河流相储层结构地震响应分析与预测

摘要

河流相储层纵向多期叠置切割,横向变化快,具有较强的空间非均质性.受限于海上大井距与地震资料分辨率,河流相储层描述与结构表征存在不确定性.本文针对河流相储层砂体发育特征,构建了反映砂体厚度变化、多期叠置与夹层发育的四类储层结构模式,基于地震正演分析了四类结构模式的地震响应特征与敏感属性,提出了河流相储层结构样式滑块模型,并建立了反映河流相储层结构主因素变化的砂岩厚度与泥质夹层数量及夹层空间位置变化的地震敏感属性响应模板.最后,利用概率神经网络算法实现了模型数据与实际Q油田储层结构预测,选取29口井作为检验井,储层结构预测的吻合率为86％.本文研究可为油田开发的井位设计与优化提供有效的技术支持.%Fluvial reservoir shows characteristics of vertical superposition and erosion,rapid lateral variation and strong spatial heterogeneity.Limited by large well spacing and insufficient seismic resolution,the fluvial reservoir description and architecture characterization is of some uncertainty.This paper,based on the features of fluvial reservoir,built four kinds of reservoir architecture models which can reflect sandbody thickness change,superposition relationship,and shale interlayer distribution,analyzed seismic responses and sensitive attributes of these models based on forward modeling,and proposed the fluvial reservoir architecture sliding block model.Furthermore,this paper established the sensitive seismic attributes response template reflecting thickness change of sandbody and shale interlayer quantity and distribution.The reservoir architecture for model data and the real data in Q oilfield is predicted using the probabilistic neural network algorithm.29 wells are selected as test wells,and it's demonstrated that the matching rate between the reservoir architecture prediction and drilled information is as high as 86％.This paper provides effective technical support for the design and optimization of oilfield development wells.