Bru'n water vs beersmith for ph correction predictions

[cfrazier77]

I have always used Bru'n water for water adjustments including ph. I use 85% phosoric acid. I tried using beersmith for the first time and got a different amount to get to the same ph, 5.4. 2 ml for 3.5 gallons Bru'n vs 5 ml beersmith. I use ro water and that was in both. I don't have the ability to test it right now, my bulb is dead on my meter, has been for years. Has anyone else noticed this? If so which one has been closer to accurate? Thanks

 

[Silver_Is_Money]

It would be quite highly naive to presume that a single example of deviance would lend any level of validity to a presumption that one is better than the other for all possible recipes and procedures and volumes and water source analyticals and mineralizations, and grist components, etc...

I have no insider knowledge whereby to verify this, but I've seen more than one mention that Beersmith applies D. M. Riffe's math models. I find his math models (or if you wish, science, which is a term much maligned these days) to be quite sound overall. The real problem (as I have often discussed) is not however one of math model soundness, but of malt data unsoundness combined with basic unavailability. If such data is unsound, and also generally to highly unavailable, then even the most sound of models will not be capable of achieving validity. Thus things like presumptions and educated guesses and data cherry picking and empirical modeling must enter in.

The other software is unknown as to its internal math modeling, as this is hidden from the public. Thus the soundness of it is fully unknown. All that can really be said here is that if it differs from modeling that is known to be compellingly sound, then ... It may however for some recipes and procedures and water analyticals and volumes and grist components, etc... make better presumptions and educated guesses and data cherry picking and empirical modeling efforts.

All of the above said, we dabblers in this art of prediction really have no idea as to what is being discussed here whereby to make an independent assessment, as zero data has been provided, whereas full data is required.

 

[CascadesBrewer]

I tried using beersmith for the first time

What version of BeerSmith are you using? I think it was in version 3.1 (which matches with the Web version) you could select between 2 models: BW, and MPH. I believe that "BW" is the same model used by Bru'n Water, and I believe that MPH was the model used in prior version of BeerSmith but it was tweaked for version 3.1.

When I first got into water adjustments about 3 years ago, I had much better luck with Bru'n Water matching up with my measured pH than I did with BeerSmith. With the release of 3.1 I found that I was getting fairly accurate results using the MPH model in BeerSmith, so I moved over to that so I did not have to enter my recipe into another piece of software.

I recently switched from using 10% Phosphoric Acid to 88% Lactic Acid and I have been finding that the MPH model in BeerSmith results in a slightly lower mash pH than my target (say 5.34 measured when my target was 5.4). This has been over 5-6 batches. Still pretty close and there is debate what is the ideal mash pH. I have sometimes just set my target a little higher or used a little less acid than recommended.

Note that I do mostly full volume mash BIAB. I am not positive how the thin mash impacts mash pH or how any of the models incorporate mash thickness.

 

[Bassman2003]

I need to revisit Beersmith's pH tool. I did not think it was accurate in the beginning but I know some changes have been made. I like how it has an offset so if you see a trend, you can account for it.

 

[tracer bullet]

I've entered various recipes into various software a few times to compare results to my meter's results, and see which one I deemed to be the most accurate. One thing I think I've found is that your base malt "actual" vs coded DIpH (or pHDI?) is going to be wrong at times, sometimes maybe substantially so, and can have a large effect.

I've learned to try and keep the same base malt in a recipe, even as the others change (for example, in a porter, let's say I stick with Golden Promise even if I change up the brown or chocolate malts). Within a type you need to probably choose a brand as well, knowing they'll vary (and of course recognize that even within a brand it's going to vary as well, but at least you can attempt to minimize this).

So I'm inventing #'s here but let's say I brew a porter, I expect a 5.4 pH per some calculator, and I measure 5.32. I'll call that a delta of 0.08 and next time add that to my target, perhaps shooting for 5.48 in that same calculator for the same or at least very similar recipe (again keeping in mind I attempt to repeat the base malt type and brand).

I'm blabbering but - it's helped me, I think. Otherwise if you change base malts and brands you'll just be all over the place and somewhat frustrated.

I made a thread about DIpH values for Maris Otter brands last night, taking a different approach to, I think, the same question as in the 1st post. So that I can at least feed in good values, or over-ride them when possible.

 

[VikeMan]

What version of BeerSmith are you using? I think it was in version 3.1 (which matches with the Web version) you could select between 2 models: BW, and MPH. I believe that "BW" is the same model used by Bru'n Water, and I believe that MPH was the model used in prior version of BeerSmith but it was tweaked for version 3.1.

I don't know what version of MpH the current Beersmith uses, but I can say for sure that MpH 4.2 was a major change from previous MpH versions.

 

[Silver_Is_Money]

For unbuffered water (which is to say more specifically "deionized water", a step up from distilled, and not uncommonly several steps up from RO as to its purity) with less than the head of a pin sized spec of a weak acid added, such as to drop its pH to say the ballpark of 4-5-ish, a full doubling of said water volume via adding additional pure and fresh pH 7 deionized water (not having been exposed to any CO2, and thus pH 7) results in a pH change of only 0.3 points, but when the system is quite highly buffered, which occurs simply via adding ones grist to it, a thinning of the mash via a doubling of the mash water volume barely moves the pH at all. (Buffering is the measure of resistance to change in pH).

Unbuffered Example:
If I have otherwise (than for a smidge of acidity) highly unbuffered DI water at pH 5.5 the initial molar H+ (acid) concentration at pH 5.5 is:
10^-5.5 = 0.000003162M
whereby: -log(0.000003162) = 5.500038 pH (with the slight pH deviation here reflecting molar concentration rounding error)

If I now subsequently double the water volume via adding DI water, I cut the molar molar concentration of H+ (free or fully dissociated acid ions) in half:
0.000003162M ÷ 2 = 0.000001581M

Whereby the resulting pH becomes:
-log(0.000001581) = 5.801

Now toss in say 5 Kg. of an aggregate malt grist with buffering strength measured at ~40mEq/Kg_pH and it may be difficult to accurately measure the movement in the pH via the doubling of the mash water due to this buffer (effectively an H+ ion absorber) vastly dwarfing the minuscule 0.000003162 molar H+ concentration of the water carrier.

Bottom line: Mash water thickness/thinness plays little to no part of consequence with regard to pH drift.

 

[Silver_Is_Money]

But what does potentially change things measurably is the degree of Malt Crush. BC's are computed for Congress Mash conditions, which call for a pulverized grist. But a crush through a mill set to 0.036" (as but an example) is nowhere near pulverized. If a pulverized base malt with a DI pH of (for example) 5.76 pH exhibits a BC (Buffering Capacity) of 40 mEq/Kg_pH then for a grist consisting of 5 Kg. of it we derive (for the case of DI water with no mineralization):

Delta_pH = mEq/(BC x Kg.)
(5.76 - 5.40) = mEq/(40 x 5)
0.36 = mEq/(200)
200 x 0.36 = mEq
mEq = 72

At 'specifically' pH 5.40 (and no other pH) the Weak Acid 85% Phosphoric delivers 14.92374 mEq/mL

Therefore:
72 mEq ÷ 14.92374 mEq/mL = 4.825 mL of 85% Phosphoric Acid required

That seems rather in line with 5 mL of 85% Phosphoric Acid.

But if the crush is poor (by comparison to that of pulverization), then some of the 'basic in nature with respect to 5.40 pH' mEq's of the grist (or effectively OH- 'like' ions) will not be liberated, quite similarly as to why some of the sugars of the grist will not get converted/liberated, leading to poor conversion vs. theoretically perfect sugar conversion/liberation for pulverized. The impact here is as if the BC was to be 'effectively' reduced (loosely akin to converting at 29 gravity points per pound vs a perfect 37 points per pound). What if your "real world crush" BC is thereby only 30 vs. pulverized at 40?

Delta_pH = mEq/(BC x Kg.)
(5.76 - 5.40) = mEq/(30 x 5)
0.36 = mEq/(150)
150 x 0.36 = mEq
mEq = 54

54 mEq ÷ 14.92374 mEq/mL = 3.62 mL of 85% Phosphoric Acid required

Not quite a low prediction of 2 mL, but it shows that crush potentially impacts things to a large degree.

And it might also potentially show that neither 5 mL nor 2 mL are highly "decent" predictions.

So for this purely hypothetical case as we have outlined, what mash pH might we expect for the case of going with 2 mL, and for the case of going with 5 mL, when the (hypothetically correct) answer is actually 3.62 mL?

First let's add 2 mL of 85% Phosphoric Acid (whereby 2 x 14.92374 = 29.85 mEq's of H+ ions):
(pH_Mash - 5.40) = (54 - 29.85)/(30 x 5)
pH_Mash - 5.40 = 0.161
pH_Mash = 5.561

Next let's add 5 mL of 85% Phosphoric Acid (whereby 5 x 14.92374 = 74.62 mEq's of H+ ions):
(pH_Mash - 5.40) = (54 - 74.62)/(30 x 5)
pH_Mash - 5.40 = -0.1375
pH_Mash = 5.263

Both are off, but within 0.2 pH of the 5.40 target pH. Albeit merely by luck!!!

 

[Silver_Is_Money]

Oops, a minor correction:

The acid strength of 85% Phosphoric Acid at pH 5.40 is actually 14.86526.

14.92374 as I've used (incorrectly) above is for 5.50 pH.

The pH prediction difference should be trivial. So I'll leave it to each of you to correct my math...

 

[Silver_Is_Money]

To do the hard math for a "real world" grist buffer of 30 mEq/Kg_pH and 5 Kg thereof, whereby to see how the doubling of the mash volume (from say 10L to 20L) might impact the pH:

Lets assume the conditions assigned a few posts above, at post #7:

Initial 10L mash water has 0.000003162 moles per liter H+. So for 10L this becomes 0.000003162 x 10 = 0.00003162 moles, which = 0.03162 millimoles (or mEq's) of H+

Diluted to 20L mash water has 0.000001581 moles per liter H+. So for 20L this becomes 0.000001581 x 20 = 0.00001581 moles, which = 0.01581 millimoles (or mEq's) of H+

In post # 8 above we determined that 54 mEq are required whereby to move a pH 5.76 grist of 5 Kg. to pH 5.40 for the "real world" case of BC = 30.

A minuscule 0.03162 mEq's and/or 0.01581 mEq's of H+ ion deviation in either direction from the hypothetically computed 54 mEq's required to hit the pH target of 5.40 is clearly not going to show up on a pH meter, when magnitudes larger computational prediction errors of 2 mL and 5 mL vs. 3.62 mL for 85% Phosphoric Acid respectively weren't even causing a measured pH difference of 0.20 pH.

You have been given the method. I'll now leave the pH deviation calculation up to each of you to complete (sort of as a homework assignment).

 

[Silver_Is_Money]

A poser/question: How can your chosen software know that the grist's "real world" BC might be closer to 30 mEq/Kg_pH instead of the software's internal "data base" grabbed presumption of 40 mEq/Kg_pH if it does not ask you to input your mill crush (in inches or mm)?

 

[CascadesBrewer]

A poser/question: How can your chosen software know that the grist's "real world" BC might be closer to 30 mEq/Kg_pH instead of the software's internal "data base" grabbed presumption of 40 mEq/Kg_pH if it does not ask you to input your mill crush (in inches or mm)?

I will go out on a limb and say that it is because the models used by most software do not account for crush!

 

[Silver_Is_Money]

Trivia: How much 85% Phosphoric Acid does it take to move 30L of pH 7.00 deionized water to pH 5.40?

10^-5.40 = 0.000003981M (or molar) H+ (acid)

30L x 0.000003981M = 0.00011943 moles of H+ = 0.11943 millimoles of H+ = 0.1194 mEq's of H+

85% Phosphoric Acid at 5.40 pH has a dissociation related acid strength of 14.86526 mEq/mL

0.1194/14.86526 = 0.00803215 mL

Answer: 0.008 mL of 85% Phosphoric Acid will move 30 Liters of freshly produced deionized water to a pH of 5.40.

Bonus Answer: 0.1095 mL of 10% Phosphoric Acid will do the same...

 

[ScrewyBrewer]

I will go out on a limb and say that it is because the models used by most software do not account for crush!

Both ezRecipe and MPH 4 definitely do account for grain crush.

 

[ScrewyBrewer]

For unbuffered water (which is to say more specifically "deionized water", a step up from distilled, and not uncommonly several steps up from RO as to its purity) with less than the head of a pin sized spec of a weak acid added, such as to drop its pH to say the ballpark of 4-5-ish, a full doubling of said water volume via adding additional pure and fresh pH 7 deionized water (not having been exposed to any CO2, and thus pH 7) results in a pH change of only 0.3 points, but when the system is quite highly buffered, which occurs simply via adding ones grist to it, a thinning of the mash via a doubling of the mash water volume barely moves the pH at all.

Yes, this is true, until you add in brewing salts used to flavor the brewing water.

 

[Silver_Is_Money]

Yes, this is true, until you add in brewing salts used to flavor the brewing water.

Precisely why pertinent brewing salts must be accounted for in units of mEq as opposed to units of ppm or mg/L, as I often argue.

 

[Silver_Is_Money]

Both ezRecipe and MPH 4 definitely do account for grain crush.

Do they ask the end user for input of the users crush setting? And then vary their prediction advice accordingly?

 

[ScrewyBrewer]

Do they ask the end user for input of the users crush setting? And then vary their prediction advice accordingly?

Yes, that they do.

 

[Silver_Is_Money]

Yes, that they do.

Glad I could help.

 

[CascadesBrewer]

Do they ask the end user for input of the users crush setting? And then vary their prediction advice accordingly?

As far as I know, there is not a value in BeerSmith that would feed grain crush into the MpH model. I don't know what version is used by BeerSmith.

I take it this is the home for the MpH model. I took a quick look at the sheet and a grain crush entry or setting did not jump out at me.
http://homebrewingphysics.blogspot.com/2020/03/mph-water-calculator-v42.html

 

[ScrewyBrewer]

Mark explained to me where he located the entry on mph4, but it has been years since I used it.

In ezRecipe it is located on the ‘Mash’ sheet tab, under grain crush.

 

[mashpaddled]

I use Bru'n Water and have for years because it seems to work out. Even if it is perpetually off, I know what I'm going to get when I target a certain ph or profile regardless of the true metrics. When I've tested the ph it is on or close enough that I worry about the difference.