(50 g L
À1 dry biomass concentration within 48 h). The result of the first OLFO
controlled simulated cultivation of S. cerevisiae is presented in Fig. 7.
In the first OLFO controlled Digital Twin cultivation of S. cerevisiae, only a low
dry biomass concentration of 4 g L
À1 could be achieved within the processing time
of 48 h, due to low substrate feed rates (max. 0.03 mL min
À1 ) determined by the
OLFO controller. A detailed analysis revealed an ethanol inhibition in the mathematical process model already starting at less than 5 g L
À1 . Consistently, the OLFO
controller calculated low substrate feed rates to avoid ethanol formation. However,
the resulting low glucose concentration limited growth.
Increasing the ethanol inhibition constant in the mathematical process model to
approx. 30 g L
À1 led to an increase in the final simulated dry biomass concentration
(15 g L
À1 ). However, the set control target could not yet be achieved. Based on
subsequent simulations with the Digital Twin, further controller model adjustments
such as modifying the metabolic rates related to the Crabtree effect, adjustments of
uptake rates, etc. were performed. The intervals for model adaptation and subsequent
substrate feed optimisations were varied and different compositions of the nutrient
medium were examined via simulations with the Digital Twin.
In the sixth OLFO controlled cultivation simulated on the Digital Twin, the set
control target eventually was exceeded by reaching a final biomass density of
80 g L
À1 within 48 h (Fig. 8) and less than 10 g L
À1 ethanol.
The resulting OLFO controller (developed on the Digital Twin) was transferred to
the real process for experimental validation. Figure 9 shows the results of the OLFO
controlled S. cerevisiae cultivation.
In the OLFO controlled real cultivation, a dry biomass concentration of more than
50 g L
À1 was achieved within 48 h. Both, the substrate feed rates (Fig. 9b) and the
Fig. 7 Result of the first OLFO controlled simulated cultivation of S. cerevisiae using the Digital
Twin “SSF-BC-Simulator”
Digital Twins for Bioprocess Control Strategy Development and Realisation
87
Précédent

- 94/260

Suivant