CFD enabled design improvement of a submersible pump

Background

This project involved the numerical analysis of a submersible pump where the first step was to accurately define the problem. Once this was done the objective of the analysis became clear. We then identified and setup the technical approach and a process for completing the analysis followed by a comparison of data from simulations with available data from the physical testing of the pump. This helped us recommend design improvements and run further simulations on the improved design to predict the behavior of the modified pumps.

Problem definition

Initially we needed to define the problem or goal of the simulations to be performed. We began by characterizing the single stage submersible pump using CFD.

For this we performed numerical analyses on the existing pump where we matched the pressure head and efficiency with the existing available data.

This allowed us to characterize the pump over a range of flow rates. Once we were able to predict the behaviour of the pump across different parameters, we were in a position to recommend design improvements for performance enhancements.

Technical approach

We saw that the pump geometry and solution exhibited similarity in angular direction and so we modeled only the rotational sector of the complete geometry with use of periodic boundary conditions.

Conservation equations for mass and momentum were re solved in three dimensions using turbulence models from k-ε family.

The impeller rotation was specified using two approaches mainly the Mixing plane model (MPM) which uses rotating reference frame and the Time accurate sliding mesh model (SMM) which captures the rotational motion of the impeller.

To get accurate results, the settings for the simulation software needed to be accurate. For this particular simulation, a Grid independence analysis was performed to come up with suitable mesh type and size. Different turbulence models were compared to make the most appropriate choice. The total pressure developed and torque supplied was calculated to predict pressure head and efficiency.

Software tools used were Pro-E for geometry creation, Gambit/ Tgrid for mesh generation and Fluent for CFD analysis.

Comparison with available experimental data

Simulations were performed on the existing pump and a characterization was done for the existing design. The simulations were performed over a range of flow rate conditions based on best practices.

The graph of Overall Efficiency (%) v/s Flow (GPM) was developed to visually correlate the simulation results with the experimental results. The graph of Pressure rise (PA) v/s Flow (GPM) was developed to visually correlate the simulation results with the experimental results.

Recommendations made for improved design

The mandate of the project was to suggest design improvements using CFD. This was to be achieved mainly through improvements in Pressure head and increased efficiency. The constraints we faced in the project were the Rotational speed of the impeller and Maximum outer diameter of the stator.

The philosophy behind the design improvements was based on the study of Meridional pressure distribution in the system at desired operating points for:

  1. Reducing the entry loss at the stator
  2. Reducing the loss during fluid flow in the stator
  3. Generating more pressure head across the impeller

Based on the recommendations we made the stator and impeller geometries were modified to meet the objective of higher pressure head and higher efficiency.

We then performed three dimensional, transient CFD simulations to evaluate the performance of the new design. Many attempts were required to come up with the shape that eventually met the desired expectations.

Conclusions

  • The CFD analysis of a single stage of submersible pump was successfully performed and the results obtained showed good agreement with the experimental data.
  • The efficiency curve in the new design was much flatter around the operating point.
  • The new design showed significant performance improvements around the desired operating point and was able to meet the requirements set at the start of the project.