Behavior of Electric Submersible Pumps (ESP) handling two-phase flow is a subject of primary concern, especially in the petroleum industry, where significant amounts of free gas may be found in oil wells production. Several attempts have been made in order to predict the performance of such kind of pumps, nevertheless, limited success has been achieved due to the complexity of the flow dynamics inside the impeller. Geometry, gas void fraction (GVF) and suction pressure seem to be the main parameters affecting ESP performance. Furthermore, the higher the GVF of the mixture is, the higher the degradation of head that it is experimented by the pump. This complex phenomena is not well understood so far. In this work, a two fluid model is used in 3D CFD simulations carried out in order to obtain the pressure, liquid velocity and gas velocity fields as well as the GVF distribution in an ESP impeller of known geometry; using flow rates, bubble diameter and GVF at the suction as independent variables and an incompressible fluid hypothesis. The gas pocket in the impeller blade reported from other researchers is obtained and comparison with experimental results has shown good agreement. The obtained variables from the simulations are the cornerstone that allows the prediction of the performance curve of the pump for different GVF and in this way, estimate the head degradation of the pump.

A numerical model of transient two-phase in a wellbore after startup of electrical submersible pump (ESP) was developed in the present study. The model consists of three modules. The first module, the wellbore model, predicts two-phase gas-liquid flow in the wellbore. The second module describes performance characteristics of centrifugal pumps. The third module, the reservoir model, predicts the variable well inflow after startup of the pump. The wellbore model is based on a drift-flux model (three equation model) of two-phase flow. The relative velocity is determined using a flow regime-dependent slip correlation or a mechanistic steady flow model. Field data collected during a startup of an ESP well were analyzed. This analysis showed that the response of the system to the ESP startup is governed by dynamic interacticn between the well and reservoir. The results obtained can be used for the development of production startup or restart procedures for wells equipped with ESP systems.