how precise does fermentation temperature control need to be before the diminishing returns make it not worth the effort?

[nathanscrivener]

This may be somewhat yeast dependent, so I'm not talking about kveik for example here.

I started out making extract beer on a WilliamsWarn brewmaster, which is essentially a glycol jacketed stainless unitank. It was capable of maintaining fermentation temperature to quite a high degree of precision. Recently I sold that that and decided to go the brew in a keg in a temperature controlled fridge way, as I figured I had unnecessary capital tied up in the brewmaster and could achieve the same result with cheaper equipment.

The first brew I just placed the temperature probe for the temperature controller in my fridge and let it work on ambient temperature. I found that there was a fair bit more variability in the beer temperature (measured by my Tilt hydrometer), up to 3c warmer than the ambient. (I think that's about 5f). Despite this, the beer (a pilsner) turned out great. One of my best yet in fact.

With my current brew I've strapped the thermometer probe to the side of the keg and put some bubble wrap over it for insulation. This is now able to keep it within 0.5 degrees C even during high krausen. Whatever variation is occurring is within the margin of error of the Tilt.

I've also built a DIY brewpi controller which promises even more accurate control, down to 0.1 degree accuracy. Haven't tried it yet because it requires installing a thermowell in my kegs.

All of this has got me wondering, what level of precision is necessary before any further improvements are not going to make a difference in the quality of the finished beer? My best guess is that there MIGHT be some difference in SOME fermentations between ambient and 'strapped to the side' level of control. But anything more than that is unlikely to provide any further benefit to the beer. Which would make the brewpi a bit of a waste of time (although the ability to track and control via a web interface is nice).

Thoughts?

 

[day_trippr]

I can't state what precision is needed, but the problem with control via "ambient" is "ambient" typically has an excessively elastic relationship with wort temperature. Pretty much the opposite of "precise control", there can be a good 5°F differential between chamber ambient and wort and that's into the range of significance with many if not most strains vs running them consistently in their optimal range.

fwiw, I run all my chambers using wireless BrewPi "minions", usually with strapped probes. Recently added a thermowell to see how closely the strapped probes track. Over multiple batches the strapped sensor with 1" of closed cell foam insulation tracked within 1°F of the thermowell'd sensor located dead center (radius and height) in the wort in a 6.5g glass carboy.

That's the level of control I can live with and avoid dealing with thermowells entirely...

Cheers!

 

[DVCNick]

fwiw, I run all my chambers using wireless BrewPi "minions", usually with strapped probes. Recently added a thermowell to see how closely the strapped probes track. Over multiple batches the strapped sensor with 1" of closed cell foam insulation tracked within 1°F of the thermowell'd sensor located dead center (radius and height) in the wort in a 6.5g glass carboy.

That's the level of control I can live with and avoid dealing with thermowells entirely...

Cheers!

Interesting...
I've been considering experimenting with this myself recently.

I currently use a Big Mouth Bubbler with a thermowell and the temp probe located almost dead center of the beer. The "problem" (if it is one) is as follows:

-Set temp to 68, for example
-Inkbird clicks on my chamber at 69
-By the time the probe in the dead center of the beer registers 68 again, there is enough residual cool air in the chamber to keep dropping and it actually will overshoot by a good bit, down to 65 or even 64.x depending on how active the fermentation is.

I've wondered if these frequent and rapid 4-5 degree temp swings are ideal during active fermentation. I was thinking of comparing internal temp vs. a taped/insulated probe on the outside of the carboy in some upcoming batches just to see what the registration difference was, and my theory is that possibly, the outside probe will still be more influenced by the ambient temp and will possibly cycle the chamber more often, but in shorter bursts and keep the beer closer to the target temp... but first I need to understand the difference between internal and taped to the outside of the carboy.

 

[bkboiler]

I'd say its heavily yeast dependent...
If you've got a conical with a cooling coil and jacket and thermowell, you'll have spent a lot of $$$, but be confident you can set the temp on the controller and know the beer is at that temp--thus you can ferment just as on your recipe from a book or elsewhere...
For me that's too much $$, so i have a metal fermenter and just dangle the sensor nearby...i may duct tape it to the fermenter with some foam over top eventually...it seems to work pretty well, but the beer may be 1-3F off of the sensor readings in some transient conditions (like when high-krausen is just kicking off), but that's only for a short time...so the affect on the final beer flavor is neglible for such small batch sizes (5g), imho...
For a 10bbl fermenter the temp difference at the "core" of the beer could be more substantial if we were only intending to rely on natural convection...

 

[nathanscrivener]

Interesting...
I've been considering experimenting with this myself recently.

I currently use a Big Mouth Bubbler with a thermowell and the temp probe located almost dead center of the beer. The "problem" (if it is one) is as follows:

-Set temp to 68, for example
-Inkbird clicks on my chamber at 69
-By the time the probe in the dead center of the beer registers 68 again, there is enough residual cool air in the chamber to keep dropping and it actually will overshoot by a good bit, down to 65 or even 64.x depending on how active the fermentation is.

I've wondered if these frequent and rapid 4-5 degree temp swings are ideal during active fermentation. I was thinking of comparing internal temp vs. a taped/insulated probe on the outside of the carboy in some upcoming batches just to see what the registration difference was, and my theory is that possibly, the outside probe will still be more influenced by the ambient temp and will possibly cycle the chamber more often, but in shorter bursts and keep the beer closer to the target temp... but first I need to understand the difference between internal and taped to the outside of the carboy.

You’ve identified why thermowells are problematic. If you have a single probe, it is much better to tape to the side with some insulation over it. The probe then is influenced by both the chamber and the beer temp and you don’t get overshoot.

the only way a thermowell is advisable is if you’ve got a system that can prevent overshoot. Brewpi uses two thermoprobes - one in a thermowell and the other measuring ambient - and has sufficient smarts to prevent overshoot.

 

[day_trippr]

Correct - BrewPi eventually characterizes the way the chamber and beer respond to cooling and heating and after awhile will figure out when to turn things on and off to hold a flat line...

Cheers!

 

[TheMadKing]

My OPINION and not backed by any research, is that a temp swing of 3-5 degrees has no tastable effect on your finished beer. I've wondered about that with my cooling coil though - it's a pretty sudden temperature drop experienced by the yeast in proximity to the coil. It makes me wonder if I'm shocking the yeast thats touching the glycol coil when my pump kicks on. It's not as gentle as an ambient temperature change.

I wonder what kind of thermal lag the big boys get in a 1600bbl jacketed fermenter - there has to be some stratification going on, as well as the sheer volume takes a long time to even out the temp unless they have a mixer in the fermenter like Carlsburg has started doing.

 

[DVCNick]

BrewPi: Just had to look that one up... looks expensive.
My other current idea besides trying the taped external probe, is to hook up my second Inkbird to the heater (unfortunately one Inkbird is not configurable enough for this to work). Set the temp to, say, 66 or 67. Would probably take some trial and error, but my idea is to use the heater to cut off the cooling overshoot when it gets to a certain point, but without heating all the way back up to the optimal set temp, or the cooling would just come right back on and it would be running one or the other almost constantly. With some trial and error I'm guessing I could keep it within 2 degrees with a reasonable total duty cycle, rather than the current 4-5 degree range.

 

[Dgallo]

My OPINION and not backed by any research, is that a temp swing of 3-5 degrees has no tastable effect on your finished beer

This maybe true as long as you’re within range but if you’re driving the fermentation, a 3-5 spike higher can have a huge impact. Take LAiii for example. I drive it to 74* and I get great esters. My temp control failed while fermenting a batch not so long ago, ans it rose to 78 for about a day. I got a huge amount of Fusel alcohol.

 

[nathanscrivener]

BrewPi: Just had to look that one up... looks expensive.
My other current idea besides trying the taped external probe, is to hook up my second Inkbird to the heater (unfortunately one Inkbird is not configurable enough for this to work). Set the temp to, say, 66 or 67. Would probably take some trial and error, but my idea is to use the heater to cut off the cooling overshoot when it gets to a certain point, but without heating all the way back up to the optimal set temp, or the cooling would just come right back on and it would be running one or the other almost constantly. With some trial and error I'm guessing I could keep it within 2 degrees with a reasonable total duty cycle, rather than the current 4-5 degree range.

Brewpi has DIY options. I built mine with a raspberry pi and an arduino.
The idea you are suggesting is overkill - tape your probe to the side of your fermenter and put some bubble wrap over it. It will work fine and a lot better than using a thermowell.

 

[Gnomebrewer]

I wonder what kind of thermal lag the big boys get in a 1600bbl jacketed fermenter - there has to be some stratification going on, as well as the sheer volume takes a long time to even out the temp unless they have a mixer in the fermenter like Carlsburg has started doing.

In a big conical, fermenting beer is constantly moving (quite rapidly) in a circular motion - up the centre then down the sides.

 

[Gnomebrewer]

I use two cheap STC1000 controllers. One measures the beer temperature (probe in the ferment). If it's high, it switches on controller two which runs the fridge. Controller two will only run until the fridge temp (ambient) is 3C below my intended ferment temp (a bit more for something aggressive like a hefe yeast) - thus no major undershoots. Reverse of course in winter in a heating chamber.

 

[DVCNick]

Ok, I'll try that on the next brew, or at least, observe the difference, and if I see only one degree, will definitely try it.

I use two cheap STC1000 controllers. One measures the beer temperature (probe in the ferment). If it's high, it switches on controller two which runs the fridge. Controller two will only run until the fridge temp (ambient) is 3C below my intended ferment temp - thus no major undershoots. Reverse of course in winter in a heating chamber.

Well that is definitely another interesting way to utilize two that I hadn't thought of... will have to keep that in mind as another possibility. If the internal temp and taped probe are really only one degree apart though, I'm guessing the taped probe is going to be easiest and likely plenty good enough. I'll be watching to see how they compare on the next brew.

The carboy heater wrap I use is pretty tame and doesn't overshoot much.

 

[Argyll Gargoyle]

Are there any folks equipped with both a thermowell and strapped sensor who are willing to try a step response in temperature - first using the strapped sensor as the control input, then repeating the test using the thermowell?
Control theory says that the strapped sensor will produce less overshoot, all things being equal.

Edit: Could do this test using plain water, by the way

 

[DVCNick]

I'm going to be running both and paying close attention on my next brew, but it is going to be a Belgian Wit finishing probably in the low 4's ABV, so likely not representative of how much internal differential you could possibly get... I still want to see it for myself though.

 

[day_trippr]

Are there any folks equipped with both a thermowell and strapped sensor who are willing to try a step response in temperature - first using the strapped sensor as the control input, then repeating the test using the thermowell?[...]

I believe this plot shows what you want for the strap-controlled scenario.

Two 6.5g carboys holding halves of the same batch, both with strapped probes (red and yellow traces), with Ferm 1 (the red trace) being the control sensor. The thermowell (green trace) is in the center of Ferm 1 as well. All of the probes were dead-nuts correlated over a number of days (I picked these 4 from the 10 under test) so what's there should be "real". Use the color key to reference which channel is which, the "Blk Dot" "Red Plus" etc are physical labels on the probes themselves. Bad planning on my part

Note the thermowell slope vs that of the two strapped fermenters. Darned near a step function, but keep in mind the scale (~.5°F) and time frame (a bit over an hour?) But it does illustrate what one would expect

I have not tried configuring my BrewPi setup to use the thermowell yet. I only did the one run so far, but from this data I'm inclined to stick with the strapped sensors...

Cheers!

 

[Jayjay1976]

A couple of degrees +/- won't mean much to the quality of your finished beers. The point of diminishing returns is in your rear view mirror.

An alternative to a two-probe system to dampen temp swings and prevent overshoot would be to set a thermowell into a smaller volume of liquid that will equilibrate at a reasonable rate but not as slow as the fermenting beer. I use a mini bottled water with the probe insterted through a small hole in the lid. The temp in the water bottle swings a few degrees up and down at a reasonable rate to kick the compressor on and off, while the mass of beer in the fermenter lags behind seeing much milder temp swings.

When I first started temp controlled fermentation I used an inkbird logger and I had the probe just dangling in the fridge. Looking at the resultant data it was clear that the probe needed thermal mass but the thermowell resulted in overshoot. Those first couple of batches tho, it was like chasing chickens with a school bus.

 

[BlutoA10C]

All, quick question on the topic regarding cooling during fermentation (and conditioning). I'm very new to brewing, have a catalyst conical fermenter (and 2 standard bucket fermenters, 7.5gal). I don't have significant space for brewing. My set up is in my basement where the temperature is a VERY consistent 72 degrees F during the summer and 68 degrees F during the winter. Is there an economical way to control fermentation temp, is it necessary? I realize brewing a lager is likely out, but will there be issues with other brews? I've done a Hefe, Oktoberfest, and I'm working on an IPA.

Any advice/thoughts is much appreciated!

Regards,
Dave

 

[Ralphie0523]

All, quick question on the topic regarding cooling during fermentation (and conditioning). I'm very new to brewing, have a catalyst conical fermenter (and 2 standard bucket fermenters, 7.5gal). I don't have significant space for brewing. My set up is in my basement where the temperature is a VERY consistent 72 degrees F during the summer and 68 degrees F during the winter. Is there an economical way to control fermentation temp, is it necessary? I realize brewing a lager is likely out, but will there be issues with other brews? I've done a Hefe, Oktoberfest, and I'm working on an IPA.

Any advice/thoughts is much appreciated!

Regards,
Dave

I used to live in a similar house and here are a few of my learnings:

1. 34/70 yeast is your friend for lagers.
2.Try doing the wet t shirt on the fermenter trick. Basically use a wet towel or t shirt around the fermenter to let evaporation cool things a bit.
3.Select yeasts that can tolerate the slightly higher temps like WLP008.
4. In terms of necessary it is up to you. If your beer is satisfactory then don’t worry about it. If you feel this is the next major step then try it.

Best of luck.

 

[BlutoA10C]

Sorry, maybe a stupid question - 34/70 yeast? Do you have any suggestions for a home built cooler system? Not sure I'm interested in spending $1000 for a cooling system or dismantling an ac system to create a "home built" system - any suggestions appreciated! Thanks!

 

[SanPancho]

a mixer in the fermenter like Carlsburg has started doing.

link please.

 

[day_trippr]

fwiw, two of my "fermentation chambers" are 17cf top-freezer fridges obtained via Craig's List for a total of $225.
While I went a different way via BrewPi over Bluetooth for controllers, one could use a $35 Inkbird ITC-308 and do the same thing.
A lot less than $1000...

Cheers!

ps: 3rd chamber cost me $50 grand. There was a new kitchen/dining room involved

 

[SanPancho]

Sorry, maybe a stupid question - 34/70 yeast? Do you have any suggestions for a home built cooler system? Not sure I'm interested in spending $1000 for a cooling system or dismantling an ac system to create a "home built" system - any suggestions appreciated! Thanks!

mini fridge cooler man. easiest there is. probably score one for free off craigslist if you're patientt. thats simple cheap and effective. (assuming your fermenter fits inside)

 

[Ralphie0523]

Sorry, maybe a stupid question - 34/70 yeast? Do you have any suggestions for a home built cooler system? Not sure I'm interested in spending $1000 for a cooling system or dismantling an ac system to create a "home built" system - any suggestions appreciated! Thanks!

34/70 is the Fermentis offering of a lager yeast that many other yeast houses have. Long story short, it is a lager yeast that seems to tolerate ale temperatures well.

I personally use a CoolStix cooling rod in my fermenter now, thanks to my wonderful wife. I have some water in a cooler that I chill down with two two liter bottles of ice I change out every 12 hours. There are more thrifty options that are similar. To me it is more of a middle road between a dedicated fridge that generally can cost a bit money and space and a wet cloth that is essentially free. I never got good results with the wet cloth option but others have reported good results.

Keep in mind there is always a do nothing option. If you are happy and don’t think you need the cooling then don’t worry about it. I sense you are hitting the question of what is the best improvement you can make given your newer brewer status.

Temp control matters in my opinion but it is more about being in a good range than having very tight control around a specific value. If there is something more pressing in your mind to improve I would say prioritize those items and revisit temp control another time. Just my opinion however, you know your situation better than I ever could.

 

[DVCNick]

So that "Belgian White" type beer I mentioned above quickly blew off and with my setup that means I have to remove the thermowell to install a blowoff tube... didn't get much data.

Situation right now: I have a normal strength ale with US-05, about 25 hours post-pitch of a dry pack right. Most of my recipes are similar enough to this that I would expect them to behave almost exactly the way this one does. This may or may not blow off, but I should have some more time even if it does. There is about an inch of krausen now and it is visibly active.

I have two Inkbirds, one is in a thermowell and one is taped to the side under an ordinary kitchen sponge for insulation. These normally read .5-1 degree apart when I'm trying to calibrate them against one another.

I think the chamber has clicked on a couple times and right now they are registering <1 degree apart:

 

[DVCNick]

The external probe is actually reading half a degree hotter right now.
Edit: forgot to mention, temp is being controlled by the external probe right now.

 

[eric19312]

I looked at Brewpi and really liked the dual temperature probe strategy but ended up going with this unit from Auber which also has dual temp probes. The Auber is not PID controlled and so won't learn how to respond and won't hold temps within 0.1F but it was a less expensive turnkey system and works great at holding temps within 1 degree F.

https://www.auberins.com/index.php?main_page=product_info&cPath=8&products_id=671
I have one probe in the thermowell, other probe is taped to inside of my fermentation freezer about half way up. I have an always on fan running inside the freezer pulling air from floor, past a ceramic bulb heater.

I will add I got into temp control issues when I started brewing in a 15 gallon spiedel tank. I just could not get same control I was getting with 6 gallon glass and plastic carboys with the extra volume and thicker/more insulating fermentor walls. I've since moved on to a stainless conical and it seem easier to control than the plastic tank ever was I think due to better heat exchange property of the steel.

As for original question how much fluctuation is too much...I measured temperature fluctuations of about 4-5 degrees .... going 1-2 degrees above set point and then overshooting to about 3 degrees below setpoint when I measured it with a data logger. I got the data logger because I felt I was tasting temp control issues in my beer and these did go away when I got temp back under control so I'd say fluctuations of 4 degrees were noticeable to me when brewing IPA style beers with Chico yeast.

 

[DVCNick]

Spot checking today, this is going very well. Internal temp has been 68.2-68.4 every time I've looked, and the chamber is being run significantly.

 

[bkboiler]

@Dgallo touched on one of the most important methods used in historical brewing technique.
If you place your beer in a much larger enclosure with a large thermal mass (an isotherm) then the beer just dumps heat to that...
target the lower end of the range for the start of fermentation and take notes on how that yeast did in your enclosure...