Mash pH adjustment for an English Brown Ale

[vinpaysdoc]

I'm currently putting the numbers into Bru'n Water for the BCS Southern English Brown recipe. If I put all of the crystal and roast malts in the mash I end up with a screamingly low predicted pH. I have very soft water and have thus far only had to use gypsum, CaCl2, epsom, and occasional baking soda to work out my pH on paper for most of my ales. When I select 'Brown Malty' for the target water profile I don't seem to be able to use enough baking soda to bring the pH up to an acceptable range for this recipe.

How would be the best way to go about achieving the correct mash pH?

1. Leave the roast malts out of the mash? (gets me close with some bicarb)
2. Leave the roast and crystal malts out of the mash?
3. Use Pickling Lime to adjust the pH (I've not yet had to use it)?

Also, if you steep the roast and/or crystal malts at the end of the mash would that not affect the pH into the kettle? Is that of concern?

I initially adjusted the crystal malts in the recipe to lighter crystal malts so that I could get the pH to work out. Is this a decent option?

TL-DR - Among all the variables you can control, what is the best way to handle a recipe in which the grain bill results in a very low calculated mash pH?

Thanks

 

[mchrispen]

How would be the best way to go about achieving the correct mash pH?

So this is really an issue of preference, not a best or recommended method. Some breweries keep the dark malts from the mash - steep them separately and add into the boil. Martin discusses this as a method used by Guiness in a recent article. Others would mash it all together and use some alkaline addition to balance the mash pH.

Pickling lime is my preferred alkaline addition as it doesn't seem to add any flavors (baking soda adds sodium), dissolves easily and gives a very quick mash pH response. I personally would use this, and withhold a portion of the roasted malt for a separate steep. Then add any remaining additions into the sparge, without any acidification in the sparge.

Something I am exploring is cold steeping my darkest malts to remove the harsh bitterness. I have some now (Black Patent) in the fridge and it tastes surprisingly like coffee with little of the acrid/burned character I have gotten in the mash. I plan to pour the whole thing into the mash just before sparging.

 

[ajdelange]

If you post your water chemistry and the recipe we can put together a spreadsheet that shows us which mash component is doing what which may suggest some things to do. All the things you have suggested are possibilities. Getting a workable solution will depend on us having detailed and accurate information on each of the malts which we won't have but at least we can probably get an idea as to what the long pole in the tent may be.

 

[vinpaysdoc]

Existing Water Profile

Na - 15
Sulfate - 22
Bicarb - 32
Ca - 0
Mg - 0
Cl - 0

5 gal batch grain bill

Maris Otter 6 lb 15 oz
Crystal 80 1 lb
Crystal 120 10 oz
Special Roast (50 L) 8 oz
Pale Chocolate (180 L) 6 oz
Carafa II (430 L) 4 oz

From BCS Southern English Brown

My equipment is a 10 gal cylindrical cooler for a mash tun and a 10 gal aluminum pot for the boil.

Thanks for your help!

 

[ajdelange]

I can work with those but do you happen to know alkalinity and pH? And where does this water come from??

 

[vinpaysdoc]

Alkalinity - 26
pH - 7.88

 

[ajdelange]

OK. If I make models for those which are, admittedly, crude but based on some measurements of malts I've done myself and some of Kai Troester's measurements I can come up with a pH estimate. I assumed you mashed the grains with 4 gallons of the specified water. The predicted mash pH would be 5.34. The proton deficits, all with respect to that pH are as follows:

Water: Low alkalinity of 26 ppm as CaCO3 implies 7.32 mEq protons needed to hit 5.34
Base Malt: MO 6.94 pounds modeled as having DI pH of 5.62 and buffering of -40.7 mEq/kg at DI pH would require 39.87 mEq protons to get to pH 5.34. This is the long pole in the tent and the major determinant of mash pH. Were the DI pH of the base malt 5.52 (0.1 pH lower) the overall mash pH would shift down 0.07 pH to 5.278.
Chocolate malts: 6 Oz of 180 L chocolate and 4 Oz 230 L Carafa. Modeled as 600L chocolate with DI mash pH 4.7 and buffering 76.49 mEq/kg at DI pH. This is probably conservative on the acid side. These malts are estimated to supply 12.80 mEq protons in going to pH 5.34.
Crystal 80 Malt: 1 #. Modeled as an 80L caramel malt I have detailed data on. DI pH 4.766, buffering 89.7 mEq/kg. This malt would supply 17.67 mEq of protons. This is the major 'acid' in the mix.
Crystal 120: 10 Oz. From Kai's data. Modeled as DI pH 4.75; buffering -48.4 mEq/kg. Supplies 7.25 mEq protons.
Special Roast 50L: Interpolated from Kai's data. Modeled as DI pH 4.84, buffering -44 kg/kg. Protons supplied 4.37 mEq.

You can certainly argue about the models for any of these malts as not a single one of them represents data taken on the malts you are actually using but they shouldn't be that far out of whack and thus I don't see how you get 'screamingly low' predicted pH. IOW based on this analysis, you should be OK - unless I happen to be off in the high pH direction on everything. This is definitely a case where a test mash would be of benefit and really the only way to put your mind at ease.

I wouldn't want to advise taking steps to raise pH without pH meter verification that it is necessary but the following actions would do that. Adding 1 mmol of sodium bicarbonate to each liter of water would increase the sodium content by 23 mg/L and increase the estimated pH to 5.404. Adding a mmol of Ca(OH)2 (no bicarbonate) to each liter would increase it to 5.43. Halving the amount of 80L crystal would raise it to 5.39.

 

[mchrispen]

For the sake of a counter point, and not a critique of AJ's advise.

That recipe in Bru'n Water with your water profile and 4 gallons mash is predicted to a Room-Temp Mash pH of 4.9. (no minerals added). With a separate steep of the roasts, I get 5.1. Some alkalinity will need to be added regardless. I see where you might be concerned about very low pH, but if you look at the profile it does ask for ~80 ppm bicarbonate, expecting some moderate alkalinity in the water - your's is quite low so expect to add some pickling lime or baking soda.

Using the targeted Brown Malty profile, I get the following to hit 5.2 Room Temp Mash pH:

0.8 grams Gypsum
1.8 grams Calcium Chloride
1.8 grams Pickling Lime

- above added to mash - this overshoots the calcium load by a bit vs the profile, but results in the minimum pH, and relatively close sulfate and chloride levels, with minimal sodium additions. You could swap in some Epsom for the Gyspum and reduce the Ca a little bit if you wanted a bit of magnesium in the beer.

an alternative and perhaps more true to the profile would be:

0.3 grams Gypsum
0.8 grams Epsom
2.0 grams Calcium Chloride
1.2 grams Pickling Lime

Based on this - I would perhaps cold steep either the Pale Chocolate or Special Roast (or both) separately and add that at the end of the mash for sparge. This will impact the flavor somewhat - so you need to consider that. Removing them from the mash will increase the mash pH slightly to a safer 5.3. I personally would only steep them separately if I thought that much lime might be problematic, as would that much baking soda. And would check mash pH throughout the mash process.

Measure the lime separately and be careful adding the pickling lime into the mash even with a pH meter, it seems to react quickly, overshoot slightly then settle into the target over a few minutes. Meaning I might see a 5.0 mash pH, correct with lime, and see the pH rise as high as 5.4 quickly, but settle closer at 5.3 in the next few minutes (assuming I used the correct amount).

 

[vinpaysdoc]

Matt's numbers below are exactly what I was getting.

 

[vinpaysdoc]

Thanks Matt. I have yet to purchase a pH meter. It's next on the list. I guess it's time to look into pickling lime. I already ordered lighter crystal malts for this time around. I will probably have a go at using the pickling lime next time around.

 

[mchrispen]

Look at your grocery for pickling lime by the canning supplies. It's cheap and will last you for years... Unless you also make lots of pickles!

Good luck to you!

 

[ajdelange]

For the sake of a counter point, and not a critique of AJ's advise.

Please don't worry about that! There is a huge discrepancy here and I want to get to the bottom of it.

That recipe in Bru'n Water with your water profile and 4 gallons mash is predicted to a Room-Temp Mash pH of 4.9. (no minerals added).

I've been over and over this and don't see how you could possibly get something that low with this grain bill and water. You'd need 124 mEq of acid (1.3 mL lactic acid) to move those as I modeled them to pH 4.9. Not that I am not claiming that my models are that great but I can't be that far off as the bulk of the grains here are quite similar to ones on which I did extensive and careful measurement.

I have to be missing something. I put

Maris Otter 6 lb 15 oz
Crystal 80 1 lb
Crystal 120 10 oz
Special Roast (50 L) 8 oz
Pale Chocolate (180 L) 6 oz
Carafa II (430 L) 4 oz

with 4 gal of water with alkalinity 26 ppm as CaCO3 into the EZ spreadheet and it predicts 5.52 (which is probably too high but then EZ assumes MO DI pH of 5.77. Just lowering that to 5.62 would drop the estimate to 5.4 which is more in line with what my model predicts but still way high WRT what 2 of you are getting from Bru'n.

But then if you think about it a minute and look at Troester's data you'll see that his highly colored malts have DI pH's in the high 4's. If mash pH is only 0.1 or 0.2 higher than this these malts don't contribute much acidity and the alkalinity of the base malts go up. At pH 4.9 the 6.9 lbs of base malt, as I modeled it, has a deficit of 130 mEq - almost all of which is taken up by the 124 mEq of strong acid I said I'd have to add. The colored malts only contribute 12 mEq. Thus I conclude that 4.9 simply cannot be close to the actual mash pH for this malt and water. It doesn't pass the common sense test. Either someone has slipped a decimal point or there is some acid addition that I have missed.

 

[vinpaysdoc]

Perhaps Martin will chime in at some point. If I missed a decimal point, I've done it several times and Matt seems to have as well. No, there must be some discrepancy with how Bru'n Water comes up with the projected pH and how you are doing so. As I recall, Martin's program uses the Lovibond rating to calculate the acid load of each grain.

Now, if I only had a pH meter we could settle this......

 

[mchrispen]

I reran the numbers today and got a differing result - between 5.0 and 5.1 without additions. So I entered something incorrectly... however with the additions I recommended, arrived at the same 5.3 prediction.

Regardless, you are going to need some alkalinity in that beer. FYI, If I dropped out the crystal as opposed to the roasts as a separate steep, the mash jumped to 5.4/5.5 with the additions. Something to consider looking at. I still have a lot of faith in the spreadsheet - it has been pretty accurate to date for me.

 

[ajdelange]

Well I tried the same numbers in Brewer's Friend and got 5.43. Note that its prediction of basemalt DI pH is 5.7 and that my model used 5.62. Decreasing the DI pH to 5.62 (which I can do in Brewer's Friend by increasing the color) would lower the pH estimate by about 0.07 to 5.36 which is awfully close to the 5.34 which I'm predicting. Remember that certainly Brewer's Friend relies on Kai's titration data and that, with the exception of the base, 80L and chocolate malts in my model which are based on my measurements are also based on Kai's data and I thought Bru'n Water was also based on Kai's measurements. Kai's calculator definitely uses color to predict malt properties and, where I don't have data from my own or Kai's measurements I interpolate using color.

 

[mabrungard]

I'm curious how the error was incurred here. Using the OP's reported 26 ppm alkalinity, a mashing water volume of 4 gal, and that grain bill, I show a mash pH of 5.16. That is what I would expect for what is in essence a porter grain bill with a LOT of crystal malt.

Minor additions of alkalinity are very much required to neutralize the malt acidity since the tap water provides little. It appears that 1.2 g of lime and CaCl2, and 0.2 g of gypsum brings the mash pH to around 5.4.

These results don't surprise me, but I'm really wondering how there have been such vacillations in the reported results.

 

[ajdelange]

I still think 5.16 is a bit low but at least now we are in the range where reasonable variation between models is a viable explanation. Someone has fat fingers!

 

[vinpaysdoc]

Mashing with 1.25 qt/lb (roughly 3.05 gal) is where I get the 4.9 pH associated with this grain bill. I'm using 3.5 L Maris Otter. The decreased water accounts for a good bit of the difference.

I substituted Crystal 10 for both Crystals in the recipe and found it much easier to balance the water profile. I guess I'll know if that was OK or if I should use the pickling lime next time in about 4 weeks. The OG sample tasted pretty good.

 

[mchrispen]

Willing sign up for user error on my part. Ran everything a third time using the subscriber version and got the second result with 5.1.

I just wish I could replicate the previous result to determine what I did wrong.

 

[vinpaysdoc]

You may have used a lower water volume. Mine came up with 4.9 when I was using 1.25 qt/lb. I increased the water/grain ratio to 2 qt/lb and that helped. I was able to get the mash to 5.4 if I kept the roasted grains out of the mash.

I also was thinking that I needed to mash somewhere between 1 and 1.25 qt/lb. I'm not sure where I got that in my head, but, re-reading 'How to Brew', it looks like I can take the mash water to grain ratio up to 2 qt/lb. That gives me a lot more room to work with the water chemistry. Sorry for the noob error guys. It was instructive.

 

[ajdelange]

Mashing with 1.25 qt/lb (roughly 3.05 gal) is where I get the 4.9 pH associated with this grain bill. I'm using 3.5 L Maris Otter. The decreased water accounts for a good bit of the difference.

That's a clue. There is so little alkalinity that the volume of water shouldn't make much difference. The water's deficit is about 2 mEq/gal. That's compared to a deficit of 41 mEq for the base malt and surfeit of about equal magnitude for the colored malts. Taking out or adding a gallon of water should shift the pH by 0.01 pH.

 

[ajdelange]

In looking at the numbers again I am beginning to think I may understand the low pH estimates here. When Kai did his pioneering measurements (upon which I think, but am not certain, Bru'n Water is based) he assumed that buffering is independent of pH. It isn't. Looking at the data set in his paper the average buffering of the whole set is -41 mEq/kg-pH and only one malt shows a buffering capacity over 50. For the malts I measured the buffering is a function of pH at at pH 5.4 is 48 for Pils, 81 for chocolate and 61 for caramel. In the better model the dark malts will be more powerful as acids than Kai's model suggests but, more significantly, the base malt is a more powerful base. Because of the greater pH difference between malt DI pH and dark malt DI pH the base malt dominates and the pH calculates higher.

 

[mabrungard]

Bru'n Water only uses Kai's grain data to formulate a correlation to the equivalents of acidity created for various grains. The rest of the mash pH prediction is based on the empirical results of myself and a select group of brewers that I trust.

To give a data point with which to check other models; I brewed the Reaper's Mild recipe which is known on this forum. It turns out to be one of the most acidic grists I've ever encountered in brewing. You will notice that it has a lot of crystal malt.

75% Marris Otter 4L,
20% Crystal 60L, and
5% Pale Chocolate 265L.

The mash thickness was 1.4 qts/lb and Ca:16, Mg: 5, and HCO3: 16 ppm.

The observed mash pH was 5.00. Following that reading, lime was added to raise the mash to 59 ppm Ca and the pH to 5.27. The Bru'n Water prediction for the mash without the lime was 5.1 and 5.4 with the lime addition. If anything, Bru'n Water tends to point a brewer to using a little less alkalinity since I am a firm believer that a beer is more likely to be better at a slightly low mash pH than a slightly high pH.

A similar brew was repeated a few months later and the low pH was again duplicated and resolved with a lime addition. I'm hoping AJ's model also duplicates this prediction. I know that the Brewer's Friend model did not match this result and I concluded it was flawed. I'm not sure if Kai subsequently revised his model. If AJ's model does not predict this result, I'm afraid his model is also suspect. The good thing is that most brewing IS NOT performed with such extreme grists and most models tend to agree for more typical mashing conditions. It's the very low pH range where it seems some models may falter.

 

[ajdelange]

The observed mash pH was 5.00.

I'm sure you saw what you saw but this stretches credibility a bit. It's asking 1/4 of the grist in transitioning 0.3 - 0.4 pH to pull 3 times as much mass 0.6 pH. I'm not saying it can't happen but it certainly is attention getting. Looking at the base malt I measured in detail you would need 251 mEq to move 7.5 kg of that malt (DI pH 5.62 - note that this is actually lower than what is often reported for MO and lower than many other lots of Weyermanns Pils I have worked with) to pH 5. That's equivalent to 2.6mL of 88% lactic acid (at 1.4 qts/lb we're talking 7.9 gal water here). Half a kg of the 600L malt I measured can only supply about 12 of those and 2.5 kg of the 80L caramel I measured can only supply 39. I have no reason to think there is anything unusual about these three malts as the data more or less agree with what Kai found for similar malts with the exception of the fact that they exhibit somewhat higher buffering capacities than his data because of the way he measured buffering. There is also some variation beacuse there is variation in malts. I already have observed that the Weyermanns Pils I measured is atypically low in DI mash pH.

It's interesting to see what sort of a malt might be needed to produce a mash pH of 5.0 at 20% presence with this base malt and 5% chocolate. As noted Kai's method underestimates buffering (the highest buffering he reports is -60 mEq/kg-pH) so I used -90 which is closer to what I observed. Note that acidity is the product of buffering and pH shift so the lower the DI pH and the higher the buffering (magnitude) the more acid the malt is. A malt that makes up 1/5 of the grist with (linear) buffering capacity of -90 mEq/kg-pH would need a DI mash pH of 3.67. The lowest DI pH measured by Kai was 4.48 so you'd need a malt appreciably more acidic that that. The only thing that comes close to meeting these numbers that I know of is Sauermalz (DI pH 3.44 according to Kai).

A similar brew was repeated a few months later and the low pH was again duplicated and resolved with a lime addition. I'm hoping AJ's model also duplicates this prediction.

It does match (est 5.1) if I use caramel malt with DI pH of 4.08 and buffering of -90 mEq/pH but of course I have no idea what the numbers for any of the malts you actually used in these brews were. If I were to put together a mash in those proportions using malts for which I do have data I would get a more reasonable pH of around 5.31. If I put this data into Brewer's Friend it comes up with a prediction of 5.41 Note that his estimate for the buffering capacity is -34 whereas the more detailed model gives -56 and that he predicts DI mash pH for the base malt as 5.66 whereas the malt I measured had a DI mash pH of 5.62.

I know that the Brewer's Friend model did not match this result and I concluded it was flawed.

It doesn't but I'm not sure that is the right conclusion.

f AJ's model does not predict this result, I'm afraid his model is also suspect.

All models are suspect. That's why we work so hard to validate and refine them. The model I am using is certainly robust WRT to the underlying science (it uses the chemist's 'proton condition') but it is not robust in the sense that I don't have the data on any of the malts a particular brewer is actually using. And of course this is a new model (proton condition certainly isn't new but application of it to brewing is to me). Surely someone else must have thought of it before I did.

The good thing is that most brewing IS NOT performed with such extreme grists and most models tend to agree for more typical mashing conditions. It's the very low pH range where it seems some models may falter.

There is no reason why they should. A purely empirical model may get a bit shaky in regions where the data is thin (e.g. low pH mashes) but if one has a robust model for an acidic malt there is no reason that the use of as much of it as desired can't be used to predict very (relatively) acid mash pH. Unless we are suggesting that activity is increasing faster than hydrogen ion concentration and I don't think we are anywhere near that even at low mash pH.

I forgot to mention the upward drift of mash pH over time. It takes time for water to penetrate the grist and time for the grist to release its solubles into the water. That's why, technically, we should always ask not only about the temperature at which mash pH was measured but when. The time drift is especially noticeable when one of the malts is acidic as is apparently the case here. I wonder if that might be responsible for the low measured pH.

 

[ajdelange]

The obvious thing for me to do here is a test mash but the pump in my DI water machine was kaput and so couldn't until I put a replacement in today. As noted in my last post the predicted pH for the malts I have data on was 5.31. In the test mash I did I used those three malts in the 75/20/5% ratio but I had 2 mVal alkalinity and 2 mVa each Ca and Mg hardness as well. These raise (RA = 57) the estimated pH to 5.36. Effective water to grist ratio was 1.2 qts/lb. The pH of the test mash measured at 48 °C but corrected to room temperature were 5.46, 5.46 and 5.47 at, respectively, 20, 25 and 30 min and, when the mash was cooled to room temperature and measured at room temperature, 5.46. Hot measurements were made against a hot buffer calibration and room temperature measurements against a room temperature buffer calibration. So I'm low by 0.1 pH which I can't really explain except that it could be aging of the malts (the titrations were done about 9 months ago). I wish I had done this experiment back then. It was hard to measure oout the tiny amounts of salts necessary to to prepare 0.1 L of mash water but as noted above the difference between 57 RA and 0 RA is only 0.05 pH and I wasn't off by that much.Is 0.1 pH as close as we can hope to get in a pH prediction even with careful malt measurement? I fear that may be the case with things like ageing, lot to lot and maltings to maltings variations being the killer. Can't draw too many conclusions from a single experiment, though.

The burning question remains as to how Martin could see pH 4.9 for a similar mash? And more than once. He has to have had a base malt with much lower DI pH and much lower buffering capacity and dark malts with much lower DI pH and much, much higher buffering capacity or both.

 

[ajdelange]

So I'm low by 0.1 pH which I can't really explain except that it could be aging of the malts...

Actually I can explain it and it isn't aging. It came to me this morning at 3:30AM when the dog decided he had to go out. The malt I used to make the test mash wasn't the malt I measured. I remembered that I had used the rest of the bag I tested in a brew and had opened a new sack with the unused portion being what was in the Vittles Vault from which I grabbed malt for the test mash. First thing this morning I checked the pH of a cold DI water mash with the new Hach meter: 5.92! I don't thing I've ever measured a malt's DI pH that high. Had to go off to a meeting and when I came back I did a hot water DI mash on that malt and got 5.91 (referred to room temp). I'm thinking I have two batches of Weyermann's Pils one that comes in at pH 5.62 and the other at 5.91 (both referred to room temperature). Indeed I do but the high pH one is the floor malted stuff. Not only does it have a much higher DI pH but it's buffering is about 70% of what the pneumatically malted Weyermann Pils shows. Put the right malt model into the spreadsheet and the prediction rises to 5.449. What I observed at mash temperature referred to room temperature yesterday was 5.457, 5.462 and 5.466 (the 3rd decimal place isn't real but I can calculate it and do so that I can see small differences in relative pH) at 20, 25 and 30 minutes respectively and a sample cooled to room temperature measured 5.46.

So the model is definitely vindicated though we can't draw global conclusions from this one test.

 

[mchrispen]

I can't really explain except that it could be aging of the malts..

AJ, any thoughts on this process as malts are stored? I wonder if environmental moisture impacts DI pH over time with a given batch as well. Theory being that bacteria *might* increase population as grain hydrates in storage. I realize it may be minimal, and I am not suggesting grain that spoils from too much moisture. I routinely freeze bags of base malt to kill any weevils, and I am fairly sure that process introduces at least some amount of moisture in the process.

I am wondering to what extend, compounded through packaging, distribution, storage at LHBS, repackaging (possible grind) may have relative to a given DI pH... could it be accounted for, or would it require a specific test mash before each brew (in which case, I assume we would have a true full recipe, not just a single malt)?

 

[ajdelange]

In this case, at least, it turned out not to be aging but misidentification. The caramel and chocolate were measured back in March, the floor Pils yesterday. So unless one of the colored malts drifted in one direction and the other in the opposite direction age was not a factor here. I really have no feel for how much aging, within reason, might change DI pH (or buffering) except for the gut feeling that properly stored (Vittles vault, sealed package under N2 etc) it probably isn't much. Now moisture pickup is a different matter. If I measure 50 mEq/kg-pH on a malt that is an 'as is' number. If the moisture content is 2% then the dry basis buffering is 49 mEq/kg-pH and if I try to use the number measured as is on 2% malt with malt with a moisture content of 5% obviously there will be error.

Clearly if one has to do the elaborate measurements required to characterize a malt before each brew there will be little value to this method. There will be value of we find that it is relatively robust WRT malt characteristics. It is encouraging that I only got 0.1 pH error by using a dramatically different base malt in the first estimate I posted for this mash. I hope we will find that if we have, for example, data for a representative 80L caramel malt that will be good enough to represent 80L caramel malt in general. If that does not turn out to be the case then a test mash with the actual grains to be brewed with will be necessary.