IPTC-19264-MS

Experimental Study of Sand Erosion in Multistage Electrical Submersible

Pump ESP: Performance Degradation, Wear and Vibration

Haiwen Zhu, Jianjun Zhu, and Zulin Zhou, University of Tulsa; Risa Rutter, Michael Forsberg, and Shawn Gunter,
Baker Hughes, a GE company; Hong-Quan Zhang, University of Tulsa

Copyright 2019, International Petroleum Technology Conference

This paper was prepared for presentation at the International Petroleum Technology Conference held in Beijing, China, 26 – 28 March 2019.

This paper was selected for presentation by an IPTC Programme Committee following review of information contained in an abstract submitted by the author(s).
Contents of the paper, as presented, have not been reviewed by the International Petroleum Technology Conference and are subject to correction by the author(s). The
material, as presented, does not necessarily reflect any position of the International Petroleum Technology Conference, its officers, or members. Papers presented at
IPTC are subject to publication review by Sponsor Society Committees of IPTC. Electronic reproduction, distribution, or storage of any part of this paper for commercial
purposes without the written consent of the International Petroleum Technology Conference is prohibited. Permission to reproduce in print is restricted to an abstract of
not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of where and by whom the paper was presented.
Write Librarian, IPTC, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax +1-972-952-9435.

Abstract
As one of the most widely used artificial lift methods, electrical submersible pumps (ESPs) have been
improved gradually since the 1910s. However, its performance and run life are affected by many problems
such as gas lock, high viscosity fluid, corrosion, and erosion. With the development of horizontal well
drilling and multistage hydraulic fracturing, sand production from unconsolidated sandstone and proppant
backflow often cause severe damage to ESPs resulting in reduced operating lifespan. Measuring wear in
an ESP pump and monitoring performance degradation is not only very difficult in field cases, but also in
experimental studies. The results are precious for understanding the wear mechanism inside an ESP as well
as guiding the ESP design and simulation. At the same time, vibration and performance data can provide
significant guidance to ESP failure diagnosis, which can potentially reduce the time and cost of well service
and extend ESP run life.
Wear processes inside an ESP can be classified by different modes of mechanisms. Erosive wear can
be observed in the primary flow channel of the impeller (rotor) and diffuser (stator). Particle strike shroud
surfaces and the scratched material is flushed away by fluids. Various semi-mechanistic erosion equations
are available to be coupled with Computational Fluid Dynamics (CFD) to predict the erosion in ESPs.
In the secondary flow region, balance chamber and sealing rings, particles are presented between the
stator and the rotating rotor. Therefore, abrasive wear is believed to dominate the wearing process. Unlike
erosion, abrasion is more complicated and abrasion equations are highly depended on geometries, physical
mechanism and load between particle and target surface. In this study, a sand wear test flow-loop is designed
and constructed to investigate wear in ESPs. Performance degradation, Erosion pattern, abrasion rate, and
stage vibration of an ESP were recorded in a 64-hour sandy flow test.

Introduction
Wear by hard particles occurs in many different situations such as on slurry pumps, pipelines, rock drilling,
rock or ore crushers, and extruders or chutes. According to Figure 1, the wear processes can be classified by
2 IPTC-19264-MS

different modes depending on the kinematics and by mechanisms depending on the physical and chemical
interactions between the elements of the tribosystem. (Zum Gahr, 1999)

Figure 1—Wear type (Zum Gahr, 1999)

Figure 2 (a) shows a typical combination of an ESP, while components are shown in Figure 2 (b) and
(c) are used to describe wear locations in ESPs. Erosion takes place on the primary flow region in ESP
impeller and diffuser flow channel and abrasion can be observed on pump bearings, bushings, seals and
sleeves. (Hashish, 1988 and Morrison et al., 2015). The leakage through the secondary flow region is caused
by head created by the impeller. (Zhu et al., 2017) Usually, leakage can be found through skirt clearance,
balance ring clearance and the diffuser bore clearance to the impeller balance holes. Since the clearance
is very small and can be treated as annular seals for a new pump, the head loss generated by leakage can
be ignored. (Zhu et al., 2018b) However, once sand is produced with fluid, the pump boosting ability and
efficiency will be gradually reduced by abrasion on seals, while the erosion on primary flow channel can
eventually destroy the pump.

Figure 2—Components of ESP, (a) entire ESP (http://www.franklinwater.com),

(b) impeller front view, (c) diffuser front view (Takacs, 2009)
IPTC-19264-MS 3

Erosion in ESP
Erosion in ESPs can cause material detached from blades, vanes, shrouds and balance holes depending on
the fluid type, and pump geometries. Minemura and Zhong (1996) investigated the wear in pump casings for
wear-resistant materials. Kruger et al. (2010) studied the erosion of a radial centrifugal pump and divided
erosion into shock-like and friction-like processes. From erosion tests on a mixed type ESP conducted by
Morrison et al. (2015), the erosion is more obvious on the second stage. Shi et al. (2017) and Shi et al.
(2018) illustrated and validated the importance of the inlet effect of a multiphase pump test. From the flow
pattern photo taken by Zhu et al. (2018a), the gas core is concentrated to the outside shroud of the impeller.
Since the solid phase usually flows with the liquid phase, the three-phase erosion tests on a split-vane ESP
conducted by Morrison et al. (2017) observed severe erosion on impeller blade and bottom shroud close to
the turning area between impeller and diffuser.
Although the detailed particle trajectory and erosion pattern is hard to be obtained, it can be easily
predicted by commercial software with low cost (Zhang, 2018). Therefore, validated by experimental
results, CFD simulation is a widely used method to predict the erosion rate and analyze the erosion
mechanism inside a complex geometry. Kruger et al. (2010) used the empirical models developed by Gülich
(2008) along with CFD simulations (Euler- Euler approach) to predict the erosion in centrifugal pumps.
Later on, Marsis and Russell (2013) applied Discrete Phase Models (DPM) and Eulerian-Granular models
to above CFD erosion models to predict erosion rates inside an ESP. In addition, Pirouzpanah and Morrison
(2014) conducted experimental and CFD studies of the ESP and developed a new empirical-numerical
model was developed to predict the erosion rate. Basaran (2017) conducted 2-stage ESP erosion simulations
under different viscosity fluid. Zhu at al. (2018) studied the effect of different erosion models, pump types,
particle density, and diameter by CFD simulations.

Abrasion in ESP
From a five-year investigation of sand abrasion phenomena in conventional electrical submersible pumps
(ESP) conducted by King et al. (1983), three-body abrasion of pumps operating under abrasive conditions
is the most destructive cause of downhole equipment failure. Worn pumps cause vibration and shaft seal
leakage, which eventually leads to electrical failure of the motor or power cable. Wear on down-thrust
washers and stages seals can be substantially reduced by the use of special sleeves and inserts without
prohibitively increasing manufacturing costs of the pump. In addition, erosion is a problem only under
severe corrosion and abrasive conditions, which can also be controlled using cast irons with good fluid
channel design. The material can be expensive but it is worthy if workover costs are high.
Later on, Wilson (1990) conducted a set of experiments on ESPs to analyze the size and quantity of
abrasives with fluid flow through the pump, in which sleeves with hard materials can help reduce the wear
on impeller hubs, stage seal rings (skirt and balance rings) and thrust washers. However, small particles (2.3
μm) increased the wear on thrust washers during the 4-hour tests. In his study, the shaft was stabilized by
hard sleeves. Large particles were too big to migrate into the thrust washer and lubricant seal rings, resulting
in low abrasion. However, when the pump was running under low thrust conditions, the seal's clearance
between impellers and diffusers was enough to allow larger abrasive grains to enter. In addition, the pump
started rotating eccentrically when impeller hub and sleeves were worn out. Then, particles larger than the
lubrication layers were wedged under the edge of the thrust washer, causing high abrasion.
In addition, Zheng (2013) and Morrison et al. (2015) conducted 117 hours of water-sand tests followed
by 68 hours water-air-sand tests on a 3-stage ESP. They concluded that the abrasion took place on the
pump bearings, and corresponding bushings, impeller downstream seals, impeller upstream seals, and their
corresponding diffuser stators, which reduced the stiffness of the entire system.
4 IPTC-19264-MS

Experimental Setup
A closed-loop experimental facility has been designed and constructed at TUALP to study the sand erosion
process in an ESP. The experimental facility consists of a closed 2" carbon steel pipe flow loop and a 12-
stage ESP. A 60-hour erosion test under 200 psia loop pressure is divided into several short interval tests
(2, 4 & 8 hours). After each interval test, gas-liquid flow rate, temperature, pump performance (pressure
increment, thrust, and torque), stage imbalance and vibration are measured. The pump is then disassembled
and wear rate is measured. Sand, which is replaced when significant changes are found, is reused in the
loop and monitored by AFS grain fineness test and SEM pictures. (Meng et al., 2018 and Meng et al., 2019)
After 60 hours, a new set of 12 stages will replace the damaged ones and a new test will start.

Experimental Facility
The facility specifications and the schematic of the facility layout for the TUALP ESP sand erosion flow
loop are shown in Figure 3. As seen, the sand erosion flow loop is composed of the main flow loop, gas
loop, and the disposal tank and water supply.

Figure 3—Schematic of TUALP ESP Sand Erosion Flow Loop

As shown in Figure 3, pure water is supplied by the water tank and recycled in the main flow loop.
Sand was added at the top of the gas separator. Absolute pressure and differential pressure transmitters,
temperature transmitters, vibration proximity probes, and a torque sensor were installed on the ESP bench.
After each interval test, pump geometry changes (seal clearance decrease and weight loss), performance
degradation (pressure increment, torque, efficiency) and vibration were measured and recorded. Pump
stages were coated to visualize the most severely eroded location. In each interval test, sand was added
through the separator and replaced every 2 hours. After a 2 hours’ recirculation, the wasted sand was drained
and the loop was washed with pure water. SEM and AFS tests were done to monitor the average sand
properties. The process was repeated until the interval test has been finished.
The pump intake pressure was maintained by a gas pressure regulator that connected the gas separator to
the gas injection loop. The slurry flow rate was controlled by a downstream manual gate valve after ESP,
which can be replaced easily if eroded. Flow rates were measured by a Coriolis flowmeter (Endress Hauser
Promass 80F), which is protected by a bypass line. At the beginning of the test, the relationship between
IPTC-19264-MS 5

the total pressure gain of the pump and flow rates was recorded, then the flowmeter is bypassed and the
pressure drop was used to indicate flow rates.
The data acquisition system was originally made by Compact FieldPoint module from National
Instruments (NI). However, FieldPoint modules were replaced by CompactDAQ modules to obtain a high
updating rate. (Chi et al. 2017 and Zhou et al. 2018a&b) NI 9208 was used for pressure, flow rate and
temperature signals with a range of 4~20 mA. The internal control signals (4~20mA) generated by NI 9265
output module were sent to VSD and control valves. In addition, NI 9228, a high updating rate voltage
input module, was used for proximity probes and torque sensor, which requires high scanning frequency.
The data processing center is a Dell computer (Optiplex 9020). The DAQ program was written in graphic-
programing language Labview.

ESP test bench

The ESP motor, thrust chamber and other equipment needed for the operation were assembled in an ESP
test bench. The skid for the ESP test bench is shown in Figure 4. As seen, the torque sensor was installed
between the motor and thrust chamber, and 8 vibration proximity probes were mounted on pump stages.
Performance, torque and vibration information was recorded to reflect the pump condition. A 12-stage ESP,
which is designed to resist erosion in practical field conditions, was installed. The best efficiency point
(BEP) is Q = 3100 bpd and N = 3600 rpm with the boosting pressure of 9.8 psig per stage. The maximum
open flow rate at N = 3500 rpm is 5000 bpd, while the maximum shut-in pump head is 16.7 psi.

Figure 4—Skid for TUALP ESP Sand Erosion Flow Loop

The ESP delivers unmatched performance for flow rates from 4000 bpd to 1200 bpd. As shown in Figure
5, the tested ESP is composed of two types of stages, SSD stages and Floater stages (general stages). All
stages have the same type of impellers. However, the diffuser bore of an SSD stage is made of Tungsten
carbide, as well as the corresponding sleeves. On the other hand, the diffuser bore and sleeve of a floater
stage is made of carbon steel. In addition, both the bottom and top bearings are made of Tungsten carbide,
which offers more stiffness at the pump inlet and outlet. Four testing stages are shown in red dash box in
Figure 2.3 (f).

Figure 5—Combination of stages

6 IPTC-19264-MS

Seals clearance and measurements

Figure 2 (b) and (c), the inside diameter (ID) and outside diameter (OD) of the hub, bottom shroud, balance
ring and skirt ring on impellers and diffusers were measured after every interval test. In addition, the outer
diameter of sleeves and weight of impeller, diffusers, and sleeves were also recorded in the test. Skirt ring
clearance can be calculated by subtracting impeller skirt ring OD from diffuser skirt ID. Similarly, balance
ring clearance is the difference between impeller balance ring OD and diffuser balance ring OD. It should
be noted that the inter-stage clearance is obtained from impeller hub OD and diffuser bore ID. Besides
the measuring clearance, pump vibration amplitude and impeller orbits can also be used to validate and
represent abrasion condition. Fast Fourier Transform (FFT) was used to filter the noise from the high-
frequency vibration signal. (Chen et al. 2015)

Experimental Results
Sand Characteristics
Sand particle used in our experiment is fracture sand SDS from a local company with a specific density
of 2.637. The mean diameter of particles is 150 microns with a range of 50 to 250 microns as shown in
Figure 6 (a).

Figure 6—Sand particle size distribution

The sand in the slurry fluid was sampled every 2 hours recirculating time during the total 64-hours long
test. Sand properties at 2, 32 and 64 hours’ testing time were compared with the original sand without
recirculation. As shown in Figure 6, the AFS number of three samples after 2 hours slurry flow is close
to 200, while the original value is around 100. The size distributions of the three samples are comparable,
indicating that half of the original particles were broken into smaller pieces. Similarly, a 50% decline of
IPTC-19264-MS 7

particle diameter is shown in Figure 7 after 2 hours’ slurry test, while the sharpness keeps consistent.
Therefore, it is presumable that the averaged sand characteristics are identical in all tests. AFS number of
150 (diameter 100 microns) and the sharpness factor is kept unity in this study.

Figure 7—SEM scans, (a) original, (b) 2 hours, (c) 32 hours, (d) 64 hours

Performance degradation
No obvious differences can be found on four measured stages. Therefore, only averaged head and efficiency
curves are shown in Figure 8. The obvious decrease on both the stage head increment and pump efficiency
in Figure 8 (a) is caused by higher friction loss in the primary flow channel and more leakage through the
secondary flow region. In low flow rate region, higher leakage is caused by larger head increment. More
turbulence is caused by increased leakage, which generates an additional turbulence head loss in the turning
area between impeller and diffuser. Other than pressure increment, more efficiency loses around the best
efficiency point. Overall, the average total head loss is around 10% and average efficiency loss is 12% after
64-hour test.

Figure 8—Head and efficiency comparison, (a) average head increment, (b) pump efficiency
8 IPTC-19264-MS

Stages vibration
As shown in the previous section, the system is stiffer at the top and bottom bearings, which is made by
Tungsten carbide. Among the four tested stages, stage 12 has the highest stiffness since it is an SSD stage
next to the top bearing. In the middle of the pump. stages 6 and 9 have the least stiffness. Especially, stage 6
that is made by carbon steel. The pump stage next to the top bearing. In the middle of the pump, stages 6 and
9 have the least stiffness. Especially, stage 6 that is made by carbon steel. The pump stage vibration after each
interval test is shown in Figure 9. Fast Fourier Transform (FFT) was used to process the vibration signals
and reduce the noises. The peak of FFT curves is the stage vibration amplitudes (mils). Similarly, with the
stiffness analysis, the vibration amplitude of stage 12 is the lowest, while it is higher on stage 6 and 9.

Figure 9—FFT of vibration in vertical and horizontal direction: (a) stage-3 horizontal direction, (b) stage-3
vertical direction, (c) stage-6 horizontal direction, (d) stage-6 vertical direction, (e) stage-9 horizontal
direction, (f) stage-9 vertical direction, (g) stage-12 horizontal direction, (h) stage-12 vertical direction,
IPTC-19264-MS 9

Same to the vibration data, orbits of stages 3, 6, 9 and 12 increased gradually at the first 16-hour slurry test,
as shown in Figure 10. In addition, the original orbits had an irregular shape with sharp edges, which became
round circles later on. It is presumable that extruded points on the original sealing rings or sleeves were
eroded after few hours of testing. Because of the higher stiffness on stages 3 and 12, their orbit diameters
are smaller than that of stages 6 and 9.

Figure 10—The orbit of impeller center

Geometry changes and weight loss

Geometry changes on diffusers and impellers caused by abrasion are shown in Figure 11. No obvious
changes can be found on the impeller hub ID, as can be seen in Figure 11 (a) since the internal surface of
the impeller hub is stationary to the shaft surface. Due to a larger clearance, the changes of ID and OD of
the outside shroud on the impeller and diffuser shown in Figure 11 (e) and (i) is much lower compared to
other seal surfaces. In addition, the skirt slots in diffusers are protected by impeller down-thrust washers,
which were completely worn out after 16 hours in test case 1. Without a washer, the clearance between skirt
slots on diffusers and skirt ring top surface is much larger than seal clearance. Therefore, the abrasion on
skirt depth in a diffuser is negligible.
Seal surfaces between stators and rotors, shown in Figure 11 (b), (c), (d), (f), (g) and (h), were eroded
rapidly. The normalized dimension in Figure 11 (k) and (l) is calculated by subtracting the original averaged
geometries from measured averaged geometries. Besides the inter-stage clearance of SSD stages, the
clearance increasing rates shown in Figure 12 are surprisingly similar. As mentioned before, the pump is
made by SSD and floater stages. The inter-stage surfaces are protected by Tungsten carbide diffuser bore and
sleeves. Similarly, the wear on the SSD sleeve OD in Figure 12 (c) and diffuser bore ID is also insignificant.
Similarly, the weight losses of SSD sleeves are much smaller than that of floater sleeves. However, the
weight losses of impellers and diffusers in SSD stages and floater stages are comparably similar. As shown
in Figure 12 (d), all weight losses are in a linear relationship with time, which is different to the abrasion
rate shown above.
10 IPTC-19264-MS

Figure 11—Geometry changes of every stages, (a) impeller hub ID, (b) impeller hub OD, (c) impeller
balance ring OD, (d) impeller skirt ring OD, (e) impeller outside shroud OD, (f) diffuser skirt ring ID,
(g) diffuser hub ID, (h) diffuser balance ring ID, (i) diffuser outside shroud inlet ID (j) diffuser skirt
ring depth (k) normalized geometry changes in impeller, (l) normalized geometry changes in diffuser
IPTC-19264-MS 11

Figure 12—Clearance changes and weight losses, (a) average clearance increases, (b) normalized clearance
increases, (c) normalized sleeves OD decreases, (d) normalized weight losses of impellers, diffusers, and sleeves

According to Archard (1953), the abrasion rate is related to the load on particles. The load between solids
and stage surfaces decreases with the increase of the clearance and orbit radius, which leads to the mitigation
of abrasion. Therefore, the changes of vibration amplitude, orbits radius and the sealing clearance are faster
at the beginning of the test. On the contrary, the erosion in ESPs only related to the primary flow region,
whose geometry changes can be ignored during the test. Therefore, it is convincible that the erosion rate in
ESPs is in a linear function with time. As shown in Figure 12 (d), weight losses in impeller and diffusers
are mainly caused by erosion.

Paint-removal test
After 64 hours’ testing, the wear pattern can be found on ESP impellers and diffusers. The abrasion
conditions on impeller seals are shown in Figure 13 (a) and (b). The seals on the first stage were damaged
more compared to that on the second. Similar to the seals, the erosion on impeller blades in prior stages is
worse than the downstream stages, as can be observed from Figure 13 (a) to (b). On the contrary, the erosion
pattern is clearer on each diffuser outside shroud in Figure 13 (d), which proves that the erosion rate on the
diffuser is higher than that on the impeller.
To capture the detailed erosion pattern and detect the most severe erosion location, the coated stages were
tested under the same flow conditions as shown above, except the flow case of rotation speed N = 2400
rpm. Different paints and primers were tested, where the brown primer on stage 5 is selected to present
the results. 6 hours slurry test was performed to obtain a clearer pattern as presented in Figure 14 (b), (d)
and Figure 15 (b), (d).
12 IPTC-19264-MS

Figure 13—Wear on ESP, (a) 64 hours 1st stage impeller, (b) 64

hours 2nd stage impeller, (c) original diffuser, (d) 64 hours diffuser

Figure 14—Paint-removal top view, (a) impeller original, (d) impeller

6 hours slurry, (c) diffuser original, (d) diffuser 6 hours slurry
IPTC-19264-MS 13

Figure 15—Paint-removal side view, (a) impeller original, (b) impeller

6 hours slurry, (c) diffuser original, (d) diffuser 6 hours slurry

From Figure 14 (b) and Figure 15 (b), the paint lost on impeller tips and impeller inside shroud are most
prominent. Some paint-removal can be detected on the outside shroud and blade surfaces that are close
to impeller outlet. As shown in Figure 14 (d) and Figure 15 (d), the erosion is obviously located on the
pressure side and the outside shroud in a diffuser, while the interaction area surprisingly suffers from less
paint-removal.

Conclusion
Before the slurry test, original ESP head, efficiency, and stage vibration under single-phase water flow
condition fit well with catalog curves. Later on, pump head and efficiency deteriorated with the sandy flow.
The significant decrease in total efficiency and the head curve can be caused by lost pressure rise due to
friction loss in primary flow region and leakage loss in the secondary flow region. High leakage can be
generated due to the high turbulence caused by a low flow rate and high-pressure increment. The averaged
head and efficiency degradation is 10% and 12% after 64 sandy tests respectively.
In erosion tests, the impeller down-thrust washer has worn out after 16h under 3100 bpd liquid-sand flow
rate with a rotation speed of 3600 rpm, and after 4h under 3100 bpd flow with 2400 rpm. Obvious abrasion
can be found and measured on sealing rings and shafts. Once the total weight loss is measured, the abrasion
and erosion weight loss can be calculated.
Abrasion can be observed and measured at the stage hub and sealing surfaces in the secondary flow field.
Both stage trajectory and clearance measurement show a higher abrasion rate at the first 16 hours test, after
which the vibration and abrasion rate becomes more stable. It is presumable that the small clearance and
pump shaft eccentricity caused higher load between particles and target surfaces, which resulted in higher
abrasion rate at the beginning of the test.
Erosion can be observed on the impeller blades, balance holes and diffuser shroud in the primary flow
passage. From the paint-removal on the selected stage, the erosion is more likely to be observed on impeller
blade tips, the inside impeller shroud close to the balance hole and the outside shroud close to impeller
outlet. High turbulence and shear flow in ESP can be the reason. The obvious erosion locates on diffuser
vane tips, flow channels and interaction zones between the impeller and diffuser. The face weighted average
erosion rate of a diffuser is dozens of times higher than that of an impeller. The paint-removal photo can be
used as a reference in CFD erosion simulations in mixed type ESPs.
14 IPTC-19264-MS

Acknowledgments
The authors appreciate Baker Hughes a GE company, GE, and Schlumberger for their pump information,
as well as Tulsa University Artificial Lift Projects (TUALP) members’ financial and technical support.

Reference
Archard, J.F., 1953, "Contact and Rubbing of Flat Surfaces." Journal of Applied Physics, 24(8), pp. 981–988.
Basaran, B., 2017, "CFD Simulation for the Erosion on Electrical Submersible Pump due to Viscosity and Air presence,"
In MS thesis, Texas A&M University, College Station, Texas.
Chi, Y., Zhou, S., Daraboina, N. et al. 2018, "Experimental Study of Wax Deposition under Two-Phase Gas-Oil Stratified
Flow. Proc.," The 11th North American Conference on Multiphase Production Technology, Banff, Alberta, Canada,
June 6-June 8.
Chen, P., Gao, D., Wang, Z., Huang, W., 2015, "Study on Multi-Segment Friction Factors Inversion in Extended-Reach
Well Based on an Enhanced PSO Model" Journal of Natural Gas Science and Engineering, 27, pp. 1780–1787.
Gulich, J., 2008, "Centrifugal Pumps", Berlin Heidelberg: Springer-Verlag
King, D., Traylor, F., & Stewart, R., 1983, "Abrasion Technology for Electric Submergible Pumps," In SPE Annual
Technical Conference and Exhibition. San Francisco, California: Society of Petroleum Engineers.
Kruger, S., Martin, N. and Dupont, P., 2010 "Assessment of Wear Erosion in Pump Impellers", Proceedings of the 26th
International Pump Users Symposium, Houston, Texas, USA.
Marsis, E. and Russel, R., 2013, "A State-of-the-Art Computational Fluid Dynamics Simulation for Erosion Rates
Prediction in a Bottom Hole Electrical Submersible Pump," In SPE Heavy Oil Conference, Calgary, Alberta, Canada
Meng M., and Z. S. Qiu, 2018, "Experiment study of mechanical properties and microstructures of bituminous coals
influenced by supercritical carbon dioxide", Fuel, 219, pp. 223–238.
Meng, M., Zamanipour, Z., Miska, S., Yu, M., Ozbayoglu, E.M., 2019, "Dynamic stress distribution around the wellbore
influenced by surge/swab pressure", Journal of Petroleum Science and Engineering, 172, pp. 1077–1091
Minemura, K. and Zhong, Y. 1995, "Numerical Prediction of Erosion Wear on Pump Casing under Solid-Water Two-
Phase Flow," Advances in Multiphase Flow, pp. 561–572
Morrison, G., Carvajal, N., Saleh, R., Bai, C., 2015, "the Measured Impact of Erosion on the Rotodynamic and
Performance Characteristics of a Mixed Flow ESP", In 44th Turbomachinery & 31st Pmp Symposia, Houston, Texas,
USA.
Morrison, G., Chen, Y., Steck, D., Chen, Y., Bai, C., Patil, A., 2017, "Effect of Gas Presence on Erosive Wear of Split-
Vane Electrical Submersible Pump", In 46th Turbomachinery & 33rd Pump Symposia, Houston, Texas, USA.
Pirouzpanah, S. and Morrison, G.L., 2014, "Predictive Erosion Modeling in an ESP Pump," 4th Joint US-European Fluid
Engineering Division Summer Meeting, Chicago, Illinois, USA
Shi, Y., Zhu, H., Zhang, J., Yin, B., Xu, R., Zhao, J., 2017, "Investigation of Condition Parameters in Each Stage of a
Three-stage Helico-axial Multiphase Pump via Numerical Simulation," In International Ocean and Polar Engineering
Conference, San Francisco, CA, USA.
Shi, Y., Zhu, H., Zhang, J., Zhang, J., Zhao, J., 2018, "Experiment and Numerical Study of a New Generation Three-Sage
Multiphase Pump," Journal of Petroleum Science and Engineering, 159, pp. 471–484.
Wilson, B.-L., 1990. "The Effects of Abrasives on Electrical Submersible Pumps," SPE Drilling and Engineering, 5 (02)
Zhang, J., 2018, "Improvements to a Semi-Mechanistic Erosion Prediction Procedure," In Ph.D. Thesis, University of
Tulsa, Tulsa, Oklahoma 74104.
Zheng, Dezhi, 2013, "Three-Phase Erosion Testing and Vibration Analysis of an Electrical Submersible Pump," In M.S.
Thesis, Texas A&M University, College Station, TX 77843.
Zhou, S., Jiang, K., Zhao, Y., Chi, Y., Wang, S., & Zhang, G. 2018a. Experimental Investigation of CO2 Hydrate Formation
in the Water Containing Graphite Nanoparticles and Tetra-n-butyl Ammonium Bromide. Journal of Chemical &
Engineering Data, 63(2), pp 389–394.
Zhou, S., Yan, H., Su, D., Navaneethakannan, S., & Chi, Y. 2018b. Investigation on the kinetics of carbon dioxide hydrate
formation using flow loop testing. Journal of Natural Gas Science and Engineering, 49, 385–392.
Zhu, J., Zhu, H., Zhang, J., Zhang, H.-Q., 2017, "An Experimental Study of Surfactant Effect on Gas Tolerance in Electrical
Submersible Pump (ESP)," In International Mechanical Engineering Congress and Exposition, Tampa, Florida, USA.
Zhu, J., Zhang, J., Zhu, H., Zhang, H.-Q., 2018a, "A Mechanistic Model to Predict Flow Pattern Transitions in Electrical
Submersible Pump under Gassy Flow Condition," In SPE Artificial Lift Conference and Exhibition, Woodlands, Texas,
USA.
Zhu, J., Zhang, J., Zhu, H., Zhang, H.-Q., 2018b, "A Mechanistic Model to Predict Flow Pattern Transitions in Electrical
Submersible Pump under Gassy Flow Condition," In SPE Artificial Lift Conference and Exhibition, Woodlands, Texas,
USA.
IPTC-19264-MS 15

Zhu, H., Zhu, J., Zhang, J., Zhang, H.-Q., 2018, "Sand Erosion Model Prediction, Selection and Comparison for Electrical
Submersible Pump (ESP) using CFD Method," In 5th Joint US-European Fluids Engineering Division Summer
Meeting, Montreal, Quebec, Canada.
Zum Gahr, K.-H., 1998, "Wear by hard particles," Tribology International, 31(10), pp.587–596