Some starter calculations & BeerSmith

[More_Hops_Please]

Okay, so I did some kitchen counter science in the interest of seeing just how accurate BeerSmith 2 is at recommending starter sizes & pitch rates. As I had suspected, the results were interesting.

I was using 2nd generation yeast washed from a Robust Porter brew for these calculations. The yeast was Safale US-05 which had been slurried and decanted twice then left to flocculate and stored at 40° F for 40 days. The slurry was stored in 4 one-quart mason jars.

For the purpose of this starter, I only used one of the jars, which I estimated to contain about 50b viable yeast cells. I calculated that by multiplying the volume of my yeast solids(15mL) by 4.5b cells per mL(according to Jamil). I then calculated a 90% viability(about right for good washed yeast), and added a 20% fudge factor on account that at least some of my solids are non-yeast proteins and/or trub. Here's my math:

15mL * 4.5b c/mL = 67.5b c
67.5b c * .9 viability = 60.75b c
60.75b * .8 estimated yeast density = 48.6b viable cells​

That's a VERY rough number, particularly because I don't know the actual ratio of yeast solids to non-yeast solids and I was just going with an educated guess. I do feel confident that if anything, I've OVER calculated my beginning yeast count, as I used very optimistic viability and density factors. Which is to say, I have at most 50b viable yeast cells, probably less, but certainly not more.

Okay, on to the starter. I pitched my flocculated 50b yeast cells to a 1.5qt(1400mL) un-hopped starter with an OG of 1.045 with no stir plate. Per BeerSmith 2, when the starter is finished, this should result in 127.8b viable yeast cells. So how many did I end up with? After all starter yeast had flocculated, I measured my yeast solid volume and calculated cells in the same manner. I had a final volume of 87mL yeast solid. Which calculates to:

87mL * 4.5b c/mL = 391.5b c
391.5b c * .9 viability = 352b c
352b c * .8 estimated yeast density = 281.9b cells!​

Over 280 billion viable yeast cells! BeerSmith 2, you're a damn pessimist! But wait, this number is actually too low for one major reason: yeast solid to non-yeast solid ratio.

You see, I know the volume of solids I started with, and I know the volume of solids I ended with, but more importantly I know that the delta is almost entirely yeast solid. No trub or hop sludge was added to the equation, so any change in my solid count can only be attributed to yeast reproduction. Knowing that, lets take the yeast density out of the equation entirely and just look at how many NEW yeast cells we've produced, since that's the number we can have the most confidence in, and is going to give us a more accurate representation of our final cell count:

87mL(final solids) - 15mL(starting solids) = 72mL delta
72mL * 4.5b c/mL = 324b c
324b c * .9 viability = 291.6 b new cells!​

Counting new viable cells only(and 90% viability is actually rather low considering the age of these new yeast), we're up near 300 billion cells, which is very nearly what BeerSmith 2 calculates using a stir plate(remember, no stir plate was used here) in a 1.5qt starter.

So you're probably now thinking the same thing I was thinking... that seems way too high, something has to be wrong with these numbers. Well lets consider which numbers could be off? My volume measurements were taken with painstaking care, so I am confident in those. Maybe Jamil's estimate of cells per mL was too high? Could be, and there's no way to check without a hemacytometer, which I have to admit I do not own.

So lets say Jamil's number is too high for this particular strain of yeast... what if it was a very conservative 2.5 billion cells/mL? Well then I'd only have about 190 billion viable cells... which is still over 50 billion more than BeerSmith 2 predicts. The only explanation is that BeerSmith is WAY too conservative in their starter calculations(at least for this particular strain at this particular volume).

You know, before I started this, I was considering buying a stir plate. But now I'm thoroughly convinced that I don't need one.

 

[winvarin]

Your math looks right to me.

Where did you get the cells/ml calculation? Is it from Jamil's "Yeast" book? I admit I have not read all the way through my copy. I am trying to determine the cell density of clean starters from DMS wort. My plan is to begin growing starters from White Labs vials of yeasts I use frequently, then decanting and washing the yeasts with canned, sanitized water. I will get X milliliters of MOSTLY clean yeast since I will be doing only DME (no hops and will leave the break material from my canned starter wort behind).

I am trying to get a good idea of the number of cells contained in that X milliliters so that I can use my date and cell amounts for calculating cell growth when I use these harvested jars to make starters for beers later down the road.

Is 4.5 billion/ml a good #, or should I estimate lower (like your 2.5 estimate). I plan on letting yeastcalc handle the viability calculation based on the density and dates I feed in.

 

[More_Hops_Please]

Yeah, that number comes from Jamil's Yeast book. It is the estimate he cites for compacted, floculated yeast. The thing is, we obviously know that while that's probably a good ballpark figure, it's going to vary pretty widely depending on the strain of yeast. Different strains tend to floculate into different "colony structures" which would affect the density of the cake. Also to get to a compacted state, I'd say you have to let that yeast sit in a cold temp for quite a while(mine was dormant for about 6 weeks, which I figured was long enough).

My 2.5b was mostly a hypothetical to prove that even at a less dense state, the calculators are REALLY conservative. How conservative is tough to tell without having a real solid cell density number which admittedly I don't. I'd guess that its higher than 2.5, but probably a little lower than Jamil's "ideal" number.

My suggestion for you, since you're coming straight out of a vial would be to maybe set that virgin yeast in a fridge for a week or two so you know its nice and dormant and settled, then measure your yeast solids volume and compare it to the White Labs cell count for your particular vial size. That would probably give you a pretty good number for the particular strain you're using. Your starter yeast won't vary that much because you won't have a lot of trub in a pure DME wort.

 

[winvarin]

Maybe I am missing something. Are you saying that if white labs says there are 100 billion cells in their tube and roughly half of the tube compacts to solid, then I just need to figure out what volume half of the tube is, then apply that ratio to my yeast start?

 

[winvarin]

Hit send too soon. So if I determine that 20 ml is solid in that vial, and I wind up with 100ml of solids after fermenting the starter, then I can assume I have 500 billion cells?

 

[More_Hops_Please]

Maybe I am missing something. Are you saying that if white labs says there are 100 billion cells in their tube and roughly half of the tube compacts to solid, then I just need to figure out what volume half of the tube is, then apply that ratio to my yeast start?

Exactly. It's just a suggestion, but I think that would give you a real good cells/mL count to go with for your White Labs strains.

To measure volume of the yeast solid accurately, I marked with a very fine felt tip where my cake ended. Then after pitching, I refilled to that line with water, then dumped the water to a graduated cylinder and measured the volume of water. Graduated cylinders are generally just a couple of dollars at your LHBS.