逐步回归分析方法在储层参数预测中的应用

Application of stepwise regression analysis method in reservoir parameter prediction

目前,储层参数(孔隙度和渗透率等)分布规律和储层非均质性研究是油气藏描述的核心,储层参数是油层评价的重要依据,储层参数预测在油气勘探开发中具有重要意义。文中通过采用某一个点的测井曲线或地震数据推测出该点的孔隙度或渗透率,基于多种测井信息的多元线性回归方法已成为储层孔隙度定量预测的主要方法,多元逐步回归分析方法的理论正好适用于这种实际问题。它是利用通过特殊仪器测量的测井曲线数据参数与岩芯属性参数(例如孔隙度),建立测井曲线数据参数与多个岩芯属性参数之间的线性关系,这种方法比较简单实用。因此文中采用逐步回归分析方法作为预测方法,重点介绍了回归分析中的逐步回归的基本思想以及具体计算步骤。最后,提出油气勘探中预测孔隙度的问题,并用逐步回归分析优化回归方程并用此方程预测岩芯属性参数。研究表明,该方法预测精度高,方法稳定有效,逐步回归较好的解决了部分测井勘探的实际问题,基于多种测井信息的多元线性回归方法已成为储层孔隙度定量预测的主要方法,该方法可以把非线性问题转化为线性问题,大大减少了技术上的难题。

At present,the reservoir parameters （porosity and permeability,etc.） distribution and the study of reservoir heterogeneity is a core of reservoir description,reservoir parameter is an important basis for reservoir evaluation,reservoir parameter prediction is of great significance in the oil and gas exploration and development.The one point through the use of logging curve or earthquake data is to infer the point porosity and permeability,multiple linear regression method based on a variety of logging information has become the primary method to quantitatively predict reservoir porosity,multiple stepwise regression analysis method of theory is applicable to the practical problems,which is used by special instrument to measure parameters of well log data and core attribute parameters（porosity,for example）,set up parameters with log data linear relationship between multiple core attribute parameter,this method is simple and practical.So the stepwise regression analysis method is used as a prediction method.Regression analysis is mainly introduced the basic idea of stepwise regression and the specific calculation steps.Finally,put forward the problems to predict porosity in the oil and gas exploration,and stepwise regression analysis optimization of regression equation and the equation of forecasting core attribute parameters.Research shows that the method is of high accuracy,method is stable and effective.Stepwise regression better solve the actual problem of partial logging exploration,multiple linear regression method based on a variety of logging information has become the primary method to quantitatively predict reservoir porosity,this method can transform the nonlinear problem for linear problems,greatly reducing the technical difficulties.