A laboratory-scale single stage flash desalination process plant: Design, modeling, and automation using an intelligent controller

A laboratory-scale single-stage flash desalination plant was designed, modeled, built, and automated using a fuzzy logic controller. Various aspects of this experimental test-bed allow it to be computer controlled, and thus it will allow testing of other modern control architectures and approaches.; The laboratory-scale plant has been designed from basic principles of flash desalination process. The only major difference between the laboratory-scale plant and a commercial plant is the brine heater. An electric water heater has been used instead of a heat exchanger that uses extra steam from a power plant to supply thermal energy to incoming seawater. The plant design was done in such a way that most of the electrical and mechanical parts in the plant have been bought off the shelf.; Existing flash-based desalination plants are generally very large scale; development of control systems, therefore, have been mostly limited to simulations. These plants, also called Multi-Stage Flash (MSF) desalination plants, conventionally operate with old control schemes and they are not necessarily robust. Classical controls require an explicit mathematical model of how control outputs depend on control inputs. Fuzzy controllers differ from classical model-based controllers. Fuzzy controllers do not necessarily require a mathematical model of how control output functionally depends on controller input. The uncertainties and nonlinearities inherent in a large-scale real time system such as an MSF plant make it an ideal candidate for model-free control schemes such as fuzzy logic control.; This research and development effort derived the following contributions to the state-of-the-art in desalination and process control: (i) a novel approach for the intelligent process control, (ii) the design and implementation of the flash desalination process plant and the control system, (iii) a fuzzy logic controller (FLC) for such a process plant is also the first, (iv) a mathematical model for the most critical component of the desalination process plant; this has made possible a comparison of performances derived from a fuzzy logic controller with that of a classical PID controller, (v) an algorithm for fuzzy logic controller design for flash desalination process plant, and (vi) a user friendly software interface for controlling the plant. (Abstract shortened by UMI.)...