Advanced Combustion Facilities and Diagnostics

Page 3

maximum heat input. A water cooled load is provided by cooling coils placed on the furnace floor. This furnace is lined with a low density fibrous refractory with a maximum operating temperature of 1650°C (3000°F). Heating the furnace from ambient conditions to thermal equilibrium at 1650°C typically takes 4 to 6 hours. The furnace is instrumented with 8 thermocouples, 5 heat flux meters, continuous flue gas analysis (CO. CO;. NOx . 0:. SOx) and flue gas temperature monitoring. Furnace control and data collection are conducted through a Siemens P L C system. With this instrumentation, heat balance and heat transfer efficiency from the flame is obtained in real-time. Air infiltration is avoided by sealing all openings and operating the furnace at a positive pressure. To conduct optical and probe measurements on the furnace, a series of opposed ports are located along the furnace centerline. Additional access through the crown is also available near the burner and before the flue duct. In particular, the furnace roof is equipped with a water-cooled periscope that gives a complete v iew of the flame. For optical measurements, it is generally better to move the flame rather than the optics because of alignment and reproducibility difficulties. With this in mind, a translation stage for the burner was implemented that prov ides + 25 c m vertical translation. The second furnace at the Countryside research center has a vertically fired chamber with internal dimensions of 0.3*0.3*1.0 m (see Figure 4). and a maximum heat input of 30 k W (0.1 MMBtu/hr). This newly implemented furnace is used for developing optical diagnostic techniques, studying fundamental combustion, and testing small-scale concepts. Heat extraction through the furnace walls can be varied by using either water cooled panels or a fiber board insulation material. Each wall has a 30*10 c m vertical opening, thereby allowing simultaneous measurements to be conducted. Guide rods and a hydraulic piston enable vertical translation of the whole furnace. This design allows all surrounding optical components to be stationary while obtaining spatial flame measurements. The furnace instrumentation includes thermocouples mounted in the walls, fuel and oxidizer flow control, and wall heat flux monitoring from the cooling supply inlet/outlet conditions. IN-FLAME MEASUREMENT TECHNIQUES These techniques are currently used in our pilot furnaces to determine oxy-flame characteristics. Sections 3.1 and 3.2 present diagnostics that help characterize flame geometry, composition, and temperature. The detailed in-flame measurements are also used for modeling validation. Sections 3.3 and 3.4 are concerned with soot-related diagnostics that help address the role of soot in heat transfer from oxy -natural gas flames. Gas sampling probe In-flame gas composition measurements are performed with a high quenching rate, miniaturized sampling probe. The front-end part of this water-cooled probe is reduced to an outer diameter of 20 m m and features a 90° elbow bringing the probe tip parallel to the main flow direction. This sampling method with a suction hole facing the streamlines is recommended to minimize flow disturbance [5-6]. Application in high temperature oxy-flames with high level of dissociated species brings special requirements on reaction quenching. Since recombination of free radicals like H, O. O H cannot be avoided, the best compromise is to preserve most of the stable molecular species while promoting radical termination on the probe walls [6-8]. The quenching section consists of a stainless steel diverging conical chamber with an inlet orifice of 2.5 m m in diameter. The absolute pressure in the suction line is maintained below 100 mbar to promote a choked flow, fast temperature and pressure drop, and effective removal of radicals at the wall. The probe aerodynamic quenching is estimated to reach a cooling rate or 108 K/s or higher [9J. Confirmation of efficient quenching was obtained by experimental results, which showed, for high temperature oxv-flames, peak C O and H : concentrations above 25 and 2 0 % (dry), respectively. The N O / N O x ratio, typically above 9 0 % . is also indicative of efficient radical quenching [7]. Since modern industrial flames are by design not necessarily axi-symmetric, the sampling probe can be traversed both in vertical and horizontal planes. Figure 5 displays C O concentration contours in a vertical plane crossing the burner axis for the flat flame A L G L A S S F C ™ burner operating at 1.0 M W in the 1.5*2.0*6.0 m furnace. A detailed description of this burner is presented by Legiret et al. [10]. Because the objective was to investigate the interaction between the flame and the heat extraction load, measurements were performed between the burner axis and the furnace floor at six axial positions and five to ten vertical positions per traverse. The C O contours reveal a broad reaction zone with low vertical gradients, a line of m a x i m u m C O concentrations below the furnace axis, and flame contours (based on 5000 p p m level) 20 c m abov e the furnace floor.