Model based NOx emission monitoring in natural gas fired hot water boilers

Due to new emission regulations, coverage of monitoring requirements extends to new boiler classes. In Europe, Industrial Emission Directive brings new 15-50 MW boilers under the directive in special cases, so inexpensive NOx emission monitoring tools are appreciated. This paper presents models to estimate NOx emission in two similar natural gas fired municipal hot water boilers. The models utilize online process measurements also available in relatively simple set-ups. In the context, the performance of linear regression model and nonlinear neural network model are introduced. Despite the similarity, the results state that the examined boilers behave differently in terms of linearity and emission levels. Therefore, every boiler should be identified separately to exploit the simplest models in practical installations.