On-site wastewater treatment plants are usually unattended, so undetected failures often lead to prolonged periods of reduced performance. To stabilize the good performance of unattended plants, soft-sensors could expose faults and failures to the operator. In a previous study, we developed soft-sensors and showed that soft-sensors with data from unmaintained physical sensors can be as accurate as soft-sensors with data from maintained ones. The quantities sensed were pH and dissolved oxygen (DO), and soft-sensors were used to predict nitrification performance. In the present study, we use synthetic data and monitor three plants to test these soft-sensors. We find that a long sludge age and a moderate aeration rate improve the pH soft-sensor accuracy, and that the aeration regime is the main operational parameter affecting the accuracy of the DO soft-sensor. We demonstrate that integrated design, monitoring, and control are necessary to achieve robust accuracy and to obviate case-specific fine-tuning. Additionally, we provide a unique labelled dataset for further feature and data-driven soft-sensor development. Our approach is limited to sequencing batch reactors. Moreover, nitrite accumulation and alkalinity limitation cannot be detected. The strength of the approach is that unmaintained sensors drastically reduce monitoring costs, enabling the monitoring of plants hitherto unchecked.
Deep learning-based morphology classification of activated sludge flocs in wastewater treatment plants
