Nonlinear Dynamic Modeling and Parameter Identification of Power Boiler: A Case Study

The 320 MW boiler of Bandar Abbas power plant has been subjected to parametric variations due to its long lifespan and numerous renovations, so a relatively accurate dynamic model is needed to retune the characteristics of its control system, simulate events, and evaluate and optimize its performance. Due to the lack of standard models for boilers with different structures, this paper deals with the modeling of this forced circulation subcritical boiler. As a result, a ninth order multivariable nonlinear state space model is developed using the physical modeling method. Due to the limitation of the measured variables and the lack of sufficient data with dynamic specifications suitable for identification algorithms, a computational procedure that uses only steady state measurements of the process is introduced to determine the unknown parameters of the model. The resulting model presents reasonable step responses and its ability in predicting the boiler outputs is confirmed using the operational data of the power plant during a sudden gas fuel pressure reduction event. Finally, the accuracy of its parameters is evaluated by performing sensitivity analysis.