Obstacles in modelling formation of nitrogen oxides in combustion systems

Obstacles in modelling formation of nitrogen oxides in combustion systems

Paper in proceedings (conference paper)

The present work deals with the possibilities and obstacles of modelling the formation of nitrogen oxides. The approaches discussed include both empirical regression modelling and computational fluid dynamics (CFD). Basis for the discussion is provided by a parametric study of a low-NOx gas burner with staged gas supply and 1000 kW firing capacity. The measured data were obtained in a new, advanced burner testing facility, described in the text. Statistical analysis is applied to choose independent variables influencing the formation of NOx among numerous burner design parameters. Structure of the resulting regression model is discussed in detail. Assessment of similar models for practical use closes the first part of the article. The approach based on CFD modelling is analysed with regard to several aspects, namely tuning of computational models using measured data, influence of simplified modelling assumptions, solution accuracy and computational time. The potential to reduce computational time by certain changes of the burner geometry while improving flame stability is discussed using information gathered from both the experimental and numerical work.

The present work deals with the possibilities and obstacles of modelling the formation of nitrogen oxides. The approaches discussed include both empirical regression modelling and computational fluid dynamics (CFD). Basis for the discussion is provided by a parametric study of a low-NOx gas burner with staged gas supply and 1000 kW firing capacity. The measured data were obtained in a new, advanced burner testing facility, described in the text. Statistical analysis is applied to choose independent variables influencing the formation of NOx among numerous burner design parameters. Structure of the resulting regression model is discussed in detail. Assessment of similar models for practical use closes the first part of the article. The approach based on CFD modelling is analysed with regard to several aspects, namely tuning of computational models using measured data, influence of simplified modelling assumptions, solution accuracy and computational time. The potential to reduce computational time by certain changes of the burner geometry while improving flame stability is discussed using information gathered from both the experimental and numerical work.

Modelling, design of experiment, CFD, combustion

Modelling, design of experiment, CFD, combustion

BĚLOHRADSKÝ, P.; KERMES, V.; STEHLÍK, P.

2010

24.08.2008

Orgit s.r.o.

Praha

978-80-02-02047-9

Conference CD - CHISA 2008

Conference CD

1

7

7

http://hdl.handle.net/