以田口方法改善高鹼度FGD廢水中重金屬干擾研究

Abstract

分析及量化含有複雜基質水樣中之重金屬是非常有挑戰性的，尤其是對於燃煤發電廠所產出的尾氣除硫((flue gas desulfurization, FGD)廢水的水樣更是困難，因為不僅背景干擾甚高且多，所需定量之濃度卻相當微量。然而，不久的將來，放流水標準中的燃煤發電廠FGD廢水放流水中砷、硒、鎘與汞將會加嚴至更低的濃度，在這情形下，在重金屬分析量化上的干擾則可能造成更明顯之誤差影響。為解決此種問題，本研究旨在探討對於FGD廢水中干擾重金屬量化的主要因子，以火焰式原子吸收光譜儀(FAAS)搭配氫化裝置分別對於模擬及真實廢水進行分析與探討。為了減少本研究進行實驗的數量，應用田口方法於實驗設計，將欲探討之造成干擾之離子或基團，在高中低不同濃度下之干擾進行探討，此外對於去除干擾之水樣之前處理測試，亦使用此田口方法設計並預測最適合之前處理程序。本研究中探討的因子由產學合作廠商提供之歷史數據中選出，分別為氨氮、硝酸鹽、硫酸鹽、氯鹽、硼、鈣、鈉、鎂等，選用L₁₈(2¹x3⁷)之田口直交表代入上述可能干擾因子，進行設計後分析其中砷、硒、鎘與汞的濃度。結果顯示主要的影響離子有硝酸鹽、硫酸鹽，為因應此結果，前處理方法則是選用四種前處理方法及三種水準帶入L₉(3⁴)田口直交表以找出對於FGD廢水中最佳的前處理程序，在最佳前處理條件、品保品管(QA/QC)及偵測極限上的結果皆呈現一個相當可行之結果，本研究成果可提供相關的方法制訂單位、檢測單位及研究單位一重要參考的依據。

Heavy metal quantification in complex matrix is a very challenging task, especially in flue gas desulfurization (FGD) wastewater from steam power generators. Since the discharge standards on arsenic and selenium will be further tightened to lower ppb levels, the interference can cause significant errors on quantification results. This study investigated the interference of major constituents in FGD wastewater on quantification of the aforementioned heavy metals by flame atomic absorption spectrometry (FAAS) coupled with a hydride generator in both synthetic and real FGD wastewater samples. To minimize the numbers of experiments, Taguchi method was employed to determine the most influential ions or ligands based on the interfering extent as well as effective pretreatment steps that can reduce the interference. Eight major ions or ligands were chosen based on historical records, i.e., ammonia nitrogen, nitrate, sulfate, chloride, boron, calcium, sodium, and magnesium. An orthogonal table L₁₈(2¹x3⁷) was used to check the interfering strengths of these eight controlling factors on the quantification of arsenic and selenium. The result showed that influential factors are nitrate and sulfate. Furthermore, in this study, four different pretreatment steps with three different levels were tested using an orthogonal table L₉(3⁴) to define an optimum pretreatment procedure to reduce the interference on quantifying real FGD wastewater samples. As a result, these pretreatment steps were determined and these heavy metals can be detected consistently in a satisfactory recovery range and all QA/QC parameters such as RSD and method detection limits are satisfactory.