利用地震属性、多元统计分析理论和ANFIS预测碳酸盐岩储层裂缝孔隙度

摘要

The carbonate reservoir is characterized by strong heterogeneity and anisotropy. It is difficult to determine fracture porosity space distribution in carbonate reservoir accurately and completely just using a single method. Introduced is a comprehensive method to predict the frac-ture porosity in carbonate reservoir such as seismic attributes, multi-analysis and Adaptive Neu-ro-Fuzzy Inference System (ANFIS). Imaging logging data is used to identify the positions of fracture porosity accurately, and calculate the pore size. The ANFIS is used to build a function relation model between the borehole fracture porosity and the seismic attributes data around the well. Then, the quantitative evaluation of the carbonate reservoir is made by combining the seis-mic attributes data of the whole work area. This method is used in.predicting the fracture porosi-ty distribution in a work area in the middle of Tarim basin. The result shows that the predicted fracture porosity developing zone is in good accordance with the high quality reservoirs drilled. It is proved that this method is adaptable in the area.%碳酸盐岩储层非均质性和各向异性强,利用某种单一技术难以准确定量地刻画出裂缝孔隙度空间分布情况.利用地震属性、多元统计分析理论和ANFIS(自适应性模糊神经网络)综合预测碳酸盐岩储层裂缝孔隙度.用成像测井资料准确识别裂缝孔隙发育位置.并计算出裂缝孔隙度的大小,用自适应性模糊神经网络建立井孔裂缝孔隙度和井旁地震属性数据体之间的函数关系模型,结合全区地震属性数据体对碳酸盐岩储层进行定量评价.利用该项技术预测塔中某工区碳酸盐岩储层裂缝孔隙度分布,预测结果与钻遇优质储层的井点吻合,说明该方法在该区有一定的实用性.