Fermentables,Attenuation, and Specific Gravity Mathematics

Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum

Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

SMc0724

Well-Known Member
Joined
Mar 7, 2011
Messages
143
Reaction score
3
Location
Beaumont
The goal of this post is to propose a couple of ideas about the mathematics of fermentables and attenuation: One can find several posts here, and on other forums, where home brewers make some assumptions in there brewing calculations about attenuation, but I can find no proposed set of algorithms or equations available for public criticism and consumption. Am I wrong?

Before delving into this thread, please visit the page https://www.homebrewtalk.com/wiki/index.php/Understanding_Attenuation, because this post assumes a working understanding of specific gravity, original gravity, final gravity, PPG, etc., as well as knowledge of such definitions as apparent attenuation, actual attenuation, fermentable, among others.

Further, this post assumes the basics of attenuation are understood, that is the mathematical relationships of attenuation, OG, FG, as well as specific gravity both of water and ethanol.

One more thing, before delving into the ideas, the goal is to establish a quantitative methodology that can estimate FG and ABV, not to derive a equation that calculates FG and ABV with absolute certainty. This post intends to suggest a methodology that will get final gravity points to no more than plus or minus 4 FGPs. The premise of this post is a realization that predicting the FG of a beer is like predicting the weather, but it is a start.

Leading to the question of why predict FG and ABV? IMHO, I believe that most would say, primarily for the purpose of flavor, more than anything else. Most, especially seasoned brewers, use these numbers largely to compare against given style values to see if a decent beer is being created. Flavor in beer is a tradeoff among sweetness, ABV, IBU, malt character, yeast profile, among others. Getting a better prediction of FG and ABV, along with IBUs, etc., just provides a better idea of what lies beyond the brew.

Finally, these are my words and notions, though derived from others, please feel free to contribute. I'm just looking for a "better mouse trap."

Enough with the prelims....

It seems safe to say that attenuation is primarily influenced by these three operations: mashing, lautering, and fermenting; and these two substances: type of starches/sugar(s), type of yeast. (Some may want to introduce other factors here, so be it. For one, I can think of the presence of enzymes from base malts. IMHO, I suggest that only the most important factors be introduced, and only those that can be quantified. And by all means, please suggest a method for mathematically including these factors. Thanks.)

Starches/Sugar: Typically, the starting point to calculate the final gravity of wort is the PPG, a measure of specific gravity points contribution per gallon, called Max PPG. To predict FG, this post implements the idea (suggested by others) of splitting of PPG into three parts: simple (fermentable) sugars, complex (but potentially fermentable) starches/sugars, and unfermentable (not fermentable under normal brewing conditions) starches/sugars. Once split, this post intends to quantitatively track these three values through the brewing process. To do this, let's define Fermentable Percentage (FP) as the fraction of PPG that is both fermentable (simple) and potentially fermentable (complex). For example, extract from crystal malt grains have a high percentage of unfermentables (30-100%), and other sugars, like lactose, are 100% unfermentable. The Fermentable Percentage would be what remains if the unfermentables are removed.

Now perform a "PPG trace" through the brewing operations:

(1) Mash: In this operation, a "PPG trace" requires a measure for the enzymatic conversion of the complex starches/sugars to simple (fermentable) sugars, a conversion that varies with both temperature and grain mixture. This measure can be hard to find in published sources, but it is around. Palmer, for example, has a huge discussion on these topics in his book (pp. 144ff.) and even shows a figure in his book on "Apparent Attenuation Limit." For this post, let's define this variable as the Limits of Attenuation Percentage (LAP). Note that this value is defined in terms of apparent attenuation, that is OG and FG.

(3) Lauter: For this Operation, a "PPG trace" needs a measure for the efficiency of extracting sugar from grain. This measure is widely available, and is typically around 0.75 for all-grain and about 0.65 for partial mashes, but 100% for extracts. Let's call this variable the Extraction Percentage (EP), again defined in terms of apparent attenuation.

(4) Ferment: Here the yeast's ability to ferment is needed, commonly measured and known as attenuation, YA. It is important to remember that yeast attenuation assumes that the yeast does not stay mixed with the wort, but settles or rises.

Remember that all of these values are defined in terms of apparent attenuation.

Now the challenge is to use these values mathematically. Begin with the common equation for the original specific gravity:
Original Gravity Points, OGP = MaxPPG x W x EP / V;
where W is the weight of grain/extract, and
V is volume of wort.​

Next, separate OGP into simple, complex, and unfermentable, using both the Limits of Attenuation Percentage (LAP) and the Fermentable Percentage (FP).
Unfermentable OGP, UOGP=OGP x [1-(FPxLAP)],
this calculation would be performed for each sugar used,
then added together.​

And the fermentable gravity points are:
Fermentable OGP, FOGP = OGP – UOGP.​

Fermentable original gravity points should be a much better value to use when estimating final gravity points, than what is currently suggested, total original gravity points.
Fermentable FGP = 1 + [(FOGP/1000)*(1 - YA)].​

The above variables, FOGP and FFGP, should provide a better estimate of ABV:
ABV = use your favorite equation.​

And Final Gravity is now calculated with:
= 1 + [(FOGP/1000) * (1 - YA)] + (UOGP/1000).​

THOUGHTS?
 
The only other thing that I would add to my original post is that High ABV beers may need some gravity adjustment by a point or two since published attenuation values seem based on lower ABV values.
 
Gavagai, as I noted in the OP, there is published information out there for what I call LAP (I saw your referenced Web page before posting the OP). However, that page does not demonstrate how to incorporate the mathematics of all the unfermentables.
 
It would be difficult to make a prediction that would be true for everyone due to the mash. Everyone mashes differently in a variety of vessels. There is a heat gradient created, even though there is a target temperature. This gradient affects the efficiency of enzyme activity in different areas of the mash, and therefore affects the ratio of sugars left over after mashing (monosaccharides, disaccharides, dextrins, oligo, etc..) which in turn affects fermentability. Their water pH also affects enzyme activity. I think that your equation would have to include an experimental number of some kind which is related to mash efficiency of an individual's system, and everyone would have to do a little experiment to find it out for their own system.
 
Gavagai, that post on beertech is also available in homebrewtalk at https://www.homebrewtalk.com/f128/testing-fermentability-crystal-malt-208361/?referrerid=0.

Based on his work, it seems reasonable to make some simple assumptions about fermentability based on Lovibond value. Also, I have complied these ideas from http://forum.northernbrewer.com/viewtopic.php?f=1&t=82885:

"...my recollection is that Pale Malt was roughly 80% fermentable (which is about what I get when mashed around 148-150F.) I believe toasted malts, like Munich were about 12% less fermentable and Crystal Malt was about 50% unfermentable, in their experiment, IIRC. There was a lot of variability between specific malts and Lovibonds, though, so I rounded and averaged a few numbers to get something I could carry around in my head."

and
"I'm looking through some notes and it looks to me like the values maybe be more like 10% unfermentables in base malts, maybe 15% in kilned malts and 20-30% in crystal. This comes closer to what I'm seeing in the way of FG in some all-malt beers."
 
dbsmith,
thanks, very good idea. Another option would be to decrease/increase the value of LAP based on the personal system. It is a bit of a guess anyway. :)
 
THOUGHTS?

If you are serious about this you should throw away the amateurs' PPG concept and work in the advanced amateurs'/professionals Extract. This is the percent by weight of everything dissolved in wort/beer which is dissolved in it. Thus a sucrose solution which has specific gravity of 1.040 contains 10% extract by weight.

One monitors efficiency in the brew house by comparing the weight of grain employed to the amount of extract obtained e.g. if you have 100 L of wort of specific gravity 1.040 (adjust by the density of water for precision) you have 104 kg of wort containing 10% extract by weight or 10.4 kg. If you mashed 15 kg of grain your efficiency is 100*10.4/15 %. The efficiencies published by the maltsters are all in terms of kg extract per kg grain and usually run aroun 80% for base malts, less for specialty, roast etc. The estimated extract in the mash tun is then a simple sum of terms each of which is the Congress mash extract (what the maltster reports) times the weight of the grain times the brew house efficiency.

You can then move on to use Ballings formula to estimate the amount of alcohol produced by calculating the amount of extract the yeast consume. Note that attenuation here refers to RDF = 'Real Degree of Fermentation' not the ADF (Apparent Degree of Fermentation) that home brewers usually talk about. RDF is the percentage of extract consumed - not the change in 'points'.

There is nothing mysterious about this - the details are in most brewing texts. Measuring TE (True Extract) is harder than AE (Apparent Extract) because the alcohol has to be removed from the beer before its SG is measured but that's not hard to do. There is also a little math but not something that isn't easily handled by a spreadsheet.
 
aj,
Would you agree with the following statement?
Use of Balling's "formula is fine for those who wish to go to the trouble to compute TE (whose real value lies in determining attenuation) which is only a small fraction of brewers. Others want a simpler, quicker route to determining alcoholic strength. This lies in Tabarie's Principle which states that the depression of specific gravity in beer to which ethanol is added is the same as the depression of water to which an equal amount of alcohol (on a w/w basis) has been added. Use of Tabarie's principle lets us calculate the true extract of a beer with apparent extract."
[http://en.wikipedia.org/wiki/Gravity(alcoholic_beverage)]
 
If you are serious about this you should throw away the amateurs' PPG concept and work in the advanced amateurs'/professionals Extract. This is the percent by weight of everything dissolved in wort/beer which is dissolved in it. Thus a sucrose solution which has specific gravity of 1.040 contains 10% extract by weight.

+1 to that. The "brewer's point" based approach assumes that there is a linear relationship between extract and gravity. That is not exactly true but works for most applications in home brewing.


Others want a simpler, quicker route to determining alcoholic strength. This lies in Tabarie's Principle which states that the depression of specific gravity in beer to which ethanol is added is the same as the depression of water to which an equal amount of alcohol (on a w/w basis) has been added.

I would think so. But maybe A.J has insight that suggests that this is not the case.

Kai
 
https://www.homebrewtalk.com/wiki/index.php/Understanding_Attenuation, because this post assumes a working understanding of specific gravity, original gravity, final gravity, PPG, etc., as well as knowledge of such definitions as apparent attenuation, actual attenuation, fermentable, among others.

This one might be more up to date: http://braukaiser.com/wiki/index.php?title=Understanding_Attenuation

It seems safe to say that attenuation is primarily influenced by these three operations: mashing, lautering, and fermenting; and these two substances: type of starches/sugar(s), type of yeast.

I don’t think lautering does much in changing the fermentability. If it does anything it would be in the noise of other factors.

To predict FG, this post implements the idea (suggested by others) of splitting of PPG into three parts: simple (fermentable) sugars, complex (but potentially fermentable) starches/sugars, and unfermentable (not fermentable under normal brewing conditions) starches/sugars.

There is fermentable and unfermentable extract. The types of sugars that can be fermented depend on the organism you are using. I don’t think that there is a pontentially fermentable group although some of the sugars are easier to ferment than others.

Once split, this post intends to quantitatively track these three values through the brewing process. To do this, let's define Fermentable Percentage (FP) as the fraction of PPG that is both fermentable (simple) and potentially fermentable (complex).

That sounds reasonable. In existing work this Fermentable Percentage is called fermentability, limit of attenuation or attenuation potential. For brewing yeast it depends only on the how the grains were mashed and adjuncts that were added. It is generally fixed through boiling.

Now perform a "PPG trace" through the brewing operations:

Don’t trace ppg. Trace extract percent, or even better, fermentable and unfermentable extract weight.

(1) Mash: In this operation, a "PPG trace" requires a measure for the enzymatic conversion of the complex starches/sugars to simple (fermentable) sugars, a conversion that varies with both temperature and grain mixture. This measure can be hard to find in published sources, but it is around. Palmer, for example, has a huge discussion on these topics in his book (pp. 144ff.) and even shows a figure in his book on "Apparent Attenuation Limit." For this post, let's define this variable as the Limits of
Attenuation Percentage (LAP). Note that this value is defined in terms of apparent attenuation, that is OG and FG.

This will be your biggest uncertainty for determining fermentability. I have done some work on this (http://braukaiser.com/wiki/index.ph...ity_and_efficiency_in_single_infusion_mashing) and there are a lot of factors that are difficult to quantify.

(3) Lauter: For this Operation, a "PPG trace" needs a measure for the efficiency of extracting sugar from grain. This measure is widely available, and is typically around 0.75 for all-grain and about 0.65 for partial mashes, but 100% for extracts. Let's call this variable the Extraction Percentage (EP), again defined in terms of apparent attenuation.

This has been called Lauter efficiency and you may want to take a look at this : http://braukaiser.com/wiki/index.php?title=Understanding_Efficiency

(4) Ferment: Here the yeast's ability to ferment is needed, commonly measured and known as attenuation, YA. It is important to remember that yeast attenuation assumes that the yeast does not stay mixed with the wort, but settles or rises.

The yeast’s effect of attenuation could be quantified by how close to the attenuation limit yeast will be able to get under given conditions for a given strain. There are many factors here and the composition of the fermentable extract is also one of these factors. Some yeasts are better than others when it comes to metabolizing maltotriose, for example.

THOUGHTS?

Your approach is sound and has been brought up before. The reason why there is no good implementation is not the difficulty of the math, which is simple, but the number of factors that affect attenuation. Many of them are known qualitatively but are difficult to be assessed quantitatively. And they also interact with each other which makes for a really complicated model.

Kai
 
I don’t think lautering does much in changing the fermentability. If it does anything it would be in the noise of other factors.
Agreed, and it does, as you state, effect the quantity of fermentables, hence it effects the math. :)

There is fermentable and unfermentable extract. The types of sugars that can be fermented depend on the organism you are using. I don’t think that there is a pontentially fermentable group although some of the sugars are easier to ferment than others.

The yeast’s effect of attenuation could be quantified by how close to the attenuation limit yeast will be able to get under given conditions for a given strain. There are many factors here and the composition of the fermentable extract is also one of these factors. Some yeasts are better than others when it comes to metabolizing maltotriose, for example.
Again agreed, I am introducing the idea of "potentially fermentable" as a quantitative variable, for lack of a better term (if you prefer another term, I'm open), to describe that set of fermentables that are unfermentable to a set of certain organisms (yeast). Consider, if we eliminate all of the known unfermentables from the wort, we are left with fermentables, a portion of which are unfermentable to "beer" yeast. However, they may become fermentable depending upon the interaction with certain enzymes, a process dependent upon grain bill and mash temperature. This can and has been quantified, and (as I understand it) does not vary much among the various "beer" yeasts.

For the grain bill, a separate calculation can be performed to determine diastatic power. No need to intermingle that already powerful calculation with this one. For mash temperature, I am suggesting using the temperature of the mash to set a maximum attenuation value. This has also been loosely quantified, as covered in the first few posts of this thread.

If there is a great deal of variation of attenuation among yeasts, I believe we will find that quantitatively in the "apparent attenuation" value, correct?

Don’t trace ppg. Trace extract percent, or even better, fermentable and unfermentable extract weight.

This will be your biggest uncertainty for determining fermentability. I have done some work on this (http://braukaiser.com/wiki/index.ph...ity_and_efficiency_in_single_infusion_mashing) and there are a lot of factors that are difficult to quantify.

This has been called Lauter efficiency and you may want to take a look at this : http://braukaiser.com/wiki/index.php?title=Understanding_Efficiency
Is that your wiki? Cool!

I would think so. But maybe A.J has insight that suggests that this is not the case.
Yes, agreed, what is it? So please allow me to clarify my question. Why use a complicated, but imprecise model to forecast a process where a less complicated, but equally imprecise model will work perfectly fine?

This one might be more up to date: http://braukaiser.com/wiki/index.php?title=Understanding_Attenuation
Your approach is sound and has been brought up before. The reason why there is no good implementation is not the difficulty of the math, which is simple, but the number of factors that affect attenuation. Many of them are known qualitatively but are difficult to be assessed quantitatively. And they also interact with each other which makes for a really complicated model.

Kai
Thank-you. This math is simple to me as well, but I try not to underestimate the difficulty of math for others. :)

With an engineering background, making quantitative processes for qualitative procedures is what I do. Predicting attenuation reminds me a lot of predicting concrete strength. Both are complicated processes, both are not purely scientific, both require some trial-and-error. And, we make both beer and concrete every day. So, for now, I am optimistic.

Thanks for your input!!! :)
 
However, they may become fermentable depending upon the interaction with certain enzymes, a process dependent upon grain bill and mash temperature. This can and has been quantified, and (as I understand it) does not vary much among the various "beer" yeasts.

In most practical brewing no active enzymes are carried over from malt to pitched wort. The effects of enzymes brought in by the fermentor (organism that ferments the wort) are considered in the types of extract that is considered fermentable

For the grain bill, a separate calculation can be performed to determine diastatic power. No need to intermingle that already powerful calculation with this one. For mash temperature, I am suggesting using the temperature of the mash to set a maximum attenuation value. This has also been loosely quantified, as covered in the first few posts of this thread.

There is more than distatic power and temperature to wort fermentability although they are part of the main players. The other main player is pH. What you would need to do is to run a number of experiments that quantify the effect of them both independent from each other and then how they play together. You’ll then need to develop a model that captures their effect sufficiently precise for practical use.

If there is a great deal of variation of attenuation among yeasts, I believe we will find that quantitatively in the "apparent attenuation" value, correct?

The yeast factor may be even more difficult to model than the mash part given the dependency on the state of the pictched yeast which can vary widely from brewer to brewer. An initial approach could use expected attenuation to attenuation limit deltas for different yeasts.

Is that your wiki? Cool!

yes and thanks. You may want to look around there since it is targeted to more technical inclined brewers like you.

Yes, agreed, what is it? So please allow me to clarify my question. Why use a complicated, but imprecise model to forecast a process where a less complicated, but equally imprecise model will work perfectly fine?

The model only has to be as complicated as needed for the desired precision in a given boundary for parameters. The bounding box for the parameters has to be large enough for the model to be useful for many brewers. I.e. coming up with a model that predicts the wort fermentability when using only Weyermann Pilsner malt , a mash pH of 5.5, a mash thickness of 4 l/kg would not be all that useful.

I know that Beersmith attempts to predict fermentability from mash temperature but I haven’t seen the scientific work that this is based on. In many cases brewers trust predictions solely b/c a few others have said that they work for them.

Thank-you. This math is simple to me as well, but I try not to underestimate the difficulty of math for others. :)

If this is packaged into a tool or spreadsheet than gives the user easy to understand inputs and outputs that match the user’s expectation for a given input it doesn’t matter how complicated the math is. If someone wants to peer review your work they can always dig into that math.


Predicting attenuation reminds me a lot of predicting concrete strength. Both are complicated processes, both are not purely scientific, both require some trial-and-error. And, we make both beer and concrete every day. So, for now, I am optimistic.

good luck. The intend of my mash experiments was to find a way to predict attenuation but the large number of unexpected attenuation values throughout my practical brewing showed me that this is not an easy task.

EDIT:

If you plan to develop a model you should start taking good notes on your mashing:
- temp profiles
- malt’s used (including an indication if malts came from the same back or might be from a different batch)
- pH
- mill setting
- mash thickness
- etc.

Then make sure you do the Fast Ferment Test to isolate attenuation limit from the actual attenuation you got. This will separate mashing effects from yeast effects.

You want to have a long list of well documented batches to validate your model one you have one.

Kai
 
aj,
Would you agree with the following statement?
Use of Balling's "formula is fine for those who wish to go to the trouble to compute TE (whose real value lies in determining attenuation) which is only a small fraction of brewers. Others want a simpler, quicker route to determining alcoholic strength. This lies in Tabarie's Principle which states that the depression of specific gravity in beer to which ethanol is added is the same as the depression of water to which an equal amount of alcohol (on a w/w basis) has been added. Use of Tabarie's principle lets us calculate the true extract of a beer with apparent extract."
[http://en.wikipedia.org/wiki/Gravity(alcoholic_beverage)]


As I wrote it it's hardly likely I'd disagree with it.

But I will point out that when I checked Tabarie's principal out with sucrose solutions I did not find that it held very closely.
 
I would think so. But maybe A.J has insight that suggests that this is not the case.

The only concern is that, as I noted in my previous post, Tabarie's principal seems to be a fairly loose approximation.

The homebrewer's approach will depend very much on the home brewer. Those who are interested in the details of the science will chase the quarks while others are content to have enough control to insure repeatability and satisfy themselves that repeatability of measurements is enough to ensure consistently good beer.

It is quite possible for a homebrewer to measure specific gravity, true extract and alcohol content quite accurately with equipment no more sophisticated than simple distillation apparatus, a pycnometer (volumetric flask designed to insure repeatable fillings as opposed to accurate filling to a particular volume) and analytical balance. Few, except those aformentioned geeks would likely want to undertake the labor.

One aspect of this I always find interesting is that given original gravity and ADF or RDF one can estimate alcohol. But what is OG? Is it the gravity found in the kettle? Or in the fermenter after transfer? Or the latter adjusted for evaporation? On the other hand if you have alcohol content and TE you can estimate OG. This is the 'effective' original gravity. I find that effective OG and OG measured in the fermenter seldom agree closely (fraction of a Plato degree).
 
Wow, this has turned into a pretty serious discussion. Gregg Doss from Wyeast did some work on this recently:
http://www.homebrewersassociation.o...pdf/2012/1616-04 Attenuation - Gregg Doss.pdf


And I've been dabbling with it and come up with things that "get the job done" but would probably fail a closely scrutinized by a scientist.

Putting it all together:
http://woodlandbrew.blogspot.com/2012/12/final-gravity-in-recipe-formulation.html

Fermentability of crystal malt:
http://woodlandbrew.blogspot.com/2012/12/fermentability-of-crystal-malt.html

Mash temperature effects on fermentability:
http://woodlandbrew.blogspot.com/2013/01/measured-mash-temperature-effects.html

Batch sparging efficiency:
http://woodlandbrew.blogspot.com/2012/12/when-more-grain-doesnt-add-more-sugar.html
 

Latest posts

Back
Top