Abstract: Due to the complex microscopic pore structures of low-permeability carbonate reservoirs in the Middle East, the mechanisms of gas injection and water-alternating-gas (WAG) flooding remain unclear. In this study, multi-scale digital core samples were constructed using micro-focus computed tomography (CT) scanning. A pore network model was extracted based on the representative elementary volume to quantitatively characterize microscopic pore structure features. The online nuclear magnetic resonance experiments under high-temperature and high-pressure conditions were conducted to elucidate flow mechanisms of crude oil. Furthermore, the Volume of Fluid (VOF) algorithm was innovatively applied to establish an oil-gas-water three-phase flow model, thereby clarifying the impact mechanisms of different injection media, miscibility degrees, and injection strategies. The results indicate that the average porosity of the reservoir is 16.3 %, with an average permeability of 5.7 mD, and it is dominated by medium and small pores with radii below 1.5 μm. The pore-throat radius exhibits a multimodal distribution, indicating good connectivity but pronounced heterogeneity. Compared to continuous CO₂ miscible flooding, the WAG process increases flow resistance through the Jamin effect induced by water occupying larger pore throats, which forces the gas to divert into medium and small pores, thereby enhancing oil recovery in these smaller pores from 2.1 % to 18.4 %. Enhancing the miscibility is key to suppressing viscous fingering. The WAG process effectively reduces the proportion of difficult-to-mobilize, isolated droplet-shaped residual oil through synergistic mechanisms, thereby inhibiting dominant flow channels and delaying gas breakthrough. The experimental insights guided the classified adjustment of development strategies for the M reservoir based on controlling factors. Specifically, the injection strategy in Zone Ⅰ was shifted to miscible WAG with extended water injection cycles, while Zone Ⅱ was switched to continuous CO₂ miscible flooding. The injection-to-production ratio was optimized for Zone Ⅲ. These adjustments have demonstrated good performance in field applications.

Key words: low-permeability carbonate reservoir, microscopic pore structure, gas injection miscible flooding, oil displacement mechanism, digital core, microscopic flow simulation

摘要： 中东地区低渗碳酸盐岩储层微观孔隙结构复杂,注气和水气交替驱油机理不明。采用微焦点CT扫描技术构建多尺度数字岩心,基于表征单元体提取孔隙网络模型以定量表征微观孔隙结构特征;开展高温高压在线核磁共振实验,揭示原油动用规律;应用流体体积(VOF)算法建立油-气-水三相渗流模型,阐明不同注入介质、混相程度及注入方式的影响机制。研究结果表明:①储层平均面孔率为16.3 %,平均渗透率为5.7 mD,孔喉以1.5 μm以下的中孔、小孔为主。孔喉半径呈多峰分布,连通性良好但非均质性较强。②相比连续CO₂混相驱,水-CO₂交替混相驱利用水相占据大孔道产生的贾敏效应增加了渗流阻力,驱使气体转向进入中、小孔隙,将小孔隙内原油采收率从2.1 % 提升至18.4 %。③增强混相程度是抑制指进现象的关键之一,水气交替混相驱通过协同机制有效减少了难动用的孤滴状剩余油比例,抑制优势通道发育并延缓气窜。④深化微观驱油机理认识,指导M油藏基于主控因素开展开发方式的分类调整。Ⅰ区转为长注水周期水气交替混相驱,Ⅱ区转为CO₂混相驱,Ⅲ区优化注采比,矿场应用效果良好。

关键词: 低渗碳酸盐岩油藏, 微观孔隙结构, 注气混相驱, 驱油机理, 数字岩心, 微观渗流模拟