A New Model for the Accurate Prediction of Critical Liquid Removal Based on Energy Balance

In order to selecting artificial lift technologies and optimizing technology parameters, the critical gas velocity and flow rate for liquid removal are important reference indexes. The conventional models of critical gas rate proposed by several scholars are obtained by stress analysis of droplets in a wellbore and subsequent empirical adjustment. Due to different empirical parameters adopted, there are tremendous variances among different model's predictions. Based on the balance relation between total droplet surface energy and gas turbulent kinetic energy, a new model without empirical parameters is proposed in this paper which adopts Turner's basic concepts, while taking into account influence of droplet diameter. Considering that the gas flow influences the diameter of droplets carried by gas, sequentially influences the efficiency of liquid removal, the new model achieves the purpose of accurately predicting the critical gas velocity and rate by calculating the droplet diameter. By comparison of case calculations using different conventional models, we find out that each model can only accurately distinguish the liquid loading situation of gas wells in a given range of liquid production rate. This explains the reason of the differences among critical gas rates calculated by different models. Well data from Li's paper are used to verify the proposed model; the distinguishing results of the new model are coincidence with the field data. In addition, the predictions of Turner's, Coleman's, Li's and the proposed model are compared with the well data collected from four gas reservoirs in China. In high production gas/water ratio (GWR > 1×104m3/m3) conditions, the deviation of the predictions of Turner's and Coleman's model is higher; in low production GWR (GWR < 1×104m3/m3) conditions, the deviation of the predictions of Li's model is higher. However, good agreement between the new model's predictions and the actual field situations is obtained in different GWR conditions. The result shows that the proposed model is superior to three former empirical models, and has the characteristics of strong adaptability and high accuracy. The new model of the critical gas velocity and flow rate for liquid removal proposed in this paper can accurately distinguish whether the gas well is liquid loading.