OCR Text |
Show effects..,t. 01 due to oxidation, by spiking relatively clean exhaust samples with a known /I concentfation of BTEX. Figure 5 shows a chromatogram from such a sample, taken while operating on RFG3 at 50% excess air and spiked with 4 ppb of benzene, ethylbenzene, toluene, and -each of the xylenes. Each of the spiked species is recovered in the sample, providing confidence in the measurement procedure. Quantitative retention of BTEX on the Tenax sorbent was verified by sampling a relatively high concentration (1 ppm) of calibration gas for intervals of varying length. These data showed a linear relationship between the measured quantity and total volume, confmning that breakthrough was not occurring for concentrations up to ppm levels. Based on system calibration and sorbent trapping tests, measurement accuracy is estimated to be ±10% or 0.5 ppbv, whichever is larger. RESULTS 1 Flame Structure Figure 6 shows changes in the flame structure with fuel type and excess air using visible flame emission. Starting in the upper left comer with the Methane flame at 10% excess air (Xair), the three-fmgered flame structure set up by colliding pairs of main fuel jets is clearly visible. Moving to the right, as excess air is increased the visible flame length is decreased, commensurate with the increased availability of oxygen and the increased momentum for mixing. Moving down to the RFG6, each of the flame fmgers becomes broader and less distinct and Fig. 6 the flame luminosity becomes less uniform than with the neat methane fuel. Moving further down to the RFG3 fuel, an additional broadening of each finger is observed such that their distinct character is almost lost. Note that the luminosity appears at two distinct levels. In the near-burner region a bright, localized luminosity is observed. This is followed by an extended region downstream where a lower-level, more distributed luminosity with an orange hue is seen. - 8 - |