摘要
应用多元线性回归分析方法,以准噶尔盆地2008年底之前所钻探井试油井段的气测录井资料为分析样本数据,结合试油结论对比,运用SPSS统计分析软件进行多元线性回归,分别建立了标准油层、油水同层、水层(含油水层、含气水层)及干层(含油层、含气层)的多元线性回归模型。模型的检验分析反映其显著性明显,样本数据回判分析表明油层和干层的正判率较高,水层和油水同层的正判率较低。利用该方法对2009-2011年新钻探井部分试油井段进行解释评价,总体解释评价符合率为85%。
Using multiple linear regression analysis method, taking gas logging information of oil testing sections of wells drilled before the end of 2008 in Junggar Basin as the analysis sample data, combined with oil testing conclusions' contrast, using SPSS statistical analysis software to carry out multiple linear regression, the authors respectively established multiple linear regression models for standard reservoirs, oil-water layers, the water layers (oily water layers, gas containing water layers) and dry layers (oil layers and gas-bearing layers). The examination and analysis of the models reflected the obvious significance, sample data back analysis showed that the reservoirs and dry layers had a higher positive rate, while the positive rates of water layers and oil-water layers were lower. Gas logging multiple linear regression method was used to interpret and evaluate some oil testing intervals among new wells drilled in 2009-2011, the overall interpretation and evaluation were in line with the rate of 85%.
出处
《录井工程》
2012年第4期22-25,80,共4页
Mud Logging Engineering
关键词
SPSS统计分析
多元线性回归
模型
相关性
F分布表检验法
气测解释
样本回判
SPSS statistical analysis, multiple linear regression, model, correlation, identifying method of F distribution list, gas logging interpretation, sample regression
