| Summary: | 博士 === 國立成功大學 === 環境工程學系 === 88 === This study investigates the relationship between volatile organic profiles in the atmosphere and emission sources in ozone episode region. Dynamometer and industrial stacks sampling work in the industrial complex were conducted to obtain the fingerprints of vehicle emissions (2-stroke, 4-stroke motorcycles and gasoline passenger cars) and stationary sources (oil refinery, synthetic organic chemical manufacture factories, and industrial coating manufacture). In addition, full-scale field sampling on the main VOC concentration in monitoring stations using USEPA method TO-14a on photochemical and non-photochemical event days were done by canisters. Tedler bags and canisters were analyzed by a gas chromatograph/ mass spectrometer (GC/MS) with a cryogenic trap and a thermal desorber unit. The concentration and emission rate database of fifty-six individual VOCs has been established. The contributions of VOC emissions from different sources were also estimated by a back-trojectory model, factor analysis (FA) and chemical mass balance model (CMB 8.0).
There is a clear relationship between VOC characteristics from emission sources and atmosphere environment. The air mass passes through the regions (Kaohsiung metropolis, Renwu, Linhai, Dafa and Linyuan industrial complexes) with complex emissions in southern Taiwan. The transport of VOC pollutants affect the air quality of Chaochow monitoring station, which is an apparently unpolluted area located in the leeward side of the ozone episode region.
For the city and suburban monitoring stations, the exhausts of motorcycles and passenger cars are the major contributors of VOC in southern Taiwan. The contributions of total non-methane hydrocarbon (t-NMHC) from the above vehicles among five stations are between 40-70% by CMB simulation during different periods, obviously affects the atmosphere of city monitoring stations during rusher hours.
Industrial stationary sources and coatings (solvent) are next in important contributors in southern Taiwan. They account for 10-30% and 15-45% of t-NMHC respectively basing on their contributions. The proportions of t-NMHC contribution from two types of emission sources are drastically different among individual monitoring stations. The contributions of t-NMHC in the atmosphere of Linyuan station are estimated 40-60% for refinery complex. Also, the results of FA and CMB from the biogenic contributions among all stations are between 0.1~2.4%, similar to the overseas researches and being an outstanding characteristic at noon. The proportions of biogenic contribution are underestimated by CMB simulation due to higher photochemical reactivity and short life time of biogenic VOCs.
Because of the difference of VOCs profiles between CMB_twvoc (built by this research) and Pall (the inbuilt VOC database of CMB 8 model, not including motorcycle profiles) databases, CMB simulated results by Pall database deviate the true valuse of emissions on subdivided vehicles (higher diesel exhaust and liquid gsoline emissions). On the contrary, CMB simulated results by CMB_twvoc database are authentically similar to the emission inventory of Taiwan. Again, three fingerprints of CMB_twvoc database from the local refinery are suitable for simulating the air samples of Linyuan station, whose varying concentrations were not properly simulated using Pall database. Since the result indicates that the local emission characteristics clearly affect the CMB model simulation, the study recommends that a regional database be established either by creating one or by referring to the past local researches before using the CMB model simulation to reduce errors.
|
|---|
