应用地震波形分析技术预测裂缝的方法探讨

探讨了利用地震道波形变化特征识别和预测裂缝的方法。介绍了该方法的基本原理和实现过程，首先从过井地震剖面出发，通过井震标定，建立裂缝的地震响应模式，总结裂缝发育层段(时窗段)的特征波形；然后利用神经网络技术逐道进行月的层段的实际地震道对比，把地震波形状分类成离散的“裂缝信息”，再根据拟合度准则进行分类，最终得到裂缝地震反射异常体的空间展布。在青西油田利用地震波形分析技术进行了裂缝预测，并根据预测结果部署了5口井，钻探结果与预测结果相吻合，位于裂缝发育区的Q11井获得了高产油流。

Fracture prediction by seismic waveform analysis

Seismic waveform analysis can depict lateral variation of seismic signal in detail by contrast to real seismic trace one by one in a layer phase according to waveform feature of seismic trace, in order to get spatial distribution of abnormal seismic reflection. On basis of qualitative identification of fracture and seismic trace analysis using nerve network of seismic trace waveform identification, we can classify the shape of seismic trace signal into discrete fracture information. Then we may calibrate object by integrating drilling core and the explanation results of well fracture, confirm the fracture type and its spatial distribution. The foundation, identification fracture theory and realization process by seismic waveform analysis technology, and an actual example of Qingxi oilfield in Jiuquan basin are presented systematically. Seismic waveform analysis is a good means to predict complicated lithological fracture reservoir.