多元判别分析方法在复杂油水层识别中的应用

大庆油田某区块地质条件复杂,储层非均质性强、物性差。随着开发的深入油水层识别越来越困难。在缺少岩电分析资料的情况下,本文通过优选预处理后能反映含油性、储层物性的电阻率、声波测井曲线和泥质参数,引入多元判别分析方法,将多变量问题降维并求取两个判别函数,建立二维油水判别图版,并与常规测井参数交会图版进行对比,分析了油水识别困难的成因机理。应用新方法建立的油水识别图版在太30区块判别油水层经试油验证,符合率为85.7%,效果良好。

The Application of Multivariate Analysis Method In Identification of the Complicated Oil and Water Layers

The geological conditions of a block in Daqing oilfield are complex with strong reservoir heterogeneity and poor physical property .Identification of oil and water layers becomes more and more difficult with further development.In lack of rock electroanalysis data, the multivariate analysis method is introduced in this paper to make multivariable problem to be dimensionality reduced to get two discriminant functions after pretreatment of the resistivity, acoustic logging curves and shale parameters which can reflect oiliness characteristics of the reservoir. Then a two dimensional plate is established to distinguish oil and water, which is compared with conventional log parameter cross-plot to analyse the mechanism cause of formation in oil and water identification.This new plot-plate was verified to be good by oil testing in Tai 30 Block. Its coincidence rate was 85.7%.