Dynamic global sensitivity analysis in bioreactor networks for bioethanol production

Abstract

Dynamic global sensitivity analysis (GSA) was performed for three different dynamic bioreactor models of increasing complexity: a fermenter for bioethanol production, a bioreactors network, where two types of bioreactors were considered: aerobic for biomass production and anaerobic for bioethanol production and a co-fermenter bioreactor, to identify the parameters that most contribute to uncertainty in model outputs. Sobol’s method was used to calculate time profiles for sensitivity indices. Numerical results have shown the time-variant influence of uncertain parameters on model variables. Most influential model parameters have been determined. For the model of the bioethanol fermenter, μmax (maximum growth rate) and Ks (half-saturation constant) are the parameters with largest contribution to model variables uncertainty; in the bioreactors network, the most influential parameter is μmax,1 (maximum growth rate in bioreactor 1); whereas λ (glucose-to-total sugars concentration ratio in the feed) is the most influential parameter over all model variables in the co-fermentation bioreactor.