Winemaking Laboratory
Alcoholic fermentation / Fermentation kinetics — simulator

Chapter III · Alcoholic fermentation

Fermentation kinetics — simulator

How fermentation proceeds day by day: how long the sugar takes to ferment, when nitrogen runs out, how fast CO₂ is released — and how to spot a sluggish or stuck fermentation in time.

Must and conditions

Model (Boulton type): yeast growth is limited by nitrogen (1 mg N/L ≈ 6·10⁵ cells/mL), sugar uptake is inhibited by ethanol, and cell death rises exponentially with alcohol and temperature. Density: SG = SG₀ − 0.000428·ΔS − 0.00042·A (empirical approximation). Parameters are illustrative.

Fermentation ends—
Peak rate—°Oe/day
Alcohol / residual sugar—% vol
Nitrogen runs out—days
Peak yeast—million/mL
CO₂ at peak—L gas / L must / day

Your measurements

Measure density with a hydrometer — during fermentation alcohol distorts the refractometer. The first line is the initial must. Temperature, nitrogen and yeast are taken from the simulator mode; the model is fitted to your data with a speed factor.

Speed factor—×
Forecast: end—
Latest rate—°Oe/day
Status—

Current state

Rule-based assessment: each factor gets a risk score (0–3). A stuck fermentation is usually a combination of causes — start with the highest score.

Sugar left—g/L
Alcohol now—% vol
Rate—°Oe/day
Assessment—

Intuition

Fermentation is a process unfolding in time, and its curve — density by day — tells us more than anything else about what is happening in the qvevri or tank. A healthy fermentation has a characteristic shape: a delay of a few hours, a rapid fall in the first week, and a slow, gradually tapering "tail" at the end. If the curve departs from this shape — falling too slowly or flattening out at one level near the end — that signals a problem.

Four factors shape the curve: how many yeast cells grow (nitrogen decides this), how fast each cell ferments sugar (temperature), how strongly the accumulated alcohol inhibits them, and how many cells die toward the end of fermentation. On this page we combine these factors into one mathematical model, compare your measurements with it, and look for the causes of a slowdown.

History

Quantitative study of fermentation began in the 20th century. In 1913 Leonor Michaelis and Maud Menten derived the rate equation for enzyme reactions, and in 1942 Jacques Monod linked bacterial growth rate to nutrient concentration — both equations still underpin fermentation models today.

The first complete kinetic model of wine fermentation was built in 1980 by Roger Boulton (University of California, Davis). It described yeast growth, sugar consumption, ethanol inhibition and fermentation heat together, and was used to plan the cooling of large tanks. Later models (Cramer and co-workers 2002, Coleman and co-workers 2007) refined the role of nitrogen: it turned out that the most common cause of a sluggish fermentation is nitrogen shortage and the reduced cell numbers that go with it.

Model equations

The model tracks four variables: viable cells $X$, sugar $S$, assimilable nitrogen $N$, and alcohol $E = (S_0 - S)/16.83$.

Yeast growth is limited by nitrogen (Monod form) and inhibited by alcohol; death rises exponentially with alcohol and temperature:

$$\frac{dX}{dt} = \left[\mu(T)\,\frac{N}{N + K_N}\left(1 - \frac{E}{E_g}\right) - k_d\right]X, \qquad k_d = k_0\,e^{a(E - E_0)}\,e^{b(T - 20)}$$

Nitrogen is consumed only by growth: each new cell needs a certain amount ($Y_{XN} \approx 6\cdot 10^5$ cells/mL per mg N/L):

$$\frac{dN}{dt} = -\frac{1}{Y_{XN}}\,\mu(T)\,\frac{N}{N+K_N}\left(1 - \frac{E}{E_g}\right)X$$

Sugar consumption is proportional to the number of cells; each cell's rate depends on sugar concentration (Michaelis–Menten form) and is inhibited linearly by alcohol:

$$\frac{dS}{dt} = -q(T)\,\frac{S}{S + K_S}\left(1 - \frac{E}{E_{\max}}\right)X$$

Temperature acts on growth and sugar uptake through Rosso's cardinal model (see "Yeast: biology and selection"). The equations are solved numerically with a 0.1-hour step.

What the model says

Nitrogen is used up in 1–3 days during the growth phase — it is built into new cells. But it is the population grown in those days that ferments the sugar to the end. So a nitrogen shortage at the start shows itself at the end of fermentation: fewer cells cannot withstand the rising toxicity of alcohol, and fermentation slows or stops.

Temperature acts both ways. In warm must yeast grows and ferments quickly, but the toxicity of alcohol rises too — the membrane becomes more fluid and ethanol enters the cell more easily. That is why a hot, high-sugar fermentation starts fast but often stops with residual sugar. A cold fermentation is slow but finishes reliably — unless the temperature is close to the yeast's minimum.

Density during fermentation

During fermentation density falls for two reasons: sugar decreases, and ethanol is lighter than water (0.789 g/mL). The calculator uses an empirical linear approximation:

$$SG = SG_0 - 0.000428\,\Delta S - 0.00042\,E$$

where $\Delta S$ is the sugar fermented (g/L) and $E$ the alcohol (% vol). A 22 Brix must (SG ≈ 1.090) falls to ≈ 0.993 when fully fermented. Fermenting 1 g/L of sugar lowers density by ≈ 0.45 °Oe.

A refractometer is no use during fermentation: ethanol's refractive index is higher than water's, so the instrument reads "sweet" even when almost no sugar is left. Density must be measured with a hydrometer, at the same temperature or with correction (see "Sugars and must density").

Fermentation rate and CO₂

Fermentation rate is conveniently expressed in °Oe per day — the difference between two hydrometer readings. In a healthy fermentation at 18–22 °C the peak rate is 10–20 °Oe/day (25–45 g/L of sugar a day), falling to 1–2 °Oe/day at the end.

One gram of sugar gives 0.489 g of CO₂, or ≈ 0.27 litres of gas at 20 °C. At the peak, at 30 g/L/day, each litre of must releases ≈ 8 litres of CO₂ a day — a 2000-litre qvevri 16 m³. This is the basis for calculating ventilation of the marani.

Sluggish and stuck fermentation

A sluggish (slow) fermentation runs far below the normal rate but is still going. A stuck fermentation stops with residual sugar (> 2–4 g/L) and the density does not change for several days. Residual sugar feeds bacteria, so a stuck wine is exposed to acetic and lactic acid bacteria.

Main causes: nitrogen shortage (especially at the start); alcohol toxicity at high temperature; too low a temperature or a thermal shock; high initial sugar (osmotic stress); too much SO₂; very low pH; over-clarified must (yeast needs solid particles as a surface and to adsorb toxic fatty acids); a weak wild population in spontaneous fermentation.

A restart is difficult: a stuck wine already contains 10–13% alcohol, and fresh yeast cannot work in it directly. The standard procedure: detoxify the wine with yeast hulls (0.2–0.3 g/L), rehydrate an alcohol-tolerant strain and acclimatise it step by step — first to diluted, then to more and more of the stuck wine (on the "mother" principle) — and finally add it to the main mass.

Key quantities

QuantityTypical valueNote
Lag phase6 – 24 hdepends on temperature and SO₂
Nitrogen exhaustion1 – 4 daysat the end of the growth phase
Maximum population10⁸ cells/mLproportional to nitrogen
Peak rate (18–22 °C)10 – 20 °Oe/day25 – 45 g/L sugar
Fermentation duration10 – 30 dayscool white takes longer
SG change per 1 g/L sugar≈ 0.45 °Oeincluding alcohol
CO₂0.27 L gas / g sugar20 °C
Sign of sticking< 1 °Oe/day, sugar > 4 g/Lfor several days

Connections

"Yeast: biology and selection" describes the metabolism of a single cell and microbial succession; this page describes the behaviour of the population over time. "Preparing the must" calculates nitrogen and the yeast rate — the starting conditions of this model. The next page, "Thermodynamics", computes the heat released from the fermentation rate and the temperature change for each type of vessel — running the two models together gives a complete picture of a real fermentation.

Application — Georgian practice

Traditionally, fermentation in a qvevri was monitored by taste, smell and the behaviour of the pomace cap: when the cap "settled" and the must no longer "boiled", fermentation was considered finished. An experienced winemaker reads this well, but cannot detect a few grams of residual sugar — and it is precisely on this sugar that bacteria can start working over winter. So in a qvevri too it is worth measuring density daily with a hydrometer (on a sample taken from below the pomace cap) and recording the curve.

The particular feature of qvevri fermentation is temperature: in a large qvevri it rises to 28–32 °C, which speeds fermentation but increases the risk of sticking in hot years with high sugar. With the simulator you can assess in advance how reliable fermentation is for a given sugar, nitrogen and temperature, and if needed add nutrient or cool in time.

Glossary

TermDefinition
Fermentation kineticsQuantitative description of fermentation rate and how it changes
Monod equationDependence of growth rate on nutrient concentration
Michaelis–Menten equationDependence of enzyme rate on substrate
Ethanol inhibitionSlowing of growth and fermentation by alcohol
Sluggish fermentationA fermentation far slower than normal
Stuck fermentationFermentation stopping with residual sugar
Residual sugarSugar left after fermentation (g/L)
Yeast hullsYeast cell walls — an adsorbent of toxins and a nutrient
RestartReviving a stuck fermentation with acclimatised yeast
Speed factorThe model's fitting parameter for your measurements

Practical use in calculations

A must at 22 Brix, 20 °C, assimilable nitrogen 200 mg N/L, yeast 5 million cells/mL. The simulator shows: nitrogen runs out in 1.9 days, the population reaches 125 million/mL, the peak rate is 13.9 °Oe/day (31 g/L of sugar a day, ≈ 8 L of CO₂ per litre a day), and fermentation finishes in 15 days at 12.8% alcohol. If nitrogen is only 100 mg N/L, the population halves and fermentation drags on to ≈ 29 days; adding 0.3 g/L of DAP when a third of the sugar has fermented shortens it to ≈ 20 days. At 14 °C the same must needs ≈ 32 days, at 28 °C ≈ 9.

A 26 Brix must at 28 °C stops at about 15% alcohol with ≈ 9 g/L of residual sugar; at 20 °C and with more nitrogen the residual sugar falls to ≈ 3.6 g/L, but fermentation still does not quite finish. Such a must needs an alcohol-tolerant strain and a moderate temperature.

A winemaker measures density daily: 1.090, 1.088, 1.078, 1.064, 1.050, 1.039, 1.030. The model fits these data with a speed factor of 1.08 — fermentation is normal and will finish around day 13. In another cellar density fell from 1.105 to 1.020 and changed by only 1 °Oe over the last two days; the temperature is 31 °C and the initial nitrogen was 140 mg N/L. The diagnosis shows ≈ 60 g/L of sugar left, alcohol 11.2%, fermentation stuck; the likely causes are overheating, alcohol toxicity at high temperature and nitrogen shortage. The first step is cooling, then a restart with an acclimatised, tolerant yeast.

For beginners — controlling fermentation day by day

1. In the "Simulator" mode enter the must sugar, fermentation temperature, nitrogen and amount of yeast (these are calculated on the "Preparing the must" page). The upper chart shows how density should fall, the lower one the fermentation rate, yeast numbers and nitrogen.

2. Try lowering nitrogen or changing the temperature and see how the duration of fermentation changes. If a card warns you in red, there is a risk of a stuck fermentation under these conditions.

3. Once fermentation has started, measure density with a hydrometer every day at about the same time. In the "Track your fermentation" mode type the day and density, one per line (for example: 3 1.064).

4. The model fits your measurements and tells you how fast fermentation is compared with what is expected and when it will finish. If the status shows "slow" or "stuck", go to the diagnosis.

5. In the "Stuck fermentation diagnosis" mode enter the initial and current density, the change over the last two days, the temperature and other conditions. You get the probable causes ranked by risk and the first steps to take.

For professionals — limits of the model

1. The model parameters are typical literature values and do not reflect a particular strain or must. The "speed factor" is the only fitted parameter; if you have data from several years, its typical value is a useful guide for your cellar.

2. The model treats temperature as constant. In a real fermentation, especially in a large vessel, temperature rises — this is calculated on the next page, "Thermodynamics". A factor fitted to a large qvevri will be above 1 precisely because of this warming.

3. Oxygen is not in the model. A little aeration at the start of fermentation (racking, stirring the pomace) lets the yeast synthesise sterols and unsaturated fatty acids and raises its alcohol tolerance — one of the best preventions of a stuck fermentation.

4. The empirical density formula assumes constant dry extract. In a very extract-rich wine (made on the pomace) the final density can be 0.995–0.998 even when dry — confirm the end of fermentation with a residual-sugar analysis (under 2 g/L = dry).

5. The diagnosis rules are general. Often several causes act together — nitrogen shortage and high temperature, for example, amplify each other. Before a restart, check the wine microbiologically (microscope, volatile acidity), because bacteria multiply quickly in a stuck wine.

1. Why is the amount of assimilable nitrogen at the start of fermentation decisive, if yeast uses all of it within the first 1–3 days?

In the growth phase nitrogen is built into new cells. When it runs out, growth stops, and the existing population ferments the remaining sugar. Less nitrogen → fewer cells → slow fermentation and risk of sticking.

2. Why does fermentation slow down at the end, even though sugar remains?

Ethanol changes membrane fluidity and inhibits sugar transporters; its toxicity rises at high temperature. That is why hot, high-sugar fermentations carry the greatest risk of sticking.

3. Why should density be measured with a hydrometer rather than a refractometer during fermentation?

Ethanol's refractive index is higher than water's, so during fermentation the refractometer reads "sweet" even when almost no sugar is left. A hydrometer measures density, which falls with alcohol.

4. Density stopped at 1.020, the temperature was 31 °C, alcohol ≈ 11%. What is the most likely cause?

At high temperature ethanol is far more toxic to yeast, and fermentation can stick even at 11%. Cooling and a restart with a tolerant, acclimatised yeast are needed.