A new approach to predicting the extraction of malvidin-3-glucoside during red wine fermentation at industrial-scale

Abstract

Red wine quality is largely determined by the concentration of phenolic compounds such as anthocyanins that are extracted from skins and seeds during alcoholic fermentation. As such, the ability to control this extraction process is a key consideration for winemakers. Current mathematical models based on regression analysis methods that are used to describe phenolic extraction during fermentation can provide insight into the effect of changing process conditions, however their use as predictive tools for process control and management of future fermentation scenarios is limited. In this study, red wine fermentations undertaken at a large commercial winery are compared with simulations of malvidin-3-glucoside extraction (the predominant red wine anthocyanin), using a first principles mass-transfer model with incorporated response surface equations for calculating relevant mass transfer parameters throughout fermentation. In general, the model employed to simulate this mass transfer process showed consistent extraction patterns to that measured experimentally in commercial large-scale fermentations, and displayed good predictive capabilities for malvidin-3-glucoside concentrations at the latter stages of fermentation, highlighting its potential usefulness as a new tool for quality management and process control during winemaking.