The prediction of the cell and substrate in a fed-batch fermentation system by modification of models

Abstract

This objective of this study is to predict cell masses and substrate consumption in a fed-batch system after switching from a batch mode. A 2.0 litre reactor vessel was used to monitor the growth characteristics of Candida utilis in a bioreactor with the aid of a computer. The study enabled the development of two major mathematical models and their respective modified models for the prediction of the cell and substrate throughout the fermentation cycle. In the first model, the overall yield coefficient, Yx/s, obtained by multiple linear regressions relating yield to the input variables for the batch model was assumed to be constant throughout the fed-batch mode. With this yield, the modified Bergter and Knorre model is applied to determine the specific growth rate over the entire fed-batch fermentation cycle to predict both the cell and substrate concentrations. The second model assumes that the fed batch mode operates at maximum specific growth rate, µmax, after switching from batch mode. The value was obtained by using multiple linear regressions to relate the maximum specific growth rate, µmax to the input variables from the batch mode. Both the models were modified and have obtained reasonably good results when compared with the experimental data.