基于测录井资料的环西—彭阳地区延安组储层流体性质识别方法研究

A Study on the Logging-Based Identification Method for Reservoir Fluid Properties of the Yan’an Formation in the Huanxi–Pengyang

针对鄂尔多斯盆地环西—彭阳地区延安组储层流体性质识别困难的问题,从测井、录井常规解释难点入手,引入SPSS软件中的Fisher判别方法,利用大数据理念,对测录井资料进行了深度挖掘。从判识率与分离度2方面考虑,筛选出15项对储层流体性质敏感的参数,进行Fisher非标准化判别函数分类识别,流体识别准确率可达92.8%。选用2个贡献率较大的函数F1和F2进行交会,建立了解释图版,图版返判率为89%。利用解释图版对环西—彭阳地区9口井10个储层进行了流体识别评价,符合率达80%,且未遗漏有价值的储层。应用结果表明,Fisher判别方法适用于类似油藏的储层流体性质识别,流体识别准确率高,具有推广价值。

In view of the difficulty in identifying the reservoir fluid properties of the Yan’an Formation in the Huanxi–Pengyang Area,beginning with the conventional interpretation difficulties of logging and mud logging,the Fisher discrimination method in SPSS software was introduced,with the concept of Big Data,the logging/mud logging parameters were deeply mined.Considering the recognition rate and separation degree,15 parameters of reservoir fluid sensitivity were selected to identify the classification of Fisher non-standardized discriminant function,and the accuracy of fluid identification could reach 92.8%.Functions F1 and F2 were selected for their large contribution rate and intersected to plot the interpretation chart,and the identification rate of the chart was 89%.The interpretation chart was applied in the fluids interpretation and evaluation of 10 layers in 9 wells of Huanxi–Pengyang Area with a coincidence rate of 80%,and no valuable layers were missed.The application results showed that the Fisher discriminant method is suitable for high accurate identification of reservoir fluid properties in similar reservoirs,and it is worthy of wide adoption.