This article, written by Assistant Technology Editor Karen Bybee, contains highlights of paper CSUG/SPE 134089, ’Experimental Investigation on the Viscous Effect on Two-Phase Flow Patterns and Hydraulic Performance of Electrical Submersible Pumps,’ by F.E. Trevisan, SPE, and M.G. Prado, SPE, University of Tulsa, originally prepared for the 2010 Canadian Unconventional Resources & International Petroleum Conference, Calgary, 19-21 October. See Journal of Canadian Petroleum Technology, April 2011, page 45.

Using a visualization prototype built from original electrical-submersible-pump (ESP) components and with minimal geometrical modifications, an experimental procedure was developed and conducted to address the viscous effect on liquid/gas two-phase flow through ESPs. On the basis of dimensionless groups that govern centrifugal-pump single-phase performance, two-phase experiments were conducted at different shaft speeds, nonslip void fractions, and viscosity values while liquid rates were kept constant at 60% of the maximum rate at the defined shaft speed.


An ESP is one of the many existing artificial-lift techniques used in the petroleum industry. It consists of a series of small-diameter diffuser-type casing centrifugal pumps, mainly known for their capacity for handling high volumes. Their applications vary from producing high-productivity oil wells onshore and offshore to dewatering coalbed-methane and gas wells.

The use of ESPs in liquid single-phase applications is well understood. However, the presence of a compressible phase, such as natural gas, will modify the hydraulic performance of this type of pump. The presence of a second phase often will degrade the ability of the pump to deliver pressure, and depending on the amount of gas flowing through the system, instabilities may occur. These, also referred to as surging, eventually lead to extremely low pressure generation, also known as gas lock.

Experimental Description

An ESP visualization prototype was designed, built, and installed in a closed-circuit loop. Fig. 1 in the full-length paper provides a graphical representation of the experimental facility. This closed circuit consisted of a 65-gal tank to store the liquid, from which the fluid passed through a booster pump and a liquid-measuring unit before flowing to the ESP prototype. Air was provided by a 1-hp compressor delivering pressures up to 10 psig and flow rates up to 10 scf/hr. Gas was injected at the liquid line close to the prototype intake and measured by rotameters. After passing through the ESP prototype, the fluids returned to the liquid storage tank that also acted as an open-air separator. Both ESP and booster-pump motors were controlled by individual variable-speed drives linked to the acquisition system. This system also received information the differential-pressure and temperature transducers.

This content is only available via PDF.
You can access this article if you purchase or spend a download.