The Study On The Microfacies Distribution Of The Shallow Delta With Well And Seismic Data Based On Intelligent Method

There are three sandstone bodies located in the II oil group of the lower Minghuazhen Formation of the Bozhong 28-2S Oilfield.They are in layer II-1,II-3 and II-4,separately.All of them are regarded as the main objects of hydrocarbon exploration and development.However,the detailed facies characterization is limited by the huge wells spacing and sparse well logging data.Moreover,the water proportion of the produced liquid from the wells in the study area is very high with the low recovery degree.To enhance the hydrocarbon production of this area,the depositional microfacies should be identified with high accuracy.Hence,in this study,a reservoir characterization method based on artificial intelligence model with well logs and seismic data is proposed by us to identify the sandstone bodies in the lower Minghuazhen Formation.The study results can be taken as an important support for the development scheme adjustment.We first confirm the stratigraphic feature of the II oil group by using the seismic –well interpretation.The structure of the study area is generally higher in the north and lower in the west,southwest,and east.The structures of different layers show obvious continuity.The thicknesses of these layers are nearly the same.All the layers are thicker in their north part and thinner in the south part.The thickest one of them is layer II-4while the thinnest one is II-1.We then predict the horizonal spatial distribution of the sandstone bodies by using the support vector machine(SVM)model with the optimized sensitive seismic attributes.The prediction shows that the sandstone bodies located in this formation as a continue slices.The layer II-3 occupied the largest horizonal space while II-1 and II-4 are smaller than it.Guided by the shallow-water delta depositional model,the vertical assemblage modes of each layer and the corresponding microfacies types of each assemblage mode are analyzed through single-well data,and the seismic waveform characteristics corresponding to different vertical assemblage modes of sand bodies are analyzed.The seismic waveform clustering results of the self-organizing neural network algorithm are calibrated with this,and the distribution rules of different types of sedimentary microfacies under the constraints of sand body distribution are described.The II-1 layer and the II-3 layer are typical lobed shallow water delta.The underwater distributary channel sands are distributed contiguously as a whole lobe.Most of them are overlapped in a cut-and-stacked way.Some inter-channel deposition can be found at the edge of channels and the space between channels.Layer II-4 is mainly branch-like shallow water deltas.The main sedimentary body of layer II-4 is closer to the lake shore.The underwater distributary channel sediments are also generated in the form of stripes.The inter-channel sediments are exhibited everywhere in this layer.The river channel is often overlapped on the mouth bar sediments.Also,the sheet sand can be seen in this layer.The artificial intelligence assisted reservoir characterization method based on well logs and seismic data provides the spatial distribution prediction for the sandstone bodies and makes the feature of the microfacies inside more clearly.All these achievements can be taken as an important support for the following hydrocarbon development in this study area.