Sie sind auf Seite 1von 15

56 Oileld Review

Advancing Fluid-Property Measurements


Soraya Betancourt
Cambridge, Massachusetts, USA
Tara Davies
Ray Kennedy
Edmonton, Alberta, Canada
Chengli Dong
Sugar Land, Texas, USA
Hani Elshahawi
Shell International Exploration and Production
Houston, Texas
Oliver C. Mullins
John Nighswander
Houston, Texas
Michael OKeefe
Hobart, Tasmania, Australia
For help in preparation of this article, thanks to Gretchen Gillis
and Don Williamson, Sugar Land, Texas; and Lisa Stewart,
Cambridge, Massachusetts.
CHDT (Cased Hole Dynamics Tester), Fluid Proling, LFA (Live
Fluid Analyzer), MDT (Modular Formation Dynamics Tester),
Oilphase-DBR, PVT Express, Quicksilver Probe and RealView
are marks of Schlumberger
OLGA is a trademark of Scandpower AS.
Reservoir-uid properties play a key role in designing and optimizing well completions
and surface production facilities to manage reservoirs efciently. Therefore, accurate
uid characterization is a vital part of any oil or gas production project. Advanced
uid-analysis techniques provide the high-quality data required to develop appropriate
production strategies.
Discovery of an oil or gas accumulation imme di -
ately prompts questions about its economic
viability. Operators want to learn about the
extent of the reservoir, the types of fluids that
would be produced, expected production rates
and how long production might be sustained.
Fluid analysis is a critical part of the process by
which engineers perform reservoir characteri -
zation, determine the reservoir architecture and
decide whether an oil or gas accumulation is
worth developing. High-quality samples are
essential, because erroneous data could lead
engineers to misinterpret production parameters
such as drainage volume, flow rates, reserves and
facilities design and completion. Clearly, poor or
misleading fluid data can have a severe negative
financial impact.
If the reservoir analysis is positive, engineers
begin designing a production system that will
efficiently transport the reservoir fluids from the
formation through wells, flowlines, production
facilities and beyond. During this journey,
reservoir fluids experience temperatures and
pressures far different from their initial in-situ
conditions. These variations may induce
physical-state changes that would inhibit or
interrupt production if not understood prior to
designing tubulars and facilities. Therefore, to
determine how the fluids will respond to
production conditions, engineers may want to
collect and analyze fluid samples from each
potentially productive layer in the reservoir.
Traditionally, fluid samples are collected and
sent to offsite laboratories for testing, a process
that delays data access and impedes an opera -
tors ability to make time-sensitive development
decisions. Today, sophisticated formation
sampling and testing tools allow data collection
1. Ratulowski J, Amin A, Hammami A, Muhammed M and
Riding M: Flow Assurance and Subsea Productivity:
Closing the Loop with Connectivity and Measurements,
paper SPE 90244, presented at the SPE Annual
Technical Conference and Exhibition, Houston,
September 2629, 2004.
2. For more on scale-removal techniques:
Crabtree M, Eslinger D, Fletcher P, Miller M, Johnson A
and King G: Fighting ScaleRemoval and Prevention,
Oilfield Review 11, no. 3 (Autumn 1999): 3045.
>
Typical deepwater Gulf of Mexico oil-phase diagram. During the journey from
the reservoir to the flowline, the oil temperature and pressure decline, and may
cross phase boundaries at which asphaltenes (purple), waxes (blue) and
hydrates (green) will tend to separate and form solid deposits. Gas begins to
separate from the oil as it passes through the bubblepoint boundary (red).
Wax
Hydrate
Bubblepoint
Reservoir
16,000
14,000
12,000
10,000
8,000
6,000
4,000
2,000
0
0 50 100 150
Temperature, F
P
r
e
s
s
u
r
e
,

p
s
i
200 250 300
Asphaltene
Flowline
60336schD7R1.qxp:Format Layout V7.2 12/4/07 7:11 PM Page 56
Autumn 2007 57
much earlier in the exploration process,
including the ability to conduct real-time down -
hole Fluid Profiling characterization of
reservoir-fluid properties and quantification of
their variation. This technology gives operators
the ability to evaluate the reservoir while the
sampling tool is in the borehole, and acquire
additional data if the reservoir is more complex
than previously thought. In addition, engineers
and fluid-property specialists can better
determine where and when to sample, and how
many samples to collect. As a result, the quality
of fluid samples brought to surface is
substantially improved.
In the laboratory, chemists determine fluid
compositions, the temperatures and pressures at
which phase transitions occur, and how each
phase behaves as a function of temperature and
pressure. Accurate fluid characterization and
knowledge of pressure-volume-temperature
(PVT) behavior are critical for making
appropriate, cost-effective decisions about well
planning, well construction, production and
monitoring. When initial PVT screening and
thermodynamic modeling identify nonstandard
phase behavior (such as emulsions, wax or
asphaltene precipitation, hydrates and scales),
specialized testing is often performed to better
understand reservoir-fluid behavior. All these
activities fall under a general umbrella called
flow assurance.
When confronted with potential flow-
assurance problems, engineers have several ways
to mitigate or prevent difficulties.
1
These
methods include thermal management (hot-fluid
circulation, electrical heating and insulation),
pressure management (pumping, boosting and
blowdown) and chemical treatments. These
techniques adjust the pressure-temperature path
that hydrocarbons experience during production
or, in the case of chemical treatments, alter the
fluid composition to prevent phase changes or
disperse solid particles when precipitation
occurs. In addition, there are physical
remediation techniques such as pigging, jetting
and cutting.
2
As E&P companies venture into increasingly
remote production environments, particularly
deep water, flow assurance is critically
important. Deepwater reservoir fluids follow a
tortuous PVT path from the formation to the
production facility, increasing the probability of
flow-assurance difficulties (previous page). Flow
assurance may also be a concern in arctic
environments where thermal differences
between the reservoir and surface facilities can
be extreme. Accurate knowledge of PVT behavior
is vital because reservoir-fluid problems in these
remote locations could threaten the economic
viability of a project.
This article describes the roles of chemistry,
geology and thermodynamics during reservoir-
characterization and flow-assurance studies at
the wellsite and the laboratory. Also presented
are two offshore-field case studies that demon -
strate how these activities benefit well-completion
design and operation.
Condensate
Median-weight oil
Water
Hydrocarbons
45
4
0
5
0
6
0
7
0
8
0
9
0
40
35
30
45
40
35
30
25
20
15
10
5
0
25
20
15
10
5
5
4
3
2
1
0
0
60336schD7R1.qxp:Format Layout V7.2 12/6/07 4:43 PM Page 57
Reservoir-Fluid Sampling
Scientists and engineers of various disciplines rely
on fluid-sample data when making decisions about
reservoir development. For example, reservoir
engineers use the data to determine reservoir
architecture, estimate reserves, perform material-
balance calculations and analyze fluid flow in
porous media. Geologists need accurate informa -
tion to perform reservoir correlations and
geochemical studies. Refining and marketing
personnel make decisions about product yield and
value. If erroneous data are used, unanticipated
and expensive consequences could result
during production.
3
A wide range of fluid behaviors can influence
a sampling and analysis program. A reservoir-
fluid system can be roughly categorized by its
vapor-liquid phase behavior; the classifications
range from dry gas, wet gas and retrograde gas to
volatile oil, black oil or heavy oil (left).
4
Another
consideration is hydrocarbon solid-phase
behavior. Wax and hydrate formation is
predominantly induced by a temperature
decline, and pressure reductions or fluid
commingling generally cause asphaltenes to fall
out of solution (next page, top).
5
Engineers collect formation-water samples to
determine whether calcite, barite or halite
scales will form within flowlines. Corrosive and
toxic substances such as carbon dioxide [CO
2
]
and hydrogen sulfide [H
2
S] must be detected and
measured because they influence tubular-alloy
selection and the design of safety and
environmental systems from the wellhead to the
surface production facility. Water pH is also an
important parameter governing scale and
corrosion, and may be measured downhole to
avoid uncertainties.
6
58 Oileld Review
3. Nagarajan NR, Honarpour MM and Sampath K:
Reservoir-Fluid Sampling and CharacterizationKey to
Efficient Reservoir Management, Journal of Petroleum
Technology 59, no. 8 (August 2007): 8091.
4. McCain WD Jr: The Five Reservoir Fluids, in The
Properties of Reservoir Fluids (2nd Edition). Tulsa:
PennWell Books (1990): 147164.
5. Akbarzadeh K, Hammami, A, Kharrat A, Zhang D,
Allenson S, Creek J, Kabir S, Jamaluddin A, Marshall AG,
Rodgers RP, Mullins OC and Solbakken T: Asphaltenes
Problematic but Rich in Potential, Oilfield Review 19,
no. 2 (Summer 2007): 2243.
6. Raghuraman B, Gustavson G, Mullins OC and Rabbito P:
Spectroscopic pH Measurement for High Temperatures,
Pressures and Ionic Strength, AIChE Journal 52, no. 9
(2006): 32573265.
Xian C, Raghuraman B, Carnegie A, Goiran P-O and
Berrim A: Downhole pH as a Novel Measurement Tool
in Formation Evaluation and Reservoir Monitoring,
Transactions of the 48th SPWLA Annual Logging
Symposium, Austin, Texas, June 36, 2007, paper JJ.
7. Riemens WG, Schulte AM and de Jong LNG: Birba Field
PVT Variations Along the Hydrocarbon Column and
Confirmatory Field Tests, Journal of Petroleum
Technology 40, no. 1 (January 1988): 8388.
8. Ruiz-Morales Y, Wu X and Mullins O: Electronic
Absorption Edge of Crude Oils and Asphaltenes Analyzed
by Molecular Orbital Calculations with Optical
Spectroscopy, Energy & Fuels 21, no. 2 (2007): 944952.
>
Generalized pressure-temperature (PT) diagram for reservoir fluids. The
diagram contains two principal regions: single-phase (green to orange) and
two-phase (beige). The boundary between these regions is called the
saturation envelope; it has three principal features. The bubblepoint locus is
the portion at which gas begins to separate from liquid. The dewpoint locus is
the segment at which liquid begins to condense from gas. The critical point is
the location where the loci meet. The cricondentherm is the highest
temperature on the saturation envelope, and the cricondenbar is the highest
pressure on the saturation envelope. Reservoir fluids are classified according
to their in-situ reservoir and production behavior in the PT scheme. Dry gas
does not enter the two-phase region at any point during the production PT
path. Wet gas remains a single-phase system in the reservoir regardless of
pressure depletion; however, during production, it crosses the dewpoint locus
and forms a liquid phase. Retrograde gas resides in the single-phase region at
temperatures between the critical point and the cricondentherm. During
pressure depletion at reservoir temperature, liquid forms within the reservoir
itself, and persists throughout the production PT path. Volatile oil resides in the
single-phase region just to the left of the critical point. Gas liberation occurs as
the fluid crosses the bubblepoint locus during production. Black oil exists in the
single-phase region at reservoir temperatures far lower than the critical point.
Gas evolves during production, but the relative proportion of gas is small
compared with that of volatile oil. Heavy oil is a subset of black oil that contains
very low quantities of gas, and the liquid phase is predominantly composed of
high-molecular-weight components.
Oil
Gas
50%
Critical
point
40%
Cricondenbar
C
r
i
c
o
n
d
e
n
t
h
e
r
m
20%
10%
5%
Temperature
Gas Field
with Retrograde
Condensation
Gas Field with
No Retrograde
Condensation
Oil Field
with Dissolved Gas
0%
D
e
w
p
o
i
n
t

l
o
c
u
s
P
r
e
s
s
u
r
e
B
u
b
b
l
e
p
o
i
n
t
lo
c
u
s
Oil and gas
>
Crude-oil samples from a single column of oil in a reservoir. The continuous color change is a vivid illustration of compositional grading. (Photograph is
courtesy of Shell.)
Increasing depth
60336schD7R1.qxp:Format Layout V7.2 12/4/07 7:11 PM Page 58
Autumn 2007 59
Another concern is the variability of
reservoir-fluid composition within a field or
formation. Petroleum reservoirs may consist of
compartments that are isolated from one
another. Independent flow elements may have an
enormous impact on an operators ability to drain
the reservoir. As an analogy, consider a single
reservoir compartment to be a sponge. Like a
sponge with its open-cell structure, the entire
contents can be drained by a single hole or well.
Carrying the analogy further, independent
reservoir compartments are similar to a spool of
plastic bubble wrapa closed-cell system
through which the contents of one bubble cannot
flow to another. If a single hole is punched
through the spool, drainage occurs only from
cells that are penetrated. The bubble-wrap spool
is thus highly compartmentalized.
An additional consideration is the existence
of large compositional variations of hydrocarbons
vertically and laterally within a compartment.
Compositional grading is often caused by gravity,
or nonequilibrium forces of biodegradation,
temperature gradients, current charging, charge
history or incompetent sealing shales.
7
The
magnitude of compositional variation can vary
greatly, depending on the geological and
geochemical history of the reservoir (previous
page, bottom).
8
A technically robust fluid-sampling program
is vitally important when reservoir compart men -
talization, compositional grading or both exist.
Reservoir-formation properties influence the
ability to collect representative fluid samples.
Sample collection requires fluid flow into the
borehole, which occurs only when the wellbore
flowing pressure is lower than formation
pressure. However, if the flowing pressure falls
below the fluid saturation pressure, a gas phase
(in the case of volatile or black oil) or a liquid
phase (in the case of retrograde gas) will form
(right). The relative mobility of each fluid phase
is different; because of unequal flow, the
composition of the fluid exiting the formation
will not be the same as that in the reservoir. This
effect can be minimized or eliminated by
sampling at flow rates and pressure differentials
that are as low as feasible.
Finally, accurate reservoir-temperature
measure ments are vital. Errors of just a few
degrees during PVT testing could result in
misinter pretation. For example, what is conden -
sate in the formation may behave like a volatile
oil at an incorrect temperature in the laboratory.
This error could result in costly production-
design errors.
>
Common deposits that form in tubulars during hydrocarbon production.
Wax and hydrate deposition mainly result from a temperature decrease,
while asphaltene precipitation may be triggered by changes in pressure,
temperature and composition. Inorganic scales arise from changes in
pressure, temperature and composition of aqueous fluids that accompany
hydrocarbon production. (With kind permission of Springer Science and
Business Media.)
Asphaltene Wax
Inorganic scale Hydrate
>
Pressure-decline effects during reservoir-fluid sampling. If the reservoir
pressure falls below the bubblepoint while sampling oil, gas separation will
occur, creating a two-phase system (top). Similarly, if the reservoir contains
retrograde gas, liquid will form if the reservoir pressure falls below the
dewpoint. When phase changes occur in the reservoir, the high-mobility phase
flows preferentially because of relative permeability effects, and the sample is
nonrepresentative. Maintaining the reservoir pressure above the bubblepoint or
dewpoint during sampling preserves single-phase behavior and ensures
collection of a representative sample (bottom).
P
r
e
s
s
u
r
e
Reservoir pressure
Original
reservoir
fluid
Original
reservoir
fluid
Bubblepoint
Wellbore flowing pressure
Free gas
One-phase
representative
sample fluid
P
r
e
s
s
u
r
e
60336schD7R1.qxp:Format Layout V7.2 12/4/07 7:11 PM Page 59
There are two principal hydrocarbon-sample
acquisition methodsbottomhole and surface
sampling. Bottomhole sampling involves convey -
ing a sampling tool on a drillstem-test (DST)
string, wireline or slickline to the producing zone
or zones. In an open hole, sampling can be
performed by the MDT Modular Formation
Dynamics Tester and the Quicksilver Probe tool for
focused extraction of pure reservoir fluid. Cased-
hole sampling devices include the CHDT Cased
Hole Dynamics Tester, single-phase reservoir
sampler (SRS) and SCAR sampling tool.
9
Surface
sampling, which is most frequently performed at
the separator under stable flow conditions,
involves collecting gas and liquid samples.
Engineers may acquire surface samples during
exploration if downhole methods are unavail able,
and may continue to do so throughout a wells
lifetime to monitor fluid-property evolution.
10
Bottomhole fluid samples must be extracted
from locations that will yield the most relevant
information for making decisions. To aid in this
endeavor, todays sampling and testing tools
include an array of instruments that can perform
downhole fluid analysis (DFA). DFA tools provide
real-time fluid-property measurements at
reservoir conditions, allowing engineers to
analyze samples before they are gathered.
60 Oileld Review
9. For more on formation-fluid sampling devices:
MDT tool: Colley N, Ireland T, Reignier P, Richardson S
and Joseph J: The MDT Tool: A Wireline Testing
Breakthrough, Oilfield Review 4, no. 2 (April 1992): 5865.
Andrews RJ, Beck G, Castelijns K, Chen A, Cribbs ME,
Fadnes FH, Irvine-Fortescue J, Williams S, Hashem M,
Jamaluddin A, Kurkjian A, Sass B, Mullins OC, Rylander
E and Van Dusen A: Quantifying Contamination Using
Color of Crude and Condensate, Oilfield Review 13, no.
3 (Autumn 2001): 2443.
Quicksilver Probe: Akkurt, R, Bowcock M, Davies J,
Del Campo C, Hill B, Joshi S, Kundu D, Kumar S,
OKeefe M, Samir M, Tarvin J, Weinheber P, Williams S
and Zeybek M: Focusing on Downhole Fluid Sampling
and Analysis, Oilfield Review 18, no. 4 (Winter
2006/2007): 419.
CHDT tool: Burgess K, Fields T, Harrigan E, Golich GM,
Reeves R, Smith S, Thornsberry K, Ritchie B, Rivero R
and Siegfried R.: Formation Testing and Sampling
Through Casing, Oilfield Review 14, no. 1 (Spring 2002):
4657.
SRS and SCAR tools: Aghar, H, Carie M, Elshahawi H,
Gomez JR, Saeedi J, Young C, Pinguet B, Swainson K,
>
Schematic diagram of a MDT Modular
Formation Dynamics Tester, using the Quicksilver
Probe tool for focused extraction of reservoir
fluid. The focused sampling probe is set against
the borehole wall to withdraw formation fluids for
Fluid Profiling characterization and sample
collection. The downhole LFA Live Fluid Analyzers
provide real-time quantitative measurements of
density, viscosity, GOR, hydrocarbon composition
and formation-water pH.
Sample
flow
Power cartridge
Sample-bottle
module
Pump module
(sample)
LFA fluid analyzer
(sample)
Hydraulic module
Focused sampling
probe
LFA fluid analyzer
(guard)
Pump module
(guard)
Guard
flow
>
Reservoir section (left) and real-time DFA GOR measurements (right) of
reservoir fluids in a deepwater exploration well. There is good agreement
between the GOR values and the reservoir structure. The FCA analysis
suggested locations for fluid sampling (blue dots, right). The reservoir fluids
vary significantly from dry gas (Fluids A and B) and condensate gases (Fluid C)
at the top, to black oils with different GORs (Fluids D through J) at the bottom.
At the bottom of the oil column (Fluids H, I and J), GOR variations indicate a
gentle fluid-composition gradient. On the other hand, a GOR inversion is
evident between Fluids E and F; Fluid F is deeper than Fluid E, but has a higher
GOR. A similar inversion occurs between Fluids G and J, suggesting a complex
reservoir structure with a potential flow barrier at sampling station J.
S
e
a
l
S
e
a
l
Reservoir section
Water Rich gas and condensate
Oil
0
B
C
D
E
F
G
J H
I
500
100 1,000
Depth, ft
10,000 100,000
DFA GOR, ft
3
/bbl
1,000
1,500
2,000
2,500
3,000
3,500
Dry gas
A
Takla E and Theuveny B: The Expanding Scope of Well
Testing, Oilfield Review 19, no. 1 (Spring 2007): 4459.
10. For more on surface sampling: Aghar et al, reference 9.
11. For more on optical DFA methods:
Betancourt S, Fujisawa G, Mullins OC, Carnegie A,
Dong C, Kurkjian A, Eriksen KO, Haggag M, Jaramillo AR
and Terabayashi H: Analyzing Hydrocarbons in the
Borehole, Oilfield Review 15, no. 3 (Autumn 2003): 5461.
Crombie A, Halford F, Hashem M, McNeil R, Thomas EC,
Melbourne G and Mullins OC: Innovations in Wireline
Fluid Sampling, Oilfield Review 10, no. 3 (Autumn 1998):
2641.
Dong C, Hegeman PS, Carnegie A and Elshahawi H:
Downhole Measurement of Methane Content and GOR
in Formation Fluid Samples, SPE Reservoir Evaluation &
Engineering 9, no. 1 (February 2006): 714.
12. Betancourt SS, Fujisawa G, Mullins OC, Eriksen KO,
Dong C, Pop J and Carnegie A: Exploration Applications
of Downhole Measurement of Crude Oil Composition and
Fluorescence, paper SPE 87011, presented at the SPE
Asia Pacific Technical Conference on Integrated
Modeling for Asset Management, Kuala Lumpur,
March 2930, 2004.
13. Dong CM, OKeefe M, Elshahawi H, Hashem M,
Williams S, Stensland D, Hegeman P, Vasques R,
Terabayashi T, Mullins O and Donzier E: New Downhole
Fluid Analyzer Tool for Improved Reservoir
Characterization, paper SPE 108566, presented at the
SPE Offshore Europe Oil and Gas Conference and
Exhibition, Aberdeen, September 47, 2007.
14. Venkataramanan L, Weinheber P, Mullins OC,
Andrews AB and Gustavson G: Pressure Gradients
and Fluid Analysis as an Aid to Determining Reservoir
Compartmentalization, Transactions of the 47th SPWLA
Annual Logging Symposium, Vera Cruz, Mexico,
June 47, 2006, paper S.
15. Dong C, Elshahawi H, Mullins OC, Venkataramanan L,
Hows M, McKinney D, Flannery M and Hashem M:
Improved Interpretation of Reservoir Architecture and
Fluid Contacts through the Integration of Downhole Fluid
Analysis with Geochemical and Mud Gas Analyses,
paper SPE 109683, presented at the SPE Asia Pacific
Oil and Gas Conference and Exhibition, Jakarta,
October 30November 1, 2007.
60527schD7R1.qxp:60527schD7R1 1/8/08 9:21 AM Page 60
Autumn 2007 61
DFA methods include in-situ optical
absorption spectroscopy, optical reflectance,
fluorescence and some nonoptical measure ments
including density, viscosity and pH. The
spectrometer operates in the visible to near-
infrared rangeat wavelengths between 400 and
2,100 nm. Spectra are recorded in real time,
revealing the proportions of methane [C
1
], ethane
to pentane [C
25
], hexane-plus [C
6+
] and CO
2
fractions, as well as a gas/oil ratio (GOR) estimate.
Additionally, differences between the reservoir-
fluid and drilling-fluid spectra indicate the level of
sample contamination.
11
Downhole fluorescence
measurements provide fluid-phase information
that is especially useful for retro grade
condensates and volatile oils.
12
Fluorescence is
also sensitive to liquid formation in a condensate
gas when the flowing pressure falls below the
dewpoint, allowing sampling engineers to monitor
fluid-phase separation in real time, and ensure
that representative single-phase samples are
collected (previous page, top left).
13
Fluid Profiling reservoir-fluid characteri -
zation by DFA can diagnose compositional
grading and help identify reservoir compart -
ments. For example, abrupt fluid-composition or
GOR changes between zones in a single well or
between neighboring wells may indicate compart -
mentalization. To confirm that perceived
fluid-property differences are truly significant,
engineers must first consider measurement
uncertainties. A recent method to evaluate
uncertainties is the fluid comparison algorithm
(FCA).
14
FCA uses parametric models to estimate
GOR and coloration uncertainties as a function
of optical-density (

) and mud-contamination
(

) measurement variability. The algorithm


compares measurements acquired from two
fluids and calculates the probability that differ -
ences are statistically significant. When the FCA
result indicates that the fluids are different,
sample acquisition for detailed surface analysis
is justified. The following case study demon -
strates how engineers employ DFA and FCA
to characterize a reservoir and determine
sampling locations.
Sampling and Reservoir Characterization in a
Deepwater Accumulation
In a deepwater exploration well, Shell employed
real-time DFA and FCA to characterize the
reservoir architecture and decide where to
collect fluid samples.
15
As the MDT assembly
traveled down the well, the spectrometers
measured the GOR at several locations (previous
page, top right). Based on FCA analysis, fluid
samples were collected at 10 different depths.
The analysis revealed wide variations in
reservoir-fluid composition, ranging from dry gas
and condensate gases at the top to black oils with
different GORs at the bottom. At the bottom of
the oil column, the GOR varied gradually with
depth in the bottom sand, indicating a fluid-
composition gradient. GOR inversions were also
detected between the top and bottom sands,
suggesting the presence of flow barriers and a
complex reservoir structure.
DFA and FCA analysis showed that fluids
above and below the inversion had a greater than
99% probability of existing in different
compartments. A pressure discontinuity between
the compartments confirmed the absence of
hydraulic communication.
Shell and Schlumberger fluid specialists
explored the oil column in detail, comparing the
GOR information with petrophysical, formation-
pressure and mud-gas logs, and performing FCA
analysis (below). The gamma ray and pressure
>
Expanded log presentation and FCA analysis of an oil column at the bottom of a deepwater
reservoir. Correlation of the GOR log (top left) with the gamma ray log (green, top center) reveals that
Fluids F and G come from one sand zone, and Fluids J, H and I reside in another. In the upper sand,
Fluids F and G have the same GOR, while in the lower sand, Fluids J, H and I display a GOR decrease
with depth. The fluid densities derived from pressure gradients (blue) reveal two principal featuresa
fluid-density inversion between Fluids G and J, and a gradual density increase from Fluid J to Fluid I
suggesting no vertical communication between the top and bottom sands. The mud-gas log (top right)
offers further supporting evidence. At the depth of Fluid J, the
13
C value (red) falls abruptly and
increases gradually with depth. The relative methane concentration (blue) also fluctuates sharply at
Fluid J, then decreases gradually with deptha feature consistent with the fluid-density increase.
Final confirmation resulted from applying the FCA technique (bottom). The FCA algorithm generates a
number representing the probability that two fluids are statistically different. As measurement
uncertainties () decrease, confidence in data quality increases. Therefore, low values indicate a
high probability that perceived fluid-property differences are real. The contour plot indicates that
Fluid G has a 99% probability of being different from Fluid J, and a 95% probability of being different
from Fluid H. Discovery of reservoir compartmentalization led Shell to reevaluate the field and make
significant strategic adjustments in reservoir development.
F
G
H
I
J
D
e
p
t
h
,

f
t
2,750
2,800
2,850
2,900
2,950
3,000
3,050
3,100
3,150
5,500 Pressure, psi 6,000
0 Gamma Ray, API 150
Relative Methane Concentration
Higher concentration of
C1 corresponds to lower
density of Fluid J
Increasing
contribution
of thermogenic
charge
0 2 4
55
13
C methane, parts per thousand 50
900 GOR, ft
3
/bbl 1,500
Top sand
Bottom sand
J
H
I
G
F
O
i
l
-
b
a
s
e

m
u
d

c
o
n
t
a
m
i
n
a
t
i
o
n
,

Optical density,

0.10
0.09
0.07
0.08
0.06
0.05
0.04
0.03
0.02
0.01 0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10
Fluids G and H
P = 0.80
P = 0.99
P = 0.95
P = 0.90
Fluids G and J
60336schD7R1.qxp:Format Layout V7.2 12/4/07 7:11 PM Page 61
logs showed that the top portion of the oil resides
in one sand lobe with a relatively constant GOR
and pressure gradient. The rest of the oil is in a
lower sand lobe, where the GOR decreases with
depth. The pressure log was consistent, indi -
cating a significant fluid-density difference
between the upper and lower lobes.
Recent advances in mud-gas logging have
given engineers another tool to perform real-
time detection of seals and permeability barriers,
lithological variations and fluid contacts.
16
Gases
gathered at the surface during drilling or
separated from fluid samples can be analyzed for
isotopic content. The isotopic signature,
13
C, is
the value of the
13
C/
12
C methane-isotope ratio in
a sample relative to a standard, expressed in
parts per thousand. When plotted alongside a
standard mud-gas log, larger
13
C values may
indicate higher concentrations of biogenic gas in
the reservoir. Trending
13
C values may indicate
nonequliibrium methane distributions, and a
clear break in the methane-isotope signature in
mud gas may imply the presence of a seal.
17
Engineers noted a
13
C discontinuity at around
2,950 ft [899 m], suggesting yet another flow
barrier. FCA analysis provided confirmation,
calculating a 95 to 99% probability that fluids
above and below the discontinuity were
dissimilar and from different compartments with
no communication.
The discovery of the fluid compartments in
this deepwater field led Shell engineers to adjust
their reservoir models and development plan -
ning. Reservoir compartmentalization increases
design complexity and cost because engineers
must treat each zone independently. Production
forecasts, reserve calculations and enhanced
recovery schemes became proportionately
more complex.
Improved regional understanding of
subsurface architecture impacted short-term
decisions on sidetrack objectives. From the
location of the borehole, Shell also reasoned that
compositional grading in the lower reservoir
likely extended downward from the penetrated
zone; as a result, production-facility plans were
altered to anticipate a GOR reduction with time.
Access to real-time fluid analyses allowed
Shell to make decisions much earlier in the field-
development process, and accelerated the
project by at least six months. At todays oil
prices, the time saved was worth hundreds of
millions of dollars.
Laboratory Fluid Preparation and
Sample Chain of Custody
The preceding case study demonstrates the
considerable effort and care that engineers
apply during the sampling process. Nevertheless,
the harsh downhole environment and the
nature of well operations may render DFA
equipment, sampling devices and subsequent
analysis suscep tible to fouling, failure and
other inaccuracies.
Schlumberger engineers addressed this
problem by implementing a chain-of-custody
procedure, a concept borrowed from forensic
science.
18
Evidence must make the journey from
a crime scene to the courtroom in a validated
and secure manner; otherwise, it may not be
admissible in court. Similarly, chemists at a
remote testing laboratory should be able to
determine whether the chemical composition of
a field sample has been preserved. DFA provides
a convenient way to establish a chain of custody
for fluid samples, because chemists have the
opportunity to compare analytical data acquired
downhole with those from the corresponding
samples that reach the laboratory.
62 Oileld Review
>
Chain-of-custody investigation of a valid crude-oil sample. Chemists
compare visible near-IR spectra from DFA and the laboratory sample. Both
spectra were measured at downhole conditions15,000 psi [103 MPa] and
250F [121C]. The spectral analysis (top) compares the downhole discrete
data (red circles) with the continuous spectrum from the laboratory sample
(blue). Data regression (bottom) of the LFA optical densities versus those
from the laboratory shows excellent agreement, evidenced by the near-
perfect overlay of the regression line (red) over the X = Y line (blue). The fluid
sample is well-preserved and suitable for further laboratory studies.
0.9
1.0
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
-0.1
1,000 1,100 1,200 1,300 1,400 1,500
Wavelength, nm
Laboratory optical density
1,600 1,700 1,800 1,900 2,000
O
p
t
i
c
a
l

d
e
n
s
i
t
y
L
F
A

o
p
t
i
c
a
l

d
e
n
s
i
t
y
1.5
1.0
0.5
0
0 0.5 1.0 1.5
Laboratory sample
LFA tool
X = Y
Data regression
1,600
1,920 1,445
1,671
1,290
1,725
1,070
60527schD7R1.qxp:60527schD7R1 1/8/08 9:22 AM Page 62
Autumn 2007 63
In the field, after tool retrieval, engineers
remove the reservoir-fluid samples. At this point,
a PVT Express onsite well fluid analysis system
may be available to conduct preliminary
measurements that determine whether the
collected-sample properties agree with those
measured by DFA.
19
If onsite analysis is
unavailable or more sophisticated testing is
required, the samples are shipped to a remote
testing laboratory in the original sample bottles
or transferred to an approved shipping container.
When fluid samples arrive at the Schlumberger
laboratory, chemists restore the fluid inside the
sample container to the original reservoir
temperature and pressure, and allow the fluid to
equilibrate by agitating it continuously for up to
five days. The restoration process is intended to
redissolve precipitated asphaltene and wax
particles, ensure a homogeneous fluid through -
out the sample cylinder and provide a
single-phase representative fluid for testing.
The visible near-infrared (IR) spectrum is an
effective hydrocarbon fingerprint.
20
The DFA
spectrometer performs sample analysis as the
crude-oil sample is acquired, providing a direct
fluid-property measurement under downhole
conditions. In the laboratory, chemists perform
the same measurement with a research-grade
spectrometer at the downhole temperature and
pressure. Differences between the DFA and
laboratory spectra may indicate that the
laboratory sample has been compromised. For
example, if the methane concentration is lower
in the laboratory spectrum, then sample-bottle
leakage or a fluid-transfer error may have
occurred during sampling or transport to the
laboratory. The examples presented below
illustrate the chain-of-custody technique.
The first example involves a fluid sample
acquired from an offshore oil field. There is
excellent agreement between the downhole
and laboratory spectra (previous page). This
indicates that the sample is well-preserved and
suitable for further laboratory studies.
The second example involves another sample
from an offshore oil field (right). The downhole
and laboratory spectra have good overall
agreement, but there are subtle differences in
the wavelength region above about 1,600 nm.
Chemists investigated the spectral differences
further by employing algorithms to calculate
GORs and probabilities of similarity between the
two samples.
21
The calculations showed that the
GOR from the downhole spectrum, 580 ft
3
/bbl
[103.3 m
3
/m
3
], was significantly higher than that
from the laboratory spectrum, 320 ft
3
/bbl
[57.0 m
3
/m
3
]. This difference corresponded to a
16. Venkataramanan L, Elshahawi H, McKinney D,
Flannery M, Hashem M and Mullins OC: Downhole
Fluid Analysis and Fluid Composition Algorithm as an
Aid to Reservoir Characterization, paper SPE 100937,
presented at the SPE Asia Pacific Oil and Gas
Conference and Exhibition, Adelaide, Australia,
September 1113, 2006.
17. Berkman T, Ellis L and Grass D: Integration of Mud Gas
Isotope Data with Field Appraisal at Horn Mountain
Field, Deepwater Gulf of Mexico, AAPG Bulletin 86,
no. 13 (2002): supplement.
18. Betancourt SS, Bracey J, Gustavson G, Mathews SG and
Mullins O: Chain of Custody for Samples of Live Crude
Oil Using Visible Near-Infrared Spectroscopy, Applied
Spectroscopy 60, no. 12 (December 2006): 14821487.
19. Aghar et al, reference 9.
20. For more on visible and near-IR spectroscopy:
Crombie et al, reference 11.
21. Mullins OC, Beck G, Cribbs MY, Terabayashi T and
Kagasawa K: Downhole Determination of GOR on
Single-phase Fluids by Optical Spectroscopy,
Transactions of the 42nd SPWLA Annual Logging
Symposium, Houston, June 1720, 2001, paper M.
Venkataramanan L, Fujisawa G, Mullins OC, Vasques RR
and Valero H-P: Uncertainty Analysis of Near-Infrared
Data of Hydrocarbons, Applied Spectroscopy 60, no. 6
(June 2006): 653662.
>
Chain-of-custody investigation of a compromised crude-oil sample. Both
spectra were measured at downhole conditions20,000 psi [138 MPa] and 200F
[93C]. The visible near-IR spectral analysis (top) shows subtle OD differences in
the region above about 1,600 nm. At 1,671 nm, indicating methane, the
laboratory-sample OD is lower than that measured by the LFA tool. At 1,725 nm,
the methylene oil peak, the laboratory sample OD is higher than that measured
by the LFA tool. These differences are also visible in the linear regression plot
(bottom). The ratio between the methane and oil peaks can be used to calculate
the GOR. In this case, the algorithms indicated that the laboratory-sample GOR
was significantly lower than that of the field sample. Therefore, further analysis
would be necessary before this sample could be trusted.
1,000 1,200
0 0.5 1.0 1.5
1,400 1,600 1,800 2,000
Wavelength, nm
Laboratory optical density
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
1.5
0.5
0
1.0
0
O
p
t
i
c
a
l

d
e
n
s
i
t
y
L
F
A

o
p
t
i
c
a
l

d
e
n
s
i
t
y
X = Y
Data regression
Laboratory sample
LFA tool
1,600
1,671
1,820
1,280
1,445
1,725
1,070
60336schD7R1.qxp:60527schD7R1 1/14/08 5:06 PM Page 63
93% likelihood that either one (or both) of the
spectra were in error, or the laboratory sample
was compromised. Further investigation of
surface procedures and DFA-tool performance
would be required before this sample could be
used with confidence in the laboratory.
Laboratory Techniques for Flow Assurance
In the laboratory, chemists determine fluid
compositions and measure fluid properties
related to flow assurance. For compositional
analysis, an accurately measured volume of fluid
is isobarically and isothermally transferred to a
pycnometer to measure mass and density. Then,
the pycnometer is connected to an apparatus in
which the oil sample is cooled to ambient
temperature and decompressed. The volume of
gas liberated by this procedure allows
calculation of the GOR.
Chemists employ gas chromatography to
determine the vapor-phase composition up to
C
15+
and the liquid-phase composition up to C
36+
.
The bulk crude-oil composition is calculated by
summing the individual contributions from each
phase (above left). This procedure ensures that a
consistent reservoir-fluid composition is avail able
for subsequent fluid-property characterization
and reservoir-production simulation.
The bubblepoint pressure is determined by
performing a constant composition expansion
(CCE) test. Technicians place a known volume of
equilibrated fluid in a PVT cell at reservoir
temperature and pressure (left). The fluid is
initially single phase, and testing begins by
reducing the pressure isothermally and
monitoring the fluid-volume change. Eventually,
the fluid separates into two phases. Pressure
reductions continue in increments ranging from
100 to 500 psi [0.69 to 3.45 MPa], the vapor and
liquid phases are equilibrated at each step, and
phase volumes are measured and plotted as a
function of pressure. For a black oil, the
intersection of the single- and two-phase lines in
the PV plot defines the bubblepoint.
The flow-assurance testing protocol depends
on the nature of the crude oil. For example, when
asphaltenes are of concern, then saturate,
aromatic, resin and asphaltene (SARA) analysis
and paraffinic solvent titration with dead oil are
principal screening techniques.
22
It is also
common to measure the asphaltene-precipitation
pressure on a live-oil sample. If testing identifies
an asphaltene-precipitation problem, additional
studies are conducted to map out the asphaltene
phase diagram and evaluate the effectiveness of
chemicals or coatings as prevention strategies.
23
64 Oileld Review
>
Typical crude-oil compositional analysis, determined by gas chromatography. The plot shows the
hydrocarbon contribution from C
1
to C
30+
, as well as the CO
2
concentration.
100.00
10.00
1.00
Components
C
O
2
H
2
S
N
2
C
1
C
2
C
3
0.10
0.01
I
-
C
4
N
-
C
4
I
-
C
5
N
-
C
5
C
6
M
C
Y
C
-
C
5
B
e
n
z
e
n
e
C
Y
C
L
-
C
6
C
7
M
C
Y
C
L
-
C
6
T
o
l
u
e
n
e
C
8
C
2
-
B
e
n
z
e
n
e
M
&
P
-
X
y
l
e
n
e
O
-
X
y
l
e
n
e
C
9
C
1
0
C
1
1
C
1
2
C
1
3
C
1
4
C
1
5
C
1
6
C
1
7
C
1
8
C
1
9
C
2
0
C
2
1
C
2
2
C
2
3
C
2
4
C
2
5
C
2
6
C
2
7
C
2
8
C
2
9
C
3
0
+
W
e
i
g
h
t

p
e
r
c
e
n
t
>
Pressure-volume-temperature (PVT) cell and bubblepoint determination. The fully visual PVT cell
allows direct confirmation of bubblepoints at various temperatures and pressures (top). A magnetic
mixer provides vigorous agitation to maintain phase equilibrium. A video-based cathetometer
measures fluid levels in the cell for phase-volume calculations. The oven heats the PVT cell to the
test temperature, and technicians monitor and plot phase volumes as a function of pressure. The
inflection point in the curves as pressure decreases defines the bubblepoint. In the case depicted
here, the bubblepoint is approximately 5,000 psi [34.5 MPa] (bottom).
High-
pressure
pumps
O
i
l
S
o
l
v
e
n
t
Side view
Magnetic mixer
Cathetometer
16,030
30
2,030
4,030
6,030
10,030
12,030
14,030
8,030
P
r
e
s
s
u
r
e
,

p
s
i
Volume, cm
3
25 30 35 40 45 50 55 60 65 70 75
Tres = 176F
T = 120F
T = 75F
60336schD7R1.qxp:Format Layout V7.2 12/4/07 7:11 PM Page 64
Autumn 2007 65
Waxy crude oils pose different production and
transportation challenges. Wax deposition inside
tubulars and pipelines reduces the effective flow
area, increasing the pressure drop and potentially
causing complete blockage. Therefore, it is
important to fully understand the oils behavior
throughout the pressure and temperature path
from the formation to the production facilities.
Temperature is the dominant parameter affecting
waxy crude-oil viscosity, gel strength, pour point,
wax crystallization and deposition. Although dead
oils such as stock-tank oil (STO) can be used to
generate preliminary data, it is important to
include live oils in the testing program because
pressure and dissolved gases may strongly
influence wax solubility.
The first characterization step is to measure
the amount of wax that can precipitate and
deposit on a solid surface. Live-oil filtration and
high-temperature gas chromatography (HTGC)
are common methods to measure the wax
content. HTGC is more valuable because it
provides the n-paraffin composition at high
carbon numbers (from C
60
to C
100
)information
chemists enter into thermodynamic models to
predict wax behavior.
The wax-appearance temperature (WAT) is
one of the most important flow-assurance
measure ments, indicating the temperature at
which wax crystals begin to form in a crude-oil
sample. This measurement provides a preliminary
assessment of the likelihood of wax-related
deposition problems. Laboratory workers place
dead oil on the stage of a cross-polar microscope
(CPM) and block light transmission by adjusting
polarized prisms at opposite ends of the sample.
When illuminated by polarized light, crystalline
materials disturb the polarization plane;
therefore, as the fluid sample cools, wax-crystal
formation is clearly visible as bright spots appear
against the black background (above right). Some
laboratories have high-pressure CPM instruments
that can measure the WAT in live oils. The sample
cell operates at pressures up to 20,000 psi
[138 MPa] and temperatures up to 392F [200C].
Formation and growth of wax crystals may
affect the rheological behavior of the crude oil.
Above the WAT, most hydrocarbon systems
behave as Newtonian fluids; however, non-
Newtonian behavior such as shear thinning may
commence as the fluids cool and phase changes
occur. This behavior must be properly quantified
to allow engineers to design a suitable
production system.
24
Rheological properties of
interest in the context of waxy crude are pour
point, apparent viscosity and gel strength.
The pour point is the temperature below
which a fluid is no longer pourable because of
viscosification, gelation or solids formation.
Pour-point testing with dead oils conforms to
ASTM Standard D97.
25
For live oils, engineers use
an apparatus containing a visual sapphire cell
mounted on an automatic, vibration-free pivoting
bracket that resides in a programmable
convection oven. The oil sample is heated to the
reservoir temperature, and slowly cooled until
fluid movement in the cell ceases. Most waxy
crude oils begin to display non-Newtonian
behavior near the pour point. If the pour point
exceeds anticipated oil-production temperatures,
engineers may add pour-point-depressant
chemicals to maintain flow.
The apparent viscosity of waxy crude oils
generally increases dramatically as the
temperature and shear rate decrease,
particularly at temperatures near the pour point.
Schlumberger engineers measure viscosity with a
rheometer that can operate at 6,000 psi
[41.4 MPa] and 302F [150C], allowing work
with live oils.
Waxy crude oils tend to form gels at
temperatures below the pour point. In the event
of a production shutdown, high pumping pres -
sures may be required to break the gel and
restore flow. Therefore, yield-stress data from
waxy crude oils are necessary to properly design
flowlines and avoid production problems. The
22. Live oil contains the dissolved gases present in the
reservoir. If the gases are released, or flashed, at the
surface or in the laboratory, the residual liquid is called
dead oil. Stock-tank oil (STO) is also a dead oil.
23. For a detailed discussion of asphaltene flow assurance:
Akbarzadeh et al, reference 5.
24. At constant temperature and pressure, Newtonian fluids
display constant viscosity at all shear rates. The
viscosity of non-Newtonian fluids is not constant at all
shear rates. Apparent viscosity is the viscosity of a fluid
at a given shear rate and temperature.
25. ASTM D97-06 Standard Test Method for Pour Point of
Petroleum Products. West Conshohocken, Pennsylvania,
USA: ASTM International, 2006.
>
Determination of wax-appearance temperature (WAT) by cross-polar microscopy (CPM). The
microscope is equipped with a heated stage between two polarized prisms (top). At the beginning of
the test, the prisms are adjusted to block light transmission. As the stage cools, wax-crystal formation
changes the light polarity, and the video camera sees the appearance of bright spots. In this example,
no light is transmitted at 111F [44C] (bottom, right panel) because the fluid temperature is above the
WAT. Spots begin to appear when the fluid cools to the WAT (center panel), and the image becomes
brighter as the fluid cools to 0C (left panel). This method is accurate to 2F [1.1C].
0C 44C WAT = 42C
1
0
5
0
Temperature 48.6
IR filter
Polarizer
Analyzer
20
360
rotatable
stage
Cooling
gas
Hot stage
top view
Charged
coupled
device
Hot stage
60336schD7R1.qxp:Format Layout V7.2 12/4/07 7:11 PM Page 65
yield stress of live fluids can be determined by a
model pipeline test (MPT) (left).
26
Another important measurement in the
context of production-system design and flow
assurance is the wax-deposition rate. The
principal controlling parameters are fluid
temperature, heat loss through the pipeline wall,
wax content, shear rate and fluid viscosity. The
wax-deposition rate can be determined under
simulated pipe-flow conditions in a wax-
deposition flow loop (WDFL) (below left).
The following case study illustrates how
operators use laboratory measurements to develop
operational strategies that prevent, mitigate or
remediate wax deposition and gel formation.
Optimizing Subsea System Design
in West Africa
A West African field, located at a water depth less
than 1,000 ft [300 m], has a reservoir
temperature and pressure of 170F [76.7C] and
3,180 psi [21.9 MPa]. The seabed temperature is
55F [12.8C]. The operator planned to produce
oil through a looped 6-in. ID subsea tieback to a
host facility 2 miles [3.2 km] from the reservoir.
With the large temperature difference between
reservoir and seabed, and the long distance that
fluids would flow along the cold seafloor, the
operator needed assurance that solids deposition
would not impede flow. Schlumberger collected
downhole fluid samples and sent them to the
Oilphase-DBR fluid sampling and analysis
laboratory in Edmonton, Alberta, Canada for a
flow-assurance study.
27
Compositional analysis of the reservoir fluid
revealed a black oil with a GOR of 230 ft
3
/bbl
[41.0 m
3
/m
3
] and an API gravity of 36.3.
Compositional analysis indicated that the
fraction with a carbon number above C
30
was
35.8%. Further characterization of the C
30+
fraction and n-paraffin distribution revealed that
the crude oil contained about 13.1 wt% C
17+
n-paraffins.
28
The relatively high n-paraffin
concentration was cause for concern that the
66 Oileld Review
26. The yield stress,
y
, is calculated by the following force-
balance equation:
y
= where P
y
is the hydraulic
pressure necessary to cause fluid movement, D is the
inner diameter of the coil and L is the coil length.
27. Alboudwarej H, Huo Z and Kempton E: Flow-Assurance
Aspects of Subsea Systems Design for Production of
Waxy Crude Oils, paper SPE 103242, presented at the
SPE Annual Technical Conference and Exhibition,
San Antonio, Texas, September 2426, 2006.
28. Paraffin is a common name for a group of alkane
hydrocarbons with the general formula C
n
H
2n+2
, where
n is the number of carbon atoms. The simplest paraffin
molecule is methane, CH
4
, a gas at room temperature.
Octane, C
8
H
18
, is liquid at room temperature. The solid
forms of paraffin are heavier molecules from C
20
to C
40
.
Linear members of the series (those with no branches or
cyclic structures) are called n-paraffins.
>
Model pipeline test (MPT) apparatus. Stock-tank oil circulates through a
coil of tubing immersed in a temperature-controlled bath. Flow through the
coil stops when the bath reaches the test temperature, allowing the fluid to
age and form a gel structure. After the aging period, laboratory workers
measure the nitrogen pressure necessary to initiate flow in the coil and
calculate the gel strength from a simple force-balance equation.
N
2
pressure to break gel
System charging pump
C
o
n
v
e
c
t
i
o
n

o
v
e
n
High-pressure
circulation pump
Fluid
sample
cylinder
Backpressure
regulator
Heated lines
Temperature-
controlled
bath
>
Wax deposition flow loop (WDFL). The WDFL is a miniature flow loop that exposes stock-tank oils
(STO) to a range of heat fluxes and shear rates that would be expected in the flowline during actual
production. The deposition section of the flow loop is a 39-in. [1-m], 0.375-in. OD stainless steel tube.
A 0.53-galUS [2.0-L] oil reservoir maintains the oil above the WAT throughout the test. The oil exits the
reservoir and flows at a controllable rate through 50 ft [15 m] of coiled copper tubing in a tempering
bath for temperature adjustment. The deposition loop is immersed in a cooling bath to simulate heat
loss in the pipeline. As the oil flows through the deposition loop, a data-acquisition system monitors
the wall temperature, oil and water temperatures, the pressure drop between the inlet and outlet of the
coil and the flow rate. Wax deposits constrict flow inside the deposition loop, increasing the pressure
required to maintain flow. The increase in pressure allows calculation of the amount of deposited wax.
Thermocouples
Pressure transducer
Reservoir
Remelting bath Tempering bath
Pump
Flowmeter
Thermocouples
Deposition section
Cooling water
Cooling bath
Thermocouples
P
y
D
____
4L
60336schD7R1.qxp:Format Layout V7.2 12/4/07 7:11 PM Page 66
Autumn 2007 67
fluid might exhibit flow-assurance problems
related to wax deposition. Therefore, the
Oilphase-DBR team objective was to generate
measurements that would provide guidance
concerning ways to mitigate and remediate wax
deposition during steady-state and transient
events during crude-oil production.
Constant composition expansion (CCE)
testing measured a bubblepoint pressure of
700 psi [4.8 MPa] at the reservoir temperature.
WAT and pour-point tests were performed with
both live- and dead-oil samples (right). As the
fluid pressure fell below the bubblepoint to
ambient conditions, dissolved gas escaped, the
average sample composition shifted toward
heavier hydrocarbons, and both the WAT and
pour point increased approximately 20F [11.1C].
Rheological testing revealed gel-structure
formation in live crude oil at low shear rates
when the fluid temperature fell below the pour
point and approached the seabed temperature
(below). As the rheometer pressure decreased,
the shear stress required to break the gels
increasedbehavior consistent with the loss of
lighter hydrocarbons. Around the pour point,
>
Wax-appearance temperature (WAT) and pour-point (PP) data measured
with live and dead West African crude oil. The dead-oil measurements are
shown at atmospheric pressure. As the live-oil fluid pressure decreased from
700 psi to atmospheric pressure, both the WAT (red) and pour point (blue)
increased approximately 20F. The increases resulted from the loss of
dissolved gases in the fluid.
160
40
70
100
130
T
e
m
p
e
r
a
t
u
r
e
,

F
Pressure, psi
0 200 400 600 800
WAT
Pour point
1.0E+11
1.0E+10
1.0E+09
1.0E+08
1.0E+07
1.0E+06
1.0E+05
1.0E+04
1.0E+03
1.0E+02
1.0E+01
1.0E+00
V
i
s
c
o
s
i
t
y
,

m
P
a
-
s
Shear stress, Pa
300 psi
10 100
1.0E+11
1.0E+10
1.0E+09
1.0E+08
1.0E+07
1.0E+06
1.0E+05
1.0E+04
1.0E+03
1.0E+02
1.0E+01
1.0E+00
V
i
s
c
o
s
i
t
y
,

m
P
a
-
s
Shear stress, Pa
1.0E+11
1.0E+10
1.0E+09
1.0E+08
1.0E+07
1.0E+06
1.0E+05
1.0E+04
1.0E+03
1.0E+02
1.0E+01
1.0E+00
V
i
s
c
o
s
i
t
y
,

m
P
a
-
s
Shear stress, Pa
100 psi
10 100
700 psi
10 100
Viscosity at 55 F
Viscosity at 65 F
Viscosity at 80 F
>
Rheological behavior of a live West African crude oil. Shear-stress
versus viscosity measurements were performed near and below the pour
point at 100, 300 and 700 psi [0.69, 2.07 and 4.83 MPa]. At 700 psi (top left),
elevated low-shear-rate plateau viscosities at 55 and 65F [12.8 and
18.3C] indicated the presence of gel structures. Little gelation occurred
near the pour point at 80F [26.7C]. Similar behavior occurred at 300 psi
(top right); however, higher shear stresses were required to break the gels
at 55 and 65Fbehavior consistent with the loss of lighter hydrocarbons
in the sample. At 100 psi (bottom left), the 80F sample exhibited strong
shear thinning behavior.
60336schD7R1.qxp:Format Layout V7.2 12/4/07 7:11 PM Page 67
shear-thinning behavior with no gelation
occurred at all pressures. Further investigation
of gel strength involved aging live- and dead-oil
samples for 12 hours at the seabed temperature.
The live-oil gel strength between 100 and 700 psi
varied from 38 to 42 Pa, and the dead-oil gel
strength at ambient pressure was more than
three times greater142 Pa.
Dead-oil wax-deposition tests were
conducted in the WDFL at temperatures
between 81 and 122F [27.2C and 50.0C]
(top). The results revealed a deposition-rate
discontinuity at the higher end of the shear-rate
range. For each shear rate, chemists entered the
measured deposition rates, the n-paraffin
distribution, C
30+
composition and viscosity
profile into a standard equation-of-state model
that calculates an n-paraffin diffusion
coefficient. The model assumes that molecular
diffusion of wax molecules is the principal
driving force governing wax deposition.
29
The next step involved entering the flow-
assurance measurements, wax-deposition data
and diffusion coefficients into the OLGA
simulatora commercial multiphase-flow fluid-
transport modelto assess and predict crude-oil
behavior in various production scenarios.
Schlumberger engineers applied OLGA simula -
tions to two cases: producing at a steady state
and restarting production after a shutdown.
Because the WAT of the West African crude oil
was about 50F [27.7C] higher than the pour
point, wax deposition and gel formation were the
principal flow-assurance risks during steady-
state production. The operators design goal was
to prevent wax deposition at production rates
above 5,000 bbl [795 m
3
] of oil per day. Methods
to prevent or slow wax deposition in a pipeline
include using insulated pipe, installing heaters
and injecting paraffin inhibitors into the crude-
oil stream. OLGA simulations determined that
pipe insulation was the most suitable flow-
assurance method (next page). Having both live-
and dead-oil data proved important. Knowing
only the dead-oil WAT would have led the
operator to believe that expensive pipe-in-pipe
insulation was necessary. Calculations using the
lower live-oil WAT showed that more economical
wet insulation would be sufficient to prevent
wax deposition.
Because the oil cools as it flows through the
pipeline to the collection facility, it was essential
to predict the likelihood of gel formation.
Assuming a production rate of 5,000 bbl/d
through wet insulated pipe, the OLGA simulator
showed that live oil could flow through the
pipeline for about 20 hours before reaching
its pour pointsufficient time to reach the
stock tank.
For flow rates less than 5,000 bbl/d, the OLGA
simulator predicted rates at which wax deposi -
tion would occur in the pipeline. Without
experimental WDFL wax-deposition data,
engineers would have to use the standard model
to estimate the n-paraffin diffusion coefficient,
predict a deposition rate, and schedule wax-
removal operations. For this West African crude
oil, the standard model predicted that remedi -
ation would be necessary every two weeks. With
WDFL data, the simulator predicted a much
lower deposition rate, increasing the time
between remedial jobs to six weeks. Methods to
remove wax deposits include pipeline pigging,
68 Oileld Review
>
Wax-deposition behavior of a dead West African crude oil. Deposition
measurements at two shear rates, 170 and 511 s
1
, revealed unusual behavior.
At the lower shear rate (red), deposition slowed steadily with increasing
temperature. However, when the fluid temperature exceeded about 100F
[37.8C], the deposition rates at the higher shear rate (blue) suddenly surpassed
those at the lower shear rate. Chemists repeated the measurements to verify
the reproducibility of this behavior. The deposition rates were entered into
equation-of-state and fluid-transport models that help engineers predict how
the crude oil will behave in various production scenarios.
40
10
15
20
25
30
35
D
e
p
o
s
i
t
i
o
n

r
a
t
e
,

m
g
/
m
2
s
Temperature, F
70 80 90 100 110 120 130
170 s
-1
511 s
-1
>
OLGA calculation of flowline-restart pressures in 6-in. pipe. If the oil gels in
the flowline during a shutdown, pressure must be applied to overcome the gel
strength and initiate flow. The maximum pressure that can safely be applied is
500 psi. The simulation shows that maintaining an internal-flowline pressure
greater than 100 psi would allow engineers to safely restart the flowline.
2,000
0
250
500
750
1,000
1,250
1,500
1,750
R
e
s
t
a
r
t

p
r
e
s
s
u
r
e
,

p
s
i
Fluid pressure, psi
0 100 200 300 400 500 600 700 800
Maximum restart pressure
60336schD7R1.qxp:Format Layout V7.2 12/4/07 7:11 PM Page 68
Autumn 2007 69
hot-oil circulation and solvent treatments with
coiled tubing.
A vital flow-assurance question concerns the
ability of the system to restart after a production
shutdown. According to the operator, 500 psi was
the highest pressure that could be safely applied
to overcome gel strength and initiate flow
(previous page, bottom). Using live-oil gel-
strength data and fluid-transport parameters,
the OLGA simulator calculated that fluid flow
could be restored as long as the pressure in the
pipeline remained above about 100 psi.
This case study shows that reliance on
dead-oil experimental data and standard wax-
deposition calculations could lead operators to
make unduly conservative decisions when
designing production systems. In this case, flow-
assurance predictions based on live-oil data
allowed the operator to save millions of dollars
in flowline costs and less frequent wax-
removal operations.
Coming Advances in Waxy-Crude
Flow Assurance
Significant work is underway to continue
improving flow-assurance testing and fluid-
property surveillance during a fields productive
life. The West Africa case study demonstrated
the benefits of performing flow-assurance
experi ments with live crude oils. However, wax-
deposition testing in devices like the WDFL has
been confined to dead oils. The WDFL could be
modified to perform high-pressure tests, but
consuming two liters of live oil would be
prohibitively expensive.
Oilphase-DBR scientists overcame this
problem by designing and introducing the
RealView live solids deposition cell.
30
Requiring
only 0.04 galUS [150 mL] of oil, the cell can
operate at pressures up to 15,000 psi [103.4 MPa],
temperatures up to 392F and Reynolds numbers
up to 500,000.
31
Unlike the WDFL, the oil
resides in a cylindrical vessel. A rotating spindle
at the center induces fluid movement. The
device can simulate production conditions of
temperature, pressure, composition, pipe-
surface roughness and both laminar and
turbulent flow. Turbulent-flow testing is useful
29. Hayduk W and Minhas BS: Correlations for Prediction
of Molecular Diffusivities in Liquids, Canadian Journal
of Chemical Engineering 60, no. 2 (April 1982): 295299.
30. Zougari M, Hammami A, Broze G and Fuex N: Live Oils
Novel Organic Solid Deposition and Control Device:
Wax Deposition Validation, paper SPE 93558, presented
at the 14th SPE Middle East Oil and Gas Show and
Conference, Bahrain, March 1215, 2005.
31. In fluid mechanics, the Reynolds number is a
dimensionless ratio of inertial forces to viscous forces.
Turbulent-flow conditions exist when the Reynolds
number exceeds 3,000.
>
OLGA model simulations of West African crude-oil behavior in a flowline
during steady-state production. Engineers used the simulations as a design
tool to choose pipe insulation and avoid wax deposition during production.
The temperatures at which oil would arrive at the collection station are plotted
against production rate (top). Wet insulation (red) is more economical than a
pipe-in-pipe (PIP) enclosure (blue), but it is four times less efficient. Having the
live-oil WAT was fortunate because it showed that wet insulation would be
sufficient at production rates greater than 5,000 bbl/d. The next graph (center)
reveals the cooldown profile as oil travels through the flowline. Using wet
insulation (red), the cooldown time to gel formation would either be 12 hours
to the dead-oil pour point or 20 hours to the live-oil pour point. The latter time
period is sufficient to displace oil from the wellhead to the collection station.
The OLGA simulator also predicted two-week wax-deposition profiles along
the flowline (bottom). The wax-deposit thickness predicted by the standard
model (blue) is nearly three times higher than the value calculated from
laboratory wax-deposition tests in the WDFL (red). As a result, the frequency
of wax-removal treatments could be reduced significantly.
160
140
120
80
60
40
100
160
150
140
120
110
100
130
T
e
m
p
e
r
a
t
u
r
e
,

F
Time, h
0 5 10 15 20 25
A
r
r
i
v
a
l

t
e
m
p
e
r
a
t
u
r
e
,

F
Oil production rate, bbl/d
0 2,500 5,000 7,500 10,000 12,500 15,000 17,500 20,000
STO WAT
STO pour point
Live-oil pour point
Live-oil WAT
1.2
1.0
0.8
0.4
0.2
0
0.6
D
e
p
o
s
i
t

t
h
i
c
k
n
e
s
s
,

m
m
Flowline distance, mi
0 0.5
Wellhead Collection
facility
1.0 1.5 2.0
Recommended wax-removal thickness
PIP insulation
Wet insulation
PIP insulation
Wet insulation
Standard n-paraffin
diffusion coefficient
WDFL n-paraffin
diffusion coefficient
60336schD7R1.qxp:Format Layout V7.2 12/4/07 7:11 PM Page 69
because it simulates the shear environment at
the flowline wall (above).
Thanks to live-oil testing in the RealView cell,
significant advancements in asphaltene flow
assurance have already been realized.
32
Preliminary laboratory results show that live-oil
testing will lead to similar improvements for waxy
crude oils. A recent laboratory study tested the
behavior of a black waxy crude oil in a hypothet -
ical flow-assurance situation. Oilphase-DBR
scientists assumed the oil entered a 6-in. ID,
2-mi long pipeline at a temperature and pressure
of 170F and 3,170 psi [21.9 MPa]. The simulated
oil-production rate was 5,000 bbl/d, and the
seabed temperature was 65F [18.3C]. Within
this scenario, engineers performed deposition-
rate tests that compared the performance of
STO and live oils in laminar and turbulent flow
(above right). The results showed that wax
deposition is slower in turbulent flow, and live-oil
wax-deposition rates are far lower than those
observed with STO.
The low wax-deposition rates observed with
live oils would significantly impact the hypo -
thetical flowline design and the frequency of
wax-removal operations. OLGA simulations
demon strated that the 2-mi flowline could be
constructed from foam-insulated pipe instead of
pipe-in-pipe, potentially saving US $4 million. In
the context of flowline remediation, traditional
WDFL measurements with STO indicated that
wax removal would be necessary every two weeks.
Simulations with live-oil data predicted that wax-
removal would be necessary about once per year,
providing significant operational savings.
As a field produces, the crude-oil properties
often change. For example, as gas condensate
falls below the saturation pressure during
depletion, the condensate yield and the WAT may
fall. In a compositionally graded accumulation,
composition may change as fluids are produced
from regions that were originally distant from the
initial sampling point. The impact on flow
assurance may be significant.
Traditionally, engineers monitor fluid-
property evolution by periodically sampling from
separators or extracting live oil near perfora -
tions, and performing flow-assurance tests in the
laboratory. In offshore and deepwater fields, this
approach is costly.
33
With the advent of
intelligent completions, equipped with sensors
that transmit downhole temperature, pressure
and flow rates in real time, production moni -
toring can be performed remotely. These
completions also incorporate remote-control
pumps and valves that engineers can use to
mitigate flow-assurance problems.
Chemical sensors are being developed that
can detect fluid-composition changes. When
installed at strategic locations in the well
completion and along a pipeline, the sensors will
provide real-time data for monitoring of solids
deposition, corrosion rates and rheological
properties. As a result, interventions for
sampling or remediation will be performed only
when necessary.
Today, fluid sampling and analysis are
progressing to a point at which consistent
standards are applied along the continuum from
DFA and reservoir characterization, to sampling
and laboratory analysis, and on to production
surveillance. This integrated approach will be
increasingly valuable to operators making
reservoir exploration, development and
production decisions, particularly in high-risk,
remote locations. EBN
70 Oileld Review
32. Akbarzadeh et al, reference 5.
33. Amin A, Smedstad E and Riding M: Role of Surveillance
in Improving Subsea Productivity, paper SPE 90209,
presented at the SPE Annual Technical Conference and
Exhibition, Houston, September 2629, 2004.
>
Illustration and schematic diagram of the
RealView live solids deposition cell. The cell can
achieve turbulent flow and wall-shear conditions
that reflect those found in flowlines (top). The
inner spindle inside the cell rotates to create
fluid movement (middle). Wall temperature and
system pressure can be independently
controlled. The deposition surface and
roughness can be changed by inserting special
sleeves. The wax deposit appears on the
stationary-cylinder surface (bottom).
>
Wax-deposition behavior of waxy crude oil in the WDFL and RealView
apparatus. Engineers performed tests with dead and live oils. In laminar flow,
dead-oil wax-deposition rates were essentially the same in the WDFL and
the RealView cell. Turbulent flow in the RealView cell reduced the dead-oil
deposition rate substantially. Live-oil deposition rates in the RealView
apparatus were even lower for both flow regimes.
30
25
20
15
10
5
0
T
o
t
a
l

w
a
x
-
d
e
p
o
s
i
t
i
o
n

r
a
t
e
,

m
g
/
m
2
s
STO,
laminar
STO,
laminar
STO,
turbulent
Live oil,
laminar
Live oil,
turbulent
24.5
23.7
15.5
9.2
3.3
WDFL RealView cell
Multipoint
thermocouples
Shear cell
Variable-speed
DC motor
High-pressure
flow-control
valves
Mechanical
mounting
stand
Inner
rotating
cylinder
Outer
stationary
cylinder
Deposit
Coolant
Coolant
out
Charge
Discharge
Coolant
in
Oil
Electrical
heating
cartridge
60336schD7R1.qxp:Format Layout V7.2 12/4/07 7:11 PM Page 70