OCR Text |
Show trap, elution from the trap using cyclohexane, and analysis of the eluate on a liquid chromatograph. The exhaust gas is filtered of particulates in an oven maintained at the sampling temperature before the gaseous components are passed over the Chromosorb 102 adsorbent. The hot filtration prevents adsorbtion of the hydrocarbons on the particulates. The liquid chromatograph separates the sample into two fractions: oxygenates (LCO) which typically have a smoky-burnt odor quality, and aromatics (LCA) which typically have an oily-kerosene odor quality. Application of this procedure to automotive diesels resulted in a correlation between the detected amount of LCO and the odor level as perceived by a trained panel (Levins e_t a_l. , 1974). The basic DOAS odor sampling system, sampling analytical procedures, and techniques employed are described in detail by Cernansky e_t al_. , (1978) and Hames - et al. , (1980). However, in the present experiments, sampling procedures were suitably modified (Tumati e_t a_l. , 1983) and the sampling was performed at near isokinetic conditions. A schematic of the DOAS sampling system is shown in Figure 2. In addition, samples that produced the highest LCO levels in each profile were separated into polar and non polar fractions employing a protocol developed by Shala (1983) using silica sep-pack cartridges. The polar fractions, which contained almost all of the oxygenates, were analysed using GC/MS. Experimental Results In general, all experiments show that LCO concentration increases initially, reaches a peak near the end of the main flame zone and decays in the post flame region. The rate of decay of LCO finally approaches zero and the final steady state LCO level indicates the ultimate emission level of the burner. Representative temperature and concentration profiles are shown in Figures 3 and 4. -5- |