摘要
为了利用随钻方位电磁波电阻率仪器的测量数据确定地层界面方位和距离,给地质导向提供决策依据,须采用准确可靠的反演方法。针对随钻方位电磁波电阻率仪器,建立了地质导向应用模型并模拟了其响应特征,研究了拟牛顿反演算法和流程,反演过程中只需要较小的计算量就可以得到Jacobian矩阵,大大提高了反演速度;并利用单界面和双界面地层的反演理论模型,验证了该算法的正确性和准确度。在胜利油田草XX井的现场试验结果表明,实时反演结果与方位电磁波电阻率成像显示及后期完井录井结果一致。该反演方法能满足利用方位电磁波电阻率进行地质导向的要求,为方位电磁波电阻率实时地质导向提供了一种高效、准确的计算方法。
In order to determine the formation interface azimuth and distance by using the data measured by LWD azimuthal electromagnetic wave resistivity logging instrument using the quasi-Newton inversion algorithm.The objective was provide a decision-making basis for geosteering,an accurate and reliable inversion method must be used.For the LWD azimuthal electromagnetic wave resistivity logging instrument,an application model of geological guidance has been established and simulated its response characteristics.By studying the quasi-Newton inversion algorithm and process,the Jacobian matrix was obtained with less calculation cost in the inversion process,which greatly speeds up the inversion.At the same time,the correctness and accuracy of this algorithm were verified by using the theoretical inversion model including both single interface and a double interface strata model.The field test of Well Cao-XX in Shengli Oilfield showed that the real-time inversion results were consistent with the imaging display of azimuthal electromagnetic wave resistivity and also with the drilling/completion/logging results in later stages.This inversion method can meet the requirements of real time geo-steering by efficient and accurate calculation of azimuthal electromagnetic wave resistivity.
作者
杨震
于其蛟
马清明
YANG Zhen;YU Qijiao;MA Qingming(MWD&LWD Technology Center,Sinopec Shengli Oilfield Service Corporation,Dongying,Shandong,257064,China)
出处
《石油钻探技术》
CAS
CSCD
北大核心
2020年第3期120-126,共7页
Petroleum Drilling Techniques
基金
国家科技重大专项“低渗透油气藏高效开发钻完井技术”之课题“低渗透油气深层高温高压随钻测控技术”(编号:2016ZX05021-001)
国家重点研发计划项目“深海关键技术与装备”(编号:2016YFC0302802)联合资助。
关键词
随钻方位电磁波
电阻率测井
地质导向
拟牛顿法
实时
反演
边界距离
azimuthal electromagnetic logging while drilling
resistivity logging
geosteering
quasi-Newton method
real time
inversion
distance to interface