風險評估中環境參數解析度之研究--以醫療廢棄物焚化爐為例

Research for Resolutions of Environmental Parameters in Risk Assessmen--A Case Study of Medical Waste Incinerators

黃子玲(Tzu-Ling Huang)

指導教授 ： 馬鴻文

國立臺灣大學/工學院/環境工程學研究所/碩士(2004年)

https://doi.org/10.6342/NTU.2004.00840

摘要

在進行風險評估時，常需使用相當多的參數，而往往在人力、物力、財力有限的情況之下，無法全面收集到充足的參數，因此造成風險評估中的不確定性。資訊量愈高，代表空間解析度愈大，則不確定性愈低。本研究試圖以統計檢定的方法，來得到採樣數要多少的答案及風險評估的可信度。
本研究分為二個部分，第一部分評估的區域為分別位於桃園縣及高雄縣境內的醫療廢棄物焚化爐週界5公里範圍內，其居民承受煙道排放戴奧辛所導致的致癌風險。由焚化爐排放之戴奧辛濃度，經由大氣擴散模式得知在地面的沈降量及濃度後，藉由多介質傳輸與轉換模式，計算不同介質中戴奧辛濃度的分佈情形。之後再結合多途徑暴露模式，估計戴奧辛進入受體的累積量與致癌風險。最後進行風險的不確定性分析及參數敏感度分析。
第二部分為具空間變異之環境參數的研究，此部分為風險評估中不確定性分析的範疇，目的在降低風險評估中的不確定性。先以已知的參數數值，使用Hydro_Gen模擬出評估區域每一點的值，假設為真實場的情形，再依資訊量的多寡畫分不同的解析度，解析度的畫分方式為正方形網格及行政界，比較不同解析度時，風險值的差異性，之後再以統計檢定（t-檢定）來判斷在不同解析度時，風險值是否可視為與真實場相同。不同解析度再個別進行20組的採樣，而在這20組中符合可視為與真實場相同風險值的比例，即為風險評估的可信度。
結果顯示，在設定情境下，桃園及高雄地區境內的醫療廢棄物焚化爐排放戴奧辛致癌風險皆大於1.00E-6。且風險有超過95%來源為攝食，顯示該區焚化爐排放濃度的確過高，有降低的必要性。而若以現行法規0.5 ng-TEQ/Nm3，則大致符合1.00E-6。在敏感度分析結果中，得到具空間變異較敏感的參數皆為土壤容積密度及土壤中有機碳的比例。以這二個參數進行空間解析度的研究，分別討論不同空間解析度時，風險評估結果的變異性及可信度。另外以誤食土壤為最大的不確定性來源。
空間解析結果，風險值變異性隨解析度增加而減少，在風險累積百分比較低時，不同空間解析的差異較小，顯示若欲得累積百分比較小之風險值時，可採較少的樣本。另外，風險可信度隨採樣數的增加而增加，在畫分正方形網格時，一般建議採樣數在100左右，即有相當不錯的可信度。而依行政區畫分時，桃園地區依村里界數目採樣時，風險可信度可達100%，高雄地區則只約30%。

關鍵字

t-檢定 ； 醫療廢棄物焚化爐 ； 不確定性 ； 風險評估 ； 空間解析度

並列摘要

There are many parameters considered in risk assessment. However, it is difficult to analyze all parameters due to limited data and resources, and this causes increases in uncertainties. If more data and resources are obtained, the uncertainties may be reduced with increasing spatial resolution. In this study, a statistical method is provided to determine how much sampling is needed to reduce the uncertainty and increase the reliability of risk assessment.

This paper is divided into two parts. The first part is to evaluate the carcinogenic effects on the residents exposed to dioxins within five kilometers of medical waste incinerators in Taoyuan County and Kaohsiung County. When concentrations of dioxin from the emission of incinerators are known, dry and wet deposition and air concentration on surface can be estimated by ISCST3. Then multi-media transport and transformation model is used to evaluate the concentration and distribution of dioxin in different media. Also the results from the multi-pathway exposure model are aggregated to estimate the accumulative dioxin concentration of receptors and carcinogen risk. Finally, risk uncertainty analysis and parameter sensitivity analysis are implemented.

The second part of the paper is the study of the spatial variances involved in environmental parameters. This is to evaluate the uncertainty range and thus to reduce it in the risk assessment. By using Hydro_Gen, each point in the assessed zone is evaluated to simulate the real situations. Then number of resolutions are generated depends on how much information is obtained. There are two ways to generate different resolutions. One way is square grid method, and another one is administrative district. After comparing the cases under different resolutions and risk variability, a statistic test (t-test) is used to decide if the calculated risk could be regarded as true value in different resolution. Repeat sampling 20 times in every resolution, the risk reliability is the proportion of which could regard as real situation.

From the results, the carcinogenic risk is found to be greater than 1.00E-6, and 90% of the risk is contributed from ingestion. This indicates that the concentration of the dioxin from the emissions of the medical waste incinerators is large enough to cause an adverse effect on human health. The medical waste incinerators are therefore to be advised to reduce the concentration of dioxin to eliminate the carcinogenic risk. By using the existing risk standard of 0.5 ng-TEQ/Nm3, the risk value is evaluated to be 1.00E-6. A sensitivity analysis shows that bulk density of soil at deposition location and fraction of organic carbon in soil at deposition location are the most important parameters. Following that, spatial resolution analysis is implemented by using these two parameters to evaluate the variances and reliability of the risk assessment. However, incautious intake of soil is considered to be the most dominant source of the uncertainties.

The result from the spatial resolutions show that the risk variability decreases as the resolution increases, and the variance between difference spatial resolutions is smaller while the accumulative risk value is lower. This indicates that reducing the number of sampling may lower the accumulative risk value, and increasing the number of sampling may increase the risk reliability. If square grid method is implemented, a proposed sampling of around 100 will give a satisfactory reliability. However, if number of samples depends on the number of districts of village, the risk reliability for Taoyuan is 100% and only 30% for Kaohsiung.

並列關鍵字

medical waste incinerator ； t-test ； spatial resolutions ； uncertainty ； risk assessment

參考文獻

環保署空氣品質保護處，2000. http://www.epa.gov.tw/Dio/890106.htm

Chen, S.Y., Cheng, P.S., Ling, Y. C., 2003. Nationwide PCDD/Fs

Bedford, S. W., 1998. Fluid Mechanics. Appendix C.

Bell, J. U., 2002. Estimating the human health risks from polychlorinated dioxins and furans in stack gas emissions from combustion units: implications of USEPA’s dioxin reassessment. Waste Management 22, 405-412.

Das, B. M., 1941. Principle of Geotechnical Engineering. 9,18,47,83