OCR Text |
Show AFRC - 1996 Int. Symposium Baltimore September 30 - October 2 1996 Page 9 of 14 Figure 4 gives also some indication of the measurement accuracy. The data at 14 % oxygen and a temperature of 395°C have been collected during three separate series of experiments. The data show a standard deviation of about 1.4 g/GJ. This relatively high inaccuracy is mainly due to the fact that the NOx production is calculated from the difference between the measured NOx concentration upstream and downstream of the flame tube. In this case, the NOx production by the precombustor is high compared to the NOx production in the flame tube; therefore, this NOx production is calculated by subtracting two relatively large numbers. Figure 5 presents the NOx production due to combustion as a function of the overall air factor A at four different initial "wet" oxygen levels: 12%, 14%, 16% and, approximately, 19%. The latter case represents a situation with a 'standard' natural gas/air mixture at ambient temperature without flue-gases. 120 100 )( - a C3 ---0"1 80 c::: .U2 60 g :l "'0 e 0 )( 0 0.. 40 >< 0 z 20 0 a ~ • °0 )( u ~ • )( 0 • • ..I/ a • 04 • 0 Q • • • o. 0 00 aa Q n .. _ 1. 1 1.2 1. 3 1. 4 1.5 1.6 1.7 1.8 Overal I air factor (-) )( 19~ 02 ; 20' C a16~02; 100'C .14~02; 370 'C o 14~ 02 ; 500'C .12~ 02; 360 'C Figure 5: Experimental data on HOx production due to combustion in the flame tube versus overall air factor in the natural gas/air/flue-gas mixture. The overall air factor, used in figure 5 has been calculated from the oxygen concentration in the flue gas downstream of the flame tube. In the experiments, the overall air factor could be varied by adjustment of the percentage of natural gas in the unbumt mixture. Thus, the data presented in figure 5 are obtained at a non-constant thermal input. This is not a complicating factor since, in the experimental set-up investigated, NOx emission is not dependent on thermal input. |