Automated managed-pressure drilling (MPD) is a method to enhance downhole pressure-control performance and safety during drilling operations. It is becoming more common to use model-based simulation for the evaluation of pressure-control systems designed for MPD automation before using those in the field. This demands a representative hydraulics-simulation model that captures the relevant aspects of a drilling system. This paper presents such a model and an approach to numerically implement that model for simulation studies. The complexity of this simulation model should be limited, first, to support effective numerical implementation and, second and most importantly, to allow for the analysis of the behavior and performance of the automated pressure-control systems during the controller-design phase. To this end, aspects of a drilling system that can considerably affect the performance of the automated MPD system are captured in the model. This hydraulics model incorporates both the distributed and multiphase-flow nature of a drilling system. Moreover, it captures nonlinear boundary conditions at the inlet of the drillstring, at the drill bit, and choke manifold, and also the variations in the cross-sectional area of the flow path. Model validations against field data from real-life MPD operations and simulations of industry-relevant scenarios indicate that these aspects are effectively captured in the model and preserved during the numerical implementation.

自动管理压力钻井（MPD）是一种在钻井作业过程中提高井下压力控制性能和安全性的方法。在现场使用之前，使用基于模型的仿真来评估为MPD自动化设计的压力控制系统变得越来越普遍。这就需要一个具有代表性的水力模拟模型来捕获钻井系统的相关方面。本文介绍了这样一个模型，以及在仿真研究中以数字方式实现该模型的方法。首先，应限制此仿真模型的复杂性，以支持有效的数值实现，其次，也是最重要的是，允许在控制器设计阶段对自动压力控制系统的行为和性能进行分析。为此，在模型中捕获了可能严重影响自动化MPD系统性能的钻井系统的各个方面。该液压模型兼具了钻井系统的分布式和多相流特性。此外，它捕获了钻柱入口处，钻头处和节流歧管处的非线性边界条件，以及流路横截面积的变化。针对来自实际MPD操作的现场数据进行的模型验证以及与行业相关的情景的仿真表明，这些方面已在模型中有效捕获，并在数值实施期间得以保留。此外，它捕获了钻柱入口处，钻头处和节流歧管处的非线性边界条件，以及流路横截面积的变化。针对来自实际MPD操作的现场数据进行的模型验证以及与行业相关的情景的仿真表明，这些方面在模型中得到了有效捕获，并在数值实现期间得以保留。此外，它捕获了钻柱入口处，钻头处和节流歧管处的非线性边界条件，以及流路横截面积的变化。针对来自实际MPD操作的现场数据进行的模型验证以及与行业相关的情景的仿真表明，这些方面已在模型中有效捕获，并在数值实施期间得以保留。