Prediction of two-phase annular flow pressure drop in wellbores by incorporating liquid film modeling

Addressing the challenge of predicting pressure drop in annular flow under high gas–liquid ratio conditions, this study overcomes the limitations of conventional empirical models by proposing a novel mechanistic model based on liquid film momentum conservation. The model analytically solves the liquid film momentum equation to predict film thickness, thereby avoiding reliance on purely empirical correlations. A key innovation is the introduction of a separated-phase liquid holdup characterization method that distinguishes between the contributions of the liquid film and entrained droplets, thus resolving the issue of holdup overestimation inherent in traditional models. By integrating advanced sub-models, a unified and self-contained framework for pressure drop prediction is established. Validation against experimental and field data demonstrates the model's superior performance: its prediction accuracy for liquid film thickness significantly exceeds that of six established classical models. Based on field data from 132 annular flow wells, the mean absolute error in pressure drop prediction is reduced by 34.7% compared to the best-performing benchmark model. In case studies involving a deep well of 5775 m and ten wells under different operating conditions, the pressure prediction errors are only 0.297 MPa and 2.05%, respectively. Combining theoretical rigor with engineering practicality, the proposed model offers a reliable tool for accurate well design and production analysis in the oil and gas industry.