Abstract Condensate banking is the most challenging engineering problem in the development of gas-condensate reservoirs where the condensate accumulation can dramatically reduce the gas permeability resulting in impairment of wells productivity. An accurate assessment of condensate banking effect is important to predict well productivity and to diagnose well performance. Traditionally, Darcy law, combined with relative permeability models, has been used for modelling condensate banking effect in conventional reservoirs. This approach is also widely adopted in reservoir engineering commercial tools. However, for shale gas-condensate reservoirs, the gas flow deviates from Darcy flow to Knudsen flow due to the very small pore size in shale matrix (3–300 nm), compared to conventional reservoirs (10–200 μm). This gas flow is highly dependent on pore size distribution and reservoir pressure. In this paper, the effect of condensate saturation on Knudsen flow in shale matrix kerogen is investigated using a 3D pore network with a random pore size distribution. The Knudsen flow is incorporated at the pore level and gas permeability is evaluated for the whole network. In addition, the pore distribution effect in terms of log-normal mean and standard deviation is investigated. The concept of relative permeability in Darcy flow is extended to Knudsen flow by defining a new parameter called relative correction factor ξ r e l to evaluate the effect of condensate banking on Knudsen flow. This parameter can be employed directly in reservoir engineering tools. Simulation results showed that the relative correction factor is not only dependent on condensate saturation but also on pressure. This is due to the impact of pressure on the contribution of pore size ranges into the gas flow. In addition, results showed the effect of the pore size distribution where the standard deviation controls mainly the behaviour of Knudsen flow under condensate saturation. Disregarding this effect can lead to an overestimation of Knudsen flow contribution in well production under condensate banking effect.

中文翻译：

---

页岩储层的凝析气流动建模

摘要 凝析油淤积是凝析气藏开发中最具挑战性的工程问题，凝析油的聚集会显着降低气体渗透率，从而影响井产能。对凝析油围堰效应的准确评估对于预测油井产能和诊断油井性能非常重要。传统上，达西定律与相对渗透率模型相结合，已被用于模拟常规储层中的凝析油堤效应。这种方法在油藏工程商业工具中也被广泛采用。然而，对于页岩气-凝析油储层，与常规储层（10-200 μm）相比，由于页岩基质中的孔隙尺寸非常小（3-300 nm），气体流从达西流偏离到克努森流。这种气流高度依赖于孔径分布和储层压力。在本文中，使用具有随机孔径分布的 3D 孔隙网络研究了凝析油饱和度对页岩基质干酪根中 Knudsen 流动的影响。Knudsen 流被纳入孔隙水平，并评估整个网络的气体渗透率。此外，研究了对数正态均值和标准偏差方面的孔隙分布效应。Darcy 流中相对渗透率的概念通过定义一个称为相对校正因子 ξ rel 的新参数扩展到 Knudsen 流，以评估凝结水堆积对 Knudsen 流的影响。该参数可直接用于油藏工程工具中。仿真结果表明，相对修正系数不仅取决于凝析油饱和度，还取决于压力。这是由于压力对孔径范围对气流的贡献的影响。此外，结果显示了孔径分布的影响，其中标准偏差主要控制冷凝水饱和下 Knudsen 流动的行为。忽视这种影响会导致高估在凝析油堤效应下油井生产中的 Knudsen 流量贡献。