Lithology identification in basement rock reservoir is one of the most important basic work for reservoir evaluation. The identification results can directly influence the evaluation of reservoir physical property, oil-bearing property and the identification of effective reservoir.
The paper takes buried hills in Bongor basin as an example to analyze the characteristics of lithologic chemical components and mineral compositions by using core observation, slice identification, geochemical analysis. Based on lithology characteristics, the classification criterion of basement rock was determined. In addition, the log response characteristics were summarized and the corresponding logging identification method was established. The results show that the basement rocks of the research area composed of metamorphic rocks and magmatic rocks. With the decrease of the content of light minerals, the contents of Si and K were decreasing and the contents of Fe, Al, Ti, Mg and Ca were increasing. According to the mineral compositions and contents, the lithologies were divided into 2 categories and 9 subclasses. There were brecciated structure, clasfic structure, gneissic structure, banded structure, net veined structure, augen structure, massive structure and intrusive structure in electrical imaging logging. Meanwhile, logging response characteristics of basement rocks were divided into 6 types combined with the difference of density curve and compensated neutron curve. The established cross plots and summarized characteristics of logging curves and response values could provide a basis for lithology identification. The machine learning approach (SVM) can improve the accuracy of lithological identification.
