Adaptive data-derived anomaly detection in the activated sludge process of a large-scale wastewater treatment plant.

This work examines real-time anomaly detection and isolation in a full-scale wastewater treatment application. The Viikinmäki plant is the largest municipal wastewater treatment facility in Finland. It is monitored with ample instrumentation, though their potential is not yet fully exploited. One reason that prevents the use of the instrumentation in plant control is the occasional insufficient measurement performance. Therefore, we investigate an intelligent anomaly detection system for the activated sludge process in order to motivate a more efficient use of sensors in the process operation. The anomaly detection methodology is based on principal component analysis. Because the state of the process fluctuates, moving-window extensions are used to adapt the analysis to the time-varying conditions. The results show that both instrument and process anomalies were successfully detected using the proposed algorithm and the variables responsible for the anomalies correctly isolated. We also demonstrate that the proposed algorithm represents a convenient improvement for supporting the efficient operation of wastewater treatment plants.