Brewing & Winemaking


 

How to Make Sour Ale: an inquiry

Raj B Apte

Matadero Creek Brewery

My favorite ale is Flemish Red. The style is closely related to Lambic--in fact, the more I know, the less difference I can see other than color, wheat, and racking. New Belgium Brewing La Folie  Rodenbach Grand Cru are my two favorites. But how does a homebrewer go about making these beasts? Follows a working monograph describing my progress. Comments and suggestions are welcome. Last updated Nov 2004.

 

Background:

New Belgium La Folie, Rodenbach Grand Cru, and  Verhaeghe Duchesse de Bourgogne are Flemish red ales. This style of sour ale is fruity and sour, with a pronounced lactic, acetic and balsamic flavor. La Folie is a wood-aged ale, blended, filtered, and primed. Duchesse is a filtered and force-carbonated blend of 8 and 18 month old beer, 60/40. The malts include caramel and possibly munich or vienna. Wheat is probably used. The fermentation follows the outline of Rodenbach: dextrinou mash, ale primary fermentation for one week, racking to lactic secondary conditioning for four weeks, racking to lactic and Bretta tertiary conditioning for months or years. Large oak vessels are used, but contribute little oak flavor. Long boils might be used for color.

Introduction:

Text and slides that explain the basics of Flemish Red Ale for general audiences. An excellent general review of fermentation is Battcock and Azam-Ali. A nice monograph on American corn beer is Cutler and Cardenas.

A new set of slides for the advanced brewer gives a better overview of the tandem fermentation dynamics, including data from van Oevelen and Martens of the KU Leuven brewing school. I presented these slides at the Northern California Homebrew Festival in 2004.

My Ales:

Assuming you have understood the introductory material, I will give a sketch of my methods. A basic recipe for Rodenbach Grand Cru can be found in Wheeler and Protz. My most recent batches are even simpler:

Basic Sour Ale

20 L batch
Pale 2row (I use organic)   6.5kg
Caramel (your choice L)       1kg
white wheat, raw                  1kg

OG 1.065
3 L/kg=18 L mash liquor
20 min 40 C
40 min 68 C
40 min 72 C
mashout 74 C
sparge 85C 20L
30 L collected.

80 g aged hops from hoptech.com (with additional aging--their aging is not sufficient).

Fermentation; pitch any ale yeast or Wyeast Roeselare (which is very nice). Ferment out for 1 week (I use open). Rack to glass carboy for 4 weeks secondary. Rack to aging vessel for 12-24 months.

Optional: fruit added after no less than 6 months aging.

I ferment this gyle with a variety of primary yeasts. After primary fermentation, the beer is racked off the yeast and inoculated with souring bugs. Secondary fermentation is in glass and lasts a month or two. I have two oak casks, 10 and 15 gallons, for the tertiary fermentations. One is inoculated with every bug I can get my hands on, the other with ale fermented with Wyeast Roeselare. Tertiary fermentation lasts about a year. Batches taken too early usually have faults (DMS, insufficient sourness, &c) that smooth out in time. Alternatives to wooden barrels (expensive) include plastic buckets and wood-stoppered glass carboys (see below). Quaternary fermentation consists of racking to a glass carboy with 3#/gallon cherries or several ounces of aroma hops. Dryhopping this style is quite nice. Bottling is done with 6g/L sucrose boiled into a syrup. I use 750ml sparkling wine bottles exclusively.

Issues:

1. Funk. One of the key differences between Lambic and Flemish Red is the use of multiple fermenters. By remaining in one vessel, the Lambic is exposed to a rich bed of autolysing yeast nutrient. This gives the Pediococcus and Brettanomyces something to chew on. In Flemish Red, a more restrained ferment is achieved by two rackings. It is only in the tertiary that the beer sits and sours, with very little sediment. Clearly this is a variable that can be manipulated to control flavor.

2 Hops. I use aged hops exclusively. They are 2-3 years old at room temperature in an open bag, or may be baked at 200F for an hour or two (not browned!). Hopping is generous, 50-120gms/20L. The fresh wort has a clear and coarse bitterness, but after aging the bitterness disappears.

Dryhopping sour ale is done by Cantillon. I have tried it to good effect with Czech Saaz. Cantillon Iris uses East Kent Goldings, I think..

3. Strain Selection. The flavor notes of a Flemish Red ale can come from the primary yeast strain, from the secondary yeast and bacteria strains, from the wood, from the long boil or from the malt. Usually the wood used is very old and has little flavour of its own left.

The Verhaeghe importer says that often kegs and even bottles will begin a fermentation if stored for long, despite the fact that they are filtered. There is noticeable sediment in most bottles. There may be enough microbes to get a culture going. The good news is the Duchesse is force carbonated, so there is no bottling yeast to create confusion. There is little mention in the literature of the importance of primary yeast in final flavor development of Flemish Reds. In my experience, Wyeast Roeselare makes a more fruity and interesting primary fermentation and gives more sourness than a typical ale yeast. For those who want to restrain the sourness, what are you doing here?

4. Malt Vinegar with Brett. The levels of lactic acid must be achieved by the souring bugs in the tertiary fermentation using nutrients from the mash. Acetic acid, if it is desired, partly comes from the early fermentation and partly from the long micro-aerobic tertiary. Acetic acid is produced by Brettanomyces and Acetobacter. I find that some acetic acid is necessary to sharpen the taste of sour ale. At the same time, Lambic brewers--even some traditional ones--are moving to limit the amount of acetic acid, which is very different from the 'soft' acidity of lactic acid. To achieve more acetic acid, I keep some ale, inoculated with Brett and Acetobacter, in an open container (cloth cover to keep Drosophila out). This can be used for blending to add acetate.

5. Pellicles. Pellicles are an indicator of headspace oxygen. Brett, Acetobacter, and other pellicle-forming bugs do so when the headspace of a container is oxygen rich. I don't understand the considerations between headspace and dissolved oxygen entirely, but too much headspace oxygen can result in too much acetic acid formation. Dissolved oxygen feeds Brett, but too much of it can oxidize the taste of the beer. Maltiness can be lost.

 

 Bretta likes to form pellicles and is very efficient in this mode. Pellicle formation may depend on the presence of gaseous O2. Although all Brettanomyces exhibit the Custer effect (aka  negative Pasteur effect), it is unclear if different flavour compounds are produced in aerobic or anaerobic conditions. Anecdotally, people complain of lack of complexity in pure culture lambics that are aged in glass. I take this to suggest that Brettanomyces should be grown aerobically, and that pellicle formation may be important for allowing additional oxygen into the system and accelerated Bretta production of flavor compounds. This is purely speculation at this point, but its consistent with the Acetobacter/Brettanomyces mixed pellicle observation (vs).

6. Alcohol. What is the optimum alcohol level for sour ale? Sour ales tend to be 2.0-7.3 v% alcohol. Given that most are taxed by extract, its natural to wonder what the optimum is, regardless of money (brewing is an art, after all). Generally, the lower the alcohol level the more beasties can survive in the gyle. High alcohol beers tend to age gracefully and seem rather immune to infection. Martin Lodahl's article in BT on Old and Stock Ales suggests that infections (sourness) may simply be untasteable underneath all that malt. Above 7.3 v%, there tends to be less sour flavor in the ales I've tried. Even Petrus Aged Pale, at 7.3 v%, is not very sour tasting (or has enough complexity to balance). De Dolle, which at one time got its yeast from Rodenbach, doesn't become very sour even after three years in the bottle (Speciaal 20).

  alcohol v% pH titratable acidity tartaric w%
Verhaeghe Duchesse 6.2 3.1 1.5
Petrus Aged Pale 7.3 3.4 1.0
Cantillon Iris 5.0 3.4 1.4
Hanssens Oudbeitje 6.0 2.8 1.5
Morte Subite Peche 4.3 3.6 0.7
Kriek de Ranke 7.0 3.5 0.7
Drie Fonteinen Gueuze 2002 6.5 3.3 0.9

New Belgium La Folie

6.0 3.1  
Lindemans Cuvee Rene 1993 5.0 3.3 1.0
Rodenbach Grand Cru 6.5 3.1 1.1
Matadero Creek Cask Porter (me) 6.0 3.3 0.6
Matadero Creek Kriek (me) 6.0 3.5 0.8

At the other end of the spectrum, Berliner Weisse can be 1-3 v%, with a typical OG of 1.030 g/L. The above table suggests 1.064 g/L as a good starting point for a 'premium' ale with plenty of aging potential. One strategy for brewing sour ales depends on the observation that souring is suppressed at higher alcohol levels. Even though my house is rife with sour barm, my barleywines turn out quite sweet and malty. If this is the reason, then one way to control sour ale production is to brew, age and sour as desired, and finally add sugar to raise the gravity. By stopping the acid production, a smoother ale may result. I wonder if this is how Boon Marriage Parfait is done. My guess is that souring an 8% ale is not possible.

7. Acidity. The above table would benefit from more titratable acidity (TA) and pH measurements (donate any of the above and I'll do the experiment!). TA is comparable to wines and significantly less than common soft drinks. The relationship between TA and pH explains part (but not all) of the organoleptic qualities of different acids. Generally, acids with high pKa1, like acetic, will have a higher pH for a fixed TA. Acids with low pKa1 will have a low pH for the same TA. For example, by using strawberries, the Hanssens Oudbeitje contains much citric acid. It is a strong acid and will lower the pH for modest TA, which is exactly what is observed. This is probably even more true of the Cantillon Vignerone, which contains tartaric (grape) acid. Generally, TA is a good measure of sourness. pH is harder to pin down, but for me affects the mouthfeel and crispness. For most people it will be the tastes of particular acids that dominate the choice.

Acid pKa1 pKa2 Mass/Mole
Tartaric 3.02 4.54 150
Citric 3.03 4.74 192
Malic 3.40 5.05 134
Lactic 3.86 -- 90
Ascorbic 4.04 11.8 176
Acetic 4.76 -- 60

Two decisions are to me made: what acids may be desired intrinsically (ie, lactic and acetic) or from added fruit (tartaric, citric, malic).  Extrinsic acids are common in Lambic and Flemish Red, with sour cherries and raspberries being the most common. See Rotter.

Fruit Acid tartaric w% Sugar w%
Apple Malic 0.8 13
Apricot Malic 1.7 9
Blackberry/Raspberry Isocitric, Malic 1.6 7
Black Currant Citric 3.2 10
Blueberries Citric 0.3 11
Cherry, sweet Malic 0.5 14
Cranberry Citric Malic 3.0 4
Elderberries Citric 1.5 8
Grape, wine Malic, Tartaric 0.6 12
Orange peel Malic, Citric    
Pomegranate   1.2 13
Strawberry Citric 1.6 7

The choice of intrinsic acids depends on how much control of the ferment is possible. Lactic levels are notoriously hard to control--ropy examples may be two to three times more TA than soft ales. Acetic is easier to control since it depends so strongly on oxygen availability (vi). Malo-lactic fermentation may be important in Kriekbier development: it will raise the pH, lower the TA, and add new flavour components. ML starter cultures are readily available.

Blending of different acidity profile ales is an old and respected process. You can stretch the limits of this by fermenting different gyles under different conditions (anaerobic, microaerobic, macroaerobic) to allow blending and control of the final product (lactic, balanced, acetic). However, the additon of food grade acid is not appropriate.

8. Oxygen. The tertiary fermentation is traditionally done in scraped oak. Scraped, old oak barrels contribute limited flavor but retains the oxygen permeability of new wood. Estimates of diffusivity of wood and some plastics follow in the table (courtesy of  Flextank.com and evalca.com). Flextank is a world leader in using polyethylene for wine aging. They claim that 200L barriques admit 20cc/L.year of oxygen but that wine could use, perhaps, four times that amount of oxygen during the bulk aging without increasing volatile acidity (acetic acid). They also report that food grade PE of any density will retain its porosity in the presence of wine for many years--it resists clogging by wine components or scalping flavor from the wine.

Material

Oxygen Permeability
cc-mm/m2-day, 23C

Wood, Oak 57
High-density polyethylene HDPE 18
Polypropylene PP 20
Polycarbonate PC 36
Nylon (not oriented) 0.62
Saran 0.02
Vinyl 20
Silicone 10600
Water 34400

 

 

Micro-oxygenation v. macro-oxygenation. How equivalent are the small doses of oxygen the diffuse through the wood on a daily basis compared to a weekly or monthly ventillation of the headspace? In the wine business there seems to be preference for the former. For flemish red or lambic, I don't know. Please comment if you have tried this experiment. In the absence of a pellicle, its hard for me to see that there would be much difference. In this respect lambic aging is different from Flemish Red. Lambics casks are often so poorly stoppered that the intact pellicle on the beer in necessary to prevent over-oxidation. With good stoppering, preserving the pellicle intact  may be less important. The use of plastic fermenters, while not necessary, is very economical. Many commercial sour ales are not aged in wood (including Lambics). Nobody likes to talk about this. The good news is that the standard, food-grade HDPE bucket is the second best fermenter for sour ale.

Tank Volume [L] O2 cc/L.year
Burgundy barrel 300 8.5   
Rodenbach tank, wood, small 12,000 0.86
Rodenbach tank, wood, large 20,000 0.53
HDPE bucket 20 220      
Homebrew barrel 40 23      
Glass carboy, 30cm vinyl immersion tube 20 0.31
Glass carboy, silicone stopper 20 17      
Glass carboy, wood stopper 20 0.10

Wood stoppers are an excellent way to turn a glass carboy into a sour ale aging tank. I use tapered red oak furniture legs ($2ea). White oak would be better, but I have'nt found a source yet. I trim the hardware off and toast the pegs in the oven at 375 for a couple of hours. Then I trim them to size for a given carboy (which do not generally have well shaped mouths). Fill with cleared flemish ale or lambic wort. For flemish ale, I fit the peg in hand tight. For lambic, I let the fermentation complete with a lock, top off with water, and then fit the peg a month later. Once the lambic started re-fermenting. I feared that the carboy would explode, but it didn't. The peg started pumping beer from the open end--ie the pressure was forcing beer through the grain of the wood to the top of the peg. At its peak, the peg wept like a spanish madonna. The carboy settled down in a week and seems OK. The peg will under-oxygenate the beer, but does provide a nice place for Brett to grow. Time will tell if this is a good way to preserve maltiness in flemish ales (look for results in fall 05).

Flemish sour ale wort should be oxygenated at pitching. However, for a more extreme lambic, limited or no oxygen will produce a strongly under-attenuated gyle. This will sour spectacularly.

9. Mashing. Lambic turbid mashing is a fun process, but I can't in good conscience recommend it for sour ale generally. It represents an extreme of unfermentabable wort. But dextrines, starches, and nitrogen (FAN) are clearly needed for year-long fermentations.  Flemish red ale is usually mashed simply or, if adjuncts such as corn and wheat are used, with a decoction or American standard adjunct mash. In order to increase the starch levels in sour ale brewing, I propose a modified adjunct procedure to increase unsaccharified starch level.

Here's the variation. Start with all the wheat and maybe a handful of barley in the adjunct kettle and all the barley in the mash tun. Heat the adjunct kettle to 65C for 30 minutes, and then boil. The tun does a normal stepped mash:

50  30 min rest
65  30 min rest
70  30 min rest
75  mash out
sparge with 80C

In the normal adjunct method the goods of the 100C adjunct kettle are dumped into the main mash to raise the temperature from 50 to 65. In the variation, do the same, but don't dump all the liquid. Add the reserve at the 75 mash out. The adjunct added at 50 will saccharify to dextrins and sugars in the normal way. The reserved adjunct addition will add starch to the mash only after the enzymes are denatured. In this way, you can simply control the relative concentrations of starch, dextrins, and sugars.
The reason I like this better than the Liddil/Boon variation (see Liddil) is that no enzymes are harmed in the process: all the diastase is kept in the main mash and is not boiled until after mash out.

I also make seitan from wheat, periodically. The leftover starchy water can be boiled and added to a sour ale fermentation. I add wheat starch to the tertiary fermenter to prolong the fermentation. The starch haze seems to be reduced in less than 6 months.

Any mashing procedure for sour ale needs to balance three elements: fermentable sugars, dextrines, and starches. Fermentable sugars are consumed rather quickly and produce alcohol and some acididy. Additional acidity and Brettanomyces funk come from bacterial amylase acting on starches during aging. Dextrines are pretty much unfermentable (unless lambic culture is used including enterobacteria). Dextrines provide sweetness.

10 Sour Mashing. It would be lovely to make sour ale quickly. It is possible to do a fast lactic acid fermentation, but we don't know how to speed up the Brettanomyces, yet. So if you must sour something quickly (and this procedure makes a very nice Berliner Weisse), here's how. First, ignore all the rubbish about smelly sour mashes in the mash tun. Lactobacilli are meso- or thermo-philic and anaerobic. Use a fermentation lock. The easiest way to sour a gyle is to put it in a carboy, inoculate with pure culture or a handful of grain, and hold at 120F for a few days with a fermentation lock. You can pitch primary yeast directly on top of it when cool, or try to pasteurize it first.

 

Fast Malty Lactic Ale

This procedure is an attempt to produce a sour ale with little or no acetic acid but with much lactic acid. The procedure is based on Wyeast Roeselare slurry (a mix of microbes). Tandem lactic and ethanolic fermentations are used. To limit acetic acid, the wort is aged in glass. Because the fermented wort is racked clear into secondary, there will be very little oxygen or nutrition for Brettanomyces activity. I also make liberal use of boiled yeast slurry to provide nutrients. The high gravity makes a sweeter and fruity ale that goes well with sourness. Note that this method gets around the usual limits on alcohol and sour ale without using blending. The idea for this came from Hilde Martens' thesis and a very interesting article on chicha (see introduction).

1. Mash pale (75%), caramel (15%) and wheat (10%) malts to produce 24L@1.072 (18P). In the boil, use no hops, but 15 min before knockout pitch 100g of slurry for nutrients.

2. Run off 8L of wort to second kettle. Boil with 40g aged hops for 1 hour. 15 min before knockout pitch 20g of slurry for nutrients. Chill, oxygenate, and pitch 50g slurry for starter.

3. Main gyle from step 1 is cooled to 40C. Run off to closed fermenter. Pitch Lactobacillus. Hold at 37-45C for 12-24 hours, until thoroughly soured, under fermentation lock. When fully soured (taste it!), cool and pitch hopped starter from second kettle. Ferment at 20C (use water bath to keep cool). Rack to glass secondary after a week.

Details: 24L wort, 70% eff, 1.074 TG
Pale Malt 6.4 kg
Caramel Malt 1.3 kg
Wheat Malt 0.9 kg.
40 g aged hops (2.7% AA), 14 IBU or less (12 IBU including utilization).

Mash in 22L @ 42C.
Hold 40C for 10 min.
Heat to 60C, rest 40 min.
Heat to 70C, rest 40 min.
Heat to 74C, rest 5 min.
Sparge with 20L @ 76F.

I use a pid hot-plate for souring. It is also very useful for making cheese, yoghurt, and tempering chocolate. The trick is to keep the temperature probe loose and to tape it to the appropriate place for each application. Typcial probe locations include the plate itself or the sides of a vessel sitting on the plate. For souring wort in a glass carboy, I rest the carboy in a partly filled kettle on the hotplate (double boiler). The probe is on the kettle, and I insulate the whole thing to keep a nice, uniform, even temperature. photo 1, photo 2, photo 3.

11. Culture. Among the differences between Flemish Red and Lambic is yeast pitching. In Flemish Red a conventional (infected?) starter is pitched, while Lambic is inoculated with a very small number of cells in the brewery itself (probably from wood or other unclean surfaces). But both styles depend on small inoculations of old ale and wood to get the slow growing bugs started. Thus, pitching dregs of commercial beers should be sufficient to inoculate either style after the primary fermentation is over.

 In order to keep the secondary bugs going and--maybe--increase the consistency of taste, some have set up beer soleras. I have done this for sour ale, and I think it should work fine for lambic. But there are two strikes against it. For blending purposes, it would be better to have several separate beers that could be blended. A solera is not likely to maintain a consistent flavor in the case of sour ale (Bouckaert says specifically that each time a cask is filled, even with 10% holdover, the taste can be quite different).

Second, you still have to respect that it takes a year for the ale to achieve basic structure. Thus, you can only remove ale on an annual basis at most. Every time you tap it, you change the environment. Its a smaller effect in Flemish Red, and a huge effect in lambic. For example, the level of dissolved O2 in a flemish barrel is very low for the first six months or so and then increases for the next year or more. If you add ale every six months, you'll never get the oxidative effects.

Wyeast Roeselare is a one-stop inoculant for Flemish red ale. I have used it only once.I pitched a tube into 20L of wort in an open fermenter. It took 24 hours to get started and then produced a nice krausen. The bubbling continued for a week at 65F ambient, 70-72F in the fermenter. After a week I racked into a oak cask and observed a renewed fermentation for two days. The smell at 1 week is very Belgian, almost like a dubbel.

 

 

Concluding Remarks

This is a good ale for the lazy, patient, and unsanitary (that's me!). Brew a couple of batches and let them age a good long time. Once you 10 or 20 gallons aging and a case or two in bottle, it takes very little effort to keep this habit up--these are not session beers. Don't rush: 18 months from mash to mouth is considered right. But for those who love the ale itself, the applied microbiology and chemistry of its production, or big, aged ales in general (barleywines, export stouts) this is a fun style to brew.


Acknowledgements

Many have contributed to this working body of knowledge. If you have experience in Flemish red or lambic brewing and can correct my errors and omissions, thank you--please send me

Sorry, but a Javascript-enabled browser is required to email me.

Special thanks to Peter Bouckaert of New Belgium Brewing for making lovely ale and egging me on.

 

References

http://www.fao.org/docrep/x0560e/x0560e00.htm -- a general review of fermentation.

http://hbd.org/brewery/Library.html -- a great place to start any brewing project. Has numerous links

http://hbd.org/brewery/library/Rodnbch.html -- included in the HBD library, this is the real deal from the horses mouth. Since leaving Rodenbach for New Belgium in Colorado, Peter Bouckaert has been introducing extremely balanced and subtle versions of sour ale to America, first in La Folie, and now in Transatlantique Kriek.

Martens Hilde, January 1996, "Microbiology and Biochemistry of The Acid Ales of Roeselare." This reference, cited by Bouckaert, is a treasure trove of information about the fermentation of red ale. Available at KU Leuven library.

Guinard, J.X., Lambic, Published by Brewers Publications, 1990. Very good reference to Lambic fermentation and grist. Don't bother with Rajotte's book on Belgian ale of the same series--it contains very little information.

http://www.johbarth.com/news-scientific.htm - A good website about hops.

http://hbd.org/brewery/library/LmbicJL0696.html - James Liddil's excellent Lambic page.

http://bergsman.org/jeremy/lambic/lambic.html  -  More about Lambic. Very informative.

http://www.belgianstyle.com/mmguide/taste/oudbruin.html - a great site with tasting notes on commercial Flemish ales.

http://www.geocities.com/beer_made_in_belgium/SourAle.html   -  a solid description of brewing parameters

http://www.allaboutbeer.com/columns/mj8.html - general discussion by Michael Jackson of brown ale.

http://www.northamericanbrewers.org/that_old_brown_magic.htm - another general discussion, incl. brown ale.

Copyright Raj B Apte, 2004. All rights reserved. You may link to this page (www.parc.com/apte/flemishredale.shtml), but no part may be excerpted or copied  without permission.

 

 

 

.