Carbon dioxide is no stranger to winemakers: It floods the cellar during fermentation, serves as a blanketing layer in tanks to keep oxygen at bay and makes sparkling wines sparkle. But in still wines, at the sub-bubble level, it doesn’t get much respect.
That is a shame, since the level of dissolved carbon dioxide in the bottle can have a major impact on how a wine hits the palate. Too little can render white wines flat from day one; too much can make reds harsh and tannic. And since the level of dissolved CO2 is significantly affected by numerous environmental variables—temperature, frequency of racking, sparging with other gases—it can easily go too far up or too far down without the winemaker doing anything directly about it.
The basics of carbon dioxide’s sensory role have been well established for some time, and methods for measuring and managing it are easily available. But except for the mega-scale industrial wineries that monitor and tweak nearly every molecule in their vats, CO2 isn’t high on most small North American producers’ checklists, if it’s on there at all.
The level of dissolved carbon dioxide in wine can have a significant sensory impact, but it gets relatively little attention from most small producers.
- Different varietals and wine styles work best with different levels of CO2.
- Testing for at least ballpark CO2 levels can be helpful to winemakers, but always in conjunction with careful tasting.
Maybe it should be, since these often-overlooked leftovers from fermentation can pack a punch.
The gospel according to Peynaud
The reference text for understanding the sensory impact of dissolved carbon dioxide in wine is French enologist Emile Peynaud’s 1983 classic, “The Taste of Wine.” In one context after another, he presents CO2 as a double-edged component capable of making wine sing or making wine bite. In the proper balance, Peynaud argues, CO2can lift the aromatics and add freshness and verve, especially to whites; in big reds, on the other hand, even much lower levels of CO2 can make a wine hard, acidic and overly tannic. If dry white wines are too low in CO2, they start life dull and quickly get duller; but for sweet whites, lower levels of CO2 enhance richness.
In all these cases, the presence of CO2 ups the perception of acidity—and well it might, since CO2 in solution takes the form of carbonic acid. Dissolved CO2contributes to titratable acidity, but at the levels present in non-sparkling wines, Bruce Zoecklein of Virginia Tech thinks it isn’t likely to have any significant effect in restraining pH or lowering the dosage of free SO2 needed for protecting wine from spoilage. Yet even if it affects perception more than chemical balance, the level matters.
Peynaud also identifies at least some of the environmental factors that can influence whether CO2 stays in wine or heads off into the atmosphere. Temperature is crucial to solubility; the colder the temperature, the better the retention of carbon dioxide. Think of the nonstop chill necessary for making Moscato d’Asti, or the flatness of warm British beer. And so it is the amount of motion and agitation—anything from racking to swirling your glass—that pushes carbon dioxide on its way.
Environmental factors and dissolved CO2 do their dance through something called Henry’s Law, formulated in 1803 by chemist William Henry, which holds that the solubility of a given gas in a liquid at a given temperature is directly proportional to the pressure that same gas exerts on the liquid from outside. CO2 inside will inexorably balance CO2 outside, but that balance point depends mightily on temperature and other variables. At cool, jacketed-tank temperatures, CO2 is highly soluble—much more so than oxygen, nitrogen or argon; at warmer barrel-room temperatures, the concentration plummets rapidly.
Following Peynaud and other researchers, the ballpark numbers are pretty well known. Wine emerges from fermentation with about two grams per liter of dissolved CO2, and it declines from there. At 500 milligrams per liter, the presence of CO2 is noticeable; at 1,000 mg/L, there is a slight perception of prickliness. The textbook recommendation is that age-worthy reds should be bottled with no more than 100-200 mg/L; light, fruity reds could benefit from about 500 mg/L, and whites, depending on stylistic intent, might range anywhere from 500 mg/L to 1,800 mg/L, from slightly punched up to noticeably spritzy.
And oh, yes, at a tad under 4 grams per liter, which makes the wine officially sparkling, the U.S. tax rate goes up.
There are four basic methods for measuring dissolved CO2. At the high end, Anton-Paar and Hach both offer monitoring systems in the $12,000-$13,000 range, with extra bells and whistles for a few dollars more. Both instruments sell primarily in other beverage sectors (beer, soda, etc.) but have a niche in large-scale wine processing as well. Hach’s Orbisphere 3658 relies on measuring thermal conductivity, while Anton-Paar’s CarboQC infers CO2content from repeated volumetric expansion. Both are highly accurate, but with daunting price tags for smaller producers.
Next in line is a form of titration, the carbonic anhydrase test, which by every account is a colossal pain. Enology consultant Lisa Van de Water of Vinotec Napa describes the chemicals involved as “very nasty stuff,” and she should know, having done about as much nasty lab work as anyone in the industry.
Much easier and more common today is a hand-held gizmo known as the Carbodoseur, made in France by the Laboratoires Dujardin-Salleron—though the usual stateside pronunciation is a very un-French Carbo-doser. Enartis/Vinquiry in California distributes the Carbodoseur, which sells for a modest $260, and they sell a l ot of them.
The method of analysis starts with a measured volume of wine, which is then shaken and re-shaken (sort of like a cocktail mixer) until all the CO2 has de-gassed itself into an outer chamber; the shrunken wine volume is then re-measured, and a handy chart translates the results into milligrams per liter of CO2. From the instructions, it’s clear this method is subject to considerable user variation—temperature, the angle of the shaking, etc.—and not likely to match the fancy setups in accuracy. But for a quick, rough reading, maybe within 100 mg/L of the “true” level, it’s fine. And that may be all that most people need. Vinquiry, in fact, uses it for its own CO2 testing.
Finally, there is a time-honored measuring device, and the price is right: the human palate. For experienced tasters, a sample of wine may not reveal exact numbers, but it does indicate whether a wine is too spritzy, too flat or right in the winery’s target zone. If that prickle on your tongue is a drag, sparge with some nitrogen, which will alter the chemical state of the CO2, thus invoking Moore’s Law and pushing some carbon dioxide out. If your stainless-steel white seems unusually lifeless, sparge with CO2 and revive it.
Sure enough, method No. 4, the human mouth test, dominates the small- and medium-sized wine industry. Lisa Van de Water says that testing (she relies on a Carbodoseur) is a fine thing to do and can help with consistency from year to year, but taste is the real test. “If you sell all your wine, if people seem to like it, if it ain’t broke, don’t fix it.”
Enough of me; let’s hear from a random sample of small winemakers in the carbon dioxide trenches.
Steve Pessagno, owner/winemaker at 9,000-case Pessagno Winery in Monterey County, Calif., has an engineer’s head on his shoulders. Turns out that when he was the production winemaker at Lockwood Vineyard, making hundreds of thousands of cases of wine per year, monitoring CO2 levels via Carbodoseur was part of the standard routine. Now that he makes only a few thousand cases of ultra-premium Pinot Noir, Syrah and Chardonnay, he follows his palate, including tasting wine at both room temperature and cellar temperature, which can be two entirely different experiences. He doesn’t measure dissolved oxygen, either. “We get the same results year after year by doing things the same way, over and over,” he says.
Eleni Papadakis at Oregon’s 15,000-case Domaine Serene says they do consider CO2, but it isn’t a focus. “We try to preserve what’s in the grapes, but not with intervention. Excessive use of nitrogen to reduce oxygen can strip out CO2, too. And we try to minimize motion.” With no particular target level established, “we don’t measure, because we wouldn’t take any action.” The winery does have a Carbodoseur, occasionally employed for “edification.”
Not far away in Oregon, the Beaux Freres winery has an equally hands-off policy when it comes to CO2. And according to marketing director Kurt Johnson, the combination of a cold cellar and virtually no racking means that their Pinot Noir naturally arrives at bottling time with a slight petillance; on occasion, in fact, writers have speculated that the Beaus Freres wines were undergoing malolactic in the bottle, which they weren’t. “If you’re drinking the wine under five years from harvest,” he says, “just give it a little splash for air and it’ll open up just fine.”
I also found some winemakers who routinely measure carbon dioxide levels, though taste is still in charge. Andrew Hedley at the Framingham Wine Co. in Marlborough, New Zealand, says, “We have sort of calibrated ourselves against the numbers over the years; however, the effect/feel in the mouth and maybe on the nose is more important than the numbers.” He has developed target ranges for different wine styles: fairly elevated (1,300-1,600 mg/L) for their “Germanic” Rieslings; closer to 1,000 mg/L for drier Riesling; under 850 mg/L for their richer, weightier, Alsatian-style Pinot Gris and Gewürztraminer; and Sauvignon Blanc somewhere in the middle with 1,100-1,400 mg/L. While these targets can be reached by sparging CO2 in or out of the wine at bottling, Hedley says by far the best approach is to carefully conserve what’s left over from fermentation.
Riesling is probably the variety whose CO2 content gets the most attention worldwide, and certainly in New York’s Finger Lakes. Sheldrake Point’s Dave Breeden is among the carbon-conscious winemakers in the region, especially for dry and off-dry all-stainless wines. Breeden’s concern isn’t getting the CO2level up high enough for the signature Riesling verve; it’s getting it down a notch. “Even with all the processing the wine goes through—Bentonite, several forms of filtration, sterile filtration, bottling at over 60ºF—we don’t remove enough CO2”—by which he means getting it down to around 1,000 mg/L, measured by his trusty Carbodoseur. “If I wanted sparkling wine,” he says, “I’d make sparkling wine. I don’t like my wine to move around in the glass.” He usually needs to sparge with nitrogen before bottling.
David Whiting at Red Newt Cellars, 20 miles down the road, hasn’t yet gotten very far into the carbon dioxide measurement and management business, but he thinks he should. From conversations with Breeden and Peter Bell at Fox Run—and for that matter, from looking at the little ring of bubbles on some young German Rieslings—something he doesn’t want—he has gotten to a point he describes as “on the cusp of managing it more accurately.”
Breeden, however, has a cautionary tale: Last year he brought one particular wine down from the 1,400 mg/L or so that’s standard in his cellar to the 1,000 mg/L he prefers for bottling, “and the wine was horrible, completely stripped. I had to put some CO2 back and do some blending. Sometimes, I’m still in the dark about this stuff.”
Hey, nobody said winemaking (even mostly non-interventionist winemaking) was easy. Bruce Zoecklein says all these concerns are cyclica l; carbon dioxide was “a big deal maybe a dozen years ago, then it dropped off the radar, especially for small producers.” Maybe it’s due for a comeback.