https://scholars.lib.ntu.edu.tw/handle/123456789/170653
Title: | 雙伽碼射線量測孔隙介質多相流體問題之探討 The study of multi-phase flow in porous media using dual-gamma ray |
Authors: | 王耀霖 Wang, Yao-Lin |
Keywords: | 非破壞性量測;雙伽碼射線;DNAPL;REV.進入壓力;毛細壓力-飽和度曲線;non-destructive measuring;dual-gamma ray;REV;Pc-S curve | Issue Date: | 2009 | Abstract: | 本研究延續李居正(2006)已建立具有241Am以及137Cs雙伽碼射線之非破壞性飽和度量測設備,設計一維穩態砂柱試驗觀察重質非水相液體 (dense-nonaqueous-phase liquid, DNAPL)在地表下之分布。試驗前以射線量測砂樣時,發現光子數變化大,因射源開孔僅2.8 mm,使量測體積未達「最小代表性體積」(representative elementary volume, REV),造成孔隙率 有著不確定。由伽碼射線衰減原理,分析影響砂箱土壤內二相流體飽和度量測準確性,為光子數、土壤孔隙率以及孔隙率結合飽和度不確定性的變異數等三項的權重組合。測中,以高斯機率密度函數擬合3~4Kev能量解析度觀測值的峰值,為衰減前後的量測值,可降低光子數變異,減少飽和度估計隨機誤差;土壤異質性產生之不確定性遠大於光子數隨機變異,因此衰減係數檢定中,僅對流體進行檢定不加入試驗砂,可得到較小飽和度之95%信心區間;當量測體積約為射源開孔之10倍時可視為達到REV,增加取樣體積可減少孔隙率不確定性。雙射源對於各介質之衰減關係為,241Am對於試驗砂、水和DNAPL介質間之差異分辨率較大,適合鑑別;而 137Cs對於試驗砂、水和DNAPL介質間的差異分辨率很低。 研究以「石英砂」作為孔隙介質,HFE-7100為試驗用DNAPL,利用重力自然排水,取得「水-空氣」流體對「首次排退」的「毛細壓力-飽和度」關係曲線,得到空氣進入壓力,換算DNAPL理論進入壓力;在「水-DNAPL」試驗中觀察水的主排退(DNAPL進入)現象,取得DNAPL進入壓力,且理論和實驗之結果相近,可作為未來二維試驗設計參考,並藉由伽碼射線及視覺化觀察,可更準確知道DNAPL指狀方式進入孔隙介質中達穩態後之分布;在「水-DNAPL-空氣」三相流試驗中,由DNAPL飽和、水殘留初始狀態下水和DANPL同時排退(空氣進入)的曲線,試驗結果在接近地下水位面DNAPL的飽和度約為25%,且有因DNAPL殘留量少,而無法分辨DNAPL的殘餘飽和度之現象。議未來可透過調整試驗砂箱厚度,射源出口大小、保持光子數量大約相同的方式,使取樣體積達到REV,減少因為土壤孔隙率異質性產生之變異性,以期後續研究提高量測之精度。 This study follows Lee (2006) using the Gamma-ray attenuation theory to establish non-destructive measuring equipment. To measure nonaqueous-phase liquid (NAPL) and water contents simultaneously, a collinear dual-energy γ-radiation system is employed. The radiation sources are 241Am and 137Cs. A one-dimensional (1-D) multiphase flow experiment is conducted to investigate the distribution of dense-NAPL (DNAPL) in the subsurface at steady state. Both holder and detector possessed lead shielding with a 2.8-mm-diameter. Before the experiment we find the variation of measuring counts is large when measuring the soil sample. That is the measuring volume does not reach representative elementary volume (REV). So the porosity ( ) has uncertainty. From the Gamma-ray attenuation theory, the uncertainty of measuring saturation comes from counts, porosity and porosity combining saturation. his study uses Gauss probability density function to fit the measurements peak-value. This method can decrease the variation of counts and saturation. The largely variation is porosity. So at attenuation coefficient test, the method only measures the fluid and not includes sand which can get less prediction interval of saturation. When the measuring volume is ten times of one single measuring volume, it may be regarded as REV. Analysis attenuation of dual-gamma source to each material that 241Am can differentiate each material easily and 137Cs can’t distinguish the material.his study uses Quartz sand as porous medium and HFE-7100 as DNAPL. We can get water-air soil experiment data of capillary pressure (Pc) and saturation (S) during primary drainage. Then the measured air entry-pressure can be used to calculate the DNAPL entry-pressure. From water primary drainage (DNAPL imbibition) process within water-DNAPL two-phase fluids, we can get DNAPL entry-pressure. The experimental results are close to the theoretical values. Then we can use it to design 2-D experiment in the future. Through gamma ray measurement and visualization, we can know distribution of the DNAPL accurately. At water-DNAPL-air three-phase fluids, the initial condition is DNAPL-saturated and water-residual. We observe water and DNAPL drainage (air imbibition) at the same time then get Pc-Scurve. The saturation of DNAPL near ground water table is approximately 25%. And the DNAPL residual is few and unable to be distinguished.o sum up, we suggest that modulating the thickness of sand box and the diameter of gamma beam to maintain the quantity of counts. We can reduce the variation of heterogeneity and increase precision by measuring at REV scale. |
URI: | http://ntur.lib.ntu.edu.tw//handle/246246/187905 |
Appears in Collections: | 土木工程學系 |
File | Description | Size | Format | |
---|---|---|---|---|
ntu-98-R95521306-1.pdf | 23.32 kB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.