Brewer's Friend - IBU Accuracy?

[specialkayme]

I'm finding that Brewer's Friend Recipe Calculator's (https://www.brewersfriend.com/homebrew/recipe/calculator) IBU to be fairly off. Am I the only one?

I tried to make a Make America Amber Again (http://brulosophy.com/2016/04/11/sparge-temperature-pt-1-standard-vs-cool-exbeeriment-results/), but I"m not a fan of very hoppy beers, so I turned the IBU down using the recipe calculator from 78 IBU that the recipe called for to 50. The beer came out very malty, almost non-existent on hop bitterness. Definitely not a 50.

I just ran the MACC IPA through the calculator (http://brulosophy.com/recipes/macc-ipa-by-ray-found/) and BF is telling me the IBU is a whopping 135 (and none of that is coming from any of the dry hops), rather than the recipe's 75.

What am I missing?

 

[Smellyglove]

Take IBUs from software as big ballparks. Just brew a few beers, and use those numbers as abitrary numbers, and adjust. What the Actual IBU is can be way off, you'd need to measure your beer to find it out. But it's not only IBU's that give the perception of bitterness.

 

[specialkayme]

So I guess the question then is, do you think Brewer's Friend puts you in the right rough ball park on IBUs?

I was under the impression that with most of the recipes you find online, the creator didn't actually send the beer out to a lab to get tested. Instead, they brewed a tasty beverage that they liked, threw the ingredients into a calculator and it spit out what the IBUs were (compared to most commercial beers that you buy that have IBU on the label that did have theirs lab tested). Now if their calculator said 75, I need to get my calculator to say roughly 75 to start with in order for me to adjust my hop additions based on the differences in acid content of the hops I"m using.

Some deviation is fine, as long as it gets me into the right rough ballpark. Brewer's Friend estimates the IBU to be 136. ezRecipe estimates it to be 97. That seems like a very wide deviation between software calculations (one [unknown] at 75, BF at 181% more, ez at 129% more). Which one is putting me closer to the ballpark I'm looking for?

I've tasted some beers with IBUs in the 90's that tasted well balanced and pleasing, while I've tasted some beers with IBUs in the 60's that taste like I'm sucking on a hop pellet. So I understand it's all relative, and there are a number of variables that may make something taste more or less bitter than the IBUs that the lab reveals. I don't care if its 75 or 150, as long as it matches the recipe's intent and the style I"m trying to brew.

 

[Smellyglove]

So I guess the question then is, do you think Brewer's Friend puts you in the right rough ball park on IBUs?

I was under the impression that with most of the recipes you find online, the creator didn't actually send the beer out to a lab to get tested. Instead, they brewed a tasty beverage that they liked, threw the ingredients into a calculator and it spit out what the IBUs were (compared to most commercial beers that you buy that have IBU on the label that did have theirs lab tested). Now if their calculator said 75, I need to get my calculator to say roughly 75 to start with in order for me to adjust my hop additions based on the differences in acid content of the hops I"m using.

Some deviation is fine, as long as it gets me into the right rough ballpark. Brewer's Friend estimates the IBU to be 136. ezRecipe estimates it to be 97. That seems like a very wide deviation between software calculations (one [unknown] at 75, BF at 181% more, ez at 129% more). Which one is putting me closer to the ballpark I'm looking for?

I've tasted some beers with IBUs in the 90's that tasted well balanced and pleasing, while I've tasted some beers with IBUs in the 60's that taste like I'm sucking on a hop pellet. So I understand it's all relative, and there are a number of variables that may make something taste more or less bitter than the IBUs that the lab reveals. I don't care if its 75 or 150, as long as it matches the recipe's intent and the style I"m trying to brew.

I have never used Brewersfriend and I don't have any references to what "35" (or such) IBU tastes like other than what's been printed on labels, and the perceived bitterness varies even between two 35 IBU beers. My only reference is targeting for instace 35 IBU in Beersmith (which have a lots of different variables to adjust), and just take it from there the next batch. The more variables you are exposed to by a software the more you understand that they are just ballparks. I'd rather just follow a recipe when it comes to additions, than looking at the numbers it spits out for a first try.

 

[specialkayme]

I'd rather just follow a recipe when it comes to additions, than looking at the numbers it spits out for a first try.

So if you have a recipe that says "add 25 IBU of Magnum @12.5%AA FWH" how much would you add, and where would you go to calculate that?

 

[Smellyglove]

I'd just add 25 IBUs worth of magnum at FWH with resonable settings in the software, and take it from there the next brew. If I have no reference of what the original beer actually tastes like, I would just tailor it to my own liking in my own setup for the second brew.

If i know what the original beer tastes like, i'd adjust mye recipe towards that. The point is if you don't know what a beer tastes like and you don't know your setup "well enough", then you are shooting in the blind. But, if you know what the original beer tastes like and you know your setup, you can get closer.

Forget about cloning a beer on your first attempt. If you come close to it it's mostly pure luck. Try what the recipe gives, it will give you beer, then use that information to close in on the profile over several future attempts.

 

[Oginme]

There are several models for calculating IBU contribution from boil hops: Tinseth, Rager, Garetz, Daniels, and so forth. They are all based upon someone's IBU testing on wort made in their system using their methods. To paraphrase Glenn Tinseth on a podcast: my model is pretty accurate for the equipment I was brewing on and the process I was using.

That said, if a recipe is giving you an IBU number without reference to the model used to derive that number or the indication that it is a tested result. To get a good idea of the type of results one can expect, take a listen to the Experimental Brewing podcast and write up from their Igor's results at https://www.experimentalbrew.com/experiments/writeups/ibu-lie-kind.

So as a homebrewer, what do you do? I brew according to my perception of bitterness in comparison to commercial brews. I know from my system that I need to target about 15% higher in calculated IBU for most beers over around 30 IBU (Tinseth) to get what I perceive is the bitterness response I am looking to achieve.

It takes time, good notes, and patience to hone your numbers to meet your expectations. Once you do, you can pretty much transpose any recipe requirements to your system and come pretty close to what you wanted.

And when I see a recipe that asks for "add 25 IBU of Magnum @12.5%AA FWH", I figure the amount of hops based upon the %AA of the Magnum I have on hand and (depending upon the beer type) target about a 15% increase from that value.

 

[specialkayme]

And when I see a recipe that asks for "add 25 IBU of Magnum @12.5%AA FWH", I figure the amount of hops based upon the %AA of the Magnum I have on hand

How do you figure the amount of hops? What formula, system, or calculator would you use?

 

[Oginme]

I use BeerSmith and the Tinseth formula. The software really does not impact the basic calculations of the model used. Understand that all the models only represent the contribution from boil hop additions. Any estimation from whirlpool hopping or first wort hop additions are strictly modifications on any of the formulas and I am not sure the different software model the same studies to derive their estimations.

In the end, it does not matter which model you use. Pick one and stick to using that model and ignore the others -- they will only add to confusion. From my perspective as a ChE, the Tinseth model more properly follows a typical batch process reaction curve than the others.

I should also add that I do not usually brew printed recipes, but design my own. I also approach a new recipe (mine or others) as the first brew will be the test batch and subsequent brews will be the ones that better match my expectations.

 

[dmtaylor]

Brewer's Friend looks to be pretty far off. Try Tinseth method in other software. If the software does not allow you to specify Tinseth..... then it is not very good software that cares about being accurate.

 

[myndflyte]

Well I see a few other problems with that recipe. First of all, the recipe lists out the number of IBUs that you should be targeting from each hop addition, yet if you add those IBUs up, it comes out to 94 IBUs, not 75. So that's issue number one there.

2nd, the recipe just lists the hops, not the % alpha acid so what Brewers Friend puts in there as stock, may not be right. For example, Citra can range from 10-15% (http://www.hopslist.com/hops/dual-purpose-hops/citra/). That's why you have to check the alpha acid % of the hops you have and use accordingly to reach the number of IBUs the recipe is suggesting for each addition...which again, doesn't equal 75. Unless I'm totally missing something and the IBUs don't just add up like that.

 

[Smellyglove]

Brewersfriend lets you customize the AA.

 

[specialkayme]

the recipe lists out the number of IBUs that you should be targeting from each hop addition, yet if you add those IBUs up, it comes out to 94 IBUs, not 75.

Well I'll be darned. I didn't notice that. Good eye.

It also makes things seem a little more in line with ezRecipe's estimation of 97 IBU.

the recipe just lists the hops, not the % alpha acid

I believe that's part of the recipe in the right hand column.

 

[specialkayme]

I use BeerSmith and the Tinseth formula. The software really does not impact the basic calculations of the model used.

Brewer's Friend appears to use the Tinseth formula as well (although you can switch to the Rager formula if you want). But then again other software systems appear to use Tinseth as well, but differ from BF's calculation, which was the spark of this thread.

BF allows you to customize and change the Scale Util of the hop additions. If I cut them down to almost half of what the pre-programmed amounts are set at, I can get the IBU calculations closer to what other software programs are using. Of course, I'm just completely guessing at cutting it in half at this point. It's nice to be able to customize that information, but I don't know enough to be able to put a custom number that works in place.

I know a bunch of people that use BeerSmith. Most I know were using BeerSmith with BrewnWater. I'm not a fan of paying for something I can get for free, which is why over the past two years or so I've been using BF. But if it's that inaccurate, it may be worthwhile to pay for the service.

 

[specialkayme]

Brewer's Friend looks to be pretty far off.

Thank you for taking the time to compare and comment. I appreciate it.

 

[myndflyte]

I believe that's part of the recipe in the right hand column.

Wow, I never saw that, lol. I was so focused on the left hand side.

I still think the recipe is off. Just plugging in that 20 min hop addition in the boil, various calculators are giving me 13 or 14 IBUs, not the 10 that's listed. So I think 75 IBUs is way off, if the listed numbers are right, it's probably in the 90's but even then it may be higher the listed IBUs are low.

 

[Oginme]

Brewer's Friend appears to use the Tinseth formula as well (although you can switch to the Rager formula if you want). But then again other software systems appear to use Tinseth as well, but differ from BF's calculation, which was the spark of this thread.

BF allows you to customize and change the Scale Util of the hop additions. If I cut them down to almost half of what the pre-programmed amounts are set at, I can get the IBU calculations closer to what other software programs are using. Of course, I'm just completely guessing at cutting it in half at this point. It's nice to be able to customize that information, but I don't know enough to be able to put a custom number that works in place.

I know a bunch of people that use BeerSmith. Most I know were using BeerSmith with BrewnWater. I'm not a fan of paying for something I can get for free, which is why over the past two years or so I've been using BF. But if it's that inaccurate, it may be worthwhile to pay for the service.

The difference may be more in the process loss assumptions and utilization factors that affect the ending result of the IBU models. I started using BeerSmith 6 years ago and to me there wasn't much of a comparison to BF back then. The people doing the development of BF have done a great job at improving it over the years, often times exceeding BS in capabilities before BS upgrades to regain the ground.

My preference for BS is that I do everything on my laptop and often have time during work travel to play with recipes, update old recipes, make changes, etc. and the locally stored program and information suits that pattern of usage. Some people like a purely web/cloud based application and more power to them.

The cost of the software is not a barrier to me. It has been well worth the cost of a single recipe. If it were a few hundred dollars, it would be a different story.

 

[specialkayme]

It has been well worth the cost of a single recipe. If it were a few hundred dollars, it would be a different story.

Not to split hairs, but it is a few hundred dollars.

$14.99/yr. I'm 33, been brewing (off and on) for (*cough*) 12 years. I don't see any reason why I'd stop as I get older. Assuming I continue brewing till I'm 80, that makes for 59 years of software use. At $14.99/yr, that's $884.

Not that BS or BF will be around that long, but you get the idea. Of course you can buy the basic software rights at $35. But then you don't get the upgrades (which the past few have proven to be substantial improvements from what I've read), and can only activate 2 computers (I've gone through 6 computers in the past 15 years).

Not that it doesn't provide value. But thinking it's only one recipe's cost isn't an accurate assessment.

 

[Oginme]

One time purchase was $28 when released recently. Updates for as long as that basic version is available. BS2 was updated regularly since 2011. Odds are this version is good for at least 4 to 5 years which makes it around $7/year for 4 years of use. Chances are that it will be longer than that.

 

[dmtaylor]

I'm still using StrangeBrew 1.8 from like 2002. I think I paid a one-time fee of $10. I update styles and ingredients manually as new ones come out. But the formulae like Tinseth don't change, they stay same as always.

Furthermore.... if you want a shortcut close approximation of Tinseth anytime anywhere, here's my Taylor formula/method, which I'm sure you'll find comes very close within about 3-5 IBUs of any *good* software:

Assuming pellet hops in 5 gallons final boil volume, ~60-minute boil, @ approximately 1.060 OG...

3.6 * oz * AA% = IBUs from bittering hops added @ about 60 minutes left in the boil
1.6 * oz * AA% = IBUs from flavor additions @ 10-15 minutes
0.6 * oz * AA% = IBUs from aroma @ 5 minutes

Treat any "whirlpool" additions as if they'd been boiled for half the time. So for example, a 20-minute whirlpool uses the 10-minute factor of 1.6 above, etc.

Add the above amounts all together, and then add another +1.5 to the final total to get the final grand total estimated IBUs.

For higher gravity worts (e.g., >1.075), the above factors are reduced somewhat from 3.6, 1.6, and 0.6 to about 3.0, 1.3, and 0.6 (last one stays same). For low gravity worts (e.g., <1.045), the factors are increased to around 4.0, 1.9, and 0.6 (this third factor always stays same). So yeah, you might need to use a little intuitive interpolation in some cases.

For different batch volumes (V) other than 5 gallons, multiply the result by 5/V.

If you use whole hop cones, then multiply your final result by 0.9 (a.k.a., 90%).

FYI -- Tinseth (upon which the above stuff is based to emulate) did originally base his formula on whole hops, AND on NOT immediately chilling the wort after the boil but rather a bit more slowly than most of us do today. Fortunately for us in the 21st century, these two differences tend to balance each other out, since pellets make the beer more bitter than whole hops would, but faster chilling reduces the bitterness compared to a longer chill time.

This works great for me. I've been calculating IBUs on napkins for several years based on this, and always come within a handful of IBUs compared to software that uses the Tinseth method. Maybe you'll find it useful. It's just a matter of memorizing the 3.6, 1.6, and 0.6, really.

 

[specialkayme]

I've been calculating IBUs on napkins for several years based on this, and always come within a handful of IBUs compared to software that uses the Tinseth method. Maybe you'll find it useful. It's just a matter of memorizing the 3.6, 1.6, and 0.6, really.

Wow. That's amazing information to share. Thank you so much.

On a side note/question, wouldn't choosing to do a whirlpool addition increase the bittering effectiveness of your 60m, 15m and/or 5m hop additions above the 3.6. 1.6 and 0.6 numbers?

 

[dmtaylor]

Wow. That's amazing information to share. Thank you so much.

On a side note/question, wouldn't choosing to do a whirlpool addition increase the buttering effectiveness of your 60m, 15m and/or 5m hop additions above the 3.6. 1.6 and 0.6 numbers?

Great question. Actually yes, you are correct. So maybe increase the factors to 3.7, 1.8, and 1.0, something like that. Bet that comes out real close then.

 

[bpgreen]

Another thing to consider is that IBUs don't tell the whole story. I used to have a chart that showed the BU:GU ratios and showed bands as malty, balanced, or bitter based on IBUs and OG, since the sweetness of more malt can mask more bitterness.

Butt even that doesn't tell the while story since a variety of factors can change the FG, which chances the residual sweetness from the malt.

 

[dmtaylor]

Another thing to consider is that IBUs don't tell the whole story. I used to have a chart that showed the BU:GU ratios and showed bands as malty, balanced, or bitter based on IBUs and OG, since the sweetness of more malt can mask more bitterness.

Butt even that doesn't tell the while story since a variety of factors can change the FG, which chances the residual sweetness from the malt.

Great points.

 

[pricelessbrewing]

This calculator at alchemy overlord is probably the best implementation of a modification of tinseth based on what we *know* about how whirlpooling, flameout additons, and oil saturation affects ibu.

https://jphosom.github.io/alchemyoverlord/

 

[Yooper]

Keep in mind that those IBU calculators all take into account boil volume, SG of the boil, etc so even if they are using the same Tinseth calculator, differences in the equipment profile, the size of the boil, the efficiency settings, etc can all make a difference in the estimation of the IBUs.

 

[ScrewyBrewer]

Not to mention the Hop Storage Index (HSI) factors effect on IBUs. Since nada single homebrewer I know has a way to accurately measure IBUs we are all left guessing on which IBU prediction is the most accurate.