Editor’s Note: This article was originally published on Brulosophy.com, a website dedicated to the understanding of brewing ingredients and methods through various “exBEERiments.”
by Marshall Schott
The first beer I ever made was an Irish Red Ale kit that came with a small baggie of a fine white powdery substance. Reading the instructions, I learned it was called “gypsum” and that I was supposed to add it to the boiling wort, that it would improve the flavor of the resultant beer. In the many subsequent extract and steeping grain batches I made, I forgot this mineral addition and never really noticed a difference (as if I was paying much attention). Following my transition to all grain brewing, water chemistry was the last thing on my mind, rather my focus was on fermentation temperature control and yeast health. Rightfully so! I’m hardly alone in thinking these 2 process components have contributed more than most anything, just short of proper cleaning and sanitation procedures, to the betterment of my end product. I made a couple fantastic beers, a few decent batches, and some total flops. As I learned the language of my system and honed my technique, I began to wonder what might be left for me to tweak to improve the quality of my beers, what small things I could change that would have a noticeable impact and maybe even bump my comp scores up a bit. I’d already started buying ingredients in bulk, vacuum sealing hops, and building a library of viable yeast. The only thing left was fussing with my water.
There was just something about this aspect of brewing that was so unappealing to me. While I’d played with adding gypsum and calcium chloride to my brewing liquor or boiling wort, I was hugely intimidated by the world of brewing water, what with its calculations and scientific jargon. Apparently, I wasn’t alone. According to a recent survey of over 2,000 homebrewers, a whopping 54.6% reported they use only unfiltered tap water to make their beer. I can totally relate, I ignored my brewing water for years and still produced the best beer on Earth according to my family and friends (“you gotta open your own brewery dude, you just gotta!”). But as I progressed in the obsession, I was influenced by those discussing the topic on popular podcasts, blogs, and magazines. My feelings of intimidation began to wane and I decided to wander down the path of alkalinity, mash pH, minerals, and acids… oh my.
After making this change, I accepted that my beers improved on faith, as I’d never compared similar beers brewed with waters of differing mineral composition. I didn’t notice any huge changes in my brewing efficiency, hop utilization, or beer clarity, as I expected, but I still believed my beers got better. The time had finally come to test whether my perceptions were right or if I was merely fooling myself, victim to the plethora of biases we unwittingly engage in to justify our decisions and beliefs.
Evaluate the differences between 2 beers of the same recipe mashed with either mineralized and acidified water or filtered but otherwise untreated water.
Based on responses to a survey I recently sent out, I chose to brew a German Pils for this xBmt, the style preferred by 54% of respondents.
|Batch Size||Boil Time||IBU||SRM||OG||Est. FG||ABV|
|5.5 gal||90 min||36||4||1.047 SG||1.009 SG||5.0%|
|Franco-Belges Belgian Pils Malt||8 lbs||86.5|
|Gambrinus Vienna Malt||1 lbs||10.8|
|Melanoiden Malt||4 oz||2.7|
|Perle||~25 IBU||60 min||Boil||Pellet||8.6|
|Tettnang||14 g/3.1 IBU||15 min||Boil||Pellet||5.8|
|Perle||11 g/3.7 IBU||15 min||Boil||Pellet||8.6|
|Tettnang||30 g/4.2 IBU||5 min||Boil||Pellet||5.8|
Given the fact manipulation of the variable in this xBmt would occur prior to the boil, it required the performance of separate mashes using the same ingredients and brewing methods. The grains were carefully measured out and milled the night prior to brewing, then I collected the full volume of water for each batch. I chose to use the no sparge method for these beers because I thought it most prudent to treat the entire volume of brewing liquor for the manipulated batch. In order to emphasize any differences water treatment might have on the finished beer, I used the Bru’n Water spreadsheet in conjunction with BeerSmith to determine the adjustments I would have to make to match a Dortmund profile, known for being rather high in mineral content. The Escobarian mounds of white powder would be enough to make Tony Montana say goodbye to his lil’ friend.
According to Bru’n Water, the differences between the untreated and treated water profiles were pretty drastic:
Untreated Tap Water Profile:
Treated Water Profile:
The salts and acid were added to one kettle and gently stirred until fully incorporated. The following morning, my rad little assistant helped mash-in, we staggered the start of each batch by about 20 minutes.
The fact I hit the same exact mash temp for both batches was encouraging and provided some confirmation my grain, water, and temperature measurements were good.
The estimated mash pH for the treated and untreated batches were 5.2 and 5.7, respectively, which was pretty close to my measurements at 10 minutes in.
The same exact volume of sweet wort was collected from each batch after a 60 minute saccharification rest and added to separate kettles.
Pre-boil SG measurements at this point indicated a very small difference, with the treated batch clocking in at .001 points lower than the untreated batch, which wasn’t expected. I proceeded to boil and kettle hop each batch, keeping all aspects as close as I’d imagine any sober dude in his garage is capable. Wort volumes at the end of the boil were exactly the same as measured by my trusty $3 hand-notched dipstick.
Just about the only thing I enjoy about the winter months is the fact my groundwater gets cool enough to allow me to chill to ale pitching temps, a luxury of sorts for homebrewers residing in warmer climates.
I took an OG reading at this point and confirmed there was indeed a very slight difference between each batch.
Of course, the difference could be a function of some extraneous variable, though my neurotic attempt to keep things as equal as possible compels me to believe it was related to the variable being tested. Either way, I thought it was an interesting observation. Each labeled carboy was then filled with its respective wort.
After about 4 hours of sitting in my cool ferm chamber, both worts had dropped to my target pitch temp of 50°F, after which the rehydrated yeast was pitched and my Black Box lager fermentation profile was engaged.
The following morning, both beers were showing signs of activity, though there were some observable differences throughout the course of fermentation.
The treated batch began to develop a krausen that was much whiter in appearance than the untreated batch.
After 10 days of fermentation, with the temp of the chamber sitting at 65°F, the krausen on the treated batch had dissipated while the untreated batch maintained a relatively thick cap of brown goo.
It was at this point I took an initial hydrometer reading and discovered both had reached my target FG, the untreated batch attenuating ever so slightly more than the treated beer.
I also found it curious that the treated batch sample was noticeably more hazy than the untreated batch.
While I detected no diacetyl at this time, both beers smelled subtly of sulfur, so I kept them in the chamber a couple more days before confirming FG had indeed been reached. The beers were then cold crashed and fined with gelatin before being packaged.
I let the beers lager and carbonate in my keezer for a week before presenting them to tasters.
A huge thanks to Colin Kaminski and his beautifully bearded brewer, Jake Lake, from Downtown Joe’s for participating in this xBmt.
How rad that one of the authors of the quintessential book on brewing water agreed to put his reputation on the line for science. As with all participants, Colin and Jake were both blind to the nature of the xBmt and had no clue it had anything to do with water. I’m sure many are curious how Colin did, if he was capable of detecting the beer brewed with treated water… unfortunately I’m unable to share that info, as doing so would break confidentiality. Rather, let’s allow the results to speak for themselves.
In all, 15 people participated in this xBmt, which would require 8 (p<0.05) correct responses to suggest a statistically significant difference. Each participant was provided 3 samples consisting of 1 beer from the treated batch and 2 from the untreated batch.
A rare occurrence indeed, this xBmt produced significant results with 9 (p=0.014) participants accurately selecting the single treated sample as being different. From this, we can deduce that the participants’ ability to distinguish between the 2 different beers is probably not entirely due to random chance, suggesting it is likely at least partly a function of the variable manipulated– water treatment.
As with every xBmt, those who correctly selected the odd-beer-out completed a second survey comparing only the 2 different beers, still unaware of the nature of the xBmt. While these results have been reported in the past, this is the first time they hold statistical value. Remember, only the 9 participants who were accurate on the triangle test completed this comparative evaluation.
Regarding aroma, 7 tasters reported perceiving the different beers as being somewhatsimilar while the others thought they were not at all similar. When asked which sample they preferred, 6 tasters chose the beer treated with minerals/acid and the other 3 chose the untreated sample. One participant commented that the beer mashed with untreated water had “more malt,” while another taster noted picking up some diacetyl in the untreated sample.
When it comes to flavor, all 9 tasters noted the different samples as being somewhat similar, with 7 preferring the sample brewed with treated water.
In terms of mouthfeel, 5 believed they were exactly the same and 4 felt they were somewhat similar. Of those who noted some similarities, preference was split evenly between the 2 beers.
Overall, these tasters preferred the beer brewed with treated water by a 2-to-1 margin. When provided the opportunity to guess what was different about the beers, only 3 participants responded and none were correct. One believed it had something to do with hop flavor and aroma, another felt there were different phenol levels, and the third thought it had something to do with the malt bill. Following revelation of the nature of the xBmt, each taster was asked to guess which beer they thought was brewed using only filtered water with no mineral additions or acidification. The majority (6) wrongly selected the beer brewed with treated water, which is interesting considering this is the same number of tasters who reported preferring the treated sample– perhaps some bias was at play here, who knows?
My Impressions: Count me among those who readily noticed a difference. Despite my obvious bias, I was rather easily able to distinguish the odd-beer-out in multiple blind triangle tests. I perceived the beer treated with minerals and acid as being considerably crisper and lighter on the palate with a pleasantly sharp hop bitterness that balanced the malt character well. The untreated beer was, in my opinion, way more malt-forward and lacking almost fully in that crispness I expect from a German Pils. Was it bad? I don’t think so. Different? Absolutely. It’s interesting that one taster noted diacetyl in the untreated batch, as I actually put both through 2 tests wherein I warmed the beer to encourage the expression of this chemical, neither time produced positive results. That said, my hypothesis is that in comparison to the treated beer, the pronounced bready character from the Belgian Pils malt was perhaps misinterpreted as diacetyl. Hmm.
Excited as I am that these results are statistically significant, it’s important to remember this is still only a single point of data and that the results arguably only hold water in the precise situation in which the experiment occurred- German Pils, Dortmund water profile, no sparge method, etc. Hence, using this xBmt alone as a basis for making decisions could be erroneous. While there exists a trove of anecdotal reports from folks claiming water manipulation contributed to a noticeable improvement in their beer, the fact is that a huge majority of homebrewers don’t mess with their water at all, many of whom I would imagine are making fantastic beer and perhaps even scoring well in competitions. Still, despite the caveats I mentioned, I personally believe these results are telling and rather exciting. I’ll admit, I’ve been a fence-rider when it comes to water treatment, engaging with many homebrewers in conversations about the subject over the years. This despite the fact I’ve regularly treated my water for the past few years. I guess I just wasn’t convinced the impact would be all that noticeable, at least enough to sway one’s preference. Now, I’m much more convinced.
As mentioned earlier, the method in which the water was treated for this xBmt was one of many, we’ve got a bunch more xBmts planned to test the other approaches as well! Until then, please share your experiences with treating your brewing water, I’d love to know what you think.