Dynamic global sensitivity analysis in bioreactor networks for bioethanol production

Area of Science:

• Biochemical Engineering
• Process Systems Engineering
• Computational Biology

Background:

• Bioreactor models are essential for optimizing bioprocesses like bioethanol production.
• Understanding parameter uncertainty is crucial for reliable model predictions.
• Dynamic global sensitivity analysis (GSA) provides a robust framework for identifying influential parameters.

Purpose of the Study:

• To identify key parameters contributing to output uncertainty in dynamic bioreactor models of increasing complexity.
• To compare sensitivity results across different bioreactor configurations.
• To inform model refinement and experimental design in bioprocess engineering.

Main Methods:

• Dynamic global sensitivity analysis (GSA) was applied to three distinct bioreactor models.
• Sobol's method was employed to compute time-dependent sensitivity indices.
• Models included a single bioethanol fermenter, a network of aerobic and anaerobic bioreactors, and a co-fermentation bioreactor.

Main Results:

• Parameter influence on model outputs varied over time.
• For the bioethanol fermenter, maximum growth rate (μmax) and half-saturation constant (Ks) were most influential.
• In the bioreactor network, μmax in the first bioreactor (μmax,1) was dominant.
• For the co-fermentation bioreactor, the glucose-to-total sugars ratio (λ) was the most significant parameter.

Conclusions:

• The study successfully identified critical parameters influencing bioreactor model dynamics.
• Sensitivity analysis revealed time-variant parameter importance, highlighting the need for dynamic assessment.
• Findings provide valuable insights for parameter estimation and model validation in bioprocess development.