Author: Marshall Schott
Arguably, the single most important component of brewing, responsible for converting the boring starches in a kernel of grain into amazingly sweet fermentable wort, is the mash. Without it, we couldn’t make the immensely delicious beers we make today. As every all grain brewer is very well aware, there are many variables to consider when deciding how to approach mashing a particular beer, nearly all focusing in one way or another on the variable of temperature. Real scientists have long since understood and spread their knowledge of how mash temperature impacts the resultant wort and finished beer, and since I’ve no interest in pretending to be as smart as them, I won’t go into too much detail (geek-out here, here, and here).
For the purposes of this xBmt, my focus will remain solely on single infusion mashing, characterized by steeping milled grain at a single temperature for a given amount of time. During this so-called saccharification rest, the starches are converted to fermentable sugars by 2 main enzymes, here’s a dumbed-down version how they work:
Beta-amylase is said to be most active between about 131°F-150°F (55°C-66°C) and chops up whatever the hell gets in its way into smaller bits of mostly maltose, a highly fermentable sugar. Because of this, lower mash temps have been shown to produce more fermentable worts.
Alpha-amylase is quite different in that it is more active at warmer temps ranging from about 154°F-162°F (68°C-72°C), converting starches into maltose as well less fermentable sugars and unfermentable dextrins.
In essence, a lower mash temp purportedly produces a beer with a lower FG that’s dry with a thinner body and crisp mouthfeel, while a beer mashed warmer is said to finish with a higher SG and be sweeter with a fuller body. Obviously, there’s far more to mashing than this, but that’s not really what I’m hear to discuss, rather I’m curious about the impact mash temperature has on perceptible differences.
| PURPOSE |
To evaluate the impact mash temperature has on 2 beers of the same recipe mashed at drastically different temperatures– 147°F and 161°F.
| METHODS |
Thanks to HomeBrewSupply.com for providing the recipe kits used for this xBmt. During chats with the guys from HBS, they were a bit surprised to learn I’d never brewed Biermuncher’s Centennial Blonde Ale, which sits in the #1 position on HomebrewTalk’s Top 10 Recipes list. The kits arrived well packaged with all of the ingredients marked, making them easily identifiable, and the 2 packs of Wyeast 1056 American Ale yeast were wrapped with ice packs.
My first order of business was making a yeast starter a couple days prior to brewing, I pitched both smack packs into a single 2 liter flask with plans to split it between each fermentor after harvesting some for future use.
I milled the grain and collected the full volume of water for both no sparge batches the night prior to brewing. I awoke the following morning and began to heat both kettles of water to strike temp, which took about 5 minutes longer for the high mash temp batch. To keep things from getting too confusing, I still staggered the start of each batch by about 10 minutes.
While I’d considered comparing a simple 10°F difference in mash temp, I ended up changing my mind and settling on the more extreme ends of the single-infusion mash range.
Each mash was allowed a 1 hour saccharification rest before the sweet wort was collected and transferred to kettles.
This is where I noticed the first rather obvious difference between the batches, with the high mash temp wort developing a rather large and persistent layer of foam.
Even as the wort was boiling, the foam stuck around, making be wonder if this was an indication of how the finished beers might differ in terms of head retention.
Hops were added at the appropriate times throughout the boil then the wort was chilled and racked to 6 gallon PET carboys. A post-boil OG comparison confirmed my process was pretty consistent between the batches.
Both fermentors were placed in my temp controlled chamber and allowed to finish chilling to my target fermentation temperature, it took about 2 hours to reach 66°F, at which point I pitched the yeast. Signs of activity were present in both just 12 hours later.
I began to notice some differences in appearance at the 18 hour mark, with the kräusen on the high mash temp beer staying fairly low with larger bubbles while the low mash temp kräusen was beginning to grow.
Sure enough, the low mash temp batch ended up making a mess 24 hours into active fermentation.
I swapped the airlock with a blowoff rig and left it on until the kräusen dropped a bit, maybe 8 hours or so. The high mash temp beer was tame throughout fermentation.
I took the first hydrometer measurement once signs of fermentation activity had waned, then confirmed it hadn’t changed the following day, 6 since brewing the beer.
The .009 SG difference was initially pretty shocking to me though ultimately left me feeling confident such a highly discussed process component actually did what it’s purported to do– science works! I proceeded to cold crash, fine with gelatin, and package in kegs.
To carbonate, I hit each keg with 40 psi for 18 hours before purging and reducing to 13 psi for serving. By the time I began to collect data the following weekend, both beers were carbonated and looking nice.
| RESULTS |
A total of 20 people participated in this xBmt including many BJCP judges, obsessed homebrewers, and certifiable craft beer nerds. A discriminative sensory analysis method referred to as a triangle test was used to determine distinguishability and involved each participant being served 3 samples in different colored paper cups, 2 from the low mash temp batch and 1 from the high mash temp batch, then asked to select the different beer. In order to reach statistical significance given the sample size, 11 participants (p<0.05) would have had to accurately select the high mash temp sample, though only 9 (p=0.134) were capable of doing so, implying a general inability for people to reliably distinguish between a beer mashed at 147°F from another mashed at 161°F.
The 9 participants who made the correct selection on the triangle test were asked to complete an evaluation comparing only the 2 different beers, though they remained blind to the nature of the xBmt. It’s important to interpret the following information with caution, as the failure to achieve statistical significance means those who completed this evaluation may have done so by chance. Also, responses were all over the board on this one…
Four tasters perceived the aroma between the beers as being similar while 3 thought they smelled exactly the same and 2 felt they were not at all similar. Of those who reported noticing a difference, 5 preferred the aroma of the high mash temp beer and 1 liked the low mash temp beer more. Regarding flavor, 6 of the 9 tasters experienced the beers as tasting somewhat similar, 2 said they were not at all similar, and 1 thought they were exactly the same. Things get a little more interesting when it comes to mouthfeel, as not a single taster perceived the different samples to be exactly the same, with 6 saying they were somewhat similar and 3 believing they weren’t at all similar. A pretty striking majority of 7 tasters reported preferring the mouthfeel of the high mash temp beer compare to only 2 who liked the low mash temp beer’s mouthfeel better. Overall preference was split with 5 and 4 tasters expressing preference for the low mash temp beer and high mash temp beer, respectively.
At this point, the nature of the xBmt was revealed to the 9 tasters and they were then asked to select the one they believed was mashed warmer. Only 4 chose the correct sample.
My Impressions: I want to send a hearty cheers to Kevin “BierMuncher” Mattie for designing such an easy-drinking and flavorful beer! This was perfect to have around for the last couple weeks of summer, which is about as long as it lasted, as everyone who came over while it was on tap couldn’t keep their hands off. Regarding my experiences trying to distinguish between these beers, and even with my obvious bias, I honestly could not tell them apart. As is often the case, I was almost convinced I could detect some very slight differences when I first started pulling pints for myself, but this feather of confidence quickly drifted away as I began to attempt triangle tests served to me by others. At the very least, I expected to perceive a fuller body in the high mash temp beer, but side by side, it was essentially impossible for me to tell them apart.
| DISCUSSION|
I reckon every all grain brewer is at least somewhat aware of the impact mash temperature has on wort, most particularly that lower temps increase fermentability while higher temps decrease it. The objectively observable differences between the beers in this xBmt corroborate this scientifically valid fact– the higher beta amylase activity in the low mash temp beer resulted in a FG of 1.005 while the increased alpha amylase activity in the high mash temp sample resulted in a FG of 1.014, a strikingly vast .009 SG difference. Prior to this xBmt, if asked how 2 beers of such varying FG might differ, I would have thoughtlessly, perhaps even arrogantly, claimed the differences in body and especially perceived sweetness would be easily identifiable. It’s a line of reasoning that, on the surface, seems obvious: higher FG = more sugar = sweeter. The data just doesn’t seem to back this up, which I find incredibly surprising.
On a more anecdotal level, many of us have enjoyed commercial beers mashed on the warmer end of the spectrum that aren’t known for being cloying such as Lagunitas IPA, which is purportedly mashed at 160°F and finishes around 1.018 FG. Perhaps there’s more to this story, shit we thought we understood that we could learn more about. While I’ve no good way of measuring it, I’m curious if the types of dextrins remaining in a beer that was mashed on the warmer end of the spectrum are imperceptible to the human palate. This obviously doesn’t explain the fact mouthfeel between the beers wasn’t a dead giveaway, something that continues to stump me.
Finally, I’ve yet to comment on another aspect of this xBmt that didn’t seem to have an impact on the ability for tasters to tell a difference between these beers– level of alcohol. The low mash temp beer had a calculated 4.4% ABV while the high mash temp beer clocked in at a much lower 3.4% ABV. Whoa. Perhaps higher mash temps are just the ticket for those who prefer making flavorful examples of big beers with lower amounts of alcohol– Session Imperial Stout, anyone?
If you have any thoughts on this xBmt, please share them in the comments section below!
Support Brülosophy In Style!
All designs are available in various colors and sizes on Amazon!
Follow Brülosophy on:
FACEBOOK | TWITTER | INSTAGRAM
If you enjoy this stuff and feel compelled to support Brulosophy.com, please check out the Support Us page for details on how you can very easily do so. Thanks!
123 thoughts on “exBEERiment | Mash Temperature: Low vs. High In A Blonde Ale”
Thanks again to share this experiment. I think that to enphasis the difference between the two mash temperature, the time should have been different too since that both enzyme do not work at the same speed. 1h30@60°C vs 45min@71°C. If one day you perfrom such experiment , you will have results from extra dry mash to extra sweet (1h30@60°C, 1h@67°C, 1h@70°C and 45min@71°C). Good job btw. Thanks again.
I knew a fella who occasionally converted some of his beer to no/very low alcohol.this would be the ultimate session beer. Process was to remove some beer from the keg,yes finished and carbonated, put in a pan and heat to 185f,alcohol boils at 180, watch all the little alcohol bubbles come out, try not to inhale too much, cool and rekeg and re-force carbonate.voila, near beer imperial stout!
I have been considering trying this same thing to make lower alcohol beers, although I still don’t have equipment to force carbonate.
But I do try to “mash high” to get great malt character from relatively less malt, with correspondingly lower ABV. The experiment results seem to support this approach, although without the reduction in grain bill. Guess I have some more iterations in my quest…
It doesn’t really work that way. The alcohol-water solution boils at some temperature lower than 100 degC. The steam contains both water and alcohol with a higher percentage of alcohol than the fluid. But it’s not 100% alcohol. Guess why have to distill whisky several times to get the raw spirits to the barrels? During boiling the alcohol level in the kettle will lower, but not at an amazing speed. And you cannot boil it for a few minutes and call it alcohol free. There is still alcohol in a wine sauce even if you boil it. And you certainly cannot boil beer to an alcohol free level.
awesome! I’ve been waiting for you to do this one for awhile. I have been wondering if there is any difference in dextrin-rich beer compared with those with fewer dextrins. I wonder why the lower mash temp beer blew off yet the the high mash beer was frothier in the kettle? strange. I asked a lot of people about the dextrin question in the past. One person I asked is Charlie Bamforth. He indicated to me that it is possible that dextrin-rich beers are perceived as “fuller” due to the amylases naturally present in our mouths. He thought it is possible that the amylases break down the dextrins some as we drink the beer and give a slighlty sweeter, fuller character to the beer. It seemed like a stretch to me but possible.
I also heard Jamil Z. talk about the fact that he tasted dextrin powder and found it to be decidedly NOT sweet. He also did a Can You Brew It where it was really windy, and he was all over the map with his mash temp and boiling, etc. He said the mash temp was really high for the beer he was doing, yet when he fermented it out it didn’t taste sweet despite the higher FG.
I think the reason we associate higher FG with sweetness is when we have poor yeast health or under-pitching, there are more than dextrins left at the end, there is also maltose left, which will make the beer taste sweet, as we all know. If you ferment down all of the maltose and leave only dextrins, it will have a high FG but not be sweet.
This is a really great ExBeeriment! I was also a little surprised that the high mash temp beer was at least as high in OG as the low temp one? Would you expect a higher OG from a temperature in the mid-range (ie a temp closer to what you usually use, such as 153ish) based on your experience with other batches you’ve done? Maybe both the really high and really low temps are equally less efficient than a mid-range temp? I have been wondering for awhile if for single-infusion batches I should just always target the same temperature, whichever one give the best conversion efficiency. However, good idea on mashing really high to get the same flavor but much lower alcohol! The calories won’t be lower though I think! And, you might get more beer farts from the high FG batch!
It is a common misconception that FG is directly related to residual sweetness. That depends on the types of sugars that make up the FG, or the difference in FG. The main effect of mash temperature is the precentage of dextrins. Dextrins has very little sweetness, perhaps just 5 % of the sweetness of table sugar (per weight unit). So it is actually not that surprising that there was no clear difference in percieved sweetness.
On the other hand, if you split a wort and ferment with different yeast strains so that there is a similar difference in FG due to unfermented maltotriose (and perhaps also maltose), then there would probably be a huge difference. I have done that myself, fermenting a triple in several buckets with various yeasts. There was a clear difference between those who stopped around 1020 (due to yeast not in perfect condition) and those who stopped at 1010.
What?!? Could not get statistical significance with something as seemingly obvious as 0.009 difference in finishing gravity? This one will definitely raise some eyebrows, and I expect a lot of discussion.
My first thought is that maybe there is a “perception limit” on finishing gravity, or at least a greatly reduced sensitivity to differences outside of a range that is higher than represented by these two beers. Or maybe something about the hop or yeast character that masks these differences.
Looking forward to others’ thoughts on this…
It’s a whopping 1% difference in density (a little less, actually) and not a high OG. This doesn’t sound crazy to me.
actually, it’s 65% AA vs 87.5% AA, which is a huge difference, YOOGE!
Another FYI tidbit. Back in the mid 90s a homebrewer won many competitions on the east cost with his dopplebocks and eisbocks. He admitted a few years later he brought the alcohol level up on these beers by adding vodka to the keg…..
Mind blown.
I was SURE this was going to be the one that produced significant results, confirming that mash temp does indeed play a huge role in the final product. Alas, it doesn’t seem like that is the case.
It does make me feel better about a beer that is in my fermenter right now though. Made an oatmeal stout a week ago, and blew my mash temp by not adjusting my tun and grain temp….was going for 154, and only hit 151.5. I’ve been kicking myself for a week! Maybe it won’t make such a difference after all.
Great work, as always!
I wonder if using a recipe with less subtle malts would produce more distinguishable characteristics. I’d like to try that, but I need to research your statistical methodology. I know very little about that.
Email with any questions, happy to assist! Science is strengthened by repeatable results.
While it gets proven time to time that almost nothing matters in homebrewing i could still make a batch with at least 3 different off-tastes more than 2 years after i started homebrewing. (sourness, plasticy smell and carboard)
Here are the suspected sources:
sourness: i added huge amounts of phosporic acid to a highly alkaline water (hco3 @ 350ppm) (12ml phosporic acid into the total water amount (mash+sparge) which was 20l) to get the mash ph @ 5.2, the mash and sparge water tasted like lemons
cardboard: the bittering hops were 2 years old stored in a regular fridge and without vacuum or co2, they smeleld like caramel/raisins upon opening, that classic old IPA smell
plastic smell: i had a huge (10F, 70F to 60F to 70F in the course of 8 hours) tempswing in the second day of fermentation and i used month old cloudy Star-San to sanitize stuff although the Star-San’s ph was still around 3.
Just to clarifiy this was my latest batch, some of these suspected errors could be good for an xBmt.
I did this experiment for my book and again recently for a home-brew seminar and got similar results. I think the benefit of mashing high is that you can pack more malt into the recipe without changing the ABV%. You can make a 5% ABV Scottish 80′ with 7# of malt and mashing low or 10# of malt when mashing high. More malt=more perceived body and more malt flavor…I always like when you prove me right!
Awesome, couldn’t agree more!
This is really great information. I’d love to beef up a beer’s flavor but keep the alcohol lower.
I had used 70deg C for the same recipe above, for the same reason. Keep what little malt is in the beer, but get alcohol lower.
Mike McDole has talked about making session beers with high mash and lots of crystal to get as much malt in the beer as possible, but keeping it low alcohol.
Damn, I spent so much time and money to build a RIMS system to keep the mash temps right on set point.
My new book on how to brew beer:
Poor some very hot water,147*-161*, in bucket, add grain.
https://brulosophy.com/2015/10/12/the-mash-high-vs-low-temperature-exbeeriment-results/
Wait between 30 minutes to 60 minutes.
https://brulosophy.com/2014/09/01/does-mash-length-matter-exbeeriment-results/
Drain and boil for 20 to 90 minutes. Cool wort.
https://brulosophy.com/2015/09/14/boil-length-pt-2-pilsner-malt-exbeeriment-results/
https://brulosophy.com/2015/03/11/the-impact-of-boil-length-ale-exbeeriment-results/
Poor ALL the kettle contents into fermenter.
https://brulosophy.com/2015/03/22/the-impact-of-kettle-trub-part-2-exbeeriment-results/
https://brulosophy.com/2014/06/02/the-great-trub-exbeeriment-results-are-in/
Add some yeast, don’t worry how much.
https://brulosophy.com/2015/07/28/yeast-pitch-rate-pt-3-ale-vs-lager-in-a-kolsch-exbeeriment-results/
https://brulosophy.com/2015/06/29/yeast-pitch-rate-pt-2-viable-cell-count-vs-vitality-exbeeriment-results/
https://brulosophy.com/2015/04/20/yeast-pitch-rate-single-vial-vs-yeast-starter-exbeeriment-results/
Ferment at 50* to 76* for some time.
https://brulosophy.com/2015/05/11/fermentation-temperature-pt-2-english-ale-exbeeriment-results/
https://brulosophy.com/2015/06/22/fermentation-temperature-pt-3-lager-yeast-exbeeriment-results/
https://brulosophy.com/2015/01/19/fermentation-temperature-pt-1-exbeeriment-results/
Carbonate and drink.
For a first printing signed by me please send $14.99 to me with your address.
Brad
Need a co-author?
Well, you might be jumping the gun with that carbonation thing…
Marshall, I thought about my book and remembered something. I think you and three or four friends brewed the exact same beer. Different people, different equipment, different methods but the exact same beer. If I remember every beer was very different from each other. Any ideas as to why?
Could be any number of things– water quality differences, slight process difference, a sneeze at the wrong time… I really don’t know.
i see the same thing in my homebrew club. we brew up 12-16 gallons of beer at a meeting and then split it 3-7 ways. the resulting beers are always very different. hard to explain. i think most of the variability is due to people oxidizing the beer. also, i’m pretty confident that bottling is inferior to kegging from the experiences i’ve had, at least the way the avg guy bottles.
With all the Marshall exbeeraments with most showing it is hard to screw up or even change the taste of a beer with brew methods. It does seem odd that there is SO much difference when different people brew the exact same beer. Aaron, maybe if everyone started with the same RO water, same additives. Would be a good exbeerament.
Sometimes I think that the brewer is a bigger variable than anything else. there are a million steps that one brewer can change compared with another that all probably add up to differences in flavor. even the same brewer can’t brew the same beer a lot of times!
The Brulosophy experiments are all very tightly controlled, and have been careful to hold all of the known important variables reasonably constant, with the exception of the variable being tested. I don’t find it at all surprising that different brewers will have widely different results, since each brewer will have many slight variations in their processes.
The takeaway for me from all of this is, if you generally follow good practices, and keep most of the variables in control, screwing up in any one area will probably not have a significant impact on the final product. The more things you “screw up”, the more likely you will see a difference. At some point, presumably, the difference will result in something bad.
Gold Brad
Best comment ever. Yes – sometimes on rainy days while I am waiting for paint to dry – I will read old Brulosophy Exbeeriment comments ….
Same here.
Hi Marshall. Another great experiment. I brew this fantastic Hopped up Red Rye Ale from Two Brothers Brewery called “Cane & Ebel” They use palm sugar as an adjunct. and what seems to be a little higher amount of crystal malts then I use in other recipes. Here is a link to that recipe: http://goo.gl/YkVzFx In the 5 gal batch you use 1 lb of the palm sugar. Let the mash temp get below 151 the finished product turns out too dry. Let it get up above 159 and it turns out almost burnt tasting. It seems that when you start adding other fermentables there is more of a Goldilocks Zone in the mash temp so the fermentable sugars that are created from the mash complement the fermentable sugars from the adjunct. Or is it the higher amounts crystal malts like this recipe uses ? Or a combination of these ? This will be in the back of my mind when adding 1.375 lbs of Honey in my 1st attempt of doing a Great Lakes Christmas Ale Clone here soon. If my experience holds true, mash too high and I miss the imperial nature of that beer. Mash too low and I get too dry. Although the Christmas ale doesn’t have near as much crystal malts in it. “So much time and so little to do. Wait a minute. Strike that. Reverse it” Willy Wonka
Prost.
Fascinating data, thanks for sharing!
Wow! I’ve been spending so much time and energy nailing my cooler mash temps too. Mind blown.
If the high FG beer does have more dextrin I initially thought, “crap…I’ll just be adding more to my beer belly!” A quick internet search led me to discuss the benefits of dextrin as a dietary fiber, with benefits such as: weight loss, lowering bad cholesterol, maintaining good cholesterol, and keeping things regular. This all sounds like a plus to me. Assuming that this is all true, and the beer tastes just as good, maybe high mash temps can be seen as the “healthy” choice! Unfortunately the ABV is lower, so I might give it a corn sugar boost near the end of boil 😉
I like your style.
think again – athletes use it in sports drinks for energy:
https://en.wikipedia.org/wiki/Maltodextrin
I think people are getting messed up with what is meant by dextrins.
Maltodextrins are chains of glucose, which can vary in chain length substantially, sweetness and are digestible by the human digestive system, but not by Sac yeast. Brett is meant to be able to metabolise it.
Some dextrins are dietary fibres, something like polydextrose which is also (primarily) a chain of glucose, but with a different bond that we cannot metabolise – so is treataed like a fibre as our gut microbes can metabolise this.
the dextrins produced in beer are shorter chain equivalents to the maltodextrins, and we can digest them. depending on their structure they can be sweet or not.
To put the ‘bad’ back into balance with the good?
Some of the discussion will get back to a similar question to the DMS.
It might have differences, but they are likely below the threshold of the measurement methodology to assess. At that point, does it matter for general home brewers? It might matter for commercial and/or competition. At the commercial end, with trained sensory panel they might be able to detect a difference. But serving beer to your mates on the weekend, probably doesn’t matter.
I have made this recipe a few times to experiment with different hops, some finished out at .004 and the last one mashed at 70deg finished out at .014
I could swear there was a little difference when side by side, but it might also be belief sticking in, and definitely lower alcohol.
I wonder what the effect will be if you mashed a DIPA at the same temperatures…
you’d have a Lagunitas IIPA apparently.
I guess you mean Maximus?
As the claim is most of their beers are mashed hot.
I cant get any here, but I never hear anything about their IIPA in the general internet world.
I hear plenty about Stone, Pliny (both), Heady, etc but never Lagunitas Maximus.
Now the important thing is would people pick up the difference if you brewed a pliny the younger clone mashed at 70deg C versus 60deg C. Or what effect would mashing a maximus clone at the same temperatures. Pliny the elder is mashed fairly warm as well. Perhaps the difference is no crystal malt(tm).
Lagunitas IIPAs or IIPA-like beers that all have similar character:
Maximus
Little Sumpin’ Sumpin’
Little Sumpin’ Extra
Brown Shuga’
Lagunitas Sucks
I find that they all have a certain full flavor. They all have pretty low bitterness for the style and a ton of hop flavor/aroma. I’m getting really excited about beefing up all of the flavor components that come from just having more malt in the batch without having as much alcohol. A whole world has opened up!
They also ferment with the Fullers strain… which I always mash super low with for hoppy beers. Nuts.
I think the water chemistry *may* have a big effect on Lagunitas vs Russian River. I am assuming Russian River is hitting the sulfate pretty hard and Lagunitas is not, but that is just based upon my tasting of their beers not on any other knowledge. The Lagunitas is always much smoother, less crisp and less mineral-like in my opinion.
JC, I don’t think you’re right about the Russian River mash temp on P the E. Look at this:
https://www.homebrewersassociation.org/attachments/0000/6351/doubleIPA.pdf
I can’t believe they get such great hoppiness using the Fuller’s strain. I think the Fuller’s strain really is great for malty English styles, but I’ll have to try it in an APA/IPA I guess. It doesn’t leave the beer sweet?
Not in my experience. Firestone Walker uses it too, Union Jack is my favorite commercial IPA.
Yeah, I really love Union Jack too. It doesn’t have a ton of hoppiness when I get it, but I buy it from in NY state in cans, and I have no idea how old it is or what is has been exposed to. It is a really nice beer though. I wish I could try it on tap closer in time to the fermenter. I mean, it must have a ton of hoppiness when fresh, right? I don’t think I’ve ever used Fuller’s yeast in a hoppy beer.
Super hoppy when I get it here in Fresno (2 hours from the brewery).
Good work on this one and some interesting results! I’d been looking forward to seeing how much of an effect mash temp had on the finished product. Definitely going to mash my next session IPA higher.
Marshall, what is the highest mash temp you have used and what conversion efficiency did you get when using it? What conversion efficiency did you get with this batch?
Not around computer now, but I always hit my expected targets, so eff was the same. This was the highest I’ve ever mashed.
Huh, I’d like to see how high you can go and keep the efficiency up and also what that does to the FG.
Hey Marshall, here’s another question. Do you think I could transfer over the 65% AA you got on the high mash temp beer to a beer with an OG of 1.060, mashed at 161F? That would be amazing! It would give me a lowish gravity IPA with only 5.0% ABV!
Can’t be certain, but I’d not fear trying.
I’m definitely going to try something like this. I think I’ll try 1.060 with a mash temp of 162F and hope for 65% AA. Is US-05 more attenuative than 1056? I’d prefer to use US-05. I could also use S-04 perhaps?
I just did a quick calculation of the calorie difference between two beers brewed with OG 1.065 and FG1 = 1.026 vs FG2 = 1.010. The FG1 case is 225 calories and the FG2 case is 217 according to http://www.brewunited.com/beer_calorie_calculator.php. If both beers could taste similar, that would be an amazing 5.0% ABV vs 7.25% ABV! I’m tempted to try it! Seems kind of insane though…
Damnit….are there any variables that actually matter!?!?
No, seriously….have there been any XBMT’s that have shown a strong/very strong statistical significance? I have been reading your blog for a while and am yet to see one that i can recall, but i am pretty sure i haven’t read every one.
One quick suggestion…would you ever consider adding a chart somewhere on the site that lists each XBMT, a couple words regarding the nature as well as the stats from the results?
Thanks for all the hard work.
Every water chem xBmt has been significant.
Working in organizing prior xBmts better…
Also, the exBeeriment where the two beers were identical was significant:
https://brulosophy.com/2014/11/03/the-impact-of-expectation-on-perception-exbeeriment-results/
Where is Rod Serling, anyone?
Actually I think most have been significant if only at showing that some things we “know” just aren’t so. The one thing that pretty much all of the XBmts prove is the validity of “Relax, Don’t Worry, Have a Homebrew”.
I had a digital thermometer drift way out of calibration on me over a bunch of brews. I kept adjusting the grain bill thinking it was the grain or? Then I checked the temp and found I was about 20 degrees off. The next brew my OG was WAY over expected. Now I know those were a much higher ABV, but I’m wondering if the dexedrines that are adding to the FG in your experiment are ones that just would not be there otherwise and that the amount of fermentable sugars is actually about the same? It would be interesting to get an actual analysis of the finished beer since we are approximating ABV based on some assumptions about what is in the wort, and those assumptions may not be correct for different mash temps.
The fermentable sugars in the low mash beer result in higher alcohol. the fewer fermentable sugars (ie dextrins) in the high mash beer result in less attenuation and less alcohol. If mashed lower, the dextrins would be chewed up into maltose and converted to ethanol
Your various experiments actually inspired me to do a high vs. low mash experiment a few months ago and I got a much smaller difference of 0.003 between the beers but a small but noticeable difference in flavour.
My testing was less scientific but the local homebrew group seemed to consistently picked up a difference and preferred the high mash. I find the low mash a bit thinner and perhaps a little astringent – particularly when drinking a whole pint even doing sort-of blind tests on myself. The high mash is gone now while the low mash keg lingers on.
Much like your other experiments, my experiment has resulted in much less stress about hitting / maintaining mash temperatures. Keep up the awesome work and testing conventional assumptions.
Full recipe:
85% Gladfield Pilsner
7% Gladfield Dark Crystal
5% Gladfield Aurora (another crystal)
3% Gladfield Red Black (another crystal)
40g EKG @ 60mins
30g EKG @ 15mins
40g EKG @ 0 mins
WLP002 (English Ale)
Mashes were 63c vs 70c:
1.046 OG, 1.013 vs. 1.010 FG.
Very cool. Did you present the beers to the tasters in a blind triangle test? If not, I’d be compelled to assume the differences they noticed were at least partially influenced by expectation of a difference.
The testers were initially blind to the nature of the test but it was just a side-by-side comparison and they were expecting a difference. I was able to reliably pick up the low mash blind but that doesn’t mean much wearing a science hat.
So not scientific, will have to do the triangle test next time to have some more confidence in the results.
That brings up an interesting idea… Have you ever considered doing more triangle tastings with all of the same beer? I know you did it once, but it would be easy enough to build up a bigger sample population for this experiment, and your tasters wouldn’t know if there is a difference to expect or not.
Of course 🙂
Can you detail the starting temp and finishing temp of the mash for both batches?
I assume they didn’t stay the same temp during the mash hour.
Thanks!
You know, I didn’t (and rarely ever) measure the finishing temp, though as reference, the largest drop I’ve experienced was 4°F during the middle of winter when it was about 40°F outside; on this brew day, it was hovering around 90°F, my guess is I lost 1°F or less.
What I seem to be learning is that.
1) Hops matter
2) Malts matter
3) Yeast matters
4) Water chemistry matters
Nothing else matters so long as you don’t do something REALLY wrong.
I would be curious if a low hopped beer would make a difference, or yeast dependent or OG dependent on if you could tell a difference between a low temp mash vs a high temp mash. Or if the malts used would make a difference.
The furthest I have gone for direct comparison was an English Session ale using S04 compared to Notty (the Notty was the better beer and both were easily distinguished with all 8 people who tried my beer picking the odd one out). I have brewed a couple of recipes just changing 1 or 2 ingredients or processes, but I don’t think I’ve ever varied the mash temp between recipes (intentionally, unintentionally, sure, by 2-3 degrees, I am not much of a mash temp stickler. I generally shoot for 148F for “dry” 152F for “medium” and 158F for “full”, but maybe it is all BS).
I would think if you are going to have that much difference in FG, once you get in to larger beers, you’ll start getting more of a flavor contribution from alcohol, so a low mashed 1.080OG beer might beer perceptibly different than the same 1.080OG beer mashed high.
Maybe the next experibeer? I’d suggest not something super hoppy, but something that focuses more on the malt bill. Like maybe a Dopplebock, or a strong English Brown ale.
I don’t know, I still think a lot of stuff matters on some level… just not what sure what level that is yet.
Agreed, and I just published a post (http://fabrikorekstedt.blogspot.se/2015/10/olexperiment-och-signifikans.html) on my blog dealing with this question. It is a bit lengthy, and in swedish, so I will give a quick summary here.
Most exBeeriments seem to indicate a small effect in that more than one third of the participants correctly identified the unique beer. But since the number of participants in the test panel is quite modest in relation to the size of the (propable) effects, statistical significance is usually not reached and we cannot rule out that the outcome was due to pure chance. At least not for any particular exBeeriment. But I am pretty confident – without having done the calculations – that we can rule out that all or most of the results are due to pure chance. And I think that on an aggregated level we might see more profound effects, e.g. by combining low pitch rates and high fermentation temperatures. Just to choose one example.
I don’t disagree and tend to think stacking certain variables (e.g., low pitch rate and warm ferm temps) may produce a more noticeable difference in the beers. This is something we plan to do as soon, I promise.
I used Google translate to read the article, nice stuff! I understand where you’re coming from regarding the larger participant pool, though I think it deserves mentioning that in even the xBmts with 30+ tasters, significance was not achieved. Our participants consist of BJCP judges, experienced homebrewers, seasoned craft beer drinkers, etc. We’re not just asking our BMC drinking neighbors to give us data, hence I think it is at least fairly representative.
Thanks for the great comment, like I said, I loved the article you wrote. Skål!
Thanks and skål! I understand the practical problems of getting enough people to the test panel (I have done this myself twice), and I agree that it is more important to have people with some sensory skills. It is just important to bear in mind -which is my main point about test panel size – that given a certain non-zero effect, you will get statistical significance with enough number of test participants. The larger the effect, the fewer you need.
excellent point
Just one final comment, a bit in regard to the comment further below about picking some other level than 0.05 % for statistical significance. I think that would be slightly dubious, and it would be better to put less emphasis on the p-value, and instead highlight the actual percentage of successful participants more. In this particular case, 45 % of the participants identified the correct sample, compared to 33 % which would be the average result if everyone just guessed entirely. So there seems to probably be some effect, but the relatively high p-value suggests that it could very well be partly or even completely due to chance.
that’s 5%, not 0.05%. I like stating the percentages as you have though. That puts things into perspective better.
I meant 0.05 = 5 % of course.
Since you did these as no sparge / full volume mashes that means no mash out, correct?
You may already have this on your list, but I’d like to see an exBEERiment where the variable is the water to grist ratio. I started all grain brewing with the BIAB method and have only done full volume mashes.
That’s right… though I don’t do a typical mashout ever, even what batch sparging larger batches. I add 180°F sparge water that brings the temp of my grain bed up to about 168°F. Water:grist xBmts are definitely on the list!
I was just thinking the same factor as an interesting topic. I don’t do full volume mashes, but bizarrely have been too lazy (and too new into doing all grain) to organize myself to test water/grist ratio.
Just something to think about, but since this recipe in particular uses Cara-Pils/Dextrin malt, that may be why it was difficult to tell as much of a difference in mouthfeel.
Very interesting that this didn’t result in a difference with a 20% relative shift in alcohol content when the measurements back the science.
Only recently found the site and have thoroughly enjoyed reading all the xBMTs. Keep up the good work.
One other thing to consider with the statistics is that you are always looking for a 95% confidence that there is a statistical difference. You can always decide to lower it, although 95% is the commonly accepted target.
Any plans to reissue the ‘My Kind of Science’ T-shirt? I’d definitely be interested and only found this after the promotion had finished.
Probably start that campaign again in the next week or so, this time with hoodie option.
I did a similar experiment with a weihenstephan-like recipe six months ago, inspired by this site (so thanks!). Much lower extremes: 150F vs 156F.
This is just a theory — and I didn’t do a broad test — but anecdotally noticed a somewhat significant difference in sweetness and mouthfeel. I was surprised by the results above, however I’d offer that a recipe that focuses more on malt and yeast could potentially have different results.
When you make big changes to a beer that is hop-centered, using almost entirely 2-row pale and 1056, you’re (again, just my theory) really just comparing two varieties of something specifically designed to be background noise.
I could be wrong, but a thought.
Very cool! Did you happen to present the beers to blind tasters? I think I mentioned that I perceived a difference in the beers when initially tasting them side-by-side and fully aware of the independent variable, but in a triangle test, I was unable to distinguish them reliably. Our minds do weird things when we expect differences.
I did not, and would have liked to. Definitely with you that if there’s one person who’s going to notice a difference no matter what, it’s a guy (me!) who spent 12 or so hours moving things around hoping to justify it :).
I mean very seriously the theory part, mostly because I think so much of our beer tastes and evaluations in the US right now are entirely about hops, often regardless of style. I mostly blame 1056 ;).
Separately, though, I did get almost a full % point lower ABV in my batches as well. As someone who prefers lower ABV beers relative to my other beer-loving peers, that was perhaps the most interesting (and objective!) take away for me.
Marshall,
Great work. It’s really useful to have the large number of comparisons done on a single setup. Helps me decide where to focus my efforts.
One question. How long after kegging do you typically perform the taste test with the panel?
I start when the beers are carb’d and clear, usually 4-5 days.
@ Eksted The whole point of the statistical significance is how confident you can be in making a conclusion. If 45% of people claim to detect a difference it doesn’t really handle the potential chance that on this occasion 45% of people just guessed correctly. The p-value starts to give you a better descriptor for that.
However, I entirely back that sample size is probably the most crucial factor here. Who’d have thought you would struggle to give beer away!
I think we basically agree, my point is that you should both look at the estimated effect and its accuracy. Unfortunately, p-values are not that intuitive. A confidence interval would be better. But the most important thing is to realize that failure to reach stastistical significance is not the same thing as establishing a non-existing effect. As i point out in my blogpost, there is a problem in some parts of the social sciences and medicine with an obsessive focus on p-values while neglecting the actual effects. There is even a moniker – “sizeless science” – to this phenomenon.
seems pretty intuitive to me. it’s just the probability of getting the results you did given that your hypothesis (eg there is no difference between the two beers) is true when it is not. Are you comfortable assuming there is no difference between the two beers at 5% probability or 10% or higher? I like this idea of “sizeless science.” It always seemed so arbitrary to focus so much on 5% as some golden cutoff.
Well, many people have large difficulties grasping the idea of p-values, including professors in medicine for instance. And there are people who seem to think that this experiment implies that mash temperature has no discernable effect on beer flavor. Because statistical significance was not achieved. Instead, what this experiment implies is that there probably is a small effect, but we cannot be sure, and a large effect is very unlikely.
And “sizeless science” is not about the arbitrary 5 % level, it is about focusing to much on p-values and ignoring effect sizes. Google “the cult of statistical significance” for instance.
I wanted to add another personal data point here. I was so excited by this ExBeeriment that I immediately made a Mirror Pond Pale Ale clone at a high mash temp to see what happens:
I was shooting for a 165F mash-in but I only hit 161F. (I’ve been seeing this 4 degree drop lately, maybe something is happening b/c I condition my malt or b/c I was doing it as a no-sparge? I don’t know but it is still a data point and pretty high mash temp.) So, I started ramping it up to 165F with my HERMS. It was at 162F by 15 mins, 163F by 20 mins, 164F by 30 or a bit less, 165F by shortly after 30 mins, maybe 36 mins. I then ramped up to 170F for a mash out and rested about 10 minutes. My conversion efficiency was great at 95%.
My OG was 1.0555. I pitched in commercial ale levels of S-04, rehydrated at 65F. I then screwed up and didn’t realize how hot and intense S-04 can get! It got away from me and was at 71F at 24 hours. by 12 hours I had chilled it slowly to 67F and held it. I then slowly ramped up to 72F with some swirling to encourage the yeast. I took a sample after about 5.5 days, and it is very dry. The final gravity? Well, it’s been at 1.0268 for 2 days now and is finished!!! 52% apparent attenuation!! I was hoping it would get no higher than 60%, but this is incredible!! My blast off ferment and then slight chill down may have impacted this some so I need to re-do it with better temp control (ie put the controller on cooling mode as soon as i pitch yeast!)
If I would have used US-05 and a mash at say 152F, I would expect this beer to get down to 1.009 at least, over 80% AA and be over 6% ABV. Instead, I have a beer with just as much delicious malt in it with only 3.7% ABV! It’s still hard to believe. Let me reiterate that the beer does not taste sweet. If I had to guess, I’d maybe say it tasted like what I am used to a 1.012 FG beer tastes like? Honestly, at this point I’m beginning to think the FG is not important unless it is considered in the context of the mash temp and the darkness of the grains.
I am definitely going to experiment with high mashing to see how the beers continue to turn out. If I don’t see any other negatives of high mashing, I think I’ll start doing this a lot of the time to keep the flavor but drop the alcohol. Maybe i’ll be farting a lot from the dextrins though. That could be bad 🙁
Just an update: Now that my APA is cold and carbonated, it is incredibly dry. I’m starting to wonder if beers mashed at higher temps will taste MORE dry and bitter compared with those mashed at lower temps for a given gravity range? The yeast have a lot less work to do with a high mash temp since there are less fermentable sugars, so maybe they really crush the fermentable sugars down more than they would in a lower mash-temp beer of the same OG? This baby has no discernible sweetness whatsoever using a mash temp of 161-165. Body seems about what I’d expect for the OG of the beer.
Very interesting!
I did a 165F mash for a split batch of Pils/Saison a few months ago. WY3711 definitely has diastatic power! It still had 92%AA vs the Pils half that only got down to 60% AA. 7.0% vs 4.5% ABV for the same wort! I need to find a new saison strain that doesn’t attenuate so much!
Could you do a similar experiment with water to grist ratio being the variable? Maybe 1.0 qt/lb and 3.0 qt/lb. I’ve often heard people say that a stiffer mash maltier beers and better mouth feel, I’ve always been suspect of this and used 1.5. Would be interesting to see if there are any differences noted and could shorten brew days by adding all liquor at once versus sparging
It’s on the list! I often do full volume mashes, works fine for me.
Hi!
This is just amazing! I like it a lot.
I had to ask Weyermann malts about this, “ein Frau Braumeister” so I guess we can expect the answer to somewhat fact-based answer. I suspect she know what she is doing, after all she is a professional brewer at a maltster, right?
The respons was that it shall be possible to determine the difference, 63 C dryer and 68 C more body. They did how ever explicitly ask for the mash temperature stability.
Do you have any data on the mash temperature, like a log over time?
And thanks, keep up the nice work!
/ Mikael
I don’t have a log, but I measure post mash temp regularly, I only lose 1-3F over a 1 hour rest depending on the weather.
Best (& most shocking) thing I’ve ever read about brewing! FG means much less that OG!!! Top hats to you bro 🙂
Have been fine tuning a 5am Saint recipe shared by Brewdog on their web site, DIY Brewdog, to match my system to the recipe’s numbers. They mash it at 144F. I thought that very low so started searching the web for info and found this blog. Don’t think I will obsess much over mash temps anymore. Thanks for all the info shared.
I’ve done a ton of mashes at 165F on all manner of beer styles. I noticed I rarely have blowoffs during fermentation in these beers. I am thinking there must be some enzymes that are denatured at that temp that normally give rise to good head retention.
I think one of the main reasons as to why results are usually non-significant is the random nature of the people making up the tasting panel and their ability to discriminate various flavour compounds etc . The tasting panel is an uncontrolled variable in every experiment and seeing as it is usually the only experimental ‘measurement’ provided it is not surprising that most results do not achieve statistical significance.
I understand the practical limitations in keeping a standing and appropriately qualified and experienced tasting panel for experiments but I’m not convinced that you can infer anything representative from any of the tasting panels other than it being accurate for that given tasting panel.
Sorry for not being particularly helpful, but research labs spend a lot of money on the best ‘detection’ apparatus for whatever is of interest for a given experiment. Researchers will then (or should anyhow) use that piece of equipment exclusively for data collection, thus controlling for another variable. Experimental science is a hard gig, just getting an experimental setup to work is hard enough, let alone trying to generate meaningful results. I take my hat off to you all, but sometimes I think you are setting yourselves up for failure before you begin. At least you get beer out of though!
Apologies for the length of the rant……
Wow, really having a hard time believing samplers could not tell the difference in a beer with a FG of 1.005 and one with a FG of 1.014!
I had a hard time believing I couldn’t tell the difference myself!
Good podcast. I wondered if the beers had really fully attenuated when you took the hydrometer measurements?
“I took the first hydrometer measurement once signs of fermentation activity had waned, then confirmed it hadn’t changed the following day, 6 since brewing the beer.”
I would have taken hydrometer readings at point of consumption too – it’s possible the difference narrowed as the yeast kept slowly working after day 6.
John Palmer must hate Brulosophy
John Palmer is friends of Brülosophy.
Very interesting study. I would say that the final point you made is actually the most important. Higher Mash means lower alcohol content. So the output of fermentable sugar from the two extremes does seem to be 100% accurate. I am going to try this same test, thanks.
I calculate the amount of grain on my recipes based on the final ABV% I want. The difficult thing is always to guess the FG.
If your mash temp is high, you’ll expect a higher FG and therefore a lower ABV%, of course. If you add more grain for the same ABV% you’ll get more body and sweetness.
I think the problem of your experiment was the question was not the correct one. Give 4 liters of the high-temp-mash-beer to the participants one day and 4 liters of the other one to the same guys next week. Then ask them how did it feel.
I think the experiment would be more interesting if you use the ABV% as a constant. You change the amount of grain and do the two same mash temps. Unfortunately that’s more difficult to calculate… but hey! you have already done the first step of the experiment, now you know what attenuation to expect for this beer for those mash temps!
I’m sure 100% of the participants will be able to tell which beer is which and you’ll see the obvious difference in body and flavour.
For this style they’ll probably choose the maltier one, but if you make the experiment with a light (color) saison I’m sure most of the people is going to go for the dry one.
I think you should replicate the xBmt!
Sorry… I forgot to thank you for your xbeeriments, I find your articles very interesting and enlightening
This gave me an interesting idea. How about splitting the grain in half and mashing the first half at a lower temperature (say 143-149 F) and the second half at a higher temperature (at approx. 150-158 F)? I’ve done this before. I got a good ABV and good body and malt flavor. So, yes, you can have it all.
The other thing I did was step mash the first half batch and then use the wort from it to mash in the second half. Apparently, some brewers in Victorian England used this technique. I never raised the wort temperature from the first batch to the point of denaturing the alpha amylase. I think that helped immensely in making sure the second half-batch got completely converted.
Maybe some more ideas for more experiments?
Interesting idea, thanks!
This a long time ago, but how did this affect time to terminal gravity?
It didn’t.
I read this xBMT a while back but never commented. Such a fuss is always made by amateur brewers about maintaining specific mash temps. What then is the point, if this is the case? Why all the worry and effort?
I now heat my strike water to 160 and dont bother trying to control the temp with blankets and / or heat belt. I let the beer run through thye 150 temps and into the high 140s before I transfer to the fermenter. My recent American Styled Amber with Nottingham yeast fermented from 1.052 to 1.008 for a respectable 5.7 ABV. My expereince has been a highly fermentable wort.
The only real issue I can imagine with such a process would be to replicate the same beer twice, and maintain consistence. I wonder if in higher volumes, and with morerobust (darker) malts, does the mash temp have greatter effect.
IMHO, for my mainly amber styled beers and IPA, I get benefit of both the Alph and Beta break down by letting the mash cool down over and hour or more. Can someone tell me my assumption is flawed? Any scientists left in the USA? A little Trump humour.
The main reason i sometimes maintain care with mash temp is when i want a lower abv for a given OG. In that case i make sure to hit temp and do a mashout to prevent excessive attenuation
I think this was a problematic xbmt from the start because of the recipe, 0.5lb of crystal is going to dominate mouthfeel regardless of mash/yeast. Again hops are going to influence perception a lot. I’d suggest a true pilsener recipe.