Over the past decades, directional drilling has continuously advanced to increase hydrocarbon recovery by effectively targeting high-productivity reservoirs. However, many existing approaches primarily focus on heuristic optimization algorithms. Moreover, existing models often neglect the incorporation of petrophysical attributes that can significantly impact the selection of production targets, such as the reservoir quality indicator. This article introduces a novel application of mixed-integer programming to define directional drilling paths, considering practical aspects of interest. The paths are subject to drift angle constraints and reference coordinates that align with the optimal reservoir targets. Such targets are identified using the authors’ proposed technique of maximum closeness centrality and the geologic model of hydraulic flow units. In order to evaluate the effectiveness of this approach, a realistic model of the Campos Basin in Brazil is studied. The results reveal that the highest recovery factors obtained with the proposed methodology (17%) exceed the historical average recovery factor of the studied reservoir (15.66%). We believe this study can contribute to the ongoing efforts to enhance directional drilling and maximize the production potential of offshore oil and gas reservoirs.

在过去的几十年里，定向钻井不断进步，通过有效瞄准高产油藏来提高碳氢化合物采收率。然而，许多现有方法主要关注启发式优化算法。此外，现有模型往往忽略了岩石物理属性的纳入，这些属性可能会显着影响生产目标的选择，例如储层质量指标。本文介绍了混合整数编程的一种新颖应用，用于定义定向钻井路径，并考虑到实际感兴趣的方面。这些路径受到漂移角约束和与最佳储层目标一致的参考坐标。这些目标是使用作者提出的最大接近中心性技术和水力流动单元的地质模型来识别的。为了评估该方法的有效性，研究了巴西坎波斯盆地的现实模型。结果表明，采用该方法获得的最高采收率（17%）超过了所研究油藏的历史平均采收率（15.66%）。我们相信这项研究有助于加强定向钻井并最大限度地发挥海上油气藏的生产潜力。