Screwcaps and beyond
by guest wine writer Paul WhiteRichard who? advertise on this site email wine words news back home translations
This detailed article was first published in World of Fine Wine magazine in the August 2004 issue. I've reproduced it here with Paul's kind permission to offer a different view on this news item I posted on 16/11/04: New Zealand takes lead in screwcaps as wine closures. Other related stuff on winewriting.com: Lingenfelder goes screwy, Laroche Chablis cork v screwcap (temporarily missing).It should be increasingly clear that screwcaps aren’t at all what they’ve been cracked up to be. Evidence increasingly suggests that they are as potentially faulty, albeit from a different set of faults, as either corks or synthetic stoppers. Before I dig myself any deeper into this contentious little hornet’s nest, let me declare up front that I have absolutely no preference for cork over screwcaps over whatever. I’ve had both magnificent and absolutely dreadful wine sealed under both corks and screwcaps. Like most consumers, all I want is to have wine delivered to me fault-free with a degree of consistency. If it’s wine intended for early consumption I would want it to be fresh and ready to drink without being disjointed or excessively harsh. And where it’s a fully matured, bottle aged wine, I would hope for the sublime heights of the finest old wines I’ve tasted so far under the best corks.
All that said, what I’m never very happy about is deception and over promise. Bearing this in mind, many of the claims for perfection that have been bantered about over the last few years by screwcap advocates are beginning to ring hollow. Let’s take a typical set of claims found on one of that industry’s main promotional pages, www.screwcap.co.nz. Consumers there are offered four promises: “Total confidence that you will receive wine in premium condition - alive, abounding with flavours and a pleasure to drink. No bottle variation - each bottle of a given wine will be just as good as the one before. Elimination of cork taint and similar mould flavours that can contaminate wines bottled with other closures. Dependable cellaring - the elimination of random, premature wine oxidation.”
Although these bold assertions suggest perfection, they are not out of sync with what we’ve been told to expect from screwcaps by the vast majority of the wine press throughout the world. Not only has this lulled consumers, retailers and sommeliers into a false sense of security where stelvins are concerned, but without doubt it’s led to situations like the one the I ran into a few weeks back in Wellington, New Zealand. After ordering a bottle of wine at a smart restaurant there, the sommelier ceremoniously snapped off its cap, filled the glasses around my table, promptly sat the bottle down and left without another word. As he passed by later, I asked why we hadn’t been offered us the opportunity to try the wine and either accept or reject it. His response was that any wine sealed under a screwcap was delivered in perfect condition, so there wasn’t any need to seek our approval. Although that bottle did indeed drink very well, as have many other wines under screwcaps, I’ve known many others that haven’t.
Where screwcap advocates can accurately claim to be free of corked derived TCA, they can’t honestly claim to be free of other faults. Problems with screwcaps first cropped up during a blind tasting of thirty, newly released, 2002 New Zealand Gewurztraminers I was surveying with two other professionally qualified judges. To our amazement, we had tossed four of eight screwcapped wines out of contention for being seriously malodorous. Two were fully reduced, exhibiting reduction’s tell-tale rubbery, rotten egg-like characters, a second had an indefinable, quasi-reduced grubbiness, and the third was obviously oxidized. This fifty percent failure rate was shocking compared to the 3-10% cork taint rates commonly bandied about back then.
A few months later, again with another group of professionally qualified judges, I blind tasted a pair of Jackson Estate 2001 Sauvignon Blancs that had been especially bottled under both cork and screwcaps to demonstrate the developmental capacities of each. This was part of an ongoing trial sent to journalists by the New Zealand Screwcap Initiative in September 2001 to demonstrate the superior ageing capacity of screwcaps over cork. Logically, the stelvin sealed wine should have been younger and fresher, whereas, the cork sealed wine less fruity and more developed. Contrary to expectations, the wine under screwcap showed signs of oxidation and considerably more forward development than the cork stoppered wine. Previously I had seen instances of dented or creased screwcaps, but this time I found no visible sign of leakage or damage to the capsule. What was going on we all wondered?
None of this was supposed to be happening from what we had been told back at the original launch of the New Zealand Screwcap Initiative in September 2001. We were assured that not only had stelvin seals totally eliminated the risk of TCA taint forever, but a special bevelled corner fit between cap and glass made oxidation and leakage a thing of the past. After questioning the anaerobic nature of screwcaps and wondering whether this could create reduction issues, I was told this was as a manageable ‘teething problem’ and that reduction would never be a problem in future.
Suddenly, a year and a bit down the track, screwcaps didn’t seem to be living up to these promises. When it comes down to it, faulty levels of oxidation and reduction are just as unacceptable as ‘corkiness.’ Unfortunately, explaining these problems accurately requires some mind numbingly complicated chemistry dealing with the interplay between amino acids, redox potential and micro-oxygenation. Leaving these details to professional winemakers, here’s a grossly oversimplified explanation.
Reduction is essentially the mirror image of ‘oxidation.’ Both alter the purest expression of fruit. And just as with brettanomyces, a tiny bit can add complexity, while too much will permanently destroy a wine’s aromas and flavours. Unfortunately both can easily tip over into unacceptably ruinous levels. So as oxidation increasingly redresses wine with an unfresh, caramel-like sherry character, reduction continually forces more negative sulphurous characters into wine reminiscent of struck flint, burned match, rubber, cabbage or rotten eggs. The reductive process revolves around a sulphur compound called hydrogen sulphide (H2S) which is formed in the absence of oxygen by yeast during fermentation. Unchecked by oxygen, H2S tends to hang around, tenaciously, stinking things up. This is not to be confused with ‘free’ sulphur dioxide (SO2) that winemakers use to sterilize and preserve wine, which dissipates more readily.
Normally the presence of oxygen forms a barrier that keeps hydrogen sulphides from reducing further into negative rotten egg-like characters. However, in anaerobic (oxygen-free) environments, any hydrogen sulphide left over in wine at the time of bottling will become a permanent part of the wine’s aromas and flavours. Once these “reduced” characters are ‘locked in’ like this, no amount of aeration after the bottle is opened will get rid of them. The only preventative is to add copper before bottling, which pulls sulphur back out of the wine and cleans it up. After a reduced wine is poured from bottle the only curative is to drop in an old penny and hope for the best. Which is bad luck for Australasians, who stopped minting copper coins years ago.
All of this creates problems for screwcaps, because of their air tight nature. Dr. Alan Limmer of New Zealand’s Stonecroft Winery, whose doctorate is in chemistry, explains, “As soon as a wine starts to look a bit reductive in barrel, the traditional and logical response is to give it some air. Racking helps blow (volatilise) H2S off, and copper will do the rest. The problem comes when the more intractable (complex) sulphides have formed. Or if all the sub-threshold (undetectable) H2S is not removed. With stelvin’s lack of oxygen, you have a recipe for further reduction. Whereas under cork a small, constant input of O2 would act as an intermediate barrier to the formation of more intractable and smelly compounds.”
After judging at the Australian Alternative Varietals Wine Show in November 2003 it was clear the same problems were still dogging screwcaps. This time faultiness was discovered while blind tasting with a panel of professionally qualified, Australasian “show circuit” judges. After the show I cross-referenced my colleagues’ discussions concerning faultiness back to specific enclosure types. Six of eighteen, stelvin sealed, Pinot Gris at AAVWS were faulty. Two of these suffered from oxidation, another reduction, while three others had faults related to VA and bacteriological taint.
Stelvin enclosed Pinot Gris failed anywhere from sixteen to thirty-three percent, depending on whether VA (volatile acidity) and bacteriological contamination are considered to be enclosure determined faults (as the AWRI suggests below). Of the fifteen remaining Pinot Gris in “other” enclosures (cork, synthetic, etc.) one was oxidized, another had VA. Using similar criteria, non-stelvins failed at between six and eleven percent. In other classes, admittedly of smaller samplings, Gewurztraminer, Sangiovese and Tempranillo in stelvins failed between twenty-five and fifty percent, again reduction the culprit, with one case of VA.
Although it’s possible to continue detailing anecdotal evidence of faultiness encountered during tastings over the last couple of years, why belabour the point. The fact is that all of the problems encountered above have been -- and are still being -- actively discussed in winemaking and academic circles. Clinical investigation by the Australian Wine Research Institute (published in Technical Review No. 142, Feb. 2003) strongly supports evidence emerging at street level. After trialling screwcapped Semillons at 18 and 36 months, testing found “wine bottled with the ROTE [screwcap] closure was rated significantly higher in a character that was defined as ‘reduced’ or ‘rubber’,” than all other enclosures, including natural corks. This was “due to chemical reactions of sulfur compounds in the relatively anaerobic environment of the ROTE seal.”
In an associated study, AWRI randomly pulled a large, commercial sampling of Rieslings off retail shelves and found a quarter of those under screwcaps also showed reduced characters. AWRI concluded that “after a period in the bottle, some commercial wines can develop this [reduced] aroma, and it seems most common in wines bottled under screw cap closures.” Clearly screwcaps’ air tightness was a primary factor in the emergence of reduced characters over time and, at the very least, it was exacerbating the problem more than other enclosures.
Reading between the lines, this tossed out some fairly profound implications. We must assume the study’s Semillons and Rieslings were made by competent winemakers who had screened the wine for any residual sulphide problems before bottling. It is likely that undetected -- or possibly even undetectable -- levels of sulphides were the cause of the problem. Although perceived to be sound at bottling, ticking away like time bombs, these wines eventually developed reduced characters after a year or more down the line.
All of which makes long term cellaring problematic. Where corkiness may ruin a couple of bottles per case, stelvin’s tendency toward reduction offers a chance of screwing up the whole lot. The AWRI eventually concluded that reduced aromas were unlikely to be a significant issue in “screwcap closures for white wines intended for early consumption where all stocks of the wine would be sold and consumed in the year of release.” But they also were careful not to endorse long term cellaring of any wine, red or white.
The oenology staff at Charles Sturt University have similar concerns. An article on screwcaps in Australian and New Zealand Grapegrower and Winemaker (March 2004) relates a heated discussion whether both the CSU’s winery’s whites and reds should be placed under screwcaps. Concerns raised there dealt with reduction, the need for oxygen ingress for development and whether well-matured reds should end up tasting like newly bottled wine. One of the earliest adopters of screwcaps, Brian Croser has expressed similar concerns about screwcapped reds he made between 1978 and 1982. In an article by Tim Atkin MW of The Observer (April 25, 2004), he is quoted as saying “The wines are developed, but they're tinny and flat; like stale jam. The best wines sealed with a natural cork are better than all the screwcap wines.” In a significant move, one prominent stelvin user in New Zealand has gone back to using corks on his reds after having trialled one vintage under stelvin. All things considered, he’s not likely to be the last.
Compounding the reduction problem, AWRI noted in their Annual Report of 2003 that they had uncovered a raft of quality control issues surrounding stelvin bottling practices. These keyed primarily on failed seals that led directly to the development of oxidation, bacteriological contamination and VA faultiness in a number of different screwcapped wines. After interviewing a number of winemakers who are using stelvins, it’s clear to me that unexpected problems with bottle calibration, fill levels, head space (specifically what gas to fill it with), capsule application pressure, transport and storage have led to much more oxidation and spoilage than is being reported. Whole batches of wines have been rejected because of mechanical failure and I have seen enough uncrimped capsules, incomplete or non-existent serrations, detached foils and other problems with seals myself to suggest more than a few individual faulty seals are flowing through the system.The weight of all these faults suggests that stelvins are not delivering what they promised. There is bottle variation, wines are being aged prematurely, and have been tainted with bacteria, and clearly both reduction and oxidation are problems in wines that are sitting on the shelves and resting in our cellars right now. Setting aside mechanical failures, why are stelvins particularly vulnerable to reduction, and to a lesser degree, oxidation? The crux of the issue revolves around whether wine ages reductively or oxidatively. This is crucial because it determines how wine is treated before going into bottle and how it behaves afterwards. On one side, stelvin supporters, citing Emile Peynaud’s views, see optimal bottle maturation as essentially anaerobic (oxygen free) in nature. On the other side, many other winemakers (including Professor Roger Boulton of UDavis) are convinced that bottle maturation is dependent on small amounts of oxygen seeping through the cork into the wine. So far, no definitive scientific studies have determined the precise nature of post-bottling chemistry. A lot of long term research is still needed to resolve the issues surrounding bottle maturation. Until we have more answers than questions, nothing is dead certain.
Alan Limmer explains the current dilemma, “Once you decide to store your wines anaerobically instead of oxidatively (screwcaps vs cork). This has huge repercussions for subsequent wine development. It would be wrong to think of the stelvin environment as neutral. Wines have what is called a ‘Redox potential.’ They are born with it and it reflects the winemaking process that has gone before. Think of it as a pendulum. On one side is the oxidative phase where wine oxidises in a short time after exposure to air. On the other side, sealed off from air, is the reductive phase. The wine swings between these phases depending on what process is occurring.” “Ferments are hugely reductive (yeasts have an enormous appetite for oxygen). This is where most of the sulphide issues originate. What happens to the wine after fermentation has a lot to do with what happens to the sulfide issues. Racking and barrel ageing shifts the wine slowly toward the oxidative phase, so the redox potential moves from reductive to oxidative. After about eight months (about the time it takes for barrel breathing to neutralise the reductive potential of a ferment) the wine’s redox potential is neutral, but moving toward oxidative. Much has been written about this measurement in wine chemistry literature and there are many complex papers on the subject. Even with a (doctorate) degree in chemistry, my head hurts reading them sometimes. But most winemakers don't even know what I'm talking about when I mention Redox measurements.”
“The point is, the wine is set up for a number of reactions subsequent to bottling driven by the redox potential. Until now it has probably not been so significant as wines appear to have been stored oxidatively in cork and any issues from a reductive environment have been treated benevolently under this regime. Our New World wine making is largely reductive - gas blanketing, etc. - to escape much of the Old World syndrome of oxidative winemaking. So, to then suddenly switch to reductive storage of wines long term was always going to have some repercussion on the wine’s post bottling behaviour.” Some stelvin advocates postulate that the best corks are anaerobic in nature and behave no differently than stelvin caps. They believe the supply of oxygen needed to stave off reduction in cork stoppered wines is contained inside the cork itself (about 2-3 ml) and doesn’t come through it. Knowing a reservoir of oxygen is required to keep reduction at bay, some producers are filling up the “head space” with this amount of oxygen. This practice is anathema to previous New World techniques where fear of oxidation would have winemakers doing everything possible to protect wine from oxygen contact at bottling.
Again, Alan Limmer takes up the other side of the debate, “Everyone I know is at pains to bottle anaerobically under cork - to stop oxidation at bottling by preventing any shock O2 doses. We know from a CSIRO/SCORPEX study that over a lifetime of say three to five years, the wine will consume about five times (or more) as much O2 as is contained inside the cork - about what you see in two to four weeks of barrel ageing. The oxygen available under cork differs fundamentally to the dose stelvin users are adding before bottling. Although both are relatively small in the scheme of things, the difference is that after bottling a cork bleeds O2 over a period of time which does have a beneficial effect on the chemistry - as per micro oxygenation technique, versus macro oxidation which gives the equivalent dose all at once.” Over and above the problems with leaky seals, Limmer suspects this may account for some of the oxidation currently showing up under screwcaps. Echoing screwcap’s most famous buzz words, Limmer concludes, “‘The perfect seal.’ Maybe - but only if you know exactly what is going on and can have complete confidence in the post bottling redox process. We are a long way from that.”
So with oxygen issues still unresolved, what about the other main defence that has been thrown at reduction, copper fining. Considered by some to be a heavy handed “last resort” for sulphide problems, coppering does not sit comfortably with artisans who follow non-interventionist winemaking philosophies. Nor is it an ideal solution for wine makers intent on building in yeast lees characters (rich in sulphides) for complexity. Why go to all that trouble carefully building these characters in, just to risk stripping them out later? Moreover, certain varietals are more susceptible to copper fining than others. Where Riesling is a relatively safe bet, Pinot Gris, Syrah, Pinot Noir and Sauvignon Blanc can be more problematic. Alan Limmer summarises what I’ve heard from many other winemakers, “Most winemakers know that Sauvignon Blanc has always been a knife edge proposition regarding use of copper – knowing when is it clean and how much is too much before the wine becomes sterile? Certainly much of the Sauvignon’s character is sulfide driven and recent French research has shown that thiols contribute largely to aromatics. Thiols are one of the compounds routinely fined with copper to reduce stink in wines - next to H2S they are probably the most prevalent. Trouble is, there are nice thiols as well as stinky ones. Unfortunately, copper doesn’t’ discriminate, the good ones are taken out with the bad.”
Which may explain the puzzling stylistic differences seen in some Marlborough Sauvignon Blancs from 2003 sealed under stelvin. Across the board vintage conditions seems to have delivered some fairly intense classic ‘sweaty, cat’s pee, herbaceous aromatics. Whereas a number of stelvin bottled wines have captured these qualities beautifully, others have come through seemingly flat and denatured, lacking herbaceous characters and the edgy style that put Marlborough on the map. One has to wonder if this is the result of heavy handed coppering Limmer discusses above? Ultimately, it could be that certain varietals will turn out to be unsuitable for screwcaps in the shorter term as well.
On a more positive note, the current rivalry between stelvins, synthetic corks and real corks has forced a scientific approach to bottle maturation where previously it was more often a black art. One thing that is quickly becoming clear is that no single enclosure is likely to dominate the future. Which is a good thing. We can thank stelvins, and synthetics earlier, for blowing the field of competition wide open. No single stopper can ever rest on its laurels again. All must strive to perfect their design. Here’s the current state of play with enclosures. Zork is a new, hybrid design that cleverly incorporates screwcap’s ease of opening with the magical “pop” consumers have come to expect from a cork being pulled. Looking a lot like an expensive sealing wax capsule, Zork is both free from TCA taint and clearly much more robust than stelvins, so should suffer less oxidation from transportation and storage. It also works on any bottle, freeing it from the naff “one size fits all” bottles screwcaps are locked into. Most importantly, this enclosure can be ordered with variable grades of oxygen permeability offering a viable alternative to both screwcap and cork advocates alike.
Improved cork products are already pointing toward the days when TCA is a thing of the past and stoppers play an active role in bottle conditioning. The reconstituted granules in Sabate’s new Diamond technical cork are treated with a supercritical CO2 wash process that penetrates like a gas and then cleanses like a liquid. AWRI studies have already certified these as showing “no evidence for trace levels of TCA” with better performance than screwcaps in terms of reduced characters. Bottle variation should be negligible. Again, these come in several grades of oxygen permeability. Pro-Cork takes a prophylactic approach, placing a two way, Gortex-like membrane between wine and a real cork. This filters out any trace of TCA and other impurities, while allowing wine to continually rehydrate the cork for a tight seal. As with Zorks and Diamonds, this is designed with graded oxygen permeability and reduced bottle variation in mind.
Real cork is rallying back through a fundamental shake-up of industrial processes. Taking a holistic approach, the introduction of quality control, smarter storage, improved pre-washes and wastage of the most likely contaminated part of un-worked bark, has already lowered the total background incidence of TCA taint in pre-manufactured cork compared to a couple of years ago. Portugal’s largest producer, Amorim’s new TCA curative wash, called ROSA, has already demonstrated significant reduction of TCA taint and other impurities in their agglomerate technical corks and sandwiched disk design, Twin Tops. The next stage of development will focus on graded permeability and reduced bottle variation, allowing wine to mature in a more predictable way under cork.
While providing options for either anaerobic or oxygen driven bottle maturation is an obviously element in all these designs, the next leap forward in design looks toward matching enclosures specifically to grape varieties and styles. Sabate’s Dean Banister explains the strategy behind Diamond cork, “We see bottle development as influenced by differences in grape variety, soil types and climate. The advantage of the individually moulded cork is that we also control density and uniformity, and through this can control oxygen permeability.” This flexibility allows winemakers to fine-tune the enclosure to all these parameters. Banister again, “Extreme examples might be NZ Sauvignon Blanc compared to Barossa Shiraz, where each could require different permeability ratings to aid or minimise bottle development.” The knock on effect of all this is that it’s very likely in future that consumers will play an active roll in choosing from a variety of stoppers, each of which will offer exactly the sort of bottle conditioned characters desired out of any given wine.
Paul White