Advancing Fluid-Property Measurements

Reservoir-fluid properties play a key role in designing and optimizing well completions
and surface production facilities to manage reservoirs efficiently. Therefore, accurate
fluid characterization is a vital part of any oil or gas production project. Advanced
fluid-analysis techniques provide the high-quality data required to develop appropriate
production strategies.

Soraya Betancourt Discovery of an oil or gas accumulation immedi- efficiently transport the reservoir fluids from the
Cambridge, Massachusetts, USA ately prompts questions about its economic formation through wells, flowlines, production
viability. Operators want to learn about the facilities and beyond. During this journey,
Tara Davies extent of the reservoir, the types of fluids that reservoir fluids experience temperatures and
Ray Kennedy would be produced, expected production rates pressures far different from their initial in-situ
Edmonton, Alberta, Canada
and how long production might be sustained. conditions. These variations may induce
Fluid analysis is a critical part of the process by physical-state changes that would inhibit or
Chengli Dong
Sugar Land, Texas, USA which engineers perform reservoir characteri- interrupt production if not understood prior to
zation, determine the reservoir architecture and designing tubulars and facilities. Therefore, to
Hani Elshahawi decide whether an oil or gas accumulation is determine how the fluids will respond to
Shell International Exploration and Production worth developing. High-quality samples are production conditions, engineers may want to
Houston, Texas essential, because erroneous data could lead collect and analyze fluid samples from each
engineers to misinterpret production parameters potentially productive layer in the reservoir.
Oliver C. Mullins such as drainage volume, flow rates, reserves and Traditionally, fluid samples are collected and
John Nighswander facilities design and completion. Clearly, poor or sent to offsite laboratories for testing, a process
Houston, Texas misleading fluid data can have a severe negative that delays data access and impedes an opera-
financial impact. tor’s ability to make time-sensitive development
Michael O’Keefe
If the reservoir analysis is positive, engineers decisions. Today, sophisticated formation
Hobart, Tasmania, Australia
begin designing a production system that will sampling and testing tools allow data collection
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 Profiling, LFA (Live 16,000
Fluid Analyzer), MDT (Modular Formation Dynamics Tester),
Oilphase-DBR, PVT Express, Quicksilver Probe and RealView 14,000
are marks of Schlumberger Wax
OLGA is a trademark of Scandpower AS. 12,000
1. Ratulowski J, Amin A, Hammami A, Muhammed M and
10,000
Pressure, psi

Riding M: “Flow Assurance and Subsea Productivity: Reservoir

Closing the Loop with Connectivity and Measurements,”
paper SPE 90244, presented at the SPE Annual 8,000 Hydrate
Technical Conference and Exhibition, Houston,
September 26–29, 2004. 6,000 Asphaltene
2. For more on scale-removal techniques:
4,000 Bubblepoint
Crabtree M, Eslinger D, Fletcher P, Miller M, Johnson A
and King G: “Fighting Scale—Removal and Prevention,”
Oilfield Review 11, no. 3 (Autumn 1999): 30–45. 2,000
Flowline
0
0 50 100 150 200 250 300
Temperature, °F

> 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).

56 Oilfield Review
 45
45 40
40 35
35 30
Median-weight oil 30 25
25 20
20 15
15 10
Water 10 5
5 0
Hydrocarbons 90
0
Condensate 80

5
70

3
60

2
50

1
40

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
0
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.

Autumn 2007 57
 Oil Field Gas Field Gas Field with Reservoir-Fluid Sampling
with Dissolved Gas with Retrograde No Retrograde Scientists and engineers of various disciplines rely
Condensation Condensation
on fluid-sample data when making decisions about
reservoir development. For example, reservoir

Cricondentherm
engineers use the data to determine reservoir
Critical Cricondenbar architecture, estimate reserves, perform material-
point
Oil balance calculations and analyze fluid flow in
us porous media. Geologists need accurate informa-
loc
o int tion to perform reservoir correlations and
lep
bb 50%
Bu geochemical studies. Refining and marketing
Pressure

personnel make decisions about product yield and

40%
value. If erroneous data are used, unanticipated
20%
and expensive consequences could result

locus
Oil and gas 10% during production.3
Gas

oint
5% A wide range of fluid behaviors can influence

wp
a sampling and analysis program. A reservoir-

De
0% fluid system can be roughly categorized by its
vapor-liquid phase behavior; the classifications
Temperature range from dry gas, wet gas and retrograde gas to
volatile oil, black oil or heavy oil (left).4 Another
> Generalized pressure-temperature (PT) diagram for reservoir fluids. The
consideration is hydrocarbon solid-phase
diagram contains two principal regions: single-phase (green to orange) and
two-phase (beige). The boundary between these regions is called the behavior. Wax and hydrate formation is
saturation envelope; it has three principal features. The bubblepoint locus is predominantly induced by a temperature
the portion at which gas begins to separate from liquid. The dewpoint locus is decline, and pressure reductions or fluid
the segment at which liquid begins to condense from gas. The critical point is
commingling generally cause asphaltenes to fall
the location where the loci meet. The cricondentherm is the highest
temperature on the saturation envelope, and the cricondenbar is the highest out of solution (next page, top).5
pressure on the saturation envelope. Reservoir fluids are classified according Engineers collect formation-water samples to
to their in-situ reservoir and production behavior in the PT scheme. Dry gas determine whether calcite, barite or halite
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 scales will form within flowlines. Corrosive and
pressure depletion; however, during production, it crosses the dewpoint locus toxic substances such as carbon dioxide [CO2]
and forms a liquid phase. Retrograde gas resides in the single-phase region at and hydrogen sulfide [H2S] must be detected and
temperatures between the critical point and the cricondentherm. During measured because they influence tubular-alloy
pressure depletion at reservoir temperature, liquid forms within the reservoir
itself, and persists throughout the production PT path. Volatile oil resides in the selection and the design of safety and
single-phase region just to the left of the critical point. Gas liberation occurs as environmental systems from the wellhead to the
the fluid crosses the bubblepoint locus during production. Black oil exists in the surface production facility. Water pH is also an
single-phase region at reservoir temperatures far lower than the critical point. important parameter governing scale and
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 corrosion, and may be measured downhole to
very low quantities of gas, and the liquid phase is predominantly composed of avoid uncertainties.6
high-molecular-weight components.

Increasing depth

> 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.)

3. Nagarajan NR, Honarpour MM and Sampath K:
“Reservoir-Fluid Sampling and Characterization—Key to
Efficient Reservoir Management,” Journal of Petroleum
Technology 59, no. 8 (August 2007): 80–91.
4. McCain WD Jr: “The Five Reservoir Fluids,” in The
Properties of Reservoir Fluids (2nd Edition). Tulsa:
PennWell Books (1990): 147–164.
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): 22–43.
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): 3257–3265.
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 3–6, 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): 83–88.
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): 944–952.

58 Oilfield Review
 Another concern is the variability of Asphaltene Wax
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 operator’s 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 wrap—a 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 Inorganic scale Hydrate
magnitude of compositional variation can vary > Common deposits that form in tubulars during hydrocarbon production.
greatly, depending on the geological and Wax and hydrate deposition mainly result from a temperature decrease,
geochemical history of the reservoir (previous while asphaltene precipitation may be triggered by changes in pressure,
page, bottom).8 temperature and composition. Inorganic scales arise from changes in
pressure, temperature and composition of aqueous fluids that accompany
A technically robust fluid-sampling program hydrocarbon production. (With kind permission of Springer Science and
is vitally important when reservoir compartmen- Business Media.)
talization, compositional grading or both exist.
Reservoir-formation properties influence the
ability to collect representative fluid samples.
Sample collection requires fluid flow into the Reservoir pressure
borehole, which occurs only when the wellbore
Pressure

flowing pressure is lower than formation Original

pressure. However, if the flowing pressure falls reservoir
fluid
below the fluid saturation pressure, a gas phase
(in the case of volatile or black oil) or a liquid Free gas
phase (in the case of retrograde gas) will form Bubblepoint
(right). The relative mobility of each fluid phase Wellbore flowing pressure
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
Pressure

effect can be minimized or eliminated by Original

sampling at flow rates and pressure differentials
that are as low as feasible.
Finally, accurate reservoir-temperature
measurements are vital. Errors of just a few
degrees during PVT testing could result in
misinterpretation. 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.
reservoir One-phase
fluid representative
sample fluid
> 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).

Autumn 2007 59
 Reservoir section DFA GOR, ft3/bbl
Power cartridge 100 1,000 10,000 100,000
0
A

Sample-bottle 500 B
Se
module al
1,000

Se 1,500
al
Pump module
(sample) C
2,000 D
E
LFA fluid analyzer
(sample) 2,500
F G
Sample
flow Hydraulic module 3,000 H J
I

Focused sampling
probe
LFA fluid analyzer
Guard (guard)
flow
Pump module
(guard)
> 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.
There are two principal hydrocarbon-sample
acquisition methods—bottomhole 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
3,500
Oil Dry gas Depth, ft
Water Rich gas and condensate
> 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.
performed by the MDT Modular Formation and may continue to do so throughout a well’s
Dynamics Tester and the Quicksilver Probe tool for lifetime to monitor fluid-property evolution.10
focused extraction of pure reservoir fluid. Cased- Bottomhole fluid samples must be extracted
hole sampling devices include the CHDT Cased from locations that will yield the most relevant
Hole Dynamics Tester, single-phase reservoir information for making decisions. To aid in this
sampler (SRS) and SCAR sampling tool.9 Surface endeavor, today’s sampling and testing tools
sampling, which is most frequently performed at include an array of instruments that can perform
the separator under stable flow conditions, downhole fluid analysis (DFA). DFA tools provide
involves collecting gas and liquid samples. real-time fluid-property measurements at
Engineers may acquire surface samples during reservoir conditions, allowing engineers to
exploration if downhole methods are unavailable, analyze samples before they are gathered.

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): 58–65.
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): 24–43.
Quicksilver Probe: Akkurt, R, Bowcock M, Davies J,
Del Campo C, Hill B, Joshi S, Kundu D, Kumar S,
O’Keefe 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): 4–19.
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):
46–57.
SRS and SCAR tools: Aghar, H, Carie M, Elshahawi H,
Gomez JR, Saeedi J, Young C, Pinguet B, Swainson K,
Takla E and Theuveny B: “The Expanding Scope of Well
Testing,” Oilfield Review 19, no. 1 (Spring 2007): 44–59.
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): 54–61.
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):
26–41.
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): 7–14.
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 29–30, 2004.
13. Dong CM, O’Keefe 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 4–7, 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 4–7, 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 30–November 1, 2007.

60 Oilfield Review
 DFA methods include in-situ optical
absorption spectroscopy, optical reflectance,
fluorescence and some nonoptical measurements
including density, viscosity and pH. The
spectrometer operates in the visible to near-
infrared range—at wavelengths between 400 and
2,100 nm. Spectra are recorded in real time,
revealing the proportions of methane [C1], ethane
to pentane [C2–5], hexane-plus [C6+] and CO2
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 retrograde
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
The analysis revealed wide variations in DFA and FCA analysis showed that fluids
reservoir-fluid composition, ranging from dry gas above and below the inversion had a greater than
and condensate gases at the top to black oils with 99% probability of existing in different
different GORs at the bottom. At the bottom of compartments. A pressure discontinuity between
the oil column, the GOR varied gradually with the compartments confirmed the absence of
depth in the bottom sand, indicating a fluid- hydraulic communication.
composition gradient. GOR inversions were also Shell and Schlumberger fluid specialists
detected between the top and bottom sands, explored the oil column in detail, comparing the
suggesting the presence of flow barriers and a GOR information with petrophysical, formation-
complex reservoir structure. pressure and mud-gas logs, and performing FCA
analysis (below). The gamma ray and pressure
5,500 Pressure, psi 6,000 Relative Methane Concentration
900 GOR, ft3/bbl 1,500 0 Gamma Ray, API 150 0 2 4
2,750
F Top sand F
2,800 G G
Higher concentration of
2,850 C1 corresponds to lower
density of Fluid J
2,900
Depth, ft

collected (previous page, top left).13 2,950 J J

H H
Fluid Profiling reservoir-fluid characteri-
zation by DFA can diagnose compositional 3,000
Bottom sand Increasing
grading and help identify reservoir compart- 3,050
contribution
I of thermogenic
ments. For example, abrupt fluid-composition or I charge
GOR changes between zones in a single well or 3,100
between neighboring wells may indicate compart- 3,150
mentalization. To confirm that perceived
fluid-property differences are truly significant, –55 δ13C methane, parts per thousand –50
engineers must first consider measurement
uncertainties. A recent method to evaluate 0.10 1.0
uncertainties is the fluid comparison algorithm 0.09 0.9
Oil-base mud contamination, ση

(FCA).14 FCA uses parametric models to estimate 0.08 0.8

GOR and coloration uncertainties as a function 0.7
0.07
of optical-density (σε) and mud-contamination 0.6
0.06
(ση) measurement variability. The algorithm 0.5
0.05 Fluids G and H
compares measurements acquired from two P = 0.95 0.4
fluids and calculates the probability that differ- 0.04
P = 0.80 0.3
ences are statistically significant. When the FCA 0.03 0.2
P = 0.90
result indicates that the fluids are different, 0.02 Fluids G and J 0.1
sample acquisition for detailed surface analysis P = 0.99
0.01 0
is justified. The following case study demon- 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10
Optical density, σε
strates how engineers employ DFA and FCA
to characterize a reservoir and determine > Expanded log presentation and FCA analysis of an oil column at the bottom of a deepwater
sampling locations. 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
Sampling and Reservoir Characterization in a with depth. The fluid densities derived from pressure gradients (blue) reveal two principal features—a
Deepwater Accumulation fluid-density inversion between Fluids G and J, and a gradual density increase from Fluid J to Fluid I—
In a deepwater exploration well, Shell employed suggesting no vertical communication between the top and bottom sands. The mud-gas log (top right)
real-time DFA and FCA to characterize the offers further supporting evidence. At the depth of Fluid J, the δ13C value (red) falls abruptly and
increases gradually with depth. The relative methane concentration (blue) also fluctuates sharply at
reservoir architecture and decide where to Fluid J, then decreases gradually with depth—a feature consistent with the fluid-density increase.
collect fluid samples.15 As the MDT assembly Final confirmation resulted from applying the FCA technique (bottom). The FCA algorithm generates a
traveled down the well, the spectrometers number representing the probability that two fluids are statistically different. As measurement
uncertainties (σ) decrease, confidence in data quality increases. Therefore, low σ values indicate a
measured the GOR at several locations (previous
high probability that perceived fluid-property differences are real. The contour plot indicates that
page, top right). Based on FCA analysis, fluid Fluid G has a 99% probability of being different from Fluid J, and a 95% probability of being different
samples were collected at 10 different depths. from Fluid H. Discovery of reservoir compartmentalization led Shell to reevaluate the field and make
significant strategic adjustments in reservoir development.

Autumn 2007 61
 mud gas may imply the presence of a seal.17
Engineers noted a δ13C discontinuity at around
1.0 2,950 ft [899 m], suggesting yet another flow
0.9 barrier. FCA analysis provided confirmation,
calculating a 95 to 99% probability that fluids
0.8
above and below the discontinuity were
0.7 dissimilar and from different compartments with
0.6 no communication.
Optical density

0.5 The discovery of the fluid compartments in

this deepwater field led Shell engineers to adjust
0.4
their reservoir models and development plan-
0.3 ning. Reservoir compartmentalization increases
0.2 design complexity and cost because engineers
must treat each zone independently. Production
0.1
Laboratory sample forecasts, reserve calculations and enhanced
0.0 LFA tool recovery schemes became proportionately
-0.1 more complex.
1,000 1,100 1,200 1,300 1,400 1,500 1,600 1,700 1,800 1,900 2,000
Improved regional understanding of
Wavelength, nm
subsurface architecture impacted short-term
1.5 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
1.0
1,070 altered to anticipate a GOR reduction with time.
LFA optical density

Access to real-time fluid analyses allowed

1,725
Shell to make decisions much earlier in the field-
development process, and accelerated the
0.5 project by at least six months. At today’s oil
prices, the time saved was worth hundreds of
1,290
1,671
millions of dollars.
1,920 1,445
1,600 X=Y Laboratory Fluid Preparation and
0 Data regression
Sample Chain of Custody
0 0.5 1.0 1.5 The preceding case study demonstrates the
Laboratory optical density considerable effort and care that engineers
apply during the sampling process. Nevertheless,
> Chain-of-custody investigation of a valid crude-oil sample. Chemists the harsh downhole environment and the
compare visible near-IR spectra from DFA and the laboratory sample. Both nature of well operations may render DFA
spectra were measured at downhole conditions—15,000 psi [103 MPa] and
250°F [121°C]. The spectral analysis (top) compares the downhole discrete
equipment, sampling devices and subsequent
data (red circles) with the continuous spectrum from the laboratory sample analysis susceptible to fouling, failure and
(blue). Data regression (bottom) of the LFA optical densities versus those other inaccuracies.
from the laboratory shows excellent agreement, evidenced by the near- Schlumberger engineers addressed this
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. problem by implementing a chain-of-custody
procedure, a concept borrowed from forensic
science.18 Evidence must make the journey from
logs showed that the top portion of the oil resides gathered at the surface during drilling or a crime scene to the courtroom in a validated
in one sand lobe with a relatively constant GOR separated from fluid samples can be analyzed for and secure manner; otherwise, it may not be
and pressure gradient. The rest of the oil is in a isotopic content. The isotopic signature, δ13C, is admissible in court. Similarly, chemists at a
lower sand lobe, where the GOR decreases with the value of the 13C/12C methane-isotope ratio in remote testing laboratory should be able to
depth. The pressure log was consistent, indi- a sample relative to a standard, expressed in determine whether the chemical composition of
cating a significant fluid-density difference parts per thousand. When plotted alongside a a field sample has been preserved. DFA provides
between the upper and lower lobes. standard mud-gas log, larger δ13C values may a convenient way to establish a chain of custody
Recent advances in mud-gas logging have indicate higher concentrations of biogenic gas in for fluid samples, because chemists have the
given engineers another tool to perform real- the reservoir. Trending δ13C values may indicate opportunity to compare analytical data acquired
time detection of seals and permeability barriers, nonequliibrium methane distributions, and a downhole with those from the corresponding
lithological variations and fluid contacts.16 Gases clear break in the methane-isotope signature in samples that reach the laboratory.

62 Oilfield Review
 In the field, after tool retrieval, engineers
remove the reservoir-fluid samples. At this point,
a PVT Express onsite well fluid analysis system 1.6
may be available to conduct preliminary
1.4
measurements that determine whether the
Laboratory sample
collected-sample properties agree with those LFA tool
1.2
measured by DFA.19 If onsite analysis is
unavailable or more sophisticated testing is 1.0

Optical density
required, the samples are shipped to a remote
testing laboratory in the original sample bottles 0.8
or transferred to an approved shipping container.
When fluid samples arrive at the Schlumberger 0.6
laboratory, chemists restore the fluid inside the
0.4
sample container to the original reservoir
temperature and pressure, and allow the fluid to
0.2
equilibrate by agitating it continuously for up to
five days. The restoration process is intended to 0
1,000 1,200 1,400 1,600 1,800 2,000
redissolve precipitated asphaltene and wax
Wavelength, nm
particles, ensure a homogeneous fluid through-
out the sample cylinder and provide a 1.5
single-phase representative fluid for testing.
The visible near-infrared (IR) spectrum is an
effective hydrocarbon fingerprint.20 The DFA
1,070
spectrometer performs sample analysis as the
crude-oil sample is acquired, providing a direct 1.0
LFA optical density

fluid-property measurement under downhole 1,725

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 ft3/bbl
[103.3 m3/m3], was significantly higher than that
from the laboratory spectrum, 320 ft3/bbl
[57.0 m3/m3]. This difference corresponded to a
0.5
1,280
1,671
1,445
1,820
1,600 X=Y
Data regression
0
0 0.5 1.0 1.5
Laboratory optical density
> Chain-of-custody investigation of a compromised crude-oil sample. Both
spectra were measured at downhole conditions—20,000 psi [138 MPa] and 200°F
[93°C]. 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.
16. Venkataramanan L, Elshahawi H, McKinney D, 19. Aghar et al, reference 9.
Flannery M, Hashem M and Mullins OC: “Downhole 20. For more on visible and near-IR spectroscopy:
Fluid Analysis and Fluid Composition Algorithm as an
Crombie et al, reference 11.
Aid to Reservoir Characterization,” paper SPE 100937,
presented at the SPE Asia Pacific Oil and Gas 21. Mullins OC, Beck G, Cribbs MY, Terabayashi T and
Conference and Exhibition, Adelaide, Australia, Kagasawa K: “Downhole Determination of GOR on
September 11–13, 2006. Single-phase Fluids by Optical Spectroscopy,”
Transactions of the 42nd SPWLA Annual Logging
17. Berkman T, Ellis L and Grass D: “Integration of Mud Gas
Isotope Data with Field Appraisal at Horn Mountain
Symposium, Houston, June 17–20, 2001, paper M.
Venkataramanan L, Fujisawa G, Mullins OC, Vasques RR
Field, Deepwater Gulf of Mexico,” AAPG Bulletin 86,
and Valero H-P: “Uncertainty Analysis of Near-Infrared
no. 13 (2002): supplement.
18. Betancourt SS, Bracey J, Gustavson G, Mathews SG and
Data of Hydrocarbons,” Applied Spectroscopy 60, no. 6
(June 2006): 653–662.
Mullins O: “Chain of Custody for Samples of Live Crude
Oil Using Visible Near-Infrared Spectroscopy,” Applied
Spectroscopy 60, no. 12 (December 2006): 1482–1487.

Autumn 2007 63
 93% likelihood that either one (or both) of the
spectra were in error, or the laboratory sample
100.00
was compromised. Further investigation of
surface procedures and DFA-tool performance
would be required before this sample could be
10.00 used with confidence in the laboratory.
Weight percent

Laboratory Techniques for Flow Assurance

1.00
In the laboratory, chemists determine fluid
compositions and measure fluid properties
related to flow assurance. For compositional
analysis, an accurately measured volume of fluid
0.10 is isobarically and isothermally transferred to a
pycnometer to measure mass and density. Then,
the pycnometer is connected to an apparatus in
0.01 which the oil sample is cooled to ambient
CO2
H2S
N2
C1
C2
C3
I-C4
N-C4
I-C5

C6
MCYC-C5
Benzene
CYCL-C6
C7
MCYCL-C6
Toluene
C8
C2-Benzene
M&P-Xylene
O-Xylene
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
C20
C21
C22
C23
C24
C25
C26
C27
C28
C29
C30+
N-C5

temperature and decompressed. The volume of

gas liberated by this procedure allows
calculation of the GOR.
Components Chemists employ gas chromatography to
determine the vapor-phase composition up to
> Typical crude-oil compositional analysis, determined by gas chromatography. The plot shows the
C15+ and the liquid-phase composition up to C36+.
hydrocarbon contribution from C1 to C30+, as well as the CO2 concentration.
The bulk crude-oil composition is calculated by
summing the individual contributions from each
phase (above left). This procedure ensures that a
Side view
consistent reservoir-fluid composition is available
for subsequent fluid-property characterization
High- and reservoir-production simulation.
pressure Cathetometer
pumps The bubblepoint pressure is determined by
performing a constant composition expansion
(CCE) test. Technicians place a known volume of
Solvent
Oil

equilibrated fluid in a PVT cell at reservoir

temperature and pressure (left). The fluid is
Magnetic mixer 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
16,030 reductions continue in increments ranging from
14,030
100 to 500 psi [0.69 to 3.45 MPa], the vapor and
liquid phases are equilibrated at each step, and
12,030 Tres = 176°F phase volumes are measured and plotted as a
T = 120°F function of pressure. For a black oil, the
10,030 T = 75°F
Pressure, psi

intersection of the single- and two-phase lines in

8,030 the PV plot defines the bubblepoint.
The flow-assurance testing protocol depends
6,030
on the nature of the crude oil. For example, when
4,030 asphaltenes are of concern, then saturate,
aromatic, resin and asphaltene (SARA) analysis
2,030
and paraffinic solvent titration with dead oil are
30 principal screening techniques.22 It is also
25 30 35 40 45 50 55 60 65 70 75 common to measure the asphaltene-precipitation
Volume, cm3 pressure on a live-oil sample. If testing identifies
an asphaltene-precipitation problem, additional
> 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
studies are conducted to map out the asphaltene
mixer provides vigorous agitation to maintain phase equilibrium. A video-based cathetometer phase diagram and evaluate the effectiveness of
measures fluid levels in the cell for phase-volume calculations. The oven heats the PVT cell to the chemicals or coatings as prevention strategies.23
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).

64 Oilfield Review
 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 oil’s 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 C60 to C100)—information
chemists enter into thermodynamic models to
predict wax behavior.
The wax-appearance temperature (WAT) is
one of the most important flow-assurance
measurements, 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 392°F [200°C].
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.
Autumn 2007
Charged
coupled
device
Hot stage
top view
Analyzer
10 50
Temperature 48.6 20
Hot stage
Cooling
Polarizer 360° gas
rotatable
stage
IR filter
0°C WAT = 42°C 44°C
> 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 111°F [44°C] (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 0°C (left panel). This method is accurate to ± 2°F [1.1°C].
The pour point is the temperature below Waxy crude oils tend to form gels at
which a fluid is no longer pourable because of temperatures below the pour point. In the event
viscosification, gelation or solids formation. of a production shutdown, high pumping pres-
Pour-point testing with dead oils conforms to sures may be required to break the gel and
ASTM Standard D97.25 For live oils, engineers use restore flow. Therefore, yield-stress data from
an apparatus containing a visual sapphire cell waxy crude oils are necessary to properly design
mounted on an automatic, vibration-free pivoting flowlines and avoid production problems. The
bracket that resides in a programmable
convection oven. The oil sample is heated to the 22. Live oil contains the dissolved gases present in the
reservoir. If the gases are released, or “flashed,” at the
reservoir temperature, and slowly cooled until surface or in the laboratory, the residual liquid is called
fluid movement in the cell ceases. Most waxy dead oil. Stock-tank oil (STO) is also a dead oil.
23. For a detailed discussion of asphaltene flow assurance:
crude oils begin to display non-Newtonian
Akbarzadeh et al, reference 5.
behavior near the pour point. If the pour point 24. At constant temperature and pressure, Newtonian fluids
exceeds anticipated oil-production temperatures, display constant viscosity at all shear rates. The
viscosity of non-Newtonian fluids is not constant at all
engineers may add pour-point-depressant shear rates. Apparent viscosity is the viscosity of a fluid
chemicals to maintain flow. at a given shear rate and temperature.
The apparent viscosity of waxy crude oils 25. ASTM D97-06 Standard Test Method for Pour Point of
Petroleum Products. West Conshohocken, Pennsylvania,
generally increases dramatically as the USA: ASTM International, 2006.
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 302°F [150°C], allowing work
with live oils.
65
 N2 pressure to break gel yield stress of live fluids can be determined by a
Backpressure model pipeline test (MPT) (left).26
High-pressure regulator
circulation pump Another important measurement in the
context of production-system design and flow
Heated lines
assurance is the wax-deposition rate. The
principal controlling parameters are fluid
Temperature-

Convection oven
temperature, heat loss through the pipeline wall,
controlled
Fluid bath wax content, shear rate and fluid viscosity. The
sample
cylinder 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
System charging pump remediate wax deposition and gel formation.

Optimizing Subsea System Design

> Model pipeline test (MPT) apparatus. Stock-tank oil circulates through a in West Africa
coil of tubing immersed in a temperature-controlled bath. Flow through the A West African field, located at a water depth less
coil stops when the bath reaches the test temperature, allowing the fluid to than 1,000 ft [300 m], has a reservoir
age and form a gel structure. After the aging period, laboratory workers
temperature and pressure of 170°F [76.7°C] and
measure the nitrogen pressure necessary to initiate flow in the coil and
calculate the gel strength from a simple force-balance equation. 3,180 psi [21.9 MPa]. The seabed temperature is
55°F [12.8°C]. 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
Pressure transducer
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
Cooling water Compositional analysis of the reservoir fluid
Thermocouples revealed a black oil with a GOR of 230 ft3/bbl
[41.0 m3/m3] and an API gravity of 36.3.
Cooling bath Compositional analysis indicated that the
Thermocouples Thermocouples
Deposition section
fraction with a carbon number above C30 was
35.8%. Further characterization of the C30+
fraction and n-paraffin distribution revealed that
the crude oil contained about 13.1 wt% C17+
n-paraffins.28 The relatively high n-paraffin
Reservoir Pump concentration was cause for concern that the

26. The yield stress, τy, is calculated by the following force-

Flowmeter PyD
balance equation: τy = ____ where Py is the hydraulic
4L
pressure necessary to cause fluid movement, D is the
Remelting bath Tempering bath inner diameter of the coil and L is the coil length.
> Wax deposition flow loop (WDFL). The WDFL is a miniature flow loop that exposes stock-tank oils 27. Alboudwarej H, Huo Z and Kempton E: “Flow-Assurance
Aspects of Subsea Systems Design for Production of
(STO) to a range of heat fluxes and shear rates that would be expected in the flowline during actual Waxy Crude Oils,” paper SPE 103242, presented at the
production. The deposition section of the flow loop is a 39-in. [1-m], 0.375-in. OD stainless steel tube. SPE Annual Technical Conference and Exhibition,
A 0.53-galUS [2.0-L] oil reservoir maintains the oil above the WAT throughout the test. The oil exits the San Antonio, Texas, September 24–26, 2006.
reservoir and flows at a controllable rate through 50 ft [15 m] of coiled copper tubing in a tempering 28. Paraffin is a common name for a group of alkane
bath for temperature adjustment. The deposition loop is immersed in a cooling bath to simulate heat hydrocarbons with the general formula CnH2n+2, where
n is the number of carbon atoms. The simplest paraffin
loss in the pipeline. As the oil flows through the deposition loop, a data-acquisition system monitors
molecule is methane, CH4, a gas at room temperature.
the wall temperature, oil and water temperatures, the pressure drop between the inlet and outlet of the Octane, C8H18, is liquid at room temperature. The solid
coil and the flow rate. Wax deposits constrict flow inside the deposition loop, increasing the pressure forms of paraffin are heavier molecules from C20 to C40.
required to maintain flow. The increase in pressure allows calculation of the amount of deposited wax. Linear members of the series (those with no branches or
cyclic structures) are called n-paraffins.

66 Oilfield Review
fluid might exhibit flow-assurance problems
related to wax deposition. Therefore, the 160
Oilphase-DBR team objective was to generate WAT
measurements that would provide guidance Pour point
concerning ways to mitigate and remediate wax 130

Temperature, °F
deposition during steady-state and transient
events during crude-oil production.
100
Constant composition expansion (CCE)
testing measured a bubblepoint pressure of
700 psi [4.8 MPa] at the reservoir temperature. 70
WAT and pour-point tests were performed with
both live- and dead-oil samples (right). As the
fluid pressure fell below the bubblepoint to 40
0 200 400 600 800
ambient conditions, dissolved gas escaped, the
Pressure, psi
average sample composition shifted toward
heavier hydrocarbons, and both the WAT and > Wax-appearance temperature (WAT) and pour-point (PP) data measured
pour point increased approximately 20°F [11.1°C]. with live and dead West African crude oil. The dead-oil measurements are
Rheological testing revealed gel-structure 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)
formation in live crude oil at low shear rates increased approximately 20°F. The increases resulted from the loss of
when the fluid temperature fell below the pour dissolved gases in the fluid.
point and approached the seabed temperature
(below). As the rheometer pressure decreased,
the shear stress required to break the gels
increased—behavior consistent with the loss of
lighter hydrocarbons. Around the pour point,

700 psi 300 psi

1.0E+11 1.0E+11
1.0E+10 1.0E+10
1.0E+09 Viscosity at 55° F 1.0E+09
1.0E+08 Viscosity at 65° F 1.0E+08
1.0E+07 Viscosity at 80° F 1.0E+07
Viscosity, mPa-s
Viscosity, mPa-s

1.0E+06 1.0E+06
1.0E+05 1.0E+05
1.0E+04 1.0E+04
1.0E+03 1.0E+03
1.0E+02 1.0E+02
1.0E+01 1.0E+01
1.0E+00 1.0E+00
10 100 10 100
Shear stress, Pa Shear stress, Pa

> Rheological behavior of a live West African crude oil. Shear-stress

100 psi versus viscosity measurements were performed near and below the pour
1.0E+11 point at 100, 300 and 700 psi [0.69, 2.07 and 4.83 MPa]. At 700 psi (top left),
1.0E+10 elevated low-shear-rate plateau viscosities at 55° and 65°F [12.8° and
1.0E+09 18.3°C] indicated the presence of gel structures. Little gelation occurred
near the pour point at 80°F [26.7°C]. Similar behavior occurred at 300 psi
1.0E+08
(top right); however, higher shear stresses were required to break the gels
1.0E+07
Viscosity, mPa-s

at 55° and 65°F—behavior consistent with the loss of lighter hydrocarbons

1.0E+06 in the sample. At 100 psi (bottom left), the 80°F sample exhibited strong
1.0E+05 shear thinning behavior.
1.0E+04
1.0E+03
1.0E+02
1.0E+01
1.0E+00
10 100
Shear stress, Pa

Autumn 2007 67
 profile into a standard equation-of-state model
40 that calculates an n-paraffin diffusion
coefficient. The model assumes that molecular
170 s-1
35 diffusion of wax molecules is the principal
511 s-1

Deposition rate, mg/m2s

driving force governing wax deposition.29
30
The next step involved entering the flow-
assurance measurements, wax-deposition data
25
and diffusion coefficients into the OLGA
20
simulator—a commercial multiphase-flow fluid-
transport model—to assess and predict crude-oil
15 behavior in various production scenarios.
Schlumberger engineers applied OLGA simula-
10 tions to two cases: producing at a steady state
70 80 90 100 110 120 130
Temperature, °F
and restarting production after a shutdown.
Because the WAT of the West African crude oil
> Wax-deposition behavior of a dead West African crude oil. Deposition was about 50°F [27.7°C] higher than the pour
measurements at two shear rates, 170 and 511 s–1, revealed unusual behavior. point, wax deposition and gel formation were the
At the lower shear rate (red), deposition slowed steadily with increasing principal flow-assurance risks during steady-
temperature. However, when the fluid temperature exceeded about 100°F
[37.8°C], the deposition rates at the higher shear rate (blue) suddenly surpassed state production. The operator’s design goal was
those at the lower shear rate. Chemists repeated the measurements to verify to prevent wax deposition at production rates
the reproducibility of this behavior. The deposition rates were entered into above 5,000 bbl [795 m3] of oil per day. Methods
equation-of-state and fluid-transport models that help engineers predict how to prevent or slow wax deposition in a pipeline
the crude oil will behave in various production scenarios.
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-
2,000 and dead-oil data proved important. Knowing
1,750 only the dead-oil WAT would have led the
operator to believe that expensive pipe-in-pipe
1,500
insulation was necessary. Calculations using the
Restart pressure, psi

1,250 lower live-oil WAT showed that more economical

1,000 wet insulation would be sufficient to prevent
wax deposition.
750
Because the oil cools as it flows through the
Maximum restart pressure
500 pipeline to the collection facility, it was essential
250
to predict the likelihood of gel formation.
Assuming a production rate of 5,000 bbl/d
0 through wet insulated pipe, the OLGA simulator
0 100 200 300 400 500 600 700 800
Fluid pressure, psi showed that live oil could flow through the
pipeline for about 20 hours before reaching
> OLGA calculation of flowline-restart pressures in 6-in. pipe. If the oil gels in its pour point—sufficient time to reach the
the flowline during a shutdown, pressure must be applied to overcome the gel stock tank.
strength and initiate flow. The maximum pressure that can safely be applied is For flow rates less than 5,000 bbl/d, the OLGA
500 psi. The simulation shows that maintaining an internal-flowline pressure simulator predicted rates at which wax deposi-
greater than 100 psi would allow engineers to safely restart the flowline.
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,
shear-thinning behavior with no gelation Dead-oil wax-deposition tests were predict a deposition rate, and schedule wax-
occurred at all pressures. Further investigation conducted in the WDFL at temperatures removal operations. For this West African crude
of gel strength involved aging live- and dead-oil between 81° and 122°F [27.2°C and 50.0°C] oil, the standard model predicted that remedi-
samples for 12 hours at the seabed temperature. (top). The results revealed a deposition-rate ation would be necessary every two weeks. With
The live-oil gel strength between 100 and 700 psi discontinuity at the higher end of the shear-rate WDFL data, the simulator predicted a much
varied from 38 to 42 Pa, and the dead-oil gel range. For each shear rate, chemists entered the lower deposition rate, increasing the time
strength at ambient pressure was more than measured deposition rates, the n-paraffin between remedial jobs to six weeks. Methods to
three times greater—142 Pa. distribution, C30+ composition and viscosity remove wax deposits include pipeline pigging,

68 Oilfield Review
hot-oil circulation and solvent treatments with
coiled tubing. 160
A vital flow-assurance question concerns the
ability of the system to restart after a production 150
shutdown. According to the operator, 500 psi was

Arrival temperature, °F
the highest pressure that could be safely applied 140
to overcome gel strength and initiate flow STO WAT
(previous page, bottom). Using live-oil gel- 130
strength data and fluid-transport parameters,
120
the OLGA simulator calculated that fluid flow Live-oil WAT
could be restored as long as the pressure in the PIP insulation
110
pipeline remained above about 100 psi. Wet insulation
This case study shows that reliance on 100
dead-oil experimental data and standard wax- 0 2,500 5,000 7,500 10,000 12,500 15,000 17,500 20,000
deposition calculations could lead operators to Oil production rate, bbl/d
make unduly conservative decisions when
160
designing production systems. In this case, flow-
assurance predictions based on live-oil data PIP insulation
140
Wet insulation
allowed the operator to save millions of dollars
in flowline costs and less frequent wax- 120

Temperature, °F
removal operations.
100
Coming Advances in Waxy-Crude STO pour point
Flow Assurance 80
Significant work is underway to continue Live-oil pour point
60
improving flow-assurance testing and fluid-
property surveillance during a field’s productive
40
life. The West Africa case study demonstrated 0 5 10 15 20 25
the benefits of performing flow-assurance Time, h
experiments with live crude oils. However, wax-
1.2
deposition testing in devices like the WDFL has
been confined to dead oils. The WDFL could be Recommended wax-removal thickness
1.0
modified to perform high-pressure tests, but
Deposit thickness, mm

consuming two liters of live oil would be 0.8 Standard n-paraffin

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 392°F 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): 295–299.
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 12–15, 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.
diffusion coefficient
WDFL n-paraffin
0.6
diffusion coefficient
0.4
0.2
Wellhead Collection
facility
0
0 0.5 1.0 1.5 2.0
Flowline distance, mi
> 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.

Autumn 2007 69
 Multipoint 30
thermocouples WDFL RealView cell

Total wax-deposition rate, mg/m2s

25 24.5 23.7
High-pressure
Shear cell flow-control
valves 20
15.5
Mechanical
15
mounting
stand 10 9.2
Variable-speed
DC motor
5 3.3

0
STO, STO, STO, Live oil, Live oil,
laminar laminar turbulent laminar turbulent

Charge
Inner
rotating
cylinder
Deposit
Outer
stationary
cylinder
Coolant
in
Discharge
> 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).
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
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 26–29, 2004.
> 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
Coolant the RealView cell. Turbulent flow in the RealView cell reduced the dead-oil
out deposition rate substantially. Live-oil deposition rates in the RealView
apparatus were even lower for both flow regimes.
Electrical
heating
cartridge
Oil
assurance have already been realized.32 composition may change as fluids are produced
Preliminary laboratory results show that live-oil from regions that were originally distant from the
Coolant testing will lead to similar improvements for waxy initial sampling point. The impact on flow
crude oils. A recent laboratory study tested the assurance may be significant.
behavior of a black waxy crude oil in a hypothet- Traditionally, engineers monitor fluid-
ical flow-assurance situation. Oilphase-DBR property evolution by periodically sampling from
scientists assumed the oil entered a 6-in. ID, separators or extracting live oil near perfora-
2-mi long pipeline at a temperature and pressure tions, and performing flow-assurance tests in the
of 170°F and 3,170 psi [21.9 MPa]. The simulated laboratory. In offshore and deepwater fields, this
oil-production rate was 5,000 bbl/d, and the approach is costly.33 With the advent of
seabed temperature was 65°F [18.3°C]. Within intelligent completions, equipped with sensors
this scenario, engineers performed deposition- that transmit downhole temperature, pressure
rate tests that compared the performance of and flow rates in real time, production moni-
STO and live oils in laminar and turbulent flow toring can be performed remotely. These
(above right). The results showed that wax completions also incorporate remote-control
deposition is slower in turbulent flow, and live-oil pumps and valves that engineers can use to
wax-deposition rates are far lower than those mitigate flow-assurance problems.
observed with STO. Chemical sensors are being developed that
The low wax-deposition rates observed with can detect fluid-composition changes. When
live oils would significantly impact the hypo- installed at strategic locations in the well
thetical flowline design and the frequency of completion and along a pipeline, the sensors will
wax-removal operations. OLGA simulations provide real-time data for monitoring of solids
demonstrated that the 2-mi flowline could be deposition, corrosion rates and rheological
constructed from foam-insulated pipe instead of properties. As a result, interventions for
pipe-in-pipe, potentially saving US $4 million. In sampling or remediation will be performed only
the context of flowline remediation, traditional when necessary.
WDFL measurements with STO indicated that Today, fluid sampling and analysis are
wax removal would be necessary every two weeks. progressing to a point at which consistent
Simulations with live-oil data predicted that wax- standards are applied along the continuum from
removal would be necessary about once per year, DFA and reservoir characterization, to sampling
providing significant operational savings. and laboratory analysis, and on to production
As a field produces, the crude-oil properties surveillance. This integrated approach will be
often change. For example, as gas condensate increasingly valuable to operators making
falls below the saturation pressure during reservoir exploration, development and
depletion, the condensate yield and the WAT may production decisions, particularly in high-risk,
fall. In a compositionally graded accumulation, remote locations. —EBN

70 Oilfield Review