| OCR Text |
Show AFRC - 1996 Int. Symposium Baltimore September 30 - October 2 1996 Page 4 of 14 The mixing performance of the venturi has been examined by varying the location of gas injection upstream of the venturi throat, including asymmetric injection. There was no effect of this location on the NOx emission of the flame in the flame tube. This indicated that the gas, air and flue-gas downstream of the venturi mixer are well mixed. 3. EXPERIMENTS The amounts of flue-gas, air and natural gas in the gas mixtures under investigation were varied by adjusting the thermal input of the burner in the precombustor, the amount of dilution air and the amount of natural gas injected upstream of the venturi throat. The temperature of these mixtures could be varied by changes in the thermal insulation of the system and/or by preheating of the dilution air. The initial NOx concentration in the unburnt mixture could be varied by operating the premix burner in the precombustor at an air factor (A) between 1.0 (high NOx) and 1.3 (low NOx) , and subsequently altering the amount of dilution air used. The following parameters in the mixture have been varied: - oxygen concentration in the unburnt mixture - air factor (A = air available/ air required) - mixture temperature (at flame holder) - mixture velocity (at flame holder) 12 - + 19 1.1 - max 20 - 500 0.7 - 2.0 % °C m/s Note: an oxygen concentration of + 19 % m the mixture corresponds with a It standard It natural gas/air mixture without flue-gas. The concentrations of oxygen, CO and NOx (determined by paramagnetic, non-dispersive infrared, and chemiluminescent analyzers, respectively) were measured at two locations: just upstream of the flame holder and downstream of the flame tube. The mixture temperature is measured with a thermocouple located at the outlet of the venturi mixer, at approximately 1 cm upstream of the flame holder. Due to insulation, the walls of the venturi mixer have approximately the same temperature as the gas mixture and therefore the temperature measured by the thermocouple is not affected significantly by the presence of hot or cold surfaces in its neighbourhood. The natural gas and air flows, as supplied to the system, are measured by mass flow meters. In the experiments, dry combustion air is used. The natural gas used is Groningen gas with a LeV of 31.7 MJ/m3(n) and an air requirement of 8.43 m3(n) (dry) air per m3(n) of gas. The stoichiometric combustion of 1 m3(n) of Groningen gas with dry combustion air generates 9.39 m3(n) of flue-gas (including 1.68 m\n) water vapour). The laminar flame velocity of Groningen gas is very similar to the laminar flame velocity of Methane. For more detailed information about Groningen Gas, see Geerssen (1988). |
