Model predictive control for liquid-solid cross flow ultrafiltration membrane separator

Abstract

The petrochemical, metallurgical and processing industries discharge of oily wastewater into the sea or rivers has been under increasingly careful. A new separation of the oil out of the wastewater technology called membrane technology has been used widely in separation industry because it does not give rise to phase changed by adding chemical and heat. The cross-flow ultrafiltration membrane of an Oily-Water emulsions system combine mathematical and experimental models; the mathematical model is based on solute diffusion through membranes and the experimental model focuses on the fouling mechanisms. The control of permeated flux of water using the transmembrane pressure, a manipulated variable, can be studied in two cases. One is to control the flux at a constant set point obtained from an overall optimization. The other one is to control the flux at three interval constant set points obtained from a dynamic optimization. In this work, Model Predictive Control (MPC), one of most widely implemented advanced process control technology for chemical process recently, is applied the flux of the system and its performance is compared with those of a conventional PID controller and a Generic Model Controller (GMC). Simulation results have shown that the PID controller cannot control the permeated flux of water to the set point for both cases. Although the GMC controller is able to control the flux at the set point obtained from the overall optimization, it cannot control the flux at the set points obtained from the dynamic optimization. The MPC controller, on the other hand, can control the flux at the set points obtained from both overall and dynamic optimization. This shows that the MPC controller provides the best control response over the GMC and the PID controllers. In the presences of plant/model mismatch, the Kalman filter is incorporated into both MPC and GMC controllers to estimate unknown/uncertain parameters. Here, the viscosity and constant's fieldmodel. unknown/uncertain parameters, have been increased 50% from nominal values. It was found that both MPC and GMC controllers with the Kalman filter from both cases are still able to control the water flux as well as the nominal cases. However, the MPC provides better control response than the GMC controller. Therefore the MPC with the Kalman Filter is the most robust and effective control algorithm for the cross flow ultrafiltration membrane of the oily water emulsion system