涠西区块薄砂体储层预测中测井耦合分频反演方法及解释
Logging coupled frequency inversion method and interpretation in prediction of thin sandstone reservoir in the Weixi Block
赵天沛 1雷蕾 1郝伟航1
作者信息
- 1. 中国石化上海海洋油气分公司,上海 200120
- 折叠
摘要
对涠西地区的三角洲平原—前缘薄层砂体,利用测井耦合分频反演方法开展储层预测.该方法利用BP神经网络智能算法建立测井和地震资料研究振幅与频率(AVF)的非线性映射关系,深入挖掘研究区测井岩性敏感自然伽马测井及地震资料有效频带的低、中、高频信息,得到包含全频段信息的高分辨率储层反演结果.在反演剖面上砂体发育特征清晰,泥岩薄夹层横向展布清楚,河道砂体横向上厚度存在明显变化,预测砂体厚度与测井解释厚度误差控制在1%左右.相较于传统反演方法,测井耦合分频反演方法有效提高了薄层砂体厚度计算的准确度.
Abstract
For the thin-layer sand bodies of the delta plain-front in the Weixi Area,the reservoir prediction is carried out using a logging coupled frequency division inversion method.In this method,the BP neural network intelligent algorithm is used to establish the nonlinear mapping relationship between amplitude and frequency(AVF)of logging and seismic data research.The low,medium,and high-frequency information of effective frequency band of natural gamma logging and seismic data sensitive to logging lithology in the study area is also excavated,and the high-resolution reservoir inversion results including full frequency band information are obtained.In the inversion profile,the development characteristics of sand body are clear,the horizontal distribution of thin mudstone interlayers is distinct,and the lateral thickness of channel sand body changes obviously.The error between the predicted thickness of sand body and the logging interpretation thickness is controlled about 1%.Compared with the traditional inversion methods,the logging coupled frequency division inversion method effectively improves the accuracy of thickness calculations of thin sand body.
关键词
分频反演/自然伽马测井/BP神经网络/储层预测Key words
frequency division inversion/natural gamma logging/BP neural network/reservoir prediction引用本文复制引用
出版年
2024
NETL
NSTL
万方数据