Yeast propagation increasingly stands for a central step in beer production and emprises an important economical and technological factor in brewing practice. Vitality and quality of the propagated yeast exerts a relevant influence on the subsequent fermentation run and the resulting beer quality. Therefore, the yeast inoculum must be available at pitching time in the right amount and especially in the right quality. To guarantee this by process management tools, in particular by modelling and active process control, is a main feature of this work. Two kinetic models of yeast propagation are introduced. These modelling approaches represent the basis for a control strategy aiming on the provision of an optimal inoculum at the starting time of subsequent industrial fermentations. Both models, a Black Box model and a metabolic model, include respiratory metabolism on sugars and ethanol as well as fermentative metabolism on sugars. Limitation effects, occurring due to specific nutritional data of the growth medium beer wort, were taken into account for sugar, nitrogen, ethanol and oxygen concentrations. Correspondingly, inhibitions of the metabolism by ethanol and high sugar concentrations were formulated. The models especially represent the Crabtree-effect. For model validation, literature data were used and selected experiments within the relevant range of manipulated variables (temperature, dissolved oxygen) were conducted for different yeast strains and propagation strategies. In a sensitivity analysis three parameters were identified as particularly relevant. After adaptation of these parameters on data sets, simulations, based on the suggested models, matched these data with a deviation below 10 mmol/L (0.2% w/w for sugar concentrations and 6.2*106 cells/mL for biomass concentration). The variable parameters showed a characteristic temperature dependency, which could be described by mathematical functions. The implementation of these functions in both models allowed predictive simulations of the yeast propagation process even applying non-isothermal trajectories. During experiments it was proved that in ideally mixed fermenters the dissolved oxygen concentration affects the yeast growth only below concentrations of 0.1 ppm. Predictive simulations allowed an active process control by a precise adjustment of trajectories of both, temperature and dissolved oxygen concentration. Thus, the optimal crop time of the inoculum could be varied within a period of two days in order to maintain high fermentation activity for the subsequent anaerobic fermentation. The validity of the modelling approaches was proved for industrial propagations and fermentations in breweries as well. In particular for the industrial propagations above mentioned accuracies were achieved. Accuracies of the simulation of brewing yeast fermentations resulted in a practicable range not before a sedimentation model was included in the process models.
«
Yeast propagation increasingly stands for a central step in beer production and emprises an important economical and technological factor in brewing practice. Vitality and quality of the propagated yeast exerts a relevant influence on the subsequent fermentation run and the resulting beer quality. Therefore, the yeast inoculum must be available at pitching time in the right amount and especially in the right quality. To guarantee this by process management tools, in particular by modelling and a...
»