OCR Text |
Show focused mainly on the jet-pump capacity, as well as the split between primary and secondary air as it is determined by the position of the fuel gun. The gun position controls the exit area of the primary air nozzle, thus controlling both the pressure drop in this air nozzle as well as the momentum of the primary air jet, the driving force for the jet pump. The optimized geometry was then to be SUbjected to a spatially detailed flame characterization, to facilitate the development of a scaled-up version (6 MW or 20 MM Btu/hr, 1994-1995), and to identify regions of NOx formation which can be further reduced. 2.2 BERLfacility Although the Burner Engineering Research Laboratory (BERL) located at Sandia National Laboratory Livermore, has been exhaustively described elsewhere [2], a brief description of its main features, as well as of some of the modifications made specifically for this program, will be given here. The BERL test facility comprises a 0.9x2.1 m (3'x7') vertical furnace with an octagonal cross section. The walls of the furnace consists of 0.3 m (1') spool sections which can be fitted with interchangeable panels, which may be either refractory lined or bare metal to provide control over the temperature-time history of the combustion process (Figure 4). Special panels can be mounted to provide full access for optical and physical flame probing. The combustion air can be preheated to 800 K (1000 OF) by means of an electric air preheater. For the recirculation of hot flue-gas from the top of the combustion chamber to the burner, a stainless steel externally insulated duct was provided for this series of experiments·. The temperature of the flue-gas could be controlled to be between 450 and 1100 K by mixing it with cold flue-gas, which can be diverted from the main exhaust gas stream and cooled ·by means of a heat. exchanger on the roof. The flue gas recirculation rate was controlled by dampers in the FGR lines. Various optical and physical sampling and analysis techniques were used to characterize the flames. For the physical sampling, a sample was drawn from the flame or the· flue gas through a stainless steel, water-cooled probe. After exiting the probe, the sample was led through a heated sample line and a pump (all of which have teflon "wetted" . parts), to the continuous emissions analyzer system. In this system, conventional flue gas analyzers, including NDIR CO and C02, parmagnetic 02, and chemiluminescent NOx analyzers, were used to characterize concentrations of major species in the flame. In addition, a flame ionization detector was used for the measurement of total hydrocarbon concentrations. The sample stream was dried by means of a chiller for all continuous analyzers, except for the flame ionization detector. The optical techniques used during . these experiments include Laser Doppler Velocimetry (LDV) for velocities, and chemiluminescence and Laser Induced Fluorescence (LIF) for CH radical concentration imaging. A description of LDV and LIF can be found in [3,4]. |