Evaluation of two unstructured mathematical models for the penicillin G fed-batch fermentation

Area of Science:

• Biochemical Engineering
• Process Modeling
• Fermentation Technology

Background:

• Penicillin G fed-batch fermentation is crucial for antibiotic production.
• Existing mathematical models (Heijnen et al., 1979; Bajpai & Reuss, 1980) offer frameworks for process optimization.
• Understanding model behavior is key to maximizing penicillin G yield.

Purpose of the Study:

• To compare the Heijnen et al. (1979) and Bajpai & Reuss (1980) mathematical models for penicillin G fed-batch fermentation.
• To analyze the impact of metabolic assumptions and kinetic differences on product optimization.
• To identify limitations in parameter estimation and propose strategies for improvement.

Main Methods:

• Comparative analysis of two established mathematical models for penicillin G fed-batch fermentation.
• Detailed examination of metabolic assumptions and kinetic parameters within each model.
• Evaluation of parameter estimation reliability using simulated fermentation data.

Main Results:

• Significant differences in model behavior regarding product optimization were observed due to variations in metabolic assumptions and kinetics.
• Physical and biochemical shortcomings were identified in both models.
• Reliable estimation of model parameters is challenging with simple, low-substrate fermentation data.
• Certain model parameters were found to be critical in determining the impact of feeding strategies on final product yield.

Conclusions:

• Mathematical model selection significantly influences penicillin G production optimization strategies.
• Experimental design, particularly feeding strategy optimization, is essential for accurate parameter estimation.
• Further research is needed to refine models and improve the reliability of parameter estimation in fermentation processes.