Performance monitoring for separator system in tapioca starch industry using state-space model

This research aimed to monitor and detect the performance of hydrocyclones in a separator unit in tapioca starch industry. This research considered a case study of the hydrocyclone system with 6 sub-units. One hydrocyclone unit houses 400-500 cyclonettes. Hydrocyclone performance depends on the pressure drop and feed concentration of the system. This research proposed to use the state-space model (SSM) analysis techniques for each section of hydrocyclone. The model was validated by comparing with the starch concentration in the separation process. Mass balance technique was used to estimate feed concentration of all states that were not measured. The efficiency unit could monitor and detect the obstruction or clogging in the separator unit by observing change in performance. The monitoring efficiency had errors at ±3%.