摘要
裂缝是碳酸盐岩等储层的重要渗流通道和储集空间,由于这类储层具有强烈的非均质性,通常难以识别。成像测井和地层倾角测井等虽然对裂缝识别效果较好,但成本高导致数量少。目前裂缝评价方法多以使用常规测井为主,但由于测井响应的复杂性以及裂缝本身的复杂性,很难使用单一的常规测井资料得到对裂缝识别满意的结果。为了提高识别精度,本文提出了综合概率密度法、判别分析法、改进BP神经网络法等三种综合裂缝识别方法,将多种常规测井曲线经过一定的数学运算手段融合起来,放大各种曲线对裂缝的敏感度,并结合成像测井等直观资料进行对比分析,最终得到了较好的裂缝识别效果。
Fracture is the crucial seeping path and reserve space in particular lithological reservoir such as carbonate reservoir.These reservoirs have strong heterogeneity and are hard to identify.Imaging logging and diplog have good fracture identification effect but limited by high cost.By now,the main method of fracture identification is using conventional logging curve.However,due to the complexity of log response and fracture,it is hard to get satisfied solution with single curve.In order to improve the recognition accuracy,three integrated fracture identification methods such as integrated probability density,discriminant analysis and improved BP NN model are put forward to synthesize multiple curves with mathematical operation and FMI information.The methods are applyed in Hope Oilfield in Kazakhstan and obtained better results.
出处
《国外测井技术》
2017年第4期8-13,2,共6页
World Well Logging Technology
基金
国家科技重大专项(2016ZX05002-005)资助
关键词
碳酸盐岩
综合裂缝识别
测井评价
cabonate
comprehensive fracture identification
well logging evaluation
