pH of Maris Otter

[motobrewer]

I just recently brewed 10.5 gallons of Orfy's Mild using Warminster Floor-Malted Maris Otter. I use RO-DI water as a base.

78.3% Maris Otter
17.9% Simpsons Medium (67 SRM)
3.9% Simpsons Chocolate (430 SRM)

Bru'n water gave me the following additions (5 gallons mash water - 1.33 qt/lb mash thickness)

0.40 gram/gallon CaSO4 (2 grams in mash)
0.40 gram/gallon NaHCO3 (2 grams in mash)
0.40 gram/gallon CaCl2 (2 grams in mash)

for an estimated pH of 5.3. I cal'd my pH meter the day before at 65F.

I dough'd in at 128F for a protein rest, let it sit 10 minutes and force cooled a sample to 77F. Reading was 5.67. I added 3.6oz of acid malt (1.5% of grainbill), let it sit for 10 minutes and force cooled another sample to 85F and it was 5.24.

Is the reason for the discrepancy just an improper modeling of that Maris Otter and Simpsons crystal malts? Bru'n water is usually pretty accurate for me.

 

[ajdelange]

Is the reason for the discrepancy just an improper modeling of that Maris Otter and Simpsons crystal malts? Bru'n water is usually pretty accurate for me.

This is sort of like asking if the country is screwed up because the White House and Congress are populated by scoundrels. But to give a slightly more focused insight I can run your grain bill with the two versions of Maris Otter I have measurements on: Crisp and Muntons. With Crisp I get a predicted room temperature mash pH of 5.48 whereas with the Muntons I get 5.60 . Neither of these is very close to your estimate of 5.3. Nor are they that close to your observed 5.67 though I suppose 5.60 isn't too far off. The interesting thing about the two I measured are that their data indicates a mash pH of about 5.24 at 50 °C for both because they have quite different temperature glide coefficients. I think we can conclude that Maris Otter malts vary quite a bit. And, of course, other malts do too. That's why you can get lulled into thinking that you are getting reliable predictions when you happen to be using malts that the program does model fairly well and then get a wild prediction and big surprise when you move to a malt that it does not model well.

The other conclusion is that when a spreadsheet or calculator tells you to add alkalinity for a mash that is less than 20% colored malts be very skeptical. Do a test mash in such cases before you dump bicarb or lime into your mash or water.

 

[motobrewer]

This is sort of like asking if the country is screwed up because the White House and Congress are populated by scoundrels. But to give a slightly more focused insight I can run your grain bill with the two versions of Maris Otter I have measurements on: Crisp and Muntons. With Crisp I get a predicted room temperature mash pH of 5.48 whereas with the Muntons I get 5.60 . Neither of these is very close to your estimate of 5.3. Nor are they that close to your observed 5.67 though I suppose 5.60 isn't too far off. The interesting thing about the two I measured are that their data indicates a mash pH of about 5.24 at 50 °C for both because they have quite different temperature glide coefficients. I think we can conclude that Maris Otter malts vary quite a bit. And, of course, other malts do too. That's why you can get lulled into thinking that you are getting reliable predictions when you happen to be using malts that the program does model fairly well and then get a wild prediction and big surprise when you move to a malt that it does not model well.

The other conclusion is that when a spreadsheet or calculator tells you to add alkalinity for a mash that is less than 20% colored malts be very skeptical. Do a test mash in such cases before you dump bicarb or lime into your mash or water.

fair enough. i'll see if this action gets repeated with further mashes.

i did find it interesting that 1.5% addition of acid malt dropped pH by 0.43! This was some older acidulated that was crushed in september of 2014. does acid malt get more acidic as it sits?

 

[Hopper5000]

Did you recirculate your mash a little to make sure everything got mixed in correctly?

 

[motobrewer]

no recirc but stirred for 10 minutes

 

[qmax]

Malts vary, and no one goes to the trouble of titrating them.
At this point there is a lack of data, and all spreadsheets use some sort of approximations.
To complicate the matters even further, the same target pH at 20°C with different base malts may not be the same pH at mash temperatures.

 

[Hopper5000]

I see, I usually try to recirculate a half gallon because I pull samples from the ball valve and the false bottom has some volume loss. That being said, grains can vary in their buffering strength based on the batch from the company. 2 grams of NaHCO3 seems like a lot to me for that recipe.

 

[motobrewer]

next time i'll try recirc while doughing in

 

[Hopper5000]

It might not make a difference but it's wort trying.

 

[motobrewer]

It might not make a difference but it's wort trying.

Lol, get it?!?! WORT trying??

 

[ajdelange]

Malts vary, and no one goes to the trouble of titrating them.

I do and Kai Troester did but given the number we can do and the number of malts out there the population of titrators is, to a first approximation, 0. The next issue of MBAA TQ will contain an article summarizing mash pH prediction. It contains an appeal to maltsters to collect the titration data as part of their QC. I'm hopeful, but don't think it very likely that they will take this on.

That being said, grains can vary in their buffering strength based on the batch from the company.

That is certainly the assumption and certainly a reasonable one but I have never measured two batches from the same maltster and thus cannot back up that presumption with real data. OTOH the two samples of Maris Otter I have checked from different maltsters are very close to one another both in DI pH and buffering at 50 °C. The DI pH's are quite different from each other at 20 °C.

2 grams of NaHCO3 seems like a lot to me for that recipe.

The ROM calculations I did for this grist in #2 included the 2 grams of bicarbonate and gave 5.48. Eliminating it would drop the pH to 5.42, a shift of only 0.06 pH. The buffering of the model mash was -346 mEq/pH (approx - 50/kg•pH) and 2 grams of bicarbonate transitioned to pH 5.42 can absorb but 21 mEq so the small shift is to be expected.

 

[qmax]

-

 

[qmax]

The next issue of MBAA TQ will contain an article summarizing mash pH prediction. It contains an appeal to maltsters to collect the titration data as part of their QC. I'm hopeful, but don't think it very likely that they will take this on..

This is great news! I've never done it myself, but I think with automatic titrators available to malsters, it should be extremely simple to get all the data we could hope for.

 

[ajdelange]

Quite the contrary. It is very difficult relative to the sort of titration like the one used for alkalinity which is easily amenable to automation. In that one records the pH, adds, by computer control, a bit of acid, waits (with stirring), records pH, adds and increment of acid, records pH and loops until the end point pH is reached. If the manufacturer is savvy he will realize that he doesn't need to go to the end point pH but can, in fact, calculate the alkalinity at any pH from measurements at any 2 pH's neither of which has to be the end point pH. If he is wise and savvy he will process more but the point is that the process of adding the acid in increments can be done to the same sample. That is not the case with malt. A separate sample is needed for each increment of acid (or base) addition and pH values over time (at least half an hour) are required. Beyond that the analyst must, with a malt, measure the glide i.e. the change in DI pH with temperature. One could, of course, design a machine (large water bath, multiple pH electrodes, computer controlled acid, base and DI water dispensers) but AFAIK no such machine exists at this point in time.

Details on the method are at
http://www.wetnewf.org/pdfs/estimating-mash-ph.html

 

[qmax]

A separate sample is needed for each increment of acid (or base) addition and pH values over time (at least half an hour) are required. Beyond that the analyst must, with a malt, measure the glide i.e. the change in DI pH with temperature. One could, of course, design a machine (large water bath, multiple pH electrodes, computer controlled acid, base and DI water dispensers) but AFAIK no such machine exists at this point in time.

May it be possible to reach an equilibrium pH with a stirred pulverized mash after each acid addition and relate it to the unstirred standard grind mash at 20 minutes by introducing a coefficient? Or otherwise account for the time factor with superposition of acid additions.
I guess what I'm saying is there may be ways to simplify the procedure and make it possible to be done in one beaker at shorter time intervals, but more research needs to be done. Maybe it is impossible though...

Your article is quite large, as I understand you suggest this:
Take a number of beakers with DI water, add different amounts of acid to each, mash in, measure pH at 20 min, plot pH(mEq).
Sounds simple enough to me!

The effect of time is not a huge one in terms of absolute values, and it depends on the grind size, stir speed, etc. We probably do not want to control it to a degree where we need to know how much time the mash spent at each pH value over time.

Temperature, on the other hand, does have a significant effect on pH.
Not only do we need to know the DI pH glide, but also the effect on the entire titration curve including buffering capacity. Maybe this is not such a difficult task after all.

 

[ajdelange]

May it be possible to reach an equilibrium pH with a stirred pulverized mash after each acid addition and relate it to the unstirred standard grind mash at 20 minutes? Or otherwise account for the time factor with superposition of acid additions.

I don't think so as the behaviour over time is not the same for different malts. In some the 20 minute, 25 minute and 30 minute curves are pretty close to one another. In others they are not.

I guess what I'm saying is there may be ways to simplify the procedure and make it possible to be done in one beaker at shorter time intervals, but more research needs to be done.

When I got started in this I, naturally, began doing it as one would do any other titration. I added acid (or base) stirred, read the pH and added more acid or base. Something made me suspicious and I decided to add base, check pH and repeat a couple of times producing a titration curve and then add acid incrementally which should produce a curve that retraced the base detemined curve. It didn't! The method I evolved, while time consuming, overcomes that problem.

The effect of time is not a huge one in terms of absolute values, and it depends on the grind size, stir speed, etc. We probably do not want to control it to a degree where we need to know how much time the mash spent at each pH value over time.

What we want is to provide sufficient data such that a brewer can get a pretty good estimated based on a measurement taken at a time that he can be confident has him pretty close to equilibrium. My early thoughts were that the Taylor series coefficients should be published as a function of time. But the time history does depend on things like stirring and grind and so I've changed my thinking to that just the 20 minute or 25 minute set will do. Even though there is possibility of some change beyond that point it isn't much i.e. things are nearly at equilibrium. But in terms of what a malster making the measurement must do he must hang in there for at least 20 - 25 minutes and he must discard the sample for the next acid addition.

Temperature, on the other hand, does have a significant effect on pH.
Not only do we need to know the DI pH glide, but also the effect on the entire titration curve including buffering capacity. Maybe this is not such a difficult task after all.

This is definitely an area that requires more thought. If we have to obtain titration curves at several temperatures and include these in the model then the model becomes impractical. At first look things are a little scary. Presumably phosphate, as there is a lot of it in mash. It fact it is probably the predominant acid/base system has three pKs two of which are relevant at mash pH. The loss of the first proton by phosphoric acid is endothermic so its pK increases as temperature goes up. The loss of the second proton is exothermic so its pK goes down with increasing temperature (but not as much as pK1 goes up.) The titration curve looks like a staircase descending in the direction of higher pH with risers at approximately 2 and 7. As temperature increases the risers shift towards one another and the tread becomes narrower. That may seem scary at first but mash pH's are solidly in the middle of the tread. The difference in proton surfeit from 1 mmol of phosphoric acid at pH 5.4 at 20°C and 50°C is print Qphos(5.4,20) - Qphos(5.4,50) = 0.0011 mEq. That's not significant and the reason for it is that mash pH is distant from the pKs. If that holds true for the other orgainic acids and/or if phosphate is really the dominant system then we should be OK using the simple glide model. Of the ionizable side chains on amino acids histidine at pK = 6 is the closest to mash pH. Other possible organic acids such as citric and lactic have pH's that fall outside the range of interest. Malic acid has a pK at 5.2 but how much malic acid (if any) do we expect to find in malt?

A titrator like Hanna HI 901 looks really nice. It has an adjustable automatic dispenser, a stirrer, can measure pH at up to 105°C..,

As discussed earlier standard titration practices don't work here.

..and export the plot.

Yes, to a standard 3.5" floppy disk! Where would I get one of those? A museum?

 

[qmax]

Something made me suspicious and I decided to add base, check pH and repeat a couple of times producing a titration curve and then add acid incrementally which should produce a curve that retraced the base detemined curve. It didn't!
brewer can get a pretty good estimated based on a measurement taken at a time that he can be confident has him pretty close to equilibrium.

That's because it never reaches equilibrium in the time frame of the mash.
But if given fine enough crush and fast stirring, it may. Then the curve should retrace itself and a standard titration should work; but will it be any useful for a normal mash with a coarser crush and no stirring?

This is definitely an area that requires more thought. If we have to obtain titration curves at several temperatures and include these in the model then the model becomes impractical. At first look things are a little scary.

If the glide model works, great. If not, maybe it will be possible to fit a 3D function Malt_Alkalinity(target_pH, temperature) or create a temperature coefficient for each of a1,a2,a3,glide, although this would make things really bulky.

The model will still be practical, I'll just add to the output the resulting pH's at specific rests or plot out mash_pH(temperature).

Yes, to a standard 3.5" floppy disk! Where would I get one of those? A museum?

Haha! That's something you would not expect from $6k piece of equipment.

 

[motobrewer]

I've brewed with this particular MO malt again (this past sunday) and again found that Bru'n Water predicted mash pH of 0.20 low, ie it predicted 5.1 and my measured pH was 5.30.

 

[mchrispen]

I have noted quite a bit of variation in acidity between malt vendors. Most of that delta, perhaps not all, is with the base malt. I would adjust the Lovibond of the MO until the predicted mash pH is close to your measurement, and then conduct a test mash. A small tweak to the Simpsons Medium crystal might be needed as well. I also tend to use the lower number from malt analysis sheets for input in BWS as a starting point.