Automated Density Measurement With Real-Time Predictive Modeling of Wine Fermentations

Wine fermentation is traditionally monitored through manual measurements of density; however, such measurements become labor-intensive as the number of fermentations increases. Various
in situ
density measurement techniques are currently used in the wine industry; however, they have not been widely studied nor widely adopted. In this work, we combine
in situ
measurements of density based on differential pressure measurements with a wine kinetic model and parameter estimation routine to predict the progression of a 1500-L wine fermentation. Pressure transducers were mounted at three vertical positions, and the influence of sensor precision and vertical separation of the transducers was investigated. The transducer measurements were in agreement with the density of samples from a hand-held densitometer. The density measurements from the start of fermentation to 75, 100, 125, and 150 h were used to predict future fermentation behavior. The results show that the automated measurement methodology, combined with a wine kinetic model, can be used in commercial wineries to monitor and predict ongoing fermentations.