Matching an IPA Water Profile - Questions

I am getting ready to brew an IPA this weekend, and was planning to try to match the water profile of one of my all-time favorite IPAs. The profile is as follows (in ppm):

Ca - 163
Mg - 8.5
Na - 21
SO4 - 365
Cl - 23.5
HCO3 - 104

At this time, I do not have a pH meter. My city water is awful for brewing, so I use 100% distilled water.

Here are some questions I have:

1) In order to hit the HCO3 target, I would need to add 0.3g per gallon of baking soda to keep the Na on target while gaining some bicarbonate. Even then, the bicarbonate would be ~60ppm. Should I even be concentrating on the HCO3 target, or just focus on Ca, Cl, and SO4?

2) To the mash, I will be adding 10g Gypsum, 0.8g CaCL, and 1.2g Baking Soda. Per Bru'n Water, this puts my estimated mash pH around 5.2. If I enter the same stuff into EZ Water, it puts my estimated pH at 5.45 ONLY if I add 3oz of Acid Malt. I do not have any acid malt entered into Bru'n water.

Any ideas as to why there is such a discrepancy between the two? Is there anyone with a pH meter out there that can tell which of the two they have found to be more accurate with their mash estimates?


As you can tell, I am not a scientist. Any help or input would be greatly appreciated!

Science is hard.

I have nothing to offer except

I think, from what I have read and heard from others, EZ water becomes more inaccurate on extreme ends of the spectrum (like what you are doing). I would trust the Bru'n Water Projections.

FWIW - seems like an awful lot of additions. Out of curiosity, how do you know for sure that this is the exact water profile that is actually used to make your favorite beer? I tried to add a lot of gypsum and other additions once to mimic a profile and it was the worst beer I have brewed in the last 3 years. Tasted like my water additions.

Braufessor said:

I think, from what I have read and heard from others, EZ water becomes more inaccurate on extreme ends of the spectrum (like what you are doing). I would trust the Bru'n Water Projections.

FWIW - seems like an awful lot of additions. Out of curiosity, how do you know for sure that this is the exact water profile that is actually used to make your favorite beer? I tried to add a lot of gypsum and other additions once to mimic a profile and it was the worst beer I have brewed in the last 3 years. Tasted like my water additions.

Click to expand...

This profile is for Fat Head's Head Hunter IPA. The head brewer at Fat Head's gave the entire recipe - including water profile - to Mitch Steele for his IPA book.

I was leaning towards the Bru'n Water side of things on this one, becase the grain bill does not mention any acid malt. I would have to add acid malt, per EZ Water. I know there are many holes in that logic, but that's what my simple mind is going with!

It's definitely a lot of additions though. Unfortunately, thems the breaks when you use 100% distilled water.

Clint, you should buy a 55 gallon drum & fill up at one of the aaco member's homes. Great water profile for brewing just about anything. Good Jay has a copy of the report.

rvklein said:

Clint, you should buy a 55 gallon drum & fill up at one of the aaco member's homes. Great water profile for brewing just about anything. Good Jay has a copy of the report.

Click to expand...

I've actually discussed this with Nathan before. At the very least, sanitize and fill 2 carboys to use.

As a matter of fact, I may be dropping Nathan a line before I brew on Saturday....

yeah that would be less extreme. I have a habit of going overboard.

Clint04 said:

This profile is for Fat Head's Head Hunter IPA. The head brewer at Fat Head's gave the entire recipe - including water profile - to Mitch Steele for his IPA book.

I was leaning towards the Bru'n Water side of things on this one, becase the grain bill does not mention any acid malt. I would have to add acid malt, per EZ Water. I know there are many holes in that logic, but that's what my simple mind is going with!

It's definitely a lot of additions though. Unfortunately, thems the breaks when you use 100% distilled water.

Click to expand...

Sounds reliable then. I guess you don't know if you don't try. Hopefully someone will chime in with a more specific answer to your question in regard to the HCO3 - I don't know the answer to that. But I would stick with the Bru'n water numbers. I have never heard anyone say that EZ Water is consistently more accurate than Brun Water.

good luck

EZ Water uses the first SRM based mash pH prediction that I came up with after I did some research on how mash pH is affected by grain and water. I recently revised this formula to be independent of mash thickness based on some additional insight. That more accurate formula is implemented in my current spreadsheet: http://braukaiser.com/documents/Kaiser_water_calculatorUS_units.xls

Unless a mash pH calculator lets you enter pH related grain properties (pH and acidity, for example) it calculates these values from the color of the malt and as a result SRM based mash pH prediction is not inherently less accurate than a grist based prediction.

Can you give more detail about the actual water, volumes and the color of the beer?

Kai

Kaiser said:

Can you give more detail about the actual water, volumes and the color of the beer?

Kai

Click to expand...

Kai,

Here is the grain bill (with Color L):

6lbs 10oz Two-Row (1L)
3lbs 5oz Maris Otter (4L)
12oz CaraHell (11L)
12oz Crystal 40 (40L)
12oz Flaked Wheat (1L)
6oz CaraPils (1L)

12.56lbs total. Estimated color is 8.1 SRM.

4 gallons of distilled water in the mash (with the previously noted additions).

Let me know if that helps!

I think 5.4 is a reasonable pH prediction for this grist and is a good mash pH. I would not aim for lower than that.

The Ca is dropping the mash pH by ~ 0.13 and the baking soda is raising it by ~0.05. That's a net drop of 0.08.

Your grist has a distilled water pH of ~5.5. Just yesterday I brewed an IPA with about 4% CaraMunich I which is similar to what you have here and my mash pH was ~5.35. I had only 80 ppm Ca but also less alkalinity. So the residual alkalinity was about the same.

I put a copy of my spreadsheet with your mash salt additions here: http://braukaiser.com/download/misc_forum/Kaiser_water_calculator_for_Clint04.xls

Kai

Thanks Kai!

I will move forward with the recipe, as I had it per Bru'n Water (no Acid Malt like EZ Water).

Clint04 said:

1) In order to hit the HCO3 target, I would need to add 0.3g per gallon of baking soda to keep the Na on target while gaining some bicarbonate. Even then, the bicarbonate would be ~60ppm. Should I even be concentrating on the HCO3 target, or just focus on Ca, Cl, and SO4?

Click to expand...

No, don't worry about the bicarbonate at all and don't worry about the other targets too much either.

To start with the bicarbonate: the target level of 104 mg/L could come from water with alkalinity of 72 with 1.75 mmol/L total carbon at pH 8 or water with alkalinity of 86 with 2.10 mmol/L total carbo at pH 7 or .... Which will you shoot for?

Whatever you decide (assume that it's something in the middle like 2.0 mmol/L at pH 7.15) by the time you acidify to pH 5.4 only 9% of the bicarb will remain. The rest will have been converted to CO2 and flown off. As the amount of bicarb remaining will have little flavor impact on the finished beer (a good thing!) why even worry about it? The only answer is in that it's effect on the alkalinity will need to be taken into account in determining how much acid to use to get to pH 5.4 (assuming that is your goal). Assuming that you will use a flavor neutral acid (such as phosphoric) the effects of bicarbonate are seen to be minimal. If, conversely, you are looking for the bicarbonate as a means of offsetting acid because you want the cation of that acid (and I'm including the acids in dark malt here) then use enough of that acid (or malt(s)) to get the flavor you want and adjust pH with a flavor neutral alkali i.e. slaked lime (Ca(OH)2).

As for the other ions: most things in nature like our hearing, seeing and tasting for example, respond geometrically rather than arithmetically. Thus looking at the chloride, for example, letting it go to half it's target value of 23, i.e. down to 11.5 will probably render the beer noticeably thinner that at 23 but if it doubles to 46 that will probably make the beer taste noticeably rounder, sweeter, and smoother than at 23. Each of these changes is 0.3 log units which is generally considered the amount of sound level difference that one easily can hear and also represents the light level change of 1 stop. OTOH a change of 0.1 log unit (increase of 26% or decrease of 21%) probably won't make that much of a difference. Based on this line of reasoning you don't need to focus on getting your calcium, chloride and sulfate levels to within a few mg/L of the targets but more like within 30% or so of their targets.

What you want is crownsville hose water. swing on by! You can look at my tart of darkness!

AJ,

As always, I appreciate the detailed response. If you don’t mind, I have a couple of follow up questions for you. It’s pretty evident that I am still trying to wrap my head around all of this.

For starters, I will ignore the bicarbonate from now on. However, do you think it is necessary to add baking soda? I was going to add enough to get 20ppm Na, just to get close to this water profile that I am trying to mirror. Also, per Bru’n water the estimated pH of my mash would be 5.2 with the baking soda addition. If I remove the baking soda, the estimated pH is 5.0, which I don’t want. Per Kai’s spreadsheet he linked in the post above mine, the estimated mash pH with the baking soda is 5.46.

Given that I don’t have a pH meter yet, I am thinking it is a good idea to keep the baking soda addition, because even if I hit either the Bru’n water 5.2 or Kai’s 5.46, that seems to still be in the correct ballpark. I was just wondering if you would weigh in on what you think.

I don’t plan on adding acid at all during the process, since the calculators have me in the correct range. In your post, I’m not sure if you were advising that I use a flavor-neutral acid or just making a point, so I figured I would add that info here.

Also, thanks for the info regarding the Ca, Cl, and SO4 targets. I guess if I don’t hit them exactly, it isn’t the end of the world. I may actually shoot for a touch less than 350ppm of SO4 in this beer, just because I don’t want to take any chances of getting a sulfur taste/aroma in this beer.

Hopefully everything works out. My next brewing investment will certainly be a pH meter!

Clint04 said:

For starters, I will ignore the bicarbonate from now on. However, do you think it is necessary to add baking soda? I was going to add enough to get 20ppm Na, just to get close to this water profile that I am trying to mirror. Also, per Bru’n water the estimated pH of my mash would be 5.2 with the baking soda addition. If I remove the baking soda, the estimated pH is 5.0, which I don’t want. Per Kai’s spreadsheet he linked in the post above mine, the estimated mash pH with the baking soda is 5.46.

Click to expand...

I'd say don't add it but it is hard to tell for sure. That's where the value of a pH meter comes in. In principal it is relatively simple to predict mash pH. One models each malt's titration curve as a Taylor series expansion of two or 3 terms about the DI water mash pH and then calculates the amount of acid or base required to move it's pH to target pH tweaking target pH until the acids provided by some of the malts balance the bases supplied by others and the water. The problem is in getting those Taylor series terms. It takes a tremendous amount of meticulous, dull, repetitive lab work to get the numbers for one malt and then they pertain to one lot of one season of one color of one maltster's malt. If I use some nominal numbers (the best I can get) I find that you will require some sauermalz (acid) but my numbers shouldn't be trusted any more than anyone elses because they are at best nominal. pH predictions are valuable for roughing in but refinement must come from measurement. My take is that it is lots easier to do a test mash than the mesurements on the individual malts required for a calculation.

Clint04 said:

Given that I don’t have a pH meter yet, I am thinking it is a good idea to keep the baking soda addition, because even if I hit either the Bru’n water 5.2 or Kai’s 5.46, that seems to still be in the correct ballpark. I was just wondering if you would weigh in on what you think.

Click to expand...

Your logic is good. But if you average in my prediction that would say don't use it or use a reduced amount. I can't in good conscience tell you not to do it but personally I wouldn't. In my experience the dangers of high mash pH are more likely to arise than low.

Clint04 said:

I don’t plan on adding acid at all during the process, since the calculators have me in the correct range. In your post, I’m not sure if you were advising that I use a flavor-neutral acid or just making a point, so I figured I would add that info here.

Click to expand...

I would add a percent or so of sauermalz to this recipe if I didn't have access to a pH meter but again I am not advising you to do so. I know this isn't much help but were I in your shoes I'd probably skip the acid and skip the bicarbonate.

Clint04 said:

My next brewing investment will certainly be a pH meter!

Click to expand...

It will really improve your brewing but pH meters bring their frustrations too!

AJ,

You have me convinced to ditch the baking soda addition, and add an ounce of acidulated malt.

Since I started following the water chemistry primer, I have been adding some acid malt to every batch of light-colored beer, and my beer has certainly improved. I see no reason to stop now. Plus, per your data, Kai's spreadsheet, and EZ Water, it appears as if acid malt should be added.

Also, I just don't want to mess around with the baking soda. A little extra Na is too much Na, in my opinion. If I accidentally added a touch too much CaCl or Gypsum, that's one thing, but to my tastes, I don't want to risk it with sodium.

I'll let you know how the beer turns out!

ajdelange said:

It will really improve your brewing but pH meters bring their frustrations too!

Click to expand...

I think that the colorpHast strips are a decent alternative to a pH meter and are a good back-up to have once you have a pH meter. Just keep in mind that they read low by about 0.3 pH units (more here: http://braukaiser.com/wiki/index.ph...st_strips_for_pH_measurements_in_home_brewing)

Kai

Clint04 said:

AJ,

You have me convinced to ditch the baking soda addition, and add an ounce of acidulated malt.

Since I started following the water chemistry primer, I have been adding some acid malt to every batch of light-colored beer, and my beer has certainly improved. I see no reason to stop now. Plus, per your data, Kai's spreadsheet, and EZ Water, it appears as if acid malt should be added.

Also, I just don't want to mess around with the baking soda. A little extra Na is too much Na, in my opinion. If I accidentally added a touch too much CaCl or Gypsum, that's one thing, but to my tastes, I don't want to risk it with sodium.

I'll let you know how the beer turns out!

Click to expand...

I'd like to hear also. The alkalinity in a very hard profile like that for a pale ale is a necessary component in order to produce any semblance of an acceptable mash pH. Ditching the alkalinity addition and even adding a bit of acid is fool-hardy. The consequences of the 2 or 3 tenths reduction in mash pH will only be a reduction in beer body due to the increased fermentability and the hop expression will also be muted due to that reduced wort pH. In my experience, none of those features are things I would want my IPA to have.

I'm assuming that Clint has been dealing with a far less mineralized water profile in his previous brews. Under that assumption, a bit of acid malt might actually be a requirement. Especially if the grist is not very acidic like that for a light colored beer with little crystal.

Assuming this is a first exploration into a water for a pale ale, I would caution against moving to such an aggressive sulfate level of 350 ppm. I like 300 ppm, but it may not be for everyone. A more modest level might be appealing to some palates.

I'm no fan of baking soda, but it is effective when used in moderation. If that addition only brought the sodium concentration to 21 ppm, then that is quite low and potentially inconsequential. A little extra sodium would actually be a good thing in this beer. Just keep it modest. I wouldn't go over 50 ppm. The 21 ppm is an appropriate enhancement to flavor.

I look forward to your impressions on this finished beer with your latest water revisions. They do say that failure is a very good learning tool. With that being said, I still prefer to keep my failures to a minimum.

Martin, I definitely appreciate the input. (Now, whether or not I appreciate more information that can change my easily persuaded mind is another thing )

Yes, I usually brew beers that are far less mineralized. I brew a lot of saisons and belgian styles. The only other pale ale I did used the levels of Ca, Cl, and SO4 that Vinnie at Russian River uses (Ca 103, Cl 57, SO4 174). This is much different than that, for sure. The beer was okay, but the hops didn't "pop" like I wanted them to. Not to mention, it was a SMaSH beer in which I didn't care for new variety of hop I used.

I actually planned on toning down the SO4, as you and others have mentioned on this post. I'll probably shoot for 300-325.

Given the wealth of knowledge that has been shared in this thread - not to mention slightly differing opinions - it is probably going to take me until Saturday to decide which direction I want to go in. I'll keep you posted.

I spent a little more time thinking about the effects of the uncertainties (of which there are many) in mash pH prediction and came up with the thought of a quicky Monte Carlo to see if I could gain some insight. So I took the original recipe and modeled the individual grains DI mash pH's as follows:

Two row 5.7
Maris Otter 5.6
Cara Hell 5.3
Crystal 40 5.3
Wheat 5.67
Cara Pils 5.2

To each of these I attached a Gaussian uncertainty of standard deviation 0.05 pH.

I assumed a target pH of 5.4.
Kolbachs commonly quoted rule of thumb predicts a pH reduction of 0.2 at knockout so mash pH reduction is going to be less than that. I modeled it as 0.15 pH with standard deviation of 0.05 pH.

I modeled the malt titration curves as being linear with slope, at the DI pH's of = -22 mEq/kg-pH with the exception of wheat which I had measured as much lower than that so for wheat I used -14. I applied a standard deviation of 7 mEq/kg-pH to all slopes except wheat for which I used 4.

Any one can, of course, argue with any of these assumptions and people with different experiences and different data sets are perfectly justified in doing so. Changing any of the assumptions changes the conclusions.

I 'brewed' (or mashed) this beer 20,000 times (takes less than a second) and determined the percentage sauermalz that would be needed to hit pH 5.4 for each brew. The results are in the histogram. On average you would need 0.5% sauermalz but you might need as much as 3% (very unlikely). Or, you could need -1.5% (also very unlikely). This means adding alkali e.g. bicarbonate, equivalent to the acidity of 1.5% sauermalz.

This does not, of course, help you to resolve your uncertainty and that's really what the message is supposed to be: there is lots of uncertainty. Based on the fact that the average acid addition is small (half a percent) I think I'd just skip it but I'd skip the bicarbonate too. I think Kai's suggestion that you use pH strips, bad as they may be, is an excellent one.

Thanks AJ.

That simulation illustrates the problem very well.

My experience has been that malt buffer capacity is somewhere between 30 and 40 mEq/(kg*pH) (it shows a fairly strong correlation to the distilled water pH) and with that assumption your most likely amount for Sauermalz would be closer to 1% as is the known rule of thumb.

Kai

I like the idea of modeling the predicted mash pH distribution for a given recipe, but do you need to use a Monte Carlo? If you're assuming everything is a gaussian, then a weighted sum will still be a gaussian. I suppose you'll divide by the buffering capacity, but there are transformations that can approximate that fairly well as another gaussian provided you can assume the denominator is highly unlikely to change signs.

A nice model of the pH distribution would let you do nice things like assigning an objective function over the mash pH to make salt/acid additions that would be more likely to get you in a desired range of mash pH as opposed to just hitting a target pH.

bdh said:

A nice model of the pH distribution would let you do nice things like assigning an objective function over the mash pH to make salt/acid additions that would be more likely to get you in a desired range of mash pH as opposed to just hitting a target pH.

Click to expand...

I think it would be a good enhancement of a water calculator to propagate errors. I did this in my efficiency spreadsheet and its is very informative. I'll keep this in mind for my next water calculator design.

You are correct that simple error propagation would have worked too, But that is tedious and writing a script that does MC simulation is easier and takes less debug time.

When I first tried to figure out the relationship between SRM and mash pH I also wrote a script that randomly mashes reasonably realistic grists and estimated their mash pH (http://braukaiser.com/wiki/index.ph...recipe_creation_and_pH_and_color_calculations)



Kai

bdh said:

I like the idea of modeling the predicted mash pH distribution for a given recipe, but do you need to use a Monte Carlo?

Click to expand...

You wanted me to do actual work?

bdh said:

I suppose you'll divide by the buffering capacity,...

Click to expand...

Actually it's the other way around. I assume a target pH and calculate the shift by subtracting this from the DI pH. The quantity of protons required for each malt is then

pH_shift(pH)*Buffering_capacity(mEq/kg-pH)*weight(kg).

Actually the reason I did the Monte Carlo was because
1) I wasn't sure what distribution I was going to use at first but finally settled on the Gaussian.
2) I'm not convinced that a 2 term Taylor series (linear) expansion for the buffering curve is adequate. I've measured second and even third order terms but am not convinced that my method is robust enough to accept that they are really there.

Kaiser said:

My experience has been that malt buffer capacity is somewhere between 30 and 40 mEq/(kg*pH) (it shows a fairly strong correlation to the distilled water pH) and with that assumption your most likely amount for Sauermalz would be closer to 1% as is the known rule of thumb.

Click to expand...

In another thread (https://www.homebrewtalk.com/f128/lactic-acid-use-378392/) I figured out what the Weyermann rule implies in terms of buffering capacity. To do that I assumed 2% (which I based on seeing somewhere that it is typically between 1 and 3% IIRC) and determined the buffering to be 21.5 mEq/kg-pH. I then looked at the limited data set I have on measured buffering capacity and found the average to be 26 with a standard deviation of 8.5. If I assumed 1% that would imply lower (10.7) buffering capacity as there would be less acid in each gram of sauermalz.

As noted in previous posts here I don't have a lot of confidence in our ability to titrate malts so I'm not ready to go to the wall on my conclusions but at least my measurements, the Weyermann rule and the 2% number are consistent with one another. That lends some comfort if not a lot.