Mash Ph expirement

[amcclai7]

I wish I had taken better notes and had pictures, etc. but we did a mash Ph experiment a while back and I though I would post the, very interesting, results here.

I own and operate a small commercial brewery. We take great care to monitor our mash Ph and I wanted to see if this really made a difference.

I had always heard and read that the optimal range of mash Ph was between 5.2-5.5 and that erring on the low end of the scale, especially with IPAs, was optimal. http://braukaiser.com/wiki/index.php?title=How_pH_affects_brewing is an article that promotes this theory.

I have a friend with a PicoBrew system that is completely automated. So he brewed two batches of one of our most popular IPAs with varying Ph. One was adjusted via EZ Water Calculator to 5.2 the other aimed for right around 5.5.

The gravitates both starting and finishing were very similar. Surprisingly enough, the one with the higher Ph had a slightly higher OG.

I then gave a triangle test (two samples of beer A, one sample of beer B. They are told, here is sample 1, 2, and 3. Two are indentical one is different. Pick the odd man out. Nothing was said at all about the nature of the experiment) to 15 people over the course of several days. In total 8 of the 15 were able to pick the odd man out. Not overwhelming evidence, but better than the 5 out of 15 that would be expected with random guessing.

Here's the funny thing... Almost everyone who in someway made their living off of alcohol (This included a master distiller, several bar managers and professional brewers) about half of the 15 total, got it right! The other half which were all craft beer drinkers and fans but had jobs outside the industry, almost all got it wrong. This makes sense to a degree but I did not think the results would be that stark.

Here's where it gets really interesting: When asked which they preferred, the results were pretty much evenly split. Those who preferred the lower Ph often described it as: Crisp, bright, and dry and the higher Ph as flabby and too sweet. Those who preferred the Higher Ph described it as fruity and smooth while calling the lower Ph sharp and too bitter.

Where does all this leave us? I mean if it doesn't make a difference in preference then why go to all the fuss? I'm not quite sure what the answer is but I do have one theory that might back prevailing wisdom.

Read through this experiment if you have the time http://brulosophy.com/2016/10/03/water-chemistry-pt-6-sulfate-to-chloride-ratio-exbeeriment-results/ (btw, this is a fantastic website!!)

He did a similar experiment with an IPA but the variable was chloride to sulfate ratio. He also came up with a strikingly similar result. A decent number of people were able to tell the difference between the beers but the preference was pretty much split down the middle. The descriptors his subjects used were also strikingly similar, "the beer with a higher sulfate to chloride ratio was typically perceived as crisp, dry, hoppy, and more bitter while tasters felt the lower ratio beer was smoother."

One thing that he, and I, forgot to include in the survey was, do you like IPAs? My thought is that the people who normally like IPAs are going to like the lower Ph, high sulfate beer whereas the people who are not as nuts about them are going to want a less aggressive version of the style. If that is the case (again, this is still speculation) then it still makes sense, at least on a commercial scale, to brew beers to the prevailing wisdom given that most people ordering a given style are going to be fans of that given style.

Here is yet another experiment was very similar results: http://brulosophy.com/2015/05/18/water-chemistry-pt-2-messing-with-minerals-exbeeriment-results/

Very similar test but with a Dry Irish Stout. Beer A used a normal stout water profile, whereas Beer B used an IPA profile. Again, a good percentage of people were able to tell the difference but the preference was nearly 50/50. Beer A exhibited normal stout characteristics and Beer B was much more sharp and hop forward. Again, I can only wonder if those who selected Beer B were normally IPA drinkers.

Food for thought. I would appreciate any comments or experiences anyone would like to share.

 

[Sadu]

So what I take from those results is that one shouldn't stress out too much if the target was 5.2 and the actual result 5.5 (which happens to me somewhat regularly).

 

[RM-MN]

What I see is more of a "try different pH for your mash and choose the one you like to drink and keep using that". My tastes differ from yours. It I'm the primary drinker, I brew to my tastes. If you don't like my beers, brew your own or shut up and drink mine.

 

[amcclai7]

So what I take from those results is that one shouldn't stress out too much if the target was 5.2 and the actual result 5.5 (which happens to me somewhat regularly).

Agreed. Its not going to mess up you beer at all but it is worth shooting for a desired Ph, IMO. Recently, I've started shooting for a 5.4-5.5 Ph on dark beers.

 

[amcclai7]

What I see is more of a "try different pH for your mash and choose the one you like to drink and keep using that". My tastes differ from yours. It I'm the primary drinker, I brew to my tastes. If you don't like my beers, brew your own or shut up and drink mine.

I agree. Even brewing professionally I brew to my own tastes 95% of the time. People respect that.

 

[jmitchell3]

Yes! Dig me some brulosophy. Love your method too using the picobrew. see PS for random off-topic thoughts...

Using this calc (https://jangevaare.shinyapps.io/triangle/) with 8/15 your p-value is 0.088 which is slightly higher than the normally used, albeit somewhat arbitrary, benchmark of p <= 0.05. So based on that the brulosophy guys would call it "not statistically significant"...

 

[amcclai7]

Yes! Dig me some brulosophy. Love your method too using the picobrew. see PS for random off-topic thoughts...

Using this calc (https://jangevaare.shinyapps.io/triangle/) with 8/15 your p-value is 0.088 which is slightly higher than the normally used, albeit somewhat arbitrary, benchmark of p <= 0.05. So based on that the brulosophy guys would call it "not statistically significant"...

Like I said, I think there is some evidence that there is a perceptible difference, especially given the expertise of the people who got it right vs. those who didn't, but it wasn't overwhelming at all. Even side by side knowing which was which the differences where subtle although, at least to me, perceptible.

As others have said, I don't think it makes a huge difference but it is worth paying attention to.

 

[Morrey]

amcclai7, do you check/adjust your mash ph during the mash (if so, when?) or do you simply adjust in the beginning and let 'er rip?

I have always been one to adjust during the mash if needed, but I have been told that it is almost like shutting the barn doors after the cows have already gotten out. Lately I have been adjusting pre-mash and monitoring for data to be used in my next brew.

Your thoughts on this?

 

[rhys333]

I'm curious to know how close pH estimates are to the measured pH.

 

[RM-MN]

amcclai7, do you check/adjust your mash ph during the mash (if so, when?) or do you simply adjust in the beginning and let 'er rip?

I have always been one to adjust during the mash if needed, but I have been told that it is almost like shutting the barn doors after the cows have already gotten out. Lately I have been adjusting pre-mash and monitoring for data to be used in my next brew.

Your thoughts on this?

I had read that conversion happens quickly but nobody seemed to say what "quickly" meant so I decided to experiment with that. I BIAB and I mill my grains very small so they can gelatinize quickly and expose the starches to the enzymes so your results may be far different if you use a coarse milling of the grain.

For my first try I used a white plate with several drops of iodine that would be my indicators of the presence of starch. I doughed in and took a sample immediately as a control which turned blue when a drop was placed on the iodine so that showed starch was present. Then I waited for 5 mnutes as I wanted to do the test at 5 minute intervals to see how quickly the "quickly" was. That 5 minute test showed no starch left. Hmm... not the results I expected.

The next time I brewed I decided to take the samples at 1 minute intervals since 5 minutes was too long. Again I took a test at dough in with the same results, and the same results at 1 minute. OK, now I know that quickly means "more than one minute". At 2 minutes I tried again and.....conversion was done according to the iodine. Having a science background, I repeated this on the next time I mashed with the same results, conversion in less than 2 minutes.

Now again, it depends on your milling of the grain as the starches must gelatinize to expose them to the enxymes and larger particles take longer but I suspect that most of your conversion is done in less than 10 minutes. That means that if you adjusted the pH at 8 minutes, most of the conversion was done at the incorrect pH.

 

[Morrey]

I had read that conversion happens quickly but nobody seemed to say what "quickly" meant so I decided to experiment with that. I BIAB and I mill my grains very small so they can gelatinize quickly and expose the starches to the enzymes so your results may be far different if you use a coarse milling of the grain.

For my first try I used a white plate with several drops of iodine that would be my indicators of the presence of starch. I doughed in and took a sample immediately as a control which turned blue when a drop was placed on the iodine so that showed starch was present. Then I waited for 5 mnutes as I wanted to do the test at 5 minute intervals to see how quickly the "quickly" was. That 5 minute test showed no starch left. Hmm... not the results I expected.

The next time I brewed I decided to take the samples at 1 minute intervals since 5 minutes was too long. Again I took a test at dough in with the same results, and the same results at 1 minute. OK, now I know that quickly means "more than one minute". At 2 minutes I tried again and.....conversion was done according to the iodine. Having a science background, I repeated this on the next time I mashed with the same results, conversion in less than 2 minutes.

Now again, it depends on your milling of the grain as the starches must gelatinize to expose them to the enxymes and larger particles take longer but I suspect that most of your conversion is done in less than 10 minutes. That means that if you adjusted the pH at 8 minutes, most of the conversion was done at the incorrect pH.

WOW....this is EXACTLY what I have always asked but nobody seemed to have a conclusive answer. Thank you so much!!!

I BIAB too and mill my grains nicely fine. I experimented with the grind and am hitting 81% efficiency every time like clockwork. I'm simply saying my grind is pretty fine so it must be like yours.

I am concluding from your tests that accurate ph adjustments are darn critical from the dough in since there is almost no time to react to adjustments later in the mash. Like I said before...shutting the barn doors after the cows are out.

I use Bru'n water religiously and adjust as directed. I'll always test ph at the 20 minute mark and I'm always really close to what the program predicts per my grain bill. I adjusted for 5.45 last brew and my Hach read 5.41 at the 20 minute mark. I figured I must be close since my efficiency numbers are over 80%.

This is really really good information my friend, and I thank you so much for sharing!!!

Note: May I ask one more thing, please? Since we are confident we have conversion this quickly, what are the benefits (if any) of continuing the mash cycle for 60 minutes as an average?

 

[Morrey]

I'm curious to know how close pH estimates are to the measured pH.

I responded to this in my previous post. Using Bru'n water software along with a Ward Lab water report, I adjusted my ph (on my last brew for example) to target 5.45 ph. At the 20 minute point during my mash I read the ph with my Hach to be 5.41. This particular brew logged in at 81.05% BH efficiency. I was pleased with the reading although I don't know how closely this compares to the readings of others.

 

[mongoose33]

Yes! Dig me some brulosophy. Love your method too using the picobrew. see PS for random off-topic thoughts...

Using this calc (https://jangevaare.shinyapps.io/triangle/) with 8/15 your p-value is 0.088 which is slightly higher than the normally used, albeit somewhat arbitrary, benchmark of p <= 0.05. So based on that the brulosophy guys would call it "not statistically significant"...

My favorite Brulosophy exbeeriment is where the results were statistically significant but the tasters were split exactly 50-50 in terms of what they preferred. It's a great example of the limitations of significance as an indicator of......something.

When I read those exbeeriments, the first question that comes into my mind is "what's the actionable intellgence here?" And when the preferences are split 50-50, the actionable intelligence is not clear. It might be that the variable being tested doesn't matter, it might be that the two different beers simply appeal to two different palates.

One thing missing from these efforts is an attempt to quantify the preference. I might prefer A over B, but if I don't really like either, is A better or just less objectionable? This is a difficult thing to assess in these studies, so I'm not being critical, just that there still is a variable we're missing.

[I've been a judge at a few of my LHBC throwdowns; a couple of times I found myself judging beers whose style I do not care for. But I still could pick out the best example of the style even though I'd never put it in my own glass.]

If I had a brewery, I'd brew things to my taste, but I'd also use results like these to brew what others tend to like as well. I can imagine a taproom with, say, 20 taps, and above each tap was an indication of which type of palate is more likely to appreciate it.

i tend to like maltier brews, I'm not so much of an IPA fan. Buddy of mine likes IPAs, in fact we just brewed one Saturday for him.

Somehow--and I can't figure out why--our target PH was 5.4 but we came in at 5.08. I suspect this will accentuate the hoppiness and bitterness, which he will like (me not so much), but I can't figure out why I was off so much. I've never missed by so much, not even close. All I can guess is the malt had something weird about it, but my efficiency was normal, the starch converted, and I ended up w/ gravities right where I'd have expected them.

I'll post about that separately in the brew science forum when I get a chance, but while the beer we brewed may not have been what we thought it would be, it will still be beer.

 

[jmitchell3]

My favorite Brulosophy exbeeriment is where the results were statistically significant but the tasters were split exactly 50-50 in terms of what they preferred. It's a great example of the limitations of significance as an indicator of......something.

When I read those exbeeriments, the first question that comes into my mind is "what's the actionable intellgence here?" And when the preferences are split 50-50, the actionable intelligence is not clear. It might be that the variable being tested doesn't matter, it might be that the two different beers simply appeal to two different palates.

One thing missing from these efforts is an attempt to quantify the preference. I might prefer A over B, but if I don't really like either, is A better or just less objectionable? This is a difficult thing to assess in these studies, so I'm not being critical, just that there still is a variable we're missing.

[I've been a judge at a few of my LHBC throwdowns; a couple of times I found myself judging beers whose style I do not care for. But I still could pick out the best example of the style even though I'd never put it in my own glass.]

If I had a brewery, I'd brew things to my taste, but I'd also use results like these to brew what others tend to like as well. I can imagine a taproom with, say, 20 taps, and above each tap was an indication of which type of palate is more likely to appreciate it.

i tend to like maltier brews, I'm not so much of an IPA fan. Buddy of mine likes IPAs, in fact we just brewed one Saturday for him.

Somehow--and I can't figure out why--our target PH was 5.4 but we came in at 5.08. I suspect this will accentuate the hoppiness and bitterness, which he will like (me not so much), but I can't figure out why I was off so much. I've never missed by so much, not even close. All I can guess is the malt had something weird about it, but my efficiency was normal, the starch converted, and I ended up w/ gravities right where I'd have expected them.

I'll post about that separately in the brew science forum when I get a chance, but while the beer we brewed may not have been what we thought it would be, it will still be beer.

Sure I get that and I see where you are going in your thinking.

I guess my point is that if results are not significant, there's no proof that anyone can reliably tell a difference with that particular variable. If this is so, it doesn't matter what the preferences are for the group that can tell a difference, because the fact they chose the right odd-beer-out is not distinguishable from pure chance.

Are BJCP judges as a group more biologically able to distinguish taste differences than non-bjcp judges? Or is it just that they know what the difference actually is and can identify and comment on said difference (i.e. sensory training vs. the layman).

Further, I think I'm suggesting that trained judges as a group are probably not anymore physically able to taste differences in anything from snicker bars to cognac than non-trained judges. That is, I don't know that there is anything about trained industry people that are biologically superior to their non-trained counterparts that would allow them to be superior tasters. [I'm sure someone can make an anecdotal argument that those who choose the food or beverage industry as a profession or even a hobby may perhaps be biologically inclined (as a group) to superior physical ability in tasting and evaluating]. They surely, however, would be ones I'd look to on a preference descriptor to see how they evaluate differences they do taste.

I'm reticent to take any actionable intelligence from any experiment where there is no significant, demonstrable, statistical observation that a difference actually does exist that humans can actually detect. IMHO, taking anything away from the second question (preferences) when the former question (detectable difference) has not been proven valid may be putting the cart before the horse.

However, I would like to suggest that, like any other scientific experimentation, these Exbeeriments should be replicated over and over again with similar results before we indeed have any reason to change our behavior. Single experiments testing a single variable are hardly demonstrative of substantive information. again, IMHO.

And, of course, this all represents my personal human opinion, which, like many things, is subject to change without notice and will not in any way shape or form affect all of our pursuit of making better beer by using any and all means / information at our disposal!

 

[amcclai7]

amcclai7, do you check/adjust your mash ph during the mash (if so, when?) or do you simply adjust in the beginning and let 'er rip?

I have always been one to adjust during the mash if needed, but I have been told that it is almost like shutting the barn doors after the cows have already gotten out. Lately I have been adjusting pre-mash and monitoring for data to be used in my next brew.

Your thoughts on this?

This is pretty much what we do. We take a water Ph, which varies quite severely throughout the year, (7.8-6.5 depending on the season) put our recipe into the water calculator, adjust with salts and/or lactic acid via the water calculator, and mash in.

We will take a Ph after we mash in, but as RM-MN said, conversion happens very quickly so we don't worry about adjusting after mash in unless the Ph is way off. We just use it to inform our next calculation. For example, I know that when the water Ph is around 7 the water calculator is extremely accurate. When the water Ph is 7.8 it might be off by as much as .2 high and vice versa for low water Ph.

Fyi, the picobrew in the experiment was done at home by my friend so it was not Ph tested during the mash. This would have been great but given the fact that he was using RO water I feel fairly confident that the water calc was at least close. Also, given the fact that one got lactic acid and one didn't, I at least know that beer A was at a lower mash Ph than beer B even if it was, for example, 5.1 vs. 5.4 instead of 5.2 vs. 5.5

 

[amcclai7]

I'm reticent to take any actionable intelligence from any experiment where there is no significant, demonstrable, statistical observation that a difference actually does exist that humans can actually detect. IMHO, taking anything away from the second question (preferences) when the former question (detectable difference) has not been proven valid may be putting the cart before the horse.

I agree. In my experiment the results where at the edge of being statistically significant. However, given the fact that nearly every trained palate got it right, that tends to make me think there is something there.

Several of these people in fact where 100% sure. In a funny example I came back to one gentleman (a certified judge and craft beer bar manager) who seemed to be perplexed and was taking a long time. I said, "still not sure?" He said, no I'm 100% sure sample 2 is the different one (he was right) I'm just trying to figure out what the exact difference is. There was also a head distiller who, not only got it right and was sure of it, but she was more or less able to guess the nature of the experiment.

All this makes me think that there is a identifiable difference to at least some people and if that is the case we ought to concern ourselves with it.

 

[ajdelange]

It is very important to understand that in a triangle test you are testing the panelists and not so much the beer. Thus the composition of the panel is critically important in getting usable information out of a triangle test. I always remember a comment made by a fellow investor to the effect that he didn't really like the Saison but as the customers loved it he really didn't care. Panels made up of a bunch of guys like the investors (all are big-time BJCP panjundrums) as opposed to ones made up of the "unsophisticated" customers) would lead to very different conclusions.

 

[amcclai7]

One thing missing from these efforts is an attempt to quantify the preference. I might prefer A over B, but if I don't really like either, is A better or just less objectionable? This is a difficult thing to assess in these studies, so I'm not being critical, just that there still is a variable we're missing.

This is exactly my criticism of the brulosophy exbeeriments and my own. Our results where shockingly similar but we both forgot to ask that. Even in my slightly more informal setting I can't even remember asking the subjects, "do you normally like IPAs?" I probably did at some point but am kicking myself for not doing so in a more formal manner.

In all three experiments I listed in the OP I have a hunch that you are exactly correct. Those that preferred the more traditional taste were fans of the "correctly" brewed beers and those that weren't found the "incorrect" beer simply less objectionable. Again, just a hunch, but I smell another experiment!

 

[amcclai7]

It is very important to understand that in a triangle test you are testing the panelists and not so much the beer. Thus the composition of the panel is critically important in getting usable information out of a triangle test. I always remember a comment made by a fellow investor to the effect that he didn't really like the Saison but as the customers loved it he really didn't care. Panels made up of a bunch of guys like the investors (all are big-time BJCP panjundrums) as opposed to ones made up of the "unsophisticated" customers) would lead to very different conclusions.

Agreed. Mine was a mix and if I had it to do over again I would probably shoot for all "professional" palates.

 

[jmitchell3]

It is very important to understand that in a triangle test you are testing the panelists and not so much the beer. Thus the composition of the panel is critically important in getting usable information out of a triangle test. I always remember a comment made by a fellow investor to the effect that he didn't really like the Saison but as the customers loved it he really didn't care. Panels made up of a bunch of guys like the investors (all are big-time BJCP panjundrums) as opposed to ones made up of the "unsophisticated" customers) would lead to very different conclusions.

I feel like "usable information" is the key here. In a commercial QC setting they are actually going deeper into qualitative data based on the known abilities of the tasters. Furthermore, the controls are ongoing and repeated exhaustively also.

In the brulosophy experiments, they are simply trying to see if there is any perceptible difference. Their conclusion reads the same..."this suggests that blah blah are / are not able to reliably distinguish..."

I'm all for someone having a dedicated, highly trained SWAT tasting panel with proven abilities to taste test the crap out of any and all of these experiments over and over again until we have good data and repeatable results [that are statistically significant]. But those aren't the quality of the panelists of any of these "exbeeriments" (and I think that is your point, ultimately). But at the end of the day, the public and a VAST majority of people are not highly trained SWAT tasting panel specialists either. We as brewers don't brew for the biologically-gifted few, we brew for everyone. Such is the democratic egalitarian nature of beer! This my point...if the broad majority of people (often including the brewer!) cannot distinguish among them with statistical significance and repeatability among multiple experiments evaluating the same variable...then that variable is just not a thing and is likely not worth attending to...

 

[mongoose33]

Sure I get that and I see where you are going in your thinking.

I guess my point is that if results are not significant, there's no proof that anyone can reliably tell a difference with that particular variable. If this is so, it doesn't matter what the preferences are for the group that can tell a difference, because the fact they chose the right odd-beer-out is not distinguishable from pure chance.

Well, we're headed here in a direction that may not be appreciated by those who lionize the exbeeriments, and in so doing, I'm not trying to cast aspersions on them, or Marshall and his site.

In fact, Marshall's the only guy I know who's actively trying to test these things and publish results. (if there are others, let me know, whomever might be reading this). He has improved his processes over time, to the point where I think they're actually pretty good. Results are expressed in terms of the panel (good thing), and there's been a trend toward better testing (triangle tests, e.g.), and the way results are expressed is one with which I'm very comfortable. He's engaged in something I tell my students to do all the time: continuous quality improvement. To his credit, he keeps working to get better at this.

[I've actually used exbeeriments in a research class I teach, because they're relatively simple to explain, and we can see how the significance of the results is different from what it means. So nobody should take what I'm saying as critical of Marshall, rather, the guy should be held up as an example of how to be honest in trying to do science. And I do.]

Anyway, to your point, there's no proof that anyone can or cannot tell a difference. One problem w/ the approach is that we do not really know what the sample of testers represents. They're just a convenience sample. Not trying to be mean here, but are there regional differences in what people like? Or what they can perceive? Is the panel composed of IPA lovers, or those like me who like maltier concoctions? Is it different between craft brew aficionados and those who judge competitions? And brewers? In the end, I don't know to what the results are generalizable.

Further, the significance of the results is likely compromised by the guessing. If someone cannot tell a difference among the three samples tested, and they guess the right one (and it's not a taste-informed conclusion but rather a simple guess), should they be included when asked for a preference? Or included in the statistics? What is it we're really looking for?

On its face, this looks like this kind of research should be easy to do. It's not.

Are BJCP judges as a group more biologically able to distinguish taste differences than non-bjcp judges? Or is it just that they know what the difference actually is and can identify and comment on said difference (i.e. sensory training vs. the layman).

Yeah, that's my point about generalizability. I'm not a BJCP judge, but I can tell differences, and I know what I like. But am I typical? Is the sample typical of anything? Or is it just beer geeks?

Further, I think I'm suggesting that trained judges as a group are probably not anymore physically able to taste differences in anything from snicker bars to cognac than non-trained judges. That is, I don't know that there is anything about trained industry people that are biologically superior to their non-trained counterparts that would allow them to be superior tasters. [I'm sure someone can make an anecdotal argument that those who choose the food or beverage industry as a profession or even a hobby may perhaps be biologically inclined (as a group) to superior physical ability in tasting and evaluating]. They surely, however, would be ones I'd look to on a preference descriptor to see how they evaluate differences they do taste.

I don't know that they're biologically superior--maybe it's just training people what to look for. Some people are super-tasters, some (like me) are not. Can we identify who those people are? And what would that mean for the results?

A friend of mine is described by a local brewer as having the best palate of anyone he's known. That friend and I were tasting a beer in which apricot flavor was added. I was unable to taste it. I had someone else taste it, and she said she couldn't detect it either. But he could.

If he's in a tasting panel, and there are a lot of others like him, what would the results mean to someone like me, who can't taste that? I don't know, but it strikes me as a meaningful question.

I'm reticent to take any actionable intelligence from any experiment where there is no significant, demonstrable, statistical observation that a difference actually does exist that humans can actually detect. IMHO, taking anything away from the second question (preferences) when the former question (detectable difference) has not been proven valid may be putting the cart before the horse.

Yes--though to me, the most informative exbeeriments are ones which assess process and which determine either that there is no detectable difference or that there was a huge detectable difference *and* virtually everbody preferred one over the other.

For instance, the great Trub exbeeriment is what led me to stop straining out the break and hops from my wort as it entered the fermenter. My beer is no worse; actually, I think it's better, but that may be from my perfecting other process elements rather than the trub. But in the end I had to try it on my own system, with my own beer, and see if it made a difference.

Ingredient exbeeriments I'm less enamored of, simply because taste is taste, my palate is not your palate, yada, yada, yada.

However, I would like to suggest that, like any other scientific experimentation, these Exbeeriments should be replicated over and over again with similar results before we indeed have any reason to change our behavior. Single experiments testing a single variable are hardly demonstrative of substantive information. again, IMHO.

Agreed. Once is not enough for a high level of certainty, and with the issue of what the tasting panels are generalizable to....

And, of course, this all represents my personal human opinion, which, like many things, is subject to change without notice and will not in any way shape or form affect all of our pursuit of making better beer by using any and all means / information at our disposal!

Here's where the internet shines, when it does: when we can have these discussions without them become personal or insulting. I am in search of the truth, wherever it may be found, and as long as I don't attach my ego to my thoughts and ideas, there's no danger to me or my ego if they are proven incorrect.

We're all trying to figure out how to do this better.

 

[RM-MN]

WOW....this is EXACTLY what I have always asked but nobody seemed to have a conclusive answer. Thank you so much!!!

I BIAB too and mill my grains nicely fine. I experimented with the grind and am hitting 81% efficiency every time like clockwork. I'm simply saying my grind is pretty fine so it must be like yours.

I am concluding from your tests that accurate ph adjustments are darn critical from the dough in since there is almost no time to react to adjustments later in the mash. Like I said before...shutting the barn doors after the cows are out.

I use Bru'n water religiously and adjust as directed. I'll always test ph at the 20 minute mark and I'm always really close to what the program predicts per my grain bill. I adjusted for 5.45 last brew and my Hach read 5.41 at the 20 minute mark. I figured I must be close since my efficiency numbers are over 80%.

This is really really good information my friend, and I thank you so much for sharing!!!

Note: May I ask one more thing, please? Since we are confident we have conversion this quickly, what are the benefits (if any) of continuing the mash cycle for 60 minutes as an average?

Conversion of starch to sugar isn't the only thing happening in the mash, its just the easiest one to quantify. I've tried different mash timings based on the conversion time and what I found was that the conversion was quick but flavor extraction was not. My minimum mash time is now 20 minutes but I most often prefer 30 minutes because I like my beers to have flavor and a 5 minute mash, while getting conversion with the very fine grain particles, just didn't have much flavor. If I could stand to drink several I could get drunk because the alcohol was there but I prefer to enjoy each beer and never get drunk.:rockin:

 

[jmitchell3]

Yep, think we're on the same page.

Well, we're headed here in a direction that may not be appreciated by those who lionize the exbeeriments, and in so doing, I'm not trying to cast aspersions on them, or Marshall and his site.

In fact, Marshall's the only guy I know who's actively trying to test these things and publish results. (if there are others, let me know, whomever might be reading this). He has improved his processes over time, to the point where I think they're actually pretty good. Results are expressed in terms of the panel (good thing), and there's been a trend toward better testing (triangle tests, e.g.), and the way results are expressed is one with which I'm very comfortable. He's engaged in something I tell my students to do all the time: continuous quality improvement. To his credit, he keeps working to get better at this.

[I've actually used exbeeriments in a research class I teach, because they're relatively simple to explain, and we can see how the significance of the results is different from what it means. So nobody should take what I'm saying as critical of Marshall, rather, the guy should be held up as an example of how to be honest in trying to do science. And I do.]

Anyway, to your point, there's no proof that anyone can or cannot tell a difference. One problem w/ the approach is that we do not really know what the sample of testers represents. They're just a convenience sample. Not trying to be mean here, but are there regional differences in what people like? Or what they can perceive? Is the panel composed of IPA lovers, or those like me who like maltier concoctions? Is it different between craft brew aficionados and those who judge competitions? And brewers? In the end, I don't know to what the results are generalizable.

Further, the significance of the results is likely compromised by the guessing. If someone cannot tell a difference among the three samples tested, and they guess the right one (and it's not a taste-informed conclusion but rather a simple guess), should they be included when asked for a preference? Or included in the statistics? What is it we're really looking for?

On its face, this looks like this kind of research should be easy to do. It's not.



Yeah, that's my point about generalizability. I'm not a BJCP judge, but I can tell differences, and I know what I like. But am I typical? Is the sample typical of anything? Or is it just beer geeks?



I don't know that they're biologically superior--maybe it's just training people what to look for. Some people are super-tasters, some (like me) are not. Can we identify who those people are? And what would that mean for the results?

A friend of mine is described by a local brewer as having the best palate of anyone he's known. That friend and I were tasting a beer in which apricot flavor was added. I was unable to taste it. I had someone else taste it, and she said she couldn't detect it either. But he could.

If he's in a tasting panel, and there are a lot of others like him, what would the results mean to someone like me, who can't taste that? I don't know, but it strikes me as a meaningful question.



Yes--though to me, the most informative exbeeriments are ones which assess process and which determine either that there is no detectable difference or that there was a huge detectable difference *and* virtually everbody preferred one over the other.

For instance, the great Trub exbeeriment is what led me to stop straining out the break and hops from my wort as it entered the fermenter. My beer is no worse; actually, I think it's better, but that may be from my perfecting other process elements rather than the trub. But in the end I had to try it on my own system, with my own beer, and see if it made a difference.

Ingredient exbeeriments I'm less enamored of, simply because taste is taste, my palate is not your palate, yada, yada, yada.




Agreed. Once is not enough for a high level of certainty, and with the issue of what the tasting panels are generalizable to....



Here's where the internet shines, when it does: when we can have these discussions without them become personal or insulting. I am in search of the truth, wherever it may be found, and as long as I don't attach my ego to my thoughts and ideas, there's no danger to me or my ego if they are proven incorrect.

We're all trying to figure out how to do this better.

 

[TheMadKing]

I had read that conversion happens quickly but nobody seemed to say what "quickly" meant so I decided to experiment with that. I BIAB and I mill my grains very small so they can gelatinize quickly and expose the starches to the enzymes so your results may be far different if you use a coarse milling of the grain.

For my first try I used a white plate with several drops of iodine that would be my indicators of the presence of starch. I doughed in and took a sample immediately as a control which turned blue when a drop was placed on the iodine so that showed starch was present. Then I waited for 5 mnutes as I wanted to do the test at 5 minute intervals to see how quickly the "quickly" was. That 5 minute test showed no starch left. Hmm... not the results I expected.

The next time I brewed I decided to take the samples at 1 minute intervals since 5 minutes was too long. Again I took a test at dough in with the same results, and the same results at 1 minute. OK, now I know that quickly means "more than one minute". At 2 minutes I tried again and.....conversion was done according to the iodine. Having a science background, I repeated this on the next time I mashed with the same results, conversion in less than 2 minutes.

Now again, it depends on your milling of the grain as the starches must gelatinize to expose them to the enxymes and larger particles take longer but I suspect that most of your conversion is done in less than 10 minutes. That means that if you adjusted the pH at 8 minutes, most of the conversion was done at the incorrect pH.

this only measures starch conversion in the wort, not within the grain particles, and at what percentage does the iodine test show a perceptible positive? 80% conversion? 95%? 99%?

Conversion rate is an exponentially decreasing function, so while most conversion happens in the first few minutes, it is continuing to some degree for the whole hour.

Another experiment to try is: mash for 10 minutes and mash for 60 minutes and measure your attenuation in a forced ferment condition. I strongly suspect that your 60 minute mash will attenuate significantly lower than your 10 minute mash.

Brukaiser has also performed a number of conversion experiments and they were not consistent with your results.

 

[RM-MN]

this only measures starch conversion in the wort, not within the grain particles, and at what percentage does the iodine test show a perceptible positive? 80% conversion? 95%? 99%?

Conversion rate is an exponentially decreasing function, so while most conversion happens in the first few minutes, it is continuing to some degree for the whole hour.

Another experiment to try is: mash for 10 minutes and mash for 60 minutes and measure your attenuation in a forced ferment condition. I strongly suspect that your 60 minute mash will attenuate significantly lower than your 10 minute mash.

Brukaiser has also performed a number of conversion experiments and they were not consistent with your results.

If I only had wort for sampling this would be true but when I included grain particles they also show negative on the starch test. With small particles the starch is quickly gelatinized and converted. With larger grain particles the process continues for a longer time until with whole grains it may take hours or even days.

The 3 batches I did with 5, 10, and 20 minute mashes all seemed to attenuate the same but the shorter mash batches did not taste as good. These were all blond ales so I could make a reasonable conclusion. I don't particularly like blond ales so I will not repeat. Someone else can pick up the torch.

 

[TheMadKing]

If I only had wort for sampling this would be true but when I included grain particles they also show negative on the starch test. With small particles the starch is quickly gelatinized and converted. With larger grain particles the process continues for a longer time until with whole grains it may take hours or even days.



The 3 batches I did with 5, 10, and 20 minute mashes all seemed to attenuate the same but the shorter mash batches did not taste as good. These were all blond ales so I could make a reasonable conclusion. I don't particularly like blond ales so I will not repeat. Someone else can pick up the torch.

Did you mash the crushed grain particles with a spoon before adding the iodine?

 

[Morrey]

Conversion of starch to sugar isn't the only thing happening in the mash, its just the easiest one to quantify. I've tried different mash timings based on the conversion time and what I found was that the conversion was quick but flavor extraction was not. My minimum mash time is now 20 minutes but I most often prefer 30 minutes because I like my beers to have flavor and a 5 minute mash, while getting conversion with the very fine grain particles, just didn't have much flavor. If I could stand to drink several I could get drunk because the alcohol was there but I prefer to enjoy each beer and never get drunk.:rockin:

I am with you 100% Good post and very helpful information you shared. Thanks!

 

[Morrey]

This is pretty much what we do. We take a water Ph, which varies quite severely throughout the year, (7.8-6.5 depending on the season) put our recipe into the water calculator, adjust with salts and/or lactic acid via the water calculator, and mash in.

We will take a Ph after we mash in, but as RM-MN said, conversion happens very quickly so we don't worry about adjusting after mash in unless the Ph is way off. We just use it to inform our next calculation. For example, I know that when the water Ph is around 7 the water calculator is extremely accurate. When the water Ph is 7.8 it might be off by as much as .2 high and vice versa for low water Ph.

Fyi, the picobrew in the experiment was done at home by my friend so it was not Ph tested during the mash. This would have been great but given the fact that he was using RO water I feel fairly confident that the water calc was at least close. Also, given the fact that one got lactic acid and one didn't, I at least know that beer A was at a lower mash Ph than beer B even if it was, for example, 5.1 vs. 5.4 instead of 5.2 vs. 5.5

I am glad you shared your ranges of ph as I thought something may be off with my calculations. I use municipal water and I figured from the onset the tested values from Ward Lab would be somewhat set in stone. My report said my ph was 7.5. So on my next brew I just double checked my water ph and it was 6.9. Whoa threw me a curve ball, so I get back into Bru'n Water and input the current ph which adjusted my mash ph lactic additions. I don't have a way to spot test my other salts, so I'll hope for the best and assume they are not changing as much as the ph. Even if they are, I can at least control the ph on a brew to brew basis.

 

[mabrungard]

Complete starch conversion has long been reported to occur within 15 minutes. But that isn't the whole story. The wort is then composed of large polysaccharides that aren't very fermentable. Additional time is required to chop those large sugars to smaller and more fermentable sugars. And it doesn't matter if you are mashing at low temp, it still takes time for the enzymes to act.

15 minutes or less, is not enough time to create a good wort.

 

[skraeling]

Here's where it gets really interesting: When asked which they preferred, the results were pretty much evenly split. Those who preferred the lower Ph often described it as: Crisp, bright, and dry and the higher Ph as flabby and too sweet. Those who preferred the Higher Ph described it as fruity and smooth while calling the lower Ph sharp and too bitter.

Given how Soda is VERY acidic, and also has quite the bite to it (much like orange juice) I would agree with this finding. Frankly im not sure I could taste a few tenths of a ph difference ever either.

 

[RM-MN]

Complete starch conversion has long been reported to occur within 15 minutes. But that isn't the whole story. The wort is then composed of large polysaccharides that aren't very fermentable. Additional time is required to chop those large sugars to smaller and more fermentable sugars. And it doesn't matter if you are mashing at low temp, it still takes time for the enzymes to act.

15 minutes or less, is not enough time to create a good wort.

My test batches with very short mast times still had good attenuation. I didn't note any real difference in the fermentability between a 5 minute mash and a 30 minute mash. What I really did notice was the lack of flavor in the short mashes. By 20 minutes I seemed to have extracted the flavors too but I would caution any brewer to not mash for less than 30 minutes. Just because I can do it does not mean someone else can.:rockin: