| 摘要: | 近三十年來人工溼地（constructed wetlands or treatment wetlands）已成為一種有效率及低成本的水污染防治生態工法。槽體不透水性或保水性為人工濕地系統最重要的設計及施工項目之一，其目的乃藉由襯裏層(liners)的施作來防止地下水層受到廢水的污染，或防止地下水上升而湧入人工溼地。目前，國內尚缺乏一套因地制宜的人工溼地不透水襯底層之最適工法的評估作業流程，以提供人工溼地設計及建造之參考。本論文採自南台灣十三處不同場址的現地土壤，於實驗室中測量土壤的不同性質，包括：粒徑大小及分佈、質地性質及滲透係數，並利用土壤夯實試驗及低透水質材的添加試驗，以開發符合永續性的人工溼地不透水襯底施工方法，及建立將現地土壤融入溼地防滲工法的標準作業流程。
結果顯示十三處土壤的黏粒含量介於10~27%、坋粒含量介於13~67%、砂粒含量10~75%，有9個屬於坋質壤土、1個屬於壤土、3個屬於砂質壤土。月世界泥池之土壤滲透係數最低為6.40×10-7 cm/sec，二行村及黃金海岸的土壤滲透係數最高，分別為5.7×10-4及1.96×10-3㎝/s。土壤的滲透係數與有效粒徑(D20、D30、D40、或D50)呈幾何正相關，滲透係數隨有效粒徑增加而指數增加。滲透係數(K)與D30的相關性較高，相關方程式為K=0.01×D301.21945 (R2=0.72)，可作為其它地區土壤推估滲透係數的參考。本研究亦利用多元迴歸法將不同土壤之滲透係數與質地三成份同時建立相關經驗式，結果顯示滲透係數與三成份同時進行迴歸得到的經驗式其相關性最好(R2=0.659)，而滲透係數與兩成份或單成分迴歸所得到的經驗式其相關性較差。此多元回歸方程式
亦可應用於推估土壤的滲透係數。
選擇五種土壤進行夯實試驗，結果顯示土壤夯實時水份添加量及敲擊次數為影響夯實結果的關鍵因素。水分添加量為13.2~14.6%進行夯實時，在相同敲擊次數下所獲得的夯實土壤樣本，其滲透係數均明顯低於水分添加量僅為9.0~9.6%者。然而，水分添加量提高到17.3~18.9%，經夯實後的土壤滲透係數並無法進一步有效降低。另外，在相同水分添加量下，經夯實後的土壤滲透係數隨敲擊次數的增加而下降，然而敲擊次數在30次以上，增加敲擊次數對滲透係數的下降不明顯。低透水質材的添加試驗結果顯示，在足夠的水分添加量之下，土壤中添加少量皂土(2~5%)，相較沒有添加皂土的對照組土壤，其滲透係數降低約30%以上。並可使土壤滲透係數降低到＜1×10-6 cm/sec之標準。實驗亦證實可使用市上販售且易獲得、但所需添加量少的皂土，取代經天然開採之高黏粒含量、但是所需添加量較高的土壤，來達成降低土壤滲透性的結果，此法可作為人工溼地不透水襯底層施工上可供選擇且有效的工法。本研究並建立了人工溼地不透水襯底層之最適工法的評估作業流程，提供人工溼地選址及規劃設計之參考。
For the past three decades, constructed wetlands have become an effictive and low-cost ecological technology for water pollution control. Sealing the basin by using the liners is one of the important design and construction works of constructed wetlands. Its aim is to prevent the contamination of groundwater or prevent groundwater from infiltrating into the constructed wetlands. At the present, there is still a lack of suitbable evaluation workflow that is in line with local conditions for sealing the basins of constructed wetlands, which can be used as reference for the design and construction of constructed wetlands. This thesis collected on-site soil from 13 different sites in sourthern Taiwan and tested their various proterties in the lab, including: particle size distribution, effective size, soil texture, and hydraulic conductivity. By using soil compaction experiment and the experiment of adding material of low hydraulic conductivity, an effective method to line the constructed wetlands will be developed, and a standard workflow for evaluating a suitable lining method through considering on-site soil properties will be established.
The result of soil texture analysis shows that clay content of the soil samples from 13 different sites was about 10~27%, silt content 13~67% and sand content 10~75%. Nine of the soil samples was determined to be silt loam, one to be loam, and three to be sand loam. The hydraulic conductivity of soil in the Moon World mud pool was the lowest, 6.40×10-7 cm/sec; the hydraulic conductivity of soil in the Erh Hang village and Golden Seacoast is the highest, 5.7×10-4 and 1.96×10-3㎝/s, respectively. The hydraulic conductivity of soil was found to be exponentially proportional to effective sizes of soil grains (D20, D30, D40 or D50). The hydraulic conductivity (K) has higher corelation with D30 than other effective sizes, the corelation equation is K=0.01×D301.21945 (R2=0.72), which can be considered as an useful emperical model for estimating hydraulic conductivity of soil from other areas.
This study also examined the corelation between the hydraulic conductivity of soils and the three elements contents of soil texture properties by using the way of multi-variables regression method. The result shows that the regression equation with higher corelation coefficient was obtained when hydraulic conductivity was related with all the three elements contents simultaneously (R2=0.659), while regression equation with lower corelation coefficient was found, if hydraulic conductivity was related with contents of two elements or single element. The multiple variable regression equotion is:
Which can aslo be applied to estimate the hydraulic conductivity of soils.
Soil samples obtained from five various sites were used to conduct the soil compaction experiment. The result shows that the amount of water addition and tapping times are the key factors that affected soil compaction result, when compacting soil. When compacting at water addition of 13.2~14.6%, under the same tapping times, the hydraulic conductivity of compaction soil sample is obviously lower than at water addition of 9.0~9.6%. But when water addition amount rose to 17.3~18.9%, the hydraulic conductivity of compacted soil did not further effectively reduce. In addition, under the same amount of water addition, the hydraulic conductivity of compacted soil decreased with the increase of tapping times; however, when tapping were over 30 times, the increase of tapping times had no obvious effect on the decrease of hydraulic conductivity. The addition experiment of low hydraulic conductivity material shows that with enough water addition, compared with the soil group without adding bentonite, the hydraulic conductivity of the soil group that with little addition of bentonite(2~5%) decreased about more than 30%, and achieved to a criterion ＜1×10-6 cm/sec. The experiment also proves that using commercial and easily available material, bentonite, which needs only small amount addition, can replace the natural soil of high clay content, which needs comparably high amount addition, to reach the result of decreasing hydraulic conductivity of soil. This method can be used as alternative and effective construction method for the liners of constructed wetlands. This study finally established a workflow for evaluating the best construction method of liners of constructed wetlands, which could provide a guideline for the site selection, plan and design of constructed wetlands. |
