Friday, November 23, 2007

New barrels unlikely source of Brett

From an article on, we have a theory that new oak barrels are the source of Brett infections in wineries...that "infected wood" has been harvested, and that...

“[Consultant winemaker Matt] Thomson believes that the incidence of brettanomyces has increased in recent years. 'I think it's a relatively new thing in many Old and New World regions,' he said, adding that he thinks the increase is, 'partly down to new wine styles that are low in acidity and relatively high in residual sugar, but also due to increased demand for new oak barrels.' ”

Should this be a concern?
Doesn't nearly everyone use some new barrels somewhere in their production - so, potentially this is a huge issue, right?

For decades wine makers have noticed that Brett infections tend to be caught in wines from new oak barrels - but does that mean the infective yeast was already in the barrel to start with?

I don't think so....

This theory has been in the "folklore" of winemaking for quite a while - it's nothing new (weird that wine makers would employ the "it-can't-possibly-be-MY-fault" type of argument, but I guess that's just human nature...). In fact I ‘ve posted before upon the research work done which showed that 2/3 to 3/4 of wines produced in the EU were contaminated with Brett, and a healthy 50% of all wines were above 425 ppb (the generally accepted threshold for 4-ethylphenol detection for the "average" person), which shows that this is not a “new” problem. But let's delve a little deeper into this subject, and see what we come up with...

  • barrels are toasted to temperatures of roughly 200~400°F, and usually this is done with an open flame (though electric heating elements are also used)
  • the barrels are likely held at toasting temperatures for 10 to 45 minutes depending on the "house style" toast, or specific toast level requested by the winery, which should be plenty long and hot enough to incinerate any beasties on the wood
  • some toasting techniques DO release and/or create sugars from breaking down the wood, some of which may be caramelized by the toasting "style" employed
  • the process of toasting can also create low levels of 4-ethylphenol and 4-ehtylguaiacol which are two of the signature compounds of a Brett infection
  • these aromas are released in the highest concentrations during the "death phase" (decline) of the population, so detection of the aroma by tasting usually means it is too late in the cycle for prevention - rather you are finding it at the "corrective action" stage
  • Brett yeast does seem to like sweeter wines, and lower acidity levels, so there is a viticultural aspect to this problem if fruit is “overly ripe”
  • the offending aromas are linked to the presence of caffeic, ferulic and caftaric acids which are at higher levels in less ripe fruit
  • contaminated cooperage and other equipment can transfer the dreaded organism from one wine to another undetected, until a later date when the wines "stink"
  • vineyards can have a "natural" population of Brett, and equipment used to pick those blocks can bring it into your winery, as well as be a source for cross-contamination of other fruit picked with the same equipment if it has not been properly sanitized beforehand
  • infected, but yet undetected, wines can also contaminate larger blends when combined with otherwise uninfected wines

Having seen my share of Brett problems, I can testify to their variable nature (some wines are more noticeably “afflicted” by the yeast, and there is a vintage-to-vintage variation for vineyards as well – though my experience is that past offending vineyards tend to remain infected at some level, that is when the infection can be traced back to a single vineyard or block of fruit). Some varieties like syrah are much more prone to the infection, possibly due to levels of the precursor acids available, and the presence of more sugars attached to the tannins & cyanidins (color compounds). Certainly white varietals are almost never infected (detrimentally) as they don’t really contain those acids, tannins and cyanidins, and therefore can’t be used to produce the offending aromas by Brett.

While true that Brett is noticed more in newer cooperage, this may be two-fold: first, there are already produced compounds of the same nature that Brett produces which may help more people to detect the problem by raising the overall level, and second, there are as previously noted more sugars available for the yeasts to live on (these leach into the wine in the first year of use), which may help larger populations grow – and subsequently generate more of the offending aromas as those larger populations die off. Some varietals have higher levels of the base acids that Brett uses to form the aromas, so are more likely to produce noticeable “faults” later on, and also younger wines still have some sugars attached to their tannin and color complexes (called glycones) which also may be an energy source for Brett when they detach from said complexes as young wines age.

The overall likelihood that the yeast is traveling into the wineries via new cooperage is –at best- doubtful. While it cannot be discounted entirely, it would be surprising since it’d be likely that specific coopers would get reputations as producing those infections, which hasn’t happened to date (I would note that some coopers do have a rep for producing barrels with a Brett aroma from their house toasting styles, but that those barrels are not then always linked to wines which show Brett character or infections later on). Also extremely doubtful is the possible survival of yeast organisms through the toasting process – though that does not mean the barrels couldn’t be contaminated when they were being handled after the toasting as the heads were being put onto them, or during the inspection phase. Important to note that only the interior of the barrels are toasted, and while the entire barrel heats up, only the interior could be considered “heat sterilized” in my mind. The Brett yeasts conceivably could still be on the exterior of the barrel and be introduced by accident when the barrels are handled later. Again, we’d likely see a string of problems all consistently pointing back to a particular cooperage, which isn’t my experience.

Lastly, there is cellar sanitation and “traditional practices”…this is yet another of those modernity vs folklore conflicts…

There is a noted high-end winery in Napa which uses spent lees from their red wines to “paint” the bilges and bungholes of their new barrels red on arrival. Ostensibly, this is used to make an even covering of the barrels, and add a little visual character to their cellar (I have one report that tourists were told the practice “illustrated the winery’s commitment to pay attention to each and every barrel in its care”, whatever that means…). Unfortunately it results in each and every barrel so treated to be twice risked for further infections: once by the fact that other wines which were contaminated but not detected with Brett may now be literally spread all over the new cooperage, and twice by the fact that even if clean wines were applied to the new cooperage, the barrels would now have a food source available for any mold/rogue yeast/what-have-you organism to colonize the barrel and possibly get transferred not only into that barrel itself, but also across to other barrels and blends, etc, as Providence might desire…

It would also not come as a surprise to regular readers that this same winery in the example has been panned by me before, due to the high levels of Brett and other organisms which consistently run through their products.

Labels: ,


Anonymous Anonymous said...

Matt Thomson responds!

A couple of comments I'd like to make! Firstly the focus of the tasting presented was to show the many forms that Brett (Dekkera) can have in a wine. The comment about Brett being linked to new barrels was a small part of the presentation resulting from a question. I was also mis-quoted as saying it may have been due to higher rs wines. In fact I said that with higher harvest Brix levels, frequently winemakers transferred wines to oak with residual sugar rather than when they were dry. We are well aware (or should be) how much Brett loves hexose sugars.

The comment about new oak bringing Brett into a cellar is not without thought however. In my experience even people who think they know about Brett usually only know about the classic 4EP horse/Band-aid characters. Very few are aware of the 4EG characters, esterase effects etc. When it comes to Pinot Noir it is crucial that tasters are aware of this as 4EP is only formed very late in the piece. The AWRI have also found this in agreement with our organoleptic experience. Over the last decade we have become very aware of its many forms tasting thousands of barrels as a panel. There are a number of things that have come from this.

There does seem to be a correlation between Brett infection and cooper. We have some batches of 50 barrels+ where the only barrels showing Brett infection are new barrels from one cooper, while none of the new barrels from other coopers show it. There is one cooper this year where across all batches of wine in three wineries, 100% of the barrels show Brett to some degree, while many barrels in the same batches don't.

I suspect we are the only group going to the extent of cross-contamination infection prevention that we do. We use a different pipette to sample each barrel for instance (these are autoclaved between). All stirrers etc are sanitized between barrels.

It is of no surprise to me that you say that coopers would get a reputation for it if it was the case. Sadly in most wineries once it is in a batch it usually quickly spreads between barrels due to the absence of the above techniques. This quickly covers the source of the inoculum.

I can also give you the year that a number of Burgundy domaines and Barolo estates first showed Brett in their wines. In the case of Barolo it doesn't seem to have been assoiciated with the Slovanian oak previously used. Interesting huh?!

Your comment about the toasting process probably killing (sterilising) Brett (how about spores?) is something we have considered. There are a few things to think about here too. Firstly Brett has been found 8mm deep in staves, and bear in mind that there has been very little work done in this area. We have found that lower-toast barrels seem to have Brett more frequently if you go close to eliminating cross-contamination. Now it should be the opposite shouldn't it?! Toasting increases the desirable substrates for Brett. Unless of course the toasting process is knocking back a resident population in the oak itself?

My concern is who is looking for Brett in new oak? Certainly not the coopers as it is against there interests. To properly eliminate it as a source a large number of barrels would need to be soaked and the resulting liquid filtered and cultured. I am as yet unaware of this having been done. I am aware of several wineries having found it in new barrels through basic swab cultures as part of a QC programme. Now I'm well aware with my background that this doesn't qualify as published research, but in the absence of the latter in combination with a vast amount of "circumstantial evidence" I belive we are quite right to be wary of new barrels in this context!

November 26, 2007 4:05 PM  
Blogger St. Vini said...

Matt, many thanks for the reply!

Bummer to hear that you have a specific cooper that has an entire run which may have been Brett infected....
...then on the other hand, at least you have a potential culprit you can watch more closely in the future.

I do know of some wineries here in California which do use strict isolation techniques to eliminate cross-contamination between wines on a production level.

Actually, my experience is that lower toasted barrels would be more likely to have a higher sugar level if the temp at which caramelization hasn't been reached or held for very long. That might leave a barrel with more available sugars than a higher toast barrel.

I have seen research that Brett can colonize deeper in the cracks of wood than was previously thought, but thankfully I haven't seen that to be a problem up here. Of course the worst part for you if that is the case there, is that it could only take one infected stave in a barrel to make it go south. In that manner, one infected tree or lot of aged staves in the cooper's yard could imaginably cause a wide-spread problem. Again, this isn't something I've ever seen or heard of here in the states.
Was it a local cooper of yours which was found to be the offender?

If the barrels are toasted to a 6~9mm depth, I imagine there would be little possible yeast activity after it, as the enzymes in the organisms would have been deactivated if the yeast itself hadn't been directly consumed in the conflagration - though I say that without an actual study to support that view, just my experience.

I'm not sure what your experience with ozone generators is, but would gladly tell you that I picked one up year-before-last, and haven't had a problem reusing cooperage from previously infected lots in other wines later after having cleaned the barrel with ozone.
It may make a nice additive to your prevention list to wash/treat the barrels from the suspected cooper with ozone upon receipt as a cautionary move.

Good luck with your fight against Brett, it's an organism I would love dearly to see less and less of!


November 27, 2007 6:01 AM  
Anonymous stuart said...

I don't think most barrels are toasted to that great of depth, maybe only 4 mm depth into the wood, but a good point all the same.
If the barrels are at 300+°F for any length of time I would assume that process would kill anything on it. Has there been any work done to see how far the heat actually penetrates into the wood?
We sterilze many things here with a 180°F water bath for 30 min., is there data to demonstrate that this is also the case with the staves of a barrel? I can't seem to locate any with a quick web search.....
Plenty of hits for toasting and temperature, but nothing obvious about the penetrating depth of the heat.

And contrary to the other opinions, I think it is in the cooper's favor to try to avoid any problem that may be due to infected wood exiting their facility. If I were the copper making the barrels and a client came back with a problem like that, saying they tracked it all back to my barrels, I'd put every ounce of effort into making sure it never happened again. A reputation of letting that sort of thing happen would surely kill any business I had to start with.

Ozone is a great help, and hasn't sacrificed any of the oak flavors, which I was very concerned might happen before I started to use one.
In fact I now use ozone on all my barrels, red or white.

November 27, 2007 9:39 AM  
Anonymous Anonymous said...

Matt again:

Sadly ozone is pretty much completely ineffective in barrels. Because of it's short half-life it really doesn't have any effect below the surface. It is great for relatively smooth impenetrable surfaces such as stainless steel. We have had quite a few years work now with ozone and certainly wouldn't recommend it for this use.

I'd be interested to hear of the extent that other wineries are going to prevent cross-contamination as few wineries are aware of just how far you need to go in this respect. I'm certainly very interested in hearing what tricks others have developed!

Steam is far more promising, but we are yet to determine the time required to reduce the Brett to levels where there is no organoleptic impact from previously infected barrels. Unless you are going to autoclave the entire barrel, complete elimination of Brett is a dream.

I'm interested in your comment about Brett not penetrating into the wood where you are, as there seems to be very little research in this area. Can you please send me the data you have in this respect? How has this been assessed?

Here are some further comments for thought. Coopers are not going to come out and say that they have found Brett in their oak..... for obvious reasons (similar to cork producers not rushing to say they have found tca in their corks!).

Any research that looks at new barrels supplied in the standard sulphured condition is going to be somewhat compromised. Recent research indicates that SO2 shocks Brett into a viable but non-culturable state. Once the SO2 level drops they then become fully active. It certainly is one very sneaky beast! Let's assume that the toasting process will kill all surface Brett within the barrel (and I think that is a safe assumption). How are you going to find it by random surface assessment of SO2 treated, toasted new barrels? I think you need to go to the staves themselves. There are some coopers who incorporate a soaking process. That volume of washings would be a good place to start...if anyone will let you look!

There is no way that anyone can state with confidence that the toasting process kills all Brett to penetrable levels. From a microbiological point of view the process is not rigorous or consistent enough for that to be the case.

Sadly again, while we have one cooper where the problem is pronounced we have had others where we have found it in circumstances that are difficult to dismiss as subsequent contamination.

The whole point of where this all started was about the complexity of the Brett issue. I don't think it is all bad in all wines. There is a certain positive effect it can have on some varieties where the primary fruit flavours are stable to the esterase produced by Brett. In these wines the integration of oak and fruit and some subtle added complexity is an asset. However I would maintain that any wine where Brett is the dominant feature is damaged by its presence. I'm also of the opinion that the esterase effect in Pinot Noir and Nebbiolo is simply too high a price to pay. It isn't so much what it adds as much as what it takes. The losses of the magical violet or old fashioned rose aromas brings tears to my eyes as both a passionate consumer or anally retentive winemaker.

November 27, 2007 4:34 PM  
Blogger Wine Limo said...

Nothing is perfect in disinfection. What is quite apparent is cleanliness is most important. To use clean instruments between one barrel and another is an excellent practice. Toasting seems to help as it treats the wood. Ozone although it does not seem effective for the wood would certainly be effective for the spores.
Another good point is to not transfer wine that has not finished into wood until it does. Acidity has long been an important factor in wine stability. Old found rules often have more than one reason why they should be followed. Over ripe fruit of low acidity and high sugar has long been a source of condemnation for old school advocates!

December 06, 2007 8:04 PM  

Post a Comment

<< Home