基于测井数据小波分析的煤层识别方法
|
摘要
针对原始测井数据煤层响应特征不理想的问题,以东濮凹陷区M6、F3、A16井测井数据为样本,借助主成分分析方法,选取自然伽马(GR)、声波时差(AC)测井数据,进行小波变换处理。结果表明,一维离散小波变换下,选用Sym2小波基,五级分解后,测井数据第5层高频分量重构曲线对煤层特征的响应最为明显。小波变换后煤层测井重构数据呈现五种典型类型,主要表现为高AC值、低GR值特征,组合表现为两端尖锐的几何形态。泥岩表现为两高(高AC值、高GR值)特征,砂岩、灰岩表现为两低(低AC值、低GR值)特征。基于测井数据小波变换的煤层识别方法,既能凸显煤层高AC、低GR的测井响应特征,又能有效抑制原始曲线中部分疑似煤层信息,提高煤层识别的可靠性。
This paper proposes a coal seams recognition method based on the wavelet analysis of well logs,an approach designed to address the case where the response characteristics of coal seam from original logging data are not ideal. The study involves applying log data of M6,F3 and A16 wells in Dongpu depression,selecting GR and AC log data using principal component analysis( PCA) and thereby performing wavelet transform. The results show under the one-dimensional discrete wavelet transform,Sym2 wavelet base selection and 5-level decomposition generates the most obvious response of the reconstruction curve of high-frequency component in the fifth layer of logging data to coal seam characteristics. Wavelet transform is followed by five typical types of reconstructed data of coal seam logging,mainly characterized by "high AC value,low GR value"and the combination manifested by sharp geometry at both ends.Mudstone is marked by "two highs( high AC value,high GR value) ",and sandstone and limestone are featured by"two lows( low AC value,low GR value) ". The recognition method could enable an improvement in the reliability of coal seam recognition by not only highlighting the logging characteristics of"high acoustic time difference( AC) and low natural gamma ray( GR) "of coal seams,but also effectively suppressing the suspected coal seam information.
This paper proposes a coal seams recognition method based on the wavelet analysis of well logs,an approach designed to address the case where the response characteristics of coal seam from original logging data are not ideal. The study involves applying log data of M6,F3 and A16 wells in Dongpu depression,selecting GR and AC log data using principal component analysis( PCA) and thereby performing wavelet transform. The results show under the one-dimensional discrete wavelet transform,Sym2 wavelet base selection and 5-level decomposition generates the most obvious response of the reconstruction curve of high-frequency component in the fifth layer of logging data to coal seam characteristics. Wavelet transform is followed by five typical types of reconstructed data of coal seam logging,mainly characterized by "high AC value,low GR value"and the combination manifested by sharp geometry at both ends.Mudstone is marked by "two highs( high AC value,high GR value) ",and sandstone and limestone are featured by"two lows( low AC value,low GR value) ". The recognition method could enable an improvement in the reliability of coal seam recognition by not only highlighting the logging characteristics of"high acoustic time difference( AC) and low natural gamma ray( GR) "of coal seams,but also effectively suppressing the suspected coal seam information.
引文
[1]陶珂,朱建军.小波去噪质量评价方法的对比研究[J].大地测量与地球动力学,2012,32(2):128-133.
[2]李增学.煤地质学[M].北京:地质出版社,2009:58-66,130-133.
[3] Yu J F,Guo K,Yuan X X,et al. Advance of the application ofwell logging to coalbed methane exploration[C]//Advances onCBM Reservoir and Developing Engineering.徐州:中国矿业大学出版社,2009:486-494.
[4]叶晓斌.煤层气常规测井方法探讨[J].云南地质,2000(4):371-375.
[5]王志文.煤层含气量的测井评价技术[D].长春:吉林大学,2009.
[6]李娜娜,施龙青,李忠建,等.主成分分析法在底板突水影响因素评价中的应用[J].煤矿开采,2011,16(2):26-28.
[7]谢中华. Matlab统计分析与应用:40个案例分析[M].北京:北京航空航天大学出版社,2010:350-374.
[8]胡嘉良,高玉超,余继峰,等.基于PCA-BP神经网络的非常规储层岩性识别研究[J].山东科技大学学报(自然科学版),2016,35(5):9-16.
[9] Yu J,Guo K,Yuan X,et al. Wavelet denoising of well logs andits geological performance[J]. Energy Exploration&Exploitation,2010,28(2):87-96.
[10]余继峰,李增学.测井数据小波变换及其地质意义[J].中国矿业大学学报,2003(3):127-130.
[11]寻知锋.小波变换在储层沉积学中的应用[D].青岛:山东科技大学,2009.
[12]徐本燕.基于测井信息提取的煤层气储层评价研究[D].青岛:山东科技大学,2007.
[13]付文钊.地层多尺度旋回性分析研究[D].青岛:山东科技大学,2011.
[14]杨力华,戴道清,黄文良.信号处理的小波导引[M].北京:机械工业出版社,2002.
[15] Trauth M H.地球科学中的Matlab应用[M].第3版.北京:国防工业出版社,2015:16-110.
[16]王明祥,宁宇蓉,王晋国.基于Mallat算法的一维离散小波变换的实现[J].西北大学学报(自然科学版),2006(3):364-368.
[17]吕瑞兰,吴铁军,于玲.采用不同小波母函数的阈值去噪方法性能分析[J].光谱学与光谱分析,2004,24(7):826-829.
[18]陶珂,朱建军.小波去噪质量评价方法的对比研究[J].大地测量与地球动力学,2012,32(2):128-133.
[19]丰彦,高国荣.小波阈值消噪算法中分解层数的自适应确定[J].武汉大学学报(理学版),2005,51(S2):11-14.
[20]要惠芳,康志勤,李伟.典型构造煤变形特征及储集层物性[J].石油勘探与开发,2014,41(4):414-420.
[21]马国栋,陈同俊,崔若飞.测井曲线识别构造煤实例研究[J].地球物理学进展,2017,32(3):1208-1216.
[2]李增学.煤地质学[M].北京:地质出版社,2009:58-66,130-133.
[3] Yu J F,Guo K,Yuan X X,et al. Advance of the application ofwell logging to coalbed methane exploration[C]//Advances onCBM Reservoir and Developing Engineering.徐州:中国矿业大学出版社,2009:486-494.
[4]叶晓斌.煤层气常规测井方法探讨[J].云南地质,2000(4):371-375.
[5]王志文.煤层含气量的测井评价技术[D].长春:吉林大学,2009.
[6]李娜娜,施龙青,李忠建,等.主成分分析法在底板突水影响因素评价中的应用[J].煤矿开采,2011,16(2):26-28.
[7]谢中华. Matlab统计分析与应用:40个案例分析[M].北京:北京航空航天大学出版社,2010:350-374.
[8]胡嘉良,高玉超,余继峰,等.基于PCA-BP神经网络的非常规储层岩性识别研究[J].山东科技大学学报(自然科学版),2016,35(5):9-16.
[9] Yu J,Guo K,Yuan X,et al. Wavelet denoising of well logs andits geological performance[J]. Energy Exploration&Exploitation,2010,28(2):87-96.
[10]余继峰,李增学.测井数据小波变换及其地质意义[J].中国矿业大学学报,2003(3):127-130.
[11]寻知锋.小波变换在储层沉积学中的应用[D].青岛:山东科技大学,2009.
[12]徐本燕.基于测井信息提取的煤层气储层评价研究[D].青岛:山东科技大学,2007.
[13]付文钊.地层多尺度旋回性分析研究[D].青岛:山东科技大学,2011.
[14]杨力华,戴道清,黄文良.信号处理的小波导引[M].北京:机械工业出版社,2002.
[15] Trauth M H.地球科学中的Matlab应用[M].第3版.北京:国防工业出版社,2015:16-110.
[16]王明祥,宁宇蓉,王晋国.基于Mallat算法的一维离散小波变换的实现[J].西北大学学报(自然科学版),2006(3):364-368.
[17]吕瑞兰,吴铁军,于玲.采用不同小波母函数的阈值去噪方法性能分析[J].光谱学与光谱分析,2004,24(7):826-829.
[18]陶珂,朱建军.小波去噪质量评价方法的对比研究[J].大地测量与地球动力学,2012,32(2):128-133.
[19]丰彦,高国荣.小波阈值消噪算法中分解层数的自适应确定[J].武汉大学学报(理学版),2005,51(S2):11-14.
[20]要惠芳,康志勤,李伟.典型构造煤变形特征及储集层物性[J].石油勘探与开发,2014,41(4):414-420.
[21]马国栋,陈同俊,崔若飞.测井曲线识别构造煤实例研究[J].地球物理学进展,2017,32(3):1208-1216.