What model for IBU calculations does brewersfriend recipe calculator use?

Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum

Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
P

prankster1590

Well-Known Member
Joined
Sep 24, 2019
Messages
58
Reaction score
11
EDIT: WATERBILL IS WRONG. IGNORE PLEASE

When I calculate my hop additions I use Tinseth. When i add these number to the recipe calculator of brewers friend I always get a lower IBU than what I have calculated by hand.

For example. I wanna brew 20 liters of a lager kind of beer with 23 IBU's. The SG at 20 liters should be 1,0477 or about 124 g/L sugar. I'm gonna use Styrian gold (AA%=3%) as bitter for 60 minutes, East Kent gold (AA%=5%) as flavor for 30 minutes and Hallertau tradition (AA%=4,3%) for 5 minutes.

At the start of the boil I have 24,641 L water and at the end of the boil I have 22,916 L water (7% / hour evaporation rate).

For 60 minutes boiling I have a SG of 1,0337 and a U% of 26,7%
For 30 minutes boiling I have a SG of 1,0401 and a U% of 19,4%
For 5 minutes boiling I have a SG of 1,0414 and a U% of 5%

EDIT: THIS IS ALL WRONG. IGNORE PLEASE.

Tinseth equation I use.
U%= (1,65*0,000125^(sg-1))*((1-e^(-0,04*t(min))/4,15)

The weigh of the hops I have to add is calculated as follows with L being the intended end volume (20L):

W(grams) = (L*IBU)/(AA%*U%*10)

At 60 minutes before end of boiling I add 15,05 IBU of Styrian gold:

w =(20*15,05 )/(0,03*267) = 37,6 grams

At 30 minutes I add 6,37 IBU of East Kent gold

W = (20*6,37)/(0,05*194) = 13,13 grams

At 5 minutes I add 1,612 IBU of Hallertau tradition

W = (20*1,612)/(0,043*50) = 15 grams

15,05+6,37+1,612 = 23 IBU

And this is wat the recipe calculator give me. Only 21,5 IBU's. Why? Should be the same calculation.

2021-08-26 15_41_50-Mail.png
 
Last edited:
BrewnWKopperKat said:
In the Brewers Friend screenshot, the hop entries are for "Leaf/Whole". The "util" numbers appear to be different than the ones you are using.
Click to expand...

Tinseth is emperically derived from experiments with hop flowers. The 4,15 in the denominator is dependant on the brewsystem you use which might have a higher or lower utilization rate than Tinseth's brewing system.

But this is a computer calculator and if using the Tinseth equation i would assume that it uses thes tandard number of 4,15. I did type in my own utilization numbers but that gave some weird results. So my suspicion is that they use another model than Tinseth.

Tinseth equation I use.
U%= (1,65*0,000125^(sg-1))*((1-e^(-0,04*t(min))/4,15)
Click to expand...
gsdg.png
 
Last edited:
BrewnWKopperKat said:
Back in the original post, it looks like you estimated 23 IBUs and Brewers Friend estimated 21.5 IBUs.

eta: IBU Calculator Beer Bitterness - Brewer's Friend mentions Tenseth. It seems plausible that the same calculator is used in the 'complete recipe builder'.
Click to expand...

Good one. I notice that they just use the target SG to estimate a boil SG. But yet again different numbers. That is a big difference. 23 and 18.7. A difference between a fine elegant pilsener and a cheap budget lager.

The utilization factors seems to be corresponding to the tinseth numbers for the estimated boil SG. My post boil volume is 22.9 liters and might be used as end volume by the calculator. But 20 liters is the end volume (minus shrinking with cooling down and trub). So it could be that 20 liters is 23 IBU and 22.9 liters is less than 23 IBU.

fbfbfhdgjfgj.png
 
Last edited:
What are your settings in your profile in Brewer's Friend? Mine is Tinseth, but others choose Rager:
1630005560949.png


And of course, there is a difference in pellet and whole/leaf hops as mentioned.
And the boil SG is estimated, but maybe a different estimation, depending on the exact volume.
 
Yooper said:
What are your settings in your profile in Brewer's Friend? Mine is Tinseth, but others choose Rager:
View attachment 740315

And of course, there is a difference in pellet and whole/leaf hops as mentioned.
And the boil SG is estimated, but maybe a different estimation, depending on the exact volume.
Click to expand...

Im using Tinseth. Which is based on hopbells. Weigh of pellets is about 90% of the weigh of Hopbells for the same effect. I think that the estimated SG by brewersfriend during the boil is responsible for the difference. I calculated for every addition the SG at that moment. And maybe brewersfriend is confused about my endvolume. Or the volume that goes into the fermenter.
 
prankster1590 said:
Im using Tinseth. Which is based on hopbells. Weigh of pellets is about 90% of the weigh of Hopbells for the same effect. I think that the estimated SG by brewersfriend during the boil is responsible for the difference. I calculated for every addition the SG at that moment. And maybe brewersfriend is confused about my endvolume. Or the volume that goes into the fermenter.
Click to expand...

Yes, the SG during the boil changes as the volume decreases, so that's probably it.
 
As it stands, the Tinseth method is:
1) Only applicable to leaf, and not applicable to pellets (by Tinseth's own admission during a recent podcast).
2) Only likely to predict your actual IBU's for leaf or plug type hops to within about 1 standard deviation of the mean, or within + or - 34% of what for you will most likely be 'your' actual instrument measured IBU fact.
3) Expanding upon #1, not likely capable of being nearly +/- 34% accurate for pellets.
4) The belief that for pellets one only needs to multiply leaf by 1.1 is a pure fantasy not originated by Tinseth.

So argument as to whether or not ones particular means to the application of Tinseth's highly empirical math model leads to a result of 23 IBU's or 21 IBU's is seemingly a rather moot point.
 
Last edited:
But then 'Prankster' is an interesting choice for ones forum name. Coincidence, or correlation (within 1 standard deviation)?;)
 
Basic Brewing Radio. November 1, 2018 - IBUs vs Wort Gravity and Hop Stand Temps

IIRC, the data was based on pellet hops. It wasn't much work to compare the data to the Tinseth formula.

As an aside, the data also seems to confirm an upper limit to the number of IBUs in a volume of wort. This limit would be good to know when restructuring recipes for 'stove top' brewing (half the wort at start of boil, rest at end).
 
When a PhD Chemist and Professor named Christopher S. Hamilton, Ph.D, at Hillsdale College gave (by honored invite) a presentation on "The Effect of Temperature and Alpha-Acid Concentration on Hop Utilization in Wort" at the highly prestigious "EBC Hop Symposium" in Nuremberg, Germany, September 2018 he showed that there is clearly no valid "utilization related" correlation to Tinseth for pellets. He and his graduate students compared them over many months and many varying SG and grist and hops scenarios within a highly controlled lab setting whereby to collect the data presented at this symposium.
 
Silver_Is_Money said:
When a PhD Chemist and Professor named Christopher S. Hamilton, Ph.D, at Hillsdale College gave (by honored invite) a presentation on "The Effect of Temperature and Alpha-Acid Concentration on Hop Utilization in Wort" at the highly prestigious "EBC Hop Symposium" in Nuremberg, Germany, September 2018 he showed that there is clearly no valid "utilization related" correlation to Tinseth for pellets. He and his graduate students compared them over many months and many varying SG and grist and hops scenarios within a highly controlled lab setting whereby to collect the data presented at this symposium.
Click to expand...

Hopefully with all of this research they have or are putting out a new set of equations for us to calculate with then?

Otherwise in my opinion the equations provide a consistency to at least be able to repeat recipes that we brew.
 
To fully release all of the IBU's within leaf type hops takes on the order of 90 (or perhaps more?) minutes of boiling. One of the things Christopher S. Hamilton unveiled is that pellet hops do this in 40 minutes or less. And for shorter boil durations the speed at which pellets liberate their IBU's is even more dramatic vs. leaf types. There is no simplistic linearity multiplier that can be applied to leaf (and thereby an empirical math model built upon the utilization of leaf types) to convert the output into one suitable for pellets.
 
Last edited:
BrewnWKopperKat said:
@Silver_Is_Money : can you provide a link to that research? If it's behind a pay wall, can you excerpt the information for the claims you are repeating?
Click to expand...

The research (and the brief PPT outline for it) were once in the public domain. But they appear to have been removed from the public domain. I have copies, but if they are no longer to be considered public, then I'm not within my rights to provide a private link to my copies. What I therefore suggest is that you contact Professor Christopher S. Hamilton directly. You should be able to do this by contacting the administration at Hillsdale College (provided that he is still a Professor there).
 
An additional consideration as to IBU's is that hop utilization seems to correlate to Wort pH. If this holds true, then if you enter the boil at a higher pH you will achieve greater utilization, and thereby higher IBU's for the $buck, but at the potential cost of greater hop harshness. The opposite of this would be to enter the boil at a lower pH and suffer from poor utilization, but with potentially smoother hop bitterness.
 
hout17 said:
Hopefully with all of this research they have or are putting out a new set of equations for us to calculate with then?
Click to expand...

https://live.staticflickr.com/7891/45991029004_df99d89bc1_o.png ?

eta: HBUs FTW !! ;)

and ...
hout17 said:
Otherwise in my opinion the equations provide a consistency to at least be able to repeat recipes that we brew.
Click to expand...
and a possible problem with "better" estimating of IBUs is that one may still need the original estimating approach to confirm recipes are 'reasonable'. Or maybe not.

 
Last edited:
All of this adds to (and compounds/expands upon) the meaninglessness of one software predicting 21 IBU's vs another predicting 23 IBU's.
 
BrewnWKopperKat said:
https://live.staticflickr.com/7891/45991029004_df99d89bc1_o.png ?

eta: HBUs FTW !! ;)

and ...

and a possible problem with "better" estimating of IBUs is that one may still need the original estimating approach to confirm recipes are 'reasonable'. Or maybe not.
Click to expand...

HBU = AAU=AA% * ounces and yeah that is a great recipe feature for consistency lol I'm talking more about new equations related to the findings discounting Tinseth above surely they can come up with a more accurate utilization curve based on their research.

That pic looks like derived shortcuts to make it easier to calculate as it appears they are still comparing their calculation to Tinseth for comparison and using their magic factors.

I've already got Tinseth dialed in right or wrong for elevation and the magic factor for pellets hahah for my estimates. This is really a futile discussion lol and a rabbit hole we shouldn't travel down.

Also yeah these are truly only estimates I highly doubt anyone would be able to tell a 5 ibu difference especially when we are dealing with estimates.

But hey this is what makes each brew unique and makes brewing beer fun.

Edit: so I guess it would be good practice if when posting recipes people would define whether they used Rager or Tinseth and any magic factors lol. This really makes vintage recipes a crap shoot might need Marty and Doc to go back and gather some info for us.
 
Last edited:
I apply magical factors (OK, empirical guesswork) within MME (at least amidst the very latest of my releases) for exclusively pellets whereby to best extract what admittedly little (and likely misunderstood as well) from among what I've gleaned of the Christopher S. Hamilton data and apply some small parts of it it to at least some admittedly loose/ballpark and empirical (OK, guessing) extent. In all honesty, and in full disclosure, Christopher S. Hamilton would highly likely be abjectly appalled at my stab at interpretation and implementation of his work, (due to my admitted misinterpretation and thereby misimplementation) if truth be told. But until he completes his study and releases it, I'm at least taking a shot at it.
 
Last edited:
Silver_Is_Money said:
I apply magical factors (OK, empirical guesswork) within MME (at least amidst the very latest of my releases) for exclusively pellets whereby to best extract what admittedly little (and likely misunderstood as well) from among what I've gleaned of the Christopher S. Hamilton data and apply it to at least some extent. In all honesty, and in full disclosure, Christopher S. Hamilton would highly likely be abjectly appalled at my stab at interpretation and implementation of his work, (due to my admitted misinterpretation and thereby misimplementation) if truth be told. But until he completes his study and releases it, I'm at least taking a shot at it.
Click to expand...

MME is great and taking a stab at it is all we can ask for with current calculation methods and applying new info as it comes in. I would love to see Hamilton's study at some point once it's completed; I'm sure it has some eye opening data.
 
Silver_Is_Money said:
I was pointed to it by others on this very forum.
Click to expand...
Would you be willing to review the BBR podcast and materials to confirm that it is (or is not) similar to the presentation you have access to?

If the BBR content is "similar enough", then everyone here has the opportunity to review the material and we can have a discussion about the content.
 
BrewnWKopperKat said:
Would you be willing to review the BBR podcast and materials to confirm that it is (or is not) similar to the presentation you have access to?

If the BBR content is "similar enough", then everyone here has the opportunity to review the material and we can have a discussion about the content.
Click to expand...

Sure, just give me the link.
 
Every time I think about educated guessing within MME (or within software in general) it hearkens me back to this scene from the movie "Star Trek IV : The Voyage Home":

James T. Kirk: Mr. Spock, have you accounted for the variable mass of whales and water in your time re-entry program?
Spock: Mr. Scott cannot give me exact figures, Admiral, so... I will make a guess.
James T. Kirk: A guess? You, Spock? That's extraordinary.
Spock: [to McCoy] I don't think he understands.
Leonard McCoy: No, Spock. He means that he feels safer about your guesses than most other people's facts.
Spock: Then you're saying... it is a compliment?
Leonard McCoy: It is.
Spock: Ah. Then I will try to make the best guess I can.
Click to expand...

Two of the most critical educated guesses for grist components involve assigned BC's in conjunction with assigned pHDI's. Factual data for these two valuations (as a pair within a triumvirate, which is the only way they can possibly have validity, as such is the case that one value not expressed in conjunction with the other, as well as in conjunction with the pH targeted during the titration assessment used to generate the BC) is slim and questionable as to validity to begin with, so that leads straight to guessing. The three "inseparable" components of the triumvirate being:
1) pHDI
2) BC
3) The pH targeted whereby to assess the derived BC

We have scant and insufficient data for #1 and #2, but little to no truly reliable data as to #3. Those few who have left us with data as to #1 and #2 almost never seemed to give thought as to documenting #3 for us. But #2 is factually inseparable from #3. And that means guessing as to the method, the methodology, the integrity, and the intent of the data generator. Who may not himself/herself ever have given thought to #3. It all comes down to BC being a pH dependent variable that for decades has been mistakenly presumed by most to be a constant. One can only guess that BC data is always reliably assessed in strict association with a target of pH 5.40, but if it was derived instead with respect to a fixed mEq of acid/base quantitity of titrant, and pH was allowed to float whilly nilly thereby during the fixed mEq titration, then any BC data thus derived is (in my book at least) useless. And I suspect that much of the BC data was derived whilly nilly.
 
Last edited:
BrewnWKopperKat said:
But it doesn't have to be futile (or a rabbit hole). Share articles that others can read. Read the articles that others share. Discuss what was read. Share results of a test batch (or two).
Click to expand...

A problem (with sharing the results of a test batch) is that most people are not going to go to the expense of IBU testing. But some commercial breweries do. One is those breweries is co-owned by an acquaintance of mine, and they sent samples of several beers for testing. Their results agreed fairly closely with Tinseth. Close enough that they were able to apply a constant factor to get very close. I suspect (would bet a paycheck) the factors needed would be different for every brewery, which is why IBU prediction software should allow for it. I should note that this was for fairly low IBU beers, and didn't include the IBU range where Tinseth (and Rager and Garetz) fall apart (above about 65 Tinseth IBUs).
 
Last edited:
Last edited:

Similar threads

D
First Recipe: 1 Gallon Maris Otter / Goldings SMaSH (Comments, suggestions?)
Replies
6
Views
364
Northern_Brewer
Northern_Brewer
M
Feedback/Critique for my "Kitchen Sink" IPA - Empty my fridge
Replies
14
Views
833
Matheos
M
Zippy123
Feedback on recipe to split mash for two different beers
Replies
2
Views
426
Zippy123
Zippy123
T
HELP: misread the recipe, crazy high IBU?
Replies
21
Views
1K
Grizwold1
Grizwold1
Gadjobrinus
3 potential new - Best, Strong Bitter, Imperial Stout
Replies
14
Views
1K
Gadjobrinus
Gadjobrinus
BrewSwag 300

Latest posts

Back
Top