In the chocolate industry, product changes are performed as flushing processes, in which one product is displaced by a subsequent product. The end of the process is detected by visual assessment based on the color change of the discharged product. Due to limited process knowledge and a lack of suitable measuring systems, these processes have so far been designed with rigid process parameters and large safety factors.
First, the authors experimentally evaluated three measurement methods – surface acoustic wave-based (SAW), capacitance-based (CMS) and camera-based (CAM) – on a horizontal DN 25 test rig and compared them to a reference cross-section measurement. SAW demonstrated the highest potential, as it can distinguish between the two fluids most effectively. CMS is suitable for draining or pigging processes with air phases in the system and CAM for end-point measurements, since it permits the analysis of the residual layers close to the wall.
The authors also investigated two alternative strategies for product changeover – simple flushing after a draining process (partially filled pipe) and flushing sequences – using experiments and numerical models. All strategies were compared to a conventional industrial product changeover with identical process times of 40 s. Flushing sequences showed the greatest benefit, reducing flushing mass by up to 32 %, while the other strategy mainly enabled direct product recovery. These findings demonstrate that targeted measurement methods and new flushing strategies can substantially reduce product loss in product changes in industrial chocolate processing.