masterclass with Matt Thomson, winemaker
Thomson is a well-known
winemaker, who consults widely, with particular associations with
Saint Clair and Delta (he’s part owner) in Marlborough
. He also works in Italy
(he’s been there every vintage since 1994), in association with
David Gleave’s Liberty Wines.
The purpose of this masterclass was to look at Brettanomyces
(abbreviated here as ‘brett’), that celebrated rogue yeast
that’s responsible for making some red wines smell of animal
sheds—it’s not usually encountered in whites.
Thomson brought over with him a range of barrel samples
of three different wines, two Pinot Noirs and a Merlot, which had
been taken from barrels that he and his winemaking team had
identified as being bretty, to show us what brett tastes and smells
like in its very earliest stages.
First, he began with a run-down on the facts about
brett. It is a yeast, and the nomenclature used in the wine industry
differs from the recently devised taxonomy, which brings the
previous five ‘species’ of brett into a single species, Dekkera
bruxellensis. It has been known for about a hundred years, first
having been identified as an important component in British and
Belgian beer styles.
Indeed, when the first single culture Saccharomyces
cerevisiae yeasts (this is the species of yeast used for making
wine) were used to make British beers, people noticed that something
was missing: the imprint of brett, which in the context of a good
bitter adds real interest. Interestingly, brewers commonly refer to
brett character in beer as being ‘vinous’.
Brett is a slow-growing yeast that is pretty tough, and
can survive hostile conditions such as high alcohol and low nutrient
levels. As it grows, a small volume of CO2 is produced,
which can often be a clue as to its presence in barrels that have
already finished malolactic fermentation. Growth is stimulated by
small amounts of oxygen, such as you might find in a barrel, and
particularly one that isn’t topped up well.
The sensory effects of brett are many. The first sign
is reduced varietal character, followed by the degradation of
certain fruity aromas by esterases present in this yeast. Thus Pinot
Noir is particularly badly hit: it loses its bright cherry and
violet characters, and this loss of fruit is a good early cue for
the presence of brett in barrels.
Specific aromatic compounds are produced by brett, and
their imprint can vary significantly. Affected wines can taste
medicinal, or earthy, or mousy, or smoky, or poopy. Compounds
produced by brett include 4-ethylphenol (this is only ever produced
by brett in wines, so it is used as an indicator of brett activity),
4-ethylguaiacol, isovaleric acid, 2-phenyethanol, guaiacol,
ethyldecanoate, trans-2-nonenal, isoamyl alcohol and
Brett likes oak. It particularly likes toasted new
barrels, and has been found 8 mm deep in staves. It can feed off a
compound, cellobiose, that is formed when barrels are toasted. It
likes high pH, residual sugar, low SO2 and lees (it
enjoys cloudy, turbid wine). It can go dormant, for example after an
SO2 addition, so that culturing doesn’t reveal its
presence, and then re-emerge later on to bloom in bottle. In short,
it’s a bit of a bugger.
One misconception about brett is that it a hallmark of
wineries with poor hygiene. ‘Brett can occur in the cleanest
cellars’, says Thomson. He thinks that oak is largely to blame for
many infections, because brett can live in the oak and is almost
impossible to get out by cleaning. ‘If you use new oak, you will
get brett: it is not something you can associate just with a dirty
But Thomson goes further, suggesting that brett is not
only associated with new oak, but also that he has identified
specific coopers who have a problem with bretty barrels, although he
won’t name them. He also thinks that brett is a growing problem.
‘I am convinced that in large numbers of wineries in both the new
and old worlds, brett is a new thing.’ He has a theory that
something happened to oak in the relatively recent past.
‘Something happened with the huge demand for new oak in the 1980s.
Coopers had a boom period and started doing something different, and
there was a change’. That’s an interesting idea.
What can be done to avoid brett, according to Thomson?
This is where things get interesting, because many of the steps that
need to be taken in order to ensure clean wines run counter to the
sort of winemaking approach you’d want to take to make interesting
The first is to avoid barrels. Stainless steel can be
cleaned properly, and ‘then you can pretty much eliminate it’,
says Thomson, ‘although you have to be careful with ball
valves’. Second, you need to avoid cross-contamination. When
taking barrel samples Thomson uses plastic barrel thiefs that are
used just once, and then sterilized. He also avoids doing rack and
return where the wine would go from several barrels to be mixed up
in one tank: instead, each barrel is racked separately to tank and
returned, and the tank cleaned before the procedure is repeated with
the next barrel.
The next stage is to keep pH low, either by acidifying
or harvesting earlier. Brett doesn’t like low pH, and at low pH SO2
additions are much more effective. Other steps include avoiding lees
ageing, keeping barrels topped up, and aggressively cleaning new and
used barrels (if you decided to use them in the first place).
Thomson says that high pressure water washing can be effective, but
it takes many of the beneficial components from the oak away. He’s
looking at steam cleaning barrels, and is currently working on how
long this has to be done for to eliminate all yeasts.
Two more approaches to brett control are the use of a
chemical called DMDC, and filtration. DMDC stands for dimethyl
dicarbamate (also known by its trade name of Velcorin). This is
extremely toxic to microbes, but breaks down into harmless products
once it has done its job. It is now legal in the EU. Thomson
describes it as ‘quite promising’.
Filtration is another way of dealing with brett:
Thomson agrees that it is a hot topic and each wine needs to be
treated on a case-by-case basis. ‘You can sterile filter some
wines and you can’t tell the difference; in other cases the wine
can be absolutely stuffed’, he maintains. ‘You need to
understand when you can get away with filtering and when you
It strikes me that there are two different
philosophical approaches to the issue of brett. One is the inoculum
concept: that you want to avoid any presence of brett in your wines
in much the way that a surgeon will scrub up and try to keep the
operating theatre sterile. The other is the habitat concept:
assuming that brett is pretty much everywhere, you want to make your
developing wines the sorts of places where brett isn’t going to
thrive. This latter approach seems to me to be more appropriate.
Then we looked at some wines. There were two flights of Pinot Noir,
and one of Merlot. Each flight consisted of barrel samples from the
same wine which Thomson and his team felt were showing signs of
brett. He outlined the sorts of effects of brett on Pinot Noir as
the bloom developed:
Progression of effects in
Loss of primary fruit, particularly lifted violet and
sweet cherry characters. Loss of new oak characters.
Hints of smoke and spice appear (4-ethylguaiacol)
The wine begins to smell and taste medicinal
Any silky characters disappear as the wines lose flesh
The bones of the wine are exposed, leaving acidity and
rustic, drying tannins
Aromas of horse and bandaid appear (4-ethylphenol)
Tasting the wines was instructive. They varied from
being pretty bright and expressive to rather muted and earthy. But
if I’m honest, it was only because this was a brett seminar that I
was able to say with any degree of certainty that I was discerning
brett in these wines: the effects at this early stage were
relatively subtle. ‘These wines still have a lot to lose’, says
Thomson, although he reckons that as the brett bloom develops, they
will become quite unpleasant.
One final thought: if you are a winemaker, you want to
be able to spot brett at a very early stage, but an acute
sensitivity to it is not something a critic necessarily wants to
develop. The risk is that you end up with ‘cellar palate’, being
able to spot ‘faults’ at a thousand miles. ‘One regret I have
is that as a winemaker, you become less tolerant’, says Thomson.
This means that there are some wine styles he just can’t enjoy
because of the presence of ‘faults’.
Having said this, brett is something that the trade
needs to become more aware of. ‘Lots of winemakers still haven’t
grasped the complexity of brettanomyces’, says Thomson, ‘and
there’s still a bit of denial out there’. It’s also something
that he thinks everyone can get. ‘All decent tasters can pick the
nuances of brett once you tune into it.’ In conclusion, I’d
suggest that we need to become more aware of brett, and better at
spotting it in its early stages, while keeping an open mind about
wine styles where it adds complexity, and not becoming brett
policemen, always trying to spot it in whatever wine we’re
Article published November 2007