From the archive: Bottle variation and the best laid plans...

Of all the mysteries of wine, bottle variation can be one of the most frustrating. In this article from our archive, Richard Hemming MW investigates the science behind the phenomenon, and explains why the old adage that ‘there are no great wines, only great bottles’ often rings true.

Originally published in Decanter magazine's June 2013 issue and now available online and in full for Premium members.

A discussion ensues – they are Bordeaux, of course, and surely Left Bank. The host concurs. Agreement cannot be reached, however, about which châteaux are represented. The debate threatens to be endless, until it is revealed that they are in fact the grand vin of one château alone.

Much amazement greets this news, and with great enthusiasm they turn to identifying the vintages. The six are all quite mature, but the browner ones are surely older – yet the tannins don’t always agree – while those displaying the greatest complexity must be the strongest vintages. It is a challenging task, but consensus is finally reached, and then with great ceremony the host disrobes the bottles in front of them. Of all the mysteries in wine, bottle variation is surely the most enigmatic.

Drinking through a treasured case of even the most carefully cellared wine can be like a cross between a lucky dip and Russian roulette – although not quite so life threatening.

The old saying ‘there are no great wines, only great bottles’ is such a well-worn cliché because it’s true.

In simple terms, bottle variation describes any two bottles of wine of the same vintage, producer, appellation and cuvée that taste palpably different to one another.

Specifically, one is usually better than the other, often by a significant margin. ‘The older that wine becomes, the more bottle variation you have to expect,’ points out wine writer Neal Martin, author of Pomerol.

‘It’s just part of the fascination of drinking mature wine and so it is something I embrace.’

Pre-bottling factors

The term ‘bottle variation’ usually implies changes that occur post-bottling. Sometimes, however, it can be caused by pre-bottling factors.

Often, wines are bottled in batches throughout the year rather than all at once. ‘It was not until the 1970s, for example, that it became commonplace for Bordeaux châteaux to ensure that a uniform blend was made before bottling,’ according to Jancis Robinson MW’s wine bible The Oxford Companion To Wine.

That means that blends might differ from batch to batch, coming from different vineyards or containing a slightly different mix of varieties.

Alternatively, they might be tailored for particular markets. For example, one well-known brand of rosé reputedly has 10 times as much residual sugar in Poland than it does in the UK.

There are also stories of blends varying for more cynical reasons. In 2011, the Barossa producer Schild Estate had its 2008 Shiraz listed in an American magazine as a top 10 wine of the year. When demand subsequently rocketed, the producer reportedly bought in 5,000 cases-worth of 2008 Barossa Shiraz and released it under the same label, raising questions about integrity – although the estate’s winemaking team were keen to point out that extra wine was sourced from a local grower and would only be sold in Australia until the 2009 vintage was ready for release.

Jacques Lurton, a flying winemaker who also makes wine at his own estates in Bordeaux and Australia, is unequivocal: ‘Bottle variation can be minimised by bottling the same wine in its full amount at the same time.’

These days, lot numbers can help to mitigate this variable, albeit cryptographically. So if we assume that two given bottles of wine were perfectly uniform when they were bottled, then what causes the changes thereafter?

The five variables

How wines evolve in bottle depends on five main variables:

time bottle size;
bottle closure;
storage;
transit conditions;
level of dissolved gasses, especially oxygen and sulphur dioxide.

These variables influence the extent of hydrolysis, esterification, oxidation, reduction, yeasts and bacteria – especially Brettanomyces – which collectively account for the development of wine.

Wines can also be altered by cork taint, lightstrike (when bottles have been exposed to excessive ultrviolet light) or heat damage.

These are straightforward faults, whereas the likes of oxidation, reduction and Brettanomyces can be interpreted as either beneficial or detrimental, according to their level of influence, the style of the wine and, of course, a taster’s preference.

Hydrolysis and esterification are a vital part of how flavour evolves in bottle.

Both form new flavour compounds by reacting with the acids, alcohols and glycosides found in wine, creating the petrol character in Riesling, or the toastiness of Hunter Semillon, for instance.

It’s a complicated process, but the basic thing to know is that this is a good thing.

But hydrolysis and esterification are likely to manifest themselves differently within each bottle, and that can be a bad thing, meaning that some bottles appear less impressive than others.

Reactions over time

With time, those differences become magnified. Furthermore, increasing the length of maturation increases the potential influence of other factors such as oxygen.

‘Bottle variation is largely a result of oxidation at varying rates,’ says Bruce Tyrrell of Tyrrell’s winery in the Hunter Valley, Australia.

Too much oxygen equals oxidation, and that is the ruin of many a good wine, killing the fruit and freshness.

The key to optimal maturation is to have just the right amount of oxygen. In practice, that means an extremely gradual, controlled volume – but the ‘correct’ amount required is quite literally an unknown quantity, and will vary according to the style of the wine.

Winemakers choose the closure and level of sulphur dioxide with this in mind.

One of oxygen’s primary roles is to soften the harsh tannins of young red wines, thereby helping to establish drinkability and balance in maturity.

So if oxygen ingress varies because of the cork, for example, then this could mean differences in tannin from bottle to bottle.

Oxygen also impacts on colour of both red and white wine, turning them increasingly brown with age.

Tyrrell reports that ‘random oxidation was a bigger problem for us than cork taint, and when we were releasing Semillon under cork at seven to 10 years old, we were rejecting up to one-third based on colour.’

Those spoilt flavours are caused by aldehydes, a flavour compound formed in the wine when oxygen reacts with alcohol.

There are many types of aldehyde, some of which can be pleasant. However, the most common type, acetaldehyde, is considered undesirable in most table wines – although it is prized as the distinctive flavour of fino Sherry.

Screwcaps and sulphur

Simply put, reduction is the opposite of oxidation. In reductive conditions, often blamed on screwcap closures, sulphur compounds can develop which significantly impact on the flavour of the wine.

Sometimes they can lend the struck match, flinty notes that are widely appreciated in wines such as white Burgundy; other times they can form the rather less desirable honk of dirty drains and burnt rubber.

On the other hand, a potential advantage of reduction is to preserve a wine’s freshness, because it means oxidative deterioration is avoided.

Therefore, just as with oxidation, the impact of reduction can be either beneficial or detrimental, and it can vary from bottle to bottle – and it is the closure that is mostly responsible.

Plenty of bacteria and yeasts would quite happily populate a sealed bottle of wine. The most notorious is Brettanomyces. This unassuming little yeast produces some of the most divisive flavours in wine, with some eulogising about their earthy, feral scent while others slam them as outright faulty.

In bottle, Brettanomyces is controlled by sulphur dioxide: the more of the latter, the less of the former. Unfiltered, unsulphured wines are therefore most at risk from Brett.

Whether you like its influence or not, microbial activity is another important piece in the bottle variation puzzle.

And as if all this wasn’t enough, another factor is the susceptibility of particular grape varieties. Neal Martin feels that monovarietal wines ‘vary more than blends… I’ve found more variation in Burgundy compared to Bordeaux. I think that is to do with the sensitivity and capriciousness of Pinot Noir.’

Jamie Hutchinson, of London wine merchant The Sampler, points out that ‘Nebbiolo [the Barolo grape] is terribly sensitive to oxidation, and shows massive variability within a single case, whereas Mourvèdre is almost immune’, while Bruce Tyrrell classifies risk by wine style, suggesting that ‘lower alcohol, non-oaked wines are more affected by bottle variation’.

So then, even if two bottles of wine are identical at bottling, even if they have identical closures, are stored identically and are tasted at the same age in the same place under the exact same conditions … even then, they might still exhibit bottle variation.

Knowing why it happens doesn’t mean it can be controlled. Nor perhaps should it be: bottle variation is an integral part of wine’s mysterious allure, and some mysteries are best left unsolved.

The five influences

Influence no1: Dissolved gases

Sulphur dioxide is added to wine as a preservative. Its efficacy depends on a number of things but, basically, the less SO2 that a wine contains the more prone it will be to bottle variation. Dissolved oxygen is generally kept to a minimum for the same reason.

Influence no2: Storage and transit

Temperature has a big influence on bottle variation. Humidity also plays a part – although that is more related to cork preservation. Warmer conditions will speed up the reactions that contribute to bottle variation – and not in a good way.

Jamie Hutchinson of the London wine merchant The Sampler says: ‘Mature Italian wines that have been aged in Italy have an unacceptable amount of variation in regards to oxidation. In my opinion this is due to terrible average storage conditions, which are too warm.’

Influence no3: Closures

Closures have a central influence on bottle variation. Jacques Lurton, a winemaker in France and Australia, says: ‘With cork it is impossible to avoid bottle variation … for me, the screwcap is a guarantee that every bottle from the same case will give you the same pleasure.’

Australian winemaker Bruce Tyrrell adds: ‘Screwcaps may not be perfect, but they are the best [closure] that I have ever seen.’

The cork industry is working harder than ever to respond. Dr Miguel Cabral, head of research and development at the cork producer Amorim, cites its ongoing trial of a white Burgundy sealed with six different types of cork, which has shown no bottle variation after 41 months maturation, either by olfactory or chemical analysis.

Influence no4: Time

Time is paramount to bottle variation: it is an inexorable fact that as the dust gathers on those bottles in your cellar, their variation potential gets greater and greater and greater…

Influence no5: Bottle size

The speed of maturation changes with bottle size. For example, according to Jancis Robinson MW’s bible The Oxford Companion To Wine, half-bottles ‘hasten wine ageing, partly because they contain more oxygen per centilitre of wine since the bottle neck and ullage are the same as for a full bottle’.

When it comes to maturing in bottle, it’s a case of the bigger the better, according to wine authority Hugh Johnson.

Originally published in Decanter magazine’s June 2013 issue.