing them to go dormant.

If you cold crash a starter at the transition from log phase to stationary phase, you can safely decant off the supernatant, warm the yeast to pitch temperature, and pitch only the healthy "young adult yeast" without all the oxidized starter wort.

 

[monkeymath]

I'm not quite sure what you (or rather your translation) means by this statement. Oxygen is obviously necessary for yeast growth.

I don't see how "I need to eat" necessarily implies "I need to eat constantly". But yeah, continuous aeration is - given a sterile process - certainly helpful; I think he was mostly making a point that it wasn't absolutely essential. Hard to tell for me tbh.

This is a false statement and my only real problem with this article. Otherwise thank you for the translation, it is much appreciated.

Yeast can flocculate for a variety of reasons. One of these reasons is because the temperature suddenly drops causing them to go dormant.

If you cold crash a starter at the transition from log phase to stationary phase, you can safely decant off the supernatant, warm the yeast to pitch temperature, and pitch only the healthy "young adult yeast" without all the oxidized starter wort.

I think this issue is my mistake, I should have highlighted the word "fully" in the bold-face bit. What Ulrich is saying is, I think, the following:

If you take actively fermenting wort and chill it down - without freezing it -, you will still have a significant amount of yeast in suspension for quite a while. And these yeast cells are predominantly the most active ones. On the bottom, you get all the dead cells, the old cells, and those yeasts that were willing to go to sleep when you turned off the lights - but the real party people are still going at it.

Other people might have a different take on the subject. I don't know, and I don't really find myself in a position to evaluate which position is correct.

 

[RPh_Guy]

I don't see how "I need to eat" necessarily implies "I need to eat constantly". But yeah, continuous aeration is - given a sterile process - certainly helpful; I think he was mostly making a point that it wasn't absolutely essential. Hard to tell for me tbh.

To grow you need to eat constantly.

I thought this thread was about "The ideal starter". I'm pretty sure continuous aeration is ideal, at least until the yeast reach terminal density.

 

[Vale71]

To grow you need to eat constantly.

I thought this thread was about "The ideal starter". I'm pretty sure continuous aeration is ideal, at least until the yeast reach terminal density.

That is correct, except there is no such thing as terminal density. Yeast will stop growing either when:

- they run out of carbohydrates as an energy and building-blocks source
- their sterol reserves are depleted and there is no oxygen available to synthesize them (sterol synthesis is strictly aerobic)

Whichever of these conditions occurs first will determine when growth will inevitably stop. I left out alcohol poisoning on the assumption that either the starter's OG is not high enough for this condition to occur before carbohydrates run out or that one is running a fed-batch process were alcohol production is constantly kept near zero.
The implication of this is that if the goal is to have the highest possible cell growth for a given starter that it's always advantageous to keep aerating till the very end since for as long as there are fermentables in the starter and yeast can synthesize sterols there will be growth, albeit slower and slower.
There are secondary benefits to continuous aeration as well since the energy yield of carbohydrates increases markedly with (prevalent) aerobic metabolism and aerobic metabolic pathways that are inherently more efficient can be activated. This results in reduced stress for the yeast and an increase in both cell mass and vitality.

 

[TheMadKing]

I don't see how "I need to eat" necessarily implies "I need to eat constantly". But yeah, continuous aeration is - given a sterile process - certainly helpful; I think he was mostly making a point that it wasn't absolutely essential. Hard to tell for me tbh.



I think this issue is my mistake, I should have highlighted the word "fully" in the bold-face bit. What Ulrich is saying is, I think, the following:

If you take actively fermenting wort and chill it down - without freezing it -, you will still have a significant amount of yeast in suspension for quite a while. And these yeast cells are predominantly the most active ones. On the bottom, you get all the dead cells, the old cells, and those yeasts that were willing to go to sleep when you turned off the lights - but the real party people are still going at it.

Other people might have a different take on the subject. I don't know, and I don't really find myself in a position to evaluate which position is correct.

the speed of flocculation is highly dependant on strain, - I would suspect he was speaking with lager strains in mind, in which case, yes they take longer to flocculate than most ale strains. But as you say "will stay in suspension for quite awhile", meaning you simply need to refrigerate longer with less flocculant strains in order to get the "party people" to drop out and be able to pour off the supernatant without losing a meaningful amount of healthy yeast.

Also, considering the fact that 90% of professional breweries (I'm talking even the big ones like Sierra Nevada, New Belgium, etc) use cold crashing and then filtration, and then only repitch the cold crashed yeast, which demonstrates that there is not a significant loss of quality when using this technique.

I just don't see what you gain (maybe mathematically more healthy yeast cells at the end of fermentation?) by pitching a full yeast starter, when you can clearly make good/great/consistent beer with yeast that has been cold crashed, and eliminate the off-flavors that can arise from pitching oxidized starter wort.

 

[RPh_Guy]

That is correct, except there is no such thing as terminal density.

Thanks for the explanation!

I don't have the sources on hand but I feel like I've read in multiple scientific articles that growth drastically slows at a certain cell concentration. Of course that may in fact be due to practical limitations providing aeration, or depletion of nitrogen/carbon/phosphorus/other nutrients.

I just don't see what you gain (maybe mathematically more healthy yeast cells at the end of fermentation?) by pitching a full yeast starter, when you can clearly make good/great/consistent beer with yeast that has been cold crashed, and eliminate the off-flavors that can arise from pitching oxidized starter wort.

Pitching actively fermenting yeast (at high kraüsen, not cold crashed) decreased lag time, which decreases oxidation and other possible off flavors.

 

[TheMadKing]

Thanks for the explanation!

I don't have the sources on hand but I feel like I've read in multiple scientific articles that growth drastically slows at a certain cell concentration. Of course that may in fact be due to practical limitations providing aeration, or depletion of nitrogen/carbon/phosphorus/other nutrients.


Pitching actively fermenting yeast (at high kraüsen, not cold crashed) decreased lag time, which decreases oxidation and other possible off flavors.

In my experience Pitching the proper cell count of yeast that have been cold crashed at high Krausen (and then warmed after decanting) into well oxygenated wort results in a lag phase (as judged by CO2 emission and Krausen formation) of less than 8 hours which is exactly the same as what I saw from pitching a full starter active starter into the wort. The difference, if it exists, is small enough to not even matter whatsoever.

I challenge anyone to be able to reliably taste the difference in a beer that had a lag time of 4 hours vs 8 hours.

I CAN tell you that a Helles that had a full starter of oxidized wort pitched into it virtually guarantees off flavors.

 

[RPh_Guy]

I challenge anyone to be able to reliably taste the difference in a beer that had a lag time of 4 hours vs 8 hours.

It depends whether you're pitching it into oxidized wort. Differences in lag time become apparent when there are delicate flavors that need to be protected.

I CAN tell you that a Helles that had a full starter of oxidized wort pitched into it virtually guarantees off flavors.

World-class breweries like Weihenstephan pitch actively fermenting wort (without crashing), so it's clearly not the wrong thing to do. If you aerate the wort long after fermentation completes, it'll oxidize.

You see I'm not disagreeing with you exactly, but it all depends on the details.

 

[TheMadKing]

It depends whether you're pitching it into oxidized wort. Differences in lag time become apparent when there are delicate flavors that need to be protected.


World-class breweries like Weihenstephan pitch actively fermenting wort (without crashing), so it's clearly not the wrong thing to do. If you aerate the wort long after fermentation completes, it'll oxidize.

You see I'm not disagreeing with you exactly, but it all depends on the details.

I will leave the first comment alone.

And doesn't weihenstephan use Krausening, I.e. actively fermenting yeast from other batches? Meaning that it has not been continuously aerated? I have no issue with that whatsoever, and the technique is sound. My issue is purely with pitching 1/32 of your total batch volume of aerated starter wort.

Narzciss (per Brukaiser) recommends pitching slurry and claims this is how all industrial German breweries function btw