Beersmith vs Bru'n water mash pH predictions

[BuiltToLast]

Previously posted on the BeerSmith forum:

I just updated BeerSmith and decided to try out the new pH features. I normally use Bru'n water so I decided i'd compare the results of both programs and make sure they match up. As you might imagine by now, they did not match up at all. First, Bru'n water works great when I'm brewing 7bbls, its spot on every single time, but when I brew a 8 gallon batch, my mash pH always comes in too high. I've corrected this by shooting for a mash pH of 5.2 if I want 5.4. Now, when I set up my water profile in BeerSmith and it calculates my 88% lactic acid addition, its saying i need 7.3ml to get it down to 5.4, whereas bru'n water tells me to get down to 5.2 pH (usually gets me 5.4) i need to add 4ml. Anyone else have similar issues, or perhaps an explanation for what might be causing this discrepancy?

 

[normonster]

Wait...what? Did you say 7bbl?

I have no explanation, but it sure makes me wonder why I have the opposite problem with Brew'n. My indicated lactic acid additions always seem too high so I've adopted a practice of adding only half, checking, then adding the rest drop by drop.

But seriously though....did you say 7bbl?

 

[BuiltToLast]

yeah...i brew commercially but I brew at home so I can try new processes and ideas (theres only so much risk ill take brewing 216 gallons). In the brewery it works great, at home (same municipal water source) its always way off.

 

[brick_haus]

I'm just babbling again... But the different results by batch size sounds like the acidification curve must not be linear and the brun water calculator doesnt/cant account for this. Martin will chime in I bet and explain it.
As far as you two having opposite phenomenas, has to be improper data being entered into one calculator or the other.

 

[BrewmasterBrad]

I've been doing some comparisons between BeerSmith and Bru'n Water lately, but I need to work more with the pH tools in the new version of BeerSmith before I can really say which one is more accurate.

I will post my results here when I have more data. I am starting with RO water and building it back up. I am also testing to see which tool best predicts the actual water makeup after adding the minerals back in to the RO water.

 

[ajdelange]

I just updated BeerSmith and decided to try out the new pH features. I normally use Bru'n water so I decided i'd compare the results of both programs and make sure they match up. As you might imagine by now, they did not match up at all.

I wouldn't expect them to.

First, Bru'n water works great when I'm brewing 7bbls, its spot on every single time, ...

You have some problem with your pH measurement technique. It is impossible for any calculator to be 'spot on' (which I am interpreting as meaning ±0.02 pH) because you can't measure pH that repeatably with a real pH meter using buffers with ±0.02 pH tolerance (the tolerance of the NIST traceable operational buffers that most of us use. Even if we used, in the brewery, the techniques we use in the lab to get ±0.02 accuray none of the popular programs (neither of the ones mentioned here) have the ability to accept the required malt characteristics to allow a pH prediction that accurate. It is doable but the calculator must know the DI mash pH, at least one and more probably (depending on the malt) 2 titration curve parameters and the malts temperature glide. The programs you are referring to use malt type and color as proxies for these because no one has the full parameter set. Attempts to get the maltsters to measure them have been fruitless. Malt color isn't a bad proxy but it isn't a particularly good one either.

So how is it that lots of people report "Well I don't know how you are screwing up but Sudsawater is spot on for me every time". I don't know but I strongly suspect it is confirmation bias. You want the program to be accurate and so you dip your pH meter and watch the numbers. The pH of a mash varies quite a bit over time and thus if you wait long enough you will get to the number you expect at which time you declare victory and write down the predicted number in you logbook even though perhaps if you had waited longer the pH would have continued to drift. This doesn't always work out, of course, and when it doesn't confirmation bias protects you by making this aberration less memorable than measurements that confirm your expectations. Confirmation bias is very real and powerful. Talk to one of your friends or relatives that thinks either of the current presidential candidates is fit for the office.

...but when I brew a 8 gallon batch, my mash pH always comes in too high. I've corrected this by shooting for a mash pH of 5.2 if I want 5.4.

This makes no sense so again I expect a problem with pH measurement or the interpretation of the reading. It would be worthwhile to compare the two meters (I'm guessing that you don't use the same one at home as at work but perhaps you do) and running the stability check (see Sticky at the top of this forum). There may be a very small difference in acheiveable pH for exactly the same proportions of exactly the same malts with exactly the same water but I would guess that none of those factors is indeed exactly the same. Things tend to be more stable in a larger system as the ratio of thermal mass of the mash to the vessel surface are is larger and so that might be a factor.

Now, when I set up my water profile in BeerSmith and it calculates my 88% lactic acid addition, its saying i need 7.3ml to get it down to 5.4, whereas bru'n water tells me to get down to 5.2 pH (usually gets me 5.4) i need to add 4ml.

Each of the two programs comes up with a mash pH prediction using the descriptive data you feed it on malt type and color but they use that data in different ways neither of which is, AFAIK, analytically robust by which I mean that I believer there is a lot of empiricism mixed in. In any case the models are not the same and thus the predicted pH's are different.

Anyone else have similar issues, or perhaps an explanation for what might be causing this discrepancy?

Lots of brewers have noted these discrepancies. If you want better pH predictions lean on you malt supplier to give you the data you need on the products he sells you. Given that data it is actually quite simple to calculate a good pH estimate.

 

[laredo7mm]

I took a look at BeerSmith's forum and Brad mentioned he posted a blog about the pH estimates. Here is the link:

http://beersmith.com/blog/2016/08/10/using-the-new-brewing-water-and-mash-ph-tools-in-beersmith-2-3/

In that blog post he says he based his calculations on the work that D. Mark Riffie did. Below are the links to the Mark's blog posts about his work. He talks about the methods used in EZ water, Bru'n Water, Kaiser Water, Brewer's Friend, and how his differs from theirs.

The grain bill: http://homebrewingphysics.blogspot.com/2013/09/a-homebrewing-perspective-on-mash-ph-i.html

The water: http://homebrewingphysics.blogspot.com/2013/09/a-homebrewing-perspective-on-mash-ph-ii.html

Here is a link to the latest version of his calculator: http://homebrewingphysics.blogspot.com/2016/03/brewing-water-calculator-mph-water.html


From running through some recipes, Brad's calculations are really close to Mark's calculator with one exception...When using wheat, BeerSmith does not take into account that it has a higher pH. When using Mark's calculator, you can enter the grain as wheat and it uses a different mEq/kg value. BeerSmith just treats it as a typical base grain.

 

[CadiBrewer]

I wouldn't expect them to.

You have some problem with your pH measurement technique. It is impossible for any calculator to be 'spot on' (which I am interpreting as meaning ±0.02 pH) because you can't measure pH that repeatably with a real pH meter using buffers with ±0.02 pH tolerance (the tolerance of the NIST traceable operational buffers that most of us use. Even if we used, in the brewery, the techniques we use in the lab to get ±0.02 accuray none of the popular programs (neither of the ones mentioned here) have the ability to accept the required malt characteristics to allow a pH prediction that accurate. It is doable but the calculator must know the DI mash pH, at least one and more probably (depending on the malt) 2 titration curve parameters and the malts temperature glide. The programs you are referring to use malt type and color as proxies for these because no one has the full parameter set. Attempts to get the maltsters to measure them have been fruitless. Malt color isn't a bad proxy but it isn't a particularly good one either.

So how is it that lots of people report "Well I don't know how you are screwing up but Sudsawater is spot on for me every time". I don't know but I strongly suspect it is confirmation bias. You want the program to be accurate and so you dip your pH meter and watch the numbers. The pH of a mash varies quite a bit over time and thus if you wait long enough you will get to the number you expect at which time you declare victory and write down the predicted number in you logbook even though perhaps if you had waited longer the pH would have continued to drift. This doesn't always work out, of course, and when it doesn't confirmation bias protects you by making this aberration less memorable than measurements that confirm your expectations. Confirmation bias is very real and powerful. Talk to one of your friends or relatives that thinks either of the current presidential candidates is fit for the office.




This makes no sense so again I expect a problem with pH measurement or the interpretation of the reading. It would be worthwhile to compare the two meters (I'm guessing that you don't use the same one at home as at work but perhaps you do) and running the stability check (see Sticky at the top of this forum). There may be a very small difference in acheiveable pH for exactly the same proportions of exactly the same malts with exactly the same water but I would guess that none of those factors is indeed exactly the same. Things tend to be more stable in a larger system as the ratio of thermal mass of the mash to the vessel surface are is larger and so that might be a factor.



Each of the two programs comes up with a mash pH prediction using the descriptive data you feed it on malt type and color but they use that data in different ways neither of which is, AFAIK, analytically robust by which I mean that I believer there is a lot of empiricism mixed in. In any case the models are not the same and thus the predicted pH's are different.



Lots of brewers have noted these discrepancies. If you want better pH predictions lean on you malt supplier to give you the data you need on the products he sells you. Given that data it is actually quite simple to calculate a good pH estimate.

AJ, stealing a quote from one of your other posts regarding remember that these are the musings of an amateur, but could some of the discrepancies between his 7bbl and his 8 gallon batches be due to the the ability to more accurately measure additions and weights at the greater size batches? I've always had trouble accurately measuring out additions for my test mashes. Being a little off on a test mash swings my results greatly. Being a little off on my ten gallon batch doesn't make a discernible mistake, usually within the margin of error for the software I'm using. I can only imagine that 240 gallons would make the accuracies that much more exact.

 

[brick_haus]

AJ, stealing a quote from one of your other posts regarding remember that these are the musings of an amateur, but could some of the discrepancies between his 7bbl and his 8 gallon batches be due to the the ability to more accurately measure additions and weights at the greater size batches? I've always had trouble accurately measuring out additions for my test mashes. Being a little off on a test mash swings my results greatly. Being a little off on my ten gallon batch doesn't make a discernible mistake, usually within the margin of error for the software I'm using. I can only imagine that 240 gallons would make the accuracies that much more exact.

Very good point.

 

[ajdelange]

It would depend on how he does things but it is certainly a possibility and well worth mentioning.

 

[55x11]

I took a look at BeerSmith's forum and Brad mentioned he posted a blog about the pH estimates. Here is the link:

http://beersmith.com/blog/2016/08/10/using-the-new-brewing-water-and-mash-ph-tools-in-beersmith-2-3/

In that blog post he says he based his calculations on the work that D. Mark Riffie did. Below are the links to the Mark's blog posts about his work. He talks about the methods used in EZ water, Bru'n Water, Kaiser Water, Brewer's Friend, and how his differs from theirs.

The grain bill: http://homebrewingphysics.blogspot.com/2013/09/a-homebrewing-perspective-on-mash-ph-i.html

The water: http://homebrewingphysics.blogspot.com/2013/09/a-homebrewing-perspective-on-mash-ph-ii.html

Here is a link to the latest version of his calculator: http://homebrewingphysics.blogspot.com/2016/03/brewing-water-calculator-mph-water.html


From running through some recipes, Brad's calculations are really close to Mark's calculator with one exception...When using wheat, BeerSmith does not take into account that it has a higher pH. When using Mark's calculator, you can enter the grain as wheat and it uses a different mEq/kg value. BeerSmith just treats it as a typical base grain.

so from yours and ajdelange, would it be fair to conclude that the new BeerSmith water chemistry calculations (or should I say - "estimations") - are as good as any out there?

Like many others, I would prefer to do all my calculations in one piece of software, and despite minor issues I have with BeerSmith (mostly minor cosmetic/design related), I do love the functionality of BeerSmith over any other software.

 

[laredo7mm]

so from yours and ajdelange, would it be fair to conclude that the new BeerSmith water chemistry calculations (or should I say - "estimations") - are as good as any out there?

Like many others, I would prefer to do all my calculations in one piece of software, and despite minor issues I have with BeerSmith (mostly minor cosmetic/design related), I do love the functionality of BeerSmith over any other software.

I hope so, as I would also like to use just one piece of software. I am not a water expert at all. I still need to buy a PH meter so I have no way to know if any of them are accurate.

 

[ajdelange]

Some people swear by one, others by another. I really do suspect confirmation bias (see #4) to be at play here but it is just possible that the model used by one of the programs is a better match to the actual grain parameters for a particular style of beer because perhaps the author concentrated on a style or styles that match the ones the brewer is interested in.

 

[malaeum]

Edit: Nevermind, reading through Martin's page on this again I seem to have overlooked his discussion of this.

Nothing to see here, move along...