摘要： 介绍了基于沉积过程的储集层随机建模基本原理及方法。与传统的储集层随机建模方法相比，基于沉积过程的随机建模方法有效地将与沉积过程有关的地质信息以及一些先验的地质知识整合到模型中，能够较为真实地再现储集层构型要素（如河道砂体、点坝、天然堤、决口扇）的几何形态和相互联系以及它们的演化过程(如改道、侧向迁移)。在井数据的条件化方面，基于沉积过程的随机建模方法采用了一种新的条件化方法，该方法提高了计算速度，同时保留了沉积体的复杂几何形态和相互联系。通过理论与实验结果研究，基于沉积过程的随机建模方法包含了更为全面的地质信息，能够更精确地表征储集层的非均质性。最后，分析了基于沉积过程的随机建模方法存在的一些问题(包括河道中线的生成问题、点坝内部侧积层的刻画问题以及井数据条件化的问题)及未来发展的方向。

关键词: 基于沉积过程, 构型要素, 随机建模, 条件化

Abstract: This paper presents the principle and method of the depositional process-based reservoir stochastic modeling.Compared with conventional stochastic modeling methods，the depositional process-based method，which efficiently integrates empirical observation and expert knowledge with respect to depositional process，enables the geologically realistic reproduction of reservoir architectural elements(such as channel fill，lateral accretion，levee，and crevasse splay)geometries and interrelationship，as well as how they are deposited over time（such as avulsion and meandering fluvial migration）.As to well conditioning，a new conditional method is adopted to improve the efficiency，while retain complicated geometries and interrelationships.Through theory and experimental results analyses，the depositional process-based reservoir stochastic modeling integrates more geological information，can represent reservoir heterogeneity more precisely.Finally，some present problems and future work are discussed such as the generation of channel centerlines，the representation of lateral accretion shale beddings in point bar，and well conditioning.

Key words: depositional process-based, reservoir architectural element, stochastic modeling, conditioning

中图分类号: 

• O213