How Exact is Your Adjusted Water Profile?

Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum

Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Cardog

Well-Known Member
Joined
Mar 1, 2010
Messages
89
Reaction score
24
Location
San Diego
As the rest of my brewing process is perfected, I have been learning about water profiles. Based on my crappy water report I will just end up starting with RO water.

Now that I am pretty sure that I have figured out Bru'n Water and what it entails, it's time to perfect the water profile

Based on Bru'n Water's profile (color, malty-bitterness) I can get the water profile to the exact mineral content right down to less than 1 ppm across the board on every profile with the use of 4 or 5 mineral additions.

My question is, at what degree does everybody adjust their water to? I see a lot of Bru'n water examples that people post here and some minerals are at 0, some are 20 ppm higher or lower. Then there's ...add a teaspoon of this and a teaspoon of that.... I gotta assume the profiles of Bru'n water are pretty accurate to the beer style one wants to brew. So what gives? For $40, for a what seems like a lifetime supply of minerals, why not make it exact? I can dial the exact profile in less than 5 min of trial and error. Love to hear what people have to say why they don't.
 
I think the reason why many don't do it is because these "reference" profiles are dodgy at best. They are based on some historical sample taken God knows when, and they are often representative of un-adjusted water - not the water actually used to brew the beer in that place. Obsessing over pinpoint accuracy is like aiming for a target that's no longer there.

Plus it's totally unnecessary to match them perfectly. It's arguably more useful to your brewing skill/knowledge to learn how the minerals affect mash pH, beer texture, and - to some extent, flavor - rather than blindly matching these profiles.

Finally, if you do take a clean slate approach to water treatment and learn how the salts affect the final product, some of the profiles will reveal themselves as being pretty extreme.
 
Fortunately, the profiles in Bru'n Water were researched and verified to be at least representative of water from the areas of interest.

I never bother to get ion concentrations exactly to the levels reported in the program. For most ions, getting within 5 or 10 ppm +/- is good enough to me.
 
I am in the same boat. I am going to brew with my water at my new house for the first time, this evening. It is also the first time adjusting the water profile as I received my water report from Ward Labs, yesterday. However, I read a lot about the effects of the different dissolved ions and decided to make my own custom profile based on the style of beer I am brewing and what I have read rather than trying to match one of the preexisting profiles. Rather than blindly matching one, I figure that I can target specific goals that I set for real reasons.
 
I also don't bother with established profiles. I have a few baselines that I like to target based on the balance of the beer for sulfate, chloride, and sodium (hop forward, malt foward, roasty, balanced, and soft are my basic ones), I basically never touch magnesium, and calcium/alkalinity/acid are adjusted to hit the pH I want (which also varies depending on what I brew, dark and roasty usually higher, 5.5 room temp, pale and crisp lower at around 5.3 room temp, most beers I target somewhere in between).

And I'm also in the +/- 5-10 ppm boat. Unless you're confident that your water is consistently the same (basically DI or RO water, I wouldn't even trust purchased non-DI/RO bottled water to be that consistent), or you're testing it yourself every time, there's little point trying to be precise when you've got a moving starting point.
 
Based on Bru'n Water's profile (color, malty-bitterness) I can get the water profile to the exact mineral content right down to less than 1 ppm across the board on every profile with the use of 4 or 5 mineral additions.

Curious as to how you do that.

I can dial the exact profile in less than 5 min of trial and error.

Trial and error? Really? Yes, the profiles can be realized exactly to less than 0.0001% error if you ignore the bicarbonate (unless you synthesize at pH 8) but I'm surprised you're able to do that by trial and error!



Love to hear what people have to say why they don't.
If you are reproducing the profiles exactly you will have noticed that many of them require the use of calcium lactate or calcium phosphate and/or the sodium salts of those acids realized by adding those acids and sodium bicarbonate, calcium carbonate and calcium hydroxide. The amounts of acid required to realize the specified sodium and calcium levels are often rather large, especially at a reasonable synthesis pH, larger than I would want to put into a beer. Besides which it doesn't matter. As natural systems seem to respond geometrically rather than arithmetically and since it takes about a 3 db (doubling) in sound or light level to be redily detectable we hypothesize that it's the same with taste and that you wouldn't notice the difference between 100 mg/L sulfate and 150.

Yes, its cool to be able to punch the button and have the exact (except bicarb that you don't care about) synthesis fall out in 20 sec when you used to have to wait overnight for that to happen as was the case when I first started matching profiles, but it really isn't necessary.
 
Well AJ, with a little trial and error on the mineral input it's not hard to get it perfect. Here is an example.....

BrunWater.png

While I agree with McKnuckle that is it more useful to use your brewing skill/knowledge to learn how the water profile affects the beer taste for certain styles, I am not quite there yet. So therefore using Bru'n Water's built in profiles would be a great start.

Although some of you use a +/- 5-10 ppm rule as the water may not be consistently the same every time anyways. So why not reduce your margin of error. I mean if the water is off 5 ppm one day and your target if off 10 ppm in the other direction that could mean a 15 ppm difference. Or maybe some of you use that rule cause it's "close enough" and not that important. Well heck, some of the profiles listed in Bru'n Water only change by 10-15 ppm.

Once you know what profile you are after, it just doesn't take that much effort to get it spot on with the proper salts on hand after starting with a clean slate like RO water.

Maybe I am a perfectionist, maybe I am overthinking it and it just isn't a big deal as I am making it out to be.......
 
Maybe I am a perfectionist, maybe I am overthinking it and it just isn't a big deal as I am making it out to be.......

It's awesome that you're paying attention to water. Not enough brewers do that. And it really does make a big difference in the beer.

Hell, those folks over at Brulosophy, brewing water was one of the few areas they actually could get a statistically significant difference under their methods.

However, I wonder if you're not losing the forest through the trees. Is your manner of measuring salts accurate enough to get you to <1ppm? Is your source of water consistent <1ppm? Are you accommodating for the absorption of moisture in your calcium chloride?

Anyone can punch things into a calculator and get an ideal result. Obtaining it in the real world is a different story.
 
However, I wonder if you're not losing the forest through the trees. Is your manner of measuring salts accurate enough to get you to <1ppm? Is your source of water consistent <1ppm? Are you accommodating for the absorption of moisture in your calcium chloride?

Anyone can punch things into a calculator and get an ideal result. Obtaining it in the real world is a different story.

My goal wasn't necessarily just to get it within 1 ppm but really to reduce the margin of error for things like absorption rate, water consistencies and I am sure we have been off by a quart of water filling our kettles.
 
Just to follow up on my post, things went well last night. Between using Bru'N Water, the additions (I added gypsum, epsom salt and canning salt) and some acid malt, I got one of the better brew house efficiency yet. I didn't do the math, but I ended up with a higher gravity than expected (including the acid malt) from a recipe I have brewed multiple times without these additions. That hasn't usually happened for me. So far so good. Now to wait until fermentation is done to see the result on the flip side.

I was measuring my additions to 0.1g resolution. I am sure there is some error there, but it should be in the right ballpark. I do agree that starting with exact numbers will reduce error stack-up. However, if a parameter does not need to be that exact, then no big deal. For example, if your sulfate ends up +10 or -10 ppm from what you wanted, are you really going to notice?
 
Well AJ, with a little trial and error on the mineral input it's not hard to get it perfect. Here is an example.....
Perfect means 0 error and it is possible to do that if you ignore the bicarbonate. If you add the salts to DI water that you specified you get close but the pH is 11.3. I don't think you want pH 11.3. At that pH the water is oversaturated with respect to CaCO3 and you may get precipitation.

If you come to me and say "A. J., I want this profile" and hand me a list of ion concentrations the first thing I am going to say is "And at what pH do you want this water". A reasonable answer for a brewer, IMO, is "Mash pH, of course". At mash pH alkalinity is no longer a factor and that is why it is a reasonable pH to shoot for. So here is a perfect synthesis for this profile at pH 5.4 from ion free water.

Salt/Acid/Base mg/gal Synth
CaCl2.2H2O 239.39
NaCl 0.25
MgCl2.6H2O 80.93
CaSO4.2H20 119.23
MgSO4.7H20 285.72
H2O (DI) Liters 0.00
CaCO3 0.00
NaHCO3 276.28
CO2 0.00
HCl 0.00
Ca(OH)2 247.90
Na2CO3.H2O 0.00
Sodium Lactate 0.00
Potassium Lactate 0.00
Lactic 892.87
Sulfuric 0.00
88% Lactic ml/Gal 0.8419

Note that a fair amount of acid is required to keep the pH from going to 11.3. Since this is a Black something profile it is quite possible that the intention is that some or all of that acid may be supplied by dark malts.

Anyway, if you are just starting out then this is probably beyond where you want to be having discussions at this point so I'll shift gears. I couldn't get sodium to match up and then noticed you have specified 8 mg/L for your RO water. Is it really that high. RO membranes are tweaked to be especially good at rejecting sodium as most systems are proceeded by a sodium cation exchanger. If your water going in has 100 mg/L sodium and the calcium and magnesium levels you specify the water going to the membrane is going to be at about 240 but the membrane should block 98% or a little more of that so your permeate sodium should be no higher than 4.8. I just checked my system which is now at least 7 yrs old and has had 24,000 L of permeate run through it. It still rejects 98.9% of sodium ions. Were your system as good as mine (and I would expect it is unless it is even older than mine) your RO water sodium would be 2.64 mg/L.


While I agree with McKnuckle that is it more useful to use your brewing skill/knowledge to learn how the water profile affects the beer taste for certain styles, I am not quite there yet. So therefore using Bru'n Water's built in profiles would be a great start.
I agree with him too. Use the supplied profiles as rough guidelines. As the remarks about pH and acid above should have made clear, those are not real profiles. They are representative and as such useful.


Although some of you use a +/- 5-10 ppm rule as the water may not be consistently the same every time anyways.
When working with water chemistry you should use a 5 - 10% rule. 10 ppm error in reproducing a sulfate level of 20 is very different from a 10 ppm error in shooting for one at 300. An error of 5% is much more meaningful and as I noted in my previous post natural systems seem to respond to stimuli geometrically, not arithmetically.


So why not reduce your margin of error. I mean if the water is off 5 ppm one day and your target if off 10 ppm in the other direction that could mean a 15 ppm difference. Or maybe some of you use that rule cause it's "close enough" and not that important. Well heck, some of the profiles listed in Bru'n Water only change by 10-15 ppm.
And you could be off by 6 ppm on sodium simply because of uncertainty in what your RO system does. Another factor to consider is that errors don't add arithmetically. They add in the mean square sense. Thus if there is an uncertainty of ±20 mg/L in your water's sulfate content and an uncertainty of ±20 mg/L in the sulfate level of your synthesis the uncertainty in the treated water is going to be 28 mg/L. Now if you go for a perfect synthesis so that unceratainty is 0 your overall uncertainty is still ±20 mg/L. Put another way, the largest error source dominates the error budget.


Maybe I am a perfectionist, maybe I am overthinking it and it just isn't a big deal as I am making it out to be.......
The main thing to keep in mind is that it is kind of fun to come up with algorithms for optimization. I've got a very elaborate spreadsheet for finding exact (and I do mean exact here) salt additions. I can decide whether to minimize the geometric or arithmetic error, whether I want to consider calcium more important than chloride and by how much if I do and so on. Thing is I never use it except to answer questions here.
 
Anyway, if you are just starting out then this is probably beyond where you want to be having discussions at this point so I'll shift gears.

This is definitely true so bare with me


I couldn't get sodium to match up and then noticed you have specified 8 mg/L for your RO water. Is it really that high. RO membranes are tweaked to be especially good at rejecting sodium as most systems are proceeded by a sodium cation exchanger. If your water going in has 100 mg/L sodium and the calcium and magnesium levels you specify the water going to the membrane is going to be at about 240 but the membrane should block 98% or a little more of that so your permeate sodium should be no higher than 4.8. I just checked my system which is now at least 7 yrs old and has had 24,000 L of permeate run through it. It still rejects 98.9% of sodium ions. Were your system as good as mine (and I would expect it is unless it is even older than mine) your RO water sodium would be 2.64 mg/L.

I am glad you confirmed exactly what I was thinking and will make that input adjustment. The example shown was the default for RO water in the Bru'n Water program which why it shows salts left in.


When working with water chemistry you should use a 5 - 10% rule. 10 ppm error in reproducing a sulfate level of 20 is very different from a 10 ppm error in shooting for one at 300. An error of 5% is much more meaningful and as I noted in my previous post natural systems seem to respond to stimuli geometrically, not arithmetically.


This I would agree more so with percentages than ppm.

And you could be off by 6 ppm on sodium simply because of uncertainty in what your RO system does. Another factor to consider is that errors don't add arithmetically. They add in the mean square sense. Thus if there is an uncertainty of ±20 mg/L in your water's sulfate content and an uncertainty of ±20 mg/L in the sulfate level of your synthesis the uncertainty in the treated water is going to be 28 mg/L. Now if you go for a perfect synthesis so that unceratainty is 0 your overall uncertainty is still ±20 mg/L. Put another way, the largest error source dominates the error budget.

Sure, you can not calculate those uncertainties or unknowns but at least you can eliminate the error on what you do know with a program like Bru'n Water.

I didn't want this to be a technical discussion but wanted to point out why anybody wouldn't adjust the mineral content to the profile they want exactly. I guess what I am trying to say is with the proper salts on hand it's really not that hard or take any more effort to go from getting within 10% to less than 1 ppm..... so why not? Do people here just lick their finger hold it up and say F**K it close enough, let's brew! I mean it takes another 1 min tweaking with Bru'n Water to get spot on. You have a program that Martin graciously produced (BTW, I will be a paid supporter) so why not take advantage to eliminate those known errors.
 
Well there are limits. Bru'n water approximates the relationship between alkalinity and bicarbonate by bicarbonate = 61*Alk/50 in some places and apparently not in others. It doesn't ask you what temperature you want analysis/synthesis done at and it doesn't take ionic strength of solutions into account (and neither do any of the other calculators I know of). Certainly in terms of available computing power those small errors could be dealt with but this would add so much to the complexity of the programs that you, the average user, would be appalled.

The thing with the profiles, however, is that the the targets they represent aren't necessarily good targets. For each beer you brew there is an optimum ion profile but it isn't the one that's in the spreadsheet (or any of them). It is the one that gives you the beer you like best and you are going to have to find that on your own.
 
I am sure we have been off by a quart of water filling our kettles.

This is a detail that should be emphasized... because if one measures a perfect weight of various salts, but one pours water haphazardly into vessels, one might as well just roll the dice.

I weigh my water in kilograms (which equal liters by the way), rather than relying on visual volume indicators. I consider the margin for error much smaller when doing it this way. This is not just for achieving a mash pH and various mineral concentrations of course; it benefits all aspects of brewing that rely on water measurement.
 
Looks like I'll have to start cooling my water to 4 C and get a kg scale that measures to 6 sig figs ;)
Considering my water-noob-ness, I'm generally happy with most ions being close (a bit beyond +/- 10 ppm), with an appropriate pH.
 
if you do precision density measurements (e.g. determine the alcohol content of your beer) you know that number as well as you know your wife's birthdate. BTW, I am all for weighing water and for a closely related reason. A kg of water is a kg of water at whatever temperature even though it is only close to a liter at 3.8 °C.

I again encourage you to think of tolerances in %. A 10 ppm error re 300 mg/L sulfate is trivial. A 10 ppm error in sulfate in a Bohemian pils could be a disaster as it may represent 100% error.
 
Back
Top