Deep marine carbonate reservoirs are important field of oil and gas exploration and development,and the identification and division of sedimentary microfacies is the basis and key for conducting detailed reservoir description.To establish an automatic identification method for sedimentary microfacies in complex reef beach carbonate reservoirs based on logging data,taking the Changxing formation reef beach reservoir in Yuanba as an example.By combining conventional logging,electrical imaging logging,and geological data,the logging response characteristics of different microfacies are clarified,and an intelligent identification sample set covering the differences in rock physics and logging curve morphology of different sedimentary microfacies is established.A new decision tree method based on Bayesian principle is proposed to address the complex structure and low accuracy of the existing C4.5 decision tree algorithm for missing attribute samples,which is suitable for missing value problems.This method reduces the uncertainty of tree construction and improves operational efficiency.Seven wells have been tested and the recognition coincidence rate is over 90%.The new method enriches and improves the exploration and development technology system of deep marine carbonate reservoirs,providing a reliable basis for using logging data to classify complex reef beach carbonate sedimentary microfacies.
关键词
元坝气田长兴组/沉积微相测井识别/决策树法/碳酸盐岩/改进的决策树法
Key words
Changxing formation of Yuanba gas field/sedimentary microfacies log identification/decision tree/carbonate rock/upgraded decision tree
维普
万方数据