Optimization of operating parameters of MSF plants through automatic setpoint control

Abstract

Controllers in an MSF process are usually designed based on simplified process models, and the focus is on obtaining reasonable dynamic responses of process outputs for setpoint and load changes. The setpoints are normally fixed by the plant operator based on previous operating experience and perceived ideas about the operating condition of the plant. This manual setpoint adjustment can never be expected to result in an optimum running condition of the plant because of the complex processes taking place inside an MSF evaporator. Operating the plant at its optimal point means substantial savings in operating costs as well as increased safety and security of the plant. This can be achieved by on-line setpoint control. This paper describes the architecture of a typical on-line setpoint controller and the main components of the setpoint optimization system. The mathematical model used for process steady-state simulation is explained, and the economic model for cost optimization is outlined. To demonstrate how the mathematical and economic models can be used to optimize the operating of a MSF plant, the MSF plant at UANE (Ext.) was selected. Typical design and cost parameters are used for this case study. It was shown that running the plant at operating conditions different from the optimal results in a substantial economic penalty. The economic penalty for not operating the plants at the optimum T max, for example, at a production level of 500 kg/s, the minimum operating cost was estimated at Dh 1.766 per ton with the optimum T max = 109.6°C. Operating the plant at 100°C instead of the optimum T max results in an increase of 6.9% in operation costs, thus incurring additional operating expenses of Dh 1.89 million per year.