Facilitating control of fed-batch fermentation processes by monitoring the growth rates of Saccharomyces cerevisiae

In this paper we present a growth rate controller for a fed-batch bioprocess. An observer estimates the growth rate. The observer is based on knowledge about the stoichiometric relations of the process. Furthermore, the observer needs online measurements of the oxygen uptake rate and the carbon-dioxide production rate. The control structure is a simple one, and the fed-batch fermentation process is inherently nonlinear. A growth rate dependent precompensation is introduced to linearise the process. The remaining control actions can then be tackled using a normal PI-controller. Simulation studies show good performance for oxidative growth on glucose. Deterioration of the control is expected if the growth rate becomes bigger than the critical growth rate. Furthermore, some non-modelled dynamical effects occur during set-point changes. To some extend these effects can be modelled and the model can be incorporated in the controller. The overall performance of the controller increases. Yet this improvement is at the cost of a more complex structure. A trade off has to be found between control performance and controller complexity