An improved model to predict liquid holdup in vertical gas wells

• Published in: Journal of petroleum science & engineering Vol. 184; p. 106491
• Main Authors: Luo, Chengcheng, Zhang, Liehui, Liu, Yonghui, Zhao, Yulong, Xie, Chunyu, Wang, Lang, Wu, Pengbo
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.01.2020
• Subjects: Empirical models; Gas wells; Liquid holdup; Mechanistic models; Pressure gradient; Mechanistic models; Pressure gradient; Liquid holdup; Empirical models; Gas wells
• ISSN: 0920-4105; 1873-4715
• DOI: 10.1016/j.petrol.2019.106491

Accurate prediction of the pressure gradient is crucial for gas-production design and gas well optimization. The most important role in modeling the pressure gradient is generally to calculate the liquid holdup accurately. Currently, numerous mechanistic and empirical models have been developed and modified to deal with all kinds of flow conditions. Due to the chaotic feature in the churn flow, it is difficult for mechanistic models to model this flow pattern accurately. Simplified assumptions for this pattern may lead to large deviations. For empirical models, the range of experimental parameters limits their application. In this study, a variation of liquid holdup with superficial gas velocity is analyzed in order to simplify the liquid-holdup correlation and improve its accuracy. Based on the annular/churn-flow and bubble/slug-flow transition points, a new empirical model is established using a power-law curve. Since the superficial gas velocity and liquid holdup corresponding to the two transition points are obtained from analytical or semi-analytical models, the proposed model is not subject to the experimental ranges, unlike other empirical models. In addition, it avoids dealing with modeling churn flow. The new model has been validated against the laboratory and field data from published studies. For comparison, some widely used models are also evaluated. The result shows that the proposed model, which is capable of predicting the pressure gradient at different pipe sizes, liquid velocities, and gas velocities, has better performance compared with other models. •A new power-low model was established for predicting liquid holdup in vertical gas wells.•The proposed empirical model isn't subject to the experimental ranges, unlike other empirical models.•The comparison against lab and field data shows that the proposed model has better performance compared with other models.