Hop Utilisation Comparison

[buttersd70]

Over the course of time, there has been an occasional murmur on this site in relation to the way Beersmith estimates IBU. Some people have noticed that the numbers don't match what is calculated by the recipe database, and others have also noticed that the numbers don't correspond to what is calculated by Promash. This anomaly has also been raised in the Beersmith support forums several times. However, it tends to be dismissed with a somewhat casual 'well, it's only an estimation anyway. And everyone's system is different, so the actual utilisation, and the actual IBU in your beer, will never be specifically what is calculated anyway.'

Well, yes, this is true.all systems are different, and the utilisation will vary, and there are other considerations that effect the perception of the bitterness in our beers, such as grist, attenuation, clarity, etc. And, of course, it's up to each brewer to brew the recipes, and then tweak it for their own personal tastes anyway. However, when considering recipes that have been formulated on one platform, and duplicated on another, shouldn't we compare apples to apples, and oranges to oranges? I think so. So here are my thoughts on what causes this discrepancy, and what can be done to overcome it.

Beersmith and Promash both calculate hopping using the Tinseth formulation as a default. However, the original Tinseth formula uses flower as the baseline hop type.

Promash also uses flower as the baseline hop type. The hop utilisation factor, which is default at 100% in Promash, therefore means that for utilisation of pellets, you are looking at utilisation of 110%. Conversely, in Beersmith, it considers the hop utilisation for the baseline hop type to be for pellet; and a flower is utilised at -10% of the baseline. The HUF in Beersmith is also default at 100%; however this is relative to pellets. This means that the utilisation for flowers in Beersmith is in fact 90%. Therefore, in order to align the hop calculations in Beersmith with that of Promash, the HUF setting should be increased in Beersmith in order to raise the utilisation for the flower from 90% to 100%. To increase 90% to 100%, the HUF needs to be increased by a factor of 1.11r. (i.e., multiplied by 111%). By doing this, it realigns Beersmith to use flower as the baseline for the calculations (even though leaving pellet as the default choice).

This can be demonstrated by inputting a recipe identical in ingredients, batch size, losses, and equipment into both programmes and comparing the IBU calculation. If the Beersmith HUF is then increased to 111%, it will show that any hopping placed into this recipe will give (roughly) equal IBU value to what is entered in Promash, when Promash is left in its default state.

Further to this, calculating on the AHB database, or manually using Tinseths' formula, will give the same (or at least very similar) results as Promash; so an increase in Beersmith's HUF to 111% will correct this as well.

So, thoughts? Comments? (Without going down the YMMV path, or the 'just change programmes' path. )

 

[AndrewQLD]

Welllll, all I can say is,

Your basing your whole argument on Promash, and your assuming that promash is more accurate than BeerSmith.

As far as the recipe database here on AHB is concerned it doesn't take equipement into account and that does affect final IBU estimates so we can forget the recipeDB as a source of data, and let's not forget there are several other calculations on here that don't seem to gel with other programs as well such as SG and FG calculations.

Don't forget that Promash is a very old program (Pre 1998) and it's last minor update was 2003, it's fairly safe to say that it has fallen behind the times and is not as up to date as it could be. Although still an excellent program.

As far as BeerSmith is concerned it is a newer program, utilizing the latest changes to brewing science as far as calculations are concerned and it gives identical IBU results to BeerTools Pro ( another new, modern program) as far as I can tell after inputting several identical recipes.

I don't disagree that there are obvious differences between the way Promash calculates IBU estimates and the way BeerSmith and BeerToolsPro do, but to assume that BeerSmith is wrong and promash is right I can't agree with.

Seems you have 2 New programs agreeing and one old program disagreeing, I think I'll stick with the new ones.

Andrew

 

[jonocarroll]

Your basing your whole argument on Promash, and your assuming that promash is more accurate than BeerSmith.

...

Seems you have 2 New programs agreeing and one old program disagreeing, I think I'll stick with the new ones.

Ugh. I'm going to sympathise with buttersd70 about people not reading this stuff properly and discounting it off hand.

My understanding of the above (and I've checked the working) is that both PM and BS calculate the same - both using Tinseth, but they use a different baseline for the calculations for the utilisation. What buttersd70 has presented is a method of unifying the calculations so they give the same results in all cases - flowers, plugs, or pellets. It's not a question of accuracy or agreement at all, just scaling.

If people are interested, I believe buttersd70 has the reverse calculations for converting the other way. It's not a 'one program does it right, the other does it wrong' thing.

As for 'other things affecting the IBUs' I'll reiterate what was said earlier today in another thread - you should still be able to calculate the theoretical value, regardless of whether or not you'll hit it. Hell, if that weren't the case I'd be out of a job! (a physicist can tell you which horse will win the race... in the limit of spherical horses B) ).

As for PM being 'outdated' I don't see how the science has changed since it was released, so that's a pretty odd comment.

 

[AndrewQLD]

QB, the implication was that BeerSmith didn't do it right but Promash did, if Butters has a reverse equation then that should have been presented as well and I wouldn't have responded.

As for PM being 'outdated' I don't see how the science has changed since it was released, so that's a pretty odd comment.

Not really, you say it yourself here

both PM and BS calculate the same - both using Tinseth, but they use a different baseline for the calculations for the utilisation.

And perhaps that is were the science has changed and is possibly more accurate.

I didn't disagree with what Butters has stipulated or his conversion, I only disagreed that he assumed promash to be correct and not the other way around.

 

[jonocarroll]

QB, the implication was that BeerSmith didn't do it right but Promash did, if Butters has a reverse equation then that should have been presented as well and I wouldn't have responded.

...

And perhaps that is were the science has changed and is possibly more accurate.

see here;

By doing this, it realigns Beersmith to use flower as the baseline for the calculations (even though leaving pellet as the default choice).

This can be demonstrated by inputting a recipe identical in ingredients, batch size, losses, and equipment into both programmes and comparing the IBU calculation. If the Beersmith HUF is then increased to 111%, it will show that any hopping placed into this recipe will give (roughly) equal IBU value to what is entered in Promash, when Promash is left in its default state.

It's a workaround for not having to change the default in BS, but to get the same IBU calculation. People getting the wrong result is due to using poor settings in BS.

Read the OP again with the intention above (that this is a way to get the default results the same) and see if it gels better. I believe you have taken the wrong interpretation.

 

[AndrewQLD]

see here;


It's a workaround for not having to change the default in BS, but to get the same IBU calculation. People getting the wrong result is due to using poor settings in BS.

Read the OP again with the intention above (that this is a way to get the default results the same) and see if it gels better. I believe you have taken the wrong interpretation.

Ah, that must be the case then.

Andrew

 

[buttersd70]

The point of my post was not to put one programme over another; but rather to illustrate why there are differences between the two. Conversly, changing the HUF in Promash to 90% would reallign it with beersmith; it works both ways.

The database was mentioned, not because I believe it to be particularly accurate; but rather, because it uses the same default hop type (ie flowers) in the calculation as promash does. The main comparisons I have done is between Promash and Beersmith; with Tinseths calculator on his home page as a check. Neither Promash or Beersmith align absolutely with Tinseths java calculator; this would, of course, be due to no equipment settings, other than volume and boil gravity being available in Tinseths script. Promash is much closer than beersmith, though.

The intention of the post is soley so that people, when trying to emulate a recipe submitted by another brewer, which has been created on a different programme, can make adjustments (if they should choose to do so), and compare apples to apples, and oranges to oranges.

How many times does brewer A say "I tried some of brewer B's beer and liked it, so I brewed brewer B's recipe, and it came out more/less bitter than when brewer A does it, and I don't understand why that is."?

And just for the record - I personally use Beersmith. All programmes have pro's and cons......this is just one of them.

 

[Fourstar]

How many times does brewer A say "I tried some of brewer B's beer and liked it, so I brewed brewer B's recipe, and it came out more/less bitter than when brewer A does it, and I don't understand why that is."?

Just to throw a spanner in the works, most brewers make the mistake of following gram weights of hops and not IBU count/AA of the hop when calculating their recipe in their own program. So they end up using the same hop, different AA and it goes from being 32IBU to 43. a very easy thing to get wrong. Or the issue of not having the same formula set in their program. Other perceptions such as yeast in suspension, tannins, serving temp and water profile can also change the perception of hop bitterness. Out of interest, how much does this baseline throw a recipe out by? 2-3IBU? If so, not enough to make a valid difference in bittering perception if this is the case. Considering most cant always tell the difference in bitterness between beers of a 5IBU difference let alone 2-3IBU.

Saying that, i agree with your original point butters, they should be follwing the same baseline for all calculation types, just for standardisation purposes.

 

[buttersd70]

Just to throw a spanner in the works, most brewers make the mistake of following gram weights of hops and not IBU count/AA of the hop when calculating their recipe in their own program. So they end up using the same hop, different AA and it goes from being 32IBU to 43. a very easy thing to get wrong. Or the issue of not having the same formula set in their program. Other perceptions such as yeast in suspension, tannins, serving temp and water profile can also change the perception of hop bitterness. Out of interest, how much does this baseline throw a recipe out by? 2-3IBU? If so, not enough to make a valid difference in bittering perception if this is the case. Considering most cant always tell the difference in bitterness between beers of a 5IBU difference let alone 2-3IBU.

Saying that, i agree with your original point butters, they should be follwing the same baseline for all calculation types, just for standardisation purposes.

Your points are all valid, fourstar. The difference in the level of the IBU depends on how bitter it is to start with...so for 25 IBU it would be 2.75IBU difference....as you say, not likely to be percieved in a moderate OG, moderate IBU beer....it's much more likely to be percieved in a low gravity beer with a fine balance, such as a mild, or a table beer.