一种基于地震振幅属性预测砂岩厚度的误差校正方法OA

With the expansion of hydrocarbon exploration and development targets toward subtle reservoirs(e.g.,lithologic and stratigraphic reservoirs),the accurate prediction of sandstone reservoir thickness has become a critical link in hydrocarbon exploration and development,as well as one of the core objectives in this field.After decades of technological advancement and practical application in the industry,three major categories of sandstone reservoir research techniques based on seismic data have been gradually established,namely seismic attribute analysis,seismic inversion,and AVO analysis.In recent years,with the emergence of machine learning technology,researchers have also begun to explore the innovative application of artificial intelligence(AI)technology in reservoir prediction.However,as one of the classic prediction techniques,the prediction of sandstone reservoir distribution using seismic amplitude attributes is still widely adopted and demonstrates significant advantages under specific geological conditions,such as channel sand bodies.However,traditional prediction methods based on seismic amplitude attributes have inherent limitations:specifically,they can only predict reservoirs near the tuning thickness,but exhibit certain or even substantial errors in predicting thin or thick reservoirs that deviate from the tuning thickness,particularly for thin sandstone reservoirs.Through research,this study proposes an error correction method that first calculates the difference between the sandstone thickness predicted by seismic amplitude attributes and the actual drilled thickness from multiple wells,then establishes an error function based on this difference and the actual drilled thickness,and finally applies this error function to correct the seismic amplitude attribute-based prediction results.Application results with actual data indicate that this method can improve the prediction accuracy of thin sandstone reservoir thickness,thereby providing a more reliable basis for the optimization of exploration targets and the deployment of development wells.