Got Steam? Understanding ESP Steam-Handling Capabilities in the Centrifugal Pump

Abstract

Summary The successful development and implementation of high-temperature electrical-submersible-pump (ESP) technology for steam-assisted-gravity-drainage (SAGD) applications have enabled operators to reduce their flowing bottomhole pressures and achieve higher production rates. However, operating under these conditions brings the pump-intake pressure (PIP) closer to the saturation pressure of steam, which can result in live-steam production through the pump. The effect that live-steam has on pump performance is not well-understood, and has been a key challenge for operators when designing and optimizing ESP systems for SAGD applications. In early 2011, ConocoPhillips, Baker Hughes, and C-FER Technologies (herein referred to as the operator, manufacturer, and independent laboratory, respectively) embarked on an experimental test program to determine the consequences of producing live steam through a centrifugal pump. This new program was meant to build on multiphase work that had begun more than a decade ago at the University of Tulsa, where researchers had focused on experimentally measuring the two-phase-flow performance of ESP stages with air at moderate temperatures (Pessoa and Prado 2003). The University of Tulsa work ultimately resulted in a wave of new technology aimed at increasing ESP gas-handling capabilities. Following a similar testing and ESP-instrumentation philosophy, this new collaboration looked to build upon the University of Tulsa experiments and expand the test fluids to include live steam, water, and air at higher temperatures. This ultimately involved the design and construction of a unique high-temperature-steam flow loop that enables live-steam injection into a centrifugal pump, while monitoring both head and performance degradation. This paper will reveal some of the unique test results collected with the first pumping system, including snapshots of the stage-by-stage pressure contributions captured in real time as air or air and steam migrated through the ESP being tested. These results also demonstrate the impact that other gases can have on steam flashing and the importance of considering gas- and steam-vapor effects in SAGD-ESP designs.