Reservoir diagenetic facies logging identification of Huagang formation in central-northern of the central anticlinal belt of Xihu sag

Artificial intelligence recognition of diagenetic facies in the area without coring wells based on logging data has become an inevitable trend. The diagenetic facies of Huagang formation reservoir in the north-central anticline zone of Xihu sag are divided by using core casting thin sections and scanning electron microscopy data and six kinds of typical diagenetic facies types are summarized. The corresponding logging responses of different diagenetic facies are obviously different. The results show that the log series of natural gamma ray, compensated neutron, acoustic time difference, deep lateral resistivity, shallow lateral resistivity and density are sensitive curves for diagenetic facies discrimination. Comparing the slice data of cores, it is shown that the probabilistic neural network model can accurately identify the diagenetic facies of the uncored interval, and the recognition rate can reach more than 90%.