摘要
页岩密度是页岩气甜点评价的重要参数,但大量的水平井不开展密度测井甚至不测井,导致评价参数匮乏。通过分析岩石密度的求取方法与多元自适应回归样条的原理,设定调优参数开展模型训练,建立利用元素录井数据计算页岩密度的计算模型。计算结果对比表明,基于多元自适应回归样条算法的模型计算精度高,与实测数据差异小。应用结果表明,页岩密度计算模型能为页岩气水平井段提供重要的评价参数,结合气测等常规录井数据,能有效用于开展页岩气水平井分段评价,为储集层评价与测试选层提供了更多依据。
Shale density is an important parameter for shale gas sweet spot evaluation,a large number of horizontal wells do not carry out density logging or even logging,resulting in the lack of evaluation parameters.By analyzing the calculation method of rock density and the principle of multivariate adaptive regression splines,the model training was carried out by setting the tuning parameters,and the calculation model of shale density was established by using element logging data.The comparison of the calculation results shows that the model based on the multivariate adaptive regression spline algorithm has high calculation accuracy and little difference from the measured data.The application results show that the shale density calculation model can provide important evaluation parameters for the horizontal section of shale gas.Combined with conventional mud logging data such as gas logging,the model can be effectively used for sectionlization evaluation of shale gas horizontal wells,and provide more basis for reservoir evaluation and test layer selection.
作者
欧传根
唐诚
王崇敬
梁波
OU Chuangen;TANG Cheng;WANG Chongjing;LIANG Bo(Geological Logging Branch of Sinopec Southwest Petroleum Engineering Co., Ltd., Mianyang, Sichuan 621000, China)
出处
《录井工程》
2021年第1期1-5,共5页
Mud Logging Engineering
基金
中石化科技部项目“基于XRF的页岩地层多参数实时求取方法及应用”(编号:P17014-9)
中石化石油工程技术股份服务公司项目“基于机器学习的页岩气录井解释参数模型研究”(编号:SG19-20Q)。
关键词
水平井
元素录井
页岩密度
多元自适应回归样条
数学模型
horizontal well
element logging
shale density
multivariate adaptive regression splines
mathematical model