Dynamic modelling and system identification of coagulant dosage system for water treatment plants

In this study, dynamic modelling of a coagulant dosage system using the principle of continuously stirred tank reactor (CSTR) is proposed for effective coagulation control in water treatment process. The multi-input multi-output (MIMO) system dynamics of the model are estimated using prediction error method (PEM) and autoregressive moving average with exogenous input (ARMAX) technique. Simulations results and comparative analysis show that PEM model estimates the system parameters better than the ARMAX model. In addition, the proposed dynamic model will serve as a viable alternative to the widely used data-based modelling techniques for coagulation control. It will adequately support the development of an efficient multivariable control strategy that will outperform the existing standard Proportional Integral (PI) feedback control strategy commonly used in water treatment coagulation process.