Summary: As an economical and efficient artificial lift method, plunger lift can be used to unload the accumulated liquids from the bottom of gas wells, which helps lower the bottomhole pressure, resulting in higher gas production rate. However, the transient flow behavior of the plunger-lift-aided production system is still not well understood due to the lack of a reliable and accurate prediction model. In this study, a transient mechanistic model is developed to simulate the comprehensive dynamic process of a plunger-lift system that is cyclically paced by a surface control valve.Starting from the Gasbarri and Wiggins (2001) dynamic plunger-lift model, four stages in the cyclic movement of a plunger can be identified and calculated using a set of specific governing equations. Considering the gas flows with a plunger in the tubing, the model can calculate the instant velocities of the plunger during its rising and falling movement. The classical inflow performance relationship (IPR) is employed as the reservoir model to obtain the fluid flow rates from the reservoir to the wellbore. The proposed new model can capture the essential parameters of plunger-lift cycles, including plunger velocity/acceleration, tubing/casing pressure, production rates, etc. Compared to previous models, the predicted rising and falling speeds of the plunger are improved. The hydrocarbon mixture properties in the gas well are computed by a compositional model in this study, which provides more accurate and reasonable predictions of tubing and casing pressure.Several parametric studies are presented in the paper. These studies will help to understand the influence of different parameters on the process of plunger lift. An appropriate combination of casing and tubing pressure should be taken into consideration. A reservoir coefficient term is introduced and defined. A larger reservoir coefficient will improve the ultimate profitability of the well by increasing the production rate at the beginning and accelerate the depletion of gas wells. If the gas/liquid ratio (GLR) is too low, liquid loading may be triggered. The parametric study shows that an adequate GLR is necessary for reliable plunger-lift performance.

Electric Submersible Pumps (ESPs) play a pivotal role in the petroleum industry, but their performance is significantly affected by factors such as oil viscosity, gas–liquid ratios, and solid content. Traditional performance prediction methods, including polynomial fitting and mechanistic modeling, often lack adaptability and efficiency, requiring extensive empirical testing. This study leverages experimental data from the viscous and gas–liquid flow tests reported in the literature to benchmark various prediction methods. This research provides a comparative analysis of traditional curve-fitting methods, mechanistic modeling, and seven machine learning approaches. A key innovation of this study is an in-depth sensitivity analysis of different machine learning methods, especially focused on neural network parameters, such as activation functions and training configurations, to assess their impact on prediction accuracy and identify optimal network designs. Furthermore, a pump testing methodology is introduced to significantly reduce testing costs while maintaining a high prediction accuracy. The findings demonstrate the advantages of machine learning over traditional methods, including an enhanced prediction accuracy, practical guidelines for efficient parameter tuning, and the ability to address incomplete pump curve data. These contributions not only highlight the value of integrating machine learning into ESP modeling and operational workflows but also pave the way for future advancements in universal modeling frameworks for diverse ESP applications.

In this study, fluid viscosity effects on LSP performance in terms of boosting pressure were numerically investigated. A water–glycerin mixture with different concentrations corresponding to varying apparent viscosities was flowed through an in-house manufactured LSP under various flow conditions, e.g., changing flow rates, rotational speeds, and fluid viscosities. The pressure increment between the intake and discharge of the LSP was recorded using the differential pressure transducer. The same pump geometries, fluid properties and flow conditions were incorporated into the numerical configurations, where three-dimensional (3D), steady-state, Reynolds-averaged Navier–Stokes (RANS) equations with a standard SST (shear stress transport) turbulence model were solved by a commercial CFD code. With the high-quality poly-hexcore grids, the simulated pressure increment was compared with the corresponding experimental measurement. The internal flow structures and characteristics within the cavities contained by the LSP impeller and diffuser were also analyzed. The good agreement between the numerical results against the experimental data verified the methodology adopted in this study.

Summary: As a natural gas well ages, liquid loading is frequently encountered, leading to the decrease of gas production rate and many other side effects, which may in turn cease the gas production. Thus, to accurately predict liquid loading onset is of significant importance in gas wells for the sake of stable production. With years of research and development in the natural gas industry, the liquid loading onset prediction models prevail in the existing literature. Based on varying mechanisms (e.g., droplet falling back, liquid film reversal, etc.), the critical gas velocities or flow rates corresponding to flow pattern transitions in gas wells can then be calculated. However, a universally validated model, whether empirical or non-empirical, that is applicable to predict the onset of liquid loading in versatile gas wells conditions (e.g., horizontal, vertical, and inclined) is, as yet, still unavailable. In this paper, we conduct a complete literature review and investigation of these existing liquid loading onset prediction models. First, we obtained detailed information of more than 600 gas wells, including well geometries, gas properties, operation conditions, and so on, from different gas fields. Then, we evaluate the validity of various liquid loading onset prediction models by use of a novel model ranking approach. To fully account for the effects of gas well properties (including but not limited to production, wellhead pressure, and pipe diameter) to the model prediction accuracy, the proposed method in this paper employs data clustering and normalization techniques, as well as the statistical relative error analysis, to rank and select the best suitable model for each specific gas well. Extensive comparison and verification of the ranking approach indicate that the proposed method provides a good reference for the rational production allocation and stable production of gas wells.