A State-of-the-Art Computational Fluid Dynamics Simulation for Erosion Rates Prediction in a Bottom Hole Electrical Submersible Pump

Abstract Electrical submersible pumps (ESP) are used extensively in the production of oil and gas. A multi-vane pump is a special type of ESP designed to handle high gas volume fractions and eliminate gas lock. Erosion in ESP systems due to sand production is a major problem that can cause equipment failure. Extensive numerical study has been performed to simulate the flow of sand particles inside a multi-vane ESP pump. This study utilizes two computational fluid dynamics (CFD) erosion models. The first CFD erosion calculation model, ERC-2003, was empirically developed in 2003 by Russell et al. (Russell 2004), and applied in this paper to predict erosion rates inside an ESP. The second CFD erosion model, ERC-2008, is a modification developed by Ronnie Russell in 2008 to broaden the accuracy range to low-velocity flow. A state-of-the-art erosion prediction technique is presented in this paper. The generalized CFD code utilized in the simulations in this paper was ANSYS Fluent for erosion prediction. Discrete phase models (DPM) and Eulerian-Eulerian granular models were compared, and the pump pressure head was used to validate the hydrodynamic models. Results show that the pump diffuser casing and impeller blades' pressure sides consistently have the highest erosion rates. In addition, the Eulerian-Eulerian granular model shows a more homogenous distribution of sand concentration in the pump compared to the DPM model. The main difference between the two models is that the Eulerian- Eulerian granular model considers the two phases and solves for each phase separately from the beginning of the simulation, while the DPM model stops the fluid iterations to carry out its discrete particle tracking iterations. Therefore, the Eulerian- Eulerian granular model can predict erosion in turbomachinery more accurately compared to the DPM model.