OCR Text |
Show EXPERIMENTAL MEASUREMENTS The liquid-fuel burner described in the previous section was tested with gas oil under various conditions to examine CO and NOx emissions as well as heat-transfer characteristics. All experiments were conducted at stoichometric oxygen/fuel ratios. Atomisers were positioned at 10, 15,25,30 and 35 mm from the tip of the burner. In-flame measurements were made using a 1.8m by 0.9m by 0.6m, refractory lined, rapid-heating laboratory test furnace. The heat transferred to a thermal load has been simulated with a sectional water-cooled base. Six gassample ports were located along the length of the furnace, (at an equal spacing of 300 mm) enabling on-line, rapid and continuous analysis of combustion products using a water-cooled quartz-lined probe. CO and C02 concentrations were measured using non-dispersive infra-red analysers, and NO and N02 concentrations were measured using a chemiluminescence detector. Pt-PtlI3% Rh, 25~m tip thermocouples monitored wall and flame temperatures, and Ktype thermocouples measured water inlet and outlet temperatures of the base. Flame lengths were measured photographically. Exhaust compositions were measured to investigate the effect of nozzle positioning on pollutants formation and heat transfer to the load. Exhaust gas samples were extracted through an "X" shaped gas-sampling probe in order to obtain a mean sample of the exhaust gases. Evaluation of the burners in an industrial production environment was accomplished on a 27 m2 surface-area recuperative unit-melter for borosilicate glass. The furnace was equipped with six bubblers which were located approximately three meters from the throat. The furnace was fired with 22 burners utilising preheated combustion air. UltimatelY,3 of the 22 burners were replaced by oxy-fuel burners during our tests. Overall temperature control was accomplished with a single-loop temperature controller adjusting the oil flow to maintain a target throat temperature of 1415 °C ± 2°C. Firing rates of the individual burners were manually adjusted so that burners near the center of the furnace had the highest firing rate while burners near the throat had the lowest firing rate. The burners on the batch end were fired at a low rate to avoid excessive dust carry over. Batch material typically covered 25% of the melt surface with thick foam extending further along the walls. The entire center portion of the glass melt was covered with "flat foam". Clear glass was observed only in a small area surrounding the bubbler line. RESULTS AND DISCUSSION Bench-Scale Studies. Injector positioning ,which alters the spray-size distribution, was used as variable. Fig. 4a shows the droplet-size distribution for a 50° angle nozzle at a distance of 1 cm from the burner head. Without air flow, the results are represented as histograms of volume of liquid present as droplets in a certain size band and as cumulative size distribution curves. These results indicate that the distribution of droplet is fairly broad, ranging from 10 to 300 ~m. The SMD (032) is 59 ~m. The effect of air flow on top and below the fuel spray on size distribution is shown in Fig. 4b It is interesting to note that droplet size changed and measured SMD values are reduced to 56 ~m compared to the case without air. Increase in air velocity did not produce a significant change in the central peak of the distribution. Fig. 4c shows the SMD (032) for 4 different fuel nozzle locations within the burner body at a distance of 10 - 35 mm from the burner tip. Results indicated that a more uniform droplet-size distribution can be generated with a nozzle at location "20" and "30" (mm from the front burner face). Pollutant Formation: The results suggest that burner performance was affected by spray characteristics. The wide drop-size distribution was seen to be narrowed significantly by the presence of the sandwich oxygen jets. These sizedistributions were used for the CFO modelling presented here. Fig. 5a shows the effect of fuel jet-atomiser positioning on measured exhaust NOx concentration. When the atomiser was placed at position 25, the NOx concentration reduced significantly. Visual observation also indicated that the flame became more stable and heat |