Tuesday, July 31, 2012

Fermenter experiment

Here's the completed write-up on the fermenter experiment. We'll see if it ends up in Zymurgy or on Brew Strong.

Does Fermenter Geometry Affect Beer Quality?

Author: Joe Dunleavy
July 2012

Does fermenter geometry affect beer quality? I make beer in different fermenters all the time and I often wonder if different flavor profiles occur with different fermenters. I’ve brewed long enough to have a collection of different vessels to ferment in: (conical, carboys, better bottles, buckets). I’ve even fermented in a stainless steel square. When I bought the conical I went with the fourteen-gallon capacity as opposed to the seven-gallon capacity. I bought the larger conical so I could ferment both five and ten gallon batches. I usually ferment five-gallon batches and wonder if the yeast acts differently than the ten-gallon batches. With the five-gallon batches the majority of the wort is located in the cone of the conical. Does the fermenter’s shape, size, pressure on yeast, affect the flavor profile? Can anyone really tell the difference between beers fermented in different vessels?

I listen to various podcasts and read every magazine and book I can find on brewing. I’ve read and heard various things about fermenter geometry affecting the flavor profile, but most of this material pertained to the commercial level. The research that I have reviewed on the homebrew level just had too many uncontrolled variables for me to feel comfortable with. Does the yeast in my small batch size act the same as in a large commercial vessel?

I reached out to some home brewing royalty in hopes of getting some information based on some of their comments I’ve heard over the years. I emailed Gordon Strong and Jamil Zainasheff hoping to get some feedback on my fermenter questions. I also mentioned an experiment I was thinking of performing regarding fermenting the same beer in different vessels. Gordon responded to my email and volunteered to evaluate my beers if I performed the experiment. Jamil stated he would review my experiment as well.  I then mentioned my experiment to a coworker named Mike Guth who just became a BJCP judge. Mike thought that some of his brew club members would be interested in my experiment, particularly David Houseman. Mike then asked David if he would like to review the beers as well. David said he was in.

All of a sudden, I had three top members of the home brewing scene to review my beer. I’m not a BJCP judge so I thought this was a great development, but I was a bit intimidated sending my beer out to reviewers of this caliber. I also knew I couldn’t pass up this opportunity. Thank you to the reviewers for taking the time to review my experiment. I was amazed at your willingness to share your time and knowledge with another home brewer. Thank you for all each of you does for home brewing!

Here are the details on the fermenter experiment. Each batch was brewed and fermented in my brewery. The yeast I decided to perform the experiment on was a German Hefeweizen strain. I felt the German Hefeweizen strain would be a good candidate because yeast manufacturers document that the flavor profile can be altered. However, I haven’t seen anything documented about altering the flavor profile by changing the fermentation vessel.
My brewery capacity is not large enough to do this in one boil so I ended up brewing three different times on two different days. The two different days were a week apart and the weather on both brew days was about the same.
Brewery capacity mandated three separate batches and boils.

Here is each batch and a description of each fermentation vessel:
Batch 1 (D,E,B):       better bottle, carboy, and 1/2 conical
Batch 2 (A,F):            full conical, open ferment 4G
Batch 3 (C):               bucket
A – 9 gallons in a fourteen-gallon conical
B - 4.5 gallons in a 14-gallon conical
C - 4.5 gallons in a 6.5 gallon bucket
D - 4.5 gallons in 6-gallon Better Bottle
E - 4.5 gallons in a 6.5-gallon carboy
F - 3.75 gallons in a 4.5 gallon 6" deep open stainless steel pan - loosely covered with tinfoil during ferment and sealed with tinfoil as fermentation slowed.

The following parameters were the same:
·      grain sacks
·      grain percentages (50% German wheat, 50% German pilsner)
·      mash temperature rests(120,152,160,168) 
·      boil time (90min)
·      ibu level(13.4) - same hop batch from Hop Union 
·      carbon filtered water 
·      water treatment (calcium 91, magnesium 5, alkalinity as CAC03 30, sodium 20, chloride 98, sulfate 94), Residual alkalinity -38 - sulfate to chloride ratio balanced
·      oxygen level - 60 seconds of bottled oxygen
·      original gravities(1.046)
·      finishing gravities (1.012-1.013)

Each fermenter was sanitized with the same 5-gallon batch of an acid based sanitizer cycled between the fermenters. Each batch received one fresh package of German Hefeweizen yeast. The full conical received two fresh packages because it was twice the volume of the other batches. I normally would create a starter, but took that variable out of this experiment. This resulted in a slight under pitch of yeast, but I was more interested in the flavor profile differences than making the best Hefeweizen to style.

Each batch was pitched at 60 then held at 62 for a week in a fermentation chamber.  The temperature controller probe was against the conical cone insulated with bubble wrap. Each batch was fermented in the same fermentation chamber (a converted freezer). The yeast blew through fermentation in a few days. An interesting note is that krausen formed at different times between the fermenters with batch E starting a few hours before D or B. Batch A, F, C seem to krausen at the same time. After a week, the fermenter was moved out of temperature control into a room temperature environment and the yeast dropped out naturally. Then each batch was siphoned or transferred under CO2 to a CO2 purged keg and carbonated to approximately two volumes. Each batch was then bottled off the keg into new, washed and sanitized bottles. The bottles and bottling equipment were sanitized with an acid based sanitizer prior to each batch bottle fill.

My tasting notes:
I tasted these beers directly from the keg. I used coded cups without knowing which beer was which. I did a taste test at 38 degrees and did not pick up many differences between (A,B,D,E). C and F were different with F being the odd ball. I was sure that F was the open ferment. I then let the beers warm up 10 minutes and did a second taste test. 
This is when I could pick up the differences better.
Batch A (full conical): Clove, banana
Batch B (half conical): less clove and banana than A but fruitier, clearer
Batch C(bucket): dull, thin, slight puckering 
Batch D(better bottle): clove, banana, tastes a lot like A, clearer than others
Batch E(carboy): Balance of clove, banana, clearer than others, preferred
Batch F(open ferment): Huge fresh sour dough yeast, mouth puckering. This is a different beer than A-E - something got in this.
I then looked at the labels and tasted each beer again because I wanted to compare A and B, D and E, A and D. I preferred A over B - different flavors. I preferred E over D - close though. I preferred D over A - close though.
Batch C was thinner tasting than other beers - maybe oxygen pickup from bucket or something else. Batch F was like sticking my head into fermenting sour dough bread. Could this have been wild yeast?
My preferences were E, D, A, B, C, F.

Jamil Zainasheff tasting notes:
I would rank the beers in terms of fermentation quality: E B D A C F
E has a lighter fruity character and balancing clove. Some malt character and much better head form and retention. Clearly the best fermented of the bunch, but not necessarily a fantastic hefeweizen. I could drink a pint of this.
Others exhibited various forms of sulfur and acid (C F).
Some very high banana and juicy fruit character (B).
F was undrinkable, it was down right horrible. Acetone, sourness, and a fart stink.
C wasn't much better. It was sour and exhibited signs of skunkiness.

David Houseman tasting notes:
A - banana, clove, light, smokey
C - lager like, less banana and phenols, bright
B - bit less banana & phenols than A but more bitter and astringent
E - lemon, clove, not as much banana/ester as A & C. Medicinal, lemon rind 
D - dull, lemon, less banana clove, some coconut notes, oxidized
F - off, diacetyl, dms, fusels.

Gordon Strong tasting notes:
All styles had similar color (light gold) and clarity (some haze).
A: strong perfumy notes, noticeable banana. most intense of all the samples.  probably the cleanest/purest of flavors and strongest aromas. med-full body. clean flavor matching aroma (banana, spice/clove/pepper).  My clear favorite, 1st place.
B: moderate perfumy aroma, not totally clean -- slightly dirty/funky, hint of sour. medium body. moderate banana and spice with light solvent. 3rd place.
C: moderately low perfumy notes. Clean, just subtle. Over time, the aroma faded and became less clean; hmm.  Full body. Clean flavor, subtle, mild fruit and spice.  5th place (3rd place earlier before the aroma died).
D: moderately strong perfumy notes. clean, mostly banana with a hint of solvent. full body. mild fruit and spice, clean flavor similar to sample C. Finish has a touch of acidity/sharpness/alcohol, which gave it a slight bite that balanced the malt.  2nd place.
E: moderately low perfumy notes. Not totally clean -- a little dirty. Kind of like a bad homebrew hefeweizen (warm ferment).  Full body. Some increased phenol/solvent. Moderate banana. More spice than others. Slight acidity.  4th place.
F: low perfumy notes. quite neutral in the nose; not much aroma at all. medium-full body. a little snappy in the flavor -- green apple? moderate banana. had the most acidity. not too clean, sharper in finish, hint of vinegar? most unpleasant aftertaste.  My clear bottom, 6th place.
So the 1st and 6th places were pretty obvious to me, while the other ones could easily move up or down relative to each other depending on the moment. The overall intensity of the aroma varied quite a bit.  Surprisingly, the body seemed to have some differences as well.   Also saw the acidity change between samples.  
I thought E had the highest phenols, and F had the most issues. I focused on the differences since there were very many similarities.
Note: Gordon tasted the sample 4 weeks after everyone else.

Preference comparison:
David:             A, C, B, E, D, F
Gordon:         A, D, B, E, C, F
Jamil:             E, B, D, A, C, F
Joe :               E, D, A, B, C, F

I was surprised with the results. I expected all the reviewers to pick the same first and last beer, but that did not occur. My expectations were that the full and half conical would taste the same, they didn’t. I expected the similar shaped vessels (carboy, Better Bottle, bucket) to produce the same flavor profile, they didn’t. Each flavor profile was similar across batches, but they were differences in each one. It was kind of like attending a family picnic; you know you’re all related but each one is different in their own special way.

So what does all this mean to me? It was clear to me that beers do differ when fermented in different vessels.  I now can utilize fermenter shape as a variable during fermentation to work towards a particular flavor profile. This experiment has also opened up a few new questions for me. This experiment was performed with a Hefeweizen strain and it performed differently between the fermenters. Can other saccharomyces yeast strains be manipulated to alter the flavor profile? Does brettanomyces produce different flavor profiles when fermented in different vessels? How do bacteria (pediococcus, malolactic, and lactobacillus) behave in different vessels? How does your favorite yeast behave when different fermentation vessels are used? I would suggest conducting your own experiment with your favorite yeast and see where it takes you. You might be surprised with the results as well!

1 comment:

  1. Really interesting experiment...just found it via The Beer Engine Blog. One potential element that stood out to me is the following: What if you went through this process but fermented the same beer side-by-side in the same shaped vessel...will that create some variation as well? My inclination is that it might. To what extent, who knows? Yeats is a funny thing and there is probably some margin for variation even when all the process and parameters seem to be exactly the same. Perhaps on a larger commercial scale there is less margin for stark variation due to the mass of batch size, where on a homebrewing scale that margin for variation increases.