Adding Chalk/Spreadsheet variations

I've been using Kaiser's Water Calculator spreadsheet as well as Martin Brungard's Bru'n Water in preparation for brewing an alt beer. These spreadsheets are great applications and have simplified the water treatment process for me. My tap water averages from the water company's website are as follows:

Alkalinity (as CaCo3): 60ppm
Calcium: 73ppm
Magnesium: 60ppm
Sulfate: 124ppm
pH: 9.6

My residual alkalinity is -27, and I would need to add some alkalinity to brew a darker beer like an Alt. My plan was to dilute with 75% RO water to get sulfate and calcium concentrations within range before any salt additions. Before I add any salts, both spreadsheets predict a mash pH of 5.3. But after I add 4 g. of dissolved chalk, they vary widely, with Kaiser's spreadsheet predicting a 5.5 pH and the Bru'n Water predicting a 5.8.

Has anyone had experience with adding chalk and recorded the change in pH when adding 1-4 g. of chalk? Has anyone reported a chalky aftertaste when adding as much as 4 g. of chalk? I plan on using Kaiser's method of dissolving the chalk in water with CO2 pressure.

The problem with using chalk in Bru'n Water is that the program assumes that the brewer has fully pre-dissolved the chalk into water. That won't happen in plain water and only progresses partially in the mash. As you mention, the method of dissolving chalk with CO2 is a necessary step if you want to see the prescribed alkalinity.

There are numerous cautions on the use of chalk in brewing in the Bru'n Water instructions and in the Water Knowledge sections. The bottom line is that you shouldn't use chalk unless you're going to pre-dissolve it fully in water. Most brewers would find that to be a pain in the a**. The better way of adding alkalinity and calcium to water and mash is to use pickling lime. It dissolves easily and delivers the alkalinity indicated by Bru'n Water without special measures.

You mention 4 g of chalk. Is that 4 grams per gallon or is that 4 grams in a 5 gallon batch (0.8 grams per gallon) ? That might be a difference. You enter your mineral additions as grams per gallon in Bru'n Water. There is a calculator on the right side of that sheet that shows you the total mineral addition based on the volume of treated water.

I'm not sure how Kai is calculating mash pH, but he has tested a lot of things out. I would not be surprised if he has accounted for the limited solubility of chalk in his program.

I did account for the grams per gallon and the total mineral addition was 4 grams of chalk. I suppose the pickling lime would be easier, and I'm seriously considering using it instead. Since I am new to using salts to increase alkalinity, I was wondering which methods work the best for experienced water tinkerers and how much the salts, dissolved chalk or pickling lime, raise pH per gram addition (in your experience, as there is no universal standard). You suggest that pickling lime would be better. I will go back and use your spreadsheet with the pickling lime additions as well. I will conduct my own mashes with some conservative additions and report back my pH readings. Thanks for the feedback and thanks for the spreadsheet!

philbert119 said:

I was wondering which methods work the best for experienced water tinkerers and how much the salts, dissolved chalk or pickling lime, raise pH per gram addition (in your experience, as there is no universal standard).

Click to expand...

Experienced tinkerers do not in general, add chalk to the water as to get it to dissolve requires acid, the spreadsheets don't calculate the amount of acid you'd need to dissolve and reach a particular pH and last, but by no means least the two major rules of thumb in brewing water treatment are 1. Alkalinity is bad and 2. Don't add alkalinity to brewing water but instead add it to the mash and do this if and only if a reliable pH measurement in the mash indicates that the pH is too low. In that case add some lime in small increments until the mash pH is correct. Lime is preferable because of rule 1. Alkalinity is bicarbonate in brewing water and if you add carbonate it becomes bicarbonate and while most of that is removed when you get to proper mash pH what's left isn't doing flavor much good.

If you do want to add alkalinity to water using the acid method calculating the amount of acid to add is difficult and must be done iteratively over the "master variable" pH. This is why most spreadsheets don't deal with it. Excel will do the iteration for you through the Solver but setting up Solver is an added complexity which takes away from the utility of a spreadsheet.

But then it occurred to me that the spread sheet doesn't have to do this nasty iterative business if somebody does it once and tells the world what the answer is. I am that somebody and here are the answers for 3 acids.

If you put calcium carbonate in water and acid to the point where all the mineral is dissolved and the pH comes to a particular value the alkalinity and the amount of acid it takes to get to the specidfied pH depends on the amount of chalk per liter and the pH. It also depends on the end point pH used to define alkalinity. Assuming end point pH is 4.3 the alkalinity of mgpL chalk dissolved with hydrochloric acid (or another strong acid such as sulfuric) and brought to pH value pHs is (approximately):

Alk = mgpL*1.0685/(1 + exp((6.3664 - pHs)/6.48654))/100 mEq/L. Multiply by 50 for ppm as CaCO3. Accuracy is good to about 1 ppm as CaCO3 (this is obviously a curve fit).

The amount of acid required to do this is 2.0408 - Alk i.e a constant minus the alkalinity calculated in the last formula in mEq/L. The amount of acid is also in mEq/L

Thus if I put 50 mg/L chalk into my liquor and add acid until the pH is 7.2 the alkalinity from the added chalk will be 0.45 mEq/L (22.5 ppm as CaCO3) and I'd need 1.59 mEq HCl per liter of water.

The only reason, in my mind, for adding alkalinity to water is if you want to duplicate the water profile of some brewing center of reknown so that you can go through all the steps the brewers of yore did. IOW you are striving for the ultimate in authenticity. You will never match real profiles with chalk and hydrochloric acid because that's not what nature uses. She uses carbonic acid. Using carbonic acid the alkalinity is

Alk = mgpL*2.031*(1 - exp((4.3 -pHs)/.42783))/100

Again this is in mEq/L and should be multiplied by 50 for ppm as CaCO3.

This time the amount of acid isn't so important as you don't measure out CO2 with a teaspoon. So I've calculated the pressure of CO2 required to bring the mix to the specified pH with the specified chalk addition. It is

PaCO2 = 10^(5.0938 - pHs)

Thus if in the same example as above except using carbonic acid I would get 1.014 mEq/L alkalinity (50.7 ppm as CaCO3) and require a CO2 pressure of 0.0078 atmospheres do get to this pH. Note that the CO2 pressure depends only on the pH. Note that in this case I got alkalinity (in ppm as CaCO3) about equal to the mass of the chalk I dissolved in ppm also obviously as CaCO3. This is the origin of the "as CaCO3" unit. This will be the case for pH's as low as about 5.75. Also compare to the strong acid where one gets about half as many ppm alkalinity as the CaCO3 he dissolved.

It's a bit tricky to dial in 0.0078 atmospheres from your CO2 bottle so what you do is sparge CO2 through the water or put the water in a Corny keg under modest pressure and when all the chalk is dissolved take the water out of the Cornie and let it stand monitoring pH as the extra CO2 escapes into the air. If you are thinking ahead you will reason that water carbonated in this way will lose CO2 and quite probably precipitate the chalk you went to so much trouble to dissolve as soon as it is heated in the HLT. That's what happens with hard carbonaceous waters to a commercial brewer and that's what will happen to you. Another reason not to bother.

Finally, I also did lactic. Lactic is interesting in that is has protons to give up (is acid) but also has alkalinity at the pH's we are intersted. So it's curve isn't quite as simple as the others and takes a pretty high polynomial to fit. The alkalinity is

Alk = mgpL*(((((((-.000741886191102494*pH+.0310611616245953)*pH-0.532742116315294)*pH+4.76952863143326)*pH-23.2787519265881)*pH+56.9307290160207)*pH-44.2861550535682)*pH-35.0097530587467)/100

This would be a bear to type into a spreadsheet but is in a form easy to cut and paste. Then substitute the appropriate cell number for pH or define pH as a variable. All the seemingly superfluous digits are there because up to the seventh power of pH is being calculated.

Again, this is in mEq/L. Multiply by 50 for ppm as CaCO3.

The amount of acid required to dissolve the chalk and attain the desired pH is 2.7808 - 1.3631*Alk but this time in mmol/L.

Continuing the example the realized alkalinity from 50 mg/L chalk dissolved with lactic acid and brought to pH7.2 is 0.608 mEq/L (30.4 ppm as CaCO3) and requires 1.95 mmol/L 100% lactic acid to achieve this.

So it turns out it is quite easy to allow for pH and correctly calculate the effects of chalk additions to brewing water after all! It should be pretty easy for the spreadsheet authors to add this into their spreadsheets and I hope they will consider doing this.

A final note: my approach involves setting my spreadsheet for ideally dilute chemistry. I cannot come up with simple formulas for alkalinity vs. pH alone if ionic strength is considered. I guess I could take the same approach and come up with fits against pH parametric in ionic strength but then the spreadsheets would have to calculate ionic strength and, AFAIK, none of them do.

My spreadsheet assumes that only half the added and not pre-dissolved chalk dissolves in the mash and thus raises alkalinity. This is pretty much what I have observed in my own experiments. You'll notice that on the advanced tab of my spreadsheet it allows you to enter chalk dissolved with CO2 and it will also allow you to estimate what CO2 pressure you'll need to dissolve that chalk in a smaller volume of water if you plan to go that route.

I think the addition of undissolved chalk to water is fine. We have done that for a long time w/o detrimental effect on the beer. Now we know that not all of it dissolves and that we have to compensate by adding a bit more. I still like using chalk since it is a bit safer to handle than pickling lime.

Kai

Adding chalk to water is fine if you want to go to the trouble and if you calculate the amount correctly it will all dissolve.

I've been fiddling with the EZ spreadsheet and have it almost wickered up to the point where it will handle chalk additions, compute the alkalinity they produce correctly, display the CO2 pressure or amount of hydrochloric, sulfuric, lactic or phosphoric acid required to get the chalk to dissolve and bring the pH of the water to a selected value (will be 7 in nearly all cases) and warn you if the resulting solution is super saturated WRT CaCO3. It's fine to be supersaturated - many waters are but as soon as you heat them the chalk drops out. That's one of the reasons people usually don't bother with it.

Still working on the bicarbonate in EZ - it's actually a more difficult problem than carbonate but only if you go for target pH greater than 8.3 though I can't imagine anyone wanting to do this.

And yes, I have had people report that following the recommendations of spreadsheets that tie water chemistry too heavily to beer color has lead to huge chalk additions with the result that the beers tasted chalky, pasty, alka selzer like. Chalk has its place in brewing but it needs to be used sparingly and carefully.

ajdelange said:

Adding chalk to water is fine if you want to go to the trouble and if you calculate the amount correctly it will all dissolve.

Click to expand...

I think that the simple approach of adding chalk and allowing it to be partially dissolved by the acids in the mash should still be supported. After all that's what we have been doing for a long time and it seems to work. As much as I like to convince brewers they should dissolve the chalk with CO2 first I have to admit that this might be a solution in search for a problem. If anybody has an example where better beer was made because the chalk was dissolved, I'd like to hear about this since then we might be up to something here.

I'm assuming that the chalk that did not get dissolved stays behind in the spent grain. This might not be a given if the wort run off from the mash is very turbid or BIAB is used. In this case chalk could carry over into the BK where it continues to rise the pH and calcium content of the wort.

It's fine to be supersaturated - many waters are but as soon as you heat them the chalk drops out. That's one of the reasons people usually don't bother with it.

Click to expand...

I have built many waters with CO2 dissolved chalk and anything with an alkalinity above 50 ppm CaCO3 could be considered supersaturated at atmospheric CO2 pressure. While I have seen more CO2 escape as usual when heating these waters I have not seen any chalk precipitation before I added the malt. It would take boiling and or extensive splashing of the water to drive out enough CO2 for the chalk to start precipitating.

As for dissolving the chalk with acids other than carbonic, doesn't this create the problem of adding other anions that would otherwise not have to be added to the water?

Still working on the bicarbonate in EZ - it's actually a more difficult problem than carbonate but only if you go for target pH greater than 8.3 though I can't imagine anyone wanting to do this.

Click to expand...

Why does bicarbonate vs. carbonate matter so much? I'm working with alkalinity which encompasses what we are looking for: the water's ability to raise pH.

And yes, I have had people report that following the recommendations of spreadsheets that tie water chemistry too heavily to beer color has lead to huge chalk additions with the result that the beers tasted chalky, pasty, alka selzer like. Chalk has its place in brewing but it needs to be used sparingly and carefully.

Click to expand...

This is because they have been using an outdated SRM to RA relationship. Based on my experiments with malts I have reworked that relationship and found that even the darkest beers don't need nearly as much alkalinity as predicted by the older models. There is more in this here: Beer color, alkalinity and mash pH. There is some math that may appear dense (not to you A.J.) but in the end I approximated the SRM/RA relationship from simulated grists. I'm also working on an analytical solution of the problem, but as expected, it yields about the same results.

I also need to point out that the SRM based RA or pH prediction works just as fine as a grist based one as long as both models predict malt acidity data from the malts color. Only when you have specific pH and/or acidity measurements from malts can you expect to get significantly better accuracy with a grist based pH prediction model.

Kai

Kaiser said:

There is more in this here: Beer color, alkalinity and mash pH. Kai

Click to expand...

Kai, on you table below, when you say "roasted malt portion of special malts" for the columns, do you mean % of roasted malt to the total grain bill or to the amount of non base malts? Not sure I made my question clear.
Would there be a formula one could use to give max and minimum RA values from SRM and % of roasted malt?

Thanks

Kaiser said:

I think that the simple approach of adding chalk and allowing it to be partially dissolved by the acids in the mash should still be supported.

Click to expand...

I do too. But the main reason, IMO, for adding it to water is in order to duplicate a given profile. Sometimes people want to do that. I've even done it though I don't really feel it's necessary. If you are going to try to duplicate a profile you need to model carbonate and bicarbonate correctly. I've been doing that but could never find a way to model things simply enough that they could be plugged into a spreadsheet usable by people just starting out. I think I've figured out how to do that (precomputation). I may decide, after trying to stuff this into the EZ sheet that it is still too complicated.

Kaiser said:

After all that's what we have been doing for a long time and it seems to work. As much as I like to convince brewers they should dissolve the chalk with CO2 first I have to admit that this might be a solution in search for a problem.

Click to expand...

There are a couple of schools here. The more practical one says just make the best tasting beer you can. To do that you add calcium chloride and acid or dark grain to soft water, adjust the pH with chalk, lime, CRS or whatever and add calcium sulfate to taste. The other says "It needs to be as authentic as possible". There are guys who are just that way. They want to prepare genuine Munich water and then decarbonate it as a Munich brewer would. I think that's a waste of time but there are people who want to follow that path. If you want to model Munich water you have to deal with the carbonate and you will have to use CO2 to get it to dissolve.

Kaiser said:

I'm assuming that the chalk that did not get dissolved stays behind in the spent grain.

Click to expand...

That's what I thought too but I remember, in particular, the face the guy who runs the LHBS made when describing how chalky his stout tasted. I've seen other people post similar things in various fora. The "alka selzer" comment I got from Gordon Strong.

Kaiser said:

I have built many waters with CO2 dissolved chalk and anything with an alkalinity above 50 ppm CaCO3 could be considered supersaturated at atmospheric CO2 pressure. While I have seen more CO2 escape as usual when heating these waters I have not seen any chalk precipitation before I added the malt. It would take boiling and or extensive splashing of the water to drive out enough CO2 for the chalk to start precipitating.

Click to expand...

I think you are fine at 50 ppm. In fact you can go as high as 100 ppm and still be undersaturated if you have set pH 7.5 (not an unreasonable pH) but such a solution will be supersaturated WRT CO2. I'll always remember a batch of water I did for Burton ale for a class. It took me a day to get the CaCO3 to dissolve (sparging atmospheric). After throwing the heat to it (single step indfusion strike temperature) there was lots of precip on the bottom of the HLT but I figured the same thing would happen at Burton and that was part of the excercise. I do not do this except when I need demo beers for a class.

Kaiser said:

As for dissolving the chalk with acids other than carbonic, doesn't this create the problem of adding other anions that would otherwise not have to be added to the water?

Click to expand...

Definitely and blows the authenticity argument right out the window because you cannot duplicate what mother nature did with limestone and CO2 with chalk and phosphoric acid. But if you want to do calculations the amount of phosphoric acid required to set alkaline sparge water to a particular pH this method will let you do it.

Kaiser said:

Why does bicarbonate vs. carbonate matter so much? I'm working with alkalinity which encompasses what we are looking for: the water's ability to raise pH.

Click to expand...

Again, it's for the authenticity fans. If matching a profile is not your goal I see little reason to add sodium bicarbonate to beer. The sodium is at best flavor neutral and if you want to add bicarbonate the way to do it is with chalk and acid.

The reason I mentioned it is because when you add chalk you are adding carbonate. The effective pH of the added chalk is much higher than any brewer would ever contemplate as a suitable pH for his treated water. But when you add bicarbonate to water most, but not all of it, stays bicarbonate (some converts to carbonate and some to carbonic). More significant is that if someone decides he wants to synthesize water at pH 9 using bicarbonate you don't need acid to get to target pH, you need base. And in terms of the model, the question becomes "which base". No one would ever do this (well, someone probably would) and so for the spreadsheet to work at any allowable entered pH I need to have a model for it. That took some though. The easiest solution is to just assume that the base is sodium carbonate and that seems to work. In my hyper complicated spreadsheet I can use any base I want but the goal here is to see if a simple spreadsheet can handle alkalinity properly. If it winds up getting too complicated, I've failed.

Kaiser said:

This is because they have been using an outdated SRM to RA relationship.

Click to expand...

Yes. Several of the later spreadsheets have turned the gain down considerably on that (but not, AFAIK, the original). They seem to work fine for lighter beers (where the low color doesn't call for appreciable RA increase).

Kaiser said:

I also need to point out that the SRM based RA or pH prediction works just as fine as a grist based one as long as both models predict malt acidity data from the malts color. Only when you have specific pH and/or acidity measurements from malts can you expect to get significantly better accuracy with a grist based pH prediction model.

Click to expand...

Yes, the answer is only as good as the underlying model. I'm not convinced, given the variability of titratable acidity among grains nor the variability of beer colors that can be produced that a good enough model can be realized. Pearson's (I always try to work him in because Gosset worked in his lab before going to Guiness) r is too small. But I could be convinced if enough of a data set were collected. There will always be outliers but if r could be gotten up over, say 90%, such a model would be useable. Even poorer models can be informative though. I definitely don't want to discourage research into either color based or malt acidity based models.
[/QUOTE]

nilo said:

Kai, on you table below, when you say "roasted malt portion of special malts" for the columns, do you mean % of roasted malt to the total grain bill or to the amount of non base malts? Not sure I made my question clear.

Click to expand...

Yes, I get your question. This is most likely the most confusing part of this model. The initial definition of this value was "percentage of roased malt among the specialty malts". I.e. if you have 5% crystal and 5% carafa, the roasted % would be 50%. I'm currently rethinking this and may change is to % of color that comes from roasted malts. But either way, it gives you the ability to reduce the increase in RA that would be needed to keep up with the increase in SRM when roasted malts are used. It's the reason why a 25 SRM stout brewed with pale malt and roast barley needs as much RA as a 12 SRM amber that was brewed with pale malt and crystal malts.

Would there be a formula one could use to give max and minimum RA values from SRM and % of roasted malt?

Click to expand...

The model itself tries to estimate the distilled water pH of the grist. From that you would calculate the RA based on the mash thickness. There is a formula in the article, but its based on regression analysis and thus not as neat as an analytical formula.

A.J., so we are basically talking about authentic water vs. water/mash treatment that does basically the same thing to your beer. I agree with you on the authenticity factor. You need to dissolve chalk with CO2 to make Munich water. Or take my tap water, which is pretty close. I do like the idea of dissolving CO2 and by doing it under pressure in soda bottles it is not too difficult. But I do lack the arguments to strongly recommend this procedure to others.

With respect to the beer color and acidity models, how precise can we be and how precise do we have to be. I know that pH changes of 0.1 or even 0.2 don't affect taste all that much. I recently collected data from recent batches in a spreadsheet and compared the actual vs. predicted pH values and I have to say that most of the time the prediction was within 0.1 pH units:



green is <=0.1, yellow is <= 0.2 and red is >0.2

The red outliers are beers where lots of the pH is determined by the base malt and the correlation between base malt color and pH is not as strong as it is for specialty malts.

The Alt also had something odd going on. After adding acid to bring the pH down I got a much larger then expected pH drop.

The table shows 2 slightly different models. The model on the right hand side is currently implemented in my water spreadsheet.

Kai

Kaiser said:

A.J., so we are basically talking about authentic water vs. water/mash treatment that does basically the same thing to your beer.

Click to expand...

Not exactly. For example, Bohemian Pils was traditionally made with very soft water and was certainly a good beer and you can make a good Boh Pils that way but you can make it even better if you add calcium chloride beyond the levels found in Pilsen or Cseke Budejovice water. I recall my experiments with ales brewed with Burton and much softer water. The Burton water ale was (on both occasions I did this) deemed more authentic but the softer water beer was deemed a better beer.

I think the styles that evolved were a consequence of trial and error by brewers doing the best they could with the water they had. Today we can have any water we want and potentially do better than they did.

OTOH Michael Lewis apparently used to advise his students to work with the water they had - to make it a part of what made their beer their beer. He also wrote a paper called "pH pHooey".

Kaiser said:

I agree with you on the authenticity factor. You need to dissolve chalk with CO2 to make Munich water. Or take my tap water, which is pretty close. I do like the idea of dissolving CO2 and by doing it under pressure in soda bottles it is not too difficult. But I do lack the arguments to strongly recommend this procedure to others.

Click to expand...

I don't recommend it either and seldom use it myself. My main thinking is that if you are going to have a spreadsheet that allows you to add carbonate and bicarbonate to water it might as well do the calculations correctly.

Kaiser said:

With respect to the beer color and acidity models, how precise can we be and how precise do we have to be.

Click to expand...

Good questions. I would think model that could produce a standard error of less than 0.1 pretty good - about 70% of estimates would be less than 0.1, 28% would be between 0.1 an 0.2 and the remaining 3% greater than 0.2. Is that acceptable? Guess it depends on how you intend to use the info. For planning I think that's plenty accurate as long as you are checking actual pH values on brew day.

Kaiser said:

I know that pH changes of 0.1 or even 0.2 don't affect taste all that much.

Click to expand...

My experience was a little difference. I've monitored pH religiously for years but never did anything about it. When I got agressive the change was only 0.1 to 0.2 and I found the beers were dramatically better. All bright flavors emerged. Not that the old stuff was bad - the new is just way better.

Interesting data.

ajdelange said:

I think the styles that evolved were a consequence of trial and error by brewers doing the best they could with the water they had. Today we can have any water we want and potentially do better than they did.

Click to expand...

Agreed. I myself questioned the idea of brewing beer with water that tastes like drywall when I brewed a Bitter with much more gypsum than I usually use. While the beer turned out fine it got comments that it tasted a bit salty. I guess that was my last time trying to mimic Burton water.

I don't recommend it either and seldom use it myself. My main thinking is that if you are going to have a spreadsheet that allows you to add carbonate and bicarbonate to water it might as well do the calculations correctly.

Click to expand...

This may bring us back to the topic of this post. How well can a spreadsheet estimate the bicarbonate/carbonate contributions of undissolved chalk if it does not completely dissolve and the amount that dissolves depends on a number of factors that we may not be able to accurately account for?

Good questions. I would think model that could produce a standard error of less than 0.1 pretty good - about 70% of estimates would be less than 0.1, 28% would be between 0.1 an 0.2 and the remaining 3% greater than 0.2. Is that acceptable?

Click to expand...

Yes, that's what I would call good enough.

When I got agressive the change was only 0.1 to 0.2 and I found the beers were dramatically better. All bright flavors emerged. Not that the old stuff was bad - the new is just way better.

Click to expand...

where those beers done as side-by-side experiments where only the pH changed?

I have done one such batch, a decocted Maerzen, and while the mash pH differed by 0.13 pH units I could not tell them apart in a triangle test. But knowing which beer was which I thought the beer with the higher mash pH was a bit maltier. I plan to repeat this experiment with a larger pH difference. For some reason the pH difference was not as big as I planned.

Kai