Red burgundy, Filtering, and Brettanomyces

Writing from San Francisco, Thomas Layton is a winemaker and lover of Burgundy. There are certain things he doesn't like to smell in his glass...

For years I was skeptical about red burgundies that had been filtered because I thought this meant they could be stripped of flavor. And I noticed that many red burgundy producers (not to mention new world producers of pinot noir) made a big deal of being “unfined and unfiltered” which seemed an important badge of honor.

Now more than 10 years later, I am pretty sure I got it all wrong. While filtering can have deleterious effects on pinot noir, I now believe that in many cases it is worth this risk especially for bottles purchased for storage and medium to long term consumption. The reason is Brettanomyces (a.k.a. Brett).

A SHORT INTRODUCTION TO BRETT AND ITS BY-PRODUCTS

Brett is yeast that is found in wine regions around the world. It apparently can grow in wine production facilities and on wine making equipment that is not routinely sterilized. Old barrels and to a lesser extent new barrels are thought by many to be a primary source for the introduction of Brett into wine production facilities. Old barrels are more often Brett contaminated but if new barrels become infected they can support massive Brett populations because of their high levels of cellulose, and other wood sugars, which serve as Brett food.

My layman’s understanding of Brett suggests that Brett organisms themselves do not have a sensory impact on wine. Instead, the spicy, smoky, metallic, horsy/mousy, and band-aid sensory qualities associated with Brett are caused by various by-products of the Brett organisms. These by-products are also cited for diminishing the fruit qualities of wine and thinning the palate over time.

While there are several such by-products, the ones most frequently cited are 4-ethyl phenol (4-EP) and 4-ethyl guaiacol (4-EG). In this post, I will ignore the many other by-products and focus exclusively on 4EP and 4EG because the presence and amount of these substances are easily tested by a wine lab and – importantly -- because the only way these substances end up in wine is via Brett.

One important aspect of Brett is that it can grow/bloom in bottle. This means that every bottle with Brett is essentially a unique Brett microcosm with different potential levels of future 4-EP/4-EG creation. The result in some cases is tremendous bottle variation.

When a wine is sterile filtered, the Brett organism is generally removed but any 4-EP / 4-EG already created is not removed. So filtering can stop the creation of Brett by-products, but filtering cannot remove any by-products already created. (There is one company that claims they can remove 4-EP with a novel resin. See www.winetech.com)

For those interested in learning more about Brett, here are two articles I found informative:

http://www.makewine.com/makewine/Brett.html

http://www.aromadictionary.com/articles/Brettanomyces_article.html

MY PERSONAL EXPERIENCE WITH BRETT

I first became interested in Brett after experiencing the evolution of the 93 Dujac Clos de la Roche. I remember several bottles of the wine consumed in 1999 and they were really fabulous. I started noticing by 2003 that the wine had changed and the red fruits had become muted, and the wine had taken on an odd spicy character. And this change in the wine seemed to me more significant than I would have expected from normal aging of burgundy. In 2004 I sent a sample to a wine lab and learned that it had 4-EP/4-EG.

I had another formative Brett experience with the 99 Roumier Chambolle Musigny. Upon release I found this to be one of the best young village wines I had ever consumed. More recently, I decided to check in on this wine and see how it was evolving. The wine had a very Bretty nose with muted fruits and it had become thin on the palate. So I sent it to a wine lab and learned that this bottle had Brett and 4-EP/4-EG.

These and many other experiences with Brett in red burgundy have led me to conclude that Brett is often more damaging to wine than beneficial, at least to my tastes. I can see how others might like the complexity provided by the smoky, spicy, animal and/or barnyard characteristics caused by Brett but I prefer wines with little to no 4-EP/4-EG (or more accurately, I prefer wines with 4-EP/4-EG levels near or below my sensory thresholds). These direct sensory impacts of Brett are unattractive to me, but what I really want to avoid is Brett’s tendency to cause a wine to become lean over time (probably via some diminution of fruit esters).

Compounding these sensory effects is the bottle to bottle variation that Brett often introduces to wine. Evolution of wine can be variable, and the result not always what we hoped for or anticipated, but to think that you could try a wine, love it, buy a significant quantity to cellar, and then watch it unravel unnecessarily via a Brett bloom over time is for me painful.

WINE LAB TESTS FOR BRETT AND 4-EP/4-EG

Out of curiosity I started sending wine samples to ETS Labs in St. Helena, California to have them tested for both (a) Brett via a DNA test for the organism, and (b) 4-EP and 4-EG. I was fascinated to learn that each of the 22 bottles I tested had some level of 4EP/4EG which means definitively they each wine had traces of Brett at some point in time. The lab results are summarized below.

It is important to note that these tests are of specific bottles and given that Brett and its byproducts can show substantial bottle variation these results do not necessarily have implications for other bottles of the same wine. Also, I want to make it clear that I only have a layman’s understanding of Brett. I am sharing this data so that others can learn what they will from my (not inexpensive) lab tests. And I look forward to the shared contributions of the more knowledgeable on this topic.

About a third of these wines were selected because I wanted to test wines from certain producers to see what I would find. The other two thirds were selected because I suspected that they had Brett. In other words, this is not a random test of red burgundy bottles and thus these results do not necessarily reflect rates of Brett incidence in red burgundies.

(1) 4-EP/4-EG levels below sensory thresholds; no or very little Brett

2002 Hudelot Noellat RSV
2001 DRC La Tache
1999 Bachelet Gevrey-Chambertin Corbeaux
1999 Fourrier Gevrey-Chambertin Clos St. Jacques
1999 Mugnier Chambolle-Musigny Amoureuses
1997 Mugnier Chambolle-Musigny Amoureuses
1996 DRC Grands Echezeaux
1964 DRC Romanee Conti

(2a) low levels of 4-EP/4-EG but above sensory thresholds; no or very little Brett (indicating that this wine might have been filtered or the bottle is old enough that the Brett died when it ran out of sustenance)

2002 Nicolas Potel RSV
1999 Dujac Clos de la Roche
2002 Clos des Lambrays

(2b) same as (2a) but with more Brett and thus the ability to create more 4-EP/4-EG

2000 Dujac Gruenchers

(3a) 4-EP/4-EG levels well in excess of sensory thresholds and little to no brett (indicating that this wine might have been filtered or the bottle is old enough that the Brett died when it ran out of sustenance)

1964 Leroy Grands Echezeaux
2000 Dujac Clos de la Roche
2002 Lignier-Michelot MSD Faconnieres

(3b) same as (3a) but with more Brett and thus the ability to create more 4-EP/4-EG

2002 Dujac Clos de la Roche
2001 Dujac Clos de la Roche
2000 DRC RSV
2000 DRC Richebourg
2000 Rochioli West Block
1999 Roumier Chambolle-Musigny

HOW WINEMAKERS MANAGE BRETT

From what I have learned (and again I look forward to the contributions of those who know much more about managing Brett in wine production than I do) it appears that winemakers have at least 3 categories of choices in dealing with Brett and none of these guarantees zero Brett and its by-products in the finished wine. These are listed in order of effectiveness in controlling Brett:

- Filter. Filtering removes most if not all of the Brett organisms, but it does not remove any 4-EG/4-EP already produced by these organisms. By removing Brett, filtering generally prevents Brett from blooming in bottle.
- Use 100% new oak. While new barrels are less likely to have Brett than old barrels, they still can have it. Thus using 100% new oak is not a guarantee of no Brett. And if a new barrel gets Brett then look out because the wood sugars in a new barrel can support massive Brett populations.
- Use a collage of Brett inhibiting strategies including steam cleaning of both new and used barrels, low temperature barrel storage, fanatical topping up of barrels, fanatical cleaning and sterilization of all winemaking equipment and facilities (tanks, valves, hoses, walls, floors, wine thieves, etc.), large SO2 additions post malo, acidulation to maintain a low pH which is inhospitable to Brett, and cask by cask bottling (to avoid spreading Brett from one or more barrels to the entire production).

MY CONCLUSION

Brett in a wine bottle is a ticking time bomb. As a result, I have decided to reduce my purchases from producers that do not filter (unfortunately, this is difficult to determine since most burgundy producers do not like to admit that they filter). Yes, I may miss some great burgundies by re-orienting some my purchases towards producers that filter, but my choices will still be pretty good since a number of great producers filter (among them Rousseau, D’Angerville, and Drouhin). To me, filtering seems preferable to a multi-decade game of Russian roulette with every bottle of red burgundy in my cellar.