多元逐步回归法在致密砂岩储层矿物与孔隙度关系分析中的应用

矿物类型及含量是影响储层致密的关键因素，为定量表征二者与储层孔隙度间的相关关系，以鄂尔多斯盆地姬塬地区长8致密储层为例，利用多元逐步回归分析方法建立了长8储层矿物含量与孔隙度之间的多元回归方程模型：y=\mathrm{18.500}-\mathrm{0.126}{x}_{{}^{\mathrm{3}}}-\mathrm{1.017}{x}_{{}^{\mathrm{4}}}-\mathrm{0.577}{x}_{{}^{\mathrm{7}}}-\mathrm{0.674}{x}_{{}^{\mathrm{8}}}（孔隙度大小=18.500-0.126×岩屑含量-1.017×绿泥石含量-0.577×硅质胶结物含量-0.674×铁方解石含量）。通过该多元回归模型得到的预测孔隙度与实测孔隙度拟合性好，地质解释合理并与储层特征契合。综合认为，基于储层各矿物类型及含量的多元逐步回归法可以较为准确地预测储层孔隙度的大小，尤其在受多类型矿物影响的致密储层储集性能研究中存在一定的优越性。

Application of multiple stepwise regression method in the analysis of the relationship between porosity and tight sandstone： Case study of Chang 8 reservoir in Jiyuan area， Ordos Basin

The type and content of minerals are the key factors affecting the reservoir density. In order to quantitatively characterize the correlation between them and reservoir porosity， taking Chang 8 tight reservoir in Jiyuan area of Ordos Basin as an example， the multiple regression equation model between mineral content and porosity of Chang 8 reservoir was established by multiple stepwise regression analysis：y=18.500－0.126x₃－1.017x₄－0.577x₇－0.674x₈ （porosity=18.500－0.126×cuttings－1.017×chlorite－0.577×siliceous cement －0.674×ferrocalcite）. The calculated porosity obtained by the multiple regression model is well fitted with the measured porosity， and the geological interpretation is reasonable and consistent with the reservoir characteristics. It is concluded that the multiple stepwise regression method based on the types and contents of various minerals in the reservoir can accurately predict the size of reservoir porosity， especially in the study of tight reservoir performance affected by multiple types of minerals.