Problem with Bru'n Water and EZ Water spreadsheets

[pensphreak]

I put in (what I think) are the same values for an upcoming Oktoberfest that I'm brewing and the output from both of these spreadsheets is pretty different (5.49 vs 5.2 pH). Can someone take a look at my spreadsheets and tell me what I'm doing wrong?

Background: I'm brewing with 100% RO water and normally I just follow the primer (1g CaCL2/gal of water). I batch sparge with (in this case) 4.88gal of mash water and 5.69gal of sparge water.

Amt 	Name
5.00 g 	Calcium Chloride
5 lbs 	Pilsner (2 Row) Ger (2.0 SRM)
4 lbs 	Munich Malt (9.0 SRM)
3 lbs 	Vienna Malt (3.5 SRM)
1 lbs 	Caramunich Malt (56.0 SRM)

Here's my recipe: http://pensphreak.com/homebrew/Oktoberfest%202.0.html
Here's the 2 spreadsheets: Bru'n Water and EZ Water

 

[DSmith]

Your using the tools correctly. Been through at least 12 mashes looking at both calculators. For me, brewing with 100% RO water, EZ Water has been closer to what I measure with a calibrated pH meter with 0.01 pH resolution. No spreadsheet has been consistently accurate and it depends on the malt/brand used.

I advocate for a very small (100 gram total) mash with your brewing water to test for mash pH and change your acidification as needed. It can be done while heating mash water and have caught potential mash pH issues before committing to acidifying the mash. It takes a fine gram scale, coffee grinder, small thermos & calibrated pH meter. Testing the mash pH becomes a formality rather than a measurement that you need to react to with some corrective measure.

 

[mabrungard]

That RO water has little alkalinity and the CaCl addition pushes RA down pretty low. With the color of that grist, I'm not surprised with the prediction of a low mash pH. I hope you have a pH meter and can report back here with your result. From my experience, its much more likely that the mash pH is going to be low than high.

 

[diegobonatto]

Did you converted the color unit for Bru'n Water? As you showed in the Bru'n Water spreadsheet, you put the malt color in SRM (EBC) units instead off Lovibond. You can convert the units using the formula:
SRM=1.3546*Lovibond-0.76

 

[pensphreak]

That RO water has little alkalinity and the CaCl addition pushes RA down pretty low. With the color of that grist, I'm not surprised with the prediction of a low mash pH. I hope you have a pH meter and can report back here with your result. From my experience, its much more likely that the mash pH is going to be low than high.

I converted it to Lovibond and it was closer, but still not 100% equal. On my last brew, I checked my pH (cheapo Milwaukee pH600) and it was 5.3. Bru'n said it would be 5.3 while EZ Water said it would be 5.46. So I'm thinking if I go with Bru'n Water again for my next brew (est. 5.3), I should be fine without any acid malt additions right?

 

[diegobonatto]

I converted it to Lovibond and it was closer, but still not 100% equal. On my last brew, I checked my pH (cheapo Milwaukee pH600) and it was 5.3. Bru'n said it would be 5.3 while EZ Water said it would be 5.46. So I'm thinking if I go with Bru'n Water again for my next brew (est. 5.3), I should be fine without any acid malt additions right?

Yes, you will be fine without acid malt.

 

[pensphreak]

Yes, you will be fine without acid malt.

Hopefully my next batch pH matches Bru'n water again

 

[DSmith]

No acid malt seems pretty safe for this recipe and 100% RO water. Even if your pH matches EZ Water, no acid malt doesn't make for that high of mash pH.

Please complete the thread with your measured room-temp pH with calibrated meter.

 

[ajdelange]

You shouldn't be surprised or concerned. Spreadsheets and calculators use models and while some models are better than others in particular situations the converse may be true in other situations. No model is perfect. In brewing there are simply too many degrees of freedom to allow for perfect prediction of mash pH.

 

[pensphreak]

So I made a small change to the Bru'n water spreadsheet, and the expected pH was 5.4. My pH600 meter measured the mash right on at 5.4

 

[diegobonatto]

Nice!

 

[ajdelange]

That's the nice thing about models. If you don't like the answer you just change the model or the input data (I used to do this for the gov't which is how I got to be so cynical).

Also bear in mind that a reading from an instrument whose precision is 0.1 cannot be more accurate that 0.028 even if the instrument carries out calibration to infinite precision on perfect buffers. Most people use technical buffers accurate to 0.02 pH (you did recalibrate just prior to taking the measurement, didn't you?). A perfect meter with 0.1 precision using 0.02 buffers is not capable of better accuracy than 0.035 pH. This means that if the instrument reads 5.4 the actual reading is between 5.36 and 5.435 with probability 68% at best and between 5.33 and 5.47 with probability 98%. And in fact the accuracy is worse than that because a meter than can calibrate to better than 0.1 does not display it's answers rounded to 1 decimal place.

Thus your mash pH is not 'right on' at 5.4 nor is Bru'n Water's prediction. Unfortunately most people think that if a device measures a value that this is truth. I really think some exposure to this should be mandatory in schools.

The real message here is that while a pH meter that has a precision of 0.1 is more valuable to a brewer than pH strips which have a precision of 0.3 and a bias of 0.3 better than 0.1 is really needed before a pH meter becomes useful to a brewer.

[Edit] Just checked on the calibration procedure for this unit. It is a single point calibration at pH 7 and the unit does not have ATC. Single point calibration is OK for solutions near pH 7 but for values near 5.4 if the temperature of the test solution is different from whatever temperature they (the manufacturer) chose when picking a slope you could have appreciable error especially as the electrode ages.

 

[pensphreak]

That's the nice thing about models. If you don't like the answer you just change the model or the input data (I used to do this for the gov't which is how I got to be so cynical).

Also bear in mind that a reading from an instrument whose precision is 0.1 cannot be more accurate that 0.028 even if the instrument carries out calibration to infinite precision on perfect buffers. Most people use technical buffers accurate to 0.02 pH (you did recalibrate just prior to taking the measurement, didn't you?). A perfect meter with 0.1 precision using 0.02 buffers is not capable of better accuracy than 0.035 pH. This means that if the instrument reads 5.4 the actual reading is between 5.36 and 5.435 with probability 68% at best and between 5.33 and 5.47 with probability 98%. And in fact the accuracy is worse than that because a meter than can calibrate to better than 0.1 does not display it's answers rounded to 1 decimal place.

Thus your mash pH is not 'right on' at 5.4 nor is Bru'n Water's prediction. Unfortunately most people think that if a device measures a value that this is truth. I really think some exposure to this should be mandatory in schools.

The real message here is that while a pH meter that has a precision of 0.1 is more valuable to a brewer than pH strips which have a precision of 0.3 and a bias of 0.3 better than 0.1 is really needed before a pH meter becomes useful to a brewer.

Yes, it was calibrated with 7.01 and 4.01 solutions prior to testing. Also, since I don't really care THAT much about the EXACT number, the inaccuracy of this meter is good enough for me. Even if it reads 5.4 and the accuracy is +- 0.1, that still puts me in the ballpark for a good mash. And that's all I really care about right now.

 

[ajdelange]

That's fine and much better than strips. I just really want people to understand how this works.

You mention calibration with 4.01 buffer. Since the unit only takes a single point calibration I'm a little curious as to how you used the 4.01 buffer. If you set the offset (what you do when you adjust for 7.0) and then go into the 4 buffer and the meter reads 4.0 then you are lucky - the slope set into the electronics matches the slope of the electrode. In such a case you can be pretty confident that the readings half way between 7 and 4 are as good as can be expected given the precision (if temperature of test solution and buffers are the same or close).

 

[pensphreak]

That's fine and much better than strips. I just really want people to understand how this works.

You mention calibration with 4.01 buffer. Since the unit only takes a single point calibration I'm a little curious as to how you used the 4.01 buffer. If you set the offset (what you do when you adjust for 7.0) and then go into the 4 buffer and the meter reads 4.0 then you are lucky - the slope set into the electronics matches the slope of the electrode. In such a case you can be pretty confident that the readings half way between 7 and 4 are as good as can be expected given the precision (if temperature of test solution and buffers are the same or close).

Yeah, I calibrated it with the 7.01 and then used the 4.01 to check.

 

[ajdelange]

You can even do a de facto 2 point calibration. When you adjust in pH 7 buffer you set the offset. When you do the 4 check the meter should read 4.0 i.e. the slope should be 1 pH/pH. If the meter reads something other than 4.0 the slope isn't 1 pH/pH but something else. If, for example, as the meter ages, you might read 4.2. The slope at this point is (7-4.2)/3 = 0.933 pH/pH. As the meter reads

reading = 7 + slope*(pH - 7)

the true pH given a particular reading is

pH = (reading - 7)/slope + 7.

Thus if you read sample as 5.4 after a pH 4 buffer reading of 4.2 the pH is (5.4 - 7)/0.933 + 7 = 5.28. There are still plenty of error sources but at least you have taken care of slope error. And the temperature of the 2nd buffer and the sample should be the same.