地质统计学反演在薄砂体储层预测中的应用

Application of geostatistical inversion in thin sandstone reservoir prediction

 张义¹ 尹艳树 2 秦志勇³

长江大学地球科学学院，湖北 武汉 430100 山东省沉积成矿作用与沉积矿产重点实验室，山东科技大学，山东 青岛 266590 中国石油大港油田滨港集团博弘石油化工有限公司，天津 300280

School of Geosciences, Yangtze University, Wuhan 430100, China Provincial Key Laboratory of Sediment Mineralization and Sediment Mineral, Shandong University of Science and Technology, Qingdao 266590, China Bohong Oil Petrochemical Co. Ltd., Bingang Group, Dagang Oilfield Company, PetroChina, Tianjin 300280, China

 研究区储层具有厚度薄、单砂体厚度小、砂体相变快等特点，常规约束稀疏脉冲反演受制于地震频带，很难精确地刻画薄储层。地质统计学反演结合了地质统计学建模和地震反演的优势，充分发挥了地震横向覆盖面广、分辨率高、测井纵向采样密集的特点，可以获得高分辨率的反演结果。文中在测井资料质控和储层岩石物理特征分析、约束稀疏脉冲反演、地质统计学参数分析、随机模拟的基础上，确定了运用以地质统计学反演为核心的储层预测方法对核三段Ⅳ—Ⅵ油组进行储层预测，并与确定性反演、随机模拟方法生成的预测结果进行了对比分析。结果表明，地质统计学反演极大地提高了预测结果的垂向分辨率，可以有效识别出薄层砂体，为研究区有利储层预测提供了依据。

 With the characteristics of thin reservoir, small thickness of single sand body and fast sandstone facies change, the limited earthquake-band of conventional constrained sparse pulse inversion cannot depict thin reservoir accurately. Combining the advantages of geostatistical modeling and seismic inversion, using the wide coverage, high resolution and fine logging vertical resolution of transverse wave, geostatistical inversion ensures high resolution inversion results. Therefore, based on the evaluation of well logging formation and analysis of lithology characteristics of the reservoir, constrained sparse spike inversion, and geological parameter analysis of statistics and stochastic simulation, geostatistical inversion is cored for reservoir prediction and is applied to He 3 Member 4-6 oil groups, and the results are contrasted with constrained sparse spike inversion method and stochastic simulation. Results show that geostatistical inversion significantly improves the prediction results of vertical resolution, which can effectively identify the thin sandstone and provide powerful guidance for the favorable reservoir prediction.

10.6056/dkyqt201505005