Fermenting Belgian-Style Beers

Famed beer writer Michael Jackson tells a particularly colorful story about how German, English and Belgian brewers differ, painting an image of a Belgian brewer as circus performer. “You imagine a lion tamer in a cage,” he said one time, stepping back and lifting an imaginary whip. “They love the danger of working with wild yeast.”

He was talking not only about the sort of wild yeasts we associate with sour beers, but also Belgian Saccharomyces cerevisiae in general.

Consider these facts:
The Trappist breweries Westmalle, Westvlet-eren and Achel all ferment with yeast top cropped at Westmalle. Westvleteren and Achel use it just as fresh as Westmalle, picking up recently harvested yeast the day they brew.

At Achel, yeast is pitched at 63–64 °F (17–18 °C) and climbs to 72–73 °F  (22–23 °C) during fermentation in cylindro-conical tanks. At Westmalle, yeast is pitched at 64 °F  (18 °C) and rises only to 68 °F (20 °C) during fermentation in closed squares. At Westvleteren, yeast is pitched at 68 °F (20 °C) and reaches 82–84 °F (28–29 °C) in open fermenters. Same yeast, three very different schedules.

At  Brasserie Caracole, brewers of Saxo, Troublette and Nostradamus, yeast is pitched at 77 °F (25 °C).  The brewers allow it to ferment as it will, depending on the season. In the summer it may reach 86 °F (30 °C) and in the winter it will fall to 68 °F (20 °C).

At American microbreweries, the usual pitching rate is 1 million cells of yeast per milliliter of wort per degrees Plato (cells/mL/°P). (One degree Plato is roughly equivalent to 4 gravity “points” on the specific gravity scale.)

Common advice for homebrewers calls for boosting the pitching rate by 50% for higher gravity beers. In contrast, Westmalle pitches 5–6 million cells per milliliter for its 19.6 °P (1.081) Westmalle Tripel — just over 0.25 million cells/mL/°P.

When working with these and other Belgian yeast strains, good fermentation practices can’t be ignored. However, these strains are different from British and American ale yeasts — and from one another.

Ron Jeffries — founder and brewer at Jolly Pumpkin Ales in Michigan — speaks like a brewer who has entered into a partnership with his yeast, rather than expecting it to obey his orders.

“I usually let (fermentation) start in the upper 60s (Fahrenheit, around 20 °C),” he says, “and finish in the mid-80s (~29 °C). I try not to mess with it. For me, all the best beers I’ve made with Belgian yeast have been the ones I’ve done the least with. The yeast is almost always one step ahead of me. I’ve learned, don’t slow them once they start. If you try (to dial down the temperature), what you think is under control isn’t. Once the temperature jumps up, step back.”

Belgian Yeast Strains

Belgian yeast strains are different. They tolerate higher alcohol levels than many other beer strains, attenuate well and generate a range of phenolics and esters. Chris White of White Labs sees cells that have a smaller surface area than other ale yeast when he looks at them under a microscope. Dave Logsdon of Wyeast Laboratories finds, “Belgian yeasts have a lot in common with wine yeasts. They produce phenolic compounds that are similar to wine yeasts.”

We understand that Belgian yeasts, aided by a proper mashing schedule, will attenuate well and produce high levels of alcohol. On the down side, we know that many show a low degree of flocculation. As such, filtration or an extended amount of conditioning time may be required to get them to clear.

The complex aromas — with scents reminiscent of fruit and spices — are what set Belgain ales apart. These characteristics stem from the esters, higher alcohols and phenols generated during fermentation. Take a good whiff of a Belgian ale and you may smell pears, apples, tangerines, oranges or strawberries. In beers that contain darker malts or sugars, you may also detect raisins, plums, figs or prunes. Peppery, perfumy and roselike characteristics may also be found in many Belgian beers.

Belgian yeasts exhibit some similarities to other wheat yeasts, although the phenolic character exhibited in Belgian beers is not the same as that found in Bavarian wheat beers. In 2003, Wyeast and Microanalytics Corporation tested a variety of wheat and Belgian yeast strains using a gas chromatograph, a piece of analytical equipment that separates mixed gases into their component parts and gives their relative concentrations. Dave Logsdon and Larry Nielsen (Microanalytics) presented their findings at the 2003 Craft Brewers Conference.

Levels of 4-vinyl guaiacol (4VG), the molecule responsible for the signature flavor of a Bavarian wheat beer, were higher in Wyeast 1214 (Belgian Ale), 3787 (Trappist High Gravity) and 3522 (Belgian Ardennes) than they were in Bavarian wheats. The panel identified them as spicy and singled out the clove in 1214 and 3522. On the other hand, Wyeast 1762 (Belgian Abbey II) showed only trace amounts and the panel cited no clove or spice characteristics.

Styrene — which has a resiny flavor,  perceived as phenolic by some — was found in Wyeast 1214, 3787 and 3522 at roughly the same levels as in a wheat beer. As with the 4VG, Wyeast 1762 showed only trace amounts. However, 1762 registered levels of phenyl ethyl alcohol and phenyl ethyl acetate, which result in rose and honey notes, closer to the other Belgian strains. Phenyl ethyl alcohol is necessary for the recognized flavor of beer, and may stand out more in beers fermented with 1762 because of lower levels of clove and spices.

It’s important to note that the percent of wheat in the grist (40%), and how it was mashed, surely affected how the beers were perceived.

Esters and Higher Alcohols

Esters are the most important aroma compounds in beer. In general, they impart a fruity character to beer. Most esters are desirable, but ethyl acetate — which is perceived as solventy, like nail polish remover — is not. Likewise isoamyl acetate, which smells like bananas, may or may not be desirable.

Higher alcohols, sometimes called fusel alcohols, are produced during primary fermentation alongside ethanol, although at much lower concentrations. Some can be converted to softer esters during conditioning. Those that don’t can contribute harsh, solvent-like flavors over a certain threshold. However — in the right beer and at the right level — higher alcohols may also increase the complexity of a beer, and those that soften in secondary add spicy, perfumy and roselike aromas.

Most of the beer produced worldwide is lager beer. Consequently, most of the studies on esters in beer have been in lagers. However, some of that research may be applicable to ales.

Gregory Casey, director of brewing services at Coors Brewing Company says, “My belief is that, directionally, many of the findings (regarding esters) with lager beer would be applicable to ale. In the case of esters and higher alcohols, the pathways leading to their formation are ‘core’ pathways for Saccharomyces yeasts in general.”

Casey gave a talk on fermentation at the 2005 Rocky Mountain Microbrewing Symposium. There, he related that yeast growth and higher alcohol production are directly correlated — i.e. more growth equals more fusels.

As with higher alcohol production, most sources claim that ester production is increased by yeast growth. However, Casey presented evidence that — at least in some circumstances — ester production may be inversely related to yeast growth.  He did show, however, that this association could be altered by the fermenter design or amount of trub carried over into the wort. Casey also cited a study that found that temperature had a much greater effect on ester production than pitching rate did; increased pitching rates lowered ethyl acetate levels and increased temperature increased ethyl acetate levels.

Fermentation Temperatures

On their posters and at their websites, yeast producers offer suggested fermentation temperatures for all of their strains. The temperatures listed for many Belgian strains might be considered “fool proof,” but are lower than the temperatures reached at some Belgian breweries. Yeast suppliers don’t want to see homebrewers, enchanted by reports of high temperatures at which some Belgians ferment, losing control of their fermentation. Logsdon explains, “For homebrewers, the problem is lack of control. If they start at 75 °F (24 °C) and let it go, then they are going to get lots of higher alcohol and solventy character.”

Once the fermentation temperature has risen beyond a certain point, it may not be possible to correct the problem and still yield good beer. Brother Joris of  Westvleteren will still try to slow the rise in fermentation temperature if he suspects it exceeds 84 °F (29 °C), even if it means rising in the middle of the night. He knows that if he tries to reduce the temperature beyond that point, his yeast may crash. Others report White Labs WLP530 (Abbey Ale) and Wyeast 3787 Trappist High Gravity), both Westmalle offspring, acting the same way for them.

“When you cool them, they stop,” White said. “They go into survival mode. You can try rousing them, raising the temperature, but they won’t start again. You just have to add a new yeast. You don’t want to let it spike, and that can be hard to control in a homebrew situation.”

Obtaining the flavor profiles listed in the literature of yeast producers requires considering many variables associated with fermentation. It helps to recognize at what stage of fermentation different flavors are produced. “You get more phenolics at lower temperatures,” White said. “The absence of esters makes them stand out more. If you continue to suppress the esters, then you will continue to perceive the phenolics. You are looking for a balance.”

American craft breweries such as Russian River Brewing and Allagash Brewing found similar balance through trial and error. Both breweries now let their temperature rise during the fermentation process. This allows them to retain the esters and attenuation they wanted without getting solventy notes.

In Belgium, Duvel Moortgat pitches yeast at 61–64 °F (16–18 °C) and will let it rise to as high as 84 °F (29 °C) during five days. “One of the things that starting cooler does, is it leaves some of the fatty acids for ester production otherwise utilized early by yeast growth,” Logsdon said.

So what’s the take-home message for a homebrewer interested in Belgian-style beers? Although Belgian brewers ferment their beers at higher temperatures, you still have to be able to control the fermentation. As a starting point, you need to know the temperature of your fermenting wort. At Westvleteren, the monks don’t measure the ambient temperature of their fermentation room — they measure wort temperature directly from the middle of their open tanks. If more homebrewers had probes inside their fermentation vessels, they might be surprised. “At a minimum, you’d expect a temperature rise in moderate fermentation to be 7 °F (~4 °C), and it might get a lot hotter,” Logsdon said. A strip thermometer on the side of a glass carboy will be more accurate than measuring the ambient temperature, but glass is a good insulator.

To increase the amount of air the wort is exposed to, try using multiple fermenters. This will reduce the height-to-width ratio of your wort and possibly put a damper on surges in fermentation temperature. “I’d go shallow, and I wouldn’t even put an airlock on,” White said. As an added benefit, it is easier to top crop yeast  from a plastic bucket than a glass carboy. Top-cropped yeast can be used for another fermentation or as bottling yeast.Higher temperatures are part of the reason that many Belgian beers are so well attenuated, usually more than yeast company profiles promise. For instance, Duvel is 93% attenuated, Westmalle Dubbel 87% and Chimay Blue 89%.

When Wyeast and White Labs provide guidelines for apparent attenuation they base them on all-malt beers, usually not fermented at the top of the suggested temperature range. Beers with sugar providing more than 10% of their fermentables will attenuate further, and further still at higher temperatures. “It’s really important that brewers let them reach terminal gravity,” Logsdon said. “I have heard too many brewers who say ‘I’m going to stop it here,’ because they’ve calculated what the attenuation should be. The worst thing you can do is get incomplete fermentation.”

Pitching Rates

As mentioned at the beginning of the article, American microbreweries usually pitch around 1 million cells of yeast per milliliter of wort per degree Plato. For example, Brewery Ommegang pitches 18.5 million cells per milliliter for its 1.076 (18.5 °P) Ommegang Ale — right at the “standard” rate.

In contrast, Duvel Moortgat in Belgium, which owns Ommegang, pitches just 7.5 million cells per milliliter in fermenting Duvel, a beer with an original gravity of 16.9 °P (1.069) — or 0.44 million cells/mL/°P.  Ommegang Brewmaster Randy Thiel and Moortgat Brewing Technical Director Hedwig Neven have discussed lowering the rate at Ommegang, but experiments yielded a beer that didn’t attenuate completely.

For the past year, Moortgat has brewed Ommegang Ale while Ommegang expanded capacity. Both versions were brewed to the same specifications using the same recipe and basically the same ingredients. The only difference is that Moortgat pitched 8 million cells per milliliter. In blind triangle tastings, only half of consumers could pick out the odd beer, just what you’d expect by chance.

Of course, these beers contain a good percentage of sugar. Simple sugars are easier for yeast to process, and the breweries are confident about the viability of their yeast. Many are using freshly top-cropped yeast — not impossible for homebrewers, but not common either.

White understands. “On the professional level, the norm is pitching 2 L (of yeast) per barrel (34 gallons/129 L),” he said. “Belgians are below that. I’ve talked some (American) brewers into cutting back on their pitching level, and they are surprised their fermentations are stronger. By pitching a little less, if your yeast is healthy, flavor is going to be spit out during growth.” He added a warning: “Of course, if you don’t pitch enough, you get solventy. The Belgians know where that balance is.”

Finding the Balance

Unfortunately, there is no guarantee that if you pitch a certain amount of yeast into a wort with a given original gravity and ferment it at a specified temperature that you will produce a perfect beer.

As Peter Bouckaert of New Belgium Brewing says, “Brewing is a compromise. You have to take into account so many factors. You can’t look at the temperature as a sole factor. It’s an interaction. You need to see any beer you create as a holistic thing.”

Key Variables (Revisited)

When planning your own Belgian-style beer fermentation at home, here are some of the key variables to consider:

Belgian yeasts typically produce more esters than British ale yeasts, and also some molecules associated with wheat yeasts. Logsdon adds, “Fusel alcohol raises perception of isoamyl acetate (banana). It isn’t detected as strongly when fusels are lower.” Higher original gravities, higher attenuation and inadequate aeration leads to more esters. (See the sidebar on page 45 for more.)

Increased fermentation temperature increases ethyl acetate levels, floral and fruity esters, and may be necessary for some of these yeasts to finish attenuating. Lower temperatures promote perception of phenols.

Higher pitching rates lower ethyl acetate levels. Very high or very low pitching rates increase ester levels.

Taming the beast begins with finding a balance. Defining balance can be another matter.

Consider Logsdon back on the subject of pitching rates. “Boosting the (pitching) rate reduces esters and creates more acetaldehyde (which have a green apple character). Reducing aeration increases esters,” he said.

He paused, then laughed. “Everybody has a different opinion of optimal profile.”


Yeast Origins

White Labs and Wyeast don’t officially list the source of their Belgian yeast for good reason.

Wyeast acquired its 1214 more than 20 years ago from Chimay, and has monitored the yeast to make sure it hasn’t changed since. Chimay, on the other hand, “cleaned up” its yeast in the 1990s, so it’s not the same as it once was.

Then there’s the “abbey” yeast that New Belgium Brewing uses for its Abbey Ale. Founder Jeff Lebesch cultured that from bottles of Chimay in the 1980s. “What I learned later is that Chimay could get kind of wild,” Lebesch said, “so who knows how reflective what I got out of that bottle was of Chimay. I was doing all my culturing from bottles then, keeping them on plates in the house. Somewhere in the early 1990s, I did a major cleanup of our yeast. It really changed the character of the beer.”

The bottom line is that a source is just that, a source. And here they are:

Wyeast: 1214 (Chimay), 1762 (Rochefort), 3522 (Achouffe), 3787 (Westmalle), 3864 (Unibroue), 1388 (Duvel), 3538 (Corsendonk-Bocq).

White Labs: WLP500 (Chimay), WLP510 (Orval), WLP530 (Westmalle), WLP540 (Rochefort), WLP550 (Achouffe), WLP570 (Duvel).


The Shape of Your Fermenter

Studies — mostly done with lager beers — show that  the levels of ester production decrease when CO2 levels increase around the yeast in a fermenter.

Coors Brewing Company Director of Brewing Services Gregory Casey says, “A classic example of a design impact became apparent with the introduction of cylindro-conicals in replacement of traditional box fermenters. Due to the greater height-to-width ratio of the former, many of the early lagers coming out of cylindro-conicals were much less estery than the box-fermented counterpart (all other things being equal). This ‘washed-out’ ester character was linked to CO2 inhibition, a finding which has since been applied in designing cylindro-conicals to more reasonable ratios (lower height to width).”

Abbaye d’Orval converted to cylindro-conical tanks in 2004 and found that it could produce a beer it considered basically unchanged in flavor, but there were some measurable differences in the levels of esters that occurred above threshold levels.

Of course, the depth of a homebrew bucket or carboy is less than even the smallest of shallow commercial fermenters. And, the pressures that inhibit ester production are not even approached. As such, experimenting with fermenter shape may not yield the same results as at a commercial scale with regards to esters.


Those Elusive Esters

Yeast growth stimulates the production of esters. Yeast growth inhibits the production of esters. If you read various brewing texts and listen to different experts — or read this issue of BYO — you can find both opinions expressed. How can this be?

Almost everyone agrees on a few facts about ester production during beer fermentation. Esters are formed by a reaction between an alcohol and a molecule called acyl CoA. (Ethanol and acyl CoA form the common ester ethyl acetate.) Different yeast strains produce different levels of esters. In addition, higher temperatures, higher original gravities, higher levels of attenuation and an inadequate amount of wort lipids (from trub) all lead to increased ester production.  Likewise, on a commercial scale it’s well-established that deeper fermenters — much deeper than a bucket or carboy — decrease ester production.

Inadequate aeration is known to increase ester production and adequate aeration lowers it. Sources differ on the effect of increasing levels of aeration to very high levels. Likewise, smaller pitching rates (requiring correspondingly more yeast growth) have long been cited as a factor that stimulates ester production.

Recently, however, some beer scientists have claimed that yeast growth inhibits ester production (because it depletes the pool of acyl CoA).

So what’s a homebrewer concerned about esters to do? Biological systems are complex and figuring out the role of yeast growth and ester production may depend on other, as yet unrecognized, variables. Likewise, there may not be a linear relationship between the variables in question — it’s at least theoretically possible that adequate yeast growth limits ester production, while too much or too little stimulates it.

From a practical viewpoint, if you are interested in controlling the ester level in your beer, you should be able to do so by selecting a suitable yeast strain and running a good fermentation at an appropriate temperature. Anecdotal evidence strongly suggests that underpitching, poor aeration or high fermentation temperatures are the main culprits when excessive esters are present in homebrew.
— Chris Colby

Issue: July-August 2006