(340r) Advanced Control Strategies for Improved Subcritical Power Plant Cycling

Recently, the number of power plants required to operate in load-following condition has been growing due to the penetration of renewables into the electric grid system. One of the characteristics of renewables, such as solar and wind, is that they are intermittent and cannot deliver a constant power amount to the grid. Additionally, power demand varies significantly during the course of a day. The combined effects of varying demand and renewable power generation require fossil-fuel-based power plants to cycle their load. However, few coal-fired power plants were originally designed to operate under load-following conditions, going from minimum to full load conditions.

Given that cycling is the new normal operating mode for coal-fired power plants, it is critical to properly control the plant to avoid damaging of the system. To operate in different power dispatching modes, a power plant will need to produce steam at certain flows, temperatures and pressures as required by the turbine. The highly variable plant dynamics will require predictive control algorithms capable of maintaining the process in the operating regions of interest. The objective in this work is to evaluate the performance of different advanced control strategies for a subcritical coal-fired power plant under cycling conditions.

Specifically, a subcritical coal-fired power plant model was employed to evaluate classical and advanced control strategies under typical or off-design operating conditions. The subcritical coal-fired power plant model was developed in MATLAB/Simulink and included the boiler, steam attemperator and turbine subsystems, in which advanced Model Predictive Control (MPC) strategies were applied.

The analyzed control strategies included a combination of classical PID and advanced model predictive control (MPC) structures, such as the Dynamic Matrix Control (DMC) and the novel modified-SQP-based approaches [1]. The modified SQP algorithm was built upon a backtracking line search framework employing a group of relaxed step acceptance conditions. This novel MPC approach is reported to be effective for addressing power plant cycling of a supercritical coal-fired power plant. This algorithm was extended to subcritical power plants to address off-design instabilities that are unique to this predictive control application. Results of the implementation of these control strategies will be discussed in terms of assessing their behavior and computational efficiency at specified power plant operating regions associated with load following.

Reference

[1] X. He and F. V. Lima. A modified SQP-based model predictive control algorithm: application to supercritical coal-fired power plant cycling. Submitted for publication, 2020.