新竹科學園區砷空氣污染之空間性影響及風險評估

Abstract

新竹科學園區為台灣的科技重鎮，然而區內公司不少製程中使用含有高毒性的砷化物，且在近期公佈了數年前曾偵測到高濃度的砷化物，因而引起大眾的重視，除了應建議園區廠商儘速推動無砷製程及公佈監測資訊，更應儘速建立一套模式供分析砷等具毒性空氣污染對周遭民眾的影響及風險評估模式。本研究因而發展一套空氣污染模式與健康風險評估方法，以供評估科學園區空氣砷污染排放物對於周界居民的健康風險影響。 本研究主要採用AERMOD模式模擬砷污染分析，在建立模式所需資料後，即以園區固定污染源排放資料模擬各模式網格的濃度增量分布，並提出方向性污染程度時數、方向性各污染程度影響區域及方向性最大小時濃度影響區域三種方法探討不同風向下砷污染物的濃度分布情形，及以等去除率法分析達成一些預期周界濃度所需的去除率，且以致癌風險及致癌負荷評估健康風險的空間分佈與差異。 本研究分別以2002年、2008年排放資料及結合兩者為WA、WB及WC三種情境，並以2002年及2007氣象資料進行模擬，模擬結果發現遠低於監測值，顯示排放清單並不完整，環保單位有必要建立更完整的清單。而在方向性污染衝擊分析中，發現高污染濃度部分以東風下的時間最長。在各風向平均濃度分析中，WA以東南東風下影響區域最廣，於WB及WC則以西南西風影響區域最廣。而各風向最大小時濃度除了靜風外，主要以南風至北北西風等8個風向影響最廣。以等去除率進行周界預期濃度分析，發現需要高達80%去除率方能降至0.2 ng/m3，加上排放清單不完整，預期所需要的減量將更大。致癌風險分析則發現部分地區所承受風險較高。

In Hsinchu Science-based Industrial Park (SBIP), semiconductor and opto-electronic factories use a significant amount of chemicals with arsenic compounds in their manufacturing processes. High ambient arsenic concentration measured in few yeas ago were recently reported and attracted great attention from the public. Other than promoting non-arsenic manufacturing processes and reporting regular monitoring data, analyzing air toxic pollution distribution and assessing residential exposure and health risks are desired. The AERMOD model was thus applied in this study to establish a model to simulate spatial arsenic pollution distribution. Three methods of directional pollution level hours, directional impact area, and directional hourly maximal concentration are developed for assessing the directional impacts of the arsenic emission from SBIP on adjacent residents. The uniform removal method is applied to estimate the required percentage reductions of source pollutants for achieving several expected ambient criteria. Two risk assessment measures are then applied to estimate the residential exposure and health risks for different scenarios. This study uses two arsenic emission data sets in 2002 and 2008 to establish scenarios WA and WB, and scenario WC for both data sets is also analyzed. The pollution distributions of the three scenarios are simulated based on 2002 and 2007 meteorological data. The simulated results are significantly lower than those sampled from random monitoring at surrounding locations. The emission data sets are obviously incomplete and should be exteneded. Based on the directional impact analysis, the largest number of hours with high pollution occurs under the eastern wind direction. According to the average concentration distribution in each direction for scenario WA, the ESE wind influences the largest area, while the WSW wind influences the largest area for scenarios WB and WC. For the maximal hourly concentration distribution in each direction, except the calm wind, the areas influenced by eight wind directions such as sourth and NNW are significantly larger than those for other wind directions. For achieving the expected ambient arsenic concentration lower than 0.2ng/m3, about 80% reduction of all pollution sources is required. Since the emission data sets are incomplete, a higher removal rate is required for the complete set. Regarding the health risk, certain parts of the studied area have higher risks than those in other parts of the area.