利用多重水力試驗方法評估場址尺度非受壓含水層異質性特徵;Systematic assessment of field-scale unconfined aquifer heterogeneity using multiple hydraulic test methods

NCUIR > College of Earth Sciences > Graduate Institute of Applied Geology > Electronic Thesis & Dissertation > Item 987654321/81761

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/81761

Title:	利用多重水力試驗方法評估場址尺度非受壓含水層異質性特徵;Systematic assessment of field-scale unconfined aquifer heterogeneity using multiple hydraulic test methods
Authors:	陳芝儀;Chen, Chih-Yi
Contributors:	應用地質研究所
Keywords:	抽水試驗;水力剖面掃描;多深度微水試驗;流通係數;水力傳導係數;異質性;Pumping test;Hydraulic tomography;Multi-level slug test;Transmissivity;Hydraulic conductivity;Heterogeneity
Date:	2019-07-25
Issue Date:	2019-09-03 16:59:53 (UTC+8)
Publisher:	國立中央大學
Abstract:	精確又有效率地下水污染整治仰賴對場址含水層特性之了解。本研究主要目的是評估現地場址尺度下，各類水力試驗及分析模式的場址參數調查差異，從分析結果獲得一般化的使用原則。本研究試驗場址位於桃園市觀音區，為非受壓含水層(unconfined aquifer)，在現地19口井執行跨孔式抽水試驗(cross-hole pumping test)和多深度微水試驗(multi-level slug test)，取跨孔式抽水試驗當中抽水井洩降資料，作為單井抽水試驗，以Neuman解析解進行分析；取其中一抽水井與其相對應的觀測井之洩降資料，作為複井抽水試驗；本研究同時將所有抽水井、觀測井資料一併代入SSLE(sequential successive linear estimator)數值模式，則視為水力剖面掃描(hydraulic tomography)；多深度微水試驗則透過Hvorslev解析解分析，並利用克利金內插得到場址三維水力傳導係數(hydraulic conductivity, K)分布。透過計算得到現地場址材料之平均流通係數(transmissivity, T)值為66.81m2/day，變異數為0.067，材料之相關長度為3.3m。以此統計結構代入SSLE模式當中作為反推估的參數。結果顯示四種水力試驗能得到相同級數(order)之K值，在資料共享的狀況下，單井、複井抽水試驗以及水力剖面掃描能反推出T值之趨勢，然對應到多深度微水試驗，分布趨勢雖不盡相同，但單井抽水試驗推得T值換算成K值之平均為7.62E-05 m/sec，落在多深度微水試驗推得之K(z)區間內，符合過往文獻提供的一般化材料特性。本研究場址主要材料為卵礫石，各式水力試驗推得之K值雖未落直接落於礫石合理的K值範圍內，但現地地質材料尚含有粉土等細顆粒，因此K值小於礫石之合理範圍實屬合理。研究結果顯示在井數量許可下，水力剖面掃描最能描繪含水層異質性的分布狀況，然而本場址尚屬均質，建議執行單井、複井抽水試驗即可；而多深度微水試驗顯示各井垂直向變異程度很小，建議執行傳統微水試驗即可。本研究亦於數值模式當中針對影響地質材料異質性之參數如變異數、相關長度等進行敏感度測試，針對數值模式的生成場和推估場進行誤差計算；其結果顯示在現地場址異質性條件下，在數值模式當中誤差會落在0.0016~0.0543m2/min之間，相差約一個級數，但是礫石的K值相差至兩個級數都還算合理，因此推測數值模式在現地場址異質性條件下之反推結果屬於合理範圍。;Accurate and efficient remediation policies highly rely on the understanding of hydrogeological conditions at sites. This study aims to conduct assessment of aquifer heterogeneity using multiple hydraulic tests at the same well field, comparing the estimations by different data analysis methods, and then carrying out the insight into the use of hydraulic surveys at the field-scale site. Cross-hole pumping tests and Multi-level slug tests(MLST) are conducted at the well field. The well field has 19 installed wells, including five 4’’ wells that are considered to be the pumping wells. The pumping and slug tests produce multiple types of observations from well field. The single- and multiple-well pumping tests are analyzed with Neuman(1975) model. All the cross-hole hydraulic test data are involved in the sequential successive linear estimator(SSLE) model to estimate the transmissivity(T) distribution. Drawdown data from MLSTs are analyzed with Hvorslev(1951) to obtain the hydraulic conductivity(K) and the results of MLSTs are interpolated with the Kriging method to obtain 3D K distribution. Results of the single-well hydraulic tests show that the mean T value is 66.81m2/day, the variance of ln(T) is 0.067, and results of the MLSTs show that the correlation length of the T field is 3.3m. These results are the input parameters in SSLE for estimating 2D T at the well field. Results show that the K values from four different hydraulic tests are in the same order. Under the same tests data source, the patterns of T heterogeneities are similar for single-well pumping tests, multi-well pumping tests, and SSLE hydraulic tomography. The MLST shows the patterns slightly different from other pumping tests. The results show that the average K value from the single-well pumping tests (7.62E-05 m/sec) is between the minimum and the maximum K(z) values obtained from MLST. The well field is mainly composed of gravel and embedded with fine sand, loam, and silt layers. The obtained K values at the site show relatively smaller than typical K values obtained from previous studies. Results also show that the SSLE can provide the most detailed information of the aquifer heterogeneities. However, the study site is relatively homogeneous. The single and multi-well pumping tests might be sufficient to resolve the aquifer properties at the site. Additionally, the MLSTs show that the lnK variance in the vertical direction is very small, indicating that the depth-averaged slug test can be used to for the well hydraulic test. The sensitivity analysis of SSLE model shows that the T estimated errors for the site is between 0.0016 and 0.0543 m2/min, which is relatively small.
Appears in Collections:	[Graduate Institute of Applied Geology] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	275	View/Open

社群 sharing

Loading...