CFD Analysis of Effect of Viscosity and RPM on an Electrical Submersible Pump
MetadataShow full item record
Computational fluid dynamics analysis (CFD) has been employed to study the flow behavior, and dependency of pump efficiency and head on the operating conditions of the pump and fluids pumped. Commercially available ANSYS Fluent is the tool used in this study for simulation of flow through the pump. A single stage of an ESP (Electrical submersible pump) WJE-1000, manufactured by Baker Hughes Ltd., is modeled and investigated. A three dimensional single phase flow has been considered for the numerical simulations to study pump performance, over a range of flow rates, viscosities and rotational speeds. It is shown that the pump speed does not affect the head coefficient and efficiency when plotted against flow coefficient, or in a more general sense, dimensionless parameters, when plotted against each other, are not affected by a change in rpm. Also, efficiency for all the cases can be represented on a single curve which includes flow coefficient and rotating Reynold’s number. It is also shown that careful selection of the CFD model is indeed very important and more work needs to be done in that regard. The ramifications of these results are very significant. The affinity laws have been modified to include the effects of viscosity. Hence a single head coefficient curve and a single efficiency curve (power required to operate the pump) can represent the pump performance over the entire flow rate and pump speed envelop. This will allow operators to be able to predict changes in pump performance with varying fluids and pump speeds. An additional benefit is that a pump need not be tested over a wide range of fluids, but only at two viscosities in order to obtain power law coefficient on Rew. Once the relationship has been determined for a specific pump design, it can be published and utilized for all operating conditions.
Agarwal, Rahul (2017). CFD Analysis of Effect of Viscosity and RPM on an Electrical Submersible Pump. Master's thesis, Texas A & M University. Available electronically from