XML Print


Abstract:   (44 Views)
The boiler drum process is a nonlinear, complex and multivariable process which includes significant time delay. Therefore, the control on the water level in the drum is not easy and ideal. The first objective of this paper is to model the drum water level referring to 210 MW power unit for Nassiriyah thermal power plant. The second objective is to study the water level controller operation with its performance investigation. Firstly, the drum water level process has been modelled based on first principles by two models: the proposed simplified linearised model and the complicated nonlinear model. Then, a comparison between the extracted practical plant data and the water level results simulated by the two models demonstrate the validity of both models with very good approximations. Secondly, Proportional Integral (PI) controller based on three element water level control strategy and used in this plant, has been described and simulated by Matlab/Simulink. The controller parameters have been selected according to practical considerations. These considerations are minimizing as possible, a number of the close and open commands to the feedwater flow control valve to extend its lifetime with maintaining the drum water level on a set point. The controller has been tested to evaluate its performance for different values of proportional gain (Kp), integral gain (Ti), gain of steam flow signal (Gx2), and gain of mass feedwater flow signal (Gx3). Firstly, the results show that selection of Kp is difficult because of the tradeoff between fast dynamic response and steady state performance. Secondly, the results show selection of Ti affects only steady state performance. Finally, the results show that selection of Gx2 and Gx3 plays an important role in stability of the drum water level.
Full-Text [PDF 2337 kb]   (13 Downloads)    
Type of Study: Research Paper | Subject: Systems Modeling and Simulation
Received: 2018/03/23 | Accepted: 2018/11/26 | Published: 2018/11/26

Creative Commons License
© 2018 by the authors. Licensee IUST, Tehran, Iran. This is an open access journal distributed under the terms and conditions of the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) license.