Accurate modeling of annular gas-water flow across diverse inclination angles using an advanced drift-flux correlation

Abstract

Drift-flux models are widely used for analyzing two-phase flows but often fail to accurately represent annular flow dynamics due to a conceptual mismatch. Traditional models assume phase dispersion characteristics that do not align with the velocity gradient-driven behavior of annular flows. This study introduces an adapted drift-flux model, redefining the drift velocity based on gas critical velocity, better reflecting annular flow mechanics. The proposed formulation is particularly suited for annular flows in inclined pipes, a critical consideration in industries such as oil and gas, chemical processing, and nuclear applications. The proposed drift-flux model exhibits excellent predictive capability, achieving an average error of 1.1 % when validated against experimental data and 1.5 % when benchmarked against the data generated by the Unified Mechanistic Model, within gas and liquid Reynolds number ranges of 38,250 - 1183,200 and 150 - 8000, respectively. Furthermore, statistical evaluations across both experimental and synthetic datasets confirm the model’s robustness, as reflected by the lowest mean absolute error (0.01 and 0.01), root mean square errors (0.01 and 0.03), standard deviations (0.01 and 0.02), and narrow 95 % confidence intervals (−0.008 ± 0.001 and 0.010 ± 0.001). To assess its generalizability, the proposed correlation was tested on blind experimental datasets featuring pipe diameters three times larger than those used during development, where it attained the lowest average error of 0.7 %. When applied to synthetic datasets covering a broad diameter range of 10–200 mm, the model consistently delivered the highest accuracy, maintaining an average error of 1.5 %.