Fundamental shift in yeast starter preparation?

[troxerX]

Had maintenance starters to build so I prepared the following. I didn’t read the research but did listen to the podcast so this is what I decided to do. Some amounts were based or around product label recommendations (except wyeast). Might be a good starting point for modifications and simplification.

Starting Volume: 1 1/2 gals (5.7L)
One large batch to be split in 2 starters
OG: between 1.008 - 1.012 (est.)

DME - 3oz - used Organic Maltoferm 10001 (1.044 PPG)
Dextrose - 1.5 oz - (some simple sugar in addition of DME)

Magnesium Sulfate - 3.0g
Zinc Sulfate - 0.10g
CaCl2 - 2.8g (1/2 tsp)

GoFerm - 2.0g (label recommendation of 1.0 - 1.5g/gallon)
Fermaid O - 2.0g (label recommendation of 1.0g/gallon)
DAP nutrient - 5.7g (1 tsp) (label recommendation of 1tsp/gallon)
Fermax - 10g (2.25 tsp) (label recommendation of 1.5 tsp/gallon)
Wyeast nutrient - 5.7g (1 tsp) (label recommendation of 1/2 tsp/5gallons)
Baking Yeast - directly added to boil (boiled between 5 - 10 mins) - 10g
WL Servomyces - 1 capsule

Definitively a witches brew starter.. obviously the color was much lighter. Given the low amount of malt added, I decided to supplement with various products given the lack of details from the manufacturers. But as this evolves there will be definitively a way to simplify.

The starters took a bit longer to become active, and now are active but at a much slower pace than usual which seems to be aligned with the fact there is less sugar to go through..

 

[VikeMan]

I didn’t read the research but did listen to the podcast so this is what I decided to do. Some amounts were based or around product label recommendations (except wyeast).

Then you'll certainly have no pre-conceived notions about what the research found.

 

[Bassman2003]

Let us know how it turns out.

 

[troxerX]

Then you'll certainly have no pre-conceived notions about what the research found.

VikeMan your’e certainly right!, I went though it now and is exactly what I expected - superb work by the people involved, but in a scientific lab under controlled environment, equipment, materials, procedures and measurements out of reach, if not impractical, for the average brewer. However; from the podcast, and research paper, the key take-aways are still the same, if you agree;

1) Wort supplemented with yeast extract is a nutrient rich medium for yeast propagation.
2) Average brewer should make step improvements to lowering sugar content (max 3P) and increasing nutrient supplementation to their starter worts.
3) doing so, the C:N ratio will be lowered where yeast metabolism is tipped to more cellular respiration than fermentation thus physiologically preparing the yeast for better fermentation performance.

Given the lack of supplements product data (or data accuracy) from the different vendors it will be difficult for the average brewer to replicate the same conditions during the study, but - we have to start somewhere, right?

 

[VikeMan]

1) Wort supplemented with yeast extract is a nutrient rich medium for yeast propagation.

Sure. If you add more nutrients, it's more nutrient rich.

2) Average brewer should make step improvements to lowering sugar content (max 3P) and increasing nutrient supplementation to their starter worts.

They showed that a lower C:N ratio resulted in a higher number of yeast cells per gram of sugar, for worts of about 4.5P or lower. They did not show that an average brewer (or anyone else) should simply use a lot less sugar and load up the nutrients in a standard volume starter. If you want to get the same amount of cells as you normally get with a standard starter, but at lower gravity, you'd have to make a bigger starter. If you're going to experiment, I'd recommend using the math from figure 2-B as a reference.

3) doing so, the C:N ratio will be lowered where yeast metabolism is tipped to more cellular respiration than fermentation thus physiologically preparing the yeast for better fermentation performance.

Yes, they showed (not that I think anyone ever doubted it) that you can reduce the crabtree effect by decreasing sugar concentrations and ensuring adequate nitrogen is available. The result of that is more cells produced per gram of sugar. But no, I would be very careful about concluding anything about "better fermentation performance." Faster and more attenuative doesn't necessarily imply better, if the goal of fermentation is beer that meets expectations of what a particular style/recipe should be.

But I will go out on a limb here and predict that there will be some who listen to the podcast, make one liter starters with a 2 or 3P wort, successfully ferment beer with the result, and declare the beer to be amazing. A new "best practice" will have been born.

 

[bionicbelly]

"But I will go out on a limb here and predict that there will be some who listen to the podcast, make one liter starters with a 2 or 3P wort, successfully ferment beer with the result, and declare the beer to be amazing. A new "best practice" will have been born."

I think everyone talking about doing this is starting with the same amount of canned wort, and just doing a more dilute prep of that. Maybe someone else would make the same volume of a lower gravity starter, but I don't think you are discussing anything with them at the moment.

"But no, I would be very careful about concluding anything about "better fermentation performance." Faster and more attenuative doesn't necessarily imply better,"

I would be careful about concluding anything about this, either way. Faster and more attenuative doesn't imply anything really, but it could be better, or it could be worse. How you gonna know unless you try?

 

[bionicbelly]

Given the lack of supplements product data (or data accuracy) from the different vendors it will be difficult for the average brewer to replicate the same conditions during the study, but - we have to start somewhere, right?

This. I think I'm just going to do one starter like I always do, and do another Low gravity, high nutrient starter, (of unknown nutrient concentrations, but the same amount of sugar)) and run those through a split batch of beer, and see if there is a perceptible difference. Should be fun!

-I have a couple other batches I have to get brewed first, but I'll do this after those are done.

 

[VikeMan]

I think everyone talking about doing this is starting with the same amount of canned wort, and just doing a more dilute prep of that. Maybe someone else would make the same volume of a lower gravity starter, but I don't think you are discussing anything with them at the moment.

I'm really discussing it with anyone who might listen to the podcast(s) and/or read certain posts in this thread and who might therefore think that a low gravity, same volume starter was somehow indicated by the study. It's good to know that that's not everyone. And it's probably not @troxerX , whose post I was replying to, either. But I could see someone reading some of his words and assuming otherwise. e.g.:

"2) Average brewer should make step improvements to lowering sugar content (max 3P) and increasing nutrient supplementation to their starter worts."

Note no mention of volume there, and if I didn't already know better, I'd probably take that "lowering sugar content" to imply same volume, less sugar.

So I'll keep pointing to figure 2-B in the study whenever it seems called for.

 

[Spivey24]

This is just purely anecdotal, but I ended up making a lower gravity, higher volume starter this time. I was short on concentrated wort, so ended up using about 25% less total sugar and increased the volume 25%, and added yeast nutrient by WhiteLabs. I ended up with wort around 1.020. I used a refrigerated previously overbuilt starter of Wyeast 1056 from about a month ago. I did the normal shake of the flask to aerate, but just to add yet another variable into this mess, I ran my aquarium air pump with filter into the flask on the stir plate. I’ve done the normal strength starter this size and yeast maybe 25 times so I am pretty familiar with the performance. This starter produced a decent amount of yeast, same or more than usual, and I pitched it 1.5 days ago and we are at 33% attenuation already, (according to Tilt) which is about 8-10 points faster than any other batch I looked up after the same amount of time. So I will see how it finishes out and tastes of course. But I might play with this setup and see if I can get consistent results. And yes, this went completely off the rails for what this paper was discussing.

Starter size:5L on stirplate for 2 days
initial yeast cel count: maybe 150 million by calc
OG of IPA: 1.054
batch size: 11 gallon
pitch: 2/3 of starter decanted (1/3 saved for next time)

 

[troxerX]

Alright, I’ve been working on the math and something doesn’t seem right based on the expectation we will need a much much larger starter volume, @VikeMan could you please double check?.

So let’s say I want to double the cell count of a Wyeast smack pack - so I start with the standard Wyeast promise of 100B healthy cells per pack - the target is to double the population so we want to produce an additional new 100B “Physiologically Tuned” cells.

Assuming we meet a C:N ratio of 100, looking at Figure 2-B (graph below), the experiment found that at 2P, the cell production efficiency is approximately 1.6x10^10 cells per grams of fermentable sugar (16B cells/g of sugar). So I divide the target Output (new cells) by the efficiency (efficiency defined as Output/Input) to determine the Input (in this case the g of sugar) - so 100B cells / 16B cells per g sugar = 6.25 g of fermentable sugar. (0.22oz of sugar).

Lets say for simplicity I’m using dextrose with a PPG of approximately 1.046, so to make a 2P (1.008) starter of dextrose with 0.22oz, using a factor of 5.7, I will need approximately 0.18 gallons = 0.68L. At this point the calculation is linear so if I want to produce 500B cells I just multiply by 5 both the sugar and volume, so I will need 1.1oz sugar in 3.4L etc etc. Given the high efficiency found during the experiment then we probably don’t need such large starters (and much nitrogen/nutrient) to apply this work unless there was an error in the experiment or my math above.

 

[VikeMan]

At this point the calculation is linear so if I want to produce 500B cells I just multiply by 5 both the sugar and volume, so I will need 1.1oz sugar in 3.4L etc etc.

I haven't checked the math up to this point, but why do you assume it's linear, regardless of inoculation rate? Chris White's data says it's not, i.e. multiplying starter size by 5, and holding starting cells constant, wouldn't increase new cells generated by a factor of 5.

But if we assume that it is linear (as I believe Kai Troester's experiments implied), the current study's data still implies that a lower gravity starter would have to be bigger than a higher gravity starter.

Given the high efficiency found during the experiment then we probably don’t need such large starters (and much nitrogen/nutrient) to apply this work unless there was an error in the experiment or my math above.

I don't think low C:N starters would have to be huge. I do think they would have to be bigger. The only way to find out how big, under homebrew-starter-friendly conditions with possibly unknown nitrogen qtys, is for someone to run the experiments, and count the cells. That someone won't be me, because it's only a first step. Once that's dialed in, assessing the resulting beer quality would be the hard part. It's not a simple as "I tried this new thing and I think my my beer is fabulous." If attenuation will be higher (as suggested by the study), I don't see how that would make beer better or even as good. i.e. getting a lower FG than normal for any particular recipe doesn't seem like a win, even if I save a buck on DME. So I'm personally not motivated to do the work. But I'd encourage anyone who is motivated to do it, in a controlled, repeatable way. Any other way will inevitably lead to premature declarations of new best practices. And IMO we've had too much of that in the past few years.

 

[Bassman2003]

From a rough, off the cuff relation - 2L to 3.4L sounds about where I would expect the lower sugar starter volume to end up for equal amounts of cell production compared to a high sugar starter. You should get plenty of bicycle riding fresh yeast cells with this sized starter. Let us know how they do!

 

[troxerX]

@VikeMan, agreed, for biological systems is probably not linear but for the average brewer should be good enough approximation. My gut feeling tells me the same that you are proposing - the need for bigger starters - even when my horrible math shows that the volumes may not be that far from what homebrewers already prepare. One thing we are still missing is that the experiment is switching some of the sugar for nutrients (contrary to what has been done historically with only sugar) and that may play a role in helping keep minimum changes to the volume while still helping increase cell production efficiency.

As for claims of fabulous beer, +1, the benefits from adopting the study conclusions will be around decreasing the likelihood of stuck fermentations, improved attenuation and increased likelihood of faster fermentation. I don’t see any significant benefits in flavor although we will still have people biased around that.

 

[VikeMan]

the benefits from adopting the study conclusions will be around decreasing the likelihood of stuck fermentations, improved attenuation and increased likelihood of faster fermentation.

I'm curious as to why you (or any homebrewer) would see "improved" (i.e. more) attenuation as a general benefit.

Regarding stuck fermentations...if that turned out to be the case, I guess that's a benefit. OTOH, I've brewing for a long time and have never experienced a truly stuck fermentation. I'd wager that most people who might go to the trouble of experimenting with starter C:N ratios probably weren't having stuck fermentations anyway, i.e. they were pitching adequate amounts of healthy yeast cells into wort with adequate nutrients.

 

[troxerX]

Same as you, I’ve never had issues with a stuck ferment but have seen a lot of those here in this and other forums.

Having a fit yeast increases the likelihood of that yeast attenuating to its label specifications. From my point of view, I would think this is an advantage as it helps minimize variability to meet your final gravity target. This will give you better predictability on your FG. Again none of this will ever be perfect but will increase the likelihood of success against your goal. There is nothing more I hate than setting a target mash temp, hitting the OG I’m looking for and then having the beer coming up 5 points sweeter cause of a yeast issue. Again it’s all on tipping the likelihood and probability scale on your side.

 

[VikeMan]

Having a fit yeast increases the likelihood of that yeast attenuating to its label specifications.

Ah, but the claim in the paper is higher attenuation (not more consistent).

Just a sidebar on yeast strain published attenuation ranges... IMO they are almost meaningless, because they know nothing about the fermentability of the wort you will be pitching into. Every strain can and will attenuate both more and less than advertised, and in my experience, it's all fairly predictable.

There is nothing more I hate than setting a target mash temp, hitting the OG I’m looking for and then having the beer coming up 5 points sweeter cause of a yeast issue.

If this was happening to me, I'd be taking a serious look at my process (thermometer calibration, etc.). I predict my FGs using my BrewCipher, and I can't recall ever being off by as much as 5 points. Not once in hundreds of batches. I don't know about your particular batches that have come out with higher than expected FGs, but just in general, it's pretty common for people to wonder what happened with "under" attenuation, when they in fact got just what should have been predicted, because they just accepted the stated target FG from a random internet recipe, or they simply applied a yeast manufacturer's number, or they used software that doesn't consider grain bill (for example).

 

[troxerX]

Yeah I re-use yeast multiple times. Knowing the yeast is fit would a positive for me to know what to expect.

We hear brewers mentioning that their house strains can only go up to certain number of generations before the beer quality is impacted, maybe this can also help with that. Also once the yeast starter is consumed, you will end with less alcohol in solution which is ultimately good for the yeast itself. My point being that there can be positives that we haven’t thought of yet.

Now that we have an idea on sugar and volume, I’ve been digging into the N side of things. Found a couple references where the total N composition for yeast extract (YE) is estimated to 9-12%. Some found 10%.

https://www.researchgate.net/post/What_is_the_composition_of_yeast_extract
This N is present in various forms:
Average Protein Content of extract - 73-75%
From the proteins:
Free Aminoacids: 35-40%
Di, Tri, Tetrapeptides: 10-15%
Both amounting up to 55% of total protein fraction with remaining amount being oligopeptides at 45%.

https://0eb15d32-a-62cb3a1a-s-sites...1TJaasAkPS5jni4tgHWQeC96zA4g==&attredirects=0
We know that glucose is 40% carbon. If we were to look at specific atomic fractions then it will be N = 12% (assuming we use YE) and C = 40% (assuming we use glucose) per gram, if we go with protein fraction then N = 75% and C = 100% per gram. Again not sure which one is best but this should help running some numbers for N. Of course we will have to assume most nutrient products available for brewers are YE based.

 

[TheMadKing]

So was browsing the March/April issue of Zymergy and this article seems relevant to the discussion.

The author also seems to come to a misleading claim by narrowing the focus of the question to make it a true statement IMO

 

[DuncB]

Well Zymurgy isn't a peer reviewed journal, so anything the editor agrees to gets in.
That bit about forcing respiration in the last line of the 3rd paragraph from the end is a bit like discussing improving oil extraction from a well with refining petrol to improve engine performance. Not that helpful.

 

[TheMadKing]

Well Zymurgy isn't a peer reviewed journal, so anything the editor agrees to gets in.
That bit about forcing respiration in the last line of the 3rd paragraph from the end is a bit like discussing improving oil extraction from a well with refining petrol to improve engine performance. Not that helpful.

Agreed, and I wasn't putting it forward as a definitive write up, just found it interesting that this has popped up more than once.

And here it also seems to disconnect ideal yeast biomass production from beer production.

 

[Craiginthecorn]

With the passage of time, I'm wondering about real world homebrew experience with this method, particularly from brewers who have used it a number of times.

My personal experience is limited to a single batch using a then new Bootleg Biology beta yeast hinted as being the Blauges strain. Attenuation was lower than expected but properly dry for the style. The batch took months to clear and longer for off-flavors to clean up, and it had a metric crap ton of sulphur that took even longer to clear. I have no other experience with this strain, so I have no idea whether to attribute the performance to the the starter method or the yeast itself. In the end, the beer, which was used a recipe shared directly by Blauges for their spelt saison, Saison d'Epeautre, tastes great and was sparklingly clear -- very close to the original, which I have had several times.

 

[Spivey24]

I posted earlier in this thread about using lower gravity starters. I settled on diluting 4L 1.035 starters to 5L so ending up with around 1.028, and pumping in air for the first 24 hours. Nothing too drastic but my flask is only 5L so can’t go any further. I have done about 12 batches since using this method - 6 ale and 6 lager 10 gallon batches. Keeping with my old recipes and mash schedule, I have seen consistently better results (better starts, faster ferments, higher attenuation), but only with the ale yeasts. No improvement in the lagers at all. Again, this is far from scientific and purely anecdotal, but it’s an easy change for me so I will stick with it. I still haven’t done a side by side test, but really lack the equipment to do a fair comparison.

[edited to say higher attenuation]

 

[cactusgarrett]

lower attenuation

Lower FG or lower attenuation percent (causing higher FG)?

 

[Schlenkerla]

One thing I have yet to see on this thread was anything about the starter vessel size verses the starter volume. Supposedly a limit of 10% utilization is necessary to ensure the aeration is occuring in the vessel. Assuming you want a bias for respiration over fermentation.

Keep in mind they are talking about a shaking table vs a stir plate. The concept is the same regardless.

https://www.infors-ht.com/en/blog/5-ways-to-increase-biomass-in-your-shake-flasks/

 

[SanPancho]

i have taken to oxygenating my starters. typically once when i initially set the stir plate in motion to get things working. then a few hours later, later that evening, etc. etc. to make sure any growth will also have access to o2 as well. i dont do counts or anything, but i'd have to guess this should help solve the problem of low aeration.

 

[Schlenkerla]

Worth noting, a 10% vessel fill limit puts that at 500ml on a 5L flask. So, it seems like the respiration growth approach has a limiting factor with a practical vessel size.

Per the Sui Generous Brewing suggestions or rationale of stepping up 1.040. - Anything under 500ml should be at 1.008 with nutrients, anything over 500ml should be done at 1.040. Also the final step prior to pitching should be at 1.040 for the yeast recoding so the lag time is shorter.

 

[MaxStout]

Hey, @Schlenkerla, good to see you back!

 

[Schlenkerla]

Hey, @Schlenkerla, good to see you back!

Thanks!

 

[Craiginthecorn]

With the passage of time, I'm wondering about real world homebrew experience with this method, particularly from brewers who have used it a number of times.

My personal experience is limited to a single batch using a then new Bootleg Biology beta yeast hinted as being the Blauges strain. Attenuation was lower than expected but properly dry for the style. The batch took months to clear and longer for off-flavors to clean up, and it had a metric crap ton of sulphur that took even longer to clear. I have no other experience with this strain, so I have no idea whether to attribute the performance to the the starter method or the yeast itself. In the end, the beer, which was used a recipe shared directly by Blauges for their spelt saison, Saison d'Epeautre, tastes great and was sparklingly clear -- very close to the original, which I have had several times.

I just did a little half-@ssed experiment. I’m brewing a 1.057 OG beer that called for Irish Ale Yeast. I had a pack…18 months old and a full year past its Use By date. I made a 500ml 1.013 starter and added 1.25g of FermFed. I then made a second identical starter, but 2x the volume, decanted only about half of the previous starter, then combined. The second starter attenuated down to 1.002. Seemed like it was “working”. Finally, I got cold feet and made a conventional 1.040 1L starter, decanted the 2nd starter. All of these were placed on a stir plate. End result as the fastest start to a fermentation that I’ve ever encountered. Three hours max in lag phase.

Now, the reality is that what I did is actually pretty close to the existing conventional wisdom. When propping up a weakened yeast, do so starting with a low gravity wort to reduce osmotic stress, then repeat with successively higher-gravity wort. So, in the end, I didn’t ”prove” a thing, but it’s something of a data point that I thought I’d share anyway.

 

[Bassman2003]

Thanks for sharing. It is all about getting the little yeast cells to do what we want. Makes complete sense that a low gravity wort allows them to work through easier when their numbers are lower. Basic physical exhaustion.