Controller tuning constants for the Advanced Power Reactor 1400 Feedwater Control System based on M_s-Constrained Integral Gain Optimization Method

Highlights: • Obtaining data sets from the thermal-hydraulic model by SPACE. • Designing the PI controller of the FWCS by Ziegler-Nichols and MIGO methods. • The MIGO approach is selected as the primary approach to design the feedwater PI controller. • Applying the PI MIGO controller to the overall schematic diagram. • The simulations prove the improvement of the FWCS when using the PI MIGO controller. - Abstract: Failure to control the steam generator (SG) water levels causes unexpected shutdowns in the Advanced Power Reactor (APR) 1400 nuclear power plants. This impacts maintenance, operational life, and safety critically. Therefore, it is vital to stabilize the SG water level at different power levels, particularly at low powers, by improving the performance of the proportional integral (PI) controller in the Feedwater Control System (FWCS). The PI controller designed using M_s-Constrained Integral Gain Optimization (MIGO) mitigated this problem significantly. In order to design this PI controller for the FWCS, the SG model was developed as a process model with a system identification algorithm in MATLAB. The algorithm processed data sets obtained from the thermal-hydraulic model by the Safety and Performance Analysis Code (SPACE). The MIGO and the Ziegler-Nichols methods were each used to produce a design for the PI controller, and the controllers were compared. Simulations were run in MATLAB for resulting designs. The PI controller designed by the MIGO is more stable than the controller using the Ziegler-Nichols approach. Finally, the PI MIGO controller was applied to an overall schematic block diagram of the FWCS. The simulations of the diagram in MATLAB Simulink demonstrate the PI MIGO controller offers superior control and enhanced robustness in the FWCS.