Gas transfer–driven kinetic model for aerobic poly-4-hydroxybutyrate production under dissolved oxygen–limited conditions

• Kinetic modeling framework constructed from routine industrial operation dataset • High prediction accuracy under dissolved oxygen–limited and non-limited conditions • Oxygen transfer rate-driven regulation of substrate utilization rate • Oxygen uptake–based logistic correction for metabolite yield coefficients • Model simulations enable k L a estimations for biotic processes in various scales The modern bioindustry is increasingly focused on replacing fossil fuel–based chemical production with greener processes, with poly-4-hydroxybutyrate (P4HB) emerging as a promising alternative to conventional petrochemical plastics. Advanced kinetic models helps to efficiently design, optimize, and scale-up novel bioprocess, such as aerobic P4HB production. Notably, dissolved oxygen depletion is a frequent occurrence during P4HB production by Escherichia coli owing to the extremely high oxygen utilization rate, which limits microbial activity and causes unexpected metabolic shifts, depending on the specific operating conditions. This study proposes an explainable model for aerobic bioprocesses that integrates material balances, transport phenomena, and microbial physiology. The model considers microbial kinetics and cellular metabolism based on oxygen transfer rate (OTR)-related terms. Employing OTR-driven substrate consumption kinetics and dynamic yield coefficient variations, the model can precisely track substrate depletion, biomass growth, P4HB yield, and time-course oxygen uptake across a range of aeration conditions. Simulation with independent datasets confirmed precise prediction of the aforementioned indicators at reactor scales of up to 3 kL. Furthermore, the versatility of the model is demonstrated through estimation of the volumetric mass transfer coefficient ( k L a ) under actual biotic conditions without gas analysis data. The framework facilitates robust bioprocess design and optimization, scalable k L a estimation, and provides the basis for a soft sensor implementation.