Mash pH prediction is different between calculators

[Silver_Is_Money]

I've been playing around with various programs which try to predict your mash pH, and so far it seems as if the sources give varying pH feedback as follows:

1) Bru'n Water generally gives the lowest predicted pH's
2) Kaiser is close, but generally a tad higher pH predictions
3) Brewer's Friend generally comes in a tad higher than these two in pH
4) EZ Water Calculator almost always comes in noticeably higher then all of the above

Is there any one of these that consistently comes closer in mash pH projection vs. reality than the others? Perhaps it is my technique in utilizing these various tools that is the problem here.

And if I'm seeing a range from 5.33 to 5.42 for the first 3, and 5.56 for the last one, should I even be worried? At what low and high pH points are adjustments required?

 

[Big Monk]

The key to accurate estimations using software is being able to track the distilled water pH of the malt in the spreadsheet. I use a purpose built sheet with water calculations that are dead on accurate 100% of the time with 1/100th of a pH point precision.

 

[Silver_Is_Money]

5.33 pH is what the actual batch delivered. Bru'n Water was the closest.

 

[dmtaylor]

So many folks swear by the accuracy of Bru'n Water that I would trust that one. Personally I use my own calculations and they turn out "good enough" as well. If it's off by 0.1, who cares.

 

[BrunDog]

Use BnW and life is good.

Going to assume RPIScotty is kidding!

 

[Big Monk]

Use BnW and life is good.

Going to assume RPIScotty is kidding!

Kidding about what? My software? Very serious!

 

[dmtaylor]

RPI does not kid.

 

[BrunDog]

I guess not. But with a claim of 100% accuracy to pH of 0.01... he should be.

 

[ajdelange]

If he's not kidding us he's kidding himself.

 

[Big Monk]

I guess not. But with a claim of 100% accuracy to pH of 0.01... he should be.

I kid not. By tracking distilled water pH of individual Malts and using all the known estimation methods including malt color, mineral acid, biological acid, salts and Sauermalz, the water portion of our spreadsheet is accurate to +\- 0.01 pH units EVERY time.

I am a relatively funny guy, but I am not joking on this one. This has been verified 50+ times thus far using the Extech model ph100.

I am contemplating breaking this simple little water section out of the main sheet.

 

[ajdelange]

That does tax one's credibility as one can hardly be expected to be able to measure the pH of anything, let alone a complex mix of components that you haven't fully characterized (it is not enough to know the DI mash pH of the malts; you must also know their buffering as a function of pH in order to work out the acid/base chemistry) using buffers with a tolerance of ±0.02.

You are definitely kidding yourself.

 

[day_trippr]

 

[Big Monk]

All I can say is that the estimates given are within +/- 0.01 EVERY time using Malts, mineral acid, sauermalz, Sauergut, salts or antioxidants or any combination of all of them as verified by the equipment we have at our disposal.

Using YOUR work as a reference. And Kai's. You can balk all you want but thems is the breaks.

 

[ajdelange]

Well I hope you can understand why I am so skeptical. I use an electrode that costs more than 4 times what that meter did with a meter that cost 4 times what your meter did and process the results in a computer using algorithms that estimate the calibration parameters, including the iso electric pH based on hundreds of calibration readings and make the pH readings over a 20 - 30 minute period and find the pH changes over a few minutes by several time 0.01 pH. Plus, as mentioned above, bufffer tolerances of ±0.02, temperature measurement error, uncertainly in pHi and instrument noise, even with the lab grade meter I use make it extremely unlikely, in fact I'll say impossible, to measure pH consistently to 0.01 50 times in a row.

I have learned over the years that if it sounds too good to be true it isn't and this definitely sounds too good to be true. Nothing in the real world works EVERY time. Something is glaringly amiss here. Of course I am curious and would be happy to see if I can help you figure out what it is.

 

[dmtaylor]

Gents... perhaps we should instead be discussing why the frak accuracy to 0.01 even matters.... To most of us, I would say it does NOT. Moot argument as far as many/most are probably concerned.

 

[Big Monk]

Well I hope you can understand why I am so skeptical. I use an electrode that costs more than 4 times what that meter did with a meter that cost 4 times what your meter did and process the results in a computer using algorithms that estimate the calibration parameters, including the iso electric pH based on hundreds of calibration readings and make the pH readings over a 20 - 30 minute period and find the pH changes over a few minutes by several time 0.01 pH. Plus, as mentioned above, bufffer tolerances of ±0.02, temperature measurement error, uncertainly in pHi and instrument noise, even with the lab grade meter I use make it extremely unlikely, in fact I'll say impossible, to measure pH consistently to 0.01 50 times in a row.

I have learned over the years that if it sounds too good to be true it isn't and this definitely sounds too good to be true. Nothing in the real world works EVERY time. Something is glaringly amiss here. Of course I am curious and would be happy to see if I can help you figure out what it is.

Think of it this way: you know how people are always doing the "how do the water softwares compare to each other" bit? And they are always different but one is closest? They say BW was closest, with actual being 5.29 and estimated being 5.34.

Well ours estimates 5.29 and is always +\- 0.01.

I'm not trying to upset the world here. This has been verified with the equipment we have at our disposal 50+ times. I understand your equipment may be more sophisticated but I'm not trying to do chemistry experiments or any crazy analysis. I just know that are estimates are dead nuts every time. More so than any other estimation software. Take that for what it is but I'm not trying to diagnose errors. I'm just stating our empirical findings. There is nothing amiss as far as I'm concerned.

 

[ajdelange]

To try to get back to the basic question:

As no one has ever taken all the programs and tested them over a large ensemble of mashes there is no way to determine whether one is more accurate than another. Some feel that a particular one is better than another based on their experiences with them. We do see the "Sudso is always spot on for me" comments from time to time but as fundamental experiments in mash pH prediction show that measured mash pH drifts over time one has to ask when these guys are taking the measurements. We strongly suspect that good old Confimation Bias may be at work in these cases IOW the guy's chosen program predicts 5.47 so he dunks he meter, watches the pH drift and waits till it hits 5.47, declares victory and writes 5.47 in his log.

Putting that aside one of the problems with accurate pH prediction is that it requires accurate knowledge of the titration characteristics of the malt and it takes a lot of effort to get that. Accordingly the different programs use different models and it is quite possible that one program's models better fit the type of brewing (the styles...) you do than another in which case that program may work better for you than another one.

 

[ajdelange]

There is nothing amiss as far as I'm concerned.

There's part of the problem right there. If you knew what you were talking about (sorry, I just can't think of a more polite way to say it) you would never accept 50 pH readings within ±0.01 of a calculated value without launching a major investigation as to how that could happen. To give you the benefit of the doubt there is something you do not understand. To be more cynical I am afraid I have to conclude that you are pulling our legs here in which case shame on you for doing it and shame on me for being taken in.

 

[Big Monk]

What a tough crowd.

I thought it would be a good thing to share my work and our findings. Evidently that was false.

 

[ajdelange]

Gents... perhaps we should instead be discussing why the frak accuracy to 0.01 even matters.... To most of us, I would say it does NOT. Moot argument as far as many/most are probably concerned.

You are right. It doesn't. The reason for dwelling on it here is that a sequence of measurements consistently better than the accuracy of which a measurement system is capable is an indication of a problem that needs to be addressed.With a pair of ±0.02 NIST traceable buffers, a good meter and careful measurement we would expect about 42% of the measurements to be within ±0.01, about 73% will be within ±0.02 and about 90% within 0.03. Thus, even with scrupulous care and the best equipment we have to accept that we will get occasional measurements off by 0.03 in either direction. Allowing for the fact that when we brew we are not using laboratory equipment or practices I'd say our realizable accuracy is probably ±0.04. So if you are getting agreement between your prediction program and your meter readings of better than 0.08 or so I'd say you should be happy with that.

BTW the probability of getting 50 readings better that ±0.01 relative to a known (predictor is perfect) pH is .42^50 = 1.45367e-19. That's way less than winning the lottery.

 

[Big Monk]

For all the literalists out there:

We have been able to measure, within the ability of our equipment and with the best of intentions, a pH that matches our excel sheets estimate within +\- 0.01. There isn't any operator error I'm aware of.

What I can tell you is that I doubt anyone is measuring pH with the level of accuracy and detail that AJ is. We are using a meter slightly above average compared to most homebrewers and we believe our sheet bests the others out there. I'm not looking for anything other than to share that with people.

 

[dmtaylor]

Are you sharing your spreadsheet with the OP?

Has the OP's question been answered to his satisfaction?

Why are we still here?

 

[Big Monk]

Are you sharing your spreadsheet with the OP?

Has the OP's question been answered to his satisfaction?

Why are we still here?

http://www.********************/wp-content/uploads/2016/10/Low-O2-Brewing-v5.2.xlsx

If you have any questions please feel free to PM me anytime. The sheet is a full brewing software but you can use it like a standalone water software as well.

Wasn't trying to derail or start a war, just stating my observations backed up by our readings and advocating for this being a good resource.

 

[Big Monk]

There's part of the problem right there. If you knew what you were talking about (sorry, I just can't think of a more polite way to say it) you would never accept 50 pH readings within ±0.01 of a calculated value without launching a major investigation as to how that could happen. To give you the benefit of the doubt there is something you do not understand. To be more cynical I am afraid I have to conclude that you are pulling our legs here in which case shame on you for doing it and shame on me for being taken in.

Skepticism is something I promote wholeheartedly. Dismissing what I'm saying and questioning my integrity when all I am doing is providing insight into data produced from a piece of hard work I distribute freely (and free of charge) with no agenda is the real shame.

 

[ajdelange]

I thought it would be a good thing to share my work and our findings. Evidently that was false.

No, it's true but your claim sounds like an investment opportunity that carries no risk but has an annual return of 20 -25%.

I stand by my offer to try to help you figure out what's going on.

 

[Big Monk]

No, it's true but your claim sounds like an investment opportunity that carries no risk but has an annual return of 20 -25%.



I stand by my offer to try to help you figure out what's going on.

I sent you a PM.

I assure I am not looking to benefit in any other way other than to help people get the most accurate and consistent pH estimates possible with Excel. No risk whatsoever and nothing but positive results.

As crazy as it may sound it is true. We distribute the sheet freely and without cost and I assure you I am modest enough of a man not to be pumping this up for an attaboy.

 

[BrunDog]

OK, RPIScotty opened the Kimono... let's look at the math and see if its legit.

 

[Big Monk]

OK, RPIScotty opened the Kimono... let's look at the math and see if its legit.

The math is legit. The readings are legit. I think the aspect being questioned is whether all the various tolerances AJ listed previously make it even possible to measure with the accuracy I describe.

I can tell you that what we've seen is more accurate than other water software out there when entered correctly. It may be that the limitations of the test equipment used means we are not ACTUALLY +\- 0.01 in reality (tolerances, temp error, etc.). To me all that means is that other programs are MORE off than we are. And I'll take that.

Lab grade pH measurements may disprove my claims of accuracy, and that's fine. Given that a very good consumer grade meter (Extech pH110) was used and we see consistent readings across batches, I believe that at the very least out sheet is more accurate than others and dependable.

 

[ajdelange]

I assure I am not looking to benefit in any other way other than to help people get the most accurate and consistent pH estimates possible with Excel. No risk whatsoever and nothing but positive results.

In the sense that those offering such investment opportunities are usually Bernie Madoff types I can see how you might have thought that my intention was to impugn your character but it wasn't. The intention was to illustrate in a perhaps more familiar way than pH measurement data that if you make an outrageous claim people will suspect your motives. It appears that in this case the intentions are good and the inflation of the claims is due to over enthusiasm fueling confirmation bias.

As crazy as it may sound it is true.

It doesn't sound crazy. I've got a lot of experience working with guys who get so excited over the things they are working on that they are blind to results they don't want to see. Scientists are well aware of the ways in this can happen and have formalized procedures for preventing it from happening but sometimes it can be very insidious.

We distribute the sheet freely and without cost and I assure you I am modest enough of a man not to be pumping this up for an attaboy.

I'll try to take a look at it. If it is indeed based on what I published in TQ then I don't think you can do better and I know that method does produce (as illustrated in the article) results in the ±0.02 region but only under the best of circumstances (when one makes a grist up from malts he has done detailed titration measurements on). Under these circumstances it is the best model out there for sure because of its sound scientific basis. When using measurements, no matter how detailed, made by someone else (or you) on malts that differ (maltster, batch number, season, cultivar....) from the measured malts accuracy naturally goes down.

So if you have implemented the algorithm correctly, and as it is so simple I don't see how you could fail at that, then your spreadsheet is indeed better than the others but subject to the huge caveat that in order to be better than the others it must be fed accurate malt data .

 

[Big Monk]

What we typically do AJ is modify what we call the "malt override" cell (listed as °L (pH) for base Malts) based on the distilled water pH found in the malt analysis data sheet.

We account for the pH drop using Sodium Metabisulfite, Antioxin SBT (hasn't made it in formally yet), Sauermalz, Sauergut, Mineral Acid (lactic and phos) and salts.

 

[ajdelange]

What we typically do AJ is modify what we call the "malt override" cell (listed as °L (pH) for base Malts) based on the distilled water pH found in the malt analysis data sheet.

Where does one put in the buffering capacity info and where does the user get that data?

 

[Big Monk]

Where does one put in the buffering capacity info and where does the user get that data?

(Edit) Are we talking about DI pH values for the base Malts? You would need to enter values for °L in the spaces shown below while removing all substances that reduce or add pH. Typically I do it by finding the °L value that gives me DI pH when the malt is set to 100% of the grist. We get the DI pH off the lot analysis or batch analysis for the malt:

View attachment ImageUploadedByHome Brew1483975492.062175.jpg

If I'm misunderstanding or unintentionally sidestepping your real question please let me know. You'll have to bear with me, I'm an electrical engineer and not a chemist.

 

[ScrewyBrewer]

It is my experience that most of these discrepancies are directly related to default DI pH values of the grain types and the real DI pH values of the actual maltster's grains used in the mash. That and the fact that some calculators rely on the color of the grains rather than DI pH values. There is also some level of user misinterpretation, and/or pH measurement process to take into account.

 

[Silver_Is_Money]

How do I select custom water and percentages of RO/custom?

 

[ajdelange]

(Edit) Are we talking about DI pH values for the base Malts? You would need to enter values for °L in the spaces shown below while removing all substances that reduce or add pH. Typically I do it by finding the °L value that gives me DI pH when the malt is set to 100% of the grist. We get the DI pH off the lot analysis or batch analysis for the malt:

No, we are talking about the acid absorbing capacity of the malt. If you have a kg of malt A mashed in DI water and you add a mEq of acid (whether that acid comes from a bottle, sauergut or patent malt) it will bring about a pH shift. For Weyermann's regular pils, for example, it takes about 41 mEq of acid to change the pH of a kg by 1 pH unit. But for Weyermann's floor pils it takes only 31. You must account for this somehow. Many of the programs are not aware of this and use some fixed number derived from Kolbach's observation or something based on color etc. If you do this, and apparently you are doing something like it, you are no better, algorithmicaly, than any of the other spreadsheets out there and accuracy predictions of ±0.01 appear even less likely. If you want to realize accuracy in the ±0.02 region you pretty much have to tell the program the actual buffering capacities of the malts you are actually using on that brew day.

If I'm misunderstanding or unintentionally sidestepping your real question please let me know. You'll have to bear with me, I'm an electrical engineer and not a chemist.

I had assumed that you were familliar with the MBAA TQ paper I referenced in an earlier post. I'll send you a copy. An electrical engineer (with only Chem 101 and 102 in his past) ginned this up so you should have no trouble with it.

 

[ajdelange]

Thought I'd find the e-mail on the website but didn't so PM that to me.

And another thing I forgot to mention is that I often advise users of the popular spread sheets to use the trick of entering a lighter or darker color for the malt in order to fool the program into thinking that a malt is more or less acidic than the program's model thinks it is.

The fact that you advocate this tells me that you are using a color model for proton demand/surfeit.

 

[Big Monk]

Thought I'd find the e-mail on the website but didn't so PM that to me.

And another thing I forgot to mention is that I often advise users of the popular spread sheets to use the trick of entering a lighter or darker color for the malt in order to fool the program into thinking that a malt is more or less acidic than the program's model thinks it is.

The fact that you advocate this tells me that you are using a color model for proton demand/surfeit.

We are using a revised color model for this sheet. It has been working well but requires a little work to get things setup. Although I have had a planned revision on the docket for a while now, which will likely be accelerated by this conversation.

I'll PM you but I think I have all your stuff saved.

 

[Big Monk]

How do I select custom water and percentages of RO/custom?

If you unhide and Unprotect the "water" sheet, there is a place for a custom profile.

We use RO/Distilled exclusively so we don't have a mix percentage cell in the sheet.

 

[Silver_Is_Money]

If you unhide and Unprotect the "water" sheet, there is a place for a custom profile.

That worked. Thanks!!!

 

[Big Monk]

This was originally a purpose built custom sheet for me and my collaborator so it differs from some of the established softwares. I'm working on incorporating some thing in a new revision.