判别分析在海拉尔油田试油中的应用

Application of Discriminant Analysis in Well Testing Interpretation in Hailar Oilfield

为了解决海拉尔油田试油解释难题,针对地下复杂的油水分布状况,收集整理了研究区2000年至2008年194份的试油地层水样品数据。考虑了能反映地层水化学性质的7个成分(CO23-、HCO-3、Cl-、SO24+、Ca2+、Mg2+、K++Na+)含量参数,应用了数学地质中判别分析的原理和方法,建立了研究区出水层和非出水层的定量判别模型。经过回判检验,正判率达到了95.9%,表明判别函数模型是显著的。利用建立的判别函数模型,对2009年17个地层水试油样品进行判别,正判率达到88%,与实际符合很好,表明判别模型可靠,可以在试油解释中使用。

In order to solve the difficult problem of well testing interpretation in Hailar oilfield,194 copies of formation water well testing data from 2000 to 2008 in the study area are collected and arranged,considered seven formation water chemistry component variables such as CO2-3,HCO-3,Cl-、SO2＋4,Ca2＋,Mg2＋,K＋＋Na＋,applied the principle and method of discriminant analysis in mathematical geology and established the quantitative discriminant model for determining water layer.By back sentencing and testing,the right judgment ratio of the model was 95.9% and it proved the model is significant.17 formation water well testing samples in 2009 was convicted by the discriminant model and its right judgment ratio is 88%.It shows the model is reliable and can be used in well testing interpretation.