Role of carbon dioxide in still wines

Published by Sally on May 30, 2009
1 Comment

A version of this article first appeared in Harpers Wine and Spirit, 2004.

When spritz is an issue for still wine

For better and worse, carbon dioxide (CO2) is implicated in several aspects of still wine making and flavour profile. Apart from imparting a spritz to a still wine, are high levels of CO2 implicated in other issues of reductive winemaking? How does one separate the good from the bad, leaving the ugly purely as a matter of personal opinion?

Impact on wine flavour

Carbon dioxide (CO2) is produced by yeast as a natural product of fermentation.  It’s also a permitted additive under EU (European Union) regulations.  It is an odourless, colourless gas, which is regarded as inert for winemaking.  It’s great for keeping oxygen away from must and wine.

Received wisdom has it that CO2 becomes perceptible to the human palate at around 1g/l as a slight spritz on the tongue, akin to the sensation given by Muscadet sur lie. At around 2g/l wine is legally semi-sparkling (see below). 

The legals:

  Pressure ~ CO2 @ 20°C EU Excise Duty  
Still < 1 bar <2 g/l Still Vinho Verde
Semi sparkling 1 to 2.5 bar 2 to 5 g/l Still Prosecco
Sparkling 3 bar + > 6 g/l Sparkling All

Source:  Wine and Spirit Trade Association. 

Its moderate presence helps to freshen a wine and gives a perception of slightly elevated acidity.  Sam Harrop MW, industry consultant says: “higher CO2 gives a crisper wine, with lower dissolved oxygen, less flavour intensity, and an addition to texture.” Geoff Taylor, of Corkwise says a bit of CO2 helps to preserve the wine a little, so you can lower sulphur dioxide (SO2) levels fractionally.    

At too-elevated levels it makes the wine spritzy and can lead to reductive notes especially in reds. CO2 management activist, Abrie Bruwer, of Springfield Estate in South Africa, says “CO2 on the palate belittles a still wine.  It makes the wine thinner and masks the nose. Think how difficult it is to assess the nose of sparkling wines.”   

What level exactly?

Appropriate levels are colour and style dependent.  Harrop said Marks & Spencer’s worked within certain bands: “Reds are generally 400-500mg/l CO2, with carbonic maceration wines slightly higher – 750-800mg/l. CO2 and malolactic characters do not fit well together – racking etc, working with oxygen avoids CO2 retention.”  Philip Shaw, MD of Cumulus Wines, added:  ” Chardonnay is likely to be 500 to 750 mg/l; being soft, you don’t want it fighting with the acid effect, you want a silk appeal.”  

Julian Grubb, head winemaker of Chilean VIA Wines has winemaking objectives of “1000 – 1100 mg/l CO2 in sauvignon blanc and aromatic whites, 800mg/l in Chardonnay and less than 500mg/l in reds. A touch of CO2 is great in a cold fresh sauvignon blanc as it gives it a little pique. However that same touch is quite unpleasant in a red served at 18 degrees.”    CO2 can make tannins sharper, more aggressive.

A more oxygenated winemaking process for reds explains their significantly lower levels.  Much CO2 is lost during pumping, racking, filtration etc.  A more reductive process appears to equate to higher levels.

Why is it important?

Greater emphasis on reductive winemaking for many white wines and everyday wines of all colour has changed the way we perceive our wines.  Combine this with the trend to make wines for drinking earlier, dissolved carbon dioxide is becoming a facet of winemaking requiring conscious attention and specification.   Whereas consumers in hotter countries like South Africa or Australia may regard a slight spritz in still wine as attractive and refreshing, consumers in cooler UK have less ‘need’ for this attribute.

Sainsbury’s are one of the few UK supermarkets to have had a CO2 policy for many years stemming from 1990s consumer perception that anything slightly fizzy in bottle was probably re-fermenting.  Howard Winn, their quality manager said: “Levels should be appropriate for the style of wine or organoleptically not apparent, so an oak-aged chardonnay would be expected to have a low level of CO2, but some of the Italian whites would have higher levels for added freshness.  Our specifications are about a maximum 400mg/l for reds, 600-800mg/l for whites, depending upon wine style.  Certain styles have intrinsically elevated levels – Muscadet at about 1g/l; Vinho Verde about 2g/l and Beaujolais Nouveau about 1g/l.”

Reductive winemaking

Reductive winemaking – winemaking which actively avoids oxygen – generally results in higher levels of CO2 in the wine, from retained fermentation-derived CO2 and from external sources of CO2 which have dissolved into the wine. 

During fermentation, naturally produced CO2 protects the fermenting must from oxidation.  After fermentation, a deliberate management policy must be adopted, depending on wine style.

Carefully controlled use of CO2 in the winery can optimise levels of oxygenation in the wine making process. Blanketed over the surface of wine, CO2 can help prevent oxidation and the growth of spoilage organisms.  But it dissolves more readily in wine at the lower temperatures usually used by winemakers who work reductively. This makes it a potential problem for red wines, which need much less dissolved CO2.  Additionally, as wine warms up, CO2 comes out of solution and pressure may build if there is no escape route.

Jürgen Hofmann, director of winemaking at Germany’s Reh Kendermann explains their reductive philosophy.  “We use CO2 in bottles to work more reductively. We add it in the tanks, in trucks, during filtration etc., to remove oxygen. We also use a lot of dry ice.  Its temperature is very low, so it is good for both cooling down grapes and keeping oxygen away; we also use it during skin contact, as it is very gentle, and for cold maceration in reds.   Dry ice is also great for reductive working in smaller wineries because it is easier to handle than moving heavy CO2 bottles around all the time. 

“One of the challenges of working very reductively is of CO2 is getting into the wine. Also we ferment quite cold which means more CO2 stays in the wine.  After fermentation you have between 1 and 2 g/l CO2 in the wine depending on the fermentation temperature. Additionally, white wines stored cool retain more dissolved CO2.”

Reh Kendermann are big exporters to the UK.  Hofmann says: “we are looking for around 1 g/l CO2 in the wine for the bottling of whites and less than 500 mg/l for reds.  We learned UK consumers are very sensitive to CO2.  We found the maximum level they tolerate is about 1g/l [the threshold of perception].  In Germany the consumer looking for more freshness, and a slightly spritzy style up to 1.3 g/l.”

Abrie Bruwer is quite fanatical about the use of CO2 with his sauvignon blancs; he pays very close attention to it throughout the winemaking and bottling processes, saying how diligent, careful people who work gently do find CO2 is an issue.  He says: “working with CO2 is a lifestyle thing.  You must think reductively all the time”

“The pyrazine flavours of sauvignon blanc are non-existent if you don’t use a lot of reductivity.  So, in the vineyard we use ascorbic acid (vitamin C) with SO2, which in combination scavenge oxygen.  Oxygen in the must destroys the pyrazines.  Ascorbic acid scavenges very quickly, binding with SO2.  This technique results in the subsequent wine being slightly bleached, which is why wines made using this method are pale.  But it also kills the yeast which gives us a window of 4-5 hours while the grapes reach the cellar.  Then we use CO2.

“We sparge with CO2 at the crusher. The maceration tanks for skin contact are filled with CO2.  All the pipes etc are filled with CO2, if we rack from tank to tank.  The cellar equipment is filled with CO2 the whole time.  CO2 is the easiest inert gas to use in the cellar because it’s heavy.

“Very little CO2 from fermentation escapes.  We use bottled CO2 at start of harvest, then we use the CO2 produced by fermentation.  You close the lid of the tank and fit a pipe from this tank to next one, making sure there are no kinks in the pipe, otherwise the pressure will build to exploding point.  We had a tank blow up one time when the CO2 couldn’t escape – the winery was very sticky!

“If you pump slowly and cold, the wine will still have 1500 to 1600 mg/l CO2.  We try to bottle below 1300 mg/l, so sometimes we need to blow off some CO2.  You don’t want to heat wine, it’s best to do it 10-12°C, but it doesn’t come off much below 13°C, so you could heat the wine above this or sparge with nitrogen, controlled with a flow meter.  A lot of care is needed because nitrogen can blow off flavours too.  But it’s a trade off.  You work gently and slowly to preserve flavour then when you want to bottle the CO2 is too high.  So you need to do something drastic like sparge with nitrogen, so maybe you pump slightly warmer, but you lose flavours.  

“We test CO2 before bottling and will blow off some if it is above 1400mg/l.  In Germany and hot countries, people are tolerant to a slight prickle on the tongue.  The UK palate is a little more sensitive – you prefer warm, still beer!”

Style specificity

Harrop emphasises a style-specific approach to CO2.  “Although some grape varieties e.g. riesling, sauvignon blanc, appear to support relatively high CO2 levels, a CO2 strategy should be predicated on wine style and vintage characteristics.”

Bruwer says the sauvignon blanc pyrazines are easily blemished by oxygen, but on his chardonnay, he pumps without CO2.  “We oxygenate the must pre-fermentation to encourage glycerol formation.  Some oxygen gives fuller, fatter wine.”

On reds, Bruwer uses CO2 during the cold soak – it means volatile acidity takes a little longer to form.  But during fermentation, Bruwer’s policy is to remove CO2: “we pump over with a fan blowing onto the cascade to give O2.  Yeast under hot conditions produces glycerol if O2 is present.  We want glycerol, rather than just alcohol, so we give O2.”

The 2003 Sancerre vintage provided Jean-Christophe Bourgeois of Henri Bourgeois with interesting choices.  He removed less CO2 in order to preserve more crispness and freshness, rather than add acidity, which had been permitted by special derogation in this hot vintage.  Bourgeois said: “Sancerre normally has about 1000mg/l CO2, but in 2003, we had about 1250mg/l CO2.  The higher CO2 left in the wine was a way of not adding acidity and to avoid adding more SO2.  The CO2 feels like freshness.  We taste before bottling to determine the best CO2 level.”

The case for red wine

Taylor confirms that noticeable CO2 in still red wine is most likely indicative of a problem. It might also indicate genuine attempts by wineries to make a wine as good for the consumer as possible, perhaps when the buying has been done on significant price pressure.

Noticeable CO2 could indicate a secondary fermentation, especially with the apparent trend to bottle some high volume reds with sufficient residual sugar to stimulate a refermentation.

Taylor says: “an increasing area is that of bottling red wine before it’s finished the malolactic fermentation.  This can cause big problems.  Not only do you get a slight spritz, but the bacterial damage can be significant.  A heavy deposit is formed which might remain in suspension giving a cloudy wine. As the bacteria utilise the available SO2, oxidation occurs.  There is much individual bottle variation.  This is occurring right across the quality spectrum.  Likely origins include the pressure to get wines to market earlier, or logistical pressure on tank space.”

Taylor sees this phenomenon on some very nice wines.  The bacteria can remain viable in bottle for some time, 12 to 18 months; it takes maybe a combination of temperature and a drop in SO2 to a point where the bacteria can work.

Pressure bottling

Philip Bailey, technical director at Corby Bottlers says their CO2 strategy depends on the winemaker.  “As a UK bottler, we liaise with the winemaker to fill sympathetically with the winemaking aims.  If a winemaker was to suggest e.g. 900 mg/l dissolved CO2 for their wine, then, depending on CO2 levels on arrival, we might add a little or take some out by nitrogen sparging, immediately prior to bottling.

“There is a safety aspect at bottling.  Glass manufacturers specifications typically suggest dissolved CO2 levels be kept below 1.2g/l to 2g/l depending on the bottle. Above this and there may be an increased risk of the glass exploding.” Above 1.2 grams per litre the gas tends to force the corks  (where used) out of bottled wines.

Bailey warned of the potential impact of high CO2 levels for bag-in-box.  “Above 600-800mg/l and the bag may swell when the temperature rises as CO2 comes out of solution. From the outside, it looks like a refermentation.” 

Regarding screwcaps, Harrop suggests virtually no CO2 escapes, so a wine bottled at 1.2g/l stays that way, so with a noticeable spritz, but “with 1.2 or 1.1g/l at bottling under cork, after shipment a wine generally loses 100-150mg over 4-6 weeks, depending on the quality of closure, diameter of bottle, temperature etc., so there’s no over-spritz.”

The significance of the temperature of storage and logistics has been reported elsewhere – as wine heats, for example when containers move over the equator, CO2 comes out of solution and risks pushing out driven closures.

Active use of CO2 is both the friend and the enemy of reductive winemaking and needs to be critically monitored throughout the winemaking process to ensure optimum wine health.

When must and wine is treated reductively, it becomes especially vulnerable to oxidation, so every process after the first reductive process must be done reductively.  CO2 use is an integral part of oxygen avoidance, as is cold temperature storage where CO2 retention is higher and risks include reductivity in the wine.  If left unchecked, the risk is for reductivity to progress to irreversible mercaptan form.

With the trends to make more reductive styles, to minimise the time from harvest to market and to drink the youngest wines available, the use of potentially more reductive closure types has implications for the final wine analyses immediately prior to bottling. If “too” reductive a closure is used any present reductivity may persist and develop further. A bit more time after bottling or a “less” reductive closure may allow more opportunity for gases to equilibrate inside and outside the bottle.   

A CO2 policy for still winemaking and bottling becomes increasingly relevant.


One Response to “Role of carbon dioxide in still wines”

  1. Sue Massey Says:

    Well said

Jump to the top of this page