主成分分析在塔中地区奥陶系鹰山组碳酸盐岩岩性识别中的应用

Application of Principal Component Analysis in Carbonate Lithology Identification of the Ordovician Yingshan Formation in Tazhong Area

塔中地区奥陶系碳酸盐岩岩性复杂,鹰山组取心资料较少,基于常规测井分析方法难以很好的识别。为了解决这一问题,选取鹰山组岩心薄片、测井和录井等地质资料较为齐全的7口井作为标准井,进而优选出对岩性变化反应敏感的自然伽马(GR)、补偿中子(CNL)、密度(DEN)、冲洗带电阻率(RXO)和声波时差(AC)等5条测井曲线。在此基础上,利用主成分分析法构建了Y1—Y5共5个综合变量,其中2个变量Y1和Y2的方差贡献率占总贡献率的85.83%,可以作为2个主成分有效地反映原有5个变量的有效信息。根据Y1和Y2主成分做交会图分析,可以有效地识别出颗粒灰岩、云岩以及藻灰岩。利用该特点,对塔中地区奥陶系鹰山组地质资料不全井的碳酸盐岩岩性进行了准确标定。与常规测井分析法相比,主成分分析法能够整合更多的测井参数,进而提供更好的能适应于碳酸盐岩岩性的测井定量识别方法。

It is difficult to recognize complex lithology of the Ordovician carbonate in Tazhong area by conventional logging analysis with too little core data of Yingshan Formation.To solve this problem,seven wells with complete core slice,borehole and logging data of the Yingshan Formation were selected as standard wells.We chose five logging curves of GR,CNL,DEN,RXO,AC which have sensitive reaction to lithology change.Using the principal component analysis,we constructed five comprehensive variables including Y1-Y5.Among these variables,the variance proportion of Y1 and Y2accounts for 85.83% of the total contribution ratio,so they can replace the original five variables efficiently as two principal components.Through the analysis of crossplot of Y1 and Y2,we can identify the grain stone,dolomite and algal limestone effectively.We depicted the carbonate lithology of non-complete wells from the Ordovician Yingshan Formation in Tazhong area precisely.Compared with the conventional logging analysis,the principal component analysis can integrate more parameters and adapt to the quantitative recognition of carbonate lithology.