Content of PetroWiki is intended for personal use only and to supplement, not replace, engineering judgment. SPE disclaims any and all liability for your use of such content. More information
Use of ESPs in harsh environments
- Multiphase fluids or high GOR wells
- Fluids with abrasive particles
- Viscous fluids
- High-temperature wellbores
- Corrosive fluids
- Scale and asphaltenes
In these environments, the demands on the equipment design functions, materials, and operational processes increase. The run-life of the entire system can be affected if proper designs for these applications are not used.
The presence of free gas in the produced fluid affects the performance of the ESP pump. Generally, a pump is designed to handle incompressible fluids (liquids), and a compressor is designed to handle compressible fluids (gases). The performance or efficiency of both will suffer if they are required to handle a multiphase fluid (liquid and free gas). Typically, as the amount of free gas to total volume of the pumped fluid increases, the pump-stage head and flow both deteriorate. 
The amount of free gas that an ESP pump can handle is a function of the following variables:
- Pump-stage geometry.
- Operating point of the pump stage
- Control by a fixed-speed or variable-speed drive
- Pump-intake flowing pressure
- Wellbore geometry
Pump-stage geometry. The gas handling capability of a centrifugal pump stage increases with flow rate or stage specific speed—a nondimensional design parameter. In other words, as the stage style moves from radial to mixed flow (Fig. 1), the gas-handling capability increases.
Pump operating point. The most stable operating region for a pump stage on gassy fluid is from the maximum recommended flow rate back to its BEP. As the flow rate moves from the BEP toward the minimum recommended operating point, the potential for gas interference affecting pump performance is increased.
Variable speed controllers (VSC) operation. The VSC allows for some additional flexibility and reduction in unit shutdowns that are related to pump gas locking. Tests have shown that the pump gas-handling capability increases slightly with increasing speed. If the pump load decreases and the motor amps drop, indicating an initiation of gas lock, the VSC can be programmed to speed up for a short period to attempt to clear the gas-lock situation. If it clears and the load picks back up, the VSC would then return to its set operating frequency. If it does not clear, the unit would then shut down on an underload situation and restart on the time out delay.
Pump-intake pressure. The gas handling capability of the pump is very sensitive to pump-intake pressure. An empirical correlation for the relationship of the amount of free gas a pump can handle and the fluids flowing pressure was established from numerous tests on a variety of pump stages. A graphical representation of this correlation is shown in Fig. 2. The area under the curve represents stable operation, and the area above indicates potential gas-interference and -locking regions.
Fig. 2-Modified Turpin correlation (data from Ref. 5) [after Centrilift Graphics, Claremore, Oklahoma (1995)].
Wellbore geometry. The natural and mechanical separation of free gas from the flowing fluid is a function of the wellbore geometry. The annular area between the casing and the ESP unit and the fluid flow rate determines the flowing fluid velocity. The natural annular gas separation decreases as the velocity increases. Also, whether the casing is horizontal, inclined, or vertical determines the flow regime of the multiphase fluid and influences its natural separation characteristics. The efficiency of annular separation is still unknown, and additional research must be done in this area.
Optimal ESP configurations for gas handling
Optimal ESP configurations for gassy applications are listed below. Depending on the severity of the application, they can be used individually or in multiple combinations.
Tapered pumps. Tapered pumps utilize several different sets of pump stages in the same pump housing or pump string. Generally, the first section of stages is mixed-flow style because they can handle a higher percentage of free gas. As the gassy fluid is pressurized through each of these first stages, the total fluid volume decreases because of the compression of the free gas. When the flow rate nears the BEP flow rate of these stages, a second set of stages is selected. Generally, a good design can be accomplished with two or three sets of stages in the taper.
Mechanical separation. The vortex and rotary separation intake components, which were discussed in the pump section, are used here to add centrifugal separation to the gassy fluid that enters the intake section. Because there are so many variables that affect their effectiveness or efficiency, the manufacturers should be contacted for separation efficiency values or guidelines. These units can also be used in tandem to accomplish series separation.
Motor shrouds. Not only can motor shrouds be used to raise the velocity of the production fluid by the motor for increased cooling, they can also be used to assist with annular separation of the free gas in the produced fluid. The different styles and their uses were discussed in the section on motor shrouds and recirculation systems.
Recirculation pump. In a completion scheme where there is insufficient clearance to run a shrouded unit below perforations, a recirculation pump can be used. A recirculation pump bleeds a small portion of the pumped fluid off and circulates it down below the motor by a small-diameter hydraulic tube. This establishes a small flow in the rathole where the motor is set. By properly designing the bleed flow, cooling flow by the motor can be maintained. Since the perforations are above the unit and pump intake, natural annular gas separation can be maximized.
Although there are guidelines from the manufacturers for ESP configurations for gassy applications, the area still remains somewhat of a black art. Since there are so many variables that affect an ESP’s ability to perform in a gassy application, the best method is for the operator to select what they feel is the best solution, based on prior field experience or the manufacturer’s guidelines. Once the equipment is operational, field tests on each wellbore can be conducted to test that specific ESP configuration under those specific wellbore conditions.
The standard ESP pump does not tolerate abrasive particles in the pump fluid. The amount of tolerance is directly related to the aggressiveness of the solids or sand. The aggressiveness is a function of the percentage of the solid substance that is harder than the material of the pump components, the size and shape of the particles, and the concentration of solids in the fluid. The most aggressive solids are those with a high solids concentration ( > 1% by weight ), a large percentage of the solids sample being quartz (harder than the base stage and bearing material), a majority of the sample under a 100-mesh sieve size (able to get into the bearing and sealing areas easier and faster), or quartz grain shapes that are angular or barbed. On the other extreme, there are cases where very round, smooth, soft sands are relatively benign to the operation of the pump.
Performance impact of abrasives
There are three types of wear that impact the pump stage and its performance. They are listed next and prioritized in order of importance or impact.
Radial wear. As the slurry wears the radial-support bushing system of the pump, it loses its lateral stability. This allows the rotating parts to start interfering with the stationary parts. Vibration increases, and it starts impacting the top of the seal section where the first mechanical shaft face seal is located. Once vibration and radial movement start to influence the face seal, leakage starts across the sealing face. This initiates a path for the well fluid to progress toward the motor.
Downthrust wear. On the floating-style stages, the abrasive slurry migrates into the downthrust bearing pad area of the pump stage. The stationary diffuser thrust pad starts boring into the impeller thrust washer area. Once it breaks through the lower shroud of the impeller, the impeller loses part of its work to recirculation flow. As the diffuser pad bores further into the impeller passageway, it also blocks a portion of the impeller flow path, thus restricting the remaining flow.
Erosion wear. As with any abrasive-slurry flow along a twisting path, erosion wear takes place. Although it is not usually associated with the failure of the pump, it is a potential failure mode and a concern, especially when modifications have been made to the pump to address the radial and downthrust wear modes. Erosion wear not only damages the stage pieces, it also wears any surface with which it comes into contact. Severe cases have resulted in the wear perforating the pump or production-tubing walls and dropping units in the well.
Optional ESP configurations for abrasives
Depending on the severity of the application, the following design options can be used individually or in combination.
Compression pumps. For many years, this was the answer for abrasive applications. In a compression or "fixed-impeller" pump, the impellers are fixed to the shaft or stacked hub to hub so there is no axial movement. With all the impellers fixed relative to the shaft, the whole impeller stack can be raised slightly so that it does not run into contact with the downthrust or upthrust pads on the diffuser. This pump design eliminates the downthrust wear mode. When it is used in conjunction with hardened journal bearings, it also addresses radial wear problems. There are several issues with compression pumps. First, they are very difficult to assemble properly. Because an ESP pump is a very long, multistaged assembly, it is very difficult to locate all of the impellers and still have the needed minimum shaft axial movement. Also, now, the thrust of each impeller is transferred to the shaft and is added to the normal shaft thrust produced by the discharge pressure on the top area of the shaft. The thrust bearing in the seal-chamber section is required to carry this additional thrust. Additionally, as the sealing areas of the pump stage wear, the downthrust also increases. Therefore, the selection of the proper thrust bearing is critical, and the anticipated thrust must be calculated on the basis of the maximum thrust seen from worn stages.
Thrust and radial protection. In this modification, the base material in the radial and downthrust areas of the stage is replaced with inserts of hardened materials. The materials are usually tungsten or silicon carbides, or ceramics. This results in a pump with both radial and downthrust protection but is built in a floater style.
Erosion protection. Currently, this area is under development, but some limited-to-moderate success has been achieved with:
- Heat treatments
- Surface hardening
- Hard-material liners
Generally, the abrasive production fluid does not impact the motor and seal-chamber section. There could be minor erosion worries because of the flow velocities of the production fluid by the outside surfaces of both units. Also, if the top shaft’s mechanical face seal in the seal-chamber section is exposed and operates in the production fluid, hardened stationary and rotating seal faces are recommended.
Viscous crude and emulsions
ESPs are also used to lift viscous fluids, commonly referred to as heavy and extra-heavy crudes. Viscosity is defined as the resistance of a fluid to movement as a result of internal friction. Resistance causes additional internal losses in a centrifugal pump. The increases in internal losses of a centrifugal pump affect each performance parameter.
Performance impact of fluid viscosity
Effect on flow capacity. Flow capacity of a given pump stage diminishes rapidly with a relatively small increase in viscosity. The rate of correction tends to moderate as viscosity continues to increase. The amount of correction is also dependent on stage geometry, and the decrease in capacity is more exaggerated for radial flow stages.
Effect on head. The total dynamic head at the best efficiency point (BEP) diminishes on a moderate curve as viscosity increases. It is affected to a lesser extent than flow capacity. The head at zero flow remains relatively constant. Fig. 3 shows various head vs. flow-rate curves for an ESP pump stage rated for about 2,100 B/D on water.
Effect on horsepower. BHP increases rapidly with increasing viscosity but tends to level off because of diminishing flow rate and total dynamic head (Fig. 4).
Effect on efficiency. Efficiency decreases in proportion to the changes in flow capacity, TDH, and HP, in terms of Eq. 1 (Fig. 5).
There are several published methods for estimating the effect of viscosity on the head, flow rate, and BHP of a centrifugal pump. These "standard" correction factors are usually not accurate for the specific small-diameter, multistage design of ESP pumps. Therefore, most manufacturers have established corrections through testing for each pump stage type in their product line. These correction factors are based on dead-oil viscosity values for the fluid at pump-intake conditions. When applying these corrections to the pump, the following should also be considered.
Effects of gas. When gas saturates into the crude, it reduces the viscosity of the fluid. Some amount of gas is helpful in reducing fluid viscosity, but an excessive amount of free gas is disruptive to well fluid production. Gas tends to migrate out of highly-viscous fluid slowly. Therefore, a higher percentage of gas tends to pass through the pump with the produced well fluid. In an application with gas, the designer must be aware of two viscosity values:
- Dead-oil viscosity - the viscosity of the crude at dead or completely degassed conditions.
- Live-oil viscosity - the apparent viscosity of the gas-saturated crude and the viscosity that affects the pump performance in a well with gas.
There are several dead-oil and saturated-oil viscosity correlations that can be used during the design process. The correlation selection should be based on modeling of the actual wellbore performance.
Effect of temperature. Temperature has a dramatic effect on the viscosity of the crude oil. Therefore, it is critical to the ESP design process that the fluid temperature in the wellbore at the pump setting depth is known. This determines the fluid viscosity and pump-performance correction factors at the first pump stage. Additionally, the inefficiency of the pump results in additional heat loss to the fluid and surrounding wellbore. This incremental elevation in temperature from stage to stage through the pump moderates the impact of the fluid viscosity on the total pump performance. Therefore, the designer should, at a minimum, use an average viscosity for the fluid through the pump for sizing applications. A more accurate method is to calculate the performance on a stage-by-stage basis, using the fluid input conditions to each stage. Most design software programs use this method.
Effects of water. With the incursion of water or brine into the wellbore, the viscosity of the oil/water mixture can increase, sometimes dramatically when emulsions occur. The shear forces on the fluid mixture, as it flows through the formation, perforations, or centrifugal pump, can cause an emulsion. Because thousands of molecular structures with different chemical and physical properties exist in crude oils, it is virtually impossible to predict viscosity characteristics on the basis of oil and water cuts. A default viscosity correction factor for emulsions, referenced in many petroleum engineering textbooks and references, has been used for many years with questionable results. The correction factor is shown graphically in Fig. 6. The curve provides for a progressive increase in the viscosity multiplier, up to 15, as the water cut increases. It then drops to 1, indicating the emulsion has inverted or become water-wetted. Use of this correction factor in viscous applications has indicated that it is too severe. Recent work has shown that because of the complexity of emulsion characteristics, it is best to run carefully controlled baseline laboratory tests on reservoir crude and brine samples to develop an emulsion correction curve. 
ESP options for fluid viscosity
Several options are available for improving the performance of the ESP pump when applied to viscous crudes.
Dilution. Some success has been achieved with diluent injection. In this process, a lighter crude or refined product, such as diesel, is injected from the surface via a separate hydraulic line to a point below the ESP or directly into the pump intake. This effectively cuts the viscosity of the wellbore fluids. The amount of injected diluent depends on the desired final mixture viscosity. This type of viscosity reduction also reduces the surface flowline losses, which reduces the required wellhead pressure or the need for diluent injection at the wellhead. Using a diluent fluid is an effective, but expensive, approach.
Temperature increase. The temperature of the reservoir or near-wellbore area can be artificially raised to make the viscous crude more mobile. The most successful method for adding heat has been through steam injection or soaking, although trials have been made with resistive, induction, and microwave technologies. This reduces the viscosity of the crude, but it also raises concerns in high-temperature operations.
Chemical injection. Viscosity-reduction and emulsion-breaking chemicals can be injected from the surface by hydraulic injection lines. This impacts the fluids in the annulus and through the pump but not very far back into the reservoir.
Water injection. When emulsions are encountered through a certain water-cut range, additional water can be injected to increase the water cut of the produced fluid, moving it out of the high-viscosity correction area. Field trials on this concept were conducted in the mid-1980s and were successful in reducing the fluid viscosity and increasing the ESP performance.
Another trend has been the application of ESPs into higher-temperature reservoirs. Typically, these are reservoirs that are either deeper or artificially heated. Standard ESP systems are commonly applied to well ambient temperatures of 250°F (121°C). Even with a velocity greater than 1 ft/sec, the temperature rise above ambient conditions will be about 50°F (10°C) for water and 90°F (32°C) for oil—higher if fluids contain gas.
Systems that have minor modifications or optional features are applied in ambient temperatures up to 350°F (177°C). Additional research and field testing is being done on systems that operate in ambient temperatures above 350°F. For these units, the motor and seal-chamber section have the most significant design changes. The design areas of concern in the motor include:
- Insulation system
- Mechanical bearing system
- Internal lubrication and cooling system
The seal-chamber section modifications include:
- The mechanical journal and thrust bearing systems
- Internal lubricating system
- The elastomeric bag(in positive-barrier styles)
Also, the selection of a power cable rated for elevated temperature service is critical.
Since the early 1990s, the focus on the application of submersible motors in high-temperature wells has not been entirely on the ambient wellbore temperature, but rather the internal operating temperature of the motor. This is because, even in what would be considered a relatively cool well, a misapplied design can possibly operate at dangerously high internal operating temperatures. Most of the ESP application software programs calculate the expected motor operating temperatures, or the manufacturer can be contacted to provide this information. The calculation of the motor operating temperature involves many variables, which are discussed in ESP motors. Historically, this calculation has been made at the stabilized design operating point of the motor. Recently, new computerized programs have allowed the ESP system operating conditions to be dynamically modeled from the static startup condition to the stabilized-flow operating point. This has allowed the designer to identify potentially dangerous transient operating conditions and to provide for options to eliminate or reduce their impact.
If the operator is operating an ESP at elevated motor operating conditions, it is suggested that a downhole motor-temperature monitoring system be run. This monitor provides warning of any high-temperature excursions of the downhole system so that remedial action is taken before potential catastrophic damage occurs.
The application expansion of ESPs into more corrosive wells has required the industry to provide enhanced corrosion-protection options. In the early years, the normal protection scheme for mild-corrosion applications was the use of protective coatings. These were either epoxy paint, babbit spray, or stainless-steel/high-alloy metal flame spray. Each of these has the disadvantage of the potential for mechanical damage during the installation handling process and deployment through the casing. The need for a higher-level corrosion-resistant ESP was first recognized with the application of units into carbon dioxide (CO2) enhanced-recovery reservoirs in the late 1970s. From this need, the first-generation corrosion-resistant ESP unit was developed. Current units use high-chromium alloys in the components exposed to the wellbore fluids.
Another source of corrosion is hydrogen sulfide, H2S. The H2S mainly attacks copper-based alloys of the pump, seal-chamber section, and cable. This type of corrosion can be controlled by replacing the copper-based alloy components with suitable materials or by isolating them from exposure to the well fluid and gases. When CO2, H2S, and hot brine are combined, unpredictable corrosion results may appear. With small changes in the concentration of CO2 and/or H2S and temperature, corrosion could even vary significantly from well to well within the same reservoir.
Another corrosion mechanism that has been around the oil field for years, but has been misunderstood or misdiagnosed, is microbiologically influenced or induced corrosion.  It is caused by sulfate-reducing bacteria, as well as other forms of anaerobic and aerobic bacteria. The four common types found in oil wells and affecting ESPs are:
- Anaerobic sulfate-reducing bacteria (SRB)
- Anaerobic acid-producing bacteria
- Aerobic acid-producing bacteria
- Slime-forming bacteria
The SRB and anaerobic/aerobic acid-producing bacteria species attach themselves to the surface of the ESP components and cause direct and indirect corrosion and severe pitting. The slime-forming species can cause some minor corrosion but is noted more for downhole formation and equipment plugging.
Scale and asphaltenes
If the well has scaling or asphaltene-forming tendencies, these can be detrimental to the performance and run-life of the total ESP system. Because of the characteristics of the ESP system, there are pressure and temperature changes, which provide a mechanism for scales to form or precipitate out of solution. Typically, scales cause two problems. They plug the flow passageways of the pump stages, reducing or stopping the flow entirely. They also adhere to the outside surfaces of the motor and seal-chamber section, reducing the heat-transfer rate, causing both units to run hotter. Asphaltenes generally only cause plugging of the pump stages. Both problems can be reduced, but not totally eliminated, by applying synthetic coatings to the surfaces affected or by using a downhole inhibitor-chemical treatment.
|C||=||constant = 3,960, where Q is in gal/min, and TDH is in ft [= 6,750, where Q is in m3/D, and TDH is in m]|
|Q||=||flow rate, B/D [m3/d]|
- Lea, J.F. and Bearden, J.L. 1982. Effect of Gaseous Fluids on Submersible Pump Performance. J Pet Technol 34 (12): 2922-2930. SPE-9218-PA. http://dx.doi.org/10.2118/9218-PA
- Lea, J.F. and Bearden, J.L. 1982. Gas Separator Performance for Submersible Pump Operation. J Pet Technol 34 (6): 1327-1333. SPE-9219-PA. http://dx.doi.org/10.2118/9219-PA
- Dunbar, C. 1989. Determination of Proper Type of Gas Separator. Paper presented at the 1989 SPE Artificial Lift Workshop, Long Beach, California, 16–17 October.
- Wilson, B.L. 1993. ESPs and Gas. Paper presented at the 1993 SPE Gulf Coast ESP Workshop, Houston, 28–30 April.
- Turpin, J., Lea, J., and Bearden, J. 1980. Gas/Liquid Flow Through Centrifugal Pumps—Correlation of Data. Paper presented at the 1980 Intl. Pump Symposium, Texas A&M U., College Station, Texas, 1 September.
- Bearden, J. and Sheth, K. 1996. Free Gas and a Submersible Centrifugal Pump—Application Guideline. Paper presented at the 1996 SPE Gulf Coast ESP Workshop, Houston, 1–3 May.
- Wilson, B.L. 1988. The Effects of Abrasives on ESPs. Paper presented at the 1988 SPE Gulf Coast ESP Workshop, Houston, 28–29 April.
- Wilson, B.L. 1990. Sand Resistant ESPs. Paper presented at the 1990 SPE Gulf Coast ESP Workshop, Houston, 30 April–May 2.
- Woelflin, W. 1942. Drilling and Production Practices. Washington, DC: API.
- Patterson, J., Henry, J., and Dinkins, W. 2002. Emulsion Viscosity Testing with ESPs. Paper presented at the 2002 SPE Gulf Coast ESP Workshop, Houston, 1–3 May.
- Adams, D.L. 2002. Biocorrosion of Electrical Submersible Pump Components by Sulfate Reducing Bacteria or Other Bacterium. Paper presented at the 2002 SPE Gulf Coast ESP Workshop, Houston, 1–3 May.
Noteworthy papers in OnePetro
Use this section to list papers in OnePetro that a reader who wants to learn more should definitely read
Use this section to provide links to relevant material on websites other than PetroWiki and OnePetro
Jose Caridad, BSME & MSc ME