Link between RO+minerals and loss of yeast character?

[Silver_Is_Money]

OK, my recently bottled (2 weeks ago) Bohemian Pilsner still needs more bottle conditioning time to bring up the carbonation, but I can report that from my early sampling it is malty and deliciously flavorful, despite my use of only RO water and added minerals. Absolutely no lack of character here. I went heavy on the chloride and light on the sulfate. And I used Avangard Pilsner malt, which I can report is very good. Single infusion, mashed at 151 degrees for 60 min. 146 degrees at the end of the mash. Matches the orange color of Pilsner Urquell to a tee.

10 lbs. Avangard Pilsner
1 lb. Briess Veinna
8 Oz. Weyermann Melanoidin
6 Oz. Weyermann Acidulated

~7.6 gallons boiled, and ~6.1 gallons to the fermenter. Rehydrated W-34/70 yeast.

Each 5 gallons of high quality RO water (only 6 to 8 TDS per 'Clearwater Systems') received:
2.1 g. CaCl2
0.65 g. CaSO4
0.8 g. NaCl

Hops:
1/2 Oz. Magnum, 60 min boil
1 Oz. Sterling, 15 min boil
1 oz. Saaz, 5 min boil.

OG = 1.053
FG = 1.010

 

[user 246304]

I'm messing with Bru'n water, Silver. With 90% RO and an "amber dry" for my strong bitter, I have no need for any acid additions, and am getting an estimated mash pH of 5.32. I presume your salt additions above were also it, right - no acid addition?

Looks good, would love to see a pilsner full.

 

[Silver_Is_Money]

I'm messing with Bru'n water, Silver. With 90% RO and an "amber dry" for my strong bitter, I have no need for any acid additions, and am getting an estimated mash pH of 5.32. I presume your salt additions above were also it, right - no acid addition?

Looks good, would love to see a pilsner full.

I added 6 Oz. of acidulated malt to the mash, which is the nominal acid equivalent of adding about 4.75 mL of 88% lactic acid to the strike water.

 

[user 246304]

OK, thanks. What was your mash pH?

Edit: Silver, I'm asking, because acidifying at the same rate - or less, by a bit, 8.2 ml over 10.3 gallons of 88% lactic, I'm getting an estimated mash pH of 4.91. I know my salts are higher so need to compare apples to apples - just wondering, if starting with pure or close to pure RO and the use of mineral salts - do we really need acids?

Edit 2: For 10.3 gallons 90% RO mash, I add 7.2 g gypsum, 1.2 g CaCl, 0.05 epsom, 0.05 table salt. 2.6 ml 10% HCL (though don't need it), yielding mash pH of 5.29 with my grist (strong bitter, small amount of chocolate, debittered black; moderate in C120, 77).

Program has me sparging with some acid, but it's not necessary.

Anyway, it occurred to me - when we use pure RO, and add in Ca salts - why would we need any acid contribution at all?

 

[Silver_Is_Money]

OK, thanks. What was your mash pH?

I didn't measure it.

 

[user 246304]

OK, thanks. I'm going to run with this setup, to try it. I'm matching fairly closely Martin's "amber dry" profile, the obtained mineral figures seem reasonable and the mash pH estimation lines up so eager to try it.

Interesting recollection...when I brewed those years ago on a well with heavily carbonate water - I didn't know any better. I just boiled, used salts in the mash and phosphoric in the sparge, measured with a pH meter, without knowing the math behind what I was doing ("1.5 ml per gallon...") and the beer came out really well. Something like the Primer, maybe...brew, sensory evaluation, brew....

 

[Silver_Is_Money]

Anyway, it occurred to me - when we use pure RO, and add in Ca salts - why would we need any acid contribution at all?

Because Pilsner malt is quite basic with respect to the mash target of pH 5.4. European Pilsner malts can often be found to be between 5.8 and 6.0 in DI_Mash pH. Avangard specifies a minimum DI_Mash pH (Congress mash) of 5.8 for their Pilsner malt.

 

[user 246304]

OK, sure, makes sense. I think I was thinking that your Vienna and Melanoidin would have brought it within range, but obviously not. Appreciate your sharing, Silver, useful stuff.

 

[Silver_Is_Money]

OK, sure, makes sense. I think I was thinking that your Vienna and Melanoidin would have brought it within range, but obviously not. Appreciate your sharing, Silver, useful stuff.

Only the Melanoidin is acidic with respect to pH 5.4. And there are only 8 Oz of it vs. 11 Lbs. of the more basic malts.

 

[user 246304]

Only the Melanoidin is acidic with respect to pH 5.4. And there are only 8 Oz of it vs. 11 Lbs. of the more basic malts.

More to learn, because again I'm correlating any color at all with a drop in pH (hence, Vienna, being just a touch darker, etc....). This itself is another world to learn - malts and mash acidity.

 

[Silver_Is_Money]

Data I have from Briess indicates that (at least sometimes) their Vienna base malt is slightly more basic than their 2-Row Brewers base malt.

 

[user 246304]

I'm puzzled as to why...?

 

[Silver_Is_Money]

I'm puzzled as to why...?

Malt color (L) correlates poorly to DI_Mash pH.

 

[user 246304]

Yes, sorry, working on that one now. Jettisoning the notion of SRM and pH effect will take some doing, I'm afraid! AJ has been kind enough to provide some material and I've been reading over the last few days. Popular wisdom says, with highly alkaline water, when making stouts, porters, etc., don't worry about acidification - the mash pH will fall in fine. My experience from those years ago, IIRC correctly (very possibly, "no"), comports with that notion, more or less. I recall acidifying the sparge with phosphorus and using salts to some extent in the mash, but not for any adjustment to pH. Is this simply wrong, from a scientific point of view?

 

[Silver_Is_Money]

Whether or not to acidify or add base to strike water all comes down to the amount of acid in the malts that comprise the grist overall vs. the alkalinity and calcium/magnesium in the strike water, and the target mash pH. Deep roasted malts are acidic, but not nearly to the extent that caramel/crystal malts are acidic (color for color). I'm looking over Briess data as I type this, and their 120L Caramel malt is more acidic than roughly half of their appreciably darker deep roasted malts.

Alkalinity in the sparge water on the other hand virtually always benefits from acidification.

 

[Silver_Is_Money]

Here's another oddity. Deep roasted grains that are not malted are not as acidic (color for color) as deep roasted grains that are malted. No software that I'm aware of has attempted to address this anomaly to date. Including my own.

 

[user 246304]

Thanks Silver. Many old assumptions, need to dive in. Appreciate your explanations very much.

 

[Silver_Is_Money]

All of this makes me chuckle when people endlessly claim that by simply entering their grist components Lovibond colors and assigning them to a few nominal (and insufficient) malt classes they can make mash pH software hit measured actual mash pH virtually on the nose every time. This sort of reasoning must always be met with a great degree of skepticism. And I'm not excluding my own software here. Just trying to be as honest as I can be. Faith in software technology and confirmation bias in pH measurement and its interpretation are powerful forces to overcome.

 

[user 246304]

In other words, modeling of whatever sort using color (and/or malt classes, as well), is undependable. I'm guessing, grab your meter, brew, and adjust?

 

[Silver_Is_Money]

In other words, modeling of whatever sort using color (and/or malt classes, as well), is undependable. I'm guessing, grab your meter, brew, and adjust?

Adjust the next (essentially recipe identical) batch you mash based upon data gleaned from actual mash pH measurements of the current batch.

This is not to say that such software (mine included) will be useless for all applications. Provided that the user can confidently enter data properly and accurately for all required inputs it is certainly in most cases (though not 100% all cases) better than nothing. If it turns a 5.7 or higher pH mash into a mash that lands somewhere within the ballpark of a 5.2 to 5.6 pH mash it is better than nothing. If it turns a 5.1 pH or lower mash into a ballpark 5.2 to 5.6 pH mash it is better than nothing. And if it guides you into the avoidance of astringency due to its compensating for the alkalinity in sparge water it is potentially doing even more good for you overall than it may be accomplishing in the mash. Plus they are good tools for determining your mineralizations ppm with respect to the pertinent flavor and pH impacting ions.

 

[user 246304]

As I just told Martin, I appreciate that, Silver. Sorry, too, to both of you as I wasn't intentionally being an idiot and trying to slam your work, or Martin's work, in devising these programs.

You know what's crazy? 20 years ago, when I brewed on that farm, with my well and its carbonate water? Using boiled water (for pale), only Promash and, like I said, some suggested amount of phosphoric in the sparge; otherwise, yes, I was basically trying to emulate the Promash Yorkshire profile so I added in some gypsum and chloride. The crazy thing is that with rare exceptions, my targets - mostly concerned with mash pH and OG - were really, really tight. Just pulled an "ESB" session printout from 2001. Gypsum and chloride at ppm's in MLT and HLT, phosphoric in HLT to bring it to 5.7 (I always used 5.7). Intended OG, 1.060, and that's what I got, exactly. Mash for me was always 5.2-5.4 (I think I was also reading Kunze then, in addition to M&BS).

My point is, there's a lot of approaches and as usual I'm overcomplicating the hell (can I say that here?) of it. I do want to learn the science, really do. But I like the simplicity of the Primer, and also like the empiricism it engenders - which is what I did, basically, with that boiled water and phosphoric additions.

Anyway....

 

[Silver_Is_Money]

Here is but a single example of the problem at hand.

You buy 10 Lbs. of a very light colored base malt. The package says 1.8L. You are making a 6 gallon (to the fermenter) batch of SMaSH lager or ale. You decide to add 3.5 grams of Calcium Chloride beads to your strike water. You decide to target a mash pH of 5.4.

1) You have absolutely no clue as to the actual DI_Mash pH for your particular lot of light colored base malt. The actual DI_Mash pH of which can span from extremes of as low as about 5.5 to as high as perhaps 6.1, or anywhere inbetween.
2) You also have absolutely no clue as to the actual color of your base malt, which might perhaps be 1.2L or might be 2.4L or anywhere inbetween.
3) You have no clue as to the actual buffering characteristics of this specific lot of malt, so you don't know how much the minerals you decide to add to the strike water will actually reduce the pH of the mash.
4) Unless you are using distilled water you have no clue as to the actual alkalinity of your water, as it can change virtually from day to day (even for RO). If you use the value given to you by Ward Labs, it was precise for only the sample you sent to them.
5) You have no clue as to whether or not your calcium chloride came to you as anhydrous or in the dihydrate state, or if it has been opened to the air enough times in the past that it is now as much as perhaps quad-hydrate. So you have no idea as to how much actual Ca++ or Cl- is in your 3.5 grams.

Here's the rub: Since you have no clue as to any of the above variables, your chosen software doesn't either.

If you used distilled water (to eliminate major problem #4 above) your SMaSH's mash (strike water) may actually require (as extremes) anywhere from essentially no acid addition, to perhaps as much as 8 mL of 88% lactic acid to hit your 5.4 mash pH target on the nose. There is absolutely no way that you can state under these radical yet potentially very real extremes (which would be worse for any water source other than distilled, as even RO water has some degree of alkalinity variability and/or alkalinity unknown) that your chosen software will get you within 0.1 of your target mash pH every time.

 

[mabrungard]

There are so many variables at play in mashing and brewing, that its truly impossible for a prediction to be dependable and accurate. However, many brewing water programs can get you closer to a 'better' result. I like to say that brewing water software is like a hand on the brewer's shoulder to help and guide their treatment. If you use a pH meter or your taste preferences, you can then make studied and reasonable 'adjustments' to your previous batch's water treatment and more likely find success.

 

[ajdelange]

Here's another oddity. Deep roasted grains that are not malted are not as acidic (color for color) as deep roasted grains that are malted. No software that I'm aware of has attempted to address this anomaly to date. Including my own.

I thought they all did. In Kai's pioneering paper he separates the malts into 3 groups: base, colored and roasted and does fits to each (IIRC). It then only makes sense to estimate from color within the individual groups. An attempt to fit his entire data set at once would not give a meaningful result. As most software (including mine when I don't have measured titration data) relies on Kai's work I would assume that their author's would have picked up on this important point.

 

[Silver_Is_Money]

I thought they all did. In Kai's pioneering paper he separates the malts into 3 groups: base, colored and roasted and does fits to each (IIRC). It then only makes sense to estimate from color within the individual groups. An attempt to fit his entire data set at once would not give a meaningful result. As most software (including mine when I don't have measured titration data) relies on Kai's work I would assume that their author's would have picked up on this important point.

A.J., what I'm referring to here is restricted to deeply roasted malts in the generally 300L to 600L range. Unmalted barley within this darkness range has less acidity than does malted barley within this darkness range. And wheats in this L range are a complete unknown to me.

And, as I mentioned in one of my posts above, much lighter in color caramel or crystal class malts can and do often have more acidity than a good percentage of the various available deeply roasted (300L to 600L, give or take) malts.

My data is from a major maltster, and I have formulated my own best fit regressions (linear and non-linear) in an attempt to correlate (by Lovibond color) math models of the various malt classifications to hard DI_Mash pH data, rather than use Kai's earlier regressions that are based upon his own measurements. I'll admit that finding precision within the correlation is indeed difficult, and data culling (meaning bias on my part as to what data to eject from the regressions) is required.

 

[user 246304]

Thanks guys. One of the things I do marvel at, and always have since beginning this craft, is that any of us can take such widely variable, living things, and make something approximating a consistent product (if that's our bag). So Silver, I hear you. Martin, I appreciate the image of the brotherly guidance of a model while one proceeds.

I just combed through a few more sessions from 2000-2001, and they confirm my memory. All but one of them nailed the intended OG out to two decimal places (we could argue whether there's some reading bias here). They spanned OG's from 1.050 out to 1.085, a big Baltic porter. The outlier was intended to come in at 1.060, and I got 1.063. Unfortunately - and I don't recall when I brought in a pH meter, over papers - I don't have much notation among my brew sessions (these are all on ProMash) on mash pH or pH at other process points. The ones I did note were within 5.2-5.4, and that's my recollection generally. I actually can't think of any mash outside the optimal conversion range, but it's very likely, I'm sure.

Again, we drew from a very carbonate well water - I do recall that. And for pales, I did boil and naively treat with salts to try and copy a profile (usually, Yorkshire). And to the sparge, the salts and phosphoric (I think I may have begun with citric or lactic, but I know I eventually moved, and stayed, with phosphoric for sparge acidification).

So the big lacuna for me, the persisting question, is that across all this variability, I ended up with such consistency - and with such primitive tools? This is ABSOLUTELY not a leading question, I mean it. I think I was lucky then, that perhaps our water was so consistent that we had a rare thing going, and I was too naive to know it. I haven't yet brewed here in Madison, but I do know we draw from different wells, and change entirely from summer to winter, so Silver, yes, absolutely - the figures I use are pretty lousy, as they're just reported averages for our town and I've never done a Ward Labs with them.

Which is why I'll probably just go to 90% RO. I still hold on to a kind of romantic notion of trace minerals (thanks for that mention too, btw, Martin), like AJ, so can't let go entirely. If life were to open up its plenty and rain it down? It would be the country brewery, with a deep well where I could proudly say something along the lines of "....crystal clear Yorkshire Dales water...."

 

[ajdelange]

1) You have absolutely no clue as to the actual DI_Mash pH for your particular lot of light colored base malt. The actual DI_Mash pH of which can span from extremes of as low as about 5.5 to as high as perhaps 6.1, or anywhere inbetween.

It's a pretty simple matter to measure the DI mash pH and doesn't take much more effort to get the slope at the DI pH (just add a bit of acid to a second malt sample and measure the pH again). It probably makes more sense to compose your grist based on desired flavor characteristics, blend the malts in the proportions determined this way and then do these two tests on a small amount of the blend. This gives you the DI pH and buffering of the mash from which it is a simple matter to compute proton deficit or surfeit at a desired mash pH. Note that this is effectively doing the 'test mash' that we have always advocated.

2) You also have absolutely no clue as to the actual color of your base malt, which might perhaps be 1.2L or might be 2.4L or anywhere inbetween.

With pHDI and slope you don't need to know color. The color model is irrelevant.

3) You have no clue as to the actual buffering characteristics of this specific lot of malt, so you don't know how much the minerals you decide to add to the strike water will actually reduce the pH of the mash.

With the second measurement described under 1) you do know the buffering (or at least the linear term). If you don't want to do the second measurement just assume buffering is 40 mEq/kg•pH. Error in mash pH estimate is much more sensitive to pHDI than it is to buffering.

4) Unless you are using distilled water you have no clue as to the actual alkalinity of your water, as it can change virtually from day to day (even for RO). If you use the value given to you by Ward Labs, it was precise for only the sample you sent to them.

This difficulty is simply overcome by acidifying the water to the desired mash pH. You don't need to do this for RO as RO water's alkalinity is completely swamped by malt's.

5) You have no clue as to whether or not your calcium chloride came to you as anhydrous or in the dihydrate state, or if it has been opened to the air enough times in the past that it is now as much as perhaps quad-hydrate. So you have no idea as to how much actual Ca++ or Cl- is in your 3.5 grams.

It is a simple matter to accurately measure CaCl2•0H2O by making a solution and measuring its specific gravity. See the sticky on this subject.

So do some or all of these things and you should be able to get a pretty good pH estimate.

 

[user 246304]

And, as I mentioned in one of my posts above, much lighter in color caramel or crystal class malts can and do often have more acidity than a good percentage of the various available deeply roasted (300L to 600L, give or take) malts.

Why? I'm seeing the poor correlation now, but could you, Martin or AJ indicate some of why a more lightly colored malt might have more acidity? Is this about phosphate precip.?

This difficulty is simply overcome by acidifying the water to the desired mash pH. You don't need to do this for RO as RO water's alkalinity is completely swamped by malt's.

Getting a touch of it, AJ. There's no need for acid either in the mash or the sparge if using RO or DI water. However, if using one's own water - say, highly carbonate, properly boiled water - you wouldn't know what's left but you could certainly bring it to proper pH for mashing, by making up a cistern totalling the mashing and sparging water - or, 20 gallons, whatever - and acidify it to 5.2-5.4. Is that correct?

ps: Kind of keyed with the notion of doing test mashes. I just saw your discussion somewhere about this, and will look for it.

Trying to recall - deClerck's recommendation you mentioned, 3 samples. Is this what you're talking of here? I once owned his book(s), as well as Kunze's. I plan on getting both in the new year, as I miss what was in them.

 

[Silver_Is_Money]

In Kai's "Kaiser Water Calculator" he makes you enter the percentage of SRM color in your recipe that comes from deep roasted malts so his math modeling can reduce their color based acidifying impact upon the grist accordingly.

 

[ajdelange]

A.J., what I'm referring to here is restricted to deeply roasted malts in the generally 300L to 600L range. Unmalted barley within this darkness range has less acidity than does malted barley within this darkness range. And wheats in this L range are a complete unknown to me.

If you are looking for an explanation of this I haven't got one but I will venture a guess. The terminal -COH in a sugar can get oxidized to -COOH which converts the sugar to a carboxylic acid. As malted products have had their long starch molecules broken down into shorter sugar molecules there are more -COH ends available to be oxidized and this seems a reasonable, though not verified correct, explanation.

And, as I mentioned in one of my posts above, much lighter in color caramel or crystal class malts can and do often have more acidity than a good percentage of the various available deeply roasted (300L to 600L, give or take) malts.

That's recognized, starting with Kai's data, and that's what I was talking about.

My data is from a major maltster, and I have formulated my own best fit regressions (linear and non-linear) in an attempt to correlate (by Lovibond color) math models of the various malt classifications to hard DI_Mash pH data, rather than use Kai's earlier regressions that are based upon his own measurements. I'll admit that finding precision within the correlation is indeed difficult, and data culling (meaning bias on my part as to what data to eject from the regressions) is required.

If you haven't seen it (but I expect you have) check mu posts in another thread that try to illustrate the consequences of a 'good' (r = 0.88) fit.