Guest Article by Sean Sullivan, Northwest Wine Report https://www.northwestwinereport.com/
I always find it curious that winemakers spend so much time, effort, and money to make the best wine that they possibly can. Then, in the very last step, many ruin a substantial percentage of the bottles by using a closure with a high failure rate.
But how big of a problem is cork taint really? Let’s break it down with some numbers.
In my experience tasting wines for review both here and when I was a contributing editor at Wine Enthusiast from 2013 to 2022, between 3-6% of wines I have tasted that were closed by natural cork showed signs of taint by TCA or some other moldy contaminant. (Read everything you ever wanted to know about cork taint.) These findings are largely consistent with those from the Cork Quality Council, which has found an average of 3% of corks contaminated by TCA.
Okay, 3%. That’s not great, but it could be worse right? 3% is actually really, really bad.
First, let’s look at it on a micro scale. Many wineries in the Pacific Northwest are small production. Say that a winery makes 500 cases of a particular wine. At a 3% contamination rate, that leads to 180 cork tainted bottles of that wine. Potentially 180 customers whose experience has been negatively impacted to a greater or lesser extent. If that winery makes 3,000 total cases annually, that’s 1,080 cork tainted bottles per year. Wow. As a small winery, can you really afford to potentially disappoint over 1,000 customers per year?
Now let’s look at it on a macro scale. The US consumes roughly 4.3B bottles of wine per year. Let’s assume 70% of those bottles are closed by natural cork. At a 3% taint rate, US consumers are opening up over 90M bottles of corked wine per year. That’s over 247,000 TCA-tainted wines opened per day, just in this country! That, my friends, is a very, very serious problem.
Why do wineries allow this problem to persist? There are a variety of reasons I have written about before.
Some wineries, especially ones making higher tier wines, think that this problem does not apply to them. They think that they are taking sufficient steps to reduce the problem, checking bales, using higher grade corks, or even using individually tested corks. Or perhaps they think that cork taint is just a problem in inexpensive wines using lower quality corks. These assumptions are, unfortunately, completely false.
I’ve talked to wineries that test every bale of cork they use. Frankly, I do not see significant differences in the percentages of corked wines I receive from these wineries compared to other wineries. I know wineries that use individually tested corks (by human or machine). I see plenty of tainted bottles from those producers too, much to their exasperation.
Additionally, in my experience, cork taint has nothing whatsoever to do with wine price. One year, the average price of a bottle of cork tainted wine I sampled was $48. No one thinks that’s an inexpensive bottle.
Many producers also think that the size and scope of the problem is much smaller than it is. I’ve had some say the cork taint percentage is 1% or even 0.1%. The reasons for these misperceptions are likely, in part, due to suppliers misrepresenting the numbers.
For example, one could truthfully say that 99% of corks have TCA levels below 2 parts per million. (See the Cork Quality Council data linked above.) That sounds good! Unfortunately, 2 parts per million is still within the sensory threshold of many people. Even with those people who cannot smell it, research tells us it still negatively impacts their experience of the wine.
Ultimately, the Cork Quality Council data make very clear that the percentage of corks contaminated by TCA is about 3%. Additionally, this is not looking at tribromoanisole (TBA) and other moldy contaminants that would drive that number higher.
Now let’s look at the impact. For the consumer, TCA ruins the experience of wine. With that comes wasted money, spoiled special occasions, and disappointment with bottles that, in some cases, people have cellared for decades. You can read my Top 10 worst cork taint experiences here and here.
For the winery, TCA taint can make a consumer think that their wine is not very good. For the critic, TCA taint can cause substantial harm to one’s reputation. If someone purchases a bottle based off a recommendation and that bottle is faulted by cork taint, consumers are unlikely to detect the issue. Rather, they will believe the critic’s recommendation was bad – and ultimately that the winery doesn’t make very good wine.
Some have asked me over the years, what level of cork taint is acceptable? The answer is zero.
All that is needed to make that a reality is the will. Cork companies, seeing a substantial loss of marketshare to the increasing number of wineries using alternative closures, have spent millions of dollars to try and address the problem. Will those efforts finally pay off? Time will tell. To date, they have not.
In the meantime, wineries that still use natural cork should continue to put pressure on companies that are selling them a consistently faulted product – or even charging them exorbitant sums of money in an attempt to make sure that the corks that they are already paying for are sound! Similarly, consumers should continue to pressure wineries by bringing corked bottles of wine to their attention and asking for replacements or refunds.
As I have written before, cork taint is a choice. It is a winemaking decision to accept a high percentage of faulted bottles. It’s time to be rid of cork taint once and for all.
37 Responses
Cork taint is rising again : From near 1995-2015 I saw a large decrease in the percentage of bottles that became cork tainted. Today my assessment is cork taint and TCA issues are back on the rise. This needs research as to why. The problem is back and increasing. Tom Payette
Hi Tom, thanks for weighing in. I think the conversation is worth pursuing, and will ask the author Sean Sullivan to comment, as well as some of my winery clients.
Thank you for this article. As a wine writer/blogger and wine drinker, my experience doesn’t support 3%, it is probably at 1% or less level (I’m not saying it is good, just presenting the numbers). There is also big dependency on the regions in my experience – I hardly ever have a corked California wine. I had been writing quite a bit and tasted a good number of Oregon wines – I never encountered any corked wines from Oregon. I also never had a corked wine from Washington. Most of my corked wine experience would be coming from Italy and France. Again, I’m simply sharing my experience of the past 12-15 years.
Screwtop is great, and I understand the danger of the corked wine on many levels – money, disappointment, frustration etc – nevertheless, as I like my wine aged, I have a strong preference for the wines closed with the cork.
Hi Anatoli, nice to hear from you. I know you get wines from both Europe and west cost, so this is an interesting perspective.
Hi Anatoli Levine, author of the article here. First, let me say that individual sensitivity to TCA and other musty aroma compounds can vary as much as 200X. That means what one person detects at 1X, another person might need it at 200X to detect. There is also daily variation in individual sensitivity that can vary as much as 3X.
I would expect that there would be regional differences in the percentage of cork tainted wines, largely tied to the percentage of wines that use natural cork. Personally, I have never seen price as a factor so believe that it is more closure-choice driven. I would expect the relative number to be about the same.
The Cork Quality Council’s data shows that 3% of corks are rejected by GCMS (https://www.corkqc.com/collections/audit-results/products/incoming-natural-corks-scores-less-than-2ppt-through-q3-2016). (Note, the link says through 2016. It’s much more recent.) This should be considered a floor, not a ceiling, as we know from company’s using one by one testing on corks that GCMS routinely fails to detect cork taint, even at fairly high levels.
You can read additional articles I’ve written about cork taint here.
https://www.corkqc.com/collections/audit-results/products/incoming-natural-corks-scores-less-than-2ppt-through-q3-2016
Thanks for the thoughtful response, Sean, and links to more on the subject.
Sean, I used the link you provided, and the data as of March 2023 show (I’m quoting): “Based on these screening results, approximately 1% of incoming natural cork lots were rejected by the CQC members prior to acceptance into inventory”. Based on the chart located at the same link, it seems that 2016 was really an anomaly. So it is 1% and not 3%.
I also have seen wineries move back to the corks after using the screwcaps, at least for some of the wines. And then there is a sustainability factor in play which pushes winemakers back to the cork, even in Australia: https://www.abc.net.au/news/rural/2022-08-08/winemakers-embrace-corks-again-for-sustainability/101303874
Cork is here, and I hope it will stay for the foreseeable future.
Anatoli
Point and counterpoint perspective appreciated, Anatoli.
Anatoli, this is interesting. It looks like CQC updated the data on their site on May 7th, which had previously been through September 2021. It is now through March 2023.
Previously, the 99% number that CQC cited was for corks below 2ppt TCA. That’s a pretty low bar. I can’t attach an image, but you can see it yourself if you use Google to view the cached page, which read:
“Current Results from Screening of Incoming Cork Shipments Show 99% with Releasable TCA below 2ppt…Based on these screening results, approximately 3% of incoming natural cork lots were rejected by the CQC members prior to acceptance into inventory.”
This was through September 2021. This is to say, in 2021, the number CQC was seeing above 0.5ppt (the limit of GCMS) was 3%. They are now seeing approximately 1%. I’ll have a sense if what I see agrees with that when I roll up this year’s numbers in 2024, though there is a lag.
Bottom line, it might be 1% for bottles that will be released now or in a couple of years, but as recently as two years ago, which includes a lot of wines hitting the shelves now, it was 3%.
Again, I have to emphasize that GCMS numbers should be considered a floor, not a ceiling. They routinely miss tainted corks, as seen by producers using one-by-one tested corks.
There’s no question 2016 was a high frequency year for cork taint. In my tastings, 6.21% of wines closed by natural cork showed signs of contamination, the highest I’ve ever seen. However, contrast that with the staggering 9% that CQC saw above the 0.5ppt detection level!
Personally, I’m not surprised to see that the CQC data indicate that the percentage is decreasing. Starting in 2021, Amorim and Cork Supply both added processes that they said would remove TCA *and other moldy contaminants* prior to punching the cork (https://www.winemag.com/2021/03/03/wine-cork-technology-taint/). The latter is particularly significant because the 3% number CQC cites is just looking at TCA, not other compounds that we know can cause issues.
If that technology works as they say – which is always a dubious assumption – we should have seen a dramatic decrease in cork taint starting in CQC’s 2022 numbers. The data is consistent with that, which is encouraging. However, I would argue that, even if the percentage of contaminated wines going forward will be 1%, that number is still too high.
More broadly speaking, I do recommend wine professionals have their sensitivity to TCA tested. Most believe that if there is a corked bottle in front of them, they will detect it. This is not the case due to very large individual differences in sensitivity. Many professionals have had the experience of going to distributor tastings where a TCA-tainted bottle of wine is three-quarters poured through and no one has said a word. Even in my tasting group – all professionals – there are individuals who are very sensitive to cork taint, some who are in the middle, and others who can only detect it at very high levels.
Whether natural cork will continue to be the dominant closure or not remains to be seen. Companies like DIAM have been taking a *huge* percentage of market share from cork companies in recent years. I hope to publish some data on what that change looks like in Washington some time this year. It has been a dramatic shift.
Part of the reason for that has been the high level of contamination of TCA in natural corks, which is why Amorim and others have invested millions in processes to try and lower taint levels. However, there is still the issue of OTR and TPO differences in individual corks that can lead to significant bottle variation. That is a much harder problem to solve and I believe will lead to alternatives continuing to eat away at the market share, particularly if they can be price competitive with natural cork.
Consumer education is an ongoing process, and closure expectations for premium priced wines will continue to evolve, and I imagine as will winery product decisions.
I feel bad for wineries – large or small – that don’t have a clue when they send samples of “corked” wine to revieweres – but even worse when a consumer’s first impression of that wine is faulty. This is a potential downfall for a winery.
Indeed Meridith! And so do I, as a publicist creating sampling lists for clients. I take it you are saying this is an issue from your perspective? We have gotten away from sending 2 bottles to mitigate cork issues over the years as mentioned by Tom Payette in his comments. Do you think its appropriate to reconsider 2 bottle sampling then?
Meredith May, author of the article here. I routinely receive samples from wineries sending in wines for the first time where one of the bottles is cork tainted. It definitely gets them off on the wrong foot.
Meridith, You receive samples from all over the world at all price points. I’m curious to hear your experience compared to Sean’s.
Lets be clear – if a wine has TCA, which the winery can’t possibly know, the wine was not flawed, the TCA ruined it. No wine writer or reviewer actually being paid is not going to hold that against the winery.
True that William. Although, professional critics understand this issue, and hopefully will request additional sample bottles.
W, author here. I would disagree with that. Cork taint is a choice that wineries are making.
As I critic, I look at that decision as causing a potential existential threat to my career. I know that a significant number of bottles that consumers are tasting will under-perform due to cork taint. I also know that most consumers will not identify the wine as corked and therefore will just think it was a disappointing wine. If that was a wine purchased based off a recommendation I made, that has the potential to hurt my reputation, maybe seriously.
Still with my reviewer hat on, I do not penalize the winery when it comes to scores and reviews. In fact, I encourage wineries to send along two bottles of each submitted wines, in case one is cork tainted or there is some odd bottle variation. If anything, that enables wineries to continue selling a product using a closure that they know to have an unacceptable failure rate. Some have argued all critics should just accept one bottle. It’s something I’ve considered. Also, I am tasting wines blind, so prior history has no impact on scores.
However, I do send wineries that have tainted wines information about the numbers that I see. I also point out to wineries that have had repeated cork taint issues that their numbers appear higher than the norm.
There was one case where I removed a winery from Hugh Johnson’s Pocket Wine Book (I write the WA and ID sections) due to a high contamination rate over a two year period. If a winery’s rate is higher than baseline (which is already higher than it should be), yes that could absolutely impact whether I include that winery in roundups and other recommended winery lists. Again, it would not affect their scores.
As a consumer, I very much hold it against wineries. I’ve dropped out of multiple wine clubs due to having cork tainted wines. I also make purchasing decisions in part based on closure type. I’m much more likely to buy a wine and buy it in some quantity if I know I don’t have to worry about cork taint. Personally, I’ve had enough cork tainted wines for one life time.
Sean, I was tempted to ask you how your really feel about cork taint, but this makes it quite clear. Thanks again for sharing your words and this guest article.
Sean,
Your comments regarding the significance of seemingly small percentages are spot-on. How many of us would get on an airplane if they had a failure rate of even 1/10th of 1%? 1% is a huge defect rate; actually, in most industries when you talk about QA, 1/10th of 1% is 100X more than a acceptable level. Likewise, you are also correct in stating that the right target is “zero”. I’m reminded of a plant safety meeting from years ago; someone asked the question: “What’s an acceptable number of lost time accidents per year?” Answer: ZERO.
The essence of the problem is consistency. As you point out, it’s not just issues of dramatic, easily-recognized, levels of taint. There can be other less obvious flavor and aromatic deductions from the desired wine experience. And ultimately that is what it is all about– the consistent delivery of the desired wine experience. Often the fastest car on the race track is not the winner– consistent performance is the key to winning races.
Tom, thanks for your perspective.
Tom Mortimer, amen!
Don’t encourage him 😉
I agree with Anatoli Levine. The percentage of corked wines I receive for review does not reach three percent. Perhaps one percent would be more accurate, as he says. Such a circumstance is always disappointing, of course, especially if it’s a special wine the producer made a big deal about, but it happens rarely enough that I don’t consider it a huge problem. Well, yes, ANY cork taint is a problem, and certainly when a consumer opens a bottle and wonders what the hell is wrong with this wine and why does it smell like damp cardboard. Speaking personally, though, as a writer who has received thousands of review samples over many years, the number of bad bottles is insignificant. On the other hand, perhaps everyone should just go to metal screw-caps, as the Australians mostly did years ago.
I’m intrigued Fredric, by the contrasting experiences of the reviewers who have weighed in so far. Let’s see where we’re at in the next week or so. Thx for the comment.
Fredric Koeppel, as I wrote to Anatoli Levine above, we know from the Cork Quality Council testing that the percentage of TCA tainted wines is, at minimum 3%. That is the percentage of corks that they are rejecting using GCMS testing (https://www.corkqc.com/collections/audit-results/products/incoming-natural-corks-scores-less-than-2ppt-through-q3-2016; note, the data is actually more recent than the html link indicates). We also know that GCMS by no means catches 100% of tainted corks, so the actual number is surely higher. Many of the cork tainted wines I’ve had that have gone through GCMS testing are far from light levels. Individual sensitivity can also vary as much as 200X from person to person, which can account for differences in what different people see. I’ve been closely tracking closure percentage and cork taint percentage the last nine years.
I recently looked at the number of cork tainted wines I’ve had from 2015 to 2021. It was 357.
In terms of experiences and whether or not it’s a big deal, here are my top 10 cork taint experiences. All I can say is none of them were fun.
https://www.northwestwinereport.com/2022/11/my-top-10-cork-taint-stories-numbers-10.html
https://www.northwestwinereport.com/2022/11/my-top-10-cork-taint-stories-numbers-5.html
Other cork taint articles that might be of interest here: https://www.northwestwinereport.com/cork-taint-resources
I agree that producers should consider TCA-free alternatives.
Data, stats and scientific analysis, indeed. Thanks Sean!
That’s why we switched to screwcaps back in 2009. Even one bottle of wine going bad is not good for the customer, let alone 3%.
Brave move Wayne, and it appears to have paid off.
Over the years I have seen a few wine bottle stoppers (not corks) made what looks like rubber/silicone, with circular ridges to assure a good grip. Would this kind of material be a better choice? The cork has bee around for millennium’s, why haven’t alternatives, besides the screw cap, been developed sooner?
And other formats such as cans, bag in box, etc. are gaining ground for quality wines. Perception is the issue still.
The bottle math looks more like 30% than 3%.
Not sure I get you Marianne. Is your comment in response to another comment?
Memphis folks think alike. I’m with Fredric, use a screw cap and be done with cork taint. The “romantic notion” that a bottle has to be uncorked seems to have diminished. As writers/consumers/critics I think we have to spread the screw cap gospel in that it’s a perfectly suitable closure. It’s possible to sample or enjoy enough wines that 3% taint occurs, for me it’s rare. Perhaps 1% or less. But, it sure is a heartbreaker when it does happen. If consumers would accept it, more wineries would use the screw cap. But I’ve spoken to winemakers who’ve said they have to use cork, their customers wouldn’t accept score cap. It’s an uphill climb to convince consumers screw caps are worth it – but I’m trying!
James, great points. Its all about educating the consumer. On premise buyers don’t seem to mind for glass pours, so maybe it depends who your target purchaser is. Additionally, from what I understand, screwcap closures allow the producer to dial in O2 exchange for aging. Older vintage wines on screwcap are a way to proof out the tech.
Coming from a quality background and thinking of cork failure – I have heard estimates of 1% to 10% failure rates – though I am skeptical of high single digit rates. I have been skeptical of anything above 1-2% now especially in our micro-agglomerated wine cork closure and screw cap world.
Though at wine events and wine competitions the “failure” rate increases surprisingly several fold. I have seen and experienced at competitions where corks were judged as failed when in fact they had not and I was amongst many people who judged the cork as having not failed. If cork failure rates were at a 5 or 6σ rate I think it is impossible to get to this metric, perhaps a 4σ (.62%) rate is possible. Rare in our product world to get to a 6σ rate. Statistically, there will be failures and I am not sure a failure rate of 1 to 3% (between a 3 and 4σ) is out of scope for other product or service categories including wine closures.
James, I appreciate the product QA comparison and that it applies to wine closures as well. Interesting that judges call out cork issues at a higher rate during wine competitions. What would account for this? Confirmation bias and uncertainty?
The numbers for “corked” wines at events/competition I have seen at 10% or higher – I have confirmed more times that not that a wine isn’t corked. I am not the only person to note this for this setting has an abnormally “corked” rate.