Twenty years after the phenomenon started causing fear and frustration in so many fine wine collectors’ hearts, there is still no clear-cut answer, no single-bullet theory about the exact causes of the premature oxidation that affected so many bottles of top white Burgundy. The debate and the reasons that have come to the surface, however, do tell a cautionary tell that should be taken into account by all makers of white wine.
One of the main aspects of the “premox” crisis—and one of the reasons it was so feared by collectors of fine wine—is the random character of that premature oxidation. Why would one bottle be fine and the next from the same lot be decrepit and tired?
Bottling and Closures
For Luc Bouchard, head of Bouchard Père et Fils, one of the most important houses in Burgundy, there is no question that the handling of wine in the cellar, especially at bottling, is an important culprit in premature oxidation issues. “If one bottle is awesome, another is just okay and a third is dead and gone, then obviously there are issues that have to do with bottling and closures and such,” he points out.
Winemaker David Croix, from Domaine Camille Giroud and Domaine David Croix, in Burgundy, also believes that premature oxidation is largely linked to bottling. He points to Burgundy’s often small scale production, especially with high-end wines: “Many producers work with very small parcels in small crus or appellations, he says, so in the cellar, they often work with very small batches, as small as one, two or three barrels.”
Croix points out that this presents particular challenges at bottling time, given bottles filled early or late during the run. Going back to the mid-1990s, in particular, when the problem arose, very few people were bottling under neutral gases, and attention to oxygen exposure during those steps was not looked at terribly carefully.
Beginnings and Endings
When a bottling run starts, if no neutral gas is used, then the whole system is filled with air, meaning that the first wine that comes in will meet with high amounts of oxygen. By contrast, the bottles in the middle of the bottling run sees less, while the ones at the end of the run will once again see more because the wine pulled from the tank or barrel has been in contact with more air in the container itself, and may be exposed to even more as the very last drops of wine go through the bottling system. A bottling study conducted by Nomacorc in California has shown that a “U-shaped” curve of oxygen ingress in bottling runs (with more oxygen at the beginning and end) is sadly a regular feature in many wineries, meaning variability in the way individual bottles will age post-bottling. The study also clearly showed that a large portion of the oxygen trapped in a bottle is actually present in the headspace, so that simple measurement of dissolved oxygen is not sufficient for optimizing bottling conditions.
Add to that variations in the permeability of natural corks, which have been well documented in several studies and trials by Nomacorc and others, and there are even more reasons for bottles showing up differently as time goes by. A few extra milligrams of oxygen per liter can mean the equivalent of an extra year or two of aging, since the amount of oxygen ingress is so small post-bottling.
Such variations should always be avoided, but they are certainly more apparent and more important in short bottling runs than longer ones, creating more variability within the lot. Let’s say that two or three cases are significantly affected at each end of a run—that is a higher percentage for a run of two barrels’ worth, or about 600 bottles, than for a run of 10,000 bottles. And lo and behold, shorter runs often go hand in hand with the high-end Burgundies that are most prized, as a producer will often have only a few rows of vines in a prestigious grand cru. “The tendency to fragment bottlings into very small lots has its risks”, says Louis-Fabrice Latour, head of another major Burgundy producer, Louis Latour, pointing out that working with ten or twenty barrels is always easier, in his opinion, than with two or three.
What About Before Bottling?
It’s hard to blame bottling practices and short runs for all the woes of prematurely oxidized Burgundies, however. Before the 1990s, most producers were likely hand-bottling high-end premier crus and grand crus, and certainly all of them used natural cork since other solutions were essentially unavailable before then. So what gives?
Many producers point to the fact that after the late 1990s, Chardonnay grapes have tended to be picked earlier in Burgundy, so they would keep better acidity. A lower pH is indeed an important component of oxygen resistance in wines, and pushing ripeness too far is not a way to increase cellaring potential, as the pH can rise rapidly as maturity increases. “In general, we can see white wines from 2009 evolving faster than those from 2008, says Louis-Fabrice Latour, and that goes with the acid levels.”
Latour also points to higher yields as a possible culprit as well. Latour has cut down on yields in its top vineyards over the last couple of decades, and results have been positive in terms of ageability, he says. This only seems logical: If the concentration of a wine is higher, as it should be with lower yields, then whatever components help a wine resist oxidation should be present in higher concentrations as well. Of course, not all components present in must and wine are beneficial with regard to oxygen resistance, but careful winemaking and proper aging should ensure that the right stuff is present in the finished wines.
Luc Bouchard also expresses the idea that earlier picking has been beneficial for reducing the occurrence of premature oxidation in white Burgundies, but he also points to more careful cellar management and the avoidance of certain practices which had become more popular in the 90s. In particular, he points to bâtonnage—the stirring of the lees in barrel—which was done frequently to build fatness in wines, and which has been essentially eliminated at Bouchard Père et Fils nowadays. Although the effect of lees stirring on oxygen resistance is a complex matter, it seems likely that opening barrels and stirring wine more regularly is likely to introduce more air into a wine overall.
Another element seems to be SO2 additions, which had tended to go down in the 1990s. Lower sulfite levels and higher pH levels can certainly be a deadly combination for many wines, since wines consume more SO2 when their acidity is lower, meaning that less sulfites remain available to actually protect the wines against oxygen over time. With insufficient control of bottling conditions and the effects of inconsistent closures, it’s easy to see how this would become yet another ingredient for bottle-to-bottle variation.
Today, between closures allowing controlled and consistent oxygen ingress and new tools to analyze dissolved and headspace oxygen (or even the phenolic composition of a certain wine), there are many ways to minimize randomness and variability in wines, and to adjust practices to the actual characteristics of a given wine. Producers like Domaine Laroche, in Chablis, have invested significantly in recent years to ensure that their SO2 additions are well adjusted to the requirements of the wine, in good part by carefully measuring things like dissolved and headspace oxygen levels.
Whatever the complete set of factors involved in premature oxidation, something that should be rather obvious has to be taken into account as bottles are being tasted and evaluated: It’s normal for wine to change over time. Premature oxidation is a real problem, but as an article by Burgundy expert Clives Coates points out, it’s possible to be overly worried about it and to confuse aging with the “pox,” as it’s sometimes abbreviated:
But a story from Dominique Lafon. In April 2011 he found himself in a bistro in Burgundy. His meal was interrupted by a stentorian (American) pronouncement—’Prematurely oxidized’—from the other side of the room. He looked up and saw that the wine was his. He called the wine waiter over. The wine was his basic Meursault 2004. He tasted it. Not a bit brilliant. And now aged. But not ‘prem-ox’. Just a bit old.
Lafon went to the table himself and offered a bottle of a higher-end wine from a younger vintage to make sure the customers were happy. In wine tasting, a lot goes to expectations: Is it possible that some tasters, having been made aware of the perils of premox, could tend to see it everywhere?
If you are tasting a 10-, 15-, or even 20-year-old white wine, and its fruit has lost some of its freshness, its color has darkened a fair bit, and its aromas seem closer to old parchment than bright flowers, it’s good to keep in mind that evolution is the norm, not the exception. Not all wines and vintages, even in the most prestigious appellations, are meant to be kept forever. A bottle that sings after four or five decades is a magical thing, and that may well be the case because it can only be something rare.