Topics Topics Help/Instructions Help Edit Profile Profile Member List Register  
Search Last 1 | 3 | 7 Days Search Search Tree View Tree View  

Visit The Brewery's sponsor!
Brews & Views Bulletin Board Service * Brews and Views Archive 2006 * Archive through June 11, 2006 * Predicting final gravity (the Holy Grail?) < Previous Next >

  Thread Last Poster Posts Pages Last Post
  ClosedClosed: New threads not accepted on this page        

Author Message
 

Michael Reed
New Member
Username: Notioncommotion

Post Number: 7
Registered: 04-2006
Posted From: 71.231.183.151
Posted on Tuesday, May 16, 2006 - 06:23 am:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

I have been trying to determine how to best predict the final gravity of my beer. There seems to be little on the topic, but then I read a couple posts on this site (Fredrick, you are the Man!). I don't need to be exact, but I would like to be close.

I am under the belief that the follow factors have the most significant effect to reduce attenuation (opposite is true to increase):
1. High percentage of non-fermentable sugars in wort based on wort ingredients and process (based on high mash temperatures, increased use of specialty malts, and DME and LME with high percentage of non-fermentable sugars).
2. High initial SG as the resultant increased alcohol production will limit yeast activity. It appears that only the SG attributed from fermentable sugars should be taken into effect as the non-fermentable sugars will not produce alcohol (and thus high gravity beers might be more alcohol limited and less governed by the creation of non-fermentable sugars).
3. Yeasts with low attenuation percentages (because they just don’t like to attenuate much). Using the values provided with the yeast alone appears to be useless.
4. Yeasts with high flocculate because the yeast settles before full conversion can occur.
5. Inadequate yeast starter, under nutrient wort, bad yeast, fermentation temperature, fermentor geometry, fermentation time, enzyme level in grain, mash tun geometry, mash pH, etc, probably have an effect, but I am hoping for arguments sake that they can be assumed negligible.

Do my first 4 items appear to be the most significant? Are there any I missed? Are there any relatively simple equations that will get me in the ballpark?

Thanks
 

Fredrik
Senior Member
Username: Fredrik

Post Number: 3146
Registered: 03-2003
Posted From: 62.20.8.114
Posted on Tuesday, May 16, 2006 - 09:50 am:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

Hello Michael,

I think you got the main factors!

FWIW here are some comments...

The starting point is always the wort, and more specifically the extract composition which determines the maximum potential fermentabiliy. If you don't know your starting point (extract composition) it makes litte or no sense to talk about the relative attenuation of yeast strains.

So once your wort is made, there is a maximum attenuation limit which no strain will violate. The exception to this rule is when you add enzymes, have infections, and also some odd strains like brettanomyces where some strains can to a certain extent further degrade the otherwise unfermentable part of the extract.

To predict the extract profile from the receipe and mash scheduele proves to be a challange! No doubt it can be done, but it is not easy since there are a multitude of factors. I am working on this myself and so far i have no finished reliable formula, but I rely on a synthesis of my own experience and theory.

Once your wort production is done and set, the next factor is how much of the potentiall fermentable extract your fermentation will be able to convert into ethanol.

Here there are many factors as well. Similarly with the mashing, it is not overly easy to predict this. Factors are the ones you mentioned... pitching rate, aeration, state of pitching yeast, stirring, temperature, stress factors like osmotic stress, alcohol etc, and nutritional levels.

At best, the attenuation numbers from the yeast manufactures is a measure of the "relative attenuation". But even the relative attenuation can be seriously tweaked, so you still have a significant impact on pitching parameters and fermentation parameters.

Given that you know your wort fermentability, there is a simple formula I use for guesstimates.

FG = (a*OG + b) ( RDFW - MTW*(1-MTD)) + (c*OG +d)
a = -1.180875689
b = 1.179398127
c = 0.993948859
d = 0.006096669

FG = specific gravity of the beer @ 20C, relative to water @ 20C.
OG = specific gravity of the wort @ 20C, relative to water @ 20C.
RDFW = The fraction of the wort extract by mass that consists of fermentable sugars.

MTW = the fraction of the wort extract by mass consisting of maltotriose.
If you don'w know this value, 10% is a good wild guess.

MTD = (MaltoTrioseDepletion) the fraction of the wort maltotriose that is fermented.

MTD = 100% means you have a DRY, completely attenuated beer.
MTD = 0% means you have a low attenuation, and a pretty high residual sweetness.
MTD < 0, means you are likely to have significant residuals of maltose as well,
which is probably rare, except for sluggish ferments of course.

RDFW is a function of your wort. IMO, knowing your wort, means knowing this value first of all.

Once you know this value, predicting FG is all about controlling yeast attenuation.

One problem when you know only OG and FG, is that while you know how much residual extract you have, you do not necessarily know how much fermentables that's left. But I think there are two ways. Fermentables impart far more sweetness than dextrins, also I think you might be able to guesstimate from the reducing sugar concentration.

A low reducing_sugar(RS)/residual_extract(RE) ratio, indicates a complete and dry ferment, a higher value probalby indicates a low attenuation with more residual sweetness.

If you want a dry ferment, you should aim for higher performance in the fermentation. More O2, more yeast, and even stirr during the finish.

If you want a sweeter finish, do not use too much much or too much oxygen. Avoid rousing.

/Fredrik
 

Bill Pierce
Moderator
Username: Billpierce

Post Number: 5409
Registered: 01-2002
Posted From: 24.57.224.243
Posted on Tuesday, May 16, 2006 - 10:22 am:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

"Given that you know your wort fermentability, there is a simple formula I use for guesstimates.

FG = (a*OG + b) ( RDFW - MTW*(1-MTD)) + (c*OG +d)
a = -1.180875689
b = 1.179398127
c = 0.993948859
d = 0.006096669

FG = specific gravity of the beer @ 20C, relative to water @ 20C.
OG = specific gravity of the wort @ 20C, relative to water @ 20C.
RDFW = The fraction of the wort extract by mass that consists of fermentable sugars.

MTW = the fraction of the wort extract by mass consisting of maltotriose.
If you don'w know this value, 10% is a good wild guess.

MTD = (MaltoTrioseDepletion) the fraction of the wort maltotriose that is fermented."



Fredrik, isn't this an oxymoron?
 

Fredrik
Senior Member
Username: Fredrik

Post Number: 3147
Registered: 03-2003
Posted From: 62.20.8.114
Posted on Tuesday, May 16, 2006 - 10:51 am:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

Maybe almost but not quite :-)

It relates a property of the wort, and a property of fermentation into a single number (FG), given that you know OG.

It is a relation. The relation itself is not an oxymoron, but it is true that you need to know something. If you don't know what is in your wort, you will consequently have no clue what will be your FG.

I meant to reply in a general sense. One could go deeper and analyse the yeast attenuation and discuss a mashing model but that is getting quite complex. I have guesstimates and some formulas I use myself, but they are at this point not clean enough to be presented.

Normal RDFW values for wort are 60-75%. The values are clearly related to the receipe, and mashing... but the local rules I have found will not make sense for someone using other grains, and other techniques. I think it takes a model that accounts for all the factors. Even, then we may lack data from the maltsters, so I think in the end the practical approach is a combination of experience (brewing trials) and then you extend that data with a good theory so that you need only a few trial points to get a decent predictive power. The better theory, the less data points would I expect.

I don't yet have such a model, but it will come in time, I am sure.

/Fredrik
 

Doug Pescatore
Senior Member
Username: Doug_p

Post Number: 1916
Registered: 10-2002
Posted From: 141.232.1.1
Posted on Tuesday, May 16, 2006 - 12:43 pm:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

I do it much the same way the guys at the Univ. of Denver predict hurricanes. I start with an educated guess, then I check my gravity after a couple of days of fermentation, re-adjust my prediction, check gravity a couple days later, re-adjust my prediction, and so on until fermentation is complete. By the time I am usually 100% accurate.

The point is, who cares? 1 point, 3 points, 7 points. If it tastes good and carbs up in the bottle (if you bottle) then it is good.

BTW, I rarely take gravity readings on my normal strength beers now a days.

-Doug
 

Graham Cox
Advanced Member
Username: T2driver

Post Number: 529
Registered: 11-2004
Posted From: 68.32.248.92
Posted on Tuesday, May 16, 2006 - 12:57 pm:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

What Doug said.

I admit I am a control freak, but once you do all you can to properly prepare the wort and the yeast and to provide them with a stable, appropriate temperature environment, you become a spectator. Why worry, so long as your FG is in the ballpark?

The average homebrewer don't have the sophisticated laboratory equipment necessary to measure wort composition, precise pitching rate, etc., so using a complex formula is not going to yield terribly accurate results. As far as that goes, the average homebrewer doesn't even have a terribly accurate instrument in his hydrometer or refractometer.
 

Fredrik
Senior Member
Username: Fredrik

Post Number: 3148
Registered: 03-2003
Posted From: 62.20.8.114
Posted on Tuesday, May 16, 2006 - 02:44 pm:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

I think wether one choose to care or not is a separate queston.

If you don't measure, don't want to know, and you get good beer and you consider that "end of story", that makes perfect sense to me, no doubt.

I just don't belong to that category. I believe in the opportunities of hidden details. Details that appear to make no sense at all, until you start to ask questions. Then and only then does something appear. The future does not lay itself before my feet. It is up to me, to create it. This is getting into the usual philosophical considerations again.

On the practical side, would you notice the difference between say FG 1.008 and 1.012 anyway?

I'd look at an example:
OG 1.060, RDFW = 72%

Ferment this completely dry, and you get FG=1.008, RE=4.4P, EtOH=6.9 vol%
Ferment this to FG=1.012 and you get RE=5.3P, EtOH = 6.3 vol%

So the residual extract increases by 20%, the alcohol drops by 10%.

Would you notice that? I am not sure. I doubt the 10% differencein alcohol. About the residucal extract along, maybe, maybe not.

But at the same time I'd estimate that the residual sweetness from the residual carbohydrates increases by 75%. I'd expect this to be noticable.

Apart from that I wouldn't rule out either that the last attenuation makes a difference to the yeast flavours and maturation.

All I am saying, is that you guys that doesn't care, hasn't convinced me to do the same :-) I still see potential.

If I were in the above mentioned category I would now ask, that even if there IS a slight difference in flavour? who cares, as long as it's good? What is the difference between two good beers anyway? ;-)

And I'd say the reason is the same as above. How do you motivate a question for someone else? It's hard. I ask my own questions, and I let everyone else ask theirs. Sometimes we ask similar questions and we can benefit from each other :-)

/Fredrik
 

Tim Wi
Intermediate Member
Username: Riverkeeper

Post Number: 419
Registered: 03-2005
Posted From: 170.141.68.99
Posted on Tuesday, May 16, 2006 - 05:40 pm:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

/Goldberg

 

Nathan Jordan
Junior Member
Username: Njordan

Post Number: 61
Registered: 06-2004
Posted From: 70.236.212.123
Posted on Tuesday, May 16, 2006 - 08:27 pm:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

My gut tells me fermentation temperature has a significant effect on attenuation. I didn't look it up in a book, though, I looked it up in my gut.

Plus I've heard that many Belgian brewers steadily raise fermentation temperature, often from the mid-60s to the mid-80s, specifically to increase attenuation.
 

Dan Listermann
Senior Member
Username: Listermann

Post Number: 2828
Registered: 03-2004
Posted From: 216.23.59.245
Posted on Tuesday, May 16, 2006 - 09:07 pm:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

Nathan may be on to something here. It is my theory that attenuation is related to floculation with less floculent yeasts attenuating more than yeasts that hit the bottom quickly. It is a matter of "air time." A higher fermentation temperature would be more vigorous and provide for greater "air time," but then again some seem to enjoy calling me, repeatedly, an idiot so you can take it or leave it.

Dan

--This space is again being left intentionally blank.-


 

Michael Reed
New Member
Username: Notioncommotion

Post Number: 8
Registered: 04-2006
Posted From: 208.195.70.39
Posted on Wednesday, May 17, 2006 - 03:18 am:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

Thank you all for your feedback.

Maybe this is way to simplistic, but It would seem to me that yeast would convert everything with any possibility of being fermented given an infinite duration of time. They would convert all types of fermentable sugars simultaneously, however, would be more likely to convert an easy sugar at any given moment analogous to quantum mechanics (reference Heisenbeer uncertainty principle – "one cannot exactly measure yeast attenuation while simultaneously enjoying a cold beer"). Yeast does not, however, have an infinite duration of time because they begin to shut down when they "sense" the environment is becoming inhospitable either due to limited easily convertible sugars or high alcohol levels. Based on the flocculation of the given yeast strain, they might hang around for a while longer and do their thing a little bit more. Correct me if I am wrong, but it appears your (Fredrick's) hypothesis is that maltotriose is considered to be more difficult to ferment, and therefore happens last and is the limiting factor. Does this conflict with my hypothesis that all sugars are converted simultaneously?

On a more applicable note, I noticed that your formula doesn't include the typically listed yeast attenuation percentage or low/medium/high flocculation rating. Are these factored in when you assign a MTD value? Also, I wasn't sure if alcohol value used to limit attenuation. Is this taken into account in the constants?

Lastly, can I can make a very loose FG estimate by just multiplying the RTFW by the high side of the yeast's published apparent attenuation, while also limiting the minimum based on a maximum alcohol tolerance of the given yeast? Is RTFW calculated as the anticipated sum of all the fermentables in each ingredient? Would this be equal to the associated points of each ingredient multiplied by the expected % fermentability (listed by manufacturer as around 80% for DME & LME, maybe 0% for spatiality malts, and somewhere between 70% and 85% based on my mash process for all grain mash)?

Thanks again. Michael

(Message edited by NotionCommotion on May 17, 2006)
 

Fredrik
Senior Member
Username: Fredrik

Post Number: 3150
Registered: 03-2003
Posted From: 213.114.44.200
Posted on Wednesday, May 17, 2006 - 05:52 am:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

I'm no expert but to give my point of view as I currently understand it...

I have read many papers on this, and each sugar has it's own utilisation channel so to speak, and they tend to be consumed in a "loose sequence", however the story is the same as with amino acid sequencing, the sequences are not strict. They overlap, morever, different strains has in fact been reported to consumes the different sugars with different rates. Sometimes the sequenece order is even different for different strains.

So the sequencing is a concept. Reality is always a bit fuzzy, but it's a useful way of thinking of it.

Typically(granted there may exist execeptions):

glucose, fructose and sucrose are the first group of sugars to be consumed.

When this group is consumed the maltose and maltotriose uptake is very low, perhaps not strictly zero, but LOW.

So when the level group of easy sugars start to fall maltose and maltotriose systems kick in at a higher speed.

Here there is also some slight differences between strains. Some strains seems to utilize maltose and maltotriose almost in paralell, but I think it is more common that even though tha maltotriose uptake starts in parallell, in some strains it sometimes runs on a low rate until the maltose level starts to drop.

There are lots of papers discussing the transporters of maltose and maltotriose in detail. There seems to be both the case specific maltotriose transporters, and a limited cross transport throuhg the maltose transporters.

Anyway I think the simplified concensus is that, the yeast treat the various wort sugars differently, and usually glucose, fructose and sucrose tend to get consumed first, then comes maltose and maltotriose. And more often than not, most of the residual fermentables left over is maltotriose, and SOMETIMES small amount of maltose, and sometimes possibly even traces of other sugars. This is not something I totally made up, it is obviously based on my current understanding but it's backed up by lots of brewing science papers I have read on this topic - papers written by people who have access to lab equipment.

A typical breakdown of extract may look like this

% of extract (RDFW=68%)
45.9 maltose
11.1 maltotriose
7.7 glucose
1.4 fructose
1.8 sucrose
32.2 unfermentable extract, dextrins and proteins

If you look at the fermentables, maltotriose is the second most abundand sugar after maltose. This is why maltotriose residuals can have a significant effect on FG and residual sweetness. Maltotriose appears to have the sweetness 4 times that of the average unfermentables/dextrin.

I too think temperature and flocculation are quite likely factors. I would absolute add to the list, pitching rate, aeration and stirring.

I have specifically read one paper which did a real test where the measured maltotriose depeltion and factors increasing attanuetion were:

(+) picthing rate, aeration, increased temperature, stirring

negative factors were found to be
(-) extremely low pH (much below 4), high osmotic pressure, high ethanol concentrations

The sensitivty of osmotic pressure and ethanol are cleraly correlated to pitching rate and aeration. So it is completely consisteny that these factors appear in different groups.

If anyone is interested to get some papers here are a couple... but there are many out there, seek and you will find.

"Factors Influencing Maltotriose Utilization
During Brewery Wort Fermentations"
[American Society of Brewing Chemists]
-- http://www.asbcnet.org/journal/pdfs/backissues/52-0041.pdf

"Maltotriose Utilization by Industrial Saccharomyces Strains: Characterization of a New Member of the alpha-Glucoside" Transporter Family"
[Applied and Environmental Microbiology]
-- http://www.pubmedcentral.gov/picrender.fcgi?artid=1214619&blobtype=pdf

"Molecular Analysis of Maltotriose Transport and Utilization by Saccharomyces cerevisiae"
[Applied and Environmental Microbiology]
-- http://aem.asm.org/cgi/reprint/68/11/5326.pdf

I do not think it is hard to get complete attenuation as such, I think the problem is to get the right attenuation. I am not so sure that fermentation to complete dryness is desirable at all time. Some residual sweetness may probably make the beer alot smoother. That's my theory at least.

I have a feeling that many of my last beers has been fermented to dryness, due to very complete and performant fermentations.

With my new tool of the blood sugar meter I may now have the tool I need to actually get some loose mesurement of this.

/Fredrik
 

Fredrik
Senior Member
Username: Fredrik

Post Number: 3151
Registered: 03-2003
Posted From: 213.114.44.200
Posted on Wednesday, May 17, 2006 - 06:02 am:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

> On a more applicable note, I noticed that your > formula doesn't include the typically listed
> yeast attenuation percentage or low/medium/high > flocculation rating. Are these factored in when > you assign a MTD value?

Yes, exactly. It's in the MTD value. Of course the MTD not only depends on the relative yeast attenuation, it also depends on pitching rates aeration etc (all factors discussed above!).

But to summarize. MTD = 100% means complete fermentation to dryness.

MTD = 0% means a very low attenuation - wether this in a particular case, is mostly due to the strain, or due to underpitching, underaeration or osmotic or alcohol stress is obviouls impossible to give a generic answer to. All factors count.

> Also, I wasn't sure if alcohol value used to
> limit attenuation. Is this taken into account
> in the constants?

No, this is not specifically accounted for in the formulas, but it's in the MTD! Extreme high alcohol, would probably lower the MTD value.

> Lastly, can I can make a very loose FG estimate > by just multiplying the RTFW by the high side
> of the yeast's published apparent attenuation,

I am not sure what you mean? RDFW is NOT the same as the AA values listed my yesat manufacturers. RDFW is a property of the wrot ONLY. This is the point with the formula. It separates the effect of wort and yeast.

> while also limiting the minimum based on a
> maximum alcohol tolerance of the given yeast?

If you reach the alcohol tresholds the MTD will go down. Then you need more yeast, more oxygen and maybe more fan, and maybe a sequential feeding.

> Is RTFW calculated as the anticipated sum of
> all the fermentables in each ingredient?

Yes, that's correct. In my recepie sheet, I use a sum of grainbill strategy, similar to what you do with extract. But it contains lots of guesstimates of course.

...more to come later... need to go :-)

/Fredrik
 

Fredrik
Senior Member
Username: Fredrik

Post Number: 3152
Registered: 03-2003
Posted From: 62.20.8.114
Posted on Wednesday, May 17, 2006 - 08:36 am:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

Here is the principle how I sum over extract contributors (the numbers are obviously just an example)

pi = the % of the total extract by mass, made up by component i

(ie p1 + p2 + ... = 1 )

RDFWi = the potential fermentability by mass, of the extract contributed by component i

RDFW = p1*RDFW1 + p2*RDFW2 + ...

An example.

Receipe

2 kg muntons light DME -> gives ~ 1.9 kg dry extract @ 65% RDFW
0.2 kg sucrose -> gives 0.2 kg dry extract @ 100% RDFW
0.5 kg crystal malt -> gives perhaps 0.25kg dry extract @ 25% RDFW

totalt extract = 1.9 + 0.2 + 0.25 = 2.35 kg

This means

p_DME = 1.9/2.35 = 80.9%
p_sucrose = 0.2/2.35 = 8.5%
p_crystal = 10.6%

So RDFW = 80.9% * 65% + 8.5% * 100% + 10.6% * 25% = 63.7%

This means that the major factor of uncertainty in this part is knowing your ingredients and your methods. If you mash, you need to estimate the resulting fermentabiltiy of your mash. If you can do this, you need only to control your fermentation and you should be dead on your target FG, or maybe +/- 1 point.


The RDFW values you pull out of experience and/or data you can find.

My experience tells med muntson light DME is usuall around 65% (corresponding to 80% apparent)

Crystal and roasted malts I usually assing 20-30% - this is consistent with data I have seen.

Base maltes obviously depends on brand and mashing, but usually in the 65-75% range.

Since the basemalt is the largest % of the bill, this is the large sensitivity.

This is why I am working on the mashing model, and this is what I am going to use the blood sugar meter for!

But even lacking the mashing model, these calculations offer some structure to the madness. Enabling you to do guesstimates based in incomplete data, and you can also to some limited back-calculations, estimating the mash output from fermentation output and so on.

Once you know the basic numbers for the basic components for a given scheduele, one should be able to write down and arbitrary receipe from scratch and make a pretty good estimate of the RDFW.

It is also useful for tweaking existing and proven receipes in a predictable way. you can use it to calculate close to exactly how much sugar, or exactly how much maltodextrin you need to add to change the FG in a previously proven and tested procedure.

The number one factor here, is to predict the mash.
Number two is to predict the yeast performance.

/Fredrik
 

Fredrik
Senior Member
Username: Fredrik

Post Number: 3153
Registered: 03-2003
Posted From: 62.20.8.114
Posted on Wednesday, May 17, 2006 - 10:56 am:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

When talking about FG, this is an illustration of the sweetness model I am currently using. Maillard stuff comes notop of this.


This models sweetness as fermentation proceeds, for a typical 1.060 / 70% RDFW wort.

The sweetness treshold I have measured myself, so it's my own personal treshold in room temp tapwater.

The sweetness value is calculated from wort composition and expected specific sweetness.

The point at 1.016 is approximately the point where the estimation says all fermentables, except maltotiose is depleted. (in reality there may be an overlap)

/Fredrik
 

Mike Reynolds
New Member
Username: Biernewbie

Post Number: 12
Registered: 04-2006
Posted From: 12.203.216.69
Posted on Thursday, May 18, 2006 - 08:52 pm:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

Wow, Fredrick relax and have a beer now!~
 

Michael Reed
New Member
Username: Notioncommotion

Post Number: 9
Registered: 04-2006
Posted From: 71.231.183.151
Posted on Friday, May 19, 2006 - 05:55 am:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

Fredrik,

I played around with your formula for a while. Pretty cool! Below is my understanding along with a couble of questions. If you disagree with something that I appear to state as a fact, please let me know your views.

I agree with your statement about starting with knowing your wort first. I have seen percent fermentable values on various LME and DME listed around 75% to 80%, and I take it that this should be used as the RDFW, but converted from apparent to real. You stated that specialty malts range around 20-30% real (do you have any data for various types of specialty malts, or do you just split the middle with 25%?). I would expect sucrose would be 100% real. Would dextrine be 0% real? Do you have any other values for other adjuncts? You stated that mashed base malts typically range around 65% to 75% based on brand and mash schedule. Brand is going to be tough (and maybe it is seasonal?), but do you have any benchmark data based on mash temperature?

I am a little concerned about the difficulty to determine MTW. Maybe one can find the numbers with extracts, but what about grain? You give a “wild guess” of 10%, but a going from 10% to 20% gives an 18% in FG.

Lastly, do you have any benchmarks or models in predicting MTD? If the yeast manufacturer listed one yeast as 75% and another yeast at 80%, how would this number change? How are other conditions incorporated in this number?

Thank you, Michael
 

Fredrik
Senior Member
Username: Fredrik

Post Number: 3157
Registered: 03-2003
Posted From: 62.20.8.114
Posted on Friday, May 19, 2006 - 09:32 am:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

> I am a little concerned about the difficulty to determine MTW.
> Maybe one can find the numbers with extracts, but what about grain?
> You give a “wild guess” of 10%, but a going from 10% to 20% gives
> an 18% in FG.

The wort from extract and AG is not that different. Good extract is nothing but AG wort, except that it's spray dried or canned.

It is true that any lack of knowledge implies further error sources. I don't know for sure, but I have never seen any data on normal wort having as high maltotriose level as 20%. So ther is a limit to this error. I think you just don't get a normal AG wort with 30% maltitriose, unless you have barley grown on another planet.

I'd suspec that normal AG wort typically have say 9-14% of extract. Though I am currently working on a correlation, so see how the % maltotriose can be predicted. It's true though there is an error here.

There is a variation but most data I have seen in papers indicaes that normally the % maltotriose of fermentables is 16-17% of fermentables (not extract).

This is probably a decent guess. But if you use odd ingredients or odd scheudles, the variation might be a bit larger.

> Lastly, do you have any benchmarks or models in predicting MTD? If
> the yeast manufacturer listed one yeast as 75% and another yeast at
> 80%, how would this number change? How are other conditions
> incorporated in this number?

This has at least as much to do with the fermentation and yeast control as it has with the strain. In a very crude estimate, you will see that the yeast numbers tend to range from about 67% to 80%.

Relatively speaking I'd interpret
80% as a high attenuator MTD ~ 100%
67% as a low attenuator ~ MTD 0%

But I have noticed from my own trials, that you can easily coach a medium attenuator to attenuate high! So things are again more complex. It's just that a high attenuator are in the wild case, more likely to end up with a higher MTD value than a low attenuator. But the exact value depends on your fermentation! pitching rate, temps, aeration everything.

This is why it takes a high degree of control of everything to predict MTD in a real batch. I am still working on this too. I hope to get more answers to come in the future.

The problem is that MTD is hard to measure, even knowing OG and FG. Low MTD would probably taste a little sweeter, high MTD would taste dry(non sweet).

But if you add to OG and FG, the residual reducing sugar RRS, one could guesstimate MTD a little better. I am currently working on this and I have a ballpark formula f1,f2 to give you

MTD = fq(OG,FG,RRS)
RDFW = fq(OG,FG,RRS)

Using only OG and FG, you can calculate RDF, but you can NOT fully resolve RDFW and MTD, but you can find some limits. You can also do guesswork based no your yeast performance.

Btw, the relation here is RDFW = RDF + (1-MTD)*MTW

OG and FG determines the relation, but can not resolve the two variables.

But adding another dimenssion (RRS), you can actually roughly speaking resolve RDFW and MTD! Which is damn neat.

I am currently working on calibrating this formula and I have currently measured on a few beers that I'm trying to fit. There are some measurement errors, but this looks promising. It's not perfect, but MUCH better than nothing.

I will post more on this as I make progress.

/Fredrik
 

Fredrik
Senior Member
Username: Fredrik

Post Number: 3158
Registered: 03-2003
Posted From: 62.20.8.114
Posted on Friday, May 19, 2006 - 09:36 am:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

> I agree with your statement about starting with knowing your wort
> first. I have seen percent fermentable values on various LME and
> DME listed around 75% to 80%, and I take it that this should be
> used as the RDFW, but converted from apparent to real.

Yes that's kind of correct, however I would not put full trust in the manufacturers loose data. I alot of this data is pretty lame, and is only approximate values. It is a good starting point, but for best results I would adjust your assumed value after a trial, or after the first time using it if necessary. I tried to get the specs from muntons for their dme but it was very lame. I foudn that the best data i have is from my own trials using it.

> You stated that specialty malts range around 20-30% real (do you have any
data for various types of specialty malts, or do you just split
the middle with 25%?).

All malts are different, so without malster analysis (which you do not typically have.), you can either use "typical" data, or make your own trials, or adjust your data as our experience tells you to.

I have seen reports on real measurements of speciality malts and these are the ranges I've seen. But this is complexed with the mash schedyuele, so it's still hard to nail.

> I would expect sucrose would be 100% real.

Correct.

> Would dextrine be 0% real?

Yes, the _dextrins_ themselvs are "0% real" by definition,
however "dextrin malt" or "maltodextrin powder" is not 0% real,
beceause even dextrin malt usually contain small amount of fermentables.

It would vary from brand to brand, but I'd guess these varieties are 10-20%. However things are complexed by that fact that dextrins may be further degraded during the mash. So when you go AG, it is not a straight sum game. You have to estimate, based on whatever you have at hand.

> Do you have any other values for
> other adjuncts? You stated that mashed base malts typically range
> around 65% to 75% based on brand and mash schedule. Brand is going
> to be tough (and maybe it is seasonal?), but do you have any
> benchmark data based on mash temperature?

I am going to make more testing on this in the future. I obviously have ballpark data from my previous mashes and temperature alone is insufficient. You need to know the rest time, thickness and alot of things.

There is alot of work to be done here. I don't have the answer you are looking for, but I think we agree on the question :-)

/Fredrik
 

Fredrik
Senior Member
Username: Fredrik

Post Number: 3160
Registered: 03-2003
Posted From: 62.20.8.114
Posted on Friday, May 19, 2006 - 01:47 pm:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

I have tested one of the own beer, and some commercial beers. On the commercial beers, I have based on the nominal carb.contents and alcohol value, backtracked OG and FG. On my own beer these are measured/known.

Then I used my new formula and from OG,FG,RRS calculated the estimated MTD value.

The results of calculated MTD and residual sweetness are

MTD RSW
Beer1: 101% 24%
Beer2: -8% 68%
Beer3: 2% 45%
Beer4: 58% 61%
Beer5: 102% 15%

Now, the interesting part is that Beer 1 (my own beer) and beer5 (commercial cheap german low alcohol lager) are preceived as the 2 least sweet beers (fairly dry), indicating a complete and dry ferment - completely consistent with theory.

Beer 2 and 4, are the only two beers preceived as having having a clear sweetness.

Beer 3 is in fact inbetween. Not obviously sweet, but not overly DRY either.

The calculations from the model seems to make perfect sense and it correlates with my perception of residual sweetness.

Btw, Beer#4 is leffe blond, which definitely has a sweetness to it.

The reason MTD and RSW doesn't correlate is because the beers are of different gravity. A higher gravity beer with the same MTD rating will accordingly be sweeter.

/Fredrik
 

Fredrik
Senior Member
Username: Fredrik

Post Number: 3165
Registered: 03-2003
Posted From: 213.114.44.200
Posted on Monday, May 22, 2006 - 06:07 am:   Edit Post Delete Post View Post/Check IP    Ban Poster IP (Moderator/Admin only)

I am currently trying to revise the few malting/sugar profile analysis I've found.

Perhaps the AA/BA ratio may be a parameter that can be guesstimated and allow for limited corrections. This in turn is a function on choice of grain as well perhaps as % adjuncts. Of course this is generall unknown, but some typical differences can still likely be made, comparing various base grains. And typical data is better than no data.

High AA/BA seems to favour higher maltotriose and glucose levels (and thus lower maltose levels).

The problem has been that the data I have seen is very imcomplete and the small series reported in various papers sometimes conflict with each other. I'm currently trying to see if there is a way to interpret them consistently. I think the AA/BA is a key.

I am sitll equally convinced that the proper prediction is impossible without a proper mashing model, but what I hope to find at this point with less effort is a litter better maltotriose prediction.

/Fredrik