| 摘要: | 台灣車籠埔鑽井計畫於大坑鑽掘了兩口兩公里的科學鑽探井,並於井孔中500 至1900公尺處施測了地球物理井下測量。為了標定剪力區的位置,我們檢視了波速 資料和密度和電阻率等資料,發現剪力區內的波速和電阻率皆有降低的趨勢。而在 1111公尺的剪力區,藉由其波速和密度的最大變化。界定為最有可能和1999年集集 大地震所造成的車籠埔斷層於錦水頁岩內的滑動相關連。而因應力變化所造成的井 壁崩蹋可由井壁影像紀錄中觀測得知,依此資料可確認其井下平均應力場的方向為 東南155度向西北, 和台灣主要的大地應力方向平行。而雙偶聲波井測的測量結果 得知最大的應力方向變化發生於1110公尺處,也和井壁影像紀錄相符合。 井測的資料顯示了在靠近斷層帶的應力方向,和大地應力方向相比,有將近 90度的旋轉。我們使用水力破裂實驗加以測量井孔中不同位置的最小水平主軸應力 的量值。藉由我們所建構的應力旋轉模型,模擬地震發生前和地震發生後的應力規 模和方向。模擬的結果得知,地震前之水平主軸應力量值約略相同(45Mpa),而遠 大於垂直應力(26MPa)。而地震後最大水平應力方向有將近90度的顯著旋轉,造成 最大水平主軸應力和最小主軸應力互換。而低摩擦係數的參數設定和完全的應力釋 放也解釋了車籠埔斷層在我們模擬的區間範圍內為一弱斷層。 而在井孔的幾何參數上,斷層帶上的摩擦係數也是影響應力規模的重要因素之 一。在模型中得知摩擦係數0.4遠低於Byerlee’s law的範圍(0.6~1)。低摩擦係數的 發現也暗示我們在斷層錯動時可能存在高孔隙水壓並潤滑斷層的動態破裂過程,也 因此造成車籠埔斷層北端有較高的錯動量。 而在考慮其區域尺度的應力影響(500公尺x500公尺),以及近乎完全的應力釋 放下,才會得到符合裂隙閉鎖試驗的結果。而此結果也可解釋由於破裂後最大水 平應力90度的方向轉變,車籠埔斷層由主要的南北破裂轉成東西向的破裂。The Taiwan Chelungpu-fault Drilling Project (TCDP) drilled a 2-km-deep research borehole in Dakeng, Taiwan. Geophysical logs of the TCDP were carried out over depths of 500–1900 m in two boreholes. In order to identify the shear zones, a shear zone at a depth of 1110 meters is interpreted to be the Chelungpu fault, located within the Chunshui shale. Stress-induced borehole breakouts were observed over nearly the entire length of the wellbore from image logs (FMI). These data show an overall stress direction (~N115°E) that is essentially parallel to the regional stress field and parallel to the tectonic stress direction. The Dipole Sonic logs (DSI) also analyzed in this study, the data shows that the most dislocation of fast shear azimuth is close to the depth 1110 meters and consistence with the borehole breakout rotation. The logging data show that near the fault, the azimuth of the maximum horizontal principal stress (SHMAX) changes by about 90o from the regional tectonic stress direction (N130oE). Hydraulic fracturing tests were used to determine the magnitude of the minimum principal stress (S3) at multiple depths. Through dislocation modeling, we simulated the abrupt stress rotation observed in the image logs at the depth of the Chelungpu Fault. In addition, the modeling indicates that the magnitudes of the minimum horizontal principal stress Shmin changed markedly during the earthquake. In order for the co-seismic stress changes to result in a ~90o stress rotation near the fault, the state of stress prior to the earthquake had to have been a reverse faulting stress regime (as expected), but with SHMAX ≈ Shmin >>Sv. The modeling in the stresses with low frictional coefficient and a near complete stressdrop after the earthquake suggests a weak Chelungpu fault, at least in the northern part of the fault. |
