pH Calculator Discrepencies

[troglodytes]

I've been doing a ton of reading on water over the past year, and now with a lot of help from these forums my beer is definitely better than it previously was. But now as I try to hone in acid additions in different garin bills and styles I am coming across some huge calculated pH discrepancies across different calculators.

I know the best thing I can do is get a pH meter, but it just isn't in the budget right now so sadly I'm just attempting to get the best estimate. I also know that different maltsters will have different DI water pH even when comparing 2-rows or ryes; however the discrepancies I'm seeing would, in my opinion, be the difference between an ok beer and a ruined one, so I just wanted to get an opinion.

Below is my grain bill and water adjustments from Bru'n water (what I always use to determine my additions). This is for a Rye IPA ( a fairly big one).

Now if I enter this exact data in Brewer's Friend and EZ Water I get the same results between those 2; however; they are very far off on pH from Bru'n water. Both BF and EZ say I need at a minimum 6 mL of lactic acid to get to 5.3 mash pH. This seems very high to me so I entered it into Bru'n water to see what I would get. According to Bru'n if I added 6 mL of lactic to my mash I would be at a mash pH of 4.8 which I think is pretty much unacceptable.

What would you do in this situation?

 

[MSK_Chess]

Go with Bru'n'water, its shown to be time and again the most accurate. You want to be adding the least amount of additives to achieve your mash pH anyway. 6ml? wow.

 

[troglodytes]

Ok thank you. I thought I was going crazy. 6 mL seemed like a lot of lactic acid for an IPA with a roasted malt addition.

 

[Silver_Is_Money]

To my knowledge Brewers Friend was developed with the assistance of Kai Troester, who within roughly the past decade actually performed a rather extensive series of his own lab experiments including titrations and pH monitoring and DI mash pH testing along with a multitude of various other tests across a broad range of various types and categories of grains and their mashes at various thicknesses in order to directly observe what was really going on therein before he developed his own modern equations to "best fit model" what he had directly observed. In a number of areas Kai's observations do not seem to match well with Bru'n Water. For but one example, his direct experimental observation was that calcium and magnesium mineral additions to the strike water (mash) do not lower mash pH to the extent that is seen in Bru'n Water. This leads me to believe that in this specific case direct observation does not sustain textbook expectations that date back to pre WWII. Perhaps all of the differences similarly boil down to the difference between modern real world measurement and observation vs. aging/classic brewing textbook equations, all of which can be openly found within the public domain. The nature of today's grains is also quite different from the grains that were available to the classic 1930's to 1950's era brewers who's formulas are now public domain, and this must also be carefully considered. If I'm understanding him correctly, A.J. DeLange has observed that modern grains buffer against pH alteration (shift) to a far greater extent than did the grains that existed a few generations ago (when the classic formulas now found within the public domain were generated).

That said, from what I have observed Brewers Friend best matches Kai's data only when used in its final batch SRM color mode. The free to download 'Kaiser Water Calculator' spreadsheet is Kai's direct spreadsheet offering, and it matches Brewers Friend only when BF is used in the SRM color mode, and not when the actual grist bill is entered in BF.

But there is hope! Brewers Friends output when in the grist bill mode is (in my opinion) substantially improved when DI Mash pH's are entered rather than Lovibond colors for each of your grist bills component malts/grains. This leads me to believe that the default set of DI mash pH's which are internally calculated from Lovibond color and grain classification type within BF are highly suspect of being way off.

Due to the old vs. new grains buffering issue alone I'm inclined to tentatively believe that the actual truth likely in most cases lies somewhere in-between the results witnessed in Brewers Friend and Bru'n Water.

 

[troglodytes]

So the two things that I can think of that would drive such a difference are:

1) How wheat and Rye are handles between the two models. I've heard both have a higher than 5.7 DI mash pH which would imply a higher mash pH in my brew than a subsequent one that uses only 2-row, vienna or similar.

2) My 29 ppm bicarb would buffer the first bit of lactic acid added to the mash, and I would need to enter more to actually bring the mash pH down.

This makes me think there is a grain of truth to me adding more than the planned 1.5 mL of lactic, but if I'm wrong and I do add 6 mL, I afraid it could be a pretty bad time for this brew, that I really don't want to ruin. I get a limited budget in both time and money for brewing and would really prefer to not ruin a batch.

Edit: Thanks Silver, I just saw your further explanation and when I enter the Brewer's Friend data based on color only, the results become a lot closer (within .05 pH to Bru'n). I'll trust those numbers then.

 

[Silver_Is_Money]

I have even further expanded upon my above post.

 

[troglodytes]

That expansion is something I didn't even consider. One day I will get a pH meter but as long as I see similar results when looking at the estimation based on srm, I feel comfortable.

 

[ajdelange]

If I'm understanding him correctly, A.J. DeLange has observed that modern grains buffer against pH alteration (shift) to a far greater extent than did the grains that existed a few generations ago (when the classic formulas now found within the public domain were generated).

I just want to be perfectly clear on this. The malts that I have measured and the malts that have been measured by my correspondent tend to have buffering capacities that exceed the single number published by Kolbach in his famous (to brewers) paper. This sort of implies higher buffering in modern malts but doesn't really prove that this is indeed the case.

 

[Silver_Is_Money]

I just want to be perfectly clear on this. The malts that I have measured and the malts that have been measured by my correspondent tend to have buffering capacities that exceed the single number published by Kolbach in his famous (to brewers) paper. This sort of implies higher buffering in modern malts but doesn't really prove that this is indeed the case.

Thank you for clarifying this A.J.! Would it be fair to state that it is most probable the real answer lies somewhere between 30 and 50? If you were to place probability bands (a +/- range of likelihood) upon the buffering value, what range would you consider as being viable?