Real-time monitoring of enhanced biological phosphorus removal in a multistage EBPR-MBBR using a soft-sensor for phosphates

Abstract

Enhanced biological phosphorus removal (EBPR) from municipal wastewater has been achieved in a multistage Moving Bed Biofilm Reactor (MBBR) configuration. The process operations can be further optimized by real-time monitoring of water quality parameters in the individual chambers of the EBPR-MBBR process. This work presents a hybrid, soft-sensor as a cost-effective monitoring option for real-time estimation of phosphates (PO43−-P) and soluble COD (sCOD) concentrations in the anaerobic chambers of a multistage EBPR-MBBR pilot plant. The soft-sensor is developed by implementing an Extended Kalman filter on a reduced-order nutrient removal model. The hybrid model is constructed by combining mechanistic elements of phosphorus release kinetics in anaerobic conditions, and a statistical model correlating PO43−-P and sCOD concentration with conductivity measurements. A systematic method for developing, calibrating a reduced-order model and tuning of the Kalman filter parameters have been discussed in this work. The drift in soft-sensor performance was studied and practical solutions were suggested for re-calibrating the model utilizing data from periodic lab measurements. The estimation results are successfully validated against standardized lab measurements to demonstrate the accuracy of the soft-sensing algorithm.