The knowledge of the effect of uncertainties on kinetic parameters is fundamental, when a model is used to improve the performance of an industrial bioreactor. In this work a 100 m3 industrial bioreactor operating as a fed-batch is simulated by computational fluid dynamic (CFD) methods to investigate the effects of varying kinetics describing the growth of S. cerevisiae. The results of the simulation show that the kinetic parameters used to describe the anaerobic metabolic state have the largest influence on the glucose distribution causing a maximum deviation of 25% of the glucose profile corresponding to the base case. The uncertainties of the parameters describing aerobic metabolism affect the glucose with 2% to 5%. The sensitivity of the kinetics is also investigated for four different volumes of the bioreactor operating as a fed-batch. The uncertainty of the kinetics has the largest impact at the beginning of the process, where the mixing time is relatively low and a small impact is found at the end where the mixing time is high. Finally, the influence of a varying yield coefficient on the development of the cell concentration is investigated. Only a small deviation in the cell concentration is found. However, by comparing the CFD simulation with a simulation where ideal mixing is assumed, a deviation of the biomass concentration of approximately 10 g/L or 33% was predicted for the final volume. This work illustrates the possibility to evaluate the effect of variations in kinetic parameters using CFD, and gives an insight on the effect of potential strain improvements on the performance of a fed-batch process.
- Computational fluid dynamics
- Large scale industrial fermentation
- Uncertain growth kinetics