新疆石油地质 ›› 2026, Vol. 47 ›› Issue (1): 116-125.doi: 10.7657/XJPG20260113

• 应用技术 • 上一篇    

基于内部充填模型的断控型油藏建模方法及应用

田元1(), 王佳乐2, 乐平1(), 赵黎明2, 张莹2, 范庆振1, 耿洁1, 牟昱1   

  1. 1.西南石油大学 a.石油与天然气工程学院;b.油气藏地质及开发工程全国重点实验室成都 610500
    2.中国石化 西北油田分公司 采油一厂新疆 库尔勒 841000
  • 收稿日期:2025-06-17 出版日期:2026-02-01 发布日期:2026-01-23
  • 通讯作者: 乐平 E-mail:592102307@qq.com;yuepingaa@126.com
  • 作者简介:田元(2001-),男,甘肃天水人,硕士研究生,油气藏数值模拟,(Tel)18200535838(Email)592102307@qq.com
  • 基金资助:
    四川省科技计划项目(2024NSFSC2012)

Modeling Method for Fault-Controlled Reservoirs Based on Internal Filling Model and Its Application

TIAN Yuan1(), WANG Jiale2, YUE Ping1(), ZHAO Liming2, ZHANG Ying2, FAN Qingzhen1, GENG Jie1, MOU Yu1   

  1. 1. Southwest Petroleum University, a. School of Petroleum Engineering; b. State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation, Chengdu 610500, China
    2. No.1 Oil Production Plant, Northwest Oilfield Company, Sinopec, Korla, Xinjiang 841000, China
  • Received:2025-06-17 Online:2026-02-01 Published:2026-01-23
  • Contact: YUE Ping E-mail:592102307@qq.com;yuepingaa@126.com

摘要:

塔河油田奥陶系碳酸盐岩储集层储集空间复杂且非均质性强,为储集层表征和建模带来挑战。提出断控型油藏储集体内部充填模型建模方法,采用“岩性-构造双约束、分级建模、分类融合”思路构建高精度模型。首先基于断控型油藏成因与储集体特征,将储集体划分为溶洞、溶孔和裂缝3类结构单元分别雕刻。溶洞与溶孔采用截断值确定性建模与人工修正建立轮廓,结合测井岩相划分储集层,并通过深度神经网络与地震反演数据融合实现充填结构精细表征。断裂及裂缝系统按尺度分级刻画:大断裂通过蚂蚁体追踪与相干属性识别;中—小断裂基于绕射波张量蚂蚁体结合体积约束建立;裂缝采用离散裂缝网格建模精细表征。继而依据成因机制-储渗功能-工程应用三元耦合原则进行多尺度模型融合,最终建立基质-裂缝双重介质地质模型。以塔河油田某单元为例,建立的高精度充填模型在静态储量与生产动态验证中与生产数据高度吻合,有效支撑了剩余油挖潜与开发调整模拟,显著提升了此类油藏数值模拟的可靠性。

关键词: 塔河油田, 断控型油藏, 地质建模, 内部充填模型, 神经网络, 地震反演

Abstract:

The complex and highly heterogeneous reservoir space in the Ordovician carbonate reservoirs in Tahe oilfield poses significant challenges to reservoir characterization and modeling. This paper proposes a modeling method for fault-controlled reservoirs based on internal filling model, enabling the construction of a high-precision model with the methodology of lithology-structure dual constraints, hierarchical modeling, and categorical integration. First, depending on the genesis of fault-controlled reservoirs and the characteristics of reservoir architectures, the reservoirs are divided into three types of structural units: vugs, pores, and fractures. Vugs and pores are delineated by using deterministic modelling with cutoff value and corrected manually to define their boundaries; combined with log-derived lithofacies, the reservoirs are identified, and the internal architecture is finely characterized by integrating deep neural network with seismic inversion data. Faults and fractures are characterized at different scales: large faults are identified through ant-tracking and coherence attributes; small-medium faults are defined by diffraction tensor ant-tracking with volume constraints; and fractures are finely described via discrete fracture network (DFN) modeling. Next, based on the coupling of genetic mechanism, storage-permeability function, and engineering application, multi-scale model integration is performed. Finally, a matrix-fracture dual-medium geological model is established. Application in a unit of the Tahe oilfield has demonstrated that the high-precision filling model yields the validated results of static reserves and production performance in good agreement with the actual production data. The proposed model can effectively support the simulations of remaining oil recovery and development adjustment, and significantly enhance the reliability of numerical simulation for such reservoirs.

Key words: Tahe oilfield, fault-controlled reservoir, geological modeling, internal filling model, neural network, seismic inversion

中图分类号: