Mechanistic Models for Process Development and Optimization of Fed-batch Fermentation Systems

This work discusses the application of mechanistic models to pilot scale filamentous fungal fermentation systems operated at Novozymes A/S. For on-line applications, a state estimator model is developed based on a stoichiometric balance in order to predict the biomass and product concentration. This is based on on-line gas measurements and ammonia addition flow rate measurements.

Additionally, a mechanistic model is applied offline as a tool for batch planning, based on definition of the process back pressure, aeration rate and stirrer speed. This allows the batch starting fill to be planned, taking into account the oxygen transfer conditions, as well as the evaporation rates of the system.

Mechanistic models are valuable tools which are applicable for both process development and optimization. The state estimator described will be a valuable tool for future work as part of control strategy development for on-line process control and optimization.