TOC estimation from logging data using principal component analysis

Abstract

Total organic carbon (TOC) content is one of the most important parameters for characterizing the quality of source rocks and assessing the hydrocarbon-generating potential of shales. The Lucaogou Formation shale reservoirs in the Jimusaer Sag, Junggar Basin, NW China, is characterized by extremely complex lithology and a wide variety of mineral compositions with source rocks mainly consisting of carbonaceous mudstone and dolomitic mudstone. The logging responses of organic matter in the shale reservoirs is quite different from those in conventional reservoirs. Analyses show that the traditional △logR method is not suitable for evaluating the TOC content in the study area. Analysis of the sensitivity characteristics of TOC content to well logs reveals that the TOC content has good correlation with the separation degree of porosity logs. After a dimension reduction processing by the principal component analysis technology, the principal components are determined through correlation analysis of porosity logs. The results show that the TOC values obtained by the new method are in good agreement with that measured by core analysis. The average absolute error of the new method is only 0.555, much less when compared with 1.222 of using traditional △logR method. The proposed method can be used to produce more accurate TOC estimates, thus providing a reliable basis for source rock mapping.