Automatic identification of activated sludge disturbances and assessment of operational parameters

Activated sludge systems are prone to be affected by changes in operating conditions leading to problems such as pinpoint flocs formation, filamentous bulking, dispersed growth, and viscous bulking. These problems are often related with the floc structure and filamentous bacteria contents. In this work, a lab-scale activated sludge system was operated sequentially obtaining filamentous bulking, pinpoint floc formation, viscous bulking and normal conditions. Image processing and analysis techniques were used to characterize the contents and structure of aggregated biomass and the contents of filamentous bacteria. Further principal component and decision trees analyses permitted the identification of different conditions from the collected morphological data. Furthermore, a partial least squares analysis allowed to estimate the sludge volume index and suspended solids key parameters. The obtained results show the potential of image analysis procedures, associated with chemometric techniques, in activated sludge systems monitoring.

► Automatic image analysis methodology was evaluated to monitor activated sludge.
► Principal component analysis and decision trees identified different operating conditions.
► Partial least squares estimated sludge volume index and total suspended solids.
► An invaluable method in quality assessment of activated sludge was established.