基于地震岩石物理分析与叠前地质统计学反演技术的齐家地区致密薄储层预测

Tight thin-bed reservoir prediction using rock physics analysis and prestack geostatistical inversion in the Qijia area

松辽盆地北部齐家地区青山口组钻遇致密砂岩油,但储集层岩石的矿物组成复杂、物性差,薄互层砂泥岩声阻抗叠置,井间薄储层地震预测因此难以开展,进而制约了致密油"甜点"区优选和水平井目标钻探的优化设计。针对上述问题,通过岩心声学测试及测井资料岩石物理分析,建立了适用于研究区地质特点的岩石物理模型,明确了不同岩性和"甜点"在纵波阻抗-纵横波速度比岩石物理解释图版中的弹性参数特征,为叠前反演统计分析和储层解释提供了基础资料。基于岩石物理分析,将叠前AVO反演和地质统计学建模相结合,采用高分辨率叠前地质统计学反演技术进行了致密薄储层预测及水平井优化设计,部署实施的5口水平井油砂钻遇率平均95%以上,表明基于地震岩石物理分析的叠前地质统计学反演技术是致密薄储层预测的有效手段。

The reservoirs of the Qingshankou formation in the Qijia area have developed tight sand oil in Songliao basin,but the reservoirs exhibit complex lithology and poor petrophysical properties.The reservoirs are thin interbed, and the distribution of the P-wave impedance of sandstone and shale under investigation overlaps to some extent. The reservoir characterization of the tight thin-bed reservoirs using cros-swell seismic data faces great challenges. Moreover/4sweet spots delineation and horizontal well placing are limited In view of the situa-tion, through the integrated analysis of core plug ultrasonic laboratory results and well logging data, a rock physics model suitable for the tight sand reservoirs of interest is defined Under the model,a rock physics template is established,and the characterizations of different li-thologies and “sweet spots” are revealed in the crossplot of the P-wave impedance and The results greatly support not only thegeostatistical analysis in the prestack inversion procedure, but also the quantitative interpretation of the outputs of the seismic inversion. Through combining prestack AVO inversion and geostatistical modeling, a high-resolution prestack geostatistical inversion is used for tight thin-bed reservoir characterization and horizontal well optimum design in the region of interest. Results showed that the average drilling rate of oil-bearing sandstones in the five horizontal wells was above 95 %, which verified the effectiveness of the integrated rock physics analysis and prestack geostatistical inversion for tight thin-bed reservoir prediction.