OCR Text |
Show A16SE713.T 8 After examining the TGNMO emissions results from the gas turbines, it was detennined that TGNMO emissions may contain significant amounts of ethane. The defmition of VOCs specifically excludes methane and ethane (40 CFR 51.100, Subpart F). In order to detennine actual VOC levels (Le., excluding both methane and ethane), ethane emissions should be measured directly and excluded from TGNMO. Formaldehyde Emissions Figure 6 presents the fonnaldehyde emissions measured from the seven gas turbines. Fonnaldehyde, like benzene, toluene, methane and TGNMO, generally decreases sharply with increasing load, and appears to increase exponentially at operating points below 50% load. Fonnaldehyde was not detected at the 100% load point tested for the Solar Mars T12000 gas turbine, and overall results from the other units range from 0.9 to 2,360 ppb at 100% load. The maximum fonnaldehyde value measured at 25% load was 10,718 ppb at 15% O2 from the General Electric LM1500 gas turbine. Using a basis of tons per year (tpy) , the maximum value for fonnaldehyde was 5.34 tpy from the Rolls Royce Avon gas turbine. Health Risk Assessment A health risk assessment (HRA) using generic meteorology data was prepared for the two utility turbines in order to provide perspective on the results. The HRA was perfonned using full load emissions data, and incorporated risks by inhalation, ingestion of soil and plants, and uptake of substances through the skin. It should be noted that standard HRA methodology approved by California regulators was used, and that this is different than the more complex methodology being used by EPRI in its PISCES HRA efforts. The results of the HRA are shown in Figures 7 and 8. The total cancer risk for the turbines were both approximately 0.1 x 10-6 , with 80 to 90% of the calculated risk resulting from chromium emissions based upon 6 % of the total chromium being of the fonn Cr(Vl) . There is no fIxed level of risk at which regulators take action, but the lowest action level implemented by a regulatory agency is 1 x 10-6 , and an action level of lOx 10-6 is more typical. Therefore, the risk levels calculated for the two utility turbines are one to two orders of magnitude below those that typically trigger regulatory action. CONCLUSIONS The major conclusions drawn from the program are: 1) The load at which a given gas turbine is operated has a strong effect on the emissions of criteria pollutants (NOx, CO) as well as on the emissions of HAPs and VOCs (benzene, toluene, methane, TGNMO, fonnaldehyde) . This effect of load is primarily due to the design of a given turbine and its combustion characteristics. Emissions of these species may also be influenced by the HAP and VOC constituents present in the natural gas fIred in the turbine. |