MULTIVARIABLE SUPERVISORY CONTROL OF A 4-EFFECT FALLING-FILM EVAPORATOR

This paper describes the second part of a project that NIZO and SETPOINT-IPCOS have defined. The project goal is to evaluate tools for dynamic modelling and supervisory multivariable control design of multiple-effect falling-film evaporators, with the NIZO four-effect evaporator as case. In the first part the results from system identification from plant data and physico mathematical modelling are compared. This part discusses the controller design and performance. The multiple control objectives are discussed and motivated. The design techniques are briefly highlighted. Practical results from tests on the NIZO installation are presented for two designs: one based on an identified model and a second based on a physical model. The practical results show that both models are accurate enough to obtain an excellent performance from modern multivariable controller designs. The resulting disturbance rejection and fast settling of multiple process outputs to their desired values at start-up, will enable industry to largely improve their process operation and benefit economically from larger yield, improved product quality and reduced energy consumption and material waste.