Performance of Vertical Transient Two-Phase Flow Models Applied to Liquid Loading in Gas Wells

Abstract

Abstract Liquid loading in gas wells is triggered when the produced gas loses the ability to lift the co-produced liquids up the tubing. These co-produced liquids accumulate at the bottom of the wellbore, causing a higher back-pressure on the formation which reduces production from the reservoir, and may ultimately kill the well. As this phenomenon is transient in nature, it requires transient modeling for proper characterization of its associated flow features and prediction of future well performance. However, there is a lack of dedicated models that can mimic the transient behavior which is typical of liquid loading. This paper describes the modeling effort carried out to investigate the liquid loading sequence in a synthetic gas well using a commercial package for transient multiphase flow modeling and two research codes (one for steady-state flow, and one for transient flow). The results of the simulations for the pressure gradient were compared against experimental data. The experimental runs were performed using a facility which has a transparent vertical test section of 43 m in length, and 0.04859 m ID. Air and water were used as working fluids. The results highlighted the capabilities and limitations of these simulators when evaluating liquid loading in gas wells. While good agreement was observed among all three codes for the modeled pressure drop, considerable divergence was noted in terms of the recognized flow regimes and the modeled liquid holdup. According to the literature and field observations, liquid loading is fundamentally related to the transitions between flow regimes and the associated transient flow mechanisms; thus, miscalculating them would inevitably lead to an erroneous prediction of well performance. Based on the results from the simulations, the potential effect of mis-modeling liquid loading on flow assurance and production optimization are investigated, and the need for further developments in transient multiphase modeling is discussed.