基于概率核地震属性融合的砂体预测方法研究

Research on sand body prediction method based on seismic attributes integrating of probabilistic kernel

随着油气田开发程度越来越高,勘探难度越来越大,如东部的老油田已经进入开发的后期,如何识别薄层砂体是非常重要的工作之一,解决这些难题这势必需要更先进的技术.地震属性能够很好的反映砂体横向展布特征,但是单一属性无法定量预测砂体厚度,而多属性之间又存在多解性,因此有必要提炼地震属性之间的共同点,将地震属性进行信息融合,形成新的融合属性.针对这一问题,本文提出首先利用高频谐波提高地震数据的分辨率,在此基础上着重研究基于概率核的地震属性融合方法,融合了几种常见的地震属性,并结合波阻抗反演方法,预测了N873区块沙三6-3小层砂体厚度.结果显示该方法能够很好的反映砂体横向展布特征,避免了地震属性多解性问题,为提高砂体预测的精度,提供了新的思路和方法.

With the increasing development of oil and gas fields,exploration is becoming difficulty.The old oil fields have entered the late stage of development in the east china.How to identify thickness sand body is one of the more important tasks.Solving these problems will inevitably require more advanced technologies.Seismic attributes can reflect the lateral distribution of sand body,but the single attribute cannot quantitatively predict the thickness of sand body,and multiple attributes existence multi-results.Therefore,it is necessary to refine the common points between seismic attributes.Integrating seismic attributes and forming new attributes.In response to this problem,this paper proposes to use high-frequency harmonics to improve the resolution of seismic data.Based on this,it focuses on the study of seismic attribute integrating methods based on probability kernels,integrating several conventional seismic attributes and combines impedance inversion method to predict the thickness of sand body on 6-3 layer on Sha-2 in N873 work area.The results show that this method can reflect the lateral distribution characteristics of sand body well,avoid the multi-results problem of seismic attributes,and provide new ideas and methods for improving the prediction accuracy of sand body.