الفهرس 
الغلافOnly 14 pages are availabe for public view
 |  | from | 108 |  |  |
| | | from | 108 | | |
Abstract
Most of the electrical submersible pumps available at the market currently are designed using water as fluid base during design phase, however while they are used to handle viscous fluids they are suffered from significant reduction at the head per stage and over all pump efficiency. Using viscous fluid as base fluid during the design phase, will leads to adapt the geometrical parameters to get best efficiency while handling of viscous fluid like heavy oil. This well resulted in significant reduction for production cost of heavy oil and will open new potentials and bring more reserves to the economic limit and add more oil production.
The aim of this study is to select the optimum pump geometrical parameters: Impeller ratio r1/ r2, horizontal blade Impeller angles β1 and β2, horizontal blade Diffuser angles β1’ and β2’ and vertical blade Diffuser ɣ1’ and ɣ2’ in order to enhance ESP performance while pumping viscous fluid specially like heavy crude oil. Increasing ESP efficiency will reduce operating cost.
A mathematical model was built in order to predict pump performance by calculating pump theoretical head and friction losses across the stage to get the actual pump flow rate - head curve. The hydraulic power was calculated and brake power from the shaft torque by using fluid velocity components. Validation was performed for this mathematical model against actual pump performance to prove the reliability of the mathematical model, then a viscosity correction is obtained to get pump performance while handling viscous fluid.
The optimization for pump geometrical parameters was performed by Surrogate method using MATLAB 2017a. The first step to build is to select the design of a published experiment to get data to feed the model and check the validity of surrogate model.
Once the model created, the results were used to get the relative importance of each parameter on pump efficiency by conducting screening process using Morris method and get the relation between most effective parameters on the pump efficiency.
Finally, the geometrical parameters optimization at three different values of viscosities was conducted using generic algorism.
Results show clearly the improvement of pump efficiency after optimization: pump efficiency increased at best efficiency point by 20%, 10% and 8.5% at viscosity 100, 300 and 500 (c.St) respectively.