| Summary: | The operation of dual purpose power station and desalination plant requires careful planning of operating strategies, to ensure optimum plant operation with variable product demand and plant availability. In order to provide practical tools for use by plant operations management, two methods of optimisation are employed - mixed linear/integer programming and simulation. Using computer based techniques, a static model of the plant is constructed in linear programming format. The plant is defined in the form of relationships, or constraints, for mass and fuel balances, and limits of operation. The objective function is the fuel consumption of the plant, for any given operating time. A dynamic model of the plant is then developed with a time cycle of 1 hour, using the static model as a basis. Incorporated in the model are reliability characteristics of major plant components, unit start-up and plant response simulation. Using the dynamic model, possible operating strategies are simulated over a 12 month period and an optimum strategy is developed for continuous and economic operation. The static and dynamic models were originally developed on a mainframe computer in Fortran. With the advent of low cost microcomputer systems, a need was foreseen to apply the optimisation procedures to a microcomputer. The application of a microcomputer enables the procedures to be available to plant management, when access to a mainframe computer does not exist. Despite some initial difficulties, a number of benefits result from the use of the microcomputer. A simulation graphics model is developed, demonstrating the use of the model as an aid to training. Further developments include ambient temperature and related performance simulation, with a plant expansion study to demonstrate the model as a plant development tool.
|
|---|
