Parametric identification of integrated model of a coal-fired boiler in a thermal power plant

Abstract

Retrofit or replacement of few units in a subcritical facility may not only improve overall efficiency of conversion of energy in a power plant but also support sustainability issues. The primary objective of this article is to identify model parameters of a coal-fired integrated boiler and to present a comparative study on three different identification methods. This leads to select most suitable models that are applied for the developed model of the boiler of 210 MW coal-fired thermal power plants. The mathematical models of economizer, drum, and super-heater assembly are derived using mass balance and energy balance equations. The derived multi input–multi output model is then validated, and the model parameters are identified using three different identification methods namely nonlinear least square technique, maximum likelihood estimation, and expectation maximization algorithms. Identification of the plant model will essentially help to frame a good controller. In this article, parameter estimation has been carried out from real-time plant as it provides selective tool through quantitative comparative study of the three methods. The expectation maximization method has been found to provide suitable results compared to the other two methods. Model parameters of integrated boiler of a comprehensive structure have been obtained for the first time using expectation maximization method.