Artificial lift plays a fundamental role in several international plays.

This
month, artificial lift in oil-sand and deepwater developments is highlighted.
In both cases, artificial lift is not just accelerating production but is vital to the
economic success of the overall development.
Today, the Canadian oil-sand developments have followed these different
approaches: mining, cyclic steam injection, and steam-assisted gravity drainage
(SAGD). As reservoir depth increases, artificial lift, rather than mining, is used
to bring hydrocarbons to the surface. While cyclic steam injection and SAGD
both push the high-temperature envelope for pumping, new techniques are being
proved for this harsh environment. During the production phase of cyclic steam
injection, the wells initially flow to the surface. Later production moves into the
flumping stage (simultaneous flowing and pumping) and then into pump-
ing. While sucker-rod pumping has been dominant, progressing-cavity pumps
and electrical submersible pumps (ESPs) are being proved under these extreme
conditions. In addition to downhole pumping, there is widespread use of surface
multiphase pumps. Technologies that allow pumping high-temperature fluids
without cooling represent opportunities to retain heat and improve the overall
process efficiency. Surface pumping of a high-temperature multiphase mixture
has become an established technology in Alberta.
For deep water, artificial lift plays a strategic role, enabling companies that pos-
sess advanced technology to develop fields that other companies may determine
to be too difficult. The ability to deliver an advanced artificial-lift system is the dif-
ference between booking reserves and walking away from deepwater discoveries.
Only recently have we seen rapid growth in the application of subsea multiphase
pumping. In addition to seafloor boosting, several other artificial-lift methods are
applied in deep water, including gas lift and wellbore ESPs. In all cases, the ulti-
mate success of these developments depends on the reliability of the artificial-lift
system combined with the cost of intervention.
Artificial Lift additional reading available
at the SPE eLibrary: www.spe.org
SPE 113904 Recent Advances and Practical Applications of Integrated
Production Modeling at Jack Asset in Deepwater Gulf of Mexico by Umut
Ozdogan, Chevron Energy Technology Company, et al.
IPTC 11594 A New Approach to Gas Lift Optimization Using an Integrated
Asset Model by Fernando Gutierrez, Schlumberger, et al.
SPE 110234 Overview of Beam-Pump Operations by O. Lynn Rowlan,
Echometer Company, et al.
Additional reading available
at OnePetro: www.onepetro.org
OTC 18820 The Use of Subsea Gas Lift in Deepwater Applications by
Subash S. Jayawardena, Shell Global Solutions, et al.
(See JPT June, 2008, page 64.)
Artificial Lift
TECHNOLOGY FOCUS
64 JPT JULY 2008
JPT
Stuart L. Scott, SPE, is an Artificial-Lift
Consultant for the Shell EP Americas
Region where he manages the artificial-lift
R&D program. He also is Chairman of the
Board of Petroleum Emerging Technology
Corporation where he leads technology
roundtables worldwide. Previously, Scott
was an Associate Professor and holder of
the John E. and Deborah F. Bethancourt
Professorship in Petroleum Engineering
at Texas A&M University. He was
awarded SPE Distinguished Membership
in 2006 and is a member of the Board
of Directors for the SPE Gulf Coast
Section and serves on the JPT Editorial
Committee. Scott holds BS and PhD
degrees in petroleum engineering and an
MS degree in computer science, all from
the University of Tulsa.
The full-length paper presents a review
of the artificial-lift (AL) and flow-assur-
ance challenges faced in the exploita-
tion of Jubarte heavy oil, starting from
the features of the Pilot Phase. The chal-
lenges posed and innovations proposed
and implemented for Phase 1 field
development, as well as expectations for
the subsequent Phase 2, are detailed.
Introduction
The Jubarte field is in the north part of
the Campos basin, approximately 80 km
offshore from the state of Esprito Santo,
at a water depth between 3,280 and
4,922 ft. The field discovery occurred
in January 2001, when an interval con-
taining 17API oil was identified in
sandstones of Maastrichtian age.
During the evaluation phase, a study
based on the concept of value informa-
tion supported the decision to drill a
horizontal appraisal well. In February
2002, the well was drilled, with a 3,510-ft
horizontal length, and was completed
with an openhole gravel pack. The result
was a well productivity index (PI) 13
times greater than the PI obtained with
the vertical well. An extended well test
(EWT) was proposed and approved. In
the EWT, which started in October 2002,
Petrobras made use of the floating pro-
duction, storage, and offloading (FPSO)
vessel Seillean, a dynamically positioned
vessel with a rig capable of perform-
ing light workover operations. The well
was connected to the FPSO through a
5
5
/8-in.-inside-diameter drillpipe riser.
Using an innovative solution, a 900-hp,
25,000-B/D-capacity electrical submers-
ible pump (ESP) was installed above
the wet tree. In case of ESP failure, the
drillpipe riser could be recovered by the
FPSO for pump retrieval and replace-
ment. The good results obtained in the
EWT led to its continuation in a Pilot
Phase, which lasted until January 2006.
After the Pilot Phase, the Jubarte
Phase 1 production was started with
the FPSO P-34. This new development
phase comprises four production wells,
and a 60,000-B/D oil-production rate is
expected. One of the wells represents an
innovative installation of the ESP on the
seabed, another well has a downhole ESP
installed, and the remaining two wells
produce with gas lift (GL). GL backup
can be used in the ESP wells while await-
ing repair in case of pump failure.
The ultimate Jubarte production sys-
tem (Phase 2), to be implemented in
2011, will comprise 11 new wells plus
the four wells from Phase1, and seven
water-injection wells, connected to the
FPSO P-57. During this production
phase, all production wells will run
ESPs installed on the seabed as the
main AL method, with GL as backup.
Phase 1
In Jubarte Phase 1, four horizontal sat-
ellite wells produce to the FPSO P-34.
This production phase does not com-
prise water injection into the reservoir.
Reservoir Characteristics. The Jubarte-
field sandstones are characterized by
their high porosity and permeability.
The porosity and the horizontal perme-
ability have average values of 23% and
1,200 md, respectively. The viscosity of
the oil at reservoir conditions is 14 cp,
and reservoir temperature is 169F.
These characteristics led to expectation
of a high PI for the well that produced
during the EWT/Pilot and for all other
wells of Phase 1. The extrapolated origi-
nal static pressure was 4,124 psi. After
1 year of production in the Pilot Phase,
this value did not experience any signifi-
cant change. The bubblepoint pressure
of the fluid determined in the laboratory
is 2,613 psi. The large difference between
the static pressure and bubblepoint pres-
sure opens the possibility of having ESPs
installed on the seabed. The Jubarte fluid
is considered heavy oil with a low gas/oil
ratio, which favors the use of ESPs.
Well Engineering. In Phase 1, the
well had a 6
5
/8-in. production string.
Long horizontal lengths, approximately
3,280 ft, were targeted to provide high
PIs. Sand control was by openhole
gravel pack. This technology proved to
be effective during the Pilot Phase.
Subsea Layout. The subsea layout was
designed to avoid interference with the
installation of the ultimate unit (Phase 2),
thus facilitating the future transition
between the two phases. Phase-1 flow-
lines followed a downhill profile, which
favors the occurrence of terrain-induced
slugs. All production flowlines are 6-in.-
diameter insulated flexible pipes. Service
lines are 4-in.-diameter flexible pipes.
For the wells using ESPs as the lift-
ing method, a special umbilical inte-
grated with power cable was developed.
Beyond its conventional functions of
hydraulic control, chemical injection,
and electric-signal carrier, the structure
carries the electrical-supply cable. This
new concept minimizes the number of
This article, written by Assistant
Technology Editor Karen Bybee, contains
highlights of paper OTC 19083, The
New Deepwater Oil and Gas Province
in Brazil: Flow Assurance and Artificial
Lift: Innovations for Jubarte Heavy
Oil, by Giovani Colodette, Carlos
A.G. Pereira, Czar A.M. Siqueira,
Geraldo A.S.M. Ribeiro, Roberto
Rodrigues, Joo S. de Matos, and
Marcos P. Ribeiro, Petrobras, origi-
nally prepared for the 2007 Offshore
Technology Conference, Houston, 30
April3 May. The paper has not been
peer reviewed.
Copyright 2007 Offshore Technology
Conference. Reproduced by permission.
Innovations for Deepwater Jubarte Heavy Oil
ARTIFICIAL LIFT
The full-length paper is available for purchase at OnePetro: www.onepetro.org.
JPT JULY 2008 65
66 JPT JULY 2008
risers in the FPSO and optimizes the
subsea layout.
AL. ESPs provide a higher production
potential than GL. They also make it
possible to accelerate production and,
consequently, profits. Another advan-
tage is that ESPs usually are indicated
in applications where it is necessary
to lift large amounts of water and
heavy, viscous oils. The Jubarte-fluid
characteristics favor ESP applications.
However, the definition of the opti-
mum AL method throughout field pro-
duction life is a complex task that
involves a large number of variables and
uncertainties. Phase 1 was designed to
serve as an intermediate stage for data
acquisition to determine the optimum
concession development. The main
objective of Phase 1 is to evaluate
uncertainties related to reservoir behav-
ior and to determine the efficiency of
the different technologies available for
AL, (i.e., ESPs and GL).
For ESPs, the good results obtained
in the Pilot Phase made personnel look
for alternatives that would make their
use possible technically and economi-
cally in the Jubarte development. Two
distinct ESP configurations were evalu-
ated in two Phase-1 wells:
1. An ESP outside the well, in a pump-
ing module on the seabed (Fig. 1).
2. An ESP inside the production well-
bore (Fig. 2).
In both arrangements, GL could be
used as a backup while awaiting ESP
repair in case of a pump failure. Both
ESP wells were equipped with man-
drels and orifice gas lift valves. This
redundancy is to minimize production
losses, improving the operational factor
of the lifting system. For the other two
wells of Phase 1, the AL method used
was GL.
ESP on Seabed Inside a Pumping
Module. The AL technology selected
for Well 7-JUB-02 HP was a 1,200-hp
ESP mounted inside a pumping mod-
ule, installed vertically in a 130-ft-
long 30-in.-diameter hollow pile. The
pumping system is composed of an
adaptive base and a recoverable pump
module mounted 650 ft away from
the well. The pumping module was
designed to allow bypass of the ESP,
if necessary. The pumping systems
parts, base, and module can be installed
separately or together. The developed
system also allows pigging operations
to be performed if necessary by use of
the bypass.
The original conceptual design con-
siders the possibility of the module
(ESP) to be recovered and/or installed
by cable, using a special boat. However,
studies are still being conducted to
make this flexibility possible. The
main benefits of the system are the
reduction of the costs and time of
intervention and elimination of work-
over tasks in case the pump needs to
be replaced. Another advantage of the
system design is that it makes possible
the use of ESPs with greater power
and lift capacity (larger diameter)
without increasing the production-
well diameter. The test of the first
prototype will occur during the pro-
duction period with P-34.
High-Power and High-Reliability ESP.
Another producer, Well 7-JUB-06 HA also
was equipped with an ESP. However, it
was installed downhole. This production
well consists of a horizontal satellite well
connected to P-34 through flexible pipes.
A horizontal tree, developed specifically
for this project, was used. The ESP
installed in this well is part of a research-
and-development project focused on the
development of a high-reliability, high-
power engine-pump set (1,200 hp) with
the capacity to lift large amounts of
liquid in deep and ultradeep waters. The
expectation is that the equipment will
operate without failure for at least 4 to
6 years. To ensure this performance and
equipment lifetime, a rigorous quality-
control program for the pumping system
has been prescribed, considering factory
acceptance tests, stackup tests, and string
tests. A new methodology for testing,
with rigorous test conditions, was speci-
fied, including vibration analysis of the
assembled equipment.
The tracking and performance analy-
sis of the downhole ESP during Phase 1
will be performed by means of measured
variables by the equipment sensors and
by the parameters of the variable-speed
drive. An important feature of the Well
7-JUB-06 HA system is the existence of
a manifold for chemical injection con-
nected at the sensor base. This device
makes it possible to inject chemical
products upstream of the pump intake
by means of a capillary tube incorporat-
ed into the power cable. Currently, the
use of a multiple-purpose product that
combines the functions of demulsifier
and scale prevention is foreseen.
Fig. 2Production system for Well 7-JUB-06 HA. Fig. 1Production system for Well 7-JUB-02 HP.
JPT
The need for high-temperature electri-
cal-submersible-pump (ESP) systems is
growing as the oil industry matures.
Canadas nonconventional oil reserves
are estimated at more than 1 trillion
bbl and Suncors heavy-oil reserves in
northern Alberta, Canada, are esti-
mated to have a potential production
of 14 billion bbl of crude oil, but the
traditional mining methods of recovery
do not make them all economically
reachable. It is estimated that less than
one-fifth of the oil-sands resource is
mineable. To deal with this, Suncor has
turned to in-situ steam-assisted gravity-
drainage (SAGD) operations as a key
part of its plans to increase bitumen
supply to its upgraders.
Heavy-Oil and Oil-Sands
Resources
The Canadian petroleum industry
is facing dwindling light-crude-oil
reserves. However, at the same time,
an abundance of established heavy-oil
and bitumen deposits remains virtu-
ally untapped. According to govern-
ment statistics, Canadas oil sands con-
tain nearly 175 billion bbl of crude-oil
reserves that can be processed with
todays technology, thus putting the
oil sands second only to Saudi Arabia
in crude-oil reserves. It is thought that
further technological improvements
would allow the recovery of more than
350 billion bbl of bitumen. The Alberta
Energy and Utilities Board estimates
annual bitumen production in Canada
will more than double from 1.1 mil-
lion B/D (December 2005) to 2.6 mil-
lion B/D by 2014. Currently, bitumen
and synthetic-oil production represents
approximately 50% of Canadas total
crude output.
Three major deposits of bitumen,
also called oil sands, are found in the
province of Alberta, Canada. These
deposits are the Athabasca, Peace River,
and Cold Lake oil sands. Despite the
technical challenges of producing and
processing bitumen, several factors
have made investments in oil sands
very attractive, given world oil prices.
There are no finding costs because
the oil sands are well delineated, there
is ready access to the largest market
in the world (the US) by established
pipelines, and new technology has
reduced operating cost by at least a
factor of two.
Surface Mining
The Athabasca oil-sand deposits occur
from the surface to a depth of 750 m
true vertical depth. Surface-mining
exploitation at depths of up to 100 m
historically has been the technique
commonly used by several producers
in the area. This form of exploitation
and development goes back to 1967
when Suncor Energy (Great Canadian
Oil Sands then) started the construc-
tion of the first oil-sands commercial-
mining and -production operation.
Since then, more than 3 billion bbl
has been produced. This successful
story has required the development
and improvement of technology such
as truck-and-shovel cold-water extrac-
tion, slurry pipelining, mechanical
separation, and the potential recovery
of byproducts.
In-Situ Operations
Approximately 80 to 90% of the
Canadian oil sands are too deep to be
This article, written by Assistant
Technology Editor Karen Bybee, con-
tains highlights of paper SPE 110103,
Pushing the Boundaries of Artificial-Lift
Applications: SAGD ESP Installations in
Canada, by F. Gaviria, SPE, Suncor,
and R. Santos, SPE, O. Rivas, SPE, and
Y. Luy, Schlumberger, originally pre-
pared for the 2007 SPE Annual Technical
Conference and Exhibition, Anaheim,
California, 1114 November. The paper
has not been peer reviewed.
SAGD ESP Installations in Canada
ARTIFICIAL LIFT
Fig. 1Well arrangement for SAGD production.
For a limited time, the full-length paper is available free to SPE members at www.spe.org/jpt.
JPT JULY 2008 67
68 JPT JULY 2008
surface mined economically. In-situ
thermal techniques are used to reach
and recover these reserves. Several
commercial projects in the Athabasca,
Cold Lake, and Peace River deposits
have a fairly long history of in-situ
operations. So far, cyclic steam stimu-
lation (CSS) and SAGD have emerged
as the two more suitable and proven
techniques. Vapor extraction, which is
the solvent analog to SAGD, is still in
the pilot phase.
In the last decade, some other
techniques and variations have been
under field trial and evaluation. Those
techniques include liquid addition to
enhance recovery, in which a small
amount of solvent (in this case, natural-
gas condensate) is added to the steam
being injected into the CSS to increase
bitumen recovery. Also included are
fast SAGD, a modification of the SAGD
process, which makes use of additional
single horizontal wells alongside the
SAGD well pair to expand the steam
chamber laterally, and a solvent-aided
process (SAP) that combines the ben-
efits of using steam with those of sol-
vents. In SAP, a small amount of hydro-
carbon solvent is introduced as an addi-
tive to the injected steam during SAGD.
Three more examples of innovative and
advanced technologies are toe-to-heel
air injection, in-situ combustion, and
reservoir electrical heating.
SAGD
The SAGD process, as shown in Fig. 1,
consists of two aligned horizontal well-
bores. Steam is injected into the top one,
and reservoir fluid is produced from the
bottom one. The process is noncyclic;
that is, steam is injected continuously
and fluids are produced continuously.
Around and above the injection well,
a steam chamber grows. The injected
steam flows into the steam chamber
and eventually comes into contact with
oil sand at its edge. The steam then
releases its latent heat to the oil sand,
the oil heats up, its viscosity drops, and
it flows (with water condensate) under
gravity down the inclined chamber
edge to the production well. The length
of the wells usually is between 500 and
1200 m. Typically, the interwell spac-
ing between the injection and produc-
tion wellbores is 5 m, and the produc-
tion well is located a few meters above
the bottom of the pay zone. The quality
of the injected steam should be as high
as possible at the sandface because any
condensate in the injected fluids falls
under gravity from the injector toward
the producer and does not deliver a sig-
nificant amount of heat to the oil sand.
To initiate the steam chamber, steam
circulation in each of the wellbores
often is carried out for up to 3 months
or more to heat the formation conduc-
tively between the two wellbores. After
the temperature of the oil between
the wellbores is sufficiently high, the
injector and producer are switched to
regular SAGD mode.
Economic efficiency of the SAGD
operation is dictated by natural-gas use
and water recycling and treating. The
most expensive activity in SAGD is the
generation of steam. Often, the thermal
efficiency is measured by the cumula-
tive steam/oil ratio (CSOR). The higher
the CSOR, the higher the steam usage,
which means more natural gas com-
busted per unit volume of produced
bitumen (as well as higher emissions
of greenhouse flue gases from combus-
tion) and, consequently, a less econom-
ic process. The lower the CSOR, the
lower the steam use per unit volume of
the produced bitumen, which implies
a more economical process. The eco-
nomics of SAGD is more sensitive to
the CSOR than to the oil-production
rate, and the higher the gas price, the
more important the CSOR is in deter-
mining the economics of SAGD.
Natural gas currently is the fuel of
choice for steam generation, upgrad-
ing, heat, and power. This comes at a
time when natural-gas supplies have
reached their peak and are expected to
decline. Currently, oil-sands operations
consume 5% of Canadas natural-gas
supply. With growth in production and
without fuel substitution, it is expected
that oil-sands operations will be using
approximately 1 billion cf/D of natural
gas, or the major part of the Arctic gas
expected to come to market over the
next 10 years.
There is a significant dependence on
water used for separation of oil from
sand in surface-mined operations and
for in-situ steam generation. To pro-
duce a barrel of bitumen or synthetic
oil, 10 bbl of water is required for min-
ing operations and 3 bbl of water for
in-situ operations. Although most of
the water is recycled, there is still
approximately 20% of potable makeup
water that is required, and this creates
concerns over the need for conserva-
tion and sustainability, especially con-
sidering the extensive use of water in
waterflooding of conventional oil and
in the energy industry as a whole.
Artificial-Lift (AL)
Systems at Firebag
SAGD operations have challenges simi-
lar to those in conventional oil produc-
tion: Increase production and recovery,
lower lifting cost, and reduce green-
house-gas emissions while achieving
these goals. Optimized performance,
run life, and operational flexibility of
the AL system are crucial for the reduc-
tion of the lifting cost, which is the out-
of-pocket cost per unit of production to
operate and maintain wells and related
equipment and facilities. The AL cost is
a key component of the economic fea-
sibility of a SAGD project. Developing
and applying the right technology will
enable SAGD wells to be produced
longer and in a more stable condi-
tion as well as enabling the reduction
of production losses and the number
of costly workover and well interven-
tions and the frequency of equipment
replacement.
Suncor, like the other SAGD opera-
tors, soon realized that the production
of SAGD wells on natural-flow mode,
which involves progressive increments
of steam-injection rates and pressures,
would be unsustainable. The possible
steam-chamber leakoff, keeping high
steam-chamber pressure so that the
produced fluids could be lifted to sur-
face, would eventually be restricted
by maximum installed steam-gener-
ation and -injection capacity at the
plant. Production rates then will start
to decline to a point at which no more
fluids can reach the surface. Therefore,
there was the need to look for a suitable
AL technology that could be installed
under the unique downhole conditions
of the SAGD wells.
The initial thought was that a high-
temperature multiphase pump would
be required to lift the very hot and
steamy fluids that are characteristic of
SAGD production. With this in mind,
Suncor worked with a Canadian AL
supplier to adapt an existing surface
multiphase pump for downhole appli-
cations. The first multiphase pump was
installed at Firebag in January 2005
after almost 2 years of pump engi-
neering changes and adjustments. Two
more of these pumps were installed in
May and December 2005. Technology
suitability and low power consumption
JPT JULY 2008 69
were proven; however, run-life targets
were not achieved.
After the multiphase-pump trial,
the AL strategy at Firebag switched
to ESPs. It was considered that these
pumps could be used in SAGD applica-
tions if the mixed subcool (the differ-
ence between local temperature along
the production liner and the saturation
temperature at the local pressure) at
the intake of the pump could be con-
trolled and optimized so the amount
of free steam/vapor was as low as pos-
sible. If the subcool gets too low, the
live steam is produced and the energy
efficiency of the process drops. Steam
breaks into the liner would cause liner
failures, possible sand production, and
pump cavitation if the intake pressure
gets lower than the specified net posi-
tive suction head.
With these considerations in mind,
the first ESP was installed at Firebag
in June 2005. Since then, 21 SAGD
producer wells have been completed
with ESPs. Pump performance has
been satisfactory, and run life has been
within the expectations and limitations
of existing technology.
The fieldwide implementation of AL
at Firebag has included the formation
of a formal project-management team
from different disciplines that covers
all the surface and subsurface aspects.
Facilities, production, and process
engineers remain in permanent contact
with reservoir, drilling, completions,
and AL engineers to monitor progress
and performance. This team also is in
charge of looking for opportunities for
continuous improvement of the AL.
ESPs at Firebag
ESPs are currently the preferred AL
method at Firebag. At least 80% of
current production is being lifted with
ESPs. The other 20% is produced by
SAGD natural flow. In the near future,
at least 100 to 120 wells will be com-
pleted with ESPs to achieve the in-situ
production target that Suncor is fore-
casting from now to 2012.
The first ESP system installed at
Firebag in June 2005 was developed
specifically for SAGD operations.
As a result of extensive engineering
research, design, and rigorous testing
for qualification in a hot-loop test well,
and successful field trials, this ESP sys-
tem is rated with maximum operating
temperature of 218C. This ESP system
has been in use in SAGD productions
since 2003, with cumulative run time
equivalent to 144 years of experience
in SAGD operations.
The use of ESPs at Firebag has
followed a very fast and progressive
learning curve and adjustments, since
the first installation in June 2005.
This rapid change and optimization
has been dictated by the high-quality
reservoir and very-prolific production
wells. As a result of this, more-efficient
pump installation and well-workover
procedures have been implemented.
New technology has been run for
downhole instrumentation. The size
and capacity of the pumps have gone
from 1000 to 2500 m
3
/d. Motor size
also has gone from 150 hp to 250 hp.
The same upsizing process has been
implemented for the surface electri-
cal gear. The intake configuration
has evolved from straight intake to
bottom-feeder gas separator (BFGS),
advanced gas handler (AGH), and a
combination of BFGS and AGH. JPT
The technical book resource for the
petroleum engineering professional
The new SPE Bookstore
undreds o addiLional LiLles
hew search eaLures
0iscounLed nenber pricing
h
f
w
I
I
I
L
f
S
&
f
f
A
I
u
k
f
S
www.spe.org/sLore
Society of Petroleum Engineers
Finding a reliable artificial-lift (AL)
pumping system for heavy-oil thermal
recovery has been a challenge, mainly
because of the high operating temper-
atures (greater than 150C). Available
options such as rod pumps and electri-
cal submersible pumps (ESPs), which
are well proved in the industry, are
not well suited for thermal production.
While rod pumps offer high-tempera-
ture service, they are limited in the
flow rate they can deliver. ESPs can
handle high volumes of low-viscos-
ity fluids but are still limited in terms
of maximum operating temperature.
Metal progressing-cavity-pump (PCP)
technology has been developed to
meet the high-temperature require-
ment of steam-assisted gravity drain-
age (SAGD) and other thermal-recov-
ery processes.
Introduction
Recovery of the worlds large reserves
of extraheavy oil (found mainly in the
oil sands of Canada and Venezuela) by
thermal processes has been increasing
because of the high price of oil. Key
to this recovery process is AL, which
is required because of the very high
density and viscosity of the crude
and the relatively low reservoir pres-
sures. Although gas lift remains an
AL option when high pressure gas
is available, pumping techniques are
more popular because of their rela-
tively higher efficiency and ability to
generate more pressure at the sur-
face for delivery to the treating plant.
However, the main challenge with
hot pumping is the high temperatures
often required [as much as 260C for
SAGD and 350C for cyclic steam
stimulation (CSS)].
The dominant pumping technolo-
gies available are beam/jack pumps,
ESPs, and elastomer PCPs. All these
pumps have their peculiar limita-
tions for hot production. While beam
pumps offer high-temperature service,
they are limited in the flow rate they
can deliver. ESPs can handle high vol-
umes of low-viscosity fluids, but they
are still limited in terms of maximum
operating temperature. For PCPs, the
limitation is operating temperature of
the elastomer (maximum 150C).
Metal-PCP Description
PCPs are known for their simplicity of
design and operation. The heart of the
pump is composed of two parts: the
stator and the rotor (Fig. 1). The sta-
tor has a dual helical profile, while the
rotor (which rotates inside the stator)
has a single helical profile designed
to mate with the stator profile. The
rotating action of the rotor (sitting
inside the stator) creates progressing
cavities from bottom, displacing the
fluid through each successive cavity,
hence the pumping action. PCPs are
nonpulsating pumps and will deliver
a constant flow rate for a given rotor
speed. In the conventional elastomer
PCP, the part of the stator with the
helical profile is made of elastomer
and is glued to an external metallic
tube. The rotor fits the stator with
negative clearance. For the metal PCP,
the stator is fully metallic and able
to withstand very high temperatures.
The metallic helical profile is pro-
duced by hydroforming. The stator is
composed of three 9-ft-long elements
welded together. Both rotor and stator
are specially coated for high tempera-
ture and wear resistance, but the rotor
serves as a sacrificial element.
Fig. 2 illustrates a typical rod-driv-
en-PCP assembly as set into a well.
The stator usually is run first with the
production string. The rotor is run
This article, written by Assistant Tech-
nology Editor Karen Bybee, contains
highlights of paper SPE 110479, Worlds
First Metal-PCP SAGD Field Test Shows
Promising Artificial-Lift Technology for
Heavy-Oil Hot Production: Joslyn Field
Case by Jean-Louis Beauquin, SPE,
and Felix Ndinemenu, SPE, Total E&P;
Gilles Chalier, Total E&P Canada;
Lionel Lemay and Laurent Seince,
PCM; and Alex Damnjanovic, KUDU
Industries, originally prepared for the
2007 SPE Annual Technical Conference
and Exhibition, Anaheim, California,
1114 November. The paper has not
been peer reviewed.
First Metal-PCP SAGD Field Test Shows Promise
for Heavy-Oil Hot Production
ARTIFICIAL LIFT
Delivery
pressure
Rotor
(steel)
Stator
Fig. 1PCP principle.
For a limited time, the full-length paper is available free to SPE members at www.spe.org/jpt.
70 JPT JULY 2008
JPT JULY 2008 71
afterward and set into the stator. The
rotor is connected to a drive head at
surface by either a standard sucker-
rod string, a continuous (coiled) rod,
or a hollow-rod string.
Advantages of the metal PCP are the
following.
Easy flow-rate control (propor-
tional to speed)
Easy to install, similar to conven-
tional PCP
Wide operating temperature range
(up to 350C)
Accepts low or high viscosities
Low net positive suction head
(NPSH) (i.e., operates with low bot-
tomhole pressure)
Nonshearing and no formation of
emulsions
Easy initial startup at higher vis-
cosities
The metal PCP currently is designed
to have:
A lifetime of 1 year minimum
(8,000 hours)
Ability to handle sand contained
in oil (up to 5%)
Development Status
Research efforts for metal-PCP devel-
opment through hydroforming tech-
nology were launched in the mid-
1990s. Several processes were tested
for developing a full metallic stator.
Only the hydroforming process was
successful in term of industrialization,
performance, and cost. By 2005, two
industrial prototypes were produced
and bench tested in hot conditions.
The tests comprised:
Performance tests at different
speeds (maximum 400 rev/min), pres-
sure drop (maximum 135 bar), and
temperature (maximum 200C)
Endurance test at 150C at maxi-
mum operating speed (400 rev/min)
and pressure drop (130 bar) for
6 weeks
Even for the first industrial proto-
types, the bench tests showed encour-
aging results, with overall efficiency
reaching 65%. They also confirmed
one of the strengths of the pump,
namely a broad viscosity handling
capability, which is an issue for beam
pumps and ESPs. For high viscosity,
the metal-PCP overall efficiency is
not very sensitive to pressure differ-
ential but improves with speed. For
lower viscosity (high temperature),
the overall efficiency decreases with
increasing pressure drop as a result of
internal fluid slippage, but this also is
improved by higher speed. Abrasion-
resistance testing also was performed
with various coatings on both rotor
and stator, leading to a satisfactory
industrial solution. Following these
performance, endurance, and abra-
sion tests, some improvements have
been made in the manufacturing pro-
cess and in the thickness of coat-
ings, providing superior performance
and resistance.
Since 2006, three models of the
metal PCP have been available to
cover a wide range of flow rates for
heavy-oil production. They are the
400MET1000, 550MET750, and the
1000MET500. The first number gives
the maximum rate in m
3
/d at zero
head and 500 rev/min, while the sec-
ond number gives the nominal head
capacity in meters of water equiva-
lent. The pumps are rated to 350C.
Currently, 20 metal/metal PCPs have
been installed in wells under CSS and
SAGD. Out of these, four failed, one
after 6,500 hours of operations and
the other three prematurely. Failure
analysis indicated that unbalanced
movement of the rod string contrib-
uted to the fatigue failure.
Joslyn Field Trial
The Joslyn lease is a very shallow
(less than 100 m true vertical depth
TVD) low-pressure oil-sands field in
the Athabasca oil sands of Alberta,
Canada. Crude oil is 8API with a
viscosity of approximately 1.7 million
cp at original reservoir conditions. In
a dedicated area of the lease, SAGD is
used to produce this bitumen through
horizontal-well pairs. Injected steam
heats the bitumen, which then flows
toward the production well under-
neath. SAGD production (Phase 1) at
Joslyn started in September 2004 with
a pilot well pair initially equipped with
a beam pump that was replaced a few
months later with an ESP because of
capacity limitations.
For Phase 2 of the project, 17 addi-
tional well pairs were drilled in 2006.
The first three well pairs to come
on stream in September 2006 were
equipped with ESPs. During that time,
a fourth well (Well A) was converted
to SAGD production using the first
metal PCP. A few months later, fol-
lowing an ESP failure in one of the
three wells running with an ESP, a
second metal PCP was installed in this
well (Well B). Currently, out of the
10 wells on production, nine are pro-
ducing with metal PCPs (including the
initial wells where the elastomer ESPs
have been replaced with metal PCPs)
and one well with high-temperature
elastomer PCP. All pumped wells are
equipped with topside and downhole
instrumentation to obtain real-time
well- and pump-performance data.
First Metal PCP. This first pump
model was a 550MET750, with a
theoretical capacity of 1.1 m
3
/d per
rev/min. The PCP was landed at
327 m measured depth (MD) at an
86 inclination with 4-in. produc-
tion tubing. A topdrive system was
used, with 1.9-in.-outside-diameter
(OD) hollow rods for motorization.
The PCP started producing mid-Octo-
ber 2006 at 150 rev/min. Initial volu-
metric efficiency was approximately
55%. During this startup phase of the
SAGD well pair, downhole tempera-
ture was less than 120C and bitumen
viscosity generated some resistance
to flow. It was during that period
that the lowest intake pressure at the
pump was seen, at 440 kPa. With the
increase of pressure communication
between injector and producer, pump
Electrical
Motor Drive Head
Rod
Centralizers
Sucker Rods
Wellhead
Stop
Bushing
Pump
(Stator/Rotor)
Torque
Anchor
Fig. 2Typical rod-driven-PCP
equipment.
intake pressure gradually increased
and pump speed was ramped up from
150 to 340 rev/min. Bottomhole tem-
perature increased to approximately
160C. Pump volumetric efficiency
remained relatively constant at 45 to
50%, but became very sensitive to
the bottomhole temperature above
160C. Approximately 5.5 months
after startup, pump efficiency seemed
to decrease slightly to approximately
40% and pump speed was increased
to 350 rev/min, and then 360 rev/min,
to compensate. This speed was regard-
ed as the maximum allowable speed to
avoid pump vibration problems. At the
time the full-length paper was written,
this PCP was running at 360 rev/min
with a 35% volumetric efficiency. This
pump has been running for 9 months
with almost no downtime. Considering
that it was the first installation, the
performance of this pump has exceed-
ed expectation.
Second Metal PCP. Well B was first
started with an ESP and produced
approximately 2.5 months before the
ESP failed. During this initial phase,
it was difficult to keep the ESP run-
ning continuously and impossible to
increase the total flow rate to more
than 50 m
3
/d. This ESP was oversized
for the initial well-production rate
and was working at the very low end
of its operating range. Additionally, it
is believed that the strong contrast of
viscosity of the produced fluid and the
limited capabilities of ESPs to handle
erratic inflow were the main reasons
for the poor performance of the ESP.
However, these are the typical well
conditions on production startup and
are even more acute at Joslyn because
of the limited achievable reservoir
heating before initiating production
and the unavoidable cool down of the
wellbore during pump installation.
The other two wells with ESPs showed
the same production problems.
When the ESP failed, the opportu-
nity was taken to try another metal
PCP to see if a PCP could resolve these
production issues. The second pump
model was a 400MET1000, with a
theoretical capacity of 0.8 m
3
/d/per
rev/min. The PCP was landed at 309 m
MD at an 88 inclination with 4-in.
production tubing. A topdrive system
was used, with 1.9-in.-OD hollow rods
for motorization. This PCP started
producing mid-February 2007, and
pump speed was quickly ramped up
to 155 rev/min and then gradually to
210 rev/min. Pump volumetric effi-
ciency was consistently very high, at
approximately 80%. Liquid-flow rate
reached 150 m
3
/d, and bottomhole
temperature was 145C with a pump
intake pressure of approximately
800 kPa.
Approximately 2 months after
startup, pump efficiency seemed to
drop to approximately 45 to 50%.
However, the well continued to warm
to 164C at the pump. When the full-
length paper was written, this PCP
was running at 330 rev/min with a
45% volumetric efficiency. This pump
has been running for 5 months with
almost no downtime.
JPT
More than 45,000
E&P technical papers
Resources for career
enhancement
Respected
conferences,
workshops,
forums
Your search for knowledge begins here.
You can research more than 50 years of
industry advancement in our vast eLibrary.
Find new resources for your projects in the
E&P Software Toolbox. Enhance your career
through networking and online training
opportunities. And thats just the beginning.
Its the E&P universe in large focus.
Search it at www.spe.org.
SPE.ORG
Industry search engine
for technical knowledge
SOCI ETY OF PETROLEUM ENGI NEERS
Search
JPT JULY 2008 73