Implementation of a multivariable modelbased predictive controller for superheater steam temperature and pressure control on a large coal-fired power plant

Large temperature excursions are known to be a major contributor to plant life reduction through increased stress and creep life damage on boilers. The performance of conventional (PID) controllers in minimising the impact of these excursions is limited, especially for plants that are required to operate flexibly over their full load range. Significant performance improvements have been achieved through the application of model-based predictive control (MBPC) technology. This paper summarises the development and implementation history, and results achieved from using a multivariable MBPC-based controller for steam temperature and pressure control on a large coal-fired boiler.