Analysis of a Commercial Red Wine Fermentation Dataset with a Wine Kinetic Model

Abstract

The adoption of sensors to monitor wine fermentation enables the collection of large datasets that relate the initial juice chemistry, density and temperature patterns during fermentation to fermentation outcomes. Wine kinetic models are now being applied to commercial fermentations in real time to identify abnormal or sluggish fermentations. In this work, 222 red wine fermentations from five harvests at two commercial wineries were evaluated by a wine fermentation model. The model parameters, initial juice chemistries and fermentation outcomes were analyzed for trends and relationships between them. While the fermentations with higher initial assimilable nitrogen concentrations had higher maximum fermentation rates, this did not guarantee successful fermentation outcomes in the tailing stage of the fermentation. Neither the initial, final, minimum and maximum temperatures, nor the initial pH, titratable acidity, measured yeast-assimilable nitrogen and primary amino nitrogen concentrations had any significant correlation with the maximum fermentation rate or successful completion of the fermentation. These results suggest that the initial juice-assimilable nitrogen measurements for these juices are of limited use in predicting slower and incomplete fermentation outcomes.

Keywords

• wine fermentation
• modeling
• incomplete and sluggish fermentation
• Industry 4.0