Subsurface Electrical Centrifugal Pumps

Abstract

Summary The first subsurface electrical centrifugal pump for oilwell service in the U.S. was installed in the Russell field, KS, in 1926. Since that time many improvements have increased the efficiency of the pump at various pumping rates and depths in a variety of casing sizes. Each oil well has a different producing environment that the design engineer must consider to optimize the pumping installation for maximum service life. This paper discusses the major items involved in selection of a subsurface electrical centrifugal pump for a specific application. Additionally, installation, operating, and servicing practices are discussed. Introduction The basic arrangement of downhole equipment of the subsurface electrical pumping system has not changed from that used in the first installation at Russell field. However, engineering improvements have been made to the pump, motor, seal, cable, and surface control systems to handle greater depths, corrosive fluids, higher temperatures, and low gas-liquid ratios (GLR's) to increase the efficiency and economic life of the pumping system. Installation and operating techniques have also changed. Engineers must understand the design limits of each component to make the optimal selection of the equipment. The operator must understand these limits to obtain maximum system life without a premature failure. Although there are other equipment installation arrangements, such as cable-suspended pumps, inverted pumps, and shrouded motors, this paper discusses only the standard equipment arrangement shown in Fig. 1 where the base of the pump motor is located approximately 30 m [98 ft] above the top of the producing formation. Discussion Pump The multistage centrifugal pump is located at the bottom of the tubing string (Fig. 1) with the pump intake located at the base of the pump. The pump consists of the multistage centrifugal impellers on a single shaft inside a single tubular housing of diffusers approximately 9 m [30 ft] long. Tandem pump systems may join two or more single-shaft pumps in series through the use of splined couplings between the shafts and a flange bolting for each housing. The total length of a tandem pump can be as long as 21.95 m [72 ft]. A single-stage centrifugal pump consists of one impeller and one diffuser. Fig. 2 shows a typical pump stage with the general nomenclature for the impeller and diffuser. Most impellers used in subsurface electrical pump designs have a nondimensional, specific speed number in the range of 1,500 to 6,000. The U.S. Hydraulic Inst. defines specific speed, Ns, of a centrifugal pump as the speed in revolutions per minute (rev/min) at which a geometrically similar impeller would operate if it were of such a size as to deliver 0.23 m3/hr [1 gal/min] against 0.3m [1 ft] of head. Head is defined as the lift that one stage at a given rev/min and capacity will generate (in meters [ft]). The following formula shows the relationship of a nondimensional design index. (1) where N = pump speed, rev/min, Q = capacity at the peak efficiency point, m3/min [cu ft/min], andH = total head per stage at the peak efficiency, m [ft]. Radial flow impellers with lower specific speeds develop head principally through centrifugal force. The higher-specific-speed pumps develop head through centrifugal and axial forces. The impeller illustrated in Fig. 2 is an example of a mixed-flow impeller (centrifugal and axial) with a specific speed of approximately 4,000. The capacity of a single stage is determined by the average OD of the impeller, the rev/min that the impeller turns, the space between the upper and lower shrouds, the average diameter of the eye space, and the curvature of the impeller vanes. Therefore, the greater the speed, OD, and shroud spacing, the greater the capacity. The eye diameter and curvature of the vanes affect the efficiency of the system. An additional impeller adds to the head but does not change the capacity of the first impeller. Each impeller develops a head that mathematically can be converted into pressure by multiplying the head generated on test by 100 times the specific gravity of the fluid being pumped (pure water is 1.000 at 20 deg. C [68 deg. F] at standard 10 MPA [1.0 bar] atmospheric pressure). Therefore, the discharge pressure of the impeller will act on the surface of the impeller exposed to it. Fig. 3 shows a cross section of an impeller and the top and bottom views of the impeller thrust areas. JPT P. 645^