Abstract: The key to increase the efficiency of exploration and development of oil fields in Eastern China is the lateral reservoir prediction and description by using comprehensive method combining seismic data with supplementary geologic data including drilling and well logging data.According to the sedimentary features of thin interbedded sandstones in fluvial facies (meander), the prediction of lithology and favorable play and the lateral prediction of hydrocarbon-bearing region were completed by using comprehensive parameter transformation, artificial neural network, wavelet transform, etc.The results were applied to various aspects of prospective prediction, prospecting target selection, prospecting well location, reserve estimation, exploitation plan design, etc., which were confirmed by the practical exploration.

Key words: fluvial facies, reservoir bed, comprehensive parameter transformation, artificial neural network, wavelet transform, seismic description, Chengdao Oilfield

摘要： 以地震资料为主,结合钻井、测井等地质资料,采用多种方法对河流相的储层作出预测描述是提高我国东部油田勘探、开发效益的关键.根据埕岛油田河流相(曲流河)薄互层砂岩沉积特点,开展了以综合参数转换法、人工神经网络、小波变换等方法的岩性、有利富集带和含油气面积的预测描述,应用于该油田的前景预测、勘探目标选择、探井部署、储量计算和开发方案设计等不同层次的要求中,经实际的勘探实践证实,皆达到了预期的地质效果.

关键词: 河流相, 储层, 综合参数转换法, 人工神经网络, 小波变换, 地震描述, 埕岛油田