论文详情
储层孔隙结构刻画是预测有利储层的关键。伊拉克M油田发育礁滩相碳酸盐岩储层,受成岩作用影响,储层孔隙结构复杂,具有相同孔隙度的储层渗透率存在明显差异。核磁共振测井及取心井岩心资料分析结果表明深、浅侧向电阻率差异与储层的孔隙结构具有相关性,基于电阻率与孔隙度参数构建了储层孔隙因子,提出了地震多属性孔隙因子参数反演方法,其算法实现分三步:①利用波阻抗反演识别储层与基质;②应用专家优选与自动优化组合方法进行孔隙因子参数敏感地震属性优选;③利用概率神经网络算法对优选的地震属性进行地震多属性孔隙因子参数反演。实际应用效果验证了该方法的有效性,有利储层预测结果为油田开发井位部署提供了支持。
Characterization of reservoir pores is essential for the prediction of a high-permeability reservoir.M oilfield in Iraq is a carbonate oilfield with developed reef-shoal facies;its reservoir exhibits strong heterogeneity caused by diagenesis.A reservoir with similar porosity can show different permeability.Analysis of NMR logging and core data demonstrates that the resistivity difference between deep and shallow lateral logging is related to reservoir pore structure.A new parameter named pore-sensitive factor has been formulated,and the seismic multi-attribute inversion for a pore-sensitive factor is proposed.First,impedance inversion is used to distinguish the reservoir and matrix.Next,expert optimization and automatic optimization are combined to sort the seismic attributes sensitive to the parameters of the pore-sensitive factor.Finally,a probabilistic neural network algorithm is used to implement the seismic multi-attribute inversion for pore-sensitive factor using the selected seismic attributes.Testing on field data indicated the validity of the method as the effective reservoir prediction results provided support for well location deployment in oilfield development.
国家科技重大专项“海外重点油气田开发钻采关键技术”(2017ZX05032-004)资助。