Model Predictive Temperature Control for Calorimetric Measurements in Baker’s Yeast Fermentations

We intended to establish a system for calorimetric measurements on a pilot scale. The on‐line calculation of model equations of the heat balance around the reactor is supposed to be simple at a constant fermentation temperature. To obtain higher accuracy in temperature control, a model predictive control (MPC) system is developed. MPC with the accuracy of 0.1 °C is much more precise than the classical single loop temperature control (SLC) with 1 °C. Calorimetry has a very high dynamic behavior compared with other on‐line determined parameters also representing metabolic activity. In a purely oxidative fed‐batch fermentation of baker’s yeast the accuracy of calorimetry is verified. The combustion enthalpy of glucose, calculated using on‐line data of microbial heat production and substrate consumption, corresponds very well with the data in the literature. It is now possible to determine the fraction of substrate uptake, utilized for energy (factor k) and biosynthesis (factor a) metabolisms on‐line.