Checklist for TCA Clean really well and use a peroxide-based product. Avoid chlorine. Use hot water. • Don't use chlorine anywhere where the liquid or vapors might contact wood, wine, or paper products. Don't use chlorinated TSP. • No wood pallets in the winery. • No cardboard near barreled wine. No wood palates in any combined space, they may bring in either precursors or chloroanisoles. Do not bring anything with the precursors into the winery and close the doors. It can change and then volatilize into the atmosphere. "If you are topping, racking, you can, for example, have a pickup from storing your cardboard bases next to your barrels." • Regularly have analysis of the water supply and the atmosphere in the cellar, caves, or winery buildings performed using GC/MS. • Test the composite of your wine for TCA prior to bottling. • Ventilate: Increase the ventilation of all confined spaces to prevent accumulation of chloroanisoles. • TCA can be formed from any wood product: This includes open bags of corks. wbm |
When Hanzell Vineyards, the Sonoma Valley producer of Chardonnay and Pinot Noir learned it had an issue with "taint" in some wine, the winery gathered as much information as it could and quickly got out in front of the issue. It's been a learning experience. Lessons learned concern sensory perceptions of TCA and the nature of TCA itself.
"This is an issue that wine producers must confront because, in one way or another, it comes with the territory," said Jean Arnold, president of Hanzell Vineyards. "Hanzell is a small winery, and we had a mighty struggle on our hands in understanding and dealing with this quickly."
The chemical 2,4,6-trichloroanisole (TCA) is produced by the metabolism of trichlorophenol (TCP) by a broad range of molds. These molds are pervasive and ubiquitous. They can grow in and on any wood or paper product. Like most fungi, they require a fairly moist environment like a winery. Trichlorophenol may be formed by the reaction of chlorine with phenols including wood and wood-products. In Hanzell's case, the primary source of taint was not a bad batch of corks but was an old drain in the fermentation room. The evidence isn't conclusive evidence, but Arnold suspects a couple cellar workers in their eagerness to clean well may have used a little more chlorine than usual to scour spots off of the winery's cement floor (The winery has switched to using a peroxide-based cleaner).
An Unexpected Discovery
Wine Spectator columnist and critic James Laube contacted Hanzell last summer after his palate detected "taint" in blind tastings of the 2000 Chardonnay. There hadn't been complaints, said Arnold, not even from the customers at a San Francisco restaurant that had just consumed 30 cases of 2000 Chardonnay at $24 per glass.
Laube's suspicions were confirmed when Hanzell had 25 samples of the 2000 Chardonnay and 2001 Pinot Noir vintages tested at ETS Laboratories of St. Helena, California: Results showed TCA levels of 2.6 and 3.2 parts per trillion, respectively.
Hanzell briefly halted sales of the wines. "We hit the pause button," said Arnold.
After extensive consideration and after consulting with flavor chemist Leslie Norris to test consumer TCA rejection thresholds of its wines, Hanzell re-released the 2000 Chardonnay.
Hanzell is expanding to an annual production of about 5,000 cases. Fortuitously, the winery was in the midst of building new production facilities with new caves when the taint was discovered.
Checklist for TCA
Clean really well and use a peroxide-based product. Avoid
chlorine. Use hot water.
• Don't use chlorine anywhere where the liquid or vapors might contact wood, wine, or paper products. Don't use chlorinated TSP.
• No wood pallets in the winery.
TCA is a relatively common (5-10 percent of all cork-sealed wines) taint that is found in both white and red wines. It is well known to influence the flavor/aroma and therefore the acceptability of such wines. However, while recent research has shown the concentration at which TCA can be detected (~ 1.0 ppt), and the concentration at which TCA may impact different attributes of a wine (1-2 ppt) there have been no reliable estimates of the concentration of TCA that is required before consumer acceptability of the wine is adversely affected. It can be expected that it will be at a higher concentration than the absolute (detection) threshold. Wine is the most complex flavor system that exists. The attributes that make up a wine are a result of interactions occurring between the flavor chemicals, non-volatiles, and alcohol. As a result, the absolute threshold, the recognition threshold and the consumer rejection threshold of TCA in wines will differ. We have seen this with the detection thresholds of TCA is wine reported anywhere from 1-5 ppt. FlavorSense conducted a pilot study in New Zealand to measure the "consumer rejection threshold" of TCA in a simple white wine selling for $20NZ. We recruited white wine drinkers: those who consumed wine once per week or more $15-$25 NZ price range. We identified the consumer rejection threshold significantly higher than the consumer detection threshold, and the trained panel detection threshold. (Our study will be published soon.) FlavorSense measured the consumer rejection threshold of Hanzell Vineyards 2000 Chardonnay. The consumer rejection threshold measures the concentration of TCA that results in a change of acceptability of the wine. We showed that the CRT of the Chardonnay was 4.4 ppt, significantly higher than the concentration found in the wine as measured analytically. Thus, the concentration of TCA in the Hanzell Chardonnay was shown not to alter the acceptance of the wine. Interestingly, we also observed a trend that there are concentrations of TCA that may actually "blend in" with the wine, and thus be more difficult to detect. FlavorSense nicknamed this "the bimodal nature" of TCA. From the graph, we can see that the consumers did not prefer the non-spiked wines at 1ppt and 8ppt. We believe that for each wine there are concentrations of TCA that "unbalance" the wine and some that do not unbalance the wine. A food related example that has similar properties might be butter flavor on popcorn. Microwave popcorn has a small amount of butter flavor compared to movie popcorn, but consumers still eat the "under flavored" and "over flavored" popcorn. |
• Regularly have analysis of the water supply and the atmosphere in the cellar, caves, or winery buildings performed using GC/MS.
• Test the composite of your wine for TCA prior to bottling.
• Ventilate: Increase the ventilation of all confined spaces to prevent accumulation of chloroanisoles.
• TCA can be formed from any wood product: This includes open bags of corks. wbm
Crash Course in TCA
Arnold sought advice from local experts and Pascal Chatonnet of Lab Excell (Merignac, France).
A winemaking consultant Arnold declines to name charged $500 per hour for advice, another reason she's keen to share what she learned.
Hanzell increased ventilation, changed its water source, added a UV filter; installed new hoses, rubber gaskets and silicon bungs; sandblasted the exterior of its French oak barrels, moved all wood palettes outside; and hired a new assistant winemaker to oversee the new, non-chlorine cleaning and sanitation protocols and to promote rigorous maintenance training for workers.
Arnold is quick to point out that TCA is difficult, if not impossible to eliminate because the precursors can come to the winery on wood palates, on cardboard, or even in barrels if they've been rinsed with chlorinated water. In one case it was determined that wine was contaminated with TCA from a mobile bottling line because the filter wasn't checked.
Wineries need to make sure they are not using chlorine anywhere on the property and must make sure that their water is not chlorinated, stressed Arnold. "You have to set up testing protocols to make sure that you are monitoring anyplace you might be at risk for TCA to enter your atmosphere."
Hanzell tested its silicon barrel bungs, rubber gaskets and rubber hoses and switched from using rubber hoses to using plastic ones. Rubber may be more of a sponge for TCA than plastic.
Going forward, Hanzell plans to check a composite blend of wine from various barrels for TCA.
Precursers to TCA can linger on the outside of one or two-year-old barrels and trichlorophenol was found on the outside of some of Hanzell's barrels, thus the sandblasting. Hanzell is eliminating the chestnut hoops on its French barrels too. They're aesthetically pleasing, but can be a sponge for TCA. The new cave now includes a partition meant to increase ventilation.
Arnold learned that plastic absorbs TCA, though the TTB does not permit putting plastic into wine to remove it (it does allow bag-in-the-box containers). Even the plastic liners on screwcaps can absorb TCA.
Treating wine with milk products introduced the correct way will reduce TCA because of absorption by the butterfat content. There are several materials that can be used to reduce TCA concentrations in wine. In general, the fining operation introduces a compound that has a higher chemical affinity for TCA than found in the wine solution.
"This science has only been available in the last five years, and only recently can you detect down to one part per trillion (due to new GC/MS analytical techniques)."
The "new science" consists of new analytical methods that introduce internal standards based on isotopes of TCA. These standards allow more accurate measurement of extremely small TCA concentrations.
Atmosphere traps can be used to check for chloroanisoles in the environment. They can be applied to determine whether a problem exists, or if an issue is identified, to identify the actions one can take to fix the problem.
"We are taking a philosophical stand that we want to manage our risk down to zero tolerance," Arnold said.
Sensory Testing
Estimating the "Rejection Threshold" for TCA in White Wine Prescott, J.1, Norris, L.2 and Kunst, M.3 1 School of Psychology, James Cook University, Cairns, Australia; 2 Flavor Sense, San Rafael. USA; 3 Wageningen University, The Netherlands. It is estimated that 5 to 10 percent of wines are tainted with 2,4,6-Tricholoroanisole (TCA). [This estimate is considerably higher than statistical research published to date, which show wine tainted from TCA to be in the 1 percent to 2 percent range, ed.]. Also known as cork taint, TCA produces odors described as musty, dank or earthy that are unacceptable to many wines drinkers. Although previous estimates of TCA in wine have put the threshold as low as 0.5 parts per trillion (ppt), it is not clear at what levels TCA begins to render a wine unacceptable. It is therefore difficult, even if one knows the distribution of TCA levels in commercially available wines, to assess the economic impact of this wine taint. We developed a method to address this question by using a paired preference test within a typical method of limits threshold procedure. The aim was to determine the point at which wine consumers wouldbegin to reject a wine containing TCA, which we termed the consumer rejection threshold (CRT). Fifty-eight regular white wine consumers (Ss) received pairs of samples of white wine and were asked to taste each and indicate which of the samples was preferred. Ss received replicate series of 8 pairs, in which one wine sample was "spiked" with TCA at the following concentrations: 0, 0.5, 1, 2, 4, 8, 16, 32 ppt. To determine if the CRT was related to sensitivity to TCA, we also determined absolute thresholds (AT) for TCA in wine. Best estimate thresholds for the CRT and AT were 6.3 and 2.8 ppt, respectively. CRT and AT were also significantly positively correlated (r = 0.43). Few of the Ss were "serious" wine drinkers as assessed by a wine knowledge test. Nevertheless, both CRT and AT were significantly negatively correlated with the score on a series of questions that assessed knowledge about TCA. These results provide a rational basis on which to assess the real impact of TCA in white wine, and estimate what levels of TCA should be regarded at unacceptable. The results also provide evidence for the utility of this method for determining CRTs. |
Sensory analysis established that the low levels of TCA in the Chardonnay were not observed or recognized by consumers and that they didn't change the acceptability of the wine for qualified, wine knowledgeable consumers who were tested. Leslie Norris, of Flavorsense has studied consumer rejection thresholds in New Zealand.
She conducted consumer testing for Hanzell to determine the concentration of TCA that would result in a change of acceptability of the wine from the viewpoint of a consumer (see "FlavorSense: TCA Data Summary). ETS assisted Flavorsense by spiking samples with a range of TCA concentrations.
"We know there are wine experts trained to detect extremely low levels of TCA," said Arnold, "but our tests of educated wine consumers, held mostly at (Copia: The American Center for Food Wine and the Arts), indicated the Chardonnay was well below their thresholds for TCA detection. That reassured us that we could re-release the Chardonnay. While we are not officially re-releasing the Pinot Noir, we will supply our customers who simply demand it. Many of our customers have tried the Pinot and continue to request it. In any case, as has always been our policy, we will stand behind our wine and guarantee a refund to anyone who is not happy."
With brettanomyces, sensory perception of it is linear: The more brett there is, the more one can taste it. Flavorsense tests showed that TCA is bimodal in its interaction with wine at low levels. Arnold, for instance, perceives it in the 2000 Chardonnay at 1 ppt, doesn't sense it at 2 ppt, but picks it back up at again 3 ppt. "There is a point, where it is not diminishing fruit and is not perceivable," she said.
"It really is a threshold issue of where this becomes a flaw," said Arnold.
So is 2 ppt acceptable?
"In this wine at 2 ppt, I believe it works in synergy with the wine and does not diminish from the fruit or the wine," said Arnold. wbm
