埕岛地区中生界储层岩性和裂缝测井识别应用

Lithology and fracture logging identification for mesozonic reservoir in Chengdao area

埕岛地区中生界储层岩性复杂,既有沉积岩又有火山岩。储层裂缝较发育,测井响应特征变化大。为增强灰色关联法在复杂岩性测井识别中的适用性,对斜率关联度的算法进行了改进,用各序列变化量的平均值代替了序列值本身,在该区岩性识别中的应用取得较好效果。就储层裂缝识别而言,单一常规测井方法识别多解性强,为此,综合多种测井信息,利用综合概率法建立了现有取心井的裂缝指示曲线;经过反复试算确定了最佳的BP网络参数,并对网络的固定学习速率进行了自适应调整,提高了该区裂缝识别精度。

Having complex lithology, mesozonic reservoir in Chengdao area has sedimentary rocks and volcanic rocks, and the log- ging response characteristics change greatly since the existence of fractures. In order to strengthen the adaptability of grey correla- tive method in complex lithologic logging identification, the improved gradient grey correlation which replaces sequence value with the average of sequence variable quantity gets good effects in the lithology identification. For reservoir fracture identification, single conventional logging method has multiple solutions in identifying fractures, therefore, fracture indications of current core wells are established by comprehensive using multiple logging information and comprehensive probability method. By repeated trial calcula- tion, the best parameters of BP network are determined and the fixed learning rate has adaptive adjustment, thereby improving the fracture identification accuracy.