Making one of the world’s favorite wine styles is a long and complicated process, Tom Jarvis reports.
If there is one wine associated with success and celebration it is, of course, Champagne.
At parties, weddings, and hundreds of other get-togethers, corks are popped by the million, as people express their joy by releasing a stream of bubbles through the neck of a bottle. And it makes no difference whether it’s a non-vintage, vintage, blanc de blancs, blanc de noirs, or zero dosage style of Champagne – it might be Dom Pérignon or it might be a supermarket brand – but the production method has remained the same for the past two centuries.
The «fizz» in traditional method sparkling wines (Méthode Champenoise until the EU outlawed the term on wine labels) is created during a second fermentation in the bottle in which it is sold.
This contrasts with basic carbonation (soda method), the Charmat method, which uses pressurized tanks, and the transfer method, where sparkling wine goes from bottle to tank for sweetness adjustments and clarification before being re-bottled. Méthode ancestrale is similar to traditional method but leaves some spent yeast in the bottle.
The Base Wine
It is often written that the base wine actually has more bearing on quality than the methods used to add carbon dioxide. However, base wine production may just involve the processes over which the winemaker has most control and greatest understanding.
Due to their northern location, Champagne makers routinely chaptalize (add sugar to the grape must) to reach 10.5 percent alcohol by volume in the base wine. In warmer regions, consistently higher grape sugar levels mean this is not required.
Grapes are usually pressed by variety, and in Champagne by vineyard origin also. Normally press houses are sited close to vineyards, to minimize phenolic extraction from crushed fruit sitting in picking bins. Whole bunches are pressed extremely gently so that minimal color, or bitter phenolics are released. The latter can inhibit the second fermentation – a big part of why relatively few sparkling red wines are made. Champagne was historically produced using wide, shallow Coquard basket presses, but modern, computer-controlled pneumatic machines are now the norm.
Grapes are pressed in several stages; juice from the first pressing (called the cuvée, which may be several cycles if using a pneumatic press) contains the most sugar, acid and delicate flavor compounds, and the least solids. The second pressing (première taille) has lower acidity and higher tannins, but richer fruit flavors. It is used either as a larger component of less expensive wines, or in smaller percentages to add body to finer bottlings. Subsequent press juice (aka the deuxième taille) can no longer be used in Champagne.
The varietal blend a large effect on base wines. Of the three principal Champagne grape varieties, Chardonnay wines can show lemon, biscuit and toasty characters and have high acidity which is important for wines spending a long time aging on their lees. Chardonnay is felt to contribute the most upfront intensity and lift, but the least length. Pinot Noir reverses these last two attributes, while Pinot Meunier is the middle ranker, producing wines that are round and fruity. The blender’s job is to deliver the optimum combination of upfront intensity and length of flavor, along with the desired balance of flavors, and create a consistent style.
Champagne can also be made from Arbanne, Petit Meslier, Pinot Blanc and Pinot Gris. However these collectively account for only 0.3 percent of production.
Further varieties are used in other regions. One example is Cava, which originated in the 1870s in Penedès when only local grapes were available, though Parellada, Macabeo and Xarel-lo can make serious wines. Deutscher Sekt traditional method sparkling wine can be made from Riesling in Germany (standard Sekt uses the Charmat method on base wines which may originate throughout Europe). Local speciality grape varieties can be used throughout France in regional Crémant wines, so called because they were historically bottled at lower pressures so the palate was more «creamy» than «sparkling».
The blending of the various base lots is a crucial stage in determining quality. Typical larger Champagne houses will have hundreds, including older wines from different grape varieties or locations. Even a small grower may work with 20 or more components, with varying aromas, flavors and structural profiles. Numerous trial combinations are made to select the best blend. For non vintage releases this takes into account the need for a consistent house style across releases.
Most non-vintage cuvées are blends based on current vintages, blended with older reserve wines with developed characters to add complexity and consistency. A standard cuvée may feature 80 percent or more of the current vintage, while the same producer may produce a higher priced wine with a larger reserve wine component. A few grower Champagne firms produce what is essentially a non-vintage Champagne from a single vintage, differentiating their vintage wine by only producing it from the best base wines in the best years. In New World countries, it is more likely that non-vintage wines simply combine consecutive recent vintages, as the growing conditions across harvests are more consistent.
Krug Grande Cuvée is an example of a multi-vintage wine, an NV that combines more equal proportions of several reserve wines, usually from better years. These wines tend to be priced on a par with vintage wines, with which they share some characteristics.
Most base wines today are fermented and aged in stainless steel. Bollinger are unusual in aging older (reserve) base wines under slight pressure in magnums.
© CIVC | Riddling (left) helps break down the yeast, before the bottle is topped up with a mix of wine, sugar and acid during dosage.Oak
Oak fermentation was once the norm for Champagne, old barrels that impart minimal oak influence. Krug still ferment wines in traditional 205-liter barrels called pièces de Champagne. According to wine writer Michael Edwards, only a handful of producers used oak in the 1990s, but now perhaps 100 producers use it, fully or in part, either to ferment wine, age reserve wines, or even in the production of the wine component of the dosage addition.
Some producers believe the use of new oak barriques has no place in the region’s wines. Champagne authority Tom Stevenson has stated that new oak can have a positive role if limited to 5-15 percent of a skilfully-blended final overall blend, with larger oak vessels preferable to barriques or pièces. Stevenson has tasted very oaky wines, aged for just a few months in small casks, which were 15 to 20 years old. According to Stevenson oak fermentation (even with new barrels) usually provides subtler more complex aromas than oak aging in used wood. Krug and Bollinger Vintage are fermented in small oak barrels, but respectively aged in tank and magnum bottles.
The Henri Giraud house is known for boldly oaked wines but, uniquely, uses wood from the nearby Argonne forest. This enhances traceability and sustainability, and the oak is deemed gentler than that from more famous forests.
Malolactic Fermentation and Lees Stirring
A proportion of base wines may go through malolactic fermentation to add aroma and flavor complexity, and to soften acidity. Care is needed to avoid overpowering the subtle aromas from yeast autolysis and fruit development. Houses such as Krug and Bollinger prefer to avoid MLF, to help to retain a degree of freshness to balance long aging and the use of oak. Lanson also avoids it, balancing any tartness with an above-average dosage (sugar addition) in the Brut cuvée, whose fruity yet tangy style can excel in tastings. Some experts feel that oaked wines should also undergo malolactic fermentation in wood to maximize flavor integration.
During the aging period for the base wine, lees stirring can maintain high protein levels to help autolysis in bottle. It also adds weight and texture to the finished wine.
Demand for pink fizz has risen considerably since 2000. This may account for variation in quality, despite being priced at a premium over comparable white cuvées. Better examples show clear aromas of red fruits rather than only visual differences.
Two main methods of rosé production may be used in Champagne. The more traditional and most widely used approach is rosé d’assemblage, where between 5 to 20 percent (15 seems to be the average) of locally produced red wine is added to the blend. Bigger houses usually make their own red wine and, in response to the rise in demand, marques such as Veuve Clicquot and Billecart-Salon have recently invested in vineyards earmarked for this purpose. But many still buy it in from specialists, usually located in the Aube in the south of the region. The wine is specially vinified for low tannins to minimize effects on the second fermentation.
Secondly, the saignée (bled) method creates a pink base wine; the fermenting juice spends the first few hours of fermentation macerating on its skins before it is bled off. Laurent-Perrier is the best-known practitioner.
A few producers, usually Blanc de Blanc specialists like André Jacquart, will assemble a blend of white and rosé components to make a paler rosé. More generally, Chardonnay is (perhaps counter-intuitively) an important component in many rosés, valued by winemakers for aromatic lift and freshness.
To add sparkle, the finished still wines are filled into heavy pressure-tested bottles and a liqueur de tirage is added. This is a solution of selected yeasts, sugar, grape must or wine in the correct proportion to produce the desired carbon dioxide content. Around 24g to 26g of sugar per liter achieves the typical 5-6 bars of pressure in a Champagne bottle (three bars of pressure is the legal minimum for EU sparkling wine, rising to 3.5 bars for «Quality Sparkling Wine»).
Small amounts of bentonite and sulfur may be added to ensure that yeast does not stick too securely to the bottle, and to prevent harmful oxidation and biological degradation. It is not thought the type of sugar consumed by yeast effects the final taste.
The choice of stopper dictates the degree of oxidation during maturation on lees. Nowadays a specially designed crown cap is normally used, allowing minute amounts of oxygen to enter and carbon dioxide to escape. Even crown caps vary in permeability, so a change of supplier might affect finished wines.
Usually a bidule is attached to the cap; this is a small empty polythene cup which improves the seal and prevents metal contacting the wine. It also helps disgorging, as lees will have been collected in the cup for easier removal.
Traditionally a cork with a large single clip called an agrafe was used. One or two houses persist with these for their top wines; Henri Giraud’s Fût de Chêne is also sold with an agrafe. These corks are less permeable than crown caps, but are more expensive and difficult to remove by machine.
Bottle storage during second fermentation and lees aging should be cool, with minimal temperature fluctuation and lighting – essentially the same conditions under which consumers are advised to keep purchased bottles. During the second fermentation carbon dioxide forms and stabilizes, and stable temperatures are key to develop quality mousse.
The rate of the second fermentation depends on a complex interaction of numerous factors including yeast type, cellar temperature, and any number of base wine characteristics. Champagne house cellar temperatures tend to be around 10 to 12 degrees, when second fermentation may last six weeks. In a cellar at 15 degrees the process may only take a fortnight. The fermentation progress is constantly monitored by analysis of sugar reduction and internal pressure. During second fermentation, the alcohol level by volume of the base wine increases by around 1.5 percent.
© Wikipedia Commons/CIVC | After riddling, the yeast remains are removed in the disgorgement process.The Bubbles
The finesse of the mousse is widely associated with quality in sparkling wine. Mousse is a chain of bubbles of trapped carbon dioxide, which nucleate around tiny solid particles (such as cleaning cloth fiber) in the wine or on the bottle or glass surface. Trace levels of detergent in a glass can inhibit mousse, while good Champagne flutes have a scratch in the bottom of the cup to aid bubble formation.
When a bottle is opened, the pressure drops from around six atmospheres to ambient atmospheric pressure. This reduces carbon dioxide solubility and liberates around five liters of carbon dioxide per bottle. The gas is metastable, liable to disturbance by solid particles triggering bubble nucleation. It escapes gradually as free energy is limited.
Achieving a fine, elegant mousse is connected to extended lees aging (which many higher priced wines undergo). The process is thought to enhance mousse stability by breaking up polymers, and increased complexity of protein and sugars present in a wine after lees aging. This may give some strength and elasticity to a bubble, limiting the likelihood of fusion with neighbors, so keeping bubbles small.
A standard Champagne cuvée may have finer bubbles than some traditional method wines from other countries. But the latter may achieve comparable finesse if aged on yeast for the same period of time in a cool environment.
Bubbles act as a flavor delivery system, as well as providing fizz on the tongue. Dozens of different active aromatic or aroma precursor compounds are much more likely to be found in the bubbles than in the liquid of the wine itself. Thus differences in pressure and mousse quality also impact on flavour perception.
Following secondary fermentation, most quality traditional method sparkling wines are aged for several years in contact with the spent yeast. The yeast autolysis that occurs in this period is a key component in the character of the finished product.
Autolysis is triggered when the yeasts added for the second fermentation have consumed all the sugar and nutrients from the liqueur d’expédition and base wine, and then used up their own internal energy resources. Yeast cells then die and degrade, releasing numerous compounds from the cell walls and cytoplasm (the gel within the cell). These products have varying levels and characteristics of aroma and flavor, and include peptides, lipids, fatty acids and esters. Research continues into the precise mechanism and products released, and the attribution of specific aromas to these products.
The process creates aromas and flavors associated with nuts and yeast, bread and brewing products, which usually become more obvious with extended lees aging. These yeast-derived aromas and flavors are usually accompanied by developed fruit aromas such as honey and vegetal notes, as the wine slowly oxidizes via the stopper.
Legally, Champagne must be aged in bottle for 12 months on lees and 15 months total before release. Even less expensive non-vintage Champagnes will likely see 2-3 years; for quality labels 3-4 years is common. A vintage wine from a large house may mature for six years on lees, or 8-10 years or more for deluxe cuvées. Bollinger RD (Récemment Dégorgé / Recently Disgorged) is a very expressive example of extended lees maturation.
Yeast choice is important both for base wines and the secondary fermentation. Cultured yeasts developed in Champagne generally have a minimal effect on wine aromatics, and flocculate well (stick together for easy removal). However performance in developing a fine mousse is widely regarded as the key attribute.
Rémuage, Dégorgement, Dosage (Clarifying and Sweetening)
Riddling (rémuage) moves sediment to the neck of the bottle for removal, clarifying the wine. In the original manual process bottles are inserted into a hinged pupitre (desk). They are slowly tipped towards upside-down vertical over 15 days, with one-eighth turns of the bottle being made by the riddler to allow the yeast to break contact and slip down the neck. Nowadays this is mainly practiced for tourists, and instead automated gyropalettes perform the tipping and turning simultaneously.
Eventually bottles which rested on their side during second fermentation are stored neck down, awaiting the next process. Bottles may be stored this way for extended periods as yeast-wine contact is limited, emphasizing secondary characters from aging.
Once rémuage is completed, disgorgement (dégorgement) takes place. The sparkling wine is refrigerated to 4-10 degrees Celsius to avoid carbon dioxide leakage, and inverted. The bottle neck is dipped into a brine solution, freezing the sediment in the neck and bidule (if used). The bottle is set upright and opened; pressure ejects the bidule and ice plug, and with them the sediment.
After disgorgement the bottle is topped up with the liqueur d’expédition – typically a mixture of wine, sugar, sulfur dioxide, citric acid and copper sulfate. Each firm will have its own formula. The sweetness of the Champagne is set via the sugar level (dosage) of the liqueur. Early 19th Century Champagnes were mostly sweet; the fashion for drier styles may have developed in Britain around mid-century. For sweetness levels of Brut and other Champagne designations, see our Champagne Label Information page.
Some wineries will blend base wines with a particular sweetness level, and particular cuvées in mind. But some producers, perhaps in higher-volume markets, modify an intended Brut cuvée to produce a sweeter style as demand requires.
Recently, houses such as Bollinger and Jacquart have installed jetting systems (used for a decade by Ferrari in northern Italy), which inject small amounts of wine just before corking. The foam generated expels oxygen in the space above the fill level. This ensures consistency in oxygen content, which can vary widely across bottles after disgorgement, and reduces the need for sulfur dioxide additions which can lead to off aromas.
Nearly all Champagnes undergo second fermentation and rémuage in a 75cl bottle, or occasionally magnums. For other larger and smaller formats, transversage takes place, where the bottle is disgorged and the wine is moved to a tank for dosage before being re-bottled in a different-size format. The process is similar to the transfer method for starting the second fermentation.
While non-vintage cuvées are not always associated with aging, «fresh» batches can suffer from bottle shock and the effects of shipping. Most benefit from being drunk within a year of disgorgement.