Modeling and Optimization of a Steam System in a Chemical Plant Containing Multiple Direct Drive Steam Turbines

Abstract

Steam systems in some of China’s chemical plants usually contain multiple direct-drive steam turbines that provide mechanical power to pumps/compressors. When optimizing this system, a certain degree of deviation is found in the theoretical models of steam turbines. A more realistic steam turbine model is developed by improving the traditional thermal model using industry data. This model characterizes efficiency variations under different conditions. Boiler and other unit models are then simplified to allow the use of this model in optimization. By incorporating the models, a mixed-integer nonlinear programming (MINLP) model is formulated to perform the operation optimization. The proposed model considers electric power as the alternative energy source for lower-level mechanical power demands. Using the proposed optimization model on an ethylene plant, a maximum of 8.01% reduction in the total operation cost is achieved compared with the original operation strategies. This case study shows a successful application of the MINLP model in optimizing an actual chemical plant.