Experiments, Performance Analysis and Predictive Modeling of Electric Submersible Pumps Under Viscous Oil Flow Condition

Abstract Accurate performance prediction of Electric Submersible Pumps (ESPs) operating under viscous fluid conditions remains a significant challenge in artificial lift operations. This study evaluates ESP performance across a broad range of oil viscosities (1–600 cP) and varying water cuts, aiming to develop and validate improved predictive models that address limitations of existing empirical correction methods. Two ESPs were experimentally tested at multiple rotational speeds, following a rigorous procedure involving oil property characterization, systematic loop testing, and repeated validation trials. Key parameters, including pump flow rate, developed head, and brake horsepower (BHP), were systematically recorded and analyzed. Results indicate substantial shortcomings in conventional empirical correction methods and dimensionless analyses, particularly at low flow conditions. Both tested ESPs exhibited clear performance degradation with increased oil viscosity, notably highlighting limitations in pumps with fewer stages. This study also introduces a new high-viscosity ESP performance calibration model based on Turzo (2000) model structure. The novel model developed in this study, validated using experimental data of nine ESPs from Tulsa University Artificial Lift Projects (TUALP) and literature sources, showed significantly improved accuracy in predicting ESP behavior for viscosities up to 650 cP. The model provides critical insights for improved ESP design, operational decision-making, and predictive performance modeling under challenging, high-viscosity conditions. The investigation further identified important variability in head and BHP factors with flow rate and viscosity. The findings underscore the need for additional testing with higher-viscosity fluids and diverse ESP designs to further enhance model robustness.