摘要
塔中地区奥陶系溶孔裂缝型储层非均质性严重,又受火成岩影响,其预测难度大,地震属性蕴含丰富的缝洞发育带及其分布特征的信息,由于属性种类很多,单因子判别存在片面性,而多属性人工综合工作量大而且受主观因素影响,为此,本文把多属性线性和非线性融合技术应用于塔中碳酸盐岩缝洞储层预测中,对与缝洞发育敏感的振幅变化率与反映流体性质敏感的频率衰减参数进行线性融合,对振幅变化率、相干值、均方根振幅、振幅方差、20Hz时频属性、能量半衰时等6种地震属性,基于人工神经技术进行地震属性非线性融合来综合预测储层发育程度及其储层厚度,其结果与钻遇储层发育情况对比得比较吻合,说明该方法可明显提高储层预测的精度。
Reservoir prediction of carbonate fracture and cavity reservoir in Tazhong area is difficult due to serious reservoir heterogeneity and effects of igneous rock. Seismic attributes include rich cavity information, the sidedness of single factor, intense workload and other factors. This paper applies liner and nonlinear combination of multi-attributes to car- bonate fracture and cavity reservoir in Tazhong area. Amplitude change rate which is sensi- tive to cavity development and frequency attenuation parameter which is sensitive to fluid property are combined linearly. Amplitude change rate coherence value, RMS amplitude, amplitude variance, 20Hz time-frequency attributes and energy half-time attributes are combined nonlinearly by artificial neural network technology. These methods are used to predict reservoir development and thickness, and the results are well consi-stent with drill-ing finding. Thus it indicates that this method could improve the accuracy of reservoir prediction impressively.
出处
《工程地球物理学报》
2010年第1期19-22,共4页
Chinese Journal of Engineering Geophysics
关键词
多属性融合
碳酸盐岩缝洞
储层预测
时频分析
人工神经网络
multi-attribute fusion
carbonate fracture and cavity
reservoir prediction
time-frequency analysis
ANN