Académique Documents
Professionnel Documents
Culture Documents
Dwindling oil supply, high energy prices and the need to replenish reserves are
encouraging oil companies to invest in heavy-oil reservoirs. Heavy and viscous
oils present challenges in fluid analysis and obstacles to recovery that are being
surmounted by new technology and modifications of methods developed for
conventional oils.
Mohamed Beshry
Paul Krawchuk
Total E&P Canada
Calgary, Alberta
George Brown
Southampton, England
Rodrigo Calvo
Jess Alberto Caas Triana
Maca, Brazil
Roy Hathcock
Kyle Koerner
Devon Energy
Houston, Texas, USA
Trevor Hughes
Cambridge, England
Dibyatanu Kundu
Bombay, India
Jorge Lpez de Crdenas
Houston, Texas
Chris West
BP Exploration (Alaska) Inc.
Anchorage, Alaska
For help in preparation of this article, thanks to Cosan
Ayan, Jakarta, Indonesia; Hany Banna, Bombay, India;
Teresa Barron, Syncrude Canada Ltd., Fort McMurray,
Alberta, Canada; Celine Bellehumeur, Jonathan Bryan and
Apostolos Kantzas, University of Calgary, Alberta; Sheila
Dubey, Shell Global Solutions (US), Houston; Maurice
Dusseault, University of Waterloo, Ontario, Canada;
Joelle Fay, Gatwick, England; Abul Jamaluddin, Rosharon,
Texas; Robert Kleinberg, Ridgefield, Connecticut, USA;
David Law and Allan Peats, Calgary; Trey Lowe, Devon
Energy Internatlional, Houston; David Morrissey and Oliver
Mullins, Houston; and Tribor Rakela and Ricardo Vasques,
Sugar Land, Texas.
AIT (Array Induction Imager Tool), CMR-200 (Combinable
Magnetic Resonance), DSI (Dipole Shear Sonic Imager),
Hotline, LFA (Live Fluid Analyzer), MDT (Modular Formation
Dynamics Tester), PhaseTester, Platform Express,
Quicksilver Probe, REDA, VDA (Viscoelastic Diverting Acid)
and Vx are marks of Schlumberger.
THAI (Toe-to-Heel Air Injection) is a trademark of Archon
Technologies Ltd.
34
Oilfield Review
Summer 2006
Heavy oil
15%
Conventional oil
30%
Extraheavy oil
25%
> Total world oil reserves. Heavy oil, extraheavy oil and bitumen, make
up about 70% of the worlds total oil resources of 9 to 13 trillion bbl.
Height, m
Calgary
Fort McMurray
3,000
Oil sands
and heavy-oil
deposits
2,000
1,000
Sea level 0
No oil
or gas
Oil
and gas
Oil
and gas
Oil
and gas
1,000
No oil or gas
2,000
3,000
Precambrian
basement
2,800 m
below sea level
4,000
British Columbia
Alberta
Saskatchewan
> Geological setting of one of the worlds largest deposits of heavy oil. During mountain-building
events, foreland basins are formed in front of the mountain range by bending of the Earths crust.
Marine sediments in the basin (purple) become source rock for hydrocarbons (dark brown) that
migrate updip into sediments (orange) eroded from the newly built mountains. Microbes in these
relatively cool sediments biodegrade the oil, forming heavy oil and bitumen. Where the overburden
is less than 50 m [164 ft], the bitumen can be surface-mined.
35
oil/water contact, where conditions are conducive to microbial activity. The depositional
environment, the original oil composition, the
degree to which it has been degraded, the influx
of, or charging with, lighter oils and the final
pressure and temperature conditions make every
heavy-oil reservoir unique, and all of them
require different methods of recovery.
10,000,000
1,000,000
Viscosity, cP
100,000
10,000
1,000
100
10
1
0
50
100
150
200
250
300
Temperature, C
36
Recovery Methods
Heavy-oil recovery methods are divided into two
main types according to temperature. This is
because the key fluid property, viscosity, is highly
temperature-dependent; when warmed, heavy
oils become less viscous (left). Cold production
methodsthose that do not require addition of
heatcan be used when heavy-oil viscosity at
reservoir conditions is low enough to allow the oil
to flow at economic rates. Thermally assisted
methods are used when the oil must be heated
before it will flow.
The original cold method of heavy-oil recovery
is mining. Most heavy-oil mining occurs in openpit mines in Canada, but heavy oil has also been
recovered by subsurface mining in Russia.4 The
open-pit method is practical only in Canada where
the surface access and volume of the shallow oilsand depositsestimated at 28 billion m3
[176 billion barrels]make it economic.5
Canadian oil sands are recovered by truck
and shovel operations, then transported to
processing plants where warm water separates
bitumen from sand (right). The bitumen is
diluted with lighter hydrocarbons and upgraded
to form synthetic crude oil. After mining,
the land is refilled and reclaimed. An advantage
of the method is that it recovers about 80% of
the hydrocarbon. However, only approximately
20% of the reserves, or those down to about 75 m
[246 ft], can be accessed from the surface. In
2005, Canadian bitumen production was
175,000 m3/d [1.1 million bbl/d]. This is expected
to grow to 472,000 m3/d [3 million bbl/d] by 2015.6
Some heavy oils can be produced from
boreholes by primary cold production. Much of
the oil in the Orinoco heavy-oil belt in Venezuela
is currently being recovered by cold production,
as are reservoirs offshore Brazil.7 Horizontal and
multilateral wells are drilled to contact as much
of the reservoir as possible.8 Diluents, such as
naphtha, are injected to decrease fluid viscosity,
and artificial lift technology, such as electrical
submersible pumps (ESPs) and progressing
cavity pumps (PCPs) lift the hydrocarbons to the
surface for transport to an upgrader.9 An
advantage of the method is lower capital
> Bitumen recovery from oil sands. When overburden is less than 50 m, bitumen can be mined
from the surface. The process, depicted in
photographs oriented from top to bottom, starts
by recovering oil sands by truck and shovel
operations. The sands are transported to
processing plants where warm water separates
bitumen from sand. The bitumen is diluted with
lighter hydrocarbons and upgraded to form
synthetic crude oil. Finally, the land is refilled
and reclaimed. (Images courtesy of Syncrude
Canada Ltd.)
Oilfield Review
Summer 2006
> Slurry produced by cold heavy-oil production with sand (CHOPS). This tank-bottom sample was
recovered from a tank farm at an oil-cleaning battery near Lloydminster, Saskatchewan, Canada, and
is composed of approximately 10 to 20% fine-grained clay and silica, 20 to 30% viscous oil and 50 to
60% water. (Photograph courtesy of Maurice Dusseault.)
37
Stage 1:
Steam Injection
Stage 2:
Soak Phase
Stage 3:
Production
38
Oilfield Review
Summer 2006
Sand
Steam
Steam
injection
Heated
heavy oil
flows to
well
Shale
> Steam-assisted gravity drainage (SAGD, pronounced sag-dee). A pair of parallel horizontal wells is
drilled, one above the other. Steam is injected into the upper well to heat the heavy oil, reducing its
viscosity. Gravity causes the oil to flow down toward the producer.
B E A U F
O R T
S EA
km
Gamma Ray
0
100
miles
100
aruk
Rive
r
Milne
Harrison Point Prudhoe
Bay
Bay
Kup
Caprock
Oil
Deep Resistivity,
Induction
Alaska USA
CANADA
L-sand
Ugnu
M-sand
N-sand
Schrader
Bluff
O-sands
39
Technology availability
Completion costs
Drilling costs
Costs
Facility costs
Fuel requirements
CO2 emissions
Permafrost characteristics
Gas hydrates
Surface
Artificial lift configuration
Fluid characterization
Rock-mechanical properties
Geology
Subsurface
Petrophysics
Poor/critical
Reservoir recovery
Fair/important
Excellent/less important
Performance/Knowledge
Key Performance
Indicators
Production
> Sensitivity matrix from the Ugnu screening study, quantifying the sensitivity of each recovery
method to production, subsurface, surface and cost factors. Each matrix block was colored
according to factor sensitivity to performance or knowledge importance. In terms of performance,
green means excellent, yellow means fair and red means poor. In terms of knowledge importance,
green means less important, yellow means important and red means critical. For example, in the
production categories, CSS was rated excellent performance for production rate per well, reserves
per well and reserves recovery. Of the subsurface factors, for example, fluid characterization and
rock-mechanical properties are rated of critical knowledge importance for every EOR method
assessed. In the interactive version of the matrix, clicking on a box accesses the reports and studies
behind the evaluation.
40
Oilfield Review
Summer 2006
Measured viscosity, cP
1,000,000
Measured viscosity, cP
10,000,000
1,000,000
100,000
10,000
1,000
100
10
100,000
10,000
1,000
100
10
0.1
0.1
0.1
1
10
100
1,000
10,000
0.1
Oil T2, ms
Oil RHI
10,000,000
10,000,000
Measured viscosity, cP
10,000,000
Measured viscosity, cP
100,000
1,000
10
0.1
100,000
1,000
General
Tuned
10
0.1
0.1
10
1,000
100,000
NMR viscosity, cP
10,000,000
0.1
10
1,000
100,000
10,000,000
NMR viscosity, cP
> Correlation between measured viscosity and viscosity calculated using an empirical relationship
based on NMR parameters T2 and RHI. The correlation between measured and calculated viscosities
(left) is good, but improves when tuned to fit viscosity data acquired over a range of temperatures
(right). (Adapted from Bryan et al, reference 20.)
41
Small-Pore Porosity
Capillary-Bound Fluid
Density Porosity
0.6
m3/m3
Density Difference
-50
kg/m3
950
PEF
0
10
Neutron Porosity
gAPI
150
Caliper
125
mm
375
Bit Size
125
mm
Measured depth, m
0.6
Gamma Ray
375
m3/m3
Free Fluid
Invaded-Zone 0.6
Resistivity
T2 Distribution
1 ohm.m 1,000
AIT 90-in.
Investigation
m3/m3
0.6
m3/m3
0 0.3
m3/m3
ms
3,000
cP
1,000,000
T2 Cutoff
0 0.3
ms
3,000 1,000
cP
1,000,000
Shale
Volume
Fraction
Sw
1 vol/vol 0
X60
X70
X80
X90
> A continuous oil-viscosity log produced from Platform Express data and CMR-200 measurements, calibrated to laboratory oil-viscosity
values. From X64 to X80 m, the continuous viscosity log (Track 5) shows a viscosity gradient, with oil viscosity increasing from 30,000 to
300,000 cP.
42
Oilfield Review
Summer 2006
Optical Channels
Oil Fractions
Water
Water-Base Mud
Gas Fraction
Pressure
Temperature
Pumpout
Volume
50,000
Time, s
> Fluid-sampling log from South America, using the MDT dual-packer module and the spectroscopic
LFA Live Fluid Analyzer. The optical channels of the LFA measurement (top) are color-coded by optical
density, which corresponds to hydrocarbon-component chain length. Channel 1 (black) corresponds
to methane. Channel numbers increase upward. In this example, all optical channels show high
amplitudes, indicating an opaque heavy oil. For comparison, LFA results for a light oil are shown to
the right, with low amplitudes in most channels. In the water, water-base mud (WBM) and oil-fraction
track, blue corresponds to mass fraction of water, green represents mass fraction of oil, and reddish
brown corresponds to slugs of WBM. The absence of gas readings in the gas-fraction track is
another characteristic of heavy oil.
43
50
PA
KI
A
ST
C H I N A
45
NEP
AL
40
Rajasthan
35
Contamination, %
BANGLADESH
30
25
20
15
10
km
500
miles
SRI LANKA
500
Sample number
Barmer Basin
Shakti,
API 15 to 19
Bhagyam,
API 21 to 30
Mangala,
API 22 to 29
Aishwariya,
API 29 to 32
Kameshwari,
API 45 to 52
Raageshwari gas
Vijaya,
API 28 to 35
Vandana,
API 28 to 35
Saraswati,
API 40 to 42
Raageshwari oil,
API 32 to 36
Guda,
API 40 to 42
> Bhagyam field in the Barmer basin in Rajasthan, India, where Cairn Energy
produces crude oils with widely varying API gravities.
44
Oilfield Review
Summer 2006
25
Contamination, %
20
15
10
5
0
10
11
12
13
14
15
16
17
18
Sample number
> The Schlumberger Reservoir Fluids Center (SRFC), in Edmonton, Alberta. At SRFC, experts carry out
both research and engineering activities, focusing on areas of phase behavior, flow assurance,
enhanced oil recovery and heavy-oil production.
45
35
Composition, % by weight
30
25
Resin
Asphaltene
20
15
10
5
0
0
10
20
30
40
50
API gravity
> Correlation between API gravity and resin and asphaltene content from SARA analysis (bottom).
The heavier the oil, the greater the content of resin and asphaltene. In the photograph (top), the flask
contains the heated heavy-oil sample. The standing tube contains resin, the Petri dish contains
asphaltene, and the other tubes contain saturates and aromatics. (Data taken from Table 1 in
reference 30.)
46
Oilfield Review
Summer 2006
47
48
100,000
Viscosity, mPa.s
To test the effectiveness of the fluidrecombination technique, the fluid derived from
the recombination procedure can be tested
against wellhead samples for bubblepoint and
viscosity. When PVT and viscosity measurements
on recombined fluids give comparable results to
the wellhead samples, engineers are able to
generate an accurate field-specific model for
predicting the properties of the heavy oil.
In one case, an oil company was concerned
about the presence of emulsified water in some
South American live heavy oils.33 Most heavy oils
are produced along with water, whether the
water occurs naturally in the reservoir or has
been injected in the form of water or steam.
During the production process, shear forces
stemming from high flow rate through pumps or
flow constrictions may be great enough to cause
the water to become emulsified in the heavy oil,
leading to a rise in viscosity. This, in turn, will
affect the efficiency of artificial lift, dramatically
increase the energy required to transport the
heavy oils and, in some cases, impact the choice
of production equipment.
The viscosity and stability of oil-water
emulsions depend on water cut and on which
phase is continuous. The viscosity of oilcontinuous, or water-in-oil, emulsions may
increase by more than an order of magnitude over
the dry-oil viscosity. The viscosity of a water-in-oil
emulsion increases with water cut up to the
emulsion inversion point, beyond which the
continuous phase changes to water, producing an
oil-in-water emulsion. In oil-in-water emulsions,
viscosity decreases with water cut.
Characterizing the stability and viscosity of
the South American heavy-oil emulsion required
development of new experimental techniques at
SRFC. Most experimental work on emulsions is
performed on stock-tank oil samples. However,
live oils contain dissolved gases that may affect
the viscosity of the oil and emulsion. SRFC
engineers developed a technique to generate
emulsions in live oils by recombining stock-tank
oil samples with gas to create a live oil. The live
oil was then blended with water at various water
cuts in a high-pressure, high-temperature
(HPHT) shear cell. The shear cell generated
emulsions with an average droplet size of 2 to 5
microns. Visual inspection and drop-size analysis
confirmed that the live-oil emulsions remained
relatively stable up to the inversion point.
The apparent viscosity of the resulting
emulsions was measured at two pressures using
an HPHT capillary viscometer (top right). The
viscosity of the emulsified live heavy oil is clearly
higher than the water-free heavy oil, up to five
10,000
1,000
100
0
10
20
30
40
50
60
Choke
manifold
Steam exchanger
PhaseTester Vx 29
Separator 1440
> Surface equipment for testing a heavy-oil well in Brazil. By including the PhaseTester Vx multiphase
well testing technology, accurate three-phase flow measurements are possible. In conventional
systems, flow is measured only after being separated by the separator. The orange line represents
three-phase flow, with oil, water and gas. The separator outputs three individual phases.
Oilfield Review
Summer 2006
Cleanup
1st
flow
1st
buildup
Various
rates
2nd flow
1st rate
5,000
4,000
3,000
2,000
PhaseTester data
Separator data
1,000
0
0
24
48
72
96
120
Time, h
> PhaseTester oil flow-rate measurements (blue) for Devon Energy in Brazil, showing more stability
than flow-rate measurements made by traditional phase separators (red). Flow rates are in barrels
per day at stock-tank conditions.
Buildup 1
Buildup 2
Pressure, psia
2,900
2,500
2,100
1,700
3/12/2005
3/13/2005
3/14/2005
3/15/2005
3/16/2005
3/17/2005
Date
1,000
100
10
1
0.001
0.01
0.1
1.0
10
Time, h
49
Buildup 1
Buildup 2
Pressure, psia
2,900
2,500
2,100
1,700
3/12/2005
3/13/2005
3/14/2005
3/15/2005
3/16/2005
3/17/2005
Date
1,000
100
10
1
0.001
0.01
0.1
1.0
10
Time, h
Production
tubing
Artificial lift
Completion equipment,
liner packers and tools
Reservoir monitoring
and control
> A proposed completion for a horizontal CSS or SAGD well. Thermal liner hangers provide pressuretight seals to increase effectiveness of steam injection. The REDA Hotline 550 artificial lift system
operates continuously at up to 550F internal motor temperature. Distributed temperature sensing
(DTS) systems monitor temperature changes during steam injection and oil production.
50
Oilfield Review
Summer 2006
40
30
20
10
0
Oct - 01
Production rate
Steam/oil ratio
0
Apr - 02
Oct - 02
Apr - 03
Oct - 03
Apr - 04
Oct - 04
Apr - 05
Oct - 05
Date
> Decreasing the steam/oil ratio (SOR) while maintaining or building production rate. Reducing the
SOR decreases the energy required to heat heavy oil, lowers the volume of water produced and
also reduces water-treatment expenses. [Data are from Encana Investor Day, November 7, 2005
http://events.onlinebroadcasting.com/encana/110705/pdfs/oilsands.pdf (accessed July 28, 2006.)]
Producer
Well
Steam
Injector Well
Monitor Well
OB1AA
Guide string
Monitor Well
OB1B
Injection
tubing
Guide string
Production
tubing
ESP
Liner
hanger
Monitor Well
OB1C
Slotted
liner
Instrument string
Liner
Slotted
liner
> Total E&P Canada SAGD pilot project, with producer-injector SAGD pair of horizontal wells and three
monitor wells to record temperatures in the injector-producer region.
51
52
, C
Temperature
150
100
12/31/2004 11:14
12/14/2004 00:57
800
11/25/2004 15:15
11/07/2004 07:33
Tim
e
600
,m
pth
De
10/19/2004 22:52
10/01/2004 14:10
Temperature
> DTS data acquired for three months starting in October 2004, showing warming of
the injector-producer region. Depth increases from the heel to the toe. One zone near
the heel of the well did not warm as much as the rest of the intrawell region.
03/31/2005 18:39
03/25/2005 12:17
Time
03/18/2005 17:48
03/11/2005 23:19
03/05/2005 04:50
02/26/2005 10:21
ratu
re,
160
140
120
pe
Tem
100
Liner
600
800
Depth, m
Temperature
> DTS data acquired after the workover, showing rewarming of the injectorproducer region.
42. Freed DE, Burcaw L and Song Y-Q: Scaling Laws for
Diffusion Coefficients in Mixtures of Alkanes, Physical
Review Letters 94, no. 6 (February 17, 2005): 067602.
Freed DE, Lisitza NV, Sen P and Song Y-Q: Molecular
Composition and Dynamics of Diffusion Measurements,
in Mullins OC, Sheu EY, Hammami A and Marshall AG
(eds): Asphaltenes, Heavy Oils and Petroleomics.
New York City: Springer (in press).
Oilfield Review
0
100
200
300
Liner
installed
in January
2005
400
February
2005 spiral
tubing effect
Depth, m
500
February
2005 top of
well flowing
temperature
600
700
October 2004
bottom of well
temperature
800
900
1,000
80
100
120
140
160
180
200
Temperature, C
> Close-up of DTS data acquired after the workover, showing an highfrequency oscilation of up to 20C (red curve). In comparison, DTS data
from before the workover are much smoother (blue curve). The temperature
oscillation in the data acquired after the workover is caused by spiraling of
the DTS instrumentation. The temperature oscillation represents the
difference in temperature between the top and bottom of the borehole.
265
260
Elevation, m
255
250
Depth of injector
245
Depth of producer
Temperature gradient at the producer
240
235
0
50
100
150
200
250
Temperature, C
Summer 2006
53