Ridvan Akkurt

Saudi Aramco
Dhahran, Saudi Arabia

Focusing on Downhole Fluid Sampling

and Analysis

Martin Bowcock
BG Group
Reading, England

A new focused-sampling device allows acquisition of downhole uid samples of

John Davies
Chevron
Houston, Texas
Chris Del Campo
Sugar Land, Texas, USA
Bunker Hill
Rosharon, Texas
Sameer Joshi
Dibyatanu Kundu
Mumbai, India
Sanjay Kumar
Cairn Energy India Pty Ltd
Gurgaon, India
Michael OKeefe
Hobart, Tasmania, Australia
Magdy Samir
Aberdeen, Scotland
Jeffrey Tarvin
Cambridge, Massachusetts, USA
Peter Weinheber
Houston, Texas
Stephen Williams
Hydro
Bergen, Norway
Murat Zeybek
Al-Khobar, Saudi Arabia
For help in preparation of this article, thanks to Stephane
Briquet, David Nunez and Ricardo Vasques, Sugar Land,
Texas; Kre Otto Eriksen, Statoil ASA, Stavanger;
Noriuki Matsumoto, Sagamihara, Kanagawa, Japan; Moin
Muhammad, Edmonton, Alberta, Canada; Oliver Mullins,
Houston; Tribor Rakela, Caracas; John Sherwood, Cambridge,
England; and Dag Stensland, ENI Norge, Stavanger.
Thanks also to ConocoPhillips (UK) Ltd.
CFA (Composition Fluid Analyzer), InterACT, LFA (Live Fluid
Analyzer), MDT (Modular Formation Dynamics Tester),
MRX (Magnetic Resonance eXpert), Quicksilver Probe,
Platform Express, PVT Express and SlimXtreme are marks
of Schlumberger.

unprecedented purity, and in a fraction of the time needed with conventional sampling
technology. The method also gives superior results for downhole measurements of
formation-uid properties.

Understanding the properties of fluids contained

in a hydrocarbon reservoir requires measurements on fluid samples. Sample analysis helps
identify fluid type, estimate reserves, assess
hydrocarbon value and determine fluid
properties, so production can be optimized.
Using fluid-analysis results, oil companies decide
how to complete a well, develop a field, design
surface facilities, tie back satellite fields and
commingle production between wells.
Fluid analysis is also important for understanding the properties of formation water,
which can have significant economic impact.
Often, the most crucial goals are to identify the
corrosive properties of the water for the purpose
of selecting completion materials and to measure
scaling potential for avoiding flow-assurance
problems. In addition, log analysts want to
quantify the salinity of the water for petrophysical evaluation, and geologists and reservoir
engineers want to establish the water source for
evaluation of reservoir connectivity.
Formation-fluid samples can be acquired
using one of three main techniques. First,
wireline formation testers deployed in open hole
can acquire fluid samples and also perform downhole analysis of fluids, ensuring optimal sample
acquisition and the possibility of analyzing fluids
early in the life of the well. These testers provide
a cost-effective method of acquiring early fluid
samples, with performance now often equal to or
above that achievable with the second method,
drillstem tests (DSTs). In the past, DSTs,
typically designed to test production and investigate reservoir extent, have produced samples
with less contamination than openhole sampling.
DSTs require early planning and a well comple-

tion that can withstand production pressures,

and can cost much more than openhole
sampling, especially in offshore wells. In a third
method, samples can be acquired by wireline
tools deployed in a cased, producing well.
An important aspect of fluid sampling is
analysis of the fluids at reservoir conditions. This
helps validate sample quality during the sampling
process, but also enables the mapping of vertical
variations in fluid properties, allowing interpreters to determine zonal connectivity and
define reservoir architecture early in field life.
Uncontaminated fluid samples allow accurate
measurement of fluid properties both downhole
and at the surface.
After samples are acquired, they typically are
analyzed in laboratories, where they undergo a
series of tests depending on what the client
needs to understand. Standard analyses for
hydrocarbon samples include chemical composition to C30+, gas/oil ratio (GOR), density, viscosity, and phase properties such as saturation
pressure, bubblepoint, pour point and stability of
asphaltenes.1 Several measurements can now be
performed downhole, using optical spectroscopy
to characterize formation fluids under reservoir
conditions.2 These include density, optical
density, GOR and chemical composition to C6+.
Laboratory and downhole fluid measurements
both require pure, uncontaminated samples.
Contamination occurs when miscible drillingfluid filtrate that has invaded the formation
mixes with the formation fluid being sampled.
For instance, hydrocarbon samples are contaminated by oil-base mud (OBM) filtrate, and water
samples are contaminated by water-base mud
(WBM) filtrate.

Oileld Review

To reduce contamination during sample

collection, engineers rely mostly on increasing
the volume of fluid pumped from the reservoir by
pumping longer or at a higher rate. Downhole
analysis of contamination level can determine
when fluid flowing through the sampling-tool
flowline is clean enough to be collected.3
However, long pumping time increases rig time
and associated costs, and may increase the risk
of downhole tool sticking. Depending on the
reservoir permeability, high pumping rates can
cause the reservoir fluid to drop below saturation
pressure. If this happens, the downhole samples
will not be representative of the reservoir fluid.
In the case of unconsolidated formations, high
pumping rate may induce sand production. Also,
in settings involving high vertical permeability,
even long pumping times and increased pumping
rates do not guarantee clean samples.

Winter 2006/2007

Fluid-analysis experts have worked to understand and mitigate the effects of contamination
on samples. Some methods attempt to derive the
composition or GOR of a pure sample knowing
the composition of the OBM contaminating the
collected sample.4 However, uncertainties and
errors accompany fluid properties estimated in

this manner. Researchers have quantified the

errors caused by contamination on some
measurements. For example, the pressure at
which asphaltenes precipitate from solution in
crude oil decreases in the presence of OBM
contamination. In one case, just 1% OBM
contamination by weight caused asphaltene-

1. The phrase composition to C30+ indicates that

compounds of up to 29 carbon atoms are separately
discriminated, with the remainder combined into a
fraction indicated as C30+.
Pour point is the minimum temperature at which oil
pours or flows.
2. Fujisawa G, Betancourt S, Mullins OC, Torgersen T,
OKeefe M, Terabayashi T, Dong C and Eriksen KO:
Large Hydrocarbon Compositional Gradient Revealed
by In-Situ Optical Spectroscopy, paper SPE 89704,
presented at the SPE Annual Technical Conference
and Exhibition, Houston, September 2629, 2004.
3. Mullins OC, Schroer J and Beck GF: Real-Time
Quantification of Filtrate Contamination During Openhole
Wireline Sampling by Optical Spectroscopy,
Transactions of the SPWLA 41st Annual Logging
Symposium, Dallas, June 47, 2000, paper SS.

Mullins OC and Schroer J: Real-Time Determination of

Filtrate Contamination During Openhole Wireline
Sampling by Optical Spectroscopy, paper SPE 63071,
presented at the SPE Annual Technical Conference and
Exhibition, Dallas, October 14, 2000.
Dong C, Mullins OC, Hegeman PS, Teague R, Kurkjian A
and Elshahawi H: In-Situ Contamination Monitoring and
GOR Measurement of Formation Fluid Samples, paper
SPE 77899, presented at the SPE Asia Pacific Oil and
Gas Conference and Exhibition, Melbourne, Australia,
October 810, 2002.
4. Gozalpour F, Danesh A, Tehrani D-H, Todd AC and
Tohidi B: Predicting Reservoir Fluid Phase and
Volumetric Behaviour from Samples Contaminated with
Oil-Based Mud, paper SPE 56747, presented at the
SPE Annual Technical Conference and Exhibition,
Houston, October 36, 1999.

Power of transmitted light

Saturation
pressure

Asphaltene-precipitation
onset pressures

Contamination
(live-oil basis)
2.6 % by weight
7.6 % by weight
14.2 % by weight
19.4 % by weight

Pressure

Contamination level

> The effect of oil-base mud (OBM) filtrate contamination on asphaltene-precipitation onset pressure.
Laboratory analysis on live oils with varying amounts of added OBM filtrate shows a decrease in
asphaltene-precipitation onset pressures with increased OBM contamination. Live oils are oils that
contain dissolved gas. Asphaltene precipitation is detected by light transmittance; precipitates scatter
light and decrease transmittance. These and similar experiments show on average that for 1% by
weight OBM contamination, asphaltene onset pressure decreases by 100 to 150 psi. The 19.4%
contamination sample reached saturation pressure before it reached the asphaltene-precipitation
pressure. (Adapted from Muhammad et al, reference 5.)

Acceptable sample
Time

> Conventional formation-fluid sampling with a wireline formation tester.

The tester forces a packer to seal against the borehole wall, then presses
a probe through the mudcake and against the formation (right). Formation
fluid is blue-gray and filtrate is light brown. The probe (left) has a single
intake port. When pumping begins, fluid is highly contaminated (graph
inset), but decreases gradually with time. However, even with long
pumping times, the contamination level may not reach an acceptable limit
in some formations.

5. Muhammad M, Joshi N, Creek J and McFadden J:

Effect of Oil Based Mud Contamination on Live Fluid
Asphaltene Precipitation Pressure, presented at the
5th International Conference on Petroleum Phase
Behaviour and Fouling, Banff, Alberta, Canada,
June 1317, 2004.

6. Sherwood JD: Optimal Probes for Withdrawal of

Uncontaminated Fluid Samples, Physics of Fluids 17,
no. 8 (August 2005): 083102.
7. Tarvin JA, Gustavson G, Balkunas S and Sherwood J:
Sampling Fluid from a Two-Dimensional Porous Medium
With a Guarded Probe, submitted to Journal of
Petroleum Science and Engineering.

precipitation onset pressure to decrease by

100 to 150 psi [0.7 to 1.0 MPa] (left).5 Thus,
measurements on contaminated samples underestimate asphaltene-precipitation onset pressure,
and may negatively affect flow-assurance and
production predictions. These results emphasize
the need for extremely low-contamination samples.
A new sampling apparatus designed to reduce
filtrate contamination focuses fluid intake so
that reservoir fluid flows into one sampling line
while filtrate flows into a separate line. With this
innovative tool, mud-filtrate contamination can
be separated efficiently from formation fluid in
the early stage of the sampling process. A clean
reservoir-fluid sample can be acquired much
faster than with conventional sampling techniques. This article describes the advantages of
the new, focused-sampling tool through field
examples of hydrocarbon and water sampling
from the Gulf of Mexico, the North Sea, India and
the Middle East.
Quicker and Cleaner
To fully appreciate the advantages of the new
sampling method requires a brief overview of
conventional downhole fluid-sampling technology.
In the typical scenario, overbalanced drilling into
a permeable formation will facilitate invasion of
drilling-fluid filtrate into the formation and the
creation of filtercake on the borehole wall.
During conventional formation-fluid sampling, a
wireline formation tester deploys a packer
against the borehole wall to isolate the sample
probe from borehole fluids and hydrostatic
pressure. The probe is then pressed through the
mudcake and against the formation (left). A
successful seal connects the sampling tool with
the formation while isolating the tool flowline
from borehole fluid and pressure.
As the sampling tool withdraws fluid from the
formation through the probe, the first reservoir
fluid to enter the flowline is contaminated with
filtrate from the drilling fluid. The level of
contamination, monitored in real time by
downhole spectroscopic analyzers, decreases as
pumping continues. Depending on formation
permeability, anisotropy, amount of invasion,
formation-fluid viscosity, and pumping time, rate
and pressure drawdown, the contamination level
may or may not decrease sufficiently to allow
collection of a fluid sample that is representative
of the formation fluid. Filtrate contamination
from deeply invaded zones may continue to feed
into the sampling probe, and in cases of poorly
formed mudcake, borehole fluid may continue to
invade the formation at a relatively significant
rate. Achieving sufficiently low levels of
contamination may require pumping for

Oileld Review

With Guard

extended periods of timemany hourswhich

can be expensive in terms of rig time and
increased exposure to sticking in open hole.
Seeking ways to improve sample quality and
reduce sampling time, researchers investigated
the effects of different probe configurations.
To test the idea that focused flow into a probe
could reduce sample contamination and shorten
sampling time, a scientist at Schlumberger
Cambridge Research in England simulated flow
into modified probes.6 The modeling results
helped determine optimal probe size.
Researchers at Schlumberger-Doll Research in
Connecticut, USA, conducted 2D experiments on
laboratory models to determine the potential
benefits in sample cleanup (above).7 The
modified probes had three openings: side openings, called guard probes, drew contaminated
fluid away from the central area of the probe, and
a central opening, called a sample probe,
collected low-contamination fluid. Experimental
results indicated that cleanup with the guard
probes active proceeded much more quickly than
without, achieving lower contamination levels
with less fluid volume pumped (right).

Winter 2006/2007

> Setup and visual results of laboratory experiments simulating focused

flow. The experimental setup (top right) consisted of a 2D formation made
of glass beads, surrounded by a single oil with an optical index identical
to that of the glass beads, all held between two vertical glass plates. A
bottom portion of oil was dyed red to represent the filtrate-invaded zone.
Above this, the oil was left transparent. A sample and guard-probe
assembly at the bottom of the formation extracted fluid (inset). A camera
monitored the cleanup in the formation directly in front of the probe
assembly. After image processing, the time-lapse visual images (left) show
large differences in the area cleaned up by the sample probe alone (left)
and the sample and guard probes together (right). The sample and guard
probes clean up a large area in front of the sample probe, ensuring that
only uncontaminated fluid enters the sample probe.

Without guard
With guard

100

Dye concentration, % of maximum

Without Guard

10

20

40

60

80

100

120

140

160

180

200

Volume pumped, ml

> Contamination reduction with and without guard probes. Laboratory

measurements detected decreasing contamination levels with increasing
volume of fluid pumped, corresponding to increasing pump time. Sampling
without the guard probe (blue) never achieved contamination levels less
than 1%.

Formation and filtrate fluid

(supply side)

Sampling and conductivity

(measurement side)
Conductivity
meters

Metering
valves
Accumulators

Flowmeters

Metering
valves

Flowmeters

> Engineering experimental setup to investigate the feasibility of focused sampling. A

formation interface tester, containing a 15.5-in. diameter, 12-in. tall sandstone core, is in
contact with two fluid manifolds and a laboratory-prototype focused-sampling probe.
Simulated formation fluid is supplied at the base of the core, and simulated filtrate is
supplied in a ring around the core. Flow into and from the tester is controlled and
monitored by pumps and valves, and contamination level is calculated from electrical
conductivity measurements on the flowlines. (Adapted from Dong et al, reference 8.)

100
Guard flowline
Total flow
Sample flowline

90
80

Contamination, %

70
60
50
40
30
20
10
0
10

10

20

30

40

50

60

Elapsed time, min

> Focused-sampling experimental data. The decrease in contamination inferred from

electrical conductivity measurements on the guard and sample flowlines demonstrates
fluid cleanup in this sampling test. The contamination level in the sample flowline (green)
decreased rapidly, while the contamination level in the guard flowline (red) decreased
gradually. Summing the flow from both flowlines produces the total flow (blue), the flow
that a traditional probe would have measured. (Adapted from Dong et al, reference 8.)

Further engineering tests at Sugar Land

Technology Center in Texas extended the 2D
results in simulated formations to three
dimensions and actual rock formations. In these
experiments, a downhole probe prototype of a
new focused-sampling tool drew fluids from a
large sandstone core in a test apparatus (left).8
The 15.5-in. diameter, 12-in. tall core contained
aqueous sodium chloride [NaCl] formation fluid
and mud-filtrate fluid of different known
conductivities. Fluid flow to the guard and
sample flowlines was controlled and measured
with metering valves and high-pressure
flowmeters. Calibrated electrical conductivity
meters on the flowlines leading from the guard
probe and the sample probe recorded the
cleanup history of each sampling test. With
focused sampling, the contamination levels of
the fluid in the sample flowline decreased
rapidly, while the contamination level in the
guard flowline decreased gradually (below left).
A traditional probe would have measured the
combined flow, and would not have achieved
contamination less than 10%.
The key to acquiring such low-contamination
samples is the focusing effect achieved by the
multi-intake probe.9 This innovative design has
been implemented in the Quicksilver Probe
wireline sampling tool, a new module of the MDT
Modular Formation Dynamics Tester tool. In
some ways, the configuration of the Quicksilver
Probe module is similar to that of traditional
samplers, in that a packer seal isolates the fluidsampling zone from the borehole. However,
within the fluid-sampling zone, a cylindrical
guard probe on the periphery of the sampling
zone surrounds the innermost sampling area
(next page, left). An additional packer seal
separates the guard intake from the sample
intake. The inner and peripheral areas are
connected to separate flowlines, called the
sample and guard flowlines, respectively. Two
pumps in the tool, one above the probe and one
below, can draw fluid into the two flowlines at
different rates, and spectroscopic analyzers
determine the composition of fluid in each
flowline (next page, right). The focusing effect of
the method is somewhat analogous to the way
laterolog devices use guard electrodes to focus
current into a formation to measure resistivity.10
The Quicksilver Probe focused-sampling tool
pumps fluid from the formation through the
central and peripheral areas of the sampling
zone simultaneously. Initially, commingled contaminated fluid flows into both areas, but this

Oileld Review

Contamination level

Power cartridge
Acceptable sample
Time

Sample-bottle
module

Sample pump
module

> Formation-fluid sampling with the Quicksilver Probe focused-sampling tool. The probe
(left) has two intake ports, with the guard intake surrounding the sample intake. Packers
surround and separate these probes and seal against the borehole wall (right).
Formation fluid is blue-gray and filtrate is light brown. When pumping begins, fluid
flowing through the sample intake is highly contaminated (graph inset), but decreases
quickly with time. Soon, contamination levels are at an acceptable value.

Sample fluid
analyzer

Hydraulic module

fluid is not collected. Fluid flow is then

separated, or split, between the guard and
sample flowlines. Fluid flow into the guard intake
can be increased, and in a short time, all
contaminated fluid is drawn into the guard
flowline, allowing low-contamination formation
fluid to flow into the sample flowline. This
technique accentuates the difference in contamination level between clean and contaminated
fluid, making it easier to identify a time at which
a clean sample can be collected. Case studies
from several environments show the sample
quality that can be obtained using the new
focusing technology.
Exploring in the Gulf of Mexico
In 2004, Chevron drilled an exploration well into
the emerging Lower Tertiary play in the
deepwater Gulf of Mexico. These wells are
typically difficult to drill and complete, with
water depths down to 10,000 ft [3,000 m] and
total well depths exceeding 25,000 ft [7,600 m].
More than 20 exploration and appraisal wells
have been drilled so far in this play, and more
than half were discoveries, many with thick oil

Winter 2006/2007

columns. However, in such conditions, well tests

usually are extremely expensive, typically costing
US$ 70 million or more. For this reason DSTs are
rarely performed in this region.
Drilling in this play in Walker Ridge Block
759, Chevron and partners announced discovery
of more than 350 ft [110 m] of net-pay oil sands
in Jack 1, the first well of the Jack prospect, in
September 2004.11 The subsalt prospect is
8. Dong C, Del Campo C, Vasques R, Hegeman P and
Matsumoto N: Formation Testing Innovations for Fluid
Sampling, presented at the Deep Offshore Technology
Conference and Exhibition, Vitoria, Espirito Santo, Brazil,
November 810, 2005.
9. Del Campo C, Dong C, Vasques R, Hegeman P and
Yamate T: Advances in Fluid Sampling with Formation
Testers for Offshore Exploration, paper OTC 18201,
presented at the Offshore Technology Conference,
Houston, May 14, 2006.
10. Doll HG: The Laterolog: A New Resistivity Logging
Method with Electrodes Using an Automatic Focusing
System, Petroleum Transactions of the AIME 192 (1951):
305316.
11. ChevronTexaco Announces Discovery in Deepwater Gulf
of Mexico, http://www.chevron.com/news/press/2004/
2004-09-07.asp (accessed September 22, 2006).

Focused-sampling
probe

Guard fluid
analyzer

Guard pump
module

> The Quicksilver Probe toolstring. Fluids enter

the tool at the focused-sampling probe.
Contaminated fluids flow downward through the
guard fluid analyzer and pump. Clean fluids flow
upward through the sample fluid-analyzer and
pump modules to the sample-bottle module. The
configuration may change for different sampling
jobs. For example, the pumps may be located
upstream of the fluid analyzers for some
applications. (Adapted from Del Campo et al,
reference 9.)

approximately 270 miles [430 km] southwest of

New Orleans and 175 miles [280 km] offshore (right).
To further evaluate the prospect, Chevron
drilled a second well, Jack 2, in Walker Ridge
Block 758 to a total depth of 28,175 feet
[8,588 m]. Departing from typical procedures,
Chevron planned a well test, which would make
Jack 2 the only Lower Tertiary well ever tested in
the Gulf of Mexico. Acquiring a pure sample of
the formation fluid prior to the production test
would aid significantly in reducing the fluid
uncertainties in the test design and therefore
enhance the value of this expensive endeavor.
A unique MDT sampling toolstring configuration allowed collection of traditionally
acquired fluid samples at two stations with an
extralarge-diameter (XLD) probe, and focused
samples at two stations with the Quicksilver Probe
module.12 Real-time analysis of flowline fluid
acquired at one station with the XLD probe shows
GOR increasing but not leveling off, even after
8 hours of pumping (below left). Nevertheless,
samples were collected at 30,000 seconds.

Houston

New Orleans

T
Jack

G U L F

O F

M E X I C O

150

km
miles

150

Start pumping

200,000

5,000

10,000 15,000 20,000 25,000 30,000

Elapsed time, s

> Cleanup plot of flowline fluid acquired with a

single extralarge-diameter probe in the Chevron
Jack 2 well. The volume of fluid pumped during
sampling is shown in the top track. Real-time
analysis of optical density measured with the LFA
Live Fluid Analyzer tool leads to quantification of
the volume fraction of C6+ components,
essentially liquid hydrocarbons (second track),
and gas/oil ratio (GOR) (third track) as flowline
fluid becomes cleaner. GOR (blue) continues to
increase, indicating cleaner sampling, but does
not level off, even after 8 hours of pumping.
Laboratory analysis of samples collected at
30,000 seconds showed the contamination level
to be greater than 10%. A data-quality flag track
(bottom track) is green when data quality is high,
and brown when data quality is lower.

10

100,000
0
1.0
0.5
0

Start pumping

200,000

Volume fraction Pumpout volume, cm3

300,000

300,000
200,000
100,000
0
1.0
0.5
0

GOR

0.5

GOR

Volume fraction Pumpout volume, cm3

0
1.0

Start pumping

100,000

GOR

Volume fraction Pumpout volume, cm3

> The Lower Tertiary play in deepwater Gulf of Mexico, where Chevron discovered the Jack field
in 2004. Other wells in the Lower Tertiary play are shown as dots.

2,500 5,000 7,500 10,000 12,500 15,000

Elapsed time, s

2,500 5,000 7,500 10,000 12,500 15,000

Elapsed time, s

> Pumpout volume, volume fraction and GOR plots for sample-line (left) and guard-line (right) fluids
obtained with Quicksilver Probe focused sampling. As seen in the pumpout-volume track (top), only
the guard-line pump (red) operates from 0 to 7,340 s. Then, the sample-line pump (brown) is activated
and pumps until 11,500 s, at which time both pumps operate synchronously but at different rates.
Cleanup can be seen by the increase in GOR (blue) in the guard flowline from 0 to 7,340 seconds,
while the sample line is idle. Then, the guard pump stops and the sample-line pump starts. The GOR
seen by the sample-line LFA module increases gradually at first, and then, when flow is split at
11,500 s, the sample-line GOR increases dramatically and reaches a plateau, indicating that the fluid
is clean. The sample acquired at 14,000 s had a contamination level that was too small to measure.

Oileld Review

Winter 2006/2007

High-strength
wireline cable
High-strength wireline
dual-drum capstan

Depth, ft

Laboratory analysis later showed that these

samples had more than 10% OBM contamination.
Samples acquired with the Quicksilver Probe
module were less than 1% contaminated after
4 hours of pumping. Pumpout volume and GOR
plots from the guard and sample flowlines show
the stages of fluid cleanup (previous page,
bottom right). From 0 to 7,340 seconds, the lower
pump initiated fluid movement into the borehole
by pulling commingled fluids through the guard
flowline. Pumping early fluids through the guardline pump helps optimize sample acquisition.
Fluids that flow early are most likely to contain
mud solids that might potentially plug the pump.
If plugging is going to occur, field engineers
prefer the guard pump to plug instead of the
sample pump. Although flow will not be focused
if pumped with a single pump, samples can still
be taken if the sample pump is functioning.
Samples cannot be collected with only the guard
pump operating.
In this example, the guard pump stopped at
7,340 seconds and the sample-line pump started.
The GOR seen by the sample-line LFA Live Fluid
Analyzer module started to build. The GOR seen
by the guard-line LFA tool is ignored, since no
fluid is flowing through it.
At about 11,500 seconds, the flow splits. The
sample-line GOR increases rapidly, indicating
fast cleanup as the contaminated filtrate is
directed away from the sample probe.
Additionally, the GOR flattens immediately,
indicating that the fluid is as clean as it can be.
The GOR on the guard line is still increasing; its
fluid is still cleaning up. The sample acquired at
14,000 seconds was later analyzed in the
laboratory and found to have a contamination
level that was too small to measure.
Use of the Quicksilver Probe device allowed
high-quality fluid samples to be compared with
those of the discovery well, and provided
improved viscosity estimation compared with
samples obtained from more conventional tool
configurations. The lower contamination samples
also improved the DST design by quantifying
GOR for separator and surface-facilities requirements, predicting PVT behavior and enhancing
reservoir characterization.
The Jack 2 production test, completed in
September 2006, was the deepest successful well
test in the Gulf of Mexico.13 During the test, the
well sustained a flow rate of more than
6,000 bbl/d [950 m3/d] of crude oil from about
40% of the wells net pay. Chevron and partners
Statoil ASA and Devon Energy Corporation plan
to drill an additional appraisal well in 2007.

Standard Schlumberger
wireline unit

30 to 60 ft recommended

10,000

20,000
Weakpoint

> High-tension logging system for logging deep wells and sticky hole
conditions with heavy toolstrings. The system comprises a standard
Schlumberger wireline unit, a high-strength dual-drum capstan and
high-strength wireline cable. The capstan increases the pull that can
be exerted on the wireline, so even heavy toolstrings can be retrieved,
reducing the risk of sticking. An electrical fusible weakpoint, located
in the logging head, is normally used with high-strength cable to allow
maximum pull on the tools if needed. The weakpoint can be broken
only electrically from surface. Extremely deep wells, such as those
deeper than 28,000 ft [8,500 m], require a modified wireline unit.

Sampling at High Pressures and Temperatures

High-pressure and high-temperature (HPHT)
conditions present a challenging environment
for all aspects of drilling, logging, completing and
producing a well. HPHT wells are often deep and
require the use of the most capable rigs, which
can be costly. Wireline logging generally requires
multiple runs to obtain the necessary information, and often encounters tool sticking, so
expensive and lengthy fishing operations are
potential risks. Pipe-conveyed logging is often
not the best option because it subjects tools to
longer exposure to high temperature.
Traditional wireline logging with standard
cables may not provide enough pull to deploy
long toolstrings in deep wells, and may lack
the extra pull necessary to get a tool out if

sticking occurs. A recently developed deployment method using a high-tension cable and a
high-tension reduction unit, or capstan, can be
effective in these cases (above).14 This allows for
the rapid deployment of the logging string and
much higher overpull, which reduces the risk of
tool sticking.
12. Weinheber P and Vasques R: New Formation Tester
Probe Design for Low-Contamination Sampling,
Transactions of the SPWLA 47th Annual Logging
Symposium, Veracruz, Mexico, June 47, 2006, paper Q.
13. Chevron Announces Record Setting Well Test at Jack,
http://www.chevron.com/news/press/2006/200609-05.asp (accessed September 22, 2006).
14. Alden M, Arif F, Billingham M, Grnnerd N, Harvey S,
Richards ME and West C: Advancing Downhole
Conveyance, Oilfield Review 16, no. 3 (Autumn 2004):
3043.

11

Pumpout volume, cm3

Volume fraction

75,000

50,000

25,000
0
1.0
0.5
0

GOR, ft3/bbl

1,500

1,000

500
0
1,000

2,000

3,000

4,000

5,000

6,000

7,000

Elapsed time, s

> Monitoring sample cleanup using the Quicksilver

Probe focusing device. At 2,600 seconds, flow
was split between the sample and guard probes,
with the guard probe pumping at a higher rate
(brown, top track). GOR increased quickly from
850 ft3/bbl to around 1,500 ft3/bbl, eventually
reaching 1,550 ft3/bbl (third track). PVT Express
onsite well fluid analysis showed the oil sample
contains almost no OBM contamination and 600
to 900 ppm of H2S. Results from onshore laboratory
analysis found OBM-filtrate contamination to be
1%. The data-quality flag (bottom track) is green,
indicating high-quality data.

In the UK North Sea, HPHT wells are usually

drilled with OBM. Conventional wireline fluid
sampling in OBM-drilled wells normally requires
long pumping times. This extended time
increases the chances of tool failure related to
high temperature, and can yield unsatisfactory
samples. Shorter cleanup time and better sample
quality are needed for successful fluid sampling
in HPHT conditions. Some examples from HPHT
wells in the North Sea show how the focusing
action of the Quicksilver Probe tool reduces
cleanup time, minimizing tool exposure to high
temperatures, improving sample quality and
reducing the risk of tool sticking.
In one HPHT example from the central North
Sea, ConocoPhillips (UK) Ltd applied new and
proven technologies to overcome difficulties in
15. A safety sample is a sample that may be less than ideal,
for example, one that is somewhat contaminated, but it
is acquired anyway in case the tool fails and no further
sampling is possible.
16. A single-phase sample bottle maintains the single-phase
nature of a fluid sample as it is brought to surface.

12

this hostile environment. Experience had shown

that in addition to high temperatures and
pressures, wells in the area were prone to sticky
hole conditions; the effects of depth, filtercake
properties, hydrostatic overbalance and wellbore
tortuosity combined to hinder wireline-conveyed
reservoir evaluation. Obtaining any data under
these conditions required assuring the drilling
team that appropriate steps would be taken to
reduce the risk of logging tools becoming stuck in
the hole. Studies were conducted to confirm that
any hydrocarbon fluids pumped into the borehole
during fluid analysis would not destabilize the
mud column. Pipe-conveyed wireline logging was
ruled out for safety reasons.
ConocoPhillips (UK) Ltd had successfully
logged other challenging HPHT wells in the
region with a high-tension logging package. In
these cases, wireline logging tools were lowered
into and raised from the wells using high-strength
cable and a capstan. The capstan, placed between
the drill floor and the wireline unit, increases the
wireline pull from 9,700 to 15,500 lbm [4,400 to
7,030 kg], ensuring that even long, heavy toolstrings can be retrieved.
In this near-field exploration well,
ConocoPhillips (UK) Ltd planned to log the well
and acquire pressures to determine mud weights
for drilling deeper and to characterize the
formation fluid. Sampling would not be required
in the primary reservoir, because the fluids in the
nearby producing fault blocks were known.
Bottomhole temperatures were expected to
reach 365F [185C] and formation pressures
could exceed 14,000 psi [97 MPa]. SlimXtreme
slimhole HPHT logging tools, including the new
analog borehole seismic tool, would be run on
high-strength cable with the capstan unit.
The well encountered an unanticipated
secondary reservoir above the primary target,
introducing uncertainty into the understanding of
fluid properties. The logging program was
immediately modified to include fluid sampling in
these newly discovered zones. A Quicksilver Probe
module was readied to run in the hole and a PVT
Express onsite well fluid analysis system was
installed on the platform to perform surface
analysis of samples collected from the four
new zones.
In the first sand sampled using the
Quicksilver Probe focusing device, the tool
operated smoothly, and flow through the tool
began in commingled mode. Flow was split
between the sampling and guard probes after
2,600 seconds of pumping, giving rise to an
abrupt GOR increase from 850 ft3/bbl

[150 m3/m3] to around 1,500 ft3/bbl [270 m3/m3]

(left). GOR leveled off at 1,550 ft3/bbl
[279 m3/m3] and remained high while samples
were collected. Initial wellsite fluid analysis
showed high-purity oil. Final onshore laboratory
results found contamination to be 1%.
PVT Express fluid-analysis experts on site
identified 600 to 900 ppm of hydrogen sulfide
[H2S] in the shallowest sand layer, which would
be incompatible with the completion design.
These levels of H2S in the first sandstone led the
ConocoPhillips (UK) Ltd subsurface team to
upgrade their scrutiny of the other unexplored
layers. Additional tool runs were scheduled for
the remaining sandstone intervals in the 8-in.
section, including many more sampling stations,
assuming the Quicksilver Probe module could
withstand the high temperatures.
At another station, Quicksilver Probe operation began and remained with the tool bypass
valve in the open position. This means that the
guard and sample flowlines were hydraulically
connected inside the tool, mimicking traditional,
single-probe sampling. Pumping continued for
more than 14,000 secondsabout 4 hoursat
which point samples were collected, because the
fluid was not getting any cleaner (next page,
top). Wellsite analysis determined contamination
to be 22%, which was confirmed by the onshore
laboratory result of 23% three weeks later.
Before moving away from this sampling
station, the field engineer managed to close the
bypass valve and establish focused flow. Fluid
flow split into the guard and sampling flowlines,
GOR increased dramatically, and contamination
decreased. PVT Express onsite analysis indicated
that contamination levels fell from 22% to 1.5%.
The fluid samples collected here showed 1%
contamination in later laboratory analysis.
Normal levels of H2S were found here and in the
remaining layers. In all, 27 fluid samples were
acquired with good-quality results in every layer.
The well was subsequently completed,
perforating only those layers that had been shown
to have low H2S levels compatible with the tubing
metallurgy. Being able to acquire a suite of
uncontaminated downhole samples as part of a
rapidly evolving logging program was vital to the
success of the development of this secondary
reservoir. If high-quality samples had not been
taken and the well completion had proceeded
without this data, high concentrations of H2S
would have damaged the production tubing and
entered the production facility. Mitigating that
would have required shutting in production and
performing a costly well workover to identify and
shut off the H2S-prone zone.

Oileld Review

Winter 2006/2007

25,000
0
1.0
0.5

50,000

25,000

GOR, ft3/bbl

0
1,500
1,000

500

Start pumping

50,000

Volume fraction Pumpout volume, cm3

Start pumping

Volume fraction Pumpout volume, cm3

GOR, ft3/bbl

0
1.0
0.5
0
1,500
1,000

500

0
0

2,500

5,000

7,500 10,000 12,500 15,000

2,500

Elapsed time, s

5,000

7,500 10,000 12,500 15,000

Elapsed time, s

40,000

Start pumping

Contamination continued to decrease, and

when GOR leveled off, a single-phase sample
bottle was opened, filled and retrieved to
surface.16 PVT Express analysis on the rig
quantified extremely low contamination,
indicating that the sample was sufficiently pure,
and the tool could be redeployed to the next
deeper and hotter level. Independent onshore
laboratory testing conducted a few weeks later
detected no contamination in this sample.
Three low-contamination samples were
successfully acquired at the next hydrocarbonbearing zone, but after that, mud-check valves in
the sample flowline started to show signs of
plugging. However, the deepest and hottest zone
remained to be sampled. There, at the watersampling station, slugs of borehole mud and OBM
filtrate were detected by the LFA module,
indicating that unexpected fluid movement was
occurring through the fluid-exit port. After some
time, this movement of fluid from the borehole
had cleared the mud-check valve, and
synchronized pumping to the guard and sample
lines proceeded normallydespite the 361F
[183C] bottomhole temperatureallowing
acquisition of formation-water samples.

Pumpout volume, cm3

> Quicksilver Probe operation in a ConocoPhillips (UK) Ltd North Sea well, with the tool bypass valve
in the open position (left). With the guard and sample flowlines hydraulically connected inside the
tool, the effect is the same as conventional single-probe sampling. The pumpout-volume track (top)
shows only the sample pump operating (blue). Cleanup is gradual, as seen by the slow increase of the
GOR with time (third track). After more than 14,000 seconds, the fluid was not getting any cleaner, so
samples were collected. According to PVT Express wellsite analysis, contamination was 22%. After
the field engineer closed the bypass valve (right), fluid flow was split into the guard and sampling
flowlines at around 15,500 seconds. Both the sample-line pump (blue) and the guard-line pump
(brown) were pumping (top track), with the guard-line pump operating at a higher rate. GOR (third
track) jumped to about 1,500 ft3/bbl, indicating a reduction in contamination. Onsite analysis with the
PVT Express system quantified a contamination drop from 22% to 1.5%.

20,000

GOR

HPHT Sampling Instead of DST

BG, drilling in another HPHT area of the central
North Sea, made a discovery with multiple
hydrocarbon-bearing zones. This exploration well
was designed not to have a DST, saving costs on
two levels. First, a DST in this region would have
cost US$ 10 to 20 million. Second, additional
savings came from installing a less expensive
completion. A DST would require heavier 978-in.
casing to withstand the pressures of the well test,
and a different well-test tree for supporting the
well-test equipment, totaling an additional
US$ 4 million. Also, producing no reservoir fluids
to surface would avoid environmental risks.
Since no DST would be run, it was crucial to
acquire high-purity samples by wireline. To allow
real-time analysis of formation fluids, the
Quicksilver Probe tool was configured with LFA
and CFA Composition Fluid Analyzer modules on
the sample flowline. A PVT Express system
installed on the rig analyzed contamination at
the wellsite. Shore-based experts were able to
participate in logging and sample analysis in real
time through the InterACT real-time monitoring
and data delivery system. By confirming sample
purity at the wellsite, engineers would know if
the quality of the acquired sample was adequate,
or if a new sample was required. BG fluid experts
were hoping for samples with less than 5% OBM
contamination. In addition to hydrocarbon
samples, the tool would acquire water samples if
it could sustain the high temperatures deeper in
the reservoir. Pressures were anticipated to be at
least 13,000 psi [90 MPa], and temperature was
expected to surpass the 350F [177C] stated
limit of LFA and CFA operability.
In the first and shallowest hydrocarbon
interval sampled, the temperature was already
340F [171C]. Quicksilver Probe operation
proceeded normally, starting with the guard
probe and sample probe connected through the
inner bypass valve. The upper pump was used to
pump fluid through the sample line. The flow was
split into guard and sample lines after
3,050 seconds of pumping, at which time a jump
in GOR on the LFA plot indicated a significant
decrease in contamination of the fluid in the
sample line (right). Less than two minutes later,
when contamination reached an estimated 10%,
the one-gallon sample bottle opened to collect a
safety samplea standard practice in difficult
wells.15 This proved prudent, because soon
afterward, the sampling-flowline pump stalled,
but started again.

2,500

5,000

7,500

10,000

12,500

Elapsed time, s

> Quicksilver Probe operation in BGs HPHT North

Sea well. Before 3,050 s, fluid flowed through the
sample-line pump (blue, top track). At 3,050 s, the
guard-line pump (brown) and sample-line pump
operated synchronously, with the guard-line pump
operating at a higher rate. This split the flow,
giving rise to an abrupt increase in GOR (blue) at
3,050 s (second track). The sample acquired at
8,700 s was found to contain no contamination.

13

Strain-gauge pressure

1.0
0 0
6

LFA
Optical Channels

Volume
Fraction

Pressure

Quartz-gauge pressure

1,000

2,000

3,000

4,000

Time, s

> Detecting OBM contamination in water samples using the LFA module in the BG HPHT North Sea well.
The top track shows quartz-gauge pressure and strain-gauge pressure along with unscaled resistivity
(pink) and pump strokes (blue and green). The second track shows volume fraction of C6+ components,
indicating OBM-filtrate contamination (green), OBM and solids (red) and water (blue). At 2,600 s, soon
after the guard and sample pumps start to pump synchronously, the sample line receives formation
water (blue). The third track contains the 10 LFA optical channels. Channel 0 (black) detects methane.
Channels 6 and 9 (darker blues) detect water. The volume-fraction track detects water (blue) when LFA
channels 6 and 9 reach high amplitudes.

PA

T
IS

C H I N A
NEP
AL

Rajasthan

Barmer Basin
BANGLADESH

100

0
0

km

500
miles

SRI LANKA
500

Contamination
weight %, 39.5

Mole compositions

10

1.0

0.10

Uncontaminated sample, WL
Contaminated sample, 1035 Well A
Well 1, Quicksilver Probe sample
0.01
en de ne ne ne ne es ne es ne ne ne ne es es es es es es es es es es es es es
og xi ha pa ta ta an ze an ue ze le ze an an an an an an an an an an an an an
dr dio et Pro -Bu Pen Hex Ben ept Tol ben o-xy ben dec dec dec dec dec icos icos cos cos cos ont ont ont
M
n in Tri ta ta na ne Tr nta pta na iac iac iac
H
Hy on
lh
p o
n
r
r
r
o
hy Ort i-Me U
rb
Pe He N ent Trit ntat
Et
Pe He N He
Ca
Tr
H
Pe

> Composition from laboratory analysis of viscous-oil samples acquired for Cairn Energy India in
the Bhagyam field in the Barmer basin. Samples, all from the same location, were collected using
a monophasic wireline (WL) sampler in cased hole (green), a conventional wireline formation tester
in open hole (blue) and Quicksilver Probe focused sampling in open hole (red). The conventional
wireline formation tester yielded a contaminated sample, while the openhole sample acquired using
the focusing method compared favorably with the one obtained in cased hole.

14

Determining OBM contamination level of

water samples downhole relied on interpreting
data from the color channels of the LFA module
(left).17 Borehole mud, OBM filtrate and
formation water each have distinctive signatures
in the visible and near-infrared frequency ranges
measured by the tool. The two water samples
collected at this level contained some OBM
filtrate, but this was not a problem, because the
OBM is immiscible in formation water.
BG estimates savings of up to US$ 24 million
by using the Quicksilver Probe focused-sampling
method instead of a drillstem test to acquire
zero-contamination samples.
Sampling Viscous Oils
Viscous oils can be especially difficult to sample
using traditional sampling technology. With its
relatively lower viscosity, OBM filtrate flows
preferentially to sampling devices, increasing
sample contamination and often leaving highviscosity formation fluids in the formation.
Cairn Energy India Pty Ltd experienced such
problems acquiring oil samples in their Bhagyam
field in northwest India. The Bhagyam field is
one of 19 fields in the Barmer basin tapping the
high-permeability Fatehgarh sandstone. Oil
reserves in the reservoir are currently estimated
at 1.5 billion bbl [240 million m3].18 Oil properties
vary from field to field within the basin, and oils
within the Bhagyam field exhibit compositional
grading from crest to oil/water contact. With a
better understanding of oil properties, Cairn
plans to optimize field development and surface
facility design.
Bhagyam oils have high wax content, giving
them high pour point and high viscosity at
reservoir temperature. Acquiring representative,
PVT-quality samples has been a challenge.19
Before the arrival of focused-sampling technology in India, most samples acquired by
Schlumberger and other service companies using
traditional openhole formation testers were too
contaminated to yield correct PVT properties
during laboratory analysis.20
To obtain contamination-free samples, Cairn
had resorted to collecting samples from cased
wells using monophasic wireline-deployed
samplers. Samples acquired in this way may have
low levels of contamination, but can be collected
only after the well has been completed.
In a campaign designed to improve sample
quality, the Quicksilver Probe device collected
samples in two Bhagyam wells.21 Of the
18 samples acquired, 15 were of PVT quality. Six
of these showed no contamination. One such

Oileld Review

sample was analyzed for chemical composition

and compared with a sample acquired at the
same location by a traditional openhole formation
tester and one obtained in cased hole (previous
page, bottom). The traditional openhole method
produced a sample that was clearly contaminated
with several OBM components. The sample
acquired in cased hole showed an overall
composition that was similar to the uncontaminated Quicksilver Probe sample, with small
concentration variations in a few components.
Following the successful acquisition of
contamination-free samples and the availability
of detailed fluid PVT data, Cairn has better fluidproperty data for carrying out field-development
studies that involve reserves estimation,
facilities design and flow assurance. These
studies will make a significant contribution to
production at Bhagyam and demonstrate the
potential for improved field-development
studies worldwide.

indicate incomplete sweep. The Quicksilver

Probe module identified and sampled fluids at
four stations in the gas zone and one station in
the oil column (below). All zones show the
characteristic increase in GOR when flow
through the guard line is split from the sample
line. Several gas samples were captured with no

Caliper
4

in.

Resistivity
12

Gamma Ray
Formation Pressure, psi

OBM-filtrate contamination and with no mobile

oil, indicating highly efficient recovery.
In the oil zone, the GOR measured by the LFA
module was within 1% of the GOR already known
for the field. The pumping time required for a
clean oil sample in this zone was about 1,600
seconds, roughly one-third of the time normally

gAPI 200

Depth, ft

0.2

ohm.m

Density
20,000

Drawdown Mobility
0.2

mD/cP

1.95

g/cm3

2.95
100

Neutron Porosity

20,000 0.45

vol/vol -0.15

50

X,X90
X,X00

0
30,000
20,000
10,000
0
5,000

10,000

X,X10
X,X20
X,X30

5
4.5
4
3.5

X,X40

3
2.5

Sample line

X,X50

1.5
1

Sampling in Mature Fields

The Quicksilver Probe device is also a valuable
tool for evaluating the efficiency of hydrocarbon
recovery in mature fields. Examples from
complex, mature fields in the Middle East show
how focused sampling acquired pure samples in
high- and low-permeability formations and
helped assess gas-sweep efficiency.
The first well is in a reservoir that is under
gas-cap expansion drive and water drive.
Recently, several evaluation wells were drilled to
monitor sweep efficiency. The wells were drilled
with OBM and logged with the Platform Express
integrated wireline logging tool and nuclear
magnetic resonance tools for openhole analysis.
The Quicksilver Probe module of the MDT tool
was used to collect fluid samples.
The objective was to evaluate the efficiency of
the gas-cap expansion in the main reservoira
heterogeneous sandstone formation with permeability exceeding 1 darcy. Although low oil
saturations from MRX Magnetic Resonance
eXpert measurements indicated highly efficient
sweep of the oil by expansion of gas, the
formation tester was run to confirm that there
was no mobile oil in the swept zone. The
identification of any remaining mobile oil would
17. Betancourt S, Fujisawa G, Mullins OC, Carnegie A,
Dong D, Kurkjian A, Eriksen KO, Haggag M, Jaramillo AR
and Terabayashi H: Analyzing Hydrocarbons in the
Borehole, Oilfield Review 15, no. 3 (Autumn 2003):
5461.
18. Wireline Sampling Technology Enables Fluid Sampling
Without Contamination, JPT 5, no. 9 (September 2006):
32, 34.

Winter 2006/2007

X,X60

Guard line

0.5
0
5,000

X,X70
X,X80
X,X90
X,X00

10,000

100

50

0
40,000
20,000
0
5,000

10,000

X,X10
750

X,X20
500

X,X30
250

X,X40
0
1,000

2,000

> Pressures and sampling results from a high-permeability zone in a complex Middle East field.
Formation pressures appear in Track 1, with gas identified as orange circles and oil as green circles.
Gamma ray and caliper appear in the depth track. Track 3 contains resistivity curves and drawdown
mobility (circles). Track 4 plots density and neutron porosity. Focused sampling stations with the
Quicksilver Probe module are indicated with small probe insets. At the top sampling station, gas was
sampled from a high-permeability zone. Downhole fluid analysis at this station (top right) shows
volume fraction and GOR values from the CFA Composition Fluid Analyzer. In the volume-fraction plot,
yellow, red and green indicate C1, C2 to C5, and C6+, respectively. GOR values (magenta) point to
sampling-line cleanup. At the third station, optical densities from the LFA Live Fluid Analyzer in the
sample line (dark blue) are greater than those in the guard line (light blue), showing cleaner fluid in
the sample line. Analysis of CFA results at the same station displays similar results. At the fourth
station, in a high-permeability oil zone, LFA measurements of GOR detect cleanup in the sample
flowline (bottom right).

19. PVT-quality samples are those that have sufficiently low

contamination, such that PVT properties measured in the
laboratory correspond to those of an uncontaminated
sample. The maximum allowable contamination varies
by company and laboratory. A general rule is 7%
contamination for this basin.
20. Alboudwarej H, Felix J, Taylor S, Badry R, Bremner C,
Brough B, Skeates C, Baker A, Palmer D, Pattison K,
Beshry M, Krawchuk P, Brown G, Calvo R,
Caas Triana JA, Hathcock R, Koerner K, Hughes T,

Kundu D, Lpez de Crdenas J and West C:

Highlighting Heavy Oil, Oilfield Review 18, no. 2
(Summer 2006): 3453.
21. Kumar S and Kundu D: Fluid Sampling in Oil Based
Mud Environment: Quicksilver Probe Aids Acquiring
Contamination Free Samples in a Challenging
Environment, to be presented at Petrotech 2007, the
7th International Oil & Gas Exhibition, New Delhi, India,
January 1519, 2007.

15

800

750
GOR, ft3/bbl

Split flowlines

Quicksilver Probe sampling

700

Traditional probe sampling

650

600
0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

5,000

Time, s

> Comparison of pumping times to acquire clean oil samples in a Middle East field using
the Quicksilver Probe module and a traditional probe. In this high-permeability
sandstone, it took the Quicksilver Probe tool (red) only about 1,600 seconds of pumping
time to draw low-contamination fluid into the sampling line, while the traditional probe
(blue) pumped about three times as long to obtain a low-contamination sample.

required for this zone in other wells (above).

Laboratory analysis determined that contamination in the oil sample acquired by the
Quicksilver Probe tool was too small to measure.
In a second Middle East example, an evaluation well penetrated six reservoirs, including a
discovery. The objectives were to obtain pressure
profiles, identify fluids downhole and acquire
clean fluid samples. In addition, the operator
wanted to establish flow from low-porosity zones
that previously had not been directly tested for
their potential producibility. The operator
selected the Quicksilver Probe device because its
large probe area was better equipped than
conventional large-diameter probes to establish
flow and obtain samples from the low-porosity,
low-permeability formations.
The well was drilled with OBM, and caliper
data showed good hole condition. Fluid sampling
points were selected using free-fluid porosity
readings from the nuclear magnetic resonance
log. On the first pressure and sampling run, the
MDT tool, deployed on wireline, acquired
pressure profiles using the large-diameter probe.
On the second descent, the Quicksilver Probe
module and sampling units were run on drillpipe,
and sampled at five stations (next page, top).
At the second sampling station, downhole
fluid analysis identified oil in a previously
untapped interval. Fluid mobility was so low in

16

this tight zone that a pressure measurement

could not be acquired. However, the LFA module
in the Quicksilver Probe tool successfully
monitored formation-fluid cleanup and found
a low GOR for the liquid, leading to additional
oil reserves.
Downhole Fluid Analysis
In most reservoirs, fluid composition varies with
location in the reservoir. Fluids may exhibit
gradations caused by gravity or biodegradation,
or they may be segregated by structural or
stratigraphic compartmentalization. One way to
characterize these variations is to collect
samples for surface analysis. Another way is to
analyze fluids downhole, without bringing them
to surface. Downhole fluid analysis (DFA) is
emerging as a powerful technique to characterize fluids downhole. DFA helps determine the
best intervals for sample collection, if necessary.
Analyzing fluid composition while the tool is still
in the hole also allows more detailed fluid
characterization, because interpreters can
modify the fluid-scanning program in real time to
investigate unexpected results.22
The ability of the Quicksilver Probe module to
supply uncontaminated fluids ensures optimal
DFA results, and the faster cleanup time allows
several DFA fluid-scanning stations to be
conducted efficiently without the long station
times associated with conventional sampling. A
combination of DFA and sample collection

helped a Norwegian operator understand

fluids in a well drilled on the Norwegian
Continental Shelf.23
The well was drilled as a final appraisal
before development of an oil field. Because of
environmental restrictions, a production test was
not planned, so it was critical to obtain uncontaminated samples and fully characterize fluid
variations within the reservoir. The fluid analysis
would be used in the material selection of subsea
pipeline and surface facilities, process design
and production planning. Because of the high
priority to capture representative hydrocarbon
samples without miscible contamination, the
well was drilled with water-base mud (WBM).
The Quicksilver Probe tool was run in the
1214-in. and 812-in. sections to collect samples of
gas condensate, oil and formation water, and
filled 19 sample chambers from many levels. An
example from one of the more challenging zones,
sampling oil in a relatively tight zone with
mobility of 17 mD/cP, shows how the focusing
technology results in an uncontaminated sample.
Fluid cleanup began with commingled flow
first through the guard flowline, then through the
sample flowline. After 1,300 seconds, flow is split
and focusing is achieved by increasing the flow
rate in the guard probe (next page, bottom). The
real-time GOR detected by the CFA module
stabilized at around 2,300 seconds, indicating
that the fluid was clean. However, pumping
continued, and a sample was acquired at 2,800
seconds. The spikes in the GOR curve indicate
the presence of produced fines from the
formation, confirmed later when the sample was
analyzed at surface. Wellsite analysis showed
some sand in the samples, but no detectable
level of WBM filtrate.
In the same well, the focusing method created
optimal conditions for DFA. The spectroscopic
analyzers that indicate when fluid in the flowline
is pure enough to sample also characterized the
fluid composition in terms of three component
groups: methane (C1), ethane to pentane (C2
to C5), and hexane and heavier (C6+). This allows
in-situ compositional analysis without collecting
a sample and retrieving it to surface.
22. Fujisawa et al, reference 2.
Mullins OC, Fujisawa G, Elshahawi H and Hashem M:
Coarse and Ultra-Fine Scale Compartmentalization by
Downhole Fluid Analysis, paper IPTC 10034, presented
at the International Petroleum Technology Conference,
Doha, Qatar, November 2123, 2005.
23. OKeefe M, Eriksen KO, Williams S, Stensland D and
Vasques R: Focused Sampling of Reservoir Fluids
Achieves Undetectable Levels of Contamination, paper
SPE 101084, presented at the SPE Asia Pacific Oil and
Gas Conference and Exhibition, Adelaide, Australia,
September 1113, 2006.

Oileld Review

Depth, ft

Porosity
Formation Pressure, psi

Drawdown Mobility
0.01

mD/cP

10,000

0.5

vol/vol

Oil
Water
1.0
0.5
0.0
0

250

500
0

5,000

10,000

15,000

40
30

X,300

20

10
0
150
125
100
75
50
25
0

4,000

6,000

8,000

10,000

1.0
0.5
0.0
300

Y,100

200

100

0
0

2,500

5,000

7,500

10,000

12,500

> Sampling a new discovery in a mature Middle East field. This evaluation well penetrated six reservoirs, including a discovery. In addition
to obtaining pressure samples, the operating company performed tests in bypassed low-porosity zones. Formation pressures appear in
Track 1, with oil identified as green circles and water as blue circles. Open circles indicate pressure measurements that do not fall on any
gradient. Stars are pressures measured with the Quicksilver Probe tool. Track 2 contains drawdown mobility. Track 3 plots porosity and
pore-fluid content with red for oil and blue for water. The second sampling station, at X,300 ft, was a discovery. LFA volume fraction and
GOR results are plotted to the right of the porosity track (top right). Pump rates and GOR values for the third station are also shown to the
right of the porosity track (middle right). A low-contamination sample was also acquired at the fifth station, at Y,100 ftthe first time oil had
flowed from this low-porosity formation. The GOR from this interval (bottom right) was found to be 250 ft3/bbl.

< Fluid cleanup in an oil well offshore Norway. Quicksilver

Probe tool operation began with commingled flow through
the guard flowline, as seen by the increase in guard-flowline pumpout flow rate (light green, top track), then through
the sample flowline (dark green, top track). After 1,300 s,
flow is split and focusing is achieved by increasing the
pumping rate in the guard probe. The GOR (bottom track)
responds by stabilizing at around 2,300 s, indicating that
low-contamination fluid is flowing through the sample
flowline. Sample flowline GOR is red, and guard flowline
GOR is blue. A sample was acquired at 2,800 s, and was
found to contain no detectable WBM. The spikes in the GOR
curve indicate the presence of produced fines from the
formation sand. (Adapted from OKeefe et al, reference 23.)

Flow rate, cm3/s

15

10

Guard pump
Sample pump
5

GOR, ft3/bbl

0
1,000

500

Sample flowline
Guard flowline
0
500

1,000

1,500

2,000

2,500

3,000

Elapsed time, s

Winter 2006/2007

17

Depth = X,X45.5 m
GOR, m3/m3 Weight%

GOR, m3/m3 Weight%

Depth = X,X46.0 m
100
50
0
5,000
2,500

100
50
0
5,000,000
2,500,000
0.75

Water
fraction

Water
fraction

0.75
0.50
0.25
0

Data-quality
flag

1,100

1,200

1,300

1,400

1,500

1,600

1,700

1,800

0.50
0.25
0

Data-quality
flag

700

800

900

1,000

Elapsed time, s
C1 weight%

1,100

1,200

1,300

1,400

1,500

1,600

Elapsed time, s

C2 to C5 weight%

C6+ weight%

C1 weight%

C2 to C5 weight%

C1

C2 to C5

C6+

GOR

C1

C2 to C5

8.9 weight%

0.1 weight%

90.9 weight%

133 m3/m3

91.0 weight%

2.7 weight%

C6+
4.6 weight%

C6+ weight%
GOR
56,602 m3/m3

> Downhole fluid analysis to identify the gas/oil contact between two fluid-scanning stations separated by 0.5 m. Measurements with the CFA module
found the pure fluid at X,X46.0 m (left) to be a black oil. The composition track (top) indicates 91 wt% C6+ (green) and 8.9 wt% C1 (yellow), with a GOR of
133 m3/m3. At the station 0.5 m higher (right), the compositional analysis yields C1 content (yellow) of 91 wt%, and C6+ (green) less than 5 wt%, indicating
a dry gas with a GOR of 56,602 m3/m3. (Adapted from OKeefe et al, reference 23.)

Two DFA fluid-scanning stations straddled

the gas/oil contact, only 0.5 m [1.6 ft] apart
(above). The lower station at X,X46.0 m indicated
a black oil with apparent density of 0.868 g/cm3
and a low GOR of 133 m3/m3 [740 ft3/bbl]. Half a
meter higher, at X,X45.5 m, the fluid composition
shows a dry gas with apparent density of
0.126 g/cm3 and C1 content in excess of 91 wt%.
The DFA measurements defined the gas/oil
contact within 0.5 m, which was a higher
accuracy than could be achieved using pretest
pressure gradients in this well. The focusing
capability of the Quicksilver Probe tool ensured
that the fluids being analyzed were representative of reservoir fluids, adding confidence to
the DFA results. Similar measurements at 15
additional DFA stations helped quantify reservoirfluid composition and delineate fluid contacts.
Collecting Formation Water Free
of WBM Contamination
In another example from offshore Norway,
focused sampling helped acquire formationwater samples in an exploration well drilled with
WBM. In this well, water characterization was a
key factor in reservoir description and economic
evaluation. The operator, taking special care to
study water composition at different times in the
life of the well, collected water samples using the
Quicksilver Probe tool, then also with a dualpacker probe in casing, and finally with a
drillstem test.24
24. OKeefe et al, reference 23.
25. OKeefe et al, reference 23.

18

To facilitate quantification of WBM filtrate

contamination, tritium, a naturally occurring
isotope of hydrogen, was added to the WBM as a
tracer. Formation waters do not contain tritium
in measurable amounts, so tritium levels
detected by laboratory testing could be easily
converted to contamination levels.
The first sample was taken with the
Quicksilver Probe focused-flow tool. Flow was
split after 18,700 seconds, and the sample was
collected after 24,960 seconds of pumping.
Laboratory analysis of tritium content showed
the sample to have 0% contamination.
The well was then cased with a 7-in. liner and
perforated over the zone of interest. A wireline
formation tester was run inside the liner, along
with inflatable dual packers to isolate the flow
interval. The increased flow area provided by
the packers would minimize the drawdown
required to extract samples and so reduce the
risk of tool sticking.
After a long cleanup time24 hoursduring
which 1,700 liters [450 galUS] of fluid were
pumped from the formation into the borehole, the
formation tester collected two samples. Laboratory
analysis indicated that the samples contained
elevated concentrations of tritium, potassium,
calcium and bromide, indicative of contamination
by completion brine and mud filtrate.
The operating company then performed a
DST to test a gas zone above the water zone.
Water flowing with the gas was collected for
analysis, but was found to be heavily contaminated
with completion brine and also contained 46%
hydrate inhibitor.

The Quicksilver Probe samples proved to be

the purest water samples ever collected from the
field, surpassing the quality obtainable from
conventional-probe sampling, cased-hole sampling
or a DST. Analysis of the samples revealed
unexpected compositional characteristics that
were difficult to believe at first, but further
analysis of core and logs corroborated the new
water-composition results.25
In another water-sampling example from
offshore Norway, both focused and conventional
methods applied to the same formation helped
compare cleanup performance. This exploration
well was drilled with water-base potassium
chloride [KCl] drilling fluid, adding difficulty to
the water-sampling program. Because the
formation water and the WBM had similar optical
properties, real-time qualification of contamination relied not on spectroscopic measurements
but on resistivity differences. For quantitative
determination of contamination levels, the
concentration of potassium in the sample,
corrected by subtracting the level assumed present
in the formation water, was divided by the known
concentration in the WBM filtrate at each depth.
In the first sampling sequence, samples were
collected at three times during focused flow, at
1,050 seconds, 7,050 seconds and 7,800 seconds,
and showed 8.35%, 0.02% and 0% contamination,
respectively. Temporarily switching off the
guard pump shows the corresponding effect
on contamination (next page, top). An additional sample collected at 1,550 seconds, after
the guard pump had been stopped, yielded
33.4% contamination.

Oileld Review

0.032

X16

Quartz-gauge
pressure

X14

0.031

X10

Resistivity
X08

0.030

Guard-pump
motor speed

X06

Sample-pump
motor speed

X04

Resistivity, ohm.m

Pressure, bar

X12

0.029

X02
X00
400

600

800

1,000

1,200

1,400

1,600

1,800

0.028
2,000

Time, s

> Focused sampling of formation water in a WBM-drilled well offshore

Norway. Because the formation water and the WBM had similar optical
properties, contamination was determined by changes in resistivity (blue),
which leveled with time. This portion of the sampling log highlights the
increase in fluid resistivity when the guard-line pump (black) is temporarily
switched off, operating only the sample-line pump (green). Sample-flowline
pressure is shown in red. (Adapted from OKeefe et al, reference 23.)

X50

0.060

X45
X40

Pressure, bar

X35
X30

0.050

X25

Resistivity, ohm.m

0.055

Sample-pump motor speed

X20

0.045

X15
X10
0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

0.040
8,000

Time, s

> Conventional sampling of formation water in a WBM-drilled well offshore

Norway, for comparison with the results of the previous figure (above).
The resistivity sensor was coated with mud initially, but began to respond
(blue) partway through the sampling program. Resistivity never leveled off,
indicating that water samples were still contaminated. Samples acquired
at this station after the same cleanup times as the three samples in the
focused sampling sequence showed contamination levels of 26.2%, 8.6%
and 8.2%. (Adapted from OKeefe et al, reference 23.)

For the next run, the field engineer shifted

the toolstring 3.5 m [11.5 ft] higher in the same
formation and used a conventional single-probe
tool with a large-diameter packer to sample
formation fluid without focusing (above).
Sampling at this station was designed to allow

Winter 2006/2007

direct comparison between the performance of

the focused probe and the conventional probe.
Three sample bottles were filled at this station
after the same cleanup times as the three
samples in the focused-sampling sequence. The
conventionally acquired samples showed con-

tamination levels of 26.2%, 8.6% and 8.2%. Not

only were the focused samples cleaner at every
time comparison, but the focused sample
acquired at 1,050 seconds (17.5 minutes) was
cleaner than the conventional sample after
7,050 seconds (2 hours).
Focusing on the Future
The focused-sampling capability of the Quicksilver
Probe tool provides higher purity fluid samples in
much less time than traditional sampling tools.
Benefits include higher quality fluid-property
data, access to accurate fluid information earlier
in the reservoir-characterization process, reduced
risk of tool sticking, enhanced capabilities for
downhole fluid analysis and all the savings
associated with getting the fluid characterization
right the first time.
For some E&P operators, especially those
involved in deepwater activities, the technology
represents the best existing substitute for
prohibitively expensive or environmentally
unfeasible DSTs. The tool can be run without the
extensive upfront planning required for DSTs.
Other operators have used high-quality
Quicksilver Probe results to optimize DST plans.
In either case, the focused-sampling approach
increases efficiency, quality and safety in these
demanding environments.
In several cases, Quicksilver Probe sampling
has yielded surprising results. The technology is
encouraging some operators to review current
plans and resample zones where other technologies
have given unsatisfactory answers. Operating
companies that have had to drill with WBM for the
purpose of obtaining clean oil samples can now
confidently drill with OBM, safe in the knowledge
that pure samples can be obtained downhole with
no miscible contamination.
As a result of the high purity of new, incoming
samples, some laboratories have had to create a
new classification for such low contamination,
called too small to measure, or TSTM. Now that
such pure samples are available for laboratory
analysis, researchers and experimentalists may
be able to perform additional analyses and
devise new measurements to better understand
fluid behavior.
An important consequence of the ability to
obtain zero-contamination samples downhole is
the improvement in accuracy of real-time
downhole measurements. This will encourage
companies to perform downhole fluid analysis for
a more complete mapping of reservoir fluids than
is done today, and will also promote the addition
of new DFA measurements.
LS

19