| Oil-water emulsion effects on a seven-stage electric submersible pump (ESP) performance is studied experimentally and numerically with Computational Fluid Dynamics (CFD) simulation. A mechanistic model for ESP head prediction is developed based on physics. Emulsion rheology inside the ESP is also studied and modeled. At different oil-water fractions, rotational speeds, temperatures the performance of the third stage was measured. Density and mass flowrate were monitored using the mass flowmeter, while the emulsion effective viscosity was derived from the pipe viscometer (PV) measurement.;CFD simulations were carried out with estimated droplet size of the dispersed phase for oil-water flows, and the results were compared with the experiments. Due to the extended use of the ESP, a higher wall roughness was used. Results show a considerable difference compared to the experiment. The difference may be partially due to neglecting leakage losses in CFD simulations. The solver takes the two phases as dispersion instead of emulsion so that the emulsion rheological behavior is not reflected. Therefore, a new liquid mixture is defined based on the combination of oil and water fractions and properties.;The mechanistic model is based on Euler's centrifugal pump equations and includes all possible losses. The model predictions agree with the experimental results at high rotation speeds but need to be improved for low rotation speeds. Emulsion rheology was modeled by considering the droplet size, turbulence, shear and stage number effects with the correspondent dimensionless numbers. Results showed a good match to the experiments, but more data are required to further validate the generality of the model. |
