OCR Text |
Show flow rates of the fuel and the air stream. Increased volumetric flow rates led to higher velocities and better temperature uniformity . Burner A appeared to mix the furnace atmosphere better compared to Burner B, mainly because its design results in higher drive at equal firing rates. The flame from Burner B was less directional with a tendency to travel to the roof due to natural buoyancy. Therefore, the roof temperatures with Burner B were higher and the furnace temperature uniformity worse than those with Burner A under the same operating conditions. Gas Temperature Profiles - A suction pyrometer was used to measure gas temperatures at various locations inside the furnace. After statistical analysis, the data were reduced to isotherms in the central plane along the length of the flame. For reasons of clarity, only two isotherms are presented here. As shown in Figures 7 and 8, the effect of firing rate on the isotherms was much greater than the effect of oxygen enrichment. The temperature profiles obtained with Burner A are qualitatively similar to those obtained with Burner B. The data in Figures 7 and 8 indicate that low level oxygen enrichment can be used for production increase purposes without affecting the flame length or the flame shape. However, if higher enrichment levels are to be achieved, careful attention must be given to the changes in the flame length and the flame shape. NOX EMISSIONS - The measured NOx levels were corrected to O~ oxygen in the flue. The results are shown in Figure 9. For both the burners tested, NOx levels increased significantly with enrichment, and less significantly with the firing rate. Note that pure oxygen was premixed with air to achieve the desired enrichment level. NOx levels, when firing on air, were similar for both the burners, but Burner B exhibited substantially higher NOx emissions at high enrichments and at high firing rates . BURNER TEMPERATURES - One of the important considerations when employing oxygen enrichment is the physical survival of the burners. Burners A and B were instrumented to measure the temperature near the downstream face of the metallic air plates. The temperature levels measured with Burner A were all below 600°F . In contrast, the temperature levels measured with Burner B were much higher and rose sharply with oxygen enrichment. Based on metallurgical considerations, 2000°F is considered to be the limit for physical survival of metallic burner components. The results with Burner B (Figure 10) clearly indicate its nonsuitability for enrichment levels greater 168 than 9~ (30~ 02 in air). Visual inspection indicated some high temperature oxidative damage to the air plate. BURNER PRESSURES - Figure 11 shows the air pressure at the burner for both the burners. It is clear from the data that oxygen enrichment lowered the air pressures at the burner. CONCLUSIONS The data collected during performance testing of off-the-shelf air-fuel burners show the following effects of 0 to 14~ oxygen enrichment (21~ to 35~ 02 in air): Increase in furnace temperature with no changes in fuel input. Reduced furnace temperature uniformity with no changes in fuel input. Significant changes in the flame characteristics . Increase in NOx emissions. Possible failure of burner components. The specific effects depend on the burner design and the level of enrichment. Based on the results shown in this study, the maximum allowable enrichment level for a given air-fuel burner in a given furnace depends on the expected changes in flame characteristics, furnace temperature distribution, NOx emissions, and burner temperature. In addition, the effect of lower burner pressures on the operation of the burner/furnace control system needs to be evaluated. FUTURE PLANS Prototype natural gas burner systems have been developed that address all the potential problems of high oxygen enrichment identified during the burner testing described here. The performance testing of the prototypes is currently in progress at Air Products and Chemicals, Inc . Field demonstrations will follow performance testing. Acknowledgements I wish to acknowledge the key contributions made by D. C. Winchester, C. A. Ward, F. A. Milcetich and others in collecting the data presented here. I am grateful to Prof . H. C. Hottel for his valuable comments and suggestions . I also wish to thank Air Products and Chemicals and the Gas Research Institute for their generous support and for the permission to present this work. |