The scientific method

[schrodinger]

A great strength of this forum is the wealth of anecdotal evidence and impressions based on personal brewing experience. Another is the willingness of Aussies to screw the rules and try something new. Together, these are the seeds for real scientific progress: being observative and being curious.

The weakness of a lot of this knowledge is that it comes from uncontrolled experiments and is not replicated. That is, you decide to try something and it produces a good beer, so you believe it works. But who knows what else may have made that beer great, that you weren't aware of, or thinking about, or paying attention to?

As a scientist myself, I'm painfully familiar with how much can be gained by using controls and replicating experiments (I say painful because I've wasted so much effort on inadequately controlled or replicated experiments). So, I'd like to suggest that we put our heads together and do some experiments, and do them right.

For clarity: by 'control,' I mean you do at least two batches, and they differ in only one respect. Maybe it's yeast pitching rate or hop times or something else, but the key is to change nothing else whatsoever. This is easiest to pull off for cold-side factors, because you can split your wort into two fermenters, but with proper experimental design we can control for things that may randomly differ on the hot side between two batches done, say, two days or a week apart.

By replication, I mean, well, it's obvious. Ideally the same brewer would do a bunch of identical batches, but that will never happen. We can get replication by comparing results across brewers, provided enough people participate.

SO... assuming you're still reading this, I'm requesting suggestions (a) for a particular factor to test, (b) for a small number of important variables to measure in the product (these may be tricky and subjective), and (c) for anyone willing to participate. We could call it the 'Great AHB (... Insert topic here...) experiment'.

I will coordinate, design the experiment, collate and analyse the results, and accept blame when it fails to work out.

Anybody interested?

 

[Glot]

Is science. Is good.

 

[manticle]

There's been a number of side by sides I've intended to do over the years but rarely got round to any of them.

The only one I did do and the analysis was extremenly subjective (but was always going to be) was a chill vs no chill version of the same beer (actually double batch, half chilled, half not) of a highly late hopped beer.

It's difficult to eradicate all variables - in mine for example, even though I pitched the same yeast strain, one was pitched immediately and one a day or so later so they weren't really side by side fermentations.

I also wanted to compare supermarket gelatine, Homebrew shop finings, isinglass and nothing and the effect on clarity of finished beer - one of each in ambient temps and one of each with cold conditioning. Got the ingredients and some vials but never got around to it.

The great NC botulism debate of 1879 did see an experiment idea eventuate which would have been interesting - can you infect no chilled wort with botulism spores and encourage the toxins to transfer over to the finished beer. Never happened and the logistics would be close to impossible.

AG versus extract - taking the highest quality briess or weyermann malt, putting together a known recipe of mine (or could be tried and true like smurtos golden ale) using whatever spec malts and doing side by side ferment and blind triangle taste test.

Some others I've considered rolling around in the back of my head somewhere

 

[Crumpet]

I currently don't have the control or room to participate but here are a couple (some quite debated)

Fermentation on trub v.s. no trub
different forms of aeration v.s. no aeration what so ever (can measure lag time and flavours produced)
yeast pitching rates
re-hydration of yeast v.s. not
Small temperature fluctuations during fermentation v.s. steady temperatures
pitching warm then cooling v.s. pitching at fermentation temps (again lag time and flavours can be measured)

 

[Adr_0]

I like, and as an engineer I do try to measure as much as possible or at least understand the mechanisms behind a lot of what we observe.

Some things are easy:
- Two identical worts, two identical yeasts, two separate fermentation temperatures
- Two identical worts and boils, two separate cooling regimes (e.g. plate vs bath)
- etc.

Some things not so much:
- Chemical analysis
- Yeast counting and health
- Perceived flavours
- Consistency, e.g. I can set my fridge to 19°C but the wort might get anywhere from 18.5°C to 22°C vs Joe up the road who might have tighter control

I have particular interest in the following:
- Chorine, chloramines and chlorophenols in beer
- pH with and without additions of salts
- Wheat strains, temperature dependence of cloves perception, acid rest dependence on cloves perception
- DMS factors
- Diacetyl factors

 

[professional_drunk]

Sounds good but first up where will you document the method, results etc? There's nothing suitable on this website for doing such a thing.

 

[schrodinger]

Thanks, all! Pro_drunk, good question; we could simply outline the results here as a thread (using HTML tables for data and jpegs for figures). If we do a good enough job, we could shoot for an academic journal (the Journal of the Institute of Brewing or the Journal of the American Society of Brewing Chemists).


Ideas mentioned above, so far, that sound feasible and potentially interesting for a small army of homebrewers:

1. compare wort aeration rate (none, medium, high) -- measure lag time, total attenuation, and qualitative factors
2. trub or no trub -- less to measure quantitatively on this one, but with enough samples, we could get a good handle on the qualitative effects
3. steady vs varying temperature -- same measurements as in (1); could be hard to control the varying temperature regime, but worth a try
4. compare different wheat strains -- I like this because I work on a wheat research station. Question is, what would we measure?

others alluded to:

5. boil time -- effect on DMS? We'd control for hop utilisation (i.e., do the same hop schedule relative to the end of the boil). Might be difficult to measure DMS, but I have colleagues who have the necessary equipment and might be keen to contribute.
6. temperature of single-infusion mash -- measure OG, FG, and measure body, etc., qualitatively. This one is cool because we could directly test all the hooyah about alpha vs beta amylase temperature optima


Adr_0: your pH idea is perfect, but almost too easy because you'd get the result on the spot. On the other hand, maybe that's a strength, and a good place to start.


**One important point: we don't have to have replication within each brewer's setup. We can replicate across brewers. e.g., have 10 guys brew at condition A, another 10 at condition B. So there's room for just about everybody to participate. "Strength in numbers" is a very real thing in science. The larger the sample size, the more uncontrolled variation we can tolerate.


Any other thoughts/ideas?

 

[schrodinger]

manticle, your idea to compare finings and measure clarity is also great. I also like the idea of comparing extract vs AG. Maybe a good club could make extract kits from an AG recipe, then we could compare beer made from those kits vs. directly from the AG recipe?

 

[professional_drunk]

Here's my suggestion on how these experiments could be conducted:
1. Either using google sites or docs, create an experiment to outline aim, hypothesis, required equipment, method.
2. Use ahb thread to get input from people on refining this article in terms of best equipment, method etc. Also to collect participants.
3. Allocate a time and have all participants follow the experiment.
4. Collect all results.
5. Ahb thread to analyse results.
6. Write up pdf with experiment, results, conclusions and stick it back onto the thread.

So ahb thread would be for feedback, collaboration, discussion but at the end of the day I'd think you'd want a single document to outline the methods and results.

example:
Question: does fermenting wort create more heat outside of ambient and if so by what degree. This will hopefully give us some insight into the age old question of probe placement in a fermenting chamber.
Hypothesis: fermenting wort generates heat.
Test method:
-create 23lt wort of 1.050 gravity and pitch yeast.
-calibrate 3 stc 1000 so they read the same temperature.
-place 1 stc probe into wort, 1 into bottle of water, 1 to control fermenting chamber.
-measure over 2 weeks.

Equipment
-fermenting fridge or freezer
-3 stc 1000
etc.

Result
-some chart with participants, their equipment and their readings from the stc

Conclusion
-more experiments needed, formulate new hypothesis or
-probe in fermenting chamber should go into wort for accuracy or
-redo experiment because stc probe should measure air instead of container of water

IANAS (I am not a scientist)

 

[verysupple]

I currently don't have the control or room to participate but here are a couple (some quite debated)

Fermentation on trub v.s. no trub
different forms of aeration v.s. no aeration what so ever (can measure lag time and flavours produced)
yeast pitching rates
re-hydration of yeast v.s. not
Small temperature fluctuations during fermentation v.s. steady temperatures
pitching warm then cooling v.s. pitching at fermentation temps vs pitching cool and allowing to free rise to fermentation temps (again lag time and flavours can be measured)

I like these ideas because they are so hotly debated. It'd be great to finally put some nails in coffins...although even after a sucessful experiment I expect these topics will still be debated.

I have one minor suggestion in red.

 

[verysupple]

I've also been planning an experiment of my own regarding rehydrating dried yeast. I'm in the process of acquiring a microscope for counting cells and have the ingredients for alkaline methylene blue pH 10.6 (for reasonably accurate measurement of viability even at low viability (in contrast to regular methylene blue which is known to be very inaccurate at viabilities below 90 %)). I don't have a decent pH meter though, so maybe people that own them could do the acidification power tests for vitality (if they have the other necessary equipment and confidence in the lab).

I'd like to test rehydrating in DI water, DI water made up to 250 ppm permanent hardness, 1.040 SG wort, and 0.9 % saline solution.

It would be great if the experiments were well designed so as to minimise interobserver variability and we could get a multiple sets of data for statistical power.

I have most of the methodological and analysis details figured out if schrodinger and others want to take this one further.

 

[GeoffN]

I'm hoping this thread goes where it is intended. I don't think I'll be able to participate as I only do kits and that infrequently as I'm not a heavy drinker, I just like the taste of beer. Only did 3 brews this year with what I had left over from last year I only went without beer for about 2 weeks.

I would like some idea of the effect of various salts on the final taste of kits. I've recently moved from regional Victoria to the sunny cost and my latest brews are a little harsher. As I only do kits it is either temperature or water. Latest brew is still conditioning and I did temperature control far better on this brew so that may fix it. I have only had diacetyls once and one toucan where the isohop bitterness was a little objectionable until now. Water is harder here than Drouin.

Please keep rolling with the suggestions.

Sent from my HDC-08 using Tapatalk 4

 

[thylacine]

Site: "Basic Brewing Radio" Just a few of their experiments...
November 7, 2013 - IPA Experiment Results
Chris Colby and Denny Conn join James to go over the results of adding increasing amounts of gypsum to an American IPA.
August 1, 2013 - Fruit Juicer Experiment
Homebrewer Jameson Parker shares his experiment exploring whether it makes a difference to run fruit through a juicer before adding to beer.
July 18, 2013 - No Chill Aging Experiment
Bob Stempski returns with another no-chill experiment. This time, how does storing unpitched no-chill wort for four months affect the final product?
July 18, 2013 - No Chill Aging Experiment
Bob Stempski returns with another no-chill experiment. This time, how does storing unpitched no-chill wort for four months affect the final product?
April 11, 2013 - Perceived Bitterness Experiment
Homebrewer Kevin Ellis shares an experiment comparing a beer with zero calculated IBUs against a more traditionally hopped beer. Brad Sturgeon supplies lab results.
March 7, 2013 - Zot's Specialty Grain Experiment
Zot O'Connor joins us to talk about his experiment in adding specialty grains to a finished beer to taste the effects.
January 17, 2013 - Specialty Grain Experiment
Homebrewer Darnell Brawner shares his experiment looking at how Carastan, crystal 40, and crystal 60 affect the same recipe.

 

[krausenhaus]

Deleted and reposting this because the paragraphing didn't come through when I typed it on my phone.

I have access to a 2L bioreactor with dissolved oxygen and pH probes, aeration/stirring rate control and PID temperature control. There is an exit gas analyser too that can measure CO2 composition in the headspace (it's meant to do O2 as well, but it's busted). Also have a UV/vis spectro and a range of other usual lab gear. No microscope in the lab but I can use one elsewhere.

I'm probably only going to be able to use the bioreactor up until around June/July. I'm happy to test some things in between my own (non-brewing-related) experiments. I can't do too many full fermentations as they require a lot of attendance and set-up/clean-up time.

What I would like to do is look into the rehydrating in water/wort vs not rehydrating. I can monitor oxygen uptake rate online which is a good indicator of cell growth, and can also use the spectro to measure growth rate indirectly. Similarly, I can also look at the effects of various levels of initial aeration

I don't have a huge amount of time to dedicate to this so if you guys want to help out with experimental design and let me know what you'd like to see, I can do the lab work when I get the time.

 

[schrodinger]

Site: "Basic Brewing Radio" Just a few of their experiments...
November 7, 2013 - IPA Experiment Results
Chris Colby and Denny Conn join James to go over the results of adding increasing amounts of gypsum to an American IPA.
August 1, 2013 - Fruit Juicer Experiment
...[

thylacine, first, I love your moniker. Second, thanks for pointing out that site! I hadn't seen it. I've just listened to two of the expts in parts so far, and they're cool. But they seem to have no replication (n=1) and to change more than one variable, at least in some cases, so they're not experiments in the modern scientific sense. I'm probably wrong about some of them, though, and will listen to them.

It's all about maximising the ratio of n/p (number of independent measurements divided by number of independent variables)...

 

[schrodinger]

I'm excited there's some interest in this idea. It seems we have several hands up for aspects of yeast pitching, which is fantastic because it's cold-side and it's something every brewer does.

Krausenhaus, that's an awesome setup. It would be great to add some of those measurements; a limitation is that only you can do them, so they won't be so replicated, but that's fine. I can measure CO2 efflux as well, so maybe n=2 there.

GeoffN, I'll bet it's your water. Let's think about experiment #2 focusing on water -- again, something else everyone can fiddle with. Also something we can get analysed for everyone's brew water samples without having to store and ship them at -80 deg C.

Verysupple, I like that idea. As with k-haus your vitality measures can inform the broader results. Let's make sure our control and treatment levels include two of the levels you're keen to look at.

Pro-drunk: good thinking on Google docs. That will make it easier to update a central source for the methods/instructions, and especially for participants to enter their results into a consistent pro forma.

What would be really cool is if we could get a lot of the resulting beers and brewers together. I'm new to the scene in Aus; is there a suitable conference or other pow-wow?

Let's keep this ball rolling. Please, tell your friends on AHB about the idea. Remember that the sciency aspect of it will come mainly from the experimental design -- people won't need to do anything more difficult than normal, except to brew a particular recipe following a particular protocol -- and from sheer numbers. (Hell, the entire science of ecology is based on collecting large sample sizes because you can't really control for anything. We can do better, and get better answers.)

Strength in numbers!!! Strength in beer!!! (Can someone translate that into Latin?)

 

[toncils]

I'm interested in how much plastic gets into your brew during no chill. Difficult to gauge, methinks.

 

[Ducatiboy stu]

BIAB V 3v AG

The perfect experiment to see the difference between 2 different production methods. You could measure things like time and overall finished product. Each method gets the same recipie sheet,yeast etc,but considerations would be needed for things like water.

A lot of BIAB & 3v brewers have some sort of temp controled fermenting fridges some that is a variable that can be reduced.

The difference between the 2 methods has produced some interesting debate over the years so it would be nice to see an actuall measureable comparison. With the end result being beer.

 

[primusbrew]

Hey shrodinger, I like the sound of this and I would be interested in participating.

A couple of topics I would be interested in having a look at:

- First Wort Hopping (FWH) versus a start of boil addition. FWH is supposed to produce a smoother bitterness so I use it but don't really know for sure.
- Yeast nutrient versus no yeast nutrient. I add it at the end of the boil but don't really know if it has much effect.
- Brewing salts versus no brewing salts. This one might be probematic as everyone will have different starting water profiles. We could possibly aim to produce two water profiles that eveyone can achieve and evaluate the differences. ie one high sulphate and one high clacium
- Different mash temperatures. Or different mash durations at the same temperature.
- I like the idea of BIAB versus 3V but I only have facilities for BIAB.

Anyway, as long as I have the gear/ability to control to particpate I'll be in.

 

[Dave70]

I will coordinate, design the experiment, collate and analyse the results, and accept blame when it fails to work out.

So, will this constitute participating members posting you stubbies labeled 'sample A' and 'sample B'?

You clever bastard.

Thats why I love science.