Integrated design of monitoring, analysis and maintenance for filamentous sludge bulking in wastewater treatment

Stable operation of activated sludge process is often compromised by the occurrence of filamentous sludge bulking in wastewater treatment. Standard maintenance is implemented as a single task, which is inefficiently manipulated based on purely chemical dosage by experiences. This paper proposes a novel and efficient maintenance framework, including a set of activities: incipient fault detection, causality analysis, remaining useful life (RUL) prediction and maintenance schedule, to develop a collective strategy for sludge bulking management. In this framework, Canonical Correlation Analysis (CCA) work together with Kurtosis (KU) and an autoregressive moving average (ARMA) to identify incipient sludge bulking and to capture the evolution of sludge bulking over a long horizon. The proposed framework was tested by the collected the field data. The results showed that the proposed framework is capable of detecting (Type I error, 2.1%; Type II error, 0%), locating (root cause: temperature), calculating RUL (32-days) and maintaining sludge bulking.