Oilfield Review
Autumn 2001
Characterizing Permeability
Improving Fluid Sampling
Global Warming
Selective Stimulation
The Schlumberger Oilfield Glossary
Do you know what the Udden-Wentworth scale is? Or what a blowout
preventer does? It’s easy to find the answer on the Schlumberger Web site.
The Schlumberger Oilfield Glossary is a unique, multidisciplinary reference that defines hydrocarbon exploration, development and production
terms for the technical generalist and expert alike. Technical experts review
all definitions, from “abnormal events” to “Zoeppritz equations.” High-quality,
full-color photographs and illustrations clarify many definitions.
Winner of an Award of Excellence from the Business Marketing
Association, the glossary currently contains more than 3600 terms, and
eventually will comprise more than 7000 definitions.
Join the “virtual crowd” and learn more about oilfield technology!
SMP-6096
50387schD05R2 12/10/01 3:14 PM Page 1
Advancing Our Understanding of Permeability
With commercial production dating back to the 1870s, the
hydrocarbon-producing industry has been in business
longer than nearly any other. The fact that we are a mature
industry does not mean we are stagnating. As articles in
this issue of Oilfield Review show, we have continually
advanced in technology, practice and understanding.
One thing remains the same, however. The goal is still to
produce hydrocarbon as fast as possible, as long as possible, and with minimal long-term consequences to environment and people. The collection of technologies amassed
to do this is impressive, but their success depends on how
well we understand the character of the reservoir that contains the hydrocarbons.
The maturity of the industry, wherein many reservoirs
worldwide have become depleted, has drawn attention to
the importance of the variability and distribution of the
properties within reservoirs. We have, in fact, been in a socalled reservoir-characterization phase of industry maturity for more than 10 years. And no reservoir property seems
to benefit more from good characterization than permeability.
Permeability is the property of a reservoir that describes
how fluid flows through it, and we know quite a bit about
it. We know that permeability is determined by the number
and size of the pores within the reservoir. The pore size, in
turn, depends on the size of the particles forming the
medium, the amount of loading on the medium, and the
amount of cements added after deposition. These complex
dependences can defy efforts to correlate permeability
with other properties such as porosity. We also know that
while permeability can be measured in the laboratory,
ways to measure it in the field are not as reliable.
Pressure-transient analysis, a mature and often successful
technology, can lead to measurements that are easily confounded by other effects, one of which is uncertainty about
the volume of investigation. Permeability also seems to be
the most variable of petrophysical properties within a
reservoir. Ranges of 1000 or more from minimum to maximum are common. All reservoirs appear to show significant permeability heterogeneity, although regions within a
reservoir can be fairly homogeneous.
We have learned a great deal about the distribution of
permeability during this reservoir-characterization period,
much of it from cores and outcrop study. We know that
sandstone heterogeneity appears to be set by the deposition of the solid material; carbonate heterogeneity, by what
happened to it after deposition. Sandstone heterogeneity
appears to be strongly correlated locally. This degree of
correlation is directionally dependent; permeability is
much more correlated horizontally (lateral or parallel to
geologic beds) than vertically (perpendicular to beds).
Heterogeneity in carbonate media is substantially greater
than in sandstones. It is far less correlated locally than in
sandstones, and the differences in correlation direction
(vertical versus horizontal) are less than in sandstones.
Both carbonates and sandstones lend themselves to layerlike descriptions. Sandstones are layer-like because of the
strong horizontal correlation in their original deposition.
Though post-deposition alterations tend to wipe out much
of the local correlation in carbonates, the low-frequency
portion that remains is strongly correlated and continues
to bear the imprint of the deposition.
These comments apply mainly to horizontal permeabilities. Much less is known about vertical permeabilities.
These decrease with averaging scale but beyond that, we
lack knowledge, primarily because of the difficulty of measuring this quantity at a scale that is meaningful for subsequent use. It is fairly obvious that the success of a horizontal well depends directly on having a large vertical permeability. What is less obvious is that vertical permeability
seems to play a significant role in all recovery predictions.
The article “Characterizing Permeability with Formation
Testers,” page 2, looks into some of the issues associated
with measuring vertical permeability.
Several questions about permeability heterogeneity
remain. For example, we do not understand why post-deposition effects should randomize permeability in carbonate
reservoirs. Nor do we understand the distinction between
fracture-dominated and stratigraphic-dominated production behavior. Work needs to be done to understand the
averaging of horizontal and vertical permeability at progressively larger scales of measurement. Horizontal averages tend to increase with scale; vertical averages tend to
decrease with scale. This issue is undoubtedly linked to
the subject of permeability distribution, which still
requires more understanding.
Larry W. Lake
Department of Petroleum and Geosystems Engineering
The University of Texas
Austin, Texas, USA
Larry W. Lake is a professor in the Department of Petroleum and Geosystems
Engineering at The University of Texas (UT) at Austin. He holds BSE and PhD
degrees in chemical engineering from Arizona State University in Tempe, and
Rice University in Houston, Texas, respectively. A prolific author, he has been
teaching at UT for 22 years. Before this, he worked for the Shell Development
Company in Houston. He has served on the Board of Directors for the Society
of Petroleum Engineers (SPE) as well as on several of its committees, and has
also been an SPE distinguished lecturer.
50387schD06R1 12/05/2001 03:06 AM Page 1
Advisory Panel
Terry Adams
Azerbaijan International
Operating Co., Baku
Svend Aage Andersen
Maersk Oil Kazakhstan GmBH
Almaty, Republic of Kazakhstan
Antongiulio Alborghetti
Agip S.p.A
Milan, Italy
George King
BP
Houston, Texas
Abdulla I. Al-Daalouj
Saudi Aramco
Udhailiyah, Saudi Arabia
David Patrick Murphy
Shell E&P Company
Houston, Texas
Syed A. Ali
Chevron Petroleum Technology Co.
Houston, Texas, USA
Richard Woodhouse
Independent consultant
Surrey, England
Executive Editor
Denny O’Brien
Advisory Editor
Lisa Stewart
Senior Editor
Mark E. Teel
Editors
Gretchen M. Gillis
Mark A. Andersen
Matt Garber
Contributing Editors
Rana Rottenberg
Malcolm Brown
Julian Singer
Distribution
David E. Bergt
Design/Production
Herring Design
Mike Messinger
Steve Freeman
Illustration
Tom McNeff
Mike Messinger
George Stewart
Printing
Wetmore Printing Company
Curtis Weeks
Oilfield Review is published quarterly by Schlumberger to communicate
technical advances in finding and producing hydrocarbons to oilfield
professionals. Oilfield Review is distributed by Schlumberger to its
employees and clients. Oilfield Review is printed in the USA.
Contributors listed with only geographic location are employees of
Schlumberger or its affiliates.
© 2001 Schlumberger. All rights reserved. No part of this publication
may be reproduced, stored in a retrieval system or transmitted in any
form or by any means, electronic, mechanical, photocopying, recording
or otherwise without the prior written permission of the publisher.
Address editorial correspondence to:
Oilfield Review
225 Schlumberger Drive
Sugar Land, Texas 77478 USA
(1) 281-285-8424
Fax: (1) 281-285-8519
E-mail: obrien@sugar-land.oilfield.slb.com
Address distribution inquiries to:
David E. Bergt
(1) 281-285-8330
Fax: (1) 281-285-8519
E-mail: dbergt@sugar-land.oilfield.slb.com
Oilfield Review subscriptions are available from:
Oilfield Review Services
Barbour Square, High Street
Tattenhall, Chester CH3 9RF England
(44) 1829-770569
Fax: (44) 1829-771354
E-mail: orservices@t-e-s.co.uk
Annual subscriptions, including postage, are
160.00 US dollars, subject to exchange rate fluctuations.
Oilfield Review is pleased
to announce the addition
of Abdulla I. Al-Daalouj to
its editorial advisory panel.
Mr. Al-Daalouj graduated
from King Fahd University
for Petroleum and Minerals
in Dhahran, Saudi Arabia,
with a degree in Petroleum
Engineering. He joined
Saudi Aramco in 1982 and
has spent his career working in the upstream sector,
predominately in petroleum engineering, producing and oil operations.
He is currently ManagerSouthern Area Producing
Engineering Department.
50387schD07R1 12/05/2001 03:10 AM Page 1
Schlumberger
Autumn 2001
Volume 13
Number 3
Oilfield Review
2
Characterizing Permeability with Formation Testers
Permeability controls reservoir performance but is difficult to determine,
often changing dramatically with scale and direction. Modern wireline formation testers, equipped with packers and multiple probes, provide cost-effective permeability data not reliably available with other techniques. Case studies show how wireline-tester data, interpreted with new models, can now
quantify the effects of small but crucial baffles and super-permeability
streaks, as well as determine vertical and horizontal permeability at a length
scale between those of cores and drillstem tests.
24 Quantifying Contamination Using Color of Crude and Condensate
Oil-base and synthetic-base mud filtrates contaminate openhole reservoirfluid samples, distorting fluid properties measured in a laboratory. These
fluid properties influence development and production decisions with
significant economic consequences. Now, monitoring hydrocarbon color
allows a quantitative measure of contamination, improving the probability
of collecting a valid fluid sample. In addition, a new, direct detection of
methane downhole makes contamination measurement possible in gascondensate zones.
44 Global Warming and the E&P Industry
The controversy surrounding global warming continues without a clear-cut
consensus as to its extent or implications. We examine the evidence and the
arguments, both pro and con, the advances being made in computer simulation of global climate systems and the proactive steps being taken by oil and
gas companies and service suppliers to reduce the impact of oilfield operations on climate change.
Observed
behavior
Comparison
and
validation
Climate-system
model
Computer
simulation
Predicted
behavior
Update and refine model
60 Isolate and Stimulate Individual Pay Zones
With coiled tubing as a conduit for proppant-laden fracturing fluids, single or
multiple zones can be stimulated consecutively during a single mobilization.
New tools selectively isolate target pay zones without conventional rigs or
wireline intervention to set mechanical plugs. Individual zones are treated
separately to achieve optimal fracture length and conductivity. Case studies
demonstrate the expanding scope and economic benefits of this technique.
78 Contributors
82 New Books and Coming in Oilfield Review
1
50387schD02R1.p2.ps 12/10/01 3:48 PM Page 2
Characterizing Permeability with
Formation Testers
We never seem to know enough about permeability. We measure it at small scales
through laboratory tests on cores. We infer it at large scales from well tests and production data. But to manage the development of a reservoir, we also need to quantify
features at intermediate scales. This is where the versatility of wireline formation
testers comes into play.
Cosan Ayan
Aberdeen, Scotland
Hafez Hafez
Abu Dhabi Company for Onshore
Operations (ADCO)
Abu Dhabi, United Arab Emirates (UAE)
Sharon Hurst
Phillips Petroleum
Beijing, China
Fikri Kuchuk
Dubai, UAE
Aubrey O’Callaghan
Puerto La Cruz, Venezuela
John Peffer
Anadarko
Hassi Messaoud, Algeria
Julian Pop
Sugar Land, Texas, USA
Murat Zeybek
Al-Khobar, Saudi Arabia
For help in preparation of this article, thanks to
Mahmood Akbar, Abu Dhabi, UAE.
AIT (Array Induction Imager Tool), CQG (Crystal Quartz
Gauge), FMI (Fullbore Formation MicroImager), MDT
(Modular Formation Dynamics Tester), OFA (Optical
Fluid Analyzer) and RFT (Repeat Formation Tester) are
marks of Schlumberger. RDT (Reservoir Description Tool)
is a mark of Halliburton.
1. In direct measurements of fluid flow in rocks, the quantity measured is the mobility (permeability/viscosity).
According to Darcy’s law, all fluid effects are accounted
for by the viscosity term, and permeability is independent
of fluid. In practice, this is not exactly true, even without
chemical interactions between rock and fluid. Absolute
permeability is also known as intrinsic permeability.
2. The term radial permeability, kr, describes radial flow
into a wellbore. In vertical wells, radial permeability is
the same as horizontal permeability. Vertical permeability
is written both as kv and kz. Spherical permeability is
written as ks.
2
Oilfield Review
50387schD02R1.p3.ps 12/10/01 3:49 PM Page 3
Grid square
A
Which Permeability?
Permeability determines reservoir and well performance, but the term can refer to many types of
measurements. For example, permeability can be
absolute or effective, horizontal or vertical.
Permeability is defined as a formation property,
independent of the fluid. When a single fluid
flows through the formation, we can measure an
absolute permeability that is more or less independent of the fluid.1 However, when two or more
fluids are present, each reduces the ability of the
other to flow. The effective permeability is the
permeability of each fluid in the presence of the
others, and the relative permeability is the ratio of
effective to absolute permeability. In a producing
reservoir, we are most interested in effective permeability, initially of oil or gas in the presence of
irreducible water, and later of oil, gas and water
at different saturations. To further complicate
matters, effective and absolute permeabilities
can be significantly different (see “Conventional
Permeability Measurements,” page 6).
Formations are usually anisotropic, meaning
their properties depend on the direction in which
they are measured. For fluid-flow properties, we
usually consider transversely isotropic formations, meaning formations in which the two horizontal permeabilities are the same and equal to
kh, while the vertical permeability, kv, is different.
Although more complicated formations exist,
there are typically not enough measurements
to quantify more than these two quantities.
Permeability anisotropy can be defined as kv/kh,
kh/kv, or the ratio of the highest to the lowest permeability. In this article we will use kh/kv, a quantity that is most often greater than 1.2
Autumn 2001
B
0
100
Depth, ft
Modern wireline formation testers bring special
knowledge about reservoir dynamics that no
other tool can acquire. Through multiple pressure-transient tests, they can evaluate vertical as
well as horizontal permeability. By measuring at
a length scale between cores and well tests, they
can quantify the effect of thin layers that are not
seen by other techniques. These layers play a
vital role in reservoir drainage, controlling gasand waterflood performance, and leading to
unwanted gas and water entries. Modern wireline formation testers can also be a cost-effective, environmentally friendly alternative to
regular drillstem and pressure-transient tests.
This article shows how permeability measurements derived from wireline formation testers
are contributing to reservoir understanding and
making an impact on reservoir development.
200
300
400
500
0
100
200
300
400
500
600
Horizontal distance, ft
700
800
900
1000
> A cross section of an idealized reservoir that exhibits large-scale anisotropy caused by local
heterogeneity. A sandstone reservoir (yellow) contains randomly distributed shales (gray). The
vertical permeability for the whole reservoir is about 104 times less than the horizontal permeability—a very large anisotropy. However, the small areas A and B are in isotropic sand and
shale, respectively. The grid square, which might represent a reservoir-simulation block, has
intermediate permeability anisotropy. Vertical permeability is close to the harmonic average of
sand and shale permeabilities, while the horizontal permeability is the arithmetic average.
[Adapted from Lake LW: “The Origins of Anisotropy,” Journal of Petroleum Technology 40, no. 4
(April 1988): 395–396.]
The next complication is related to spatial distribution. Reservoir management would be much
simpler if permeability were distributed uniformly,
but, in practice, formations are complex and heterogeneous—that is, they have a range of values
about two or more local averages. The number of
measurements needed for a full description of a
heterogeneous rock is impossibly high; moreover,
the result of each measurement depends on its
scale. For example, for an idealized reservoir comprising isotropic sand with randomly distributed
isotropic shales, there are three scales to consider—megascopic (the overall reservoir), macroscopic (the grid squares used in reservoir
simulation), and mesoscopic (individual facies)
(above). The megascopic anisotropy is very
high—between 103 and 105. However, areas A
and B are isotropic, while the grid squares
are intermediate, showing that the large-scale
anisotropy is in fact caused by local heterogeneity. Measurements at different scales and in
different locations will find different values for
both kh and kv and hence different anisotropy.
Which permeability to choose? In a singlephase, homogeneous reservoir, the question is
irrelevant—but such reservoirs do not exist.
Almost all reservoirs, and particularly carbonates, are highly stratified. For some formations,
flow properties also vary laterally. For instance,
in deltaic sandstone deposits, the world’s most
prolific reservoirs, flow properties vary laterally
because of the sorting of sediments according to
size and weight during transport and deposition.
Whether in sandstone or carbonate, as heterogeneity increases, the distribution of permeability becomes as important as its average value.
Early in the life of a reservoir, the main concern
is the average horizontal effective permeability to
oil or gas, since this controls the productivity and
completion design of individual wells. Later on,
vertical permeability becomes important because
of its effect on gas and water coning, as well as
the productivity of horizontal and multilateral
wells. The distribution of both horizontal and vertical permeability strongly affects reservoir performance and the amount of hydrocarbon recovery,
while also determining the viability of secondaryand tertiary-recovery processes.
3
50387schD02R1
11/29/01
4:59 AM
Page 4
Conduits
Giga
Baffles
Nonsealing fault
Healed fractures
Open fractures
Low-permeability genetic units
High-permeability genetic units
Low-permeability stylolite
High-permeability stylolite
Tight laminations
Small fractures
Shale lenses
Vugs
Low-permeability recrystallization
feature
High-permeability solution channel
Meso
Mega and Macro
Sealing fault
> Permeability baffles and conduits at different length scales. In each case, reservoir management can be improved by quantifying the effects of these features.
3. Weber AG and Simpson RE: “Gasfield Development—
Reservoir and Production Operations Planning,” Journal
of Petroleum Technology 38, no. 2 (February 1986):
217-226.
4. Ayan C, Colley N, Cowan G, Ezekwe E, Wannel M, Goode P,
Halford F, Joseph J, Mongini A, Obondoko G and
Pop J: “Measuring Permeability Anisotropy: The Latest
Approach,” Oilfield Review 6, no. 4 (October 1994): 24-35.
5. The so-called drawdown permeability is calculated as
kd = C qµ/∆pss in units of mD, where q is the flow rate in
cm3/s, µ is the fluid viscocity in cp, and ∆pss is the measured drawdown pressure in psi (and therefore includes
any pressure drop due to mechanical skin). C, the flowshape factor, depends on the effective radius of the
probe, and equals 5660 for the standard RFT and MDT
Modular Formation Dynamics Tester probes and the
units given.
4
6. Dussan EB and Sharma Y: “Analysis of the Pressure
Response of a Single-Probe Formation Tester,” SPE
Formation Evaluation 7, no. 2 (June 1992): 151-156.
7. Jensen CL and Mayson HJ: “Evaluation of Permeabilities
Determined from Repeat Formation Tester
Measurements Made in the Prudhoe Bay Field,” paper
SPE 14400, presented at the SPE Annual Technical
Conference and Exhibition, Las Vegas, Nevada, USA,
September 22-25, 1985.
8. Goode PA and Thambynayagam RKM: “Influence of an
Invaded Zone on a Multiple Probe Formation Tester,”
paper SPE 23030, presented at the SPE Asia Pacific
Conference, Perth, Western Australia, Australia,
November 4-7, 1991.
We might expect the buildup permeability to be higher
than kd since, by reading farther into the formation, it
should read closer to the effective permeability of the
formation to oil or gas. However, in general experience,
the buildup permeability reads lower.
The magnitude of permeability contrast
becomes increasingly important with prolonged
production. Thin layers, faults and fractures can
have a dramatic effect on the movement of a gas
cap, aquifer, and injected gas and water. For
example, a low-permeability layer, or baffle, will
impede the movement of gas downwards. A
high-permeability layer, or conduit, will quickly
bring unwanted water to a production well. Both
can significantly affect the sweep efficiency and
require a change in completion practices. Sound
reservoir management depends on knowing not
only the average horizontal permeability but also
the permeability distribution laterally and vertically, and the conductivity of baffles and conduits
(left). As has been known for a long time, reservoir heterogeneity is one of the major reasons
why enhanced oil recovery is so difficult.
Permeability heterogeneity, unexpected baffles
and insufficiently detailed reservoir evaluation
are often the reasons that these projects fail to
be economical.3
In normal reservoir-engineering practice, the
main sources of average effective permeability
are pressure-transient well testing and production tests. These are usually good indicators of
overall well performance. Cores and logs are
used, but often after some matching, or scaling
up, to well-test results. Once a reservoir has been
on production, conventional history matching
gives information on average permeability, but
cannot resolve its distribution. The presence of
high- or low-permeability streaks and their distributions are inferred from cores and logs, but this
information is qualitative rather than quantitative.
Wireline formation testers (WFTs) have stepped
into this gap, providing various measurements of
permeability from simple drawdowns with a single probe to multilayer analyses with multiple
probes. The latter were originally used mainly to
determine anisotropy.4 With recently developed
analytical techniques and further experience,
multilayer analyses now provide quantitative
information about permeability distribution.
Wireline Formation Testers
Early wireline formation testers were designed
primarily to collect fluid samples. Pressures were
recorded, so that the pressure buildups at the end
of sampling could be analyzed to determine permeability and formation pressure. In spite of the
limited gauge resolution and the few data points
available, the results were often an important
input to formation evaluation. Now, buildups
acquired after sampling are still analyzed to obtain
an estimate of permeability at little extra cost.
The Schlumberger RFT Repeat Formation
Tester tool introduced the pretest, a short test
Oilfield Review
50387schD02R1.p5.ps 12/10/01 4:43 PM Page 5
initially designed to determine whether a point
was worth sampling. To the surprise of many,
pretest pressure turned out to be representative
of reservoir pressure. As a result, pressure measurements became the main WFT application.
Permeability could be estimated from both the
drawdown and the buildup during a pretest.
Since a reliable pressure gradient required
pretests at several depths, much more permeability data became available. With tens of test
points in a single well, it became easier to establish a permeability profile and compare results
with core and other sources.
Pretests continue to be an important feature
of modern tools, although the reliability of the
permeability estimate varies. Since pretests
sample a small volume, typically 5 to 20 cm3
[0.3 to 1.2 in.3], the drawdown permeability, kd,
can be overly influenced by formation damage
and other near-wellbore features.5 Detailed analysis shows that kd is closest to kh, although it is
influenced by kv.6 The volume of investigation is
significantly larger than that of a core plug, but of
the same order of magnitude. However, kd is typically the effective permeability to mud filtrate in
the invaded zone rather than the absolute permeability as obtained from core. Although some
good correlations between the two have been
found, kd is generally considered to be the
minimum likely permeability.7 Nevertheless, it
can be computed automatically at the wellsite,
and is still used regularly as a qualitative indicator of productivity.
Pretest buildups investigate farther into the
formation than drawdowns, several feet if the
gauge resolution is sufficiently high and the
buildup is recorded long enough. Except in lowpermeability formations, buildup time is short, so
that the tool may be measuring the permeability
of either the invaded zone, the noninvaded zone,
or some combination of the two.8 As in the interpretation of any pressure-transient data, flow
regimes are identified by looking for characteristic gradients in the rate of change of pressure
with time. For pretest buildups in which the flow
regimes are spherical and occasionally radial,
consistent gradients often prove hard to find, and
even then may be affected by small changes in
the pretest sampling volume. For reliable results,
each pretest must be analyzed—a time-consuming process. Today, the analysis of short pretest
buildups for permeability is rare, mainly because
there are much better ways to obtain permeability with modern tools.
Modular Wireline Formation Testers
The third-generation WFT is the modular tester.
This tool can be configured with different modules to satisfy different applications, or to handle
varying conditions of well and formation (below).
6.6 ft
8 ft
~3 ft
Input
port
2.3 ft
A
B
C
D
E
F
G
H
Usually
ks
kh,kv
kh,kv
kh,kv,φCt
kh,kv,φCt
ks and/or kh
kh,kv
kh,kv
Sometimes
kh
φCt
φCt
> Typical MDT tool configurations for permeability measurements: single probe with sample chamber and flow-control
module (A); a sink, normally the bottom probe, with one (B) or two (C) vertical observation probes; dual-probe module
with one (D) or two (E) vertical probes; mini-DST configuration with dual-packer and pumpout module (F); dual-packer
module with one (G) or two (H) vertical probes. The flow-control module, sample chamber and pumpout module can be
added to any configuration. When only one pressure transient is recorded, as in (A) and (F), permeability determination
depends on identifying particular flow regimes, type-curve matching or parameter estimation using a forward model.
With one or more vertical probes, as in the other configurations, it is possible to perform a local interference test, also
known as an interval pressure-transient test (IPTT). With these tests, interpreters can determine kv and kh for a limited
number of layers near the tool. Storativity, øCt, can be determined with the dual-probe module, and sometimes when
three vertical transients are available, as in (C) and (H). With other configurations, it must be determined from other
data. Pretest drawdown and buildup permeabilities can be determined with the dual-packer module and each probe
in all configurations.
Autumn 2001
5
50387schD02R1.p6.ps 11/17/01 6:01 PM Page 6
Conventional Permeability Measurements
6
Water-wet
Relative permeability
1.0
0.8
kro
0.6
0.4
krw
0.2
0
0
0.2
0.4
0.6
0.8
Sw
B
Oil-wet
1.0
A
1.0
Relative permeability
Pressure-transient analysis, production tests, history data, cores and logs are all used to estimate
permeability. Each measurement has different
characteristics, advantages and disadvantages.
Core data—Routine core measurements give
absolute, or intrinsic, permeability. In shaly
reservoirs with high water saturation or in oilwet reservoirs, the effective permeability can be
significantly lower than the absolute permeability (right). Core data are taken on samples that
have been moved to surface and cleaned, so that
measurement conditions are not the same as
those made in situ. Some of these conditions,
such as downhole stress, can be simulated on
surface. Others, such as clay alteration and
stress-relief cracks, may not be reversible.
To be useful for reservoir characterization,
there should be enough core samples to capture
sufficiently the reservoir heterogeneity—various
statistical rules exist to determine how many
samples are required. But it is not always possible to capture a statistically valid range of samples even in one well. Highly porous samples
may fall out of the core barrel, while cutting
plugs from very tight intervals is difficult. Some
analysts prefer permeameter measurements
because more samples can be taken.1 Averaging,
or scaling up, is another tricky issue. For layered flow, the arithmetic average, kav =[∑ki hi/
∑hi], is the most appropriate for the horizontal
permeability. For random two-dimensional flow,
it is the geometric average, kav =[∏ki hi / ∑hi],
while for the vertical permeability, the harmonic
average, kav =[∑ki-1 hi/ ∑hi]-1, is important.2
Log data—Logs measure porosity and other
quantities that are related to pore size, for
example irreducible water saturation and
nuclear magnetic resonance parameters.3
Permeability can be estimated from these measurements using a suitable empirical relationship. This relationship normally must be
calibrated for each reservoir or area to more
direct measurements, usually cores, but sometimes, after scaling up, to pressure-transient
results. The main use of log-derived permeability
is to provide continuous estimates in all wells.
On the economic side, cores and logs have many
applications, so that the extra cost of obtaining
permeability from them is relatively small.
0.8
kro
0.6
krw
0.4
0.2
0
0
0.2
0.4
B’
0.6
Sw
0.8
1.0
A’
> Typical relative-permeability curves for oil and water in a water-wet
reservoir (top) and an oil-wet reservoir (bottom). Effective permeabilities
are relative permeabilities multiplied by the absolute permeability. Points
A and A’ represent the typical situation for a wireline formation tester
drawdown measurement in water-base mud. In a water-wet reservoir, the
filtrate flows in the presence of 20% residual oil and has a relative permeability of 0.3. Points B and B’ represent the typical situation for pressuretransient analysis in an oil reservoir. In a water-wet reservoir, the oil flows
in the presence of 20% irreducible water and has a relative permeability of
0.9. Points A, A’, B and B’ are also known as endpoint permeabilities. Some
engineers refer relative permeabilities to the effective permeability to oil
rather than the absolute permeability, as shown here.
Well tests—Pressure-transient analysis of well
tests measures the average in-situ, effective
permeability of the reservoir. However, the
results have to be interpreted from the change
of pressure with time. Interpreters use several
techniques, including the analysis of specific
flow regimes, and matching the transient to
type curves or a formation model. In conventional tests, the well is produced long enough to
sample up to the reservoir boundaries. Impulse
tests produce for a short time and are useful for
wells that do not flow to surface. In both cases,
but especially for impulse tests, there is not
necessarily any unique solution for permeability.
In most conventional tests, the goal is to measure the transmissivity (khh/µ) during radial
flow. The reservoir thickness, h, can be estimated at the borehole, but is it the same tens
and hundreds of feet into the reservoir where
the pressure changes are taking place? In practice, other information—geological models and
seismic data—helps improve results. With conventional well tests, the degree of heterogeneity
can be detected, but the permeability distribution cannot be determined and there is no
vertical resolution.
Oilfield Review
50387schD02R1.p7.ps 12/10/01 4:43 PM Page 7
Economically, well tests are expensive from
the point of view of both equipment and rig
time. Well tests are also undertaken to obtain a
fluid sample so that the incremental cost of
determining permeability may be small.
However, obtaining high-quality permeability
data often requires long shut-in times and extra
equipment such as downhole valves, gauges
and flowmeters.4
Production tests and production history—
An average effective permeability can be
obtained from the flow rate and pressure during
steady-state production, preferably from specific
tests at different flow rates. Skin and other
near-wellbore effects have to be known or
assumed. An average permeability can also be
determined from production-history data by
adjusting the permeability until the correct history of production is obtained. However, in both
cases, the permeability distribution cannot be
obtained reliably. In the presence of layering or
heterogeneity, this is a highly nonlinear inverse
problem, for which there can be more than
one solution.
In the absence of other data, permeability is
often related to porosity. In theory, the relation
is weak—there are porous media that have
been leached to give high porosity with zero
permeability, and others that have been fractured to give the opposite. However, in practice,
there do exist well-sorted sandstone reservoirs
with a consistent porosity-permeability relation.
Other reservoirs are less simple. For carbonate
rocks in particular, microporosity and fractures
make it almost impossible to relate porosity and
lithofacies to permeability.
1. Zheng S-Y, Corbett PWM, Ryseth A and Stewart G:
“Uncertainty in Well Test and Core Permeabilty Analysis:
A Case Study in Fluvial Channel Reservoirs, Northern
North Sea, Norway,” AAPG Bulletin 84, no. 12 (December
2000): 1929–1954.
2. Pickup GE, Ringrose PS, Corbett PWM, Jensen JL and
Sorbie KS: “Geology, Geometry, and Effective Flow,”
paper SPE 28374, presented at the SPE Annual Technical
Conference and Exhibition, New Orleans, Louisiana, USA,
September 25-28, 1994.
3. Herron MM, Johnson DL and Schwartz LM: “A Robust
Permeability Estimator for Siliclastics,” paper SPE 49301,
presented at the SPE Annual Technical Conference and
Exhibition, New Orleans, Louisiana, USA, September 2730, 1998.
4. Modern Reservoir Testing. SMP-7055, Houston, Texas,
USA: Schlumberger Wireline & Testing, 1994.
Autumn 2001
Some of these modules are particularly relevant
for permeability measurements. The descriptions
of the modules below refer to the Schlumberger
MDT Modular Formation Dynamics Tester tool,
unless otherwise specified.
The single-probe module—This module provides communication between the reservoir and
the tool. It consists of the probe assembly,
pretest chamber, strain and quartz pressure
gauges, and resistivity and temperature sensors.
The probe assembly has a small packer, which
contains the actual probe. When a tool is set,
telescoping backup pistons press the packer
assembly against the borehole wall. The probe is
pressed farther through the mudcake into contact
with the formation. Special probe-assembly
designs are available for difficult conditions.9
Communication is established with the formation
by a short pretest, after which the module can
withdraw fluids for sampling or act as a passive
monitor of pressure changes.
The dual-probe module—This module consists of two probe assemblies mounted in fixed
positions on the same mandrel. In the Halliburton
RDT Reservoir Description Tool, the probes are
mounted above one another, separated by a few
inches and facing the same way.10 One probe,
known as the sink probe, withdraws fluids, while
the other monitors the pressure transient. In the
MDT tool, the two probe assemblies are
mounted diametrically opposite each other on
the mandrel.11 One probe is a sink while the other,
known as the horizontal probe, is solely a monitor with no sampling capability. The main purpose of the dual-probe module is to combine with
a vertical probe to determine kh, kv and storativity (øCt) through a local interference test or, to
use a more specific name, the interval pressuretransient test (IPTT).12 By withdrawing fluid
through the sink, three pressure transients can
be recorded at three different locations along the
wellbore, two of which are from monitor probes
and are not contaminated by the effects of tool
storage, skin and cleanup.13
The dual-packer module—This module has
two packer elements that are inflated to isolate a
borehole interval of about 1 m [3.3 ft]. Once these
are inflated, fluid is withdrawn, first from the isolated interval, and then from the formation. Since
a large section of the borehole wall is now open
to the formation, the fluid-flow area is several
thousand times larger than that of conventional
probes. This offers important advantages in both
low- and high-permeability formations, and in
other situations.
• Probes are sometimes ineffective when set in
laminated, shaly, fractured, vuggy, unconsolidated or low-permeability formations. The dual
packer allows pressure measurements and
sampling in these conditions.
• Used alone, the dual packer makes a small version of a standard drillstem test (DST) that is
known as a mini-drillstem test, or mini-DST.
Since the mini-DST opens up only 1 meter of
formation, it acts as a limited-entry test from
which both kv and kh may be determined under
favorable conditions. Used in combination with
one or more vertical probes, the dual packer
can record an IPTT.
• The pressure drop during drawdown is typically much smaller than that obtained with a
probe. Thus, it is easier to ensure that oil is
produced above its bubblepoint, and that
unconsolidated sands do not collapse. Also,
with a smaller pressure drop, fluids can be
pumped at a higher rate, so that for the same
time period, a larger volume of formation fluid
can be withdrawn and a deeper-reading pressure pulse created.
9. For the MDT tool these include: large-area packers for
tight formations; large-diameter probes for unconsolidated as well as tight formations; long-nosed probes for
unconsolidated formations and thick mudcakes; and
gravel-pack probes and a large-area filter similar to an
automobile oil filter for extremely unconsolidated sands
(the Martineau probe).
10. Proett MA, Wilson CC and Batakrishna M: “Advanced
Permeability and Anisotropy Measurements While
Testing and Sampling in Real-Time Using a Dual Probe
Formation Tester,” paper SPE 62919, presented at the
SPE Annual Technical Conference and Exhibition, Dallas,
Texas, USA, October 1-4, 2000.
11. Zimmerman T, MacInnes J, Hoppe J, Pop J and Long T:
“Applications of Emerging Wireline Formation Testing
Technologies,” paper OSEA 90105, presented at the
8th Offshore Southeast Asia Conference, Singapore,
December 4-7, 1990.
12. The term vertical interference test (VIT) is also used for
vertical wells. The terms local interference test and
interval pressure-transient test are appropriate for deviated or horizontal wells.
Storativity is the product of porosity, ø, and total rock
compressibility, Ct, which is the sum of the solid compressibility, Cr, and the fluid compressibility, Cf . When not
measured by an IPTT, Cf must be estimated from fluid
properties and Cr from knowledge of the solid framework
based on acoustic logs, porosity and other data. If there
is more than one fluid, the saturation of each fluid is estimated from logs or sample volumes.
13. Skin is defined as the extra pressure drop caused by
near-wellbore damage (mechanical skin), flow convergence in a partially penetrated bed, and viscoinertial
flow effects (usually ignored). The flow-convergence
factor can be calculated from knowledge of bed thickness and test interval.
Tool storage is due to the compressibility of the fluid in
the tool, and causes the measured flow rate to be different from the actual flow rate at the formation surface,
or sandface. Cleanup refers to the increase in flow rate
as the flow of fluids removes formation damage near
the borehole.
7
50387schD02R1.p8.ps 01/10/2002 03:52 PM Page 8
The pumpout module—This module pumps
fluid from the formation into the mud column, and
from one part of the tool to another. Pumping into
the mud column allows much larger volumes of
fluid to be withdrawn than when sampling into
fixed-volume sample chambers. The module can
also pump fluid from one part of the tool to
another; from the mud column into the tool, for
example to inflate the packer elements; or into
the interval between the packers to initiate a
small hydraulic fracture. For permeability measurements, the pumpout module is capable of
sustaining a constant, measured flow rate during
drawdown, thereby simplifying considerably the
interpretation of pressure transients. The flow
rate though the pump depends on the pressure
differential, increasing at low differential to a
maximum of 45 cm3/s [0.7 gal/min]. At very high
differential, such as in a tight rock, the pump may
not be able to maintain a constant rate.
The flow-control module—This module withdraws up to 1000 cm3 [0.26 gal] of fluid from the
formation while controlling and measuring the
flow rate. The fluid withdrawn is either sent to a
sample chamber or pumped into the borehole.
The module works in various modes such as
constant flow rate, constant pressure and
ramped pressure, and can also draw repeated
pulses of fluid from the formation. The time for
pulses to arrive at a vertical probe is an important input in the determination of kv. Since the
flow-control module can control flow rate precisely, it can regulate the withdrawal of sensitive
formation fluids into small-volume pressure-volume-temperature (PVT) sample bottles. This is
important for the sampling of condensate reservoirs. (For more on sampling, see “Quantifying
Contamination Using Color of Crude and
Condensate,” page 24).
All these features provide many ways to measure permeability, ranging from simple pretest
drawdown to multiple probes and dual packers
(right). For the most reliable in-situ determination
of permeability and anisotropy, experience has
shown that interference tests should be performed with multiple pressure transients. Results
from other methods will always be more ambiguous, but can still be useful, and even good, estimates in the right conditions. One such technique
is the mini-DST.
8
Advantages
Flow Source
Limitations
Probe
• Simplest method of establishing
communication with formation
• Multiple probes can be added in one
tool string
• Difficult to get good tests in fractured,
vuggy and tight formations (difficult to
withdraw fluids, seal failures)
• High drawdowns in low k/µ formations
may release gas, complicating analysis
Dual packer
• Easier to test fractured, vuggy
and tight formations
• At same flow rate as probe, less drawdown
helps avoid gas and sanding
• For same time period as probe, more
fluid is withdrawn, creating deeper pulse
• Fear (usually unjustified) of sticking or
of releasing gas slug into borehole
• Low drawdown may give insignificant
signals at vertical probes in high k/µ
formations
Drawdown
• Automatic computation, available during
acquisition
• Many (tens) of pretests often recorded
for pressure, allowing qualitative comparisons
• Small volume of investigation (inches)
• Measures effective permeabiliby to
mud filtrate
Buildup
• Deeper radius of investigation than drawdown
• Many (tens) of pretests often recorded
for pressure, allowing qualitative comparisons
• Small sampling volume, cleanup and
tool storage can make analysis difficult
• Measures effective permeability to mud
filtrate, formation fluid or a mixture
of the two
Probe Pretest
Single-Transient Analysis
Dual-packer
mini-DST
or extended
drawdown and
buildup with
probe
• Data available while sampling
• Gives ks and/or kh and can avoid costly DST
• Need a particular combination of
formation properties and thickness to
get both kv and kh
• Need to know φCt to get ks, and need to
know h to get kh
• Tool storage, skin, free gas and continuous
cleanup can complicate analysis
(especially with probe)
Dual-Transient IPTT
Dual packer
+ probe or
tandem probes
• Gives kh and kv
• The simplest configuration for an IPTT
• Need to have a good idea of φCt
• Sink drawdown and early buildup affected
by tool storage, skin, free gas and cleanup
Multiple-Transient IPTT
Three probe
(sink, horizontal
and vertical)
• Analysis can be done without sink drawdown
• Gives φCt as well as kh and kv
• Smaller vertical investigation than other
IPTT configurations (sometimes an
advantage)
Second vertical
probe
• Best configuration for layered reservoirs,
faults and fractures
• Analysis can be done without sink drawdown
• Longer tool
> Features of the flow sources and methods used to derive permeability from the MDT tool.
Oilfield Review
11/29/01
4:24 AM
Page 9
Mini-DSTs
In a standard DST, drillers isolate an interval of
the borehole and induce formation fluids to flow
to surface, where they measure flow volumes
before burning or sending the fluids to a disposal
tank. For safety reasons, many DSTs require the
well to be cased, cemented and perforated
beforehand. The MDT tool, in particular the dualpacker module, provides similar functions to a
DST but on wireline and at a smaller scale.
The advantages of the mini-DST are less cost
and no fluids to surface. Cost benefits come from
cheaper downhole equipment, shorter operating
time and the avoidance of any surface-handling
equipment. On offshore appraisal wells, cost savings can be more than $5 million. With no fluids
flowing to surface, there are no problems of fluid
disposal, no surface safety issues and no problems with local environmental regulations. MiniDSTs are much easier to plan and can test multiple
stations on the same trip—usually a sufficient
number to sample the entire reservoir interval.
The mini-DST has disadvantages: it investigates a smaller volume of formation due to the
smaller packed-off interval (3 ft versus tens of
feet), and withdraws a smaller amount of fluid at
a lower flow rate. In theory, we may be able to
extend the tests and withdraw large amounts of
fluid, but in practice, there may be a limit to how
long the tool can safely be left in the hole.14 The
actual depth of investigation of a wireline tester
depends on formation permeability and other factors, but is of the order of tens of feet, rather than
the hundreds of feet seen by a normal DST.
The smaller volume of investigation is not
necessarily a disadvantage. A full DST reveals the
average reservoir characteristics and assesses
the initial producibility of a well. Permeability
variations will be averaged, and although they
contribute to the average, they are neither
located nor quantified. With the help of logs, the
smaller volume mini-DST can evaluate key intervals. The procedure for interpreting pressure transients from mini-DSTs is the same as for full DSTs
and the same software can be used for both.
TotalFinaElf used mini-DSTs in the Arab reservoir of an aging Middle East field to look for zones
with moveable oil and to calibrate the permeability anisotropy used in a simulation model.15 Since
the packed-off interval rarely covers the whole
reservoir, a mini-DST is a limited-entry, or partially
penetrating, well test. To determine formation
parameters, interpreters need to identify flow
regimes in the buildup. In a homogeneous layer,
there are three flow regimes: early radial flow
around the packed-off interval, pseudospherical
flow until the pressure pulse reaches a boundary,
Autumn 2001
and finally total radial flow between upper and
lower no-flow boundaries. Rarely are all three
seen because tool storage effects can mask the
early radial flow, while the distance to the nearest barrier determines whether or not the other
regimes are developed during the test period.16
However, it has been common to observe a pseudospherical flow regime, and occasionally total
radial flow in buildup tests (below). On a log-log
plot of the pressure derivative versus a particular
function of time, spherical flow is identified by
a slope of –0.5, and radial flow by a stabilized
horizontal line.
Spherical permeability, ks= 3√(kh2 kv) can be
estimated from a pressure-derivative plot during
spherical flow or from a separate specialized
plot.17 Horizontal permeability, kh, can be estimated from a pressure-derivative plot during
radial flow, or from a specialized plot of pressure
versus Horner time, provided the thickness of the
interval is known.18 In this case, the thickness
was obtained from openhole logs, particularly
images from the Schlumberger FMI Fullbore
Formation MicroImager tool. When both spherical- and radial-flow regimes occurred, the interpreters could estimate vertical permeability, kv,
from kh and ks. These initial estimates were combined with the geological data to build a model of
formation properties. Different analytical techniques, such as type-curve matching, were then
used to match the full pressure transient and
improve the permeability estimates.
Measured pressure difference
Measured derivative
Model pressure difference
Model derivative
1000
Pressure difference, psia, and derivative
50387schD02R1
100
10
1
Spherical
flow
Radial
flow
0.1
0.01
0.1
1
10
Time since end of drawdown, sec
100
1000
Type-curve parameters:
kh = 39 mD
kv = 24 mD
µ = 1 cp
Thickness of zone = 8 m
Mechanical skin = 1.3
> Pressure difference and the derivative of pressure with
respect to a function of time for the buildup at the end of a typical mini-DST. The pressure difference is between the measured
pressure and a reference taken near the end of the drawdown
period. The derivative is calculated from d∆p/dln[(tp+∆t)/∆t]
where tp is the producing time and ∆t is the time since the end
of the drawdown. We identify spherical flow by the slope of
–0.5 on the log-log derivative, and radial flow by the slope of
0 (horizontal). The solid lines are the results of a type curve, or
model, computed with the parameters in the table.
14. In one recent job, the pumpout module was run continuously for 36 hours. In another job, the dual-packer module was in the hole for 11 days.
15. Ayan C and Nicolle G: “Reservoir Fluid Identification and
Testing with a Modular Formation Tester in an Aging
Field,” paper SPE 49528, presented at the 8th Abu Dhabi
International Petroleum Exhibition and Conference, Abu
Dhabi, UAE, October 11-14, 1998.
16. Tool storage includes the compressibility of the fluid
between the packers. A common model is to relate the
sandface flow rate, qsf, to the measured flow rate, q, and
the rate of change of pressure by a constant, C: qsf =
q+24Cdp/dt. The very early part of a buildup is dominated
by wellbore storage, also called afterflow. C can be estimated from the rate of change of pressure at this time.
17. On a specialized spherical plot, the slope, msp during
spherical flow is given by: msp = 2453qµ(√µøCt)/ks3/2 in
oilfield units, where ø is usually taken from logs, and q,
the flow rate, is measured or estimated. The viscocity, µ,
is determined from the PVT properties of the mobile fluids. If there is more than one mobile fluid, their saturations are estimated from logs or sample volumes.
18. Horner time is [(tp+∆t)/∆t] where tp is the drawdown
time, and ∆t is the time since the end of the drawdown.
The slope, mr , during radial flow is given by mr =
162qµ/khh, where h is the thickness of the formation
interval, and the other terms are defined in reference 17.
9
50387schD02R1.p10.ps 11/17/01 6:02 PM Page 10
TotalFinaElf recorded ten tests in two wells,
one of which was cored. Both kv and kh were subsequently measured on core plugs sampled every
0.25 or 0.5 m [9.8 or 19.6 in.], and compared with
the mini-DST results (below). Care was taken to
scale up the core data to the mini-DST interval
and to convert from absolute to effective permeability. For some of the tests, pressure-transient
data were also available from two probes in the
MDT tool string, making it possible to compare
mini-DST results with results from a full IPTT as
well as from core samples. The IPTTs measure
larger volumes of formation, yet the results generally agree with the mini-DST, especially for the
near probe. The fact that the different measurements agree suggests that the formations may be
relatively homogeneous, or that the scaling up of
the core data was appropriate. While this good
agreement validates the use of a mini-DST in
these conditions, it is inadvisable to assume the
same degree of homogeneity in other formations.
Horizontal Permeability
600
Permeability, mD
500
Mini-DST
Core
IPTT (V1)
IPTT (V2)
400
300
200
100
Vertical
permeability
0
0
1
2
3
4
5
Test number
Vertical Permeability
35
30
Mini-DST
Core
IPTT (V1)
IPTT (V2)
Permeability, mD
25
20
15
10
5
0
0
1
2
3
4
5
Test number
> Comparison of the horizontal (top) and vertical (bottom) permeabilities measured by mini-DSTs, cores and IPTTs. The core data were
averaged over each mini-DST test interval and converted to effective
permeability using relative-permeability curves. Arithmetic averaging
was used for horizontal permeabilities, and harmonic averaging for
vertical permeabilities. The IPTT data are from the same tests as the
mini-DSTs, but using two probes: V1 at 2 m [6.6 ft] and V2 at 4.45 m
[14.6 ft] above the packer interval. The intervals tested are therefore
different. In this case, the agreement between the different measurements is generally good.
10
Cased-Hole Mini-DSTs
Phillips Petroleum, operating in the Peng Lai field
offshore China, found that cased-hole mini-DSTs
were a valuable complement to full DSTs and
openhole WFTs in evaluating their reservoir.19
Like many operators, they initially ran mini-DSTs
to obtain high-quality PVT samples, but then
found that the pressure-transient data contained
valuable information. Peng Lai field consists of a
series of stacked, unconsolidated sandstone
reservoirs with heavy oil—11° to 21° API—of
low gas/oil ratio (GOR), whose properties vary
widely with depth. Testing each reservoir in each
well with full DSTs was proving expensive, and
was not always successful. Among other factors,
the handling of the heavy oil at surface caused
each DST to last between five and seven days.
Large drawdowns, which were sometimes
needed to lift the oil to surface, caused the formation to collapse and the near-wellbore pressure to drop below the bubblepoint. As a result,
mini-DSTs were an attractive alternative for all
but the largest zones.
With a probe, the drawdowns were too high,
while unstable boreholes and high pressure differentials made openhole wireline testing with a
dual-packer module risky. Phillips’ answer was to
run the dual packer in cased holes. By the end of
2000, they had performed 27 cased-hole miniDSTs in seven wells. In one typical test, they
identified a 3-ft low-resistivity zone that was isolated from the main reservoir at the well by thin
shales above and below (next page, left). After
cement isolation was checked, a 1-ft [30-cm]
interval was perforated, and the MDT dual packers were set across it. Communication was
established, and the formation fluid was pumped
into the borehole until the oil fraction stabilized
(next page, top right). Two oil samples were
taken, and after an additional drawdown, a pressure buildup was recorded over 2 hours. The total
testing time of 16 hours would normally be considered excessive and risky in openhole conditions, but presented no problem in cased hole.
The pressure derivative during buildup shows
a short period of probable spherical flow followed by a period of radial flow (next page,
bottom right). With initial values of ks and kh from
flow-regime identification, the buildup data were
matched with a limited-entry model, assuming a
formation thickness of 3 ft with no outer boundaries. The match is excellent. The high horizontal
permeability (2390 mD) and the low vertical permeability (6 mD) were not surprising for this
zone. Overall, a zone that looked doubtful on logs
proved not only to be oil-bearing but also to have
excellent producibility.
Oilfield Review
Depth, ft
50387schD02R1.p11.ps 11/17/01 6:02 PM Page 11
X00
SP
-100
mV
0
Gamma Ray
0
API
150 1
Resistivity
ohm-m 1000 45
Porosity
p.u.
0
1800
Initial buildup
Sampling
Buildup
Pressure, psia
X10
X20
Oil
breakthrough
1700
1600
Perforations
X40
X50
Pump rate, rpm
X30
600
300-rpm constant pump rate
300
0
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Time, hr
> Pressure and pump rate during the cased-hole mini-DST from Peng
Lai field. After communication was established with the formation,
the pump withdrew invasion fluids until oil broke through. Once the
oil fraction had stabilized (as measured by the OFA Optical Fluid
Analyzer tool, not shown), two samples were taken. After one additional
drawdown, a 2-hr buildup was recorded. Minimum drawdown pressure
was 164 psi [1130 kPa], at or above the expected bubblepoint pressure,
thereby avoiding free gas. The solid pressure line is the result predicted
by the limited-entry model.
X60
> Gamma ray, resistivity and porosity logs across
a low-resistivity reservoir in the Peng Lai field,
offshore China. The mini-DST was performed in a
thin 3-ft zone that is isolated above and below by
thin shale beds (gray) within a larger reservoir.
Any oil found in this zone was expected to be
about 13º API with high viscosity.
Model parameters:
kh = 2390 mD
kv = 6 mD
µ = 300 cp
Thickness of zone = 3 ft
Skin = + 5.5
Depth of investigation = 80 ft
1000
Pressure difference, psia, and derivative
Mini-DST Limitations
In spite of these good results, the permeability
measurements have some limitations. The lack of
an observation probe means that the only pressure transient comes from the pressure sink,
which is affected by skin and tool storage. Both
skin and storage influence the early part of the
buildup and make identification of flow regimes
and interpretation more difficult. Later in the
buildup there needs to be the right combination
of formation properties and bed thickness for significant periods of both spherical and radial flow
to be observed. The radial-flow interpretation
depends directly on identifying bed boundaries,
while spherical-flow interpretation depends on
knowing the storativity. Thus, it is difficult to
determine both kv and kh simultaneously.
Finally, several factors can make a single
transient hard to interpret. These include gas
evolution near the wellbore, pressure and flowrate variations due to continuous cleanup, and
noisy drawdown pressures from pump strokes.
Pressure measurements at observation probes
are not usually affected by these phenomena.
Since these probes are higher up the string,
they also increase the volume investigated.
Pressure
difference
100
Pressure
derivative
10
Spherical
flow
Radial
flow
1
0.0001
0.001
0.01
0.1
1
10
Time since end of drawdown, hr
> Pressure difference and derivative for the buildup at the end of the
Peng Lai test. Spherical flow is identified by the slope of –0.5 on the
derivative and radial flow by the slope of zero. The solid lines are the
predictions of a limited-entry model using the parameters in the table.
19. Hurst SM, McCoy TF and Hows MP: “Using the Cased
Hole Formation Tester for Pressure Transient Analysis,”
paper SPE 63078, presented at the SPE Annual Technical
Conference and Exhibition, Dallas, Texas, USA, October
1-4, 2000.
Autumn 2001
11
50387schD02R1.p12.ps 11/17/01 6:25 PM Page 12
UAE Carbonate Permeability
kh (Core)
0.1
mD
1000
kv (Layered Model)
0.1
UAE Carbonate Porosity
mD
No Core
1000 Permeability
kh (Layered Model)
X100
0.1
X110
mD
Layer No.
1000
1
2
3
X120
4
X130
5
X140
6
X150
87 9
10
X160
11
X170
12
13
X180
14
15
Depth, ft
X190
16
17
X200
18
19
X210
20
X220
21
22
IPTTs have proved to be an effective means for
determining permeability distribution near the
wellbore; in fact, they are the preferred method
for layered systems. Mini-DSTs are usually run
when the main objective is to recover a fluid
sample, or to measure reservoir pressure, particularly in tight or heterogeneous formations.
Permeability is an additional parameter with
which to judge the producibility of the interval.
Interval Pressure-Transient Test
An IPTT run in a carbonate reservoir in the United
Arab Emirates (UAE) illustrates the sequence of
operations and methods employed in a full analysis.20 This reservoir has distinct, contrasting layers that appear to extend over large areas.
Reservoir management and the design of secondary-recovery schemes depend strongly on
knowing the vertical and horizontal permeabilities and the communication between layers. In
particular, the implementation of an injection
scheme depends on the permeability of several
low-porosity, stylolitic intervals. Will the stylolites act as baffles to injected fluid and severely
affect sweep efficiency?
The stylolitic intervals may be thinner than
1 ft, but can be observed on logs and cores (left).
However, their effectiveness as barriers is not
clear. They can be correlated between wells, but
their lateral continuity and permeability are
uncertain. Cores could not be recovered from
20. Kuchuk FJ, Halford F, Hafez H and Zeybek M: "The Use of
Vertical Interference Testing to Improve Reservoir
Characterization," paper ADIPEC 0903, presented at the
9th Abu Dhabi International Petroleum Conference and
Exhibition, Abu Dhabi, UAE, October 15-18, 2000.
23
X230
24
25
26
X240
27
X250
28
29
X260
30
X270
X280
31
< Log porosity in a layered carbonate (left). The
low-porosity streaks are stylolites. The positions
of the packer and the probes at each test location were chosen to straddle the stylolites. The
right track shows the layered model used to
interpret the IPTTs, with kv and kh from the model
and kh from core. Core permeability is generally
too high and is either absent from the stylolites
or fails to reflect the large contrasts seen by the
IPTT. The FMI image (left) shows two low-porosity streaks (white) separated by a dark interval.
The top streak is particularly patchy. The layered
model used to match the IPTT showed that the
top streak had higher kv than kh, while the center
interval had very high permeability.
X290
X300
0
05
10
15
20
25
30
35
Porosity, p.u.
12
Oilfield Review
50387schD02R1 12/21/2001 02:57 PM Page 13
20
4200
Tool
retraction
15
4000
Pressure
10
3800
Tool
setting
Pretest
Drawdown
Buildup
Flow rate, B/D
Packer pressure, psi
Flow rate
5
3600
0
0
1000
2000
3000
4000
Time, sec
4200
3930
4000
Packer
3910
Probe 1
3800
3900
Probe 2
Probe pressure, psi
Packer pressure, psi
3920
3890
3600
3880
0
1000
2000
3000
4000
Time, sec
> The sequence of events in a typical IPTT, as shown by the pressure
and the flow rate recorded in the dual-packer interval (top). After tool
setting, the pretest establishes communication with the reservoir by
withdrawing up to 1000 cm3 [60 in.3] through the packer and 20 cm3
[1.2 in.3] through each probe. During drawdown, the flow rate is constant since it is controlled by the pumpout module. During the buildup
period, the pressure is recorded for a sufficiently long time, approximately the same as the drawdown period, to ensure good pressuretransient data. At the end of the buildup period, the probes and packer
are retracted. Packer and probe pressures were recorded with CQG
Crystal Quartz Gauge pressure gauges during the IPTT (bottom). Note
the much more sensitive scale for the probe pressures. Their final
buildup pressure is lower because they are higher in the well. Note
also the distinct delay in the start of the buildup on Probe 2, due to the
low vertical permeability. The delay on Probe 1 cannot be seen at this
time scale. The packer pressure is slightly noisy due to pump movement.
many of these intervals, and, in any case, give a
very local value of the permeability. The operator
decided to investigate the stylolites with a series
of IPTTs in a new well. These could be recorded
on a single trip in the hole, allowing the complete
reservoir section to be tested efficiently.
An IPTT needs a minimum of one vertical
observation probe and a sink, either a dual-probe
or a dual-packer module. In this case, in order to
sample more layers, the MDT tool was equipped
with two vertical observation probes at 6.4 ft and
14.4 ft [1.95 and 4.4 m] above the center of the
Autumn 2001
packer interval. The dual-packer module was
chosen so as to generate a sufficiently large
pressure change at the far probe. The pumpout
module was used to withdraw formation fluids
from each tested interval. Pressures were measured by quartz-crystal and strain gauges at both
probes and packer.
Sequence of operations—Using openhole
logs, the operator selected six test locations,
with the depths chosen so that the stylolites lay
between the dual packer and near probe. At each
test location, the operator followed the same
sequence of events: set the packers and probes,
pretest probes and packer interval, drawdown,
buildup, and retract packers and probes (above).
The pretests measured formation pressure and
established communication with the formation.
Once communication was established, formation
fluids were withdrawn through the packer interval at an almost constant rate for between 30
and 60 minutes. The rate was slightly different
for each test, but remained between 15 and 21
B/D [2.4 and 3.3 m3/d]. After each drawdown, the
interval was shut in for another 30 to 60 minutes.
13
50387schD02R1.p14.ps 11/17/01 6:02 PM Page 14
31-Layer Model
Layer
Thickness
Core kh
kh
kv
Number
ft
mD
mD
mD
1
7
65
0.21
low
2
97
_
98
2
0.1
0.021
0.15
moderate
dense zone
3
6
_
610
610
0.27
high
high permeability
4
7
78
68
35
0.26
moderate
5
10
33
26
16
0.28
low
6
8
61
67
48
0.28
low
7
2
46
53
39
0.18
low
8
0.5
32
28
0.15
low
9
0.5
19
_
0.9
11.1
0.14
moderate
patchy stylolite
10
4
_
1350
725
0.27
high
superpermeability
11
12
81
75
31
0.28
moderate
12
8
30
24
14
0.26
low
13
9
8-60
46
26
0.26
low
14
2
2.7
9.9
33.8
0.2
low
15
5
16
15.6
5.4
0.29
high
Porosity
Confidence
16
7
18
11.3
12.9
0.3
high
17
2
9.3
1.4
1.3
0.11
high
18
7
13
6.7
2.3
0.29
high
high
19
6
9.4
6
3.5
0.28
20
8
12.3
7.4
7.8
0.3
high
21
3
3.3
3.5
0.25
high
22
2
12.1
_
1.3
1.1
0.19
high
23
8
_
3.2
3.2
0.2
high
24
4
8.6
7.9
6.4
0.28
high
25
1
19.1
19.8
3.8
0.2
high
26
6
16
5.4
2.3
0.28
high
27
5
10
11.4
4.6
0.29
high
28
7
6.8
3.1
0.28
high
29
1
11
_
0.1
0.89
0.19
high
30
22
11.3
4.2
1
0.28
high
31
14
1.4
0.9
0.45
0.1
high
Comments
patchy stylolite
dense zone
dense zone
patchy stylolite
dense zone
dense zone
> Model with 31 layers used for interpreting pressure transients. Each layer is assigned a thickness,
vertical and horizontal permeability, porosity, and level of confidence.
In this test, packer pressure dropped sharply
by approximately 300 psi [2070 kPa], while nearprobe pressure dropped more slowly by 10 psi
[69 kPa] and far probe by 2 psi [14 kPa]. These
responses give a first idea of permeability. The
fact that there is a response at the vertical
probes showed that there was communication
across the stylolite.
Analysis—Interpretation starts with a look at
each test independently. As with mini-DSTs, the
first step is to analyze flow regimes. Buildups are
preferred to drawdowns because they are less
14
affected by near-wellbore factors, such as
cleanup and pressure fluctuations caused by the
pumpout piston. The interpreter examined each
of the three pressure transients from the six
tests, and established some initial estimates of
permeability. Because of the highly stratified
nature of this carbonate formation, these estimates were rough averages of the permeability
near each station.
The heart of the interpretation is a realistic
model, layered in this case, with permeabilities,
porosities and thicknesses for 31 layers (above).
Initial layer boundaries and thicknesses are
determined from the logs, actually from high-res-
olution images since layers as thin as 0.5 ft
[15 cm] may play an important role. Porosity and
rock-framework compressibility are based on log
data; fluid compressibility and viscosity come
from fluid saturations and PVT analysis. Initial
horizontal and vertical permeabilities are taken
from the flow-regime analyses and other available sources—cores, logs and pretests. Initial
estimates are also needed for tool storage and
skin around the packer.21 Finally, the flow rate
during drawdown is an important input; in this
case, it was measured and was taken to be
essentially constant during most tests.
Oilfield Review
50387schD02R1.p15.ps 11/17/01 6:25 PM Page 15
Computed data
MDT-measured data
Probe
Probe
Probe
log∆t
Flow-regime
identification
and analysis
Packer
Model
definition
t
Pretest
analysis
• Formation pressures
• Drawdown permeabilities
log∆t
Skin, storage constants,
formation pressures,
flow rates
Multilayer
model
Probe
Packer
Packer
Compute
transients
from model
kh
kv
φCt
Packer
∆P,∆P’
P
Probe
kh
kv
φCt
∆P,∆P’
P
t
Pressure
transient
Flow rate
Packer
Single-layer
model
Measured data
kh
kv
φCt
Adjust model to
minimize difference
between computed
and measured data
Probe
Probe
Packer
Packer
kh
kv
φCt
Initial average
• ks, if spherical flow
• kh, if radial flow
• kv,kh, if both
Other data
Openhole
Openhole logs, Fluid analysis:
logs: φ,Sw,Cr images: layers
µ,Cf
> A typical workflow for the interpretation of an IPTT, with dual packer and one vertical probe. Each job is different, and the actual path taken depends on
a trade-off between speed, complexity of problem and accuracy of results. Quickest, but least accurate results come from analyzing individual transients.
Next may be analysis of all transients from one test with a single-layer model, then with a multilayer model. Adjusting the model to best match all the
available data may require several iterations.
With these initial estimates, the expected
pressure transients at the packer and the two
probes are computed and compared with the
measured transients during drawdown and
buildup (above). An automatic optimization procedure adjusts the model parameters to minimize
the differences over all transients. The main goal
is to obtain the best kv and kh for the layers near
the station. Bed boundaries are changed manually if necessary, while, in this case, øCt was
Autumn 2001
known well enough to be fixed. Permeabilities of
layers away from the station may affect results to
some extent but are not allowed to change significantly. Flow rate is held closely to the measured
rate, but is still computed so as to allow for tool
storage and the effect of small flow-rate changes
on the transients.
When the results are not satisfactory, the geological model is reexamined with the geologist,
redefining some layers and changing some initial
estimates. Different weights can be applied to
different time periods and different transients. For
example, the packer drawdown period might
receive less weight because, unlike observationprobe pressures, it is affected by the noise associated with production and variable cleanup.
The interpreter applied the model to each test
in turn. However, this was not the end, since
some tests were conducted close enough to each
other that changing the parameters in the vicinity
of one may have altered the results from another.
21. Since the flow rate into the probe is negligible, the skin
and tool storage at the probe can be ignored.
15
50387schD02R1.p16.ps 11/17/01 6:03 PM Page 16
4
Probe 2
Measured
Computed
Pressure difference, psi
3
2
1
0
0
500
1000
1500
2000
2500
Time, sec
12
Probe 1
Measured
Computed
Pressure difference, psi
10
8
6
4
2
Probe 2 (for reference)
0
0
500
1000
1500
2000
2500
Time, sec
400
Packer
Measured
Computed
350
Pressure difference, psi
300
Therefore, the optimized model was reapplied to
each test so as to achieve a good match between
all measured and computed transients (left).
Some layers were better defined than others
because there were more pressure transients in
their vicinity. For this reason, the confidence factor for the bottom 15 layers, for which there were
four tests, was higher than for the top 15, in
which there were only two tests.
Results—Overall, the interpreter performed a
type of history matching in which the reservoir
model was iteratively adjusted to match 18 pressure transients distributed along the wellbore.
The estimated permeabilities differed considerably from core permeability, being generally
lower and varying by several orders of magnitude,
from almost 0.02 mD to 1350 mD. No core-derived
permeability measurements were available from
intervals having these extreme values. On the
other hand, the porosity varied little, except
within stylolitic zones. As for most carbonate formations throughout the Middle East, porosity is
not a good indicator of permeability. Of the six
low-porosity intervals on the logs, only two had
permeabilities below 1 mD. Two others were
patchy with significant permeability, one with
kv > kh at X151 ft. In this particular test, the small
pressure response at the probes (less than 0.5 psi
[3.5 kPa]) could be explained only by a superpermeability layer between packer and probe. This
surprising result was supported by an FMI image
of the stylolite, which showed a conductive layer
between two dense streaks, one of which had
gaps in it (figure, page 12). None of this was
apparent from the core data.
250
200
150
100
50
0
0
500
1000
1500
2000
2500
Time, sec
> A comparison between the measured pressure-transient response
at the packer (bottom) and the two probes (top and middle), and the
response computed from the layered model after nonlinear optimization of the parameters. The good agreement validates the parameters
in the model. Other solutions may be possible, but were ruled out on
the basis of other data.
16
Oilfield Review
50387schD02R1
11/29/01
4:43 AM
Page 17
8100
The final model suggested that the layers
should communicate over time. Pressure communication was confirmed by the formation pressure
gradient from MDT pretests (left). The relatively
uniform gradient showed that the stylolites did
not act as pressure barriers. However, good pressure communication does not necessarily mean
that fluids will flow uniformly through the reservoir. As the model showed, at least two highpermeability layers can act as conduits for
injected water. This information has been used
in the full-field reservoir simulator, and to
examine unexpected water breakthroughs in
production wells.
8150
Depth, ft
0.34 psi/ft
8200
8250
8300
3840
3860
3880
3900
3920
Pressure, psi
> Pressure profile from MDT pretests
across the reservoir. The pretests were
taken at the packer and probes as part of
each IPTT. The reservoir has been on production for nearly 20 years. After this
much production, any barriers to pressure
communication should cause the pressure gradient to be much less uniform.
However, the lack of pressure barriers
does not necessarily mean that fluids will
flow vertically with ease.
Mapping Stylolites
Carbonate rocks typically form in shallow, tropical marine environments. In some cases, a formation can extend for hundreds of miles.
Carbonate sediments contain significant
amounts of the metastable minerals aragonite
and magnesium calcite; calcite itself is readily
dissolved and reprecipitated by percolating pore
fluids. Carbonate rocks are, therefore, likely to
undergo dissolution, mineralogical replacement
and recrystallization. These effects vary according to temperature, pore-fluid chemistry and
pressure. Carbonate diagenesis commonly
begins with marine cementation and boring by
organisms at the sediment-water interface prior
to burial. It continues through shallow burial with
cementation, dissolution and recrystallization,
and then deeper burial where dissolution processes, known as pressure solution, may form
such features as stylolites and vugs (below).
The resulting diagenetically altered zones,
whether of lower or higher permeability than the
surrounding formation, are frequently extensive
and affect large sections of a potential reservoir.
For this reason, such features detected by borehole measurements often can be extrapolated
some distance from the well.
The first IPTT example showed how the permeability of stylolites could be determined in a
single well. The next question is how far the layers extend across the field. The depth of investigation of an IPTT depends on transmissivity
(khh/µ) and storativity, and varies with each test.
> Large dissolution cavity. Although carbonates can have large dissolution cavities, they are
not always as large as this.
Autumn 2001
17
50387schD02R1.p18.ps 11/17/01 6:03 PM Page 18
22. Badaam H, Al-Matroushi S, Young N, Ayan C, Mihcakan M
and Kuchuk FJ: “Estimation of Formation Properties
Using Multiprobe Formation Tester in Layered
Reservoirs,” paper SPE 49141, presented at the SPE
Annual Technical Conference and Exhibition, New
Orleans, Louisiana, USA, September 27-30, 1998.
18
C
A
B
D
North pattern
E
F
South pattern
G
Y1
XI
Y2
Stylolites analyzed
XII
Y2A
XIIIA
Y3
XIIIB
Y4
XIV
Y5
XV
XVI
> Field with two pilot gas-injection schemes planned, one in the north, and
the other in the south. The design depended heavily on the properties of the
stylolites, Y1 through Y5. These zones could be easily identified on density
logs and could also be correlated fairly easily across the reservoir. However,
their properties varied, and it was not clear how effective they were as barriers to flow. IPTTs were recorded in seven wells (A through G) to quantify and
map their properties correctly.
A
B
Well
D
C
E
F
G
Y2
Stylolite
In the previous example, the depth of investigation ranged from about 20 to 30 feet [6 to 9 m].
The next example, from another field in the UAE,
examines the lateral extent of barriers by running
IPTTs in several adjacent wells (right).22 The lowporosity, dense stylolites can be correlated easily
between wells, but their actual density varies, so
it is quite possible that their permeability also
varies. The size and number of stylolites are
observed to increase towards the flanks and
toward one side of the field.
A total of 23 IPTTs was recorded in seven
wells in two areas in which pilot gas-injection
schemes were to be implemented. The main
objective was to determine the vertical permeability of four stylolites—Y2, Y2A, Y3 and Y4.
In this case, the MDT tool was configured
with four probes (next page, top). A sink probe S
creates the transient, which is measured by a
horizontal observation probe H at the same depth
but diametrically opposite the sink, and two
observation probes V1 and V2 vertically displaced
from the sink by 2.3 ft and 14.3 ft, [0.7 and 4.4 m].
With this configuration, the storativity, øCt, need
not be assumed in the permeability analysis,
since it can be determined directly from the transients. An FMI image, recorded after the tests,
clearly showed the imprint left by the probe
assemblies on the borehole wall. The tool can be
seen straddling two stylolites. In some tests, the
flow-control module was used to give a constant
flow rate. In others, formation fluids were withdrawn using the pumpout module for a longer
test. Thus, as in the last example, a measured
flow rate was generally available for each test.
In some tests, the sink probe could not withdraw fluids as it was set against a highly localized tight spot. In these cases, the operation was
changed to withdraw fluids from the V1 probe,
using S and V2 as the observation probes. More
recently, interval tests in carbonates have been
performed with the dual packer because its production interval is several thousand times that of
a sink probe. Fluid withdrawal is then possible
even with a high degree of heterogeneity and in
relatively low permeabilities.
Y2A
Y3
Y4
Permeability, mD
0
0 - 0.3
0.3 -1
1- 3
3 - 10
> 10
Not
tested
> Vertical permeability of the four main stylolitic intervals as found by 23 IPTTs run in seven wells.
Oilfield Review
50387schD02R1
11/29/01
4:43 AM
Page 19
Unmoved Oil
Moved Oil
Water
Clay
Depth, ft
Dolomite
V2
Limestone
Anhydrite
Volume, %
100
p. u.
Discontinuous
stylolite
0
Stylolite
X125
Porous
limestone
X150
Stylolite
X175
V1
The interpretation began, as before, by flowregime identification and analysis. Because of the
large volume of data, each test was initially interpreted assuming a single homogeneous but
anisotropic layer. This interpretation is quicker
and gives an average kh and kh/kv over some interval of reservoir rock containing the stylolite. Later,
a more complete study was undertaken using a
multilayer model as in the previous example.
The results showed considerable variation
between the wells (previous page, bottom).
In general, the stylolites were not absolute barriers to flow. For example, the Y2 stylolite was
found to be a barrier in the south of the area, in
Wells F and G, but very conductive in Well E. The
Y2A stylolite was also very conductive in Well E.
FMI images showed that the stylolite and its adjacent layers had a significant number of vugs, a
feature not captured by the cores. Cores generally
found a higher kh than did the IPTT but missed
the vuggy intervals entirely (below). The IPTT
quantified the degree of hydraulic communication and allowed better planning of the pilot
gasflood scheme.
< Volumetric analysis (left) and the four-probe
MDT tool (middle) set across the Y3 stylolitic
interval in Well F. The FMI image (right) was run
after the tests and shows clearly the imprint (circled in green) of the four probe assemblies at
two different tool locations.
X200
S
> Comparison of kh from core plugs with kh from the
corresponding layers of the IPTT interpretation. The
core values were obtained by arithmetic averaging
of the samples within the IPTT interval and by converting from absolute to effective permeability. In a
perfect match, points would lie on the dotted line.
Core-derived kh is generally higher. The core data do
not capture effectively the vuggy layers of Well E.
100
MDT layer permeability, mD
H
Well E - Y2
Well D - Y2
Well E - Y2A
Well G - Y2A
Well E - Y4
10
Layers with
vugs in Well E
1
0.1
0
1
10
100
Core-plug permeability, mD
Autumn 2001
19
50387schD02R1.p20.ps 11/17/01 6:03 PM Page 20
Anisotropy in Sandstones
Sandstones also pose questions about vertical
permeability and barriers to flow. Anadarko
Algeria’s plans for the development of Hassi
Berkine South field called for injection of both
miscible gas/water and possibly water-alternating-gas (WAG) in the future (left). They needed to
know the permeability anisotropy in the field to
improve confidence in the vertical sweep efficiency, and in the recovery values being predicted from numerical models. This information
was required early in the appraisal-drilling program as it affected decisions on facilities and
infrastructure. The reservoir is in the Triassic
Argilo-Gréseux Inferior (TAGI) sandstone.23 The
TAGI is fluvial in origin, with sands that are 5 to
15 m [15 to 50 ft] thick. The area of interest has
two major rock types: a fine- to very fine-grained
sand with interspersed shale laminae, and a fineto medium-grained braided-stream deposit with
discrete claystone layers (next page).
Upon reinjection, gas and water will be taken
mainly by the high-permeability layers. It was
important to determine the degree of gravity segregation expected in the TAGI, and the corresponding influence on vertical sweep, oil
recovery and future production performance.
ALGERIA
TUNISIA
Hassi
Berkine
South
LIBYA
ALGERIA
0
km
0
50
miles 30
> The Hassi Berkine South field in Algeria operated by Anadarko.
2200.00
2200.00
Probe
Pressure, psi
Pressure, psi
Probe
Probe
Pressure, psi
2200
kh/kv = 100
kh/kv = 10
kh/kv = 1
kh = 10
2193
kh = 100
2198.95
0
500
1000
500
Time, sec
1000
500
Time, sec
2200
1000
Time, sec
2200
Packer
Pressure, psi
Packer
Pressure, psi
0
Packer
Pressure, psi
2200
kh = 1000
2199.85
0
kh/kv = 100
kh/kv = 10
kh/kv = 1
kh = 10
2040
0
kh = 100
2184
500
Time, sec
1000
0
kh = 1000
2198
500
Time, sec
1000
0
500
1000
Time, sec
> The pressure response at a dual packer and a vertical probe 6.6 ft [2 m] higher during a drawdown followed by a buildup
modeled for different horizontal permeabilities and anisotropies, but the same flow rate. Note the expanding pressure scale for
each plot from low kh on the left to high kh on the right. Higher kh reduces the signal (causes a smaller pressure drop) at both
packer and probe. Higher kh/kv reduces the signal at the probe but increases it at the packer. The response is complex and
sometimes paradoxical. For example, at the end of a very long flow period, the pressure drop at the vertical probe depends only
on kh, while the drop at the dual packer depends on both kh and anisotropy. Also, no signal at the vertical probe can mean that
there is a layer of either zero or infinite permeability between it and the dual packer. These paradoxes partly explain why simple
analytical solutions are not reliable.
20
Oilfield Review
50387schD02R1
11/29/01
4:43 AM
Page 21
Water
Horizontal Mobility
from IPTT, mD/cp
1
Depth, m
Gamma Ray
0
API
Oil
3000
Anisotropy
kh /kv
Drawdown Mobility
1
140
mD/cp
3000
1
Caliper
4
in.
Probe Pressure (Quartz)
20 5110
psi
5150 1
AIT Resistivity
ohm-m
3000
1
Core
MDT
100
Sandstone
Bound Water
Clay
Volumetric Analysis
100 0
vol/vol
1
Layer 1
XX30
0.1 mm
Layer 2
XX40
0.1 mm
XX50
> The two layers of the15-m TAGI sandstone. Layer 1 is fine-grained with shale laminations; Layer 2 is a medium-grained massive sandstone with
thin claystone beds. The two IPTTs in Layer 1 both give horizontal mobilities below 100 mD/cp and moderate anisotropy. In Layer 2, both tests show
high horizontal mobility, but the top test has low anisotropy, while the bottom test has high anisotropy, most likely due to the thin clay (green highlight
in Track 4) at XX40.2 m between packer and probe. The average core anisotropy is similar, but slightly higher.
For the reservoir engineers simulating the gas
injection, the most critical parameter was the
anisotropy, kh/kv. They were not confident in the
anisotropy from cores, around 10, as this value
was unexpectedly low for such a depositional
environment. The claystone layers were a particular worry since they seemed to extend across
the field. An IPTT offered an attractive solution. It
would test the anisotropy on a much larger scale
than cores, and would provide permeability values at nearly the same vertical scale as the grid
blocks used in the numerical simulation.
Four stations were planned—two in the finegrained, lower resistivity layer; two in the
medium-grained layer, one of which was
designed to straddle a thin claystone.
Autumn 2001
Permeabilities are high, so as part of the pretest
planning it was important to check that sufficient
pressure changes would be seen at the monitor
probe. Using expected values for permeability
and other parameters, simulations showed that if
the flow-control and pumpout modules were
used as flow-rate sources, the resulting pressure
pulse at the monitor probe would be barely measurable (previous page, bottom). A higher flow
rate, and hence a larger pressure response, could
be obtained by flowing directly to a sample
chamber. This is clearly desirable unless it draws
gas out of solution or causes sanding. After further modeling and checking experience elsewhere, the operator ran tests with the dual
packer connected directly to the sample chamber.
The interpreters analyzed each test with a
single-layer model, treating the entire 15-m
sandstone as one layer. With no flow-rate measurement available, a special approach to the
analysis had to be taken. In this approach, the
probe pressure transient is used to estimate kv
and kh, while the packer transient is used to estimate the flow rate and packer-interval skin. Since
the estimates are interdependent, it is necessary
to iterate between the formation parameters at
the probe and the flow rate and skin at the packer
until the results converge.
23. Peffer J, O’Callaghan A and Pop J: “In-Situ Determination
of Permeability Anisotropy and its Vertical Distribution—
A Case Study,” paper SPE 38942, presented at the SPE
Annual Technical Conference and Exhibition, San
Antonio, Texas, USA, October 5-8, 1997.
21
50387schD02R1.p22.ps 11/17/01 6:03 PM Page 22
Well trajectory
Porosity
1
Pretest (k k ) /2
h v
1
Interval test (kh kv) /2
Pressure
6800
0.4
4200
0.3
4150
1
Porosity
0.2
Fracture test
TVD, ft
6860
Permeability, mD
10
6840
4100
0.1
4050
0
7000
4000
Reservoir pressure, psi
6820
6880
6900
0
1000
2000
3000
4000
5000
6000
Horizontal displacement, ft
> Reservoir pressure and permeability from the MDT tool in a horizontal well. Permeability is measured by both pretest drawdowns and
interval pressure-transient tests, the latter being generally an order
of magnitude higher. The pretest permeability may be low due to formation damage or because it is measuring the effective permeability
to filtrate in a water-wet reservoir. Porosity is from openhole logs.
Between 1765 and 5266 ft horizontal displacement, the pressure
is significantly lower than elsewhere, indicating higher depletion and
poorer pressure support from water injection in the reservoir.
The resulting permeabilities reflected the average properties of the formation near each station.
The results near the top two stations were similar,
with horizontal mobility (permeability/viscosity)
near 50 mD/cp and anisotropy near 10. The bottom two stations lay in the medium-grained layer.
They both showed high horizontal mobility, but
while the third station was nearly isotropic, the
fourth station showed a much higher kh/kv.
Assuming that the third station defines the properties of the clean sandstone, it seems likely that
the fourth station is affected by the thin clay at
XX40.2 m, which lies between probe and packer.
Assuming also that the clay acts as an impermeable disk lying around the wellbore, we can estimate its radius as 2 m [6.6 ft].24 By this estimate, it
is quite limited in extent.
The entire TAGI interval in this well was
cored, with horizontal permeability measurements made on plugs every 15 to 30 cm [6 to
12 in.], and vertical permeabilities about every
meter. When the core permeabilities were averaged over the 2-m interval of each MDT station,
they compared well with MDT results, both indicating anisotropy less than 100.25 When shale
laminae or claystone beds are absent, the
anisotropy is less than 10. These results were
22
further supported by five whole-core samples
from other wells in the field.
The MDT data were analyzed further with a
two-layer model, the only multilayer model available at the time. The results were similar. Ideally,
a model with at least five layers is needed to simulate the whole formation. However, in this case
of relatively homogeneous formations, the operator obtained answers that were sufficiently fit for
the purpose with the simpler single-layer model.
The MDT results increased confidence in the
anisotropy values that reservoir engineers were
using for numerical modeling, and thus also in
the predicted performance of the planned injection scheme. In fact, the MDT-measured values
were used directly in the simulator. The field has
been on production since early 1998, producing
in excess of 70,000,000 barrels [11,123,000 m3].
The MDT-derived anisotropy values continue to
be used in the simulator, since the history match
between actual field performance and predictions from the simulator have been excellent.
Although in this case the core anisotropy data
proved to be broadly correct, the confirmation on
a much larger scale was a key piece of information gathered during the appraisal of the field.
Horizontal Wells
Operators rarely acquire permeability data in horizontal wells for reservoir description. However,
horizontal wells often fail to live up to expectation. Some of the many causes are related to
reservoir heterogeneities. In one horizontal well,
6 IPTTs and 19 pretests were run to investigate
why neighboring wells had performed below par
(above).26 Two major features were observed that
could cause poor production—the variation in
reservoir pressure, dropping by as much as
100 psi [689 kPa] in the middle of the well; and
the variation in permeability from 5 to 50 mD for
fairly constant porosity. Clearly, the middle interval has been more depleted and received less
support from water injection into the reservoir.
Upon completion, the middle interval is predicted
to clean up less easily, while injection water will
probably break through first at the toe, or end, of
the well. For these reasons, it was recommended
to complete the well with a casing.
IPTTs are particularly useful for evaluating the
conductivity of faults and fractures in horizontal
wells. Interpreting conventional well tests is difficult due to strong crossflow from pressure and
permeability variations. Borehole images can
determine the location of faults and fractures,
Oilfield Review
50387schD02R1.p23.ps 11/17/01 6:26 PM Page 23
and whether or not they are mineralized. In this
well in a carbonate reservoir, images showed
many vertical fractures but could not determine
their hydraulic conductivities. Pressure differences indicated that while some were closed,
others may have been open. Open fractures could
harm production by quickly drawing water up
into the well.
To test the fractures, the MDT tool was set
with a dual-packer module straddling a set of
fractures seen at 2983 ft (below). The logarithmic
derivative with respect to Horner time for the
buildup test at the packer location indicates a
tool storage-dominated period that ends with a
short slope of –1.0 at 0.015 hr. Following the
storage period, the derivative exhibits a –0.5
slope spherical-flow regime until 0.15 hr, after
which the derivative goes downward, indicating
a higher permeability region. The probe buildup
derivative also exhibits a short spherical-flow
regime, though its value is lower than that of the
packer test. The fact that the probe derivative is
lower but ends at the same time at both packer
and probe indicates a conductive fracture to the
left of the probe. The fracture or fractures must
either be short or have a finite conductivity
because the derivative decreases only gradually.
In addition, the best match to the transients was
achieved with a positive skin—another indication that the fractures opposite the packer were
not open.
All the major fracture intervals were analyzed
in this manner. The combination of fracture analysis, permeability and pressure data is of great use
not just for predicting the performance of a particular well, but also for analyzing how the reservoir
is responding to water injection and deciding
whether to drill horizontal or vertical wells.
Conclusion
Operators are expanding their use of modern
wireline formation testers to determine permeability and help make important well-completion
and reservoir-management decisions. Compared
with conventional cores and well tests, these
testers provide cost-effective information at a
Dual-packer
module
Probe
100
Pressure derivative
Packer
Slope = 1
10
Slope = 1/2
Slope = 1/2
1
Probe
0.1
0.001
0.01
0.1
1
Time since end of drawdown, hr
> Pressure derivatives from probe and packer transients
(bottom) for the analysis of fractures in a horizontal well. The
engineer set the dual-packer (top) astride a set of fractures
that had been interpreted on FMI images (at 2983 ft, see figure
previous page), and performed an IPTT. The probe derivative
is less than the packer derivative, but spherical flow ends at
the same time on both transients. These observations along
with the positive skin are best explained if the fractures
between the packers are not hydraulically conductive, and
if there is a conductive fracture to the left of the probe.
Autumn 2001
scale intermediate between the two. This information is critical for evaluating the effect of
reservoir heterogeneities, baffles and conduits.
Wireline formation testers measure permeability in different ways, depending on the hardware configuration. The mini-DST is particularly
useful for evaluating small intervals at a fraction
of the cost of a full well test. The interval pressure-transient test provides the most reliable and
extensive permeability information from these
tools. With recent developments in software and
interpretation techniques, interval tests can now
evaluate highly layered formations, horizontal
wells, and even gas reservoirs.27 The latter have
often been considered too challenging because
of the high compressibility and mobility of the
fluid. In addition, the risk of sticking the tool—
the fear of many operators—has been reduced
through the use of risk-assessment software.28
Currently, engineers are seeking to improve
results in formations with high mobilities, heavy
oil or unconcolidated sands—all difficult but not
impossible cases. Work continues on the perennial problem of scaling up from cores to tests,
and of integrating interval-test results with other
data. Attempts are being made to measure in situ
the variation of effective permeability with water
saturation, using the fluid fractions measured
while sampling in combination with openhole
logs and interval-test data. As long as reservoirs
continue to be heterogeneous and permeability
distribution remains an issue—both virtual certainties—wireline formation testers will be
needed to evaluate them, and improvements will
continue to be made.
—JS/LS
24. Goode PA, Pop JJ and Murphy WF III: “Multiple-Probe
Formation Testing and Vertical Reservoir Continuity,”
paper SPE 22738, presented at the SPE Annual Technical
Conference and Exhibition, Dallas, Texas, USA, October
6-9, 1991.
25. A thickness-weighted arithmetic average was used for
the horizontal permeability, and a thickness-weighted
harmonic average for the vertical permeability.
26. Kuchuk FJ: “Interval Pressure Transient Testing with
MDT Packer-Probe Module in Horizontal Wells,” paper
SPE 39523, presented at the SPE India Oil and Gas
Conference and Exhibition, New Delhi, India, February
17-19, 1998.
27. Ayan C, Donovan M and Pitts AS: “Permeability and
Anisotropy Determination in a Retrograde Gas Field to
Assess Horizontal Well Performance,” paper SPE 71811,
presented at the Offshore Europe Conference,
Aberdeen, Scotland, September 4-7, 2001.
28. Underhill WB, Moore L and Meeten GH: “Model-Based
Sticking Risk Assessment for Wireline Formation Testing
Tools in the U.S. Gulf Coast,” paper SPE 48963, presented
at the SPE Annual Technical Conference and Exhibition,
New Orleans, Louisiana, USA, September 27-30, 1998.
23
50387schD03R1.p24.ps 11/17/01 8:30 PM Page 24
Quantifying Contamination Using Color of
Crude and Condensate
Establishing the level of oil-base and synthetic mud-filtrate contamination in fluid
samples is critical for obtaining meaningful data on fluid properties. New tools and
techniques now allow real-time, quantitative measurement of contamination in gascondensate and oil reservoirs.
R. John Andrews
Hibernia Management and
Development Company Ltd.
St. John’s, Newfoundland, Canada
Gary Beck
BP
Houston, Texas, USA
Kees Castelijns
London, England
Andy Chen
Calgary, Alberta, Canada
Myrt E. Cribbs
ChevronTexaco
Bellaire, Texas
Finn H. Fadnes
Jamie Irvine-Fortescue
Stephen Williams
Norsk Hydro, ASA
Bergen, Norway
Mohamed Hashem
Shell
New Orleans, Louisiana, USA
24
A. (Jamal) Jamaluddin
Houston, Texas
Andrew Kurkjian
Bill Sass
Sugar Land, Texas
Oliver C. Mullins
Ridgefield, Connecticut, USA
Erik Rylander
Belle Chase, Louisiana
Alexandra Van Dusen
Harvard University
Cambridge, Massachusetts, USA
For help in preparation of this article, thanks to Victor
Bolze, Reinhart Ciglenec, Hani Elshahawi, Troy Fields, Gus
Melbourne, Julian Pop and Rod Siebert, Sugar Land, Texas;
Peter Kelley, ChevronTexaco, Houston, Texas; and Toru
Terabayashi, Fuchinobe, Japan.
AIT (Array Induction Imager Tool), CHDT (Cased Hole
Dynamics Tester), CMR (Combinable Magnetic Resonance),
FFA (Field Fingerprint Analyser), LFA (Live Fluid Analyzer),
MDT (Modular Formation Dynamics Tester), OCM (Oil-Base
Contamination Monitor), OFA (Optical Fluid Analyzer),
Platform Express and TLC (Tough Logging Conditions) are
marks of Schlumberger. RCI (Reservoir Characterization
Instrument) is a mark of Baker Atlas. RDT (Reservoir
Description Tool) is a mark of Halliburton.
1. Joshi NB, Mullins OC, Jamaluddin A, Creek J and
McFadden J: “Asphaltene Precipitation from Live Crude
Oil,” Energy and Fuels 15, no. 4 (2001): 979-986.
In deepwater areas, an oil or gas company may
spend tens of millions of dollars drilling a well to
prove the presence of hydrocarbons, and then
plug and abandon the well almost immediately.
The operator may take years designing and building facilities before drilling another well in the
field. Exploration wells provide a narrow window
of opportunity for collecting hydrocarbon samples to make development decisions; therefore,
obtaining high-quality samples is imperative
whether the prospect is in deep water or on the
continental shelf, in China, Canada, the Caspian,
or elsewhere.
Testing well production is a good way to
obtain fluid samples, but that is not always feasible for economic or environmental reasons.
Downhole samples define fluid properties that
are used throughout field development.
Estimates of hydrocarbon volume, bubblepoint
pressure and gas/oil ratio (GOR), simulation of
reservoir flow and placement of wells all depend
on formation-fluid properties. Hydrate, asphaltene and wax formation must be controlled or
treated. Presence of corrosive gases affects the
choice of materials for flowlines and surface
facilities. These examples illustrate the wide
impact that hydrocarbon composition and behavior have on planning a new field.1
Oilfield Review
50387schD03R1
11/29/01
3:28 AM
Page 25
Openhole-wireline or drillstring-conveyed formation testers analyze selected fluid properties
downhole and acquire small volumes of reservoir
fluid for later testing in a laboratory. However,
mud filtrate invades the formation during drilling,
so these fluid samples usually are contaminated.
During the past few years, real-time methods
have been developed as part of the openholelogging suite of services to analyze sample
contamination. These methods ensure that representative fluid samples are collected and
minimize tool-sticking risks by introducing efficiencies in sample collection. Until recently,
these sampling methods were unreliable in holes
drilled with oil-base and synthetic muds or in formations with high GOR.
This article reviews the requirements and
challenges in sampling reservoirs and reports on
advances in evaluating sample contamination.
Except where explicitly stated to be contamination from water-base mud, this article discusses
oil-base or synthetic-base mud-filtrate contamination. We describe a technique for determining
the time required to collect an acceptable fluid
sample at a given sampling station and show
how proven sample-contamination measurements can be extended to high-GOR fluids and
condensates. Quantitative contamination measurement is illustrated with case histories from
offshore Newfoundland, Canada, the Gulf of
Mexico and the Norwegian North Sea.
Obtaining Downhole Fluid Samples
Formation fluid samples provide important data
to optimize operator investment in both upstream
and downstream facilities. Laboratory measurements establish standard fluid properties such as
pressure-volume-temperature (PVT) behavior, viscosity, composition and GOR. In fields destined
for subsea development, flow assurance is a
major concern, so tests are performed to evaluate gas and solids content. Hydrogen sulfide
[H2S] and carbon dioxide [CO2] in oil require special handling and materials. Temperature and
pressure changes in pipelines can lead to asphaltene and wax precipitation and deposition, and
low seafloor temperatures can induce hydrate
Pumpout module
Sample-chamber
modules
Multisample
modules
LFA Live Fluid
Analyzer module
Hydraulic-power
module
> The MDT Modular Formation Dynamics Tester
tool configured for fluid-sample collection.
Autumn 2001
Single-probe module
25
50387schD03R1.p26.ps 11/17/01 8:30 PM Page 26
OFA module
Precipitated solids
in the separator
OCM module
Contamination
measurement
> Components of MDT optical analysis modules.
Subsea wellhead
Buildup of solids
in the wellbore
Asphaltene deposition in
the near-wellbore region
> Transport hazards from reservoir-fluid constituents while flowing to surface.
Asphaltenes, waxes and hydrates can form during fluid transport to surface.
Depositing such solids clogs tubulars or blocks pores in the formation. Solids
also precipitate in separators under certain conditions. In addition, commingling fluids at wellheads can generate unstable conditions leading to precipitation of solids.
26
Gas flag
Water flag
Oil flag
LFA module
Color channels
Methane flag
Methane channel
Gas flag
Gas refractometer
Water flag
Water channels
Oil flag
Oil channel
Solids in subsea flowlines
formation. Commingling different crude oils
through satellite tiebacks can dramatically alter
fluid properties (above).
The data-acquisition process must include
fluid characterization to get the most out of every
prospect. Taking fluid samples early in the life of
a well ensures that fluid composition and properties are available for timely input to field planning decisions. If fluid properties will affect
facilities or transport, accurate fluid analysis
gives an operator the opportunity to mitigate or
eliminate problems through changes in production design, or to manage them through ongoing
treatments such as heating pipelines—a choice
between upfront capital expenditures and ongoing operating expenses.
In some fields, fluid samples can be obtained
during a drillstem test (DST) or, after a well is
flowing, a production test. In some cases, a well
must be completed before a flow test, which
can cost tens of millions of dollars in deepwater Gulf of Mexico wells. In areas such as the
Grand Banks, offshore Newfoundland, Canada,
operators want to minimize operation times to
avoid risks such as harsh seas and iceberg hazards. Environmental concerns restricting flaring
and removing fluids from the rig also restrict use
Color channels
Gas refractometer
Water channels
Oil channel
of DSTs and production tests. The cost and risk of
DSTs lead operators to use wireline tools for
fluid-sample acquisition.
A major problem in downhole fluid-sample
collection is contamination from drilling-mud filtrate entering a tool with reservoir fluids.
Contamination from water-base mud (WBM) can
be discriminated easily from reservoir oil. In
many of today’s high-risk wells, oil-base muds
(OBMs) and synthetic oil-base muds (SBMs) are
used to ensure compatibility with shales,
improve wellbore stability and increase drilling
speed. OBM and SBM filtrates mix with reservoir
crude, making quantification of contamination
much more difficult than when using WBM. Fluid
properties are often extrapolated to an uncontaminated condition by mathematically removing
the contaminant from the distribution of constituents. However, extrapolation from high levels of contamination is risky—most companies
avoid liquid-phase contamination greater than
10% on a volume-to-volume basis.
Several commercially available tools have
fluid-sampling capabilities, including the
Schlumberger MDT Modular Formation Dynamics
Tester tool, the Baker Atlas RCI Reservoir
Characterization Instrument tool, and the
Halliburton RDT Reservoir Description Tool
sonde. Most wireline formation testers press a
probe against the borehole wall at a specified
depth, pump down the formation and draw in
fluid for evaluation, and then collect samples
when desired fluid characteristics are reached.2
With a probe securely pressed against the
borehole wall, a short, rapid pressure drop
breaks the mudcake seal. Normally, the first fluid
drawn into the tool will be highly contaminated
with mud filtrate (next page, top). As the tool
continues to withdraw fluid from the formation,
the area near the probe cleans up, and reservoir
fluid becomes the dominant constituent. The time
required for cleanup depends on many parameters, including formation permeability, fluid viscosity, the pressure difference between borehole
and formation, and the duration of the pressure
difference during and after drilling. Increasing
pump rate can shorten the cleanup time, but the
rate must be controlled carefully to preserve the
reservoir-fluid condition. Because many factors
affecting cleanup time have unknown values,
determining the contamination level during a logging job is crucial to obtaining good samples.
The versatile Schlumberger MDT system performs a variety of functions, depending on which
modules are joined together. The tool’s primary
purposes are to obtain formation-fluid samples,
to measure formation pressures at given points
in the reservoir and to estimate permeability in
situ. For a description of use of the tool for permeability measurement and description of other
tool modules, see “Characterizing Permeability
with Formation Testers,” page 2.
The OFA Optical Fluid Analyzer system in the
MDT tool has provided a qualitative measure of
contamination since its introduction in 1993.
Schlumberger has developed the OCM Oil-Base
Contamination Monitor technique to predict the
time needed to achieve an acceptably low level
of contamination at a given sampling station.
This reliable new technique monitors sample
contamination quantitatively, adding confidence
to these crucial contamination measurements.
Oilfield Review
50387schD03R1.p27.ps 11/17/01 8:30 PM Page 27
2. For more on use of the MDT tool for downhole fluid sample analysis: 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.
Badry R, Fincher D, Mullins O, Schroeder B and Smits T:
“Downhole Optical Analysis of Formation Fluids,” Oilfield
Review 6, no. 1 (January 1994): 21-28.
Oil
Filtrate
t3
Oil cone
t2
OD
Filtrate
t1
Oil cone
t1 t2 t3
Time
Filtrate
The new LFA Live Fluid Analyzer module adds
a methane detector that provides a more definitive measure of gas content in the oil phase and
allows calculation of GOR. This module can be
used to ensure that the fluid remains in single
phase during sampling; dropping pressure below
the bubblepoint would make the fluid unrepresentative. The quantitative OCM contamination
measurement can be used with either an LFA or
OFA module (previous page, right).
Modular reservoir sample chambers (MRSCs)
are available to collect large samples (below).
Multiple 6-gallon [22,712-cm3] chambers can be
run at the bottom of the tool string to act as dump
chambers. Samples for PVT analysis are more
commonly collected in smaller chambers. A multisample module (MRMS) allows collection of six
easily removable sample bottles (MPSR) that are
certified for transport by the US Department of
Transportation (DOT) and by Transport Canada.
The 450-cm3 [0.12-gal] MPSR bottle is reduced
to 418 cm3 [0.11 gal] when an agitator is added
to improve fluid mixing in the laboratory. The
Schlumberger Oilphase single-phase multisample chamber (SPMC) can be used in the
MRMS when keeping a reservoir fluid sample in
> Drawing in filtrate. The MDT probe pressed against a borehole wall is the source of a pressure
drawdown, pulling fluids into the tool. Filtrate near the probe enters first, but as the pressure sink
expands, a higher proportion of fluid is reservoir fluid. The optical density (OD) increases as darker
crude oil replaces the more transparent mud filtrate.
MRSC
H 2S
MRMS
Non-H2S
MPSR
SPMC
Maximum hydrostatic
pressure
20- and 25-kpsi [138and 172-MPa] options 1
14 kpsi [97 MPa]
10 kpsi [69 MPa]
20- and 25-kpsi
options 1
20- and 25-kpsi
options 1
Sample pressure
20 kpsi
14 kpsi
10 kpsi
20 kpsi
20 kpsi
Downhole temperature
204°C [400°F]
204°C
204°C
204°C
204°C
54°C [130°F]
Not allowed
100°C [212°F]
204°C
2
Surface heating
allowed
77°C [170°F]
Volume
1- and 2.75-gal [3785and 10,410-cm3] options
1- and 2.75-gal
options
6 gal [22,712 cm3] 3
450 cm3 [0.12 gal] 4
250 cm3
[0.07 gal]
Transportable 5
No
No
No
Yes
No
Pressure
compensated 6
No
No
No
No
Yes
1
2
3
4
5
6
The 25-kpsi limit is for special high-pressure modules, and the sampling must be done in low-shock mode—the bottle is compensated to hydrostatic pressure behind the piston.
Only Schlumberger Oilphase is allowed to heat chambers above 54°C [130°F].
Six-gallon bottles must be run on the bottom of the string. Several bottles can be combined in one string.
Addition of an agitator reduces this volume to 418 cm3 [0.11 gal].
Transportable indicates US Department of Transportation Exemption and Transport Canada Permit for Equivalent Safety.
Compressed nitrogen is used to compensate the sample pressure so it does not decrease as much upon cooling when brought to surface.
> Sample bottles available for the MDT tool.
Autumn 2001
27
50387schD03R1.p28.ps 11/17/01 8:31 PM Page 28
Black Oil Isn’t Always Black
Oils have color—black, brown, red, tan and even
green crude oils have been seen. The hue and
intensity of light transmitted or reflected from
crude oil or gas condensate depend on the light’s
interaction with molecules and molecular bonds in
the fluid. Measurements of this interaction can be
used to distinguish oils of different compositions.
The unit of light absorption or optical density
(OD) is the logarithm of the ratio of incident-light
to transmitted-light intensity. Therefore, darker
fluids have higher OD, and a one-unit increase in
OD represents a factor of ten decrease in transmittance. An OD of zero indicates all light is
transmitted, while an OD of two represents 1%
transmission. A fluid’s OD varies with the wavelength of incident light.
Reduction of transmitted-light intensity can be
caused by one of two physical processes. Some
light is scattered by particles in the fluid; scattering outside the optical path to the detector
decreases intensity. Light also can be absorbed by
molecules in the fluid. The MDT optics relies on
differences in absorption in visible and nearinfrared portions of the electromagnetic spectrum
to discriminate fluids in the flowline.
Pure, light hydrocarbons such as pentane are
essentially colorless; they do not absorb light in
the visible spectrum. Condensates may be clear
or lightly shaded reddish-yellow to tan, because
they absorb more from the blue end of the spectrum than from the red end. Heavier crude oils,
which contain more complex molecules, absorb
light strongly throughout the visible region, making them dark brown or black.
Light with a wavelength in the visible or nearinfrared spectra, referred to as the color region,
interacts with a molecule’s electronic energy
bands. Compared to less complex molecules,
larger and more complex aromatic hydrocarbon
molecules, such as asphaltenes and resins,
absorb light having longer wavelengths.3 Because
heavier oils contain more aromatic compounds,
they tend to have darker coloring than less dense
oils and condensates (above). Waxes are colorless, but if the molecules are long enough, they
will scatter light and appear white.
28
3.0
2.5
2.0
Optical density
single phase from the downhole collection point
to the PVT laboratory is necessary. After the MDT
pumpout module fills a SPMC chamber at formation pressure, a preset nitrogen charge is
released. Acting through a piston floating on a
synthetic oil buffer, the nitrogen adds sufficient
overpressure to keep the fluid in single phase
during retrieval to surface.
Condensates
Black oils
Asphalts
1.5
1.0
0.5
0
500
1000
1500
Wavelength, nm
2000
2500
> Optical density of various oils. The OD spectrum of hydrocarbons is related
to the amount of aromatics, which also relates to API gravity. Gas condensates have little or no color absorption beyond about 500 nanometers (nm).
The range of oils grades through increasingly dense black oils having higher
color absorption out to asphalts, which absorb strongly even into the nearinfrared region. All oils and condensates absorb near 1725 nm. The hydrocarbon
peak from 2300 to 2500 nm is beyond the region covered by the MDT channels.
Despite the differences in optical absorption
of various reservoir oils caused by composition,
there is a common behavior. Electronic absorption
generally decreases as wavelength increases.
The OD decay in the visible and near-infrared
region can be characterized by a single parameter, which can be thought of as the color of the oil.
To understand how OD measurements can be
used to quantify contamination, it is important to
distinguish between absorption in the color
region by two kinds of hydrocarbons: complex
aromatics and saturated aliphatics. Complex aromatics contain carbon rings with both single and
double carbon-carbon bonds, which are excited
by visible and near-infrared light. Aliphatic compounds are open chains of carbon atoms. If all the
carbon-carbon connections are single bonds and
other bonds are with hydrogen, the aliphatic
molecule is termed saturated. Only high-energy
ultraviolet light can excite saturated aliphatic
molecules, so they have a low OD in the color
region of the spectrum.
Black oils contain many complex aromatic
compounds, whereas natural OBMs comprise
mostly saturated compounds; SBMs are made
only from saturated aliphatics. The difference in
chemical composition between reservoir-crude
oil and drilling-mud filtrate makes OD a good
measure of filtrate contamination in crude oil.
Exciting Molecules
Water can be distinguished from oil easily,
because it is highly absorbing at near-infrared
wavelengths around 1445 and 1930 nanometers (nm), where oil is relatively transparent
(next page, top). Oil has a strong absorption peak
around 1725 nm, where water does not. These
peaks come from the interaction of light with
vibrational energy bands in carbon-hydrogen
bonds for oil and oxygen-hydrogen bonds for
water. Molecules containing such a bond absorb
photons of the proper wavelength, and the photon energy is converted into molecular vibration.
Monitoring absorption at these three wavelengths differentiates between water and oil.
Hydrocarbon compounds comprise linked
chains, branches or rings of carbon atoms, each
having hydrogen atoms attached. Typically, a carbon atom will bond with two other carbon atoms
and two hydrogen atoms. Carbon atoms at the
end of a molecule will have three hydrogen
atoms attached, while those at a branch, connecting with three other carbon molecules, will
have only one hydrogen bond. Methane is a
single carbon molecule with four hydrogen
atoms attached.
The oil peak in Channel 8 of the OFA module
measures molecular absorption of light by carbon
atoms having two hydrogen atoms attached,
which are the primary constituents of reservoir
Oilfield Review
50387schD03R1.p29.ps 11/17/01 8:31 PM Page 29
4
Water peak
Color or electronic
absorption region
Vibrational
absorption region
3
M or
2
e co
m pl
Oil peak
es
cul
ole
ex m
Optical density
Water peak
H-C-H vibrational peak
H H H
C C C
H H H
Methane peak
H
H C H
H
1
0
Channel
number
0
1
2
3
500
4
5
6
1000
7
0'
8
9
1500
2000
Wavelength, nm
> Absorption spectrum. The MDT tool monitors light absorption starting in visible wavelengths and extending into the near-infrared. The ten channels of the OFA module, numbered 0 through 9, are shown. In the color region on the left, crude oils have a rapidly decaying absorption, caused by
interaction of light with electrons in the molecules. More complex aromatic molecules (green shapes) absorb at longer wavelengths. Channels 6
and 9 are tuned in the middle of molecular vibrational peaks for water; Channel 8 is in the molecular vibration peak for the CH2 bond in hydrocarbons. Channel 0’, which replaces Channel 0 in the LFA module, is tuned to the methane peak.
3. Mullins OC: “Optical Interrogation of Aromatic Moieties
in Crude Oils and Asphaltenes,” in Mullins OC and Sheu EY:
Structures and Dynamics of Asphaltenes. New York,
New York, USA: Plenum Press, 1998.
4. A live crude oil evolves significant quantities of gas
when its pressure and temperature are lowered. A dead
oil does not evolve gas at atmospheric pressure and
room temperature. Stock-tank oil, the liquid emerging
from the final surface separator, contains little gas.
Autumn 2001
0.8
Oil peak
0.7
Methane
n-Heptane
Methane-heptane mix
0.6
0.5
OD
oils. A high-resolution optical spectrometer
reveals this oil peak in much greater detail, showing several absorption peaks in hydrocarbon fluids
(right). Although methane has some absorption at
the oil peak, there is no absorption by hydrocarbons with more than one carbon atom at the
methane peak. This provides an ideal discriminator for methane content in live crude oils—utilized by a new MDT tool, the LFA Live Fluid
Analyzer module.4 The detection channel tuned to
that wavelength replaces the OFA module’s shortest wavelength color band in Channel 0.
Methane peak
0.4
0.3
0.2
0.1
0
Wavelength, nm
> High-resolution vibrational absorption spectrum of heptane, methane and a
mix of the two. Heptane (green) does not absorb light at the CH4 methane
peak. Methane absorption (red) at the CH2 oil peak is low. Absorption of a mixture of the two (black) is the sum of the individual absorptions, according to
the Beer-Lambert law. The LFA module has a channel set at the methane peak.
29
50387schD03R1.p30.ps 11/17/01 8:31 PM Page 30
Light-emitting
diode
Gas refractometer
Lamp
Water
Fluid flow
Fluid flow
Gas
Oil
Oil
Optical density detectors
> Optical detectors. Light passes through a sapphire window and reflects off the surface in contact with the fluid flowline into the gas refractometer. The reflection angle is set so gas reflects
much more strongly than oil or water. Another light path passes through a flowline into a series
of filters to detect absorption, or optical density, in the visible and near-infrared spectra.
0.40
0.36
OD
Channel 4
0.32
0.28
Channel 4 minus Channel 6
0.24
200
400
600
800
1000
Pumping time, sec
1200
1400
> Removal of scattering. To remove scattering from the OD signal, a nearby
channel at longer wavelength, which has less color absorption but the same
amount of wavelength-independent scattering, is subtracted. In this case,
the signal from Channel 6 (not shown) is subtracted from Channel 4 (yellow)
resulting in a data curve (red) that is fit to the OCM prediction (black).
30
The Key to Quantifying Contamination
The MDT tool includes an optical module with
two devices designed to monitor contamination
in OBM systems. A gas refractometer uses light
from a diode reflected off a sapphire window to
qualitatively identify the fluid phase in a flowline
(left). At the selected angle of incidence, the
reflection coefficient is much larger when gas is
in contact with the window than when oil or
water contacts it.5
The second detector in the OFA module uses
transmitted light to evaluate absorption characteristics of a fluid. A high-temperature tungsten
halogen lamp provides a broadband source of
light that passes along optical guides and
through a 2-mm thick optical chamber in the
flowline. The distribution of transmitted light is
recorded at 10 wavelengths in the visible and
near-infrared spectra. Two of these channels
detect the strong water-absorption peaks, indicating water content in the fluid when compared
with the strong hydrocarbon-absorption peak.
Discriminating gas and water from oil is simpler than distinguishing between crude oil and
OBM or SBM filtrate, because crude, OBM and
SBM all absorb strongly at the oil peak near
1725 nm. Fortunately, oils have different color
according to the quantity of large, complex aromatic compounds they contain. This affects
absorption in the MDT spectrometer in the
shorter wavelength channels constituting the
color region. Since SBM and OBM contain simple
aliphatic compounds, their absorption in these
channels is small.
In most cases, when the MDT tool first begins
drawing fluid from a formation, the OD is high
due to light scattering off mudcake solids in the
fluid. After a few seconds, the OD falls to a low
value, and then increases slowly as the mud filtrate drains from the formation near the probe
and is replaced by darker crude oil.
Particles of mudcake or other solid material
generate noise in the absorption channels.
Scattering caused by these particles is wavelength-independent, so the effect can be
removed by subtracting a nearby channel. In the
color region, absorption decreases quickly
enough that skipping a channel and subtracting
from the next one down removes noise due to
scattering without significantly affecting the signal (left). The result is a smoothly varying contamination curve.6
The change in OD as reservoir crude replaces
mud filtrate in the flowline follows the BeerLambert law, which states that a mixture of two
Oilfield Review
50387schD03R1
11/29/01
3:31 AM
Page 31
Sample
Pumping time
OFA contamination
Laboratory contamination
1
2
3
4
5
695 sec (12 min)
940 sec (16 min)
1264 sec (21 min)
1681 sec (28 min)
2250 sec (37 min)
17%
13%
12%
9%
8%
22%
17%
13%
11%
10%
3.0
50
3.0
Contamination
Optical density
40
2.6
30
2.2
20
1.8
2.0
Data and OCM fit
OD
Contamination, percent
OD
2.5
1.5
10
1.4
Acceptable contamination level
0
1.0
500
1000
1500
Pumping time, sec
2000
1.0
2500
Pumping time, sec
> Quantitative prediction of contamination. Fluid samples were taken at five times during cleanup. Color channel data from the OFA module are fit using
the OCM model (left) to determine contamination cleanup (right). The OCM prediction of contamination levels agrees well with laboratory contamination
measurement (table).
Autumn 2001
OD at
specific
wavelength
100%
crude
oil
OD5
OD5
OD4
OD4
OD3
OD2
OD1
OD3
OD2
OD1
Wavelength
0
0
100
0
Pumping time
% OBM contamination
100%
OBM
filtrate
Optical density
oils has an OD that is a linear, volumetrically
weighted combination of the two individual
ODs, evaluated at each wavelength. A change
in OD is directly related to a change in composition (right).
Because most OBMs and SBMs mainly contain simple aliphatic compounds, their OD is
effectively zero except in the lowest MDT channels. With one endpoint determined, quantitative
evaluation of contamination through OD requires
a method for finding the other endpoint—the OD
of uncontaminated crude. This comes from
understanding the way fluids move during
cleanup. Fluid withdrawal through the probe creates an expanding pressure sink around the wellbore.7 The OCM analysis fits the cleanup data
with a curve—having a specific shape based on
the physics of the tool and wellbore—to determine the remaining amount of filtrate contamination. In one well, five samples were captured in
the MDT tool at different times during cleanup.
The laboratory results show contamination
results consistent with the OCM model (above).8
∞
> Beer-Lambert mixing. Light absorption for crude oil (brown) is greater than for OBM filtrate (yellow)
(left). The Beer-Lambert law says that the optical density (OD) of mixtures of the two (shades from
yellow to brown) is related to the relative proportion of the two fluids. As the fluid cleans up, the OD
increases from the OBM value OD1 asymptotically to the crude-oil value OD5 (right).
5. Badry et al, reference 2.
6. Mullins OC, Schroer J and Beck GF: “Real-time
Quantification of OBM Filtrate Contamination During
Openhole Wireline Sampling by Optical Spectroscopy,”
Transactions of the SPWLA 41st Annual Logging
Symposium, Dallas, Texas, USA, June 4-7, 2000,
paper SS.
7. Hashem MN, Thomas EC, McNeil RI and Mullins O:
“Determination of Producible Hydrocarbon Type and Oil
Quality in Wells Drilled With Synthetic Oil-Based Muds,”
SPE Reservoir Evaluation and Engineering 2, no. 2
(April 1999): 125-133.
8. 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, Texas, USA, October 1-4, 2000.
31
50387schD03R1.p32.ps 11/17/01 8:31 PM Page 32
1.25
0.26
Color channel
1.20
1.15
Color OD
0.24
1.10
Methane channel
0.23
1.05
Methane OD
0.25
0.22
1.00
500
1000
1500
2000
3500
0.21
3500
3000
Pumping time, sec
> Contamination prediction in a Gulf of Mexico well. After noise is removed
from an LFA color channel (red) and the methane channel (blue), each data
set is fit to the OCM prediction (smooth curves). For this sample, the color
data predict 4.9% contamination and the methane data predict 6.2%. The 5.5%
average agrees with 4.3% contamination from a GC in the laboratory.
100
2.00
1.75
1.50
80
4
70
3
60
2
1.25
Color OD
5
1
50
1.00
0
1
0
100
0
2
C3
h4
a
n 5
n
e 6
l
n
u 7
m
b
0.50
00 c
se
,
00
e
m
ti
g
in
30
40
00
e 8
r
20
p
P
9
30
m
u
Start fit
0.25
40
20
End fit
0.75
Contamination, %
Optical density
90
0.00
10
0
0
500
1000
1500
2000
2500
3000
3500
4000
Pumping time, sec
> Avoiding long cleanup times. After the MDT tool had pumped from the formation for about an hour, the OCM software indicated about 18% contamination (blue curve), and an additional 41⁄2 hours to achieve less than 10% contamination. The inset shows the OD measurement for Channels 0 through 9
(shaded green). Channel 4, which has the greatest change in OD during
cleanup, was used for the fit after subtracting Channel 6 to remove scattering
from large particles (red curve). The vertical dashed lines at the left and right
of the plot indicate the range over which the OCM method fit the data.
32
Like the other optical-detection bands, the
methane channel of the LFA module displays a
high OD as mud solids pass through a tool’s flowline after pumping begins. Since drilling muds do
not contain methane naturally, the initial high
concentration of filtrate drawn into the MDT tool
during cleanup results in a substantial drop in the
OD recorded in the methane channel. As reservoir fluid replaces filtrate in the line, the signal
OD increases in proportion to the oil’s methane
content, generating the same curve shape as
cleanup with the OFA tool (left).
Time for complete cleanup cannot be predicted before the logging run, because there are
too many unknown reservoir variables. For example, there is not a direct relationship between
formation permeability and cleanup time.
Although fluid can be pumped quickly from a
high-permeability formation, which would imply
a short cleanup time, that high permeability may
have allowed mud filtrate to penetrate deeply
into the formation before the wireline run. In that
case, cleanup time could be long. Collecting fluids close to a shale stringer can shorten cleanup
time, since the shale provides a flow barrier,
allowing collection of less contaminated reservoir fluid farther away from the wellbore.
The ability of the OFA and LFA modules to
quantify contamination levels while pumping
allows sampling decisions to be made in real
time. The OD for all channels is transmitted to
surface at high rate, and the OCM software
updates its analysis every 20 seconds. Once sufficient data have been acquired, the software
selects the color channel that will provide the
best fit to the expected trend and shows the
degree of contamination and the time required to
achieve an acceptably low level of contamination.
In a Gulf of Mexico well, the MDT probe was
set within a massive sand, and the tool measured
a mobility of 87 millidarcies per centipoise
(mD/cp). After pumping for 71 minutes, the OCM
software predicted an additional 41⁄2 hours pumping time to achieve an acceptable level of 10%
contamination (left). Rather than wait or waste a
sample bottle on highly contaminated fluid, the
operator chose to move to another level within
the same formation.
The tool was moved 44 ft [13 m] lower in the
formation. The mobility was higher, 256 mD/cp.
Contamination dropped to 9% within 132 minutes, and samples taken at this location were
acceptable for PVT analysis (next page, bottom).
Oilfield Review
Page 33
9. Mullins O, Beck GF, Cribbs M, Terabayashi T and
Kegasawa K: “Downhole Determination of GOR on
Single-Phase Fluids by Optical Spectroscopy,”
Transactions of the SPWLA 42nd Annual Logging
Symposium, Houston, Texas, USA, June 17-20, 2001,
paper M.
0.24
0.85
Methane channel
0.22
0.75
0.20
Methane OD
0.80
0.70
0.18
Color channel
0.65
0.16
500
1000
1500
2000
2500
Pumping time, sec
0.60
3500
3000
> Wavelength-dependent scattering. The optical absorption response in
the pumping period between 1000 and 1500 seconds indicates some scattering remains even after subtracting a baseline channel. This wavelengthdependent response is stronger in the color channel (purple) than in the
methane channel (blue). The noise in the data after 2500 seconds occurred
during sample collection. The OCM method was still able to fit the data,
predicting 7% contamination based on the average of color and methane
data of 7.9% and 6.0%, respectively.
100
1.2
90
1.1
1.0
5
0.9
4
0.7
70
3
60
2
50
1
0.6
0
0.5
1
0
2
C 3
h 4
a
n 5
n
e 6
l
n
u 7
m
b
e 8
r
0.4
0.3
0.2
0.1
40
Contamination, %
0.8
80
100
0
c
20
,
00
se
30
e
m
g
20
in
75
p
00
m
u
P
9
ti
End fit
Comparing Contamination at Surface
Samples are collected to determine properties of
reservoir fluids such as PVT behavior. Mud filtrate
mixed in the sample must be accounted for to
arrive at reasonable estimates of reservoir-fluid
properties. The OFA and LFA modules measure
0.90
Optical density
Scattering Light
Scattering from particles smaller than the incidentlight wavelength—several hundred nanometers
diameter—depends on the wavelength of incident light. The intensity of this scattering
increases with decreasing wavelength. This
effect, called Rayleigh scattering, gives the sky
its blue color.
Wavelength-independent scattering is removed
by channel subtraction, but this leaves some
wavelength-dependent Rayleigh scattering.
For the OCM-color procedure, a longerwavelength channel is subtracted, but for
the OCM-methane procedure, the subtracted channel is at a shorter wavelength. Since
one procedure slightly overcorrects for wavelength-dependent scattering and the other
slightly undercorrects, averaging OCM-color
and OCM-methane contamination values from
the LFA tool tends to remove some of that
scattering effect (right).
Discrepancies between the contamination
determinations indicate the need to look more
closely at other channels to identify the cause
before collecting a fluid sample. Methane detection has been shown to be valid for fluids with
GOR as low as 700 scf/bbl [126 m3/m3].9
However, in reservoirs containing oil with low
methane content, color channels may provide
better information on contamination than the
methane channel does. For gas-condensate fluids, methane detection using the LFA module is
essential, because even in the shortest wavelength color channels, OD remains low and the
progression of cleanup using the OCM-color procedure is difficult to assess. In some cases, a
drilling-mud filtrate may be darker than the condensate, and the OCM-color procedure may not
be able to discriminate contamination from reservoir fluid. The OCM-methane detection in the
new LFA module works well in such cases.
Start fit
3:32 AM
Color OD
11/29/01
Color OD
50387schD03R1
10
0
0.0
0
1000
2000
3000
4000
5000
6000
7000
8000
Pumping time, sec
> Obtaining acceptable samples. After about two hours of pumping, contamination had dropped to about 9% (blue curve). OD for all channels is shown in
the inset (shaded green). The OCM model was fit to data of Channel 4 minus
Channel 6 (red curve) between the start- and end-fit lines (green dashed lines).
The increases in OD past the end-fit line occurred during sample collection.
Autumn 2001
33
50387schD03R1.p34.ps 11/17/01 8:31 PM Page 34
Concentration, mole %
100.0
Oil 1
Oil 2
10.0
Trend for Oil 1
3.1% contamination
1.0
0.1
C2
C4
C6
C8
C10
C12
C14
C16
C18
C20
C22
C24
C26
C28
C30+
Component
> Removing contamination. GC results indicated Oil 1 (blue) and Oil 2 (red) from neighboring wells had
similar profiles except for the contamination of C16 and C18 from synthetic drilling mud. Contamination
can be removed by developing the trend line for Oil 1 and decreasing the concentrations of C16 and C18
to the trend level. This analysis confirms that the oils came from the same source rock.
contamination in real time before collecting samples. At the rig floor or in a laboratory, sample
contamination can be analyzed further using a
gas chromatograph (GC), a gel-permeation chromatograph (GPC), tracer analysis or, less commonly and not discussed here, a nuclear
magnetic resonance (NMR) spectrometer.
In a GC, a small quantity of sample fluid is
injected into a carrier gas such as high-purity
helium. Light gaseous components are separated
using a molecular sieve and heavier components
are separated using a packed chromatographic
column. A molecular sieve relies on particle size
for separation, with smaller molecules staying in
the sieve longer. In a packed column, the gas
flows past particles coated with a fluid, termed
the stationary layer in a GC because the gas
does not mobilize it. The relative solubility of
components in the stationary layer separates
them as the carrier gas moves a sample through
the column. Chromatographs are calibrated for
sample components.
The process is similar for a GPC except the
inert carrier is a liquid, and constituents do not
separate as well at the detector. Component
peaks from a GC are typically distinct, but those
from a GPC can be smeared together. The
Oilphase FFA Field Fingerprint Analyser rigsite
device incorporates a GPC.
At the end of the column, the carrier gas or
liquid containing the sample enters a detector.
For hydrocarbons, this is usually a thermalconductivity detector or a flame-ionization
detector. Some detection methods respond to
mass and others to the number of carbon atoms in
the molecule.
The distribution of crude-oil constituents normally declines smoothly with increasing carbon
34
number beyond eight.10 OBM and SBM filtrate
contamination causes this distribution to deviate
from the expected shape. SBMs use a narrow
range of molecular weights, so contamination
can be discerned with both a GC and a GPC as a
sharp increase in the frequency of molecules
between carbon numbers of C14 to C18 (above).
OBMs with a mineral-oil base include a broader
range of compounds, perhaps ranging from C8 to
C20, and are difficult to distinguish using a GPC.
Often, these muds can be separated from the
crude-oil signature when using a GC. Drilling
muds that include produced reservoir oil cannot
be distinguished from formation oil using either
form of chromatography, unless a tracer is added
to the mud.
An OBM or SBM filtrate response also can be
removed from the GC result by separately measuring the response of the filtrate, normalizing the
two signals and subtracting.11 Drilling-mud composition must be maintained while drilling an openhole section before sampling because variations in
mud composition add error to the analysis.
Sometimes, contamination is measured using
tracers, by tagging drilling mud with an isotope
or a molecule that is not present in high concentration in reservoir oils. For isotopic tagging of
hydrocarbons, 13C replaces 12C, or deuterium
replaces hydrogen. Mass spectroscopy measures
the concentration of an isotope in a reservoirfluid sample to determine contamination.
Detected isotope concentrations must be higher
than those found naturally for this procedure to
work. Chemical tagging may use linear alpha
olefins, detected using a GC.
Tagging is an expensive procedure that must
be planned in advance. The isotope or chemical
tag must be in the mud in a constant concentration before drilling into the zone of interest and
must remain in the mud until samples are taken,
since all drilling mud that filters into the formation must be tagged to have a meaningful result.
Chemical tagging has an added problem: the
selected molecules may not behave like reservoir
crude. For example, linear alpha olefins are less
stable at high temperature than the corresponding alkanes, and may not travel through porous
media at the same rate.
Results of several contamination-measurement techniques have been performed at Hebron
field offshore Newfoundland, Canada, and in Gulf
of Mexico wells.12,13 At Hebron field, the synthetic
drilling mud was tagged with deuterium. Fluid
samples from five different zones were collected
using the OFA module. The OCM-color procedure
evaluated contamination while fluid was pumped
from the formation. The Oilphase FFA device
determined contamination using a GPC at the
rigsite. Isotope tag concentration was determined using mass spectroscopy, and a laboratory
GC determined the constituents of the fluid.
The LFA module including the OCM analysis
was compared with laboratory GC analysis on
live oils from several Gulf of Mexico wells. In
both this study and the Hebron field study, the
real-time LFA or OFA measurements generally
agree with the isotope, GC and FFA results
(next page, top).
Some discrepancy between methods is
expected, as all methods have potential errors.
The FFA device can overestimate contamination if
the mud is not synthetic; even with SBM, both the
FFA results and GC methods assume a distribution of hydrocarbon constituents to determine
contamination. Tagging is expensive and in principle can be accurate, but in practice, it may not
obtain reliable results. It is difficult to ensure that
all the drilling mud has a uniform concentration of
the chemical or isotopic tag and that the tagged
molecules have the same physical and transport
properties as the rest of the filtrate. The OCMcolor method has problems when the mud filtrate
has significant color or the reservoir oil is colorless, because the method requires a contrast
between the two. However, the LFA-OCMmethane method provides a solution for such
cases, since it is based on methane concentration.
Even if contamination-detection methods
always were correct, many errors can occur in
collecting samples. The fluid can go through
a phase transition as it is drawn into the tool,
leaving components behind in the formation, or
phases can separate in the tool. Valves can fail,
either not opening properly downhole and capturing insufficient fluid or not closing completely
and losing pressure and fluid after sample collection. At surface, every time the fluid is transferred
Oilfield Review
50387schD03R1.p35.ps 01/10/2002 03:55 PM Page 35
1a
160
1b
157
1c
2a
155
3a
Sample number
Sample number
2b
3b
3c
025
13
Gas chromatograph
OFA measurement
FFA analysis
Isotopic tag determination
4b
5a
5b
Gas chromatograph
LFA-color measurement
LFA-methane measurement
12
11
5c
0
20
40
60
Contamination, %
80
100
0
20
40
60
Contamination, %
80
100
> Comparing different methods of evaluating contamination. Contamination measurements of fluid
samples from Hebron field (left) and Gulf of Mexico wells (right) indicate agreement among different
methods for most samples.
2700
ur
ess
-pr
tion nt
ura gradie
Gas-oil contact
Sample gas-loss correction
2750
Contaminated
sample
2800
270
275
nt
radie
ure g
s
Pres
True vertical depth, m
2725
e
Autumn 2001
045
4a
2775
10. Gozalpour F, Danesh A, Tehrani DH, 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, Texas,
USA, October 3-6, 1999.
11. MacMillan DJ, Ginley GM and Dembicki H: “How to
Obtain Reservoir Fluid Properties from an Oil Sample
Contaminated with Synthetic Drilling Mud,” paper SPE
38852, presented at the SPE Annual Technical
Conference and Exhibition, San Antonio, Texas, USA,
October 5-8, 1997.
Gozalpour et al, reference 10.
12. Connon D: “Chevron et al. Hebron M-04 Contamination
Prediction Method Comparison,” Released Project
Report available at Canada-Newfoundland Offshore
Petroleum Board, St. John’s, Newfoundland, Canada,
May 1, 2001.
13. Mullins et al, reference 9.
14. Fadnes FH, Irvine-Fortescue J, Williams S, Mullins OC
and Van Dusen A: “Optimization of Wireline Sample
Quality by Real-Time Analysis of Oil-Based Mud
Contamination—Examples from North Sea Operations,”
paper SPE 71736, presented at the SPE Annual Technical
Conference and Exhibition, New Orleans, Louisiana,
USA, September 30-October 3, 2001.
150
Sat
or a sample bottle is handled, there is potential
for damaging the sample. Bottles should be
heated and agitated for about five days before
performing laboratory analyses, but not all laboratories follow this recommended procedure.
Collecting the right base oil of the drilling mud—
used to compare with spectra of contaminated
reservoir oil—is difficult because mud composition often changes during a job as components
are added to control various drilling problems.
Collection and analysis of fluid samples are
important; operators must control sources of
error to obtain the best possible data. The OFA
and LFA procedures measure properties downhole in real time before sample collection, a
distinct advantage. The few sample bottles available on the tool are not wasted storing bad
samples. Since OCM measurements are made
before any possible transport and handling
problems, they provide a check for the quality of
later measurements.
When sufficient information is available from
the reservoir, measured values of fluid properties
can be used as an additional check on sample
quality. Norsk Hydro conducted a detailed study
of oil samples taken from several North Sea
fields.14 In a reservoir with a gas cap, both chemical tags and the FFA device indicated a high
level of sample contamination, ranging from
8.9% to 25.8%. The OFA-OCM method and GC
analysis indicated lower contamination levels of
2.6% to 6.8%. The difference in these two
ranges of contamination measurement led Norsk
Hydro to investigate further.
Reservoir saturation pressure, Psat, at the
sampling depth was estimated from reservoir
pressure and density gradients starting at
the gas-oil contact (right). The reservoir saturation pressure of the sample, based on PVT
Removal
of 3%
contamination
280
285
290
295
Removal of 9%
contamination
300
305
Pressure, bar absolute
> Using reservoir properties to evaluate contamination measurements. The known gas-oil contact,
pressure gradient (blue line) and saturation-pressure, or bubblepoint, gradient (green line) intersect at
the gas-oil contact for a North Sea well. The contaminated sample had a bubblepoint pressure of
about 272 bar [27.2 MPa or 3950 psi] (dark brown). PVT modeling allowed prediction of bubblepoint
pressure of uncontaminated oil by mathematically removing measured contamination from the sample.
Removing 9% contamination, measured using isotopic tagging and the FFA procedure, produced an
unphysical result above the reservoir values (purple). Removing only 3% contamination (dark blue),
based on the OFA-OCM result, did not raise the bubblepoint enough. Assuming the sample bottle dead
volume of 2.5% was all lost gas provides another factor to adjust the PVT properties of the contaminated sample (light brown). Combining the 3% contamination correction with the gas-loss correction
brings the prediction of bubblepoint (light blue) close to the saturation-pressure gradient.
35
50387schD03R1.p36.ps 11/17/01 8:31 PM Page 36
properties determined with contaminants in the
fluid, was about 20 bar [2 MPa or 290 psi] below
the saturation-pressure gradient at the sampling
depth. These PVT properties can be mathematically corrected to remove the effect of contaminants and then compared with the reservoirgradient calculation.
When using the FFA contamination value of
9%, the resulting calculated Psat was greater
than the reservoir pressure, which is an unphysical result. When using the OFA-OCM value of
contamination, Psat was about 10 bar [1 MPa or
145 psi] below the expected value. This indicates
the sample may have lost gas before PVT properties were evaluated. Gas could separate from liquid in the formation due to near-well pressure
drawdown, but the downhole conditions were
not known well enough to evaluate this effect.
The investigation focused on what happened to
the sample coming out of the hole.
The sample bottle did not allow for downhole
pressure compensation. The fluid could enter the
two-phase region due to cooling from the reservoir temperature of 107ºC [225ºF] during transport to surface. The sample had probably cooled
below 102ºC [217ºF]—the temperature at which
pressure in the enclosed chamber decreased
below the bubblepoint—and was in two phases
by the time it reached the surface. The 450-cm3
bottle has a dead volume of 2.5% between the
isolating valve on the bottle and the valve on the
downhole flowline, which could have been filled
with gas that was lost when the valves were
opened at the surface. The PVT properties of
the contaminated samples can be corrected for
this gas loss, increasing the contaminatedsample bubblepoint pressure by 10 bar. When
the gas-loss correction is combined with removal
of the contamination, as measured by the OCMcolor method, Psat increased to within 4 bar
Color OD
End fit
Start fit
0.4
0.3
0.2
0.1
100
75
0.50
50
0.25
25
0.00
0
1000 2000 3000
Contamination, %
Methane OD
0.75
0
4000 5000 6000 7000 8000 9000 10,000
Pumping time, sec
> Cleanup curve for gas condensate. This North Sea gas condensate was
transparent. Even the OD of the shortest wavelength channel (top) showed
insufficient contrast to reliably determine OD buildup using the OCM-color
method (red). Cleanup was more reliable in the methane channel (bottom)
(pink), giving a more quantifiable OCM-methane fit to the OD data (black).
Calculated values of contamination are shown in the lower plot, with an
OCM-color curve (green), an OCM-methane curve (purple) and the average
of the two (blue). In this case, the large discrepancy is caused by the light
color of the condensate.
36
[0.4 MPa or 58 psi] of the expected reservoir
value, which is reasonable agreement. This analysis could not have been performed without the
downhole OCM-contamination measurement.
Monitoring Gas Directly
Gas-condensate fields engender additional difficulties for fluid sampling when OBM and SBM
are used. Although they contain single-phase fluids in the reservoir, gas condensates separate
into a gas phase and a liquid phase when conditions drop below the dewpoint. The liquid
derived from gas condensates is a more valuable
commodity than the gas. Surface-separator conditions are tuned to optimize the volume and
value of liquid obtained from condensates. The
separator designs often are based on fluid properties from wireline samples, so determining the
level of contamination and correcting the PVT
properties are essential.
OBM and SBM filtrate may mix only partially
with condensate in a reservoir, leaving mud filtrate in a liquid-hydrocarbon phase and a gas
phase with some of the more volatile components of the filtrate. A wireline sampling probe
draws both hydrocarbon phases into the device,
and samples collected contain both reservoir
fluid and filtrate contamination. When the fluid
pressure is lowered during laboratory testing, the
phases separate. All mud filtrate is concentrated
in the liquid phase; presence of contamination
strongly affects a sample’s dewpoint pressure.
To calculate a correct GOR and other reservoir-fluid properties, the volume of the oil phase
must be adjusted to remove contamination. That
liquid-phase contamination must be kept low to
avoid excessive correction factors, just as with a
black oil. However, to compensate for the concentration of SBM and OBM contaminants in the
liquid phase, many companies set the acceptable
level of contamination in a gas condensate
below that for a black oil. The LFA tool provides
significant new information for gas-condensate
reservoirs, improving data quality used for
designing production facilities.15
A gas-condensate prospect in the Norwegian
North Sea offered one of the first tests for the LFA
tool, used in this case without the OCM module.16
A mobile C36+ GC, capable of measuring individual
constituents up to C36 at the rigsite, indicated
contamination of 32% to 60% in the low-pressure
liquid phase. This was comparable to results from
subsequent FFA analysis onshore. The LFA timesequence data were later analyzed using the
Oilfield Review
50387schD03R1.p37.ps 11/17/01 8:32 PM Page 37
Gas Oil Fluid Color
Flag Flag
OD Channels
Methane Channel
1512
0.5
1494
1476
1458
1440
Ratio of methane peak to oil peak
0.4
1422
1404
1386
1368
0.3
1350
1332
1314
0.2
1296
1278
1260
1242
0.1
Live oils
Binary mix*0.85
Dead oil from GOR run
1224
1206
1188
1170
0
0
1000
2000
4000
3000
5000
1152
6000
1134
GOR, scf/bbl
1116
1098
1080
1062
Time, sec
> GOR measurement derived from molecular vibration peaks. In laboratory
tests, the ratio of absorption at the methane peak to the oil peak fits well with
GOR for both methane-heptane mixtures (red squares) and live oils (blue
circles). The multiplicative factor applied to the methane-heptane mixtures
accounts for the absence of other gases normally present in live oils. The dead
crude oil (orange triangle) was evaluated after gas was removed in the laboratory.
1044
1026
1008
990
972
954
936
918
900
OCM method. Mud filtrate and the reservoir fluid
were indistinguishable in the color channels.
The OCM-methane analysis provided a quantitative contamination measurement, about 8% of
the live oil (previous page). The operator had little
experience with the new tool and sought to understand the difference.
A subsequent well test proved a gas-condensate find. Surface-separator samples collected
during the flowing well test and analyzed using a
C36+ GC indicated stock-tank oil contamination of
23%. A full PVT analysis provided the GOR,
allowing correction of contamination to singlephase, downhole conditions. The result indicated
6 to 7% contamination, which agreed well with
the OCM-methane measurement on the live fluid.
During determination of fluid properties for a
gas-condensate reservoir drilled with OBM, the
buildup of methane measured with the LFA module
is essential to obtain accurate, real-time condensate-contamination measurement. The alternatives
are to conduct a DST or complete a well with
water-base mud to avoid oil contamination
altogether. Moreover, using the LFA device also
provides a simultaneous measurement of GOR.
The gas refractometer on both the OFA
and LFA tools indicates gas only when it is in contact with the detector window. Gas bubbles may
not be detected if they are in the center of the
Autumn 2001
flowstream, or on the opposite side. The refractometer detects all gases, regardless of composition, so CO2 and H2S are flagged.
The LFA module also provides a complementary gas-detection system using measurement of
OD in the methane channel. Although insensitive
to other gases, this detector monitors all
methane passing through the flowline. If live oil
is flowing, the volume percentage of methane
will be low. However, if the pressure drops below
the bubblepoint, gas evolves and methane
absorption will be high when a bubble passes the
light beam anywhere within the flowline. The
combination of the gas refractometer and
methane detector makes a robust LFA gas-detection method (right).
The ratio of the methane peak to the oil peak
in the LFA module correlates with GOR both for
mixtures of pure components and for live crude
oils (above). A multiplying factor applied to the
methane-heptane mixtures compensates for
other hydrocarbon components in the gas phase
of reservoir oils. The tool does not measure CO2 or
H2S, so the LFA-GOR measurement may be incorrect for fluids from reservoirs containing significant quantities of these nonhydrocarbon gases.
15. Mullins et al, reference 9.
16. Fadnes et al, reference 14.
882
864
846
828
810
792
774
756
738
720
702
684
666
648
630
612
594
576
558
540
522
504
> LFA gas-detection combination. After an initial
cleanup period, the color Channels 1 through 5 in
Track 4 show little absorption, confirming a gas
condensate. Channels 6 and 9 also have low OD,
which means no water is present. The oil peak in
Channel 8 is transformed into an oil flag in Track 2
(green), indicating periods when no oil flows,
particularly from 1116 to 1188 seconds and 1422
to 1458 seconds. The gas refractometer in Track 1
(red) measures all gases, but only when they
contact the sapphire window of the refractometer. It misses some periods of gas flow. The LFA
methane response from Channel 0, expanded in
Track 5, is sensitive to all methane in the flowline,
but not to other gases. The combination of the two
gas detectors is more robust than either alone.
37
3:35 AM
Page 38
LWD
CMR Standard-Resolution
Gamma Ray Resistivity Permeability
Porosity
NMR T2
Depleted
condensate
in Yellow
sand
CMR
High-Resolution
MDT
MDT-OFA
Permeability
Pressure
Fluid Typing
Pumpout at X482
40° API gravity
Pumpout at X597
40° API gravity
Pumpout at X616
35° API gravity
Blue sand
Real-Time Fluid Typing
The combination of the MDT system and the
CMR Combinable Magnetic Resonance tool
revealed new insights about a reservoir operated
by Shell in the Gulf of Mexico. The Yellow sand
unit had been depleted for two years. The new
drilling target was an underlying sandstone formation, called the Blue sand, separated from the
overlying reservoir by a thick shale.
A logging-while-drilling (LWD) resistivity log
revealed a 10-ft [3-m] water layer on top of the
Blue sand oil, which is not a gravitationally stable
situation. A thin hydrocarbon layer sat atop the
water, just below the thick shale (below). The
operator wanted to know whether water from
above had broken through. The LWD gamma ray
log and standard CMR processing did not explain
how this water could be above the oil (right).
Pressures collected with an MDT tool indicated
that the water zone was not in pressure communication with either the Yellow sand above or the
Blue sand below. Reservoir pressure in the water
zone was about 800 psi [5.5 MPa] higher than the
Blue sand, and was slightly less than original
reservoir pressure for the Yellow sand.
Yellow sand
11/29/01
Shale
50387schD03R1
Pumpout at X640
35° API gravity
> Water above oil investigated by MDT-OFA fluid typing. There is a water zone over the oil-saturated
Blue sand, located at the blue arrow pointing to the low resistivity in Track 2. The responses in the
gamma ray and standard-resolution CMR logs do not explain how this water zone can sit atop oil. A
reprocessed high-resolution CMR permeability log (Track 6) shows a thin permeability barrier, indicated by the green arrow. MDT logging shows three pressure compartments: the depleted Yellow
sand above the shale, the Blue sand below the barrier and the region between the shale and the thin
barrier at high pressure. The MDT color channels, evaluated at the depths indicated by the black
arrows, were used to type the reservoir fluids. The oil atop the water above the barrier has the same
characteristics as the oil in the Yellow sand. This caused the operator to reevaluate the boundary
between the Yellow and Blue sands in this well as being at the thin barrier rather than the thick shale.
Shale
Splinter of
Yellow sand
Target oil in
Blue sand
> Section of Yellow sand below shale. The Yellow
sand above the shale is saturated with a condensate. The oil-saturated Blue sand did not extend
to the shale, but stopped at a thin barrier (thick
black line). The splinter of the Yellow sand unit
below the thick shale had a water leg (blue) below
a thin layer of condensate.
38
The depleted Yellow sand placed a limit on
the mud weight that could be used in the borehole. This created concerns about the wellbore;
the well was not stable enough to leave the MDT
tool in place long enough for formation fluid to
clean up. The MDT tool was used instead for fluid
typing with the gargling technique developed by
Shell Deepwater Services.17 In this technique,
reservoir fluids from the formation were pumped
for a short period of time through the OFA module and out to the wellbore, without collecting
samples in bottles. An OD spectrum from the OFA
module allowed analysis of these small quantities of reservoir oil. Since oil color relates to API
gravity and GOR, the color pattern from the 10
OFA channels enabled discrimination between
the oils. In this case, the Yellow sand was a gas
condensate with an API gravity of about 40º and
a GOR of 6000 scf/bbl [1080 m3/m3], while the
Blue sand held a 35º API gravity oil with a GOR of
2000 scf/bbl [360 m3/m3]. Surprisingly, the color
spectrum of the hydrocarbon sitting on top of the
water had the same signature as the Yellow sand
above the thick shale.
The CMR log data were reprocessed to
improve resolution from 18 in. [46 cm] to about
8 in. [20 cm], revealing a thin permeability barrier
at the base of water, thought to be about 6-in.
[15-cm] thick. This led to a rethinking of the distinction between the top and bottom units. In
other wells, the Yellow sand remained above the
large shale, but in this well, a splinter member of
the Yellow sand cut below the shale. The true
boundary between the zones was the thin barrier,
which appeared to be sand on sand, undifferentiated on conventional logs.
Oilfield Review
50387schD03R1
11/29/01
3:36 AM
Page 39
100 km
62 miles
CA
NA
DA
NEWFOUNDLAND
St John’s
Hibernia field
Rift-basin
outlines
Atlantic Ocean
Water depth
200 m
2000 m
> Hibernia field, offshore Newfoundland, Canada.
Had this been an exploration well, facilities
planning would have relied on results from fluid
sampling. Depending on where the samples were
collected, the GOR could have been too high or
too low, leading to an inefficient design. If the
sample GOR measured were lower than the
actual production, the facilities would have an
undercapacity for gas production, and insufficient
compression and transmission capabilities,
resulting in lost or delayed revenues. Significant
error in GOR in the opposite direction could have
the opposite problem—an expensive overdesign
with too much capacity. MDT fluid typing is a
valuable means for detecting such situations.
In a well in the North Sea, Norsk Hydro drilled
a pilot hole through three horizons prior to drilling
a horizontal section.18 The typical log response in
this field made distinguishing the fluid type in
each formation difficult. Precise definition of fluid
compositions was not required, but rapid differentiation of gas, oil and water was imperative
because the rig was idle while the operator
awaited this fluid identification. The operator
wanted to drill a horizontal wellbore into the
deepest oil-bearing zone. The MDT sonde was
chosen to identify the fluids in real time.
Autumn 2001
Pumping fluid into the tool progressed until
the OFA-OCM method indicated contamination
had dropped below 8% in the middle zone and to
1% in the upper zone. The MDT tool indicated
that the lower zone was water-filled. The low
contamination values in the other zones gave the
operator confidence in the tool response, which
showed that the reservoir fluid was oil. A 3%olefin tracer placed in the OBM mud before
drilling the section allowed rapid confirmation of
these contamination values using a GC at the rig.
The surface contamination measurements—
5% in the middle zone and 4% in the upper—
provided reasonable agreement with the OFAOCM measurement.
Although additional fluid samples had been
collected for testing onshore, the real-time
results using the OFA-OCM analysis coupled with
a rigsite GC confirmation provided answers that
were conclusive enough to cancel the onshore
testing program. The horizontal section was
drilled into the middle horizon immediately after
completing the MDT run, resulting in a successful well.
Norsk Hydro no longer uses olefin tracers to
tag drilling mud. Recent wells have relied successfully on the combination of the OCM method
and a C36+ GC.
Fluid Compartments in Hibernia Field
The Hibernia field, discovered in 1979 and operated by Hibernia Management and Development
Company, Ltd. (HMDC), was the first significant
oil discovery in the Jeanne d’Arc basin on the
Grand Banks of Newfoundland, Canada. Oil production commenced on November 17, 1997, from
an ice-resistant, gravity-based platform in 80 m
[262 ft] of water, 315 km [196 miles] east-southeast of St. John’s, Newfoundland (above).
The structure is a highly faulted, south-plunging anticline containing approximately 3 billion
barrels [475 million m3] of oil-in-place, with an
estimated 750 million recoverable barrels
[120 million m3]. Most of these resources are in
two Lower Cretaceous reservoirs, the Hibernia,
and the combined Ben Nevis and Avalon sandstones. The Hibernia reservoir will be depleted
17. Hashem et al, reference 7.
18. Fadnes et al, reference 14.
39
50387schD03R1
11/29/01
3:37 AM
Page 40
Bonavista
platform
0
0
1
2
1
3
4
5 km
2
3 miles
N
Na
lus
fau
lt
Mur
re fa
ult
uti
> Hibernia water- and gasfloods. The 3D image indicates some of the oil-production (green), water-injection (blue) and
gas-injection (red) wells in the highly faulted reservoir (left). The structure map shows distinct fault blocks in the Hibernia formation (right). Part of the field is under waterflood (blue) and part under gasflood (red). The section line (black)
indicates the location of the cross section shown on page 42.
3550
B-16 5 MDT 2
B-16 5 MDT 3
3600
B-16 6 MDT 3
3650
B-16 2 BHS
B-16 6 MDT 1
3700
B-16 3 BHS B-16 3 MDT 3
B-16 3 BHS
B-16 1 BHS
B-16 3 BHS
B-16 1 BHS
B-16 1 BHS
3750
Depth, m
B-16 3 MDT 4
B-16 9 MDT 6
C-96 DST 4 BHS
3800
C-96 DST 3 BHS
B-16 9 MDT 3
3850
B-16 7 MDT 3
C-96 DST 1 BHS
3900
B-16 11 MDT 6
B-16 7 MDT 2
3950
4000
125
175
225
275
325
using both waterflood and gasflood processes
(above). Delineation drilling of the Ben Nevis and
Avalon formations continues; these reservoirs
will be produced under waterflood.
HMDC encountered operational problems
while drilling the first four wells using WBM.
Shifting to OBM resulted in improved borehole
conditions, few seal losses while running the
logs and decreased logging-acquisition time.
Extensive faulting makes reservoir continuity
uncertain. Early in field development, HMDC initiated a comprehensive data-acquisition plan to
determine fluid compositional variation between
fault blocks and within the vertically extensive
fluid column. Obtaining high-quality samples
with the MDT tool is an integral part of the
program for determining reservoir-fluid properties. MDT pressure measurements establish
pressure gradients and locate gas-oil and wateroil contacts.
Fluid samples were collected in three ways—
MDT samples, bottomhole samples and separator samples. The MDT string typically was
configured to obtain approximately 30 pressure
points across selected reservoir intervals and
included six MPSR sample bottles. Several wells
were sampled using 12 sample cylinders: six
375
GOR, m3/m3
> Hibernia GOR. Fluid samples from the MDT tool and from bottomhole samples
(BHS) from DSTs indicate the trend of GOR with depth. Separator samples from
Hibernia are not associated with a specific depth and are not shown here.
(225 m3/m3 = 1249 scf/bbl.)
40
Oilfield Review
50387schD03R1.p41.ps 11/17/01 8:32 PM Page 41
Gasflood Wells
Waterflood Wells
17-6
14-3
20-6
17-5
14-2
OFA-OCM measurement
Gas chromatograph
12-6
OFA-OCM measurement
Gas chromatograph
17-4
17-3
12-5
OFA-OCM measurement
Gas chromatograph
20-5
17-2
17-1
12-3
16-6
12-2
16-5
12-1
16-4
11-6
11-5
8-5
20-4
20-3
Sample number
12-4
Sample number
Sample number
Ben Nevis and Avalon Wells
16-2
16-1
9-5
9-3
8-3
9-1
8-1
7-3
5-6
7-2
20-2
20-1
19-1.11
19-1.10
6-6
5-5
6-5
5-4
19-1.09
6-4
5-3
6-3
5-2
6-1
0
20
40
60
80
Oil-base mud contamination, %
100
19-1.08
0
20
40
60
80
Oil-base mud contamination, %
100
0
20
40
60
80
Oil-base mud contamination, %
100
> Comparison of contamination measurements. The OFA-OCM measurement at the wellsite agrees well with laboratory GC measurements for the gasflood
(left) and waterflood (middle) zones of the Hibernia formation and Ben Nevis and Avalon formations (right).
MPSR cylinders and six pressure-compensated
SPMC cylinders. The variation of PVT properties
in the MDT samples helped define depth and
areal trends, which were further refined by geochemical fingerprinting of the samples. MDT
detection of OBM contamination was important
for the program. Use of OCM real-time monitoring allowed collection of high-quality gascondensate samples.
Initially, bottomhole samples from the entire
perforated interval were collected during production testing to obtain representative PVT properties. Single-phase flow conditions were
maintained downhole during sampling. Fluid
samples collected from test separators were less
expensive, allowing continued monthly sampling
to monitor compositional changes. Samples from
the three sources have shown excellent agreement in PVT studies and determination of OBMcontamination levels (previous page, bottom).
Autumn 2001
The operator uses PVT data from these
sources for well-test analysis, reserves determination, material balances, reservoir simulation,
production allocation, production monitoring and
fluid-metering factors, process simulation and
regulatory reporting.
The initial pressure in the Hibernia reservoir
was approximately 40 MPa [5800 psi]. Because
the bubblepoint varies across the field, the
company avoided sampling below bubblepoint
pressure. The MDT tool monitored pressure
during sampling, allowing minimal drawdown
and accurate bubblepoint determination from
recovered samples.
The OFA module detected sample contamination levels to estimate pumping time to achieve
cleanup. About halfway through the sample-collection program, the OCM option became available, providing a quantitative measure of
contamination in real time. The OFA results from
the previous logging runs were analyzed later
using the OCM-color methodology to determine
contamination levels (above).
The MDT sampling tool is an effective means
of collecting representative fluid samples to evaluate variations through long fluid columns. The
Hibernia group has successfully run the tool on
wireline, but because of wellbore deviations up
to 80°, the tool typically was run as part of a TLC
Tough Logging Conditions superstring. The TLC
tool usually includes the Platform Express integrated tool, including the AIT Array Induction
Imager Tool sonde, a caliper and gamma ray tool,
and the MDT modules. Logs collected on a first
pass were transmitted in real time to the company office in St. John’s where engineers picked
points for MDT pressure determination and a
sample-collection pass. With fluid columns in
excess of 400 m [1300 ft] thick in areas of the
41
50387schD03R1
11/29/01
3400
3:37 AM
Page 42
NW
C block
B block
SE
GOC
3600
B-08
Depth subsea, m
B-16
15z
B-16 10z
3800
WOC
0
1
2 km
4000
0
B-16
11
0.5
1
1.5 miles
B-16
14
5:1 vertical exaggeration
4200
400
n-C20
n-C30
n-C14
n-C15
Base oil
300
n-C20
Hibernia oil sample
n-C25
Detector response
500
n-C12
n-C10
600
n-C22
n-C18 Ph
n-C16
700
n-C17 Pr
> Cross section spanning Blocks B and C in the Hibernia field gasflood area. The Hibernia formation dips steeply,
plunging into the Murre fault in the northwest. The gas-oil contact (GOC) is shown at the crest. The water-oil contact
(WOC) is unknown in the southeast; in the northwest it lies between the two marked depths. This section line is indicated on the map on page 40.
n-C30
200
Pr Ph
100
0
0
10
20
30
40
50
60
70
80
Time, min
> Gas chromatograms of reservoir and drilling-mud base oils. The sharp peaks on the curves are specific carbon compounds, such as normal-alkane C30 [n-C30]. Pristane (Pr) and phytane (Ph) are geomarkers found in reservoir fluids. A scaling factor is applied to the base-oil spectrum before subtracting it from the reservoir-oil spectrum. The scaling factor is related to the degree of contamination.
42
Oilfield Review
50387schD03R1.p43.ps 11/17/01 8:33 PM Page 43
Bonavista
platform
0
0
1
2
3
1
4
5 km
2
3 miles
N
Na
uti
lus
fau
Mur
re fa
ult
lt
> Fluid regions in Hibernia field. Seven distinct regions are defined for Hibernia fluids, based on
constituents and physical properties determined from DST and MDT fluid samples.
field, using MDT pressures and fluid-type determination to establish gas-oil and water-oil
contacts was important (previous page, top).
A significant benefit of the MDT logging
program is real-time decision-making on sample
collection points.
MDT fluid-sample composition was determined in a PVT laboratory by the GC method. The
chromatogram of the base oil of the mud was
subtracted from the sample GC spectrum
(previous page, bottom). The resulting peakheight spectra from different blocks, coupled
with other PVT data such as the bubblepoint
pressure, GOR and formation volume factor, provided evidence to correlate oil from different
fault blocks, indicating seven distinct fluid
regions across the field (above). With this information, gasflooding and waterflooding can be
implemented more efficiently. Formation pressures from openhole MDT runs also indicated
whether offset production had drawn down formation pressure in the new locations. Other measurements made on the reservoir fluids, including
wax content, sulfur content, acid number, pour
point, cloud point and saturates-aromaticsresins-asphaltenes content, also indicated variations by fault block, impacting the production and
completion strategies.19
Autumn 2001
A Downhole Chemistry Laboratory
Distinguishing fluid phases may seem like some
of the simplest chemistry that can be performed.
Doing it from miles away, in a harsh environment,
is the significant new accomplishment of the
MDT tool. The channels of absorption information in the OFA tool have allowed correlation
with many more attributes of the fluid: oil-shrinkage factor, bubblepoint pressure, oil compressibility, oil density and average molecular
weight.20 Minimizing contamination in collected
samples and controlling phase separation during
collection to enhance the value of in-situ fluid
properties measurements is an ongoing challenge. The additional capabilities in the new LFA
module provide direct measurement of methane
content, allowing estimation of GOR and a more
robust gas flag to avoid taking the fluid into the
two-phase region.
In addition, obtaining fluid samples from
behind casing is significantly easier now. The
CHDT Cased Hole Dynamics Tester tool can drill
up to six holes through casing in one trip and, in
combination with other MDT modules, obtain
samples and monitor contamination in real time.
It then seals the hole through the casing with a
corrosion-resistant plug rated to 10,000-psi
[69-MPa] differential pressure.
Already, significant decisions are made based
on real-time downhole fluid measurements.
Continuing development will improve the range
and reliability of these measurements. —MAA
19. The pour point is the lowest temperature at which an
oil will begin to flow under standard test conditions.
The cloud point is the temperature at which paraffin
molecules first start to crystallize from oil, as
observed visually.
20. Van Dusen A, Williams S, Fadnes FH and IrvineFortescue J: “Determination of Hydrocarbon Properties
by Optical Analysis During Wireline Fluid Sampling,”
paper SPE 63252, presented at the SPE Annual Technical
Conference and Exhibition, Dallas, Texas, USA,
October 1-4, 2000.
43
50387schD01R1
11/29/01
5:23 AM
Page 44
Global Warming and the E&P Industry
The question as to what extent man-made emissions of greenhouse gases may be
causing climate change has stirred intense debate around the world. Continued shifts
in the Earth’s temperatures, predicted by many scientists, could dramatically affect the
way we live and do business. This article examines the evidence and the arguments,
and describes some of the mitigating actions being taken by the exploration and production (E&P) industry.
Melvin Cannell
Centre for Ecology and Hydrology
Edinburgh, Scotland
Jim Filas
Rosharon, Texas, USA
John Harries
Imperial College of Science,
Technology and Medicine
London, England
Geoff Jenkins
Hadley Centre for Climate
Prediction and Research
Berkshire, England
Martin Parry
University of East Anglia
Norwich, England
Paul Rutter
BP
Sunbury on Thames, England
Lars Sonneland
Stavanger, Norway
Jeremy Walker
Houston, Texas
44
Scientists use language cautiously. They tend to
err on the side of understatement. During the
mid-1990s, in the Second Assessment Report of
the Intergovernmental Panel on Climate Change
(IPCC), leading scientists from around the world
expressed a consensus view that “the balance of
evidence suggests a discernible human influence
on global climate.” In July 2001, for the IPCC
Third Assessment Report, experts took this conclusion a step further. Considering new evidence,
and taking into account remaining uncertainties,
the panel stated “most of the observed warming
over the last 50 years is likely to have been due
to the increase in greenhouse-gas concentrations.”1 The word ‘likely’ is defined by the IPCC as
a 66 to 90% probability that the claim is true.
An important and influential segment of the
global scientific community firmly believes that
human activity has contributed to a rise in the
Earth’s average surface temperature and a resulting worldwide climate change. They contend that
such activity may be enhancing the so-called
‘greenhouse effect.’ Other distinguished scientists disagree, some dismissing the IPCC view
as simplistic.
The Greenhouse and Enhanced
Greenhouse Effects
The greenhouse effect is the name given to the
insulating mechanism by which the atmosphere
keeps the Earth’s surface substantially warmer
than it would otherwise be. The effect can be
illustrated by comparing the effects of solar
radiation on the earth and the moon. Both are
roughly equidistant from the sun, which supplies
the radiation that warms them, and both receive
about the same amount of heat energy per
square meter of their surfaces. Yet, the earth is
much warmer—a global average temperature of
15°C [59°F] compared with that of the moon,
-18°C [-0.4°F]. The difference is largely due to the
fact that the moon has almost no atmosphere
while the Earth’s dense atmosphere effectively
traps heat that would otherwise escape into space.
Climatologists use a physical greenhouse
analogy to explain how warming occurs. Energy
from the sun, transmitted as visible light, passes
through the glass of a greenhouse without hindrance, is first absorbed by the floor and contents, and then reemitted as infrared radiation.
For help in preparation of this article, thanks to David
Harrison, Houston, Texas, USA; Dwight Peters, Sugar Land,
Texas; and Thomas Wilson, Caracas, Venezuela. Special
thanks to the Hadley Centre for Climate Prediction and
Research for supplying graphics that were used as a basis
for some of the figures appearing in this article.
1. Climate Change 2001: The Scientific Basis: The
Contribution of Working Group I to the Third Assessment
Report of the Intergovernmental Panel on Climate
Change. New York, New York, USA: Cambridge University
Press (2000): 10.
Oilfield Review
50387schD01R1
11/29/01
5:24 AM
Page 45
50387schD01R1
11/29/01
5:24 AM
Page 46
Because infrared radiation cannot pass through
the glass as readily as sunlight, some of it is
trapped, and the temperature inside the greenhouse rises, providing an artificially warm environment to stimulate plant growth (right).
In the natural greenhouse effect, the Earth’s
atmosphere behaves like panes of glass. Energy
coming from the sun as visible short-wavelength
radiation passes through the atmosphere, just as
it does through greenhouse glass, and is
absorbed by the surface of the earth, which then
reemits it as long-wavelength infrared radiation.
Infrared radiation is absorbed by naturally occurring gases in the atmosphere—water vapor,
carbon dioxide [CO2], methane, nitrous oxide,
ozone and others—and reradiated. While some
energy goes into outer space, most is reradiated
back to earth, heating its surface.2
An enhanced greenhouse effect occurs when
human activities increase the levels of certain
naturally occurring gases. If the atmosphere is
pictured as a translucent blanket that insulates
the earth, adding to the concentration of these
greenhouse gases is analogous to increasing the
thickness of the blanket, improving its insulating
qualities (below).
Visible energy from the sun
passes through the glass,
heating the ground.
Some reemitted infrared
radiation is reflected by the
glass and trapped inside.
> The greenhouse analogy. A greenhouse effectively traps a portion of the
sun’s energy impinging on it, raising the interior temperature and creating an
artificially warm growing environment.
Natural Greenhouse Effect
Enhanced Greenhouse Effect
Enhanced
absorption by
greenhouse gases
Absorption of outgoing
radiation by indigenous
atmospheric gases
Reradiation
into space
Outgoing
long-wavelength
radiation
Incoming
short-wavelength
radiation
Reradiation
to earth
Reradiation
into space
Outgoing
long-wavelength
radiation
Incoming
short-wavelength
radiation
Reradiation
to earth
> Natural and enhanced greenhouse effects. In the natural greenhouse effect (left), indigenous atmospheric gases contribute to heating of the Earth’s
surface by absorbing and reradiating back some of the infrared energy coming from the surface. In the enhanced greenhouse effect (right), increased gas
concentrations, resulting from human activity, improve the atmosphere’s insulating qualities.
46
Oilfield Review
50387schD01R1
11/29/01
5:24 AM
Page 47
Atmospheric
constituent
Lifetime
Source
Carbon dioxide
Combustion of
fossil fuels
and woods
Land-use changes
100 years
Methane
Production and
transport of fossil
fuels
Decomposing waste
Agriculture
Dissociation of
gas hydrates
10 years
Combustion of fossil
fuels
Combustion of waste
150 years
Chlorofluorocarbons
Production
100 years
Ground-level ozone
Transport
Industrial emissions
3 months
Aerosols
Power generation
Transport
2 weeks
Nitrous oxide
Nitrous oxide
10%
Autumn 2001
Carbon dioxide
63%
Others
3%
> Man-made emission sources and lifetimes for
greenhouse gases. Various gases and aerosols
are emitted daily in commercial, industrial and
residential activities. Carbon dioxide is the most
important, because of its abundance and effective lifetime in the atmosphere of about 100 years.
Man-made emissions of greenhouse gases
occur in a number of ways. For example, carbon
dioxide is released to the atmosphere when solid
waste, wood and fossil fuels—oil, natural gas
and coal—are burned. Methane is emitted
by decomposing organic wastes in landfill sites,
during production and transportation of fossil
fuels, by agricultural activity and by dissociation of gas hydrates. Nitrous oxide is vented
during the combustion of solid wastes and fossil
fuels (above left).
Carbon dioxide is the most important, due
principally to the fact that it has an effective lifetime in the atmosphere of about 100 years, and is
the most abundant. Every year, more than 20 billion tons are emitted when fossil fuels are
burned in commercial, residential, transportation
and power-station applications. Another 5.5 billion tons are released during land-use changes,
such as deforestation.3 The concentration of CO2
in the atmosphere has increased by more than
30% since the start of the Industrial Revolution.
Methane
24%
> Relative warming projected from different
greenhouse gases during this century. Of the
various greenhouse gases, carbon dioxide is predicted to have the greatest capacity for causing
additional global warming, followed by methane
and nitrous oxide.
Analysis of air trapped in antarctic ice caps
shows that the level of carbon dioxide in the
atmosphere in pre-industrial days was about 270
parts per million (ppm). Today, readings taken at
the Mauna Loa Observatory in Hawaii, USA,
place the concentration at about 370 ppm.4
Concentrations of methane and nitrous oxide,
which have effective lifetimes of 10 and
150 years, respectively, also have increased—
methane more than doubling and nitrous oxide
rising by about 15% over the same time span.
Both are at much lower levels than CO2—
methane at 1.72 ppm and nitrous oxide at
0.3 ppm—but they exert a significant influence
because of their effectiveness in trapping heat.
Methane is 21 times more effective in this regard
than CO2, while nitrous oxide is 310 times more
effective, molecule for molecule.5
The global-warming potential of a gas is a
measure of its capacity to cause global warming
over the next 100 years. The warming effect of
an additional 1-kg [2.2-lbm] emission of a greenhouse gas discharged today—relative to 1 kg of
CO2—will depend on its effective lifetime, the
amount of extra infrared radiation it will absorb,
and its density. On this basis, experts calculate
that, during this century, CO2 will be responsible
for about two-thirds of predicted future warming,
methane a quarter and nitrous oxide around a
tenth (above right).6
2. The description above is a simplification. In fact, about
25% of solar radiation is reflected back into space before
reaching the Earth’s surface by clouds, molecules and
particles, and another 5% is reflected back by the Earth’s
surface. A further 20% is absorbed before it reaches the
earth by water vapor, dust and clouds. It is the remainder—just over half of the incoming solar radiation—that
is absorbed by the Earth’s surface. The greenhouse analogy, although widely used, is also only partly accurate.
Greenhouses work mainly by preventing the natural process of convection.
3. Jenkins G, Mitchell JFB and Folland CK: “The Greenhouse
Effect and Climate Change: A Review,” The Royal Society
(1999): 9-10.
4. Reference 1: 12.
5. “The Greenhouse Effect and Climate Change: A Briefing
from the Hadley Centre,” Berkshire, England: Hadley
Centre for Climate Prediction and Research (October
1999): 7.
6. Reference 5: 7.
47
50387schD01R1
11/29/01
5:24 AM
Page 48
Observed
behavior
Comparison
and
validation
Climate-system
model
Computer
simulation
Predicted
behavior
Update and refine model
> Climate simulations. Scientists use sophisticated models and computer simulations of the Earth’s climate system to confirm historical, and predict future,
temperature changes. Results are validated by comparison with actual temperature measurements. Such analyses form a basis for updating and refining
the reliability of simulations.
Temperature anomalies, C
1.0
1.0
Model
Observations
0.5
0.5
0.0
0.0
–0.5
–1.0
1850
–0.5
Natural factors only
1950
1900
Temperature anomalies, C
1.0
Model
Observations
Human factors only
–1.0
2000 1850
1900
1950
2000
Model
Observations
0.5
0.0
–0.5
–1.0
1850
Human and natural factors
1900
1950
2000
> Observed and simulated global warming. Neither natural nor man-made effects alone account for
the evolution of the Earth’s climate during the 20th century. By combining the two, however, the
observed pattern is reproduced with reasonable accuracy.
48
Measuring and Modeling Climate Change
IPCC scientists believe that we are already experiencing an enhanced greenhouse effect.
According to their findings, the Earth’s global
average surface temperature increased by about
0.6°C [1.1°F] during the last century. They maintain that this increase is greater than can be
explained by natural climatic variations. The
panel believes there is only a 1 to 10% probability that inherent variability alone accounts for this
extent of warming. Most studies suggest that,
over the past 50 years, the estimated rate and
magnitude of warming due to increasing concentrations of greenhouse gases alone are comparable to, or larger than, the observed warming.7
To better understand the physical, chemical
and biological processes involved, scientists
investigating climate variations construct complex
mathematical models of the Earth’s weather system. These models are then used to simulate past
changes and predict future variations. The more
closely that simulations match historical climate
records built from direct observations, the more
confident scientists become in their predictive
capabilities (left).
Greater emphasis on diagnosing and predicting the impact of global warming has resulted in
increasingly sophisticated simulations. For example, a state-of-the-art, three-dimensional (3D)
ocean-atmosphere model developed at the
Hadley Centre for Climate Prediction and
Research in Berkshire, England, appears to replicate—with reasonable precision—the evolution
of global climate during the late 19th and 20th
centuries. This simulation matches records that
clearly show that the global mean surface air
temperature has increased by 0.6°C ± 0.2°C
[1.1°F ± 0.4°F] since 1860, but that the progression has not been steady. Most of the warming
occurred in two distinct periods—from 1910 to
1945, and since 1976—with little change in the
intervening three decades.
When factors that impact the Earth’s climate
vary—concentrations of greenhouse gases, but
also heat output from the sun, for example—
they exert a ‘forcing’ on climate (see “Increases
in Greenhouse Forcing,” next page). A positive
forcing causes warming, a negative one results
in cooling. When researchers at the Hadley
Centre and the Rutherford Appleton Laboratory,
near Oxford, England, simulated the evolution
of 20th century climate, they concluded that,
by themselves, natural forcings—changes in
volcanic aerosols, solar output and other
phenomena—could not account for warming
Oilfield Review
50387schD01R1
11/29/01
5:25 AM
Page 49
Increases in Greenhouse Forcing
Observed
90˚ N
45˚ N
45˚ S
90˚ S
180˚ W
90˚ W
–1
–0.5
0˚
0
0.5
90˚ E
1
1.5
180˚ E
2
Simulated
90˚ N
45˚ N
45˚ S
90˚ S
180˚ W
90˚ W
–1
–0.5
0˚
0
0.5
90˚ E
1
1.5
180˚ E
2
> Observed (top) and simulated (bottom) surface air temperature changes.
Computer models closely resemble the global temperature signature produced by measurements of the change in air temperature. Values increase
from negative to positive as the color scale moves from blue to red.
in recent decades. They also concluded that
anthropogenic, or man-made, forcings alone
were insufficient to explain the warming from
1910 to 1945, but were necessary to reproduce
the warming since 1976. However, by combining
the two simulations, researchers were able to
reproduce the pattern of temperature change
with reasonable accuracy. Agreement between
observed and simulated temperature variations
supports the contention that 20th century warming resulted from a combination of natural and
external factors (previous page, bottom).8
In addition to examining the global mean temperature, researchers at the Hadley Centre also
Autumn 2001
compared geographic patterns of temperature
change across the surface of the earth. They
used models to simulate climate variations
driven by changes in greenhouse-gas concentrations and compared the ‘fingerprint’ produced
with patterns of change that emerge from observation. Striking similarities are evident between
the fingerprint generated by a simulation of the
last 100 years of temperature changes and the
patterns actually observed over that period (above).
Despite many advances, climate modeling
remains an inexact science. There is concern
that, at present, simulations may not adequately
represent certain feedback mechanisms, especially those involving clouds. Researchers, like
Early this year, scientists at the Imperial
College of Science, Technology and Medicine in
London, England, provided the first experimental observation of a change in the greenhouse
effect. Previous studies had been largely limited
to theoretical simulations.1 Changes in the
Earth’s greenhouse effect can be detected from
variations in the spectrum of outgoing longwavelength radiation, a measure of how the
earth gives off heat into space that also carries
an imprint of the gases responsible for the
greenhouse effect.
From October 1996 until July 1997, an instrument on board the Japanese ADEOS satellite
measured the spectra of long-wavelength radiation leaving the earth. The Imperial College
group compared the ADEOS data with data
obtained 27 years earlier by a similar instrument aboard the National Aeronautics and
Space Administration (NASA) Nimbus 4
meteorological satellite. The comparison of the
two sets of clear-sky infrared spectra provided
direct evidence of a significant increase in the
atmospheric levels of methane, carbon dioxide,
ozone and chlorofluorocarbons since 1970.
Simulations show that these increases are
responsible for the observed spectra.
1. Harries JE, Brindley HE, Sagoo PJ and Bantges RJ:
“Increases in Greenhouse Forcing Inferred from the
Outgoing Longwave Radiation Spectra of the Earth in
1970 and 1997,” Nature 410, no. 6832 (March 15, 2001):
355-357.
those at Hadley, do not claim that close agreement between observed and simulated temperature changes implies a perfect climatic model,
but if today’s sophisticated simulations of
climate-change patterns continue to closely
match observations, scientists will rely to a
greater extent on their predictive capabilities.
7. Reference 1: 10.
8. Stott PA, Tett SFB, Jones GS, Allen MR, Mitchell JFB
and Jenkins GJ: “External Control of 20th Century
Temperature by Natural and Anthropogenic Forcings,”
Science 290, no. 5499 (December 15, 2000): 2133-2137.
49
50387schD01R1
11/29/01
Radiation
into space
5:19 AM
Page 50
Radiation
into space
Soot
Coalesced
state
Aerosol
Radiation
from Earth's
surface
Separate soot
and aerosol
constituents
(external mixing)
Radiation
from Earth's
surface
Coalesced soot
and aerosol
constituents
(internal mixing)
> Impact of aerosols and soot. Temperature
simulations that take into account an internally
mixed, or coalesced, accumulation of aerosols
and soot (right) are more consistent with observations than separate, or externally mixed,
accumulations (left).
Global-average
surface temperature
change
(1900 to 2000)
+ 0.6 C
Results:
10% decrease in snow cover
(since the late 1960s)
2-week shorter annual ice cover
0.1- to 0.2-m sea-level rise
0.5 to 1% increase in precipitation
per decade (Northern Hemisphere)
> Observed impact of global warming. The
0.6°C temperature rise observed during the last
100 years has been postulated as the cause of
decreased snow and ice cover, higher sea levels
and increased precipitation.
50
The Opposing View
Not all scientists accept the IPCC findings.
Many distinguished researchers argue that the
panel’s approach is too simplistic. For instance,
Dr Richard Lindzen, Alfred P. Sloan Professor of
Meteorology at the Massachusetts Institute of
Technology (MIT) in Cambridge, USA, suggests
that clouds over the tropics act as an effective
thermostat and that any future warming because
of increased carbon dioxide concentration in the
atmosphere could be significantly less than current models predict.
Scientists have voiced strong objections that
even sophisticated circulation models do not
adequately describe the complexity of the mechanisms at work. A group of researchers at the
Harvard-Smithsonian Center for Astrophysics in
Cambridge, Massachusetts, for example, claims
there are too many unknowns and uncertainties
in climate modeling to have confidence in the
accuracy of today’s predictions. The group argues
that even if society had complete control
over how much CO2 was introduced into the
atmosphere, other variables within the climate
system are not sufficiently well-defined to produce reliable forecasts. The researchers are not
trying to disprove a significant man-made contribution, but rather contend that scientists do not
know enough about the complexity of climate
systems, and should be careful in ascribing too
much relevance to existing models.9
New scientific studies are shedding more
light on the problem. For example, previous
investigations have concluded that the Earth’s
climate balance is upset not only by emissions of
man-made greenhouse gases during processes
such as the combustion of fossil fuels, but also
by small particles called aerosols, such as those
formed from sulfur dioxide, which cool the Earth’s
surface by bouncing sunlight back into space.
But, new findings suggest that things may not be
that simple. A researcher at Stanford University,
California, USA, states that black carbon, or soot,
emissions from the burning of biomass and fossil
fuels are interfering with the reflectivity of
aerosols, darkening their color so that they
absorb more radiation. This reduces the cooling
effect, and could mean that black carbon is a
major cause of global warming, along with carbon dioxide and other greenhouse gases.
Atmospheric computer simulations usually
assume that aerosols and soot particles are separate, or externally mixed. An internally mixed
state—in which aerosols and soot coalesce—
also exists, but no one has yet successfully determined the relative proportions of the two states.
The Stanford researcher ran a simulation in
which black carbon was substantially coalesced
with aerosols. His results were more consistent
with observations than simulations that assumed
mainly external mixing. Although this could mean
that black carbon is a significant contributor to
warming, there is a bright side to the discovery.
Unlike the extended lifetime of carbon dioxide,
black carbon disappears much more rapidly. If
such emissions were stopped, the atmosphere
would be clear of black carbon in only a matter of
weeks (left).10
9. Soon W, Baliunas S, Idso SB, Kondratyev KY and
Postmentier ES: “Modelling Climatic Effects of
Anthropogenic Carbon Dioxide Emissions: Unknowns
and Uncertainties.” A Center for Astrophysics preprint.
Cambridge, Massachusetts, USA: Harvard-Smithsonian
Center for Astrophysics (January 10, 2001): to appear as
a review paper in Climate Research.
10. Jacobson M: “Strong Radiative Heating due to the
Mixing State of Black Carbon in Atmospheric Aerosol,”
Nature 409, no. 6821 (2001): 695-697.
11. Reference 1: 2-4.
12. Reference 1: 12-13.
13. Climate Change 2001: Impacts, Adaptation and
Vulnerability: Contribution of Working Group II to the
Third Assessment Report of the Intergovernmental
Panel on Climate Change. New York, New York, USA:
Cambridge University Press (2001): 5.
Oilfield Review
50387schD01R1 12/17/01 10:04 PM Page 51
Greater exposure
to disease
Increase in frequency
and intensity
of severe weather
Decreased food
supply
Water shortages
Increased flooding
> Future impact of global warming. IPCC scientists predict a number of consequences if climate changes
track the latest simulations, ranging from water shortages to flooding and decreased food supply.
Predicting the Future Impact of
Global Warming
The IPCC has described the current state of scientific understanding of the global climate system, and has suggested how this system may
evolve in the future. As discussed, the panel confirmed that the global-average surface temperature of the earth increased by about 0.6°C during
the last 100 years. Analyses of proxy data from
the Northern Hemisphere indicate that it is likely
the increase was the largest of any century in the
past millennium. Because of limited data, less is
known about annual averages prior to the year
1000, and for conditions prevailing in most of the
Southern Hemisphere prior to 1861.
The IPCC report states that temperatures
have risen during the past four decades in the
lowest 8 km [5 miles] of the atmosphere; snow
cover has decreased by 10% since the late
1960s; the annual period during which rivers and
lakes are covered by ice is nearly two weeks
Autumn 2001
shorter than at the start of the century; and average sea levels rose by 0.1 to 0.2 m [0.3 to 0.7 ft]
during the 1900s. The report further states that,
during the last century, precipitation increased by
0.5 to 1% per decade over most middle and high
latitudes of Northern Hemisphere continents,
and by 0.2 to 0.3% per decade over tropical land
areas (previous page, bottom).11
While these changes may appear to be modest, predicted changes for this century are much
larger. Simulations of future atmospheric levels of
greenhouse gases and aerosols suggest that the
concentration of CO2 could rise to between 540
and 970 ppm. For all scenarios considered by the
IPCC, both global-average temperature and sea
level will rise by the year 2100—temperature by
1.4°C to 5.8°C [2.5°F to 10.4°F] and sea level by
0.09 to 0.9 m [0.3 to 2.7 ft]. The predicted temperature rise is significantly greater than the 1°C
to 3.5°C [1.8°F to 6.3°F] estimated by the IPCC
five years ago. Precipitation is also forecasted to
increase. Northern Hemisphere snow cover is
expected to decrease further, and both glaciers
and ice caps are expected to continue to retreat.12
If climate changes occur as predicted, serious
consequences could result, both with respect to
natural phenomena, such as hurricane frequency
and severity, and to human-support systems. The
IPCC Working Group II, which assessed impacts,
adaptation and vulnerability, stated that if the
world continues to warm, we could expect water
shortages in heavily populated areas, particularly
in subtropical regions; a widespread increase in
the risk of flooding as a result of heavier rainfall
and rising sea levels; greater threats to health
from insect-borne diseases, such as malaria, and
water-borne diseases, such as cholera; and
decreased food supply as grain yields drop
because of heat stress. Even minimal increases in
temperature could cause problems in tropical
locations where some crops are already near their
maximum temperature tolerance (above).13
51
50387schD01R1
11/29/01
5:25 AM
Page 52
Sea-level rises could threaten five parts of
Africa that have large coastal population centers—the Gulf of Guinea, Senegal, Gambia,
Egypt and the southeastern African coast. Even a
somewhat conservative scenario of a 40-cm
[15.8-in.] sea-level rise by the 2080s would add
75 to 200 million people to the number currently
at risk of being flooded by coastal storm surges,
with associated tens of billions of dollars in property loss per country.14
Africa, Latin America and the developing
countries of Asia may have a two-fold problem,
being both more susceptible to the adverse
effects of climate change and lacking the infrastructure to adjust to the potential social and
economic impacts.
The IPCC Working Group II has ‘high confidence’ that:
• Increases in droughts, floods and other
extreme events in Africa would add to stresses
on water resources, food-supply security,
human health and infrastructures, and constrain further development.
• Sea-level rise and an increase in the intensity
of tropical cyclones in temperate and tropical
Asia would displace tens of millions of people
in low-lying coastal areas, while increased
rainfall intensity would heighten flood risks.
• Floods and droughts would become more
frequent in Latin America, and flooding
would increase sediment loads and degrade
water quality.
The Working Group has ‘medium confidence’
that:
• Reductions in average annual rainfall, runoff
and soil moisture would increase the creation
of deserts in Africa, especially in southern,
northern and western Africa.
• Decreases in agricultural productivity and
aquaculture due to thermal and water stress,
sea-level rise, floods, droughts and tropical
cyclones would diminish the stability of food
supplies in many countries in the arid, tropical
and temperate parts of Asia.
• Exposure to diseases such as malaria,
dengue fever and cholera would increase in
Latin America.15
Not all impacts would be negative, however.
Among projected beneficial effects are higher
crop yields in some mid-latitude regions; an
increase in global timber supply; increased water
availability for people in some regions, like parts
of Southeast Asia, which currently experience
water shortages; and lower winter death rates in
mid- to high-latitude countries.16
52
Retreating glaciers
Thawing of permafrost
Melting of sea ice
Floods
Increased rainfall
Intense cyclones
Decreased food supply
Rising sea levels
Higher heat index
Hotter summers
Reduced water supply
Increase in forest fires
Deteriorating air quality
Floods
Droughts
Degraded water quality
Droughts
Floods
Decreased food supply
Expanding deserts
Sea-level rise
> Impact of global warming by region. All continents will be affected significantly if global warming
continues. The type and severity of specific impacts will vary, as will each continent’s or country’s
capacity to use infrastructure and technology to cope with change.
Other studies—such as the US Global
Research Program’s report “Climate Change
Impacts on the United States,” and the European
Community-funded ACACIA (A Consortium for
the Application of Climate Impact Assessments)
Project report—are consistent with future IPCC
forecasts, and provide a more detailed picture for
particular regions.
According to the US study, assuming there are
no major interventions to reduce continued growth
of world greenhouse-gas emissions, temperatures
in the USA can be expected to rise by about 3°C to
5°C [5.4°F to 9°F] over the next 100 years, compared with the worldwide range of 1.4°C to 5.8°C
[2.5°F to 10.4°F] suggested by the IPCC.17
Assuming there are no major interventions,
other predictions include the following:
• Rising sea levels could put coastal areas at
greater risk of storm surges, particularly in the
southeast USA.
• Large increases in the heat index, the combination of temperature and humidity, and in the
frequency of heat waves could occur, particularly in major metropolitan cities.
• Continued thawing of permafrost and melting
of sea ice in Alaska could further damage
forests, buildings, roads and coastlines.
In Europe, negative climate changes are
expected to impact the south more than the
north. Sectors such as agriculture and forestry
will be affected to a greater extent than sectors
such as manufacturing and retailing, and
marginal and poorer regions will suffer more
adverse effects than wealthy ones.
The ACACIA report, which provided the basis
for the IPCC findings on impacts in Europe, makes
the following predictions for southern Europe:
• Longer, hotter summers will double in frequency by 2020, with a five-fold increase in
southern Spain, increasing the demand for
air conditioning.
• Available water volumes will decrease by 25%,
reducing agricultural potential. Careful planning will be essential to satisfy future urban
water needs.
• Desertification and forest fires will increase.
• Deteriorating air quality in cities and excessive
temperatures at beaches could reduce recreational use and associated tourist income.
Predictions for northern Europe include the
following:
• Cold winters will be half as frequent by 2020.
• Northern tundra will retreat and there could be
a loss of up to 90% of alpine glaciers by the
end of the century.
• Conversely, climate changes could increase
agricultural and forest productivity and water
availability, although the risk of flooding could
increase (above).18
Oilfield Review
50387schD01R1
11/29/01
5:20 AM
Page 53
The Sociopolitical Debate and Its Impact
on Process and Technology
On balance, the potential dangers and adverse
effects of global warming far outweigh any possible benefits. Both legislative and technical
options are being explored to mitigate the
impacts of future climate change.
With its 100-year effective lifetime, CO2 concentration in the atmosphere is slow to respond
to any cut in emissions. If nothing is done to
reduce emissions, the concentration would more
than double over the next century. If emissions
are lowered to 1990 levels, the concentration
would still rise, probably to more than 500 ppm.
Even if emissions were slashed to half that level
and held there for 100 years, there would still be
a slow rise in concentration. Best estimates suggest it would take a reduction of 60 to 70% of the
1990 emission levels to stabilize the concentration of CO2 at the 1990 levels.19
Against this backdrop, there have been political attempts to grapple with the problem for
nearly a decade. These have achieved, at best,
modest results. Although an in-depth discussion
of global-warming politics is beyond the scope of
this technically focused article, conferences held
to date and their resulting protocols illustrate the
challenges that will be faced by new-generation
oilfield processes and technologies, and by business and industry in general (above).
The political movement toward global consensus began in 1992 at the United Nations
Conference on Environment and Development
held in Rio de Janeiro, Brazil. This conference
resulted in the United Nations Framework
Convention on Climate Change (UNFCCC), a
statement of intent on the control of greenhousegas emissions, signed by an overwhelming
majority of world leaders. Article II of the convention, which came into force in 1994, said the
signatories had agreed to “achieve stabilization
of greenhouse-gas concentrations in the atmosphere at a level that would prevent dangerous
anthropogenic interference with the climate system…within a time frame sufficient to allow
ecosystems to adapt naturally to climate change,
to ensure that food production is not threatened,
and to enable economic development to proceed
in a sustainable manner.” The developed nations
taking part also committed themselves to reduce
their emissions of greenhouse gases in the year
2000 to 1990 levels.
A more ambitious target was set in 1997 in
the Kyoto Protocol, an agreement designed to
Autumn 2001
Conference
_____
Outcome
1992
1997
2000
2001
Rio de Janeiro,
Brazil
_________
Kyoto,
Japan
_________
The Hague,
The Netherlands
_________
Bonn,
Germany
_________
Statement of
intent on control
of greenhouse
gases
Protocol on
reduction levels
for specific
commitment
period
Collapse of
implementation
plan for Kyoto
Protocol
Broad agreement
on rulebook
for implementing
Kyoto protocol
(except USA)
> Major international global warming conferences. A concerted effort at
addressing the sociopolitical implications of global warming in a forum of
nations began in 1992 in Rio de Janeiro, Brazil. The most recent conference,
held in July 2001 in Bonn, Germany, was the latest attempt to reach some
type of formalized agreement on reducing greenhouse-gas emissions.
commit the world’s 38 richest nations to reduce
their greenhouse-gas emissions by an average of
at least 5% below 1990 levels in the period from
2008 to 2012.20 The Kyoto Protocol put most of
the burden on developed countries, which, as a
group, had been responsible for the majority
of greenhouse gases in the atmosphere. It
excluded more than 130 developing countries,
even though many poorer nations were adding to
the problem in their rush to catch up with the
developed world. European Union (EU) countries
agreed to a reduction of 8%, and the USA
promised a 7% cutback, based on 1990 levels. To
take effect, it was agreed that the Protocol must
be ratified by at least 55 countries, including
those responsible for at least 55% of 1990 CO2
emissions from developed countries.
The targets set in Kyoto are more rigorous
than they might first appear since many developed economies have, until very recently, been
growing rapidly and are emitting greater volumes
of greenhouse gases. In 1998, for example, the
US Department of Energy forecasted that US
emissions in the year 2010 would exceed the
Kyoto target by 43%.
The November 2000 talks in The Hague on
implementing the Kyoto Protocol collapsed when
the EU rejected a request that the estimated
310 million tons of CO2 soaked up by forests in
the USA be set against its 7% commitment. The
EU suggested instead that the USA be allocated
a 7.5-million ton offset.
In July 2001, 180 members of the UNFCCC
finally reached broad agreement on an operational rulebook for the Kyoto Protocol at a meeting in Bonn, Germany. The USA rejected the
agreement. If the Protocol is to go forward, the
next step would be for developed-country
governments to ratify it so that measures could
be brought into force as soon as possible, possibly by 2002.
One issue resolved at the Bonn meeting was
how much credit developed countries could
receive towards their Kyoto targets through the
use of ‘sinks’ that absorb carbon from the atmosphere. There was agreement that activities that
could be included under this heading included
revegetation and management of forests, croplands and grazing lands. Individual country quotas
were set so that, in practice, sinks will account
only for a fraction of the emission reductions that
can be counted towards the target levels.
Similarly, storage options exist for carbon dioxide
that offer attractive alternatives to sinks under
certain conditions (see “Mitigating the Impact of
Carbon Dioxide: Sinks and Storage,” page 54).
The conference also adopted rules governing the
so-called Clean Development Mechanism (CDM)
through which developed countries can invest in
climate-friendly projects in developing countries
and receive credit for emissions thereby avoided.
(continued on page 56)
14. Reference 13: 13-14.
15. Reference 13: 14-15.
16. Reference 13: 6.
17. Climate Change Impacts on the United States, The
Potential Consequences of Climate Variability and
Change: Foundation Report, US Global Change Research
Program Staff. New York, New York, USA: Cambridge
University Press (2001): 6-10.
18. Parry ML (ed): Assessment of Potential Effects and
Adaptations for Climate Change in Europe. Norwich,
England: Jackson Environment Institute, University of
East Anglia, 2000.
19. Jenkins et al, reference 3: 10.
20. Kyoto Protocol, Article 31, available at Web site:
http://www.unfccc.de/resource/docs/convkp/kpeng.html
53
50387schD01R1
11/29/01
5:21 AM
Page 54
Mitigating the Impact of Carbon Dioxide: Sinks and Storage
In the short to medium term, the world will
continue to depend upon fossil fuels as cheap
energy sources, so there is growing interest in
methods to control carbon dioxide emissions—
for example, the creation of carbon sinks and
storage in natural reservoirs underground or in
the oceans.1
Carbon sinks—Carbon sinks are newly
planted forests where trees take CO2 from the
atmosphere as they grow and store it in their
branches, trunks and roots. If too much CO2 is
being pumped into the atmosphere by burning
fossil fuels, discharge levels can be compensated for, to some extent, by planting new trees
that soak up and store CO2.
In 1995, the IPCC estimated that some
345 million hectares [852 million acres] of new
forests could be planted between 1995 and 2050
that would sequester nearly 38 gigatons of carbon. These actions would offset about 7.5% of
fossil-fuel emissions. The IPCC added that other
measures, like slowing tropical deforestation,
could sequester another 20 to 50 gigatons.
Taken together, new forests, agroforestry, regeneration and slower deforestation might offset 12
to 15% of fossil-fuel emissions by the year 2050.
An attractive feature of this approach is that, if
implemented globally, it buys time during which
longer term solutions can be sought to meet
world energy needs without endangering the
climate system.
There are, however, other factors that must
be considered, such as how to quantify the
amount of carbon being sequestered, how to
verify sequestration claims and how to deal with
‘leakage.’ Leakage occurs when actions to
increase carbon storage in one place promote
activities elsewhere that cause either a
decrease in carbon storage (negative leak) or
an increase in carbon storage (positive leak).
Preserving a forest for carbon storage may, for
instance, produce deforestation elsewhere (negative leakage) or stimulate tree planting elsewhere to provide timber (positive leakage). The
carbon-sink process is reversible. At some
future date, some forests could become unsustainable, leading to a rise in CO2 levels.
Carbon storage—Carbon dioxide is produced
as a by-product in many industrial processes,
54
Sleipner
West
Sleipner
East
Statfjord
Gullfaks
NORWAY
Frigg
Heimdal
Stavanger
Sleipner
Ula
Ekofisk
NORTH SEA
DENMARK
UNITED
KINGDOM
GERMANY
> Sleipner field location.
usually in combination with other gases. If the
CO2 can be separated from the other gases—at
present, an expensive process—it can be stored
rather than released to the atmosphere. Storage
could be provided in the oceans, deep saline
aquifers, depleted oil and gas reservoirs, or on
land as a solid. Oceans probably have the greatest potential storage capacity. While there
are no real engineering obstacles to overcome,
the environmental implications are not adequately understood.
For years, carbon dioxide has been injected
into operating oil fields to enhance recovery,
and normally remains in the formation. The use
of depleted oil or gas reservoirs for CO2 storage,
however, has a further advantage in that the
geology is well-known, so disposal takes place in
areas where formation seals can contain the gas.
The first commercial-scale storage of CO2 in
an aquifer began in 1996 in the Sleipner natural
gas field belonging to the Norwegian oil company Statoil. The project is named SACS (Saline
Aquifer CO2 Storage) and is sponsored by the
EU research program Thermie. A million tons,
a year of CO2 production, are removed from the
natural gas stream using a solvent-absorption
process and then reinjected into the Utsira
reservoir, 900 m [2950 ft] below the floor of
the North Sea (above). According to a report by
the Norwegian Ministry of Petroleum and
Energy, the Utsira formation is widespread
and about 200 m [660 ft] thick, so it can theoretically accommodate 800 billion tons of
CO2—equivalent to the emissions from all
northern European power stations and major
industrial establishments for centuries to come
(next page, bottom).
Oilfield Review
50387schD01R1
11/29/01
5:21 AM
Page 55
To monitor the CO2-injection area,
Schlumberger is conducting four-dimensional
(4D), or time-lapse, seismic studies that compare seismic surveys performed before and during injection. A survey acquired in 1994, two
years before injection began, served as the baseline for comparison with a 1999 survey acquired
after about 2 million tons of CO2 had been
injected. Higher seismic amplitudes in the 1999
survey show the location where gas has displaced brine in the Utsira formation. Another
4D survey is scheduled for late 2001 (right).
The Sleipner CO2 sequestration project
already has inspired other oil and gas companies to consider or plan similar efforts in southeast Asia, Australia and Alaska.
Sleipner CO2 injection siesmic monitoring
E-W section preliminary raw stack
1. Cannell M: Outlook on Agriculture 28, no. 3: 171-177.
> Seismic responses due to carbon dioxide injection. A 1994 seismic survey (left)
served as a baseline for a 1999 survey (right) that showed the pattern of brine
displacement by carbon dioxide following injection of 2 million tons of the gas.
1994
1999
after injecting 2 millIon tons of CO2 since 1996
no change above this level
Top Utsira formation
–250 m
Injection point
500 m
Velocity push-down
beneath CO2 cloud
Depth, m
Sleipner T
Sleipner A
0
500
CO2 injection well
1000
CO2
Utsira formation
1500
Sleipner East production
and injection wells
2000
2500
0
500
1000
1500 m
0
1640
3280
4920 ft
Heimdal formation
> Carbon dioxide injection well in Utsira. The Utsira formation is about 200 m [660 ft] thick and can hold the equivalent of all carbon dioxide emissions
from all northern European power stations and industrial facilities for centuries to come.
Autumn 2001
55
11/29/01
5:25 AM
Page 56
BP Emissions-Reduction Program
_________
Capture and reuse emissions.
Stop deliberate venting of carbon
dioxide and methane.
Improve energy efficiency.
Eliminate routine flaring.
Develop technologies to separate
carbon dioxide from gas mixtures.
> Cutting emission levels. BP has undertaken an
aggressive, multifaceted program to reduce
emissions, ranging from improved energy efficiency to elimination of routine gas flaring.
The Kyoto Protocol includes a compliance
mechanism. For every ton of gas that a country
emits over its target, it will be required to reduce
an additional 1.3 tons during the Protocol’s second commitment period, which starts in 2013.
Some reports contend that concessions made at
the conference reduced emissions cuts required
by the Protocol from 5.2% to between 0 and 3%
in 2010. The UNFCCC is more cautious in its
statements. As of August of this year, its secretariat had not calculated how the Bonn agreements might affect developed-country emission
reductions under the Kyoto Protocol, and indicated that this would not be known with any precision until the 2008-2012 target period.
E&P Company Initiatives
Today, many oil and gas companies are taking
global warming seriously, convinced that it is sensible to adopt a precautionary approach. Others
have taken a more conservative stance: they
agree that climate change may pose a legitimate
long-term risk, but argue that there is still insufficient scientific understanding to make reasonable
predictions and informed decisions, or to justify
drastic measures. All agree that a combination of
process changes and advanced technologies will
be required within the industry to meet the types
of emission standards being proposed.
BP and Shell have implemented strategies
based on a judgment that while the science of
climate change is not yet fully proven, it is prudent to behave as though it was. Both companies
have established ambitious internal targets for
reduction of their own emissions. The Kyoto
Protocol calls for an overall reduction of greenhouse-gas emissions of at least 5% by 2008 to
2012, compared with 1990. BP has undertaken to
56
reduce its greenhouse-gas emissions by 10% by
the year 2010, against the 1990 baseline. Shell
intends to reduce emissions by 10%, against the
same baseline, by 2002.
Companies are choosing to cut emissions in
several different ways. The BP emissions reduction program, for instance, includes ambitious
commitments:
• Ensure that nothing escapes into the environment that can be captured and, ideally, used
elsewhere. BP intends to stop the deliberate
venting of methane and carbon dioxide wherever possible. This may involve redesigning or
replacing equipment, and identifying and eliminating leaks.
• Improve energy efficiency. Engineers are examining all energy-generating equipment to
ensure that the company is making the best
possible use of hydrocarbon fuels and the heat
that is a by-product of energy generation.
• Eliminate routine flaring. It is better to flare gas
than vent it directly to the atmosphere, but it
is still a waste of hydrocarbons—although
some flaring may still be necessary for
safety reasons.
• Develop technology to separate carbon dioxide
from gas mixtures, then reuse it for enhanced
oil recovery or store it in oil and gas reservoirs
that are no longer in use, or in saline formations (above).
Integrated oil companies also are trying to
help customers reduce greenhouse-gas emissions
by increasing the availability of fuels with lower
carbon content and offering renewable energy
alternatives, like solar and wind-driven power.
Some companies, including BP and Shell,
have introduced internal greenhouse-gas emissions trading systems. The attraction of emissions
trading is that it allows reductions to be achieved
at the lowest cost; companies for whom emissions reductions are cheap can lower their
emissions and sell emission rights to firms that
would have to pay more to decrease emissions.
The BP emissions trading system is based on
a cap-and-trade concept, and was primarily
designed to provide BP with practical experience
dealing with an emissions trading market and to
learn about its complexities. At its simplest level,
a cap is set each year to steer the group toward
the most efficient use of capital to meet its 2010
target of 10%. Say, for example, increased production is planned from an offshore platform,
thereby causing emissions above its allocated
allowance. If the platform’s on-site abatement
costs are higher than the market price of CO2, the
company may decide to purchase CO2
allowances for that unit. Similarly, if a downstream unit has upgraded its refinery and emits
less CO2 than its allowances cover, it is economically desirable to both companies if the latter
sells its allowances to the former (below).
The operation of these systems will be
closely followed not only by other oil and gas
companies but also by governments, since the
principles behind emissions trading are broadly
the same whether trading takes place within a
single company, among companies within a single country, among companies internationally or
between nations.
Oilfield Technology Development
and Application
Working with oil and gas companies, major oilfield service suppliers have been at the forefront
in addressing a range of health, safety and environmental issues—from reducing personnel
exposure to risks at the wellsite, to application of
‘green’ chemicals that provide equal or enhanced
performance while decreasing ecological impact,
and to methods for cutting or eliminating emissions resulting from processes such as burning
oil and flaring gas during well-testing operations.
Emission limit
after trading
Units bought
Carbon dioxide emissions
50387schD01R1
–10
Units sold
40
Each company
initially is
allocated 50
permits to emit
50 tons
Company A
+10
Emission limit
before trading
50
Company B
> Emissions trading system. This process strives to reduce emissions at the
lowest cost by permitting the buying and selling of emissions rights between
various units within a given company or between companies.
Oilfield Review
50387schD01R1
11/29/01
5:26 AM
Page 57
Gas
Flaring
Series of pumps
Produced fluid
Oil
Pipeline
Water and oil emulsion
Disposal
Stage 1
Separator
Flaring
Gas
Produced fluid
Gas and oil
Neutralizer and emulsion breaker
Series of pumps
Separator
Stage 2
Oil
Broken emulsion
Skimmer
Oil
Pipeline
Surge tank
Clean water
Produced fluid
Gas and oil
Neutralizer and emulsion breaker
Disposal
Gas and oil
Multiphase flowmeter
Multiphase pump
Stage 3
Pipeline
Broken emulsion
Skimmer
Oil
Surge tank
Clean water
Disposal
> Three-stage program to eliminate flaring. A Schlumberger team in the Middle East committed to first reduce and then fully eliminate flaring of gas and
burning of oil and, at the same time, generate greater revenue for the operator by increasing pipeline throughput.
Solutions to eliminate flaring—Burning oil
and flaring natural gas during testing operations
not only are costly due to lost revenue, but also
produce large quantities of carbon dioxide. Small
amounts of toxic gases, soot and unburned
hydrocarbons are also released. Eliminating oil
burning and, ultimately, gas flaring not only creates a safer working environment, but also helps
reduce the key constituent, carbon dioxide,
thought to be associated with global warming.
Recently, a Schlumberger team in the Middle
East, working closely with a major operator in the
region, addressed the flaring problem for production testing where an existing export pipeline
was available. Considering the nature of the testing program, there were several key challenges
that had to be overcome. Wells are typically
highly deviated or horizontal, and penetrate massive carbonate formations. Large quantities of
acid are used to treat the zones, giving rise to
long cleanup periods and an erratic initial flow of
mixtures of spent acid, emulsions, oil and gas.
Autumn 2001
Traditionally, the wells were flowed until sufficient oil was produced at sufficient pressure to
go directly into the production pipeline, requiring
burning of oil in the interim. Care had to be taken
that the fluid’s pH was high enough so as not to
cause corrosion problems.
A three-stage program to eliminate flaring
and simultaneously solve associated well-testing
problems was undertaken. In the first stage,
beginning in 1998, the goal was to pump separated oil into the pipeline from the outset,
instead of burning it. This required the design of
specialized, dual-packing centrifugal pumps that
were run in series to achieve the required pressure for oil injection into the pipeline. Natural
gas was still flared, and separated water discarded. Residual oil and water emulsions
remained a problem, since a single separator
was insufficient to break them.
In the second stage of the project, a neutralizer and breaker system was designed for treatment of the emulsion phase prior to entering the
main separator. Remaining gas and oil were then
flowed through the separator. A skimmer and
chemical injection system were employed to
reduce the oil content in the water underflow
stream from 3000 ppm to less than 80 ppm,
allowing safe disposal of all residual water. Oil
produced through emulsion breaking was
pumped into a surge tank and then into the production pipeline, saving additional oil that would
have otherwise been discarded.
In the third stage, currently under way, the
goal is for complete elimination of flaring by
using advanced multiphase pumping technology
with multiphase metering. When the wellhead
pressure is insufficient to route gas back through
the line after the multiphase meter, a variabledrive multiphase pump—that can handle a variety of flow rates and pressures—would be
introduced so that both oil and gas can be
injected into the production pipeline (above).
57
50387schD01R1
11/29/01
5:21 AM
Page 58
In the first year of implementation of the initial
stages of the project, the operator was able to sell
an additional 375,000 barrels [59,600 m3] of oil
that otherwise would have been burned, generating more than $11 million in increased revenues.21
Zero-emission testing—The next frontier is a
generalized solution for zero-emission testing for
exploration and appraisal wells where an export
pipeline is not available. Here, the challenge is to
take a quantum step beyond improved burner
technology. The goal is elimination of all emissions by keeping produced hydrocarbons contained either below surface or the mudline, or in
special offshore storage vessels. Through the use
of advanced downhole measurements and tools,
high-quality test data and samples could still
be captured.
There are several approaches to downhole
containment. In particular, three options are
currently undergoing intensive investigation. The
first is closed-chamber testing. Here, test fluids
flow from the formation into an enclosed portion
of a tool or pipestring. A short flow period is
achieved as the chamber fills and its original contents become compressed. Flow stops as the
chamber reaches equilibrium, allowing analysis
of the subsequent buildup. This method, applicable to both oil and gas wells, is simple, and the
short test duration limits rig time compared with
a conventional test. But, there are drawbacks.
With only a small flowed volume due to capacity
limitations of the test string or wellbore, only a
limited radius of investigation near the wellbore
can be evaluated. Lack of thorough cleanup after
perforating can potentially affect the quality of
collected samples. If the formation is not wellconsolidated, hole damage or collapse may occur
because of high inflow rates (below left).
Surface valve
A second method is production from one zone
and reinjection into the same zone, known as
harmonic testing. Here, fluid is alternately withdrawn into a test string and then pumped back
into the reservoir at a given periodic frequency.
The reservoir signature is determined point-bypoint as a function of frequency by varying the
frequency during testing. The advantage is that a
21. The team that spearheaded this project won the
Performed by Schlumberger Chairman’s Award 2000,
the top award in a company-wide program to strengthen
the Schlumberger culture of excellence. Client team
members included Abdullah Faddaq, Suishi Kikuchi,
Mahmoud Hassan, Eyad Al-Assi, Jean Cabillic,
Graham Beadie, Ameer El-Messiri and Simon Cossy.
Schlumberger team members included Jean-Francois
Pithon, Abdul Hameed Mohsen, Mansour Shaheen,
Thomas F. Wilson, Nashat Mohammed, Aouni El Sadek,
Karim Mohi El Din Malash, Akram Arawi, Jamal Al
Najjar, Basem Al Ashab, Mohammed Eyad Allouch,
Jacob Kurien, Alp Tengirsek, Mohamed Gamad and
Thomas Koshy.
Tubing
Circulating valve
Barrier valve
Upper packer
Circulating valve
Ball valve
Downhole
pump assembly
Gas-liquid
interface
Test valve
Produced fluid
and initial
liquid cushion
Packer
Lower packer
Pressure
gauge
Sand screen and
gravel pack
> Closed-chamber testing. Test fluids from the
formation enter an enclosed space until the contents compress and reach equilibrium. This brief
flow period is then followed by a second stage of
pressure buildup.
58
Flow direction
> Continuous production and reinjection. A specially designed tool allows produced fluid from
one zone to be continuously injected into another
using a downhole pump to provide a prolonged
testing period. Samples can be retrieved, and
flow and pressure data are measured downhole
for subsequent analysis.
Oilfield Review
50387schD01R1
11/29/01
5:26 AM
Page 59
Drilling and
production unit
Storage modules and
processing facilities
Dynamically positioned
storage or shuttle tanker
Rigid production
riser
Export flowline
BOP or subsea
test tree
> Offshore storage-module concept. A vessel for storing and offloading fluids collected in closed
modules during testing operations might offer an approach to eliminate the need for flaring while
generating increased revenues.
separate zone for disposal of the produced fluid
is not needed, but defining the pressure-response
curve would require more time than for a conventional test and may not be cost-effective.
Advanced signal processing may be able to
reduce the time required, but still may not make
the process economically viable.
The third method is to continually produce
from one zone and inject the produced fluid into
another zone. Reservoir fluids are never brought
to surface, but are reinjected using a downhole
pump. Drawdown is achieved by pumping from
the production zone into the disposal zone.
Buildup is provided by simultaneously shutting in
the production zone and stopping the downhole
pump. If injectivity can be maintained, this continuous process emulates a full-scale well test. A
larger radius of investigation is possible due to
larger flow volumes, with the potential to investigate compartmentalization or even reservoir
limits. A longer flow period improves cleanup
prior to sampling. Flow and pressure are measured downhole and analyzed with conventional
methods for radial flow. It is possible to capture
small pressure-volume-temperature (PVT)-quality
samples and larger dead-oil samples downhole.
Drawbacks include a somewhat complex tool
string, an inability to handle significant quantities
of gas and no time-saving over a conventional
well test. The key factor is having a suitable
injection zone that provides sufficient isolation
(previous page, bottom right).
Two joint industry programs have been established to investigate each of the three methods in
detail, with participation by BP, Chevron, Norsk
Hydro and Schlumberger. The first, conducted by
Schlumberger, is assessing downhole tool design
Autumn 2001
and capability requirements. The second, a threeyear program at Imperial College in London,
England, is defining the interpretation packages
and procedures that would be required to capture
the maximum amount of reliable information
from the data.
Once the selection of the preferred method is
finalized, the next step will be a proof-of-concept
field experiment that mirrors the requirements of
a variety of well-test conditions. Currently, the
continuous production-reinjection option looks
most promising.
Modules mounted on the deck or in the hold
of a suitable floating vessel are being investigated for storing fluids collected offshore during
testing. Fluid-processing facilities also would be
provided onboard. Large discoveries, marginal
fields and deepwater prospects are targeted
applications. Equipment would be designed to
handle a broad range of testing conditions and
durations. The vessel would receive and store
gas and liquids, and offload the contents at the
end of the well test or at intervals during the test.
This concept could completely eliminate the need
for flaring, and generate revenues from sale of
produced fluids that would otherwise be lost. The
procedures for handling and storing liquids have
already been successfully demonstrated in
extended well tests in fields such as BP’s
Machar—proving both the feasibility and financial viability of the approach. Gas handling and
storage, however, pose additional challenges
that would probably require compression
and transfer facilities to create compressed
natural gas. This is a costly proposition and
may not be economically viable at current
gas prices (above).
With growing emphasis on eliminating all
types of gas emissions, particularly carbon dioxide, these areas of investigation are expected to
continue to receive close attention and significant industry funding.
Future Challenges
In the near future, governments around the world
will receive the IPCC Synthesis Report which will
attempt to answer, as clearly and simply as possible, 10 policy-relevant scientific questions.
Perhaps the pivotal question, as stated by the
IPCC, is: “How does the extent and timing of the
introduction of a range of emissions-reduction
actions determine and affect the rate, magnitude
and impacts of climate change, and affect global
and regional economies, taking into account historical and current emissions?”
In another five years, the IPCC is expected to
publish its Fourth Assessment Report. By then,
climatologists may have resolved some of the
uncertainties that limit today’s climate models.
They should, for example, be able to provide a
better description of the many feedback systems
associated with climatic phenomena, particularly
clouds. Greater understanding could lead to
reduced uncertainty about a causal connection
between increased greenhouse-gas concentrations and global warming. This would be a major
step forward.
In the interim, oil and gas companies,
working closely with oilfield service companies,
will continue to be proactive in developing
technologies and operational procedures for
reducing emissions.
—MB/DEO
59
50387schD10R1.p60.ps 12/7/01 8:52 PM Page 60
Isolate and Stimulate Individual Pay Zones
Coiled tubing-conveyed fracturing is a cost-effective alternative to conventional
reservoir-stimulation techniques. This innovative approach improves hydrocarbon
production rates and recovery factors by providing precise, reliable placement of
treatment fluids and proppants. What began as a fracturing service is evolving into
broad technical solutions for new completions, as well as workovers in mature fields.
Kalon F. Degenhardt
Jack Stevenson
PT. Caltex Indonesia
Riau, Duri, Indonesia
Byron Gale
Tom Brown Inc.
Denver, Colorado, USA
Duane Gonzalez
Samedan Oil Corporation
Houston, Texas, USA
Scott Hall
Texaco Exploration and Production Inc.
(a ChevronTexaco company)
Denver, Colorado
Jack Marsh
Olympia Energy Inc.
Calgary, Alberta, Canada
Warren Zemlak
Sugar Land, Texas
ClearFRAC, CoilFRAC, CT Express, DepthLOG, FMI (Fullbore
Formation MicroImager), Mojave, NODAL, PowerJet,
PowerSTIM, PropNET, SCMT (Slim Cement Mapping Tool)
and StimCADE are marks of Schlumberger.
For help in preparation of this article, thanks to Taryn
Frenzel and Bernie Paoli, Englewood, Colorado; Badar Zia
Malik, Duri, Indonesia; and Eddie Martinez, Houston, Texas.
60
Operators traditionally rely on drilling programs to
achieve peak productivity, maintain desired production levels and optimize hydrocarbon recovery.
As oil and gas developments mature, however,
reservoir depletion reduces field output and fewer
opportunities exist to drill new wells. Drilling programs alone may not effectively stem the natural
decline of production. In addition, infill and reentry drilling often become less profitable and present greater operational and economic risks
relative to their higher capital investments.
In many fields, operators intentionally and
unintentionally bypass some pay zones during
initial phases of field development by focusing
only on the most prolific producing horizons.
Cumulatively, these marginal pay intervals contain substantial hydrocarbon volumes that can be
produced, especially from laminated formations
and low-permeability reservoirs. Accessing
bypassed pay zones is economically attractive to
enhance production and increase reserve recovery, but poses several challenges.
Typically, bypassed zones have lower permeabilities and require fracturing treatments to
achieve sustainable commercial production.
Conventional well-intervention and stimulation
methods involve extensive remedial operations,
such as mechanically isolating existing perforations or squeezing them with cement and utilizing multiple runs to perforate bypassed pay.
These procedures are expensive and cannot be
justified for zones with limited production potential. In the past, fracture stimulations were not
commonly attempted on bypassed pay, especially
when multiple stringers were involved.
The mechanical condition of wellbores can be
a limitation as well. If fracture stimulations are not
anticipated during well planning, completion tubulars may not be designed to withstand highpressure pumping operations. Also, scale buildup
and corrosion from prolonged exposure to formation fluids at reservoir temperatures and pressures
can compromise tubular integrity in older wells. In
slimhole wells, workover options are further limited by small tubulars. These operational and economic constraints often mean that bypassed or
marginal pay remains untapped. Ultimately, hydrocarbons in these intervals are left behind when
wells are plugged and abandoned.
Integration of coiled tubing with fracturing
operations overcomes many of the constraints
associated with stimulating bypassed or
marginal pay zones using conventional techniques, allowing additional reserves to be tapped
economically. High-strength continuous coiled
tubing strings transport treatment fluids and
proppants to target intervals and protect existing
wellbore tubulars from high-pressure pumping
operations, while specialized downhole tools
selectively isolate existing perforations with
increased precision.
Oilfield Review
50387schD10R1.61.ps 12/06/2001 01:46 AM Page 61
> A fit-for-purpose CT Express coiled tubing unit performing a selective fracturing treatment in Medicine Hat, Alberta, Canada.
Autumn 2001
61
50387schD10R1.62.ps 12/06/2001 01:46 AM Page 62
This article describes operational and design
aspects of coiled tubing-conveyed fracturing
treatments, including enabling technologies such
as surface equipment improvements, high-pressure coiled tubing, low-friction fracturing fluids
and new downhole isolation tools. Case histories
demonstrate how this technique reduces completion time and cost, improves post-treatment
cleanup, increases production and helps tap
reserves bypassed by conventional completion
and fracturing methods.
Conventional Stimulations
Average recovery factors for most reservoirs from
primary- and secondary-drive mechanisms are
just 25 to 35% of original hydrocarbons in place.
Producible reserves also are left behind in thin,
lower permeability zones of many mature reservoirs. One North Sea study, for example, determined that more than 25% of recoverable
reserves lie in the low-permeability, laminated
horizons of Brent sandstone reservoirs.1
Matrix acidizing and hydraulic fracturing are
common reservoir-stimulation techniques used to
enhance well productivity, increase recovery efficiency and improve well economics.2 However,
effectively completing and stimulating heteroge-
neous reservoirs and discontinuous pay zones
among numerous shale intervals are challenging,
particularly when fracture stimulations are
required. Reservoir pay thickness, quality, pressure and stage of depletion, and cost to treat an
entire productive horizon all must be considered
when choosing completion strategies.
Conventional fracture stimulations attempt to
connect as many producing zones as possible
with single or multiple treatments performed during separate operations. Historically, net pay
zones over several hundred feet of gross interval
are grouped into “stages,” with each stage stimulated by a separate fracturing treatment. These
massive hydraulic fracturing jobs, pumped
directly down casing or through standard jointed
tubing, are designed to maximize fracture height
while attempting to optimize fracture length.
However, uncertainty associated with predicting
height growth often compromises the stimulation
objectives of large treatments and precludes creation of the fracture lengths required to optimize
effective wellbore radius and reserve drainage.
Proppant placement in individual zones is difficult to achieve when a single treatment is performed across numerous perforated zones
(below). Thin or low-permeability zones grouped
> Single-stage treatment diversion: radioactive tracers and production logs. With limited-entry techniques, some zones are not stimulated effectively and others may remain untreated. In this example,
six pay zones over a 300-ft [90-m] gross interval were fractured through 24 perforations. A radioactivetracer survey shows that the three upper zones received most of the treatment fluids and proppant,
while the three lower zones were not adequately stimulated (left). If an interval did not take fluid at the
beginning of a treatment, perforation erosion in other sands eliminated the backpressure necessary
for diversion. The lowest zone contributes no production; the other two contribute very little flow on
the production log spinner survey (right).
62
with thicker zones may remain untreated or may
not be stimulated effectively, and some zones are
occasionally bypassed intentionally to ensure
effective stimulation of more prolific
pay. Limited-entry perforations and ball sealers
distribute fluid efficiently during pad injection,
but less effectively during proppant placement
as perforations are enlarged by erosion or
treatment fluids flow preferentially into higher
permeability zones.3
Unintentionally bypassed and untreated
zones also are attributed to variable in-situ
stresses. In past conventional fracturing designs,
the fracture gradient, or stress profile, was
assumed to be linear and to increase gradually
with depth. In reality, formation stresses often
are not uniform across an entire geologic horizon,
and again, some zones may be difficult to treat
and stimulate effectively (next page, top).
Grouping pay zones in smaller stages overcomes some of these limitations and helps
ensure sufficient fracture coverage, but multistage treatments usually require several perforating and fracturing operations in succession.
Isolating individual zones for conventional fracture stimulations with workover rigs and jointed
tubing is problematic as well, requiring additional equipment and workover procedures.
There are fixed costs associated with each stage
of multistage fracturing operations. Conventional
fracturing operations add redundancy to stimulation operations and increase overhead costs.
Every time wireline units and pumping equipment are moved onto a wellsite for perforating
and stimulation operations there are separate
mobilization and setup charges. There are also
separate coiled tubing or slickline costs to wash
out sand plugs or set and retrieve bridge plugs,
which have to be purchased or rented. Hauling,
handling and storing stimulation and displacement
fluids for each nonconsecutive fracturing operation involve additional costs. Testing each individual stage in a well again requires multiple setups
and significantly increases completion time.
Some gas wells with several large treatment
stages may take weeks to complete. Redundant
charges accumulate quickly on wells with more
than three or four stages and significantly affect
the economics of stimulation procedures. These
higher costs typically become a major influence
on completion or workover decisions and strategies and may limit development of marginal pay
zones that cumulatively contain sizeable volumes
of oil and gas.
To stimulate bypassed zones in existing
wells, conventional fracturing requires that lower
producing zones be isolated by a sand plug or
Oilfield Review
Increasing depth
50387schD10R1.63.ps 12/06/2001 01:46 AM Page 63
> Variations in formation stress. In single, multizone treatments, pressure
changes are assumed to be linear with depth (far left). Depleted zones cause
pressure to decrease abruptly (middle left). Excessively depleted sands also
reduce pressure over extensive intervals (middle right). In some cases, formations have pressure and stress variations that make diversion of treatment
fluids and stimulation coverage during a single-stage treatment extremely
difficult (far right).
> Conventional and selective stimulations. Fracturing several zones grouped
in large intervals, or stages, is a widely used technique. However, fluid diversion and proppant placement are problematic in discontinuous and heterogeneous formations. Conventional treatments, like this four-stage example,
maximize fracture height, often at the expense of fracture length and complete interval coverage (left). Some zones remain untreated or may not be
stimulated adequately; others are bypassed intentionally to ensure effective
treatment of more permeable zones. Selective isolation and stimulation with
coiled tubing, in this case nine stages, overcome these limitations, allowing
engineers to design optimal fractures for each pay zone of a productive
interval (right).
Autumn 2001
downhole mechanical tool such as a retrievable
or drillable bridge plug. Upper perforations are
sealed off by cement squeezes that are often difficult to achieve, require additional rig time and
add to completion costs. There also is a risk that
squeezed perforations will break down during
high-pressure pumping operations.
These limitations, inherent in conventional
fracturing techniques, reduce stimulation effectiveness. Unconventional well intervention and
stimulation techniques are needed to ensure
hydrocarbon production from as many intervals
as possible, especially from zones that previously
could not be completed economically. Coiled tubing-conveyed fracturing techniques overcome
many of the limitations associated with conventional fracturing treatments (below left).4
Selective Stimulations
Combining coiled tubing and stimulation services
is not new. In 1992, coiled tubing was used to
fracture wells in Prudhoe Bay, Alaska, USA. The
31⁄2-in. coiled tubing was connected into the wellhead and left in the well as production tubing to
help maintain flow velocity. This technique
never gained wide acceptance because it was
limited to smaller intervals and lower treating
pressures in wells where a single zone was
targeted for completion.
1. Hatzignatiou DG and Olsen TN: “Innovative Production
Enhancement Interventions Through Existing Wellbores,”
paper SPE 54632, presented at the SPE Western regional
Meeting, Anchorage, Alaska, USA, May 26-28, 1999.
2. In matrix treatments, acid is injected below fracturing
pressures to dissolve natural or induced damage that
plugs pore throats.
Hydraulic fracturing uses specialized fluids injected at
pressures above formation breakdown stress to create
two fracture wings, or 180-degree opposed cracks,
extending away from a wellbore. These fracture wings
propagate perpendicular to the least rock stress in a
preferred fracture plane (PFP). Held open by a proppant,
these conductive pathways increase effective well
radius, allowing linear flow into the fractures and to the
well. Common proppants are naturally occurring or
resin-coated sand and high-strength bauxite or ceramic
synthetics, sized by screening according to standard US
mesh sieves.
Acid fracturing without proppants establishes conductivity by differentially etching uneven fracture-wing surfaces in carbonate rocks that keep fractures from
closing completely after a treatment.
3. Limited entry involves low shot densities—1 shot per foot
or less—across one or more zones with different rock
stresses and permeabilities to ensure uniform acid or
proppant placement by creating backpressure and limiting pressure differentials between perforated intervals.
The objective is to maximize stimulation efficiency and
results without mechanical isolation like drillable bridge
plugs and retrievable packers. Rubber ball sealers can
be used to seal open perforations and isolate intervals
once they are stimulated so that the next interval can be
treated. Because perforations must seal completely, hole
diameter and uniformity are important.
The pad stage of a hydraulic fracturing treatment is the
volume of fluid that creates and propagates the fracture
and does not contain proppant.
4. Zemlak W: “CT-Conveyed Fracturing Expands Production
Capabilities,” The American Oil & Gas Reporter 43, no. 9
(September 2000): 88-97.
63
50387schD10R1.64.ps 12/06/2001 01:46 AM Page 64
> Coiled tubing-conveyed fracturing with a single tension-set packer and sand plugs.
By 1996, coiled tubing-conveyed fracturing
was identified as a preferred completion strategy
for shallow gas fields in southeastern Alberta,
Canada.5 Selective placement of proppant in all
the productive intervals of a wellbore reduced
completion time and enhanced productivity. The
best candidates were wells with multiple lowpermeability zones where gas production was
commingled after fracturing. Previously, these
wells were stimulated by fracturing one interval
per well and then moving to the next well. While
a fracturing crew treated the first interval of the
next well, a rig crew prepared previous wells for
fracturing of subsequent intervals.
Extensive rig-up and rig-down times were
required to treat as many as four wells a day. In
terms of number of treatments performed, this
process was efficient, but moving equipment
from one location to another took more time than
actually pumping the fracturing treatments.
Operators evaluated the possibility of grouping
zones into stages for conventional multizone
stimulations using limited-entry perforating, ball
sealers or other diversion techniques to individually isolate zones, but could not justify these
standard industry practices economically.
One solution was to use a coiled tubing tension-set packer and sand plugs for zonal isolation.
The lowest zones were treated first by setting the
64
packer above the interval to be fractured.
Proppant schedules for each zone included extra
sand to leave a sand plug across fractured intervals after pumping stopped and before treating
the next zone. Each treatment was underdisplaced, and wells were shut in to allow the extra
sand to settle into a plug. A pressure test verified
sand-plug integrity and the packer was reset
above the next interval. This procedure was
repeated until all pay intervals were stimulated
(above). The larger coiled tubing string was rigged
down and smaller coiled tubing was brought in to
wash out sand and initiate well flow.
Coiled tubing-conveyed fracturing has since
expanded to slimhole wells—23⁄8-, 27⁄8- and 31⁄2-in.
tubulars cemented as production casing—and to
wells with open perforations or questionable
tubular integrity that prevented fracturing down
casing. Conventional workovers and stimulations
that require cement squeezes to isolate open
perforations are expensive and risky under these
conditions. Shallow gas and deeper coiled tubing
stimulations in mature oil and gas regions of the
continental region of the United States formed
the basis for CoilFRAC selective isolation and
stimulation services.
In east Texas, USA, coiled tubing was used to
stimulate wells with open perforations above
bypassed zones and wells with low-strength
27⁄8-in. production casing weakened further by
corrosion. After the target zone was perforated, a
tension-set packer on coiled tubing isolated the
wellbore and upper perforations (next page, top
left). In south Texas, bypassed pay zones
between open perforations in wells with casing
damage near the surface were stimulated successfully by setting a bridge plug below the target zone and then running a tension-set packer
on coiled tubing (next page, top right). These
fracture stimulations were performed without
cementing existing perforations or exposing production casing to high pressures.
Early CoilFRAC techniques with tension-set
packers improved stimulation results, but were
still time-consuming and limited by having to set
and remove plugs. The next step was to develop
a coiled tubing straddle-isolation tool that sealed
above and below an interval to eliminate separate operations for spotting sand or setting bridge
plugs with a wireline unit (next page, bottom). This
modification allowed coiled tubing strings to be
moved quickly from one zone to the next without
pulling out of the well.
5. Lemp S, Zemlak W and McCollum R: “An Economical
Shallow-Gas Fracturing Technique Utilizing a Coiled
Tubing Conduit,” paper SPE 46031, presented at the
SPE/ICOTA Coiled Tubing Roundtable, Houston, Texas,
USA, April 15-16, 1998.
Zemlak W, Lemp S and McCollum R: “Selective Hydraulic
Fracturing of Multiple Perforated Intervals with a
Coiled Tubing Conduit: A Case History of the Unique
Process, Economic Impact and Related Production
Improvements,” paper SPE 54474, presented at the
SPE/ICOTA Coiled Tubing Roundtable, Houston, Texas,
USA, May 25-26, 1999.
Oilfield Review
50387schD10R1 12/14/01 5:12 PM Page 65
> Coiled tubing-conveyed fracturing with a single
tension-set packer for casing and tubing protection.
> Coiled tubing-conveyed fracturing with a single
packer and mechanical bridge plugs. In south
Texas, a well with casing damage near the surface and a bypassed zone between existing open
perforations was stimulated successfully with
coiled tubing. The operator set a bridge plug to
isolate the lower zone before running a tensionset packer on coiled tubing to isolate the upper
zone and protect the casing. This technique eliminated a costly workover and remedial cementsqueeze operations.
> Multistage coiled tubing-conveyed fracturing operation with early straddle-isolation tools.
Autumn 2001
65
50387schD10R1.66.ps 12/06/2001 01:46 AM Page 66
Elastomer cup-type seals were added above a
tension-set packer to isolate perforated intervals
and eliminate separate plug-setting operations.
However, additional modifications were required
to further reduce time and cost. In Canada, an
isolation tool with elastomer cups above and
below an adjustable ported spacer assembly, or
mandrel, was developed to allow multiple zones
to be treated in one trip (right).
This version of the straddle-isolation tool,
which had no mechanical slips to facilitate quick
moves and fishing, carried shallow-gas projects
in Canada through more than 200 wells and 1000
individual CoilFRAC treatments. Continuing
improvements to this tool allow bypassed and
marginal zones to be stimulated at nominal incremental cost. Efficient isolation and stimulation of
individual sands maximized completed net pay
and made zones previously considered marginal
economically viable.
More Experience in Canada
Wildcat Hills field is located west of Calgary,
Alberta, Canada, on the eastern slope of the
Rocky Mountains in a protected grassland area.6
This area has produced natural gas from deep
Mississippian discoveries since 1958. During the
early 1990s, two Olympia Energy wells tested
shallower Viking sands. The wells initially produced about 900 Mcf/D [25,485 m3/d], but
declined rapidly to 400 Mcf/D [11,330 m3/d].
Although pressure-buildup and production tests
indicated substantial reserves, the low reservoir
pressure, poor deliverability and high completion
costs precluded development of marginal
Viking zones.
A 1998 seismic survey identified a third Viking
target in an area where the formation was
uplifted by more than 3000 ft [914 m], potentially
creating natural fractures that might enhance gas
deliverability. The 3-3-27-5W5M well encountered about 45 ft [14 m] of pay in five zones
across 82 ft [25 m] of gross interval (next page,
top). An FMI Fullbore Formation MicroImager
microresistivity log verified existing natural fractures in the reservoir, but drillstem testing indicated a low pressure of 1100 psi [7.6 MPa].
Pressure-buildup tests before setting 41⁄2-in. casing and after perforating indicated drilling-fluid
invasion into natural fractures and additional formation damage from completion fluids.
A mud-solvent treatment failed to remove the
damage, so a fracturing treatment was selected
6. Marsh J, Zemlak WM and Pipchuk P: “Economic
Fracturing of Bypassed Pay: A Direct Comparison of
Conventional and Coiled Tubing Placement Techniques,”
paper SPE 60313, presented at the SPE Rocky Mountain
Regional/Low Permeability Reservoirs Symposium,
Denver, Colorado, USA, March 12-15, 2000.
66
> Coiled tubing isolation tools. The first CoilFRAC operations used a single
tension-set packer above a zone with sand plugs or bridge plugs to isolate
below the zone (left). Subsequent versions were modified to include an upper
elastomer seal cup above the zone and a lower packer to isolate below (middle). This second-generation tool was followed by a straddle design with elastomer seal cups on the top and bottom of a ported spacer, which increased
the speed of packer moves, and reduced execution time as well as operational
costs (right). These specialty tools eliminated rig and wireline operations
because sand plugs and bridge plugs were not needed. Coiled tubing could
be moved quickly from one zone to the next without pulling out of the well.
to increase gas deliverability. Fracturing down
casing with limited-entry diversion was not an
option because the well had already been perforated. The operator evaluated diversion with ball
sealers as well as mechanical zonal isolation
with sand plugs, bridge plugs or coiled tubing.
Ball-sealer effectiveness is questionable, especially during fracturing treatments, so mechanical diversion was deemed the most reliable
method to ensure stimulation of all pay zones.
With only 13 to 16 ft [4 to 5 m] between four
zones, engineers eliminated use of sand plugs
because close spacing made it difficult to accurately place the correct sand volumes.
Conventional jointed tubing with packers and
bridge plugs for isolation involved separate operations to treat individual zones one at a time from
the bottom up. This required repeated equipment
mobilization and demobilization, redundant services for each zone and retrieving or moving
bridge plugs after each treatment—all of these
made the costs prohibitive.
Oilfield Review
50387schD10R1.p67.ps 12/7/01 8:52 PM Page 67
> Well 3-3-27-5W5M, Wildcat Hills field. Previous attempts to stimulate the Viking formation as a continuous interval were not successful because of difficulty in intersecting multiple zones with conventional
single-stage fracture treatments. Closely spaced perforated intervals prohibited isolation with a packer
and sand or bridge plugs. Selective CoilFRAC treatment placement simulated four zones individually to
increase recovery by isolating and fracturing pay that often is bypassed or left untreated. Secondary
goals were to simplify several days of completion operations into a single day and reduce cost.
> Comparison of conventional and CoilFRAC Viking completions. Coiled tubing-conveyed fracture stimulations required 58% less total proppant, reduced
overall completion operations from 19 days to 4, and improved well cleanup
and fracturing fluid recovery. CoilFRAC treatment placement and simultaneous flowback improved fluid recovery and saved Olympia Energy about
$300,000 per well in the Wildcat Hills field, which reduced cost per Mcf/D by
about 78%.
Autumn 2001
The operator selected CoilFRAC services to
stimulate each zone separately and treat several
zones in a single day. On the first day, the jointed
tubing string used to perform production tests and
the solvent treatment was pulled from the well.
Coiled tubing, fracturing and testing equipment
was moved to location on the second day while a
wireline unit set a bridge plug to isolate the lower
Viking formation. The maximum recommended
interval that the isolation tool could straddle at
that time was 12 ft [3.7 m], which was less than
the length of the lowest interval, so a tension-set
packer was used to fracture the first zone.
Three fracture stimulations were attempted
on the third day. Sticking problems required the
straddle-isolation tool to be pulled for repair of
the elastomer seal cups. A casing scraper run
smoothed the rough casing. This step is now
performed routinely before CoilFRAC treatments
as part of wellbore preparation. Annulus pressure increased while pumping pad fluids in the
second interval, indicating possible communication behind pipe or fracturing into an adjacent
zone. This treatment was cancelled before initiating proppant, and the tool was moved to the
third interval.
After the fourth interval was stimulated, the
straddle-isolation tool was pulled, so that openended coiled tubing could be used to clean out
sand and unload fluids. On the fourth day, a snubbing unit ran jointed production tubing in the well
under pressure to avoid formation damage from
completion-fluid invasion.
To eliminate the snubbing unit, coiled tubing
now is used to run a packer with an isolation
plug. After the packer is set, coiled tubing is
released and removed from the well. The packer
plug controls reservoir pressure until jointed production tubing is run. A slickline unit then
retrieves the isolation plug, initiating well flow.
Before stimulation, the 3-3-27-5W5M well
flowed 3.5 MMcf/D [99,120 m3/d] of gas at
350-psi [2.4-MPa] surface pressure. After three
of the upper four zones were fractured successfully, the well produced 6 MMcf/D [171,818 m3/d]
at 350 psi. The well continued to produce at
5 MMcf/D [143,182 m3/d] and 450 psi [3.1 MPa]
for several months. The CoilFRAC treatment
delivered an economic production gain in addition to reducing cleanup time and simplifying
completion operations (left). Minimal operations
and faster cleanup helped bring production on
line sooner by reducing completion cycle time
from 19 to 4 days.
67
50387schD10R1.68.ps 12/06/2001 01:46 AM Page 68
Olympia Energy drilled six more wells in the
Wildcat Hills field after completion of the 3-3-275W5M well. Because the Viking formation varies
from well to well, the operator selected fracturing techniques based on sand thickness, fracture
containment barriers, vertical spacing between
sands and required number of treatments. Three
of these wells contained two or three thick Viking
sands that were fractured down casing. The
larger zones required higher pump rates to optimize fracture height and length, which ruled out
use of coiled tubing because of potentially excessive surface treating pressures.
Like the 3-3-27-5W5M well, the other three
wells had similar interbedded sand-shale
sequences and 6- to 13-ft [2- to 4-m] pay zones,
so Olympia Energy used CoilFRAC selective stimulations. This approach increased productivity
and recovery by selectively treating pay that had
been bypassed or not stimulated effectively, and
it ultimately decreased operational costs.
Pre- and post-treatment production logs were
run on the 4-21-27-5W5M well to evaluate
increased production from zones in one of the
wells that was fractured using coiled tubing
(below). Prior to fracturing, the well produced
2 MMcf/d [57,300 m3/d] with flow from
two intervals. After CoilFRAC treatments on
five intervals, gas production increased to
4.5 MMcf/D [128,900 m3/d] with flow from four
of the five intervals. Olympia Energy saved
$300,000 per well on fracturing operations alone
by using CoilFRAC techniques to stimulate
Wildcat Hills Viking wells. One of the original Viking
gas wells has been reevaluated and identified as
a candidate for stimulation with coiled tubing.
At a depth of 8200 ft [2500 m], this coiled tubing-conveyed application demonstrated the
impact of combining coiled tubing and stimulation technologies on well productivity and
reserve recovery. The smaller surface footprint,
less time on location and fewer wellsite visits
combined with less gas emissions and flaring as
a result of flowing, testing and cleaning up all the
pay zones at one time make CoilFRAC treatments
particularly attractive in environmentally sensitive areas like the grasslands around Wildcat
Hills field.
> Pre- (left) and post-stimulation (right) evaluation. Production log spinner surveys in Viking Well 4-2127-5W5M confirmed that CoilFRAC selective fracturing treatments in each Viking sand improved the
production profile and total gas rate (right).
68
Fracturing Designs and Operations
Coiled tubing-conveyed fracturing is constrained
by restrictions on fluid and proppant volumes
related primarily to smaller tubular sizes and
pressure limitations. The application of CoilFRAC
services requires alternative fracture designs,
specialized fluids, high-pressure coiled tubing
equipment, and integrated fracturing and coiled
tubing service teams to ensure effective stimulations and safe operations.7
Injection rates, fluid parameters, treatment
volumes, in-situ stresses and formation characteristics determine the net pressure available
downhole to create a specific fracture geometry—width, height and length. Minimum
pump rates are required to generate the desired
fracture height and to transport proppant along
the length of a fracture. Minimum proppant concentrations are needed to attain adequate fracture conductivity.
Coiled tubing strings have a smaller internal
diameter (ID) than the standard jointed workstrings used in conventional fracturing operations. At the injection rates required for hydraulic
fracturing, frictional pressure losses associated
with proppant-laden slurries can lead to high
treating pressures that exceed surface equipment and coiled tubing safety limits. Using larger
coiled tubing reduces friction pressures, but
increases equipment, logistics and maintenance
costs, and may not be practical for small-diameter slimhole and monobore wells.
This means that treatment rates and proppant
volumes for coiled tubing-conveyed fracturing
must be reduced compared with those of conventional fracturing. The challenge is to achieve
injection rates and proppant concentrations that
transport proppant effectively and create the
required fracture geometry. Coiled tubing-conveyed fracturing requires alternative equipment
and treatment designs to ensure acceptable surface treating pressures without compromising
stimulation results.
Reservoir characterization is the key to any
successful stimulation treatment. Like conventional fracturing jobs, coiled tubing treatments
must generate a fracture geometry consistent
with optimal reservoir stimulation. The preferred
approach is to design CoilFRAC pumping schedules that balance required injection rates and
optimal proppant concentrations with coiled tubing treating-pressure constraints. Fracturing fluid
selection depends on reservoir characteristics
and fluid leakoff, downhole conditions, required
fracture geometry and proppant transport. Fluids
Oilfield Review
50387schD10R1.69.ps 12/06/2001 01:46 AM Page 69
for CoilFRAC treatments include water-base linear or low-polymer systems and polymer-free
ClearFRAC viscoelastic surfactant (VES) fluids.8
In the past, polymers provided fluid viscosity
to transport proppant. However, residue from
these fluids can damage proppant packs and
reduce retained permeability. Engineers often
increase proppant volumes to compensate for
any reduced fracture conductivity, but slurry
friction increases exponentially with higher proppant concentrations and can limit the effectiveness of CoilFRAC treatments. Increased surface
treating pressure from frictional pressure losses
is the dominant factor in coiled tubing-conveyed
fracturing, so reducing surface pump pressures is
critical in CoilFRAC applications, particularly in
deeper reservoirs.
Because of their unique molecular structure,
VES fluids exhibit as much as two-thirds
lower frictional pressures than polymer fluids
(right). Nondamaging ClearFRAC fluids may provide adequate fracture conductivity with lower
proppant concentrations at acceptable surface
treating pressures. This facilitates optimized fracture designs. These fluid characteristics make
coiled tubing-conveyed fracturing feasible at commonly encountered well depths.
Another advantage of ClearFRAC fluids is
reduced sensitivity of fracture geometry to fluid
injection rate. Height growth is better contained,
resulting in longer effective fracture lengths,
which is particularly important when treating thin,
closely spaced zones. Fluids based on a VES also
are less sensitive at downhole temperatures
and conditions that cause fracturing fluids to
break prematurely.
If pumping stops because of an operational
problem or fracture screenout, the stable suspension and transport characteristics of ClearFRAC
fluids prevent proppants from settling too quickly,
especially between the seal cups of straddle-isolation tools. This allows time to clean out remaining proppant and decreases the risk of stuck pipe.
In addition, these fluids provide a backup contingency in high-risk environments, such as highangle or horizontal wells, where proppant settling
also can be a problem.
Recovering treatment fluids is critical when
target zones have low permeability or low bottomhole pressure. Another benefit of VES fracturing fluids is more effective post-stimulation
cleanup. Field experience has shown that VES
fluids break down completely in contact with
reservoir hydrocarbons, through extended dilution by formation water or under prolonged exposure to reservoir temperature, and are
transported easily into wellbores by produced fluids. Retained permeability is close to 100% of
Autumn 2001
> Effect of friction-reducing fluids. As CoilFRAC applications expand to include
deeper wells, low-friction fluids will be a key to future success. This plot compares surface-treating pressure versus depth for 2-in. coiled tubing using a
polymer-based fracturing fluid and a ClearFRAC viscoelastic surfactant (VES)
fluid, both with 4 ppa proppant concentrations.
original permeability with VES fluids. In addition,
treating and flowing back all the zones at one
time improve fluid recovery and fracture cleanup.
High-strength, 13⁄4- to 27⁄8-in. coiled tubing is
used to accommodate higher injection pressures.
Coiled tubing for fracturing operations is fabricated from high yield-strength, premium-grade
steels with high burst pressure. For example,
13⁄4-in., 90,000-psi [621-MPa] yield strength coiled
tubing has a burst-pressure rating of 20,700 psi
[143 MPa] and can withstand collapse pressures
of 18,700 psi [129 MPa]. Coiled tubing is hydrostatically tested to about 80% of its burst-pressure
rating, 16,700 psi [115 MPa] for this 13⁄4-in. string
prior to pumping operations, and maximum pump
pressure is set at 60% of the design
burst pressure, or about 12,500 psi [86 MPa], for
this example.
Because the entire coiled tubing string contributes to friction pressure, regardless of how
much is inserted in a well, the length of coiled
tubing on a reel should be minimized relative to
the deepest interval. There has been concern
that centrifugal forces on the proppant would
erode the inner wall of spooled coiled tubing.
However, visual and ultrasonic inspection before
and after fracturing found no erosion inside the
coiled tubing and detected only minor erosion at
coiled tubing connectors after pumping as many
as nine treatments.
Operational safety is critical at the high pressures required for hydraulic fracturing treatments. For example, personnel should not be
permitted near wellheads or coiled tubing equipment during pumping operations. Coiled tubingconveyed fracturing requires specialized surface
equipment and innovative modifications to
ensure safe operations and to deal with contingencies in the event of a screenout.9 On the
surface, coiled tubing equipment, such as quickresponse, gas-operated relief valves, remotely
operated fracturing manifolds and modifications
to coiled tubing reels and manifolds, allow highrate pumping of abrasive slurries.
Precise depth control also is important for
selective stimulations. Inaccurate positioning of
coiled tubing results in serious and costly problems—perforating off-depth, placing a sand plug
in the wrong place, problems positioning straddleisolation tools or stimulating the wrong zone.
Straddle-isolation tools must be positioned accurately across perforated intervals. Five types of
depth measurements are used: standard levelwind pipe measurements as coiled tubing comes
off the reel, a depth-monitoring system in the
injector head, mechanical casing-collar locators
and two new independent systems used
by Schlumberger—the Universal Tubing-Length
Monitor (UTLM) surface measurement and the
DepthLOG downhole casing-collar locator.
7. Olejniczak SJ, Swaren JA, Gulrajani SN and Olmstead
CC: “Fracturing Bypassed Pay in Tubingless
Completions,” paper SPE 56467, presented at the SPE
Annual Technical Conference and Exhibition, Houston,
Texas, USA, October 3-6, 1999.
Gulrajani SN and Olmstead CC: “Coiled Tubing Conveyed
Fracture Treatments: Evolution, Methodology and Field
Application,” paper SPE 57432, presented at the SPE
Eastern Regional Meeting, Charleston, West Virginia,
USA, October 20-22, 1999.
8. Chase B, Chmilowski W, Marcinew R, Mitchell C, Dang Y,
Krauss K, Nelson E, Lantz T, Parham C and Plummer J:
“Clear Fracturing Fluids for Increased Well Productivity,”
Oilfield Review 9, no. 3 (Autumn 1997): 20-33.
9. A screenout is caused by proppant bridging in the fracture, which halts fluid entry and fracture propagation. If
a screenout occurs early in a treatment, pumping pressure may become too high and the job may be terminated before an optimal fracture can be created.
69
50387schD10R1.p70.ps 01/10/2002 03:59 PM Page 70
In the past, the accuracy of standard coiled
tubing depth measurements was about 30 ft
[9.1 m] per 10,000 ft [3048 m] under the best conditions and as much as 200 ft [61 m] per 10,000
ft in the worst cases. The dual-wheel UTLM surface measurement is self-aligning on the coiled
tubing, minimizes slippage, offers improved wear
resistance and measures unstretched pipe
(below).10 Two measuring wheels constructed of
wear-resistant materials, on-site data processing
and routine calibration eliminate the effects of
wheel wear on surface measurement repeatability and provide automatic redundancy in addition
to slippage detection.
The remaining factors that affect measurement accuracy and reliability are contaminants
and buildup on wheel surfaces, and thermal
effects that change wheel dimensions. An antibuildup system prevents contamination of wheel
surfaces. Downhole coiled tubing pipe deformation is evaluated using computer simulation.
For thermal pipe deformation modeling, a wellbore simulator provides a temperature profile.
The total deformation can be estimated with an
accuracy of about 5 ft [1.5 m] per 10,000 ft. The
combination of more accurate surface measurements with modeling and improved operational
procedures result in about a 11 ft [3.4 m] per
10,000 ft accuracy, and a repeatability of about
4 ft [1.2 m]. In most cases, a value of less than 2 ft
[0.6 m] is achieved.
> The UTLM dual-wheel surface depthmeasurement device.
70
> Hiawatha field producing horizons. In the Hiawatha field of northwest
Colorado (insert), pay zones historically were grouped in intervals, or stages,
of 150 to 200 ft [46 to 61 m] and stimulated with a single fracture treatment.
Thin sands were grouped with thick sands, and occasionally thin sands
were bypassed to avoid less effective stimulation of more prolific sands.
Multiple hydraulic fracture stages were still required to treat the entire
wellbore. Each fracture stage was isolated with a sand plug or mechanical
bridge plug. Justifying completion of thin sands capable of 100 to 200 Mcf/D
[2832 to 5663 m3/d] was difficult.
Oilfield Review
50387schD10R1.71.ps 12/06/2001 01:47 AM Page 71
Previously, depth correction with wireline
inside coiled tubing or memory gamma ray logging tools, “flags” painted directly on the coiled
tubing and mechanical casing-collar locators
often were inaccurate, costly and time-consuming. Schlumberger now uses a wireless
DepthLOG tool, which detects magnetic variations at joint casing collars as tools are run into a
well and sends a signal to surface through
changes in hydraulic pressure. Subsurface
depths are determined quickly and accurately by
comparison with baseline gamma ray correlation
logs. The use of wireless technology decreases
the number of coiled tubing trips into a well and
saves up to 12 hours per operation on typical
coiled tubing-conveyed perforating and stimulation operations.
In the past, separate coiled tubing services, if
required, followed fracturing operations to clean
out excess proppant. Coiled tubing-conveyed
fracturing, however, requires the combined
efforts of fracturing and coiled tubing personnel.
Initially, service crews faced a steep learning
curve as they began working together to reduce
the time required for various operations.
Subsequent CoilFRAC projects increased operational efficiency and reduced completion time. To
further increase efficiency, Schlumberger has
formed dedicated CoilFRAC teams to integrate
coiled tubing and fracturing expertise.
Revitalizing a Mature Field
Texaco Exploration and Production Inc. (TEPI),
now a ChevronTexaco company, extended
the productive life of West Hiawatha field in
Moffat county, Colorado, USA, with CoilFRAC
techniques.11 Discovered in the 1930s, this field
has 18 pay sands over 3500 ft [1067 m] of
gross interval. Gas production comes from
the Wasatch, Fort Union, Fox Hills, Lewis and
Mesaverde formations (previous page, right).
Previously, wells were completed with 41⁄2-, 5- or
7-in. casing and stimulated using conventional
staged fracturing treatments.
A common practice was to stimulate zones
from the bottom upward until production rates
were satisfactory. As a result, thin zones often
were ignored and undeveloped uphole potential
existed throughout the field. In 1999, TEPI evaluated bypassed pay in the field to identify and rank
workover potential based on reservoir quality,
cement integrity, completion age and wellbore
integrity. New drilling locations were identified
after a successful workover on Duncan Unit 1
Well 3, but the challenge was to develop a strategy that could effectively stimulate all of the pay
zones during initial completion operations.
Autumn 2001
> Evaluating single-stage Hiawatha field fracture stimulations. Without selective isolation of individual
sands, variations in fracture gradients make it difficult to optimize fracture lengths with a single conventional treatment and limited-entry perforating. For two Wasatch zones that would be grouped when
stimulating multiple intervals with a single treatment, StimCADE hydraulic fracturing simulator plots
indicate that about two-thirds of the proppant is placed in the upper interval (top). This results in a
wider, more conductive fracture and a half-length almost 50% greater than in the lower interval
(bottom). If there are more than two zones, this problem is further compounded by variations in discontinuous sands from wellbore to wellbore.
The operator chose CoilFRAC services to
selectively stimulate Wasatch and Fort Union
sands, which comprise multiple sands from 5 to
60 ft [1.5 to 18 m] thick from 2000 to 4000 ft [600
to 1200 m] deep. This approach provided flexibility to design optimal fracture treatments for each
zone rather than large jobs to intersect multiple
zones over longer intervals.
In the first drill well, individual CoilFRAC
treatments were performed on 13 zones in three
days. Seven zones were treated in a single day.
This well’s average first month production was
2.3 MMcf/D [65,900 m3/d]. The second drill well
involved eight treatments in one day. Average
production from the second well during the first
month was 2 MMcf/D. Treating pressures ranged
from 3200 psi [22 MPa] to the maximum allowable 7000 psi [48 MPa].
Zones separated by 10 to 15 ft [3 to 4.6 m]
were fractured with no communication between
stages. Pump-in tests verified that fracture gradients between zones varied from 0.73 to 1 psi/ft
[16.5 to 22.6 kPa/m]. The variation in fracture
gradient for each zone confirmed the difficulty of
stimulating multiple zones with conventional
stage treatments (above). In addition to eight
workovers with mixed success, nine successful
10. Pessin JL and Boyle BW: “Accuracy and Reliability of
Coiled Tubing Depth Measurement,” paper SPE 38422,
presented at the 2nd North American Coiled Tubing
Roundtable, Montgomery, Texas, USA, April 1-3, 1997.
11. DeWitt M, Peonio J, Hall S and Dickinson R:
“Revitalization of West Hiawatha Field Using CoiledTubing Technology,” paper SPE 71656, presented at the
SPE Annual Technical Conference and Exhibition, New
Orleans, Louisiana, USA, September 30-October 3, 2001.
71
50387schD10R1.72.ps 12/06/2001 01:47 AM Page 72
wells were drilled in Hiawatha field from May
2000 through July 2001. These new wells were
completed with CoilFRAC stimulations in the
Wasatch and Fort Union formations, and conventional fracture treatments for the more continuous Fox Hills, Lewis and Mesaverde intervals
below 4000 ft [1220 m].
To quantify coiled tubing stimulation results,
the CoilFRAC completions were compared with
wells fractured conventionally between 1992 and
1996 (right). Average production from CoilFRAC
completions increased 787 Mcf/D [22,500 m3/d],
or 114%, above historical rates. However, production from individual wells may be misleading
if reserves are drained from offset wells. Field
output will not increase as expected when there
is interference between wells; natural pressure
depletion should result in new wells producing
less, not more.
From 1993 to 1996, Hiawatha field output
increased from 7 to 16 MMcf/D [200,500 to
460,000 m3/d] as a result of the 12-well drilling
program. Production doubled again from 11 to
22 MMcf/D [315,000 to 630,000 m3/d] as a result
of workovers and new wells completed mostly
with coiled tubing-conveyed stimulations. Field
production is at the highest level in 80 years.
Stimulating each zone individually during initial
completion operations is believed to be the key
to improving production and increasing reserve
recovery in this mature field.
State-of-the-Art Downhole Tools
Isolation tools have evolved along with CoilFRAC
treatments and specific requirements generated
by various stimulation applications. Coiled tubingconveyed fracturing operations are performed
under the most dynamic reservoir stimulation
conditions. Treatments take place in live wells at
formation temperatures and pressures, and with
the completion of each selective stimulation,
these conditions change. As a result, increasingly
demanding applications in deeper wells require
more reliable, multiple-set isolation tools.
Driven by a need to minimize operational and
financial risks and reduce the impact of
unplanned events, like proppant screenout,
Schlumberger developed the CoilFRAC Mojave
line of downhole tools (next page). This improved
straddle system consists of three technologies—
the pressure-balanced disconnect, the modular
straddle assembly with ported sub, and the slurry
dump valve. In combination, these components
provide selective placement of sequential acid or
proppant fracture stimulations, and matrix acid,
72
> Analyzing Hiawatha field coiled tubing fracturing results. Production from
wells completed with CoilFRAC selective isolation and simulation treatments
(red) was compared with production from wells that were previously fractured conventionally (black). Average daily well rates for each month was
normalized to time zero and plotted for the first six months. Initial production
from the CoilFRAC completions was about 787 Mcf/D [22,500 m3/d], or 114%,
more than historical rates.
screenless sand-control or scale-inhibitor treatments in a single trip with coiled tubing.
The pressure-balanced disconnect features a
mechanical shear disconnect that is pressurebalanced to coiled tubing treating pressure. Only
mechanical coiled tubing loads are transferred to
the shear-release pins; treating pressure does
not affect the shear-pin release function. This
reduces the likelihood of leaving the tool in a
well as a result of unexpectedly high downhole
treating pressures during CoilFRAC stimulations,
such as a screenout. The pressure-balanced disconnect allows coiled tubing to be run deep
because the disconnect does not require extra
shear pins to account for pressure loads during
treatments. If the tool becomes stuck, it can be
fished by overshot or internal fishing neck.
The CoilFRAC Mojave isolation tool has
opposing elastomer cups for 41⁄2- to 7-in. casing.
The tool functions in vertical or horizontal wells
and has no mechanical slips and no moving parts.
An internal fluid bypass in the tool body permits
running to deeper depth—10,000 ft instead of
less than 4000 ft. This feature lightens coiled
tubing loads during trips in and out of wells to
reduce elastomer wear, minimize swab and surge
forces on formations and decrease the risk of a
tool sticking between zones. A modular design
and special 2-ft [0.6-m] ported fracturing sub
allow 4-ft sections to be assembled for spacing
elastomer cups up to 30 ft apart.
The CoilFRAC fracturing sub also includes a
fluid bypass and resists erosion when pumping
up to 300,000 lbm [136,100 kg] of sand. It is possible to pump up to 500,000 lbm [226,800 kg] of
less erosive resin-coated and man-made
ceramic proppants. Reverse circulation is
required to clean the coiled tubing and CoilFRAC
Mojave isolation tool when run without a slurry
dump valve. A lower reversed bottom cup
seals during reverse circulation to improve
post-treatment cleanup. A gauge port is built
into the tool for downhole pressure and temperature measurements.
Since the slurry dump valve (SDV) is flowoperated, no coiled tubing movement is required.
One SDV design in two sizes is compatible with
standard 41⁄2- to 7-in. CoilFRAC Mojave tools
and functions in vertical or horizontal wells.
Incorporating a SDV allows slurry to be dumped
from the coiled tubing between zones and facilitates stimulations in low-pressure reservoirs and
formations with fracture gradients of less than a
full water gradient, or 0.4 psi/ft [9 kPa/m].
The SDV is closed and acts as a fill valve
when running in a well. It also reduces formation
damage during multizone well treatments.
Reverse circulation is not required for coiled tubing cleanup, which reduces total stimulation fluid
requirements, eliminates the environmental
impact of slurry returned to surface, reduces
elastomer wear by equalizing pressure across
elastomer seal cups, and reduces abrasive wear
on coiled tubing and surface equipment.
Oilfield Review
50387schD10R1
12/06/2001
02:15 AM
Page 73
> CoilFRAC Mojave isolation tools. From single mechanical packers to elastomer cup and packer combinations and the earliest versions of opposing
elastomer-cup straddle tools, the suite of CoilFRAC tools has expanded to
include specially designed straddle assemblies. The effectiveness of CoilFRAC
straddle assemblies for zonal isolation has been aided by more reliable sealing technologies. An annular flow path within the assembly allows for easy
deployment and retrieval.
12. Al-Qarni AO, Ault B, Heckman R, McClure S, Denoo S,
Rowe W, Fairhurst D, Kaiser B, Logan D, McNally AC,
Norville MA, Seim MR and Ramsey L: “From Reservoir
Specifics to Stimulation Solutions,” Oilfield Review 12,
no. 4 (Winter 2000/2001): 42-60.
13. NODAL analysis couples the capability of a reservoir to
produce fluids into a wellbore with tubular capacity to
conduct flow to surface. The technique name reflects
discrete locations—nodes—where independent equations describe inflow and outflow by relating pressure
Autumn 2001
losses and fluid rates from outer reservoir boundaries
across the completion face, up production tubing and
through surface facility piping to stock tanks. This
method allows calculation of rates that wells are capable of delivering and helps determine the effects of damage, or skin, perforations, stimulations, wellhead or
separator pressure and tubular or choke sizes. Future
production also can be estimated based on anticipated
reservoir and well parameters.
Optimizing Recovery in South Texas
Samedan Oil Corporation operates North Rincon
field in south Texas, producing gas from various
zones of the Vicksburg formation at 6000 to
7000 ft [1800 to 2100 m]. The Martinez B54 well,
completed in a single 25-ft [7.6-m] zone, had an
initial production rate of 4.5 MMcf/D before
declining to 1 MMcf/D. In December 2000,
Samedan evaluated fracturing this zone for the
first time as well as completing deeper pay in the
Martinez B54 well. Openhole logs had identified
several other productive zones that had been
intentionally bypassed because of marginal economics. In February 2001, Schlumberger assembled a multidisciplinary team to integrate
petrophysical and reservoir knowledge with
completion design, execution and evaluation
services using the PowerSTIM stimulation optimization initiative.12
Samedan and the PowerSTIM team analyzed
well data to determine reservoir size and remaining reserves for the current producing zone.
These calculations indicated a 19-acre [7700-m2]
drainage area and confirmed that a nearby geologic unconformity acted as a seal. Production
and NODAL analyses matched the 1-MMcf/D
production and indicated that, based on a limited
drainage area and low formation damage,
remaining reserves could be recovered in a
few months.13 This interval was not a candidate
for stimulation.
Samedan decided to deplete the existing
zone before completing the most attractive
bypassed zones. Reinterpreted logs indicated
77 ft [23 m] of high-quality net pay with significant recoverable reserves in five deeper zones
over 700 ft [213 m] of gross interval.
Conventional stimulation techniques required
limited-entry perforating for diversion of large
fluid and proppant volumes pumped at high rates
to cover and fracture this entire interval.
The operator considered setting production
tubing and a packer below existing perforations
and completing only one or two of the uppermost
bypassed zones. This approach, however, would
leave a significant volume of additional reserves
untapped behind pipe. The PowerSTIM team recommended CoilFRAC selective isolation services
with optimized fracture designs to complete and
individually stimulate all five bypassed zones. A
2-in. coiled tubing string was selected to convey
fracturing fluids and proppant at the required
rates. An SCMT Slim Cement Mapping Tool log
confirmed cement integrity and adequate zonal
isolation behind pipe across the proposed
completion intervals. The existing perforations
were sealed with a cement squeeze prior to
CoilFRAC operations.
73
50387schD10R1.p74.ps 01/10/2002 03:44 PM Page 74
< Martinez B54 well CoilFRAC treatment stimulation results for five zones.
In May 2001, Samedan and Schlumberger
performed a five-stage CoilFRAC selective
stimulation (next page, top). On the first day, the
five zones were perforated with deep-penetrating PowerJet premium charges to maximize
perforation entry-hole size and reservoir penetration. After perforating, the commingled zones
produced 1.1 MMcf/D [31,500 m3/d] during a
prestimulation test.
On the second day, each zone was isolated
sequentially with a 5-in. CoilFRAC Mojave
straddle tool and fracture-stimulated with a nondamaging ClearFRAC fluid and 136,000 lbm
[61,700 kg] of man-made ceramic proppant. All
five zones were treated within a 24-hour period.
Pump rates ranged from 8 to 10 bbl/min [1.3 to
1.6 m3/min] with treating pressures up to
11,000 psi [76 MPa]. Because of potentially high
gas production rates, PropNET fiber additives
were incorporated at the end of the pumping
schedules to prevent proppant flowback.14
When all the zones were commingled and
tested, the well flowed more than 5.1 MMcf/D
[146,000 m3/d] and 120 B/D [19 m3/d] of condensate, which closely matched production predictions. A production log spinner survey indicated
that four of the five Vicksburg zones had been
stimulated successfully (above and left). One month
later, the well was still producing about 5 Mcf/D,
which did not follow the expected decline.
Estimated payout was three months. Samedan
engineers evaluated the next three drill wells, but
none of these new wells were viable candidates
for coiled tubing-conveyed fracture stimulation.
Completing five zones in a single trip mitigated the risk of formation damage from multiple
well interventions, and risk of fluid swabbing
associated with conventional fracturing operations, jointed tubing and standard downhole
tools. This CoilFRAC treatment took only two
days, while a conventional five-stage fracturing
job might have taken up to two weeks.
74
Oilfield Review
50387schD10R1.p75.ps 12/7/01 8:52 PM Page 75
Accurate CoilFRAC selective placement allows
scale inhibitors to be conveyed deeper into the formation during fracturing or acidizing stimulation
treatments. Integrating scale inhibitors and stimulation treatment fluids into a single step ensures
that the entire productive interval—including the
proppant pack—is treated.
Performing multiple, smaller fracture treatments is another approach to reduce scale
buildup and sand production. This method
reduces the pressure drop across the formation
face, which decreases or, in some cases, prevents scale and asphaltene formation. During
production, pressure drawdown increases the
vertical stress on producing intervals and exacerbates sand production. An alternative is to treat
smaller intervals and reduce the pressure drop
across the formation face.
> Martinez B54 well in the North Rincon field, south Texas (Courtesy of Samedan Oil Corporation).
> Unconventional coiled tubing-conveyed treatments. CoilFRAC treatments also are applicable for
chemical scale inhibition and sand-control methods. Coiled tubing places scale inhibitors included in a
preflush before fracturing or proppant impregnated with scale inhibitors more effectively than conventional treatment techniques (left). Novel screenless completions provide sand control without downhole mechanical screens and gravel packing by using technology like resin-coated proppants and
PropNET fibers to control proppant flowback and sand production (right). The primary challenge of
applying these techniques is ensuring coverage of all perforated pay zones.
Additional Applications
The combination of reservoir-stimulation and
well-treatment technologies with coiled tubing
conveyance is expanding selective CoilFRAC
techniques to include applications, like acid fracturing, and specialized completion techniques
such as scale inhibition, controlling proppant
flowback and screenless sand control (above).
With advances in friction-reducing fluids,
injection rates are sufficient for coiled tubing
and CoilFRAC tools to be used as mechanical
Autumn 2001
diversion during acid fracturing. This capability is
increasingly important in mature carbonate
reservoirs when small zones within larger producing intervals require stimulation. CoilFRAC
stimulations help operators deplete reserves uniformly across an entire hydrocarbon-bearing
interval and facilitate reservoir management.
The downhole buildup of scales, asphaltenes
or migrating fines and the plugging of perforations
and completion equipment impair permeability
and can restrict or prevent production altogether.
Screenless Sand-Control Completions
Innovative screenless completions provide sand
control without the need for downhole mechanical screens and gravel packing by using technologies such as resin-coated proppants and
PropNET fibers to control proppant flowback and
sand production. The primary challenge of applying screenless technology is ensuring coverage
of all perforated pay zones. In general, interval
length is the controlling factor. Thicker intervals
typically reduce treatment success rates. Coiled
tubing-conveyed fracturing, with the capability of
treating numerous zones, increases screenless
completion effectiveness and reduces overall
costs while increasing net pay potential.
Treatments in North America have reduced proppant flowback by five-fold.
PT. Caltex Pacific Indonesia, a ChevronTexaco
affiliate, operates the Duri field in the Central
Sumatra basin.15 Primary recovery is low, so
steam injection is used to achieve higher recovery factors. This multibillion-barrel steamflood covers 35,000 acres [14 million m2] and produces
280,000 B/D [44,500 m3/d] of high-viscosity
crude oil. Oil-bearing sands are highly unconsolidated, Miocene-age formations with permeability
14. Armstrong K, Card R, Navarrete R, Nelson E, Nimerick K,
Samuelson M, Collins J, Dumont G, Priaro M, Wasylycia N
and Slusher G: “Advanced Fracturing Fluids Improve
Well Economics,” Oilfield Review 7, no. 3 (Autumn 1995):
34-51.
15. Kesumah S, Lee W and Marmin N: “Startup of Screenless
Sand Control Coiled Tubing Fracturing in Shallow,
Unconsolidated Steamflooded Reservoir,” paper SPE
74848, prepared for presentation at the SPE/ICOTA
Coiled Tubing Conference and Exhibition, Houston,
Texas, USA, April 9-10, 2002.
75
50387schD10R1.p76.ps 12/7/01 8:52 PM Page 76
as high as 4000 mD (right). Combined pay thickness is about 140 ft [43 m] over an interval from
X430 to X700 ft. In addition to 3600 producing
wells, the operator maintains about 1600 steaminjection and temperature-observation wells.
Heat requirements are lower in temperaturemature areas where the steamflood has been in
operation for an extended period of time. Steam
injection can be reduced, allowing the operator
to convert injectors and observation wells into
producers. Low reservoir pressure causes
drilling, completion and production problems
including lost circulation, hole collapse and sand
production. Severe sanding leads to frequent
well servicing to replace worn or stuck artificiallift equipment. The marginal nature of these
wells, initially completed with 4-, 7-, or 95⁄8-in. OD
monobore casing, limits conventional gravelpacked screens for sand control. In most wells,
screens are not installed because of restricted
wellbore access, smaller pump sizes and, consequently, unfavorable production rates.
In a recent field test on several wells, the
operator in Duri field used CoilFRAC techniques
to perform screenless completions using curable
resin-coated sand and tip-screenout fracture
designs to prevent proppant flowback and migration of formation grains.16 After resin-coated sand
is placed and cured, proppant packs are locked
in place to create a stable filter against the
formation in perforation tunnels and nearwellbore regions.
Using resin-coated proppant to control sand
without mechanical screens is not new. In 1995, a
Duri field pilot project used conventional fracturing with resin-coated sand to complete Rindu
sands at about X450 ft. Single-stage tip-screenout treatments attempted to place resin-coated
proppant in multiple zones across 50 to 100 ft [15
to 30 m] of gross interval. This technique failed to
achieve acceptable results because the gross
intervals were too long and not all perforations
received resin-coated sand. In addition, produced
formation sand covered some lower zones and
steam injection did not cure the resin-coated sand
across the entire section.
The primary objectives of the most recent
field test were to ensure complete treatment
coverage of all perforations and achieve tipscreenout fractures for proper proppant packing.
Grain-to-grain contact and closure stress improve
the curing process and ensure a strong compacted filter medium. Heat or alcohol-base fluids
cure phenolic resins. The operator uses both
methods to ensure a complete resin set.
CoilFRAC selective isolation and treatment
placement provided accurate and complete perforation coverage, which made screenless
completions a viable alternative to gravel
76
> Duri field, Indonesia, producing horizons and typical well completion.
packing or frac packing with screens, and
previous screenless completions that were
attempted conventionally.
Fracture treatments and pumping schedules
were designed to achieve required fracture halflength and conductivity. Relatively low pumping
rates control vertical coverage, while higher
proppant concentrations are needed to ensure
fracture conductivity and achieve tip screenout.
The maximum rate is usually about 6 bbl/min
[1 m3/min] with proppant concentrations of
8 pounds of proppant added (ppa). The number of
treatment stages in a given well was determined
by evaluating perforated interval length and
spacing between zones.
Interval length needed to be less than 25 ft to
ensure complete coverage with a minimum of 7 ft
[2 m] between intervals to allow the isolation
tool to set properly. The operator verified cement
bond and quality to ensure isolation behind the
pipe and avoid proppant channeling. Extra resincoated sand deposited after each treatment isolated that interval from subsequent treatment
intervals. After all zones were treated, the oper-
ator left the well undisturbed for about 12 hours
to allow the resin to set and obtain adequate
strength. Partially cured resin-coated sand in the
wellbore was drilled out prior to production.
With the exception of one well, screenless
completions significantly increased cumulative
oil production during nine months of evaluation
(next page, left). Average failure frequency
before CoilFRAC screenless completions was 0.5
per well per month. The operator allocated 36 rig
days and 32,000 bbl [5080 m3] of deferred oil production for all four wells to clean out sand. After
CoilFRAC screenless treatments were performed,
failure frequency dropped to 0.14 per well per
month, resulting in an extra five months of oil
production per well per year. Screenless
16. In standard fracturing, the fracture tip is the final area
to be packed with proppant. A tip-screenout design
causes proppant to pack, or bridge, near the end of the
fractures in early stages of a treatment. As additional
proppant-laden fluid is pumped, the fractures can no
longer propagate deeper into a formation and begin to
widen or balloon. This technique creates a wider, more
conductive pathway as proppant is packed back toward
the wellbore.
Oilfield Review
50387schD10R1.p77.ps 12/7/01 8:53 PM Page 77
> Ongoing CoilFRAC operations in Medicine Hat,
Alberta, Canada.
> CoilFRAC screenless completion results in Duri field, Indonesia.
CoilFRAC treatments paid out in 35 to 59 days.
However, the use of resin-coated sand in
extremely hot steamflood conditions was found
to have limitations.
Early in the application of screenless completions, the operator recognized a need to use inert
proppant flowback control. The resin coating used
initially in CoilFRAC screenless completions was
thermally stable to 375°F [191°C], but could fail in
steam environments of 400°F [204°C]. As a result,
periodic steam injection and flowback to stimulate oil output could cause stress cycling and
proppant-pack failure that resulted in sand production. Proppant flowback control using PropNET
fibers rated to 450°F [232°C] is proving to be a
solution to this problem.
The operator selected a local sand combined
with PropNET fibers in place of resin-coated sand
for eight recent screenless completions in Duri
field. The PropNET fibers were added throughout
Autumn 2001
sand-laden treatment stages to ensure complete
interval coverage. Optimized perforating techniques also has been introduced for screenless
sand control. These wells have minimal production
data, but early production results are encouraging.
Milestones in Selective Stimulations
Selective coiled tubing-conveyed isolation and
stimulation have established a template for
future workovers on existing wells and new well
completions. The CoilFRAC methodology allows
controlled delivery and accurate placement of
treatment fluids and proppant in existing or
bypassed pay intervals at little or no additional
cost because decreased fluid volumes and elimination of redundant operations reduce mobilization, equipment and material charges.
CoilFRAC treatments are useful for fracturing
bypassed single or multiple zones, protection of
casing and completion equipment, and for
development of coalbed methane reserves. This
technique is also valuable in settings where
chemical inhibition, reservoir flow-conformance
modifications, water-control or sand-control
methods may be required. Schlumberger has
pumped more than 12,000 CoilFRAC fracture
stimulations in more than 2000 wells. Coiled tubing-conveyed treatments can now be performed
in vertical, high-angle and horizontal wells with
measured vertical depths up to 12,200 ft [3720 m].
Pumping rates can range from 8 to 25 bbl/min
[1.3 to 4 m3/min] with 5 to 12 ppa of proppant.
Coiled tubing-conveyed fracturing was originally developed for multilayered shallow-gas
reservoirs in Canada and further developed in the
USA (above). These CoilFRAC treatments, however, are being refined in applications around the
world, from Indonesia, Argentina and Venezuela
to Mexico and now Algeria.
The largest total volume of proppant placed in
a single wellbore was 850,000 lbm [385,555 kg]
for a well treatment in northern Mexico. A well in
southeast New Mexico, USA, was the first horizontal well to be fracture stimulated using a
CoilFRAC Mojave tool. Two separate zones at
9123 and 9464 ft [2781 and 2885 m] measured
depth were treated. The deepest CoilFRAC job to
date was recently performed at 10,990 ft [3350 m]
for Sonatrach in Algeria. The progress to date in
selective stimulations has been impressive.
Continued research and field experience are
expected to further extend the range of applications
and reach of this innovative technique.
—MET
77
50387schD08R1.ps78.ps 12/5/01 6:57 AM Page 78
Contributors
R. John Andrews is currently a senior staff reservoir
engineer with Husky Energy Inc. in St. John’s,
Newfoundland, Canada. He provides reservoir engineering technical expertise to support development
planning in relation to Husky’s East Coast assets. His
key responsibilities include monitoring Terra Nova
field operations, reserves evaluation and management, and assistance in the development of a dataacquisition strategy for the White Rose field. Other
responsibilities involve detailed reservoir-fluid analysis, equation-of-state modeling and reservoir simulation. John spent his first eight years in Calgary,
Alberta, Canada, on reservoir engineering assignments involving conventional oil and gas, and heavyoil and oil-sands projects. After returning to
Newfoundland in 1989, John spent seven years as a
reservoir-conservation engineer with the CanadaNewfoundland Offshore Petroleum Board, and five
years as a senior reservoir engineer with Hibernia
Management and Development Company Ltd.
(HMDC) before accepting his current position with
Husky Energy Inc. in May 2001. A Committee member
for the SPE Atlantic Canada Section for six years,
John received a BS degree in mathematics from the
Memorial University of Newfoundland in St. John’s,
and a Bachelor of Industrial Engineering degree at
Technical University of Nova Scotia in Halifax, Canada.
Cosan Ayan, Account Manager and Principal
Reservoir Engineer for the UK, is based in Aberdeen,
Scotland, where he works on interpretation and development such as transient well tests, wireline formation testers, production logging and reservoir
monitoring. Previously, he held similar responsibilities as division reservoir engineer for Schlumberger
Central Gulf division based in Abu Dhabi, covering
UAE, Qatar, Iran and Yemen. Cosan was also division
reservoir engineer for Schlumberger East
Mediterranean division in Cairo, Egypt (1991 to 1993).
He joined the company in 1990 to work with
Schlumberger Reservoir Characterization Services in
Dubai, UAE. During this assignment, he worked on
geological modeling and developed scaling-up algorithms for reservoir-simulation grid blocks. Before
joining Schlumberger, he was an assistant professor
at the Middle East Technical University in Ankara,
Turkey. Cosan holds a BS degree from Middle East
Technical University, and MS and PhD degrees from
Texas A&M University in College Station, USA, all in
petroleum engineering. The author of papers on well
testing and reservoir engineering, he is currently a
technical editor for SPE Formation Evaluation.
Gary Beck received his BS degree in geology from
Hofstra University in Hempstead, New York, USA, and
his MS degree in geology from Purdue University, West
Lafayette, Indiana, USA. He then joined Chevron in
New Orleans, Louisiana, USA, where he worked in
development geology before moving into formation
78
evaluation in 1988. In 1997 Gary moved to Vastar
Resources in Houston, Texas, where he was principal
petrophysicist, Deepwater Special Projects. After BP
acquired Vastar in 2000, he became a staff petrophysicist for BP in the North American Exploration
Deepwater Appraisal Group in Houston, Texas. There
he is involved in all aspects of petrophysics with a
special interest in mineral-based log analysis, capillary pressure, formation sampling, measurementswhile-drilling and nuclear magnetic resonance. Gary
has written and presented numerous papers on various topics at SPWLA symposia and chapter meetings
and at SPE conferences, and was awarded the Best
Paper Award at the 1996 SPWLA Annual Symposium.
He is a past-president of the SPWLA and has served
multiple positions on the SPWLA Board of Directors
during the past seven years.
Melvin Cannell has been Director of the Centre for
Ecology and Hydrology (CEH) at Edinburgh, Scotland,
since 1987. CEH is a component of the UK Natural
Environment Research Council (NERC). He began his
career in 1966 as a research scientist, and worked for
the Coffee Research Foundation in Kenya, Africa. In
1971 he joined NERC as a research scientist at the
Institute of Tree Biology in Edinburgh. Three years
later he became a senior scientist at the NERC
Institute of Terrestrial Ecology in Edinburgh.
Professor Cannell holds BS, PhD and DSc degrees in
agricultural botany from University of Reading in
England. He is a Fellow of the Royal Society of
Edinburgh and a Fellow of the UK Institute of
Chartered Foresters.
Kees Castelijns manages the Schlumberger Data
Services Center in London, England. He joined
Schlumberger in 1977 as a wireline field engineer and
spent four years in Oman, Saudi Arabia, Iran, the
Philippines, Dubai, Yemen and Egypt. In 1982 he
became wireline location manager in Kirkuk, Iraq.
After Wireline sales and marketing assignments in
Oman, India, Malaysia, Norway and The Netherlands,
he became manager of the Data Services Center in
The Hague, The Netherlands. He transferred to the
Sugar Land Product Center as domain expert for the
development of a thin-bed evaluation program in
1994. From 1994 to 1997, he was section manager of
petrophysics, in charge of developing and sustaining
petrophysical interpretation products, such as
PrePlus*, ELAN* Elemental Log Analysis and
PetroViewPlus* software. Prior to his current assignment, he was Schlumberger interpretation development manager, responsible for interpretation support
and new technology introduction. Kees obtained an
engineering degree in applied physics from Eindhoven
Technical University in The Netherlands.
Andy Chen has been a Calgary, Alberta, Canada-based
reservoir engineer with Schlumberger of Canada since
1996. He earned a BS degree in 1983 and an MS
degree in 1986, both in petroleum engineering, from
East China Petroleum Institute, and a PhD degree in
mechanical engineering in 1995 from University of
Manitoba, Winnipeg, Canada.
Myrt E. Cribbs is a senior reservoir engineer for
Texaco Exploration in Bellaire, Texas. After receiving
his BS degree in petroleum engineering from
Mississippi State University, he joined Getty Oil in
New Orleans, Louisiana. He worked as a production
and reservoir engineer until the merger with Texaco
in 1984. For Texaco, he continued to work as a reservoir engineer on shelf and deepwater properties in
the Gulf and participated in several early DeepStar
subcommittees. He also had international experience
working on carbonates. For the last four years, he has
been Texaco Exploration's deepwater Gulf of Mexico
reservoir engineering specialist, responsible for datacollection plans and reservoir evaluation, while developing a keen interest in downhole fluid sampling and
well testing. Recently, he has been responsible for the
design and execution of several international deepwater well tests.
Finn H. Fadnes is principal research engineer at
Norsk Hydro Petroleum Research Centre, in Sandsli,
Bergen, Norway. He has been involved in supervising
pressure-volume-temperature and fluid characterization since 1987. Prior to this (1983 to 1987), he was a
research engineer and then manager of the Fluid
Properties department at Rogaland Research in
Stavanger, Norway. He has also been a visiting
research associate in chemical engineering at Rice
University in Houston, Texas. Finn obtained a BS
degree in chemical engineering and an MS degree in
physical chemistry from the University of Bergen.
Jim Filas, Well Testing Joint Industry Projects (JIP)
Coordinator at Schlumberger Reservoir Completions
Center in Rosharon, Texas, is responsible for coordinating various joint industry projects with clients
(including the development of zero-emission welltesting technology). He is also involved in technical
coordination between Schlumberger business segments, client business development, and contract and
license negotiation. He began his career in 1977 as a
project engineer for a manufacturing subsidiary of
Sonat Offshore Drilling, where he worked on design,
analysis and manufacturing management of oilfield
equipment and drilling rigs. In 1982 he became a
research associate for Getty Oil Exploration and
Production Research Center in Houston, Texas. Two
years later he moved to Texaco Central Offshore
Engineering in New Orleans, Louisiana, as an
advanced petroleum engineer. He joined
Schlumberger in 1984 as a senior development engineer for logging vehicle and hydraulic system design,
structural analysis and strain gauge testing. From
1992 to 1998, he was section manager for Wireline
Engineering units in Houston and Austin, Texas. Prior
to his current position, he was product champion for
fiber-optic sensing in Paris, France. Jim earned a
BS degree in engineering science at Louisiana
State University in Baton Rouge, and an MS degree
in mechanical engineering at University of Houston.
Oilfield Review
50387schD08R1.ps79.ps 12/5/01 6:57 AM Page 79
Byron Gale is currently a senior production and operations engineer for Tom Brown, Inc. in Denver,
Colorado, USA. His main responsibilities involve new
well completions, workovers and recompletions, and
production operations in the Paradox and Piceance
basins, in Colorado and Utah (USA). He joined ARCO
Oil and Gas Company in 1986 and spent the next
decade with them and with Vastar Resources, working
in operations and analytical engineering projects in
Bakersfield, California, USA, and in Midland and
Houston, Texas. Before joining Tom Brown in 1997, he
spent a year with WhitMar Exploration Company in
Denver. Byron has a BS degree in petroleum engineering from Montana College of Mineral Science and
Technology in Butte, USA.
Duane Gonzalez, a production engineer for Samedan
Oil Corporation in Houston, Texas, works in south and
west Texas. He joined Schlumberger Dowell in 1993 as
a field engineer in Laredo, Texas and moved to their
production enhancement group three years later. From
1996 to 1998, he was a DESC* Design and Evaluation
Services for Clients engineer in Midland, Texas, working with Mobil and Texaco. He performed the same
function for Mobil and Conoco in Houston from 1998 to
2000. Duane earned a BS degree in mechanical engineering from Texas A&M University in College Station.
Hafez Hafez, Senior Reservoir Engineer with Abu
Dhabi Oil Company for Onshore Oil Operations
(ADCO) in the United Arab Emirates, deals with reservoir modeling, performance and management.
Previously, he spent five years with the Gulf of Suez Oil
Company in Egypt as an operations and reservoir engineer involved in different aspects of reservoir engineering. Hafez received a BS degree from University of
Cairo in Egypt and has written several SPE papers on
waterflooding and permeability distribution.
Scott Hall, Team Leader, ChevronTexaco, is based in
Denver, Colorado, where he manages new drilling and
workover opportunities in Wyoming, USA. He joined
the company in 1981 as a field engineer. He became
production supervisor in 1984, and a production engineer in 1985. From 1986 to 1987, he was a reservoir
engineer, and then became assistant to the vice president of exploration (1988 to 1990). For the next two
years he was a drilling engineer before moving to production engineering (1993 to 1994). He spent four
years as an asset-team engineer, before assuming his
current position as team leader for Wyoming in 1999.
Scott holds a BS degree in civil engineering from
University of Colorado in Boulder. He served as an SPE
Distinguished Lecturer in 1997.
John Harries, Professor and Chair of Earth
Observation at Imperial College of Science,
Technology and Medicine in London, England, has
held his current position since 1994. As a teacher and
researcher, he heads the Space and Atmospheric
Physics research group. In 1972, after receiving a BS
degree (Hons) in physics from University of
Birmingham, England, and a PhD degree from King’s
College in London, he was appointed senior scientific
officer at the National Physical Laboratory (NPL).
Three years later he became principal scientific officer
and head of the Environmental Standards group at
NPL. In 1980 he was appointed senior principal scientific officer and head of the Remote Sounding division
at Appleton Laboratory. Four years later he became
Autumn 2001
deputy chief scientific officer and head of the
Geophysics and Radio division, Rutherford Appleton
Laboratory, becoming the laboratory’s associate director and head of the space science department in 1986.
Since 1985 he has been a member of the HALOE
International Science team, and since 1995 has been
principal Investigator for the Geostationary Earth
Radiation Budget (GERB) experiment. Author of many
books and papers, he has also served as president of
the International Radiation Commission (1992 to
1996), president of the Royal Meteorological Society
(1994 to 1996), and as a member of NERC Council and
chair of Earth Observation Science & Technology
Board (1995 to 1997).
Mohamed Hashem, Global Consultant and Staff
Petrophysical Engineer for Shell Deepwater Services,
is based in New Orleans, Louisiana. His projects span
the globe and involve advising on fluid sampling and
pressure testing for Shell's projects worldwide, with
more than 100 sampling jobs and eight years of MDT*
Modular Formation Dynamics Tester development
experience. He joined Shell in 1990, and worked five
years in exploration and production as a petrophysical
engineer for the Shelf Division. Following that, he
worked on Gulf of Mexico deepwater exploration,
development and production projects. He worked
extensively in the Garden Banks area of the Auger
basin, with three major developments and three discoveries. Previously, he worked for Schlumberger in
various Middle Eastern and European locations as well
as in California; he also taught formation evaluation at
University of Southern California in Los Angeles.
Author of numerous publications, he received the
SPWLA Best Paper Award in 1998. Mohamed earned a
BS degree in mechanical engineering from Ain Shams
University in Cairo, Egypt; an MS degree in petroleum
engineering from University of Southern California in
Los Angeles; and an engineer’s degree in petroleum
engineering from Stanford University in California.
Sharon Hurst, Senior Reservoir Project Development
Engineer, Bohai Commercial Group, Phillips China
Inc., is responsible for engineering support of exploration activities in Bohai Bay, China, including project
evaluation and economics, as well as cased-hole logging and well-testing design, supervision and analysis.
She joined Phillips in Houston, Texas, in 1987 as a
reservoir and production engineer in the Gulf Coast
and areas across the USA (1987 to 1992). From 1994 to
1997, she was a reservoir and operations engineer for
the eastern Gulf of Mexico. She then served two years
as company well-test specialist at Phillips Research
Center in Bartlesville, Oklahoma, USA. Prior to her
current position, she was an international exploration
engineer, based in Bartlesville (1999 to 2000). In addition to her other assignments, she has served as exploration and well-test engineer and supervisor in Alaska
(USA), the Gulf of Mexico, Venezuela and China (1992
to 2000). Sharon obtained a BS degree from the
University of Texas at Austin, and an MS degree from
the University of Houston, both in petroleum engineering.
Jamie Irvine-Fortescue, Norsk Hydro ASA Production
Technology Discipline Manager for Njord field, is based
in Bergen, Norway. There he is responsible for all production technology work including production optimization. He began his career with BP Exploration in
1984 and for the next eight years held various positions including petroleum engineer, field production
engineer, commissioning engineer and production
engineer in Scotland, England and Norway. Since 1993
he has been with Norsk Hydro as a completion technologist and production technologist in Oslo, Norway,
and as manager and advisor for well testing in Bergen.
Jamie received a degree in mechanical engineering
from Robert Gordon's Institute of Technology in
Aberdeen, Scotland, and a BS degree in petroleum
engineering from Imperial College in London,
England. Author of many papers, he has served as
membership chairman and director of the Bergen
Section of the SPE.
A. (Jamal) Jamaluddin, Fluid Analysis Business
Manager-Worldwide, works at Oilphase, a division of
Schlumberger in Houston, Texas. His main responsibility is developing the company’s reservoir-fluid analysis
business globally. He began his career as a research
scientist at Noranda Technology Centre in Montreal,
Quebec, Canada, in 1990. For the next six years he
served as project leader and then program leader on
projects related to oil and gas research and technology
development. Prior to assuming his current position in
1998, he was director of technical services at Hycal
Energy Research Laboratories in Calgary, Alberta,
Canada. Jamal earned a BS degree in petroleum engineering from King Fahad University of Petroleum and
Minerals, Dhahran, Saudi Arabia, and MS and PhD
degrees in chemical engineering from the University of
Calgary. He is a coinventor of five patented processes
related to petroleum production and optimization and
has coauthored more than 70 technical papers on various subjects.
Geoff Jenkins, Head of the Climate Prediction
Programme at the Hadley Centre for Climate
Prediction and Research in Berkshire, England, has
held his current position since 1995. Previously, he
held another post at the center and at the UK
Department of the Environment. Dr. Jenkins obtained
BS and PhD degrees in physics from University of
Southampton in England.
Fikri Kuchuk, Schlumberger Fellow, is chief reservoir
engineer for Schlumberger Oilfield Services in the
Middle East and Asia. Previously, he was senior scientist and program manager at Schlumberger-Doll
Research, Ridgefield, Connecticut, USA. From 1988 to
1994, he was a consulting professor in the Petroleum
Engineering department of Stanford University in
California. Before joining Schlumberger in 1982, he
worked on reservoir performance and management for
BP/Sohio Petroleum Company. He has an MS degree
from Technical University of Istanbul, and MS and PhD
degrees from Stanford University, all in petroleum
engineering. Fikri received the SPE 1994 Reservoir
Engineering, SPE 2000 Formation Evaluation, and SPE
2001 Regional Service Awards. In 1995, he was elected
to the Russian Academy of Natural Sciences and
received the Nobel Laureate Physicist Kapitsa Gold
Medal. In 1996, he was named SPE Distinguished
Member and received Henri G. Doll Award in 1997 and
1999. He is currently SPE International Director-atLarge, SPE Northern Emirates Section Director and a
member of many SPE committees. A prolific author, he
has been associate editor of Journal of Petroleum
Science and Engineering since 1994, Turkish Journal
of Oil and Gas since 1996, and editor of Middle East
Reservoir Review since 1996.
79
50387schD08R1.ps80.ps 12/5/01 6:57 AM Page 80
Andrew Kurkjian, Schlumberger Customer Needs
and Product Strategy Manager in Sugar Land, Texas,
assesses client needs to determine an appropriate
product development strategy. In 1982 he joined
Schlumberger-Doll Research in Ridgefield,
Connecticut, as a research scientist. There he was
principal inventor of the DSI* Dipole Shear Sonic
Imager tool. From 1988 to 1990, he was engineering
manager for cross-well seismic development at
Schlumberger Riboud Product Center in Clamart,
France. He then moved to Schlumberger Cambridge
Research in England where he headed borehole seismic research. Since 1993 he has been involved with
the MDT tool as principal authority on fluid sampling
and is also a coinventor of the CHDT* Cased Hole
Dynamics Tester tool. Andrew earned a BS degree in
electrical engineering from Catholic University in
Washington, DC, USA, and MS and PhD degrees, also
in electrical engineering, from Massachusetts
Institute of Technology in Cambridge, USA.
Larry W. Lake is a professor in the Department of
Petroleum and Geosystems Engineering at The
University of Texas (UT) at Austin. He holds BSE and
PhD degrees in chemical engineering from Arizona
State University in Tempe, USA, and Rice University in
Houston, respectively. Dr. Lake has published widely;
he is the author or coauthor of more than 100 technical papers, the editor of three bound volumes and
author or coauthor of two textbooks. He has been
teaching at UT for 22 years prior to which he worked
for the Shell Development Company in Houston,
Texas. He has served on the Board of Directors for the
Society of Petroleum Engineers (SPE) as well as on
several of its committees; he has also been an SPE
distinguished lecturer. Among his many honors and
awards are the Shell Distinguished Chair, 1996
Anthony F. Lucas Gold Medal of the SPE, 1998
Election to the National Academy of Engineers and
the 2000 IOR Pioneer Award from the SPE.
Jack Marsh, Vice President of Engineering and
Business for Olympia Energy Inc. in Calgary, Alberta,
Canada, has been with the company since 1994. He is
responsible for all facets of production and reservoir
engineering as well as business development, asset
management and evaluation. Previously, from 1976 to
1994, he worked for Imperial Oil (an Exxon affiliate)
in Calgary, in positions such as wellsite geological
technologist, production and drillstem testing technologist, business development engineer and field
production engineer. He earned a diploma in earth
sciences from the Northern Alberta Institute of
Technology in Edmonton, Alberta, Canada, and a BS
degree in chemical engineering from the University of
Calgary. A director of the Canadian Gas Processors
Association, Jack is also a registered member of the
Alberta Professional Engineers Geologist and
Geophysicist Association.
80
Oliver C. Mullins received a BS degree in biology
from Beloit College in Wisconsin, USA, and MS and
PhD degrees in chemistry from Carnegie-Mellon
University, Pittsburgh, Pennsylvania, USA. After holding a research position in chemistry at the University
of Chicago, Illinois, USA, and in physics at the
University of Virginia in Charlottesville, USA, he
joined Schlumberger-Doll Research (SDR),
Ridgefield, Connecticut in 1986. He is a principal contributor to the OFA* Optical Fluid Analyzer, the SAS
Spectral Analysis System, the LFA* Live Fluid
Analyzer and to other projects currently in field testing. Oliver is currently a principal research scientist,
manager of the MDT program at SDR and Flow
Assurance Theme champion. He has coauthored about
50 articles in refereed journals, is coholder of 14 US
patents and has coedited two books on asphaltenes.
Aubrey O’Callaghan, Principal Reservoir Engineer
with Schlumberger GeoQuest in Puerto La Cruz,
Venezuela, provides technical support for reservoir
studies. His current interests include dynamic reservoir characterization through numerical simulation
and well testing. He also maintains an interest in horizontal well evaluation and advances in production logging. Since joining Schlumberger in 1979 as a field
engineer in Norway, he has held many technical positions during his 22 years with the company. He has
managed the Schlumberger Training Center in Parma,
Italy. In Nigeria and later Algeria, he was in charge of
dynamic reservoir studies and reservoir simulation.
Aubrey obtained a BS degree in engineering science
and an MS degree in mathematics from The University
of Dublin, Trinity College, Ireland. He also holds an
MS degree in petroleum engineering from Heriot-Watt
University in Edinburgh, Scotland.
Martin Parry was appointed professor of
Environmental Science and director, Jackson
Environment Institute at the University of East Anglia
in Norwich, England in 1999. From 1995 to 1999, he
was professor of Environmental Management at
University College in London; professor of
Environmental Management, and director of the
Environmental Change unit at the University of
Oxford (1992 to 1995); and professor of
Environmental Management, University of
Birmingham, England (1990 to 1992). He received a
BA degree from University of Durham, England; an MS
degree from University of the West Indies; and a PhD
degree from University of Edinburgh in Scotland. He
received the Order of the British Empire (OBE) in
1998 for services to the environment and to climate
change. He was also awarded the World
Meteorological Organization's Gerbier-Mumm
International Award in 1993, and the Royal
Geographical Society's Cuthbert Peek Award in 1991,
both for contributions to research on climate change.
John Peffer, Reservoir Manager, Groupement Berkine
(Sonatrach/Anadarko Association), is based in Hassi
Messaoud, Algeria. Since joining Anadarko in 1985, he
has held various reservoir engineering positions with
the company in Midland, Texas (1985 to 1989, and
1994 to 1996); Algiers, Algeria (1990 to 1993); and
London, England (1997 to 1998). He has been based
in Hassi Messaoud since 1999 in a management role.
John earned BS and MS degrees in petroleum engineering at University of Texas at Austin.
Julian Pop, an engineering advisor with
Schlumberger Oilfield Services in Sugar Land, Texas,
is involved in algorithm development for the MDT tool
and design and specification of wireline formation
tester interpretation-software products. Since joining
the company in 1979, he has had technical and managerial involvement in interpretation development
projects for completion and formation testers and
management of tool and interpretation software. He
also has taught at University of Texas at Austin and at
Rice University in Houston. Julian holds a BS degree
in mechanical engineering from the University of
Melbourne, Victoria, Australia, and an MS degree
from the Johns Hopkins University, Baltimore,
Maryland, USA, and a PhD degree from Rice University.
Paul Rutter, BP Group Senior Advisor on
Environmental Technology at the BP Technology
Centre in Sunbury on Thames, Middlesex, England,
has held his current post since 2000. He maintains
strong links with Imperial College in London, and
Princeton University in New Jersey, USA, and has
been involved in a number of UK government research
committees and advisory panels. He received a BS
degree (Hons) in chemistry and a PhD degree at
University of Leeds in England in 1972. After that he
worked mainly in industrial research on various projects centered on physical chemistry: toiletries development with Boots, oral microbiology with Unilever,
and biocompatible materials as a research fellow at
the London Hospital Medical School. He joined BP in
1981 to develop alternative fuels using coal. He then
worked in minerals processing and became manager
of the BP Mineral Processing R&D group in 1987. In
1990 he moved to BP Exploration technology as manager of the Production Operations branch. He started
BP’s produced water network in 1992. In 1998 he combined the group’s environmental technology programs
into “Green Operations.” This technology network and
research program covers the three BP group strategic
areas of climate change, water and biodiversity, as
well as technology programs specific to the individual
business streams. The network now has over 1200
active members throughout the company’s global
operations.
Erik Rylander, MDT Field Service Manager,
Schlumberger Gulf Coast Special Services, is based in
Belle Chasse, Louisiana. He joined the company in
1995 as a junior field engineer in Duncan, Oklahoma,
and then moved to Equatorial Guinea and Nigeria as a
field engineer (1996 to 1997). He spent the next four
years as a MDT specialist field engineer with Gulf
Coast Special Services before taking his current position in 2001. Erik holds a BS degree in engineering
with an electrical specialty from Colorado School of
Mines in Golden.
Oilfield Review
50387schD08R1.ps81.ps 12/5/01 6:57 AM Page 81
Bill Sass has been a software engineer at the
Schlumberger Sugar Land Product Center in Texas
since 1995. He has worked on wellsite MDT interpretation software and is responsible for the development of
the OCM* Oil-Base Contamination Monitor product.
He joined Schlumberger as a field engineer in 1981,
after receiving a BS degree in mechanical engineering
from the University of Western Ontario in London,
Ontario, Canada in 1981.
Lars Sonneland is research director at Schlumberger
Stavanger Research in Norway, where the focus is on
geophysical reservoir characterization and monitoring.
After receiving a degree in mathematics, computer science and physics and a PhD degree in applied mathematics from the University of Bergen, Norway, he
joined GECO in 1974. He had various technical assignments in geophysical applications until 1989 when he
transferred to Schlumberger-Doll Research,
Ridgefield, Connecticut. From 1990 to 1998, he had
several technical management positions within
Schlumberger. He transferred to Schlumberger
Cambridge Research (1999 to 2000). At the same time,
he helped establish Schlumberger Stavanger
Research. Lars has published more than 70 scientific
papers and holds a number of patents. Recipient of the
Norwegian Association of Chartered Engineers'
Technical Award and the Norwegian Geophysical
Award, Lars has played a major role in the development of 3D seismic technology, the Charisma* seismic
interpretation software system and seismic reservoir
characterization and monitoring.
Alexandra Van Dusen is currently pursuing a PhD
degree in geochemical oceanography in the
Department of Earth and Planetary Science at Harvard
University, Cambridge, Massachusetts. Prior to this,
from 1997 to 2000, she worked for Schlumberger
Oilfield Services as a wireline logging engineer first in
Bakersfield, California, and then in Bergen, Norway.
She is a graduate of Princeton University, New Jersey,
with a BA degree in geological sciences.
Jeremy Walker, Schlumberger Well Completions and
Productivity, Testing & Completions Marketing
Manager, is based in Houston, Texas. There he has
been responsible for development of the marketing
plan and strategy for testing services since 1999. He
began his career in 1980 as a field engineer with
Flopetrol International in The Netherlands. From 1982
to 1984, he was field service manager for well testing
in Al-Khobar, Saudi Arabia. Subsequent assignments
included location manager for well testing in
Aberdeen, Scotland; sales engineer for well testing in
West Africa; staff technical engineer, testing and production services for Africa and the Mediterranean
region; district manager, Schlumberger Wireline &
Testing in Hassi Messaoud, Algeria and in Port
Harcourt, Nigeria; business manager for production
services in Paris, France; district manager, testing in
Aberdeen, Scotland; and business development manager for testing in Aberdeen. Jeremy earned a BS
degree (Hons) in mechanical engineering from the
City University of London, England.
Autumn 2001
Stephen Williams works for Norsk Hydro ASA in
Bergen, Norway, as technical advisor for logging. He is
responsible for planning, execution and follow-up of
formation evaluation programs on Norsk Hydro wells
as well as related contracts. He has held this position
since he joined the company in 1998. Prior to this, he
spent 14 years with Schlumberger in various assignments in operations, technical management, training,
and management in North and South America, Europe,
Scandinavia and the Middle East. Stephen earned BA
and MA degrees in natural sciences from University of
Cambridge in England.
Warren Zemlak earned associate degrees from
Robertson and Kelsey Institutes in Saskatchewan,
Canada. He began his oilfield experience with a major
drilling contractor prior to joining Schlumberger in
1989. His career has included both field and technical
assignments throughout Canada in well cementing,
stimulation and coiled tubing. He was project leader in
several of the first directional underbalanced coiled
tubing drilling applications and was a member of the
team that installed the first high-pressure coiled tubing offshore the east coast of Canada. In 1996 he pioneered the first application of multizone fracturing
through coiled tubing. Currently based in Sugar Land,
Texas, Warren is CoilFRAC* business development
manager, responsible for the worldwide implementation and development of multizone stimulation
techniques. The author of several SPE papers, he
holds patents specific to multizone stimulation and
isolation tools.
Murat Zeybek, Senior Reservoir Interpretation and
Development Engineer for Schlumberger Oilfield
Services in Saudi Arabia, Bahrain and Kuwait, works
on interpretation of wireline formation testers, pressure-transient analysis, numerical modeling, water
control, production logging and reservoir monitoring.
Before this, he was the Schlumberger district reservoir
engineer in Doha, Qatar. He was a research associate
in the Petroleum Engineering department of the
University of Southern California in Los Angeles from
1991 to 1992 and also worked for Intera West
Consulting in California. Before joining Schlumberger,
he worked as an assistant professor at Technical
University of Istanbul in Turkey. He served as a committee member for 1999-2001 SPE Annual Technical
Conference and Exhibition and has written many
papers about fluid flow through porous media and
pressure-transient analysis. Murat holds a BS degree
from Technical University of Istanbul, and MS and PhD
degrees from University of Southern California, all in
petroleum engineering.
An asterisk (*) is used to denote a mark of Schlumberger.
81
50387schD09R1.p82.ps.ps1.ps 12/5/01 7:12 AM Page 1
Coming in Oilfield Review
NEW BOOKS
Advances in Borehole Imaging.
Operating environments for E&P
companies have become more
demanding. Oil-base and syntheticbase muds have addressed many of
the challenges endemic to these
areas. But because these muds are
nonconductive, borehole-imaging
options are limited. A new tool that
combines innovative technology
with the time-honored principle
of resistivity logging provides
microresistivity images in these
difficult environments.
Well and Platform Abandonment.
As abandonment of aging wells
and fields becomes more frequent,
responsible operators must balance
environmental and financial objectives. Remediation of deficient
plugging and abandonment (P&A)
operations exacts a toll on both
the environment and the company’s
financial performance. Many operators are revising their P&A procedures to ensure that abandoned
reservoirs are permanently sealed.
In this article, we review P&A and
decommissioning practices and
new technologies that bring new
meaning to the “permanent”
aspects of P&A work.
Seismic Depth Imaging. In many
of today’s hot exploration areas,
especially where faulting and salt
structures lead to complex seismic
velocities, traditional time-domain
processing gives misleading results;
only depth imaging reveals the true
location and shape of subsurface
features. This article explains
depth imaging and presents examples showing how oil and gas
companies use it to improve their
success rates.
Lifelong Reservoir Management
Using the Web. In the new internet-enabled economy, the ability to
act quickly with up-to-the-minute
information provides a business
advantage. Web-based tools assist
in portfolio management, including
acquisition and divestiture activities. Collaboration among multidisciplinary teams and with partners,
service providers and governmental
bodies is possible with data stored
on secure servers. Accessing applications across the net allows work
to be done from anywhere, at any
time, and creates new ways for
teams to accomplish tasks. This
article describes tools that improve
reservoir management throughout
its life.
82
GeoComputation
Stan Openshaw and
Robert J. Abrahart
Taylor & Francis
29 West 35th Street
New York, New York 10001 USA
2000. 413 pages. $85.00
ISBN 0-7484-0900-9
Combustion and
Gasification of Coal
Sedimentology and Sedimentary
Basins: From Turbulence to Tectonics
A. Williams, M. Pourkashanian,
J.M. Jones and N. Skorupska
Taylor & Francis
29 West 35th Street
New York, New York 10001 USA
2000. 263 pages. $115.00
Mike Leeder
Blackwell Science, Inc.
350 Main Street
Malden, Massachusetts 02148 USA
1999. 620 pages. $56.00
ISBN 0-632-04976-6
ISBN 1-56032-549-6
The book is a compilation of essays on
the specialties that geocomputation
comprises: computer technology,
leading-edge mathematics, visual
analysis and modeling.
Contents:
• GeoComputation
• GeoComputation Analysis and
Modern Spatial Data
• Parallel Processing in Geography
• Evaluating High Performance
Computer Systems from a
GeoComputation Perspective
• GeoComputation Using Cellular
Automata
• Geospatial Expert Systems
• Fuzzy Modelling
• Neurocomputing—Tools for
Geographers
• Genetic Programming: A New
Approach to Spatial Model Building
• Visualization as a Tool for
GeoComputation
• Spatial Multimedia
• Fractal Analysis of Digital
Spatial Data
• Cyberspatial Analysis: Appropriate
Methods and Metrics for a New
Geography
• Integrating Models and Geographical
Informations Systems
• Limits to Modelling in the Earth and
Environmental Sciences
• GeoComputation Research Agendas
and Futures
• Index
The text provides information on
new technology that may impact the
environmental effects of coal generation. Other topics are pollution and its
control and coal-gasification technology.
The book provides explanation of the
physical and chemical processes that
control the deposition of sediments. An
introductory chapter gives perspective
on how the discipline of sedimentology
fits into general earth science study.
Contents:
• An Overview of the Energy
Contribution of Coal
• Properties of Coal
• Pollutant Formation and
Methods of Control
• Combustion Mechanism of
Pulverized Coal
• Combustion Mechanism of Coal
Particles in a Fixed, Moving, or
Fluidized Bed
• Industrial Applications of
Coal Combustion
• Two-Component Coal Combustion
• Coal Gasification Processes
• References, Appendices, Index
Contents:
• Introduction
• Origin and Types of Sediment Grains
• User’s Guide to Sedimentological
Fluid Dynamics
• Sediment Transport and
Sedimentary Structures
• External Controls on Sediment
Derivation, Transport and Deposition
• Sediment Deposition, Environments and Facies in Continental
Environments
• Sediment Deposition, Environments
and Facies in Marine Environments
• Sedimentology in Sedimentary Basins
• References, Index
Throughout, the writing is at a
level that should be understandable
by general readers with modest backgrounds in chemistry.
If you need an up-to-date, comprehensive overview of depositional processes and the resulting sediments...
[the book] is an excellent value and a
good buy for its price.
It is amply illustrated with tables
and charts….Extensive reference
list…. A very good introduction to
the field.
Wenzel LA: Choice 38, no. 5 (January 2001): 937.
...my fundamental criticism: it
contains a fantastic amount of knowledge...but the book provides the reader
with neither the tools nor the perspective on how to use that knowledge for
a clear, practical purpose.
Van De Graaff WJE: Journal of Sedimentary
Research 70, no. 4 (July 2000): 970-971.
…this book provides a complete
map to the road that geocomputation
is taking to mature into a full-fledged
discipline.
Spencer LT: Choice 38, no. 5 (January 2001): 936.
Oilfield Review
50387schD09R1.p82.ps.ps2.ps 12/5/01 7:12 AM Page 2
Gas Migration—Events
Preceding Earthquakes
Applied Sedimentology,
2nd Edition
Leonid F. Khilyuk, George V. Chilingar,
Bernard Endres and John O.
Robertson, Jr.
Gulf Publishing Company
P.O. Box 2608
Houston, Texas 77252 USA
2000. 389 pages. $125.00
Richard C. Selley
Academic Press
525 B Street, Suite 1900
San Diego, California 92101 USA
2000. 523 pages. $82.50
Catherine E. Grégoire Padró
and Francis Lau (eds)
Kluwer Academic/Plenum Publishers
233 Spring Street
New York, New York 10013 USA
2000. 192 pages. $90.00
ISBN 0-12-636375-7
ISBN 0-306-46429-2
ISBN 0-88415-430-0
The book has a strong emphasis on the
applications of sedimentology, especially
in the search for natural resources.
The three main sections discuss the
generation of sediments, sedimentary
processes and structures, and the
transformation of sediment into rock.
The book contains 14 papers presented
at the 1999 American Chemical Society
Symposium on Hydrogen Production,
Storage and Utilization, held in New
Orleans, Louisiana, USA. An introduction
includes discussion of the problem of
carbon dioxide emission and potential
methods of mitigation.
The 27 chapters in this volume cover
key themes on gas migration and its
relation to seismic events. Included are
origins and sources of gas, migration of
natural gas from petroleum reservoirs,
and prediction of land subsidence and
earthquakes based on information
about the rates and contents of
migrating gases.
Contents:
• Tectonics and Gas Migration
• Events Preceding Earthquakes
• Principles of Gas Migration
• Interrelationships Among Subsidence,
Gas Migration, and Seismic Activity
• References, Indexes
The book provides a powerful conceptual basis and methodologies for
understanding and predicting natural
disasters and environmental hazards.
It is very important for environmental
engineers and scientists, civil engineers, petroleum geologists and engineers, seismologists, urban planners
and students of related specialties.
Islam R: Journal of Petroleum Science and
Engineering 29, no. 1 (January 2001): 83-84.
Contents:
• Introduction
• Weathering and the
Sedimentary Cycle
• Particles, Pores, and Permeability
• Transportation and Sedimentation
• Sedimentary Structures
• Depositional Systems
• The Subsurface Environment
• Allochthonous Sediments
• Autochthonous Sediments
• Sedimentary Basins
• Index
Its descriptions of the industrial
applications of sedimentology and
stratigraphy are found in few other
books and will have considerable value
for undergraduate and graduate students in the earth sciences.
With its emphasis on the “practical,” there is considerably more material on issues such as porosity and
permeability and far less on historical
patterns of sedimentation….
The writing is unnecessarily curmudgeonly, bordering on rude, with
remote sensing geologists termed
“mouse-masters,” interpretive diagrams “geophantasmograms,” and
even an imaginary trace fossil called
an “orgasmoglyph.”…Nevertheless,
an important contribution.
Wilson MA: Choice 38, no. 5 (January 2001): 936.
Autumn 2001
Advances in Hydrogen Energy
• The Application of a Hydrogen Risk
Assessment Method to Vented Spaces
• Modeling of Integrated Renewable
Hydrogen Energy Systems for Remote
Applications
• Index
In general, Advances in Hydrogen
Energy presents a very useful and
readable collection of articles. This
book potentially is very helpful to
researchers, students, and engineers
of the field of hydrogen energy systems.
Yürüm Y: Energy and Fuels 15, no. 3
(May/June 2001): 767.
Contents:
• Hydrogen from Fossil Fuels Without
CO2 Emissions
• Hydrogen Production from Western
Coal Including CO2 Sequestration
and Coalbed Methane Recovery:
Economics, CO2 Emissions, and
Energy Balance
• Unmixed Reforming: A Novel Autothermal Cyclic Steam Reforming Process
• Fuel Flexible Reforming of Hydrocarbons for Automotive Applications
• The Production of Hydrogen
from Methane Using Tubular
Plasma Reactors
• A Novel Catalytic Process for Generating Hydrogen Gas from Aqueous
Borohydride Solutions
• Production of Hydrogen from Biomass
by Pyrolysis/Steam Reforming
• Evaluation and Modeling of a HighTemperature, High-Pressure, Hydrogen
Separation Membrane for Enhanced
Hydrogen Production from the WaterGas Shift Reaction
• A First-Principles Study of Hydrogen
Dissolution in Various Metals and
Palladium-Silver Alloys
• Investigation of a Novel Metal Hydride
Electrode for Ni-MH Batteries
• Hydrogen Storage Using Slurries of
Chemical Hydrides
• Advances in Low Cost Hydrogen
Sensor Technology
83
D11R1schPage86 12/10/01 3:22 PM Page 1
Oilfield Review Electronic Archive
1992
1993
1994
1995
1996
1997
1998
1999
2000
About
Topics
Issues
Products
and
Services
Authors
Help
Search
Would you like to receive
archived issues of
Oilfield Review?
Oilfield Review Electronic Archive preserves the look of the printed magazine
in a format that is accessible on both PC
Windows and Macintosh platforms. Fullcolor articles can be printed or explored
on screen using keyword searches,
thumbnail pages, bookmarks and crossdocument links. New features include
the ability to search by author and
Schlumberger product or service name.
Articles published in 35 issues from 1992
to 2000 are available on Oilfield Review
Electronic Archive 1992-2000. Previous
versions of this CD have been a popular
choice among industry professionals.
Oilfield Review Electronic Archive 19922000 is a 2001 Business Marketing
Association Lantern Award winner. Copies
are available from Corporate Express at
cedpm.houston@cexp.com for US $25
(including airmail postage and handling).
Windows is a trademark of Microsoft Corporation.
Macintosh is a trademark of Apple Computer, Inc.