Experimental and numerical investigation on the influence of temporal fuel/air unmixedness on NOx emissions of lean premixed catalytically stabilized and non-catalytic combustion

The influence of temporal fuel/air unmixedness on NOx emissions in lean premixed catalytically stabilized and non-catalytic combustion of propane is experimentally and numerically investigated. Experiments are performed in a tubular, adiabatic flow reactor with catalytically active and inactive honeycomb structures inserted at the inlet to the combustion chamber. Catalytically stabilized combustion is performed as "hybrid" catalytic combustion, where only a fraction of the fuel is consumed within the catalyst and the remaining fuel is burned downstream in a homogeneous combustion zone. A special mixing section is used which allows variation of frequency and magnitude of temporal fuel fluctuations while maintaining spatially uniform conditions over the cross-section. Numerical investigations are performed by coupling a transient catalyst model with a PFR model for hybrid catalytic combustion and a PFR model for non-catalytic combustion. The catalyst model comprises a 2D model for the gas phase and a 1D model for the catalyst substrate, including one-step surface kinetics for propane oxidation. The PFR model uses a detailed reaction mechanism to model NOx formation in the homogeneous combustion zone. It is shown experimentally that NOx emissions from catalytically stabilized combustion are much less sensitive to temporal fuel/air unmixedness than non-catalytic combustion. The catalyst model reveals (1) that the catalyst lowers the mean fuel concentration and the magnitude of the fluctuations prior to the homogeneous combustion zone downstream of the catalyst and (2) that the catalyst's thermal inertia acts as a buffer for the fluctuating heat release on the surface, i.e., it prevents the development of large temperature fluctuations in the gas phase. Comparison of modeled NOx emissions with experimental data shows good agreement for non-catalytic combustion, but poor agreement for hybrid combustion. These findings are discussed in detail and improvements are suggested to improve agreement between experiment and prediction for the hybrid combustion case. (C) 2000 Elsevier Science B.V. All rights reserved.