EPA-105-U102-03-A284

105

32,000,000 千元

科技發展-環境科學及技術之研究

2016/10/20

2017/12/31

李崇德

監資處

黃健瑋

國立中央大學

環境監測及檢測

PM2.5化學成分監測、PM2.5化學成分時間與空間分布特徵、推估PM2.5污染來源及能見度影響因子

PM2.5 chemical speciation, Temporal and spatial distribution of PM2.5 chemical components, Source apportionment of PM2.5 and visibility influencing factors

周崇光、張士昱、許文昌

黃秋華

03-4226551#

cilia0110@gmail.com

本計畫執行PM2.5化學成分監測，於2016年11月底至12月初在斗六、嘉義、花蓮環保署空氣品質監測站進行連續8天任務型採樣，2017年1月至11月，在板橋、忠明、斗六、嘉義、小港、花蓮測站進行每6天一次例行性採樣。這兩種型態採樣PM2.5質量濃度與環保署監測站連續監測濃度的時間變化趨勢具有一致性，顯示採樣結果具有代表性。 各站2017年1月至11月例行性採樣，由東而北而南PM2.5日平均質量濃度分別為11、20、23、28、28、29 μg m-3。四季PM2.5質量濃度變化，大致呈現夏季最低，秋季升高，冬、春濃度高於秋季。冬季小港站擴散條件較差，又受到上風區域污染傳輸影響，濃度為6站最高，春季濃度最高站點在斗六，嘉義站濃度非常接近，這兩個站點常有相近高污染濃度發生；夏季以板橋站濃度最高，秋季則以小港站濃度最高。採樣過程PM2.5化學成分會有揮發及吸附，若僅使用單張濾紙，水溶性無機離子將會低估NH4+濃度6%~20%、低估NO3-濃度4%~54%，也會低估Cl-濃度19%~62%。碳成分在高PM2.5質量濃度時有較多的負干擾(揮發)，低濃度時則正干擾(吸附)影響較為顯著。各站各季節PM2.5質量濃度占比最高化學成分：冬季板橋與忠明站為SO42-，斗六、嘉義、小港站為NO3-，花蓮站為OC；春季板橋、忠明、花蓮站為SO42-，斗六、嘉義、小港站為NO3-；夏季板橋、忠明站為OC，其餘各站為SO42-；秋季各站都是以SO42-為主。以日平均濃度35 μg m-3為分界，採樣期間PM2.5低濃度(日平均濃度為17 μg m-3, n=266)相對於高濃度(日平均濃度為46 μg m-3, n=70)樣本，PM2.5 NO3-占比從11% 增大到24%，其他成分則縮小或不變，顯示PM2.5 NO3-前驅物NOx排放源的管制應該加強。各站PM2.5金屬元素成分最高兩種分別為Na和K，顯示受海風、地殼塵土逸散、生質燃燒影響大。計算各站金屬元素富集因子，指出地殼、燃煤、交通排放為重要排放源。彙整各次採樣PM2.5濃度大於35 μg m-3事件污染來源，得出大部分事件都是區域傳輸、擴散不佳、發生光化學反應所造成，僅少數來自境外傳輸。 綜合6個測站PMF受體模式推估結果，各測站主要污染因子都以二次硝酸鹽和二次硫酸鹽這兩類富含衍生污染物的因子為主，其中，二次硝酸鹽於冬、春季多有最高占比，二次硫酸鹽則多在夏、秋季有最高占比，突顯各季節污染源變化特性。大氣能見度推估顯示各監測地區影響能見度最大的PM2.5化學成分為NH4+或SO42-，顯示衍生污染物對於能見度的影響相當可觀。在PM2.5化學成分技術評估方面，本年度彙整34篇相關國際文獻。

PM2.5化學成分監測、PM2.5化學成分時間與空間分布特徵、推估PM2.5污染來源及能見度影響因子

EPA-105-U102-03-A284.pdf

28MB

2018/2/1 上午 12:00:00

This study conducted an assigned mission of PM2.5 chemical speciation monitoring at Douliu, Chiayi, and Hualien air quality monitoring sites of Environmental Protection Administration for continuous eight days from the end of November to early December in 2016. In contrast, regular PM2.5 chemical speciation monitoring at Banqiao, Zongming, Douliu, Chiayi, Xiaogang, and Hualien were implemented on a shift of every 6 days starting from January to November in 2017. PM2.5 mass concentrations from these two types of collection varied consistently with those of continuous monitoring stations which indicated the achievement of representative collections of PM2.5. PM2.5 daily averages, from January to November 2017, were 11, 20, 23, 28, 28, and 29 μg m-3, respectively, at the sites from east and north and south of Taiwan. Seasonal variations of PM2.5 mass concentration are with a pattern of the lowest in summer, higher than summer in autumn, and further up in winter and spring. In the winter time, the Xiaogang site was with the highest concentration due to bad environmental ventilation and the influence from upstream transport. In spring, the Douliu site ranked the highest but Chiayi was close to it especially for high concentration events. The Banqiao site became the highest in summer, while Xiaogang site claimed the highest in autumn. Volatilization and adsorption of PM2.5 chemical components were evaluated during collection. Underestimates of 6%~20% from NH4+, 4%~54% from NO3-, and 19%~62% from Cl- were found if only a single filter was installed for PM2.5 collection. Carbonaceous contents were with more negative interferences (volatilizations) in high PM2.5 mass concentrations in contrast to prominent positive interferences (adsorptions) in low PM2.5 mass concentrations. The highest levels of PM2.5 chemical component at each site for various seasons are as follows. In winter, SO4- was the highest chemical component at the Banqiao and Zongming sites, NO3- at the sites of Douliu, Chiayi and Xiaogang, and OC at the Hualien site, respectively. Whereas SO4- was the greatest component at the Banqiao, Zongming, and Hualien sites and NO3- at Douliu, Chiayi, and Xiaogang sites in spring. The OC was the highest component at the Banqiao and Zongming sites in contrast to SO4- at the other sites in summer. Interestingly, a unanimous highest SO4- level of chemical components was found for all sites in autumn. Splitting from daily average at 35 μg m-3, the ratio of NO3- to PM2.5 for low PM2.5 concentrations (17 μg m-3, n=266) relative to high PM2.5 concentrations (46 μg m-3, n=70) was elevated from 11% to 24% while other components shrank or unchanged. This indicates a need of stringent control of NOx emission sources, which is PM2.5 NO3- precursor. The highest two of PM2.5 metal elements are Na and K with the implication of the significant influences of sea-salt, fugitive crustal dust, and biomass burning. The computations of enrichment factor across all sites showed that fugitive crustal dust, coal combustion, and traffic emissions were important sources. Summarizing all events of PM2.5 concentrations greater than 35 μg m-3, this study found most events were under the influence of regional transport, bad ventilation, and photochemical reactions with only few from transboundary transport. In summarizing PMF receptor modeling results for the 6 sites, the major pollution factors across all sites are derivative types of factor such as secondary nitrate and secondary sulfate. Secondary nitrate contributed high PM2.5 concentration ratios to most sites in spring and winter, while secondary sulfate contributed mostly in summer and autumn. The estimation of atmospheric visibility showed that corrected NH4+ or SO42- was the major PM2.5 chemical species influencing visibility at each individual site. This implies that the effects from derivative pollutants on atmospheric visibility are significant. For the assessment of PM2.5 chemical speciation techniques, 34 related international papers were collected and reviewed.

PM2.5 chemical speciation, Temporal and spatial distribution of PM2.5 chemical components, Source apportionment of PM2.5 and visibility influencing factors