Soft-sensing strategy for real-time biomass and poly(3-hydroxybutyrate) estimation in methanotrophic cultures using...

Real-time estimation of microbial biomass and product formation is crucial for effective monitoring and control of bioprocesses. In this study, we present a generalized soft-sensing algorithm that utilizes commonly available dissolved oxygen (DO) and base addition signals to estimate both total cell concentration and intracellular poly(3-hydroxybutyrate) (PHB) content. Total cell concentration was calculated using a DO-based oxygen balance equation, while PHB accumulation was inferred from a nitrogen balance framework, in which nitrogen consumption was estimated based on the frequency of base additions under pH-stat control. We validated the algorithm using methane-fed cultures of Methylocystis sp. MJC1, demonstrating high accuracy in predicting total biomass concentration across various media compositions and operational conditions, provided that ammonium was used as the nitrogen source. Notably, PHB concentrations were reliably estimated in real time by subtracting the non-PHB biomass (derived from base addition signals) from the DO-based total biomass. Although some deviations in PHB predictions were observed, these were attributed to imperfect correlations between nitrogen consumption and pH behavior. The method requires minimal data preprocessing and has low computational demands, making it well-suited for real-time applications in gas fermentation as well as other aerobic bioprocesses. This soft-sensing strategy offers a simple, robust, and cost-effective approach for online monitoring of microbial growth and product formation.