Arnaud DELAHERCHE¹, Etienne DORIGNAC¹, Maryam EHSANI ¹, Marie Charlotte COLOSIO², Richard MARCHAL ³, Thomas SALMON ³, Yann VASSEROT ³
¹ Fermentis, Marquette lez Lille France
² Institut Français de la Vigne et du Vin, Nantes France
³ Université de Reims Champagne Ardenne, Reims France

Prise de mousse with SafŒno™ SPK-05


The market for yeasts selected to produce sparkling wines, as opposed to that to produce white wines, is rather limited and consists of strains with strong fermentation vigour producing low concentrations of aromatic compounds. Although vintage sparkling wines are recognized for their quality, today the market increasingly demands sparkling wines characterized by greater freshness and fruitier sensations than in previous decades.

The prise de mousse (PDM) or effervescence is the stage that allows a base wine to become sparkling or semi-sparkling and the success of this technique depends largely on the preparation of a yeast starter culture (YSC). The YSC makes it possible to reach enough yeast cells to guarantee an optimal PDM process and its preparation traditionally requires several days and several steps.

This work summarizes three years of experimentation which led to the selection of a new yeast to produce sparkling wines. Its fermentation characteristics made it possible to optimize the existing protocol for the preparation of YSC.

Results show that with SafOeno™SPK 05 (commercial name of the selected yeast), YSC preparation time can be reduced from 3-4 days in a two-step process to a single step of approximately 30-36 hours, without affecting the PDM duration or quality.


The market for yeasts selected to produce white wines is made up of a great number of references reflecting the wide variety of wines produced. The role of selected yeasts in modulating the sensory profile of a wine depending on the grape varieties, methods and winemaking techniques used, has been known for decades.

On the other hand, for sparkling and semi-sparkling wines, the range of yeasts available on the market is more limited. These yeasts are characterized by undeniable fermentation qualities and, in general, by the particularity of leaving little or no sensory fingerprint on the wine. This characteristic allows a better expression of tertiary aromas resulting from the autolysis of the yeast which appear during aging.

Although the quality of this type of sparkling wine is recognized and appreciated by many consumers, today the market also demands sparkling wines characterized by greater freshness and fruitier sensations than in previous decades.

To produce a sparkling wine, a base wine must undergo a second fermentation called prise de mousse (PDM), refermentation or effervescence. This technique requires the preparation of a yeast starter culture (YSC) containing a high amount of yeast cells capable of consuming all the fermentable sugars added to the base wine and transforming them into alcohol and carbon dioxide. The quality of the YSC is crucial for the success of the PDM process, which can take place in a bottle using the traditional method, or in a closed tank, in an autoclave, according to the Charmat or Martinotti method. However, over the years, many producers have expressed concerns about the excessive length of the classic YSC preparation protocol and the need to constantly monitor the process, even at the expense of their days off work.

This study, the result of three years of experiments, led to the selection of a new yeast strain which makes it possible to produce sparkling wines that meet the demands of modern consumers, while helping producers to simplify the process of the YSC preparing, maintaining high standards of fermentation quality and safety.

Materials & methods

Selection of strains: multiple fermentation and refermentation tests were carried out in Champagne, during 3 vinification campaigns (2019, 2020 and 2022), at pilot scale in 15-liter containers and at real scale in vats of 20 – 130 hL of Pinot Noir musts (cuvées) obtained from healthy grapes. The musts had assimilable nitrogen (YAN) levels of 240 – 280 mg/L and potential alcohol levels of 9.95 – 11.05% (v/v).

Initially, fermentation trials were carried out in the laboratory and in cellars on 5 different winemaking sites in Aube and Marne. Five strains were considered: three experimental strains, candidates for selection and not present on the market, and two commercial strains. The experimental vinifications were carried out in duplicate and included a control modality fermented with a commercially available Champagne strain.

PDM tests were then carried out to compare the new selected yeast to two commercial strains commonly used for the preparation of YSC in Champagne. In these tests, the traditional protocol proposed by the Comité Interprofessionel de Champagne (CIVC)(1) was applied for the preparation of the YSC.

All yeasts were inoculated at 1 million live cells/mL and the quantity of live cells was determined by cell counting in a Thoma chamber with methylene blue. The tirage was carried out without YAN adjustment, with 20 g/L of cane sugar and riddling aid, in 750 mL Champagne-type glass bottles. Twelve bottles were made for each test, three of which were equipped with permanent aphrometers. All bottles were placed in a room at a controlled temperature of 13°C for refermentation and the pressure was monitored twice a week.

Optimization of the yeast starter culture preparation protocol

Considering the technological characteristics of the yeast observed during the first fermentation and PDM tests, the possibility of preparing a YSC in a shorter amount of time, without the acclimatization phase required by the traditional protocol, was evaluated. The main metabolisms of S. cerevisiae, both carbon and nitrogen, were studied.

The base wines were adjusted with DAP to obtain three levels of YAN: 520 mg/L (N520), 400 mg/L (N400) and 280 mg/L (N280). Sugars, YAN, number of live cells and pressure were monitored according to the methods described above.

Champagne material & methods - Fermentis

Results and discussions

The selection process: vinification tests

Figures 1a and 1b show the fermentation kinetics of the new selected yeast (strain B) compared to a strain commonly used in the Champagne region. In laboratory and cellar vinifications, the experimental strain B demonstrated much faster fermentation kinetics than the other strains used. This observation is mainly due to a shorter lag phase. Indeed, all the fermentations were completed without leaving any residual sugars.

Figure 1
Figure 2

Figures 1a and 1b. Fermentation kinetics from two different fermentation trials on two different juices. Comparison between experimental strain B (Fermentis) and a commercial Champagne strain.

In all the tests, the experimental strain B stood out for its great ability to initiate and complete the alcoholic fermentation in a short time. In addition, the following characteristics were observed in the analysis of the base wines obtained with this strain:

  • Excellent adaptation to low pH (2.95-3.15 in these tests).
  • Low volatile acidity and total SO2 production.
  • Slightly lower total acidity and malic acid levels.
  • Significantly higher residual alpha-amine nitrogen levels at the end of fermentation.

From a sensory perspective, the base wines produced from this strain have been described as follows:

  • Expressive nose, characterized by fine fruitiness with notes of very pleasant red fruits, peach, apricot and citrus fruits such as grapefruit, typical of Pinot Noir.
  • On the palate, the wine appears fresh, with good acidity, round, smooth, with pleasant sweet notes, slightly tannic in the finish and good length.

Considering its high fermentation capacity, its low production of SO2 and volatile acidity, its lower YAN requirements compared to the other strains tested, as well as the sensory characteristics obtained in the base wines, strain B was selected for PDM trials.

The selection process: the PDM test

The first PDM tests were carried out by applying the traditional protocol for preparing yeast starter cultures (YSCs) and comparing the selected experimental yeast with two commercial strains commonly used for the preparation of YSCs in Champagne.

All PDM were carried out successfully despite the low YAN content and the low refermentation temperature (13°C).

The sparkling wines obtained were tasted after 13 months of aging on lees (at 13°C) by a tasting panel composed of cellar managers, oenologists, and staff from the URCA wine laboratory. The panel’s comments were as follows:

  • Fine perlage, light and very pleasant foam with the formation of a beautiful crown of bubbles.
  • Fresh and elegant nose with notes of strawberry, brioche, almond, white flowers, and smoky notes.
  • Lively palate with a firm and pleasant attack and a discreet effervescence. Fine and balanced wine with a slight astringency on the finish, good length, and persistent fruitiness without fermentation notes. The palate is slightly sweet with a balanced finish, a slightly evolved character of candied fruits making it ready for consumption.

The characteristics of the sparkling wines obtained with this strain can be summarized as follows: fruity and fresh aromas, mineral sensations and good persistence, fine foam, for sparkling wines ready to be consumed.

Preparing the yeast starter cultures

The traditional protocol (Fig. 2) proposed by the Comité Interprofessionnel du vin de Champagne (CIVC) lasts, in average, 3 – 5 days and consists of two steps: an acclimatization phase and a multiplication phase.

Protocol details

Figure 2: Traditional protocol for the preparation of the yeast starter culture used for secondary fermentation of sparkling wines.

Classic yeast starter culture (YSC): Phase 1 acclimatization

It is estimated that 6-10 hours are required for the initial density of the acclimatization medium to decrease from an initial value of 1040 g/cm3 to values ​​between 1025-1015 g/cm3. The results obtained in this study, however, show that this decrease occurs over a longer period. Density monitoring, illustrated in Figure 3a, shows that with an average YAN content of 400 mg/L (N400), it takes 12 to 22 hours to reach a density of 1025-1015 g/cm3.

The monitoring of fermentable sugars and PAA is shown in Figures 3b and 3c. In a classic PDC, prepared with 100 g/L of sugars and 400 mg/L of YAN (N400), approximately 57 g/L of sugars (linear model) and 207 mg/L of YAN (logarithmic model) have not yet been consumed by the yeast after 17 hours of acclimatization.

Figure 3
Figure 4
Figure 5

Figures 3a, 3b and 3c: YSC – Acclimatization phase 1. Control of density, fermentable sugars and PAA.

In other words, at the end of the acclimation phase, approximately 50% of the initial sugar and APA concentrations remained in the foot, which guarantees a margin of safety and obtaining a foot of yeast capable of actively multiplying even in the later multiplication phase. Even when preparing a foot with an initial APA concentration lower than 30% and equal to 280 mg/L (N280), similar results were observed, so that the nitrogen content at the end of the acclimation phase remains sufficiently high. These observations will be considered later when preparing a PDC without an acclimatization phase.

Classic yeast starter culture (YSC): – Phase 2 multiplication

The multiplication phase usually lasts 2 days but can last longer. At the beginning of the multiplication phase, the recommended density is between 1015 and 1025 g/cm3, while the final density at the time of inoculation of the base wine is between 998 and 995 g/cm3. During this phase, gentle and continuous agitation or one or two pumping per day are necessary.

The composition of the classic YSC at the beginning of the multiplication phase generally has a sugar content of 84.5 g/L and a mineral nitrogen content of about 100 mg/L.

YSC without acclimation phase: initial composition

In a conventional YSC, the yeast population only increases by a third during the first 24 hours of acclimatization, whereas during the estimated 6-10 hours required, the increase is much smaller. With the aim of obtaining a cell population comparable to that obtained with the conventional protocol with a 1-step YSC preparation protocol (without acclimatization step) of only 30-36 hours, a YSC was prepared as follows:

  • 9 kg of ADY/1000 hL of base wine, or 9 g of ADY/hL for an inoculation rate of 3% (v/v).
  • 84 g/L of sugars, simulating the sugar concentration found at the start of the multiplication phase in a classic YSC.
  • 100 mg/L of mineral nitrogen simulating the YAN found at the start of the multiplication phase in a classic YSC.
Protocol 2

Figure 4: YSC protocol without acclimatization (36 hours) for the inoculation of 1000 hL of base wine.

Classic YSC and YSC without acclimatization phase: evolution of parameters

The two YSC preparation methods were compared in terms of yeast populations, sugars, density, and YAN. The tests were carried out in duplicate.

YSC without an acclimatization phase resulted in a live yeast population of 110 million cells/mL in 30-36 hours (Fig. 5a). With this number of cells, an inoculation of only 1% (v/v) in the base wine is necessary to obtain a cell concentration of 1.1 million live cells/mL. This is equivalent to using approximately 4 to 5 g of ADY per hL of base wine.

living cells/ time graph
living cells/ time graph
density over time graph

Figure 5: Comparison of the two YSC preparation protocols. YSC with acclimatization (B1 and B2 – duplicates) and without acclimatization phase (B3 and B4 – duplicates). Monitoring of live yeast population (Figure 5a), sugar content (Figure 5b), mineral nitrogen (Figure 5c) and density (Figure 5d).

In the YSC without an acclimatization phase, the decrease in sugar content was continuous and regular and did not show a slowdown (Fig. 5b). After approximately 30 hours, the YSC without an acclimatization phase had a sugar level of 50 g/L and a density of 1010 g/cm3 (Fig. 5d), which allows the yeasts to continue to multiply in the case where the YSC must be maintained for several hours, even a whole day, before the inoculation of the base wine.

Regarding YAN, we observe a regular decrease in its concentration in the YSC without acclimatization, unlike the classic protocol which shows a sudden drop in YAN when the yeast enters the multiplication phase (Fig. 5c). For the classic two-phase protocol, lasting 72 hours, the medium becomes deficient in YAN after 60 hours, whereas with the YSC protocol without an acclimatization phase, 30 mg/L of YAN remain after 36 hours. It is therefore essential to pay a particular attention to this parameter because a nitrogen deficiency can lead to a strong reduction in yeast activity, cell multiplication speed and fermentation kinetics.

The results of these tests show that it is entirely possible to prepare a YSC in a single step and in less than half the time of the conventionally used protocol.
To evaluate the behaviour of the yeasts obtained with a YSC protocol without an acclimatization phase, two YSCs were started, one at 14°C and the other at 18°C (Fig. 6). The fermentation kinetics of the newly selected strain (strain B) was similar to that recorded with the commercial Champagne strain (Fig. 6). Although there was a slight longer lag phase with the YSC carried out at 14°C with strain B, the fermentation kinetics were almost identical and made it possible to reach 5 bars of pressure simultaneously with the Champagne strain in approximately 20 days at 18°C and in approximately 40 days at 14°C.


Figure 6: Monitoring of bottle pressure during YSC, at 18°C and 14°C, using permanent aphrometers.


This work made it possible, during the first years of experimentations, to select a new Saccharomyces cerevisiae yeast, commercially available as SafOenoTM SPK 05, capable of giving sparkling wines the sensorial characteristics sought by modern consumers. Furthermore, considering the fermentation characteristics of this strain, an innovative technical proposal was carried out for the preparation of the YSC, which demonstrates how the commonly used process can be simplified without compromising the success of the PDM. Reducing the YSC preparation time from several days to 30-36 hours gives producers greater freedom in organizing their work, increasing availability of multiplication tanks by 50% and, from an environmental point of view, helps reducing energy costs in the cellar.
The project was carried out thanks to the contribution of several cellars in the Champagne region who collaborated in the organization and implementation of trials on a production scale.
Pilot tests are underway in several cellars to validate the sensory quality of sparkling wines produced with this protocol, which will be followed over several years.