On-line Optimization of Drum Boiler Startup
Mads Rode, Klaus Kr · 2003
On-line optimization of industrial processes is increasingly important to minimize cost and environmental impact of a plant during its operation. The new Modelica.Media and Modelica.Fluid base libraries allow the dynamic modeling of process systems [4]. Their application to dynamic optimization is discussed on a tutorial example. The optimization is based on a simple non-linear drum boiler model from the literature [1]. The model is implemented in Modelica using the new Modelica.Media and Modelica.Fluid base libraries. The model exhibits three control inputs: feed water flow rate, heat input, and position of a valve at the steam outlet. A PI control is embedded into the model for the feed water flow. The remaining two control inputs are optimized. Optimization results are compared with a straightforward control strategy. On-line optimization based on a sophisticated boiler model is currently being applied to a 700 MW coal fired power plant.