The Study On Reservoir Architecture Of Well Logging And Seismic Data Combination Of The Upper Guantao Formation In The Central Area Of Chengdao Oilfield

The upper Guantao formation in the central area of Chengdao oilfield is fluvial facies deposition,with rapid transition between different reservoir facies and complex sedimentary evolution.It is difficult to identify the stacked relationships between sand bodies between wells.In addition,there are many directional wells,which results great differences in well spacing,so it is difficult to predict the reservoirs.Therefore,performing well seismic combined with reservoir prediction and fine characterization of reservoir architecture is of great significance to oilfield development and provides necessary geological basis for further efficient development of oilfield.With the guidence of fluvial sedimentary architecture models,this thesis uses intelligent seismic attribute analysis technology,through frequency band optimization,attribute optimization and intelligent attribute fusion,the quantitative distribution of single-layer sand body in the study area is predicted,and the prediction accuracy of sand body is improved.The artificial intelligence inversion technology is applied to characterize the superposition relationship of sand bodies,and summarize the reservoir distribution characteristics and distribution style in the study area.On the basis of reservoir prediction,types of architecture units and their logging and seismic responses in the study area are clarified.The idea of"combination of well and seismic,mode guidance and hierarchical constraints"is adopted to successively complete the fine characterization of channel belt,identification of point bar and internal architecture anatomy of point bar,and analyze the distribution characteristics of each architecture unit.The results show that:（1）The fusion method of multiple frequency-decomposed attributes realizes the accurate prediction of sand bodies with different thickness,the attribute fusion method to reduce the interference of neighboring zones solves the problem of complex reservoir prediction,and the intelligent seismic inversion method improves the prediction accuracy of superimposed sand bodies and thin-layer sand bodies;（2）The application conditions of the reservoir prediction method of well logging and seismic data combination are different,according to the characteristics of sand body development,different methods are selected for reservoir prediction,so as to improve the accuracy of reservoir prediction;（3）In the main layer of the study area,Ng4₂¹single layer develops a north-south extending banded channel sand body,and Ng5₂²single layer develops a northwest-southeast extending channel sand body,which is distributed in a banded continuous manner;（4）The study area is a meandering river deposit,which mainly develops channels,overbank sand and flood plains.The logging response characteristics of each architecture unit in the study area are summarized,and the seismic attribute response characteristics of each architecture unit and the seismic inversion response characteristics of different sand body superposition modes are summarized;（5）The well seismic combined with multi-level architecture anatomy method is explored.The single-layer channel belt of Ng4₂¹is distributed in a serpentine shape,and two narrow strip distributary channels are developed in the middle.A total of7 single meander zone point bars are identified in the main channel;The Ng5₂²single-layer channel belt is distributed in a wide strip.A total of 14 point bars are identified in the channel,and the point bars are distributed in a"wide strip"to"narrow strip";Try to dissect the internal architecture of the typical point bar.Eight lateral-accretion sand bodies are developed in the typical point bar of Ng4₂¹single layer.