Modelling Comparison and Health Risk Assessment of Volatile Organic Compounds in the Atmosphere of Petrochemical Industrial Park

• Main Authors: Cheng-Bin Chen, 陳正彬
• Other Authors: Chung-Shin Yuan
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/30608354787419685952

碩士 === 國立中山大學 === 環境工程研究所 === 101 === Petrochemical and petroleum refining industrial complexes are gathered in northern Kaohsiung, including China Petroleum Company (CPC) refinery plant, Dazher and Renwu Petrochemical industrial parks. Large quantities of volatile organic compounds (VOCs) emitted from this area has caused long-term poor ambient air quality. In recent years, many studies have measured the VOCs concentration in this area, but had only limited relevant information to clearly identify the causes of poor ambient air quality. Thus, this study aimed to investigate the main species and the contribution of VOCs in the petrochemical industrial complexes. The results of this study will be valuable for establishing air quality management and air pollution control strategies for environmental authorities to evaluate and improve ambient air quality. This study conducted five times of field sampling and chemical analysis of VOCs at northern Kaohsiung petrochemical industrial complexes in the wet and dry seasons from 2009 to 2012. For the first sampling period, we focused on the characteristics of VOCs concentration and fingerprints in different elevation levels (0 m, 100 m, 300 m, and 500 m), and determined the apportionment of fugitive, mobile, and stationary sources by using a chemical mass balance (CMB) receptor model (CMB8). For the second and the third sampling periods, eight sampling sites located at the petrochemical industrial complexes were selected to characterize the chemical fingerprints and the diurnal concentration variation of VOCs. We further used the CMB model to determine the apportionment of VOCs emission sources. For the fourth and the fifth sampling periods, VOCs were sampled at either schools or governmental agencies in the nearby densely inhabited districts. Furthermore, we utilize an air pollution dispersion model (ISCST3) to simulate the spatial distribution of VOCs concentration and compare them with those obtained from the chemical mass balance receptor model. Finally, we assessed the health risk based on the field measurements of VOCs. The vertical profile of VOCs showed that the species of VOCs were different from the ground and the aerial. It was mainly attributed to the emission of VOCs from elevated stacks located at the petrochemical industrial complexes. The concentration of VOCs in the dry season was generally higher than that in the wet season. The concentrations of VOCs at sampling sites which are far away from the petrochemical industrial complexes ware generally lower. Among the VOCs measured in the ambient air of the petrochemical industrial complexes, toluene and acetone were two major species of VOCs. The dispersion of VOCs surrounding the petrochemical industrial complexes was mainly influenced by prevailing winds. If the near-surface wind speed was not too high, VOCs emitted from elevated stacks would disperse to the region within the distance of 1-2 km downwind the petrochemical industrial complexes. Chemical mass balance receptor modeling results showed that, at the ground level (H=0m), stack emissions contributed about 42 %, while fugitive sources and mobile sources contributed about 26 % and 25 %, respectively, of VOCs. At the altitude of 100 m, stack emissions contributed about 56 %, while fugitive sources and mobile sources contributed about 22 % and 13 %, respectively, of VOCs. At the altitude of 300 m, stack emissions contributed about 60 %, while fugitive sources and mobile sources contributed about 15 % and 6 %, respectively, of VOCs. At the altitude of 500 m, stack emissions contributed about 68 %, while fugitive sources and mobile sources contributed about 15 % and 7 %, respectively, of VOCs. We further investigated the emission sources in the daytime and at nighttime. Stack emissions contributed about 46 %, while fugitive sources and mobile sources contributed about 30 % and 8 %, respectively, in the daytime. At nighttime, stack emissions contributed about 57 %, while fugitive sources and mobile sources contributed about 28 % and 6 %, respectively. Correlation analysis between CMB receptor model and dispersion model showed that the simulation at high altitudes was less accurate than that at low altitudes, dry season was better than wet season, and the Renwu industrial park was better than the Dazher industrial park. Air temperatures in the wet season were generally higher than those in the dry season, which could accelerate the atmospheric photochemical reaction. Emission sources at the Dazher industrial park are more complex than the Renwu industrial park, which resulted in lower VOCs concentration at the downwind sites than those located at the upwind sites. Results from health risk assessment showed that the cancer risks on the 95 % confidence interval ranged from 4.42 × 10-4 to 1.58 × 10-6, which were higher than the reference value of 10-6. Cancer risks on the 50 % confidence interval ranged from 8.85 × 10-5 to 6.29 × 10-7 and partial sampling sites had higher cancer risks than the reference value of 10-6. The results of non-carcinogen risk assessment showed that the hazard index of sampling sites RW4, DS3, DS4, and DA3 were higher than unity on the 95 % confidence interval, while the hazard index of sampling sites RW4, DS4, and DA3 were higher than unity on the 50 % confidence interval. Moreover, all sampling sites had demonstrated no acute toxicity hazards during the VOCs sampling periods.