A Review of Methode Champenoise Production (How Champagne is Made)
A Review of Methode Champenoise Production (How Champagne is Made)
Centuries of experience have enabled the sparkling wine producer to refine the art of bottle-fermented sparkling winemaking to the system known as méthode champenoise. This system, however, is not a rigid one. Certain steps are prescribed by law in France, while few are required in America. Within certain guidelines there is considerable variation in production philosophy and technique regarding méthode champenoise. Stylistic decisions are vast and include viticultural practices, cultivars, maturity, pressing vs. crushing, types of press and press pressures, press fractions, phenol levels, use of SO2 and the oxidative condition of the base wine, yeast for primary and secondary fermentation, barrel fermentation and aging, fermentation temperatures, malolactic fermentation, post primary fermentation lees contact, age of cuvée, reserve wine, blending, time spent sur lie, nature of the dosage, and CO2 pressure. This publication describes production philosophy and practices of méthode champenoise producers.
Author: Bruce Zoecklein, Associate Professor and Enology Specialist, Department of Food Science and Technology, Virginia Tech
Table of Contents
Liqueur de Tirage
Some Terms Used in Méthode Champenoise Production
The array of viticultural parameters affecting méthode champenoise palatability is broad. Environmental and viticultural factors influencing cuvée chemistry include meso climate, canopy climate, soil moisture, temperature, berry size, rootstock, asynchronous development, fruit maturity and leaf area/fruit weight or fruit weight/pruning weight. For the producer, understanding the relationships between vineyard management and wine quality may be even more difficult for sparkling wines than for table wines. Cuvées are evaluated and blended when they have the better part of their lives ahead to age and develop. This requires considerable insight and may tend to obscure the relationships between vineyard management activities and sparkling wine palatability.
With the exception of the Mosel of Germany, the Champagne region is the most northern significant grape producing macroclimate in the world. Epernay is on the 49° parallel, the same as the Washington/Canadian border. The total degree days in Champagne average around 1890, as compared to 2,340-2,610 for Napa and 2,160-3,600 for Sonoma County. The daytime temperatures seen in Napa are higher with cooler nighttime temperatures than in Champagne. As a result of increased solar radiation, grapes tend to ripen more quickly and potentially reach a higher level of maturity in Napa than in the Champagne region. This is true for most other regions in California as well. In California, grapes suitable for producing highly palatable sparkling wines are generally grown in regions I to III (Amerine et al., 1980). In warm regions, great care must be given to harvesting early enough to retain desirable acidities and pH values. A primary problem in warm climates is the production of a base wine that is not too heavy in body or varietal character, too alcoholic, or too colored. Warm climate wines, by and large, offer more definitive fruit flavors, less complexity and lower acidity than Champagne, and they develop more quickly.
Among the viticultural options affecting grape components either directly or indirectly, mesoclimate (site climate) is considered one of the most important. Mesoclimate has been divided into two general temperature zones, Alpha and Beta (Jackson, 1987). In Alpha zones, maturity occurs just before the mean monthly temperature drops to 10°C (Jackson, 1991). Specifically, Alpha zones are those where the mean temperature at the time of ripening, for a particular variety, is between 9-15°C. In warm climates, the length of the growing season is more than adequate to ripen most grape varieties which, therefore, mature in the warm part of the season. In Alpha zones, day temperatures are moderate and night temperatures are usually cool, creating desirable conditions for the development of important secondary grape metabolites. On the other hand, in Beta zones the majority of grapes ripen well before temperatures begin to drop. Specifically, Beta zones are those with a mean temperature above 16°C at the time of ripening for a particular variety.
It is generally accepted that a cool climate that allows the fruit to stay on the vine longer while retaining desirable acidities is important in the production of base wine which will develop the needed complexity during aging sur lies. If the field temperatures and heat summation units were the sole parameters affecting the grapevine climate, then we need only consider the macroclimate in analyzing the temperature effects on quality. The real situation, of course, is not that simple. Solar radiation, wind velocity, and to a lesser extent, sky temperature can give ranges of berry temperatures of more than 15°C above to 3°C below the air temperature (Kliewer and Lider, 1968). These variables are further influenced by row orientation, training system, trellis height and vine vigor. There are several reasons why comparisons between climates, secondary metabolite production, and grape and wine quality have been confounded. First is the effect of crop load. Crop load, and most significantly, the ratio of exposed leaf area-to-crop load, can have a profound effect on the rate of maturity. Fruit maturity and the rate of fruit maturity can influence grape and wine quality. Another factor often overlooked is asynchronous growth (either berry, cluster or vine) (Due, 1994). This will also delay maturity, yet few comparisons of climate and wine quality have taken this into account.
To some m<éthode champenoise producers, a high malic acid level in the grape is considered a desirable characteristic. Malic acid is principally influenced by maturity, crop level, and temperatures (day and night). Short term exposure to high temperatures is significant to fruit malic acid levels, to say nothing of the effects on phenols and aroma components. The effect of brief exposure to high temperature may raise serious doubts about how one integrates, over time, climatic parameters such as heat summation to fruit composition. For a review comparing climate factors see Bloodworth (1976), Jackson (1995), Poinsaut (1989), Pool (1989), Reynolds (1997), and Riedlin (1989).
Some of the many cultivars utilized in various growing regions for méthode champenoise are given in Table 1. Chardonnay, Pinot noir, Pinot meunier, and Pinot blanc are among the more popular varieties. The concentrations of amino nitrogen, acetates, diethyl succinates, and organic acids are strongly affected by the varieties used in base wine production.
Grapes used in the Champagne region are almost exclusively Pinot noir, Chardonnay, and Pinot meunier. There is a tendency for Pinot meunier to be replaced by Chardonnay or Pinot noir, both of which give greater yield and produce higher quality (Hardy, 1989). Chardonnay gives life, acid, freshness, and aging potential to méthode champenoise. Care must be taken to avoid excess maturity (in warmer climates particularly), which produces a dominant, aggressively varietal character. Warm climate Chardonnay cuvées may suffer from a narrow flavor profile, high "melony" aroma notes, and lack of freshness, liveliness and length. Additionally, rich fertile soils can cause this variety to produce foliage and grassy aromas. When combined with Pinot meunior, Chardonnay has a greater capacity to age harmoniously and for a longer time (Hardy, 1989).
Table 1. Varieties Used for Méthode Champenoise
|Cool Regions||Warm Regions||Hot Regions|
|Pinot noir||Chenin blanc||Parallada|
|Pinot blanc||Meunier||Pinot noir|
Source: Dry and Ewart (1985). Regions based on UCD heat summation units.
Pinot noir adds depth, complexity, backbone, strength, and fullness (what the French call "carpentry") to méthode champenoise wines. These generalizations are broad and become nebulous when one considers, for example, that there are over 82 different clones of Pinot noir in the Champagne viticole. Clonal selection continues. Pinot noir is seldom used by itself, even in Blanc de noirs. Uneven ripening in Pinot noir is often a problem for producers trying to minimize excessive color extraction. Pinot noir at the same degrees Brix as Chardonnay generally has less varietal character.
Pinot blanc, like Pinot meunier, is a clonal variant of Pinot noir. It is generally neutral, but has some Chardonnay traits with a bright fruit character that is somewhat thin. Pinot blanc, like the Pinot meunier used in France, ages more quickly than Chardonnay, yet adds fullness, body and length to the finish. It may be a desirable blend constituent. Pinot blanc has a tendency to drop acid more quickly on the vine and, like Pinot meunier, usually has a lower titratable acidity than Chardonnay. It is, therefore, harvested somewhat early.
The chemistry at maturity of several California sparkling wine cultivars is given in Table 2. Grape harvests should be based upon a determination of desired style. Méthode champenoise producers harvest based upon the flavor and aroma of the juice, as well as analysis of °Brix, acid and pH. Producers are generally striving for base wines that are clean, delicate, not varietally assertive, yet not dull or lifeless. A desired cuvée is one with body, substance, and structure. Immature fruit produces wines that are green or grassy in aroma. Overripe fruit can produce a base wine that is excessively varietal or assertive. Often the producer is looking for bouquet in the finished product, not for extensive varietal aroma. This is a stylistic consideration. However, the winemaker should never lose sight of the effect carbon dioxide has on one's perception of wine character. The "sparkle" significantly magnifies the odorous components of the wine. Early harvest in warmer climates helps minimize excessive varietal character, which can be overpowering. Changes in aroma range from low intensity, green-herbaceous characters toward more intense fruit characters. Chardonnay aroma can be described as melon, floral, pear or smokey; Pinot noir as strawberry floral, tobacco, toffee; and Pinot meunier as confectionery. In warm climates, mature fruit aromas/flavors can be noted when the sugar concentrations are low (- 16°Brix). The CIVC bases its picking decisions on sugar: acid ratios with the preferred ratio between 15-20. This means grapes reach optimum maturity at 14.5 - 18°Brix and a titratable acidity of 12-18 g/L (tartaric). At this acidity, the malic acid is 50-65% of the total acid content. The traditional importance of acid may be partly the result of the fact that, in Champagne, sugar addition is legal, but acid addition is not. At bottling, 11.5% alcohol (v/v) is desired. Alcohol helps foam and bubble retention. Also, in warm climates, a sugar: acid ratio of 15-20 may be reached after some mature fruit flavors have developed (Jordan and Shaw, 1985).
Table 2. Fruit Chemistry of Some California Grapes for Méthode Champenoise*
|Chardonnay||Pinot Noir||French Columbard||Chenin Blanc|
|Titratable Acid g/L||11.0-14.0||10.0-13.0||12.0-14.0||10.0-11.0|
*Average of several viticultural regions.
Return to Table of Contents.
The desirable chemical attributes of the cuvée usually include alcohol (between about 10.5-11.5), high acid, low pH, low flavonoid phenol content, low aldehydes, low metal content, low volatile acidity, and little color (See Tables 3 and 4). Many producers carefully hand-harvest into small containers (30-1000 pound boxes or bins) to avoid berry breakage and then bring the fruit in from the field as quickly as possible. The least possible skin contact is sought, particularly with red varieties used for Blanc de Noirs. Proximity to the processing facility is, therefore, important. This aids in preventing undue extraction of phenolics from berries possibly broken during transport. Oxidation will reduce desirable aroma/flavor and provide excessive phenols which may cause bitterness and reduced aging capacity. Grapes must be harvested as cool as possible to avoid excessive phenolic pickup and loss of fruit quality. This makes long transport of warm, machine-harvested fruit undesirable for méthode champenoise.
Grapes are weighed and either pressed or crushed and pressed. Crushing and pressing may be satisfactory, provided the contact of the skins with the juice is brief. For premium méthode champenoise, however, the grapes are usually pressed rather than crushed and pressed. Lack of skin contact produces a more elegant, less varietally dominant base wine. Skin contact releases more aroma, but may also extract courser undesirable components. There is, of course, a yield reduction by pressing the fruit rather than crushing and pressing. The economics, the targeted market, and the style desired must be carefully reviewed.
As Figure 1 indicates, here are three juice zones in the grape berry: the juice of the pulp (Zone 1), the juice of the pulp area around the seeds (Zone 2), and the juice from just beneath the skins (Zone 3). In order to obtain the desirable cuvée chemistry, traditional producers of méthode champenoise press rather than crush and press. The point of rupture is usually opposite the pedicel. The intermediate zone (1), which contains the most fragile cells, is first extracted before the central zone (2) and finally the peripheral zone (3) (Dunsford and Sneyd, 1989). The concentration of tartaric acid is highest in zone 1 and lowest in zone 3, and hence should be extracted initially. Malic acid concentration decreases from the center (zone 2) to the skin, and so is also extracted fairly quickly. By contrast, the concentration of potassium, the dominant cation, is highest in zone 3, which is extracted last. A juice extracted from the first two zones will, therefore, have the highest acidity, lowest potassium, lowest pH and the lowest susceptibility to oxidation which will result in a wine of greater freshness.
The goal is usually to preserve the integrity of the berry so that the components of the different zones are not mixed. Thus, mechanical harvesters and crushers are not used. Owing to the way in which the sugars and acids are positioned in the grape, the juice flowing out of the berry comes from the juice of the pulp during the early stages of pressing and is usually better suited for méthode champenoise. Conveyors and delivery systems that may break the berries prior to either pressing or crushing and draining tend to extract more phenolics and may be considered undesirable. One California sparkling wine house developed a vacuum system capable of moving 20 tons/hour of whole grapes into the press. This prevents berry breakage and can reduce the phenol level by 100 mg/L G.A.E. or more (Fowler, 1983a, b).
Table 3. Composition of Eight Successive Fractions From Chardonnay Grapes in a Champagne Press
acid tartrate (g/L)
|Vin de cuvée||1.||200.||193.||7.9||2.98||6.12||4.71|
|Vin de taille||Deuxieme taille||7.||2.70||191.||5.1||3.43||4.10||6.55|
Source: Francot (1950).
Table 3 shows the chemistry of various press fractions from a study conducted in Champagne (Francot, 1950). In Champagne, only the first 2,666L (70 gal) extracted from a marc (4,000 kg or a little more than 8,800 lbs) has the right to the appellation. At least several press fractions are taken, fermented and aged separately. Some of the later press fractions may be blended with the primary fractions as a result of economic and/or sensory considerations.
Table 4. Method of Fractionating a 4,000 kg Lot of Champagne Grapes.
|The 1st Taille||400||103|
|The 2nd Taille||200||52|
Source: Hardy (1989)
Table 4 summarizes the volume breakdown of the fractions frequently separated in Champagne. The first fraction contains dust and residues and is frequently oxidized as a result of inadvertent bruising during harvest. The cuvée portion is the best for sparkling wine production, being the least fruity, highest in acidity, and sweetest while not being oxidized. Fast pressing risks higher extraction of polyphenols. Juices extracted slowly at low pressure to give low solids are therefore less vulnerable to oxidation. The integrity of the pressing can be measured by comparing the differences in titratable acidity (ΔTA) between the fractions (Dunsford and Sneyd, 1989).
ΔTA (Cuvée - 1st taille)
= ΔTA (1st - 2nd taille)
== 1.5 g/L tartaric acid
Table 5. California Pinot Noir Press Fractions*
|Total Phenols (mg/L)|
*Data averaged from several sources.
Table 5 gives press data for a California Pinot noir. Segregation of press fractions is frequently based upon taste, which is affected by the significant drop in acidity with continued pressing following approximately 110 gallons per ton. Each press fraction differs in acid, pH, and phenolic and aroma/flavor components. In years of Botrytis degradation of greater than 15% of the berries, a first press fraction of about 10 gallons per ton is also separated. Crusher-stemmers mix the juice fractions and can result in < or = to 100 mg/L more phenolics than pressing whole grapes.
The trend in the sparkling wine industry is to employ tank presses, champagne ram presses, and traditional basket presses. The champagne basket press of cocquard is still used by some houses in Europe. This unit is unique in that it has a very shallow maie or press basket, rarely over two feet deep, with a diameter of 10 feet. The shallowness of the base relative to its width allows for grapes to be spread out in a fairly thin layer which reduces skin contact with the juice as it flows through the pressed mass of grapes. Thus, less press pressure is required.
The level of total phenols and the types of phenols present are a function of the design of press and press pressures among other factors. White wines with a total phenol count of 200 mg/L G.A.E. can expect to have approximately the following constituents: 100 mg/L nonflavonoid caffeoyl tartrate and related cinnamates; 30 mg/L nonflavonoid tyrosol and small molecular weight derivatives; 50 mg/L flavoinoids - especially catechins (flavor 3 diols)-and flavon polymers (tannins); and 15 mg/L SO2 and other interferences (Singleton, 1985). The nonflavonoid fraction is relatively constant in the initial pressing of white and red grapes because these compounds are present mainly in the easily extracted juice. The nonflavonoid fraction of cuvées not exposed to wood cooperage totals about the same as that in the juice. There is, however, considerable modification of phenols, and some may be lost or gained with aging (Singleton et al., 1980). Most nonflavonoid phenols are individually present below their sensory threshold, but their additive effects are believed to contribute to bitterness and spiciness.
Flavonoids such as catechins are extracted from the skins with increased press pressure and may vary with the type of press employed. Catechins account for most of the flavor in white wines with limited skin contact. Vin de cuvées (first press cuts) produced by low press pressures and thin layer presses can be low in total phenols, and particularly in flavonoid phenols, resulting in low extracts. This is an important production consideration. In Bruts especially, finesse must be in balance with the liveliness and the body of the wine. An extract of approximately 25 g/L gives body without heaviness (Schopfer, 1981). Moderate pressures or combining portions of later press fractions are methods of stylistic input that can affect such things as the tactile base of the aroma/flavor character of the cuvée. Most producers are looking for delicate aroma/flavors, which are associated with the initial juice extracted. Thus, a low volume gives a base wine that is low in extract and may, therefore, be elegant but lack depth.
No separation of the stems need occur before pressing. The stems insure efficient and improved draining and pressing of the whole grapes at lower pressures. Ultimately, this aids in obtaining a higher quality, more delicate first-cut press juice. Francot (1950) found that the Williams press produced juice with composition similar to the traditional basket press. Unlike the basket press, newer tank presses are pneumatic, give complete control, higher yields, produce less nonsoluble solids, low phenols, and require much lower press pressures (Downs, 1983). Low pressure minimizes the chance of macerating the stems and releasing bitter compounds into the juice. Gentle pressing of cool fruit extracts fewer flavonoid phenols. These compounds are responsible for astringency, bitterness and color. The juice near the skins and seeds, released by heavier press pressures, has more intense aroma/flavors and more flavonoid phenols. A tank press can press to dryness at two atmospheres or less and take press cuts. The rules of thumb in Champagne for pressure maxima during pressing are:
the cuv ée extraction at < 1 bar;
the first taille (1°T) at < 1.2 bar;
and the final fraction (2°T) at < 1.4 bar
Many ram-type presses require higher pressures to reach dryness. Filling the press with whole clusters reduces the press load. For example, a Bucher 100 RPM tank press that is rated for a charge of 20 tons will hold about 12 tons of whole clusters.
Pressing Chardonnay and Pinot noir may produce an average yield of 140 and 120 gallons per ton, respectively. The Chardonnay grape contains slightly more pulp than the Pinot noir. As stated, press fractions are often segregated by taste by monitoring the reduction in juice acidity. For Chardonnay and Pinot noir, a dramatic drop in acidity occurs between the extraction of 110-120 gallons/ton.
For red varieties such as Pinot noir and Pinot meunier, care is often taken to avoid excessive color extraction. Excess color will affect the sparkling wine character, degree of foaming, and rate of secondary fermentation (Schanderl, 1943). Color extraction is minimized by pressing cool fruit and segregating pressing fractions. The ability to increase the extraction of colored vs. noncolored phenols may be an advantage in producing sparkling rosés. In the production of rosé by cuvasion it is essential that color extraction occur without extraction of excess astringent phenols. The use of cold soak with or without pectinolytic enzymes helps to attain this goal (Zoecklein et al., 1995). The other method of producing a sparkling rosé is by rougissament, or blending. Subsequent color modifications may occur in the dosage stage to produce a sparkling rosé which is said to "reflect the color of rubies."
The Premier taille (Table 3) is fruitier, less fresh and less elegant than the Vin de cuvée. The later press fractions possess the following attributes: high pH, excess color, high total phenolic content, often excessive varietal character, harshness, higher nonsoluble solids, and a lesser quality aroma. The harshness, color, and nonsoluble solids of later press fractions can be reduced by fining with protein agents, occasionally in conjunction with bentonite and kieselsol. All or portions of the second press fractions may be blended with the primary fraction due to sensory and economic necessity. The third fraction is seldom employed in premium méthode champenoise production. For a review of m»thode champenoise grape handling, see Hardy (1989) and Dunsford and Sneyd (1989).
Sulfur dioxide is added to the juice expelled from the press but never directly into the press in order to avoid extraction of phenols. Although it is considered desirable to use SO2 to help control oxidation, there is no industry consensus regarding optimum amounts. In the United States, 30 mg/L is added to the first cut press fraction, though such a decision must be based upon the freedom from rot, juice chemistry, temperature and malolactic fermentation desires.
Phenols are oxidized in the absence of sulfur dioxide and, therefore, some pass from the colorless to the colored or brown form. This results in some juice browning. Less soluble or insoluble phenols precipitate and may be removed during fermentation due to the absorbent capacity of yeast. Muller-Spath (1981) originally suggested the desirability of low sulfur dioxide additions (20-25 mg/L) to the juice under the right microbiological and temperature conditions to encourage some oxidation. Singleton et al. (1980) showed that oxygenation of must for white table wine production increases resistance to further browning but results in less fruity wines. The use of sulfur dioxide in base wine production may be important to minimize oxidative loss of aroma precursors needed for bottle aging (Hardy, 1989).
The press juice fractions are often cold-settled (debourbage) or centrifuged to reach a nonsoluble solids level of between 1/2-2 1/2% prior to fermentation. The primary press fraction from a thin layer press, such as a Bucher, may already be sufficiently low in nonsoluble solids. Grape solids are removed to minimize extraction of phenols that may occur during fermentation. This is frequently accomplished with the aid of pectinolytic enzymes. Bentonite is usually not used in the primary juice fractions (Munksgard, 1998). There is a significant reduction of yeast levels from centrifuged juice (95%) vs. cold settled juice (50-60%) (Linton, 1985). The ability to rapidly settle is the result of the low pH in the primary press fractions. Some producers use prefermentation juice fining to aid settling and to modify the palate structure of the base wine (Zoecklein et al., 1995). The 1sttaille often receives 60-70 mg/L SO2 and 50 g/hL bentonic/casein (Hardy, 1989).
The lower the nonsoluble solids content and the cooler the fermentation, the greater the production and retention of fatty acid esters (Williams et al., 1978). These compounds are responsible for the fruity, floral, aromatic nose of wines produced under such conditions. Some producers choose to ferment their cuvées warm (65-70°F) to reduce the floral intensity, thus making a more austere product. Elevated fermentation temperatures are desirable if a malolactic fermentation is sought. Vinification at 55-60°F is not uncommon in this country. Many producers check the nitrogen status (total and NH4 N) of juice prior to fermentation and make adjustments accordingly (Zoecklein et al., 1995). A standard addition of 5-10 g/hL of diammonium phosphate is widely used in Champagne. An addition of 10-25 g/hL of bentonite is made during the primary fermentation of the cuvée by some (see protein stability/bubble size section, pg. 9). Higher additions of up to 150 g/hL of a bentonite/casein mixture is often added to the "tailles" or to the first cuvée fraction when a significant amount (greater than 15% of the berries) of rot is present.
The yeast employed is occasionally the same for the primary and secondary fermentation. Sparkling wine yeasts are selected for their ability, among other things, to produce esters. Using the same yeasts for both fermentations can result in an end product that is too floral and too high in volatile components. Those yeasts often used for primary fermentation include Montrachet UCD 522, Pasteur Champagne UCD 595, and California Champagne UCD 505, among others. Yeasts infrequently used for primary fermentation include Epernay -2, Steinberg, and French White (Bannister, 1983).
The primary fermentation is generally conducted in stainless steel. Some European houses use small wooden casks and barrels to ferment all or part of the cuvée. Those who suggest that greater finesse and elegance results from wood are countered by the majority who fear the wine will pick up excess tannin and color. Barrel fermentation results in added structure, often without significant harshness or astringency. Henry Krug ferments their entire vintage slowly at low temperatures in oak vats, believing this to add more bouquet. This is consistent with their desired style, which is full flavored, mature tasting, and complex.
For product consistency and temperature and biological control, some producers blend a percentage of the previous year's cuvée into the fermenting juice. Reserve wine can also be added during assemblage or blending and may be a component of the dosage. Such practices are based upon production and vintage dating considerations. In the United States, vintage labeling requires that at least 95% of the wine comes from the vintage year.
Following primary fermentation, the goal of many méthode champenoise producers is to process the cuvée for the secondary fermentation as rapidly as possible. This enables the wine to reach the consumer sooner and also takes advantage of the nutrient-rich young cuvées that support the secondary fermentation. Others counter that there is no need to rush the cuvée into the second fermentation. These winemakers usually prefer to allow their base wines to age and develop, noting that the secondary fermentation is a rejuvenating step.
Protein and protein-like fining agents can be used to clarify and lower the phenolic content of the base wines. Isinglass and gelatin are the most common agents. Schanderl (1962) recommended the use of polyvinyl-pyvrolid one (PVP) to remove polyphenolic compounds from the base wine. It should be noted that juices are much more forgiving of the harsh action of protein fining agents than are wines. (For a detailed discussion of fining and fining agents see Zoecklein et al., 1995). The total phenol content, as well as the phenol fractions, can be determined by a number of analytical procedures such as HPLC, Folin Ciocalteu and permagnate method. (Zoecklein et al., 1995). Schanderl (1962) recommended a simple pH 7 test for the determination of polyphenol levels in juice and wine (see Zoecklein et al., 1995 for details).
Potassium Bitartrate Stability
Most producers stabilize their base wines to prevent bitartrate precipitation which can influence taste (KHT is both salty and bitter) and gas release from sparkling wines. There is wide variation in the exact procedure used by producers to determine KHT stability. A freeze test relies on the formation of crystals as the result of holding wine samples at reduced temperatures for a specified time period. Often a sample is frozen and then thawed to determine the development of bitartrate crystals and whether or not those crystals return to solution. Zoecklein et al. (1995) discussed some of the problems associated with using a freeze test to predict bitartrate stability. Several winemakers use a slight variation of the freeze test. Realizing that the prise de mousse will create anywhere from 1.1 - 1.5% additional alcohol (in mouseux production), they will fortify a small quantity of their cuvée and perform a freeze test on the fortified sample. Alcohol, among other factors, affects KHT precipitation. Fortification may be a desirable change to the freeze test procedure, but the inherent problems of the freeze test still exist even when the sample is fortified. An electrical conductivity test is a much more accurate method of determining bitartrate stability (Zoecklein et al., 1995).
Protein Stability/Bubble Size, Retention and Foaming
Carbon dioxide is available in two forms; free gas, and CO2 electrostatically bound to constitutants such as proteins, polysaccharides and lipids (see Figure 2). Makers of sparkling wine must manage their cuvée protein levels to obtain a product with minimum protein precipitation in the bottle while not detrimentally affecting carbonation. Precipitation of protein is affected not only by the exposure temperature, but also by the duration of heating. Since all cuvée proteins may be precipitated by heat, there are varying degrees of heat stability with regard to proteins. For example, heating a sample at 40°C for 24 hours precipitates about 40% of the wine proteins, whereas holding at 60°C for 24 hours precipitates 95-100% of the proteins (Pocock and Rankine, 1973). The time necessary for haze formation decreases with increasing temperature.
Several winemakers use a heat test and recommend chilling the wine sample following heat treatment. Visible haze formation is slightly greater than that seen in a sample without subsequent cooling. Protein precipitation, like potassium bitartrate precipitation, is affected by alcohol. Winemakers may choose to fortify their cuvée blends by 1.1-1.5% alcohol in the laboratory prior to running a heat test. This is to duplicate the alcohol level which will be achieved in the bottle. Precipitation tests such as the TCA procedure are not uncommon methods for determination of protein stability. The makers of sparkling wines must look beyond stability to the effects proteins have on bubble size, bubble retention and foaming. Indeed, the influence of cuvée proteins, fermentation rate, and yeast autolysis products may be greater than that of such traditional parameters as alcohol on bubble size, retention and foaming. Gauging optimum cuvée protein is a matter of experience. Those using bentonite as a riddling aid may want to not fine with bentonite or purposely underfine the juice or cuvée, knowing that additional protein will be bound in tirage. Little has been published about the influence of tirage fining agents on bubble and mousse. Munkegard (1998) noted the increase in mousse quality with the addition of tirage tannin. This may relate to protein tannin interaction (for additional information on bubble and foam quality, see page 16).
Because it is rare that a single wine of a single vintage from a single vineyard will be perfectly balanced in composition and flavor for a premium sparkling wine, blending is often performed. Blending is considered by most to be the key to the art of méthode champenoise. The selection of the cuvée components is conducted with three main objectives in view: the production of a sparkling wine of definite consistent flavor and quality; the enhancement of the quality of the individual wines; and the production of a base wine of sufficient quantity. Blending is an important tool that produces a result that is greater than the sum of the parts. The art of blending depends in part on chemical formulae, but also relies heavily on the gift and talent of the blender. The winemakers must blend wines for sparkling wine production when the wines have the better part of their lives yet to come. This requires considerable insight. It is difficult to predict the final results of blends that will be consumed years later.
The first decision to make is whether the new wines are of sufficient palatability to produce méthode champenoise. The magnifying effect of carbon dioxide on sparkling wines significantly highlights any enological flaws in the product, so wines for cuvée selection should be tasted at room temperature and on several occasions.
The decision of whether the cuvée is to be non-vintage or vintage dated is an important one. Non-vintage products rely on product consistency and usually require vin de reserve (cuvée blending from previous years). Generally, at least one eighth of the new wine is put into reserve for this purpose in Champagne. Reserve wine is either stored in magnums (as is the case with Bollenger) or in bulk, sometimes under a gas environment.
Some makers prepare cuvée blends prior to stabilization. When wines of different ages, grapes, and origins first meet, bitartrate and protein precipitation can occur. Cellar treatments such as fining and filtration can remove colloidal protectors, and thus affect potassium bitartrate stability. Due to the character of the wine, many prefer to make cuvée blending decisions following stabilization. It is essential that protein and bitartrate stability be evaluated just prior to cuvée bottling.
Technology dictates that producers rely on the chemical composition of the cuvée, as well as its taste, for the blending determinations that aid in production consistency. For example, wines with high alcohol, low pH and/or low level of assimilatable nitrogen cuvées may have difficulty completing the secondary fermentation, while low alcohol cuvées produce sparkling wines with poor bubble retention (Amerine and Joslyn, 1970). Many producers add a source of nitrogen such as DAP (24 g/HL) prior to tirage.
The primary requisites for a cuvée are a high titratable acidity (7.0 g/L or more expressed as tartaric), low pH (less than 3.3), low volatile acidity (less than 0.60 g/L), and moderate alcohol level (between 10.0 and 11.5% v/v). The cuvée should be light in color, with a balanced, fresh aroma. Many are looking for base wines with no single varietal character dominating, but with body, structure, substance, and length. Wines with a low acetaldehyde (< 75 mg/L), low copper (< 0.2 mg/L), and low iron (< 5 mg/L) content are sought. Additionally, wines with a relatively low phenolic content are often desired. An extract of 25 g/L adds body without making the wine heavy.
The concentration of aldehydes is a gauge by which general sparkling wine quality can be measured. Aldehyde concentration is primarily a function of the extent of oxidation but also of the quantity of SO2 added during primary and secondary fermentation. Concentrations of acetaldehyde greater than about 75 mg/L may add an overripe, bruised apple aroma (Zoecklein et al., 1995).
Another important blending consideration is the amount of second-cut press material to employ. This affects the phenolic content and is both a production and economic question. The goal is often to produce a cuvée that is delicate and 'clean' and has structure to provide the framework for bottle bouquet. For 'Vintage' years and Petillants, the alcohol level of the wine is usually somewhat higher (11-11.5% (v/v). Cuvée alcohols greater than about 12.6% can lead to sticking of the secondary fermentation. The base wine should be low in free sulfur dioxide content (< 20 mg/L) to ensure the ability to referment. Additionally, both the total and free sulfur dioxide content must be kept low if a malolactic fermentation is desired.
Chardonnay alone can be highly perfumey and somewhat candy-like, with intense richness. Excessive varietal character is often reached in California. This is not a problem in the eastern U.S., which may make Chardonnay production for sparkling wine quite suitable for the region. Pinot noir often produces a light, earthy, strawberry aroma. Our European colleagues use the analogy: the Pinot noir is the frame; the Chardonnay, the picture; and the Pinot meunier, the dressing for their Champagnes. Pinot noir, Pinot blanc, and Pinot meunier age more quickly than Chardonnay. Some generalizations regarding palate profiles can be made of young wines produced in Champagne. Chardonnay is detected at first with its intensity and perfume. This is followed by Pinot meunier with broad mid-palate flavors, and finishes with Pinot noir which adds length and intensity. Both Pinot noir and Chardonnay take more time to develop than Pinot meunier. Often meunier is utilized to a greater degree if wines are aged 1 year or less sur lie. With increasing tirage age, Pinot noir will increasingly dominate the nose and palate. The lack of knowledge as to which cultivars to use and which blends will age needs particular attention.
In California, the prevalent attitude is that a high malic acid level in the cuvée, coupled with a low pH, add life and freshness to the sparkling wine. Malolactic fermentation is avoided because the wine then stays fresher and ages less quickly. Some French producers, however, believe that a malolactic fermentation of the cuv»e or a component part can broaden and lengthen the finish and flavor. An elevation in pH and a reduction in acidity changes the palate structure. In Champagne there are climatic differences that help explain a preference for malolactic fermentation. The days are warmer, the nights cooler, and the light intensity greater in Napa (Maudiere, 1980). Grapes ripen faster in California and generally have higher sugars and lower titratable acidity than in Champagne. Many French houses put their sparkling wine bases through a malolactic fermentation. The result is a wine with the same acidity as a California product in which the bacterial fermentation has been prevented.
Table 6. Méthode Champenoise Analysis
|Total phenols mg/L GAE||209||294||261||261||245||340||317||300|
|Tartaric acid g/L||3.12||3.45||1.99||3.56||2.76||4.15||1.22||2.15|
|Malic acid g/L||4.78||2.03||2.79||0.33||3.32||0.25||1.00||2.96|
|Citric acid g/L||0.18||0.16||0.79||0.17||0.23||0.22||1.61||0.22|
|Lactic acid g/L||0.15||2.06||0.15||3.80||0.12||3.12||0.24||2.02|
|Acetic acid g/L||0.45||0.28||0.16||0.37||0.23||0.30||0.18||0.44|
|Succinic acid g/L||0.15||0.33||0.27||0.21||0.37||0.52||0.28||0.63|
Table 6 provides some analytical data from the Enology - Grape Chemistry Laboratory at Virginia Tech comparing European and American méthode champenoise. A major difference illustrated is the high malic acid content (low lactic acid) of some of the finished products. When malolactic bacteria grow in wine they can reach population levels of 106 - 108 cells per milliliter. Such titers are equivalent to yeast populations during active fermentation. It seems likely that the significant production of proteases, lipases, and esterases caused by malolactic fermentation could significantly alter the finished product. Some méthode champenoise producers appear to be utilizing malolactic fermentations of the cuvée to control the palate structure. A malolactic fermentation may modify the sweet-sour perception one experiences occasionally with méthode champenoise produced from low pH, high acid cuv»es. Malic acid is rather aggressive, while lactic acid is much softer on the palate. An increased number of American producers are now experimenting with partial or complete M/L fermentations of their cuvées (Zoecklein, 1986b).
Immediately prior to bottling, many producers filter their cuvées. This occurs, of course, before yeasting. The purpose of such an operation is twofold: to help prevent malolactic fermentation and to begin the secondary fermentation with "clean" wine. Some, such as Krug, do not filter at all, but simply clarify once with isinglass (Duijker, 1980). Malolactic fermentations can easily transpire under pressure, such as might occur during the secondary fermentation. Such a bacterial fermentation reduces malic acid, increases lactic acid, raises pH, and increases the titer of bacteria. The latter, particularly, results in riddling difficulty and possible loss of product palatability. The general nature of the cuvée usually helps prevent a spontaneous malolactic fermentation. Grapes are brought to the sparkling wine house at low pH levels and often pressed, avoiding skin contact, thus aiding in reducing the likelihood of a spontaneous fermentation. Those concerned with the possibility of a malolactic fermentation in the bottle generally sterile filter their cuvées. If a malolactic fermentation has been completed, a D.E. filtration, pad filtration or no cuvée filtration may occur. An additional advantage of a completed malolactic fermentation of the cuvée is that it will not occur during secondary fermentation or storage.
Sparkling wine yeasts are available on slants, in liquid, and in active dry forms. The yeast volume employed for the secondary fermentation is usually a 2-5%-activity growing culture. Many traditional sparkling wine houses build up an active yeast innoculum from slant cultures by either a step-wise volume increase or by the use of yeast generators with or without oxygen sparging. Yeast preparation for bottle fermentation is of obvious importance. Some believe it desirable to culture yeast under stressful conditions such as higher SO2 levels, and cooler temperatures (the so-called step down theory), so that when the secondary fermentation begins the yeast will be more vigorous. Others have expressed the desirability of conditioning the yeast to the exact same conditions (except CO2 pressure) that will be found in the bottle. Research continues in this regard.
A common preparation method is given: (Bannister, 1983) 500 milliliters of a solution of sterile wine (the cuvée to be fermented) and sterile water are diluted to 7% alcohol. To this, 5% sugar and 1 2 grams of yeast extract are added. This media is inoculated from a slant yeast culture using strict aseptic techniques and incubated at approximately 80°F. When the sugar is 1/2 utilized, the culture is transferred directly into 1 1/2 liters of undiluted wine to which 5% sugar has been added. This is repeated using a 10% inoculum into a new-wine volume that has 5% sugar added. Transfers are made at 2 1/2% sugar. This is repeated again until a 5% inoculum volume has been produced (5% of the cuvée volume that is to be fermented). Care must be taken not to allow the culture to go to dryness prior to transfer because the alcohol level will increase and begin to inhibit the yeast. When all the sugar has been depleted in the media, the yeasts rapidly begin the death phase. Transferring the growing culture at 2 1/2% sugar will acclimate the yeast to be able to grow in a 2 1/2% sugared cuvée. Additionally, during the transfers it is desirable to go from inoculation temperature to the temperature at which the cuvée will be fermented.
Aeration will produce yeast cell membranes rich in ergosterol which will result in increased alcohol tolerance. Optimally, the producer will examine the starter culture to assure that the culture is actively growing and not contaminated. A large percentage of budding yeast (70-80%) is desired. It is essential that the culture be free of contamination. Some use a methylene blue test to monitor yeast growth (Zoecklein et al., 1995; Fuglesang, 1997) for stain preparation.
To insure secondary bottle fermentation, a minimum of 1 million cells per milliliter should be added to each bottle (Geoffroy and Perin, 1965). An actively growing culture is usually about 1x106-8cells per milliliter. From 0.8 to 2.5 x 106 cells per milliliter is usually added for the secondary fermentation. Yeast cell titers can be determined as described by Fuglesang (1997). Some producers prefer to simply add lyophilized yeast directly to the cuvée. Active dry yeast contains 20-30x109 live yeast cells per gram (Berti, 1981). If equipment is limited, the use of active dried yeast may be considered easier. It is preferable to feed and grow several generations of active dried yeast prior to the addition into the cuvée. This allows the producer to train the yeast to go in the cuv»e as well as monitor yeast viability and possible contamination. An increase in the number of yeast cells in the cuvée may give a fuller character and flavor to the sparkling wine (Berti, 1981). Care must be used, however, to avoid rapid secondary fermentation and the development of hydrogen sulfide and other off-odors. (For additional information regarding yeast culture preparation, see Fuglesang, 1997.)
For the secondary fermentation (prise de mousse), a yeast with the following attributes is desirable: pressure tolerance, alcohol tolerance, cold tolerance, SO2 tolerance, produces little SO2 , ferments to dryness, dies or becomes inactive following fermentation, does not stain the wall of the bottle, desirable flocculating or agglutinating ability, produces no off flavors or odors, and has a desirable effect on carbonation.
Because the demands on the yeast are very specific, the vintner must be specific in yeast selection. For example, Chardonnay is sometimes difficult to ferment to dryness; therefore, a strong fermenter may be desirable. Some yeasts are very delicate, others assertive or defined, regarding the character they impart to the sparkling wine. This is another stylistic consideration. There is significant variation in the ease of riddling with different yeast (Geoffroy, 1963). Several °champagne strains° of Saccharomyces cerevisiae and S. bayanus (formerly oviformis) have many of the above-mentioned properties including enhanced agglutinating ability. S. bayanus has a slightly greater alcohol tolerance than the S. cerevisiae. Additionally, some producers use S. unarium for the secondary fermentation. Epernay, AKA Prise de Mousse, is a highly flocculent yeast with good riddling ability. It is fairly assertive and is therefore usually not employed to carry out both the primary and secondary fermentation. This yeast is the same as Epernay 2, which is a low-foaming strain often employed when a sweet finish is desired. The Geisenheim strain of champagne Epernay does not produce SO2 during fermentation, does not stick to the bottle, ferments at relatively low temperatures, and is sandy in its agglutinating ability (Becker, 1978). Both California Champagne (UCD 505) and Pasteur Champagne (UCD 595) are popular yeasts for secondary fermentation. Both are available in dehydrated form. California Champagne (UCD 505) is a good flocculator and may be considered to be more delicate than Pasteur Champagne (UCD 595). Some sparkling wine producers use mixed cultures for the secondary fermentation believing that such a procedure adds complexity. Many sparkling-wine houses employ their own proprietary yeast strains. New or prospective producers should do some "in-house" experimentation to determine the merits and deficiencies of different yeasts under their own conditions.
To enhance riddling ability, disgorgement, and possibly wine palatability, some vintners add riddling aids at the time of cuvée bottling. Such aids (fining agents) may enhance the riddler's ability to convey the yeast to the neck of the bottle. When there is sedimentation of the yeast with the proper fining agent, riddling can be much easier. Some common riddling aids are:
- Sodium and calcium bentonite
- Clarifying Agent C
- Adjuvant H
- Diatomaceous earth
Clarifying Agent C is a proprietary bentonite preparation used with phosphate mazure; Adjuvant H is a proprietary bentonite-based agent used with tannin; Colvite is a proprietary isinglass; and Botane is a proprietary tannin formulation. All are of European origin.
Bentonite is, perhaps, the most popular riddling aid in this country. It is added at the time of cuvée bottling in levels seldom exceeding 6 g/HL (2 pound/1000 gallons). In Europe, calcium bentonite (3.5 g/HL (1/4 lb per 1000 gallons) is frequently used. The choice of riddling aids should also be based upon the expected time sur lie. Clays are often preferred for young wines, while gelatins are for aged or older wines.
The major disadvantage with the use of riddling aids is that their effects on both riddling ease and sparkling wine palatability are not predictable. Riddling aids may influence foam and/or bubbles as well as wine clarity. Tirage tannin, for example, may positively influence mousse quality (Munksgard, 1998). Further research in the area is needed. Because each cuvée is different, the winemaker must wait until riddling and disgorgement to review the merits or deficiencies of the riddling aid(s) employed. Bentonite is the most common riddling aid because of its relatively inert nature. It seldom has a detrimental effect on product palatability at the levels employed (usually less than 6 g/HL or 2 pound/ 1000 gallons). Care must be taken to avoid the addition of too much riddling aid, which can make riddling, and particularly disgorgement, difficult (Zoecklein, 1987a).
Return to Table of Contents.
Liqueur de Tirage
Different wineries use various sugar sources for the prise de mousse (secondary fermentation). Bottler-graded sucrose or dextrose are perhaps the most common in this country; however, larger operations may choose to employ sugar syrups. Many French producers use high quality beet sugar. Some use a 50% sugar solution - 500 grams/liter of sugar in wine, with 1.5% citric acid frequently added to invert the sugar if sucrose is used.
Theoretically, 4.04 grams of glucose or 3.84 grams of sucrose upon fermentation will yield 1.00 liter CO2 (760 mm and 0°C) weighing 1.977 grams (Berti, 1981). The actual yield is less due to production of small amounts of aldehydes, volatile and fixed acids, glycerol, and other entities produced by the yeast. In actual practice, sparkling wine producers estimate that 4.0 to 4.3 grams of sugar per liter is needed to produce one gas volume (ATM) of carbon dioxide (4.3 grams of sugar per liter is equal to 1 pound of sugar in 27.3 gallons). If, for example, 6 gas volumes of CO2 are required, then approximately 4.2 grams times 6 atmospheres or 25.2 grams of sugar per liter are added. This will produce between 1.1-1.5% additional alcohol (v/v). If the cuvée already contains fermentable sugar, this must be taken into account.
In this country, sparkling wines are those that contain 0.392 grams CO2 per 100 mL or more at 60éF. A wine containing this amount of CO2 will exert about 15 psi pressure at 15.56°C. In Europe, the minimum pressure for sparkling wines recommended by l'Office International de la Vigne et du Vin is 51 psig at 20°C in bottles over 250 mL capacity. Accurate determination is therefore critical. Carbon dioxide pressure in the U.S. is more a stylistic consideration. Petillants possess about 2-2.5 atmospheres pressure at 1°C and have