EPA-103-GA13-03-A230

103

13,200 千元

2014/06/23

2015/09/22

吳素慧

土污基管會

張志瑋

台灣中油股份有限公司

土壤及地下水

環境法醫、市售柴油、生物標誌物、指紋圖譜、診斷比值、統計分析技術、檢測技術、氣相層析質譜儀

Environmental forensics, Diesel, Biomarker, Chemical fingerprint chromatogram, Diagnostic ratio, Multivariate statistical analysis techniques, Measuring Technologies, GC/MS

高振山

吳素慧

037-356150#201

048682@cpc.com.tw

本計畫因市售B2生柴油品停售已依規定申請計畫變更，並於103年12月18日獲環保署來文（環署土字第1030107554號）核准在案。本計畫依變更後共完成下列主要工作： 一、在文獻蒐集部份，本期計畫持續蒐集到近三年的相關研究文獻與報告14篇，亦納入前期已建立的Excel簡易資料庫系統內。連同包含前期計畫執行所蒐集的相關研究文獻與報告共計245篇。關於油品鑑識技術相關文獻研析，除了對柴油鑑識技術進行研析之外，主要進一步針對汽油相關鑑識分析技術、是否添加生質柴油之柴油產品的風化作用探討以及油田洩漏事件鑑識技術之應用。另一方面，本計畫除進行文獻研析外，亦從之前承接委辦計畫報告的專有名詞的用語蒐集開始，初步完成149個專有名詞中英用語的比對工作與資料表建置，並協助環保署初步完成土壤與地下水油品污染鑑識執行架構及標準作業程序(草案)的研訂。 二、完成涵蓋北中南東地區由中油與台塑供應的加油站，共126個市售柴油(即高級柴油)樣品之氣相層析質譜儀(Gas Chromatography/Mass Spectrometry,GC/MS)分析，並完成每個樣品至少5個群組指紋圖譜分析與診斷比值計算之數據處理，同時建置分析油品化學指紋圖譜資料庫。並藉由研析結果，進一步篩選出具市售柴油供應源鑑識適用性之41診斷比值與診斷能力值(Diagnostic Power, DP)大於5之29個診斷比值，同時運用主成份分析(Principal Component Analysis, PCA)統計分析方法，將126個樣品群聚分析結果，可明顯的將其區分為中油[CPC Corporation, Taiwan (CPC)]與台塑(Formosa Petroleum Corporation, FPC)供應源群組。進而運用本期市售高柴油品與前期計畫市售B2柴油共267個樣品彙總解析結果，驗證前期所建立之市售柴油供應源鑑識技術具有其適用性與可行性。 三、完成由中油與台塑煉油廠所提供的72個加氫脫硫柴油樣品(含煉製成品與出廠前油品)之GC/MS分析，及其至少5個群組指紋圖譜分析與診斷比值計算之數據處理。並完成市售柴油與煉油廠加氫脫硫柴油之特性綜合研析。 四、在模擬混合柴油方面，運用台塑與中油市售柴油配置3種不同比例(75:25、50:50、25:75)混合柴油，完成配置81個混合柴油的GC/MS分析，及其至少5個群組指紋圖譜分析與診斷比值計算之數據處理。比對指紋圖譜與診斷比值的實際計算與模擬計算之差異性結果，運用實際計算與運用數學式模擬計算的41個診斷比值，分別進行主成份分析所繪製的散布圖，可看出其3種混合柴油落點依混合比例含量大小，由左而右(75:25、50:50、25:75) 依順序呈現相似的位移模式。 五、本計畫共計完成6處實際污染場址33個污染土壤樣品的前處理與GC/MS分析，並完成至少5個群組指紋圖譜分析與診斷比值計算之資料數據處理。並由本期與前期彙整篩選出的鑑識市售柴油供應源之41個與DP值>5的21個診斷比值中，完成刪除受汽油混合污染干擾、溶解作用與菌蝕生物降解作用影響之診斷比值，初步篩選出柴油洩漏污染供應源鑑識適用性的36個、31個與DP值>5的18個診斷比值。並進而經由四個確認供應源，並分別受到包括汽柴油混合污染、經過整治與混有不同時期洩漏污染干擾的實際柴油洩漏污染場址的污染土壤樣品的群聚分析驗證，指明所建立的柴油供應源鑑識技術得以適用於受汽油混合污染干擾、溶解與輕度至中度菌蝕生物降解作用影響之柴油洩漏污染供應源之鑑識。惟仍有待更多處於不同狀態的污染場址樣品分析進一步的驗證，以真正釐清其適用性並進一步修正。 六、本計畫共計完成三次已建立之鑑識技術移轉。並參加2015/8/3-8/6於加拿大舉辦的2015 International Network of Environmental Forensics (INEF) 研討會，綜合本計畫與前期計畫執行成果論文發表。

環境法醫、市售柴油、生物標誌物、指紋圖譜、診斷比值、統計分析技術、檢測技術、氣相層析質譜儀

EPA-103-GA13-03-A230.pdf

85MB

2016/2/1 上午 12:00:00

The final report is divided into the following parts: In this project, fourteen papers relevant to spilled-oil identification are collected and added into the previously established Excel Database which now amounts to 245 documents. Besides diesel spilled-oil identification papers ,we reviewed about gasoline identification techniques, weathering processes of biodiesel ,and different oil spill identification papers. In addition to documents analysis, the project scrutinizes 149 bilingual professional terms which were collected from those used by previous project, website and mainland China. The project also provides Environmental Protection Administration (EPA) with a draft of standard operating procedure (SOP) and execution framework of spilled-oil identification techniques for contaminated soil and underground water. We collected 126 diesel samples in different gas stations around Taiwan supplied by either CPC Corporation, Taiwan (CPC) or Formosa Petroleum Corporation (FPC). A Gas Chromatography/Mass Spectrometry (GC/MS) analysis of these samples was carried out. The data processing of at least five groups of chemical fingerprint distributions and diagnostic ratios were performed for all samples. In the meantime, we established a chemical fingerprints database for future reference. Based on the analysis result, we screened out 41 diagnostic ratios and among them there are 29 with the value of Diagnostic Power (DP) bigger than five. We also conducted the cluster analysis for 126 samples through Principal Component Analysis (PCA). The results of the study can discern the difference of the diesel oil source between CPC and FPC. With the analysis results of 267 gas station diesel samples which includes some diesel samples used in this project and others (B2 diesel) in previous project, we prove the practicability and feasibility of the identification techniques for diesel oil source. We completed the GC/MS analysis of 72 diesel samples including hydrogen desulfurization refined products and those before leaving factory in both CPC and FPC Refinery. Among these samples a data processing of at least five groups of chemical fingerprint distributions and diagnostic ratios was carried out. A comprehensive analysis of the character of marketing diesel, hydrogen desulfurization diesel and those before leaving factory was executed We used diesels from CPC and FPC to produce three different mixed diesels with the ratios of 75:25, 50:50 and 25:75. We performed eighty-one GC/MS analysis and conducted the data processing of at least five groups of chemical fingerprint distributions and diagnostic ratios. The study conducted correlation of chemical fingerprint distributions, diagnostic ratios calculation, PCA of above-mentioned eighty-one mixed diesel samples. We also carried out simulation calculation of diagnostic ratios of above-mentioned eighty-one mixed oil samples through the diagnostic ratios of pure diesel samples and performed PCA cluster analysis of simulation calculation of diagnostic ratios. The correlation between PCA cluster analysis of real mixed oil samples and that of simulated mixed oil samples indicates that they are characteristic of similar cluster distribution. This project completed pretreatment and GC/MS analysis of 33 contaminated soil samples collected from six diesel contaminated sites and carried out the data processing of at least five groups of chemical fingerprint distributions and diagnostic ratios. Based on the study of this and previous projects, we screened out the 41 diagnostic ratios of forensic marketing diesel sources and among them there are 21 diagnostic ratios with the value of DP bigger than five. We also deleted the diagnostic ratios that were interfered by mixed gasoline contamination, dissolution and biodegradation. Furthermore, we screened out 36 diagnostic ratios which are suitable for forensic identification and among them there are 18 with value of DP bigger than five. The PCA of contaminated soil samples from four verified spilled-diesel sites indicates that the technology of diesel source identification is suitable for identifying the spilled-diesel contamination which was influenced by disturbance of mixed gasoline contamination, dissolution, and light to moderate biodegradation. As the four above-mentioned contaminated sites could not represent all, screening out more sites with various contamination factors is needed to verify its suitability and improve it for future use. This project completed the transfers of forensic identification technologies three times separately after preliminary report, mid-term report and before final report. The results of this and previous projects have been presented in 2015 International Network of Environmental Forensics (INEF) held in Canada during Aug 3-6, 2015.

Environmental forensics, Diesel, Biomarker, Chemical fingerprint chromatogram, Diagnostic ratio, Multivariate statistical analysis techniques, Measuring Technologies, GC/MS