Big Monk
Trappist Please! 🍷
- Joined
- Dec 24, 2015
- Messages
- 2,192
- Reaction score
- 1,152
I have been graciously gifted the Barth and Zaman paper first mentioned in post #186 above. Some amazingly interesting things emerge from it, but I must seriously question their choice to sample at only 5 minutes into the mash for each of their multiples of mash pH tests. There is indeed a wealth of interesting information to be pondered within this paper, and I will list a couple of their "shocking" findings here:
1) They discovered that for the specific case of measuring at 5 minutes into a single infusion mash the Kolbach "3.5" divisor for Ca++ is actually (on average, and rounded) ~14.8 for Pilsner, ~7.2 for Pale Ale and ~12.2 for Munich malts respectively.
2) They discovered that the multiple point titration "curve" for their Pilsner malt was essentially linear to within 5% precision across a rather broad titration range, and not highly exponentially curved as for the Pilsner malt titration chart seen in post #108 above. They admit that they were quite surprised to find the buffering to be so linear.
If, as per Kolbach, the 3.5 divisor for calcium ions emerges post the boil (which is where he actually measured and thereby determined it, presumably for a German Pilsner malt as my guess), and at 5 minutes into the mash this divisor is a (minuscule as to its pH shifting impact) 14.8, then I speculate intuitively (while cautiously admitting that intuition generally leads to bad science) that the impact of calcium ions upon the downward shift of pH is gradual, and only minimally noticed at only 5 minutes into the mash. This leads me to wonder what the divisors value would be at 15 min., 30 min., 45 min., and 60 min. into the mash.
If we linearly regress Pilsner malts 14.8 divisor at 5 minutes and 3.5 divisor at 120 minutes (presuming here a typical 60 min. mash and 60 min. boil) the simple linear equation that evolves is:
Ca_Divisor_Value = -0.0982609(Time_Min) + 15.2913
Solving linearly for the Ca++ ions divisor value at a few evolving times within the brewing process thereby gives us:
5 min. = 14.8
15 min. ~= 13.8
30 min. ~= 12.3
45 min. ~= 10.9
60 min. ~= 9.4
120 min. = 3.5
Disclaimer: It is purely speculation that this phenomenon is linear with respect to time, and it far and away most likely is not.
A few thoughts:
1.) I too am curious about the 5 minute measurement. Let's assume for a second that this paper represents a valid change in our thinking on the downward pH shift from Ca (they don't measure for the effect of Mg here but we can assume something comparable). Obviously, the larger the Ca content, the more drastic this drop in mineralization change is. Very curious findings and very interesting.
2.) Like I said in our PM, this should not be shocking (or "shocking" ). We can safely assume that the a2 and a3 titration values, if measured for most base malts, would be very small compared to their crystal, roast and Sauermalz counterparts. These values are really what sway the linear vs. polynomial characteristics of the malt.
With respect to the mineralization portion, I have what I think is a parallel in my industry (electric utility): Load Masking. Load masking is a phenomenon by which distribution feeder loading (residential, commercial, etc. customers) is "hidden" by the fact that we have a considerable amount of photovoltaic, wind, or waste gas generators on the circuit. It is important to us to "unmask" the "hidden" load in order to be able to do some of our important background calculations for transformer loading , etc.
What we may have here, particularly for batches produced with hefty source water calcium or added calcium, is a major overestimation of the downward pH shift from Ca (and assuming Mg, although we know the effect of Mg is much in general because of lower values in source water and lower additions of Mg salts). This could have the effect of "masking" other errors in a pH algorithm, especially is the modern denominators are much larger than derived by Kolbach.