Optimization of methods for liquid loading prediction in deep condensate gas wells

• Different distribution of the critical flow rate and surface tension are considered.
• New model to calculate high GLR wellbore temperature and pressure is established.
• Methods to calculate low GLR wellbore temperature and pressure are chosen.
• Modified Li's model and kinetic energy factor method are chosen to predict loading.

Liquid loading is a serious problem in hydrocarbon production from gas wells in the Yakela-Dalaoba deep condensate gas reservoir. The applicability of frequently-used methods for predicting liquid loading still needs to be demonstrated. For this purpose, the applicabilities of critical flow rate methods and the kinetic energy factor method are evaluated and optimized. The fundamental principles of each method for predicting liquid loading and differences among them are discussed. Different distributions of the critical flow rate and surface tension along the wellbore are considered in order to improve original critical flow rate models. For the Yakela-Dalaoba gas reservoir, the modified pseudo single-phase coupling model used for calculating the pressure and temperature distribution in high gas-liquid ratio wellbores is established, and methods suitable for calculating the pressure and temperature distribution in low gas-liquid ratio wellbores are optimized through error analysis. Results show that the modified pseudo single-phase coupling model is suitable for calculating the pressure and temperature distribution in high gas-liquid ratio wells, while Hagedorn-Brown's method and Hasan's method are suitable for low gas-liquid ratio wells. The modified critical flow rate models increase precision of liquid loading prediction compared to original models. The modified Li's model and the kinetic energy factor method which sets threshold value as 6 are of highest precision. Both of them are suitable for predicting liquid loading of gas wells in the Yakela-Dalaoba deep condensate gas reservoir.