Most of the industrial fields, especially oil and gas, involve the application of electrical submersible pump (ESP) for flow transport. The pump performance will deteriorate depending on conditions of usage. The main parameters that affect or influence the pump performance curve are pump types, fluid flow rates, gas fractions, the impellers’ geometry, rotational speeds and fluid properties. In this work, the pump performance curve was studied as the governing parameters such as flow rates, meshing elements and turbulence models were varied. A computational fluid dynamics simulation (CFD) was applied and the findings were compared with the manufacturer’s data for single-phase flows. The main purpose was focused on getting the right technique to investigate this problem. The various turbulence CFD models were analysed and sources of errors were explained. The corresponding pump head losses were discussed and the main improvement or the criteria to obtain the highest efficiency were suggested. The results show that the k-epsilon with enhanced wall treatment is the best CFD technique since it produces the most accurate results and the least errors by allowing flexibility in large pressure gradient and rapid changes in flow properties.Keywords: ESP, CFD simulation, head loss, performance curve, turbulence models
Read full abstract