基于加权马氏迭代法的储层定量分类评价方法
|
摘要
为实现低渗储层的定量分类,采用基于k-均值聚类分析的加权马氏迭代法,对表征储层孔隙结构和渗流特征的七个特征参数进行聚类运算,并结合贝叶斯判别分析法,确定储层定量判别公式,建立先分类、后判别的储层定量分类评价方法。应用该方法对渤南油田义176区块沙四段砂岩储层开展分类评价,获得不同参数的储层分类门限。结果表明,研究区储层可分为三类,Ⅰ～Ⅲ类储层物性逐渐变差,孔隙半径逐渐减小,排替压力、退汞效率逐渐增加。该方法适用性强、区分度明显,可保证评价结果的准确性和定量性。
This paper is an attempt to achieve quantitative classification of low permeability reservoirs. The study involves clustering the seven characteristic parameters of reservoir pore structure and seepage characteristics using the weighted Mahalanobis iterative method based on k-means clustering analysis; determining the reservoir quantitative determination formula,combined with Bayesian discriminant analysis; and developing a quantitative classification and evaluation method for reservoirs with initial classification and subsequent discrimination. The study also describes the process of applying the method to classifying and evaluating the sandstone reservoirs of the fourth member of the Shahezi formation in the Block Yi176 of the Bonan oilfield,and thereby obtaining the reservoir classification thresholds with different parameters. The results show that the reservoirs in the study area can be divided into three categories,exhibiting gradually deteriorating physical properties,gradually decreasing pore radius,and gradually increasing displacement pressure and mercury removal efficiency. The method capable of a stronger applicability and an obvious discrimination could ensure the accuracy and quantitativeness of the evaluation results.
This paper is an attempt to achieve quantitative classification of low permeability reservoirs. The study involves clustering the seven characteristic parameters of reservoir pore structure and seepage characteristics using the weighted Mahalanobis iterative method based on k-means clustering analysis; determining the reservoir quantitative determination formula,combined with Bayesian discriminant analysis; and developing a quantitative classification and evaluation method for reservoirs with initial classification and subsequent discrimination. The study also describes the process of applying the method to classifying and evaluating the sandstone reservoirs of the fourth member of the Shahezi formation in the Block Yi176 of the Bonan oilfield,and thereby obtaining the reservoir classification thresholds with different parameters. The results show that the reservoirs in the study area can be divided into three categories,exhibiting gradually deteriorating physical properties,gradually decreasing pore radius,and gradually increasing displacement pressure and mercury removal efficiency. The method capable of a stronger applicability and an obvious discrimination could ensure the accuracy and quantitativeness of the evaluation results.
引文
[1]杨正明,张英芝,郝明强,等.低渗透油田储层综合评价方法[J].石油学报,2006,27(2):64-67.
[2]孟祥超,陈扬,谢天寿,等.煤系砂岩储层粒度定量评价图版建立及油气富集区预测———以准噶尔盆地玛湖凹陷西斜坡Aih12井区八道湾组为例[J].东北石油大学学报,2017,41(6):19-20.
[3]刘奕蔚.油气储层成岩作用及定量评价技术[J].石化技术,2017(12):155.
[4]赵加凡,陈小宏,张勤,等.灰关联分析在储层评价中的应用[J].勘探地球物理进展,2003,26(4):282-286.
[5]刘吉余,李艳杰,于润涛.储层综合定量评价系统开发与应用[J].物探化探计算技术,2004,26(1):33-36.
[6]吕红华,任明达,柳金诚,等. Q型主因子分析与聚类分析在柴达木盆地花土沟油田新近系砂岩储层评价中的应用[J].北京大学学报(自然科学版),2006,42(6):740-745.
[7]付殿敬,徐敬领,王贵文.基于Q型聚类分析和贝叶斯判别算法研究储层分类评价[J].科技导报,2011,29(3):29-32.
[8]闫明,张云峰,李易霖.基于K-均值聚类与贝叶斯判别的储层定量评价———以大安油田泉四段储层为例[J].深圳大学学报(理工版),2016,33(2):212-217.
[9]王维喜,曹天军,朱海涛.压汞曲线在特低渗油藏储层分类中的应用[J].重庆科技学院学报(自然科学版),2010,12(3):18-20.
[10]袁志发,周静芋.多元统计分析[M].北京:科学出版社,2002:241.
[11]韩胜娟. SPSS聚类分析中数据无量纲化方法比较[J].科技广场,2008(3):229-231.
[12]纪友亮.油气储层地质学[M].北京:石油工业出版社,2015:330.
[13]柳娜,南珺祥,刘伟,等.鄂尔多斯盆地湖盆中部长7致密砂岩储层特征[J].西安石油大学学报(自然科学版),2014,29(4):6-13.
[14]李婷婷,王凤琴,令狐松,等.低渗透储层分类方法研究[J].国外测井技术,2010(6):23-26.
[15]吴香华,牛生杰,吴诚鸥,等.马氏距离聚类分析中协方差矩阵估算的改进[J].数理统计与管理,2011,30(2):240-245.
[16]胡雷芳.五种常用系统聚类分析方法及其比较[J].浙江统计,2007(4):11-13.
[2]孟祥超,陈扬,谢天寿,等.煤系砂岩储层粒度定量评价图版建立及油气富集区预测———以准噶尔盆地玛湖凹陷西斜坡Aih12井区八道湾组为例[J].东北石油大学学报,2017,41(6):19-20.
[3]刘奕蔚.油气储层成岩作用及定量评价技术[J].石化技术,2017(12):155.
[4]赵加凡,陈小宏,张勤,等.灰关联分析在储层评价中的应用[J].勘探地球物理进展,2003,26(4):282-286.
[5]刘吉余,李艳杰,于润涛.储层综合定量评价系统开发与应用[J].物探化探计算技术,2004,26(1):33-36.
[6]吕红华,任明达,柳金诚,等. Q型主因子分析与聚类分析在柴达木盆地花土沟油田新近系砂岩储层评价中的应用[J].北京大学学报(自然科学版),2006,42(6):740-745.
[7]付殿敬,徐敬领,王贵文.基于Q型聚类分析和贝叶斯判别算法研究储层分类评价[J].科技导报,2011,29(3):29-32.
[8]闫明,张云峰,李易霖.基于K-均值聚类与贝叶斯判别的储层定量评价———以大安油田泉四段储层为例[J].深圳大学学报(理工版),2016,33(2):212-217.
[9]王维喜,曹天军,朱海涛.压汞曲线在特低渗油藏储层分类中的应用[J].重庆科技学院学报(自然科学版),2010,12(3):18-20.
[10]袁志发,周静芋.多元统计分析[M].北京:科学出版社,2002:241.
[11]韩胜娟. SPSS聚类分析中数据无量纲化方法比较[J].科技广场,2008(3):229-231.
[12]纪友亮.油气储层地质学[M].北京:石油工业出版社,2015:330.
[13]柳娜,南珺祥,刘伟,等.鄂尔多斯盆地湖盆中部长7致密砂岩储层特征[J].西安石油大学学报(自然科学版),2014,29(4):6-13.
[14]李婷婷,王凤琴,令狐松,等.低渗透储层分类方法研究[J].国外测井技术,2010(6):23-26.
[15]吴香华,牛生杰,吴诚鸥,等.马氏距离聚类分析中协方差矩阵估算的改进[J].数理统计与管理,2011,30(2):240-245.
[16]胡雷芳.五种常用系统聚类分析方法及其比较[J].浙江统计,2007(4):11-13.