| Summary: | 碩士 === 臺灣大學 === 職業醫學與工業衛生研究所 === 98 === In Taiwan, VOCs emitted from fixed and fugitive pollution sources (such as various industrial processes in petrochemical industries) are being charged due to a regulation by the EPA. This fee is calculated based on the estimated emission rate of VOCs using the U.S. EPA’s Method AP-42 Compilation of Air Pollutant Emission Factors. However, these emission factors, which come from equations based on averages of available data, incorporate many assumptions and are not plant specific. The Vertical Radial Plume Mapping (VRPM) technique described in the Other Test Method 10 (OTM-10) of the U.S. EPA, which was released recently in 2006 and applies open-path path integrated – optical remote sensing (PI-ORS) system with multiple beam configurations, is a potential alternative to estimate the emission rates. Since the estimated emission rates from VRPM are calculated from measurement data at specific plants, rather than from assumed emission factors, these estimates should be more reliable. In this study, we evaluated the VRPM of OTM-10 in complicated petro-chemical industries. Open-Path Fourier Transform Infrared (OP-FTIR) spectroscopy was used to measure the emission rates in VRPM. Four factories in a petro-chemical industrial area were selected for field experiments. The VRPM configuration is a 2-D vertical plane. Length of beam paths and height of retroreflectors in the VRPM configuration were set up according to objective manufacturing process. The wind direction and wind speed were measured simultaneously. To verify the accuracy of experiment, we release the amount of the tracer gas of SF6 we know during manufacturing process. The amount of emission was obtained by integrating the data of concentration obtained from OP-FTIR measurement and reconstruction algorithm then multiplied by wind speed. Concordance Correlation Factor (CCFPIC) between measured and predicted path integrated concentration (PIC) was used to judge the fitting quality of the reconstruction results. PIC data which fit specific pattern (Among three beam paths the concentration of ground level is the highest, top level the lowest, middle level falls between the top level and ground level.) indicated that the CCFPIC were mostly higher than 0.8. In the first two field campaigns, the calculation result of the amount of emission is quite close to the true value, around 20% was underestimated. Although the last two field campaign provided a highly underestimated result, some important factors were derived: wind direction, arrangement of tracer gas, length of beam paths (compared with objective manufacturing process) and height of retrorelectors.
|
|---|
