Technical University of Munich, Freising, Germany
The actual fermentation process of beer is based on brewer’s individual recipes and not consequently linked to precise process monitoring and control. Nevertheless, robust and adequate control of fermentation plays a decisive role to guarantee defined processes and optimal product quality. Due to the lack of inline measurement systems, varying raw material qualities, and different yeast activities, there is always an inevitable variation in processes causing quality losses and production rejects in the worst case. Actual methods for monitoring fermentation activity during beer production are not inline capable and therefore not suitable for adaptive control of the fermentation process based on the actual process state.
In this work, we present a novel ultrasound-based sensor technology to monitor fermentation activity during beer production on a technical scale. The changing fermentation conditions, different growth of yeast cells as well as the formation of gas bubbles influence the propagation of ultrasound signals. These mentioned changes during fermentation (including temperature changes) showed a high impact on the ultrasound signals. An ultrasound sensor setup for measurements in pulse-echo mode based on a piezoelectric ceramic at the fundamental frequency of 2 MHz was constructed. Offline-based reference measurements to characterize the fermentation activity were performed and correlated with the ultrasonic signals. The ultrasound signals were analysed by features based on physical and signal dependent effects using a hybrid evaluation approach.
With the results of the offline-based reference measurements and the use of multivariate data analysis of the ultrasonic signals, correlations with R² > 0.9 between alcohol production, wort-extract concentration and measured ultrasonic signals were achieved. Thereby, the fermentation activity was continuously evaluated and monitored. Furthermore, the developed ultrasonic sensor can be implemented in industrial production scale serving as an example of the applicability of the developed approach.
Michael Metzenmacher studied Pharmaceutical Bioprocess Engineering at the Technische Universität München, where he graduated as a M. Sc. in 2016. He has worked with Prof. Dr.-Ing. Thomas Becker since 2017 as a member of the scientific staff in the BioPAT and Digitization Work Group at the Chair of Brewing and Beverage Technology. His work focuses on using ultrasound to characterize high-viscosity, multiphase systems of substances.