OCR Text |
Show and hypothetical extremes, which are not considered representative of normal field operation but which were investigated in the research furnace to elucidate organic H A P emission mechanisms. The majority of detected data for normal operating conditions lies just above the limits of detection (represented by the shaded area on the figure), with concentrations generally on the order of 1 to 100 parts per billion or less. With the exception of formaldehyde, significantly higher concentrations were encountered only in the hypothetical extreme cases. Note, data which were measured but not detected are not shown on this figure. It can be seen that benzo[a]pyrene, for example, was not detected under normal operating conditions and was rarely detected even during tests at hypothetical extreme operating conditions. In the P E R F program, a "baseline" operating condition was established within the normal operating range. Tests at this condition were repeated numerous times throughout the program to verify the absence of undesired changes in operating or measurement conditions and to define the inherent variability of the results. Average emissions of benzene and P A H at the baseline conditions were essentially at the detection limits and formaldehyde emissions were slightly above detection limits (Figure 3). Total P A H in Figure 3 includes the sum of eighteen individual substances8. Naphthalene is a decomposition product of the sorbent used to collect P A H samples and therefore a potential contaminant. It is often detected well above the analytical detection limits in both the sample and in quality control blanks. Therefore, naphthalene was excluded from the total in this and the following figure to enable a more valid comparison to field data. Undetected data are included individually or in average results at the full detection limit in Figure 3 and the following charts, unless noted otherwise. Results showed that organic H A P emissions were the same as baseline levels even with poor fuel/air mixing (including simulated air staging conditions) or simulated mechanical failures. Organic H A P emissions increased only under extreme substoichiometric air/fuel ratios at very high excess air levels. Figure 3 also shows "smoke spot" readings taken from the filters used in the sampling for P A H . The dark smoke spot indicates that a smoke plume would be visible under extreme air starvation conditions, a condition which would be quickly corrected in an actual facility. Extreme air quench conditions also would be corrected quickly, since this produces very high thermal efficiency losses. In addition, either extreme air starvation or extreme excess air represent unsafe practical operating conditions because they could result in the onset of unstable combustion creaating a potential explosion hazard Benzene, formaldehyde and total P A H emissions generally remained near or below detection limits for the range of normal operating conditions (Figure 4). With rare exceptions, P A H other than naphthalene were not detected. Formaldehyde was detected but remained near detection levels in most tests. In one test (labelled "No H2, Increased H H V (1)"), benzene and formaldehyde showed an unexpected increase. Formaldehyde results were not reproduced when this test was repeated (labelled "No H2, Increased H H V (2)") but benzene results were consistent. This illustrates how formaldehyde measurements using the D N P H method can occasionally be subject to unexpected variability even under the best circumstances. Nitrogen oxides ( N O x), carbon monoxide (CO), and total hydrocarbons (THC) were measured concurrently with the organic H A P measurements during the P E R F tests. The low-NOx burner 4 |