Discrepancy between EZ calculator and actual pH readings?

[snowveil]

Pardon my ignorance....I'm just starting to work through Palmer and Kaminski's Water book.

I recently purchased a pH meter (Milwaukee pH56 meter, as shown here) and I've been getting consistently lower pH readings than the EZ water calculator predicts I will get.

For example, the last beer I made, a German Schwarzbier (with a late-mash addition of Briess Blackprinz) had a predicted pH of 5.55 after an addition of 4g CaCl2 and 2g of MgSO4 to the mash (using RO water).

My pH readout came to 5.2 after a 15 minute settle of the mash, and about 5.15 after 30 minutes of mashing. I calibrated the meter using pre-packaged 4.01 and 7.01 solution immediately prior to taking the readings, and chilled the wort samples down to 75-ish degrees F.

The grist was as follows:
7lb German Pils
2lb German Munich
8oz Caramunich (60L)
6oz Briess Special Roast
10oz Breiss Blackprinz (Late mash addition, after taking pH readings)

Mashed at 149*F for 75 minutes with 4 gallons of water (1.6 qt/lb)


Could it be that the EZ Water Calculator spreadsheet was just that far off with predicted base malt pH? I understand that 5.2 is a well acceptable mash pH, but based on this prediction I don't want to predict a beer will come out at 5.2 and have it turn up at 4.9.

Thoughts?

 

[Hermit]

Have you tried other calculators? I don't use that one so don't know how well it does. If the error is consistent it really doesn't matter though. Adjust accordingly.

 

[ajdelange]

Could it be that the EZ Water Calculator spreadsheet was just that far off with predicted base malt pH?

Yes, easily. There is a robust algorithm for computing mash pH but the calculators don't use it and one of the reasons they don't is because the robust algorithm requires robust (accurate) malt titration data for the actual malts you are using (type, maltster, lot number). This data is hard to get. Try other calculators. Bru'n water is likely to come in low (relies on malt color, I believe, to estimate buffering). Brewers Friend, the most robust of the three most popular ones, will probably be closest to the truth because he uses a reasonably robust model (linearized buffering based on actual malt measurements but buffering is not linear). Also he doesn't have the data for your malt - just something more or less like it.

This is why it is best not to rely on calculators for mash pH prediction. Guidance, understanding of which does what yes but when it comes to brew day if you have a meter make a test mash and scale the necessary acid additions to the actual brew length.

 

[mabrungard]

While AJ believes that the Brewers Friend model is more robust, he is misapplying his understanding of malt buffering. While buffering is not linear across a wide range of pH, when we look at the very narrow range of pH that a typical mash exists at, that buffering is virtually linear. My results show that a linear assumption is quite valid and also provides insight as to why programs like Brewers Friend consistently overpredict the mash pH when operating in the high 4 to very low 5 pH range.

To the OP, I suggest that a review of your mash with the Bru'n Water model is likely to confirm your pH observation.

 

[MNDan]

I won't be using EZ water after this experience:

https://www.homebrewtalk.com/f128/mash-ph-over-time-does-drop-484454/

Brewer's Friend comes closest for me, and will be what I use in the future for initial estimates. I'll be recording the pH of future batches and comparing it to what the 3 water spreadsheets estimate to see how well it holds up, but so far it's Bru'n water -.25, Brewer's Friend +.04 and EZ Water 3.0 +.3 - fun stuff!

 

[ajdelange]

While AJ believes that the Brewers Friend model is more robust, he is misapplying his understanding of malt buffering. While buffering is not linear across a wide range of pH, when we look at the very narrow range of pH that a typical mash exists at, that buffering is virtually linear.

Of course it's linear in a narrow region. That's why the robust model is robust with only three terms in the Taylor series. But we are not only concerned with the narrow region around the desired mash pH. In order to compute proton deficit (or surfeit) we are concerned with the span from the intrinsic (distilled water mash pH) to the target (desired) pH. See Fig. 21 in Palmer's book. The non linearity isn't terrible and indeed some malts do exhibit a nearly linear titration curve. In the Palmer book examples on p91 the chocolate malt buffering is pretty constant (meaning the linear model is good) between it's DI pH and pH 5.6 which covers most target pH values of interest. Other malts are less linear. We also need to keep in mind that we aren't solely concerned with calculated values in the target range. If we do something foolish like use too much of a dark malt which would result in a calculated pH of 4.8 then we would like some assurance that the actual pH would be around 4.8 though obviously it is more important that we have accuracy with respect to the pH's we would actually hope to implement.

Of course the question is 'What are the implications if I ignore the linearity?' That's easy enough to do for a particular mash. Suppose I mash 80% Weyermanns pneumatic pils with 20% Weyermanns 60L caramel using water with alkalinity of 1.8 mEq/L and calcium hardness of 3.24 mVal (just some numbers that happen to be in my spreadsheet). Predicted mash pH is 5.31. Turning off the non linear coefficients in the malt titration model changes predicted pH to 5.34. Not a big deal in this case but other cases will be different depending on the malts.

My results show that a linear assumption is quite valid...

The linear assumption is OK, as we have seen above and works pretty well in a lot of cases. I even use it in my own spreadsheet when I want to see the effect of a small change in something(e.g. lets try 2.5% sauermalz instead of 3). The linear assumption is pretty good in those cases and is usually reliable enough. For accuracy I don't use the linear assumption. There is a cost for this, of course, and that is that an iterative solution is required. In Excel the Solver makes this trivial but mention it and people run screaming from the room.

..and also provides insight as to why programs like Brewers Friend consistently overpredict the mash pH when operating in the high 4 to very low 5 pH range.

I can remember one interesting case we discussed where Bru'n was predicting 4.9 and Brewer's Friend was just about right on (according to what the guy asking the question measured). A little analysis showed that even if the guy had used all sauermalz instead of one of his specialty grains he couldn't reach 4.9. That is, of course, 1 example but based on it it seems Bru'n water underestimates mash pH at the low end and this is the impression I have gathered from other peoples' posts but it is only an impression. I have not done side by sides and I do not know in detail what the algorithms that any of these calculators use. OTOH most people seem to find it quite good in the mid ranges.

To the OP, I suggest that a review of your mash with the Bru'n Water model is likely to confirm your pH observation.

I suggest checking both Bru'n and Brewer's Friend. But first I suggest checking your pH meter using the protocol spelled out at https://www.homebrewtalk.com/f128/ph-meter-calibration-302256/. Running some numbers on the specified grain bill suggests that using Weyermanns Floor Pils a mash pH of 5.59 might be expected whereas with their pneumatic pils it would be more like 5.47. Thus 5.55 isn't that bad an estimate given that we expect EZ to come in a bit high. Now please note that I have no idea which "German Pils" is being used (though both the ones I ran qualify) nor have any idea about what the characteristics of the other grains may be except to use the DI mash pH and estimates of the linear approximation to the buffering from Kai Troester's measurements on similar grains. But 5.2 is low enough that a cal check on the meter is, IMO, warranted.

I don't claim great accuracy. I haven't the temerity. This is really where the lack of robustness comes into play - none of the algorithms have good malt data to work with. I believe Kai understands the chemistry and he has at least made measurements. That's why I call his model the most robust. If I knew the details of any of the three I might have a different opinion.

 

[ajdelange]

I'll be recording the pH of future batches and comparing it to what the 3 water spreadsheets estimate to see how well it holds up, but so far it's Bru'n water -.25, Brewer's Friend +.04 and EZ Water 3.0 +.3 - fun stuff!

Excellent! What you are reporting so far is consistent with my impressions. I guess I should make clear that the 'Excellent' goes with your promise - not just that your data agrees with my impressions.

 

[-TH-]

Hi guys.

Based on my experience, usually when there is a moderate to large discrepancy between EZ's predicted ph and the measured ph (assuming the measurement is accurate ) it is due to the variation in the distilled water mash ph's of base malts from one maltster to another. For example, read these two posts from the EZ3.0 thread:

Some five months later and I am ready to share my findings. This process has been great and one that I have learned a great deal from.

Basically I took a bunch of my most commonly used grains and did single grain type mini-mash tests. 100g of lightly crushed grain were mixed in schott bottles with 250ml of 70C brewing water, bringing the temp to approx 67C. I placed the bottles in a water bath that kept it a constant 67C. I waited around 30 mins and then took samples from each. Let the samples cool to room temp and then took pH readings on a good quality (accurate and calibrated) pH meter. Some of the more interesting results:

Pils malt: EZ Water lists it at 5.75. I recorded: Bestmaltz Pils 5.64, Bests Boh Pils 5.59, Dingeman Pils 5.72.

Maris Otter: EZ Water lists it at 5.77. I recorded: Simpson Maris Otter 5.52.

There are plenty of other variations in other grains but these are the most important for me as I brew a lot of pale beers with my super soft water.

The end result: All mash pHs have been +/- 0.1 of the EZ Water calc with the new figures placed in it. I started using acid malt to drop the pH, but now I have graduated to using 88% lactic acid (calculated on EZ Water). The last brew I did EZ water calculated 5.43 after acidification and I measured 5.42. Couldnt be happier! I highly recommend this to anyone looking to fine tune their mash pH.

It also means I am making all the salt additions in the kettle (except for a little Gypsum in the mash) which I have found does great things for the taste. I am also looking a lot now at the final beer pH, and finding that with good mash acidification and sparge water acidification I am ending up with beers in the 4.0 - 4.4 range which seems to be the accepted industry range.

As always, it is a work in progress...but thanks EZ Water!

I modify the base grain pH in distilled water also based on 100g test mashes on my common base malts. The 2 that I've been using the most lately are [Crisp] Maris Otter & Castle Pilsner, both about 5.55. I've tested Rahr 2-Row and found it is also in the 5.6 range like EZ water makes note of. These are great base malts for brewing with RO water because a little specialty malt and brewing salts (CaCl2/CaSO4) bring the mash pH to the 5.4-5.45 range with RO water.

You probably knew this already but you can modify distilled water mash ph values in your copy of the spreadsheet by changing the values like shown:

Of course it would be best to do some experimenting of your own to determine what the distilled water mash ph's of the malts that you use are, but maybe you can search to see if anyone else has done this already. Then adjust your copy of the spreadsheet based on your findings and I think it will be much more accurate for you going forward.

Maybe in v4 I will have enough data so that the user can select base malt type AND malster. But until then...

Cheers!

 

[ajdelange]

Based on my experience, usually when there is a moderate to large discrepancy between EZ's predicted ph and the measured ph (assuming the measurement is accurate ) it is due to the variation in the distilled water mash ph's of base malts from one maltster to another.

With lighter colored (base) malts all having linear terms in the mid 30's (mEq/kg-pH) and with base malts being the majority of the malt in a mash it is little surprise that base malt DI pH has a major effect on mash pH (calculated or real) so that...

you can modify distilled water mash ph values in your copy of the spreadsheet by changing the values like shown...

...makes a big difference in the robustness of this algorithm. I did not know this was possible. Now let the user change the buffering coefficients and you are almost there.

Then adjust your copy of the spreadsheet based on your findings and I think it will be much more accurate for you going forward.

The observed +0.3 bias suggests that you might want to revise the default DI mash pH's down somewhat. Obviously lowering the defaults given by 0.3 would give some pretty ridiculous DI pH's which suggests that buffering capacities are too high. What are you using for them?

 

[dmr]

I won't be using EZ water after this experience:

https://www.homebrewtalk.com/f128/mash-ph-over-time-does-drop-484454/

Brewer's Friend comes closest for me, and will be what I use in the future for initial estimates. I'll be recording the pH of future batches and comparing it to what the 3 water spreadsheets estimate to see how well it holds up, but so far it's Bru'n water -.25, Brewer's Friend +.04 and EZ Water 3.0 +.3 - fun stuff!

Sounds like a good experiment. You may also like to compare your measurements with predictions from a mash pH calculator that I wrote, MpH Water Calculator 2.0 (an Excel spreadsheet), available at homebrewingphysics.blogspot.com (link below). As documented in a couple of papers on the blog, the calculations in MpH are based on Kai Troester's experiments. I find that MpH typically predicts mash pH close to the Brewer's Friend calculator. I'd be interested in any experimental results that you come up with.

FWIW, the above mentioned papers also discuss the mash pH models used in EZ Water and Brun Water.

Cheers!

 

[MNDan]

In the next month I hope to get to a Dubbel, a darker Pumpkin beer, and a 3rd try at a tasty Pale Ale that hits all my numbers for a lighter-colored beer. I'll let you know the results...

 

[snowveil]

Yikes! I didn't expect such a detailed reply. Thank you.

I'll look in to experimenting with the other software titles available, and also making some test mashes with my base malts (I typically purchase pale and pils malt by the sack).

My (albeit very basic) understanding of optimizing water chemistry on the easy level is:

1) Hit the Chloride : Sulfate ratio (keeping in mind ppm minimums for effectiveness) to accentuate either malt or hop character.
2) Adjust your mash pH using acid/salts in the mash as needed to hit the desired mash pH.

Does this sound about right, or am I missing out on other major factors? The Water book is kind of making my head spin with a plethora of information that, while I'm sure useful to some, seems overly complicated at my scale.

P.S. A.J., While reading the Water book on my lunch break today I found your name referenced. Awesome!

 

[ajdelange]

1) Hit the Chloride : Sulfate ratio (keeping in mind ppm minimums for effectiveness) to accentuate either malt or hop character.

Forget this one. Sulfate and chloride each have their own set of flavor influences. They are not antipodal. You cannot makeup for a chloride surfeit by adding more sulfate nor conversely. You really would be better off if you had never heard about this. The is something to it in the sense that two degrees of freedom can be expressed by giving two numbers or by giving one number and telling you it is in a given ratio to the other but focusing on establishing a particular Cl:SO4 ratio has the potential to lead you astray. Learn the proper level of sulfate for a particular beer and the proper level of chloride. The ratio will be what it will be.

2) Adjust your mash pH using acid/salts in the mash as needed to hit the desired mash pH.

This one is critically important.

P.S. A.J., While reading the Water book on my lunch break today I found your name referenced. Awesome!

Yes, I was able to help the guys with a couple of things.