Alkalinity on Swedish water reports

[mrmekon]

I'm trying to learn some brewing chemistry, and I'm stuck on my own local water report.

Stockholm publishes a water report here: http://www.stockholmvattenochavfall...t/kval-dekl-lov-2017-med-logga-2018-03-30.pdf

I think you can figure out the translations. The confusing factor for me is Alkalinity, which is specified as 72 mg/l HCO3. Is it actually reporting bicarbonate there, or is it reporting "Alkalinity as HCO3"? If the second one, which is what I suspect, why is it different from every other water report (mEq or 'as CaCO3')?

Secondly, somebody linked to this same thing in a sticky post and pointed out that the standard that Sweden uses for measuring alkalinity apparently involves titrating only to pH 5.4: https://www.sis.se/api/document/preview/18780/

I'm not even sure if that ISO standard applies to this water report, though.

So is this report actually specifying alkalinity as "the equivalent amount of HCO3 when titrating from pH 8.5 to pH 5.4 with a strong acid"? If that's the case, then simply translating it back to mEq (1.18 mEq?) will give me an alkalinity measurement with a different end point than all of the calculators expect, right? How does that affect things? I saw that pH 4.3, 4.4, and 4.5 are all used as references, but I haven't figured out if that matters or what calculations it affects.

Or does it just mean bicarbonate?

 

[ajdelange]

Unfortunately the second link delivers a pdf that cuts off before they get to the part where they tell you what to do with the titration results. In the titration a certain number mEq of acid are used to get the pH to 5.4. The next step is the one we need to be sure of what they are doing. If in that last step it says to multiply the mEq/L by 61 then they are expressing alkalinity as bicarbonate (Martin should love this). Assuming that this is the case (and I assume it is) you need to convert this to alkalinity as CaCO3 re pH 4.5 (the ISO standard). To first, convert the alkalinity as bicarbonate back to mEq/L : 72/61 = 1.18033. That's the number of mEq of acid required to lower the pH of 1 L of the sample from its presenting pH of 8.5 to pH 5.4. The key to solving problems like this is knowing the charge on a mole of carbo (= the sum of the moles of carbonic acid, bicarbonate ion and carbonate ion) at each of the pH's of interest. Qc(5.4) = -0.094789287 and Qc(8.5) = -0.991652811. The difference between those two ( 0.896864 mEq/mmol) is the amount of acid it takes to move one mmol of carbo from 8.5 to 5.4. Since your required 1.18033 mEq/L it is clear that you water contains 1.18033/0.896864 = 1.31606 mmol/L carbo. The charge on 1 mmol of carbo at pH 4.5 (the ISO pH) is Qc(4.5) = -0.013011082 and the amount of acid to move 1 mmol from pH 8.5 to pH 4.5 is then -0.013011082 - (-0.991652811) = 0.978642 mEq/mmol. As you have 1.31606 mmol your alkalinity WRT pH 4.5 is 1.31606*0.978642 = 1.28795 mEq/L (64.4 mg/L as CaCO3) which is not a lot higher than the 1.18033 mEq/L WRT pH 5.4

 

[mrmekon]

Great, thanks ajdelange! Then it seems I did understand it, and it's just odd. I found another Swedish paper from the concrete industry that says explicitly "the target alkalinity is 244 mg/l as HCO3" and references EN ISO 9963-2, so I guess that's what they do here.

Your conversion math is much wiser than mine. I rebased the reference pH by iteratively solving for a carbo amount that worked...

In your 'Estimating Mash pH' article you use pH 4.3 as the reference. Any reason why you used 4.5 above? Now I understand your bit about how the Residual Alkalinity formula is not a function of pH, and that is weird. I guess, as long as the reference pH is in the 4.3-4.5 range, the differences are just too small to matter considering that it is all estimates anyway?

One more question: I'm calculating Q using the formulas you posted in that article and around this board, but my calculations are a tiny bit off from yours. I get Q(5.4) = -0.0944539663278 and Q(8.5) = -1.00558438197 using pK1 = 6.3817 and pK2 = 10.3756. Are you using different pK values?

Ultimately, I need to figure out how to measure my actual water by titrating 80% lactic acid (the only thing I can get), since I guess these water reports are pretty much useless.

 

[ajdelange]

Great, thanks ajdelange! Then it seems I did understand it, and it's just odd. I found another Swedish paper from the concrete industry that says explicitly "the target alkalinity is 244 mg/l as HCO3" and references EN ISO 9963-2, so I guess that's what they do here.

When my mother would get mad at my father she would call him a "square headed Swede". Maybe she knew something. In fact the man was Norwegian.

Your conversion math is much wiser than mine. I rebased the reference pH by iteratively solving for a carbo amount that worked...

If it worked, who cares?

In your 'Estimating Mash pH' article you use pH 4.3 as the reference. Any reason why you used 4.5 above? I guess, as long as the reference pH is in the 4.3-4.5 range, the differences are just too small to matter considering that it is all estimates anyway?

When I first started out exploring brewing water chemistry information was pretty thin. The best treatment of it I could find was a chapter by Manfred Moll in Hardwicke's Handbook of brewing. Manfred (who turned out to be a very nice fellow) stated in that chapter that alkalinity titration was carried out to pH 4.3 and I threw that number about for years until I came across the AWWA (Standard Methods) procedure which said to use any pH you like but report it with the alkalinity number. I subsequently discovered the ISO ISO 9963-1 which says to use 4.5 and then someone (from Sweden) brought up ISO 9963-2 which says use 5.4. Ward Labs uses 4.4. Indeed the titratiion curve is pretty flat between 4.3 and 5.4 so that results anywhere in that range are going to be pretty close (barring the presence of other cations). It turn up below 4.3 (as 10^-pH becomes appreciable) and down above pH 5.4 and r1 = 10^(pH - 6.38) becomes appreciable.

Now I understand your bit about how the Residual Alkalinity formula is not a function of pH, and that is weird.

I found out about RA from that same chapter in Hardwicke, thought it was pretty cool and posted about it a lot. I even made up a chart of hardness vs. alkalinity with constant RA lines on it and plotted the water of several brewing cities on it that you may have seen. And that's what RA is for - comparing waters. This stuff gets pretty complicated and so home brewers tend to grasp at straws. RA became one of those straws and I have to accept at least some responsibility for that. It really doens't have that much utility other than what Kolbach intended it for (comparing waters) but is, nevertheless, considered by some to be the holy grail of brewing water chemistry. It is very important to understand that the definition of it, as given by Kolbach, refers to calcium's and magnesium's effect at knockout - not in the mash. How, then, can we use it to predict mash pH?

One more question: I'm calculating Q using the formulas you posted in that article and around this board, but my calculations are a tiny bit off from yours. I get Q(5.4) = -0.0944539663278 and Q(8.5) = -1.00558438197 using pK1 = 6.3817 and pK2 = 10.3756. Are you using different pK values?

Slightly. I rounded to 6.38 (corresponding to 20.3 °C) and 10.38 (corresponding to 19.6°C) because I can remember those numbers.

Ultimately, I need to figure out how to measure my actual water by titrating 80% lactic acid (the only thing I can get), since I guess these water reports are pretty much useless.

The problem isn't that you can't get an ISO 9963-1 value (or value referred to any other pH) from an ISO 9963-2 report (or one referred to any other pH). As you have discovered what really characterizes a water WRT alkalinity is its pH and carbo content. The real problem with these reports is that last year's reported average alkalinity may not match what is coming out of your tap today. One way around this is to measure the alkalinity before every brew. Another is to use the Zero Effective Alkalinity technique in which you simply add acid to the brewing water until its pH is the desired mash pH (and in so doing implicitly measure the alkalinity and treat it rather than doing so explicitly).

Measuring alkalinity with lactic acid is going to be a little tricky. Lactic acid (with its pK of 3.86) isn't quite strong enough for an alkalinity titration. This means the normality (mEq/mL) of that 80% lactic acid is going to vary from 9.7 if titrating to 4.3 to 12.8 if titrating to 5.4 - if the stuff in the bottle labeled 80% is really 80%. You would have to standardize any acid you bought and the only thing I can think of to standardize it against would be chalk (no water of hydration so that 100 grams of it dried in an oven would contain 100 grams CaCO3), But drying in an oven is a pain and and main problem is that chalk reacts slowly with acid so that you would have to wait hours for the acid to neutralize the chalk.

Can you not obtain an alkalinity test kit from an aquarium supply shop?

 

[mrmekon]

When my mother would get mad at my father she would call him a "square headed Swede". Maybe she knew something. In fact the man was Norwegian.

That might be the single most insulting thing you can say to a Norwegian

It really doens't have that much utility other than what Kolbach intended it for (comparing waters) but is, nevertheless, considered by some to be the holy grail of brewing water chemistry.

In all of the reading I've done this week, one thing I have not put together is what a more accurate way is to calculate calcium/bicarbonate cancellation. Kolbach's formula has clearly been shown to be a good rough estimate, but that's all it is. I was just reading your paper on the interaction between phosphates and calcium, wondering if the secret is in there, but the chemistry is over my head.

I'm teaching myself this stuff by experimenting in Python with your suggestion of charge cancellation between the water alkalinity and a grist. I implemented one of your polynomial curves describing the charge absorption of a malt. I'm at a loss for what to do with the Calcium/Magnesium content in this model, though, other than just calculate RA and accept that it has high error.

Can you not obtain an alkalinity test kit from an aquarium supply shop?

I didn't know about those... apparently I can!

 

[ajdelange]

In all of the reading I've done this week, one thing I have not put together is what a more accurate way is to calculate calcium/bicarbonate cancellation. Kolbach's formula has clearly been shown to be a good rough estimate, but that's all it is. I was just reading your paper on the interaction between phosphates and calcium, wondering if the secret is in there, but the chemistry is over my head.

I assume you are referring to the Cerevesia paper? Yes, the chemistry in that paper is real mess - simultaneous solution of all the phosphate equations, carbonate equations, proton condition, charge neutrality. I think I did get some numbers which sort of agreed with Kolbach's finding that says perhaps that is the right approach but I certainly don't think that chemistry represents a practical approach to the problem. Clearly only a fraction of available reactants actually come together. How would we know what that fraction is? Kolbach's emprirical observation may only really apply to German lagers but it might be possible to find other numbers for other beer styles.

I'm teaching myself this stuff by experimenting in Python with your suggestion of charge cancellation between the water alkalinity and a grist. I implemented one of your polynomial curves describing the charge absorption of a malt. I'm at a loss for what to do with the Calcium/Magnesium content in this model, though, other than just calculate RA and accept that it has high error.

i just go with Kolbah's definition and assume that alk - [Ca++]/3.5 - [Mg++]/3.5 means that 3.5 mmol/L calcium releases 1 mEq/L H+ ions at knockout and that 7 mmol/L Mg++ does the same. As I know the calcium and magnesium reactions continue at least to some extent in the kettle I consider that somewhat less than 1 mEq/L H+ is released by, respectively, 3.5 mmol/L and 7 mmol/L. I have a cell in my spreadsheet where the user can enter the factor (3.5 or greater) he wants to use to model this.

I didn't know about those [aquarium test kits]... apparently I can!

That's great. I think they read in ° dH but the conversion of those to mEq/l should be easy to find.