Flue gas recirculation is one of the most effective ways to reduce nitric oxides (NOx) emission in conventional industrial furnaces. To design an effective control scheme, one has to understand the dynamic relationships among different furnace inputs and outputs. This paper concentrates on the construction of such dynamic models for an industrial furnace using numerical simulations and frequency domain system identification techniques. The numerical simulations are based on the conservation equations of mass, momentum, and energy. The inputs to the furnace consist of the pressure head of the flue gas recirculation fan, the temperature of the combustion air, and the flow rate of the combustion air. The outputs considered herein are NOx and oxygen (O2) concentrations. To obtain a dynamic model for this multi-input and multi-output system, low amplitude sinusoidal signals of different frequencies are administrated at the furnace input. The dynamic relationships among the inputs and outputs at these frequencies are established in terms of frequency responses (magnitude and phase) around a particular furnace operating point. These frequency responses are further processed by a least squares based system identification technique to convert them to a set of parametric models. The result of the system identification is a set of the transfer functions with the order ranging from 3rd to 6th. Studies have been carried out to verify the validity of these dynamic models by comparing the responses generated from these models with those obtained from the full-scale numerical solution. These dynamic models provide a starting point for the design of realtime optimal feedback control systems for minimizing NOx emission.

You do not currently have access to this content.