Loading...
Onset of liquid loading in large diameter inclined pipes
Rastogi, Ayush
Rastogi, Ayush
Citations
Altmetric:
Advisor
Editor
Date
Date Issued
2020
Date Submitted
Collections
Research Projects
Organizational Units
Journal Issue
Embargo Expires
Abstract
The challenges related to liquid loading have been observed during flow-back after hydraulic fracturing, as well as during the production phase, and are further aggravated with the high inclination angles found in deviated wellbores. Accurate prediction of the onset of liquid loading is of great importance in terms of production design and operation optimization. An experimental study was carried out to investigate the onset of liquid loading and a unified mechanistic model was developed to predict critical gas velocity for large diameter pipes. The experimental setup includes a 6-inch diameter acrylic test section which can be inclined from 0° to 90°. The study involves two-phase air-water flow in low-liquid-loading conditions to simulate a gas well. The critical gas velocity associated with the onset of liquid loading shows a strong function with the inclination angle and liquid flow rate in the current experimental study. A comparison with previous experimental data reveals that it also depends on the gas density and pipe diameter – i.e., it decreases with increasing gas density and increases when pipe diameter increases. A comprehensive model evaluation was conducted in the current study. It showed a large discrepancy for inclination angles higher than 45° and only a few existing models capturing all the effects of inclination angle, liquid flow rate, pressure, and pipe diameter. The experiments in this study provide new insights into the onset of liquid accumulation in large diameter deviated wells. The new mechanistic model fills the critical gap to enhance accuracy when predicting the onset of liquid loading especially for deviated and large-diameter wells. Accurate prediction of the onset of liquid loading requires a good prediction of pressure gradient and liquid holdup for segregated flow. In addition to the mechanistic model for the onset of liquid loading, the current study proposes a new hybrid-physics-data-driven algorithm to predict pressure gradient and liquid holdup for segregated flow. The current most widely accepted model for segregated flow is a two-fluid-model which treats gas and liquid phase separately by incorporating gas and liquid mass and momentum conservation equations. However, due to the empirical nature of closure relationships such as the wall and interfacial friction factor correlations, its application is restricted to the range in which the experiments are conducted. Using machine learning algorithms, another model is proposed which successfully captures the complex, dynamic and non-linear relationships between the friction factors and the flowing conditions. The model couples the data-driven wall and interfacial friction factors with the two-fluid model, keeping the physics of the segregated flow. The results are found to be a big improvement under a wide range of conditions when compared with the existing friction factor correlations found in the literature. Using the hybrid model, a detailed statistical evaluation of multiple studies is performed which demonstrates an improvement in prediction accuracy of critical gas velocity. This work provides a workflow to reduce the dependence on empirically derived relationships and adds value by determining pressure gradients and volume fraction more precisely for the optimal design of pipeline systems.
Associated Publications
Rights
Copyright of the original work is retained by the author.