Quantcast

Energy and Utilities Virtual Conference

Proceedings

Plants Going Green: Intelligent Optimization for Power Plants

Houda Karaki, EUtech Scientific Engineering GmbH

Read abstract

Using model predictive control, the intelligent optimizer reduces emissions and cuts costs all while improving combustion efficiency. This session demonstrates the end-to-end optimization solution, developed with a model-based approach and, more importantly, presents the tangible results from a large-scale power plant. The results underline the environmental and financial gains that paid off the system within a year.

The overall performance and availability of a power plant is predominantly affected by the steam-generating unit and the combustion process. Even though conventional plant control systems ensure safe and reliable operation, they do not rigorously optimize boiler operations or take care of special combustion problems. In addition, much of the information gathered by modern IT systems and advanced monitoring equipment remains untapped.

The presentation features a complete, ready-to-integrate control system for a fossil-fired power plant. EUtech applies Model-Based Design and relies on widely available tools like Simulink and Model Predictive Control Toolbox® as well as our own Thermolib, which dedicated to modeling thermodynamic systems. While it certainly is not trivial, designing and building a fully fledged intelligent controller is no longer the Herculean task it once was.

The solution devised is modular, highly customizable, and robust. Most of all, however, it requires neither major overhead on the plant operators' part nor infrastructural changes. We show the significant improvements realized in a large-scale coal-fired power plant in Germany, covering financial savings, optimized combustion, and reduced NOx and CO emissions.