Cantillon Analytics

[Thorpe]

Absolutely! Existing literature shows a fair bit of ABV variation between barrels of spontaneously fermented beer. What's strange is that these two beers have the same OG, the same FG, but a different ABV. Most people treat this is a simple three-variable system: if you know two of the variables {OG, FG, ABV}, then the third one is fixed. Obviously reality can be more complicated, but for most fermentations the three-variable simplification works really well. Given that alcohol makes up a large mole fraction of the finished product and has a markedly different density than water, SourBrewer's results indicate that there must be some other fermentation product whose concentration varies considerably between the two beers and has a markedly different density as well. In essence, a hidden fourth variable.

Couldn't the answer be that some of the organisms that get into the wort are converting sugar into something other than CO2 and alcohol?

 

[olympuszymurgus]

In the Cantillon testing OG is estimated from FG & ABV. In Olympuszymurgus's case it is a known OG and FG with different ABV, but that could be influenced by the barrels from things like residual grape sugars and evaporation. Would be a better experiment to ferment a batch in stainless to reduce variables.

Lets not forget residual alcohol from the wine soaked into the barrel. A big Syrah can top 15% abv, a gallon of syrah coming out of the oak could add .3% abv, and I bet a barrel could hold far more than a gallon of wine in the oak.

 

[duketheredeemer]

Couldn't the answer be that some of the organisms that get into the wort are converting sugar into something other than CO2 and alcohol?

Yes, exactly what I'm wondering. Or contributions from the barrel , as per what olympuszymurgus and SourBrewer said, or changes in volume due to evaporation as SourBrewer noted (particularly if the rate of evaporation for water and ethanol were different, as they are wont to be).

 

[duketheredeemer]

In the Cantillon testing OG is estimated from FG & ABV. In Olympuszymurgus's case it is a known OG and FG with different ABV, but that could be influenced by the barrels from things like residual grape sugars and evaporation. Would be a better experiment to ferment a batch in stainless to reduce variables.

Definitely. I'd be surprised if there were that much sugar or alcohol left in the barrel, but I could be wrong. In any event, those would all qualify as a fourth variable.

 

[Pneill3161]

Something tells me you guys are close to the answer. I'm gonna reread this as I think it applies to this thread and might hold the answer;

 

[drgarage1]

Something tells me you guys are close to the answer. I'm gonna reread this as I think it applies to this thread and might hold the answer;

That's a good book!

 

[Gestalten]

Great thread!
I'm not surprised by St. Lam. The grape sours always seem to be funkier and more acidic. I think it's that they sugars present in those fruits are so easily metabolized by brett and bacteria.
As to the bitterness, I'm also not really surprised. Despite what you may hear, aged hops are pretty intensely bitter.
Lambic wort is generally bitter as hell. Of course, that bitterness fades substantially even after 6 months (though I could still taste aged hop in Cantillon's 13 month old Lambic). I also would assume overall bitterness varies substantially from bottling to bottling, depending on it's balance with the other flavors.

The whole alpha:beta acid thing is pretty interesting, especially when you add age into the mix...

Beta acids

These compounds are not actually bitter, but will turn into bitter compounds when they oxidize during storage. The alpha:beta ratio is considered important in gauging how a hop will provide bitterness to beer as the hops age. The bittering potential from alpha-acids reduces with time but the bittering potential from oxidized beta-acids increases. In a hop with a 2:1 ratio of alpha:beta the bittering potential may remain fairly constant. The oxidation reaction will take place to an even greater extent during kettle boiling., and again the chemistry will be discussed during the wort boiling article. Beta-acids consist of lupulone, colupulone, adlupulone and other substances, and like alpha-acids differ in the structure of the side chains. Again there is a difference of opinion in the brewing world as to the character of bitterness derived from beta-acids compared to that of alpha-acids. In Germany oxidized beta-acid bitterness is preferred while in Japan it is considered too "harsh."

http://byo.com/grains/item/848-hop-chemistry-homebrew-science

The shocker to me, was the low attenuation overall. That's really odd considering how dry most of these taste. I'd say it's probably the high hopping slowing down the metabolism of the wild yeast/lab, possibly combined with pretty low cellar temps.

 

[thefleshlight]

Interesting. I recently received some data on NMR analysis of a slew of barrels. Same wort added to each barrel, all around the same FG. ABV varied from 5.8-7.4% ABV.

OP, shoot me a PM. I'm in socal and doing some of this science with help from our local uni.

NMR will not be able to quantify ABV accurately whatsoever. It has an inherent 5% variance in the first place with the most pure of samples. Your best bet is either GC or HPLC as is accurate into the hundreths of a percent at the worst case, ten thousanths of a percent in best case.

 

[olympuszymurgus]

NMR will not be able to quantify ABV accurately whatsoever. It has an inherent 5% variance in the first place with the most pure of samples. Your best bet is either GC or HPLC as is accurate into the hundreths of a percent at the worst case, ten thousanths of a percent in best case.

I will humbly disagree.

If we were talking the NMRs of a decade ago, you would be more than correct. New technology has allowed NMR to be a very effective and comprehensive method of quantification, allowing very precise and accurate data. There are more than a few examples in the literature of NMR quantification being used for beer to measure much less abundant analytes than alcohol.

 

[thefleshlight]

I will humbly disagree.

If we were talking the NMRs of a decade ago, you would be more than correct. New technology has allowed NMR to be a very effective and comprehensive method of quantification, allowing very precise and accurate data. There are more than a few examples in the literature of NMR quantification being used for beer to measure much less abundant analytes than alcohol.

That's if whoever is doing the analysis has that hardware, software, and are running high-field instruments, most of which cost far more than $1M. But I seriously doubt anyone would use an NMR. While it being far more expensive, it is in fact much faster, but it's far cheaper to do a ~15min GC/GCMS/HPLC run than a ~2min NMR run. I use a 500MHz daily and cannot be taken alone for over 95% purity. GCs, MS, and HPLC are far more accurate, even with the a 1GHz NMR.

 

[Beerontwowheels]

NMR will not be able to quantify ABV accurately whatsoever. It has an inherent 5% variance in the first place with the most pure of samples. Your best bet is either GC or HPLC as is accurate into the hundreths of a percent at the worst case, ten thousanths of a percent in best case.

I will humbly disagree.

If we were talking the NMRs of a decade ago, you would be more than correct. New technology has allowed NMR to be a very effective and comprehensive method of quantification, allowing very precise and accurate data. There are more than a few examples in the literature of NMR quantification being used for beer to measure much less abundant analytes than alcohol.

That's if whoever is doing the analysis has that hardware, software, and are running high-field instruments, most of which cost far more than $1M. But I seriously doubt anyone would use an NMR. While it being far more expensive, it is in fact much faster, but it's far cheaper to do a ~15min GC/GCMS/HPLC run than a ~2min NMR run. I use a 500MHz daily and cannot be taken alone for over 95% purity. GCs, MS, and HPLC are far more accurate, even with the a 1GHz NMR.

Just kidding, of course!

 

[olympuszymurgus]

Just kidding, of course!

In a bit we'll be pulling out our pipettes and measuring them.

 

[SourBrewer]

Here are the final numbers from the Bruery lab:

 

[SeaWatchman]

Here are the final numbers from the Bruery lab:

Purple drank get u drunnnnnk

 

[Rowingbrewer1]

Wait so cantillon lied to me when they said all their beers were 5%?

 

[SourBrewer]

Wait so cantillon lied to me when they said all their beers were 5%?

I was expecting the abv to be higher on all of them, but they are not super attenuated. I need to add these numbers to the spreadsheet:

Apparent Extract (Ea):
CG-2.59P (1.010SG)
SL-1.83P (1.007SG)
GC-2.90P (1.011SG)

Real Extract (Er):
CG-4.59P (1.018SG)
SL-4.32P (1.017SG)
GC-4.82P (1.019SG)

Original Extract:
CG-12.99P (1.053SG)
SL-14.81P (1.060SG)
GC-12.89P (1.053SG)

 

[SourBrewer]

Improved version:

 

[drgarage1]

Improved version:

I love this work soooooo much. Can't wait to see follow-ups (other brewers, different vintages, etc.) on this fantastic project.

 

[olympuszymurgus]

 

[PDXAmbassador1]

Great thread!
I'm not surprised by St. Lam. The grape sours always seem to be funkier and more acidic. I think it's that they sugars present in those fruits are so easily metabolized by brett and bacteria.
As to the bitterness, I'm also not really surprised. Despite what you may hear, aged hops are pretty intensely bitter.
Lambic wort is generally bitter as hell.

I'll throw in my two cents:

(1) lambic wort is bitter largely from oxidized Beta Acids, not from the isomerized Alpha Acids which are what IBU tests measure right? I've experimented with this and found that if you dry hop with aged hops you can get a super bitter result - without even boiling, since oxidized beta acids are soluble in beer at room temp.

(2) As I understand it simple sugars (as in grape must) would result in less funk rather than more complex ones (like the long chain sugars and dextrines made by turbid mashing in lambic production). The metabolic pathway for turning simple sugars into alcohol is quick and produces no funky byproducts. But when microorganism shave to break down long chain carbs it is the stress and effort that creates funk.

(3) Heterofermenting bacteria create negligible amounts of alcohol compared to any yeasts present (0.3% ABV contribution according to this milk the funk article). Reports of 100% lacto fermentation are now believed to be due to lacto cultures contaminated with yeast cells. Anyway, I'm pretty sure as other have said, that it's the residual alcohol present fresh wine barrels Jean Van Roy uses for St Lam that contribute to the higher ABV in that beer (while the Gueuze and Bruocsella he brews use 2nd-use or 100th-use barrels which wouldn't change the ABV at all since whatever is soaked into the walls of the barrel has the same ABV of the new beer).

(4) as for why lambic has high residual gravity...my theory relates to the fact that enteric bacteria present in the initial spontaneous fermentation go through an explosive growth stage and dominate the fermentation during its early stages. The result is that they gobble up much of the essential nutrients from the wort - before the sacchromyces and brettanomyces have a chance to get going - so by the time these yeasts start growing they are in a seriously nutrient-depleted environment (I've read this somewhere but don't remember where). Anyway maybe that is what stops yeast from taking it down all the way to FG 0.000 - they don't have the nutrients they would use to produce the metabolic reactions needed to break down the longest chain carbohydrates, so they poop out early.

 

[Gestalten]

1) Yes, that is correct as far as I know. I hadn't thought about the bitterness being more soluble. That explains some things!

2) Pure speculation but I always get this funk that to me says grape skin. Could be anything but I wonder if the skins do play a role in that specific taste.

3) I've never heard of St. Lam using fresh barrels. The site isn't specific, so I'd think not. Jean is usually pretty specific about that sort of thing.

4) That's an interesting theory. I guess it could be. That's an aspect of lambic production that is often overlooked. I've always thought it was probably way more important that people seem to think. I know there are flavor precursors developed early on by enterobacter.

 

[quirkzoo1]

(total noob here but)

1 - That is cool, I didn't know about the solubility piece..

2 - My understanding is that what we call "funk" has less to do with the sugars that are fermented and more to do with "other" products (Phenols, Esters...) being broken down and metabolized by Brettanomyces and turned into other things. The best example of this would be wine that is infected with Brett. That wine will typically have zero residual sugar and yet still develops the characteristic brett funk. I would guess that the funk in St. Lam definitely has to do with the introduction of different compounds from the grapes.

3 - I agree with Gestalten almost all the barrels at Cantillon are stripped clean before being filled. In the rare instance he uses a "fresh" barrel, that barrel character is very clearly indicated (for example in the various Cognac barrel treatments they have released).

4 - "they don't have the nutrients they would use to produce the metabolic reactions needed to break down the longest chain carbohydrates, so they poop out early." This makes sense and is why I assume a turbid mash is important.

 

[TripleSixHoppia]

I'm not surprised by St. Lam. The grape sours always seem to be funkier and more acidic.

Also grapes are acidic as ****. And some varietals are pretty dusty/funky especially when they have brett in them.

Isn't St. Lam with grapes from Saint Émilion? That region can be pretty dirty/tobacco barn.

 

[Gestalten]

Also grapes are acidic as ****. And some varietals are pretty dusty/funky especially when they have brett in them.

Isn't St. Lam with grapes from Saint Émilion? That region can be pretty dirty/tobacco barn.

Ah, I didn't know that about the wine varietal. Sounds really interesting.