雲林地區地層下陷特性調查監測與模擬

Abstract

雲林縣因長年超抽地下水導致部分地層下陷，嚴重威脅其内包括高速鐵路 在内之公共工程安全，使用數值模型配合現場監測以了解下陷機制並預測 地層下陷分佈及趨勢乃是必須。此區域位於濁水溪沖積扇南翼，土層厚度 在300 m以上，地層包含4個含水（粉砂土層）層和3個阻水（黏土）層， 其中粉/砂土含有大量高壓縮性雲母礦物，壓縮性遠高於常見之石英砂土， 使此區域分層沉陷分佈與一般黏土/砂土互層結構不同，為確認沉陷機制， 分層沉陷量測與考慮此效應之數值模擬為必要之作為。現有監測地層下陷 是使用Sondex系統，將磁感應環錨定於鑽孔内不同深度，使用銦鋼尺連 接探測棒以手動方式放入孔内偵測磁感應環深度，來判定地層下陷隨深度 變化關係，使用此半世紀以上之手動技術無法對地層沈陷做即時監測。現 有壓縮土層内應變估計平均在1%以下。未來如果控制地下水抽取，此應 變量可以更小，屬於土壤力學所考慮之中小應變，是土壤從彈性轉變為塑 性之過渡階段。現有數值模型大多從地下水模型來考慮，沒有將上述土壤 特性反應在數值模型中，其正確性存疑。本計晝之研究内容包括：選定監 測地點進行鑽探與非擾動取樣，現地土壤透水性試驗(子計晝三)與剪力波 速量測（子計晝二、四），使用非擾動黏土與粉/砂土試體進行一系列中高 壓單向度壓密(子計晝一)與中高圍壓下K。壓密三軸試驗(子計晝二），決定 深層土壤在中高壓與中小應變情況下壓縮與力學特性。以光纖感測技術， 安裝可在同一鑽孔内做自動化分層水壓、鑽孔傾斜與軸向上浮/沈陷監測系 統（子計晝三）。使用時頻分析觀念分析地層水壓與沈陷分佈監測數據（子 計晝四），確認地層下陷與時間/空間之關係。使用優化觀念，根據室内與 現地試驗結果，建立數值模型輸入參數（子計晝五）。根據上述輸入參數 與現地監測所得結果，對現有地層下陷數值模型做必要之修正與驗證（子 計晝六）。最後整合研究結果，提出地盤沉陷因應對策。

As a result of excessive ground water pumping, parts of Yun-Lin County are settling. The ground subsidence seriously threatens the safety of infrastructures that include the high speed rail, passing through this area. Numerical simulations coupling with field monitoring are imperative for the understanding of the mechanism and predicting the trends of ground subsidence in this area. Yun-Lin is located in the south wing of the Zhuo-Shwei river delta. The depth of alluvial soil is believed to be exceeding 300m in depth. The soil deposit consists of four aquifers (silts and sands) and three aquitards (clays). The silt/sand soils contain significant amounts of muscovite, a highly compressible mineral, making them much more compressible than conventional quartz sands. Because of these reasons, the distribution of consolidation among soil layers is significantly different from that of conventional interlayered sand/clay soil deposits. To ascertain the mechanisms of the layered soil consolidation, it is important that the field monitoring that measures individual soil layer settlement as well as numerical simulation consider the effects these soil characteristics. The ground subsidence profile is currently monitored using the Sondex system. A series of magnetic rings are installed at various depths in a borehole. The depth location of the magnetic rings was measured by lowering a sensing probe at the end of a graded indium steel tape to the borehole by hand. The half-century old technology is not able to provide the readings on a real-time basis. The average field soil compression is expected to be less than 1% in its axial strain. If ground water pumping is controlled in the future, the strains can be even smaller. Under such conditions, soil behaves in a transient category that spans from elastic to plastic. The currently available numerical simulations consider ground subsidence mainly from hydrological point of view, and rarely incorporate the elastic-plastic soil behaviors. Hence, their accuracy can be questionable. This collaborative research project will perform drilling to perform undisturbed soil sampling (sub-project 3) and field soil testing that includes permeability (sub-projects 2 and 4) and shear wave velocity measurements at the selected test site. A series of mid to high stress one-dimensional consolidation (sub-project 1) and Ko consolidated triaxial tests (sub-project 2) will be performed on the undisturbed silty clay and silt/sand soils to determine their compressibility. The research team will install a series of fiber optic sensors capable of measuring pore water pressure, ground subsidence/heaving and inclination profile of the sensors (sub-project 3).

Using the time frequency concept, the correlation among pore water pressure, ground movement and depth will be analyzed (sub-project 4). The testing and monitoring results will be coupled into the numerical modeling and their impacts evaluated (sub-projects 5 and 6). Based on the overall evaluation of the results from this research, recommendations to mitigate ground subsidence in the region will be made.