Model-based control of the ammonia nitrogen removal process in a wastewater treatment plant

Abstract

This chapter investigates what benefits could be achieved from model-based ammonia nitrogen controllers in wastewater treatment plants (WWTPs) in comparison with PI controllers. The controllers addressed are a model-based feedforward-feedback controller and a model predictive controller (MPC). Validation of the controllers was performed both in simulation and on a real plant in the context of various performance criteria related to an effluent ammonia limit violation and air consumption. The simulation results indicate that the feedforward-feedback controller and MPC result in reduced air consumption compared to the PI controller while achieving the same or better ammonia removal. When model-based ammonia controllers are applied to a full scale pilot plant they have to be simplified and properly accommodated. The application of controllers confirmed that the feedforward-feedback controller and MPC give better results than the PI controller. The main reason for this lies in the fact that they use the additional measurable disturbance of the influent ammonia. Interestingly, at the real plant poorer results are obtained with the MPC than with the feedforward-feedback controller. The explanation for this could be associated with the various limitations in applying a more complex MPC in a real environment.