A地区页岩气储层总有机碳含量测井评价方法研究  

Research on logging evaluation method of TOC content of shale gas reservoir in A area

作　　者：熊镭[1] 张超谟[1,2] 张冲[2] 谢冰[1] 丁一[1] 韩淑敏 

机构地区：[1]长江大学地球物理与石油资源学院,武汉430100 [2]长江大学油气资源与勘探技术教育部重点实验室,武汉430100 [3]中国石油集团东方地球物理有限公司,河北涿州072750

出　　处：《岩性油气藏》2014年第3期74-78,83,共6页Lithologic Reservoirs

基　　金：湖北省自然科学基金项目"基于等效岩石单元模型的渗透率测井评价方法研究"(编号:2013CFB396)资助

摘　　要：页岩气储层中总有机碳含量(TOC)反映了页岩的生烃潜力,准确获取页岩气储层TOC含量对页岩气的开发具有重要意义。利用测井资料的连续性和纵向分辨率高等特点,建立精度较高的TOC测井评价模型。在分析几种常用TOC测井评价方法限制因素的基础上,结合A地区岩性变化复杂的实际情况,建立了BP神经网络预测TOC、拟合方法计算TOC、基于干酪根含量计算TOC共3种模型,并对该地区X井页岩进行了TOC含量评价。结果表明:在A地区采用BP神经网络预测TOC模型其精度最高,可为岩性复杂地区的TOC含量评价提供技术支持。Total organic carbon （TOC） content of shale gas reservoir reflects the hydrocarbon generation potential of shale rocks. It has an important guiding significance for shale gas development to obtain the TOC content accurately by use of conventional logging data which has characteristics of continuousness and high vertical resolution. Therefore, it is especially important to establish highly precise model of the organic carbon content evaluation. Combining the limiting factors of TOC content evaluation methods with the reality of complex lithological changes of A area, we established three kinds of TOC content models to evaluate shale rocks from X well in A area. They are BP neural network, uranium and kerogen. It is concluded that the BP neural network model is with the highest precision to forecast the total organic carbon content, and provide technical support to TOC content evaluation in complex lithology areas.

关 键 词：页岩气 总有机碳 BP神经网络 自然伽马能谱 干酪根 

分 类 号：P631.8[天文地球—地质矿产勘探]