Dynamics and Control of Falling Film Evaporators with Mechanical Vapour Recompression

Inferential cascade control of multi-effect falling-film evaporator

“…In this algorithm, the secondary process variable is the product concentration exiting the first effect and primary process variable is product concentration exiting the third effect and the manipulated variable is steam flow rate. The principle disturbances arise due to variation in the feed concentration and it was shown (Winchester & Marsh, 1999) that the affects of other disturbances (product feed temperature and condenser cooling water temperature) on the concentration were insignificant. The other key control loop on the evaporator is regulation of the third effect temperature by manipulation of the flow rate of cooling water to condenser.…”

“…Winchester and Marsh (1999) studied a single pass falling-film evaporator with mechanical vapor recompression and concluded that it was not controllable (as defined by Skogestad & Postlethwaite, 1996) thus it was not possible to achieve a high enough disturbance rejection bandwidth in the concentrate total solids loop. They further concluded that the disturbance rejection bandwidth would become more inadequate for multi-pass evaporators due to the increase in process time delay.…”

“…A three-effect falling-film evaporator at Isfahan Milk Company was used for this study. The dynamic model of three-effect falling-film evaporator was nonlinear and developed based on principles of model reported by Winchester and Marsh (1999). This paper reports on the design of inferential cascade controller based on total solid concentration estimation in the first effect by using Kalman filter.…”

Cascade controller design for concentration control in a falling-film evaporator

“…Winchester and Marsh (1999) studied a single pass falling-film evaporator with mechanical vapour recompression and concluded that it was not controllable (as defined by Skogestad & Postlethwaite (1996)) thus it was not possible to achieve a high enough disturbance rejection bandwidth in the concentrate total solids loop. They further concluded that the disturbance rejection bandwidth would become more inadequate for multipass evaporators due to the increase in process delay.…”

“…A two-effect falling-film evaporator at Fonterra Ingredients-Whareroa, Fonterra Coop Group Ltd. was used for this study. The dynamic models were developed based on the principles used by other researchers (Choudhary, Ali, & Das, 1996;Runyon, Rumsey, & McCarthy, 1991;Winchester & Marsh, 1999) with some improvements to existing models. This paper reports on the design of a cascade controller for this situation to improve the disturbance rejection bandwidth with the aim of rejecting disturbances up to the disturbance gain crossover frequency.…”