Research And Application Of Low-amplitude Structure Identification And Reservoir Seismic Prediction

With the deepening of exploration research,the proportion of low-amplitude structural oil and gas reservoirs has gradually increased,and the research and application of low-profile structural identification has received increasing attention.Low-profile structural identification generally includes low-profile ridge structures and low-order fracture identification.Therefore,this paper studies the identification of low-amplitude structure and intelligent detection of low-order fracture in view of solving difficulties such as small trap area,difficult identification,reservoir thin and so on,completing the clear depiction of the low-level underground structure,and providing a reliable visualization basis for the exploration and development of unconventional oil and gas.Firstly,in order to reduce the thin layer tuning effect,improve the resolution,and provide more detailed broadband seismic data for the subsequent low-profile structural identification and low-order fault tomographic detection,this paper uses log data with higher vertical resolution to carry out multi-well joint constrained extension frequency processing on the original seismic data.Secondly,for the identification of low-profile structures,the paper uses the combination of trend decomposition method and wavelet decomposition method to enhance the local micro-structural features and make them prominent on the structural plan.Furthermore,for the identification of low-order fractures,the paper uses edge detection as the prior information,and performs all-round intelligent optimization detection on the coherent data body to improve the visibility of small-scale information and achieve the purpose of finely identifying low-order faults.Finally,the three main methods of the thesis are applied to a block in the slope zone of the Orient Basin,and the feasibility of the low-structure automatic identification and the intelligent detection method of fracture is verified by the good identification effect of the actual data,developing a set of automatic identification program for it.