Batonnage

Batonnage is the process of stirring the fine lees remaining in the barrel of unfinished wine after the initial settling (debourbage.) While some producers do not engage in any batonnage (allegedly this group includes Leroy/d'Auvenay, Jadot (in 1996 only), Remi Jobard, Latour, Ramonet and perhaps Matrot), the great majority of white burgundy producers engage in batonnage to at least some degree in vintages where the lees are healthy (i.e. free of rot, mildew, botrytis or other problems affecting the grape skins.) Many positive results are claimed for batonnage by its proponents, including that it “enriches” the wine with better flavors, promotes reductive as opposed to oxidative reactions in the barrel, allows the wine to fall clear, etc. Those who oppose using batonnage at all, aside from oxidation concerns, contend that it can induce “off” or “animal” aromas in the wine or that it tends to blur terroir differences and to “make all wines taste the same.”

In its traditional format, batonnage is practiced by removing the bung from the barrel and inserting a device to vigorously stir the lees located at the bottom of the barrel to re-mix them with the liquid. The device used for batonnage can vary from something as unsophisticated as a broom handle to a device with motor sitting atop a long stem which turns two wing-like devices at the end of the rod. Batonnage can also be performed by rolling a barrel without removing the bung. The latter method of batonnage, at least until very recently, has been relatively rare and Bouchard was the most well-known user of this method. John Gilman, author of the The View from the Cellar, reports in his July-August 2007 issue that the use of batonnage is now diminishing and that a greater number of those who do batonnage are now rolling closed barrels to accomplish it. There are also new barrel racks which hold barrels and permit spinning the barrel to conduct batonnage.

Batonnage as traditionally practiced is inherently oxidative. Batonnage “degasses” the wine in the sense that the stirring stimulates the release of free SO2 and CO2 trapped in solution in the wine (which help protect the wine from oxidation), and, due to the open bung and the process of stirring, simultaneously adds oxygen into solution with the wine and into the remaining space in the barrel not filled by liquid. While Pierre Rovani and Jean-Marie Guffens have claimed that batonnage is a reductive rather than oxidative process (allegedly because the lees themselves are allegedly reductive), this seems obviously to be contrary to scientific fact. While winemaking has sometimes been referred to as a process of controlled oxidation, it seems clear that the more batonnage that is done, the more dissolved oxygen that will added to the wine (even during the ML fermentation process when the wine is producing substantial quantities of CO2 from the fermentation process) and the more acetaldehydes that will be formed. While it is intellectually appealing to assume that there must be a linear relationship between batonnage frequency and perceived oxidation, there are no known empirical studies on this subject to date. Moreover, the oxidative effects of batonnage may be offset to some degree by maintaining SO2 at higher levels (although anecdotally this does not appear to have been done during the period when the 1995-2004 vintages were vinified.)

The batonnage regimens in use in Burgundy vary considerably. Some producers stir the lees during the primary barrel fermentation. Some stir during the period between the completion of the primarly fermentation and the start of the malo-lactic (ML) fermentation. Others stir during the ML fermentation. Still others (though apparently a minority) continue to stir the lees for some weeks or months after the ML fermentation is completed. Some producers stir the lees during more than one of these periods and, of course, some producers do all of the above. Vintage conditions can also directly affect the amount of batonnage done. For example, for those producers whose custom is to stir the lees through the completion of the ML fermentation, in a vintage such as 1996, where the ML fermentations lasted longer than normal, by definition more batonnage would occur.

Stirring frequency is another variant. After the primary fermentation is completed, some producers report that they stir the wine a few times before the ML fermentation is completed. Others stir the lees as many as three times a week through the completion of the ML fermentation and, in some instances, for up to 15 months from the date of the harvest. Some producers follow a standard recipe for batonnage. Others vary the frequency and length of the lees stirring according to varying vintage conditions. Information on the individual batonnage regimes of the various producers, to the extent known, is set forth at the top of each page describing the wines from each producer.

Batonnage was done extensively in the 1995, 1996 and 1999 vintages. It apparently reached its zenith in the 1996 vintage because of the very high yields, the presence of ultra-clean fruit that would supposedly permit a lot of batonnage, and the very high acid levels inherent to the vintage. Many winemakers reported to Steven Tanzer, Pierre Rovani, and others that they thought that they needed to “enrich” their wines via batonnage as much as possible because of the high acids and high yields. Many producers, including Coche-Dury, admitted to having done batonnage “like never before” in 1996. On the other hand a few producers like Lafon reported having done less batonnage in 1996 than in preceding vintages due to oxidation concerns. Some producers, like Jean-Marie Guffens of Verget even publicly proclaimed that they were doing more batonnage in lieu of adding SO2, asserting that batonnage was an effective substitute for adding sulphites.

According to John Gilman, subsequent to the 1996 vintage some burgundian producers, including Coche-Dury and Lafon, substantially reduced or ceased doing batonnage after they saw what happened to the 1996 vintage. Verget has similarly changed its practices. Jean-Marie Guffens has acknowledged his prior winemaking errors resulted in prematurely oxidized wines from the 1994, 1995 and 1996 vintages and has stated that starting with 1997 he has added more sulfur and now restricts the amount of batonnage and the time in barrel. In 1999 Guffens claims that he did no batonnage. Others, such as Bouchard, changed their method of doing batonage – henceforth simply rolling the barrels back and forth to mix the lees with the wine, instead of repeatedly opening the barrels in order to stir the lees. Le Moine “stirs” the lees by injecting CO2 gas into the bottom of the barrel via a tube rather than stirring. But the change in batonnage regimes by the likes of Coche-Dury, Lafon, Verget and Bouchard was by no means universal. In 1999, with its very high yields and healthy grapes, many winemakers again saw a perceived benefit from doing extensive batonnage to “enrich” the wines. Some producers (e.g. Sauzet and Pierre Morey) actually increased their use of batonnage, or further extended the period of stirring, versus the 1996 vintage. Indeed Jadot, who claims to have done no batonnage on the 1996 vintage, engaged in batonnage in 1999.

In the opinion of some burgundy collectors/drinkers, (including Don Cornwell who compiled this section), there is an apparent correlation between those producers that engaged in the most extensive use of batonnage, particularly in the post-ML completion period, and premature oxidation (e.g. Verget and Sauzet). Gerard Boudot, the wine maker at Sauzet, recently admitted that the extensive battonage and failure to properly sulphite the wines during this oxidative process was the cause of the premature death of his 1996 vintage. Quoting from Issue 134 of the International Wine Cellar (Sept/Oct 2007), Mr. Boudot stated: “In '96 in particular, the malos were long and I did a lot of lees stirring. Because of the late malos I did not dose with SO2 until it was too late.” According to John Gilman, in his July-August issue of The View from the Cellar, Mr. Boudot has recently substantially curtailed his use of batonnage. Conversely, some of those producers who used little or no batonnage seem to have very few prematurely oxidized bottles (e.g., Leroy and D'Auvenay).

But batonnage is clearly not a complete answer because at least two producers who apparently did no batonnage in 1996 (Jadot and Ramonet) have significant incidence of prematurely oxidized wines from the 1996 vintage. In the case of Ramonet, the estate blamed the oxidation observed in several wines from the 1996 vintage on cork problems and perhaps residual peroxide remaining in certain cork batches at the time of bottling (however Ramonet has continued to have oxidation issues in the 1999 and 2002 vintages as well). In the case of Jadot it is most likely the “ultra low” usage of S02 bragged about Jadot's winemaker.

[Text and opinions by Editor Don Cornwell. © Don Cornwell 2005-2010]

I wrote my Master of Wine dissertation on the use of oak in winemaking and in my researches I learned that lees stirring is essentially reductive - it changes the redox potential of the wine. It enables winemakers to use less SO2 as a consequence. The assumption that it must increase the level of oxygen in a wine due to “de-gassing” does not follow; the release of SO2 and CO2, if true, would actually work to prevent oxygen ingress to the wine - CO2 would settle on the wine's surface being heavier than air) and the SO2 would scavenge some of the oxygen whilst being released.
Most producers in 1994 did not stir the less due to the low acidity of the wine as lees stirring helps to encourage the malo-lactic fermentation. In a year such as 1996 with high-acid wines lees stirring would have been carried out extensively as a means to help soften acidity in the wines.
I am not aware of the problem as stated as I do not taste enough white Burgundies but I would be far more inclined to blame faulty corks (very very little oxygen passes through any cork) or else lower SO2 levels than usual. Wines do not actually oxidise as they age in bottle - the ageing reactions take place only at the wine/container interface (see David Bird MW's book on wine science) - but deepening of colour and changes in aromatics are normal. Any dry, white wine of age will deteriorate noticeably if opened and poured for a few hours.
I don't wish to in any way suggest that there isn't a problem but I'm not certain, from the descriptions mentioned, that oxidation is the problem. I hope this helps in some way.
[Dermot Nolan MW - nolan_mw@compuserve.com]