Download Oilfield Review Spring 2009

Oilfield Review
Spring 2009
Electromagnetic Geophysics
Deepwater Project Planning
Volcanic Reservoirs
09-OR-0002
Marine CSEM: Evolution of a Technology
Interest in marine controlled-source electromagnetics
(mCSEM) emerged during the cold war era. A postulated
high-resistivity layer within the oceanic lithosphere was proposed as a means for secure communication in the event of
nuclear war. In a more practical sense, mCSEM became a
reality because pioneering measurements in the 1970s by
Charles Cox and coworkers at the Scripps Institution of
Oceanography showed that the natural electric field noise
level at frequencies around 1 Hz is extremely low (~1 pV/m).
Cox realized that weak fields induced within the earth by a
near-seafloor, horizontal electric dipole source could be
detected many kilometers away. He also understood that
mCSEM is preferentially sensitive to relatively resistive
zones under the seafloor. In contrast, the widely used magnetotelluric method detects conductive material. Thus,
mCSEM made it possible to study the earth in a new way.
Subsequently, Cox and his colleagues designed and built
apparatus to carry out mCSEM surveys in the deep ocean.
First applied in 1979, it continued in use through the 1990s.
Other mCSEM programs evolved at the University of
Toronto, Canada, in the early 1980s and the University of
Cambridge, UK, in the late 1980s. The former was primarily
focused on shallow targets of geotechnical interest, while
the Scripps and Cambridge groups used mCSEM for the
academic study of mid-ocean ridges and related features.
I joined Cox as a postdoctoral researcher in 1980. At that
time, we recognized that mCSEM had potential industrial
applications on continental shelves. In particular, like
many others, we appreciated that petroleum-bearing formations are typically resistive compared with the substrate,
giving mCSEM advantages over magnetotellurics for hydrocarbon surveys. Concomitantly, Len Srnka at Exxon investigated the method. After a brief flurry of interest, industrial
mCSEM dwindled after the mid-1980s—the result of a
combination of low oil prices, the emergence of 3D seismic
reflection technology and a primary focus on drilling in
water under 300 m deep, where mCSEM is problematic
because of interference from a strong component propagating along the sea surface–air boundary.
Interest re-emerged in the late 1990s, resulting in field
demonstrations in 2000–2002 over producing fields off
Angola and West Africa by Statoil and ExxonMobil, respectively. The evaluations showed the viability of mCSEM for
direct detection of hydrocarbons through their low resistivity. Startup companies to provide mCSEM services to the
petroleum industry were spawned, including EMGS, OHM
and AOA Geomarine. The latter was acquired by Schlumberger in 2004. Unfortunately, the resurgence of interest
also precipitated exaggerated claims for the effectiveness
of mCSEM and wild predictions by stock analysts that the
technology would consume 25% of the industry’s marine
exploration budget by 2009. The reality is closer to 5%.
Nevertheless, I predict a bright future for mCSEM as one
of a range of tools that a marine petroleum explorationist
will employ. Rather than being a panacea that yields clear
discrimination of producing discoveries from dry holes,
mCSEM is evolving into one component of an integrated
exploration approach involving seismic reflection and other
technologies. This conjoining of technologies will be especially important as exploration moves into deeper water
where the cost of drilling is extreme. As a consequence,
petroleum companies that do much of their data analysis
and interpretation in-house, such as ExxonMobil, and integrated services companies, such as Schlumberger, will have
a business advantage over specialist companies that provide
only mCSEM. I think consolidation in the industry is likely.
None of this would have been possible in the 1980s.
Although the apparatus in current use by industry is essentially that of academia 20 years ago, incremental improvements have been incorporated as technology has evolved.
The biggest difference is one of scale: The number of
receivers that can be deployed has increased dramatically,
and sources are substantially more powerful. However, the
most important technical improvements have occurred in
interpretation. In the 1980s, modeling was limited to 1D
structures, and 2D analysis was at the cutting edge. Today,
3D modeling is becoming routine and is rapidly evolving in
sophistication. Careful 3D modeling is essential for the
interpretation of mCSEM in hydrocarbon applications. Coupled with other exploration measurements, increasing 3D
capabilities will move mCSEM into the mainstream of
hydrocarbon exploration.
Alan D. Chave
Woods Hole Oceanographic Institution
Alan Chave is a Senior Scientist in the Deep Submergence Laboratory at the
Woods Hole Oceanographic Institution. He holds a BS degree in physics from
Harvey Mudd College, Claremont, California, USA, and a doctorate in oceanography from the MIT-WHOI Joint Program, Woods Hole, Massachusetts, USA.
Alan was involved in early development of mCSEM and maintains an active
experimental research group focused on marine electromagnetics, optics and
ocean observatory technologies.
1
Schlumberger
Oilfield Review
Executive Editor
Mark A. Andersen
Advisory Editor
Lisa Stewart
Senior Editor
Matt Varhaug
Editors
Rick von Flatern
Vladislav Glyanchenko
Tony Smithson
Michael James Moody
4
Electromagnetic Sounding for Hydrocarbons
Deep-reading electromagnetic surveys examine subsurface
resistivity, providing information that is complementary to
seismic data. This article introduces magnetotelluric and
controlled-source electromagnetic technologies. Marine
examples include case studies from the Gulf of Mexico,
offshore Brazil and Greenland.
Contributing Editors
Rana Rottenberg
Glenda de Luna
Design/Production
Herring Design
Steve Freeman
Illustration
Tom McNeff
Mike Messinger
George Stewart
Printing
Wetmore Printing Company
Curtis Weeks
20 Near-Surface Electromagnetic Surveying
Electromagnetic surveys deliver information by probing the
resistivity of the earth. One method employed in land surveys
uses an inductive loop to investigate the near surface. Two
case studies from the United Arab Emirates illustrate its use.
One study mapped an aquifer for a water-storage project.
The other identified the bottom of the low-velocity layer in
an area of dunes, which helped determine how to apply static
corrections for a seismic survey.
On the cover:
WesternGeco specialists install a
navigation beacon onto the transmitter
antenna cable during a controlled-source
electromagnetic survey. The yellow
floats provide buoyancy to keep the
cable at a specified distance above the
seabed. Toisa Valiant, the vessel in the
inset, is equipped for performing these
electromagnetic surveys.
Useful links:
Schlumberger
www.slb.com
Oilfield Review Archive
www.slb.com/oilfieldreview
Oilfield Glossary
www.glossary.oilfield.slb.com
2
Address editorial
correspondence to:
Oilfield Review
5599 San Felipe
Houston, Texas 77056 USA
(1) 713-513-1194
Fax: (1) 713-513-2057
E-mail: [email protected]
Address distribution inquiries to:
Tony Smithson
Oilfield Review
12149 Lakeview Manor Dr.
Northport, Alabama 35475 USA
(1) 832-886-5217
Fax: (1) 281-285-0065
E-mail: [email protected]
Spring 2009
Volume 21
Number 1
26 A Plan for Success in Deep Water
Advisory Panel
Abdulla I. Al-Kubaisy
Saudi Aramco
Ras Tanura, Saudi Arabia
Deepwater E&P operations present the upstream industry
with unprecedented technological and economic challenges.
Addressing them requires a fundamental change in the way
the offshore oil and gas industry operates. The projects in
these environs are best viewed as a single, integrated effort,
from exploration to production and even beyond.
Dilip M. Kale
ONGC Energy Centre
Delhi, India
Roland Hamp
Woodside Energy, Ltd.
Perth, Australia
George King
Independent consultant
Houston, Texas, USA
Eteng A. Salam
PERTAMINA
Jakarta, Indonesia
Jacques Braile Saliés
Petrobras
Houston, Texas
36 Evaluating Volcanic Reservoirs
Richard Woodhouse
Independent consultant
Surrey, England
Volcanic rock can contain oil and gas in commercial quantities,
but evaluating these reservoirs is not straightforward.
By extending techniques designed for evaluating sedimentary
reservoirs, some companies find profitable opportunities in
areas that others might deem unworthy of consideration.
Examples from China and India demonstrate successful
petrophysical evaluation of volcanic formations using neutroncapture spectroscopy, nuclear magnetic resonance, borehole
resistivity images and conventional logging technology.
48 Contributors
51 New Books and Coming in Oilfield Review
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: [email protected]
www.oilfieldreview.com
Annual subscriptions, including postage,
are US $200, subject to exchangerate fluctuations.
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.
© 2009 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.
3
Electromagnetic Sounding for Hydrocarbons
James Brady
Tracy Campbell
Alastair Fenwick
Marcus Ganz
Stewart K. Sandberg
Houston, Texas, USA
Marco Polo Pereira Buonora
Luiz Felipe Rodrigues
Petrobras E&P
Rio de Janeiro, Brazil
Chuck Campbell
ACCEL Services Inc.
Houston, Texas
Leendert Combee
Oslo, Norway
Arnie Ferster
Kenneth E. Umbach
EnCana Corporation
Calgary, Alberta, Canada
Tiziano Labruzzo
Andrea Zerilli
Rio de Janeiro, Brazil
Edward A. Nichols
Clamart, France
Steve Patmore
Cairn Energy Plc
Edinburgh, Scotland
Jan Stilling
Nunaoil A/S
Nuuk, Greenland
Oilfield Review Spring 2009: 21, no. 1.
Copyright © 2009 Schlumberger.
For help in preparation of this article, thanks to Graeme
Cairns, George Jamieson, Jeff Mayville, Fred Snyder and
Xianghong Wu, Houston.
MMCI and Petrel are marks of Schlumberger.
4
Recent advancements in identifying subsurface features by resistivity contrasts
have added a significant tool in the quest to locate hydrocarbon resources.
The electromagnetic sounding technique comprises two related technologies,
magnetotelluric and controlled-source electromagnetic surveys, that provide distinctly
different insights into the subsurface. Their ability to clarify structures and to help
identify possible hydrocarbon deposits before drilling is exciting explorationists.
The Sun provides us with energy in many forms.
A surprising connection between exploration for
energy resources and the Sun is becoming increasingly significant for the E&P industry. Ions emitted
by the Sun experience a complex interplay with
the Earth’s magnetic field, generating propagating electromagnetic fields that penetrate the
Earth and interact with its conductive layers. As
the industry’s search for hydrocarbon resources
intensifies, more geoscientists are relying on these
electromagnetic fields to probe areas that are
difficult to image with seismic methods.
The study of electrical currents in the Earth,
called tellurics, is not new. Conrad Schlumberger,
one of the founders of Schlumberger, used the
phenomenon in early surface studies that he
directed in the 1920s, prior to his start in wireline logging.1 Louis Cagniard, a professor at the
Sorbonne in Paris, first reported combining a
measurement of electric and magnetic fields,
termed magnetotellurics (MT), for exploration
of the Earth’s subsurface in 1952.2 However, MT
has become an important tool for explorationists in the E&P industry only within the past few
years—thanks to advances in 3D modeling and
inversion technology. Now, MT results can be
combined more efficiently with seismic and gravity surveys, resulting in a more-calibrated model
of the earth.
Although Cagniard also discussed a method
related to MT that uses an artificially imposed
electromagnetic field, techniques for generating and detecting a signal strong enough for
use in the E&P industry came decades later, in
the 1960s on land and then in the 1980s in the
marine environment. This method is now termed
controlled-source electromagnetics (CSEM).
The interaction of the earth with impinging electric and magnetic fields is complex.
Two important factors in MT analysis are the
frequency spectrum of the fields and the resistivity (or its inverse, the conductivity) of the
particular medium through which the field waves
propagate. Analyzing data from the frequency
spectrum helps obtain an apparent resistivity as
a function of frequency.3 This apparent resistivity can be related to the true resistivity of the
formation at various depths. If the subsurface is
homogeneous, the measured apparent resistivity is the same as the true resistivity, but if the
resistivity changes with depth, apparent resistivity is a conflation of measurement effects and
some average of the resistivities. Through data
analysis, interpreters can determine the depths
of bodies with contrasting resistivities, providing
a result termed an MT sounding.
Oilfield Review
This article discusses the physics of these
electromagnetic interactions and how they are
interpreted to give information useful in basin
and reservoir evaluation. It also describes the
equipment used to detect and, in the case of
CSEM, to generate relevant electromagnetic
fields. Case studies from the Gulf of Mexico, Brazil
and Greenland illustrate these technologies for
offshore salt mapping and reservoir illumination. A companion article describes near-surface
applications of CSEM on land (see “Near-Surface
Electromagnetic Surveying,” page 20). The next
section focuses on natural electromagnetic fields
and their interactions with the Earth.
Blowing in the Wind
The solar wind is a stream of positive and negative ions emitted by the Sun. Wind intensity
varies, increasing during periods of high sunspot
activity. This ionic wind “blows” through space;
auroras manifest its interaction with the Earth’s
magnetic field in spectacularly colorful ways.4
Although most solar ions are deflected by
the magnetic field in a region known as the
magnetopause, which is several Earth radii out
in space, some ions leak in. Those that reach
the upper atmosphere can ionize particles in
the ionosphere, which ranges from 75 to 550 km
[50 to 340 mi] above the surface of the Earth. In
the ionosphere, the particle velocities are high
enough and the particle density low enough that
charged ions do not immediately recombine
into neutral atoms and molecules: They form a
plasma of charged particles. This plasma makes
the ionosphere a conducting layer, unlike the
nonconducting layers of the lower atmosphere
where the particle density is too high to maintain
charged ions for a significant period of time.
The motions of charges in the ionosphere are
constrained by the Earth’s magnetic field, whose
lines of force stretch from pole to pole. When
solar ions enter the plasma within this magnetic
field, they generate electromagnetic (EM) pulses
1. Leonardon EG: “Some Observations Upon Telluric
Currents and Their Applications to Electrical
Prospecting,” Terrestrial Magnetism and Atmospheric
Electricity 33 (March–December 1928): 91–94.
2. Cagniard L: “Basic Theory of the Magneto-Telluric
Method of Geophysical Prospecting,” Geophysics 18
(1953): 605–635.
3. Apparent resistivity is a volume average of the true
resistivities of the media within the volume measured
by a device, such as a resistivity or induction tool, or a
magnetotelluric receiver.
4. For a recent discussion about the origin of the auroras:
Brown D and Layton L: “NASA Satellites Discover What
Powers Northern Lights,” NASA News & Features,
http://www.nasa.gov/home/hqnews/2008/jul/HQ_08185_
THEMIS.html (accessed March 2, 2009).
Spring 2009
5
10
Magnetic field spectral amplitude
1
0.1
0.01
0.001
0.0001
0.00001
0.000001
0.001
0.01
0.1
1
10
Frequency, Hz
100
1,000
10,000
> Typical magnetic-field amplitude spectrum from the atmosphere. The
ionospheric signal originating from interactions of the Earth’s magnetic
field decays rapidly with increasing electromagnetic frequency. Lightning
generates signals in a region called the Schumann bands in the spectrum
between about 7.8 and 60 Hz.
that resonate in the ionosphere, traveling along
the magnetic field lines. The result is analogous
to plucking the string of a guitar; just as the
string resonates at characteristic frequencies, so
too does the ionosphere resonate electromagnetically. The complex interaction of magnetic field,
atmospheric plasma and solar ions results in a
broad spectrum of EM frequencies, including the
visible-light phenomena of the aurora borealis
and the aurora australis. The spectral range useful for E&P-related MT extends from frequencies
of about 0.001 Hz to 10 kHz; for studies extending to the Earth’s mantle even lower frequencies
are used (above). Frequencies above 1 Hz are
severely attenuated through conductive seawater
and thus create no subsea earth response, making this the effective upper-frequency limit for
marine MT.
The amplitude and frequency spectrum of
the signal is highly variable.5 The fluctuations
in the solar wind reflect the 11- to 14-year cycle
of sunspot activity. The spectrum also depends
on the season and time of day, since sunlight
influences the degree of polarization in the
ionosphere. Signal levels in equatorial regions
are low, whereas they are high in polar regions.
Geomagnetic index
16
12
8
Annual 2007
average
EM_FIGURE 01
2005
2006
2008
4
0
1/1/08
4/1/08
7/1/08
10/1/08
12/31/08
Date
> Electromagnetic activity. Planetary electromagnetic activity is estimated
from measurements of a geomagnetic index taken by the US National Oceanic
and Atmospheric Administration (see reference 5) at several locations. The
activity fluctuates both annually and weekly, as shown for 2008 (black). The
solar cycle is currently in a quiet period.
6
This stronger signal near the poles or near
the peaks of the solar-activity cycle results in
higher-quality MT data; conversely, obtaining
data from deepwater equatorial areas, especially
during low-activity periods, is more challenging
(below left).
A part of the frequency spectrum is influenced
by lightning. A lightning discharge can generate
current in the range of 20 to 50 kA, which initiates a strong interaction in the ionosphere. The
charge pulse follows the magnetic field lines
around the Earth, reflecting near the poles and
playing its own resonance notes.6 The EM fields
resulting from a lightning strike are global.7
The lower atmosphere is a poor electrical
conductor, so the EM waves propagate with virtually no attenuation.8 This lack of attenuation
allows radio broadcasts to be heard far from the
source when atmospheric conditions are right for
refracting them to listeners. In contrast, once
the waves reach the surface layers of the Earth,
they interact with seawater and formations that
are electrically conductive to a greater or lesser
extent. Conductive bodies attenuate EM waves.
Most of a rock’s solid matrix conducts electricity poorly. However, various saturating fluids
have differing conductivities. Brine conducts
well, but oil and gas have high resistivities.
Adjacent formations with a marked resistivity
contrast—such as a hydrocarbon-bearing zone
surrounded by brine-saturated strata—affect the
propagating EM field in different and potentially
measurable ways. The resistivity contrast is also
high between brine-filled sedimentary layers and
some specific lithologies, such as salt, basalt and
resistive carbonates.
The EM waves interact with conductive formations and induce a response wave that propagates
back to the surface. Although the geometry of
signal and response is sometimes depicted as
analogous to that of a seismic reflection, the EM
effect has a different physical origin and different behavior than a reflected seismic wave.9 The
time-varying EM signal induces a current loop
in the conducting layer. The properties of this
induced eddy current depend on the resistivity of
the conducting formation and the magnitude and
time rate of change—or the frequency—of the
source signal. The eddy current, in turn, induces
a magnetic field, which propagates from the
formation. Sensors on the surface measure this
response field.
Oilfield Review
ω=2
ω=5
ω = 10
ω = 10
° ~0
° ~0
° ~ 10
° ~1
Skin depth
Skin depth
Skin depth
Skin depth
> Skin effect. A downward-moving electromagnetic field (blue curve)
leaving a highly resistive medium, such as air, begins to decay when it
enters a more-conductive medium, such as rock. Lower-frequency waves
(left ) propagate farther than higher-frequency waves (center left and
center right ), and waves propagate farther in less-conductive media
(right ). The amplitude has an exponential decay (red) that is a function of
the conductivity of the medium, σ, and the frequency of the wave, ω. The
skin depth is the distance at which the amplitude has decayed to 1/e of
the incident value. The wave in the conductive medium also experiences a
gradual delay in the phase. Since the phase change is difficult to see in this
example, one illustration (far left ) also shows an attenuated wave without
the phase change (violet). Frequency and conductivity values are relative
among these examples.
The eddy current in the conducting formation
opposes the change in the source field. The result
of the eddy current and the transfer of energy to the
response signal is attenuation of the incoming EM
wave. Thus, as the wave passes successively deeper
into the conductor, the eddy current becomes
incrementally weaker, making the response field
smaller also. As this process continues, the incident signal decays, while weaker response signals
form at each successive increment of depth within
the conducting formation. This decay is known as
the skin effect (above).
A characteristic distance for penetration of the
signal into a conductor, termed the skin depth, is
obtained by determining when the field amplitude
drops by a factor of 1/e, the inverse of the exponential function. Attenuation is frequency dependent;
high frequencies attenuate more rapidly than low
frequencies. It is also a function of the formation
conductivity: In more-conductive formations the
impinging field induces greater current flow that
partially cancels the source field. In a typical geological section, the natural frequencies used in
MT have skin depths of a few tens to a few tens of
thousands of meters. The high-frequency components useful for detecting thin, shallow formations
are present only for land-based (or extremely
shallow-water) surveys because of attenuation by
conductive seawater. The deeper a target structure
is buried, the larger it must be to enable detection
through MT evaluation; this basic MT-resolution
problem at depth is more severe than resolving
small, deep features using seismic waves.
The response signal contains information in
the impedance value about the resistive properties of the formations. Impedance is a complex
term—comprising real and imaginary parts—
EM_FIGURE
that designates the
difficulty 04
of propagating the
EM energy through a medium. It is determined
from the amplitude and phase relationship that
exists between the measured electric and magnetic fields.10 It is also a tensor quantity that can
be related to the apparent resistivity of the formation. Impedance varies with the frequency of
the incoming signal.
Because the source is so distant, the MT
fields impinging on an E&P survey area can be
approximated over a wide bandwidth as vertically incident plane waves with the electric field
horizontally polarized.11 MT fields are sensitive
to large conductive features, making them useful in studies of large salt, basalt and carbonate
bodies due to the contrast of these resistive
features with the conductive surroundings.
However, the attenuation of the MT fields with
depth—the skin effect—makes them insensitive
to resistivity contrasts of thin formations such as
5. Data are available from the US National Oceanic and
Atmospheric Administration, http://www.swpc.noaa.
gov/ftpmenu/indices/old_indices.html (accessed
May 5, 2009).
6. This response to lightning is termed a Schumann
resonance, after German physicist Winfried
Otto Schumann, who predicted the resonances
mathematically in 1952.
7. Active storms generating lightning seem to be linked:
Synchronized lightning strikes from widely spaced
geographic locations have been observed from the
NASA Space Shuttle. For more on synchronized lightning
strikes: Yair Y, Aviv R, Ravid G, Yaniv R, Ziv B and Price C:
“Evidence for Synchronicity of Lightning Activity in
Networks of Spatially Remote Thunderstorms,” Journal
of Atmospheric and Solar-Terrestrial Physics 68, no. 12
(August 2006): 1401–1415.
8. Electromagnetic waves propagate through a vacuum
with no attenuation.
9. EM energy in a conductive medium has a diffusive nature
rather than a wave nature.
10.The phase of a wave describes where it is in its
amplitude cycle of maximum to minimum and back to
maximum as the phase angle goes from 0° to 360°. The
electric and magnetic fields of a propagating wave are
not necessarily at 0° phase at the same time, and the
difference between the two is also referred to as the
phase angle.
11.The waves impinge vertically because air is not
conductive. The uniformity of signal for MT surveys is
based on the large distance to the ionosphere compared
with the length of a survey line. However, if the signal
comes from a lightning strike that is close to the survey
area, the plane-wave assumption does not hold and the
local geometry influences the interpretation.
Spring 2009
7
Ex (t)
Marine MT
Hy (t)
Ex =
H
V
Marine CSEM
Passive (atmospheric) source
Active controlled source
Plane waves, vertically incident
Localized dipole source
Basin scale
Reservoir scale
Detection of structure and lithology
Detection of resistivity contrast, such
as that caused by a resistive pore fluid
against a conductive background
Wave-field sensitive to conductors
Wave-field sensitive to resistors
> Comparison of marine MT and CSEM survey technologies.
V
L
X
L
Ex
Z=
= iρ ° µ
a
Hy
Y
ρa =
° =
µ=
Z=
E=
H=
t=
V=
L=
i=
formation resistivity
frequency
magnetic permeability
formation impedance
electric field
magnetic field
time
voltage drop across dipole
dipole length
–1
> Sensing impedance. A vertically incident
EM wave interacts with the Earth through the
formation impedance, Z. The Z value can be
determined by measuring the horizontal electric
field, E, and the magnetic field, H, at the surface
or on the seabed (tan). The apparent resistivity,
ρa, is the aggregate resistivity of the formation
layers beneath the electric dipole antenna and
the magnetometer coil of a sensor (yellow). In
the case shown, E and H are in phase; if the
zero crossings of the two fields were out of
synchronization, there would be a phase angle
between the two fields.
hydrocarbon-bearing sediments. Generally, to be
resolved by MT, the layer’s thickness should be at
least 5% of its burial depth, and the layer should
be more conductive than its surroundings. These
limitations led to the development of the CSEM
EM_FIGURE 07
method (above right).
The CSEM method imposes a powerful, artificially generated EM signal. The source is a
localized electric dipole with a controlled signal
that extends over a narrow bandwidth, often just
a few fundamental frequencies and their harmonics. The EM fields generated by such a source are
not plane waves. The composition and geometry
of the signal are chosen to make it more sensitive
for detection of a thin formation at a particular
8
hypothesized location and with a resistivity value
that contrasts to that of surrounding formations.
This difference between MT and CSEM source
signals affects the method of processing the data
and impacts the type of structures that can be
detected by the two methods, as discussed in the
next two sections.
to obtain a model of the earth. The result is not
unique, so the process iterates until the result is
acceptable. Many algorithms are in use for converging the inversion on a particular model.
A key step of preacquisition planning is to
determine if different models will be distinguishable in the data. This is typically accomplished
by first forward modeling the response of various
Deep Vision with MT
predicted scenarios, then possibly employing
The atmospheric source for MT signals varies inversion on modeled synthetic data. To invesrandomly in time, but at any given time the ver- tigate whether the original model can be
tically incident waves are uniform over a large recovered, the synthetic data include noise reparea. The wavefields are planar and vertically resenting expected background or measurement
incident on the surface of the Earth; the electric noise. This step can help justify the usefulness of
field has only horizontal components, as does the a proposed survey or, alternatively, advise against
orthogonal magnetic field. As a matter of nomen- its application. Acquisition parameters such as
clature, the portion of the electric field that can the location of instruments and how long they
be resolved along the strike of a geologic feature must remain on the ground are also results of this
is termed the transverse electric (TE) mode; the process. In CSEM surveys, the optimal frequenportion across the strike is the transverse mag- cies can also be determined through modeling.
netic (TM) mode.
Recent interest in MT measurements has
Because of the vertical and planar geometry focused on evaluations in marine environments,
of MT, the impedance of the subsurface can be driven by the increasing costs of drilling in deep
obtained by taking the ratio of the horizontal water and the complexity of imaging below salt
electric field in one direction to the horizon- and basalt. As a result, technologies that increase
tal magnetic field in the orthogonal direction the chance of economic success after locating
(above left).12 This calculation removes the tem- drilling targets have great value.
As with seismic surveys, EM surveys require
poral variability of the incident signal, leaving
EM_FIGURE
05
deployment of equipment, either on land or at
only the desired formation response.
The complex impedance can be calculated to sea. Marine MT surveys are acquired using small
obtain the apparent resistivity, ρa, of the underly- vessels and small crews. CSEM surveys need
ing formations and the phase angle, φ, between larger vessels to handle the source equipment
the electric and magnetic fields. Geoscientists and larger crews to operate and maintain that
use these results to interpret the subsurface equipment. Typically, both MT and CSEM surstructure through forward modeling or through veys are targeted, examining specific ambiguous
inversion.13 Forward modeling assumes a struc- structures or promising anomalies on a seismic
ture and certain properties, such as layer depth section. Thus, the duration and areal scope of
and resistivity, and predicts the earth’s elec- these studies are typically smaller than those for
tromagnetic response to the assumed model. seismic surveys.
Subsea EM measurements—both MT and
Comparing or normalizing processed data against
this model assesses its goodness of fit. Inversion CSEM—are similar to land measurements aside
is the reverse of forward modeling, using the data from the vast difference in the resistivities of seato step backward through the physical process water and air. At the air/land interface there can
Oilfield Review
be no vertical electric current because the air is
not conductive, but on the seabed a vertical electric current can exist in the conductive water.
The consequence of this difference is subtle. On
land, the electric field responds significantly to
changes in resistivity in subsurface layers, but
the magnetic field has much less variation. In
contrast, in the marine environment it is the
magnetic field rather than the electric field that
displays the greater variation with change in
subsurface structure, although both fields carry
information on structure.14
Measuring the Signal
The two basic devices for measuring EM fields
are a pair of electrodes to sense an electric field
potential difference and a magnetometer to sense
magnetic field variations. The pair of electrodes
forms an electric dipole, allowing measurement
of the potential voltage difference between them.
A magnetometer is a coil of conducting wire that
generates a measurable current based on the
changing magnetic flux through the coil.
When only two sensors of one type are used,
they are oriented to measure the orthogonal field
components in the horizontal plane. The vertical
component of the field is measured only if a third
sensor is used.
The primary recent interest in the E&P industry has been offshore, and considerable effort has
been made over the past decade to develop a
sensor for marine use. The Scripps Institution of
Oceanography in La Jolla, California, USA, developed the basic electric field sensor that is used
by WesternGeco today. The magnetometers were
developed by Electromagnetic Instruments Inc.,
which was acquired by Schlumberger in 2001.15
In the WesternGeco device two horizontal
electric dipoles are formed by silver–silver chloride electrodes placed at the ends of four long
fiberglass tubes, extending from each of the four
sides of the receiver frame (above right). The present configuration includes a vertical dipole with
a length of 1.82 m [6 ft]. Its length is restricted
by the need to maintain orthogonality and
stability—a longer dipole is more susceptible to
seabed currents that move the dipole antenna
and introduce noise into the measurement within
the frequency bandwidth of interest.
The magnetometers, multiturn coils in a nonmetal housing, are used to sense the magnetic
flux. The magnetometer tubes are secured horizontally into holes in the frame. The operating
range is from 0.0001 to 100 Hz.
Calibration of both types of sensor is critical. The WesternGeco sensors and amplifiers are
individually calibrated far from electromagnetic
Spring 2009
Instrumented
strayline float
Dipole for
electric field
Gas flotation
Induction coil
magnetometers
Logger
Acoustics
Burnwire release
mechanisms
Concrete
anchor
> CSEM receiver. Orthogonal dipole antennas on the receiver measure Ex and Ey and two induction coil
magnetometers measure Hx and Hy . Each tube containing an antenna is 3.6 m [12 ft] long; coupled with
the dimension of the frame, the electric dipole length formed by a pair pointing in opposite directions
is 10 m [32.8 ft]. A concrete anchor carries the receiver to the seafloor, where it remains throughout
the test. The electronic logger records for a set time. At the conclusion of the test, an acoustic signal
from the ship triggers a mechanism to burn through the wire holding the device to the anchor. Airfilled glass spheres raise the receiver to the surface, where it is retrieved and the data are captured.
In some cases, the receiver includes a vertical dipole to measure the vertical electric field, Ez (not
shown). (Image courtesy of Scripps Institution of Oceanography.)
noise in a remote part of the Norwegian countryside. In addition, data quality requires strict
adherence to deployment procedures on the
survey ship.
A concrete block attached to the bottom of
the receiver frame provides weight to take it to
the ocean floor. This concrete anchor also helps
to stabilize the instrument against forces from
sea currents; antenna rotation as tiny as 1 μrad
can easily be detected by the magnetic induction coil moving in the Earth’s magnetic field. At
the conclusion of the survey, an acoustic signal
from surface triggers release from the block, and
air-filled glass spheres lift the receiver to surface
for retrieval.
The cost and logistics of establishing electrical connections with multiple receivers placed
on the seabed in deep water are prohibitive, so
engineers designed the receiver to operate independently and to be retrieved at the end of the
test. Each receiver carries a data logger that
controls operation and records the signals on a
compact flash card. High-resolution data from
the dipoles and magnetometers come from 24-bit
analog-to-digital converters, which accurately
record time so that the signals can be synchro-
nized later with the source record and with
each other.
The unit has several independent battery
packs. One provides power to the data-logger
electronics. A separate battery powers the anchorrelease devices, and another powers an acoustic
positioning beacon that indicates the unit’s location on the seabed. The battery pack that powers
the data logger lasts up to 40 days; the long
battery life provides time to deploy the sensors
and then acquire data. The anchor-release battery
pack lasts more than a year, in case circumstances
prevent immediate removal of the device after
the survey.
12.Cagniard, reference 2.
13.For more on inversion: Barclay F, Bruun A, Rasmussen KB,
Camara Alfaro J, Cooke A, Cooke D, Salter D, Godfrey R,
Lowden D, McHugo S, Özdemir H, Pickering S,
Gonzalez Pineda F, Herwanger J, Volterrrani S,
Murineddu A, Rasmussen A and Roberts R: “Seismic
Inversion: Reading Between the Lines,” Oilfield Review 20,
no. 1 (Spring 2008): 42–63.
14.Constable SC, Orange AS, Hoversten GM and
Morrison HF: “Marine Magnetotellurics for Petroleum
Exploration, Part I: A Sea-Floor Equipment System,”
Geophysics 63, no. 3 (May–June 1998): 816–825.
15.Webb SC, Constable SC, Cox CS and Deaton TK:
“A Seafloor Electric Field Instrument,” Journal of
Geomagnetism and Geoelectricity 37, no. 12 (1985):
1115–1129.
Constable et al, reference 14.
9
The seabed orientation of the horizontal sensors is random. The measurement directions are
resolved to a desired orientation during processing. The newest devices have a compass, but in
the past the orientation for each receiver was
obtained either by comparison with land-based
sensors or by orientation based on the direction
of a towed source in a CSEM survey.
Towfish
Cable to
survey vessel
Streamer Antenna
Neutrally
buoyant cable
Tow-cable termination
300-m dipole
Transponder
A
Electrode 1
B
Electrode 2
2.5 m
Strain relief
Instrument
suite
Transponders
20 m
A: Telemetry and signal conditioning
B: Transmitter power section
> CSEM transmitter. The transmitter comprises a towfish—the head section containing power and
instrumentation—and a streamer antenna with dipole electrodes at the ends of two cables. The
dipole is the source of the CSEM signal. The signal transmission and waveform parameters are set
from the survey vessel during operations, and results are telemetered to the operators for real-time
quality control of the signal. The photograph (top) shows a towfish being removed from the ocean,
with the antenna trailing in the water.
1.5
ω0
Five-term sum
1.0
Amplitude
0.5
9ω 0
3ω 0
5ω 0
7ω 0
0
–0.5
–1.0
–1.5
0
1
2 09
EM_FIGURE
3
4
Time, s
Square wave (ω 0) =
4
°
sin(ω 0t) +
sin(3ω 0t)
3
+
sin(5ω 0t)
5
+
sin(7ω 0t)
7
+
sin(9ω 0t)
9
+ ...
> Square-wave components. A square wave (magenta) can be broken into
an infinite series of sine waves by using the Fourier transform (equation). The
fundamental frequency, w 0 , has the greatest amplitude; each subsequent odd
harmonic has a lower amplitude. Even-numbered harmonics are not included
because of the symmetry of the square wave.
10
CSEM: Focusing on Hydrocarbon Detection
MT measurements are not sensitive to thin resistive layers, so they are not well suited for evaluating potential hydrocarbon reservoirs. Over the
course of a few decades starting in the 1980s,
several research institutes and companies developed the equipment, modeling and interpretation
tools that became the marine CSEM technique
(see “Marine CSEM: Evolution of a Technology,”
page 1).16 The systems are now widely available.
Since the same receivers function for both
CSEM and MT measurement, both responses
can be recorded during a survey. The CSEM
technique focuses on measuring and interpreting the response from the controlled source,
while between those measurements, MT data
are recorded. The processed and interpreted
MT data establish a background model for the
CSEM interpretation or inversion.
The typical marine CSEM transmitter source
is a long horizontal dipole (above left). The
source comprises two neutrally buoyant antenna
cables, each terminating in an electrode, thereby
forming a dipole. The electrodes are pulled
through the water behind a streamlined sensor platform, called a towfish, that is towed by
the ship at a nominal speed of 2.8 to 3.7 km/h
[1.7 to 2.3 mi/h or 1.5 to 2.0 knots] at an altitude
of 50 to 100 m [160 to 330 ft] above the seabed. To
provide accurate values for processing, the towfish measures seawater conductivity, local sound
velocity and altitude above the seafloor.
The strength of the dipole source is given by
its dipole moment. This value is the product of
the magnitude of the electric current flowing
through the electrodes—given by the strength of
the first harmonic of the output signal—and the
distance between the electrodes.
The power to generate a high-current, lowvoltage source signal and propagate it along
16.The first development was by Charles Cox of Scripps
Institution of Oceanography: Cox CS: “On the Electrical
Conductivity of the Oceanic Lithosphere,” Physics of
the Earth and Planetary Interiors 25, no. 3 (May 1981):
196–201.
For a recent overview of the history of CSEM: Constable S
and Srnka LJ: “An Introduction to Marine ControlledSource Electromagnetic Methods for Hydrocarbon
Exploration,” Geophysics 72, no. 2 (March–April 2007):
WA3–WA12.
Oilfield Review
[6.2 mi] away, the electric-field magnitude is
small, less than 1 nV/m. For the typical 10-m span
of a seabed receiver dipole, the measured 10 nV
is about 80 million times smaller than a AAA
battery’s 1.2 V. The response magnetic-field magnitude at that distance from the source is about
0.0001 nT, which corresponds to about 2 parts in
a billion of the Earth’s DC magnetic field.
The controlled source typically generates
square waves or sequences of square waves at
user-defined fundamental frequencies. Fourier
analysis resolves the square wave into sinusoidal
waves of many frequencies (previous page, bottom left). The strongest components are the
primary frequency w0 and the odd harmonics 3w0, 5w0 and 7w0, each with sequentially
decreasing magnitudes. The combination of the
skin-depth relationship to frequency and use of
multiple frequencies means this process samples
at several depths and with several resolutions.
several kilometers of cable is typically provided
by a 250-kV.A system on the ship. Transformers
convert this to a low-current, high-voltage signal
sent along the cable. In the towfish the signal
is transformed back to the high-current, lowvoltage signal.
The towfish generates a designed waveform
based on commands from the ship. The actual
current waveform transmitted by the source electrodes is measured and recorded by a data logger
in the towfish and transmitted to the vessel in
real time for quality control via high-speed telemetry. Because the waveform transmitted by the
antenna is affected by antenna impedance and
wear and by water salinity, accurate monitoring
of the actual waveform is required to correctly
resolve the survey data.
Although the power emitted at the source
is large—nominally 50 kW—the signal decays
rapidly with distance. At a receiver placed 10 km
The data from the receivers are collected as
time-series data, but for the CSEM method, they
must be synchronized to the source square-wave
signal through accurate time measurement.
Thus, in addition to the source GPS synchronization, each receiver has a high-precision clock
that is GPS synchronized upon deployment and
recovery. The instantaneous dipole-source position and orientation must also be captured for
accurate inversion. Acoustic transponders in
several locations along the antenna give this
information by transmitting their positions at
1- to 4-s intervals. Accurate measurement of the
feathering or tilt of the antenna is important for
correct processing.
The measurements of the fields are timedomain data, but these are typically converted
to the frequency domain using a Fourier transform (below). The data are stacked by overlaying
responses from multiple, sequential square-wave
Ex
Ey
Hx
Hy
Time, min
5
10
15
20
25
30
35
40
–8
Frequency, Hz
0.0625
0.1875
0.25
0.315
0.4375
0.75
1.25
1.75
Scaled electric amplitude, V/(A.m2)
–9
–10
–11
–12
–13
–14
–15
–16
–10
–9
–8
–7
–6
–5
–4
–3
–2
–1
0
1
2
3
4
5
6
7
8
9
10
Source-receiver offset, km
> Converting time-domain measurements to amplitude versus offset. Each receiver records data for two horizontal electricand magnetic-field measurements (top). A Fourier transform converts these time-domain signals into the frequency domain.
Fourier conversions of similar measurements at many receiver locations allow development of a frequency-dependent
amplitude versus offset relationship (bottom). This can be developed for each measured component of the electric field (only
one is shown) and the magnetic field. The resistivity of the subsurface affects the shape of these curves.
Spring 2009
11
Air-wave signal
Direct sig
Water
nal
Conductor
Geologic signal
Resistor
10-3
10-4
Electric field, V/m
10-5
10-6
-7
10
10-8
Reservoir
10-9
Background
(no resistive formation)
10-10
10-11
10-12
0
1
2
3
4
5
6
7
Source-receiver separation, km
8
9
10
> Paths from marine source to receivers. Signal energy from the marine source reaches the receivers
by following three types of paths. A direct signal passes through the water to the receiver; this signal
is strongest at the near-offset receivers. Signal energy that enters the subsurface interacts with layers
of varying resistivity and generates a response signal containing geologic information that travels up to
the receivers. Signal energy that reaches the air interface travels along the interface as an air wave,
which also travels to receivers. In shallow water or at long source-receiver offsets in deep water, the
air-wave signal is strongest.
series, called a time gather, to improve the signal/
noise ratio. The window for the time gather must
be short enough that the source movement does
not significantly alter the sampled volume of
the subsurface.
Since the objective of E&P prospecting is
to detect hydrocarbons, the signal from the
CSEM source is optimized to find thin, nonconducting layers (possible hydrocarbon-bearing
formations) in a conducting background (waterbearing formations). The discussion on skin
depth pointed out that detecting thin formations
requires higher-frequency components than
available using MT. The typical frequency range
of the CSEM signal is between 0.05 and 5 Hz; 1 Hz
is the effective upper limit for marine MT studies.
As a first-order approximation, the signal can
take three general paths between the source
and the receivers (above). When the sourcereceiver offset distance is short, the direct path
through the water dominates the signal. The
strength of the signal decreases rapidly with
12
distance because of its attenuation in conductive water. A second contribution comes from
the air wave. The electromagnetic field travels
to the water surface, where it encounters highly
resistive air. The resistance contrast forces
the wave propagation to follow the air/water
interface. In deep water, the air-wave signal
dominates only at long offsets, normally beyond
10 km, because, unlike the signals following the
other two paths, the signal at the air/water interface has little attenuation.
The third portion of the signal travels through
the subsurface. Under the proper conditions of
frequency, water depth and subsurface conductivity, there is aEM_FIGURE
range of offsets
11 for which the
third path dominates the signal. For this path,
waves propagate into the subsurface, where they
interact with resistive formations and generate a
response field; some of that energy travels back
to the seafloor receivers. This response signal
appears at receivers at offset distances that are
typically greater than the reservoir depth below
the seabed, but at even greater offsets it attenu-
ates so much that the air-wave signal overwhelms
it. Since the waves propagate more easily though
a resistive than a conductive formation, the
presence of a reservoir enhances the received
signal compared to a uniform subsurface lacking a resistive layer. Geoscientists can identify
resistivity anomalies and therefore infer geologic
information by analytic means through comparing the observed data with predictive models or
by numeric means through inversion.
At a certain offset distance, the natural noise
limitation of the receiver exceeds the strength of
the signal that originated at the source transmitter, presenting an effective limit on the depth of
investigation in the subsurface. This limitation,
or noise floor, varies with frequency and depends
on the characteristics of the receiver and its
environment—such as mechanical noise generated by water waves moving the antennas. The
noise floor can be lowered through improved
instrumentation, such as quieter electronics or
more-stable mechanics, or through intelligent
signal processing to remove motion noise or
coherent noise across the survey.
The source, receiver and environmental
characteristics can be incorporated into a presurvey analysis to determine whether a resistive
target at a certain depth can be detected (next
page). Carbonates, which are resistive, present
a problem: A trap with low oil saturation inside
a resistive carbonate host may have insufficient
detectable contrast.
The receiver data can be presented as electricor magnetic-field amplitudes and phases that are
functions of the offset distance between source
and receiver. The effect of a resistive anomaly
can be highlighted by several methods: analytic
methods using only measured data, modelbased methods derived during survey planning,
and inversion.
Oilfield Review
Spring 2009
0
1.5
1.4
Seawater
500
1.3
1.2
1.1
1,000
Frequency, Hz
1.0
1,500
2,000
2,500
0.9
0.8
0.7
0.6
Amplitude
ratio
6
0.5
5
0.4
4
3
0.3
3,000
2
0.2
Depth, m
1
0
3,500
2,000
1.5
4,000
Basalt
4,000
6,000
8,000
Transmitter-receiver offset, m
10,000
12,000
1.4
1.3
4,500
1.2
1.1
5,000
1.0
Frequency, Hz
One of the analytic methods normalizes the
electric- and magnetic-field amplitudes versus
offset response over the anomaly to the response
of a distant receiver that does not sense the
anomaly. A second analytic method compares the
normalized response of the inline measurement
with the crossline measurement, essentially
comparing the two horizontal components of
the electric field, Ex and Ey. The presence of an
underlying resistive structure, such as a hydrocarbon-bearing formation, has greater effect on
the inline response because of the polarization
of the signal.
A third analytic method converts the field
data to apparent resistivity in a 2D or 3D pseudosection plotted as a function of source-receiver
offset and signal frequency.17 When the dataset
is normalized to a section space that contains
no anomaly, the anomalous apparent resistivity
values appear as deviations from unity.
Alternatively, presurvey models can be built
when seismic data or data from nearby wells
are available. Typically, a WesternGeco survey
includes at least two 3D models that are based
on the target properties and survey geometry.
One model incorporates a resistive body; the
other uses a uniform earth without a resistive
body. Response curves are extracted from the 3D
models for each receiver-site and tow-line combination. Once data are acquired, the observations
can be normalized to each of the models to determine which provides the best fit.
Beyond these analytic and model-based
methods, CSEM inversion is a powerful way to
derive the earth’s resistivity profile from observed
data. However, like most inversion methods, the
solution is not unique. Forward-modeling codes
are run iteratively with model parameters perturbed until the output result matches the data
within an acceptable range. Jointly inverting as
many significant channels and frequencies as
possible constrains the possible solutions, but
at a cost of longer processing time. Additional
constraints—such as placement of known geologic structures—are sometimes introduced. Log
and seismic data provide a starting model to help
constrain the inversion.
MT data also have limited resolution, so the
modeling step benefits from information based
on other types of measurement. Seismic interpretations often serve as constraints. Gravity
surveys provide an independent constraint, as
do well logs. The WesternGeco MMCI multi­
measurement-constrained imaging technique
uses an iterative approach with gravity, MT and
seismic data to improve inversion results, leading
to a final, more-constrained depth image.
5,500
6,000
0.9
0.8
0.7
Phase
difference, °
0.6
40
30
20
10
0
–10
–20
–30
–40
0.5
6,500
0.4
0.3
0.2
7,000
1
10
Resistivity, ohm.m
100
0
2,000
4,000
6,000
8,000
Transmitter-receiver offset, m
10,000
12,000
> Presurvey modeling. To optimize CSEM acquisition parameters, the subsurface is modeled as a
series of resistive layers (left ). Two models having identical geometries are compared. One model
incorporates a layer of highly resistive basalt (yellow and brown); the other model assigns that layer
a lower resistivity (yellow only). The two models have different phase and amplitude responses to a
simulated CSEM pulse. The amplitude ratio between the models (top right ) is maximum (red) at an
offset—distance from source to receiver—of about 7,000 m and at a frequency of about 0.7 Hz. The
phase difference (bottom right ) has a maximum (red) at about 8,500 m and at a frequency less than
0.1 Hz, and another maximum (violet) at long offset and high frequency. Based on the information in
both plots, geoscientists determined that the optimal offset to maximize the chance of detecting this
anomaly is about 8,000 m, at frequencies of 0.5 and 0.125 Hz. The contour lines indicate various levels
of receiver noise floors (labeled by the power of 10), which depend on the sensors, electronics and
the environment. Although the noise floor in some environments may be as poor as 10-14, these plots
extend to a noise floor of 10-15, which can usually be achieved.
Although marine MT and CSEM receivers Mexico, offshore Louisiana, USA.18 Exploration
have been used in studies since the 1990s, the companies have had an interest in evaluating
industry’s interest has risen rapidly in the last hydrocarbon potential in subsalt formations
few years, resulting in a rapid increase in the in this area. The seismic data available at the
total number of sites evaluated. A large, multi­ time, a legacy survey called E-Cat, had been
phase study recently performed in the Gulf of reprocessed recently over Garden Banks, but it
Mexico included more marine MT receivers than had insufficient resolution to reliably determine
EM_FIGURE
05a
the base of a salt intrusion. The objective of the
the total deployed worldwide to that date.
new study was to integrate marine MT, fulltensor gravity and seismic measurements using
Finding the Base of Salt
In 2006, WesternGeco began a test of the MMCI an MMCI evaluation to improve the interpretaconcept in the Garden Banks area of the Gulf of tion of the base of salt.
17.A pseudosection uses approximate or pseudo spatial
coordinates. It provides a semiquantitative way to look
at spatial data.
18.Sandberg SK, Roper T and Campbell T: “Marine
Magnetotelluric (MMT) Data Interpretation in the Gulf
of Mexico for Subsalt Imaging,” paper OTC 19659,
presented at the 2008 Offshore Technology Conference,
Houston, May 5–8, 2008.
13
Line 5
Line 4
Line 3
Line 2
Line 1
N
Seafloor
depth, m
750
975
1,200
Tamara well
1,425
1,650
Line 6
The Garden Banks study included 171 seabed
receivers, more than any previous marine MT
survey, although surveys of this density are more
common today. The survey utilized five parallel north-south lines of receivers, about 2.5 km
[1.6 mi] apart, and an east-west cross line
(right). Additional receivers placed between
these lines provided denser coverage near the center of the survey area. Bathymetry data indicated
seabed expressions of the underlying salt domes.
During the course of the project, two events
provided additional data for this investigation. During October and November of 2007,
WesternGeco acquired a multiclient wideazimuth (WAZ) seismic survey over the area, which
provided significantly better resolution for base of
salt than did the previous E-Cat narrow-azimuth
survey. However, even with WAZ illumination, the
base of salt was poorly resolved in some areas.19
The second event occurred near the end of
2007, when BP released logging data from its
Tamara well in Garden Banks Block 873. This
well was drilled through the central portion of
the survey area. The gamma ray log indicating
the base of salt became available after most of
the MT interpretation was completed, providing
a base-truth point for comparison.
An approach combining 1D models for each
receiver station detected the salt body, but the
details of the structure were incorrect because
of its complex geometry. Several 2D approaches
14
1,875
2,100
0
0
km
10
mi
10
> MT survey in Garden Banks area. The MT receivers (inset) were placed in
five north-south lines and one east-west crossline. Additional receivers were
placed in the central area, near the Tamara well. The color-coding indicates
seawater depth from bathymetry.
19.For more on WAZ surveys: Camara Alfaro J, Corcoran C,
Davies K, Gonzalez Pineda F, Hampson G, Hill D,
Howard M, Kapoor J, Moldoveanu N and Kragh E:
“Reducing Exploration Risk,” Oilfield Review 19, no. 1
(Spring 2007): 26–43.
20.An autochthonous formation is one that was deposited in
its current location. This salt would be the source of the
shallower salt bodies that moved to their current positions
because of density difference and salt plasticity.
EM_FIGURE 13
Oilfield Review
Spring 2009
Gravity
survey
response
Density
2.1 kg/m3
2.7 kg/m3
Volume = 3,600 m3
3,600 m3
2,800 m3
> Nonuniqueness of gravity survey. A gravity survey responds to the mass of an
anomaly. A solution can propose one object or many, or have different density
and size, as long as the mass and the center-of-mass location for the anomaly
are the same. In this example, all three readings measure the same mass.
The success of this proof-of-concept study
was the impetus for a large-scale multisurvey
MT project that has been active since May 2007
in other key areas in the Gulf of Mexico. For
example, in the Keathley Canyon area, deter-
mining base of salt from seismic data alone was
difficult. Gravity data provided improvement, but
several alternative interpretations could not be
distinguished. By adding MT data and combining
all the information through the MMCI approach,
Distance, km
–80
–40
0
40
2,500
5,000
80
120
SE
Resistivity, ohm.m
NW
Depth, m
were also used, but the results of all the 2D
inversions indicated thinner salt bodies than
shown by data from the Tamara well. The threedimensional nature of the body dictated a 3D
approach to modeling.
The first 3D approach taken by the study
team was to fit the MT data independent of seismic and gravity data. The model started with a
homogeneous and isotropic resistivity below the
seabed. During iterations, each cell resistivity
was allowed to change to match the apparent
resistivity and phase measurements. A smooth
inversion algorithm ensured that the resistivity
changed as smoothly as possible from cell to cell.
Agreement above the main salt body was good
among the WAZ seismic result, the MT fit and the
gravity model. In addition, the interpreted base of
salt is within a few hundred feet of the log-derived
base of salt in the Tamara well—a good match.
However, the gravity model required adjustments
to fit the measured gravity data. Similar gravity
measurement results can be obtained for different configurations (above right). In this case, salt
could be added either to the salt layer within the
model space or to the autochthonous salt—which
was mostly below the maximum depth of the
seismic-velocity volume—or the subsalt formation
densities could be altered to match the result.20
A second approach used the interpreted
seismic survey to provide a starting point for the
shape of the salt body. Resistivity for this a priori
model was initially set at 50 ohm.m inside the salt
body and at 1.2 ohm.m in the surrounding sediments. The inversion changes the values of the
resistivity in the grid blocks to fit the measurement data while preserving the initial model as
much as possible.
The best interpretation used MMCI procedures, incorporating all available information,
including MT, gravity and WAZ seismic data.
Porosities were computed from the WAZ velocity field using local knowledge of the sand/shale
ratio of the sedimentary section, and densities
were computed from the matrix densities of
sand and shale, the density of seawater and the
velocity-derived porosity. Density in the salt was
assumed constant at 2.16 g/cm3.
The 3D model result in the Garden Banks
area had an improved interpretation of the base
of salt compared with that based on seismic data
alone (right). Resistivity data indicated that a
large lobe suggested by the seismic interpretation is not a part of the salt, but belongs to an
underlying formation.
10
1
EM_FIGURE 16
7,500
10,000
> Confirmation by drilling. MT measurements detected a high-resistivity salt intrusion (pink). The
Tamara well, drilled near MT receiver Line 3, provides a point of reference for interpretations of base of
salt. The base-salt interpretation (gray) of the best WAZ data available shows a lobe to the southeast
that is not supported by the MT resistivity data; the 35- to 50-ohm.m area of resistivity (pink) excludes
that lobe from the salt. The 3D MMCI interpretation of seismic, gravity and MT data indicates a base
of salt (white) within a few hundred vertical feet of the base determined from the well gamma ray log
(turquoise). At the base of salt, the well log resistivity (orange) decreases significantly. MT receiver
locations are shown on the seabed (white squares).
15
the analysis team obtained a consistent interpretation of the structure, including the base
of salt (below). In parts of the survey area, the
difference in interpretation of the base of salt
was almost 3,000 m [9,700 ft].
EM Studies Offshore Brazil
Marine MT surveys have also improved depth
imaging in other parts of the world. The Santos
basin, offshore Brazil, contains recent subsalt
discoveries made by Petrobras. High-resolution
Distance, km
128
136
144
152
160
NW
SE
2,500
Depth, m
5,000
7,500
10,000
12,500
Distance, km
128
136
144
NW
152
160
SE
2,500
Resistivity, ohm.m
10
Depth, m
5,000
1
7,500
10,000
MMCI base salt
12,500
Seismic base salt
> Keathley Canyon interpretation. The base of salt is difficult to find in the WAZ seismic section (top).
The best pick based on the seismic data had a thick section to the right of middle (white outline, bottom).
MT resistivity data (colors) add significant new information. Combining seismic, MT and gravity data in
the MMCI evaluation improves the previous interpretations of the base of salt and gives interpreters
greater confidence in their result (yellow dashed line).
16
seismic imaging has mapped the stratigraphy
of hydrocarbon-producing turbidite reservoirs
and the geometries of salt structures, including a thick sedimentary sequence in a syn-rift
structure beneath the salt.21 The lithology of this
sequence was defined by the first discovery well
of the Tupi area. An MT survey northwest of Tupi
confirmed the complex structure and demonstrated
the utility of marine MT surveys to Petrobras.22
To the east of the Santos basin MT survey
just described, Petrobras and WesternGeco performed a marine CSEM survey in the Tambuatá
block of the basin as part of a cooperative
project (next page, top).23 The survey location was
about 170 km [106 mi] south of Rio de Janeiro.
The water depth was taken from bathymetry data,
and processing also included the variation in
seawater resistivity as a function of depth.
The survey used 180 receivers spaced
approximately 1 km [0.6 mi] apart and deployed
on the seabed over known reservoirs. A vessel
towed the source over the receiver lines. Data
acquisition used 0.25- and 0.0625-Hz square-wave
signals that are also rich in odd harmonics of
these frequencies.24
Analysts processed multicomponent electricand magnetic-field responses for all frequencies
in the survey using an advanced workflow based
on instantaneous measures of dipole length,
dipole moment, dipole altitude, feather angle and
dip. The data interpretation proceeded in stages,
starting with generating a background model to
compare with the processed measurements.
Borehole measurements provided information on background resistivities, but the log data
have more detail than CSEM measurements can
discriminate. Thus, analysts reduced the number of layers in the resistivity model to reflect
the resolving power of CSEM, but they ensured
the resampled well logs retained the same CSEM
response as the detailed log-based layering
would have. To determine where the boundaries
had to be placed, both the cumulative resistance
and cumulative conductance were calculated
from the well logs and coupled with stratigraphy. This not only clarified the locations of the
layer interfaces but also determined the resistivities of the layers and the anisotropy caused
by interbedding low- and high-resistivity layers.
Analysts conducted detailed 3D modeling based
on the blocked well log resistivities and based
on model geometries derived from seismic sections without incorporating any reservoirs.
The resulting models generated reference
background fields, which provided a basis to
normalize processed multicomponent field data
at each receiver location.
Oilfield Review
Risking Prospects in the Arctic Frontier
As operators move into increasingly difficult
environments, the Arctic beckons as one of the
last mostly unexploited frontiers. In 2008, the
US Geological Survey (USGS) estimated undiscovered resources north of the Arctic Circle at
14 billion m3 [90 billion bbl] of oil and 47.8 trillion m3 [1,669 Tcf] of gas—of that total the
province west of Greenland and east of Canada
had an estimated 1.1 billion m3 [7 billion bbl] of
oil and 1.5 trillion m3 [52 Tcf] of gas.25
21.Syn-rift refers to events that occur at the same time as
the process of rifting. A syn-rift basin is formed along
with, and as a consequence of, the rifting process. In the
Santos basin, the rifting refers to the early stages of the
separation of the South American and African continents.
22.de Lugao PP, Fontes SL, La Terra EF, Zerilli A, Labruzzo T
and Buonora MP: “First Application of Marine
Magnetotellurics Improves Depth Imaging in the
Santos Basin–Brazil,” paper P192, presented at the
70th EAGE Conference and Exhibition, Rome,
June 9–12, 2008.
23.Buonora MP, Zerilli A, Labruzzo T and Rodrigues LF:
“Advancing Marine Controlled Source Electromagnetics
in the Santos Basin, Brazil,” paper G008, presented
at the 70th EAGE Conference and Exhibition, Rome,
June 9–12, 2008.
24.The strongest harmonics are 0.75, 1.25 and 1.75 Hz
for the 0.25-Hz signal, and they are 0.1875, 0.3125 and
0.4375 Hz for the 0.0625-Hz signal.
25.Bird KJ, Charpentier RR, Gautier DL, Houseknecht DW,
Klett TR, Pitman JK, Moore TE, Schenk CJ, Tennyson ME
and Wandrey CJ: “Circum-Arctic Resource Appraisal:
Estimates of Undiscovered Oil and Gas North of the Arctic
Circle,” USGS Fact Sheet 2008-3049 (2008), http://pubs.
usgs.gov/fs/2008/3049/ (accessed March 31, 2009).
Spring 2009
–46°
–48°
–44°
–42°
Altitude, m
1,365
–22°
N
662
Rio de Janeiro
0
MT
São Paulo
–135
–2,286
B
Selected tow lines were interpreted using a
2.5D inversion. The 2.5D analysis incorporates a
2D geological model and solves for multiple transmitter positions simultaneously, but the sources
and receivers are not confined to the plane of
the geological model. Thus, realistic acquisition
geometries can be simulated (bottom right).
The known reservoir underlying the survey area
appeared in the EM response as a zone of higher
resistivity than the surrounding formations.
As with the MT project farther west in the
Santos basin, the CSEM project also shows promise for adding considerable value in upstream
applications. Both projects underscore the
need for advanced integrated interpretation to
improve the result over individual seismic, well
log and electromagnetic measurements. They
also advance the case for the industry to include
these novel integration paradigms in standard
applications. Petrobras has a technical collaboration agreement with Schlumberger to develop
technology that integrates marine EM into other
technologies for enhanced depth imaging and
reservoir characterization.
a
s
rea
ya
rve
u
S
–24°
CSEM
s
nto
Sa
e
ad
ci
–3,784
Ocean
depth, m
0
km
0
100
miles
–26°
–48°
Tupi area
100
–46°
–44°
–26°
–42°
> Marine MT and CSEM surveys, offshore Brazil. Three lines of receivers for the MT survey (red)
extended offshore toward the southeast and into deeper water. The main line was about 148 km [93 mi]
long, starting about 42 km [26 mi] offshore, and the two adjacent lines were each about 54 km [34 mi]
long. The CSEM survey lines (white) to the east of the MT survey covered the Tambuatá block (red).
The map shows the ground elevation and ocean depth.
0
LTAM1
N
N
0
0
km
10
mi
10
Resistivity,
ohm.m
40
EM_FIGURE 17
10
1
0.4
> Combined analysis for the Tambuatá block. Reservoirs (green and pink outlines, top) identified by
seismic interpretation were the targets of a CSEM and MT study. Receivers (white triangles) were laid in
orthogonal sets, and the CSEM source was towed along the same lines (black). A 2.5D MMCI inversion
based on EM and seismic data resulted in a section color-coded for resistivity, with seismic data providing
texture (bottom). Along tow line LTAM10 N, a 20-ohm.m resistive anomaly (red) is clearly distinguished from
the more-conductive background of about 1.2 ohm.m (green). Seismic results constrained the anomaly
shape—by defined control points (white circles, bottom)—for the data inversion.
17
Prospect without
resistive anomalies
Volcanic flows
Volcanic
flows
Resistivity,
ohm.m
20
18
16
14
12
10
8
6
4
2
Prospects with
resistive anomalies
N
> Prospects with resistive anomalies. Several prospects in a block west of
Greenland were interpreted from seismic data (green outlines). The survey
design placed lines of CSEM receivers (white icons) along the source tow
lines (white lines) above the seismically determined prospects. The CSEM
study distinguished the structures with vertical resistive anomalies (oranges
and yellows) from those with no anomaly (representative locations labeled).
Volcanic flows above the target formation are also identified along the lines.
In this view, resistivities less than 10 ohm.m are not shown. The contour
lines indicate depth of the seismic horizon of the target; each contour
line represents a 100-m [328-ft] depth difference (also represented as the
background color sequence).
EnCana Corporation and its joint-venture exploration targets. For more information on
(JV) partners Nunaoil A/S and Cairn Energy volcanic formations, see “Evaluating Volcanic
have exploration prospects in two blocks in the Reservoirs,” page 36.
Before conducting the CSEM survey,
frontier basin offshore Greenland, 120 to 200 km
[75 to 124 mi] west of the capital city, Nuuk. WesternGeco performed extensive 3D resistivity
The ocean depth over the prospects ranges from modeling over each prospect. This step confirmed
250 to 1,800 m [820 to 5,900 ft]. Geologists that the survey could help define the presence of
believe this area’s rifting and sedimentary-fill hydrocarbon-bearing reservoirs at up to 3,000 m
history is similar to that of the productive North below the seafloor. Synthetic data were used in
Sea basins. However, the nearest well control is forward-modeling and inversion methods. Based
more than 120 km away, and there are no proven on well log data from key distant offset wells,
petroleum systems in the basins. The JV needed a simplified starting model was created that
a way to lessen the risk of drilling dry holes, so a included a reasonably uniform, 1.5-ohm.m clastic sedimentary
section from the seafloor to the
CSEM survey was acquired to help identify potenEM_FIGURE
26
target depth, a deeper layer with 4-ohm.m resistial hydrocarbon-bearing features.26
Sedimentary filling of the basin following rift- tivity extending to the basement, and a 60-ohm.m
ing created a fairly simple geology, with the major basement formation.
As part of this presurvey analysis, geosciencomplication coming from Paleocene volcanic
activity. The volcanic flows are easily identifiable tists optimized the design for target sensitivity,
geologically, seismically and magnetically. These presence of volcanic cover, reservoir proximity to
volcanic rocks are the only known resistive litho- basement and signal waveform, as a few examlogic units in the survey area above basement, ple parameters. This optimization helped the
and they are well separated from the Cretaceous EnCana JV plan a survey covering the vast area
in a cost-efficient way.
18
The survey layout based on this analysis comprised 24 transmitter lines and 182 receivers. The
tow-line geometry generated data from multiple
angles on the receivers. The resulting vertical
resolution was designed to be 50 m [164 ft] for
the Cretaceous targets at depths of 3,500 m
[11,500 ft] below the seafloor.
A high-quality CSEM dataset was obtained in
the summer of 2008. Processing the electric- and
magnetic-field measurements yielded amplitude
and phase responses at each receiver. Starting
with electric-field responses, geoscientists
analyzed these data using a complex 3D anisotropic-resistivity model. The starting geo­metry
used the JV’s seismic interpretation and well log
resistivity information, but no potential reservoirs
were included. The 3D inversions required considerable computation time and interpreter input.27
The results were numerically stable with electrical
models that were geologically consistent.
The inversion process identified resistive
anomalies over 8 of 14 prospects. The team used
Petrel seismic-to-simulation software to visualize
the resistivity volume data for these eight anomalies with geologic, seismic, gravity, magnetic and
marine MT data (left). The results were insensitive to reasonable variations in the starting
model, with each variation converging to a similar resistivity solution.
The known Paleocene volcanic rocks provided another indication that the inversions were
robust and geologically meaningful. Although the
isolated volcanic features were not included in
the starting models for the inversions, the inversion procedure located them correctly.
The EnCana JV’s objective for obtaining the
CSEM study was to improve the assessment of
the probability that the structures were charged
with hydrocarbons. With firm data lacking prior
to the study, the hydrocarbon-charge probability
was indeterminate and the JV assigned it an initial value of 50% for each of the eight prospects.
The team’s analysis increased the probability of
hydrocarbon charging for several features and
decreased it for others.
The prospect with the greatest probability
for hydrocarbon charging displays many of the
characteristics the geoscientists looked for in
the analysis. Its resistivity anomaly conforms
well with the target interval. The CSEM inversion
resistivity within the anomaly increases upward
from 10 ohm.m at the base of the structure to
35 ohm.m at the crest. Finally, the anomaly
base is flat, which could suggest a hydrocarbon/
water contact.
Oilfield Review
> Deployment of CSEM receiver. Each receiver is assembled on the deck using defined deployment
protocols. Then the receiver is hoisted and dropped at a specified location.
EnCana and its partners are now prioritizing
their prospects to identify the most prospective
drilling candidates based on the geology, the
geophysical mapping and the CSEM 3D model
inversion results. The risk for exploration in this
frontier Arctic basin remains great, but CSEM
technology offers promising potential to reduce
dry holes.
Sounding for the Next Generation
Although MT and CSEM surveys have been performed for many years, commercial use of the
marine technology in the E&P industry is relatively new. The industry is still in its infancy in
interpreting this electromagnetic survey data
and combining the information with that of seismic surveys.
26.Umbach KE, Ferster A, Lovatini A and Watts D:
“Hydrocarbon Charge Risk Assessment Using 3D
CSEM Inversion Derived Resistivity in a Frontier
Basin, Offshore West Greenland,” CSPG CSEG CWLS
Convention, Calgary, May 4–8, 2009.
27.Mackie R, Watts D and Rodi W: “Joint 3D Inversion
of Marine CSEM and MT Data,” SEG Expanded
Abstracts 26, no. 1 (2007): 574–578.
28.National Petroleum Council (ed): Hard Truths: Facing the
Hard Truths about Energy. Washington, DC: National
Petroleum Council, 2007. Also available online at http://
www.npchardtruthsreport.org/ (accessed May 5, 2009).
29.WesternGeco regularly performs 3D modeling studies
and offers 3D CSEM inversion including the use of
algorithms in which the MT data are jointly inverted to
help constrain the CSEM inversion.
Spring 2009
The seabed receivers used by WesternGeco
follow the basic design developed by Scripps
Institution of Oceanography, but the devices and
methodologies are continually being upgraded
to improve instrument efficiency and reliability.
In addition to changes in materials used in the
manufacture of the dipoles and magnetometers
and their packaging, new equipment has been
added to the receiver pack, such as a highprecision compass.
The dipole source for CSEM is also under­
going improvement by the industry. Equipment
vendors have worked to refine the timing
synchronization of the source waveform and the
precise positioning of the source antenna.
Major obstacles to marine EM efficiency are
the cost and time involved in data collection.
Seismic measurements over large 3D areas are
efficient because vessels tow multiple receiving
streamers and source array guns. In contrast,
CSEM surveys cover less area because either
sources or receivers, or both, are deployed individually, with receivers remaining stationary
during the survey and then recovered (above).
The development of a purely surface-towed,
deep-reading EM system is likely at the forefront
of R&D activities at many geophysical companies.
The problems are noise inherent in the motion
of sensors through the water and signal attenuation in seawater, which dramatically reduce
the coupling of the source with the seafloor and
the amplitude of the response field. The dipole
antennas are long, and even with the present
seabed configuration, currents can move the
antennas and impact data quality.
The National Petroleum Council (NPC), an
industry body that advises the US government,
studied several advancements related to CSEM,
rating them as highly significant for exploration
activities.28 To secure energy resources for the
future, this expert group identified two improvements in CSEM technologies needed over the
short term. Development of fast CSEM 3D modeling and inversion could reduce the number of false
positives, or resistive anomalies that currently
may be misinterpreted as a commercial petroleum
response. These include hydrates, salt bodies and
volcanic lithologies. The second short-term goal is
integration of CSEM with structural information
from seismic surveys to improve the resolution
of the EM data. As discussed in the case studies
in this article, this work is currently underway
through efforts such as the MMCI method.29
Over a longer term, the NPC experts also
rated advancing the realm of CSEM studies into
shallow water, onshore, and deeper formations
as highly significant. The signals in shallow
water and onshore are much noisier than in
deep water because of the air wave. Signal
strength now limits the depth of the CSEM
surveys, but the NPC group saw that developments leading to evaluating deeper formations
would extend the application to new basins.
Alternative acquisition geometries might play a
role in ultradeep reservoirs.
The term “electromagnetic sounding” is not
yet commonly heard in the E&P industry, but
impressive results from this generation of tools
and interpretation methods have already sent
a clear message. With commercial success will
come further advances in technology and a wider
variety of applications. —MAA
19
Near-Surface Electromagnetic Surveying
The E&P industry typically focuses on deep formations, but frequently the nearsurface layers also need to be evaluated. Land-based electromagnetic surveys
provide insights into this often complex zone. The interpreted resistivities of these
layers help map and define features for applications as diverse as seismic studies
and aquifer delineation.
Mohamed Dawoud
Environment Agency–Abu Dhabi
Abu Dhabi, UAE
Stephen Hallinan
Milan, Italy
Rolf Herrmann
Abu Dhabi
Frank van Kleef
Dubai Petroleum Establishment
Dubai, UAE
Oilfield Review Spring 2009: 21, no. 1.
Copyright © 2009 Schlumberger.
For help in preparation of this article, thanks to Marcus
Ganz, Houston.
1. The principle describing the changing magnetic field
is Faraday’s law of induction. Including the sign of the
induced current is Lenz’s law. The converse principle
involving a changing current or electric field is Ampère’s
law. These laws are included in Maxwell’s equations.
20
The near surface of the Earth is a complex place,
the result of dynamic action by wind, water and
other forces of nature. In those first tens of
meters below the surface, the jumbled detritus of
weathering is gradually buried. The near-surface
layers, like those below, vary in resistivity according to their mineral and fluid compositions. This
property allows their investigation using electromagnetic (EM) surveys.
Often, surveys are performed using an artificial source of EM radiation, rather than the
magnetotelluric (MT) radiation resulting from
the interaction of the solar wind with the Earth’s
magnetosphere. On land, there are two general
methods of controlled-source electromagnetic
(CSEM) measurement for generating the signal
and detecting the response. The grounded-source
method requires burial of source and receiver
electrodes in electrical contact with the earth.
The inductive-source method uses a current loop
on the surface to induce a variable magnetic
field, and the same or another loop to detect the
response signal.
The grounded-source method is efficient and
sensitive to horizontal resistive targets because
the electric field has a vertical component.
The grounded receivers measure the response
electric field; response magnetic fields are also
measured to provide control during modeling.
On land, however, the surface conditions must
be suitable for creating and maintaining electrical contact. This prerequisite excludes the
practical application of this method in large, arid
dunes—sand grains are nonconducting. But in
some areas, contact can be improved by drilling
patterns of shallow holes for source and receiver
electrodes and wetting the soil as the hole is
refilled. Deeper investigation into the Earth
requires stronger current sources, among other
factors, and the high contact resistances on land
mean this requires high-voltage systems to drive
that current.
The inductive-source method does not require
electrical contact, since the current loop generates a magnetic field through a time-varying
signal. This field generates a response electric
field, but because the electric field is largely
horizontal, the process is not as efficient for
imaging horizontal layers of resistive hydrocarbons as the direct injection of current using the
grounded-source method. Again, both the electric
and magnetic response fields can be measured
using the inductive-source technique. Coils
for the current loops are square and for nearsurface investigation range from about 10 to 300 m
[30 to 1,000 ft] on a side. Far larger loops have been
used for deeper, but low-resolution, investigation.
A companion article (see “Electromagnetic
Sounding for Hydrocarbons,” page 4) describes
the basic physics of the EM interaction with the
Earth and discusses marine EM studies. It also
covers MT in detail, because the objectives of
those studies are similar for land and marine environments. This article focuses on investigations
Oilfield Review
using the inductive-loop method for nearsurface imaging, illustrated by two WesternGeco
cases from the United Arab Emirates. One study
mapped an aquifer in Abu Dhabi for a waterstorage project. The second determined nearsurface resistivity variations in Dubai sand dunes,
providing valuable input for making static corrections in a seismic survey of the area.
Stacking Time Sequences
Maxwell’s equations describe the basic physics
behind the interplay of electric and magnetic
fields in a time-varying current loop. A current
loop generates a magnetic field. If the current
changes, the induced field also changes. The
opposite is also true: Changing the flux of a magnetic field within a conducting loop induces a
changing current.1 A simple way to generate such
a current is to move a magnet toward or away
from a wire loop. The movement changes the flux
through the loop, inducing a current. This current induces a response magnetic field oriented
to oppose the change in flux through the loop
caused by the moving magnet.
No actual magnet is required for this effect
to take place. One coil with an imposed timevarying current sets up a time-varying magnetic
LAND EM_OPENER
Spring 2009
21
> Basic induction loops. An alternating current
passing through a set of coils (blue) induces a
cyclic magnetic field. When this field passes
through a second set of coils (red), it induces
a cyclic current in that circuit. Thus, energy
passes from one circuit to another without a
direct electrical contact. This is the basis for a
transformer, which is a device that converts from
an input voltage to a different output voltage by
having different numbers of loops in the two coils.
Ramp time
Batteries and
signal electronics
Transmitter current and
primary magnetic field
Ramp time
Time on Time off Time on
field in response. Current is induced in a second
coil positioned close enough to experience the
changing flux. This is the configuration of a transformer (left). Energy passes from one circuit to
the other through the changing magnetic field.
One method that uses inductive measurement for evaluating the near surface is a
time-domain electromagnetic (TDEM) survey.
A conducting loop of wire set out in a square
on the surface of the Earth acts as the first
coil, and the second loop forms in conducting formations of the Earth itself. The primary
magnetic field from the transmitter loop generates horizontal currents, called eddy currents,
immediately beneath the loop. These currents
induce a response field that can be detected at
a surface receiver loop, but this field also travels
farther into the subsurface, generating progressively weaker eddy current loops with larger
radii and smaller response fields (below).2
Source current loop
Measurement
period
Induced electromotive force
in nearby conducting layers
Eddy currents
induced by
field change
LAND EM_FIGURE 03
Depth
Response magnetic field
induced by eddy currents
Receiver-coil voltage from
response magnetic field
Time
Secondary
magnetic field
Eddy currents
at later times
> TDEM inductive method. A large square loop is placed on the surface at the sounding site (top right).
Passing a current pulse through this loop generates the primary magnetic field. The field induces
a secondary, eddy current loop in the ground, as described by Faraday’s law (middle right ). This
secondary current induces a response magnetic field, which can be recorded by a receiver loop on
the surface. In this case, the same loop is used as source and receiver. The primary field decays with
depth into the ground, generating response fields from each subsequent depth (bottom right ). The
timing of the signals and response fields is also shown (left).
22
The transmitter and receiver loops can
be the same wire coil when an appropriate
time sequence of current steps is applied, or
coaxial but separate coils in a typical setup.
In all inductive-loop TDEM surveys, the time
sequence begins by turning the current on to a
constant DC value. Sufficient time elapses for
transient responses in the subsurface to decay.
Then, electronics shut the current off in a rapid,
controlled ramp, inducing a known electromotive force in the immediate subsurface. The
transient electromotive force generates eddy
currents, producing a secondary magnetic field
that decays with time. The secondary field is
detected by the receiver coil. After enough time
has elapsed, the sequence repeats with opposite polarity. Stacking many repeated responses
improves the signal/noise ratio.
As the eddy currents move progressively
deeper into the subsurface, the response field
contains resistivity information about deeper
layers. Fundamentally, the resistivity variation
with depth determines the rate of decay of the
transient; higher conductivity results in slower
decay. Inversion of the stacked data from a
TDEM sounding reveals the distribution of nearsurface resistivity.
The measured signals are very small, so land
surveys must consider and avoid, if possible, any
sources of noise. Electric trains, power lines,
electric fences, buried utility cables, pipelines
and water pumps distort the local measurement;
large temperature variations and wind impact
stability; and variations in soil moisture and permeability affect uniformity.3
TDEM is not commonly used directly in oil
and gas exploration, although it has utility in
evaluating surface statics for seismic studies.
The method is applied widely for exploration
within the mining industry, employing both terrestrial and airborne sources. It is also a tool for
environmental and water-resource management,
as shown in the first of the following case studies
from the Middle East.
Sounding for Water Storage
A recent land-based EM application helped
to locate potential water-storage sites in the
UAE.4 The Environment Agency–Abu Dhabi
(EAD) is managing a study for the government
to evaluate storage plans for 30 billion British
imperial gallons [136 million m3, 36 billion
galUS] of fresh water in the northeast area of
the Emirate.5 The country wants a freshwater
reserve for emer­gency periods and to meet peak
Oilfield Review
summer demands. The water for this aquifer
storage and recovery (ASR) project will be transported via pipeline from a water-desalination
plant in the Emirate of Fujairah.
EAD retained Schlumberger to identify and
test a potential ASR site. This involved defining
the subsurface storage zone and surrounding formations, aquifer thickness and related hydraulic
parameters. Schlumberger selected a preferred
site and constructed three pilot wells, which
were tested to determine the aquifer’s potential.
Geologic studies of the area found that
deep-seated faults had been reactivated in the
Late Tertiary Period by the northeastward displacement of the Arabian Peninsula, resulting
in a series of tightly folded faults. The overlying
layer of Quaternary Period sediments, consisting of eolian sands and alluvium, were primarily
deposited along the reactivated faults and in the
synclines between them, giving the sediment
thickness a directional bias. The directionality
can influence groundwater flow, creating a preference for flow parallel to the structure.
Schlumberger evaluated this geologic structure in 2006 during the drilling, logging and
testing periods. Logs from the pilot wells indicated a resistivity contrast between the targeted
sand-and-gravel aquifer and the underlying clayrich layer. Because TDEM data are useful for
aquifer characterization, the evaluation over the
ASR site included a survey to define the lateral
extent of the aquifer—necessary to compute the
potential water-storage volume. In a time-domain
survey, the apparent resistivity of the underlying
formations is determined from the time variation of the electric and magnetic response fields.
For the ASR study, the same set of coils placed
on the surface was used for both current and
receiver loops. The survey covered a 6- by 7-km
[3.7- by 4.4-mi] area.
A 1D Occam inversion at each receiver yielded
resistivity input for constructing a 3D model.6 The
maximum depth obtained by the inversion was
about 250 m [820 ft]. Resistivity logs from the
three wells were compared with the inversion
results at adjacent sounding locations (right).
The top of the underlying clay unit is clear in the
eastern part of the survey area but less obvious in
the western part.
2. Nabighian MS: “Quasi-Static Transient Response of a
Conducting Half-Space—An Approximate Representation,”
Geophysics 44, no. 10 (October 1979): 1700–1705.
3. Constable SC, Orange AS, Hoversten GM and Morrison HF:
“Marine Magnetotellurics for Petroleum Exploration, Part I:
A Sea-Floor Equipment System,” Geophysics 63, no. 3
(May–June 1998): 816–825.
Spring 2009
Sounding S049
Depth, m
TDEM Resistivity
ohm.m 100
1
Well SWS17
1
Log Resistivity
ohm.m 100
Sounding S071
1
Well SWS15
Log Resistivity
TDEM Resistivity
ohm.m 100
ohm.m 100 1
Sounding S013
1
Well SWS16
TDEM Resistivity
Log Resistivity
ohm.m 100 1
ohm.m 100 Depth, m
–280
–280
–270
–270
–260
–260
–250
–250
Water table
–240
Water table
Water table
–240
Bottom of aquifer
–230
–230
Bottom of aquifer
–220
–220
–210
–210
Bottom of aquifer
–200
–200
–190
–190
> Comparison of TDEM soundings with resistivity profiles from logged wells. The TDEM resistivity
measurements from soundings correlate closely with the resistivity logs from adjacent wells. The
soundings at S049 and S013 show reasonable correlation for the contact between the aquifer and
the clay-rich unit below (violet), which is true for most other soundings in the eastern part of the
investigated area. The contrast is not as clear at S071, where the more resistive lower unit does not
provide a sufficient contrast for the TDEM measurement. This trend follows for most of the soundings
in the western part of the survey. The top of the water table (blue dashed line) was determined from a
map of the water depth by interpolating between wells in the area.
4. For more on water storage: Black B, Dawoud M,
Herrmann R, Largeau D, Maliva R and Will B: “Managing
a Precious Resource,” Oilfield Review 20, no. 2
(Summer 2008): 18–33.
5. A British imperial gallon is equivalent to 1 galUK.
6. An Occam inversion is a smooth inversion that does not
predefine the number of layers.
23
2 km
80 m
2 km
N
> Resistivity discontinuity in the bottom of an aquifer. The discontinuity is a band with 15- to 20-ohm.m
resistivity (yellow and green), which contrasts with the 1- to 10-ohm.m resistivity (blue and violet)
elsewhere in the resistivity cube. The resistivity of the clay at the bottom of the aquifer (violet) to the
east of the discontinuity is lower than that to the west. This discontinuity roughly aligns with a thrust
fault (tan) that was identified at about 3,000 m [9,800 ft] by a seismic interpretation. Outcrop studies
performed to the south of the survey area (not shown) support the surface expression of the fault
being slightly west of the deeper seismic interpretation, and those observations are consistent with
the location of the near-surface resistivity discontinuity. The syncline axis (blue) from the seismic
interpretation, another thrust fault (purple) and some wells are also shown. (Seismic lines adapted
from Woodward and Al-Jeelani, reference 7.)
LAND EM_FIGURE 24
The TDEM data clearly show a discontinuity in the resistivity distribution in the clay unit
(left). There is also a difference in resistivity
between the eastern and western compartments;
the western part exhibits significantly greater
resistivity at a given depth. The anomaly aligns
with a seismically mapped thrust fault.7 The
seismic interpretation was based on a survey
acquired in the early 1980s and reprocessed in
1992 to highlight the shallow structures.
The shallow units above the anomaly are
expected to show some structural complexity; they
exhibit rapid variations in saturated thickness and
possibly no saturated thickness in some areas. On
the eastern side of the discontinuity, the horizontal extent of saturated thickness is suitable for an
ASR unit. The western side of the discontinuity
shows some potential, but has a larger risk: The
interpretation in that area has a greater uncertainty because of the poor resistivity contrast
between the saturated sands and the underlying clay. The discontinuity in the clay formation
should not be seen as a complete hydraulic barrier
in the shallower aquifer layer. Paleochannels or
tear faults—those striking perpendicular to the
overthrust fault—are expected to provide preferential flow paths from east to west across the line
of the discontinuity.
This TDEM study suggests that around 4 billion imperial gallons [18 million m3, 4.8 billion
galUS] of water can be stored at this site, giving
it the capacity for daily production of more than
20 million imperial gallons [91,000 m3, 24 million
galUS] for 200 continuous days.
Mapping the Dunes
Within the same regional setting as the waterstorage site, a 2D seismic survey was conducted
for Dubai Petroleum Establishment (DPE). The
same clay layer forming the base of the storagesite aquifer is present in this area; it provides
a marker for the base of the weathered surface
layer. The depth of the clay varies across the survey area, and lines of dunes add local variation
7. Woodward DG and Al-Jeelani AH: “Application of
Reprocessed Seismic Sections from Petroleum Exploration
Surveys for Groundwater Studies, Eastern Abu Dhabi,
UAE,” paper SPE 25538, presented at the SPE Middle East
Oil Show, Bahrain, April 3–6, 1993.
8. An uphole is a shallow well used to determine nearsurface velocities for a seismic survey.
9. Colombo D, Cogan M, Hallinan S, Mantovani M,
Vergilio M and Soyer W: “Near-Surface P-Velocity
Modelling by Integrated Seismic, EM, and Gravity Data:
Examples from the Middle East,” First Break 26
(October 2008): 91–102.
> Desert terrain at the ASR site. The blue cable is part of a TDEM sounding
loop. The building houses a submersible pump and water tank for the nearby
wells (blue caps) of the ASR project.
24
Oilfield Review
240
North
Resistivity,
ohm.m
South
150.0
200
92.8
UH-08
UH-09
Elevation, m
160
57.5
UH-06
35.6
22.0
120
13.6
8.4
80
5.2
40
3.2
2.0
0
4
5
6
7
8
Distance, km
9
10
11
> Soundings along a seismic line. Interpolation between sounding points yields a detailed 2D model of
resistivity along a seismic line. The sharply layered model at each sounding point (filled black square)
is shown as a narrow column (top). Uphole sites (UH-09, -06, -08) contain domains of constant velocity
in the weathered layer (yellow stippling) of about 1,400 m/s [4,600 ft/s] and in the underlying clay and
limestone (gray stippling) of greater than 2,000 m/s [6,560 ft/s]. The variation in properties both within
the dunes and in the lower layer is evident along the entire seismic line (bottom). The higher resistivity
of the bottom layer (south end of the seismic line) indicates a relatively clay-poor region.
The survey comprised 505 sounding sites
to the depth of the weathered layer. Sand dunes
generally exhibit low seismic velocity, and defin- using square loops of 50 m [164 ft] on a side,
ing the velocity variation and thickness of the except for a few sites that used 75-m [246-ft]
surface layer is crucial for deriving a long-wave- square loops for deeper penetration.9 Spacing
length static correction for the seismic data.
between sounding points was generally about
The seismic crew drilled several upholes to 1,000 m [3,280 ft]; GPS was used to position the
log the velocity of the surface and underlying sites. The effective time for decay ranged from
clay layer.8 The upholes typically were at seismic- 0.01 to 10 ms; the pulse repetition rate was 6.3 Hz.
Given the subhorizontal nature of the zone
line intersections and, for practical reasons,
away from the higher dune crests. However, this of investigation and its shallow depth compared
pattern often does not sample the near-surface with the TDEM station spacing, 1D resistivity
variations found in sand dune areas, so finer sam- inversion modeling was selected for the analyTwo 1D resistivity inversion methods were
pling was desired. DPE elected toLAND
use aEM_FIGURE
TDEM sis.MARGHAM
resistivity survey to map the area because it applied. The first one incorporated about 15 laywould be more cost-effective than drilling more ers extending to a depth of about 200 m [650 ft].
upholes and would avoid additional drilling on Layer thickness increased logarithmically with
depth. Resistivity was a free parameter, and
the environmentally sensitive dunes.
this inversion yielded a detailed, smooth variation of resistivity.
Spring 2009
The detailed fit provided a starting point for
the second inversion. Termed a layered fit, it used
the minimum number of layers required to fit the
data to less than 5% root-mean-square misfit. This
was typically two to five layers. Analysts selected
the starting definition of these layers from the
detailed fit. The layered model generated stronger resistivity contrasts than the detailed one.
Interpreters created a 2D model with grid
blocks 200 m wide and 5 m [16 ft] deep along
a seismic line. They used the sounding sites
along the seismic line to evaluate the surface
layering. Model resistivity values were obtained
by interpolating between the 1D smooth inversions at those sounding sites (left). The result
is a detailed description of the location of the
bounding clay layer at the bottom of the lowvelocity zone. The resistivity data were not
calibrated to seismic velocities.
The seismic processing team used these maps
during the estimation of the surface static corrections. The velocities for the surface zone were
interpolated from velocity measurements taken
at the upholes. The TDEM approach gave the
seismic interpreters a geologically consistent
way to remove the effects of the laterally varying sand velocities. The resistivity analysis also
highlighted variations within and below the lowvelocity weathered layer.
Sounding Deeper
Both case studies used TDEM methods to examine near-surface features. However, because
using inductive-source techniques is inefficient
for defining deep targets, the method is not the
exploration tool of choice for examining deeper
structures. The industry is improving techniques
for using the alternative, grounded-source method
to inject current into the earth.
The source for the grounded method must
be able to inject a large current at a voltage
that is sufficient to overcome contact resistance
in areas where the soil is dry. This combination
has been difficult to achieve. The method using
direct injection of current is more sensitive to
resistive targets, making it a more likely method
than the inductive-loop option to provide a direct
hydrocarbon indication. —MAA
25
A Plan for Success in Deep Water
Deepwater oil and gas are conventional resources in an unconventional setting;
operations are notable mainly for their high risk and high reward. Because of the
scope and complexity of projects beyond the continental shelves, the difference
between success and failure often hinges on good planning.
Adwait Chawathe
Umut Ozdogan
Chevron Corporation
Houston, Texas, USA
Karen Sullivan Glaser
Houston, Texas
Younes Jalali
Beijing, China
Mark Riding
Gatwick, England
Oilfield Review Spring 2009: 21, no. 1.
Copyright © 2009 Schlumberger.
Petrel and SMC are marks of Schlumberger.
For help in preparation of this article, thanks to Robert Clyde,
Debra Grooms, Scott Scheid and Drew Wharton, Houston;
Nils A. Solvik, Framo Engineering, Bergen, Norway; Merrick
Walford, Pau, France; and Jeremy Walker, Rosharon, Texas.
1. Curole MA and Turley AJ Jr: “Mars Debottlenecking
Project,” paper SPE 69199, presented at the SPE Annual
Technical Conference and Exhibition, New Orleans,
September 27–30, 1998.
2. Wetzel RJ Jr, Mathis S, Ratterman G and Cade R:
“Completion Selection Methodology for Optimum
Reservoir Performance and Project Economics in
Deepwater Applications,” paper SPE 56716, presented at
the SPE Annual Technical Conference and Exhibition,
Houston, October 3–6, 1999.
Amin A, Riding M, Shepler R, Smedstad E and
Ratulowski J: “Subsea Development from Pore to
Process,” Oilfield Review 17, no. 1 (Spring 2005): 4–17.
In 2001, while constructing its massive deepwater Mars tension-leg platform, Shell concluded
the plans for the facility required major adjustments. The changes were needed to take
advantage of just-introduced advances in well
completion technology that would boost production beyond the original design parameter of a
maximum 1,750 m3/d [11,000 bbl/d] of oil per
well. Fortunately, because the Mars team
comprised experts from many project disciplines,
it was aware of overall project parameters, and
Shell was able to implement the necessary
changes in the construction yard before the giant
floating platform sailed off.1
This Shell experience clearly demonstrates
the case for planning practices that consider the
development as a whole—from subsurface
modeling to completion strategies to first oil and
beyond. By considering every aspect of development at the planning stage, operators are more
likely to find they still have viable options before
or during deployment and operational phases.
Such flexibility may become critical as
new information about a reservoir, the available
Load and interpret
LWD data
during drilling.
Plan a new
well based on an
up-to-date 3D
reservoir model.
Make live
updates to the
3D model with
the new data.
Plan follow-up steps
(completion, next well)
with latest subsurface
views at hand.
> Iterative processes. Project plans are fine-tuned constantly as a field is
developed. Beginning with a 3D reservoir model, drilling experts select
drilling locations, target zones and trajectories. Model updates take place
as wireline and LWD measurements are acquired, enabling changes that
reflect the most recent information. This process repeats throughout the
development drilling program.
26
Oilfield Review
technology or any number of related parameters
becomes apparent during project commissioning,
drilling, completion or production. The penalty
for inefficient or incomplete planning could be
an inability to change designs or accept
compromises in critical elements such as well
location, completion type, well size or field
configuration once work has begun. The result
could be a less-than-optimal development, which
almost always translates to negative outcomes
such as reduced ultimate recovery, lower productivity rates and significantly higher capital and
operating costs.
Adoption of proper deepwater project
planning practices will probably require more
cultural change than technological innovation.
This is because the upstream industry traditionally treats the various operations that make
up field development as separate tasks performed in series by experts working
independently. More importantly, oil industry
operators, contractors and service companies
have a long history of working from out-of-date
plans or those too general to be of much use. This
mode of operation forced them to deal with
individual problems in a reactionary mode rather
than planning in advance for potential problems
and possible solutions. In deep water, where
stakes are high and the time between concept
and first oil can be as much as a decade,
segregation of responsibility and use of static
plans that cannot be adjusted to respond to
changing circumstances are no longer options. It
is, therefore, essential that experts of all
disciplines take a longer, more integrated view.
For example, it is usually desirable to begin a
drilling program by considering the type of
completion required to best exploit the reservoir.
While this reservoir-driven approach is common,
the ultimate goal of a thought-out plan is the
profitability of the overall project. This being the
case, the well’s productivity becomes only one
factor in the selection of completion type.2 Other
considerations include cost-sensitive factors and
the risks associated with overall project expense,
interventions, well longevity, sand production
and flow assurance.
In recent years, many drilling and completion
engineers have made progress toward a more
integrated approach. But deepwater project
planning requires extension of that practice
beyond well construction to connect the entire
enterprise—from early exploration to final
production—while using each step in between to
refine the process.
Spring 2009
O
Therefore, a typical deepwater project plan
not only includes each of the following elements,
but also considers their influence on each other:
• subsurface reservoir model
• drainage strategy and bottomhole locations
• field development plan
• well design engineering and technology
• intervention methodology
• pipeline and platform design and installation.
From a practical standpoint, planning begins
at the exploratory stage. Once the reservoir has
been characterized through seismic data
interpretation, petrophysical information about
the target formation is gathered during the
drilling process using such tools as wireline logs,
LWD operations and dynamic testing (previous
page). The resulting combination of data about
27
> Deepwater drilling units. The complex, dynamically positioned units capable of
drilling in extreme water depths are relatively rare. The cost to build and outfit
them has been reported to be as high as US $750 million, and despite a recent
spate of new construction, demand outstrips supply. The investment needed to
drill in water depths greater than 1,800 m [6,000 ft] is reflected in the unit lease
rate—often US $1 million per day. (Photograph courtesy of Transocean Ltd.)
the reservoir matrix, fluid properties and
producibility serves as a basis for the many
decisions that will be made about the field
throughout its life.
One such technical decision is well trajectory
within a reservoir. Since efficient reservoir
drainage—using as few wells as possible to access
and produce the maximum volume of oil and gas
at the most advantageous rate—is key to
profitability in deepwater project planning, well
angle and reach are decided early in the process.
However, in a holistic approach these calculations
must include more than maximum reservoir
exposure—the most common driver for the use of
extended-reach wells. Completion designs for
these deepwater wells must also consider the
optimal flow rate for the long term, which requires
a balance between maximizing ultimate recovery
through prudent production practices and
maximizing immediate returns through high
flow rates.
These decisions both drive and are driven by
available drilling technologies and their
associated completion configurations. Operators
3. Perforation density is the number of holes per linear foot
of borehole, reported as shots per foot (spf). Perforation
phasing refers to the angle at which the perforations are
offset from the toolstring axis. Thus, in a 30° phasing,
each perforation is separated by 30°.
4. Iledare OO: “Profitability of Deepwater Petroleum
Leases: Empirical Evidence from the US Gulf of Mexico
Offshore Region,” paper SPE 116602, presented at the
SPE Annual Technical Conference and Exhibition,
Denver, September 21–24, 2008.
5. Mouawad J and Fackler M: “Dearth of Ships Delays
Drilling of Offshore Oil,” http://www.nytimes.com/
2008/06/19/business/19drillship.html (accessed
December 11, 2008).
Ultradeep water depths are considered by industry to be
those beyond about 1,800 m [6,000 ft].
6. Lifting cost is the operator’s total financial outlay for
bringing oil and gas to the surface and is generally
calculated in US dollars per barrel of oil equivalent.
7. Cullick AS, Cude R and Tarman M: “Optimizing Field
Development Concepts for Complex Offshore Production
Systems,” paper SPE 108562, presented at the SPE
Offshore Europe Oil & Gas Conference and Exhibition,
Aberdeen, September 4–7, 2007.
8. For more on inversion: Barclay F, Bruun A,
Rasmussen KB, Camara Alfaro J, Cooke A, Cooke D,
Salter D, Godfrey R, Lowden D, McHugo S, Özdemir H,
Pickering S, Gonzalez Pineda F, Herwanger J,
Volterrani S, Murineddu A, Rasmussen A and Roberts R:
“Seismic Inversion: Reading Between the Lines,”
Oilfield Review 20, no. 1 (Spring 2008): 42–63.
28
may choose to develop their fields through a few
extended-reach wells, numerous vertical wells,
multilateral wells, intelligent wells or some
combination of these and other scenarios.
Throughout exploration, assessment and
development drilling, virtually all development
parameters—such as well location, completion
type and flow rates—may be altered as the
reservoir model is refined by information
gathered from new wells.
Real-time data and the actions taken in
response to confirmation of, or changes to,
assumptions about a reservoir are used throughout
the life of the field. Improved knowledge about
rock stresses, for instance, impacts such vital
details as perforation density and phasing and
choice of sand control system.3 Updated models of
porosity, permeability and fluid characteristics do
not just shape the drilling and completion
program, they are also fundamental inputs to key
decisions about flow assurance and facility design.
Modeling fluid parameters over the life of a
deepwater project is itself a far more complex
undertaking than for onshore or shallow-water
fields. In deep water, economics dictate that
multiple reservoirs—often with different and
changing characteristics—share facilities, pipelines and other infrastructure.
Because of the complexity-driven risk and
large reserves potential involved, deepwater
developments are economically more sensitive
than most other E&P endeavors. According to an
analysis of Gulf of Mexico lease data conducted by
the US Minerals Management Service, both risk
and reward rise significantly with increased water
depth.4 Given that relationship, it is clear that
operations planned in waters deeper than 3,050 m
[10,000 ft] have increased the stakes to such a
level that even seemingly minor missteps may
conspire to quickly overwhelm project economics.
Current extreme costs in the deepwater play
are rooted in two major outlays: the costs of
facilities, pipelines and other infrastructure and
high rental rates—dayrates—that contractors
must charge to make a reasonable return on
their investment in rigs. For a rig capable of
operating in ultradeep water, that investment is
about US $500 million in construction costs alone
(above left).5 As a consequence, the spread
rate—dayrate plus all other required equipment
and services for any given operation—for such a
drilling vessel is approximately US $1 million per
day, or nearly US $42,000 per hour. In deep water,
well construction typically accounts for about
50 to 60% of the total lifting costs, split evenly
between drilling and completion.6 A field infrastructure often requires capital expenditures of
more than US $1 billion.7
Oilfield Review
Surface Seismic
Data Optimization
• Seismic cube, line
• Seismic gathers
• Seismic velocities
Reservoir
Characterization
Well Services
• Fracture design
• Microseismic
Petrophysicists
Geologists
Fracture
Characterization
Drilling Measurements
• Seismic while drilling
• Sonic
Inversion specialists
Geophysicists
Rock physicists
Reservoir engineers
Logs
• Dipole sonic
• Density
• Geology, reservoir
• Borehole seismic
Reservoir Operations
Pore-Pressure
Prediction
Geohazards
• Field development planning
• Reservoir modeling
• Mechanical earth modeling
• Wellbore stability and planning
• Production enhancement
> Integrated reservoir characterization workflow. Schlumberger uses a multidisciplinary team with expertise in petrophysics,
geology, geophysics, inversion and rock physics to evaluate new reservoirs. Through attribute analysis and inversion of seismic
traces and log data, the team derives the acoustic and elastic parameters required to discriminate hydrocarbons, estimate rock
properties and characterize fracture systems. Collaboration with geomechanics and field development planning experts results
in a predrill seismic assessment of drilling hazards and computation of a pore-pressure cube by high-grading the surface
seismic velocities. As well data are collected by logs, drilling measurements and well services, they are used to refine initial
reservoir characteristics.
While these absolute costs are significant,
well construction costs typically include 24 to
27% nonproductive time (NPT)—a time loss that
is aggravated by working in reactive mode.
Subsea architecture and installation of production facilities routinely incur 30 to 35% NPT. It is
clear, given the investments involved, that these
percentages represent a significant amount of
money and underscore why minimizing NPT is a
key goal of operators.
This article looks at the many obvious and notso-obvious parameters considered in deepwater
project planning. In some instances, the project is
the entire venture from seismic survey to
abandonment. In others, the project is more
specific—testing, cementing or some other major
component of a larger operation. Deepwater case
histories from the Gulf of Mexico, West Africa and
the North Sea demonstrate why and how
deepwater operators and service providers must
take a long-term integrated view in this challenging environment.
Spring 2009
From the Bottom Up
Reservoir drainage strategy essentially drives
deepwater projects. Engineers must have a
thorough understanding of the reservoir before
they can optimize well locations and make
informed decisions about wellbore size, sand
control, artificial lift, perforation and all other
facets of a drilling, completion and production
program. They must also make decisions as to
wellhead type, pipeline and manifold configuration and the type of host platform to be used.
As with all modeling systems, a wrong first
step in deepwater project planning endangers all
decisions that follow. In the case of deepwater
field developments, the earliest planning steps
occur at the seismic-to-reservoir simulation stage.
Good reservoir simulators have been available for more than 20 years, but for much of that
time, preparing input and analyzing the results
of reservoir simulation were difficult tasks. A
lack of integration between pre- and postprocessing tools and the need for many manual,
time-consuming data transfers and dataformatting steps frequently caused operators to
avoid what was often cumbersome simulation
work even while making critical business
decisions about their developments.
Today’s software overcomes this hurdle to
best practices by clarifying the intersections
of seismic data and reservoir modeling.
Geoscientists now interpret and quantify
reservoir properties using processes that
integrate seismic data with all available
petrophysical data through inversion and
reservoir modeling. An important prerequisite to
this process is the conditioning of both seismic
and reservoir data for inversion and integration
into a reservoir characterization workflow. Well
logs and vertical seismic profiles provide
calibrated earth properties that enable well-tie
processing and testing of inversion models.8
The objective of this exercise is to
characterize the reservoir and obtain estimates
of reservoir lithology and fluid distributions by
estimating rock properties such as porosity, sand
and shale volumes, density and water saturation
(above). One operator drilled three Gulf of
Mexico deepwater wells with mixed results
29
> Taking the proper angle. Seismic gathers that are flat out only to an angle of 30° (left ) are of poorer
quality than the ones made after data optimization, which makes the gathers flat out to an angle of
54° (right ). Optimization resulted in a good density determination.
before turning to Schlumberger to manage the
risk and cost of drilling subsequent wells. While
some wells were successful, others drilled in
promising formations found only residual gas
instead of commercially viable accumulations—
a difficult distinction to make from seismic
amplitude data alone.
The solution was to calculate density through
an amplitude variation with offset (AVO) process.
To determine the percent gas in a reservoir, a
geoscientist must first model the densities of
both the rock and fluids. Doing so requires the
ability to see the variation in the far angles of the
seismic gathers. This information in turn allows
interpreters to perform a three-term AVO
inversion calculation that includes density. The
outputs of the three-term inversion process are
relative acoustic impedance, shear impedance
and density volumes.
Shared Earth Model
Planning
The lithology and fluid types, using properties
derived from log measurements in nearby wells,
along with varying percentages of gas are
modeled to see the effects on the seismic AVO
signature. The workflow process then compares
the modeling results with the inverted seismic
volumes—acoustic impedance, shear impedance, density volumes and Poisson’s ratio—and
the results are the basis for generating lithofacies
and fluid saturation prediction volumes. Using
these probability volumes, with statistical
uncertainties, geoscientists can give better
predictions of reservoir quality and distribution.
These calculations require data to a fairly
high-angle range in the seismic gathers. In this
instance, the Schlumberger team was able to
extend the usable seismic angle from 30 to 54°,
enabling accurate density determinations
(above). Calibration of the seismically derived
Execution, Real-Time Monitoring, Replanning
Evaluation
G&G,
RE, PE
Scope
risk
Offset
wells Key historical
information
Detailed
engineering
Actual
versus
plan
Drilling
operations
TD
Final
well
report
Lessons
learned
Replan
> Refining the plan. A shared earth model is the basis for collaborative well planning. From this
starting point, geoscientists (G&G), reservoir engineers (RE) and production engineers (PE) define
subsurface targets. Using offset well correlations or simulations from reservoir analysis software, this
model delivers profiles of pore pressure and rock fracture strength with depth. The drilling engineer
uses these inputs plus well objectives for conceptual design or scoping. The output assists in detailed
engineering decisions on rig selection, drilling engineering technical risk and probabilistic time and
cost estimates. Operators undertake drilling operations and modify them based on actual drilling
performance. As each well reaches total depth, the team incorporates lessons learned with other
offset well data and uses the update to modify the conceptual model and plan the next well. This
iteration process repeats throughout the development-well drilling phase.
30
reservoir properties of porosity and hydrocarbon
saturation with measurements from an existing
wellbore validated the seismic facies predictions.
With this otherwise unattainable piece of
information, the operator was able to determine,
without first drilling the well, that the prospect
would not produce. Further, the company is using
the seismically derived reservoir measurements
to evaluate other prospects in the field and to
manage development by ranking drilling locations according to risk and the probability of
commercial success.
Once seismic data have been used to characterize the reservoir, reservoir modeling integrates
geophysics, geology and reservoir engineering to
build one earth model. Reservoir engineers use
this model to predict drainage patterns and design
injection strategies. Drilling and production
engineers use it to plan well trajectories.9
The latest versions of these tools, such as
Petrel seismic-to-simulation software, enable
seismic- and geomechanics-centered disciplines
to build shared earth models, rendering a more
accurate subsurface picture than that created by
one of them working independently. Changes in
the seismic interpretation or geological model
easily cascade through to the reservoir
simulation model and back. The workflows upon
which these software packages are constructed
are increasing the role of seismic data in
understanding dynamic reservoirs (below left).
Surface Work
The better the understanding of the target
reservoir, the less potential there is for surprises
such as noncommercial volumes of hydrocarbons,
flow assurance issues, early water breakthrough,
sand production and changes to fluid makeup.
Similarly, a well-planned overall field development
strategy—completion configurations, well locations, processing facility types and sizes, and
intervention decisions—is key to efficient
ultimate oil and gas recovery.
This is because the consequences of poor
planning are often not felt until near the end of
the field’s projected life span. Significant
revenue is lost when fields are abandoned
prematurely because the costs of remediation or
operating expenses are greater than the value of
the remaining reserves.
Avoiding these risks requires bypassing the
pitfalls created by strict separation of
engineering disciplines. Traditionally, reservoir
engineers have focused on well count, well
placement and recovery mechanisms; production
Oilfield Review
and completion engineers on well design; and
facilities engineers on the subsea layout, facility
size and topsides design.
These seemingly disparate groups must
instead be convinced to perform their tasks
independently yet understand the connection
imposed by production and therefore the
economics of the project. In turn, these
economics are dependent on the physical
limitations of the overall system. To function
properly, each discipline must be aware of how it
impacts the work of others; each member or
team involved in a development must work from
a common forecasting system.
One such system that comprises dynamically
linked models of the field’s subsystems—
reservoir, well and facility—is called an
integrated production model or integrated asset
model (IAM).10
During development planning stages and
before operations are undertaken, asset teams
use these integrated models to analyze the
interaction of proposed subsystems within a
project. IAMs represent a break with traditional
field development practices that are more likely
to be centered on capital expenditure and
focused on implementing modifications that
drive down costs. A common pitfall of the
traditional approach is a failure to properly
quantify the effects of changes on system
deliverability that in turn may ultimately lead to
suboptimal designs.
In contrast, an IAM uses a reservoir simulation
model to calculate fluid movement and pressure
distribution. Then, at the subsurface coupling
point—the well locations in the reservoir
model—these factors are put into the well model,
which establishes conditions at the sandface. The
sandface condition is used as a boundary to
compute the fluid rates or pressures at the surface
coupling point—the wellhead—where the well
model is linked to the surface facility.11
The interaction of well–surface boundary
conditions makes it possible to calculate the
backpressure of the production system for each
well. This is then conveyed back through the
system to the reservoir. The process iterates to
balance the full network. The result is stabilized
solutions for fluid flow from the reservoirs into
9. Hopkins C: “Go Beyond Reservoir Visualization,” E&P 80,
no. 9 (September 2007): 13–17.
10. For more on integrated asset modeling: Bouleau C,
Gehin H, Gutierrez H, Landgren K, Miller G, Peterson R,
Sperandio U, Traboulay I and Bravo da Silva L: “The Big
Picture: Integrated Asset Management,” Oilfield
Review 19, no. 4 (Winter 2007/2008): 34–48.
Spring 2009
Modeling
Framing
Production
engineer
Flow
assurance
engineer
Static QC
Facilities
engineer
Decision
analyst
Reservoir
Middle
engineer
management Simulation modeler
IPM
Initialization
Dynamic QC
Material balance
Simulation
> Integrated production model (IPM). Chevron engineers planning the
company’s ultradeepwater Jack field used a five-step workflow to build an
integrated production model for field development. In the framing step, the
entire team describes the project in terms of objectives, time frames, givens,
assumptions and deliverables. Once all parties were aware of these inputs,
subsurface, wellbore and network models were created during the modeling
step. A static QC was conducted by comparing the reservoir-to-separator
model inputs with available data from logs, cores, fluid samples and other
measurements. The wells in the subsurface (material balance or simulation)
model were linked to their pairs in the surface facility model during
intialization. The linked system was then run to check if the model converged
to a solution. Following this initialization, the whole system was run for the
entire prediction period. At the end of the period dynamic pressure and
temperature responses at separator, manifold, wellhead and bottomhole
were plotted during the dynamic QC step to help the team understand the
operating pressures and temperatures of seafloor boosters, wellheads and
bottom hole. Field schedules and constraints including water-handling
capacity, well maximum drawdown and minimum bottomhole pressure were
then checked to see if they were honored by the model.
the well and from the well into the surface
system and then to sales points. In this way, the
IAM technique considers the response of the
surface system in fluid-flow calculations.12
Chevron engineers used integrated production
management as a forecasting tool to couple
models of the subsurface with a surface network
via a wellbore model at their deepwater Gulf of
Mexico Jack field. A steady-state model calculated
temperature and pressure changes within the
wellbore. The surface-network model included the
subsea and surface elements such as manifolds,
seafloor pumps, wellheads, risers, flowlines and
separators. The surface and subsurface models
were linked at a bottomhole node.13
The Chevron model was constructed using a
five-step workflow (above). Those steps included
• defining the problem in terms of objectives, time
frames, givens, assumptions and deliverables
• modeling
• quality-checking of reservoir-to-separator
model input data against available data under
static conditions
• linking the surface and subsurface models
• quality-checking the full system for the entire
prediction period under dynamic conditions.
Once the workflow steps were completed, the
integrated project model determined that seafloor boosting, coupled with downhole artificial
lift, would best exploit reservoir deliverability. A
Chow CV, Arnondin MC, Wolcott KD and Ballard ND:
“Managing Risks Using Integrated Production Models:
Applications,” Journal of Petroleum Technology 52, no. 4
(April 2000): 94–98.
11. Tesaker Ø, Øverland AM, Arnesen D, Zangl G, AlKinani A, Torrens R, Bailey W, Couët B, Pecher R and
Rodriquez N: “Breaking the Barriers—The Integrated
Asset Model,” paper SPE 112223, presented at the SPE
Intelligent Energy Conference and Exhibition,
Amsterdam, February 25–27, 2008.
12. Tesaker et al, reference 11.
13. Ozdogan U, Keating JF, Knobles M, Chawathe A and
Seren D: “Recent Advances and Practical Applications
of Integrated Production Modeling at Jack Asset in
Deepwater Gulf of Mexico,” paper SPE 113904,
presented at the SPE Europec/EAGE Annual Conference
and Exhibition, Rome, June 9–12, 2008.
31
second study using experimental design then
allowed the Chevron team to identify the key
parameters of the artificial lift system.14 The
operator considered time of installation, seafloor
boosting inlet pressure, ESP horsepower and
setting depth. The company concluded that the
ESP horsepower was the most significant design
parameter across various recovery mechanisms.
The operator then switched from an
integrated to a modular approach—using the
wellbore model alone—that would support
design of the Jack field water-injection facility
and calculate the recovery trade-offs for various
topsides pressures and injection rates. Topsides
pressures were converted into bottomhole
pressures (BHPs) using the wellbore model.
From the resulting recovery contouring study the
team concluded that maximum recovery would
require a range of BHPs and injection rates. A
BHP of 21,000 psi [144.7 MPa] was infeasible
because the proposed high injection rates and
topsides pressure requirements limited available
pressure. Modeling, on the other hand, showed
that lowering the injection rates would not result
in significant production loss given the field
recovery response.15 Further developments to the
integrated model were used to optimize the
number of flowlines, number of seafloor boosting
pumps and platform location.
Driven to Succeed
Deposited on a seafloor of steep slopes, gullies and
canyons, hydrocarbon-bearing formations beyond
the continental shelves may be quite unlike those
found in shallow water. So it follows that project
plans are driven by considerations specific to deep
water. Because they are submersed in nearfreezing water, subsea flowlines, mudline
wellheads and manifolds pose flow assurance
concerns. As a consequence, subsea surveillance
instrumentation is installed throughout critical
points along the well-to-surface flow path. The
surveillance data are fed into fluid models that
enable engineers to take preemptive steps to
prevent hydrate or paraffin blockages (next page).
Because of the ability of hydrate, paraffin
and asphaltine deposition to impact project
economics, flow assurance is a primary
consideration in many deepwater project plans.
For example, given the wide separation of the five
fields and comparatively small reserves base that
make up BP’s Western Area Development (WAD)
in Angola’s deepwater Block 18, project planners
paid special attention to flow assurance issues
and system deliverability.
32
At the planning stage, engineers applied a
numerical method coupled with engineering
software to calculate multiphase thermalhydraulic behavior in an IAM.16 The goal was to
avoid potential flow disruptions caused by the
solidification of gas hydrates in the flowlines
traversing the cold seafloor.
In performing the thermal analyses of the
subsea systems at the WAD, the operator
considered conventional wet insulation, pipe-inpipe systems and flexible pipelines to determine
the time required for the coldest location in the
production flowline system to fall to temperatures at which hydrates form. Known as cooldown
time, this parameter indicates how long the
operator has before hydrate-prevention measures
must be taken following an unplanned shut-in.
Analysts also used IAM to perform deliverability calculations for numerous field
architectures and to investigate the effects of
tubing and pipeline sizes, looping pipelines and
subsea multiphase boosting. Their findings
enabled BP to combine representative production
profiles with capital and operating expense
models to derive a full economic evaluation and
screening of the company’s options.
Evolution or Revolution
Industry response to the unique challenges of
deep water has done more than spawn procedural
changes; it has also generated a nearly overwhelming burst of innovative drilling and
production hardware in a relatively short period
of time. As demonstrated by the Shell Mars
platform experience, this onslaught of new
technology has at times threatened to outpace
engineers’ efforts to keep abreast of it.
When coupled with the economic requirement that deepwater fields be developed with
as few wells as possible, this flood of new tools
and procedures has made it particularly difficult
for completion engineers to be certain of
delivering optimal solutions throughout the
project. The effort to use the most-effective
equipment is also thwarted by the time elapsing
between project commissioning and well
completion installation—often more than five
years—during which tool design and availability
may change radically.
To deal with these issues, experts have
designed modeling techniques to facilitate
quantitative analysis of wellbore-reservoir
interactions. These methods allow engineers to
plan individual wells based on surface and
subsurface characteristics as well as the current
state of technology. All this is done while taking
into account the constraints and characteristics
imposed on and by the associated disciplines of
geology, drilling, reservoir evaluation and
production operations.
One such method treats well design in much
the same way the processing industry handles
engineering-design problems. Process plants
are designed with feed and effluent flow
streams, using a flow diagram to capture the
process. This diagram then becomes the basis for
detailed design, component specifications and
other considerations.17
Using this method, engineers divide the well
design into conceptual and detail phases with
iterations between the two. The conceptual design
phase includes simple diagrams that highlight the
impact of critical surface and subsurface
attributes on well architecture. These attributes
are examined by the interdisciplinary team so that
alternatives may be considered for the key
components of well design—well trajectory,
formation completion and wellbore components.
Choice of well trajectory is a function of local
geology, reservoir properties and drilling capabilities. Formation completion refers to the
interface between the wellbore and the reservoir;
its configuration is determined by such factors
as rock lithology, mechanical properties, grainsize distribution and operational constraints in
the process.
Wellbore components are important elements
in the architecture of all wells. But they are
especially critical in deep water, where modification after the fact can be costly and technically
challenging and where conventional technologies
are sometimes insufficient. One way to choose the
appropriate technology for a particular completion is to sort available hardware options into four
categories by function—packers, valves, pumps
and sensors—and then use simple block
diagrams to discern the optimal type of each.
14. Experimental design is a branch of statistics that
outlines the way in which experiments should be carried
out so that statistical information can be maximized with
the minimum number of trials.
15. Ozdogan et al, reference 13.
16. Watson MJ, Hawkes NJ, Pickering PF, Elliot J and
Studd LW: “Integrated Flow Assurance Modeling of
Angola Block 18 Western Area Development,” paper
SPE 101826, presented at the SPE Annual Technical
Conference and Exhibition, San Antonio, Texas,
September 24–27, 2006.
17. Sinha S, Yan M and Jalali Y: “Completion Architecture:
A Methodology for Integrated Well Planning,” paper
SPE 85315, presented at the SPE/IADC Middle East
Drilling Technology Conference and Exhibition, Abu Dhabi,
UAE, October 20–22, 2003.
Oilfield Review
Sensor
Systems
Acquisition
Systems
Fluid-Property
Models
Process
Models
Operations
Changing parameters
Facilities
simulator
Multiphase
flowmeters
Dynamic dataacquisition system
Distributed
temperature sensors
Thermodynamic
models
Pressure and
temperature gauges
Multiphaseflow models
Electrical
submersible
pump monitors
Static-data
storage system
Flowline
simulator
Monitoring
Wellbore
simulator
Optimization
Deposition
models
Model conditioning
> Prevention not remediation. Flow assurance strategies are an integral part of production operations in deep
water. Surveillance instrumentation such as multiphase meters, distributed temperature sensors, pressure and
temperature gauges, an SMC subsea surveillance system and ESP monitors installed in the flow path (bottom)
delivers data in real or near-real time. This data stream feeds predictive fluid-property models for solids deposition,
corrosion, rheology and thermodynamics. The planning team uses these models to create process models that
include facilities, flowline and wellbore simulators. Over time, continuous monitoring of changing parameters
closes the loop and includes fluids data, flow models and real-time measurements (top). This loop drives the
optimization of mitigation strategies and should be included in the production system design as early as possible.
In this way, overtreating by a chemicals-delivery system set to worst-case scenarios can be minimized. Monitoring
and modeling also provide the information for decisions on the most appropriate proactive, preventive treatments
or remedial techniques—thermal, chemical or mechanical—to be put into place to prevent plugging and other
impediments to flow, thus avoiding expensive rig-based interventions.
Spring 2009
33
Reservoir Geology and Geometry
Reservoir Property
Yes
Surface setting:
subsea
Large
Yes
Property:
kv /kh > 0.25
Geometry: flat
areal structure
No
Small
Horizontal,
highly deviated wells
No
Poor
mobility
Vertical,
low-deviation wells
Vertical,
low-deviation wells
Vertical,
low-deviation wells
> Well completion technology. In complex deepwater developments, where most wells are subsea completions, each well is
a critical contributor to overall production. In the simplified decision tree shown, basic geological and reservoir attributes are
used to determine, generally, the appropriate wellbore angle—low to vertical or high to horizontal. Knowledge of necessary
well angle and reach informs operator decisions about requirements for drilling rig capabilities and directional drilling. In a
small areal structure, low-angle or vertical wells are usually sufficient to drain the reservoir. This may allow the use of lessrobust and therefore less-costly equipment for development drilling. In larger structures, wellbore deviation severity is
subject to secondary parameters of formation permeability and fluid mobility.
The selection process begins with an
evaluation of basic geological and reservoir
attributes that constrain options for wellbore
trajectories (above). Similar exercises can be
used to decide the advisability and type of subsea
wellbore configurations such as multilateral
completions, stimulations, sand control and
artificial lift.
The resulting conceptual well design can be
used for more detailed analysis that is then
added to a project workflow. This workflow
considers a quantitative assessment of well
performance, preliminary economic analysis and
detailed design.
Planning for the Unexpected
As subsea wellheads increasingly become the
completion design of choice in deep water,
operators strive to contain costs by minimizing
interventions. The high cost and technical
uncertainty associated with entering wellheads
located on the ocean floor beneath thousands of
feet of water were, in fact, early motivations for
developing intelligent completions in the 1990s.
Although remote downhole monitoring and
operations, along with more-robust completions,
have done much to reduce their frequency, rigbased deepwater well interventions are unlikely
to be eliminated entirely (next page).
34
The links between intervention, production
and costs have been demonstrated clearly and
often. Frequent, immediate remedial action to
repair failed components yields higher production rates but increases operating expense. A
policy of fewer or delayed maintenance operations is less costly but results in lower production
volumes and revenues.
The challenge then is to develop a plan that
strikes a balance of the two options, which is
commonly done by dividing interventions into
immediate, performance-based or campaign
(numerous interventions performed within a
field in sequence) repair strategies. To make
such a policy workable through the life of the
project requires including proactive maintenance during the front-end engineering and
design (FEED) stage, rather than relegating it
the traditional “as needed” role.
Norske Shell used a modified version of the
three-repair strategy to help mitigate intervention costs at its Ormen Lange project—Norway’s
first deepwater subsea development 120 km
[74 mi] northwest of Kristiansund. During the
FEED stage planners instituted a simulation
approach for estimating future intervention,
maintenance and repair costs of various
strategies. Risk expenditures included the direct
cost of the intervention to repair a specific
component plus the lost revenue incurred as a
result of the failure. This method allowed the
operator to assess the impact of different
intervention strategies throughout the life of
the project.18
The overall result of this type of modeling is
to direct intervention, maintenance and repair
efforts toward areas where they yield the highest
value. It also focuses attention on critical
equipment packages within the project that
contribute the most to risk expenditures—the
sum of the revenue lost to producing well failures
and the cost of intervention. This information
may be used to improve designs, increase
equipment reliability and initiate smarter, lessexpensive ways to fix failed units and thus reduce
risk expenditures.
Finally, such a strategy often allows the
operator to continually update forecasting
information within the model as the project
progresses into the operational phase. In the
case of Ormen Lange, for example, the model
was initially built on rough estimates of the
18. Eriksen R, Gustavsson F and Anthonsen H: “Developing
an Intervention, Maintenance and Repair Strategy for
Ormen Lange,” paper SPE 96751, presented at the SPE
Offshore Europe Oil & Gas Conference and Exhibition,
Aberdeen, September 6–9, 2005.
19. Eriksen et al, reference 18.
20. For more on the Jack field: Aghar H, Carie M,
Elshahawi H, Gomez JR, Saeedi J, Young C, Pinguet B,
Swainson K, Takla E and Theuveny B: “The Expanding
Scope of Well Testing,” Oilfield Review 19, no. 1
(Spring 2007): 44–59.
Oilfield Review
Similarly, while a great deal of progress
has been made in the past decade, the deepwater environment still holds technological
challenges for the industry. Just as operators
once purchased leases in water depths greater
than 2,300 m [7,500 ft] knowing they could not
economically produce them, today they are
asking service companies and drilling contractors
to customize tools to extend depth, pressure and
temperature barriers to 3,050 m and beyond.
For example, Schlumberger engineers were
recently asked to develop a perforating system
specifically for Chevron’s Jack field test well.20
The solution required re-engineering existing
tools based on casing size and the unique
bottomhole pressure conditions expected in this
ultradeepwater well. In the end Schlumberger
delivered a unique combination of tools that
together operated as a 172-MPa [25,000-psi]
perforating system.
Such requests will continue to be made as
operators bring their ultradeepwater prospects
beyond the exploratory stage to testing and
development. But it is important to consider that it
took nine months for Schlumberger to develop,
qualify, quality-test and deliver the necessary
system working from a similar tool developed for
Chevron’s deepwater Gulf of Mexico Tahiti field.
Given the task, nine months was a remarkably
short time in which to deliver. But it clearly
demonstrates the necessity of maintaining a broad
view if costly delays caused by technological
shortcomings are to be avoided in the complex and
unforgiving deepwater operating environment.
Such foresight demands a culture of planning
that simultaneously encompasses a long-term
vision of integrated tasks and short-term detailoriented designs.
—RvF
> Light well interventions. Rigless subsea well interventions from multiservice vessels are considerably
less expensive than those performed from deepwater drilling rigs. Light interventions commonly
support remotely operated vehicles used to replace or reset controls or sensors on subsea trees,
pipelines or manifolds. They may also facilitate rigless downhole intervention using wireline or coiled
tubing, as shown here, to perform such functions as safety valve maintenance, perforation or coiled
tubing wellbore cleanout. Currently, these types of interventions are restricted to water depths of
about 500 m or about 1,500 ft.
field layout. As the project moved forward,
the Norske Shell team continually refined the
model to assist in equipment configuration and
to predict performance, which is used in
economic evaluations.19
The Big Picture
More than any previous shift in its environment,
operating in deep water has forced the upstream
industry to change how it conducts business. This
cultural shift is due to the outsized rewards and
Spring 2009
risks attached to deepwater operations, but
perhaps even more so it is the consequence of
the unprecedented length of time between the
decision to exploit a prospect and first oil. It is
impossible to predict oil and gas prices or the
state of the global economy over such a time
span. Operators must make critical investment
decisions without the benefit of traditional
economic fundamentals that come with more
immediate returns on investment.
35
Evaluating Volcanic Reservoirs
M.Y. Farooqui
Gujarat State Petroleum Corporation (GSPC)
Gandhinagar, Gujarat, India
Huijun Hou
Dhahran, Saudi Arabia
Guoxin Li
PetroChina Exploration and Production
Company Limited
Beijing, China
Nigel Machin
Saudi Aramco
Dhahran, Saudi Arabia
Tom Neville
Cambridge, Massachusetts, USA
Aditi Pal
Jakarta, Indonesia
Chandramani Shrivastva
Mumbai, India
Yuhua Wang
Fengping Yang
Changhai Yin
Jie Zhao
PetroChina Daqing Oilfield Company
Daqing, China
Xingwang Yang
Tokyo, Japan
Oilfield Review Spring 2009: 21, no. 1.
Copyright © 2009 Schlumberger.
DMR, ECS and FMI are marks of Schlumberger.
For help in preparation of this article, thanks to Martin
Isaacs, Sugar Land, Texas, USA; Shumao Jin, Brett Rimmer
and Michael Yang, Beijing; Charles E. Jones, University of
Pittsburgh, Pennsylvania, USA; Andreas Laake, Cairo; and
Hetu C. Sheth, Indian Institute of Technology, Mumbai.
36
Hydrocarbons can be found in volcanic rock—sometimes in significant quantities.
Petrophysical methods originally developed for sedimentary accumulations are being
used to evaluate these unusual reservoirs.
In the early days of petroleum exploration, the
discovery of hydrocarbons in anything other than
sedimentary rock was largely accidental, and such
accumulations were considered flukes. Serendipity
is still part of exploration, but geologists now know
that the presence of oil and gas in such rock is
certainly no coincidence. Igneous rock—created
by the solidification of magma—hosts petroleum
reservoirs in many major hydro­carbon provinces,
sometimes predominating them.
In general, igneous rocks have been ignored
and even avoided by the E&P industry. They have
been ignored because of a perceived lack of reservoir quality. However, there are many ways in
which igneous rocks can develop porosity and
permeability.1 Far from inconsequential, igneous
activity can influence every aspect of a petroleum
system, providing source rock, affecting fluid
maturation and creating migration pathways,
traps, reservoirs and seals.2
Igneous rocks have been avoided for other
reasons. They tend to be extremely hard, although
improvements in bit technology are helping drillers cope with these tough lithologies.3 Because
they typically prevent deep pene­tration of seismic
energy, igneous layers are considered an impediment to evaluation of underlying sediments as well.
New seismic methods are advancing solutions
to this problem, but with their strong refractive qualities, igneous reservoirs remain difficult
to characterize.4
Once hydrocarbons are found in igneous
reser­voirs, assessing hydrocarbon volumes and
productivity presents several challenges. Log
interpretation in igneous reservoirs often requires
adapting techniques designed for other environments. Logging tools and interpretation methods
that succeed in sedimentary rock can give meaningful answers in igneous rock, but they often
require artful application. Furthermore, because
mineralogy varies greatly in these formations,
methods that work in one volcanic province may
fail in another. Usually, a combination of methods
is required.
This article describes the complexity of volcanic reservoirs and presents technologies that
have proved successful in characterizing them.
The discussion begins with a review of igneous
rock types and follows with an examination of
the effects of igneous processes on petroleum
1. Srugoa P and Rubinstein P: “Processes Controlling
Porosity and Permeability in Volcanic Reservoirs from
the Austral and Neuquén Basins, Argentina,” AAPG
Bulletin 91, no. 1 (January 2007): 115–129.
2. Schutter SR: “Hydrocarbon Occurrence and Exploration
in and Around Igneous Rocks,” in Petford N and
McCaffrey KJW (eds): Hydrocarbons in Crystalline Rocks,
Geological Society Special Publication 214. London:
Geological Society (2003): 7–33.
3. Close F, Conroy D, Greig A, Morin A, Flint G and Seale R:
“Successful Drilling of Basalt in a West of Shetland
Deepwater Discovery,” paper SPE 96575, presented at
the SPE Offshore Europe Oil and Gas Conference and
Exhibition, Aberdeen, September 6–9, 2005.
Salleh S and Eckstrom D: “Reducing Well Costs by
Optimizing Drilling Including Hard/Abrasive Igneous Rock
Section Offshore Vietnam,” paper SPE 62777, presented
at the IADC/SPE Asia Pacific Drilling Technology
Conference, Kuala Lumpur, September 11–13, 2000.
4. Hill D, Combee L and Bacon J: “Over/Under Acquisition
and Data Processing: The Next Quantum Leap in Seismic
Technology?” First Break 24, no. 6 (June 2006): 81–95.
White RS, Smallwood JR, Fliedner MM, Boslaugh B,
Maresh J and Fruehn J: “Imaging and Regional
Distribution of Basalt Flows in the Faeroe-Shetland
Basin,” Geophysical Prospecting 51, no. 3 (May 2003):
215–231.
Oilfield Review
Oilfield Review
Winter 09
Volcanic Fig. Opener
ORWINT09-VOL Fig. Opener
Spring 2009
37
systems. Two field examples highlight formation
evaluation in volcanic rocks. A case study from a
gas-rich reservoir in China presents a technique
that combines conventional logging measurements and image logs with neutron-capture
spectroscopy and nuclear magnetic resonance.
Plume
Ash-cloud surge
Pyroclastic flow
Traps
Eruption
column
Volcaniclastic rocks
Laccolith
exposed
by erosion
Dikes
Volcano
Granite wash
Lava flow
Dike
Plutonic
rock
Laccolith
Sill
Country rock
Pluton
Basement
> Emplacement of igneous rocks. Plutonic rocks, formed by cooling of magma within the Earth, display
well-developed crystals with little porosity. Plutons and laccoliths—bulging igneous injections into
sedimentary layers—are examples of plutonic rock. Volcanic rocks, formed when magma extrudes
onto the surface and cools rapidly, show very fine crystalline or even glassy textures. Buildup
of pressures within the Earth can cause explosive eruptions; these result in the accumulation of
fragments of volcanic material in pyroclastic deposits. Rock containing clastic fragments of volcanic
origin is termed volcaniclastic. Complex porosities and permeabilities can develop as a result of these
different processes.
Structures
Textures
Flow—Flows form when the fabric of lava aligns in
parallel rows or ropy waves.
Brecciated—Most angular particles exceeding 2 mm
in diameter are volcanic breccia. Typically, particles
form from the movement of partially solidified rock, not
from the ejection of fragments.
Pillow—Lava that erupts under water and quickly
develops a cool skin around a molten core forms pillow
structures, which are bulbous piles of rock. Pillow lava
often incorporates seafloor sediments.
Glassy—Lava that cools rapidly forms volcanic glass
such as obsidian, tonalite and pitchstone, which differ
mainly in their alkali feldspar content.
Porphyry—One of the most common porphyritic
Oilfield ReviewTuffaceous—Consolidated pyroclastic material less
structures is phenocrysts, 1- to 2-mm [0.04- to 0.08-in.]
crystals embedded in a fine-grained, often glassy
matrix. 09
than 2 mm [about 0.08 in.] in diameter is tuff.
Winter
Andesite and basalt often have olivine and pyroxene
tuff is ash. Both can be deposited far
Volcanic Fig. 1Unconsolidated
phenocrysts.
from their source. A common epiclastic, or weathered
ORWINT09-VOL
Fig. 1reservoir rock is tuffaceous sand, in which
volcanic,
Pyroclast—Pyroclasts are sharp, chiseled rock
reworked tuff accounts for less than half the volume of
fragments created during a volcanic explosion. Glass
rock. When tuff makes up more than half the rock,
shards are often a key component. Sharp shards indicate
the deposit is called sandy tuff.
rapid burial or minimal postdepositional reworking.
Vesicular—Gas expanding in cooling lava creates
pores called vesicles. Often unconnected, they are the
reason very porous volcanic rock, such as pumice, can
> Structures and textures in volcanic rocks.
float but has negligible permeability. Vesicles often
fill with secondary minerals, usually hydrated silicates
Variations in structure and texture give rise to
called zeolites. These filled vesicles, called amygdules,
the wide range of porosity and permeability
reduce intergranular porosity in the same manner as
observed in crystalline and pyroclastic rock.
clay in sandstone.
38
An example from India demonstrates the importance of incorporating borehole resistivity images
in the evaluation of oil-bearing volcanic rock.
About Igneous Rocks
Igneous rock is formed through the solidification
of magma—a mixture of water, dissolved gases
and molten to partially molten rock. Igneous
rocks vary from one reservoir to another because
their constituents have diverse chemistries,
origi­nating from magma that mixes material
from the Earth’s mantle, crust and surface—
typically oxides of silicon, iron, magnesium,
sodium, calcium and potassium. They also have
diverse structures and textures—leading to
complex porosities and permeabilities—depending on how they were emplaced. Emplacement
mechanisms include sudden explosive eruptions,
syrupy viscous flows and slow, deep subsurface
intrusions. Subsequent weathering and fracturing can further complicate rock properties.
Igneous rocks form under a wide range of conditions, and therefore display a variety of properties
(left). Molten rock that cools deep beneath the surface forms intrusive, or plutonic, rocks. Slow cooling
of deep magmas forms large crystals, resulting in
coarse-grained rock. These formations typically
have low intergranular porosity and insignificant
permeability, making them of little interest to the
oil industry. The one exception is fractured granites,
which can produce hydrocarbons.5 Magmas that
approach the surface tend to cool more rapidly.
This allows less time for the formation of crystals,
which therefore tend to be smaller, resulting in finegrained crystalline rock.
Extrusive, or volcanic, rocks are created
when magma erupts through the Earth’s surface. Magma may extrude in flows of molten
lava that, when cooled, form fine- to very finegrained crystalline volcanic rock. Sometimes,
cooling occurs so quickly that crystals cannot
form, resulting in volcanic glass, such as obsidian. When magmas contain large amounts of
water and dissolved gases, buildup of excessive
pressure under the ground can cause explosive
eruptions of volcanic material. Ejected fragments, or pyroclasts, can range in size from fine
volcanic ash to “bombs” tens of centimeters in
diameter. Once they have been ejected, individual fragments accumulate to form pyroclastic
rock. Lava flows and pyroclastic deposits may
be a few centimeters to a few hundred meters
thick, covering thousands of square kilometers.
These deposits can have sufficient porosity
and permeability to make them viable hydrocarbon reservoirs.
Oilfield Review
Spring 2009
Coarse Grained
Peridotite
Basalt
Andesite
Dacite
Rhyolite
Gabbro
Diorite
Granodiorite
Granite
100
Calcium-rich
plagioclase
feldspars
80
Quartz
Potassium
feldspar
60
Sodium-rich
plagioclase
feldspars
Olivine
40
Pyroxene
20
Biotite
Amphibole
0
45%
Increasing silica content
75%
Increasing calcium, magnesium and iron content
Increasing potassium, sodium and aluminum content
1,200°C [2,200°F]
Increasing temperature of crystallization
700°C [1,300°F]
> Classifying igneous rocks by mineral composition. Fine-grained and coarse-grained rocks of similar
composition have different names. For example, a magma containing quartz, potassium feldspar,
sodium-rich plagioclase and biotite may cool slowly and form coarse-grained granite. If the same
magma is extruded, it will form fine-grained rhyolite. Olivine-rich magmas do not commonly extrude,
but crystallize at depth, and so form only coarse-grained rocks.
Clast or
Crystal
Size, mm
Sedimentary
Clasts
Boulders
256
Cobbles
16
Pyroclastic
Fragments
Crystalline Rocks:
Igneous, Metamorphic
or Sedimentary
Blocks
and bombs
Very coarse
grained
Very coarse
crystalline
Gravel
64
Pebbles
Lapilli
Coarse grained
2
1
0.5
0.25
0.125
Coarse crystalline
Granules
Very coarse sand
Medium grained
Coarse sand
Medium sand
Fine sand
Very fine sand
Oilfield Review
Coarse ash
Winter 09grains
Volcanic Fig. 2
ORWINT09-VOL Fig. 2
Sand
4
Medium crystalline
Fine grained
Fine crystalline
0.032
Silt
0.004
Clay
Very fine grained
Mud
5. For example, recoverable oil reserves in the fractured
granite of the Cuu Long basin offshore Vietnam are
estimated at 2 billion bbl [320 million m3] or more. For
more: Du Hung N and Van Le H: “Petroleum Geology
of Cuu Long Basin—Offshore Vietnam,” Search and
Discovery Article #10062, http://www.searchanddiscovery.
net/documents/2004/hung/images/hung.pdf (accessed
April 6, 2009).
The giant Suban gas field in southern Sumatra contains
estimated reserves of 5 Tcf [140 billion m3] in fractured
granites. For more: Koning T: “Oil and Gas Production
from Basement Reservoirs: Examples from Indonesia,
USA and Venezuela,” in Petford N and McCaffrey KJW
(eds): Hydrocarbons in Crystalline Rocks, Geological
Society Special Publication 214. London: Geological
Society (2003): 83–92.
Landes KK, Amoruso JJ, Charlesworth LJ Jr, Heany F
and Lesperance PJ: “Petroleum Resources in Basement
Rocks,” Bulletin of the AAPG 44, no. 10 (October 1960):
1682–1691.
6. An acidic rock contains proportionately more nonmetallic
oxides than a basic rock and forms an acid when
dissolved in water. A basic rock contains proportionately
more metallic oxides than an acidic rock and forms a
base when dissolved in water.
7. The term “mafic” is derived from the words magnesium
and ferric, whereas “felsic” is a combination of feldspar
and silica.
Hyndman DW: Petrology of Igneous and Metamorphic
Rocks, 2nd ed. New York City: McGraw-Hill Higher
Education, 1985.
Fine Grained
Mineral composition, volume percent
The different modes of formation of igneous
rocks—cooling of lavas, either under the ground or at
the surface, and agglomeration of fragments ejected
during explosive eruptions—allow a subdivision of
igneous rocks into two groups: crystalline igneous
rocks and fragmental igneous, or pyroclastic, rocks.
A simple and common compositional classification of crystalline igneous rocks is based on
silica [SiO2] weight percentage. Rocks low in SiO2
(less than 52%) are classed as basic, rocks high in
SiO2 (more than 66%) are acidic and those with
SiO2 between 52 and 66% are intermediate.6
A parallel classification system groups rocks
by weight percent of dark-colored minerals.
Rocks rich (more than 70%) in dark minerals,
such as olivine and pyroxene, are mafic; those
containing few dark minerals (less than 40%),
and therefore more light minerals, such as quartz
and feldspar, are silicic, sometimes called felsic.7
Mafic rocks, such as basalt, tend to be basic;
silicic rocks, such as granite, tend to be acidic.
A different classification encompasses em­place­ment mechanism, crystal size and mineralogy, dividing crystalline volcanic rocks into
four main types (above right). The trend from
basalt to andesite, dacite and rhyolite forms a
continuum of mineralogy.
Pyroclastic rocks, on the other hand, are
typically classified by grain size, as are clastic
sedimentary rocks. Relative proportions of three
grain-size classes—blocks and bombs, lapilli
and ash—are used to classify a pyroclastic rock
(right). Pyroclastic and crystalline rock types
exhibit differences in texture and structure that
lead to differences in porosity and permeability
(previous page, bottom).
Fine ash
grains
Very fine crystalline
Cryptocrystalline
> Classifying pyroclastic rocks by grain size. Pyroclastic rocks are identified
based on grain size, in a similar fashion to clastic sedimentary rocks.
39
ARGENTINA
Chaitén
CHILE
Plume
Ash cover
ATLANTIC OCEAN
0
km 100
0
miles
100
> Image of the Chaitén volcano, southern Chile, from the NASA Terra satellite. The volcano, thought
to be dormant before its May 2, 2008, eruption, sent a plume of ash and steam 10.7 to 16.8 km [35,000
to 55,000 ft] into the atmosphere. This image, acquired three days after the eruption, shows the plume
extending eastward more than 1,000 km across Argentina and into the Atlantic Ocean. The volcanic
plume (white) is distinguishable from the clouds (turquoise). The land surface is dusted with tan-gray
ash. [From “Chile’s Chaiten Volcano Erupts,” http://earthobservatory.nasa.gov/IOTD/view.php?id=8725
(accessed April 6, 2009)].
Volumes of Volcanics
Petrologists have calculated that the shallow part
of the Earth’s crust contains a volume of volcanic rock—formed by the ejection of lava at the
surface—of 3.4 to 9 x 109 km3, an order of magnitude greater than the volume of sedimentary rock.
This estimate includes extrusions at seafloor rift
zones, where oceanic plates are pulling apart and
new crust is created by volcanic activity.
The presence of volcanic rocks in hydrocarbon provinces is common because volcanic
activity has taken place in or near many sedimentary basins at one time or another. Volcanism
can also affect distant basins—large volcanoes
can push pyroclastic flows up to 1,000 km [about
600 mi] from their origin and wind can carry ash
thousands of kilometers (left). Consequently,
blankets of ash and tuffs, or consolidated ash,
may be found far from their source.
Hydrocarbon-producing igneous rocks occur
the world over (below). The earliest documented oil discovery in volcanic rock may be
the Hara oil field of Japan, which began producing in 1900.8 The field produced oil from
three tuffaceous layers. Other early production was recorded in Texas, in 1915, along a
trend of seafloor volcanoes that erupted during deposition of the Austin Chalk.9 The buried
volcanic formations produced 54 million bbl
[8.6 million m3] of oil from 90 fields in more than
200 igneous bodies.
Oilfield Review
Winter 09
Volcanic Fig. 5
ORWINT09-VOL Fig. 5
Hydrocarbons associated with
igneous rocks or igneous activity
> Distribution of hydrocarbon-bearing igneous rocks. Gold dots represent locations of hydrocarbon seeps, shows and reservoirs in igneous rocks. (Adapted
from Schutter, reference 36).
40
Oilfield Review
Volcanic reservoirs may contain significant
accumulations. As of 1996, cumulative produc­
tion from the volcanic tuff and associated
layers of the Jatibarang field, West Java, was
1.2 billion bbl [190 million m3] of oil and 2.7 Tcf
[76 billion m3] of gas. Speculated reserves are
4 billion bbl [635 million m3] of oil and 3 Tcf
[85 billion m3] of gas.10 Reservoir analysis yields
porosity values of 16 to 25% and permeability up
to 10 darcies. In this reservoir, the volcanic rocks
are also source rocks.11
Christmas Tree Laccolith
Punched Laccolith
Petroleum Systems
Volcanism can affect all aspects of a petroleum
system, producing distinctive source rocks,
accelerating fluid maturation, facilitating fluid
migration, and creating traps, reservoirs and seals.
Source Rock—Although most hydrocarbons
found in volcanic rocks come from sedimen­
tary source rock, some volcanic rocks are also
source rocks. Vegetation entrained in ash flows
may contain enough water to protect it from
the heat of emplacement. Subaerial volcanism
may create lakes and swamps with kerogen-rich
sediments, and the volcanically warmed water in
these basins encourages nutrient growth, further
enhancing the production of organic material.
Maturation—By adding heat, igneous bod­
ies can accelerate hydrocarbon maturation.
Large intrusive bodies, such as thick dikes and
sills, cool slowly and may affect great volumes
of surrounding rock, causing overmaturation.12
Volcanic flows cool relatively quickly, so they usu­
ally have less impact on maturation. The impact
of igneous activity on fluid maturation can be
assessed by petroleum systems modeling.13
In addition to direct heat, the circulation
of hydrothermal fluids in the heated zone also
may affect maturation. For example, scientists
working in the Guaymas basin of the Gulf of
California have reported that hydrothermal
fluids heated to 400°C [752°F] are responsible
Traps—Igneous intrusions into surrounding
for alteration of organic matter and the creation of
petroleum.14 The process is rapid, taking hundreds sedimentary layers, called country rock, often
to thousands of years rather than the millions of result in closed structures within the intruded
formations. The Omaha Dome field in the Illinois
years typically needed to generate oil.15
Migration—There are several ways for hydro­ basin, USA, was formed by this type of trap. The
carbons that originated elsewhere to become trapping structure is a Christmas tree laccolith
produced by an ultramafic intrusion (above).16
trapped in volcanic rocks:
•Hydrocarbons can pass vertically or later­ The field was discovered in 1940 and produced
ally from sedimentary rocks into structurally about 6.5 million bbl [1 million m3] of oil from
Oilfield Review
higher volcanic rocks.
sandstones that are in contact with the intrusion.
Winter
09
•Compaction of sedimentary rocks can
force
Reservoirs—Igneous rocks share another
Volcanic Fig.characteristic
7
hydrocarbons downward into volcanic rocks.
with sedimentary reservoir rocks;
ORWINT09-VOL Fig. 7
•Hydrothermal fluids are capable of dissolving they can have primary porosity and sometimes
hydrocarbons and depositing them in igne­ develop secondary porosity. But unlike sedimen­
ous rocks.
tary rocks, igneous rocks lose their porosity quite
•If the vapor pressure in volcanic rocks becomes slowly with compaction. Primary porosity may
low enough during cooling, hydrocarbons may be intergranular or vesicular—a type of poros­
be drawn into the pore spaces.
ity resulting from the presence of vesicles, or gas
8.Mining in Japan, Past and Present. The Bureau of
Mines, Department of Agriculture and Commerce of
Japan, 1909.
9.Ewing TE and Caran SC: “Late Cretaceous Volcanism in
South and Central Texas—Stratigraphic, Structural, and
Seismic Models,” Transactions, Gulf Coast Association
of Geological Societies 32 (1982): 137–145.
10.Kartanegara AL, Baik RN and Ibrahim MA: “Volcanics
Oil Bearing in Indonesia,” AAPG Bulletin 80, no. 13
(1996): A73.
11.Bishop MG: “Petroleum Systems of the Northwest
Java Province, Java and Offshore Southeast Sumatra,
Indonesia,” USGS Open-File Report 99–50R (2000),
http://pubs.usgs.gov/of/1999/ofr-99-0050/OF99-50R/
ardj_occr.html (accessed April 7, 2009).
12.Schutter, reference 2.
13.Yurewicz DA, Bohacs KM, Kendall J, Klimentidis RE,
Kronmueller K, Meurer ME, Ryan TC and Yeakel JD:
“Controls on Gas and Water Distribution, Mesaverde
Basin-Centered Gas Play, Piceance Basin, Colorado,”
in Cumella SP, Shanley KW and Camp WK (eds):
Understanding, Exploring and Developing Tight-Gas
Sands: 2005 Vail Hedberg Conference, AAPG Hedberg
Series, no. 3 (2008): 105–136.
14.Simoneit BRT: “Organic Matter Alteration and Fluid
Migration in Hydrothermal Systems,” in Parnell J (ed):
Geofluids: Origin, Migration and Evolution of Fluids
in Sedimentary Basins, Geological Society Special
Publication 78. London: Geological Society (1994):
261–274.
Spring 2009
> Traps caused by laccolith intrusion. The trap of the Omaha Dome field in
Illinois was caused by a Christmas tree laccolith (left ) of mica-peridotite
intruding into limestones and sandstones. Traps (green) can also be caused
by punched laccoliths (right), which lift overlying layers along bounding faults.
15.Kvenvolden KA and Simoneit BRT: “Hydrothermally
Derived Petroleum: Examples from Guaymas Basin, Gulf
of California, and Escanaba Trough, Northeast Pacific
Ocean,” AAPG Bulletin 74, no. 3 (March 1990): 223–237.
16.English RM and Grogan RM: “Omaha Pool and
Mica-Peridotite Intrusives, Gallatin County, Illinois,”
in Howell JV (ed): Structure of Typical American Oil
Fields, Special Publication 14, vol. 3. Tulsa: American
Association of Petroleum Geologists (1948): 189–212.
41
Fresh basalt
Weathered basalt
Nonbasalt rocks
Fresh basalt
Payun
Payun
Weathered basalt
Basalt with sparse vegetation
Nonvolcanic sediments
0
0
km 20
mi
20
Vegetation
> Remote sensing in volcanic provinces. Satellite data from visible, near-infrared, infrared and thermal
bands help geophysicists assess topography and ground surface character before planning seismic
survey acquisition. In this example from Argentina, satellite data (bottom) from several spectral bands
are combined and color-coded to distinguish different surface characteristics. Recently erupted
basalt flows are highlighted as dark red in both satellite images. Acquisition crews use the information
to determine whether the terrain is accessible to vibrator trucks and other equipment (top). The
photograph of the survey vehicles shows the Payun volcano seen from the south.
bubbles, in igneous rock. Porosities in vesicular surface mapping of elevated structures has
basalts and andesites may reach 50%.17 Secondary revealed volcanic deposits. For example, in
porosity is important for many volcanic reservoirs Japan, rhyolitic volcanic rocks containing large
and is sometimes the only porosity present. It may hydrocarbon accumulations have been discovresult from hydrothermal alteration, fracturing ered by mapping structural highs.18 Another traand late-stage metamorphism—metamorphism ditional method, the recognition of hydrocarbon
Oilfield
Review
during the late stages of igneous activity
that
seeps at the surface, is used to find deeper reserWinter 09
alters the minerals formed earlier. Sills and lac- voirs. Oil and gas sometimes rise to the surface
Volcanic Fig. 8
coliths may become reservoirs, especially
when alongFig.
contacts
between igneous and sedimentary
ORWINT09-VOL
8
they intrude into source rocks. They may fracture rocks. Seeps in the Golden Lane area of eastern
upon cooling, providing porosity, permeability Mexico have been associated with steeply dipand migration pathways.
ping igneous rocks that have penetrated thick
Seals—Igneous rocks can provide seals. After oil-rich carbonate layers.19
alteration to clay, extrusive layers may act as
Advanced techniques are also used. Satellite
tight seals. Impermeable intruded rocks, such as imagery has been applied to evaluate the basaltlaccoliths that form traps, also may seal hydro- covered Columbia basin in Washington and
carbons in formations beneath them.
Oregon, USA.20 Geochemical analysis of groundwater in the same region has detected significant
Exploration in Volcanic Provinces
levels of methane over a large area, indicating
Hydrocarbon exploration in and around igneous potentially commercial quantities of natural gas
rocks may involve a variety of geological, geo­ in Columbia River basalts.21
physical and geochemical techniques. Traditional
42
Depending on the properties of the volcanic
rocks, gravity and magnetic techniques may be
useful. These were among the earliest geophysical approaches applied, and they contributed to
the successful exploitation of the 1915 Texas volcanic play mentioned previously. Mafic igneous
rocks—richer in dense and magnetic minerals
than felsic igneous rocks—offer better contrast
with regional sediments, so they may show
up distinctly on gravity and magnetic surveys.
Aeromagnetic surveys have been effective in
identifying prospects in mafic flood basalts in the
Otway basin, southeastern Australia.22
Magnetotelluric (MT) methods have also
been used, usually in conjunction with other
techniques, to investigate high-resistivity vol­canic
rocks as potential reservoirs (for more on MT,
see “Electromagnetic Sounding for Hydrocarbons,”
page 4). For example, MT surveys in the Yurihara
oil and gas field in Japan are aiding exploration of
areas surrounding producing reservoirs.23 On some
MT lines, resistive uplifted volcanic layers have
been identified as possible prospects. Integration
of MT surveys with surface seismic information
was valuable in characterizing the internal structure of an oil- and gas-producing basalt layer.
Seismic methods, while extremely useful for
detecting sedimentary structures, have had mixed
success in volcanic provinces. Massive basalts without internal layering have high effective seismic
quality, meaning they are not highly absorptive,
so seismic waves pass through them with little
attenuation. Seismic surveys are relatively successful in delineating the tops and bottoms of such
layers. However, layered basalts, especially those
with interspersed weathered surfaces, tend to
scatter seismic energy and may yield poor data.24
To improve the quality of seismic data in volcanic
provinces, survey planners use satellite sensing
to determine lithology and topography, and are
incorporating the results in assessments of survey
logistics, acquisition parameters and processing
requirements (above left).25
In areas with highly attenuating volcanic layers, borehole seismic surveys have shown some
promise in improving seismic image resolution.
Such was the case with an offset vertical seismic
profile (VSP) acquired in a 4,750-m [15,600-ft]
exploratory well in the Neuquén basin, Argentina.26
At the well location, the surface was covered
by approximately 150 m [490 ft] of basalt that
strongly attenuated surface seismic energy. The
VSP produced an image with higher resolution
than the surface seismic results and illuminated
other igneous bodies in the subsurface.
Oilfield Review
17.Chen Z, Yan H, Li J, Zhang G, Zhang Z and Liu B:
“Relationship Between Tertiary Volcanic Rocks and
Hydrocarbons in the Liaohe Basin, People’s Republic of
China,” AAPG Bulletin 83, no. 6 (June 1999): 1004–1014.
18.Komatsu N, Fujita Y and Sato O: “Cenozoic Volcanic
Rocks as Potential Hydrocarbon Reservoirs,” presented
at the 11th World Petroleum Congress, London,
August 28–September 2, 1983.
19.Link WK: “Significance of Oil and Gas Seeps in World
Oil Exploration,” Bulletin of the AAPG 36, no. 8
(August 1952): 1505–1540.
20.Fritts SG and Fisk LH: “Structural Evolution of South
Margin—Relation to Hydrocarbon Generation,” Oil &
Gas Journal 83, no. 34 (August 26, 1985): 84–86.
Fritts SG and Fisk LH: “Tectonic Model for Formation
of Columbia Basin: Implications for Oil, Gas Potential
of North Central Oregon,” Oil & Gas Journal 83, no. 35
(September 2, 1985): 85–89.
21.Johnson VG, Graham DL and Reidel SP: “Methane
in Columbia River Basalt Aquifers: Isotopic and
Geohydrologic Evidence for a Deep Coal-Bed Gas
Spring 2009
XS8
X,200
XS401
XS4
XS602
XS6
XS601
X,400
X,600
X,800
Y,000
Y,200
Y,400
0
km
0
Conglomerate
Shale
Upper volcanic
Sedimentary
Lower volcanic
Basalt
2
mi
2
Y,600
Y,800
R U S S I A
Daqing
N
MONGOLIA
P
A
N. KOREA
A
Beijing
C
0
km 400
0
mi
H
I
N
A
S. KOREA
J
Gas-Bearing Volcanic Formations in China
The giant Daqing field, discovered in 1959, is the
largest oil field in China and one of the largest
in the world. The field has produced more than
10 billion bbl [1.6 billion m3] from sedimentary
layers 700 to 1,200 m [2,300 to 3,900 ft] deep.
Stratigraphic wells—drilled to understand the
basin-scale relationships between the reservoirs
and the surrounding strata—encountered gas in
volcanic layers at depths between 3,000 and
6,000 m [10,000 and 20,000 ft]. Because of the
difficult environment and challenging reservoir
rocks, these reserves were not immediately targeted for development.
In 2004, PetroChina initiated a nine-well
appraisal program and entered into a joint project with Schlumberger to better understand
these deep volcanic reservoirs. The study area
covered 930 km2 [360 mi2] and incorporated 3D
seismic data along with wireline logs, borehole
images and core analyses from 15 wells. To support development decisions, analysts constructed
a workflow to evaluate these complex reservoirs
and estimate the amount of gas in place.27
The initial step in the workflow involved
building a structural model from seismic data.
The top of the Yingcheng volcanic group is a significant seismic reflector, and interpretation of
this horizon supplied the major structural control
for the model. In addition to the top of the group,
seismic interpreters distinguished three main
volcanic sequences, with interbedded and bounding sedimentary sequences (above right). Within
the structural model, each sequence was divided
X,000
Depth, m
Once a hydrocarbon-bearing volcanic deposit
is discovered, evaluating the reservoir can be
a challenge. Methods for assessing porosity,
permeability and saturation in sedimentary rocks
must be modified to work in volcanic provinces.
Case studies from China and India demonstrate
such techniques.
400
> Structure of the Yingcheng volcanic group beneath the Daqing field. Interpretation of seismic data
determined the top of the volcanic group, and integration of seismic and log data allowed delineation
of the upper volcanic, lower volcanic and predominantly basaltic sequences.
into smaller cells that were later populated with
physical properties.
The reservoir consists mainly of interlayered
crystalline rhyolites and rhyolitic pyroclastics, but a
full spectrum of volcanics was encountered, ranging from basaltic to rhyolitic in composition and
from crystalline igneous to pyroclastic in texture.
Identifying rock types within the sequences
and correlating them between wells were difficult tasks. Lithology classification for most types
of rocks relies on mineralogy, which cannot be
determined easily for the very fine-grained or
glassy textures common in volcanic rocks. This
led scientists studying volcanic rocks to focus on
chemical composition as the key factor in classification schemes. With elemental concentrations
from an ECS elemental capture spectroscopy
tool, interpreters used these chemistry-based
classification schemes to provide a continuous lithology description.28 However, chemical
26.Rodríguez Arias L, Galaguza M and Sanchez A: “Look
Source in the Columbia Basin, Washington,” AAPG
Ahead VSP, Inversion, and Imaging from ZVSP and
Bulletin 77, no. 7 (July 1993): 1192–1207.
OVSP in a Surface Basalt Environment: Neuquen Basin,
22.Gunn P: “Aeromagnetics Locates Prospective
Argentina,” paper SPE 107944, presented at the SPE
Areas and Prospects,” The Leading Edge 17, no. 1
Latin American and Caribbean Petroleum Engineering
(January 1998): 67–69.
Conference, Buenos Aires, April 15–18, 2007.
23.Mitsuhata Y, Matsuo K and Minegishi M:
27.Li G, Wang YH, Yang FP, Zhao J, Meisenhelder J,
“Magnetotelluric Survey for Exploration of a VolcanicOilfield
Rock Reservoir in the Yurihara Oil and Gas Field,
Japan,”ReviewNeville TJ, Farag S, Yang XW, Zhu YQ, Luthi S, Hou HJ,
Zhang SP, Wu C, Wu JH and Conefrey M: “Computing
Geophysical Prospecting 47, no. 2 (March 1999):
195–218.09
Winter
Gas in Place in a Complex Volcanic Reservoir in China,”
24.Rohrman M: “Prospectivity of Volcanic Basins:
Trap
Volcanic Fig. 9 paper SPE 103790, presented at the SPE International
Delineation and Acreage De-Risking,” AAPG ORWINT09-VOL
Bulletin 91,
OilFig.
and 9
Gas Conference and Exhibition in China, Beijing,
no. 6 (June 2007): 915–939.
December 5–7, 2006.
25.Laake A: “Remote Sensing Application for Vibroseis Data
28.Barson D, Christensen R, Decoster E, Grau J, Herron M,
Quality Estimation in the Neuquen Basin, Argentina,”
Herron S, Guru UK, Jordán M, Maher TM, Rylander E
paper presented at the IAPG VI Congreso de Exploración
and White J: “Spectroscopy: The Key to Rapid, Reliable
y Desarrollo de Hidrocarburos, Mar del Plata, Argentina,
Petrophysical Answers,” Oilfield Review 17, no. 2
November 15–19, 2005.
(Summer 2005): 14–33.
Coulson S, Gråbak O, Cutts A, Sweeney D, Hinsch R,
Schachinger M, Laake A, Monk DJ and Towart J:
“Satellite Sensing: Risk Mapping for Seismic Surveys,”
Oilfield Review 20, no. 4 (Winter 2008/2009): 40–51.
43
Common depth point number
600
650
700
750
800
850
900
Pyroclastic flow
Lava flow
Pyroclastic fall
Extrusive
FMI Image
50
Porosity
%
0
Facies
Lava
flow
Tuff
Pyroclastic
flow
Water laid
Outer dome-building volcanic
Pyroclastic flow
Middle dome-building volcanic
Pyroclastic fall
Inner dome-building volcanic
Surge flow
Intrusive
Upper lava flow
Middle lava flow
Surge
flow
Lower lava flow
Pyroclastic
fall
> Correlation of igneous rock types with seismic data. Rock types were identified using FMI images,
NMR T2 distributions and ECS elemental concentrations. Rock types were classified into seven
crystalline lithologies (greens, pinks and purples) and four pyroclastic lithologies (orange and yellows).
A sample correlation (bottom) shows an FMI image acquired through an interval of predominantly
pyroclastic layers. A seismic section (top) through the central well is used to extend rock types across
the field. The rock types observed in the central well are displayed at the well location using the color
codes for volcaniclastic and crystalline lithologies. Rock types extrapolated away from the central well
are displayed as semitransparent colors on the seismic section.
32.Kumar R: Fundamentals of Historical Geology and
29.Li GX, Wang YH, Zhao J, Yang FP, Yin CH, Neville TJ,
Farag S, Yang XW and Zhu YQ: “Petrophysical
Oilfield Review Stratigraphy of India. New Delhi: New Age International
Publishers Limited, 2001.
Characterization of a Complex Volcanic Reservoir,”
Winter 09
Transactions of the SPWLA 48th Annual Logging
33.Negi
Volcanic
Fig. 10 AS, Sahu SK, Thomas PD, Raju DSAN, Chand R and
Symposium, Austin, Texas, June 3–6, 2007, paper
E.
Ram J: “Fusing Geologic Knowledge and Seismic in
ORWINT09-VOL
Fig. 10 for Subtle Hydrocarbon Traps in India’s Cambay
Searching
30.Freedman R, Cao Minh C, Gubelin G, Freeman JJ,
Basin,” The Leading Edge 25, no. 7 (July 2006): 872–880.
McGinness T, Terry B and Rawlence D: “Combining
NMR and Density Logs for Petrophysical Analysis in
34.Pal A, Machin N, Sinha S and Shrivastva C: “Application
Gas-Bearing Formations,” Transactions of the SPWLA
of Borehole Images for the Evaluation of Volcanic
39th Annual Logging Symposium, Keystone, Colorado,
Reservoirs: A Case Study from the Deccan Volcanics,
USA, May 26–29, 1998, paper II.
Cambay Basin, India,” presented at the AAPG Annual
Convention and Exhibition, Long Beach, California, USA,
31.Short NM Sr and Blair RW Jr (eds): Geomorphology
April 1–4, 2007.
from Space. NASA (1986), http://disc.gsfc.nasa.gov/
geomorphology/ (accessed March 3, 2009).
44
composition is not the whole story; for example,
if a particular rock has a rhyolitic composition,
chemistry alone cannot distinguish between a
crystalline rhyolite and a pyroclastic rhyolite
tuff. Textural information from borehole images
obtained by the FMI fullbore formation microimager provided the basis for distinguishing these
rock types and tying together log data from all
the wells. Magnetic resonance T2 distributions
provided additional information to complete the
lithology classification.
By combining all available information,
geologists were able to identify 11 igneous rock
types in each well and then correlate them
across the field using seismic data and conceptual geological models from other volcanic
environments (left).
Evaluating the petrophysical properties of
each rock type was particularly challenging.29
Compared with the clastic and carbonate rocks
that form conventional hydrocarbon reservoirs,
these volcanic rocks exhibit the most problematic features of both; the complex mineralogy,
including the presence of conductive minerals
such as clays and zeolites, parallels that of the
most challenging clastic rocks, and their texture and pore structure mimic those of the most
complex carbonate rocks. This combination of
features presents difficulties for the evaluation of
porosity, permeability and fluid saturations.
A robust scheme for lithology-independent
evaluation of porosity in low-porosity, gas-bearing
formations is the DMR density–magnetic resonance interpretation method, which combines
bulk density and magnetic resonance porosity
measurements.30 A relationship between matrix
density and elemental concentrations derived
from core analysis was applied to the ECS results
to produce a continuous log of matrix density.
The matrix density provided input to the DMR
process for calculating high-quality estimates of
porosity and indications of gas saturation in each
well. To extrapolate porosity information to areas
away from the wells, interpreters developed
probability distributions of porosity for each rock
type and used them to populate the model.
Estimating gas saturation was a challenge
because the complex rock texture prevented
development of a suitable Archie-type saturation
equation, so a capillary pressure–based approach
was used to estimate saturation. Pseudocapillarypressure curves were derived from well-log
magnetic resonance T2 distributions and calibrated to mercury-injection capillary-pressure
Oilfield Review
measurements performed on cores. Saturation
values computed in this way showed a strong
dependence on pore network geometry. For
example, the core measurements showed the
air-fall tuffs—volumetrically the most significant reservoir rock type—to be microporous, or
having pore throats less than 0.5 μm in radius.
Saturation profiles across these formations exhibited long transition zones extending hundreds of
meters and covering most of the reservoir. The
saturation results, validated with gas indications
from the DMR method, downhole fluid analysis
measurements and production data, were consistent with the assumption that the reservoir was a
single-pressure system with one free-water level.
The capillary pressure–based approach was
subsequently used to populate the model with
saturation values.
Gas in place for the reservoir was calculated
by summing the gas contained in each model
cell. However, reservoir rock quality in this
field is extremely heterogeneous. In addition,
well control was limited, and the seismic data
were imperfect in guiding the distribution of
petrophysical properties. To cope with these difficulties, engineers employed a stochastic method
to populate cells with porosity and gas saturation. Nearly 60 realizations were performed to
evaluate the potential quantities of gas in place
for the study area, providing an understanding
of the range of uncertainty associated with field
volumetrics. The results of the overall study supported the decision to develop the field.
Oil in India’s Deccan Traps
The Deccan Traps were formed by Late
Cretaceous extrusion of flood basalts that today
cover more than 500,000 km2 [190,000 mi2] of
central western India. They are called traps, from
the German word treppen for step, because they
give rise to topography characterized by stepped
terraces of resistant basalt layers (above right).31
The episode of volcanism was synchronous with
the rifting of the Indian continent from southern
Africa. Although the genesis and the mechanism
of emplacement of these basalts are still debated,
the general consensus is that they erupted under
water.32 More than 40 such basalt layers have
been identified, many of them interbedded with
fluvial and estuarine limestones, shales and
sandstones. In some places, total thickness of the
traps exceeds 3,000 m.
During the last 40 years, Cambay basin, one
of the oldest hydrocarbon plays of western India,
has produced hydrocarbons from sediments
overlying the Deccan basalts.33 Until recently,
Spring 2009
PA
S
KI
TA
N
C H I N A
NEP
AL
Cambay basin
BANGLADESH
Deccan Traps
I
N
D
I
A
Mahabaleshwar
0
0
km
500
miles
500
SRI LANKA
> The Deccan Traps of India. The Deccan Traps are a sequence of approximately 40 basalt layers
covering portions of central western India. Differences between the basalts, which are competent,
and interlayered sands, shales and limestones, which are more easily eroded, give rise to the rough
terrain (right ). This photograph was taken at the Mahabaleshwar escarpment in the Western Ghats.
The Cambay basin (left) is a downdropped graben with oil-bearing sediments overlying the basalts.
Basalt outcrops are shown in orange. (Photograph courtesy of Dr. Hetu C. Sheth, Department of Earth
Sciences, Indian Institute of Technology, Mumbai.)
the top of the volcanic deposits was considered by Well PK-2 was laterally extensive. Based on
economic basement, below which commercial this model, Well PK-6 was drilled in 2005 just
hydrocarbon reservoirs were not expected to be 600 m [1,970 ft] to the southwest of PK-2, but
Oilfield
Review
found. However, in the past few years,
oil has
unfortunately it did not flow any hydrocarbon.
Winter
been discovered in these deeper volcanic
rocks.09 This unexpected result encouraged GSPC to
Volcanic Fig. 11
In 2003, Gujarat State Petroleum Corporation
update
ORWINT09-VOL
Fig. the
11 reservoir model through further data
(GSPC) initiated a six-well campaign in analysis, specifically considering the rock facies
Block CB-ONN-2000/1. The first three wells and fractures and their interplay with faults
exhibited oil shows in the volcanic layers. In within the volcanic layers.34
2004, the fourth well, PK-2, proved to be a signifiAs a first step, geologists developed a textural
cant oil discovery, testing at 64 m3/d [400 bbl/d]. classification of the volcanic layers. Three main
For planning the next well, a simplistic reservoir facies—vesicular basalt, nonvesicular basalt
model was constructed that assumed the hydro- and volcaniclastic units—were identified using
carbon-bearing topmost basalt layer penetrated borehole image logs, petrography from Well PK-1
45
Vesicular Basalt
Nonvesicular Basalt
Well PK-2
Volcaniclastic Rock
Well PK-6
Top
Basalt A
1,775
1,775
Top
Basalt B
1,800
1,800
1,825
1,825
3 cm
Top
Basalt C
1,850
1,850
1,875
1,900
1,900
Depth, m
1,875
> Textural classification of Deccan basalt facies. Images from the FMI borehole
resistivity imaging tool helped geologists identify three main rock types.
Vesicular basalts (left) exhibited vesicles in image (top), in hand specimen
sample (bottom) and also in sidewall cores from a neighboring well.
Nonvesicular basalts (center ) showed no such gas bubbles in borehole
images or in sidewall cores. Images of volcaniclastic basalts (right) showed
fine-scale layering of angular particles. (Basalt photograph courtesy of
Charles E. Jones, University of Pittsburgh, Pennsylvania.)
and hand specimens of basalt (above). Next, the aluminum, iron and titanium for Basalts A, B and
facies were correlated from well to well—an C showed that Basalt A, the top unit, is composiexercise that was far from straightforward. Lava tionally different in the key wells, while Basalts B
flows can commingle, and after solidification and C are compositionally similar (next page).
other changes can occur, such as hydrother- This suggests that the top basalt layer is disconmal alteration, weathering, cementation and tinuous laterally between the two wells, contrary
structural deformation. These changes can be to the assumption in the original model.
Following the facies analysis, the next phase
identified in outcrop, but tracking them in the
subsurface is not easy. Based on image facies and of the study involved characterizing natural fracOilfield Review
log signatures, three main basalt layers, A, B and tures, which are abundant within the volcanic
Winter 09
C, could be correlated between key wells
PK-2 Fig.layers.
Volcanic
12 In the discovery Well PK-2, the top basalt
that
flowed
and PK-6 (above right).
ORWINT09-VOL Fig.
12 hydrocarbon is thick, comprising a
In outcrop studies, volcanic rocks can be cor- nonvesicular basalt layer overlying a vesicular
related using geochemical analysis of major and basalt section with a number of fractures that
minor elemental composition. In the subsurface, appear conductive on borehole images.35 The
similar data can be acquired using the ECS tool. presence of open fractures and vesicles creates
Crossplots of elemental silicon versus calcium, a good-quality reservoir with a dual-porosity
system, and the fracture network enhances per35.In the absence of acoustic or testing data, conductive frac­
meability. In contrast, in Well PK-6, the top basalt
tures on borehole images are considered open to flow.
36.Schutter SR: “Occurrences of Hydrocarbons in and
layer, which is thinner, essentially nonvesicular
Around Igneous Rocks,” in Petford N and McCaffrey KJW
and less fractured, is not a good reservoir.
(eds): Hydrocarbons in Crystalline Rocks, Geological
Society Special Publication 214. London: Geological
In addition to facies type and the presence of
Society (2003): 35–68.
fractures, the geometrical relationship between
46
1,925
Volcaniclastics
Nonvesicular
basalt
Vesicular
basalt
Brecciated zone in
nonweathered basalt
> Initial well-to-well facies correlation. Texturebased facies classification allowed correlation
of three basalt layers between Well PK-2 and
Well PK-6. Basalt A (blue) is the producing zone
in Well PK-2, but not in PK-6. Basalts B and C are
nonproductive.
fractures and faults also seems to play a crucial
role in localizing hydrocarbon accumulations.
In Well PK-2, the open fractures occur at high
angles to a seismic-scale fault, while fractures in
Well PK-6 are aligned approximately parallel to the
fault. Interpreters developed a conceptual model
in which the seismic-scale fault facilitates fluid
communication, allowing the open fractures that
intersect it to conduct hydrocarbons to producOilfieldaligned
Reviewwith the fault are less
ing wells. Fractures
Winter
likely to intersect
it, 09
and therefore are unlikely to
Volcanic Fig. 13
conduct hydrocarbons.
This Fig.
concept
ORWINT09-VOL
13 was validated in a new well, PK-2A1, which contained
conductive fractures oriented perpendicular to
seismic-scale faults and also produced oil.
Future Volcanic Activity
Evaluation of hydrocarbons in volcanic rock presents many challenges, but creative application of
techniques designed for sedimentary reservoirs
is helping oil and gas companies characterize
and exploit these complex accumulations. The
Oilfield Review
Depth, m
Well PK-2
Gamma Ray
Lithology
Image
Logs
Elemental Concentrations, kg/kg
1,760
1,770
Ca/Si
Fe/Si
0.20
0.15
Basalt A
1,780
0.15
0.10
1,790
0.10
1,800
1,810
0.05
Basalt B
0.05
1,820
0
0.10
1,830
0
0.15
1,840
1,850
0.25
0.30
0.35
Al/Si
0.14
Basalt C
0.20
0.05
0.15
0.35
Ti/Si
0.06
0.12
0.25
0.05
0.10
0.04
0.08
0.03
0.06
Depth, m
Well PK-6
Lithology
Gamma Ray
0.02
0.04
0.01
0.02
0
0
1,760
1,770
0
0.10
0.20
0
0.30
0.10
0.20
0.30
Basalt A
1,780
1,790
Ca/Si
0.15
Fe/Si
0.20
1,800
0.15
1,810
1,820
0.10
Basalt B
1,830
0.10
0.05
0.05
1,840
1,850
0
0.10
0
0.15
0.20
0.25
0.30
0
0.35
0.10
0.20
0.30
1,860
1,870
Al/Si
0.14
1,880
1,890
1,900
Ti/Si
0.06
0.12
0.10
Basalt C
1,910
0.06
1,920
0.04
1,930
0.04
0.08
0.02
0.02
0
1,940
0
0
0.10
0.20
0.30
0
0.10
0.20
0.30
> Comparison of basalts in two wells. Elemental concentrations (right) from the ECS tool are expressed as ratios of
calcium, iron, aluminum and titanium to silicon (Ca/Si, Fe/Si, Al/Si and Ti/Si). Ratios are plotted for Basalts A (blue
oval), B (green oval) and C (red oval). In each of the ratio plots, the red and green ovals have approximately the
same relationship to each other, but not to the blue ovals. For example, in the Ca/Si plot for Well PK-2 (top), the red
and green ovals are next to each other, and the blue oval is inside the red oval. However, in the Ca/Si plot for Well
PK-6, the red and green ovals are still next to each other, but the blue oval is inside the green oval. This arrangement
indicates that Basalts B and C correlate from one well to the other, but Basalt A does not.
combination of borehole resistivity images with
neutron-capture spectroscopy and magnetic
resonance logs is becoming the new standard
data suite for evaluation of volcanic reservoirs.
With increased understanding of the capacity of
volcanic rocks to contain oil and gas, other
companies may consider reassessing volcanic
formations they have bypassed, with a view to
reevaluating their potential.
Spring 2009
Unlike their sedimentary counterparts, volcanic rock reservoirs have not been studied
systematically. In addition to the few examples
Oilfield Review
described inWinter
this article,
09 hydrocarbons occur in
or around igneous
rocks
Volcanic Fig.in14more than 100 counORWINT09-VOL
Fig. 14
tries.36 In many
instances, only
oil shows and
seeps have been documented, but further exploration may uncover significant reserves.
The presence of volcanic rocks in a basin
may not ever become a basis for exploration, but
the possibility of such basins sustaining a viable
petroleum system should be included within an
array of options. While some operators might stop
drilling after encountering “basement,” those
with a better understanding of the potential of
volcanic rocks may treat them like any other prospective reservoir rock.
—LS
47
Contributors
James Brady, Product Development Manager for
WesternGeco Electromagnetics (EM) in Houston, oversees technology development for the company’s EM
products. He joined Schlumberger in 1988 as an engineer at the Schlumberger Wireline Systems
Development Center in Austin, Texas, USA. He subsequently worked on land seismic product development
in Hannover, Germany. He has also served as geological
product development manager, product planning and
marketing manager, Petrel* integration manager and
as research and innovation program manager. James
earned an MS degree in electrical engineering at
University of Texas at Austin, and a BS degree in electrical engineering and a BA degree in economics at
University of California, Santa Barbara, USA.
Marco Polo Pereira Buonora is Potential Method
Geophysical Manager at Petrobras in Rio de Janeiro,
where he is responsible for all nonseismic data acquisition and interpretation, such as gravity, magnetic
and EM methods, particularly marine controlledsource electromagnetics (mCSEM). After earning a BS
degree in geology at the Federal University of
Pernambuco in Recife, Brazil, he began his career as a
geologist at the Brazilian Ministry of Mines and
Energy, working on gravity and magnetic data acquisition and interpretation. Later, he worked as a geophysicist on data acquisition, processing and
interpretation of magnetics and gamma spectrometry
data. In 1974, he was granted a Fulbright scholarship
and attended St. Louis University, Missouri, USA,
where he earned Master in Professional Geophysics
(MprGph) and PhD degrees in geophysics, with concentrations in gravity and magnetics. After returning
to Brazil in 1980, he joined Petrobras as a potential
method scientist and has been active in gravity and
magnetic interpretation of several onshore and offshore sedimentary basins in Brazil. In the last four
years, he has been involved with the data acquisition
and interpretation of mCSEM. He is also an associate
professor at the Fluminense Federal University in
Niterói, Rio de Janeiro, where he teaches applied gravity, magnetics and digital signal analysis. He is a member of the SEG and EAGE and of the Brazilian
Geophysical Society, serving as its president from 1989
to 1991.
Chuck Campbell is President and Senior Geoscientist
at ACCEL Services Inc. in Houston. There, he provides
interpretation services in gravity, magnetics and electrical methods. He began his career with Unocal in
1979, moving to Sohio/BP four years later. He has been
with ACCEL Services since 1992. Chuck holds a BS
degree in geology and geophysics from Western Illinois
University in Macomb, USA.
Tracy Campbell, WesternGeco Business Development
Manager in the International Multiclient group, works
to build the company’s multiclient portfolio. He joined
Schlumberger in 2004 with the acquisition of AOA
Geomarine Operations (AGO). Since then, he has been
WesternGeco EM data processing manager in Austin,
WesternGeco EM North America sales manager in
Houston, and WesternGeco EM global projects manager in Houston. Tracy received a BS degree in physics
from University of Alberta, Edmonton, Canada.
48
Alan D. Chave is a Senior Scientist in the Deep
Submergence Laboratory at the Woods Hole
Oceanographic Institution (WHOI) in Woods Hole,
Massachusetts, USA. He holds a BS degree in physics
from Harvey Mudd College, Claremont, California, and a
doctorate in oceanography from the MIT-WHOI Joint
Program in marine science. Alan was involved in the
early development of mCSEM and maintains an active
experimental research group focused on marine electromagnetics, optics and ocean observatory technologies.
Adwait Chawathé is Subsurface Team Leader (Jack
project) for the Chevron North America Exploration
E&P Deepwater Business Unit. Based in Houston, he
manages a subsurface team that evaluates the Jack
and St. Malo developments. He began his career in
1995 as a senior research associate at the Petroleum
Recovery Research Center in Socorro, New Mexico,
USA. Two years later, he joined Chevron’s Simulation
Consulting Team. Before assuming his current position in 2007, he spent more than three years in Kuwait
as Chevron Ratawi asset team leader. Adwait earned
his PhD degree in petroleum and natural gas engineering from The Pennsylvania State University,
College Park, USA.
Leendert Combee is Principal Geophysicist, Marine
Seismic and EM Acquisition, at the WesternGeco Oslo
Technology Center (OTC) in Norway. There, he is
involved with the development of electromagnetic
acquisition systems. In addition, he oversees the geophysical direction of marine seismic acquisition projects. He joined Geco-Prakla in Delft, The Netherlands,
in 1992, as a research scientist studying the geophysical specifications for the Q-Land* system. He continued his research work, extending it to the Q-Marine*
system, at Schlumberger Cambridge Research in
England. Before joining the EM system development
group in 2005, he was geophysical advisor for GecoPrakla receiver systems in Asker, Norway, responsible
for all geophysical development of marine seismic systems including Q-Marine, Q-Seabed,* Q-Fin* and 4Dsteering. Leendert obtained MS and PhD degrees in
electrical engineering and electromagnetic surveying
from Delft University of Technology.
Mohamed Dawoud is Manager of the Natural
Resources Policy Department at the Environment
Agency–Abu Dhabi, UAE. He is also Associate
Professor (on leave) at the Research Institute for
Groundwater, National Water Research Center in
Egypt. Since 1991, he has maintained research, teaching and consulting activities in Egypt, Nigeria, Saudi
Arabia and the UAE. His current research includes
analysis of water supply-and-demand issues; development of a geographic information systems database;
numerical modeling for groundwater flow and management; water management; and the role that improvements in water management can play in reducing
poverty, improving environmental quality and enhancing food security. Dr. Dawoud has a BS degree with
honors in civil engineering from Menoufia University,
Egypt; and MS and PhD degrees from Ain Shams
University, Cairo, through a joint program with
Colorado State University, Fort Collins, USA.
M.Y. Farooqui is General Manager (Planning and
Development) with Gujarat State Petroleum
Corporation (GSPC). He began his career with Oil and
Natural Gas Corporation (ONGC) and supervised operations as wellsite geologist for more than 75 exploration or development wells. Since joining GSPC as
senior geologist in 1994, he has played key roles in all
aspects of the business. He represents the company on
various operating and management committees and
has helped secure domestic and international exploration acreage for GSPC. He has written several technical papers and is an active member of the SPE. He
holds MSc and MPhil degrees in geology from Aligarh
Muslim University, Uttar Pradesh, India, with a specialization in micropaleontology.
Alastair Fenwick is Acquisition Development
Manager for WesternGeco in Houston. He has also
served as global sales and marketing manager for
WesternGeco EM and as senior account manager for
North American marine projects. After earning a BSc
degree in oceanography and sedimentology from the
University of East London in 1982, he joined OHP Ltd.
as a hydrographic surveyor in Aberdeen. He moved to
Geco-Prakla in 1990 as a navigation specialist, soon
becoming a navigation processing supervisor. Alastair
has had varied assignments with WesternGeco in
marine sales and exploration services management in
many locations around the world, particularly in
Southeast Asia.
Arnie Ferster, Exploration Manager, Greenland, for
EnCana Corporation in Calgary, manages the company’s exploration program offshore southwest
Greenland. He joined EnCana in 2000 as a staff geophysicist and worked on new ventures in Libya. He
then worked on new ventures in West Africa before
serving as team leader for projects in Ghana and later
Oman. Arnie holds a BS degree in physics from the
University of Victoria, British Columbia, Canada, and is
an active member of APPEGA and AAPG.
Marcus Ganz has been Marketing and Sales Manager
for WesternGeco Electromagnetics in Houston since
2008. His main responsibility is the commercialization
of the electromagnetics product group. He joined the
company in 1981 to work on seismic field crews. He
has also served as Schlumberger Oilfield Services manager in Argentina and Chile; WesternGeco region manager for South America, based in Rio de Janeiro; and
WesternGeco land general manager, based in Gatwick.
Marcus has a BS degree (Hons) in physics from
University of Southampton, England.
Karen Sullivan Glaser, who is based in Houston, is
Manager of the Reservoir Consulting group, a part of
Schlumberger Data & Consulting Services (DCS). She
oversees a large group of geologists, geophysicists,
petrophysicists and engineers who assist clients in
improving characterization of their reservoirs. She
joined Schlumberger GeoQuest in 1995 and subsequently worked for WesternGeco, Integrated Project
Management and DCS in various technical, marketing
and management roles. Before joining Schlumberger,
she worked for Exxon Production Research as a
research geologist focusing on sequence stratigraphy.
Oilfield Review
She has also worked for Amoco Production Company
in the Permian basin. Karen has a BA degree in geology from Colgate University, Hamilton, New York, USA;
an MS degree in petroleum geochemistry from
University of Oklahoma in Norman, USA; and a PhD
degree in geology from Rice University in Houston.
Stephen Hallinan, WesternGeco Land EM Manager, is
based in Milan, Italy, where he is involved in project
sales, operations and interpretation of magnetotellurics, controlled-source electromagnetics (grounded
dipole and inductive loop) and gravity surveys. He has
more than 15 years of experience as an EM project
geophysicist in various locations around the world.
Stephen received a BA degree in geology from Trinity
College, Dublin, Ireland, and a PhD degree from The
Open University, Milton Keynes, England.
Rolf Herrmann is a Technical Manager for
Schlumberger Water Services in Abu Dhabi, UAE. As
principal hydrogeologist, he is involved in subsurface
exploration and evaluation of aquifers and reservoirs.
He has carried out numerous projects in the assessment of hydrogeological systems and analysis of
dynamic conditions of groundwater aquifers and reservoirs. He also provides expertise in the development of
conceptual models and numerical simulations. Rolf
has served as a project manager for the exploration of
carbonate aquifer structures for underground gas storage, including planning and design, supervision and
evaluation of geological well information and 2D and
3D seismic information. His specialties include aquifer
and reservoir characterization, dynamic simulation
and evaluation of geophysical logs and all aspects of
aquifer storage and recovery systems. He has an MS
degree in geology from The State University of New
York and a BS degree in earth sciences from the
University of Würzburg in Germany.
Huijun Hou is a Senior Geologist for Schlumberger,
providing LWD support in Dhahran, Saudi Arabia. He
joined Schlumberger in 2000 as a geologist in Beijing,
working on interpretation of borehole image logs.
While there, he was part of the team that performed an
integrated characterization of a volcanic gas reservoir
in the Songliao basin. Before assuming his current position, he was operations manager in Sudan and geology
domain champion in Bangkok, Thailand. Before joining
Schlumberger, Huijun worked for the China National
Logging Company. He earned a BS degree in geology
from Jianghan Petroleum Institute in Wuhan, Hubei
province, China, and an MS degree in geosciences from
the University of Petroleum in Beijing.
Younes Jalali is a Schlumberger Reservoir
Engineering Advisor at the Beijing Geoscience Center.
He has been with Schlumberger since 1990, with
assignments in North Africa, Europe, the USA and now
Asia. Prior to joining Schlumberger, he was a member
of the Petroleum Engineering Faculty at Stanford
University, California. He has BS and MS degrees in
petroleum engineering from the University of Tulsa
and a PhD degree in petroleum engineering from the
University of Southern California. He holds a number
of patents related to reservoir and well evaluation and
is a regular contributor to SPE literature.
Spring 2009
Tiziano Labruzzo, who is based in Rio de Janeiro, has
been Senior Geophysical Programmer for WesternGeco
Electromagnetics since 2007. He is responsible for
development of new technologies for marine EM processing and for interpretation of CSEM and marine
magnetotelluric data. He began his career in 1993 as a
software engineer for the National Museum of Science
and Technology in Milan, Italy. He also worked for
Coas srl and Asforil srl, both in Milan, as a software
engineer and consultant, respectively. Before becoming an engineering intern at Schlumberger EMI
Technology Center in Richmond, California, in 2005,
he was a consultant on EM processing and interpretation software for Geoinvest srl, Piacenza, Italy. Tiziano
is a graduate of Università di Milano, with a degree in
computer science.
Guoxin Li is a Senior Petrophysicist and Director of
the Engineering Technology and Supervision
Department of PetroChina Exploration and Production
Company, where he is responsible for drilling, petrophysics and mud logging. He has a bachelor’s degree in
petrophysics and a master’s degree from the China
University of Petroleum.
Nigel Machin is a Senior Reservoir Geologist for the
Central Arabia Group with Saudi Aramco in Dhahran.
He began his career in 1993 with Enterprise Oil in
London, specializing in the application of borehole
images to reservoir evaluation. He has 16 years of
reservoir evaluation experience and worked as a consultant geologist for several operators in Indonesia
including Core Laboratories and Halliburton. He
joined Schlumberger as geology domain champion for
India in 2004 and moved to Saudi Aramco in 2007. His
areas of interest are the evaluation of deepwater and
fluvio-eolian depositional systems and fracture evaluation through borehole imaging techniques. Nigel is a
member of the International Association of
Sedimentologists (IAS) and Society for Sedimentary
Geology (SEPM) and holds a BSc degree in mining and
mineral exploration (applied geology) from the
University of Leicester, England.
Tom Neville is a Petrophysics Advisor and Acting
Research Director at Schlumberger-Doll Research in
Cambridge, Massachusetts. His research focuses on
formation evaluation of unconventional resources.
Before this, he led formation evaluation at
Schlumberger Data & Consulting Services (DCS) in
China. In his 13 years with Schlumberger, Tom has
held a variety of field positions, in engineering, in
research and at headquarters. Before joining
Schlumberger, he had six years of experience as an
exploration and development geologist with independent oil companies in Australia. Tom earned a BS
degree in geology from the University of Queensland,
Brisbane, Australia.
Edward A. Nichols is an EM Specialist at
Schlumberger Riboud Product Center in Clamart,
France. Previously, he was EM discipline manager,
EMI Technology Center, in Richmond, California,
where he was responsible for land and marine geophysical instrumentation products and for borehole
physics. He has also been project manager for technology development and adviser on EM instrumentation.
He began his career in 1977 as a geologist-geophysicist
in eastern Canada with Amax Minerals Exploration. In
1982, he became president of a Canadian consulting
group, Capital Resources. From 1985 to 2004, he
worked for Electromagnetic Instruments Inc. as vice
president R&D, president, operations manager and
consultant geophysicist. The author of numerous publications and holder of many patents, Edward holds a
BS degree in mathematics (Hons) from Mount Allison
University, Sackville, New Brunswick, Canada, and an
MS degree in geophysics from McGill University in
Montreal, Quebec, Canada.
Umut Ozdogan, Lead Reservoir Engineer in the
Chevron Angola Reservoir Management Group, heads
reservoir simulation and engineering activities in the
Takula asset, offshore Angola. He joined Chevron
Corporation in 2003 as a reservoir engineer for the
Tahiti field in the deepwater Gulf of Mexico. Since
then he has led and participated in multiple reservoir
engineering and simulation studies in the deepwater
Gulf of Mexico, West Africa and Eurasia. Umut
received a BS degree in petroleum and natural gas
engineering from Middle East Technical University,
Ankara, Turkey, and an MS degree in petroleum engineering from Stanford University, California.
Aditi Pal, Schlumberger Borehole Geology Team
Leader, is based in Jakarta. She joined Schlumberger
in 2002 in Mumbai, where she worked on interpretation of borehole image data in both clastic and carbonate reservoirs. In 2005, she participated in a multiwell
study involving facies and fracture analysis in a basalt
reservoir in India. She assumed her current position in
2007. Aditi has a bachelor’s degree in geology from the
University of Calcutta and MS degrees in applied geology and geo-exploration from the Indian Institute of
Technology, Mumbai.
Steve Patmore is Principal Explorer, Greenland, for
Cairn Energy Plc in Edinburgh, Scotland. Last year he
joined the new Greenland team with responsibilities
for geophysical operations, interpretation and regional
consulting. He began his career with Conoco in 1974 as
a geophysicist and went on to work in Chad, Norway,
China, North America, Egypt and West Africa. Since
1992, he has worked the UK Continental Shelf including the central Graben, southern North Sea and West
of Shetlands areas. After joining Cairn in 2004 as principal geophysicist working in Indian assets, he then
became asset manager for Nepal and northern India
until taking his current post last year. Steve earned a
BS degree (Hons) in geology and oceanography at the
University of Wales in Swansea.
Mark Riding, Schlumberger Deepwater Theme
Director, is responsible for deepwater corporate strategic planning, sales and technology development worldwide. He began his 27-year career with Schlumberger
as a field engineer for Flopetrol well testing and interpretation services. He transferred to Wireline openhole services in 1990 and has subsequently worked in
various field, sales and management capacities,
including district management for Wireline operations, Trinidad; business manager for Testing services,
Asia; general manager for Testing services worldwide;
and VP and general manager for subsea services worldwide. Recent corporate roles have enhanced his expertise in mergers and acquisitions, HPHT operations and
knowledge management. Mark holds a BS degree in
mining and chemical engineering from the University
of Birmingham, England.
49
Luiz Felipe Rodrigues earned a BS degree in geology
from the Federal University of Rio de Janeiro in 1996.
Since then he has worked in acquisition, processing
and interpretation of high-resolution seismic data for
engineering and exploration projects. He joined
Petrobras in 2000, where he specialized in geophysics,
with emphasis on seismic methods, then acted as seismic interpreter in the Santos basin. His responsibilities include several exploration projects. Luiz is a
member of the Brazilian Society of Geophysicists and
the SEG.
Stewart K. Sandberg, Senior Geophysicist at
WesternGeco Electromagnetics in Houston, interprets
electromagnetic data acquired offshore to map geology and to evaluate potential hydrocarbon horizons.
His more than 30 years of industry experience have
included work as supervising geophysicist, project
manager, assistant professor and private consultant
performing geophysical fieldwork, data processing
and interpretation for geological, hydrogeological and
environmental assessments. From 1996 to 2005, he
was president of Geophysical Solutions, supplying geophysical contracting and consulting services in environmental, engineering and mining exploration
applications. The author of many technical papers,
Stewart received BS degrees in mathematics and in
physics and an MS degree in geophysics, all from the
University of Utah in Salt Lake City, USA. He also
earned a PhD degree in geological sciences from
Rutgers University, New Brunswick, New Jersey, USA.
Chandramani Shrivastva is Schlumberger Geology
Domain Champion, based in Mumbai. He began his
Schlumberger career as a data management geoscientist in New Delhi, India, in 2002. In 2003, he transferred to Mumbai to work on interpretation of
borehole image data. As a borehole geologist, he was
involved in a multiwell project integrating borehole
images, openhole log data and seismic data to characterize fractures and facies in a volcanic reservoir. He
was promoted to his current position in 2007.
Chandramani has a BS degree in geology from Patna
University, Bihar, India, and an MS degree in geological engineering from the Indian Institute of
Technology in Roorkee, Uttarakhand.
Jan Stilling is Chief Geologist for Nunaoil A/S, Nuuk,
Greenland. In 1996, after beginning as a geologist
with PC-Laboratoriet in Fjerritslev, Denmark, Jan
moved to Statoil ASA in Harstad, Norway, to work as a
geologist and then senior geologist. He participated in
evaluation of exploration opportunities offshore
Norway and in the Svale development projects before
joining Nunaoil in 2001. Jan obtained an MS degree in
geology from Aarhus University, Denmark.
Kenneth E. Umbach, a Geophysicist with EnCana
Corporation in Calgary, has been responsible for geophysical interpretation and operations in Greenland
and the Middle East since 2004. He began his career
in 1978 as a geophysical interpreter with Amoco
Canada and subsequently served in Houston, Jakarta
and Calgary. In 1992, he joined PanCanadian to work
on new ventures in North Africa, Europe, the Far East
and Australia. He also worked on geophysical interpretation and operations in Australia and the Middle
East for AEC, the predecessor of EnCana.
50
Frank van Kleef has been Lead Geophysicist for the
Dubai Petroleum Establishment in the UAE since
2007. Before that he was a senior geophysicist for Gaz
de France and worked on assets in Algeria and offshore Netherlands. Frank received an MSc degree in
geology and geophysics from the University of Utrecht,
The Netherlands.
Yuhua Wang is Deputy General Manager of
PetroChina Daqing Oilfield Company. He has more
than 20 years of experience as an exploration geologist and earned a doctoral degree from the Chinese
Academy of Sciences.
Xingwang Yang, a Senior Petrophysicist and DCS
Manager with Schlumberger in Tokyo, manages business for Japan, Korea and Taiwan. He joined
Schlumberger in 1999 as a log analyst. Since then,
he has worked in China and Saudi Arabia on various
petrophysics projects. From 2004 to 2007, he was
involved in evaluating deep gas reservoirs in volcanic
rocks in several Chinese fields. Before joining
Schlumberger, he worked for PetroChina as a wireline
field engineer for two years. Xingwang has a BS
degree in petrophysics from the University of
Petroleum in Dongying, Shandong, China.
Changhai Yin is Director of the Natural Gas
Department in the Exploration and Development
Research Institute of PetroChina Daqing Oilfield
Company. He has more than 20 years of experience as
a petrophysicist and holds a doctoral degree from the
China University of Geosciences.
Andrea Zerilli is a Research Scientist with
WesternGeco Electromagnetic Services in Rio de
Janeiro. With more than 30 years in the oil industry
and worldwide experience in R&D, his current interests include emerging, deep-reading EM technologies,
new high-resolution marine EM technologies, development of integrated solutions and management of multidisciplinary R&E and multiproduct projects. Before
he joined Schlumberger in 2003, he worked for ENI as
research project leader, for KMS Technologies as
director of integrated geophysics, for the USGS as a
visiting scientist and for the Colorado School of Mines
as research associate. Invited speaker and organizer
at many technical society meetings, Andrea holds a
DSc degree in earth sciences from Parma University
in Italy.
Jie Zhao is a Vice Chief Engineer in the Exploration
Company of PetroChina Daqing Oilfield Company. He
has more than 20 years of experience as a petrophysicist and obtained a doctoral degree from the China
University of Mining and Technology.
An asterisk (*) is used to denote a mark of Schlumberger.
An asterisk (*) is used to denote a mark of Schlumberger.
Casing Drilling® is a registered trademark of Tesco Corporation.
Oilfield Review
Coming in Oilfield Review
NEW BOOKS
Guesstimation: Solving the
World’s Problems on the Back
of a Cocktail Napkin
Lawrence Weinstein and
John A. Adam
Princeton University Press
41 William Street
Princeton, New Jersey 08540 USA
2008. 320 pages. US $19.95
(paperback)
ISBN 978-0-691-12949-5
Today, the ability to estimate is a
crucial skill. This book is a collection
of problems from daily life that allows
those with basic mathematics and
science skills to quickly estimate almost
anything using plausible assumptions
and basic arithmetic.
Contents:
• How to Solve Problems
• Dealing with Large Numbers
• General Questions
• Animals and People
• Transportation
• Energy and Work
• Hydrocarbons and Carbohydrates
• The Earth, the Moon, and Lots
of Gerbils
• Energy and the Environment
• The Atmosphere
• Risk
• Unanswered Questions
• Appendixes, Bibliography, Index
One excellent way to start honing
such skills is with a few so-called Fermi
problems, named for Enrico Fermi, the
physicist who delighted in tossing out
the little mental teasers to his
colleagues whenever they needed a
break from building the atomic bomb.
… Dr. Adam and his colleague
Lawrence Weinstein, a professor of
physics, offer a wide and often amusing
assortment of Fermi flexes in a book that
just caught my eye, [Guesstimation].
Angier N: “The Biggest of Puzzles Brought Down
to Size,” New York Times (March 30, 2009),
http://www.nytimes.com/2009/03/31/science/
31angi.html?ref=science (accessed April 22, 2009).
Spring 2009
Flash of Genius: And Other
True Stories of Invention
John Seabrook
St. Martin’s Press
175 Fifth Avenue
New York, New York 10010 USA
2008. 384 pages. US $14.95
ISBN 0-312-53572-4
This book of essays is a collection of
true stories about great ideas. New
Yorker author Seabrook explores the
moment when inspiration strikes in an
otherwise average life, and what
happens when that idea takes on a life,
and commercial possibilities, of its own.
Contents:
• The Flash of Genius
• The Fruit Detective
• Game Master
• Child’s Play
• Sowing for Apocalypse
• The Tree of Me
• Fragmentary Knowledge
• Invisible Gold
• Selling the Weather
• The Slow Lane
• The Tower Builder
• American Scrap
• It Came from Hollywood
• Tremors in the Hothouse
• The Spinach King
… characters, not events are at the
heart of Seabrook’s excellent writing.
Even the technical details of the ideas
and inventions come second, almost
incidental to his first-person
explorations of the people imbued with
or affected by the inspiration.
… Seabrook fits in a surprising amount
of technical detail. You’ll learn, for
instance, how to recover scrap metal or
design skyscrapers that won’t topple
over. But… the real lesson of this book
is a human one: Flashes of genius,
no matter how small, can come from
anywhere and perhaps anyone.
Simonite T: New Scientist 199, no. 2675
(September 27, 2008): 46.
When Science Goes Wrong:
Twelve Tales from the
Dark Side of Discovery
Simon LeVay
Penguin Group
375 Hudson Street
New York, New York 10014 USA
2008. 304 pages. US $15.00
(paperback)
ISBN 978-0-452-28932-1
Neuroscientist LeVay understands the
high potential cost of erroneous
theories and bad information. This book
presents 12 stories of catastrophic
blunders in a wide variety of scientific
disciplines, from engineering geology
and volcanology to microbiology and
nuclear physics.
Contents:
• Neuroscience: The Runner’s Brain
• Meteorology: All Quiet on the
Western Front
• Volcanology: The Crater of Doom
• Neuroscience: The Ecstasy and
the Agony
• Engineering Geology: The Night the
Dam Broke
• Gene Therapy: The Genes of Death
• Nuclear Physics: Meltdown
• Microbiology: Gone with the Wind
• Forensic Science: The Wrong Man
• Space Science: Off Target
• Speech Pathology: The
Monster Study
• Nuclear Chemistry: The
Magic Island
• Epilogue, Sources
Petroleum System Modeling.
Petroleum system modeling, sometimes called charge modeling, uses
seismic interpretations, well logs,
laboratory data and geological
knowledge to model the evolution of
a sedimentary basin to determine if a
reservoir has been filled, or charged,
with hydrocarbons. This article
describes the steps in the process
and explains how this modeling
helps to establish fluid presence and
type with confidence and to assess
exploration risk before drilling.
Coalbed Methane—A Global
Resource. Commercial coalbed
methane (CBM) production was
originally a North American
phenomenon. Techniques used to
tap this unconventional resource,
many introduced in the 1980s, have
changed considerably as CBM
development has increasingly
become a global venture. This article
presents the expanding geographical
scope of CBM production and
describes recently introduced methods
for evaluating, drilling, completing
and producing natural gas from coal.
Crosswell Electromagnetic
Surveys. To better manage producing fields, operators must understand and predict fluid movements
between wells. A recently developed
crosswell electromagnetic induction
system can image resistivity distribution between wells. Formation
resistivity is, in turn, a function of
porosity and fluid saturations. This
new crosswell system illuminates
the interwell reservoir area using a
transmitter in one well and a string
of receivers in another, and it can
propagate signals up to 1 km [0.6 mi]
through a typical oilfield section.
Adding to the drama of each story
are the scientists’ own interpretation of
events. … LeVay has tried, where
possible, to interview all major players
involved. Reading the characters’ own
reflections and opinions in their own
words and then comparing that to the
“facts” makes for an absorbing read.
[The book] is written for both the
scientist and the layperson to enjoy.
Wayman E: Geotimes 53, no. 7 (July 2008): 43.
51
• Lithology and Porosity Estimation
• Saturation and Permeability
Estimation
• Index
Well Logging for Earth
Scientists
Darwin V. Ellis and Julian M. Singer
Springer
P.O. Box 17
3300 AA Dordrecht, The Netherlands
2008. 692 pages. US $99.00
ISBN 978-1-4030-3738-2
This revised, expanded edition of the
classic 1987 text provides detail on a
variety of specialized well logging
instruments used to obtain measurements from the borehole during and
after the drilling process. The book
contains information about the physical
basis of borehole geophysical measurements, as well as an introduction to
practical petrophysics—extracting
desired properties from well log
measurements.
Contents:
• An Overview of Well Logging
• Introduction to Well Log
Interpretation: Finding the
Hydrocarbon
• Basic Resistivity and Spontaneous
Potential
• Empiricism: The Cornerstone
of Interpretation
• Resistivity: Electrode Devices and
How They Evolved
• Other Electrodes and Toroid Devices
• Resistivity: Induction Devices
• Multi-Array and Triaxial
Induction Devices
• Propagation Measurements
• Basic Nuclear Physics for Logging
Applications: Gamma Rays
• Gamma Ray Devices
• Gamma Ray Scattering and
Adsorption Measurements
• Basic Neutron Physics for
Logging Applications
• Neutron Porosity Devices
• Pulsed Neutron Devices
and Spectroscopy
• Nuclear Magnetic Logging
• Introduction to Acoustic Logging
• Acoustic Waves in Porous Rocks
and Boreholes
• Acoustic Logging Methods
• High Angle and Horizontal Wells
• Clay Quantification
52
The collaboration has resulted in a
book that is both authoritative and
lucid and a suitable text for university
curricula. However, it is also an
important reference book for industry
users, describing both the fundamental
physics of well logging and a historical
development of tool design.
I recommend this book highly. The
authors have worked hard and meticulously on the text as a labor of love, as
is obvious on every page. This book
should be on the shelf of everyone who
works with logs or aspires to, as an
introduction, a refresher, and a well
logging reference source.
Doveton JH: AAPG Bulletin 93, no. 2
(February 2009): 293–294.
• Psychrometry, Evaporative Cooling,
and Solids Drying
• Distillation
• Equipment for Distillation, Gas
Absorption, Phase Dispersion,
and Phase Separation
• Liquid-Liquid Extraction and
Other Liquid-Liquid Operations
and Equipment
• Adsorption and Ion Exchange
• Gas-Solid Operations
and Equipment
• Liquid-Solid Operations
and Equipment
• Reactors
• Alternative Separation Processes
• Solid-Solid Operations
and Processing
• Waste Management
• Process Safety
• Energy Resources, Conversion,
and Utilization
• Materials of Construction
• Index
physics. He divides impossibility into
three classes: Class 1 for concepts
beyond current technology, but which
do not violate any known physical laws;
Class II for concepts beyond present
technology that also challenge interpretation of those laws; and Class III for
concepts that defy known physical
laws and would demand huge changes
in our understanding of how the
universe works.
… [the book] remains the definitive
resource for process engineering, and
is a must have for university libraries
and practicing chemical engineers
everywhere. … it will enable a trained
engineer to handle any process design
contingency with confidence.…
Highly recommended.
The study of the impossible has
opened up entirely new vistas for
science, Kaku rightly points out. It is
here that the book’s strength lies: the
impossible is a gateway for discussing
what we still do not understand, those
gray areas that are surely the most
fascinating part of physics.
King MR: Choice 45, no. 8 (April 2008): 1366.
… there is a surprising amount of
heavyweight, cutting-edge science
woven into the fabric of this book.
… [The book] is, in fact, an easy-toread physics primer in disguise.
Kaku has a huge reach as a writer
and speaker.
Perry’s Chemical Engineers’
Handbook, 8th ed.
Don W. Green and
Robert H. Perry (eds)
McGraw-Hill Professional
Two Penn Plaza, 23rd Floor
New York, New York 10121 USA
2008. 2,400 pages. US $199.00
Contents:
• Class I Impossibilities: Force Fields;
Invisibility; Phasers and Death
Stars; Teleportation; Telepathy;
Psychokinesis; Robots; Extraterrestrials and UFOs; Starships;
Antimatter and Anti-Universes
• Class II Impossibilities: Faster
Than Light; Time Travel; Parallel
Universes
• Class III Impossibilities: Perpetual
Motion Machines; Precognition
• Epilogue: The Future of the
Impossible
• Notes, Bibliography, Index
Brooks M: New Scientist 197, no. 2645
(March 2008): 52.
ISBN 0-07-142294-3
First published in 1934, this book has
long been regarded as an expert source
of chemical engineering information.
This updated classic text covers every
aspect of chemical engineering, from
fundamental principles to chemical
processes and equipment to new
computer applications.
Contents:
• Conversion Factors and
Mathematical Symbols
• Physical and Chemical Data
• Mathematics
• Thermodynamics
• Heat and Mass Transfer
• Fluid and Plastic Dynamics
• Reaction Kinetics
• Process Control
• Process Economics
• Transport and Storage of Fluids
• Heat-Transfer Equipment
Physics of the Impossible:
A Scientific Exploration into the
World of Phasers, Force Fields,
Teleportation, and Time Travel
Michio Kaku
Doubleday, Division of Random
House, Inc.
1745 Broadway
New York, New York 10019 USA
2008. 329 pages. US $26.95
ISBN 0-385-52069-7
In this book, noted physicist Michio
Kaku explores the extent to which the
technologies and devices of science
fiction, deemed impossible today, might
become commonplace in the future.
From antimatter to time travel, the
author explores the basics and the
limits of the current known laws of
Oilfield Review