SPE-185141-MS

Brazil Field Experience of ESP Performance with Viscous Emulsions and

High Gas Using Multi-Vane Pump MVP and High Power ESPs

Lissett Barrios and Marisela Rojas, Shell; Guilherme Monteiro and Neil Sleight, Shell Brazil

Copyright 2017, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Electric Submersible Pump Symposium held in The Woodlands, Texas, USA, 24-28 April 2017.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents
of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect
any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written
consent of the Society of Petroleum Engineers 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 SPE copyright.

Abstract
This paper provides field experience for the Caisson ESP Technology in subsea boosting system with
emulsions and high Gas Volumen Fraction (GVF) using the high power Electric Submersible Pumps (ESPs)
systems. Field experience and experimental performance are compared regarding the effects of high viscous
emulsion using conventional high capacity ESP systems and the effect of two phase (liquid & gas) fluids
on ESP with new technology for high GVF fields and high viscous applications.
The Electrical Submersible Pump (ESP) system is an important artificial lift method commonly used for
subsea boosting systems. Multiphase flow and viscous fluids cause problems in pump applications. Free gas
inside an ESP causes many operational problems such as loss of pump performance or gas lock condition.
The objective of this paper is to understand MVP performance for high GVF and viscous emulsions.
This paper provides a summary on the performance comparison for a high power ESP system for
viscous emulsions and Multi-Vane Pump (MVP) for high GVF wells for Shell major Projects BC-10. These
novel projects continue the long tradition of Shell’s leadership in the challenging deepwater environment.
Presented is the capability and effects of viscosity and two phase (liquid & gas) fluids using a 1025 series
pumps with a charge MVP in series; as well as a 875 series standard ESP system mixed-type pump, which
is a multistage centrifugal pumps for deep boreholes.
Extensive testing and qualification of the subsea boosting system was undertaken prior to field
application. The subsea boosting system experience for offshore operations is reported with new technology,
and the effects of viscosity and two phases in real conditions. MVP and high power pumps were proved to
be a reliable technology to use in field application managing GVF higher than 50% and high viscous fluid
as high as 1200cp as consequence of fluid emulsion. Correction factors needed to be applied to standard
design curves to ensure proper field design at opearting conditions.

Background
Shell is the operator in two major projects where there have been significant advancements in deepwater
subsea boosting technology, operations, and surveillance: Perdido in the Gulf of Mexico (on stream since
March 2010) and BC-10 offshore Brazil (on stream since July 2009). These novel projects continue the long
tradition of Shell’s leadership in the challenging deepwater environment by development and application
2 SPE-185141-MS

of a novel seafloor boosting system [1]. Electrical submersible pumps (ESP) were selected as the subsea
boosting system used by both the Perdido and BC-10 assets for increased ultimate recovery and higher
production rates. The subsea boosting system is located in a seafloor caisson. The caissons inside are used
to collect production fluids from the manifolded subsea wells. Extensive testing and qualification of the
subsea boosting system was undertaken prior to final configuration selection including construction of the
world's only 1500-hp ESP test facility capable of controlling multi-phase fluid viscosities and temperatures
in Shell Gasmer Prototype test facility in Houston [1].
Pump performance degrades due to friction losses resulting from the fluid viscosity. As the viscosity
increases, the maximum flow capacity of the pump is reduced and the required brake horsepower (BHP)
increases. Accurately predicting the hydraulic pump performance and horsepower requirement for high
output systems in deepwater can significantly affect predictions for total recoverable returns, development
needs and overall project economics [2].
To mitigate the risks further, Shell performed fullscale tests of caisson ESP systems to validate the
performance of the ESP components over the range of fluids viscosity, GVF and flow rates expected in the
field. These tests were conducted at a land-based facility rather than offshore to reduce costs and minimize
delays due to logistics. A comprehensive test plan was implemented and performance tests had been carried
out for these systems. Field pump performance were calibrated and compared with experiemental results
for validation and accuracy.
This paper considers three different pump configurations being use in Brazil Deepwater operations and
Gasmer Testing [2, 3]. Details of the ESP equipment, is given in Table 1 for each configuration. The ESP
system included a 1025 series tandem WJE 1000, 875 series WJE700 configurations and MVP system
configuration. These ESP systems used 875 series tandem seals, a single 2000 hp variable speed drive (VSD)
and a 1500 hp tandem motor. Test results of Gasmer Testing were discussed by Gilyard [1] and Barrios [2,
3]. A long shroud is used to ensure that fluids are directed past the motor to guarantee adequate cooling
of the motor.

Table 1—ESP Components in the Initial Configurations

Unit Motor Seal Chamber Section Pump Motor Jacket

5 Stage 1025 series MVP-

1600 HP, 725 Series, 4440 Tandem 875 Series With G400 with integral intake Yes, 100 foot 10.75-inch OD
Field: BC10 M0B010
Volt, 212 Amp, Tandem High Load Bearing. & 36 Stage P340 Tandem 40.5 lb/ft J-55 Casing
Compression Pumps

1600 HP, 725 Series, 4400 Tandem 875 Series With 36 Stage P340 Tandem Yes, 100 foot 10.75-inch OD
Field: BC10 MOBO 8
Volt, 212 Amp, Tandem High Load Bearing. Compression Pumps 40.5 lb/ft J-55 Casing

32 Stage WJE1000 Tandem

1500 HP, 725 Series, 4160 Tandem 875 Series With Yes, 100 foot 10.75-inch OD
Configuration "H" Compression Pump with
Volt, 212 Amp, Tandem High Load Bearing. 40.5 lb/ft J-55 Casing
Integral Intake

5 Stage 1025 series MVP-

1500 HP, 725 Series, 4160 Tandem 875 Series With G400 with integral intake & Yes, 100 foot 10.75-inch OD
Configuration "J"
Volt, 212 Amp, Tandem High Load Bearing. 26 Stage WJE1000 Tandem 40.5 lb/ft J-55 Casing
Compression Pumps

The ESP is being operated in varius conditions including, inlet pressure, viscosity, speed, flow rate,
% GVF and its effect on the ESP performance and reliability. Gasmer Testing results has provided the
guidelines for the field operations on pump correction factors for Head and Power [2, 3]. Various field
parameter envelopes are given in Table 2. Different oil emulsions are being study for various viscosities
from 1 to 1000cp. The WJE1000 and MVP pumps were tested in oil only, in order to understand the viscosity
effect on the pump performance in separate testing, up to 1000 cP viscosity and from 40 to 60 Hz speed
range. Viscosity correction factors, based on speed and flow rates were developed for accurately predicting
SPE-185141-MS 3

the pump performance in viscous applications and to further enhancing the effect of gas on viscous fluid
applications.

Table 2—Configuration G, H & J Test Parameters

Configuration H
Field: BC-10 Min - Max Range Configuration J Min - Max Range
Min - Max Range

Total Discharge Flow

15000 - 30000 11000 - 40000 11000 - 50000
Rate (oil + Gas) BPD

Intake Pressure (PSI) 700-1800 250 - 1000 500 - 1000

% GVF Up to 50% GVF Up to 50% GVF Up to 80% GVF

Viscosity (cP) 100 - 400 2 - 350 2 - 350

Speed (Hz) 50 - 62 50 - 62 50 - 62

Pump HP 1000 - 1500 300 - 1225 300 - 1500

Motor Loading (%) 45 to 100% 45 to 110% 45 to 100%

Amps 98170 - 200 98 - 235 98 - 220

Degradation effects on ESP performance were observed due to gas flow into the pump and viscosity
effects. Instantaneous oil rates and gas rates varied widely with little impact on the ESP in single-phase
performance. Deviation from manufacturer's predicted performance was observed when GVF exceeded
30-40% through the pump and they were applied in the field during design and operation. Emulsion viscosity
on the field up to 1200cp correlated fairly well with correction factors obtained during experimental work
using mineral oils.
The primary paper objectives for this work is to validate hydraulic performance of the Caisson ESP
systems in the field operations with high viscous emulsions and high GVF; and compare with testing results
in Gasmer Facility.

System Schematic
The BC-10 Caisson ESP system consisted of:

• A 300-foot caisson vessel, 36-inch OD suspended in the cased hole

• ESP suspended inside the caisson from its top tubing hanger

• Power and instrument cables running to and from the ESP, including medium-voltage, subsea
electrical power connectors
• Instruments connected to the ESP and to the production tubing to measure pressure, temperature,
flow, level and vibration at various locations within the caisson
• A control system for operating the ESP through its variable-frequency drive

• Variable-frequency drives to power the ESP(s)

Numerous experimentsl testing were performed varying pressures and temperatures, liquid flow rates,
rpm, and torque prior to field installation. This gas-liquid outdoor facility is a fully instrumented state-of-
the-art flow loop, capable of testing the ESP multistage prototype performance. For more information about
the loop refer to Barrios [2]
4 SPE-185141-MS

Analysis
Viscous Emulsions
Water-in-oil emulsions are important in the petroleum industry in production operations, where the water
cut of the emulsion can be as high as 65-70% in volume. The effective viscosity of water-in-oil emulsions
depends mainly on the volume fraction of dispersed phase (water) and temperature of the system, along
with other minor effects. Since oil is typically co-produced with water when the water breakthrough in
the reservoir, emulsions can be encountered in all phases of oil production such as inside the reservoirs,
completion, wellbores, wellheads, and topside facilities; transportation through pipelines and storage. The
formation of emulsions during oil production lead to high viscous fluid which cause costly problem such
as increased pressure drop causing lost of production.
Viscosity of emulsions can be substantially higher than the viscosity of either the oil or the water. This
is because emulsions show non-Newtonian behavior caused by the dispersed droplet at certain volume
fractions of the water phase known as water cut (WC), oilfield emulsions behave as shear-thinning, or
pseudo-plastic fluids, that is, as shear rate increases, their viscosity decreases. Figure 3 shows BC-10
viscosity emulsion up to a WC of 100%, viscosity increases up to 850cp as WC increases up to 70%. At 70%
fluids experience phase inversion. Phase inversion refers to a phenomenon that occurs when continuous
phase change, from oil to water and vice versa.
Mitigation. Injection of reusable surfactants and co-surfactants into a heavy oil reservoir in the absence
of any source of heat has been proposed for the primary use of viscosity reduction of the crude oils. The
viscosity of a W/O emulsion can be significantly reduced by adding some appropriate demulsifiers, and its
inner and outer phases can be inversed to form an O/W emulsion with lower bound water content.

Head-Flow-Power Performance
Pump Head performance correction factors (PCFs) have been a primary focus for the ESP industry. As
gas is produced through the pump the head degradation is the primary concern to ensure that the number
of stages is sufficient at the target frequency of operation. The motors are selected based on the number
of stages and the power predicted by modeling. As per general guidelines the motor may be oversized to
account for inaccuracies of the models and to allow for operational flexibility
As gas is introduced to the pump intake, the pump performance is deviating from single phase
performance due to two effects. First, increases in total flow rate and decreases in mixture density
(Homogeneous Model) and second, pump head degradation due to accumulation of gas inside the pump
blocking the flow path.
In this study we compare performance between Gasmer testing and field data. For the ESP performance
modeling, Baker Hughes AutographPC ESP software was used to match VFD and Head data provided from
both Gasmer and MOBO 10. The software allows modeling of each system and calculates losses and power
factor in the system given set manufacturer selections for motor, conductor size and cable length.
HCF is calculated using the following equation:

Eqn 1

Also, the data obtained from the VFDs is therefore compared to the predicted power requirement at the
output of the controller in the model. However, the hydraulic performance of the pump must first be matched
in the model to the actual performance given known measured flowrates for liquid and gas. Field data from
BC-10 for one of the Caisson ESPs was made available for validation of the correction factors attained from
the Gasmer test data. The data used was from the MOBO system that uses the MVP technology. Operational
testing at decreasing inlet pressures and increasing GVF was performed as part of the field commissioning
SPE-185141-MS 5

and performance verification for the system. This data was then used to create Autograph model matches
using the HPCFs that were developed. HPCF is calculated using the following equation:

Eqn 2

Data Analysis
Single Phase Viscous Flow
Pumps were tested in water at 3500 rpm (60 Hz) speed as factory accepted test (FAT) in manufacturer test
facility in Claremore OK. Viscosity correction factors for the pump were developed by testing at two speed
2625 and 3500 rpm, up to 2500 cP for WJE1000 pump and up to 1000 cP for MVP G470 pump, in the R&D
Viscous test loop of pump plant in Claremore, OK. The WJE1000 pump performance curves for the head
and BHP at 2625 rpm, up to 2500 cP is shown in the Figures 1. MVP-G470 Pump performance at 3500 rpm
(60 Hz) is shown in the Figures 2. Viscosity correction factors for flow, head and BHP were developed for
single phase viscous fluid to be used for Field operation at high emulsion viscosity.

Figure 1—WJE1200 Viscous Head-Flow & BHP-Flow Pump Performance

6 SPE-185141-MS

Figure 2—MVP G470 Viscous Head-Flow & BHP-Flow Pump Performance

Figure 3—BC-10 Fluids Emulsions

Effect of Emulsion Viscosity on Pump performance for BC-10 Operations

In general, as viscosity increases, the pump performance deteriorates and requires more horsepower for the
same conditions which emphasizethe importantance of a proper calculation of pump performance at these
conditions to ultimately avoiding production deferment or improper design/pump selection. Figure 4 shows
the HCFs for field BC-10 and Gasmer experiments for Viscosity up to 400cp. HCF for all the viscosity are
slightly higher or equal to 1 which indicate no additional correction factor beside the emulsion increased
viscosity required to match pump performance. HCFs for all viscosities are approximately the same for
BC-10 data and Gasmer experimental work for 15% GVF. Experiment at Gasmer marched fairly well with
field ESP performance.
SPE-185141-MS 7

Figure 4—HCF at different Emulsion Viscosity for Field and Gasmer Testing

BC-10 Subsea Emulsion Breaker Trial

BC-10 has suffered from high viscosity for several years as WC increased, in Q4 2014 the emulsion strength
increased dramatically requiring continuous hot oil supply to decrease the discharge pressure required to lift
the fluids to surface. Demulsifier (DM EC2314A) was tested by injection into the inlet of one of the Caisson
ESP systems at different injection rates. Tested rates of 20-150ppm (vol. oil) were used. It was observed
impact on pump operation and flow line operation with all other parameters constant such as inlet pump
pressure, WC, GVF through the pump. ESP testing parameters for the trial are:

• ESP rate: 11000BPD

• GVF at pump intake: 15%

• Pump suction pressure: 800psia

Pipesim modelling indicates average viscosity dropped from 900cP to 500cP with DM in the flowline.
Figure 5 shows the results of the testing using demulsifier. Pump inlet viscosity reduces from 1200cp
initially without DM injection to 100cp at the higher injection rate. Likewise, pump speed reduces from
57Hz initially to 48Hz at the end of the testing proving a huge improvement in pump performance due to
viscosity decreasing and reduction in boost demand due to flowline back pressure reduction. ESP/Pump
also experienced a big reduction in friction as viscosity was reduced by 90%. Using DM indicates the ESP
suction pressure could be reduced to increase production since additional power and pump speed is available
at lower emulsion viscosity
8 SPE-185141-MS

Figure 5—Apparent Emulsion Viscosity and Pump Speed as a

Function of Demulsifier Injection Rate for BC-10 Field Trial Operations

Effect of High Gas Volume Fraction (GVF) for BC-10 Operations

As gas is introduced to the pump intake, the pump performance starts deviating from single phase
performance due to these effects: increase in total flow rate, decrease in mixture density (Homogeneous
Model) and pump head degradation due to accumulation of gas inside the pump blocking the flow path.
Pump performance is obtained using Autograph PC Software to match field conditions. Equation 1 was
used to calculate HCF. If HCF is equal or greater than 1, the system has homogeneus flow regime. When
HCF starts having values lower than 1, head degradation is observed.
Shell BC-10 development uses Electric Submersible Pump (ESP) systems similar to Gasmer equipment
installed in shallow Caissons as the artificial lift solution for subsea mudline boosting. A Baker Hughes
P340 primary pump with 36 stages and five stage MVP G400 charge pump section as the inlet was deployed
in the BC-10 Field for MOBO 10 unit (Table 1). Data was collected for production with GVF of up to 55%
at the pump inlet. Figure 6 and 7 shows the HCF and HPCF comparison between the field data and Gasmer
testing data using the MVP stages for heavy oil (~100cp).
SPE-185141-MS 9

Figure 6—HCF at different GVF for Field and Gasmer Testing

Figure 7—HCF at different GVF for Field and Gasmer Testing

Figure 6 shows less degradation effect as the GVF increases, at 40% GVF the pump starts deviating from
Homogeneus behavior and HCF starts decreasing less than 1 which indicate accumulation and or bubble
size increasing inside the pump vanes. At 55% GVF the HCF is ~0.8 for field operation as well for Gasmer
testing, indicating same degradation behaviors.
10 SPE-185141-MS

Figure 7 shows better performance in HPCF for the Field which translates to lower horsepower
requirements for the field application as compared to Gasmer testing results. A possible explanation for
the variance is that the free gas at the pump inlet as part of the produced hydrocarbon is compressed into
solution as the pump adds pressure to the fluid. This decreases the density and viscosity of the liquid phase
which in turn reduces the Horse Power requirements. At Gasmer, the test oil does not readily absorb the
nitrogen gas that is used for the multiphase fluid testing [2,4].

Performance Summary
The results in this study provide a better understanding of actual head and horsepower requirement for
high output systems in deepwater applications where both high viscous emulsions and high GVF which
have a significant impact on pump performance and efficiency translated to total field production. This
can significantly affect predictions for total recoverable oil reserves, development needs and overall project
economics. For planning purposes, the accurate prediction of the pump performance in term of Head and
Power correction factors (HCF & HPCF) translates to accurate production profiles for Brazil BC-10 field
operations and any other field with high GOR and viscous emulsions.

Conclusions
• Pump manufacturers need to supply guidelines for predicting head performance and horsepower
requirements of multi stage pumps in viscous emulsion and two phase flow conditions to provide
correction factor for each pump configuration if testing is not an option.
• MVP behaves as expected for the Field application for Head and Power requiremets as compare
to the Gasmer Testing, similar correction factors were calculated to determine pump behaviors
• Viscosity correction factor using sintetic oil in the experimental lab, predicts pretty accurate the
performance of heavy viscous fluid cause by emulsions
• Using demulsifier to break emulsions highly improves pump performance by decreasing apparent
viscosity by 90%
• For subsea and other applications requiring high installation and intervention cost, the system
testing in the test loop provides valuable information for reliable system design, capacity and
operations as per Brazil Caisson ESP experience

Acknowledgments
The authors would like to acknowledge Shell Brazil and joint venture partners QPI and ONGC for
permission to publish this paper.

Nomenclature
FAT = Factory Acceptance Test
BEP = Best Efficiency Point
BHP = Brake Horse Power, HP
GVF = Gas Void Fraction (Gas volume / Total Fluid Volume)
HCF = Head Correction Factor
HPCF = Horse Power Correction Factor
cP = Centi Poise, Dynamic Viscosity
cSt = Centi Stokes, Kinematic Viscosity
DP = Differential Pressure across the pump
WC = Water Cut
W/O = Water in Oil Emulsion
O/W = Oil in Water Emulsion
SPE-185141-MS 11

MOBO = Module of Bombeo

ESP = Electrical Submersible Pump
DM = Demulsifier
MVP = Multi-Vane Pump
GOR = Gas Oil Ratio

References
1. Gilyard, D., Brookbank, E. B., "The Development of Subsea Boosting Capabilities for Deep
water Perdido and BC-10 Assets", SPE 13493, Oct 2010
2. Barrios, L. PhD, Deuel C., Knowles D., Patni S. and Scott S. PhD - Shell E&P & Ketan Sheth -
Baker Hughes. "Large-Scale Experimental Investigation of ESP Performance with High Viscosity
Fluids and Gas", ESP Workshop 2011, April 2010
3. L. Barrios, and S. Scott, SPE, Shell Exploration& Production; R. Rivera, Kinetic Pump
Innovations, LP. and K. Sheth, SPE, Baker Hughes Inc. "ESP Technology Maturation: Subsea
Boosting System with High GOR and Viscous Fluid", SPE ATCE October 2012
4. Barrios, L., Rojas M. and Scott S. - Shell and Rivera R. - KPI, LP, "Horsepower Correction
Factor Caisson ESP system in Moderate GOR Application". 28th Electric Submersible Pump
Workshop- April 22,24, 2015. The Waterway Marriott The Woodlands, Texas
5. Sheth, K. K., Crossley, A., "Viscosity Correction Factors", SPE Gulf Coast ESP Workshop, April-
May 2009.
6. Barrios, L PhD; Scott, Stuart., Rivera Robert - Shell oil E&P and Padro Mauricio - The university
of Tulsa. "Surveillance models of Large-Scale ESP performance with hig Viscosity Fluids and
Gas", SPE 152217, May 2012.
7. Stepanoff, A. J., "Centrifugal and Axial Flow Pumps," 2nd Edition, John Wiley & Sons, pp
307–318, 1957