By Dr Victoria Hughes – R&D Manager/VIC State manager, Dr Carly Gamble – Technical Manager/WA State manager, and Dr Alana Seabrook – Operations manager
Sparkling wine production starts in the vineyard. Be it Prosecco from the King Valley, Chardonnay from the Riverland or Pinot Meunier from the Tamar Valley in Tasmania, grapes are purposefully grown and picked for the end result. The soil profiles and climatic conditions are often compared with the likes of Champagne among other iconic sparkling producing regions to provide the desired palate weight, length, fruit profile, texture as well as minerality and foam formation. The calcium content of the soil and protein content in the grapes will impact winemaking practices with regards to downstream stabilisation. Seemingly simple, getting the base wine right will impact all downstream quality outcomes.
This article will go through the main steps of production for both the traditional method Champenoise and Charmat, as well as transfer and carbonation, and focus on the analytical checkpoints. This article will not focus on soil profiles and viticultural techniques.
In principal, base wine production should be similar for all methods of sparkling production. Qualitatively this may vary slightly depending on the outcome. The major grape varieties in Champagne are Pinot Noir, Pinot Meunier and Chardonnay. Generally speaking, grapes for sparkling base wine are picked at a lower baume than if they were destined to make table wine. The secondary fermentation will add approximately an additional 2 % v/v alcohol which is factored into the picking decision. Whilst a large SO2 addition may be standard practice for grapes which are transported for white production, this is a critical control point for sparkling base due to the fact that it needs to go through successive alcoholic fermentations. SO2 is often added at picking to prevent not only microbial spoilage, but in the case of white wine, phenolic oxidation once the berries have been compromised. Machine picking and SO2 can increase the skin contact and consequently oxidation and phenolic extraction. Therefore, the decision of how to harvest (machine vs handpicking) and distance to the winery can greatly impact the end result. The use of bioprotection may reduce the requirement for additional SO2 treatments in terms of microbial control.
Analytical check point – Ensuring SO2 is lower than 50 ppm. pH/TA can vary depending on variety and region, but it is not uncommon for sparkling base to have very low pH (< 3.0) and high TA (greater than 7.0 g/L tartaric acid)
Once at the winery, the press and the press cycle itself will also impact both the oxygen pick up and phenolic extraction. Gentle press cycles with increasing pressure rather than multiple rotations are often preferred. The use of a pressing enzyme with both depectinisation and gentle cellulose and hemicellulose can assist with increasing varietal aromas and critical textural elements.
Phenolic fining with either animal or vegetable-based fining is important depending on the phenolic load present. Press fractions are routinely separated to enable targeted treatment for each fraction. This may happen prior to fermentation in the juice stage, or during alcoholic fermentation.
Analytical check point – taste only. In-house it may be possible to conduct spectral measurements to compare phenolic levels, but this will be subjective.
Once the juice has been pressed it is clarified by a variety of techniques. If bentonite is added at this stage, it is important that the juice has been completely depectinised, as the addition of bentonite will inactivate any residual pectolytic activity.
Yeast selection for base wine fermentation will impact the sensory profile of the wine. Selected yeast strains are described as providing icon champagne characteristics. However, often the base wine may be fermented with different yeast strains to create a base wine with a distinctive aroma profile. A thiol or ester producing yeast will produce a more aromatic outcome in this base wine stage.
The decision to carry out malolactic fermentation (MLF) is a winemakers’ decision. Many base wines from iconic sparkling producing regions including Champagne and Tasmania may present up to 6 g/L malic acid. The base wine requires specific strains of bacteria to go through malolactic fermentation due to the often ultra low pH (under 3.0), most of which require an acclimatisation step in which the bacteria produces a beta-glucan layer around the bacteria itself. Many sparkling wine producers decide not to got through malolactic fermentation, whilst others prefer partial or full malolactic fermentation. Once the base wine has gone through MLF, there is no further risk that the wine will go through MLF during a secondary alcoholic fermentation.
The base wine at this point may be cold stabilised and protein stabilised. Any crystals that form in bottle will likely impact the size and shape of the bead in the final wine. Importantly, crystals present additional nucleation points which could lead to gushing. Protein stabilisation often involves the addition of bentonite. However, there is a layer of complexity with sparkling wine production that is not present in table wine production. The protein content is quite relevant to the amount and size of bubbles present in the final product. A wine with zero protein content would have very few bubbles. Winemakers often settle for a protein unstable wine at this point so as not to remove too much of the protein content.
Analytical check point – End of alcoholic fermentation analysis may include Glucose/Fructose, Malic acid, pH and titratable acidity. Tartrate stability and protein stability should be tested at this point. A small addition of SO2 may be required depending on the next stage of the process.
Prise de mousse (PDM)
Any process of creating carbonation via means of a secondary fermentation is considered a prise de mousse. This may happen in bottle (Method Champanoise, Transfer) or in tank (Charmat). Building up a yeast culture is an art in itself, paramount being a yeast strain optimal for the prise de mousse step. Other than fermentation kinetics, the yeast strain will be important for the sensory outcome of the sparkling wine. There are a number of Saccharomyces cerevisiae yeast strains specifically for this PDM process that will provide optimal aroma and palate characteristics of sparkling wine. The yeast is normally built up and has base wine incorporated into the culture prior to dosing into the tirage process – it is best to follow manufacturers’ instructions for that yeast strain. This is different to a standard fermentation as the wine is already starting with generally a 10-11 % alcohol percentage, often with very low pH.
Analytical check point – Yeast count and viability. Ensuring the yeast is viable and in sufficient numbers to inoculate into the wine is critical. Normally the yeast culture will end up with a yeast count of 200 million cells/mL but this may vary depending on yeast strain, base wine chemistry and propagation conditions.
The gold standard of Champagne production involves a secondary fermentation in bottle. Sucrose may be added to sparkling wine at this stage at a very specific rate. Too much sucrose may cause the yeast to produce too much carbon dioxide and risk the bottle exploding. Too little and the end product will not have sufficient carbonation. A common rate of 20-24 g/L is standard, which will produce 8 bar of pressure and just over 1% alcohol. Yeast is dosed in with an adjuvant to assist the yeast settling after it has carried out fermentation.
Charmat is a traditional method of producing larger volumes of sparkling wine, typical in Prosecco regions. This process is where the secondary fermentation is conducted in a large pressure vessel, whereby the yeast and sugar are added to the pressure tank, and fermentation is monitored. Once the sugar has been consumed, the resulting pressure is retained even through the filtration process. This is much less labour intensive but doesn’t have the same amount of lees contact time. This fermentation may be finished in 7-10 days.
This method is very similar to the traditional method Champenoise whereby the secondary alcoholic fermentation is undertaken in bottle, however rather than the bottles being riddled, disgorged and dosed individually, the transfer method involves a metal straw piercing the lid, with the contents of the bottle post fermentation (including lees and solids) being transferred into one large pressure tank. This allows for greater consistency across vintages and batches in regard to any additions being made.
Analytical check point – Yeast count and viability. Producers will vary in how much yeast they choose to add, but generally a minimum 1-2 x 10 6 cell/mL is required to ensure fermentation will be completed. This number may be significantly higher for Charmat. It is critical to ensure that the correct amount of sugar has been added in the form of sucrose to achieve optimal carbonation.
Carbonation is where the base wine is simply dosed with Carbon dioxide (CO2) directly and there is no secondary fermentation. This is often suitable for lower price point sparkling wine as the cost of carbonating is much lower than Charmat or Method Champenoise.
Analytical check point – Test for CO2 levels before and after bottling to ensure the resulting CO2 in bottle is correct.
Sparkling wines that are bottle fermented are routinely left on lees for a period of time. Generally this will range from 6 months up to several years, depending on the desired sensory outcome. Aging on lees can confer distinctive ‘yeast’ and ‘bread’ notes as well as increased weight and texture on the palate. The longer the wines are left sitting on their lees the more developed this flavour and style will become. Like any wine left without SO2 in the presence of sugar, there is always a spoilage risk, particularly from Brettanomyces bruxelllensis. If the bottles for this process, particularly with the transfer method have been rinsed and re-used, this risk is higher again as B. bruxellensis can form a biofilm on the bottles themselves. If MLF has not been carried out previously this may also happen in bottle to varying degrees, risking significant bottle variation in regards to levels of malic, lactic and consequently pH and TA.
Analytical check point – It is important to ensure primarily that the secondary alcoholic fermentation has concluded by measuring residual sugar after 6-8 weeks depending on storage conditions. If extended aging on lees in bottle is desired then periodic bottle checks should be carried out to ensure there are no obvious sensory or chemical changes. Some indicators of change are assessing the B. bruxellensis population by PCR or plating, or simply measuring ethyl phenols 4-ethylphenol and 4-ethylguacol. If MLF has not previously been conducted, measuring any change in malic acid as well as pH and TA on a periodic basis would be ideal.
After the secondary alcoholic fermentation has finished and any aging on lees desired has been carried out, the bottles are either riddled manually in racks or with automation in cages. The success of this will be impacted by the adjuvant, temperature and riddling technique. For wines that have been sitting on lees for an extended period this is an opportunity to adjust and freshen the wine with the dosage liqueur. Riddling can take anywhere from several weeks when performed manually to several days in cages. The objective of riddling is to precipitate the lees from the secondary fermentation into the neck of the bottle, at which point the neck of the bottle is frozen in brine to form an ice block which is then pushed out by pressure in the bottle. At this point the wine may have a dosage of sugar, wine or brandy and SO2 and is re-sealed. The amount of sugar will depend on the style of wine desired, anywhere from 10-15 g/L for a Brut Champagne.
Analytical check point – Wines are normally disgorged, dosed and sealed at the same time. Given there is often an addition of sucrose and SO2 it is critical to test both of those parameters in real-time in order to adjust the dosage rate if incorrect. Sensory analysis to prepare the correct dosage must be conducted at the same temperature as the desired serving temperature, as 8 g/L of sucrose will taste very different at 4°C than 20°C. Disgorging can be a fairly manual process, it is critical to check the NTU to ensure there hasn’t been any carry over from the yeast lees. A yeast count may also be useful here.
Rather than riddling the wine as in the case of MC, wines are clarified by filtration prior to bottling. These wines must be transferred to a pressure tank from either bottle (transfer method) or tank (Charmat) whereby the level of carbonation is retained. Any loss of carbonation may be factored in prior to bottling. At this stage it is possible to add SO2, sucrose and blend back any wine addition that may assist in balancing the palate or ensuring there is consistency across vintage/batches.
Analytical check point –– Post filtration CO2 will need to be measured and any adjustments made. It is critical to check basic chemistry to ensure there haven’t been any dilutions in this process as well as confirming that the level of sucrose added is optimal. Sensory is critical in the presence or absence of any additions.
As with any bottling, it is important to check the post bottling chemistry to ensure the absence of unintentional dilutions or over additions. Although larger companies may be in a position to automate, riddling and disgorging can be an extremely manual process subject to human error.
Analytical check point – Monitor all post bottling chemistry including alcohol, turbidity, Free and Total Sulfur dioxide, pH, titratable acidity, residual sugar (Sucrose, Glucose and Fructose), as well as Malic acid.
Sparkling winemaking requires an infinite amount of planning, well before the grapes are harvested. Over additions, poor handling or excessive oxygen pick up will all be amplified in sparkling wines over any other type of winemaking.
Other methods of producing carbonated wines such as Pet Nat (short for pétillant naturel) have not been discussed here, but are increasingly present. This type of wine is made by bottling the wine before primary fermentation has been completed, leaving yeast lees and sediment in bottle. For larger companies where consistency is paramount, Pet Nat does not offer this and is typically not employed as a winemaking style.
Whist some winemaking techniques can be more forgiving with regards to analytical check points, producing a consistent quality of sparkling wine has a number of critical control points that cannot be missed as discussed here. It is crucial to be aware of SO2 levels in order to carry out multiple alcoholic and malolactic fermentations. The amount of sucrose going into the fermentation will dictate whether the wine will have the perfect level of carbonation, or turn into a hazard. A cold unstable wine could present excessive nucleation points and cause gushing in the final bottle. An overly protein stable wine may struggle to achieve the desired bead. Of all wine types, analysis during sparkling wine production by any of the methods discussed is integral to a consistently qualitative outcome.
References
Dols-Lafargue, M., Lee, H.Y., Le Marrec, C., Heyraud, A., Chambat, G. and Lonvaud-Funel, A., 2008. Characterization of gtf, a glucosyltransferase gene in the genomes of Pediococcus parvulus and Oenococcus oeni, two bacterial species commonly found in wine. Applied and Environmental Microbiology, 74(13), pp.4079-4090
Gerbaux V, Briffox C, Dumont A, Krieger S. Influence of inoculation with malolactic bacteria on volatile phenols in wines. American journal of enology and viticulture. 2009 Jun 1;60(2):233-5.
Ribéreau-Gayon P, Dubourdieu D, Donèche B, Lonvaud A, editors. Handbook of enology, Volume 1: The microbiology of wine and vinifications. John Wiley & Sons; 2006 May 1.
