Application of Neural Network Automatic Identification of Sedimentary Microfacies in Thin Reservoirs of Delta Exo-front Facies

Sedimentary microfacies is important in the formation and spatial distribution of the oil reservoirs. Recognize the sedimentary microfacies again in the middle and late stage of development is of great significance to explore the residual oil and enhance oil recovery. Traditional recognition of sedimentary microfacies has a strong subjectivity and consumes a lot of time. Therefore, it is necessary to use neural network to realize automatic identification of microfacies in thin reservoirs. This article establishes the logging facies mode of sedimentary microfacies by analyzing sealed coring data in A block. Optimize four logging curves GR, RMN, RMG and AC and combine with the constraints such as the reservoir parameters and the petrophysics to establish the neural network model for different oil reservoirs. This article established 7 delta front facies sand body logging facies. According to the characteristics of the logging facies and reservoirs, the neural network model is established to automatically identify the sedimentary microfacies. Comparing the automatic identification results with the artificial identification results of thin reservoirs in A block, the identification precision can reach more than 70%. Using neural network to identify the thin reservoirs can improve the work efficiency greatly and save lots of manpower. This method can be used in more than 60 blocks in the region, which has high promotional significance. The sedimentary microfacies plane diagrams drawed by using the automatic identification results can provide a strong geological foundation for the design of the development plans.