Physics-informed neural networks (PINNs) represent an emerging computational paradigm that incorporates observed data patterns and the fundamental physical laws of a given problem domain. This approach provides significant advantages in addressing diverse difficulties in the field of complex fluid dynamics. We thoroughly investigated the design of the model architecture, the optimization of the convergence rate, and the development of computational modules for PINNs. However, efficiently and accurately utilizing PINNs to resolve complex fluid dynamics problems remain an enormous barrier. For instance, rapidly deriving surrogate models for turbulence from known data and accurately characterizing flow details in multiphase flow fields present substantial difficulties. Additionally, the prediction of parameters in multi-physics coupled models, achieving balance across all scales in multiscale modeling, and developing standardized test sets encompassing complex fluid dynamic problems are urgent technical breakthroughs needed. This paper discusses the latest advancements in PINNs and their potential applications in complex fluid dynamics, including turbulence, multiphase flows, multi-field coupled flows, and multiscale flows. Furthermore, we analyze the challenges that PINNs face in addressing these fluid dynamics problems and outline future trends in their growth. Our objective is to enhance the integration of deep learning and complex fluid dynamics, facilitating the resolution of more realistic and complex flow problems.

Drill-string axial vibration at the surface technology is proposed to reduce the friction between the drill-string and the borehole wall, and to improve load transfer efficiency, the rate of penetration (ROP), and the extended-reach limit of a horizontal well. An analytical framework utilizing the “soft-string” model is constructed. The results obtained from numerical simulations reveal that during the slide drilling operation, the drill-string experiences an axial stick–slip motion, and the weight on bit (WOB) undergoes periodic oscillations. The conventional calibration method of the WOB in the weight indicator gauge is not applicable when the ROP is low. After applying axial vibration on the drill-string at the surface, the WOB increases and becomes smooth because of a release of friction. The amplitude and frequency of the exciting force are the main factors affecting surface vibration effectiveness. There is an optimal frequency for a given case (10 Hz in this paper). This means that the conventional manual pick-up and slack-off by drillers with a high amplitude and a low frequency has little effect on friction reduction. In addition, the conventional method can bring in high risk because of its high root mean square (RMS) acceleration. Safety evaluation results indicate that the drill-string is in a safe state under most of the exciting parameters. The results verify the feasibility and advantages of the proposed technology, and lay a solid theoretical foundation for its application in real drilling applications.

Summary: Hole cleaning is an important factor in horizontal drilling; however, there is no generally accepted approach to measure the cuttings distribution downhole in practical drilling applications. This study proposes an approach to evaluate the dynamic downhole cuttings distribution quantitatively by using along-string annular pressure (ASAP) data. The relationships between cuttings distribution and pressure drop are investigated, and a pressure-driven hole cleaning model is developed by inverting the effect of cuttings on wellbore pressure drop for given flow conditions. A hybrid drilling hydraulic model is developed by combining the traditional drilling hydraulic model and a self-tuning approach to accurately calculate wellbore pressure drop without cuttings. The effect of cuttings on pressure drop is obtained by comparing ASAP data and predicted pressure data from the hybrid drilling hydraulic model. Then, the downhole cuttings distribution is obtained by inverting the pressure changes using the pressure-driven hole cleaning model. This work provides an accurate approach to evaluate real-time hole cleaning conditions and detect insufficient hole cleaning problems for real-time drilling activities if wired drillpipe is deployed. It also expands existing analysis techniques for real-time drilling data and provides new methods to make use of the expanded sensor capability from wired drillpipe.

Cuttings beds in horizontal wells significantly affect the frictional torque and drag along the drill string; however, their quantification and modeling have been relatively underexplored. To gain deeper insights into the impact mechanisms of the cuttings bed distribution on drilling mechanics, this study establishes a model linking the cuttings bed height with variations in axial and tangential forces on the drill string through experimental investigations. By integrating this model with previously developed transient cuttings transport and torque–drag models, a coupled transient hole cleaning and drill string mechanics model is constructed. This comprehensive model simulates the dynamic distribution of cuttings along the entire well trajectory and its influence on the drill string torque and drag. The results reveal that accumulated cuttings significantly reduce the weight on bit (WOB), increase the drill string torque, and cause problems related to a high equivalent circulation density (ECD). For long horizontal sections, the key to achieving effective hole cleaning lies in optimizing the design of the tripping circulation time to ensure that all cuttings are removed from the wellbore. Using the proposed coupled model, a methodology is developed to minimize the tripping circulation time by solving optimization problems within a constrained 2D domain, providing scientific guidance for drilling operations. The findings demonstrate that dynamically managing the cuttings distribution in the wellbore can significantly mitigate issues arising from insufficient hole cleaning, thereby ensuring drilling safety and efficiency. This study provides a scientific foundation for the optimized design of long horizontal well drilling operations and highlights the critical role of cuttings management in enhancing hole cleaning performance and mitigating drilling risks.