Advanced process control for power plants

Abstract

The European power market has changed substantially over the past twenty years. Increased competition, increasing use of renewables and environmental regulations contributed to a power market in which the power supply and demand fluctuates more frequently than ever. By nature, most power plants are sensitive to these fluctuations. The main reason for this is coupling between throughput (i.e. power output) and quality (e.g. pressure, temperature) and coupling amongst the quality parameters itself. As a consequence, when throughput fluctuates the quality is disturbed and oscillations occur in many of the process loops. Since these oscillations reduce power plant efficiency, increase wear and reduce plant lifetime it is desired to investigate solutions. In this study the effect of multivariable advanced process control (MIMO APC) on internal process loop oscillation and performance is investigated. Furthermore, the effect of improved process loop control on overall power plant performance is studied. To this end, a conventional drum-boiler power plant simulator (PID-controlled) is equipped with an Hinf-controller. A safe shutdown master-slave control structure is proposed to enable practical implementation. To accentuate the influence of improved internal process loop control on overall performance, the Hinf-controller will be responsible for four boiler process loops (the boiler has a central role in determining the overall behavior of power plants). This study shows that MIMO APC is a viable method for reducing oscillations and improving performance of internal process loops. It has become clear that better internal control also improves overall power plant performance. As a result, the effective maximum power production rate can be increased, which was found to have a surprisingly small effect on the achieved performance increase. These results indicate that it is possible to increase the economic potential of power plants using MIMO APC.