| OCR Text |
Show To explore methods of controlling N O x output from glass furnaces of the unit melter type featuring recuperative combustion air preheating without retrofitting flue gas treatment plants, Gaswarme- Institut (GWI) and Thyssengas G m b H decided to carry out a number of tests [5]. GWI Test Facility The test furnace depicted in Fig. 1 was used for the GWI/Thyssengas test programme. The furnace chamber is designed for a njaximum wall temperature of 1,600 °C. The chamber consists of sections to vary the furnace length between 2 and 5 m If the total length of the furnace is used, burners designed for a total njjaximum heat input of 1.5 M W can be tested. It is therefore feasible to examine burners of the size used for glass tanks featuring recuperative combustion air preheating. In the tests carried out for the GWI/Thyssengas project, the furnace length was 4 m. As heat transfer from the flame to the molten glass is critical in glass furnaces, total heat transfer total radiation were measured in addition to pollutant concentrations and furnace temperatures. A n ellipsoidal radiometer was used to determine total radiation, while heat transfer by radiation and convection was measured by a total heat flux meter. These measurements at different locations along the furnace centerline in the burner level were carried out to examine the influence of the different N O x control techniques on heat transfer from the flame to the molten glass. Various gas burners designed for heat inputs of 520 k W weTe tested. A n electric air heater was installed to preheat the combustion air. In the tests, the combustion air temperature was 690 °C. The resulting furnace surface loading was 167 k W / m 2 which is typical of glass tanks. The furnace chamber loading was 140 kW/m3. NOx Emissions and Thermal Properties of Different Burners Standard Burner To examine the influence of the mean furnace wall temperature on NOx emissions, a burner (burner A) normally used in recuperative glass furnaces was tested. Fig. 2 is a schematic of the burner. Tlie air ratio for the tests was 1.05. NOx measurements confirmed an overproportionate rise in NOx concentration as the furnace wall temperature increased. The N O x level peaked at 2.250 mg/m31 at a temperature of 1,600 °C (see Fig. 8). Glass furnace field measurements for burners of this type showed N O x concentrations of between 1,200 and 1,500 mg/m2. One reason w h y the test furnace N O x concentrations were higher is the compact design which prevents adequate flue gas recirculation in the furnace chamber. Secondly, the high N O x level was due to the absence of the molten glass which would be at a lower temperature than the furnace floor. A s the flame in the test furnace is completely surrounded by hot furnace walls, the flame temperature is higher than in a normal glass furnace. All concentrations given as NQ2 (8 % O2, dry) 3 |
