運用垂直煙流分布技術推估石化廠之儲槽揮發性有機物逸散排放量

Abstract

現行石化工業廠區的揮發性有機物(VOC)的排放量推估方法大部分是根據標準的排放係數(Emission factor, AP-42 method)所計算出。此方法由美國環保署公佈，其中的排放系數多由許多假設與過去的經驗及根據美國產業制定與建立。相對而言，垂直煙流分布(VRPM)方法運用光學遙測儀器，於下風處進行汙染物的濃度量測，同時結合風向與風速資料，計算推估上風處汙染源單位時間內的排放量。 在本研究中，挑選一石化廠區之儲槽區，透過架設開徑式紅外線傅立葉轉換儀與其所需之多面反射鏡，於下風處構成五道光徑之垂直煙流分布平面(VRPM plane)，量測通過此平面之揮發性有機物煙流(plume)並計算排放通量。此VRPM平面，其長約為190公尺，高約為30公尺。整體量測期間為14天。利用不同方法所得之氣象資料、不同大小之垂直煙流分布平面之網格面積、以及校正方案推算目標汙染物之排放量。另於監測期間，施放兩天之追蹤氣體SF6以驗證此方法的準確度。追蹤氣體實際施放速率為 0.4178 g/s，所推估之排放速率介於0.30至0.77 g/s。本研究面臨之主要限制為：當風向不理想時，排放量推估需要校正，此校正受限於風向範圍的選取；監測期間僅約為二星期的時間，結果難以代表長期之排放量；雖上風處無槽區與產品製程區，但本研究缺少上風處量測資料故難以完全排除上風處污染源之影響程度。

Conventionally the emission rate of volatile organic compounds (VOCs) is estimated by applying standard emission factors which are based on the AP-42 method of the US EPA. On the contrary, the Vertical Radial Pluming Mapping (VRPM) technique is a method designed for modeling emissions flux by directly monitoring the VOCs plumes crossing a vertical plane at a downwind site. In this thesis, a vertical plane is constructed with multiple retroreflectors and an optical remote sensing (ORS) instrument at a tank farm in a petroleum plant. The VRPM plane was approximately 190m (width) x 30m (height) at the downwind site. The total experimental period is 14 days. We aim to evaluate the effects on model performance from using different wind vector calculations, two grid sizes of the VRPM plane, and various scenarios of adjusting the emission rates. During the monitoring period, the tracer gas of SF6 was released for two days to verify the accuracy of the estimated emission rate. The true release rate of SF6 was 0.4178 g/s and the VRPM estimates ranged between 0.30至0.77 g/s. The major limitations in this thesis were as follows: (1) When the wind did not come from a proper direction, the estimate need to be adjusted. However, this adjustment was limited by the selection of applicable wind direction range; (2) The data collected over the 2-week monitoring period may not represent the long term conditions; and (3) No monitoring data were collected at the upwind site. Thus the impacts from upwind sources could not be ruled out completely.