A novel explainable kinetic model for two-stage fermentation profile

Abstract

Bio-products have significantly contributed to human society, particularly in advancing sustainability efforts. Kinetic models have been developed for better understanding fermentation dynamics, which is crucial for optimizing and scaling up the fermentation process. However, conventional kinetic models often struggle to accurately decipher two-stage fermentation profiles or fully explain the process. To bridge this gap, modified hyperbolic secant (MHS) model was proposed in this study, with three case studies to evaluate its applicability. Firstly, the MHS model was utilized to simulate the production of 1,3-propanediol fermentation, employing a co-fermentation approach with glucose and glycerol. The MHS model not only resulted in a greater fitting accuracy than conventional Logistic model but also provided essential kinetic parameters. Secondly, when applied MHS model to simulate biohydrogen production in dark fermentation, the MHS model exhibited superior fitting performance in comparison with the Monod model. Thirdly, the MHS model portrayed, when compared to Gompertz model, better methane profiles in anaerobic digestion. These case studies illustrate that the MHS model as a reliable, comprehensible, and universally applicable kinetic model. The implementation of the MHS model holds promise for supporting bioproducts production and offering valuable insights into process development.