Abstract
A study of the feasibility of perturbation-based control methods in industrial fed-batch fermentations based on experiments in industrial production scale bioreactors (>100 m3) is presented, as well as modelling of the relation between substrate feed rate and dissolved oxygen level in such a process. Several different types of perturbation-based control methods have been suggested for control of this type of process but it has been reported that perturbations in the feed rate may cause decreased productivity in fermentations. The results of this study show that perturbations in the feed rate of production scale fermentations can achieve significant dissolved oxygen level responses without decreased productivity. A model based on data for dissolved oxygen responses and a simulation using a simple observer are given, showing that it is possible to model industrial mixing dynamics in a simple way and that this can be used for perturbation-based on-line estimation of the metabolic state of the system in regard to overflow metabolism. A frequency region where the model can be used has been identified, indicating which frequencies would be suitable for perturbation-based control in industrial fermentations.