岩石熱導率對於油氣探勘以及地下熱流評估皆為重要的參數之一。由於傳統量測岩石熱導率方法不易進行，一旦缺乏岩心資料則岩石熱導率就無從得知。因此，利用井測資料所得的岩石物理參數來估算熱導率就成為較易實施且有效率的方法。為建立井測資料與岩石熱導率間的經驗關係式，本研究取用位於台灣西南外海之台南盆地以及雲林湖山水庫沉積岩，年代從中生代到新生代地層。本研究方法分成三步驟，首先在實驗室量測岩石樣本的基質密度、孔隙率、P波慢速與熱導率等物理性質，結果顯示熱導率與其它三種皆為負相關。第二步利用這些量測值與多重線性迴歸方法 (Multiple Linear Regression)， 建立熱導率與三種物理性質的經驗數學式。最後則是利用台南盆地六口研究井的井測結果所得之岩石參數，代入經驗式後計算地下連續性之岩石熱導率變化。此經驗式可以應用於估算有鑽井資料的沉積岩(砂岩、砂泥互層、泥岩)熱導率， 估算結果之準確性落在20%信賴區間。;Rock thermal conductivities are one of the essential parameters for determining crustal heat flows. If drill cores are not retrieved from boreholes, no direct laboratory measurements of thermal conductivity can be made for subsurface rocks and sediments. In this case, petrophysical properties obtained from well logging can be used to determine thermal conductivities. In this study, I firstly establish empirical relations between thermal conductivities and three petrophysical properties: matrix density, porosity, and sonic slowness, from well-log measurements at the same depth of dry rock samples. We then apply these empirical equations to borehole well-log data to obtain a continuous thermal conductivity profile for some desired rock formations. Rock samples from 6 boreholes of Cretaceous to Miocene sandstones and shales in the Tainan Basin, offshore SW Taiwan, and Pleistocene sandstones at the Hushan Reservoir, Yunlin County, are measured for determining thermal conductivity (λ), matrix density (ρma), porosity (φt) and sonic slowness (Sp). A regression analysis of the measurements yields a best fit (e.g., rms of 0.08W(mK)-1) of a linear trend line, and this trend is then applied to well-log data with continuous measurements of matrix density, porosity and sonic slowness to obtain a continuous thermal conductivity profile for rocks drilled in a particular borehole. The proposed empirical linear relations between thermal conductivity and matrix density, porosity, and sonic slowness can be readily used to obtain thermal conductivities from well-log data.
