May2016 Oilfield Review

Oilfield Review
May 2016
Marine Seismic Imaging

Microseismic Fracture Mapping
Corrosion Review
Slide DrillingFarther and Faster
Join us on Facebook.
Oilfield Review is now on Facebook. Join to learn when new
issues are published and defining series articles added and
improve your knowledge through definitions selected from
Oilfield Glossary.
For users of the Oilfield Review app, learn when new issues are
available for download.
Download the app by visiting the Apple or Google Play online store
and searching for Schlumberger Oilfield Review.
Oilfield Review
Authoritative. Relevant. Informative.
16-OR-0002
EDITORIAL
Schlumberger and Cameron: A Meeting of Minds and Technologies

In the early 1920s, Jim Abercrombie and Harry Cameron
designed and built the worlds first blowout preventer
(BOP) in their shop in Houston. In France, brothers Conrad
and Marcel Schlumberger were developing methods to
explore the Earths subsurface using electronic sensors.
The Cameron Iron Works Company was established in
Houston in 1920, just six years before the Schlumberger
brothers registered their companySocit de Prospection
lectriquein France.
Although the two companies are of similar vintage, they
pursued different paths to success and now are one.
Cameron built its reputation predominantly on the surface,
providing the equipment necessary for operators to manage
drilling and production operations. Schlumberger success
is anchored by decades developing and refining methods to
remotely explore, understand and produce from reservoirs
below the surface. As is often the case when two companies
join forces, the greatest benefit arises from a unique combination of similarities and differences.
Cameron has constantly defined and refined the surface
technologies and services essential for operators to drill,
complete and produce hydrocarbons. Schlumberger technologies help operators find, access, quantify and optimally
recover hydrocarbons. Both companies have forged much
of their reputations in the most challenging surface and
subsurface environments.
Cameron provides the means and methods to confidently
manage and control flow in all environments, pressure
regimes and temperature ranges throughout the world.
Schlumberger uses technology to help operators understand
and optimize their assets through seismic surveying, downhole logging, testing, laboratory analysis and software platforms. For Cameron and Schlumberger, the meeting place
is the wellhead.
Managed pressure drilling (MPD) is one illustration of
the Schlumberger and Cameron fit. Today, a small but
growing group of engineers is coming to view MPD not as a
specialty method reserved for wells that are otherwise
technically impossible to drill but as a conventional drilling technique. Schlumberger will leverage its research,
software and engineering analytics capabilities and downhole expertise together with Cameron manufacturing, flow
control and automation technology to enable MPD to fulfill
its potential as a standard well construction technique.
Schlumberger and Cameron also share a culture of excellence. They employ award-winning engineering fellows,
holders of PhDs and recognized industry experts. Both
companies invest substantially in research and development;
Inconel
as a consequence, they are recognized as innovators in

their respective industry sectors.
Cameron is adopting wireless technologies to track assets
and monitor equipment condition and is investigating the
advantages of nanotechnology and composite materials.
And both companies will combine existing research programs such as the use of 3D printing to print metal parts
using high-end metals such as Inconel alloy. This technology has the potential to allow engineers to design highly
complex structures that would otherwise be impossible
to manufacture.
The integration of Schlumberger and

Cameron is a story of combining expertise
and innovative technologies.
Schlumberger state-of-the-art mechanical, electronics
and sensing technologies, combined with the rigorous
application of materials science, have led to the development of high-performance downhole drilling systems.
Joining these systems with the Cameron drilling equipment portfolio and automation capabilities will create a
fully integrated land drilling system to create a step
change in operational efficiency.
The integration of Schlumberger and Cameron is a story
of combining expertise and innovative technologies. The
union creates a company that can function as a partnership
internally and with its clients. In the larger view, the integration of surface and subsurface experts can enhance the
performance of drilling and production systems and serve
as a model of the collaboration and commercially beneficial
alignment between operator and service company.
Justin Rounce
Vice President Marketing & Technology, Cameron Group
Houston, Texas, USA
Justin Rounce returns to Schlumberger as the Vice President Marketing &
Technology, Cameron Group from the position of Cameron vice president of
marketing and chief technology officer. Prior to joining Cameron, Justin was
vice president of marketing and technology for OneSubsea, a joint venture
with Schlumberger and vice president and director, mergers and acquisitions
for Schlumberger. Justin joined Schlumberger in 1987 as a testing engineer in
the North Sea region; he moved to the Wireline division in 1992 and then into
various roles within the company, including field operations, operations management, product management and new technology development. In 2003,
he joined the Schlumberger Information Solution Segment as vice president,
software products. In 2007, he was promoted to vice president, software
governance, and was responsible for all internal and commercial software in
Schlumberger. In 2009, he became vice president, marketing and technology
for the Schlumberger Production Group.
is a registered trademark of Special Metals Corporation.
Oilfield Review
May 2016
Volume 28, Number 2
ISSN 0923-1730
www.slb.com/oilfieldreview
Executive Editor
Charlie Cosad
Senior Editors
Tony Smithson
Matt Varhaug
Rick von Flatern
Editors
Irene Frgestad
Richard Nolen-Hoeksema
Contributing Editors
David Allen
H. David Leslie
Ginger Oppenheimer
Design/Production
Herring Design
Illustration
Herring Design
Pennebaker
George Stewart
Printing
RR DonnelleyWetmore Plant
Advisory Panel
Hani Elshahawi
Shell Exploration and Production
Houston, Texas, USA
Gretchen M. Gillis
Aramco Services Company
Houston, Texas
Roland Hamp
Woodside Energy Ltd.
Perth, Australia
Dilip M. Kale
ONGC Energy Centre
Delhi, India
George King
Apache Corporation
Houston, Texas
Michael Oristaglio
Yale Climate & Energy Institute
New Haven, Connecticut, USA
John Thorogood
Drilling Global Consultant LLP
Aberdeenshire, UK
Publishing
Oilfield Review is published and
printed in the USA.
2016 Schlumberger.
All rights reserved.
Articles
4Marine

Imaging in Three Dimensions:
Viewing Complex Structures

Seismic imaging may fail to resolve potential
exploration targets beneath shallow rock
layers. New developments in marine seismic
acquisition and imaging are helping to reduce
this uncertainty.
16

Hydraulic Fracturing Insights from
Microseismic Monitoring
Microseismic monitoring is used for evaluating hydraulic stimulation operations in
unconventional reservoirs. Advanced techniques and new technologies are allowing operators to use microseismic data for effective well
placement and manage stimulations in real time.
34

CorrosionThe Longest War
Oil and gas operations often provide ideal environments for corrosion to develop and grow. In
spite of the challenges created by corrosion, anticorrosion practices in use today help operators
maintain equipment integrity and safely produce
hydrocarbons.
50

A surface-mounted torque-oscillation system
provides consistent weight transfer to the bit,
improving rate of penetration and directional
control during sliding operations.
Departments
1
Editorial
Schlumberger and Cameron: A Meeting of
Minds and Technologies
57
Looking Back
Birth of La Pros: The 90th Anniversary of the
first Schlumberger Company
60
The Defining Series

Selected from the Defining Series online:
Subsea Infrastructure and Geophysics
64
Looking Back
Origins of the Technique of Wireline Logging
On the cover:
Cameron specialists with a seven-cavity TL* offshore ram-type blowout preventer (BOP)
stack at the Cameron facility in Berwick, Louisiana, USA. Cameron became part of
Schlumberger in March 2016. Although the company is most well-known for BOPs, the
first designed in 1922 by company founder Harry Cameron, the portfolio includes a
complete range of rig, wellhead, production and process equipment and services as well
as valves; OneSubsea provides subsea production and processing systems.
An asterisk (*) denotes a mark of Schlumberger.
Article Summaries
Marine Imaging in Three

Dimensions: Viewing Complex
Structures
Prospective reservoirs are typically
located beneath complex rock layers
that act as reflective barriers to
seismic signals and thus distort
images of the subsurface. The challenge to geophysicists is to peer
through the overburden to the reservoir beneath it.
To do so, geoscientists use seismic
survey data, which undergo advanced
processing, imaging, inversion and
interpretation workflows. The results
enable exploration teams to make
immediate decisions about a prospects value.
In challenging offshore frontier
environments, operators are using a
recently developed seismic technology and inversion process to provide
3D, full-bandwidth imaging of finescale structures. Full waveform
inversion results in a model of seismic velocities that can be used with
the seismic data to form an image of
the geology from the surface to the
targets of interest. Page 4.
About Oilfield Review

Oilfield Review is the Schlumberger
flagship journal of technology, innovation
and the science of E&P. Contributors to
articles are industry professionals and
experts from around the world. Those
listed with only geographic locations are
employees of Schlumberger.
Reproductions without permission are
strictly prohibited.
Hydraulic Fracturing Insights

from Microseismic Monitoring
Horizontal drilling and hydraulic
fracturing revolutionized the exploitation of tight and unconventional
oil and gas reservoirs. Engineers
can track the progress of hydraulic
fractures through a formation using
microseismic monitoring.
Armed with new techniques and
technologies, operators are able to
develop unconventional reservoirs
with reduced risk and greater understanding and certainty than ever
before. By mapping the spatial and
temporal patterns produced by microseismic events, they can refine reservoir models and create field
development strategies.
Microseismic monitoring, introduced in the 1980s, evolved as the
application of hydraulic fracturing of
horizontal wells increased since that
time. Using new methods and techniques, geoscientists are improving
the answers that can be derived from
microseismic data. Engineers can
even adjust stimulation programs on
the fly using real-time data to respond
to downhole events as they occur.
Modeling software can predict anticipated downhole responses and then
update the production predictions
based on the data they acquire from
downhole. Page 16.
Oilfield Review online

Visit www.slb.com/oilfieldreview for the
current and all previous editions, videos
associated with certain articles, the
defining series, as well as links to the
apps and Oilfield Glossary.
www.slb.com/oilfieldreview

Corrosion has brought down bridges,
downed aircraft, leveled chemical
plants, parted drillpipe and ruptured
pipelines. Given sufficient time, this
phenomenon has the potential to
degrade any material. In certain
environments, the unchecked effects
of corrosion can come swiftly, and
the consequences of failure to manage corrosion can be costly.
One North Sea operator reported
that outlays for corrosion prevention
and control averaged about 8% of its
total project capital expenditures.
Direct costs associated with those
expenses include replacement of corroded equipment and lost production
and contamination; indirect costs
arise from health, safety and environment concerns.
The environments that host oil and
gas operations often provide ideal
conditions for corrosion. Ongoing
research and advances in coatings,
cathodic protection, nondestructive
testing, corrosion analysis and inhibitors allow operators to safely produce
oil and gas in these corrosive environments. Page 34.
Oilfield Review app

Download the free app by visiting the
Apple or Google Play online store and
searching for Schlumberger Oilfield
Review. Previous issues and video
content are available.
Slide DrillingFarther
and Faster
Directional wells allow operators to
efficiently access reservoirs and
maximize wellbore exposure to productive zones. In many such wells,
directional drillers use steerable
mud motors to kick off the well,
build angle, drill tangent sections
and maintain trajectory necessary to
hit target zones.
When using mud motors, drillers
alternate between rotating and sliding
modes of drilling. In rotating mode, in
addition to the downhole motor, the
drilling rigs rotary table or topdrive
rotates the entire drillstring to transmit power to the bit. During slide
drilling, the drillstring does not rotate,
and the bit is turned by the mud
motor alone. As a consequence, less
weight is transferred to the bit, and
slide drilling is less efficient than
rotary drilling.
An automated torque control system alternates torque direction to
rock the drillstring, which increases
ROP through better transfer of weight
to the bit while in sliding mode. This
weight transfer also helps control toolface orientation. In addition, it helps
minimize the number of downhole
motor stalls and increases bit life by
preventing weight from being transferred to the bit suddenly. Page 50.
Correspondence
Oilfield Review
5599 San Felipe
Houston, TX 77056
United States
(1) 713-513-3760
E-mail: oilfieldreview@slb.com
Marine Imaging in Three Dimensions:

Viewing Complex Structures
Recent developments in multimeasurement marine seismic acquisition and full
waveform imaging enable geophysicists to compensate for distortions caused by
shallow geology and sharpen images of deep targets to reduce the uncertainty of
seismic information.
Anatoly Aseev
Moscow, Russia
Sandeep Kumar Chandola
Low Cheng Foo
PETRONAS Carigali Sdn Bhd
Kuala Lumpur, Malaysia
Chris Cunnell
Malcolm Francis
Shruti Gupta
Peter Watterson
Gatwick, England
Michelle Tham
Kuala Lumpur, Malaysia
Oilfield Review 28, no. 2 (May 2016).
Copyright 2016 Schlumberger.
For help in preparation of this article, thanks to Thomas
Ajewole, M. Nabil ElKady, M. Faizal Idris, Satyabrata Nayak
and M. Iqbal Supardy, PETRONAS Carigali Sdn Bhd, Kuala
Lumpur, Malaysia; and Richard Coates, Houston, Texas, USA.
Dynel 2D, IsoMetrix and Q-Marine are marks of
Schlumberger.
1. Christie P, Nichols D, zbek A, Curtis T, Larsen L,
StrudleyA, Davis R and Svendsen M: Raising the
Standards of Seismic Data Quality, Oilfield Review 13,
no. 2 (Summer2001): 1631.
2. Robertsson JOA, Moore I, Vassallo M, zdemir K,
vanManen D-J and zbek A: On the Use of
Multicomponent Streamer Recordings for Reconstruction
of Pressure Wavefields in the Crossline Direction,
Geophysics 73, no. 5 (SeptemberOctober 2008):
A45A49.
3. For more on full-azimuth seismic surveying and imaging:
Brice T, Buia M, Cooke A, Hill D, Palmer E, Khaled N,
Tchikanha S, Zamboni E, Kotochigov E and MoldoveanuN:
Developments in Full Azimuth Marine Seismic Imaging,
Oilfield Review 25, no. 1 (Spring 2013): 4255.
4. Inline is in the direction the seismic vessel travels and
acquires data; crossline is in the direction perpendicular
to vessel travel.
5. For more on the Q-Marine system: Christie et al, reference 1.
For more on the 3C seismic MEMS accelerometer unit:
Paulson H, Husom VA and Goujon N: A MEMS
Accelerometer for Multicomponent Streamers, paper
WeP606, presented at the 77th European Association of
Geoscientists and Engineers Conference and Exhibition,
Madrid, Spain, June 14, 2015.
Hydrocarbon exploration requires that geoscientists understand the geology of prospective reservoirs often located beneath complex rock layers.
From the geophysicists perspective, the overburden acts as a defective lens, distorting seismic
images of deeper geologic structures. As a result,
targets appear indistinct, distorted, out of place or,
in extreme cases, completely obscured. The geophysicists challenge has been to devise methods
for peering through the overburden and bringing
the underlying geology into focus.
Make-or-break decisions on project viability
often hinge on how well prospective reservoirs can
be imaged, a key factor determining exploration
risk. Operators need accurate images of reservoirs
to help them place exploration wells where they
effectively test the prospect, conduct field planning
and place development wells. In addition to imaging
reservoirs, geophysicists must correctly image the
overburdenthe layers above the reservoirto
reduce drilling risks from operational challenges
such as maintaining a stable wellbore and
controlling formation pressure.
The value that seismic data adds to the
exploration process depends on the quality of
the image produced and the cost incurred in
acquiring such data. Cost-effective seismic
acquisition requires surveying large areas
quickly without compromising data quality and
while minimizing operational and environmental
exposure. Fast acquisition helps shorten the
time frame between the decision to evaluate a
play and the decision to drill.
High-quality data enable exploration teams to
attain a clear understanding of the geology from
the seafloor to the target prospect and then to
decide whether to test and appraise the prospect.
The acquired data must also be suitable for use in
advanced processing, imaging, inversion and
interpretation workflows. These workflows provide vital inputs for geomechanical, reservoir and
basin models.
IsoMetrix marine isometric seismic technology
and full waveform inversion processing are
enabling imaging of complex structures in frontier
areas. The IsoMetrix technology allows for fullbandwidth imaging of fine-scale structures in the
subsurface in all directionsinline, crossline and
verticalfor detailed imaging from seabed to
reservoir. Full waveform inversion results in a
model of seismic velocities, which is used with the
seismic data to form an image of the geology from
the surface to the targets of interest.
This article describes surveys acquired using
IsoMetrix technology in offshore Malaysia and
the North Sea. The survey results demonstrate
the benefits of IsoMetrix technology for overcoming a challenging acquisition environment
and increasing spatial bandwidth and of applying
full waveform inversion for determining overburden and reservoir properties, specifically seismic velocities.
Improving Data and Image Quality
Good seismic imaging requires a chain of factors:
a good acquisition system, optimal survey geometry and accurate processing algorithms and
workflows. More than 15 years ago, Schlumberger
geophysicists embarked on a program to move
from conventional seismic acquisition toward discrete sensor technology. The technology includes
improvements in receiver sensitivity and positioning accuracy, steerable streamers, increased
source control and point-receiver acquisition,
which records traces from individual receivers to
provide consistently repeatable high-quality data.1
These capabilities are evolving. New measurements of the crossline and vertical gradients
Oilfield Review
10 m
Single Channel Interpolation
x
P
Crossline direction
Multichannel Reconstruction
Crossline direction
Figure 1. Streamer element. An element of the IsoMetrix streamer system

(left) combines a hydrophone (inset) that measures pressure (P ) with a
calibrated, triaxial microelectromechanical (MEMS) accelerometer that
measures the axial, or inline (x), radial, or crossline (y), and vertical (z)
accelerations. The IsoMetrix technology facilitates interpolation between
streamers. Using hydrophone streamers (top right), only the amplitude of
variations with distanceof the pressure wavefield enable the signals received from a marine
seismic shot to be processed as a full 3D wavefield rather than as a collection of 2D profiles.2 In
addition, a newly developed, calibrated, broadband marine seismic source provides improved
low-frequency signal content; no source notches,
or missing frequencies, below 150 Hz for all directions within a 20 cone from the vertical; and
cancellation of the source ghosta delayed
reflection of the source from the sea surface.
These acquisition improvements have been
complemented by innovations in marine
surveying geometriesfor example, multivessel
shooting and full-azimuth source-receiver
configurations. Together, these technologies
make it possible to illuminate targets of interest
previously obscured by folded or faulted
sediment, overlying salt layers or other complex
geologic bodies.3
Seismic acquisition and survey geometry are
only the starting points for seismic imaging.
Accompanied by onboard processing capabilities,
data reliability has vastly improved. In addition,
the application of robust seismic inversion and
imaging techniques, such as full waveform inversion and reverse time migration, allow geophysicists to deliver sharper images and estimate rock
properties for explorationists and reservoir
May 2016
the wavefield (blue) can be measured at each streamer location (black

dots). Therefore, the reconstructed wavefield (red) between streamers
is aliased and incorrect. Using multisensor streamers (bottom right), the
wavefield amplitude and gradient (cyan) can be measured at each streamer
location. Consequently, using both attributes of the wavefield, geophysicists
can reconstruct the wavefield accurately between streamers.
engineers who develop static and dynamic

models of the reservoir. These models are based
on the seismic resultsimages, velocities and
horizonsthat are integrated with well data.
Before drilling, explorationists use the models to
predict the petroleum systems present within the
seismically imaged volume, define plays and
locate prospects for drilling. Reservoir engineers
use refinements of these models to plan field
development and, later, manage hydrocarbon
recovery operations.
Imaging Between Streamers
The purpose of IsoMetrix technology is to provide a densely sampled representation of the
wavefield in all directions. An idealized seismic
acquisition system would be able to record the
seismic signals from everywhere below the surface. This capability would maximize the opportunities for separating the signal from unwanted
noise and imaging the reflectors in the subsurface. However, conventional seismic data are
recorded along only a small number of long
streamers towed behind a vessel. Thus, although
conventional seismic data are well sampled in
recording time and along the streamer (inline),
they are not recorded between the streamers
(crossline), which may be separated by large distances of 50, 75 or 100 m [164, 246 and 328 ft].4
As a result, any waves propagating in the
crossline direction may be aliased, or inadequately sampled.

Often, the focus of marine seismic imaging is to
thoroughly sample the wavefield in the reservoir.
However, good sampling of the wavefield in the
overburden is also important because these depths
must be imaged correctly to enable the geophysicist to see clearly into the reservoir. Sampling the
seabed or other interfaces that generate multiple
reflections is important because such reflections
interfere with primary reflections. Shallow depths
are important because of possible seabed and
shallow subsurface hazards to drilling.
Typical marine seismic receivers are hydrophones that record the pressure wavefield only.
Reconstruction of the pressure field between
streamers requires interpolation between
known pressures at each streamer location and
results in crossline pressure fields becoming
aliased and incorrect.
The IsoMetrix technology is based on the
Q-Marine point-receiver marine seismic system
and combines hydrophones for measuring the seismic wavefield pressure with a three-component
(3C) microelectromechanical systems (MEMS)
unit.5 The 3C MEMS unit contains three orthogonal accelerometers for measuring the full 3D vectorial motionmagnitude and directionof the
recorded wavefield (Figure 1).
Survey vessel
Streamers
line
Inli
Cross
ne
Seismic dataset
Figure 2. Marine seismic acquisition via conventional versus IsoMetrix technology. Conventional
surveys (left) are acquired using streamers of closely spaced hydrophones. The resultant seismic
dataset consists of a set of parallel vertical sections. Surveys acquired with IsoMetrix technology (right)
use streamers of closely spaced multisensor receiver units. The multiple components enable
interpolation between streamers, and the resultant dataset is a true 3D grid.
By adding 3C accelerometers, the marine

receivers record the variation of acceleration,
which is proportional to the pressure gradient, or
the spatial derivative of pressure with direction.
In an acoustic material such as water, hydrophones measure the pressure (P) fluctuations
caused by the seismic wave. Three-component
accelerometers measure the accelerations in
three orthogonal directions (ax, ay and az).
Newtons Second Law specifies the force that
results from a difference in pressure; the force is
directed from high to low pressure. The relationship between the difference in pressure with
directionthe spatial derivative of Pand the
acceleration, for example in the x direction, is
ax = P/ x, where is the material density,
and the direction of force is opposite, or negative
to, that of the pressure gradient. This type of relationship holds for each spatial direction (x, y and z)
and allows the calculation of the spatial derivative
of pressure directly from the acceleration measurement. Consequently, knowing the pressure
gradients, geophysicists can reconstruct the unaliased pressure field in all directions. Therefore,
geophysicists can estimate the 3D wavefield
around the streamers using the same spacing in all
directionsinline, crossline and vertical.
Reconstructing the Wavefield
The ability to measure the crossline wavefield
gradient enables geophysicists to acquire marine
seismic data using streamers spaced farther

apart than those in conventional surveys and to
reconstruct the 3D wavefield on a dense grid at
points between streamers (Figure 2). For example, if the actual recordings were accomplished
using eight streamers spaced 75m apart, providing a streamer spread that is 525m [1,720ft]
wide, the wavefield may be reconstructed as if it
were recorded using virtual streamers spaced a
tenth of the distance7.5m [24.6ft] apart.
When wide streamer spacing is used, areas of
exploration can be surveyed faster and more efficiently using fewer sail lines, thereby reducing
survey duration, acquisition cost, operational
complexity and exposure to adverse environmental conditions.
Recording the vertical wavefield component
improves the geophysicists ability to remove
noise, particularly ghost reflections, which are
always present in marine seismic survey recordings. Ghosts are generated when the upward traveling primary signal is reflected downward by the
sea-air interface. This downward traveling ghost is
detected by the seismic receivers and, if uncorrected, causes a frequency dependent blurring of
the final image. Using the vertical acceleration
measurements, the geophysicist can separate the
upgoing and downgoing components of the wavefield, thereby facilitating removal of ghost reflections. The ability to remove the ghosts also allows
IsoMetrix streamers to be towed deeper than
hydrophone-only streamers; towing deep often

reduces other sources of noise such as those
caused by ocean waves and by the motion of the
streamer through the water.
Generalized matching pursuit (GMP) is a processing method that can take advantage of the
multimeasurement data delivered by the
IsoMetrix technology.6 The GMP process operates
on components of the seismic wavefield that are
not confined to traveling straight from the source
to the receiver but instead have a significant
degree of propagation across the streamer
spread. These components may include seismic
reflections, diffractions, multiples or other noise
modes, and, if not treated correctly, can generate
spurious effects in the final images. For example,
any energy arriving from the crossline direction,
which had been spatially aliased previously in
conventional datasets, can now be sampled
appropriately using GMP spatially and temporally by taking advantage of the crossline and
vertical gradient measurements.
The GMP process is data driven and has
proved that it can interpolate the pressure
wavefield accurately in the crossline direction,
even in adverse situations in which the results
from conventional processing would be highly
aliased. The output from the GMP process is a
grid of data channels spaced 6.25m [20.5ft]
apart in the inline direction along virtual
streamers, which are nominally separated by
6.25m in the crossline direction.
The ability to image in 3D enables geophysicists to consider seismic survey acquisition
designs that depart from common practice, as
one operator learned when faced with data
acquisition challenges.
Challenging Acquisition Conditions
To clearly define prospects in the South China
Sea, geophysicists at PETRONAS Carigali Sdn
Bhd acquired a broadband 3D seismic survey offshore Malaysia. The survey area is an elongated
rectangle oriented NWSE. A major NS striking
fault crosses the survey area, and structural dips
6. For more on generalized matching pursuit: zbek A,
Vassallo M, zdemir K, van Manen D-J and
EggenbergerK: Crossline Wavefield Reconstruction
from Multicomponent Streamer Data: Part 2Joint
Interpolation and 3D Up/Down Separation by Generalized
Matching Pursuit, Geophysics 75, no. 6 (November
December 2010): WB69WB85.
7. Chandola SK, Foo LC, El Kaldy MN, Ajewole TO, Nayak S,
Idris MF, Supardy MI, Tham M, Bayly M, Hydal S,
Seymour N and Chowdhury B: Dip or Strike?
Complementing Geophysical Sampling Requirements and
Acquisition Efficiency, Expanded Abstracts, 85th SEG
Annual International Meeting and Exhibition,
New Orleans (October 1823, 2015): 110114.
Oilfield Review
Fault
Average dip
~40 to 50
acc
60
ess
area
30
Average dip
~25
0
Structural dip, EW direction
Major faults, NS direction
Geologic dip of surface of interest
Figure 3. Geologic structure. In the time structure map of the horizon of interest (left), the contour
interval is 100ms two-way traveltime. The black area is a major fault surface that dips to the east.
The white quadrilateral is the survey area, and a no-access area is west of it. The fault is 5 to 8km
[3 to 5mi] wide, has a NS strike and a throw of about 2.5s two-way traveltime. The horizon map on
the rightthe surface area at the prospective reservoir levelshows structural dips that have been
estimated from legacy seismic data. The dips are aligned along a WE trend.
Scenario 1
Scenario 2
Main Survey Area

EW shooting
10 streamers, 8,000 m long,
towed 100 m apart
Patch Survey Area
NS shooting
towed 50 m apart
Entire Survey Area

NS shooting
towed 100 m apart
During seismic processing, the
streamer data are processed
and output to a 6.25-m 6.25-m
grid that is 8,000 m long and
950 m wide.
6 km
m
5
km
Strik
km
Strik
6k
1
Dip
Dip
vey
Are
0 km
Enti
ur
re S
Ma
rve
n Su
y Ar
ea
ea
r
ss a
ea
rea
y Ar
ss a
urve
acce
ch S
No-
Pat
0 km
e
Zon
ault
in F
a
M
acce
No-
May 2016
Average dip
~7
N o-
as high as 50 occur in the area along a WE

trend (Figure 3). The area is bounded on the west
by a no-access zone that survey vessels are not
permitted to enter.7
Typically, optimal seismic acquisition geometry
for conventional 3D surveys requires shooting
parallel to the predominant structural dip
direction. This inline direction facilitates closespaced sampling of the seismic wavefield in the dip
direction, in this case WE, in which the geology
has the most variation. In addition, the typical
conventional seismic bin, or survey subdivision,
into which geophysicists sort seismic traces, is
asymmetric and elongated in the structural strike
direction, which is the crossline direction.
The no-access zone prohibited the vessel from
obtaining full subsurface coverage at the western
edge of the survey and presented an acquisition
challenge to geophysicists, who considered two
options (Figure 4). In the first scenario, they
could acquire most of the survey by shooting
short lines, spaced 100m apart, parallel to dip to
avoid the no-access area. Then, complete the survey using long lines, spaced 50m apart, sailing
parallel to strike adjacent to the no-access zone
boundary; the close line spacing of these strikeparallel lines ensured adequate sampling of the
structural dip. Alternatively, they could acquire
the entire survey using exclusively strike-parallel
sail lines.
The first option was inefficient because of the
two acquisition directions, which required nonproductive time during the many turns and while the
streamers were repositioned for close spacing. The
second option was more efficient for acquiring
data but risked degrading the seismic information
if acquired using conventional technology.
According to conventional wisdom, the strike-parallel survey direction, which had typical line spacing and sampling of the seismic wavefield in the
dip direction, was not ideal for imaging the subsurface and meeting the objectives of company geologists and geophysicists.
The company used IsoMetrix technology,
which enabled symmetric, isometric, or equidistant, sampling of the wavefield in the inline and
crossline directions, to acquire the survey parallel to the structural strike. In addition, the company acquired a smaller swath of data in the
direction of the dominant structural dip, which
would allow comparison and validation of the
integrity of survey shooting in strike.
The data were acquired using ten 8-km [5-mi]
long streamers spaced 100m apart. The streamers were towed at a water depth of 18m [59ft] to
minimize noise from variable currents and
inclement weather during the survey campaign.
Estimated Survey Duration:

Scenario 1 = 2 Scenario 2.
Figure 4. Acquisition options. The survey was restricted by a no-access area on its western
boundary. The company geophysicists considered two options for acquiring the seismic data. In the
first option (left ), the acquisition vessel would sail the main survey area in the dip direction and then
reconfigure the streamers and sail the patch survey area, adjacent to the no-access boundary, in
the strike direction. In the second option (right ), the entire survey area would be acquired by sailing
in the direction of geologic strike and would parallel the no-access boundary. The company chose
the second option and elected to use IsoMetrix technology, which allows for reconstruction of the
wavefield sampled equally in both inline and crossline directions, to acquire the data.
Bathymetry from Multibeam Echo Sounder
Bathymetry from IsoMetrix Technology
5 km
5 km
10 100 m spread
6.25 6.25 m binning
5 5 m sampling
Figure 5. Seafloor features. Bathymetry data (left) were acquired using a multibeam echo sounder; the black arrows indicate features such as sand dunes,
waveforms, mounds and pockmarks on the seafloor. A map of the seafloor surface (right) from the seismic data, which were acquired using IsoMetrix
technology, showed similar features.
After acquisition, the data were preprocessed

and then the full 3D wavefield was calculated
using simultaneous interpolation and deghosting
by means of the GMP method. Next, the upgoing
pressure wavefield (P-wave) was output on a
6.25m by 6.25m grid for each shot record for
further processing and imaging.
The data proved to be of high quality. For

example, a map of the seafloor surface showed
sand banks similar to those observed in bathymetry data obtained using a high-resolution multibeam echo sounder (Figure 5).
Upon comparing the dataset acquired in the
strike direction with that acquired in the dip
Control Swath
Production Volume
1s
1s
5s
5s
Dip shooting direction, inline stack

18 km
Strike shooting direction, crossline stack

18 km
Figure 6. Comparing acquisition directions. Both seismic sections (left and right) are from identical
locations but resulted from perpendicular acquisition directions. The section on the left was from
the control swath acquired in the dip, or crossline, direction and stacked in the strike, or inline,
direction. The section on the right was from the production volume and acquired in the strike direction
but stacked in the dip direction. Except for subtle differences, the sections show similar results
and indicate that the IsoMetrix technology yields similar quality data regardless of the acquisition
direction. The magenta ovals indicate structures, or features, that appear different from one another
as a result of acquisition in the strike direction rather than in the dip direction.
direction, geophysicists judged the datasets to be

similar (Figure 6). The fine spatial sampling of
the wavefield in the inline and crossline directions obtained with IsoMetrix technology enabled
the company to accomplish its geologic and geophysical objectives and achieve acquisition operational efficiency.
In addition to freeing up constraints on seismic
survey acquisition design, uniform inline and
crossline data wavefield estimation facilitates the
increase in spatial resolution and bandwidth
required to compensate for distortions caused by
shallow overburden layers and to sharpen images
of deeper targets. These improvements in resolution and bandwidth helped reduce the uncertainty
of seismic information across the operators drilling prospects.
Broadband in 3D
Oil discoveries at three locations in the southwest Barents Sea have generated significant
interest in exploration of the region. The discoveries offshore northern Norway at the Gohta
prospect in 2013 and at the Alta prospect in 2014
were both by Lundin Norway AS; those at the
Wisting Central prospect in 2013 were by OMV
(Norge) AS. The Gohta and Alta discoveries were
west of the Loppa High, a roughly 150km [90mi]
long and 100km [60mi] wide tilted fault block
that has been affected by a series of events in the
North Atlantic Ocean that include:
Paleozoic rifting
Mesozoic opening of the North Atlantic Ocean
and of the Greenland and Norwegian seas
Quaternary glaciation.
Oilfield Review
W
0.5 s
Bjrnyrenna
fault complex
Hoop
fault complex
Bjrnyrenna
fault complex
A
A
Hoop
fault complex
Loppa High
5.0 s
N
0.5 s
Asterias
fault complex
Asterias
fault complex
B
B
Variance
Low
5.0 s
High
Reflection amplitude
Figure 7. Fault system. This seismic time slice (left) at 1,100ms is through
the Loppa High; the seismic attribute is displayed to emphasize the variance
in seismic reflectivityareas of high variance values are colored from
black to red and yellow. Three major fault systems, which show up as areas
of high variance, affected the Loppa High. The WE striking Asterias fault
complex crosses the Loppa structure in the south; the southern portion of
the SWNE striking Hoop fault complex cuts across and forms the narrow
The WesternGeco seismic vessel Western

Trident acquired the East Loppa Ridge survey in
2014. The survey covered 4,777km2 [1,844mi2]
and is part of the Schlumberger Multiclient
Barents Sea program. The program used
IsoMetrix technology to record wide spatial
bandwidth datathe recorded wavefield contains the fine-scale detail necessary to represent
subsurface geology accurately.
In conventional 3D seismic surveys, a common objective is to acquire broadband surveys of
high temporaltraveltimebandwidth and
resolution. The ideal broadband survey has a
wide band, or range, of frequencies and is
acquired at a high sample rate. The objective for
maximizing temporal bandwidth is primarily to
maximize resolution in depthto image thin
beds and small faults.
Geology is best understood by observations in
three dimensions, which requires maximizing
May 2016
Loppa High graben and the Bjrnyrenna fault complex separates the
Loppa High from the Bjrnya basin (not shown) on the west. SectionsA
and B (right top and bottom) display grabens associated with the fault
systems. The northern portion of the Loppa structure is in the center of
SectionA. SectionB shows graben structures in the north associated with
the Hoop fault complex and in the south associated with the Asterias fault
complex, which separates the Loppa High from the Hammerfest basin.
spatial bandwidth in all directions. In the spatial

domain, the wavenumber (k) is the spatial
frequency, or the number of wavelengthswavecycle lengths per unit distance. The
wavenumber is analogous to the temporal
frequency (f) or the number of wave periods
wave-cycle timesT per unit time. Wavenumber
in the space domain and frequency in the time
domain are related through the phase velocity
(vp), which is equivalent to wavelength divided
by period (vp = / T ), frequency divided by
wavenumber (vp = f/k) or wavelength times frequency (vp = f ).8 Consequently, for 3D seismic imaging of geology, the notion of broadband
must be expanded to include 3D spatial bandwidth and resolution.
The East Loppa Ridge survey was acquired
using 12 streamers that were 7km [4.3mi] long,
spaced 75m apart and towed at a constant depth
of 25m [82ft]. After acquisition, the datasets
were preprocessed and then simultaneously spatially dealiased and receiver-deghosted in 3D by

means of the GMP method.
The tectonic, stratigraphic and petroleum
systems geology of the southwest Barents Sea
region is complex.9 The structural setting resulted
from several tectonic events that established a
dense mosaic of fault systems (Figure 7). The
8. In this context, phase refers to a wave of a single
frequency in a wave train. The phase could be of a
compressional (P) wave, shear (S) wave, other waves or
their associated reflections and refractions; the waves
velocity is the phase velocity.
9. Henriksen E, Ryseth AE, Larssen GB, Heide T, Rnning K,
Sollid K and Stoupakova AV: Tectonostratigraphy of the
Greater Barents Sea: Implications for Petroleum
Systems, in Spencer AM, Embry AF, Gautier DL,
Stoupakova AV and Srensen K (eds): Arctic Petroleum
Geology. London: The Geological Society, Memoir 35
(August 9, 2011): 163195.
Gernigon L, Brnner M, Roberts D, Olesen O, Nasuti A
and Yamasaki T: Crustal and Basin Evolution of the
Southwestern Barents Sea: From Caledonian Orogeny to
Continental Breakup, Tectonics 33, no. 4 (April 2014):
347373.
W
0.5 s
5 km
5.0 s
BCU
Seafloor
Tertiary
Upper Triassic-Jurassic
Lower-Middle Triassic
Late Carboniferous to
Permian carbonates
Carboniferous postrift
Evaporites
Carboniferous synrift
Pre-Carboniferous
basement
Figure 8. East Loppa regional seismic section. The seismic section (top) runs from the Loppa Ridge
in the west toward the Ottar basin in the east. The interpretation (bottom) is a balanced section,
which was modeled using the Dynel 2D restoration and forward modeling tool. This section shows
extensional rifting and synrift and postrift sediment deposition during the Carboniferous period.
During the Late Carboniferous to Permian, a carbonate platform developed and evaporate was
deposited. During the lower to middle Triassic uplifting and tilting of the Loppa High, karstification
of the carbonates and sedimentation of shales occurred. The upper Triassic and Jurassic periods
were characterized by high clastic sedimentation rates and floodplain development from rivers and
deltas. Rifting occurred again during the upper Jurassic to lower Cretaceous; the base Cretaceous
unconformity (BCU) defines the transition from synrift to postrift sedimentation. Finally, Tertiary
sediments occur above the BCU to the seafloor.
10. Gabrielsen RH, Frseth RB, Jensen LN, Kalheim JE and

Riis F: Structural Elements of the Norwegian
Continental Shelf. Part I: The Barents Sea Region,
Stavanger, Norway: Norwegian Petroleum Directorate,
NPD Bulletin no. 6, May 1990.
11. For more on the Gipsdalen Group: Larssen GB,
ElvebakkG, Henriksen LB, Kristensen S-E, Nilsson I,
SamuelsbergTJ, Svn TA, Stemmerik L and Worsley D:
Upper Paleozoic Lithostratigraphy of the Southern
Norwegian Barents Sea. Stavanger, Norway: Norwegian
Petroleum Directorate (2002).
Barents SeaCarboniferous to Permian Plays,
Norwegian Petroleum Directorate, http://www.npd.no/
en/Topics/Geology/Geological-plays/Barents-Sea/
Carboniferous-to-Permian/ (accessed August 29, 2015).
10
12. For more on petroleum systems: Al-Hajeri MM,

Al SaeedM, Derks J, Fuchs T, Hantschel T, Kauerauf A,
Neumaier M, Schenk O, Swientek O, Tessen N, Welte D,
Wygrala B, Kornpihl D and Peters K: Basin and
Petroleum System Modeling, Oilfield Review 21, no. 2
(Summer 2009): 1429.
13. For more on the Snadd Formation: Klausen TG,
RysethAE, Helland-Hansen W, Gawthorpe R and
Laursen I: Regional Development and Sequence
Stratigraphy of the Middle to Late Triassic Snadd
Formation, Norwegian Barents Sea, Marine and
Petroleum Geology 62 (April 2015): 102122.
14. Houbiers M, Wiarda E, Mispel J, Nikolenko D, Vigh D,
Knudsen B-E, Thompson M and Hill D: 3D FullWaveform Inversion at MarinerA Shallow North Sea
Reservoir, Expanded Abstracts, 82nd SEG Annual
Loppa High area contains three major fault complexes.10 The Asterias fault complex forms the
southern boundary, which separates the Loppa
High from the Hammerfest basin to its south. The
southern portion of the Hoop fault complex
strikes SWNE and cuts across the Loppa structure as a narrow graben. The Bjrnyrenna fault
complex separates the Loppa High from the
Bjrnya basin on the west. Broadband seismic
images make it possible to delineate the fault
patterns and establish the regional structural
framework within the East Loppa Ridge survey
area. The structural framework influences local
petroleum systems.
The Gohta and Alta oil discoveries were in
reservoirs located in carbonates of the Gipsdalen
Group, which were deposited in warm, shallow
marine environments during the Late
Carboniferous to Permian periods and, since
then, have been altered by dolomitization and
karstification (Figure 8).11 Additional petroleum
systems elements in the Loppa High area include
reservoir prospects in Triassic sandstones, source
rocks in Carboniferous synrift and postrift sediments and in Permian and Triassic sediments
and seals formed by Triassic and Cretaceous
shales.12 The broadband East Loppa Ridge seismic dataset offers an opportunity for detailed
interpretation of the complex geology in the
Loppa High area.
The upper Paleozoic carbonates have been
the most promising stratigraphic level for Loppa
High exploration (Figure 9). Broadband seismic
data facilitate detailed mapping, analysis and
interpretation of the carbonate morphology,
which has polygonal ridges characteristic of modern carbonate platforms.
Oil has been discovered in the upper Triassic
Snadd Formation but at locations with low reservoir quality. Within the Snadd Formation, the
broadband seismic data reveal the fluvial system and aid automated mapping, which should
reduce the uncertainty of locating higher quality reservoir sands. The data show complex
International Meeting and Exhibition, Las Vegas,

Nevada, USA (November 49, 2012).
Houbiers M, Mispel J, Knudsen BE and Amundsen L:
FWI with OBC Data from the Mariner Field, UKThe
Impact on Mapping Sands at Reservoir Level, paper
We1105, presented at the 75th European Association of
London, June 1013, 2013.
stmo S, McFadzean P, Silcock S, Spjuth C, Sundvor E,
Letki LP and Clark D: Improved Reservoir
Characterisation by Multisensor Towed Streamer
Seismic Data at the Mariner Field, paper WeP0312,
presented at the 76th European Association of
Amsterdam, June 1619, 2014.
Oilfield Review
Figure 9. Carbonate reservoir. The seismic time horizon of the top surface of
the Late Carboniferous to Permian-age Gipsdalen Group is shown in map
(left ) and perspective (right ) views. These views show a seismic attribute
that emphasizes edges on the surface. The surface in both views displays
polygonal ridges that are reminiscent of polygonal oceanic reef systems
observed in modern carbonate platform environments (inset).
fluvial and floodplain geology and reveal that

the channel system is associated with floodplain development (Figure 10).13 The data
reveal a variety of fluvial features, including
point-bar systems, clustered channel fill complexes and ribbon-channel sandstone bodies;
the ribbon channels were at depths greater
than 1,000 m [3,280 ft] and estimated to be less
than 100 m wide.
The East Loppa Ridge survey demonstrates
the imaging power of acquiring true 3D, broadband seismic data. High spatial resolution in all
directions facilitates and improves imaging of
complicated 3D geology such as fault networks,
anastomosing fluvial channel complexes and carbonate platform deposition and karstification.
The increased detail offered by broadband images
promotes improved understanding of petroleum
system geology and better discrimination of lithologies and their rock properties.
Full Waveform Inversion
Geophysicists use full waveform inversion (FWI)
for calculating horizontal and vertical seismic
wave velocities of geology from the surface to targets of interest. The result is a velocity image in
depth that reveals the sought-after structural
and depositional information.
Traditional migration produces an image of the
subsurface by attempting to reposition, or migrate,
seismic data reflection points to their correct locations in 3D space. A velocity model is almost
always an input to migration; and a refined velocity model may be a byproduct of migration.
Unlike conventional migration, FWI is a
method for building a velocity model by attempting
to match the complete recorded wavefield that
May 2016
results as seismic waves travel through the Earth

and encounter changing properties in the
subsurface geology. The starting point for FWI is
an approximate model of velocities. Geophysicists
use this velocity model to simulate the recorded
wavefield. They then subtract the simulated
wavefield from the observed wavefield to obtain
the residual wavefield. The residual wavefield
is then backward propagatedextrapolated
downward in space or backward in timethrough
the velocity model to obtain a dataset of velocity
gradients. These gradients inform where to
increase or decrease velocities but not by how
much. To calculate a velocity model update, the
gradients are multiplied by a step length, which
scales the gradients. The velocity updates are
added to the current velocity model to create a
new velocity model, and the process is repeated.
The iterations continue until the residual
wavefield is acceptably small, meaning that the
modeled wavefield closely approximates the
observed wavefield. The final model of seismic
velocities can be used as an input to migration to
produce an image that better represents subsurface rock characteristics or may be used directly to
interpret rock and fluid properties.
This technique was used in Mariner field, discovered in 1981 and located about 150km [93 mi]
east of the Shetland Islands on the UK Continental
Shelf in the North Sea. The field is under development by operator Statoil UK Limited with partners
JX Nippon Exploration and Production (UK)
Limited and Dyas UK Limited. The field consists of
two reservoirs. The shallow reservoir contains
heavy oil of 12.1API gravity and is about 1,200m
[3,940ft] below sea level in sands of the Heimdal
member of the Middle to Late Paleocene Lista
Variance
Low
High
Figure 10. Floodplain channels. This seismic

time slice at 1,100 ms is at the depth of the upper
Triassic Snadd Formation. The time slice shows
the variance in reflectivity. The dark linear and
curvilinear features are faults. The lighter gray,
sinuous and interweaving features are networks of
fluvial channels crisscrossing a floodplain.
Formation, composed predominantly of shale. The

deeper reservoir in the Maureen sandstone
member contains heavy oil of 14.2API gravity and
is at the base of the Early Paleocene Vle
Formation at depths of 1,400 to 1,500m [4,590 to
4,920ft] below sea level.
The Mariner field presents various challenges
for seismic imaging.14 The shallow overburden
above the reservoirs contains channel sands that
have higher seismic velocities than those of surrounding geologic units. These sands can be
mapped easily, but their presence causes distortions in the images of the reservoir zones beneath
them. For example, shallow, high-velocity channel sands cause pull-ups of, or apparent structural high spots in, underlying reflectors. The
Heimdal reservoir sands consist of complex channel sands as well as sand injectites, or sand intrusions; these sands are difficult to image because
of their low impedance contrast with the shales
that host them. The Maureen sandstone contains
small-scale faults and calcite layers that are
important for developing production from the
11
sandstone but are below the detection

capabilities of traditional seismic techniques.
The imaging challenges presented by the
reservoirs may be mitigated by full waveform
processing techniques that enable removal of the
distortions caused by the high-velocity channel
sands in the shallow overburden.
In 2012, the operator acquired a broadband
seismic survey at the Mariner field using the
WesternGeco IsoMetrix technology. The survey
data were acquired using eight streamers, each
3km [1.9mi] long, spaced 75m apart and towed

at a constant depth of 18m. After acquisition, the
data were preconditioned and then simultaneously interpolated and deghosted using the GMP
method. The upgoing pressure wavefield was
then output on a 6.25m by 6.25m grid for subsequent processing and imaging.15
Initial inspection of the dataset showed it to
be richer in high frequencies than in two conventional 3D seismic datasets and richer in low frequencies than in an earlier ocean bottom cable
(OBC) survey. Both qualities are important for
resolving subsurface geology and velocities
through inversion of seismic data. High frequencies enable resolution of relative velocities
between small stratigraphic and structural
details. Low frequencies facilitate determination
of absolute velocities, which are calibrated
against borehole data.
The data underwent fast-track processing,
using prestack time migration, which demonstrated the Heimdal member sands could be
imaged more reliably using the broadband data
than the earlier data.16 The operators geoscientists were able to establish the relationship
between seismic reflectors and geologic horizons
with improved confidence.17 Encouraged by these
results, WesternGeco geophysicists applied FWI to
the broadband dataset.18
The starting point for FWI is a velocity model
(Figure 11). The geophysicists began with a simple
model, using seismic velocities interpreted from
sonic logs from wells in the area of the Mariner
field, which were then interpolated laterally
between the wells along layers bounded by known
geologic horizons. Based on previous processing
studies, the overburden formations were assumed
to be anisotropic; the P-wave anisotropy parameters epsilon () and delta () were initially defined
as linearly increasing from the seafloor to the base
Cretaceous unconformity but were subsequently
updated using a multiparameter inversion step in
the FWI workflow.19
The geophysicists wanted to know whether the

results of FWI would isolate the velocities in
shallow channel sands within the overburden. As a
test, one of the known channels delineated from
legacy 3D seismic data was inserted into the initial
velocity model and given a higher velocity than its
host units. If successful, the FWI method would
sharpen the velocities within this control channel
but also pick out other channels in the area.
To compensate for velocity imprecisions
introduced by interpolation, the geophysicists
applied one iteration of common image point
(CIP) tomography to the interpolated velocity
model. Common image point tomography is an
iterative method of inverting for seismic velocities using seismic reflections. During an iteration, the amount of residual moveoutdepth
variationalong reflections in prestack depthmigrated (PSDM) CIP gathers is used to determine adjustments in the velocity model to bring
the subsequent version of the PSDM image into
better focus.20 After one iteration of CIP tomography, the velocity model was smoothed and ready
for input to the FWI process.
Next, the geophysicists started the FWI process, which, beginning with the initial earth
model of velocities, iteratively models the
observed seismic wavefield and adjusts the velocities in the earth model until there is an acceptable match between the modeled wavefield and
the recorded wavefield.21 The observed wavefield
was the upgoing P-wave wavefield that had been
isolated at an early stage of processing from the
broadband dataset. The criterion for convergence to an acceptable match between synthetic
and observed wavefields is to minimize a misfit
function that quantifies the difference between
the modeled and measured data. To ensure that
the FWI process converges on the global, or true,
minimum rather than a localized minimum, the
geophysicists conduct FWI in stages. First, they
find an acceptable fit of the low-frequency
wavefield. They then add and fit to successively
15. zbek et al, reference 5.

16. Migration is a seismic processing step in which
reflections in seismic data are moved to their correct
locations. Time migration locates reflections in two-way
traveltimefrom the surface to the reflector and back
as measured along the image ray. Depth migration
locates reflectors in depth. Mathematically, migration is
performed by various solutions to the wave equation
that describe the passage of seismic waves through
rock. Kirchhoff migration is a ray-based approximation
founded on the integral solution to the wave equation
derived by 19th-century German physicist
GustavKirchhoff.
For more on migration and imaging: Albertin U, Kapoor J,
Randall R, Smith M, Brown G, Soufleris C, Whitfield P,
Dewey F, Farnsworth J, Grubitz G and Kemme M:
The Time for Depth Imaging, Oilfield Review 14, no. 1
(Spring 2002): 215.
17. stmo et al, reference 14.
18. Gupta S, Cunnell C, Cooke A and Zarkhidze A:
High-Resolution Model Building and Imaging Workflow
Using Multimeasurement Towed Streamer Data: North

Sea Case Study, Expanded Abstracts, 85th SEG Annual
International Meeting and Exhibition, New Orleans
(October 1823, 2015): 10491053.
19. The base Cretaceous uniformity is the term applied to a
strong seismic reflection surface that is mappable over
much of the continental shelf in the North Sea. The
reflector is an unconformity that is located close to the
bottom of Cretaceous-age rocks and separates
sediments deposited before rifting of the North Sea from
sediments deposited after rifting.
Anisotropy is the variation of a physical property,
such as P- or S-wave velocity, with the direction
of its measurement. For more on elastic anisotropy:
Armstrong P, Ireson D, Chmela B, Dodds K, Esmeroy C,
Miller D, Hornby B, Sayers C, Schoenberg M, Leaney S
and Lynn H: The Promise of Elastic Anisotropy,
Oilfield Review 6, no. 4 (October 1994): 3647.
Epsilon () and delta () are P-wave parameters that
describe vertical transverse isotropy. Epsilon is the
P-wave anisotropy parameter and equal to half the ratio
of the difference between the horizontal and vertical

P-wave velocities squared divided by the vertical
P-wave velocity squared. Delta is describes nearvertical P-wave velocity anisotropy and the difference
between the vertical and small-offset moveout velocity
of P-waves. For more on seismic anisotropy parameters:
Thomsen L: Weak Elastic Anisotropy, Geophysics 51,
no. 10 (October 1986): 19541966.
20. For more on CIP tomography: Woodward M, Nichols D,
Zdraveva O, Whitfield P and Johns T: A Decade of
Tomography, Geophysics 73, no. 5 (September
October 2008): VE5VE11.
21. Vigh D, Starr EW and Kapoor J: Developing Earth
Models with Full Waveform Inversion, The Leading
Edge 28, no. 4 (April 2009): 432435.
22. For more on multiscale inversion: Bunks C, Saleck FM,
Zaleski S and Chavent G: Multiscale Seismic
Waveform Inversion, Geophysics 60, no. 5 (September
October 1995): 14571473.
23. Reference 16.
Build initial velocity model
Control test for FWI: Insert known channel

into velocity model
One iteration of common image point (CIP)

tomography to smooth velocity model
Low-frequency FWI iterations using

frequency band, 1.5 to 7 Hz
peak frequency, 2.5 Hz
Intermediate-frequency FWI iterations using

frequency band, 1.5 to 13 Hz
peak frequency, 5 Hz
One iteration of CIP tomography to refine

velocities in deepest intervals of the model
Multiparameter FWI to refine

anisotropic parameters
Final high-frequency FWI iterations to enhance

the resolution of velocities
Figure 11. Workflow for full waveform inversion.
12
Oilfield Review
higher frequency bands until there is an

acceptable fit of the full-frequency wavefield.
This sequential FWI procedure stabilizes the
inversion algorithm and ensures that the process
converges to a global minimum.22
Application of FWI to the broadband dataset
collected at Mariner field showed that a seismic
dataset acquired using IsoMetrix technology can be

inverted for a geologically relevant seismic velocity
model that is capable of sharpening the focus of
seismic images. After FWI processing, the velocity
model was input into two prestack depth migration
algorithms: a Kirchhoff depth migration (KDM) to
compare directly against legacy data volumes and a
high-frequency reverse time migration (RTM) performed directly in the natural shot domain after
GMP.23 The velocity model from FWI sharpened the
image of the control channel embedded into the
overburden of the initial velocity model and
highlighted additional channels (Figure 12).
Before FWI Processing

2,250 m/s
Depth slice, 158 m
1,700 m/s
2,025 m/s
Depth slice, 278 m
1,625 m/s
2,225 m/s
Depth slice, 844 m
2,500 m/s
1,500 m/s
2,025 m/s
After FWI Processing

2,250 m/s
Depth slice, 158 m
1,700 m/s
2,025 m/s
Depth slice, 278 m
1,625 m/s
2,225 m/s
Depth slice, 844 m
2,500 m/s
1,500 m/s
Figure 12. Comparing models before and after full waveform inversion (FWI).
Both seismic sections (left top and bottom) show the same geology to a
depth of 1,200m [3,940ft] below sea level. The depth sections are the result
of Kirchhoff depth migration (KDM); the sections are overlain by the velocity
model (colors) that was used as input to KDM. The top section resulted
from KDM using the initial velocity model. The control channel is in the top
center and was given a higher velocity than its surroundings. The bottom
May 2016
2,025 m/s
section resulted after using the velocities output after completion of FWI.
The control channel is in better focus, and the velocities of other channels
are evident. The velocities of the overburden units have become more
defined. The images on the right are depth slices at 158, 278 and 844m
[518, 912 and 2,770ft] below sea level. Compared with the before FWI
processing results, geologic features (yellow arrows) have become better
defined after FWI processing.
13
Legacy KDM Model
1,700
1,500
Onla
Two-way traveltime, ms
1,600
Isolated depositional sandstones

in the Frigg Formation mudstones
Onlap
1,800
1,900
2,000
2,100
1,600
1,700
1,800
1,900
2,000
2,100
Heimdal sandstone
1 km
Concept 1: isolated depositional sandstones

Heimdal sandstone
Sandstone intrusions crosscut the

Balder and Frigg Formation mudstones
Crestal intrusion fringe
n
trusio
-like in
Wing
Two-way traveltime, ms
1,500
Heimdal sandstone
1 km
Concept 2: sandstone intrusions crosscut mudstones
Top of the Frigg sandstone

Top of the Balder sandstone
Top of the Sele Formation
Top of the Heimdal sandstone
KDM using FWI Model

Heimdal sandstone
RTM using FWI Model
Heimdal sandstone
Figure 13. Modeling results. A clear progression of improvement occurs from the legacy velocity
model and Kirchhoff depth migration (KDM, top) to the revised KDM using FWI model (middle) to the
high-resolution reverse time migration (RTM) also using the FWI model (bottom). The progression
demonstrates improved imaging of the steep dips and signal-to-noise characteristics in the reservoir
section, discriminating the Heimdal injectite, or intrusion, features (circled) from the background Lista
shales. The inset shows conceptual sketches of how the sandstone bodies or intrusions might have
become incorporated into the Lista shales above the Heimdal formation. (Inset adapted from Huuse
et al, reference 24.)
14
The velocities in the shallow layers became

more clearly defined. Below them, the reservoir
zones of interest were less distorted. Cross sections
through the KDM image volume showed that the
velocities from FWI made a demonstrable difference in the focusing and positioning of overburden
formations, while the RTM image volume gave the
best resolution and signal-to-noise discrimination of
Heimdal injectites against the background Lista
shales (Figure 13).24
The IsoMetrix marine isometric seismic technology and full waveform imaging are enabling and
complementary technologies for increasing the
qualitative and quantitative accuracy of seismic
information. The IsoMetrix technology allows
deghosting and interpolation of the recorded
wavefield to produce unaliased seismic records. In
turn, FWI provides geologically relevant velocities
at scales that can be used to bring the overburden
into focus. Together, these techniques enable geophysicists to image reservoir targets more clearly
(Figure 14).
Advances in the sequence of steps from seismic data acquisition to final imaging are helping
operators characterize the subsurface more
distinctly. Measurements of the pressure
wavefield and its gradients using IsoMetrix
technology represent a significant development
24. For more on injectites: Braccini E, deBoer W, Hurst A,
Huuse M, Vigorito M and Templeton G: Sand Injectites,
Oilfield Review 20, no. 2 (Summer 2008): 3449.
Huuse M, Cartwright J, Hurst A and Steinsland N:
Seismic Characterization of Large-Scale Sandstone
Intrusions, in Hurst A and Cartwright J (eds): Sand
Injectites: Implications for Hydrocarbon Exploration and
Production, AAPG Memoir87. Tulsa: AAPG (2007): 2135.
Oilfield Review
Ocean Bottom Cable Survey
IsoMetrix Survey
1.7 km
in marine seismic data acquisition. The

development of circle shooting, simultaneous
firing of sources and full-azimuth source-receiver
configurations embody advances in marine
seismic survey geometry and design. Full
waveform inversion, along with reverse time
migration, is advancing geophysicists capability
to develop data-driven velocity models. The
converging improvements on all three fronts
acquisition, survey design and processing
provide the means for imaging complex geologic
structures, forecasting drilling hazards and
illuminating reservoir targets.
RCNH
Figure 14. Comparing images from ocean bottom cable (OBC) and IsoMetrix technologies. Both images
are seismic depth sections to a depth of 1,700m [5,600ft] below sea level. They show the same
geology extracted from datasets that have been processed using similar workflows through FWI
and prestack depth migration; in each case, the color overlay is the P-wave velocity model that
results after processing. For the 2008 OBC survey (left), the FWI processing was completed to a peak
frequency of 10 Hz before migration using KDM. For the 2012 survey using IsoMetrix technology (right),
the FWI processing was completed to a peak frequency of 5 Hz, followed by migration using highresolution RTM. Despite some differences in the two workflows, both used a 2.5Hz peak frequency for
the first FWI updates. After processing, the velocity model result from IsoMetrix technology has the
same, or better, resolution in the shallow overburden as the model result from the OBC survey.
Contributors
Anatoly Aseev, based in Moscow, was a Seismic
Interpreter for Schlumberger Multiclient seismic
projects from 2014 to 2016, with focus on the
Norwegian Continental Shelf area. He began his career
in 2006 as a geologist with Rosneft in Krasnodar,
Russia, and worked on exploration projects in
the Ciscaucasia basin. He joined Schlumberger
PetroTechnical Services (PTS) in 2011 and served
as a geologist and then senior geologist working on
exploration projects in the Timan-Pechora, West
Siberia, Barents Sea and West Black Sea basins.
Anatoly holds an MSc degree in petroleum geology
from the North-Caucasus Federal University, Stavropol,
Russia. He is pursuing a PhD degree in regional
geology from Lomonosov Moscow State University.
Sandeep Kumar Chandola is a Custodian of
Geophysics with PETRONAS Carigali Sdn Bhd in
Kuala Lumpur. He served with Oil and Natural Gas
Corporation, the Indian national oil company, for
more than 20 years before joining Petronas Carigali
in 2005. His work has supported the design of 3D
acquisition geometries and the introduction of new
geophysical technologies to the company. He has a
masters degree in physics from Hemvati Nandan
Bahuguna Garhwal University, Sringar, Uttarakhand,
India, and a specialized diploma in petroleum
geophysics from the Indian Institute of Technology
(IIT) Roorkee, Uttarakhand. He is a member of the
SEG, the European Association of Geoscientists and
Engineers and the Society of Petroleum Geophysicists
(India), an SEG Honorary Lecturer and an adjunct
lecturer at Universiti Teknologi PETRONAS, Malaysia.
Sandeep has authored more than 50 publications and
is a recipient of the National Petroleum Management
Programme Award for Excellence from the government
of India.
Chris Cunnell leads Technical Sales and Marketing
of IsoMetrix* for WesternGeco in Gatwick, England.
Chris, who has more than 20 years of geophysics
experience, joined Schlumberger in 1997 and has
May 2016
worked in technical, service and marketing managerial

positions in the UK, US and Egypt. Before joining
the team for IsoMetrix technology, Chris was based
in Cairo and managed advanced imaging services,
including full waveform inversion and Seismic
Guided Drilling* service, across the Middle East
and North Africa. He received an MBA degree from
the Rotterdam School of Management at Erasmus
University, the Netherlands.
Low Cheng Foo is Custodian of Geophysical Acquisition
for PETRONAS Carigali Sdn Bhd in Kuala Lumpur,
where he is involved with new technology projects
such as broadband, multicomponent, multiazimuth
and full azimuth seismic data acquisition. He has 35
years of experience with the company. Previously, he
was head of acquisition after serving as an acquisition
and processing geophysicist. He has been involved in
land, marine and transition-zone seismic acquisition
programs in various countries in Southeast Asia, the
Middle East, Suriname and Cuba. Low earned a BSc
(Hons) degree in physics, majoring in geophysics, from
the University of Science Malaysia in Penang.
Malcolm Francis is a Schlumberger Advisor and the
Eastern Hemisphere Exploration Services Manager
for WesternGeco in Gatwick. Before his current role,
he held technical and management positions as the
Eastern Hemisphere multiclient chief geophysicist,
global manager of geology and interpretation
and senior manager E&P solutions. Earlier in his
career, Malcolm managed the special processing
and interpretation departments. He began in the
industry in 1980 with Western Geophysical, where he
undertook collaborative research with Saudi Aramco.
He obtained a bachelors degree in geology from the
University of Manchester, England, and MSc and PhD
degrees in geophysics from Imperial College London.
Malcolm is a member of the European Association
of Geoscientists and Engineers, the SPE, SEG and
Petroleum Exploration Society of Great Britain and is
a Fellow of the Geological Society of London.
Shruti Gupta is an Area Geophysicist for Schlumberger

in Gatwick, England, where she provides technical
support for time and depth processing of marine and
ocean bottom cable (OBC) seismic data. Shruti, who
has more than seven years of experience in the oil and
gas industry, started her career with Schlumberger
as a field geophysicist on a land and transition zone
seismic acquisition crew in Egypt. She then worked
with the depth imaging group in Houston. She has an
MSc degree in applied geology from the IIT Kharagpur,
West Bengal.
Michelle Tham is the Technical Support Manager
for WesternGeco in the Asia Pacific region as well
as the Petrotechnical Expertise Discipline Career
Manager for the Schlumberger Asia region; she is
based in Kuala Lumpur. She began her career with
Schlumberger in Calgary and has worked in the US,
Myanmar, Indonesia, Australia, Nigeria, UAE and
Malaysia. Before her current position, she served as a
seismic data processing geophysicist, data processing
supervisor, staff geophysicist, area geophysicist,
seismic survey design and modeling manager and
geophysics global discipline career manager. Michelle
holds a BS degree in geophysics from the University
of Calgary.
Peter Watterson is the Manager of the Marine
Geosolutions Technology Commercialization group
for WesternGeco in Gatwick. His focus is on research,
engineering and marketing of various marine seismic
acquisition and processing technologies. Pete
began his career in the geophysics industry with
Western Geophysical in 1991 in London. He has held
positions in seismic data processing and technology
management in the UK, Venezuela, US, Trinidad and
Brazil and worked for several years as the regional
geophysicist for WesternGeco for South America. He
received a BSc degree in physics from the University
of Leeds, England.
An asterisk (*) denotes a mark of Schlumberger.
15
Hydraulic Fracturing Insights from

Horizontal drilling and hydraulic fracturing revolutionized the exploitation of tight and
unconventional oil and gas reservoirs. Microseismic monitoring provides operators
with crucial information to improve these operations and helps reservoir engineers
with modeling and making decisions on well placement, completion design and
stimulation operations.
Jol Le Calvez
Raj Malpani
Jian Xu
Houston, Texas, USA
Jerry Stokes
Mid-Continent Geological, Inc.
Fort Worth, Texas
Michael Williams
Cambridge, England
For help in preparation of this article, thanks to Julian Drew,
Perth, Western Australia, Australia; Tony Probert and
Ian Bradford, Cambridge, England; and Nancy Zakhour,
Callon Petroleum, Houston.
CMM, ECLIPSE, Mangrove, MS Recon, NetMod, Petrel,
ThruBit, UFM, VISAGE, VSI and VSI-40 are marks of
Schlumberger.
1. Seismic waves convey energy by means of the particle
motion of solid materials.
2. For more on fracture diagnostic techniques: Bennett L,
Le Calvez J, Sarver DR, Tanner K, Birk WS, Waters G,
Drew J, Michaud G, Primiero P, Eisner L, Jones R,
Leslie D, Williams MJ, Govenlock J, Klem RC and
Tezuka K: The Source for Hydraulic Fracture
Characterization, Oilfield Review 17, no. 4
(Winter 2005): 4257.
3. A tiltmeter measures minute rotationschanges of
inclinationof the ground in which it is embedded.
16
Operators producing from unconventional reservoir plays face many challenges. Fluid flow
through unconventional reservoir rocks is limited by matrix permeability, which is generally
several orders of magnitude smaller than that of
conventional reservoir rocks. Preexisting faults
and fracture networks often provide pathways
for the flow of hydrocarbons and play an important role in increasing reservoir drainage volumes. Hydraulic fracture stimulation treatments
can often connect the wellbore to existing natural fracture networks; however, effective stimulation requires knowledge of the distribution of
those networks.
Well completion engineers use geomechanical and fracture models to plan where to initiate
hydraulic fractures and predict their propagation
through the reservoir. These models require calibration and validation. Microseismic monitoring
has proved to be a viable means for calibrating
the models and for providing empirical data
about the effectiveness of stimulation operations.
Microseismic monitoring is a technique that
records and locates microseismic eventscollectively referred to as microseismicitywhich
are small bursts of seismic wave energy generated by minute rock movements in response to
changes of the in situ stresses and rock volume
such as those that occur during fracture stimulation operations.1 During these operations, fractures are created by injecting fluid at high
pressure. These fractures propagate and are then
held open using a solid proppant. Mapping the
spatial and temporal pattern of these events has
proved successful for monitoring the progress of
hydraulic fractures as they advance through and

alter a formation.
Engineers may employ several techniques to
determine the effectiveness of hydraulic stimulation operations.2 For instance, during stimulation
operations, microseismic (MS) monitoring and
tiltmeter measurements can indicate mechanical
changes in the subsurface that occur over a wide
area centered on the treatment well.3 Afterward,
engineers have used radioactive and chemical
tracers, temperature tools and production logs to
provide complementary indications of changes in
fluid pathways resulting from the stimulation.
Geophysical service companies often acquire
MS data, which they interpret and integrate with
other measurements to provide oil and gas operators with an understanding of hydraulically
induced fracture systems. The primary data used
for evaluating MS events are waveform measurements acquired from a network of receivers
placed either downhole or at the surface.
Geoscientists use these data to map the extent
and evolution of MS events. These maps provide
valuable information related to strain and stress
variations in the reservoir and surrounding formations and are used to guide stimulation decisions during job execution. If MS events indicate
undesired fracture growth or fault activation,
operators may choose to terminate stage pumping early, use diverter technology or skip stimulation stages.
Microseismic monitoring also provides information about the nature of the physical processesinduced fracturing of the rock or slippage
on preexisting fracturesthat occur at the
Oilfield Review
May 2016
17
Treatment well
Monitoring well
Microseismic event
Sensors
Reservoir
Stimulated volume
Anchoring
arm
Shaker
Three y
component x
accelerometers z
Isolation
spring
Coupling
contacts
Shaker
z
x
Figure 1. Hydraulic fracture monitoring from a vertical well. Multicomponent

sensors in a vertical monitoring borehole record microseismic events caused
by hydraulic fracturing (top). Event locations determined from data processing
allow engineers to monitor the progress of stimulation operations. To
acquire high-fidelity seismic data, the VSI versatile seismic imager (bottom)
uses three-axis (x, y and z) geophone accelerometers (inset ) that are
acoustically isolated from the tool body by isolation springs. The VSI service
is mechanically coupled to the casing or formation by a hydraulically
powered anchoring arm. The
acquisition
engineer can test the coupling
Oilfield
Review
quality by activating an internal
MAYshaker
16 before operations begin. The VSI-40
40-shuttle versatile seismicMicroseismic
imager allowsFig
up 1to 40sensor packages, or
shuttles, to be linked together;
however,
12shuttles
ORMAY 16 MCSMC 1 are typically used in
hydraulic fracture monitoring operations.
18
location of the MS sources. Characterization of

the population of MS sources helps quantify the
magnitudes and directions of stress and displacement variations in the affected reservoir volume
during the stimulation. To describe the magnitude
and direction of the rock movements at each
source location, geophysicists process MS waveform recordings, account for propagation effects,
determine the radiation pattern of the acoustic
emission and invert for source propertiesrock
movements and energy released.4 Reservoir engineers then combine the space-time evolution of
source characteristics with additional information to determine the state of stress and fluid flow
paths in the reservoir. From this information, they
make productivity predictions, which help operators develop and manage their reservoirs.
In this article, we review the acquisition, processing and interpretation of MS monitoring
data. Advances in these areas are described, and
workflows that integrate MS data into geomechanical modeling and reduce interpretation
uncertainty are presented. A case study from an
unconventional reservoir in Arkansas, USA, illustrates performance trade-offs for surface and
downhole acquisition geometries. Case studies
from Texas, USA, illustrate how MS monitoring
has added value to stimulation operations by
helping geoscientists identify fault interactions,
fracture growth and stage-to-stage variability of
stimulation responses.
Typical Monitoring Systems
Microseismic monitoring (MSM) is the detection of signals generated by small seismicor
microseismicevents. Engineers began using
this technique during hydraulic fracturing in oil
and gas operations as early as the 1980s.5 The
Cotton Valley Consortiuma research group
studying hydraulic fracturing of the Cotton Valley
Formation play in Texas and Louisiana, USA
used microseismic monitoring to understand
fluid flow in a Cotton Valley reservoir in 1997.6
Operators also successfully applied MSM in evaluating fracture stimulations in the Barnett Shale in
Texas, which helped improve their understanding
of the fracture network development during stimulations, avoid geohazards and enhance production.7 These early MSM operations incorporated
arrays of three-component (3C) geophones or
accelerometers deployed near reservoir depth in
a nearby vertical monitor well (Figure 1).8
Oilfield Review
Extensive hydraulic fracturing of horizontal

wells began after 1997 as a result of Mitchell
Energys successful application of the method in
the Barnett Shale; MSM from adjacent horizontal
boreholes soon followed. The use of sensor arrays
in horizontal wells next led to the evaluation of
MSM performance.
The effectiveness of the sensor array geometry
depends on the layout of the monitoring and
treatment wells. Monitoring from vertical wells
close to treatment stages results in improved
location accuracy for in-zone and out-of-zone
microseismicity. Monitoring from nearby horizontal wells often provides coverage along the length
of a stimulated lateral well; similar coverage may
be unavailable from surface arrays or vertical
monitor wells. The recording geometry may
require cost trade-offs between monitoring the
entire treatment well and detecting MS events
that may occur outside the target interval.9
Microseismic events that are detected outside
the targeted interval can indicate unintended
consequences of the stimulation program such
as breaching the reservoir seal or activating existing faults.10
Knowledge of MS measurement accuracy is
crucial for understanding the validity of MS data
interpretations.11 Survey designers have developed modeling software that predicts the minimum detectable event magnitude with respect to
the distance of the monitoring array from source
locations. The software also outputs estimates of
the associated uncertainty for locating and characterizing MS events.12 Accuracy of the estimated
event hypocenters3D locations using easting
and northing geographic Cartesian coordinates
along with the depth of event initiation points
is affected by the monitoring geometry and the
accuracy of the velocity model that is used to
transform waveform arrival times at recording
instruments to distances of the instruments from
the events.
The precision of hypocenter estimates depends
on the geophone array geometry and data errors,
which influence the determination of event arrival
time and the direction of arrivals at the receivers.13
During stimulation operations, extraneous, highamplitude noise sources are numerous. As a consequence, the low signal-to-noise ratio (S/N) is one
of the greatest challenges in the acquisition and
processing of MS data.
Early MSM from a single monitoring well provided valuable information, although it had
shortcomings. Microseismic monitoring from
single wellbores imposes the requirement that all
multicomponent sensors have the same vector
May 2016
fidelityaccuracy for measuring signal magnitude and directionbecause accurate waveform

polarization information is crucial for determining the direction to each event hypocenter.14 In
addition, seismic tools must record incoming MS
signals with the same spectral fidelityaccuracy for measuring frequency contentwithin
the typical signal bandwidth of 10 to 1,000Hz
used in these operations. When monitored from
single wells during a multistage stimulation,
some stages may be too distant from the sensors
for reliable event detection or characterization.
Sensor geometries along a single linear array are
insufficient to determine the source mechanismssize, direction, orientation and duration
of 3D rock movementsassociated with MS
events; thus, microseismic engineers seek to
record seismic waveforms from multiple observation points and azimuths.
Valid interpretation of MS data also requires
careful signal analysis. The processing of MS data
is preceded by the construction and calibration of
a model of seismic P-wave and S-wave velocities
extending from the planned stimulated volume
of interest to the receiver array. Geophysicists

calibrate the velocity model using perforation
shot, string shot, checkshot or VSP survey data.15
Analysts originally performed event detection and
localization processing using P-wave and S-wave
arrival time picks and polarization from 3C MS
waveforms.16 Today, event localization algorithms,
such as the CMM coalescence microseismic
mapping procedure, use an automatic scanning
and grid search algorithm that correlates signal
traveltimes and waveform polarizations to locate
hypocenters.17 Multicomponent waveforms are
processed to assess how well the observed timing and polarization of arrival phases across
the receiver array match the modeled values
associated with potential hypocenter locations
in the volume of interest. Arrival time picking
can also be performed automatically and then
refinedmanually.
Analysts interpret MS locations to show
induced fracture extentlength, height and
azimuth. However, stimulations in tight and
unconventional reservoirs often produce complex, nonplanar hydraulic fracture geometries.
4. The radiation pattern is a description in 3D space of

the amplitude and sense of initial motion of P and S
wavefronts as they propagate away from the initiation
position of a microseismic event. For more on seismic
sources and their radiation patterns: Lay T and
WallaceTC: Modern Global Seismology. San Diego,
California, USA: Academic Press, 1995.
5. For more on the context for microseismic monitoring:
Maxwell SC, Rutledge J, Jones R and Fehler M:
Petroleum Reservoir Characterization Using Downhole
Microseismic Monitoring, Geophysics 75, no. 5
(SeptemberOctober 2010): 75A12975A137.
6. The Cotton Valley formation is a Cretaceous-age tight
sandstone that stretches from Texas to northern Florida,
USA. The main play produces mostly natural gas and is
located in north Louisiana and northeast Texas. For more
on the Cotton Valley Consortium project: Rutledge JT,
Phillips WS and Mayerhofer MJ: Faulting Induced by
Forced Fluid Injection and Fluid Flow Forced by Faulting:
An Interpretation of Hydraulic-Fracture Microseismicity,
Carthage Cotton Valley Gas Field, Texas, Bulletin of
the Seismological Society of America 94, no. 5
(October 2004): 18171830.
7. For more on the use of microseismic monitoring in the
Barnett Shale: Maxwell S: Microseismic: Growth
Born from Success, The Leading Edge 29, no. 3
(March 2010): 338343.
8. Seismic data acquired from three-component (3C)
geophones use three orthogonally oriented geophones
or accelerometers. The early Schlumberger
microseismic acquisition system typically included a
VSI tool with eight 3C geophones.
9. For more on the accuracy of hypocenter estimates:
Maxwell S and Le Calvez J: Horizontal vs. Vertical
Borehole-Based Microseismic Monitoring: Which
is Better?, paper SPE 131780, presented at the
SPE Unconventional Gas Conference, Pittsburgh,
Pennsylvania, USA, February 2325, 2010.
10. Induced seismicity refers to earthquakes that are
attributable to human activities, which may alter the
local stresses and strains in the Earths crust and cause
rock movements that generate earthquakes.
11. Maxwell, reference 7.
12. To predict sensor network performance, Schlumberger

engineers used the NetMod microseismic survey design
and evaluation software. For more on microseismic
survey design: Raymer DG and Leslie HD: Microseismic
Network DesignEstimating Event Detection,
presented at the 73rd EAGE Conference and Exhibition,
Vienna, Austria, May 2326, 2011.
13. The hypocenter, or focus, is the point within the Earth
at which rupture starts during an earthquake or
microseismic event. The point directly above it on
the Earths surface is the epicenter.
14. For more on vector fidelity: Berg EW, Rykkelid, Woje G
and Svendsen : Vector Fidelity in Ocean Bottom
Seismic Systems, paper OTC 14114, presented at
the Offshore Technology Conference, Houston,
May 69, 2002.
15. In a checkshot survey, seismic specialists measure the
traveltime of seismic waves, usually P-waves, from the
surface to known receiver depths. A vertical seismic
profile (VSP) is a more extensive survey in which
geophones are placed at regular, closely spaced
positions in the borehole. Both surveys use a seismic
source positioned on the surface. Perforation shots and
explosive string shots serve as seismic sources in the
treatment well, and traveltimes are measured downhole
or at the surface. In all cases, the source and receiver
locations are known and, from the observed traveltimes,
velocity may be calculated.
16. The P-waves used for seismic processing are elastic
body waves, or sound waves, in which particles
oscillate in the direction the wave propagates. The
S-waves are elastic waves in which particles oscillate
perpendicular to the direction in which the wave
propagates.
17. Drew J, Bennett L, Le Calvez J and Neilson K:
Challenges in Acoustic Emission Detection and
Analysis for Hydraulic Fracture Monitoring, paper
presented at the 17th International Acoustic Emission
Symposium, Kyoto, Japan, November 912, 2004.
Drew J, Leslie D, Armstrong P and Michaud G:
Automated Microseismic Event Detection and Location
by Continuous Spatial Mapping, paper SPE 95513,
presented at the SPE Annual Technical Conference
and Exhibition, Dallas, October 912, 2005.
19
7,200
17:00
19:00
20:00
Time, hr:min
18:00
Top of Barnett Shale

7,600
Z-axis
Depth, ft
21:00
22:00
8,000
Bottom of Barnett Shale

South
7,120,000
Y-axis SN, ft
7,130,000
8,400
North
Figure 2. Microseismicity and effective stimulated volume. Located events (circles, color-coded
according to time) were generated during the stimulation of a horizontal well (red line) in the Barnett
Shale in Denton County, Texas. Analysts built 3D cells in a model of the monitored volume of the reservoir.
They counted the number of events that exceeded a predetermined threshold in each cell and calculated
the resulting stimulated volume within the cells. The green envelope gives the effective stimulated
volume, estimated here at 180millionft3 [5million m3]. Yellow and blue disks on the horizontal well
mark perforation clusters for stimulation Stages 1 and 2, respectively. Isolated events shown outside
the green envelope are not considered hydraulically connected to the stimulated volume. (Adapted from
Le Calvez et al, reference 59.)
Consequently, geophysicists compute the effective stimulated volume (ESV) as a measure of MS

activity. The size, shape and extent of the ESV is
based on a distribution of event locations and
their uncertainties (Figure 2).18 The ESV provides
information on the complexity of the hydraulic
fracture network. A long, narrow ESV is probably
dominated by a single planar through-going fracture, whereas a short, wide ESV probably consists
of a complex, multibranching fracture network.
Taking the Broad View
Engineers may conduct MSM from a single well,
multiple wells, grids of shallow wells, surface
arrays or networks of surface sensor patches. To
meet acquisition goals, they may also combine
a variety of designs (Figure 3).19 Typically, analysts employ numerical simulation techniques
that account for signal frequency content and
attenuation and that make use of source, geology and noise models. They may use statistical
analysis to predict the number of detectable
events for given monitoring geometries. Analysts
also recognize the importance of accounting
for anisotropy in the velocity models. Seismic
velocity and attenuation tomography based on
crosswell surveys can be used to constrain these
models.20 During perforating, data acquired
from surface sensors can provide calibrated
P-wave traveltimes.21
20
Monitoring from surface and near-surface

positions offers a potentially larger field of view
than that from monitoring wells alone, and it
eliminates the need for providing dedicated deep
monitoring wells.22 Surface monitoring enables
long treatment laterals to be monitored along
their entire lengths. However, because the S/N is
often low, locating and characterizing MS events
using data recorded at the surface may be difficult. To overcome low S/N and detection uncertainty, survey designers use receiver arrays
containing many hundreds to thousands of sensors. Data from multiple points can be processed
to reduce noise and accentuate the true signal.
Aided by recent improvements in signal processing, geophysicists can use these monitoring arrays
to map microseismic events more completely over
Oilfield Review
extended MAY
stimulations
than is possible from an
16
array placed
in
a
single
well.
Microseismicmonitoring
Fig 2
16 MCSMCwells
2 are available,
When ORMAY
nearby observation
downhole monitoring offers proximity to treatment well stages and ensures higher S/N than that
offered by surface monitoring. Broad bandwidth
signals recorded by downhole arrays often retain
more high-frequency content than do surface
arrays. This high-frequency content is useful for
MS event characterization. Recording P-wave and
S-wave arrivals downhole using 3C sensors also
improves localization accuracy compared with
that from surface recordings of P-waves alone.
Monitoring from multiple wells provides

observations of the source position from multiple
directions and enables more-complete source
characterization compared with that from single-well monitoring. Downhole monitoring
requires correct velocity models to reduce event
localization uncertainty, and it requires precise
well deviation surveys to determine exact
receiver positions. The models must also contain
accurate values for Qp and Qs, the quality factors
related to P-wave and S-wave attenuation during
propagation. These factors are used to determine wave amplitudes at the MS hypocenters
and reduce uncertainty in the inversion for the
source mechanism.23
Modern signal processors use nonlinear
mathematical methods for the detection and
localization of MS events. In combination with
CMM processing, these mathematical methods
have the potential to detect and locate weak MS
events automatically without prior knowledge of
the source mechanism and its radiation pattern.24
Analysts have also extended event localization
algorithms to use the full MS waveforms. Early
event localization methods used traveltimes and
polarizations of the P-wave and S-wave direct
arrivals only. However, geoscientists can use
waveform synthetics to model the source time
functions, the principal features of the waveform
time series recorded by each instrument in the
Oilfield Review
Figure 3. Sensor network deployment options. Sensors for MS monitoring

of hydraulic fracturing may be deployed in vertical (1), horizontal (2) or
deviated monitoring boreholes. Survey engineers may use a grid of shallow
wells (3) containing arrays of multicomponent sensors. On the surface, they
may deploy single component or multicomponent geophones in 2D patches
or in extensive linear arrays (4). Sensor networks that record MS waveform
data over a broad area provide data that can be used to characterize MS
sensor arrays. Analysts then extract arrival times

of direct, refracted and reflected P-wave and
S-wave arrivals. An extension of the CMM method
uses these additional arrivals to identify their
energy for event detection and characterization.25
Tracking Microseismicity to the Surface
In 2011, Schlumberger and an independent
operator acquired a comprehensive MS dataset
while monitoring hydraulic fracturing operations
in the Fayetteville Shale in Arkansas. Completion
engineers stimulated two horizontal wells using
a zipper fracture method, in which hydraulic
fracturing is conducted sequentially in side-byside wells.26 Concurrently, MS survey engineers
conducted a test to assess and quantify the event
detection capabilities, accuracy and resolution
of surface, near-surface and downhole acquisition systems. Sixteen stimulation stages were
May 2016
event compression (red ellipsoids) and dilation (blue ellipsoids) radiation

patterns (5) and estimate source mechanisms. Schlumberger engineers
use the MS Recon high-fidelity microseismic surface acquisition system
to acquire MS data at the surface. The system incorporates proprietary
geophone accelerometers (6) (inset ), ultralow-noise electronics and
a nodal-based wireless acquisition technology (7). (Adapted from
18. Effective stimulated volume (ESV), also referred to as

stimulated reservoir volume, is an estimate of the total
rock volume affected by the hydraulic fracture
stimulation.
19. For more on survey design: Le Calvez J, Underhill B,
Raymer D and Guerra K: Designing Microseismic
Surface, Grid, Shallow and Downhole Surveys,
Expanded Abstracts, 85th SEG Annual International
Meeting and Exposition, New Orleans (October 1823,
2015): 26452649.
20. For more on the use of crosswell surveys: Le Calvez J,
Marion B, Hogarth L, Kolb C, Hanson-Hedgecock S,
Puckett M and Bryans B: Integration of Multi-Scale,
Multi-Domain Datasets to Enhance Microseismic Data
Processing and Evaluation, paper BG08, presented at
the 3rd EAGEOilfield
Workshop
on Borehole Geophysics,
Review
Athens, April 1922, 2015.
MAY 16
21. For more on the use of perforation shots to build velocity
Microseismic Fig 3
models: Probert T, Raymer D and Bradford I: Comparing
ORMAY
16 MCSMC
3 Data
Near-Surface
and Deep-Well
Microseismic
and Methods for Hydraulic Fracture Monitoring,
paper PS07, presented at the 4th EAGE Passive Seismic
Workshop, Amsterdam, March 1720, 2013.
22. For more on surface microseismic monitoring:
DuncanPM and Eisner L: Reservoir Characterization
Using Surface Microseismic Monitoring, Geophysics75,
no. 5 (SeptemberOctober 2010): 75A13975A146.
23. For more on uncertainty in microseismic monitoring:

Eisner L, Thornton M and Griffin J: Challenges for
Microseismic Monitoring, Expanded Abstracts,
81stSEG Annual International Meeting and Exhibition,
San Antonio, Texas, USA (September 1823, 2011):
15191523.
24. For more on nonlinear processing methods: zbek A,
Probert T, Raymer D and Drew J: Nonlinear Processing
Methods for Detection and Location of Microseismic
Events, paper Tu 06 06, presented at the 75th EAGE
Conference and Exhibition, London, June 1013, 2013.
25. For more on full MS waveform processing: Williams MJ,
Le Calvez JH and Gendrin A: Using Surface and
Downhole Data to Drive Developments in Event Detection
Algorithms, Extended Abstracts, 76th EAGE Conference
and Exhibition, Amsterdam, June 1619, 2014.
26. Zipper fracture, or simul-frac, is a technique in which
two or more parallel wells are drilled, perforated and
stimulated via an alternating sequence of stages. This
stimulation method results in a high-density network of
fractures between the wells that increases production
in both wells.
For more on advances in fracturing technology:
RafieeM, Soliman MY and Pirayesh E: Hydraulic
Fracturing Design and Optimization: A Modification to
Zipper Frac, paper SPE 159786, presented at the
SPEEastern Regional Meeting, Lexington, Kentucky,
USA, October 35, 2012.
21
Line 3
1,000 ft
Line 4
Line 7
Line 2
Downhole array
Well M
Well 1H
Well 3
Line 5
Well 4
Well 5
Line 1
Line 6
Well 2
Well 1
Shallow well
2D patch
Well 2H
Depth, ft
Treatment well
Well M
1,000 ft
Well 1H
Well 2H
Figure 4. Fayetteville Shale MSM operation. The map view (top) shows the sensor network layout
for a Fayetteville Shale stimulation. The 4,100-channel surface seismic array consisted of five radial
lines (red, Lines 1 through 5) offset and emanating from the treatment wellhead, two crosslines (red,
Lines6 and 7) and three areal 2D patches (green squares). Data were also acquired using sensors
deployed in a deep horizontal borehole (yellow), in one deep monitoring well (green) and in five vertical
shallow wells (blue circles).The vertical section view (bottom) shows well trajectories used for the MS
monitoring. The trajectories of treatment Wells 1H and 2H are shown in gray and yellow, respectively.
WellM (green) is shown along with the vertical monitoring wells (blue). (Adapted from Schilke et al,
reference28.)
monitored across a reservoir interval at 3,600ft

[1,100m] TVD.
Survey engineers deployed a wide-aperture
borehole seismic array that extended from the
reservoir to the surface (Figure 4). This array
acquired an MS dataset at the reservoir level as
well as data that revealed how signals propagated
and how noise levels varied between the reservoir and the surface. Engineers acquired additional MS data from a deep horizontal well and
from five shallow vertical wells, each containing a
seismic array. They also recorded MS data using
an extensive surface seismic array that consisted
of five radial lines fanning and emanating from
the treatment wellhead, two parallel lines that
crossed the radial lines and three 2D surface
patches that were located about 3,000, 5,000 and
22
8,000ft [915, 1,520 and 2,440m] from the treatment wellhead. Time synchronization between
all recording systems ensured that the same MS
events could be identified on all monitoring systems (Figure 5).27
Data from this comprehensive test allowed
analysts to compare the effectiveness of nearReview
surface and Oilfield
downhole
hydraulic fracture moniMAY 16
toring. Analysts observed that surface array line
Microseismic Fig 4
segments canORMAY
mitigate16
surface-wave
MCSMC 4 noise from a
known source, such as treatment wellhead
pumps, but are less effective against distributed
or moving sources of noise, which may be dominant in areas covered by the array.28 Surface
patches2D arrays of closely spaced sensors
can effectively remove noise coming from multi-
ple directions but cannot cover the same

distances as can linear arrays.29 Sensor arrays in
shallow wells are less sensitive to noise propagating along the surface, but signal processing that
discriminates against noise is hampered by the
small number of sensors available in these
arrays.30 Surface and near-surface array designs
may be adapted to known noise conditions but
are constrained by land access, environmental
effects and cost concerns.
The results of the Fayetteville survey showed
that a downhole array could detect MS events
from nearby stage treatments better than from
other array geometries. But the downhole array
suffered reduced sensitivity and increased location uncertainty for distant stage treatments and
events. Surface and near-surface monitoring,
Oilfield Review
MSM surface lines
MSM wells
Vertical
array
Horizontal array
Vertical Seismic Array
MSM Wells
Horizontal Seismic Array
MSM Stacked Data
Figure 5. Data record from one microseismic event. A microseismic event was detected across the
downhole, near-surface and surface sensor arrays during a test in the Fayetteville Shale. The modeled
waveforms (top, purple) from the event are shown along with sensor positions (green). Waveforms
propagate from the events hypocenter, the location of which was estimated from the data. Recorded
waveforms from the event are shown from the vertical seismic array (middle left), the array in the
horizontal monitoring well (bottom left ) and the 3C arrays in the five shallow vertical MSM wells
(middle right). Five stacked traces from the vertical component waveforms recorded on the five
surface radial lines are also shown (lower right ). (Adapted from Peyret et al, reference27.)
although less sensitive to recording deep signals

than downhole monitoring, offers more uniform
sensitivity over a wider area. Surface patches are
easily deployed over wide areas and may ultimately become the preferred surface monitoring
configuration. However, their successful use will
require the recording of sufficient signal and the
effective application of noise attenuation methods. The lessons learned in the test provide the
planners of future MSM systems with a clearer
understanding of the trade-offs involved when
specifying acquisition equipment layouts.
May 2016
27. For more on the Fayetteville Shale MS test: Maxwell SC,

Raymer D, Williams M and Primiero P: Tracking
Microseismic
SignalsReview
from the Reservoir to Surface,
Oilfield
The LeadingMAY
Edge 31,
16no. 11 (November 2012): 13001308.
Peyret O, Drew
J, Mack M, Brook
Microseismic
Fig 5K, Maxwell S and
Cipolla C: Subsurface to Surface Microseismic
ORMAY
16
MCSMC
5
Monitoring for Hydraulic Fracturing, paper SPE 159670,
presented at the SPE Annual Technical Conference and
Exhibition, San Antonio, Texas, October 810, 2012.
28. For more on surface array performance: Schilke S,
Probert T, Bradford I, zbek A and Robertsson JOA:
Use of Surface Seismic Patches for Hydraulic Fracture
Monitoring, paper We E103 04, presented at the
76thEAGE Conference and Exhibition, Amsterdam,
June 1619, 2014.
29. For more on signal processing approaches applied to

surface patch data: Petrochilos N and Drew J: Noise
Reduction on Microseismic Data Acquired Using a
Patch Monitoring Configuration: A Fayetteville Formation
Example, Expanded Abstracts, SEG 84th Annual
International Meeting and Exposition, Denver
(October 2631, 2014): 23142318.
30. The amplitude of seismic surface waves decreases as
its position away from the surface increases. For more
on noise sources in the Fayetteville Shale test: Drew J,
Primiero P, Brook K, Raymer D, Probert T, Kim A and
Leslie D: Microseismic Monitoring Field Test Using
Surface, Shallow Grid and Downhole Arrays, paper
SEG 2012 0910, presented at the 82nd SEG Annual
International Meeting and Exposition, Las Vegas,
Nevada, USA, November 49, 2012.
23
Mw 2/3 (log Mo ) 6.06.
Figure 6. Seismic moment, energy and magnitude

equations. The seismic moment, Mo, of an
earthquake source is defined as the product
of the shear modulus () of the host rock that
the fault cuts through, the average shear
displacement (D) along that surface and the
affected fault surface area (A). The amplitudes of
emitted seismic waves are directly proportional
to the seismic moment. Seismologists have
also related seismic moment to the seismic
energy (Es), which is radiated when a fault slips,
resulting in a change in the static shear stress
(s) along the fault. The moment magnitude
(M w) is computed from Mo and is a logarithmic
measure of the energy released during a seismic
event. In this equation, Mo is expressed in units
of N.m. When expressed in units of dyne.cm,
10.73 is used instead of 6.06; for units of lbf.ft, the
constant is 5.97.
Making Sense of the Microseismic Cloud

Geophysicists studying earthquakes use characteristics such as the seismic moment, moment
magnitude, stress drop, stress change and source
dimensions to describe the physical processes
occurring at earthquake hypocenters.31 In earthquake seismology, a typical earthquake is caused
by shear displacementsurface parallel slip
along a preexisting fault plane. Earthquake
intensity is related to the seismic moment, MO,
which can be determined by measuring the
amplitudes of seismic waves generated during
Oilfield
the event (Figure
6).32Review
MAY
16seismologist Hiroo Kanamori
In 1977, Japanese
Microseismic
Fig 6seismic moment
used the relationship
between
ORMAY 16 MCSMC 6
and energy to introduce the moment magnitude
(Mw) scale. Today, geophysicists in the oil and gas
industry are applying earthquake seismology
concepts to analyze MS data; moment magnitude
is routinely used to characterize the size of MS
events. Individual source dimensions such as
incremental fracture surface areas and lengths
can be estimated for MS events from their
waveform spectra and source models.
The distribution of MS source locations provides an indication of the rock volume affected by
the hydraulic stimulation. Reservoir engineers
initially related, with some success, well productivity to MS activity using the ESV as a measure of
the volumetric extent of reservoir stimulation
(Figure 7).33 Pinpointing the location of MS
events, however, is sometimes insufficient for
accurately predicting reservoir performance.
This insufficiency may result from inaccurate
24
the active volume of microseismicity may be an

overestimation of the hydraulically connected volume. To reduce this uncertainty, some analysts
consider how many events occur in the neighborhood of each event location when computing ESV.
For MSM that has limited array coverage, distant
low-amplitude events may not be detected. This
phenomenon, referred to as monitoring bias,
may also reduce the computed ESV.
18
16
14
Production, %
Es = 0.5 (s / ) Mo.
estimates of the event density and the actual

extent of the ESV.
When determining ESV, geoscientists must
take into account the spatial density of interconnected fractures within the stimulated volume
and their surface area in contact with the
reservoir. The set of MS event locationsthe
microseismic cloudmay include stress-induced
events in nonhydraulically connected areas. Thus,
12
R 2 = 0.80
10
8
6
4
2
0
400
500
600
700
800
900
1,000
ESV, 1,000 ft 3
879,000
878,000
Production
data
Y-axis, ft
Mo D A.
877,000
Monitoring well
876,000
1,538,000
1,539,000
1,540,000
1,541,000
X-axis, ft
Figure 7. Eagle Ford Shale effective stimulation. The map view (bottom)
shows MS events (colored dots), effective stimulated volume (ESVs, colored
opaque envelopes) and production log results (dashed red lines) for the
stimulation of a horizontal well (dashed blue line) in the Eagle Ford Shale in
Texas. The ESVs were calculated based on the density and magnitude of MS
events for each perforated interval. The length of the bisecting red lines from
the production data are related to the contribution of individual perforation
clusters to the total flow of hydrocarbons. Engineers observed a definite
correlation (top) between production contribution from individual perforated
intervals (red circles) and the ESV derived from the hydraulic fracture model
analysis. For the plotted data, R2 is a linear regression measurement related
to the quality of curve fitting. A value of 0.80 indicates a good fit. (Adapted
from Inamdar etal, reference33.)
Oilfield Review
MAY 16
Microseismic Fig 7
ORMAY 16 MCSMC 7
Oilfield Review
0.6
0.4
0.2
0
25
Treatment pressure, psi
0.8
30
1.1
45
7,800
40
35
1.2
50
40
7,600
Pump rate
35
7,400
30
7,200
Proppant concentration
7,000
25
Cumulative moment
1.0
0.9
0.8
0.7
0.6
0.5
20
6,800
15
6,600
10
6,400
10
6,200
0.1
6,000
20
Event rate
1 am
2 am
3 am
4 am
15
Cumulative moment, N.m 10 6
1.4
1.3
55
8,000
45
Pump rate, bbl/min
Proppant concentration measured, lbm/galUS
8,200
50
1.6
1.0
Pressure
55
1.8
1.2
60
8,400
Event rate, accepted events per 2 min
60
2.0
0.4
0.3
0.2
5 am
Time of day
Figure 8. Cumulative seismic moment. Engineers plotted cumulative seismic moment (black) along with the MS event rate (gray
vertical bars) and pumping parameters to help them understand fracture stimulation job performance during the treatment of a
well in the Eagle Ford Shale. The pump rate (blue), surface pressure (red) and proppant concentration (green) are shown. Analysts
use these plots to identify the time-dependent response of MS events to the stimulation. An abrupt increase in cumulative seismic
moment indicated that deformation increased significantly about halfway through the planned pumping schedule. By comparing
multiple treatments, engineers can determine how microseismicity changes in response to adjustments to the pumping schedule
and whether it is consistent across stages. (Adapted from Downie et al, reference34.)
Microseismic events are produced when rapid

deformation occurs within the reservoir or surrounding formations in response to stress
changes arising from increased pressure during
fracture stimulation operations. The deformation
consists of slip of unknown length along failure
planes of unknown area and orientation.34
Analysts estimate the seismic moment of individual events using the amplitudes and frequency
content of received seismic waveforms.35
Geophysicists can use the seismic moment to
enhance the interpretation of MS data.36 By summing these moment values over time for all
events within distinct spatial volumes or grid
cells, analysts obtain the cumulative moment as a
function of time and space.
31. For more on source parameters: Shearer PM:
Introduction to Seismology, 2nd ed. Cambridge, England:
Cambridge University Press, 2009.
32. Seismic moments range from 105 N.m [105 lbf.ft] in the
case of the smallest detectable microearthquakes to
1023 N.m [1023 lbf.ft] in the case of great earthquakes.
33. For more on ESV versus reservoir production:
InamdarA, Malpani R, Atwood K, Brook K, Erwemi A,
Ogundare T and Purcell D: Evaluation of Stimulation
Techniques Using Microseismic Mapping in the
Eagle Ford Shale, paper SPE 136873, presented at
May 2016
Engineers may compare time series of cumulative moment and stimulation treatment data to
better understand the stimulation process
(Figure 8). An increase in cumulative moment
over time indicates progressive deformation.
Maps of final values of cumulative moment indicate the spatial distribution of seismic deformation observed during the stimulation. Because
large, detectable events contribute significantly
more moment than numerous, small, undetectable events, cumulative moment provides a measure of stimulation response that is less sensitive
to monitoring bias than is an ESV based on event
locations alone. Using 3D mapping of seismic
moment or cumulative moment provides insight
Oilfield Review
MAY 16
the SPETight Gas Completions Conference, SanAntonio,
Microseismic
Texas, November
23, 2010.Fig 8
ORMAY
16D,MCSMC
34. Downie R,
Xu J, Grant
Malpani R 8
and Viswanathan A:
Utilization of Microseismic Event Source Parameters
for the Calibration of Complex Hydraulic Fracture
Models, paper SPE 163873, presented at the
SPEHydraulic Fracturing Technology Conference,
The Woodlands, Texas, February 46, 2013.
35. Seismic sources are formally treated as displacement
discontinuities to describe the difference in motion of
material on opposing faces of fracture surfaces. This
motion need not be parallel to the surfaces.
into fracture behavior during stimulations; such

insights can be used to calibrate complex hydraulic fracture models.
Borrowing concepts from earthquake seismology, analysts use statistical measures such as
b-values and D-values to further describe groups
of detected MS events.37 The relative frequency of
occurrence of earthquakes over a range of magnitudes is described by b-values. Many more small
magnitude events tend to occur than do large
ones, and the b-value quantifies this tendency.
The statistics of the distances separating earthquake hypocenters is described by D-values.
Populations of events occurring on the same fracture and fault planes tend to present characteristic distributions of spatial separation.
36. Analysts use source models, formation material
properties and measured frequency spectra to estimate
the fracture surface area of individual events. Slip
lengths can then be inferred from seismic moment
estimates.
37. For more on b-values and D-values: Grob M and
van der Baan M: Inferring In-Situ Stress Changes
by Statistical Analysis of Microseismic Event
Characteristics, The Leading Edge 30, no. 11
(November 2011): 12961301.
25
1.0
Magnitude of completeness, Mc
Cumulative frequency
Gutenberg-Richter relationship
log10 N = a b M.
0.1
Events
0.01
2.8
2.6
2.4
2.2
2.0
1.8
Moment magnitude
1.6
1.4
1.2
1.0
Figure 9. Plot of b-values. In seismology, the Gutenberg-Richter empirical

law relates the magnitude (M) of earthquakes to their frequency of
occurrence, N, where N is the number of events of magnitude M or greater,
and a and b are constants. Events observed during a stimulation stage in the
Barnett Shale in Denton County, Texas (red dots), show that the cumulative
frequency is equal to N divided by 336, the total number of detected events.
A statistical method has been used to estimate b, which accounts for the
limited ability to detect low-magnitude events. Because data from small
magnitude events cannot be recorded reliably, events that have magnitudes
smaller than the magnitude of completeness, Mc (dashed line), are not used
in the calculation because the S/N is too low. (Adapted from Williams et al,
reference 61.)
In seismology, the Gutenberg-Richter empirical law relates the magnitude of earthquakes, M,

to their frequency of occurrence, N (Figure 9). In
microseismic studies, geoscientists have substituted moment magnitude, Mw, in this relation
and have explored how hydraulic fracture parameters affect the values of a and b, the slope and
intercept, respectively, of the log10 N versus Mw

relationship. Some evidence suggests a possible
relationship between the value of athe number of events at the intercept at Mw equaling 1
and the pump rates used in hydraulic fracturing,
for example, when the cumulative volume of
pumped fluid influences the microseismic event
2.0
2.5
log10 (C (l))
3.0
log10 (C) D log10 (l)
Oilfield Review
MAY 16
Microseismic Fig 9
ORMAY 16 MCSMC 9
3.5
4.0
4.5
5.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
log10 (l)
Figure 10. Plot of D-values. The D-value is estimated from the slope of the
line of event locations during a stimulation in the Barnett Shale in Denton
County, Texas. The cumulative distribution of interevent distances, C(l), is the
total number of MS events (purple circles) that have hypocenters separated
by a distance of l meters or less. The circles represent measured data
and show a good match with the modeled data (dashed black line), which
corresponds to D=1.86. This D-value indicates a near-planar distribution
of event locations for a range of intermediate distances around the event.
The closer and more distant events trend away from straight-line behavior.
(Adapted from Williams et al, reference 61.)
26
rate.38 Global seismology data often show the

slope b to be about 1 for tectonic earthquakes. In
some geologic settings, MS interpreters use
b-values to distinguish failure along naturally
occurring faults from that along hydraulically
induced fractures.
Determination of b-values may also provide
completion engineers with an indication of stress
changes over the course of multistage stimulations. During hydraulic fracturing treatments,
b-values greater than 1 are typically observed,
whereas b at about 1 has been observed during MS episodes dominated by movement along
faults.39 Scientists have observed relationships
between the b-value and local stress conditions.40
Some studies of MS data have shown variations in
b-values within regional shale formations being
stimulated.41 Other studies have shown that
b-values may be time dependent and vary as stress
changes throughout the stimulation process.42
Seismologists use D-values to convey the spatial statistics of earthquake hypocenter occurrence. Computed from event locations, D-values
may be used to summarize interevent distance
statistics. If the cloud of events maps onto a
point, D is expected to equal 0. A D-value of 1 is
expected if the geometric distribution is linear,
2 if planar and 3 if dispersed.43 The distribution of
MS hypocenters has the potential to reveal the
location of interconnected fracture surfaces, and
analysts have developed a variety of techniques
to extract linear and planar features from the
microseismic cloud.44
In one method, D-values were computed separately around each detected MS event location
(Figure10). Closely spaced events are more likely
to show some random scatter due to processing
effects (in which D equals approximately3).
Events occurring on the same fracture and fault
plane will tend to align (in which D equals approximately2). Interpreters identified linear and planar structures in the data by selecting only events
for which the D-value is less than or equal to 2.
Analysts have used changes in both b-values and
D-values to infer stress changes in the reservoir.
Applying additional concepts from earthquake seismology, geophysicists measure P-wave
and S-wave amplitudes across broad receiver networks to determine radiation patterns and then
invert them to estimate seismic moment tensors,
which describe the orientation, magnitude and
slip of individual MS events.45 For MSM, geophysicists use moment tensor inversion (MTI), which
is an advanced seismic processing technique, to
provide information about the mechanism of failure at fracture sites.46 They then decompose each
moment tensor into its constituents to estimate
Oilfield Review
the relative proportion of each failure mode

such as shear slip, tensile opening, expansion or
other process or a combination of themand the
orientation of local fracture planes and the direction of shear slip. Recently, mathematicians have
developed theoretical extensions of MTI in terms
of potency tensors; such extensions display
unique fracture planes and displacement vectors
(Figure 11).47
Although the deformation represented by MS
signals constitutes a small fraction of the total
deformation and fracture volume created during stimulation, MTI processing holds promise to
provide insights into natural fracture characteristics and local stress fields. Geophysicists extract
planar features for input to construct discrete
fracture networks (DFNs), which represent the
distribution, orientation, shape, connectedness
and fluid flow properties of a population of fractures. The MTI results may provide constraints to
help build these networks and calibrate hydraulic
fracture models.48
Integrating MS Data and Geomechanical
Modeling
Engineers integrate MS data with geologic models, mechanical earth models (MEMs), formation
imaging logs and production logs to characterize
reservoirs and aid their understanding of microseismicity. In the past, engineers predicted
future reservoir production based on correlations
between poststimulation production and ESV.
Analysts now use geomechanical modeling to
enhance the forecasting.
Geomechanics can aid in the design of
hydraulic stimulations to maximize the hydraulic
fracture surface area exposed to the reservoir
and to the system of natural fractures within. For
planning wells and determining in situ stress
states, engineers perform stress simulations
using static and time-lapsed 3D MEMs that are
integrated with results from reservoir simulation
models.49 Geomechanical modeling can provide
insight into the extent of fracture-to-fracture
interference between fracture stages in a treatment well, nearby treated wells and the natural
fracture system.
Knowledge of the regional stress state and the
characteristics and distribution of natural fractures in the reservoir is important for predicting
the effectiveness of reservoir stimulations.
During stimulation operations, hydraulic fractures interact with preexisting natural fractures.
Slip along natural fractures generally increases
the permeability of the stimulated fractures. The
density and orientation of the natural fracture
population are significant factors that influence
May 2016
Perforation
cluster
Well
Expansion
Opening
Slip
Figure 11. Source mechanism expansion, opening and slip. The estimated moment tensor for
each event detected during a well stimulation stage (top) has been decomposed into expansion,
opening and slip components and displayed as glyphs. For reference, the well (magenta) is shown
with perforation clusters (red disks). A glyph is composed of two disks and a wireframe sphere
superimposed over them (bottom). The wireframe sphere represents expansion if red or contraction
if blue. The thickness of the disks represents opening, and their relative displacement represents the
degree of slip. The glyphs central plane, which is parallel to the disks planar surfaces, is oriented
with respect to the strike and dip of the fracture plane activated or caused by the event. (Adapted from
Leaney et al, reference47.)
38. Shapiro SA, Dinske C, Langenbruch C and Wenzel F:

45. By analyzing the amplitudes of waveforms received
Seismogenic Index and Magnitude Probability of
at an array of recording sensors, geophysicists can
Earthquakes Induced During Reservoir Fluid
determine the location, shape, size and orientation of
Stimulations, The Leading Edge 29, no. 3 (March 2010):
the motions of the causative event. Geophysicists then
304309.
use the amplitude data to invert for the moment tensor,
a system of point-force couples, which is the best-fit
39. Cipolla C, Maxwell S and Mack M: Engineering Guide
seismic radiation pattern equivalent to that observed
to the Application of Microseismic Interpretations,
from seismic event displacement discontinuities.
paper SPE 152165, presented at the SPE Hydraulic
For more on radiation patterns and moment tensors:
Fracturing Technology Conference, The Woodlands,
Lay and Wallace, reference 4.
Texas, February 68, 2012.
46. Moment tensor inversion (MTI) is now integrated with
40. Schorlemmer D, Wiemer S and Wyss M: Variations in
the Petrel software platform and Mangrove reservoirEarthquake-Size Distribution Across Different Stress
centric stimulation design software workflows.
Regimes, Nature 437 (September 22, 2005): 539542.
47. For more on the theory, decomposition and display
Downie RC, Kronenberger E and Maxwell
SC:
Using
Oilfield Review of moment tensors: Leaney S, Chapman C and Yu X:
Microseismic Source Parameters to Evaluate the
MAY 16
Anisotropic Moment Tensor Inversion, Decomposition
Influence of Faults on Fracture TreatmentsA
Microseismic
Fig 11and Visualization, Expanded Abstracts, 84th SEG
Geophysical Approach to Interpretation,
paper
Annual
SPE 134772, presented at the SPE AnnualORMAY
Technical16 MCSMC
11 International Meeting and Exposition, Denver
(October 2631, 2014): 22502255.
Conference and Exhibition, Florence, Italy,
September1922, 2010.
48. For more on deriving DFN from MS data: Yu X,
RutledgeJT, Leaney SW and Maxwell S: Discrete41. Boroumand N: Hydraulic Fracture b -Value from
Fracture-Network Generation from Microseismic Data
Microseismic Events in Different Regions, presented
by Use of Moment-Tensor- and Event-Locationat the GeoConvention 2014, Calgary, May 1216, 2014.
Constrained Hough Transforms, paper SPE 168582,
42. Zorn EV, Hammack R and Harbert W: Time Dependent
presented at the SPE Hydraulic Fracturing Technology
b and D-values, Scalar Hydraulic Diffusivity, and
Conference, The Woodlands, Texas, February 46, 2014.
Seismic Energy From Microseismic Analysis in the
49. Schlumberger reservoir engineers couple ECLIPSE 3D
Marcellus Shale: Connection to Pumping Behavior
simulations with the VISAGE finite-element
During Hydraulic Fracturing, paper SPE 168647,
geomechanics simulator to create dynamic, time-lapse
models of stress and production history of single and
Conference, The Woodlands, Texas, February 46, 2014.
multiple wells and fields. For more on integrated
43. Grob and van der Baan, reference 37.
modeling: Alexander T, Baihly J, Boyer C, Clark B,
44. Williams MJ, Khadhraoui B and Bradford I: Quantitative
Waters G, Jochen V, Le Calvez J, Lewis R, Miller CK,
Interpretation of Major Planes from Microseismic Event
Thaeler J and Toelle BE: Shale Gas Revolution,
Locations with Application in Production Prediction,
Oilfield Review 23, no. 3 (Autumn 2011): 4055.
Expanded Abstracts, 80th SEG Annual International
Meeting and Exposition, Denver (October 1722, 2010):
20852089.
27
Data
Product
Seismic
Structure
Natural fractures and faults
Rock properties
Stress variations
Event processing
Interpretation
Visualization
Real-time Treatment
Modification
Modify pump rate
Modify proppant schedule
Continue or abort job
Model
Logs, Cores
and Petrophysics
Natural fractures
Rock properties
Stress profile and anisotropy
Reservoir properties
Completion and
Stimulation
Perforation locations
Stage sequence
Treatment data
Flow
Production logs
Production data
Earth Models
Velocity model
1D or 3D MEM
Geologic model
Reservoir model
DFN
Hydraulic Fracture Models

Fracture model calibration
Reservoir
Simulation Models
Reservoir model calibration
Hydraulic Fracture Model

Fracture geometry and
conductivity distribution
Completion and Stimulation

Modify completion
strategy
Reservoir Simulation
Production profile
Hydrocarbon recovery
Drainage architecture
Figure 12. Workflow for completion and stimulation design and field development applications
of microseismic mapping. Using seismic data, well logs and core data along with treatment and
production data (left), engineers build a series of earth models and discrete fracture network (DFN)
models (middle). They use these models to generate hydraulic fracture predictions for the fracture
geometry and conductivity distribution resulting from stimulation operations. They also use MS event
locations and deformation to calibrate the earth and fracture models. Reservoir engineers are then
able to predict fracture network geometry and conductivity and generate reservoir performance
simulations (right). (Adapted from Cipolla et al, reference 39.)
the stimulated fracture network development the UFM unconventional fracture model can then
be used to predict the fracture geometries that
and control reservoir productivity.50
Hydraulic fracturing creates a tensile frac- result from the stimulation.54
ture that opens slowly, and most of the rock
Modelers can use treatment data such as pump
deformation occurs aseismically at much lower pressure, fluid volumes and proppant loadings as
frequencies than the typical MS signal band.51 inputs to the numerical calculations and then use
Shear deformation occurs in the process zone MS data to constrain the results. These UFM simuaround the fracture tip, in the vicinity of the frac- lations yield predictions of stimulated fracture
ture face as a result of leakoff into preexisting geometry and conductivity distributions. Analysts
natural fractures and at doglegs and other geo- iteratively calibrate the model by adjusting the
metric deflections. In contrast to the aseismic input parameters to the UFM simulation to achieve
nature of tensile dilation, shear deformation a match between fracture geometry predictions
often emits sudden, high-frequency, audible seis- and observed MS event locations and deformation,
Oilfield Review
or seismic moments (Figure 14). Adjustable
mic energy.
MAY 16
Reservoir engineers have developedMicroseismic
a variety parameters
Fig 12 include horizontal stresses and propORMAY 16 MCSMC
12 DFN and fracturing fluid. Analysts use
erties of the
of approaches to characterize poststimulation
fracture networks (Figure 12).52 In one approach, volumes of fluids pumped and stresses to constrain
analysts use information from seismic reflection the fracture model to reduce the set of possible
surveys, well logs and cores to build a DFN, which solutions. They conduct sensitivity analyses to
they combine with a set of earth models to determine how much the answer changes as input
describe the reservoir and surrounding forma- parameters are varied and to ensure that non
tions.53 Fracture information may often be unique multiple solutions give results that are
derived from resistivity or ultrasonic image logs reasonable in terms of predicted production.
(Figure 13). Hydraulic fracture models such as
28
Reconciling the UFM modeling results with

the MS event patterns requires the evaluation of
multiple DFN realizations and may not result in
an exact match but rather a statistically probable
match to the MS pattern. After the UFM simulations are calibrated, the fluid flow characteristics
predicted by the UFM technique can be incorporated in a reservoir simulation. In this process,
the hydraulic fractures are explicitly gridded in
the reservoir model to honor the 3D hydraulic
fracture geometry and proppant distribution.
In addition to the reservoir grid, a fluid model,
a set of relative permeabilities, stress-dependent
hydraulic fracture conductivity profiles, historical production rates and bottomhole pressure are
input into the reservoir simulator. After the reservoir models are calibrated, analysts can use
them to forecast hydrocarbon recovery and perform sensitivity analyses. The analysts can vary
completion parameters, including the number of
stimulation stages, the number of perforation
clusters per stage and reservoir parameters such
as permeability, porosity and saturations to maximize future stimulation operations.
50. For more on fracture network development: Johri M
and Zoback MD: The Evolution of Stimulated Reservoir
Volume During Hydraulic Stimulation of Shale Gas
Formations, paper SPE 168701/URTeC 1575434,
presented at the Unconventional Resources Technology
Conference, Denver, August 1214, 2013.
51. Maxwell SC and Cipolla C: What Does Microseismic
Tell Us About Hydraulic Fracturing?, paper SPE 146932,
presented at the SPE Annual Technical Conference and
Exhibition, Denver, October 30November 2, 2011.
52. For more on an early, computationally efficient
geomechanical modeling approach: Xu W, Le Calvez J
and Thiercelin M: Characterization of HydraulicallyInduced Fracture Network Using Treatment and
Microseismic Data on a Tight-Gas Formation: A
Geomechanical Approach, paper SPE 125237,
presented at the SPE Tight Gas Completions Conference,
San Antonio, Texas, June 1517, 2009.
53. For more on the construction of DFNs: Will R, Archer R
and Dershowitz B: Integration of Seismic Anisotropy
and Reservoir Performance Data for Characterization of
Naturally Fractured Reservoirs Using Discrete Feature
Network Models, paper SPE 84412, presented at the
SPE Annual Technical Conference and Exhibition,
Denver, October 58, 2003.
Offenberger R, Ball N, Kanneganti K and Oussoltsev D:
Integration of Natural and Hydraulic Fracture Network
Modeling with Reservoir Simulation for an Eagle Ford
Well, paper SPE 168683/URTeC 1563066, presented at
the Unconventional Resources Technology Conference,
Denver, August 1214, 2013.
54. UFM processing is embedded in the Mangrove software
platform and uses the output of the VISAGE simulator.
For more on UFM processing: Weng X, Kresse O,
CohenC, Wu R and Gu H: Modeling of HydraulicFracture-Network Propagation in a Naturally Fractured
Formation, SPE Production and Operations 26, no. 4
(November 2011): 368380.
For more on UFM modeling: Cipolla C, Weng X, Mack M,
Ganguly U, Gu H, Kresse O and Cohen C: Integrating
Microseismic Mapping and Complex Fracture Modeling
to Characterize Fracture Complexity, paper SPE 140185,
Conference and Exhibition, The Woodlands, Texas,
January 2426, 2011.
Oilfield Review
Primary fracture set

n = 775
Orientation:
approximately
N54E
0
330
Start
30
300
Observations: Logs,
seismic data and natural fractures
60
DFN Realization
90
270
Fracture treatment data and

mechanical earth model (MEM)
UFM Prediction
120
240
150
210
180
Secondary fracture set

n = 175
Orientation:
approximately
N60W
Composite Fracture Set
Microseismic Events
Comparison
Match?
No
Yes
1,000 ft
Output
fracture
model
Figure 13. Construction of a discrete fracture network. For a fracture

stimulation in the Eagle Ford Shale, analysts processed logging data
obtained from a well drilled with oil-base mud using an imaging tool and
were able to detect fractures and determine their orientations from a
rose plot (top). The red dots indicate dip azimuth and inclination angle of
the poles of the primary fracture planes; blue dots are dip azimuth and
inclination angle of the poles of the secondary fracture planes. From the
inside to the outside edge of the rose plot, the dip varies from 0 to 90.
The black lines represent fracture strike orientation and the length is
related to the abundance
a given direction. Analysts identified a
Oilfieldalong
Review
primary and a secondary
fracture
set. The intensity of fracture occurrence
MAY 16
along the wellbore correlated with formation curvature determined using
Microseismic Fig 13
reflection seismic surveys. Co-krigeda geostatistical technique for data
ORMAY 16 MCSMC 13
interpolationfracture intensity and curvature were used to populate
the 3D volume of interest with fractures and build the DFN (bottom). The
SWNE trending band of fracture intensity corresponds to an area of high
formation curvature. The well trajectory (blue line) is superimposed on the
DFN. (Adapted from Offenberger et al, reference 53.)
May 2016
Figure 14. Flowchart for calibrating UFM processing and DFN simulations.
Analysts build a DFN model (top right, light gray lines) based on geologic,
geophysical, well log and core data. Preexisting discrete fractures
directly affect the hydraulic fracture system. The UFM simulations predict
fracture geometry (middle right, heavy blue lines) based on treatment
parameters and an earth model that includes the estimated stress field.
The stress field can be calculated with a 3D geomechanical simulator
using wellbore measurements as calibration points. Analysts compared
the fracture geometry predicted from UFM processing with maps of the
observed MS event pattern (bottom left, red dots), taking into consideration
the deformation represented by the seismic moment obtained from MS
monitoring. The engineers then executed an iterative calibration loop,
adjusting UFM processing inputs for multiple DFN realizations (bottom
right ) to arrive at the best overall agreement between the modeled fracture
geometries and the observed deformations. (Adapted from Cipolla et al,
reference54.)
Oilfield Review
MAY 16
Microseismic Fig 15
ORMAY 16 MCSMC 15
29
Monitor Well C
Stimulation stage
Geophones
1
9 10 11 12 13 14 15 16
Well plug
Treatment Wells
A and B
Austin Chalk
Eagle Ford
Shale
Buda
Limestone
Figure 15. Vertical section view of MS events recorded during the

stimulation of two horizontal wells in the Eagle Ford Shale. Microseismic
event hypocenters (spheres, color-coded by stage number) and the
trajectories of WellsA (orange), B(blue) and C (yellow) are shown in
conjunction with the formation tops (horizontal tan and light blue) and faults
(vertical gray and green). Treatment WellsA and B were stimulated in a
21-stage zipper fracture operation. Engineers used a 12-level geophone
array positioned in the vertical portion of WellC to monitor 16 of the 21 stages
of the stimulation operation. Microseismic events were largely confined to

the Eagle Ford Formation, bounded above by the Austin Chalk and below by
the Buda Limestone. During stimulation Stage8 (dark blue) of WellA, the
MS events aligned with the adjacent major fault and propagated downward
into the Buda Limestone. This observation suggested the fault affected
fracture growth during the stimulation. A decision was made to abort
Stages9 (teal) and 10 (crimson) and proceed with Stage11 (orange) on the
other side of the fault. (Adapted from Zakhour et al, reference58.)
6,000
6,400
6,800
Depth, ft
7,200
Major fault
7,600
8,000
Oilfield Review
MAY 16
Microseismic Fig 16
ORMAY 16 MCSMC 16
8,400
0
8,80
1H
3H
40
0
ft
2H
Figure 16. Trajectories of three wells to be stimulated in the Barnett Shale in relation to a fault system
mapped from 3D surface seismic data. Proximity of faults can influence the local stress field, affecting
induced fracture propagation and associated microseismicity. The well stimulation plans for these
operations included a buffer zone containing no perforations in Wells1H, 2H and 3H near a major fault
(aqua surface). The colored disks represent the perforation intervals for the stimulation stages; no
stimulation stages were attempted south of the fault in Wells1H and 3H. (Adapted from Le Calvez et al,
reference 59.)
30
Engineered Completions
The Eagle Ford Formation, an upper Cretaceous
marl in South Texas, is a target for oil and gas
development. Because the formation is highly
laminated and has ultralow permeability, effective completion designs are required to maximize
production. In September 2013, an operator
tested MSM as a method for guiding hydraulic
fracture operations and evaluating fault interactions during stimulation.
The operator drilled two horizontal lateral
wells in the gas-condensate window of the Eagle
Ford Shale trend in Karnes County, Texas. These
wells, which were drilled parallel to each other
and approximately 330ft [100m] apart, crossed
a major fault. After the operator drilled the wells,
measurements of petrophysical and geomechanical properties were acquired using the ThruBit
through-the-bit logging services.55 Using the
Mangrove engineered stimulation design module
in the Petrel platform, engineers developed a
completion design for one of the two laterals.56
Oilfield Review
May 2016
Well 1H
Stage 1
Stage 2
Stage 3
Stage 4
1H
2H
3H
Well 2H
Stage 1
Stage 2
Stage 3
Stage 4
Y-axis, ft
Well 3H
Stage 1
Stage 2
Stage 3
Stage 4
Stage 5
400 ft
Stimulation design engineers used reservoir

and completion quality parameters derived from
the Mangrove software to aid in optimizing stage
intervals and the placement of perforations along
the lateral.57 Flow measurements in some previous wells that had geometricevenly spaced
completions had shown unequal contribution to
production across perforations. The engineered
stages grouped perforation clusters in regions of
the lateral that had similar horizontal stress.
Completion engineers anticipated that all perforation clusters in a stage would break down and
initiate fractures simultaneously because each
cluster had similar stress characteristics.
Measurements made during stimulation confirmed that lower than average treating pressures were required for the engineered
completion than had been used for the lateral
that was stimulated using a geometric model.
Survey engineers acquired MS data during
the stimulation using a 12-receiver downhole
array in a nearby vertical well. Using these MS
data, the operator monitored hydraulic fracture
development near a fault system, identified fault
interaction and adjusted the completion design
to avoid the fault. Engineers later studied the
mechanisms of fault interaction through postjob
integration of MS data with treatment data.
Completion designers normally try to avoid
the interaction of hydraulic fractures with large
faults. Avoiding faults can prevent the loss of
treatment fluid and proppant to thief zones along
the fault. In this project, the completion design
excluded stimulation stages that were within
250ft [76m] of the identified major fault.
Monitoring revealed that MS events were
generally well bounded within the target Eagle
Ford Formation and the overlying Austin Chalk
(Figure 15). However, for some stages, MS activity and treating pressure records indicated unexpected fracture bridging and potential proppant
screenout. Analysis of MS event clusters and
b-values alerted the engineers that the hydraulic fractures had encountered the nearby fault,
which blocked and limited fracture development and led to premature stage terminations.
Real-time interpretation allowed modification of
the completion strategy, and several stimulation
stages planned near the fault were abandoned.
Recommendations were made for future completion designs to increase buffer zones from 250ft
to about 400ft [120m] on either side of major
faults to minimize the risk of pumping nonproductive stages.58
400 ft
Figure 17. Microseismic events

detected during stimulation of
three horizontal wells in the
Barnett Shale. Fault traces
(cyan) are mapped at the
depth of laterals 1H (red), 2H
(green) and 3H (yellow). The
zipper fracturing performed on
Wells1H and 3H was monitored
from Well2H. Engineers also
monitored four stimulation
stages performed on Well2H
using sensors in Well3H.
Color coding and symbols
are used to represent stages
and wells. Microseismicity
for stages closer to the fault
system tended to be compact,
an observation that was
explained later by fracture
modeling as stimulation-fault
interaction. Longer hydraulic
fracture wings occurred in the
stages that were executed
closer to the toe of the well than
those executed close to the
heel. Microseismicity overlap
observed between successive
stages indicates insufficient
fracture isolation. (Adapted from
X-axis, ft
Maximizing Recovery
In 2011, engineers and geoscientists with Teleo
Operating, LLC and Eagleridge Energy, LLC conducted a multistage, multilateral stimulation in
the Barnett Shale in Denton County, Texas. The
operators drilled three parallel horizontal wells
into the lower Barnett Shale. Well trajectories
were about 500 ft [150 m] apart; the central well
was landed about 80 ft [25 m] shallower than the
outside laterals. Because of lease boundary constraints, the wells had to be placed in the vicinity
of several large faults. The lateral sections of the
Oilfield
Review
wells were drilled away
from
the major fault and
MAY(Figure
16 16). Fault throws
through a smaller fault
varied from 20 to 100 ft [7 to 30 m]. Engineers

used a zipper fracture stimulation on the central
Well1H and eastern-most lateral 3H. The western-most lateral 2H was stimulated later.
Microseismic survey engineers monitored
stimulations on Wells 1H and 3H using a
3C-accelerometer array that was placed in horizontal Well 2H using a tractor and repositioned
according to the well and stage to be monitored.
Operations on Well 2H were later monitored
using an array deployed in a vertical section
below the 3H wellhead. During monitoring, engineers observed MS activity across all pumped
stages (Figure 17).
16 MCSMC
18 T,
55. For more on ThruBitORMAY
services: Aivalis
J, Meszaros
Porter R, Reischman R, Ridley R, Wells P, Crouch BW,
Reid TL and Simpson GA: Logging Through the Bit,
Oilfield Review 24, no. 2 (Summer 2012): 4453.
56. For more on the Mangrove service: Ajayi B, Aso II,
Terry IJ Jr, Walker K, Wutherich K, Caplan J,
GerdomDW, Clark BD, Ganguly U, Li X, Xu Y, Yang H,
LiuH, Luo Y and Waters G: Stimulation Design for
Unconventional Resources, Oilfield Review 25, no. 2
(Summer 2013): 3446.
57. For more on determining reservoir and completion
quality: Slocombe R, Acock A, Fisher K, Viswanathan A,
Chadwick C, Reischman R and Wigger E: Eagle Ford

Completion Optimization Using Horizontal Log Data,
paper SPE 166242, presented at the SPE Annual
Technical Conference and Exhibition, New Orleans,
September 30October 2, 2013.
58. For more on horizontal completion optimization across
a major fault: Zakhour N, Sunwall M, Benavidez R,
HogarthL and Xu J: Real-Time Use of Microseismic
Monitoring for Horizontal Completion Optimization
Across a Major Fault in the Eagle Ford Formation,
paper SPE 173353, presented at the SPE Hydraulic
Fracturing Technology Conference, The Woodlands,
Texas, February 35, 2015.
Microseismic Fig 18
31
Fault
100
90
Effective stress, psi
80
70
60
50
40
3H
1H
30
2H
20
10
0
100
Fault
90
Effective stress, psi
80
70
60
50
40
30
20
3H
10
1H
2H
Figure 18. Geomechanical modeling using finite element analysis. The stimulation of the Barnett
Shale Well3H, Stage5, was near a major fault. The fault plane projections (olive green) are shown
intersecting Wells1H (blue), 2H(green) and 3H(yellow). Simulations were run for materials near the
fault that had equivalent high stiffness (top) and low stiffness (bottom). The fault properties affect the
extent of the region perturbed by fracturing in the simulations. Analysts consider the region perturbed
by fracturing to be similar to the region where microseismicity occurs. In this case, failure and
microseismicity are expected to be limited in the region near the fault. The colored volumes show only
the simulation elements where the stresses are perturbed toward failure by the stimulation. The colors
correspond to the minimum in situ effective stress; purple is low compression, showing the region
where tensile failure is most likely to occur, and red is high compression. (Adapted from Williams et al,
reference 61.)
During zipper fracturing of Wells 1H and 3H, of the target depth interval. Observations
interpretation of MS event location trends indi- revealed downward fracture growth and alerted
cated the hydraulic fracture azimuth, N5055E, the operator to stop pumping to avoid fracturing
was consistent with the expected maximum hori- into the water-bearing Viola Limestone below the
zontal stress direction; however, the extent of the zone of interest. During another stage, engineers
microseismicity varied from stage to stage. The recognized that the planar alignment of MS
MS locations for the early stages, near the toe of events indicated slip along a fault and were able
Review
to stop pumping for that stage and bypass a
the wells, extended farther from the Oilfield
wellbore
MAY
16
faulted zone before resuming stimulation.
than those observed during later stages,
toward
Microseismic Fig 19
the heel of the wells. These later stages were Microseismic monitoring allowed the operator to
ORMAY 16 MCSMC 19
closer to the main fault and displayed a more modify the stimulation program during the ongocompact microseismicity pattern and shorter ing job operation.59
fracture wings away from the wellbore.
Postsurvey modeling and data integration
The MS locations observed in these later provided a more complete explanation of the
stages overlapped with those observed in some stimulation and MS responses.60 Analysts conearly stages. Several later stages near the heels of structed a complete history of the treatment
the wells displayed microseismicity that was out
32
using a range of analytical techniques and incorporated summary statistics of MS event attributes within the workflow. Event attributes
included seismic moment and moment magnitude. From these parameters, analysts determined summary b-value statistics and inferred
relative stress magnitudes for each stimulation
stage.61 They used D-value estimates to extract
fracture planes from the clouds of microseismic
eventsD-values near two indicated planar
alignmentsand then used the planes for DFN
construction. Monitoring geometries that can
provide full MTI offer an additional means to constrain DFN construction and modeling but were
not available in this study. Additional input to the
analysis included earth model parameters such
as the rock mechanical properties of layers and
natural fracture geometries along with stimulation data such as well geometry, flow rates, fluid
types and surface pumping pressures.
Engineers combined the Mangrove, UFM and
VISAGE software to model the hydraulic fracturing process, the interaction of induced and natural fractures and the stress field.62 These models
predicted the chronological development of the
interconnected fracture network and were calibrated iteratively using the observed evolution of
MS activity. They tested fracture propagation scenarios by matching time-distance relationships
within the microseismicity pattern and then
iteratively improved the interpretation by updating the location and properties of the natural
fractures. Engineers constrained the simulation
using material balance, which reconciled the
fracture volume opened during stimulation with
the volumes of pumped fluid and proppant and
the volume of fluid estimated to have leaked off
into the formation. The simulation provided a
description of proppant placement together with
a prediction of which natural fractures might
59. Le Calvez J, Xu W, Williams M, Stokes J, Moros H,
Maxwell S and Conners S: Unconventional Approaches
for an Unconventional Faulted ReservoirFrom Target
Selection to Post-Stimulation Analysis, paper P336,
presented at the 73rd EAGE Conference and Exhibition,
Vienna, Austria (May 2326, 2011).
60. Williams MJ, Le Calvez JH and Stokes J: Towards
Self-Consistent Microseismic-Based Interpretation of
Hydraulic Stimulation, paper Th 01 15, presented at
the 75th EAGE Conference and Exhibition, London,
June 1013, 2013.
61. Williams MJ, Le Calvez JH, Conners S, Xu W:
Integrated Microseismic and Geomechanical Study in
the Barnett Shale Formation, Geophysics 81, no. 3
(MayJune 2016): 113.
62. In the Barnett Shale case study, Schlumberger
engineers used Mangrove software with the UFM
complex fracture simulator to model fracture interaction
and the VISAGE simulator to model stress.
63. For more on modeling of faulted rock masses: PandeGN,
Beer G and Williams JR: Numerical Methods in Rock
Mechanics. Chichester, New York, USA: John Wiley
and Sons Ltd., 1990.
Oilfield Review
open and which ones behaved as barriers that

promoted vertical or asymmetric fracture growth.
Modeling for this Barnett Shale stimulation
helped analysts understand the downward growth
of microseismicity into the Viola Limestone. The
complex fracture simulator reproduced observations of hydraulic fracture interactions with
natural fractures, which acted as barriers to propagation in the Barnett Shale and in the underlying
Viola Limestone. By tuning the fault properties
within the finite-element geomechanical model,
analysts were able to match results with the
observed distribution of MS events.
Analysts used finite-element geomechanical
modeling to study how fault properties influenced the zone that a hydraulic fracture perturbs
toward failure (Figure 18). In this type of modeling, the rock mass, including its embedded faults
and fractures, is described as an equivalent
medium that has standard mechanical properties
such as elastic modulus and strength.63 The presence of an active fault may alter the regional
stress field in its vicinity, pushing outward, normal to its plane, and increasing the stress on fracture interfaces. The simulations revealed that
the compact distribution of MS events in stages
pumped close to the fault was consistent with
large values of fault-related stiffness. By simulating all stages and varying fault-related stiffness,
modelers demonstrated that the interaction with
the fault was consistent with microseismicity
that was more compact in the stages toward the
heel than in those toward the toe. The understanding of fault interaction gained from this
case study should benefit reservoir engineers
planning future refracturing operations in these
wells or stimulation treatments in nearby wells.
Challenges and the Future
As operators develop unconventional resources,
determining the optimum well spacing and completion strategies that maximize ultimate recovery is critical. To help operators achieve these
objectives, MSM provides key data for constraining and calibrating models used to help geoscientists with data interpretation and integration.
Microseismicity is induced as the reservoir and
adjoining formations respond to stimulation
treatments. Models used to predict how these formations should respond to stimulations must
simulate and faithfully reproduce the observed
microseismicity. Challenges for geoscientists are
the accurate measurement of MS events and
extracting of maximal information from them.
Advances in acquisition, processing, interpretation and integration of MS data are providing

unique insights into and increased understanding of stimulated reservoir behavior. Models help
engineers interpret and constrain MS data, and
geomechanical earth models help them characterize the variability of reservoir properties.
Fracture network modeling facilitates predictions of the interactions between hydraulic fractures and rock fabric. Reservoir simulations
assist in predicting field drainage patterns, and
productivity may be validated through production history matching. Monitoring microseismicity offers valuable data for validating these
models and simulations.
Methods such as MSM give operators insight
into reservoir dynamics that far exceeds what was
possible even a few years ago. Success in developing unconventional reservoirs owes much to the
pioneers working in plays such as the Barnett
Shale. Recent advances in tools and technologies are allowing operators to develop unconventional reservoirs with greater certainty, reduced
risk and deeper understanding of the nature of
these formations.
HDL
Contributors
Jol Le Calvez is a Schlumberger Geophysics Advisor
and Microseismic Domain Expert in Houston. He works
on development and commercialization of microseismic
and borehole seismic products while heading a team
of geophysicists, geologists and stimulation engineers
working on various plays around the world. He has managed the Microseismic Services Answer Product Center
and the borehole seismic processing and crosswell seismic groups in Houston since 2014. His main responsibilities are the processing and interpretation of data for
geologic, geophysical and geomechanical applications.
He also works with product centers defining and testing
software programs and with research centers on defining and testing of algorithms. Jol joined Schlumberger
in 2001, and after several years in the field acquiring
and processing seismic data, he led the microseismic
processing and interpretation team in Dallas from 2008
until 2011. He then moved to Houston to manage the
North America microseismic processing and interpretation center. He earned a BSc degree in mathematics and
physics and an MSc degree in geology and geophysics,
both from the Universit de Nice Sophia Antipolis,
France; a Diplme dEtudes Approfondies in tectonophysics from the Universit Pierre et Marie Curie, Paris;
and a PhD degree in geology from The University of
Texas at Austin, USA.
Raj Malpani is a Senior Completions and Production
Engineer with Schlumberger Technology Corporation
in Houston. For the past 10 years, he has been a part
of integrated teams that address technical challenges
pertaining to unconventional reservoirs. His interests
May 2016
include hydraulic fracture treatment design and evaluation, production data analysis, reservoir simulation,
geomechanics, microseismic monitoring, restimulation,
multiwell pad development and weak interface modeling. Raj holds a BTech degree in petrochemical engineering from Dr. Babasaheb Ambedkar Technological
University, Lonere, Maharashtra, India, and an
MSdegree in petroleum engineering from Texas A&M
University, College Station.
Jerry Stokes is the President and Owner of MidContinent Geological, Inc. in Fort Worth, Texas. He has
been a certified petroleum geologist with the AAPG for
more than 35 years. Since 1987, he has been involved
in oil and gas exploration, geologic consulting and sales
and marketing of geologic projects throughout Texas
and nearby states. As a geologist for Panhandle Eastern
Pipeline, Jerry was responsible for the early development of underground gas storage fields in Kansas,
Louisiana, Illinois and Michigan, USA. He then worked
for Rust Oil Corporation as the exploration manager
for the Permian basin. He is a member of the Society of
Independent Professional Earth Scientists and the Fort
Worth Wildcatters. Jerry has a BSc degree in geology
and geophysics from Texas Tech University, Lubbock.
Michael Williams is a Principal Reservoir Engineer in
Geophysics at Schlumberger Gould Research (SGR),
Cambridge, England. Since 2008, he has worked in the
area of interpretation of microseismicity, specifically
the accurate recovery of statistical information from
detection-limited microseismic data, and the application
of microseismic interpretation to reservoir simulation
and geomechanical modeling. He joined Schlumberger

GeoQuest in 1997 as a commercialization software
engineer and worked as project leader and team leader
in Abingdon, England. In 2002, he was a team leader in
Sugar Land, Texas, developing the first hydraulic fracture monitoring software to support the interpretation
of microseismic information in the context of fracture
stimulation. He joined SGR as a senior research scientist in 2004, where he worked in applied reservoir engineering, fluid measurements (as program manager) and
well test interpretation. Michael received a BS degree
in physics and an MS degree in geophysics, both from
Imperial College of Science, Technology and Medicine,
University of London. He also has a PhD degree in physics from the University of Wales, Aberystwyth.
Jian Xu is a Senior Microseismic Services Engineer in
Houston. He focuses on microseismic data interpretation, hydraulic fracture monitoring and stimulation
program evaluation in unconventional plays. He joined
Schlumberger in 2008 as a field engineer in Bryan,
Texas. He held various positions, including access
field engineer, production stimulation engineer and
microseismic services engineer working on several
unconventional plays in the US, all while located in
Houston. Before his current assignment, Jian was
a senior production stimulation engineer at the
Production Technology Integration Center in Houston.
He obtained BS and MS degrees in electrical engineering from Tianjin University, China, and a PhD degree
in petroleum engineering from Texas A&M University,
College Station.
33

Locations that host oil and gas operations often provide ideal conditions for corrosion.
Ongoing research and advances in coatings, cathodic protection, nondestructive
testing, corrosion analysis and inhibitors allow operators to safely produce oil and
gas in these corrosive environments.
Nausha Asrar
Bruce MacKay
Sugar Land, Texas, USA
ystein Birketveit
Marko Stipanicev
Bergen, Norway
Joshua E. Jackson
G2MT Laboratories, LLC
Houston, Texas
Corrosion validates the universal law of entropy;

everything trends toward a state of greater chaos
and disorder. The flecks of rust on an iron bar or
the green patina on a copper fixture are evidence of
the insidious effects of corrosion. These examples
may be regarded as an annoyance, but taken to the
extreme, the results of corrosion can lead to catastrophic outcomes.
Corrosion has brought down bridges, downed

aircraft, leveled chemical plants, parted drillpipe and ruptured pipelines. Given sufficient
time, this adversary has the potential to degrade
any material. In certain environments, the
unchecked effects of corrosion can come swiftly,
and the consequences of failure to manage corrosion can be costly.
Alyn Jenkins
Aberdeen, Scotland
Denis Mlot
Total
Paris, France
Jan Scheie
Stavanger, Norway
Jean Vittonato
Total
Pau, France
DS-1617 is mark of M-I LLC.
Hastelloy is a registered trademark of
Haynes International, Inc.
Inconel and Monel are trademarks of
Special Metals Corporation.
34
Oilfield Review
According to the US Federal Highway Administration, the approximate annual direct cost of
corrosion for the US in 2015 was an estimated
US$500 billion, representing around 3.1% of
the nations gross domestic product.1 This figure
amounts to six times the average annual cost of
weather-related disasters for the US, which was
about US$87 billion in 2011.2 Unlike weather
events, corrosion can be controlled or at least
managed; scientists estimate that 25% to 30% of
corrosion costs could be avoided if good corrosion
management practices and preventive strategies
were employed.3
Throughout the ages, and despite an early
lack of understanding concerning the fundamental mechanisms involved, humans have attempted
to control corrosion. In ancient times, corrosion
resistance was sometimes imparted to materials
as a matter of circumstance rather than design
(Figure 1).4 Early corrosion control methods
included the use of bitumen and lead-based
paints by the Romans in the first century. Around
500BCE, Chinese sword makers used copper
sulfide coatings to inhibit corrosion on bronze
swords. Centuries later, the copper sheathing
used on British sailing vessels to reduce biofoulingfouling of underwater surfaces by organisms
such as barnacles and algaeand increase speed
accelerated the corrosion of nails that held the
ships together.5
1. Koch GH, Brongers MPH, Thompson NG, Virmani YP and
Payer JH: Corrosion Costs and Preventive Strategies in
the United States, Washington, DC: US Department of
Transportation Federal Highway Administration,
Publication FHWA-RD-01-156, March 2002.
Jackson JE: Corrosion Will Cost the US Economy over
$1 Trillion in 2015, G2MT Laboratories, http://
www.g2mtlabs.com/corrosion/cost-of-corrosion/
(accessed January 6, 2016).
Papavinasam S: Corrosion Control in the Oil and Gas
Industry. Waltham, Massachusetts, USA: Gulf
Professional Publishing, 2014.
2. The US$ 87 billion cost of weather-related disasters in
2011 was the highest on record. The average annual cost
has been closer to US$ 10 billion in recent years. For
more on the cost of weather-related disasters: Smith AB
and Katz RW: U.S. Billion-Dollar Weather and Climate
Disasters: Data Sources, Trends, Accuracy and Biases,
Natural Hazards 67, no. 2 (June 2013): 387410.
3. Chillingar GV, Mourhatch R and Al-Qahtani GD: The
Fundamentals of Corrosion and Scaling for Petroleum
and Environmental Engineers. Houston: Gulf Publishing
Company, 2008.
4. Kumar AVR and Balasubramaniam R: Corrosion Product
Analysis of Corrosion Resistant Ancient Indian Iron,
Corrosion Science 40, no. 7 (July 1, 1998): 11691178.
Balasubramaniam R: Story of the Delhi Iron Pillar. Delhi,
India: Foundation Books Pvt. Ltd, Cambridge House, 2005.
5. Groysman A: Corrosion for Everybody. Dordrecht, The
Netherlands: Springer Science+Business Media, 2010.
6. Ahmad Z: Principles of Corrosion Engineering and
Control, 1st ed. Burlington, Massachusetts: ButterworthHeinemann, 2006.
7. Kermani MB and Harrop D: The Impact of Corrosion on
the Oil and Gas Industry, SPE Production & Facilities 11,
no. 3 (August 1996).
May 2016
PA
S
KI
TA
C H I N A
NEP
AL
New Delhi
BANGLADESH
I
Mumbai
0
0
km
500
miles
SRI LANKA
500
Figure 1. Delhi pillar. This iron pillar is located in the Qutub Complex in New Delhi, Delhi, India (inset).
It is about 9.1m [30ft] tall and weighs approximately 6,000kg [13,200lbm]. Erected in 400CE, the
pillar is essentially free of the typical rusting that would be expected to take place over 1,600years
of exposure. Reasons for the lack of corrosion include New Delhis low humidity but are primarily
attributed to the high concentration of phosphorus in the iron.
Michael Faraday was one of the most important contributors to the early understanding of
corrosion; in the early 1800s, he established a
quantitative relationship between the chemical action of corrosion and electric current.6
Although much more is known about the subject
today, scientists continue to study the mecha-
nisms of corrosion and search for methods to

manage and control it.
Combating corrosion is a significant source
of expenditures for the oil and gas industry
(Figure2). British Petroleum (BP) conducted a
study of its operations in the North Sea in 1995.7
The company found that outlays for corrosion
US Oil and Gas Corrosion Expenditures, US$ billion/year
Oilfield Review
MAY 16
Refining
Corrosion
Fig 1 Distribution
3.7
ORMAY 16 CRSSN 1 5.0
Production
1.4
Storage
7.0
Tankers
2.7
Pipelines
7.0
Figure 2. Corrosion expenditures. Corrosion expenditures in the US

oil and gas industry are about US$26.8billion/year. The downstream
segment of the industryproduction, pipelines and tankers
accounts for 41% of the total, or US$11billion/year. (Adapted from
Koch et al, reference 1.)
35
Anodic reaction
Fe0 Fe2+ + 2e
Cathodic reaction
H2O + 2e 0.5 O2 + 2OH
Fe2+ + 2OH Fe(OH) 2
OH
OH
Water
OH
Fe 2+
Fe0
Fe(OH) 2
Anode
Fe(OH) 2
Fe 0
Cathode
Electron flow
Fe0
Fe 0
Steel
Figure 3. Corrosion cell. When steel in water rusts, several reactions take
place simultaneously. At the anode, steel [Fe0] goes readily into solution to
form ferrous iron [Fe2+] and ferric iron [Fe3+] (not shown) ions, and electrons
move to the cathode. Electrons at the cathode react with water [H2O] to
form oxygen [O2] and hydroxyl [OH] ions. The OH ions combine with the
solubilized Fe2+ to form iron hydroxide [Fe(OH)2].
prevention and control averaged about 8% of

Some forms of metal corrosion are related
the total capital expenditure for its projects. to stability; for example, galvanic corrosion is
On the UK Continental Shelf, 25% to 30% of BPs an electrochemical process associated with the
operating costs were related to the control and movement of electrons between areas that have
management of corrosion. Costs associated with different electrochemical potentials. The corroreplacing corroded equipment, lost production sion cell schematically describes oxidizing corroand corrosion-related contamination contributed sion, which is analogous to a battery in which two
to overall expenditures. In addition to the direct dissimilar metals are connected by an electrolyte
costs, the company found that corrosion had a (Figure 3).9 A metal that has a higher corrosion
significant indirect cost on health, safety and ratemore unstablerepresents the negative
environmental concerns.
part of the cell and acts as the anode; a second
This article focuses on descriptions of cor- metal that has a lower corrosion ratemore
rosion, management techniques and advances stableacts as the positive part of the cell,
in corrosion abatement technologies. Field thecathode.10
examples from Gabon, deepwater Nigeria
andReviewDuring the galvanic corrosion process, metal
Oilfield
the North Sea illustrate the ongoing battle
waged
MAY 16 oxides are formed as electrons flow from the
Corrosion Fig
3 to the cathode through the electrolyte
against corrosion by oil and gas operators.
anode
ORMAY 16 the
CRSSN
fluid3in contact with the anode and cathode.
The Corrosion Process
A simplified version of iron oxidation can be used
Scientists and engineers today have a better to illustrate the galvanic corrosion processthe
understanding of corrosion processes than did actual process is more complex. The presence of
the ancient Romans and Chinese. Fighting cor- water [H2O] on the surface of the iron [Fe or Fe0]
rosion requires an understanding of the principal releases electrons to form ferrous iron [Fe+2] and
elements that cause and contribute to the corro- ferric iron [Fe+3] ions, which act as the anode in
sion. There are several categories of corrosion; for our battery analogy. The liberated electrons flow
the oil and gas industry, common types include to the cathode, where, in the presence of oxygen
exposure to carbon dioxide [CO2, sweet corro- [O2], ferrous oxide [FeO] and ferric oxide [Fe2O3]
sion], hydrogen sulfide [H2S, sour corrosion], form as scales of rust or precipitates. A byproduct
oxygen [O2] and corrosion causing microbes, of the reaction at the cathode is hydroxyl ions
referred to as microbiologically influenced cor- [OH] from the reduction of oxygenated water.
rosion (MIC).8
36
Iron can also react with CO2 to form iron carbonate [FeCO3] and with H2S to form iron sulfides
[FexSx]. In the absence of O2 but the presence of
CO2 and H2S, the cathodic reaction can generate
hydrogen gas.
These reactions can occur rapidly, but if the
reaction rate can be reduced, the overall corrosion rate will also be reduced. Many factors influence the reaction rate. These include the type
and quality of metal, electrolyte compositions,
pH, temperature, pressure, presence of dissolved
gases, liquid velocity, water salinity, application of cathodic protection and the presence of
microbes.11 To manage corrosion and corrosion
rate, knowledge of the metallurgy of the materials to be used and the environments in which
they will operate is important.
If CO2 comes into contact with water in the
producing or transportation system of an oil and
gas operation, areas typically affected include
well internals, gathering lines and pipelines. In
CO2 corrosion of iron, the products of reaction
are carbonic acid, iron carbonate [FeCO3] and
hydrogen gas [H2].12 For CO2 corrosion to occur,
the partial pressure of the gas can be as low as
21kPa [3 psi]. To prevent this type of corrosion,
operators commonly use organic films that act as
barriers and inhibitors that neutralize the acidity
of the carbonic acid generated in the corrosion
process. Operators may also use corrosion resistant alloys (CRAs), which are resistant to general
and localized corrosion, in environments that are
corrosive to carbon and low-alloy steels.
Hydrogen sulfide is often found in produced
fluids or as a result of MIC.13 Although H2S is not
corrosive, it becomes corrosive in the presence
of water.14 Sour corrosion from H2S can affect
any part of the producing system, including well
internals and oil and gas gathering lines. Oilfield
fluids are considered sour if the produced gas contains more than 5.7mg of H2S per m3 [4parts per
million (ppm)] of natural gas or produced water
has greater than 5ppm H2S.15 At the anode, the
H2S reacts with the iron to form several variants of iron sulfide [FexS] such as mackinawite
[(Fe,Ni)(1 + x)S], pyrrhotite [Fe(1 - x)S] and troilite
[FeS].16 These iron sulfide species precipitate and
can form localized microgalvanic corrosion cells.
The corrosion cells formed during sour corrosion cause pitting, sulfide stress cracking (SSC)
and hydrogen embrittlement.17 Stress corrosion
cracking is a result of tensile stress combined
with a wet environment and often causes shallow, round pits that have etched bottoms accompanied by branching cracks that can lead to rapid
failure. Hydrogen embrittlement occurs when
H2S and H2 diffuse into metal, recombine with
Oilfield Review
Corrosion Form and Appearance

The word corrode comes from the Latin corrodere
meaning to gnaw; it can carry the additional
meaning of eat or wear away gradually.21
Corrosion typically leaves a visible signature that
is characteristic of the agent and mechanism
May 2016
30
25
Corrosion rate of carbon steel, mpy
other molecules and create pressure within the

metal matrix; byproducts of cathodic protection,
galvanic corrosion and other mechanisms may
lead to hydrogen embrittlement.
The failure mode during hydrogen embrittlement depends on the steel type; for example,
low-strength steels exhibit blistering. The failure
mode of high-strength steels can be catastrophic
when the pressure of the trapped gas exceeds the
tensile strength of the metal. To control sour corrosion, operators use organic film formers, H2S
scavengers, metals resistant to SSC, flowline pigging, nitrate treatments and biocides that reduce
the growth of microbes that cause MIC.18
Oxygen-related corrosion in oil and gas
producing environments is often much more
aggressive than corrosion caused by CO2 or H2S
(Figure4).19 Corrosion by oxygen is directly proportional to the concentration of the dissolved
gas. If chlorides, CO2 or H2S are present, the corrosion rate can increase significantly.
Oxygen has the ability to induce corrosion
throughout producing systems. Inhibition of oxygen corrosion is difficult, and corrosion reduction
efforts for production and water handling facilities have usually been directed toward exclusion
of oxygen from the system and the use of oxygen scavengers. Typical oxygen scavengers are
ammonium bisulfite [NH4HSO3], sodium sulfite
[Na2SO3] and sodium bisulfite [NaHSO3].20 In
addition to scavenger stripping, vacuum deaerators are sometimes used to control the corrosive
effects of oxygen on metals.
Exposure to oxygen is also a major source of
drillpipe corrosion. While it is being run in and
out of the well, drillpipe is exposed to atmospheric oxygen. During drilling, drillpipe comes
into contact with oxygen in the mud system.
Both instances can induce corrosion. The usual
expression of oxygen-related corrosion is pitting.
Pitting can even develop under mud left on and
inside drillpipe, where pipe storage racks contact
the pipe and at crevices. Deep corrosion pits in
drillpipe can lead to the onset of fatigue failure.
Drillpipe may be coated with epoxies or resins
to stop corrosion, but the harsh downhole environment often quickly removes these protective
coatings. Pipe dope, lubricating grease applied
to threaded connections, may help prevent corrosion of these connections.
20
15
O2
CO2
10
H2S
50
100
150
200
100
200
300
400
250
300
350
400
450
500
600
700
800
900
Oxygen
Carbon dioxide
Hydrogen sulfide
Gas concentration in water phase, parts per million
Figure 4. Corrosion rates. The relative rates of corrosion in milli-inches/year (mpy) of carbon steel
show pronounced differences when the steel is exposed to varying concentrations of O2, CO2 and H2S.
At a concentration of 5ppm, O2 is almost three times more corrosive than is H2S and 30% more
corrosive than is CO2. Photographs near each curve show the effects of these corrosion agents on
metal surfaces.
15. Stewart M and Arnold K: Gas Sweetening and

8. Popoola LT, Grema AS, Latinwo GK, Gutti B and
Processing Field Manual. Waltham, Massachusetts.
Balogun AS: Corrosion Problems During Oil and Gas
Gulf Professional Publishing, 2011.
Production and Its Mitigation, International Journal of
Industrial Chemistry 4, no. 1 (2013).
16. Ning J, Zheng Y, Young D, Brown B and Nesic S:
A Thermodynamic Study of Hydrogen Sulfide Corrosion
Chillingar et al, reference 3.
of Mild Steel, paper NACE 2462, presented at the
9. Stansbury EE and Buchanan RA: Fundamentals of
NACE Corrosion 2013 Conference and Exhibition,
Electrochemical Corrosion. Materials Park, Ohio, USA:
Orlando, Florida, USA, March 1721, 2013.
ASM International, 2000.
17. Kvarekval J: Morphology of Localized Corrosion Attacks
Brondel D, Edwards R, Hayman A, Hill D, Mehta S
in Sour Environments, paper NACE 07659, presented at
and Semerad T: Corrosion in the Oil Industry,
the NACE Corrosion 2007 Conference and Exposition,
Oilfield Review 6, no. 2 (April 1994): 418.
Nashville, Tennessee, USA, March 1115, 2007.
10. Although the battery analogy is acceptable for
18. Pipeline operators send mechanical devices called pigs
explaining corrosion involving two dissimilar metals,
Oilfield
corrosion processes also take place on single
metals. Review through pipelines to clean the inner surface. This can be
done without halting flow, and the flow stream pushes
16
In single metals, the mechanism for corrosionMAY
consists
of small crystals with slightly different compositions.
Corrosion Fig 4 the pig through the piping.
19. Popoola et al, reference 8.
The anode and the cathode are located on different
ORMAY 16 CRSSN
4
areas of the metal surface and, depending on the
Chillingar et al, reference 3.
conditions, may be close to each other or far apart.
20. Chillingar et al, reference 3.
11. Heidersbach R: Metallurgy and Corrosion Control in
Care must be taken when using NH4HSO3 as an oxygen
Oil and Gas Production. Hoboken, New Jersey, USA:
scavenger. This compound is corrosive in itself and can
John Wiley & Sons, Inc., 2011.
also act as a food source for bacteria, thereby
12. At elevated temperatures, magnetite [Fe3O4] may
potentially encouraging MIC.
also form.
21. Davis JR: CorrosionUnderstanding the Basics.
13. In MIC, the H2S is produced as a byproduct of the
Materials Park, Ohio: ASM International, 2000.
activities of sulfate reducing bacteria (SRB).
14. Chillingar et al, reference 3.
37
Water
Flow
Anode
Cathode
Steel
General or Uniform
Galvanic
Erosion or Flow Induced
Crevice
Water
Steel
Force
Pitting
Intergranular
Stress
Corrosion Cracking
Corrosion Fatigue
Figure 5. Generalized categories of corrosion. Corrosion can be categorized by appearance and the
agent of causation. These eight corrosion types cover most of the observed corrosion mechanisms
for metals.
that caused it. Although not an exclusive list, corrosion usually falls into one or more of the following categories: general or uniform, localized,
galvanic, erosion or flow induced, crevice, pitting,
under deposit, cavitation, intergranular, stress
cracking and corrosion fatigue (Figure5). Other
types of corrosion include environmental, top-ofline and microbial. Based on the observed characteristics of the corrosion, engineers can adopt
appropriate preventive and mitigationmeasures.
Uniform corrosion is typical of low-alloy
steels and may be observed over an entire
exposed area. Initial evidence of uniform corrosion is surface roughness. The metal becomes
thinner as the corrosion progresses, and it will
eventually fail from internal pressure or external
forces. Because this type of corrosion is linked to
surface exposure, it may be prevented by properly protecting the surface. Uniform corrosion
may occur in equipment used for oilfield operations such as hydraulic stimulation andacidizing.
Localized corrosion occurs at specific sites
rather than over a generalized area and may be
more dangerous than some other types of corrosion because of its unpredictable nature and
38
the potential for rapid growth. Localized corrosion, of which even CRAs such as stainless steels
are susceptible, can be subdivided into pitting,
crevice and under deposit corrosion. Pitting ultimately can cause holes in metal components and
is one of the primary causes of failure in oilfield
hardware, including tubing, casing, sucker rods
and surface equipment.
Crevice corrosion occurs in constricted areas,
wherein the metal at the crevice becomes anodic
and the rest of the metal serves as the cathode.
The crevice can form where two dissimilar metals
come into contact or be created by microgalvanic
cells that may occur in certain steel alloys.
Pitting corrosion rates are often much higher
than those of other types of corrosion. Inhibitors
Oilfield
may be applied
to theReview
surface to prevent initiaMAY 16
tion, but once Corrosion
a pit has formed
Fig 5 the inhibitors are
often unable toORMAY
slow its16
growth.
CRSSN 5
Under deposit corrosion occurs when sand,
corrosives or porous solids adhere to the metal
surface. Although the area underneath the
deposit is resistant to inhibitors and can corrode
quickly, this type of corrosion can often be man-
aged by cleaning internal piping surfaces, for

example, with the use of pipeline pigs.
Galvanic corrosion can be a problem when
two dissimilar metals are in contact. The metal
that has the least resistance to corrosion acts as
the anode and the more resistant metal serves as
the cathode. The anode typically corrodes preferentially. This form of corrosion is frequently
observed in offshore platforms and pipelines.
The galvanic series, which orders metals according to their anodic or cathodic tendencies, is a
good predictor of corrosion severity (Figure 6).
Galvanic corrosion is controlled and mitigated by
use of the following:
good engineering designto ensure that corrosively active components present larger surface area than do less active components
material selectionto avoid metals far apart
in the galvanic series
isolationto provide pipelines coming from the
sea with sacrificial anodes and protect those
going into land with impressed current systems
inhibitors and coatingsto control initiation
of corrosion, although this method may be ineffective once corrosion forms.
Oilfield Review
Flow-induced corrosion occurs when liquid

flow accelerates corrosion. Wellheads and pumps
are susceptible to this form of corrosion, which
may occur as erosion or cavitation. Erosion corrosion results when fluid flow removes the protective film that forms naturally or has been
applied externally. Because of their abrasive
properties, suspended solids will accelerate the
process. Damage can be seen as grooves in the
piping that correspond to the flow direction.
Proper engineering design that allows for sufficient pipe diameter and removing solids from
flow streams can minimize this type of corrosion.
Inhibitors may be applied to replace protective
films stripped away by the flowing fluids.
Pipe
Wet gas
Condensate
Monoethylene glycol
Anodic
Magnesium
Zinc
Cadmium
Aluminum
Steel
Chromium steel
Stainless steel
Lead
Figure 7. Top-of-line pipeline corrosion. Top-of-line corrosion can

result from the stratified multiphase flow of wet gas in horizontal
pipelines. Liquidsincluding condensate and inhibitors such as
monoethylene glycolsettle to the bottom of the pipe. Wet gas fills
the pipe above the liquid line. If either CO2 or H2S are present in the
gas, along with water, corrosive byproducts form at the top of the
pipe and may not be controlled if the inhibitor remains at the bottom
of the pipe.
Tin
Nickel
Inconel
Hastelloy
Brasses
Copper
Bronzes
Monel
Chromium steel
Silver
Titanium
Graphite
Gold
Platinum
Cathodic
Figure 6. Galvanic series. Metals (not all shown)

can be described by their anodic or cathodic
tendencies arranged in a galvanic series. When
dissimilar metals are connected electrically
and submerged in an electrolyte, the anodic
metal, rather than the cathodic metal, will
preferentially corrode. The rate of corrosion is
a function of the separation between the paired
metals in the galvanic series. The series shown
here is for seawater; the order may change
based on the electrolyte.
May 2016
Oilfield Review
MAY 16
Corrosion Fig 6
Cavitation is caused by collapsing bubbles hydrogen embrittlementhydrogen enters

that occur when the pressure changes rapidly in
the metal matrix and weakens it
flowing liquids. Over time, cavitation may cause stress corrosion crackingcracks form after
deep pits to form in areas of turbulent flow, especorrosion has attacked a surface
cially in pump impellers. Low-carbon steels are sulfide stress crackinga failure of the metal
susceptible; stainless steels are more resilient.22
caused by H2S.
Material
selectionopting for materials that are
Intergranular corrosion results from corrosive attacks at metal grain boundaries in the resistant to hydrogen embrittlement and sulfide
form of cracks. The grain boundaries can become crackingis the primary avoidance technique.
anodic with reference to the cathodic surround- Low-stress design practices and stress relief by
ing surface, typically due to formation of chro- heat treatment are also commonly used, and premium carbides or nitrides. Metal impurities can venting corrosion in components subject to stress
increase the effect, as can precipitates in the is another method.
Pipelines are subject to top-of-line corrosion
metal that form during heat treatments. When
Oilfield Review
(Figure
7). Water condenses at the top of the
chromium combines with nitrogen orMAY
carbon,
16
pipe
as
the
fluid inside cools. The corrosion rate
less free chrome is available locally for
corroCorrosion Fig 7
depends
on
sion protection, and cracks can form along
the
ORMAY 16 CRSSN 7 the condensation rate and concengrain boundaries. Quenchingthe rapid cool- tration of organic acids. Generally, this type of
ing after heat treatmentsmay be effective in corrosion is controlled with inhibitors and pipereducing or eliminating intergranular corrosion. line insulation that reduces condensation.
Material selectionavoiding metals that are
22. Port RD: Flow Accelerated Corrosion, paper NACE721,
susceptible to this conditionis the most relipresented at the NACE Corrosion 98 Annual Conference,
San Diego, California, USA, March 2227, 1998.
able method to preclude intergranular corrosion.
International: Standard Practices for Detecting
Tests such as ASTM A262 can be used to evaluate 23. ASTM
Susceptibility to Intergranular Attack in Austenitic
Stainless Steels, West Conshohocken, Pennsylvania,
susceptibility of materials to this mechanism.23
USA: ASTM International A262-15, 2015.
Environmental cracking occurs when cor
rosion coincides with tensile stress. It may be
manifested as the following:
39
Separator Tank
Sidestream
Emulsion
Water outlet
Oil outlet
Water
Oil layer
Corrosion
inhibitor film
Iron sulfide
particle
Figure 8. Microbiologically influenced corrosion products. Softscale

corrosion, referred to as schmoo (right), can form in production systems
if microbes are not controlled. The photograph shows a mixture of
iron sulfide [FeS], asphaltenes and biomass that was collected at the
sidestream outlet of a separator tank (top left). Corrosion inhibitors form
protective films around iron sulfide particles (bottom left) inside the
separator and prevent softscale formation in produced waters.
Microbiologically Influenced Corrosion

From the moment of immersion in a nonsterile fluid that supports microbial development,
microorganisms begin to attach to the material
surface (Figure 8). Planktonic microorganisms
become fixed at the fluid boundary, typically a
pipe wall, or onto porous media such as preexisting corrosion.24
Attachment of microorganisms leads to the
formation of biofilmsmicrobial communities
embedded in or on an attaching surface.25 Wherever biofilms are found in producing systems,

MIC can occur, including inside production tubing, gravity and hydrocyclone separators, storage
tanks, pipelines and water injection systems.
Depending on the microbial species, the corrosion mechanisms can take various forms.
Biofilms can trap ions and create localized
electrochemical potentials analogous to a galvanic corrosion cell or may contribute to corro-
Oilfield Review
MAY 16
Corrosion Fig 9
ORMAY 16 CRSSN 9
Figure 9. Microbiologically influenced corrosion from byproducts. Biofilm on

the surface of this metal piece produced H2S that damaged the piece and
led to premature failure of the equipment.
40
sion by taking a role in the formation of cathodic

and conductive corrosion products on the metal
surface (Figure 9). Sulfate producing prokaryotes
(SPPs) are the chief offenders. Prokaryotes are
microbes that have no cell nucleus or membranebound organelles. The most prominent group of
SPPs are sulfate reducing bacteria (SRB) and
sulfate reducing archaea (SRA). They contribute
to corrosion by various means, for example, by
taking a role in the formation of cathodic ferrous
sulfide corrosion products and the formation of
galvanic cells. The production of H2S by SPPs can
also lead to sour corrosion.
Biofilms may develop local concentration
cells that are created by oxygen depletion or may
attach to a metal surface. Microbes can contribute to corrosion by the direct effects of metabolic waste products such as organic acids that
are capable of altering the local pH and forming
pH cells. Some microbes are anaerobic and can
tolerate extremes of pressure, temperature, pH
and fluid salinity. These include methanogens
microbes that produce methane as a metabolic
byproduct in anoxic conditions.
Regardless of the source of MIC, prevention
measures in most cases attack planktonic and
sessile populations.26 Methods include biocides
to kill the microorganisms, coatings to inhibit
biofilm formation, removal of nutrients from the
flow stream to control microbial populations and
mechanical removal of an established biomass
via pigging.
Corrosion Control Methods
Metallurgical solutions can be effective deterrents of corrosion, but their costs may be beyond
the economic limit of many oilfield projects.
Building every structure and tubular from iridiumthe most corrosion resistant element
might win the battle against corrosion, but incur
unsustainable expenses, and that would be
assuming a sufficient supply of iridium exists in
the world to attempt such a task. Aluminum is
a corrosion-resistant metal used in many oilfield
applications; however, it is unsuitable for highpressure and high-temperature operations. In
addition, although aluminum is considered corrosion resistant in seawater, the mechanism for
resistance relies on the formation of a thin film
of aluminum oxide on the surface of the metal. In
environments that have high levels of acidity (low
pH) or alkalinity (high pH), the aluminum oxide
can become unstable and thus nonprotective. In
many cases, steel alloys and CRAs are required to
meet both strength and cost requirements.
Oilfield Review
Although materials selection is a major part

of the corrosion control process, once the equipment is deployed, oilfield operations generally
follow three methodologies to battle corrosion.
Operators and service companies rely on surface
coatings to protect susceptible metals, cathodic
protection for active protection and inhibitors as
a low-cost treatment option.
Surface coatings provide chemical and
mechanical resistance. They may also offer thermal protection. For surface coatings to provide
maximum effectiveness, good adhesion to the
target surface is required. Coatings are available
in organic and inorganic types. Organic coatings
include epoxies, phenolic resins, polyurethanes,
polyethylenes and polyesters. Metals applied as
suspensions and electroplating are examples of
inorganic coatings; inorganic ceramics may also be
applied to protect surfaces. Although not normally
an advanced-technology solution, the cement
placed in the annulus between the wellbore casing
and the formation can act as an inorganic coating
that prevents corrosion.
Cathodic protection (CP) consists of two primary forms: passive and active (Figure 10). In
either form, it relies on a movement of electrons
(current) from an external anode to the equipment being protected, which acts as a cathode.
Both the cathode and anode must be in the same
electrolyte and electrically connected. The most
common uses of CP are protecting large structures, piping, casing and equipment exposed to
the elements. It may also be installed inside or
outside tanks and pressure vessels.
Operators often use sacrificial anodes with
CP to protect structures in areas where electrical
power sources are not readily available such as
in remote operations or on offshore structures. If
the structure can be made to serve as the cathode
in relation to an anode, the disposable sacrificial
anode will corrode while the cathode remains
unscathed. This type of CP has been referred to
as fighting corrosion with corrosion.27
The first use of CP is attributed to Sir
Humphry Davy, who described the process in a
series of articles to the Royal Society of London
in 1824.28 The technique was used in an attempt
to prevent the corrosion of nails used in wooden
oceangoing vessels. Accelerated corrosion of the
nails occurred when copper claddingused to
prevent biofoulingwas applied to the outside
of vessels. Davy found that sacrificial anodes
protected the iron nails. The actual processes
were not well understood at that time, but it is
recognized today that the contact of dissimilar
metalsthe copper cladding and the iron of the
May 2016
Galvanic DC current
milliamp
Electrolyte
Cathode
Anode
Backfill
Figure 10. Cathodic protection circuit. Cathodic protection methods may

use naturally occurring galvanic current or employ a direct current (DC)
source (impressed current) when the electrolyte is resistive. The protected
elementa pipeline is shownis the cathode. The sacrificial element,
located some distance from the cathode, serves as the anode. The DC
source may be batteries or solar panels in remote pipeline applications.
nailsled to the corrosion. Davy and his assisBecause the direct current (DC) is extertant carried out a number of experiments on cor- nally applied, this type of corrosion managerosion prevention techniques; that assistant was ment is referred to as impressed cathodic
Michael Faraday, who would later establish the protection. It is most frequently used for cases
relationship between the chemical action of cor- in which the electrolyte resistance is high, such
rosion and electric current.
as in soil or freshwater, and where a constant
In the oil field, CP was first applied to land24. Stipanicev M: Improved Decision Support Within
based pipelines, and the first documented use
Biocorrosion Management for Oil and Gas Water
Injection Systems, PhD thesis, Institut National
was by Robert J. Kuhn in 1928.29 He established
Polytechnique de Toulouse, France (2013).
a negative 850 mV potential between the steel 25. Madigan MT, Martinko JM, Bender KS, Buckley DH,
Stahl DA and Brock T: Brock Biology of Microorganisms,
pipe of a pipeline and a copper-sulfate electrode.
14th ed. San Francisco: Benjamin Cummings, 2014.
This example became the foundation of modern
Oilfield Review
26. Sessile refers to fixed or immobile organisms.
CP technology, although for many years the
effecMAY
16 27. Lehmann JA: Cathodic Protection of Offshore
Structures, paper OTC 1041, presented at the
tiveness was met with scientific skepticism.
Corrosion Fig 10
First Annual
ORMAY
10 Offshore Technology Conference, Houston,
Today, CP uses sacrificial elements
made 16 CRSSN
May 1821, 1969.
from aluminum, zinc and magnesium to protect 28. Davy H: On the Corrosion of Copper Sheeting by Sea
Water, and on Methods of Preventing This Effect; And
the steel of large structures and piping. These
on Their Application to Ships of War and Other Ships,
dissimilar metals create the galvanic coupling
Philosophical Transactions of the Royal Society of
London 114 (January 1, 1824): 151158.
that establishes a current path between the
Baeckmann W: The History of Corrosion
anode and the cathode, and, over time, the 29. von
Protection, in von Baeckmann W, Schwenk W and
Prinz W (eds): Handbook of Cathodic Corrosion
sacrificial anode rather than the protected
ProtectionTheory and Practice of Electrochemical
structure experiences metal loss. Appropriate
Protection Processes, 3rd ed. Houston: Gulf Professional
Publishing (1997): 126.
placement and distribution of the anodes is crucial to ensure that all parts of the structure are 30. Amani M and Hjeij D: A Comprehensive Review of
Corrosion and Its Inhibition in the Oil and Gas Industry,
sufficiently protected.30
paper SPE 175337, presented at the SPE Kuwait Oil
and Gas Show and Conference, Mishref, Kuwait,
October 1114, 2015.
41
Water
Water
Oil
Oil
CH 3
Alkyl
chain
CnH 2n
N+
Polar head
group
Metal surface
Figure 11. Film formers. Although they vary in composition and avenue of
protection, film formers create barriers between corrosive elements (water
and oil, top) and metal surfaces. Inhibitors may be adsorbed on the surface
(alkyl chains, middle) or form a strong bond by sharing charges with the
metal (polar head group, bottom). When molecules of the polar head group
of film formers attach to the surface of the metal, a portion of the molecule
extends into the fluid. This usually oil-soluble tail is hydrophobic, repelling
water away from the metal surface.
source of current is readily available. The use is to interrupt the electrochemical process by
of solar panels in remote locations has greatly which the corrosion cell forms between the metal
increased the potential applications of impressed and the liquids in and around the equipment.
Inhibitors can be a flexible and cost-effective
cathodicprotection.
In the impressed CP technique, current of method of fighting corrosion, and the inhibiseveral amps from a low-voltage rectifier passes, tor application can be altered when conditions
or is impressed, from an inert anode (for exam- change. Although acquiring and delivering the
ple, graphite or iron) to the structure being pro- inhibitor incur an ongoing cost, the lower costs
tected, which acts as the cathode. The anode associated with using less corrosion resistant
is attached to the positive terminal of the DC low-carbon steels usually more than make up
source, and the cathode is attached to the nega- thedifference.
Inhibitors fall into four main categories: scavtive terminal. The anode and cathode are often
some distance from each other, separated by engers, reactive agents, vapor phase and film
formers. Oxygen scavengers are frequently used
anelectrolyte.
To counteract corrosion, sufficient current in operations in which oxygen poses a corrosive
density must be supplied to all parts of the pro- threat. These agents not only reduce oxidizing
tected structure and the current density must corrosion, but also control the growth of microbes
always exceed what would be the measured cor- that require oxygen to thrive. Examples of oxygen
scavengers used in the oil and gas industry are
rosion rate under the same conditions.Oilfield
If theReview
corrosion rate increases, the impressed MAY
current
16 sodium sulfite, sulfur dioxide, sodium bisulfite,
Corrosion
11 metabisulfite and ammonium bisulfite.
sodium
density must be increased.31 Although the
initial Fig
ORMAY
CRSSN 11 bisulfite and sodium bisulfite are
equipment cost may be higher for impressed
CP 16Ammonium
than it is for sacrificial protection, this technique commonly used in seawater injection systems. To
may be less expensive over the long term because speed reaction rates, a catalyst may be included
sacrificial anodes do not need to be replaced. in the chemical.
Hydrogen sulfide scavengers reduce the
Impressed CP also has the advantage of providing information to the operator about the extent level of H2S in the flow stream. Examples of H2S
of corrosion over time.
scavengers are amines, aldehydes and zinc carboxylates. Common forms of amines are monoCorrosion Inhibitors
ethanolamine (MEA) and monomethylamine
Another line of defense against corrosion is (MMA) triazine. In some situations, operators
inhibitors, of which there are a variety of types may be able to regenerate MEA and MMA for
and applications. The primary goal of inhibitors reinjection and reuse.
42
Reactive inhibition operates at the cathode

level for the corrosion cell. The cations of the
inhibitor react with the cathodic anions to form
insoluble films, which adhere to the surface
of the metals and prevent O2 from coming into
contact with the metal. These films also prevent
the evolution of H2, a byproduct of the corrosion
cell. Examples are forms of calcium carbonate,
magnesium carbonate and iron oxides. Reactive
inhibitors can also serve as poisons to the corrosion cell process by interfering with the formation of H2 and reducing the reaction rates at both
the cathode and anode.
Vapor phase inhibitors are primarily used for
combating CO2 corrosion. These inhibitors neutralize CO2 and block the formation of carbonic
acid [H2CO3]. They are transported via vapor
phase in wet gas lines. To protect against future
corrosion, they may also be used during hydrostatic testing of components with water, especially when the components are to be stored after
fitness testing. Examples of these types of inhibitors include morpholine and ethylenediamine.
Film Formers
Film formers are the most widely used corrosion
inhibitors in the oil and gas industry. They create a continuous layer between the metal and the
reactive fluids, thus reducing the attack of corrosive elements (Figure 11). They may also attach
to the surface of corroded metal, altering it and
reducing the corrosion rate. Although they are
effective in reducing CO2 and H2S corrosion, film
formers are not effective against O2 corrosion.
Film formers are available in oil-soluble,
water-soluble and oil solublewater dispersible
forms. Oil-soluble inhibitors are used to treat oiland gas-producing wells. Water-soluble inhibitors are used in high water-cut flow streams,
including those found in producing wells, transmission lines and separators. Oil solublewater
dispersible inhibitors are used in oil and gas
wells that are also producing water.
Film-forming inhibitors take various chemical
forms but are typically composed of long carbon
chains with nitrogen, phosphate esters or anhydrides. Inhibitors may adhere to or be adsorbed
on the metal surface, which prevents the corrosives from attacking the metal. The most effective film-forming inhibitors create a molecular
bond at the metal surface in a process of charge
sharing or charge transfer. For effective inhibition, the surface of the metal being protected
must be fully covered; injection of the proper
concentrations of the inhibitor are crucial. After
they interact with the corrosive elements, some
Oilfield Review
inhibitors are gradually removed from the metal

surface and must be continuously replenished
with new inhibitor.
In the petroleum industry, organic inhibitors
are frequently used because they can form protective layers even in the presence of hydrocarbons.
Amides and imidazolines are examples of organic
film-forming inhibitors that are effective over
a wide range of conditions, especially in sweet
(CO2) and sour (H2S) gas corrosion environments.
They can be water or oil soluble. Amines, which
are also organic inhibitors, are effective for sweet
and sour corrosion but may exhibit biologic toxicity and are thus not as environmentally friendly
as are amides.
Quaternary ammonium salt, or quaternary
amine, inhibitors are effective against sour corrosion.32 The corrosive element formed by sour gas
is iron sulfide on the metal surface. Quaternary
ammonium cations, or quats, are positively
charged, and when they are adsorbed on the
surface of the material to be protected, they disrupt the normal corrosion cell charge. However,
at least one study indicated that quaternary
ammonium inhibitors may actually increase the
corrosion rate of sweet corrosion in the presence
of brine.33 The biocide properties of quaternary
ammonium salts may also preventMIC.
Many additional film formers are used in the
oil and gas industry, including phosphate esters,
ester quats, dimer and trimer acids and alkyl pyridine quaternary compounds. Most film-forming
applications include multiple inhibitors; laboratory testing is used to establish optimum concentrations, fluid tolerances, stability, effectiveness
and persistency of the film. Inhibitor selection
can be a complicated process and typically must
be adjusted over time to meet the demands of
changing fluid conditions.
Inhibitor Selection
Laboratory evaluation is the key to developing
an effective program in inhibitor selection for
corrosion control. Technicians begin the process
using fluid samples that replicate field conditionsactual produced fluids are best if available. Simulated and synthetic fluids are used
when produced fluids cannot be obtained. From
laboratory tests, corrosion rates can be measured and predictions can be made for largescale operations (Figure 12). Methods for testing corrosion inhibitors include the following
tests: wheel, kettle (also called linear polarization resistance (LPR) tests), rotating cylinder
May 2016
Use of corrosion
inhibitor recommended
Predict corrosion rate

from field data
Develop and execute

testing program
Rotating Cylinder
Electrode Test
Kettle Test
Autoclave Test
Recommend and
implement corrosion
inhibitor addition
Conduct field trial
Sidestream Test
Analyze field
trial results
Make final
recommendation
Figure 12. Laboratory testing of corrosion inhibitors. Operators usually

develop corrosion control plans and then test inhibitors using conditions
expected from the field. This flowchart follows a testing sequence. Three
common testing methods are the rotating cylinder electrode, kettle and
autoclave tests (middle). Even after laboratory testing, field trials should be
conducted to validate the effectiveness of the program. A sidestream test
(lower left ) acquires samples for analysis. If the proposed method provides
acceptable results, the method is adopted, although the corrosion inhibition
program must be reevaluated during the life of the well.
31. Schweitzer PA: Corrosion of Linings and Coatings:

Cathodic and Inhibitor Protection and Corrosion
Monitoring. Boca Raton, Florida: CRC Press, 2006.
32. Binks BP, Fletcher PDI, Hicks JT, Durnie WH and
Horsup DI: Comparison of the Effects of Air, Carbon
Dioxide and Hydrogen Sulphide on CorrosionOilfield
of a Low
Carbon Steel under Water and Its Inhibition by a

Quaternary Ammonium Salt, paper NACE 05307,
presented at the NACE Corrosion 2005 Conference and
Exhibition, Houston, April 37, 2005.
33. Binks et al, reference 32.
Review
MAY 16
Corrosion Fig 12
ORMAY 16 CRSSN 12
43
Rack
Metal coupon container

High-temperature autoclave
Pressure source
(hydraulic pump)
Figure 13. Autoclave corrosion testing. A high-temperature autoclave is

used to test the effectiveness of corrosion inhibitors on metal coupons
(inset). Hydrostatic pressure and temperature can be applied to simulate
downhole conditions.
Thermometer
Voltage potential, mV
Gas sparge
electrode, autoclave, jet impingement and flow

loop (Figure 13). The most common are the
wheel and kettle tests.34
The wheel test measures the loss of metal
during a specified period of exposure to corrosive liquids. Corrosives include produced fluids,
brines and refined oils. The test fixture includes
a rotating wheel inside a sealed box that keeps
the specimen, usually strips of metal or coupons,
in constant motion. Temperature can be maintained at a constant value or varied to simulate
field conditions. The samples are tested with and
without inhibitor and the results are compared.
The kettle test, or LPR test, measures corrosion rates electrochemically. Metal electrodes
are placed in the test vessel, which is heated
while the corrosive fluid is continously agitated.
Agitation attempts to replicate field conditions
mild agitation is similar to flow of two distinct
sources, and high agitation replicates turbulent
fluid flow that has dispersed hydrocarbons. To
simulate the presence of gases, CO2 and H2S can
be bubbled through the liquid in the vessel in a
process referred to assparging.35
To establish a control corrosion rate, the test
is run with the electrodes exposed to the fluids
in aqueous phase without an inhibitor and then
followed by a series of tests on solutions that have
increased inhibitor volumes. Linear polarization
is performed by controlling the voltage potential
and measuring the current then controlling the
Steel electrode
Heating mantel
Voltage
Corrosion rate
Current
Current, mA
Stir bar
Figure 14. Kettle test. To perform kettle tests, or linear polarization resistance
tests, technicians use a test fixture (left) and control the pressure and
temperature. They submerge electrodes inside the fixture into the fluids
expected downhole and then measure electrical properties of the
electrodes. The tests are performed by controlling the voltage potential
44
and measuring the current then controlling the current and measuring
the voltage. The electrolyte can be agitated using the stir bar. Gas can be
injected into the test fixture, a process referred to as sparging. From the
slope of the polarization resistance curve (right ), the corrosion rate can
becomputed.
Oilfield Review
Cape Lopez
Cathodic protection station
Section 1
Section 2
Section 3
Tchngue
AT LA N T IC O CEAN
GABON
Batanga
Cathodic protection stations

powered by two solar cells
Input temperature
~ 60 C
Rabi Field
Figure 15. Corrosion in a pipeline from the Rabi field to Cape Lopez. A
three-section, 18-in. pipeline carries oil from the inland Rabi field in Gabon
to Cape Lopez on the coast. Cathodic protection stations are located along
the pipeline. Because the incoming oil is hot (around 60C), Section 1 of the
pipeline (red and dark blue) is exposed to a higher temperature than is the
remainder of the pipeline. The elevated temperature led to the disbondment
current and measuring the voltage potential.

The data are plotted, and the slope of the line
is the polarization resistance, which is inversely
proportional to the corrosion rate (Figure 14).
This technique provides corrosion rate evaluation from external measurements, whereas other
methods require technicians to physically measure and evaluate corrosion.
The effectiveness of inhibitors is dependent
on fluid velocity. For fluids containing little or
no solid particles, high flow rates can lead to
flow-accelerated corrosion. If the flow stream
contains solid particles, the accelerated corrosion is termed erosion corrosion. Several test
methods have been developed to model corrosion
in high-flow conditions and determine a films
persistence, especially where turbulent flow is
present.36 Test methods include jet impingement,
rotating cylinder electrodes and flow looptesting.
The testing of inhibitors should determine
the following:
May 2016
of the protective outer covering of the pipeline. Engineers concluded

that corrosion observed in Section 1 resulted from a combination of the
disbonding of the protective covering and ineffective cathodic protection.
Although the pipelines safety was not compromised, the operator
implemented new procedures to prevent the corrosion from recurring.
thermal stability
emulsification tendency
foaming tendency
metal compatibility
elastomer compatibility
compatibility with other chemicals used in the
same stream.
Application methods should be evaluated as well.
Injection may be continuous, batch or squeeze.
The rate of film removal is a key concern when
determining the optimal application mode.
Corrosion in the Oil Field
A recent example of pipeline corrosion from
Gabon illustrates the need for thorough testing
and understanding of the corrosion process.37 A
pipeline transports oil from the Rabi field to Cape
Lopeza distance of approximately 234km
[145mi]. The 18-in. pipeline comprises three
sections: Section1 from Rabi to Batanga, 105km
[65mi]; Section2 from Batanga to Tchengu,
100km [62mi] and then Section3 from Tchengu

to Cape Lopez, 29km [18mi] (Figure15).
The inlet pressure at Rabi was about 40bar
[580psi], and the flowing temperature was 60C
[140F] at the inlet. Beyond the inlet, the line operates at about 35C [95F]. Impressed cathodic protection is used for the pipeline, which has sections
that have solar cells to provide current. The pipeline
was coated with three-layer polyethylene; each joint
was brush cleaned and wrapped with heat-shrink
34. NACE Task Group T-1D-34 on Laboratory Corrosion
Inhibitor Test Parameters: Laboratory Test Methods for
Evaluating Oilfield Corrosion Inhibitors, Houston, NACE
International, NACE Publication 1D196, December 1996.
35. NACE Task Group T-1D-34 on Laboratory Corrosion
Inhibitor Test Parameters, reference 34.
36. Efird KD: Jet Impingement Testing for Flow Accelerated
Corrosion, paper NACE 00052, presented at NACE
Corrosion 2000 Conference and Exhibition, Orlando,
Florida, March 2631, 2000.
37. Melot D, Paugam G and Roche M: Disbondments of
Pipeline Coatings and Their Effects on Corrosion Risks,
Journal of Protective Coatings and Linings 26, no. 9
(September 2009): 6776.
45
may have prevented cathodic current from reaching

and protecting the surface of the exposed steel.
Although engineers discovered corrosion as a result
of disbonding of coatings, based on ASME standards,
the degree of corrosion was deemed not mechanically dangerous. They also concluded that as long as
coatings remain bonded to the steel and cathodic
protection is correctly applied, monitored and maintained, no corrosion risk existed for this pipeline.
Figure 16. An example of pipeline corrosion. After the protective coating

disbonded on a pipeline in Gabon, corrosion formed as pitting (inset).
sleeves and hot-melt adhesive that overlapped the

polyethylene.38 The pipe was buried in wet, compacted sand that had a pH of approximately 5.4.
Problems began to develop in the first 15km
[9.3mi] of pipe. Routine inspections found disbondment at the sleeves where they overlapped
the polyethylene coating in the Rabi section.
Disbondment allowed water under the protective
coating, which negated the cathodic protection
and allowed corrosion to develop (Figure 16).
The remainder of the pipeline did not experience
the same level of corrosion, although significant
BURKINA FASO
BENIN
disbondment occurred in the sleeves. The main

difference between the sections that had differing corrosion levels was that the temperature in
the more corroded sections was higher. Further
testing of pipe Sections 2 and 3 found no evidence
of similar levels of disbondment or corrosion.
After a thorough examination, engineers recommended abrasive blast cleaning prior to applying
heat-shrink sleeves for future installations rather
than the standard brush cleaning of connections.
Another possible solution was liquid polyurethane
or epoxy applied at the joints. The disbonded coating
Corrosion Inhibitors in Deepwater

Deepwater projects can pose unique challenges
for corrosion control because the completions
are usually located at the seafloor and flowlines
must come to the surface or back to shore. A
deepwater field located in the southern Niger
delta demonstrates the use of inhibitors to combat CO2-induced corrosion (Figure 17).39
The production path for deepwater wells
passes through cold water, which can subject the
originally hot fluids in the flow stream to rapid
cooling. Conversely, inhibitor injection is often
through long umbilicals that are subjected to
temperature contrasts between the surface and
subsea wellheads. Injection of inhibitors is further complicated by the normally high flowing
pressures associated with deepwater production.
Temperature extremes, pressure extremes
and long umbilicals combine to affect inhibitor
FPSO vessel
NIGERIA
COTE
GHANA
DIVOIRE
Niger Delta
field
Gulf of Guinea
CAMEROON
GABON
Subsea
wellheads
Flowlines and
umbilicals
Figure 17. Niger delta subsea operations and a floating production, storage, and offloading (FPSO) vessel. Production from subsea wellheads (yellow) at a
field in the Niger delta off the coast of Nigeria (inset) is sent to an FPSO. Oil is transferred to tankers, and natural gas is piped directly to the mainland.
46
Oilfield Review
38. Roche M: The Problematic of Disbonding of Coatings

and Corrosion with Buried Pipelines Cathodically
Protected, presented at the 10th European Federation
of Corrosion, Nice, France, September 1216, 2004.
39. Jenkins A: Corrosion Mitigation in a Deepwater
Oilfield Case Study, paper IBP1194_15, presented at the
Rio Pipeline Conference and Exposition, Rio de Janeiro,
September 2224, 2015.
40. API: Attributes of Production Chemicals in Subsea
Production Systems, Washington, DC: API,
API Technical Report 17TR6, 2012.
41. Jenkins, reference 39.
May 2016
Inhibitor
Dose Rate, ppm
Uninhibited
Corrosion Rate, mpy
Inhibited
Corrosion Rate, mpy
Protection, %
DS-1617 inhibitor
10
173.01
4.18
97.58
DS-1617 inhibitor
20
156.43
0.98
99.37
Inhibitor
Dose Rate, ppm
Corrosion Rate, mpy
Protection, %
None
71.04
DS-1617 inhibitor
20
1.16
98.37
Figure 18. Corrosion testing of the DS-1617 inhibitor. Technicians conducted kettle tests with fluids
representative of field conditions to evaluate the effectiveness of the DS-1617 inhibitor (top). They
also performed autoclave testing at high temperature (bottom). The corrosion rate is in milli-inches of
penetration/year (mpy).
Engineers developed the DS-1617 deepwater

corrosion inhibitor to meet the challenges of
thisfacility.
To qualify this inhibitor, they tested the
chemicals in accordance with the API TR 17TR6
standard, which requires replicating the temperatures and pressures experienced by the inhibitor
during deployment through the umbilicals.40 The
evaluation included high-pressure flow-loop stability tests. The engineers conducted additional
tests to look at resistance to hydrate formation,
thermal aging and compatibility with seawater.
Because the operator was concerned about foaming in the glycol regeneration unit, the inhibitor
was tested for foamingtendency.
Laboratory technicians performed kettle
tests using the DS-1617 inhibitor at 20 ppm,
which is a relatively low dosage; the corrosion
rate was reduced by 99% (Figure 18). They also
performed high-temperature autoclave testing
on carbon steel coupons. The samples were subjected to test fluids that had corrosion inhibitor
and to pressurized CO2 heated to 120C [248F].

The results indicated a 98% reduction in the
corrosion rate.41 The 20ppm concentration
yielded corrosion rates of about 0.00016in./year
[0.004mm/year]. For corrosion rate, the standard industry units are milli-inches/year, or mpy.
For this test, the corrosion rate was equivalent to
0.16mpy. Test technicians reported no foaming
problems associated with the inhibitor.
The operator adopted the use of the DS-1617
inhibitor and monitored corrosion at six locations on the FPSO. No corrosion monitoring
was installed on the deepwater flowlines. The
DS-1617 inhibitor was injected at a 100-ppm
rate, which is a lower rate than the initial inhibitor that was deemed insufficient. Criteria for
corrosion protection established by the operator was a rate below 0.05mpy [0.0013mm/year].
Testing at all six locations indicated corrosion
rates below the target rate (Figure 19). Based on
the testing, the operator implemented the use of
the DS-1617 inhibitor.
6.0
Low-pressure separator A
Low-pressure separator B
5.0
Corrosion rate, mpy
stability, performance and properties. Thorough

testing of the inhibitors is required to ensure
corrosion controlling properties are maintained,
that the injected chemicals remain stable and
that the inhibitors can be reliably delivered via
the umbilicals into the flow stream.
Another risk in deepwater production is the
formation of hydratesice-like solids of water
and gas that form above the normal freezing point
of waterthat can plug flowlines. To ensure corrosion controlling properties are maintained,
inhibitors must be thoroughly tested to confirm
that the injected chemicals remain stable and
that the inhibitors can be reliably delivered via
umbilicals into the flow stream.
These conditions were faced by an operator
of a deepwater production platform in Nigeria.
The platform served nine wells drilled in water
depth of 1,030m [3,380ft]. The operator used
subsea completions that included five manifolds
and eight production flowlines and risers. The
flowlines were connected to a floating production, storage, and offloading (FPSO) vessel that
had 320,000m3 [2million bbl] of onsite storage
capacity. Produced oil flowed to the FPSO for
transfer to tankers. Produced gas was directed to
shore via pipelines.
The pipelines used to transport the oil and
gas were constructed of carbon steel. The flowing pressure from the wells averaged 80bar
[1,160psi], and the average temperature was
85C [185F]. The water cut was 45% and the
natural gas contained about 1.4% CO2. The combination of produced water (brine) and CO2
presented a high corrosion-rate potential for
the low-carbon steel. In wet gas pipelines such
as these, produced water has a tendency to condense at the top of the pipe, allowing top-of-line
corrosion; the presence of both water and CO2
accelerates corrosion.
Engineers installed chemical umbilicals of
1to 20km [0.6 to 12.4mi] to inject corrosion
inhibitor into the deepwater production flowlines. As the project progressed, engineers at
M-I SWACO, a Schlumberger company, reevaluated the initial inhibitor used in the project,
which also protected the topside piping and
storage vessels, and deemed it to be insufficient.
Bulk oil treater

Target corrosion rate
4.0
3.0
2.0
1.0
0.0
0
0.5
1.0
1.5
2.0
2.5
3.0
Time, hr
3.5
4.0
4.5
5.0
6.0
Figure 19. Corrosion monitoring at a production facility. The operator injected DS-1617 inhibitor into
the flowlines of producing wells using underwater umbilicals. The corrosion rate of the flowlines was
monitored at the low-pressure separator A (blue), low-pressure separator B (red) and the bulk oil
treater (green) as well as at three other locations (not shown). The corrosion rate dropped below the
target level (black) established by the operator. Corrosion rates remained below the threshold at all
test sites for the duration of the testing period.
47
Anode
Water Depth, m
Weight Loss, %
13
13
73
31
116
25
116
39
Figure 20. Anode corrosion after eight years of service in the North Sea.
North Sea Cathodic Protection

North Sea production platforms routinely use
cathodic protection. On one platform, the operator installed 10 sacrificial anodes below the surface of the water and left them in place for eight
years. The anodes were composed of zinc, silver
and silver chloride and were located at various
depths and locations on the plaform.42 The system
was designed to protect the structure for a minimum of 20 years. Engineers monitored the output
current from three of the anodes over the period.
The anodes were removed and inspected at the
end of eight years.
After retrieval, the sacrificial anodes were
cleaned and weighed (Figure 20). Technicians
determined changes in physical dimensions and

measured electrical properties. Four anodes
were analyzed for the study. The reduction of
anodes that had been placed in deeper water was
greater than that of those placed in shallower
water. Some of the anodes were so corroded that
visual inspection was difficult (Figure 21).
The original 20-year design projected that at
eight years, the anodes should be reduced by 40%;
however, the average weight loss of the anodes
was only 24%. The engineers concluded that the
original design, although conservative, would
protect the structure for at least 20 years. Based
on the results of the study, a model was established for periodic inspections to be performed.
New Developments in Corrosion Control

Controlling corrosion has been an ongoing
battle between humans and nature for millennia. Since scientists such as Sir Humphry Davy
and Michael Faraday discovered some of the
underlying physics that explained corrosion,
various methodologies have been adopted and
adapted. Modern scientific understanding and
new technologies are combining to improve
the tools available to fight the unending battle
with corrosion.
One area of emerging materials science is
nanoparticles and nanostructures.43 Having surface thickness of 1 to 100 nm, these coating materials have unique properties that may make them
almost impervious to corrosion. Nanoparticles
and nanostructures may be deposited on metal
surfaces as films, similar to film-forming techniques, but because of nanoparticles greater
persistence, reapplying them is unnecessary. The
surfaces also become super-slickexhibiting
low friction coefficientswhich reduces wear
and increases durability. Such surfaces are also
less likely to experience biofouling.44
The battle against corrosion will never
be won; entropy will eventually win the war.
Humans will, however, continue to search for
effective means to combat this nemesis. The
costs of ignoring the problem are too great and
the consequences of failure can be potentially
catastrophic. At least in the oil field, operators
are armed with knowledge, science and effective
tools that allow them to actively manage or mitigate the effects of corrosion.
DEA/TS
Figure 21. Cathodic protection on a North Sea platform. Anodes were recovered
after eight years of service from a North Sea platform. After the anodes were
cleaned and weighed, technicians were able to determine the effectiveness of
the anodes at protecting the structure.
42. Roche M: Offshore Cathodic Protection: The Lessons of

Long-Term Experience, paper OMC-2005-020, presented
at the 7th Offshore Mediterranean Conference and
Exhibition, Ravenna, Italy, March 1618, 2005.
43. El-Meligi AA: Nanostructure of Materials and Corrosion
Resistance, in Aliofkhazraei M (ed): Developments in
Corrosion Protection. Rijeka, Croatia: InTech (2014): 323.
48
44. Tesler AB, Kim P, Kolle S, Howell C, Ahanotu O and

Aizenberg J: Extremely Durable Biofouling-Resistant
Surfaces Based on Electrodeposited Nanoporous
Tungstite Films on Steel, Nature Communications 6,
no. 8649 (October 20, 2015).
Oilfield Review
Contributors
Nausha Asrar is the Manager for Materials Support
and Failure Analysis at the Schlumberger Houston
Pressure and Sampling and Formation Evaluation
Centers in Sugar Land, Texas, USA. He began his
career with Schlumberger in 2005 as a senior materials scientist. He previously worked for Shell Global
Solutions in the US, the Saudi Basic Industries
Corporation Technology Center and Saline Water
Conversion Corporation, both in Saudi Arabia, and
as principal corrosion engineer at the Research and
Development Center for Iron and Steel for the Steel
Authority of India, Ltd. A NACE certified material
selection and design specialist, Nausha is a member
of NACE, ASM and SPE as well as a life member of the
Indian Institute of Metals; he is the author of more
than 60 technical papers and reviews on corrosion,
phase diagrams, composite materials and failure cases.
He received an MS degree in chemistry from Aligarh
Muslim University, Uttar Pradesh, India, and a PhD
degree in materials science and engineering from the
Moscow State University.
Bruce MacKay is Client Support Manager for the

Schlumberger North America fracturing and cementing operations in Sugar Land, Texas. He has worked
as a chemical problem solver in various capacities
for Schlumberger for 10 years, spanning the R&D
spectrum from research to product development to
technology implementation. He has authored 12 peerreviewed scientific journal articles and five SPE papers
and has been granted several patents on chemical
technologies related to oilfield applications. He has
been a speaker on the importance of chemistry in oilfield development to a variety of audiences, including
the US National Academy of Sciences, the American
Chemical Society and the National Aeronautics and
Space Administration Jet Propulsion Laboratory.
Bruce was a Natural Sciences and Engineering
Research Council of Canada postdoctoral research
scholar at the California Institute of Technology,
Pasadena, USA. He earned a BS degree in chemistry
and a PhD degree in inorganic chemistry from the
University of British Columbia, Vancouver, Canada.
ystein Birketveit is Technical Manager for Production

Technologies for M-I SWACO, a Schlumberger company,
in Bergen, Norway. For the past 18 years, he has
specialized in the field of corrosion. Prior to joining
M-ISWACO, ystein worked for Statoil and for Det
Norske Veritas. He earned his MSc degree in materials
and electrochemistry from the Norwegian University of
Science and Technology, Trondheim.
Denis Mlot is a Nonmetallic Materials Expert with

the technology department of Total Upstream, in Paris.
Using his foundation of studies in polymer science,
his focus is on nonmetallic materials and corrosion.
Prior to beginning work with Total in 2003, he was a
researcher in the R&D department of Elf Atochem,
which is now Arkema, in Serquigny, France. He then
spent six years as the technical manager for pipe
coating products with the company. Denis chaired
the ISO 12736 working group on wet thermal insulation systems, was a member of pipeline coatings work
group ISO 21809 and holds certifications from the
Association pour la Certification et la Qualification
en Peinture Anticorrosion and Faglig Rd for
Opplring og Sertifisering av Inspektrer innen
Overflatebehandling . He holds numerous patents in
his field and has coauthored several papers on the
subject of coatings and corrosion. Denis has a degree
in materials science from the cole Universitaire
DIngnieurs de Lille, France, and received his PhD
degree in polymer science from Universit de Lille.
Joshua E. Jackson is the CEO of G2MT LLC as well as

the cofounder of G2MT Laboratories, LLC in Houston.
G2MT Labs is a metallurgical consulting and analysis
company that performs nondestructive materials characterization to evaluate residual stress mechanical
properties and other critical parameters including the
effects of corrosion. His scientific focus areas include
corrosion analysis, high-temperature materials, hydrogen absorption effects, failure analysis and statistics.
Joshua is the coauthor of numerous papers in the field
of materials science covering subjects including nondestructive testing, metallurgy, welding, corrosion and
hydrogen. He obtained BS degrees in both mathematics and physics from the Massachusetts Institute of
Technology, Cambridge, USA, and MS and PhD degrees
in metallurgical and materials engineering from the
Colorado School of Mines, Golden, USA.
Alyn Jenkins, based in Aberdeen, serves as the Global
Asset Integrity Manager for Schlumberger Production
Technologies. He manages asset integrity product
lines that include corrosion inhibitors, biocides, H2S
scavengers and oxygen scavengers and is responsible
for research and development projects related to
corrosion. He began his career in 1998 with Clariant
Oil Services in Aberdeen and then worked for Baker
Hughes in Liverpool, England. Alyn joined M-I SWACO
in 2005 as a corrosion specialist in Stavanger and then
served as lead integrity management specialist. Alyn
holds BS and MS degrees, both in chemistry, from the
University of Wales, Bangor.
May 2016
Jan Scheie is a Project Leader and Account Manager

for Production Technologies (PT) in Schlumberger
Norge A/S in Stavanger, where he serves customers in
Scandinavia. He has also been an account manager
for production technologies, an international sales
manager and an area manager for production chemicals in Stavanger. He has worked for M-I SWACO in
market development for the eastern hemisphere, as
technical manager in the Middle East and CIS, as sales
manager in South Asia and as principal engineer for
developing sales strategy in mainland Europe. He is a
member of TEKNA, the Norwegian Society of Graduate
Technical and Scientific Professionals, the SPE and
the National Association of Corrosion Engineers. He
received an MSc degree in chemical engineering from
the Norwegian University of Science and Technology,
Trondheim, Norway, and an MBA degree from
Thunderbird School of Global Management, Glendale,
Arizona, USA.
Marko Stipanicev is Corrosion Discipline Lead for

Schlumberger Production Technologies in Bergen,
Norway. Upon graduation from the University of
Zagreb, Croatia, he worked as an external consultant
on industry related projects at the Faculty of Chemical
Engineering and Technology, in Croatia. Beginning in
2010, he worked as a research engineer for Det Norske
Veritas in Bergen, investigating corrosion-based failures
and performing root cause analysis studies. He joined
M-ISWACO in 2013 as a corrosion specialist, working
in Bergen, and in 2015, he was named the corrosion
discipline lead. Marko is responsible for Schlumberger
corrosion products, which include inhibitors, biocides,
scavengers and nutrients. He has authored and coauthored numerous papers and publications related to
corrosion and corrosion management. He holds an MSc
degree in chemical engineering and technology from
the University of Zagreb and a PhD degree in environmental process and biocorrosion management from the
Universit de Toulouse, France.
Jean Vittonato, is Head of the Total E&P Technology
Division corrosion department in Pau, France. He
is responsible for the headquarters corrosion team
and provides technical assistance to projects and
operating subsidiaries worldwide. He started work in
1999 focusing on cathodic protection with COREXCO,
an engineering cathodic protection company, where
he was in charge of designing cathodic protection
systems for both onshore and offshore and for installation, monitoring and maintenance follow-up. In 2006,
he joined Total as a corrosion specialist and was in
charge of cathodic protection activities. He provided
support for projects for both Total E&P and operating
subsidiaries and was in charge of research projects
related to cathodic protection. He spent three years
in Republic of the Congo as the head of the Total corrosion department, where he supervised all projects
related to corrosion. Jean is a certified Cathodic
Protection Specialist with the National Association
of Corrosion Engineers and with the Centre Franais
de la Protection Cathodique and is chair of the
ISOTC67SC2GW11 working group on cathodic
protection of pipelines. He obtained an engineering degree from Institut National Polytechnique de
Grenoble, France.
49

For decades, the mud motor and bent housing assembly have played a critical role in
directional drilling; however, the technique used to drill a lateral section resulted in
slow drilling rates. A surface-mounted torque control system is helping drillers reach
farther while improving rates of penetration and toolface control.
Steven Duplantis
Houston, Texas, USA
For help in preparation of this article, thanks to
Edina Halilagic and Brandon Mills, Houston.
Slider is a mark of Schlumberger.
1. Maidla E and Haci M: Understanding Torque: The Key to
Slide-Drilling Directional Wells, paper IADC/SPE 87162,
presented at the IADC/SPE Drilling Conference, Dallas,
March 24, 2004.
Directional wells have been a boon to oil and

gas production, particularly in unconventional
plays, where horizontal and extended-reach
wells maximize wellbore exposure through productive zones. In many of these wells, steerable
mud motors have been crucial to achieving the
well trajectory necessary to hit operators target
zones. Directional drillers use a downhole mud
motor when they kick off the well, build angle,
drill tangent sections and maintain trajectory.
A bend in the motor bearing housing is key
to steering the bit toward its target. The surfaceadjustable bend can be set between 0 and 3.
This slight bend is sufficient for pointing the bit

in a given direction yet is small enough to permit
rotation of the entire mud motor assembly during
rotary drilling. This seemingly minor deflection
determines the rate at which the motor builds
angle to establish a new wellbore trajectory. By
orienting that bend in a specific direction, called
its toolface angle, the driller can change the
inclination and azimuth of the well path.
To maintain the orientation of that bend and
thus change wellbore trajectory, the drillstring
must not be allowed to rotate, and this is where
the mud motor comes into play. A mud motor is
Rotation
Mud flow
Wear pad
3.00
2.77
Protective
housing
2.89
2.38
2.77
Effective
bend
2.60
2.89
3.00
Stator
Rotor
Figure 1. Typical mud motor. The bent housing of the mud motor (left) is the key to building wellbore
deviation and controlling wellbore trajectory while the rotor turns the bit. The bend in the housing is
dialed in at the drill floor when the drilling crew makes up the bottomhole assembly; here, the bend has
been set at 2.89 degrees (middle). By selecting a larger bend, the driller is able to obtain curve having
a smaller radius. The motor, installed immediately above the bit, consists of an eccentric rotor within
an elastomer stator (right ). As drilling mud flows through the stator, it displaces the helical rotor shaft,
causing the shaft to rotate within the stators protective housing, which turns the bit.
50
Oilfield Review
a type of positive displacement motor powered

by drilling fluid. An eccentric helical rotor and
stator assembly drive the mud motor (Figure1).
As it is pumped downhole, drilling fluid flows
through the stator and turns the rotor. The mud
motor converts hydraulic power to mechanical
power to turn a drive shaft that causes the bit
torotate.
Using mud motors, directional drillers alternate between rotating and sliding modes of drilling. In rotating mode, the drilling rigs rotary
table or topdrive rotates the entire drillstring to
transmit power to the bit. This rotation enables
the bend in the motor bearing housing to point
equally in all directions and thus maintain a
straight drilling path (Figure 2). In most operations today, measurement-while-drilling (MWD)
tools provide real-time inclination and azimuth
measurements that alert the driller to any devia-
Kickoff 1
Build section, 3/100 ft
Tangent section
Marker bed
Kickoff 2
Build section, 3/300 ft
Lateral section
Reservoir
Gauge hole
Bent housing
Kickoff
Build section
Rotary
mode
Sliding
mode
Increased diameter
caused by outward
tilt of bit
Figure 2. Drilling using a bent housing. In rotating

mode, the bit carves a straight path parallel to
the axis of the drillstring, which is also rotating.
Because the bent housing forces the bit to tilt
outward by a few degrees, the bit drills a hole that
is slightly larger than the diameter of the bit. When
the driller switches to sliding mode, only the bit
rotates. The resulting hole is in gauge and follows
the axis of the BHA below the bent housing.
May 2016
Figure 3. Directional drilling trajectory. After the well is drilled vertically to the specified kickoff point,
the mud motor is used to build angle while slide drilling. When the target angle is achieved, a straightline tangent section may be drilled in rotary mode. While the BHA maintains inclination and azimuth,
the driller resorts to sliding mode only when the drilling direction deviates from the planned trajectory.
In some fields, a stratigraphic marker bed above the reservoir can be detected by LWD tools,
prompting the driller to initiate a second slide section to land the well horizontally within the reservoir.
tions from the intended course. To correct for

those deviations or to alter the wellbore trajectory, the driller switches from rotating to sliding
mode. In sliding mode, the drillstring does not
rotate; instead, the downhole motor turns the bit
and the hole is drilled in the direction the bit is
pointing, which is controlled by toolface orientation. Upon correcting course and reestablishing
the wellbore trajectory needed to hit the target,
the driller may then switch back to rotating
mode. (Figure 3).
Of the two modes, slide drilling is less efficient; lateral reach usually comes at the expense
of penetration rate. The rate of penetration (ROP)
achieved using conventional sliding methods
typically averages 10% to 25% of that attained in
rotating mode.1 Conversely, by rotating the entire
drillstring, drillers gain a substantial advantage
in ROP. This article describes an automated system that helps drillers achieve significant gains
in horizontal reach with noticeably faster rates of
penetration. Field experience in Colorado, USA,
illustrates how a torque-oscillation system can
help operators exploit unconventional plays.
Slide Drilling Challenges

To initiate a slide, the driller must first orient the
bit to drill in alignment with the trajectory proposed in the well plan. This requires the driller to
stop drilling, pull the bit off-bottom and reciprocate the drillpipe to release any torque that has
built up within the drillstring. The driller then
orients the downhole mud motor using real-time
MWD toolface measurements to ensure the specified wellbore deviation is obtained. Following
this time-consuming orientation process, the
driller sets the topdrive brake to prevent further
rotation from the surface. The slide begins as the
driller eases off the drawworks brake to control
the hook load, which, in turn, affects the magnitude of weight imposed at the bit. Minor right
and left torque adjustments are applied manually
to steer the bit as needed to keep the trajectory
oncourse.
As the depth or lateral reach increases, the
drillstring is subjected to greater friction and drag.
These forces, in turn, affect the drillers ability to
transfer weight to the bit and control toolface orientation while sliding, making it difficult to attain
51
sufficient ROP and maintain trajectory to the target. Such problems frequently result in increased
drilling time, which may adversely impact project
economics and ultimately limit the length of a lateral section.
The capability to transfer weight to the bit
affects several aspects of directional drilling. The
driller transfers weight to the bit by easing, or
slacking off, the brake; this transfers some of the
hook load, or drillstring weight, to the bit.2 The
difference between the weight imposed at the bit
and the amount of weight made available by easing the brake at the surface is primarily caused
by drag. As the horizontal departure of a wellbore
increases, so does the longitudinal drag of the
drillpipe along the wellbore.
Controlling weight at the bit throughout
the sliding mode is made even more difficult
by drillstring elasticity, which permits the pipe
to move nonproportionally. This elasticity can
cause one segment of drillstring to move while
other segments remain stationary or move at
different velocities.3
Poor hole cleaning may also affect weight
transfer. In sliding mode, hole cleaning is less efficient because there is no pipe rotation to facilitate
turbulent flow; this condition reduces the drilling
fluids ability to carry solids. Instead, the solids
accumulate on the low side of the hole in cuttings beds that increase friction on the drillpipe,
making it difficult to maintain constant weight on
bit (WOB).
Differences in frictional forces between the
drillpipe inside of casing versus that in open hole
can cause weight to be released suddenly, as can
hang-ups caused by key seats and ledges. A sudden transfer of weight to the bit that exceeds
the downhole motors capacity may cause bit
rotation to abruptly halt and the motor to stall.
Frequent stalling can damage the stator component of the motor, depending on the amount of
the weight transferred. The driller must operate
the motor within a narrow load range to maintain
an acceptable ROP without stalling.4
At the drillers console, an impending stall
might be indicated by an increase in WOB but
with no corresponding upsurge in downhole pressure to signal that an increase in downhole WOB
has actually occurred. At some point, the WOB
indicator will show an abrupt decrease, indicating a sudden transfer of force from the drillstring
to the bit.5
Increases in drag impede a drillers ability
to remove torque downhole, making it more difficult to set and maintain toolface orientation.6
Toolface orientation is affected by torque and
WOB. When weight is applied to the bit, torque at

the bit increases. Torque is transmitted downhole
through the drillstring, which turns to the right,
in a clockwise direction. As weight is applied to
the bit, reactive torque, acting in the opposite
direction, also develops. This left-hand torque
is transferred upward from the bit to the lower
part of the drillstring. Reactive torque builds as
weight is increased, reaching its maximum value
when the motor stalls. This reactive torque also
affects the orientation of the motor.
Reactive torque must be taken into account
as the driller tries to orient the drilling motor
from the surface. In practice, the driller can
make minor shifts in toolface orientation by
changing downhole WOB, which alters the reactive torque. To produce larger changes, the driller
can lift the bit off-bottom and reorient the toolface. Even after the specified toolface orientation is achieved, maintaining that orientation
can be challenging. Longitudinal drag increases
with lateral reach, and weight transfer to the
bit becomes more erratic along the length of the
horizontal section, thus allowing reactive torque
to build and consequently change the toolface
angle.7 The effort and time spent on orienting the
toolface can adversely impact productive time on
the rig.
Topdrive
Maximum
surface-applied
torque
Rotating
only
S li d
in g
s
am
plu
Dyn
to
ic
fri
ti o
no
pd
Maximum rocking depth

ri v
f ro
et
tat
o rq
Point of interference
Mud motor
ue
ion
Sliding only
Sliding plus reactive torque
Static friction
Dynamic friction
Minimum surface-applied torque
Figure 4. Torque versus friction. Longitudinal drag along the drillstring

can be reduced from the surface down to a maximum rocking depth, at
which friction and imposed torque are in balance. By manipulating the
surface torque oscillations, this point can be moved deep enough to
produce a significant reduction in drag. Similarly, reactive torque from
the bit creates vibrations that propagate back uphole, breaking friction
52
Bit
Minimum reactive torque
Maximum reactive torque
and longitudinal drag across the bottom section of the drillstring up to a

point of interference, where the torque is balanced by static friction. An
intermediate zone remains unaffected by surface rocking torque or by
reactive torque. By continuously monitoring torque, WOB and ROP while
sliding, the Slider system helps minimize the length of this intermediate zone
and thus reduces longitudinal drag.
Oilfield Review
2,000
Depth of twist, ft
4,000
6,000
No
8,000
fric
tion
red
uct
ion
React
ive to
10,000
rque
12,000
0
1,000
2,000
3,000
4,000
Surface applied torque, ft-lbf
5,000
6,000
7,000
Figure 5. Torque versus depth plot. Surface-applied torque will tend to twist the drillstring to a certain
depth depending on the drag encountered over the length of the pipe and on pipe thickness and
weight. In this model, 2,000 ft-lbf of torque applied at the surface will cause the pipe to twist to a depth
of 6,400 ft. (Adapted from Maidla et al, reference 5.)
The Slider System

Manually correcting and maintaining toolface
orientation can be a difficult process. Drilling
efficiency is largely dependent on the drillers
ability to:
transfer weight to the bit without stalling the
mud motor
reduce longitudinal drag sufficiently to achieve
and maintain a desired toolface angle
attain acceptable ROP.
The Slider automated surface rotation control
system was developed to help operators regain
some of the drilling performance of a conventionally rotating drillstring. The Slider interface
interacts with the topdrive control system to
rotate the drillstring back and forth. This torque
rocking technique reduces longitudinal drag
along part of the drillstring while slide drilling.
Rocking back and forth subjects the upper drillstring to near-constant tangential motion, producing a dynamic friction coefficient, which is
lower than a static friction coefficient created by
nonrotating pipe. Rocking can also help reduce

axial friction along the drillstring. However, this
motion is not necessarily transmitted all the way
to the bitother processes are at work.
Torque from the topdrive rotates the drillstring from the surface down to a maximum
rocking depth, where friction against the side of
the hole prevents the pipe from turning. At the
same time, as the mud motor turns the bit, it generates a reactive torque in the opposite direction.
This torque is transmitted a short distance up
the drillstring until it is overcome by friction at
some point between the bottom of the wellbore
and the BHA, referred to as the point of interference (Figure 4). Throughout the interval between
the bit and the point of interference, the velocity
component of reactive torque imposes a reduction
in longitudinal drag along the lower part of the
drillstring and possibly a change in toolface orientation. Between the depth where surface torque
is overcome by friction and the point where reactive torque is overcome by friction, the pipe does
not rotate. This section of drillstring, which has

no tangential motion, moves by sliding only and is
subject to static friction, which is greater than the
dynamic friction of pipe inmotion.
The location of the point of interference
varies with changes in the amount of reactive
torque. To efficiently minimize the sliding interval between the depth of rocking and the point of
interference while keeping the maximum rocking
depth relatively constant, an automated control
system must be used.
The amount of surface torque supplied by the
topdrive dictates in large part how far downhole
the rocking motion will be transmitted. This relationship between torque and rocking depth can
be modeled using conventional torque and drag
programs (Figure 5). However, these programs
are not needed when using the Slider system.
Using inputs from surface hook load and standpipe pressure as well as downhole MWD toolface
angle, the Slider system automatically determines the amount of surface torque needed to
2. The hook load includes more than simply the weight of

the drillstring in air; it is the total force pulling down on
the hook as it hangs beneath the derrick traveling block.
This total force includes the weight of the drillstring, drill
collars and ancillary equipment reduced by any forces
that tend to decrease the weight. These forces might
include friction along the wellbore wall and buoyant
forces on the drillstring caused by its immersion in
drilling fluid.
3. Maidla and Haci, reference 1.

4. Tello Kragjcek RH, Al-Dossary A, Kotb W and Al Gamal A:
Automated Technology Improved the Efficiency of
Directional Drilling in Extended Reach Wells in Saudi
Arabia, paper SPE 149108, presented at the SPE/DGS
Saudi Arabia Section Technical Symposium and
Exhibition, Al-Khobar, Saudi Arabia, May 1518, 2011.
5. Maidla E, Haci M, Jones S, Cluchey M, Alexander M and

Warren T: Field Proof of the New Sliding Technology for
Directional Drilling, paper SPE/IADC 92558, presented
at the SPE/IADC Drilling Conference, Amsterdam,
February2325, 2005.
May 2016
53
Weight on Bit
1,000 lbf
Differential Pressure
psi
Rotary Torque
lbf
100 0
Rotary Speed
rpm
800 0
Hook Load
1,000 lbm
20,000
150 0
Toolface
degree
360
300 0
Rate of Penetration
ft/h
50
Rotating Mode
Manual Mode
Slider Mode
Figure 6. Comparison of rotating and sliding drilling parameters. Rate of penetration (ROP) and toolface
control depend largely on the drillers ability to transfer weight to the bit and counter the effects of
torque and drag between rotating and sliding modes. The best ROP is achieved while rotating (top);
however, toolface varies drastically, as there is no attempt to control it (Track3). Hook load (Track2)
and weight on bit (WOB) remain fairly constant while differential pressure (Track1) shows a slight
increase as depth increases. To begin manual sliding (middle), the driller pulls off-bottom to release
trapped torque; during this time, WOB (Track1) decreases while hook load (Track2) increases. As
drilling proceeds, inconsistencies in differential pressurethe difference between pressures when
the bit is on-bottom versus off-bottomindicate poor transfer of weight to the bit (Track1). Spikes
of rotary torque indicate the directional drillers efforts to orient and maintain toolface orientation
(Track2). Toolface control is poor because of trouble transferring weight to bit, which is also
reflected by poor ROP (Track3). Using the automated Slider system (bottom), the directional driller
quickly gained toolface orientation. When the WOB increased, differential pressure was consistent,
demonstrating good weight transfer (Track1). Weight on bit during a Slider operation is lower than
during a manual sliding operation. Left-right oscillation of the drillpipe is constant through the slide
(Track2). Average ROP is substantially higher than that attained during the manual slide, and toolface
orientation is more consistent (Track3).
54
transfer weight downhole to the bit, thus eliminating the need to come off-bottom to make
toolface corrections (Figure 6). This results in an
efficient drilling operation and reduced wear on
downhole equipment.
System Hardware
Slider system hardware consists of a compact
package that houses the circuitry and sensors
needed to interact with the rigs topdrive control system. An interface plug is installed on the
control panel for the topdrive, and the system
is mounted at the drillers console. Installation
typically takes less than two hours with no
interruption to the drilling process. The Slider
systems connections require no alterations to
the drilling contractors topdrive mechanism or
modifications to the drilling rig. The system is
entirely surface mounted and has no downhole
equipment that might become lost in the hole. To
ensure operational safety, the system is designed
to allow manual intervention at any time.
The directional drillers interface consists of
a ruggedized notebook computer with a display
configured to enable the driller to command
the Slider system while monitoring surface and
downhole parameters (Figure 7). The Slider
system takes input such as MWD toolface angle,
surface torque and standpipe pressure from measurements already available at the rig. The MWD
toolface measurement is used to determine the
amount of correction needed to restore the toolface to the angle needed to drill the prescribed
trajectory. Surface standpipe pressure provides
an indicator of reactive torque. The Slider
software processes these inputs to determine
whether additional torque should be applied
to the drillstring to maintain the toolface angle
andROP.
To begin slide drilling, the driller can activate the Slider system and initiate the automatic
rocking action, which alternately applies torque
to the right and the left. The transfer of weight is
controlled by varying surface torque to compensate for changes in reactive torque. Corrections
in toolface angle are achieved through additional torque pulses during the rocking cycles.8
For every torque cycle to the left or right, a corresponding differential pressure peak occurs,
indicating that the weight is being transferred
to the bit. To adjust the toolface orientation, the
driller can control the magnitude and frequency
of torque pulses during a rocking cycle.
Oilfield Review
Figure 7. The Slider system graphical display. Downhole performance

parameters are monitored and controlled via a notebook computer
interface between the topdrive and Slider system. The directional driller
can configure this display to show various key parameters such as toolface
(dial, center) and torque and differential pressure (chart, bottom right ).
Torque curves show higher values for righ-hand torque (yellow) than for
Field Experience
Slider technology has been instrumental in
developing unconventional plays throughout
North America. Wattenberg field, one of the more
prolific fields in the Denver-Julesburg basin, is
located in Weld County, Colorado. There, a leading operator in the area used the Slider system to
drill horizontal wells in the Cretaceous Niobrara
gas play (Figure 8).
One of those wells, spudded in February
2016, was drilled vertically to its kickoff point,
then drilled in a westerly direction to its land8. Slider: New Level of Efficiency to Directional Drilling,
Drilling Contractor 11, no. 4 (JulyAugust 2004): 2831.
May 2016
left-hand torque (orange). Up-and-down keys allow the driller to set values
for left and right torque (upper left ). Brief torque increases above set values
can be added for one oscillation cycle by bumping left or right (middle left ).
The driller can immediately override the system by hitting the disable button
(upper right ).
Wyoming
Wattenberg
field
Colorado
Denver-Julesburg
basin
Nebraska
USA
Kansas
Figure 8. Wattenberg field. The prolific Wattenberg field lies in north-central Colorado, USA, within the
Denver-Julesburg basin.
55
Toolface
degree
Hook Load
1,000 lbm
Depth, Time,
ft
h:min 0
Weight on Bit
1,000 lbm
360 0
Rotary Torque
lbf
300 0
Rotary Speed
rpm
100 0
Rate of Penetration
ft/h
34.39 0
Pump 1
strokes/min
200
80 0
Standpipe Pressure
psi
12,000
300 0
Differential Pressure
psi
1,000
07:12
07:14
07:16
07:18
07:20
07:22
X,X25
07:24
07:26
07:28
07:30
07:32
X,X35
07:34
07:36
07:38
07:40
07:42
07:44
X,X47
07:46
07:48
07:50
Figure 9. ROP improvement. A drilling parameter display indicates that from X,X25ft to X,X35ft, the directional driller
was manually controlling the slide drilling operations. After taking 16 min to drill that interval (averaging 38ft/h ROP),
the driller activated the Slider automated system. Drilling from X,X35 to X,X47ft in 14min (averaging 51ft/h ROP), the
driller achieved a 34% improvement in ROP.
ing point in the Niobrara pay zone. Beginning

at X,X25ft, the directional driller manually controlled the slide drilling while averaging 38ft/h
[11.6m/h] ROP. Upon engaging the Slider automated system, the driller reported an average
ROP of 51ft/h [15.5m/h], for an improvement
of 34% in ROP compared with that of manually
controlled sliding (Figure9). The Slider system
was engaged several times during the course of
drilling this well. Each time the trajectory began
to drift beyond specified tolerances, the directional driller switched from rotating to sliding
modes to bring the wellbore back on course.
Comparisons between manually controlled sliding and automated sliding with the Slider system
consistently showed significant gains in ROP over
the manualapproach.
56
Faster and Farther

By sensing the amount of surface torque required
to transfer weight to the bit and by eliminating
the need to pull off-bottom to make toolface corrections, the Slider automated surface rotation
control system enables substantial increases
in ROP and lateral reach for directional wells.
Rocking, or oscillating, the drillstring back and
forth helps the driller overcome friction and thus
reduce drag on the drillstring. Along with reducing drag, operators can decrease the amount of
mud additives normally used for lubrication. The
Slider automated system typically applies less
WOB to maintain toolface control and has markedly fewer motor stalls than are experienced
while manually slide drilling. By achieving consistent toolface control, this automated torque
rocking system facilitates a longer horizontal section, which has less tortuosity, ultimately leading
to increased production.
MV
Contributor
Steven Duplantis is an Operations Manager for
Schlumberger in Houston, where he oversees day-to-day
operations, sales and development for the Slider automated surface rotation control system. Steven began
his oilfield career in 1994 and has held positions with
MD Totco, Noble Engineering and Epoch/Optidrill.
Since 2006, he has been focused on the Slider system,
working as a field coordinator and engineer involved
with Slider system testing and field trials along with
implementation and ongoing development of advanced
features for the Slider product line.
Oilfield Review
LOOKING BACK
Birth of La Pros: The 90th Anniversary

Philippe Theys
Houston, Texas, USA
Though difficult to fathom, large companies that

have tens or hundreds of thousands of employees
were once startups employing a dozen or fewer
people. Some statistics suggest that, today, nine
out of ten startups fail within the first three
years. Investigating the critical elements that
help a young organization survive those early
yearsin this case, Schlumbergeris compelling. Studying a companys beginnings can be
essential for understanding the corporate image
and characteristics that eventually emerge.
Many successful global companies that have
familiar names started small. The inventor of
xerography, Chester Carlson, conducted early
experiments in his apartment kitchen, where
occasional explosions or, perhaps, malodorous
results occurred because his methods involved
hydrogen sulfide. The Paul Allen and Bill Gates
partnership, which led to the creation of Traf-OData and then of Microsoft, is, perhaps the ultimate example of humble beginnings to megacorporation stories.
The Schlumberger beginnings, forged in the
crucible of a global search for hydrocarbons, may
be considered, in many ways, more adventurous
andexciting.
Between the early experiments by Conrad Schlumbergerthe electrical resistivity mapping of the lawn of the family property at Val-Richer in
Normandy, France, in the summer of 1912and the official registration of
Socit de Prospection lectriqueor SPE, and soon nicknamed La Pros
in July 1926, a long gestation period occurred.
The primary reason for the delay in the companys full formation was
World War I. Engineers, scientists and inventors in Europe did not have the
freedom to choose not to spend time in the army of fighting in the war. The
Schlumberger brothers were no exception. Conrad Schlumberger joined the
French artillery as an officer and his younger brother Marcel joined the
cavalry. Consequently, the Schlumberger startup history was put on hold
in1914.
Conrad was intimately involved with and horrified by the war. He began to
question his future and considered giving up his work to become a pacifist.
While in the trenches at Verdun, he said of the family textile business: Im
ashamed of belonging to a family of textile manufacturers whose employees
got up at dawn, walked for miles to reach the factory, and worked twelve or
fourteen hours a day. Not one of those workers had a life fit tolive.1
Microsoft is a registered mark of the Microsoft Corporation.
But Paul Schlumberger, Conrad and Marcels visionary father, who had
another four childrenJean, Daniel, Maurice and Paulinewith his wife
Marguerite de Witt, redirected Conrad from factory life and tied him to his
brothers future.
Paul wrote, The scientific interest in research must take precedence over
financial interests. Marcel will bring to Conrad his remarkable competence
as an engineer and his common sense. Conrad will be the wise physicist. I
will support them. He offered a sum of 500,000 francs for theendeavor.2
In 1919, 500,000 francs represented 160 kg [5,100 troy oz] of gold. Unfortunately, the franc value was cut by a factor of five from 1914 to 1926, and LaPros,
which had engineers on four continents, soon had to learn how to manage
currency variations. The business invested in trucks and equipment and sent
people to other countries, some of which were not hospitable. These early
processes meant the company had essential cash flow needs.
The survival of the young company was based on three components: seed
money, innovation and sound management of intellectual property and
people. Success came about because the company built an invincible team
and instilled exceptional values.
1. Schlumberger AG: The Schlumberger Adventure: Two Brothers Who Pioneered in
Petroleum Technology. New York City: Arco Publishing, Inc., 1982.
2. Schlumberger, reference 1.
Xerox is a trademark of the Xerox Corporation.
May 2016
57
LOOKING BACK (continued)
Figure 1. Patents filed in France, Brazil, Mexico and Australia. The first
patent was filed in France on September 27, 1912 (top left); the illustration
(top right) is from the first patent. Conrad Schlumberger filed patents in,
among other countries, Mexico and Brazil (middle) although SPE did not
58
work in these countries until 1936 and 1945, respectively. The patent at the
bottom was filed in Australia. (Documents courtesy of the Schlumberger
museum in Crevecoeur, France, and Collections cole Polytechnique,
Palaiseau, France.)
Oilfield Review
Figure 2. Socit de Prospection lectrique founders and early employees. From left to right, Paul, Conrad and Marcel Schlumberger; Henri-Georges Doll;
and Eugene Leonardon, the first engineer hired by the brothers. Other early employees include Jacques Gallois and Roger Jost. (Photographs courtesy of
the Schlumberger museum in Crevecoeur, France, and Collections cole Polytechnique, Palaiseau, France.)
Innovation and Intellectual Property

Innovation is the backbone of startups, but new techniques and technologies often fall in the public domain and quickly become available to competitors and to clients who may take advantage of this availability without
rewarding the original inventor. Chester Carlson, for example, working in
the 1930s and 1940s, wisely registered every development of his research for
the company that would eventually become Xerox Corporation.
Conrad Schlumberger filed his first patent, Procd pour la dtermination du sous-sol au moyen de llectricit (Subsurface characterization by
electrical methods) on September 27, 1912, in France. He then took care to
file patents in what seemed like unusual locations, including Mexico,
Czechoslovakia, Brazil and the Belgian Congo. By 1926, he had filed patents
in 18 countries, which would become the core of the intellectual property of
SPE (Figure 1).
The Team
Reinforced by their fathers monetary donation, Conrad and Marcel constituted an unbreakable partnership. Question the one without the other and
the answer would be: Ill talk about it with my brother Marcel . [and] Ill
ask my brother Conrad. 3
Eugne Lonardon was the first full-time engineer hired by the brothers.
He began working with Conrad in 1913 but then served in the French artillery during World War I; he was rehired in 1919. Additional field hands were
cautiously recruited. The ads in Le Journal des Mines called for ingnieurs
sportifs, athletic engineers. After Lonardon came a series of intrepid engineers (Figure 2). Jules Carr had worked with Conrad as early as 1913.
Gilbert Deschtre and Raymond Sauvage participated in the first survey for
Shell in 1928. Jacques Gallois and Pierre Baron constituted the first prospecting crew in Freeport, Texas, USA. Felicien Mailly joined in 1925 to manage the electrical research division.
Not all of these new recruits were graduate engineers. Roger Jost was
the nephew of the rgisseur, or stage manager, for events at Val-Richer. The
brothers also looked for help within the family. Nephew Marc Schlumberger
was sent to the US to scout for business. Henri-Georges Doll, who was
Paulines nephew, married Conrads daughter Annette and joined the company in January 1926 as a full-time employee. Not all of the new hires were
French. Sherwin F. Kelly, an American who graduated in France, started in
1921. Swiss geologist Edouard Poldini joined in 1922. By 1926, the company
had a staff of 17 field engineers working in Romania, Serbia, Canada, South
Africa and the Belgian Congo. Because of the distances and lack of communication, these people were given high levels of independence, initiative,
3. Schlumberger, reference 1.
May 2016
responsibility and trust that are difficult to imagine in our era of instant
texting and emailing.
All the right components were finally available; Socit de Prospection
lectrique was thus incorporated in Paris on July 1, 1926, with 200,000
francs in capital divided in 2,000 shares. SPE was now the organization
overseeing all operations, personnel and premises that were previously held
loosely by the brothers. Conrad was president. Maurice and Albert Doll, who
was Paulines husband and Henri-Georges uncle, were scrutateurs, or treasurers. Jean, cofounder of La Nouvelle Revue Franaise or NRF, and
Mauricecofounder of the bank Neuflize Schlumberger Malletwere
named minority shareholders. The headquarters, at 30 rue Fabert in Paris,
consisted of a small, five-room apartment close to Place des Invalides.
In September 1926, Conrad left France for the US to further develop that
market. In October of that year, Paul Schlumberger passed away on one of
those drizzling autumn days well known to those living in Normandy. Conrad,
Marcel and Henri-Georgeswho contributed to the success of the mission
Paul had dreamed ofjoined the family patriarch in the cemetery of the village of Saint-Ouen-le-Pin, France, in 1936, 1953 and 1991, respectively.
Of the many Schlumberger companies, SPE is the oldest. In the US, the
Schlumberger Electrical Prospecting Method, defunct as the result of the
Great Depression, was replaced by Schlumberger Well Services Corporation in
1934 and by Schlumberger Technology Corporation in 1984. In 2016, SPE still
exists and recruits young engineers to send to the far corners of the globe.
Schlumberger currently operates in more than 85 countries and has
about 100,000 employees. Wherever the drill bit is turning in the search
for hydrocarbons, the Schlumberger name can be found. The startup company formed by a few industrious and adventurous engineers has become
one of the most recognized names in the oil and gas industry. Could Paul,
Conrad or Marcel have ever imagined the enterprise that today bears their
name? Schlumberger stresses its guiding values of people, technology and
profits. Although they differ from the original principles, they echo the
seed money, innovation and team concepts that formed the company
that would become Schlumberger; together, these ideas continue to guide
the company into the future.
Philippe Theys was hired by Socit de Prospection lectrique in 1972, 46 years

after its creation. Much like the first Schlumberger employees who worked in many
parts of the world, he worked in the oil field in France, Sweden, Germany, Austria,
Western and Eastern Australia and Taiwan as well as in Louisiana and Texas in the
US before a 26-year career in marketing, research and engineering and quality. He
graduated from cole Centrale Paris in 1971, 64 years after Marcel Schlumberger,
alumnus of the same cole dingnieurs. Philippe retired from Schlumberger in 2004
and now consults for oil and gas companies.
59
THE DEFINING SERIES
Typical subsea field layout. Subsea trees positioned atop four wells contain pressure-control
valves and chemical injection ports. A jumper carries produced fluids from each tree to a subsea
manifold, which commingles production from the wells before sending it through flowline jumpers
to a subsea boosting pump station. The pump provides energy to send the produced fluids through
two pipeline end terminations (PLETs) and then through flowlines and up the risers to the
production deck of the floating production, storage, and offloading vessel (FPSO). An integrated
umbilical (green) from the FPSO supplies electric and hydraulic power for subsea tree or manifold
control functions along with chemicals to suppress the formation of scale and hydrates in the
production stream. On the seabed, the umbilical termination assembly (UTA) routes the chemical
and hydraulic fluids (white jumpers) to the manifold, which sends them to each tree. The UTA also
sends electric power to a distribution system which routes power lines (black leads) to the
manifold, boosting pump and trees.
FPSO
Production risers
Subsea tree
Manifold
Umbilical riser
Boosting pump station

PLET
Jumper
Electric power leads
Electric power
distribution unit
Injection fluid jumper

UTA
Hydraulic fluid jumper
Subsea Infrastructure
Matt Varhaug
Senior Editor
To replace dwindling reserves from onshore fields or from offshore wells
drilled in shallow waters, many E&P companies are turning to their deepwater prospects. Exploration or production in deep and ultradeep waters is
carried out at water depths of 300 to 3,050 m [1,000 to 10,000 ft] or greater.
These depths dictate that most wells be completed subsea, with wellheads,
pressure-control equipment and production equipment placed at the seafloor.
From deepwater and ultradeepwater completions, produced fluids are
sent to a processing facility by way of a subsea production system. A subsea production system consists of the subsea infrastructure used to produce oil and gas from offshore reservoirs. It encompasses one or more
subsea wells and the subsystems necessary to deliver hydrocarbons to a
fixed, floating, subsea or onshore processing facility. These subsystems can
be divided into subsea trees, production controls, manifolds, jumpers, flowlines, risers, umbilicals and processing components. Injection of water or
gas back into subsea wells is also a function of the subsea production system. Generally, oil, gas and water produced from the reservoir will flow
from a wellbore to a subsea tree and through a jumper to a manifold and
subsea flowline. Today, many operators route the flowline to a booster
pump to energize the flow as it travels between the seafloor and a riser that
carries it to the surface for processing.
60
Seabed Equipment
The subsea wellhead, installed at the beginning of the drilling phase, provides the structural foundation for the well. The wellhead is also where the
subsea tree is mounted. In some configurations, the tree contains the production tubing hanger and accommodates hydraulic and electrical lines
used for managing downhole safety valves, completion valves and pressure
or temperature sensors. The function of the subsea tree is to control and
manage pressure and flow over the life of the well and enable any necessary
intervention. The tree is the primary mechanism for shutting in the well at
the seabed and serves as the interface for well reentry operations. A subsea
control module (SCM) attached to the tree contains the instrumentation,
electronics and hydraulics connections needed for safe operation of the
subsea tree valves, chokes and downhole valves.
Sections of pipe, known as jumpers, run between subsea structures to
serve as links through which fluids are transmitted. These pipe segments
range in length from a few meters to hundreds of meters. A jumper is often
installed to carry production downstream from the tree. The produced fluid
may be routed through a multiphase flowmeter to measure production
rates and volumes.
Where multiple wells produce in a subsea development, flowline jumpers
from individual wells send produced fluids to a subsea production manifold
(Figure 1). By routing produced fluids from multiple wells to the production
manifold, the operator can reduce the number of flowlines that must be
accommodated at the next step in the production chain. Upon reaching the
Oilfield Review
manifold, produced fluids from the various wells are commingled before they
are directed to a flowline that leads to the production platform. Injection
manifolds function similarly and are used to manage the distribution of
injected water, gas and chemicals to one or more subsea wells.
Some subsea installations have a pipeline end manifold (PLEM), which
connects a flowline with another subsea structure or joins a main pipeline
with a branch pipeline. The PLEM can incorporate tie-in points with other
components such as isolation valves, diverter valves and sensor arrays.
Some PLEM designs incorporate facilities for launching pipeline pigs
devices used to clean or monitor the inside of a pipeline.
When a reservoir does not have sufficient energy to produce the fluids
from one subsea component to the next, a subsea boosting pump may be
installed. Boosting pumps function as a seafloor artificial lift system, increasing both flow rate and recovery by reducing backpressure on the reservoir.
Other recent advances in subsea processing are used to enhance field
economics. Seafloor separation and reinjection of produced water can alleviate constrained topside water handling capacity while supplementing reservoir energy through waterdrive. Subsea gas compression, including wet
gas compression, can improve viability of certain marginal developments.
Figure 1. Subsea manifold. This manifold, hoisted in preparation for installation

on the seafloor, will take produced fluids from several wells and route them to
a flowline running to a production platform.
Flow to the Surface

Flowlines tie one or more fields back to a production facilitya shorebased processing facility or fixed production platform in shallower waters
but in deeper waters, a semisubmersible, spar and floating production,
storage, and offloading vessel (FPSO) is used. The flowlines do not necessarily trace a straight course from wellhead to platform but may bend to avoid
obstacles such as existing subsea infrastructure or natural obstructions
such as underwater seamounts or canyons. As it follows the topography of
the seafloor, the flowline climbs gradually from the colder, deeper reaches
of the field upward through relatively warmer, shallower waters before
reaching the production facility.
Water depth affects temperature, which can adversely impact flow
between the subsea tree and the production facility. Upon exiting the wellhead, warm produced fluids may encounter deepwater temperatures
approaching 2C [36F] at the seafloor. Heat transfer between the produced
fluid in the pipeline and the surrounding seawater can cool the fluid to the
May 2016
High-voltage cable
Electric cable
Chemical injection
fluid line
Fiber-optic cable
Outer plastic sheath
Hydraulic fluid line
Wire cable armor
Figure 2. Cross section of an umbilical. Umbilicals supply electric power,

hydraulic fluid, chemicals and fiber-optic communications to the subsea
production system. Separate hoses, cables, injection lines and other
conduits are bundled together and enclosed within an armored outer ring
designed to withstand harsh subsea conditions.
point that gas hydrates start to form. The change in fluid temperature
beyond the tree influences the operators thermal management strategy. At
some fields, chemicals such as methanol [CH3OH] or monoethylene glycol
[C2H6O2] are injected into the system to keep the wellstream flowing then
recovered on the surface and reused. Some operators use electrically heated
flowlines; others use foam-insulated pipe. Some operators bury the flowline
beneath the seafloor for insulation, but flowlines at certain fields require no
additional heat or insulation at all. The chemistry and rheology of the produced fluids ultimately dictate which methodology is adopted.
Production flowlines run from the manifold to structures that are linked
to risers that direct the flow to the production facility. Risers transport produced fluids from the seafloor to the surface production facility. Like flowlines, many risers are insulated against cold seawater temperatures. They
offer a measure of flexibility to withstand subsurface water currents or
movement of the floating facility.
Surface Lifeline
The surface processing facility provides power, control, communication and
chemical injection services back to the subsea production system. These
services are transmitted through a subsea distribution system using umbilicals. Multiple steel and thermoplastic conduits are often bundled together
with hydraulic lines, chemical injection lines, power conductors and fiberoptic cables to form a single integrated umbilical (Figure 2). These flexible
conduits require sophisticated materials and manufacturing techniques to
withstand deep-ocean currents, pressures and temperatures. Power conductors provide electricity for subsea equipment and system sensors.
Hydraulic lines are used to open and close subsea valves. Fiber-optic lines
instantly relay sensor information and control commands between the seafloor and the surface. Some umbilical lines pump chemicals into the production stream. Umbilicals directly or indirectly service nearly every component
in the subsea production system and are critical to operating the field. The
lines typically run from the surface processing facility to an umbilical termination assembly (UTA) on the seafloor, from which services are distributed throughout the field.
Upon reaching the surface, the produced fluids are separated and
treated by the processing facility. From there, an export pipeline transports
the product to a storage and offloading installation or to an onshore refinery
for further processing and distribution.
61
THE DEFINING SERIES
Geophysics
Richard Coates
Research Manager and Scientific Advisor
Geophysics is the study of the physics of the Earth, the propagation of elastic waves within it as well as its electrical, gravitational and magnetic fields.
Although the origins of geophysics can be traced to ancient times, it was not
until the early 20th century that scientists began applying geophysical concepts and techniques to the search for hydrocarbons and minerals and to
evaluate geothermal energy resources. Now, geophysics plays a critical role
in the petroleum industry because geophysical data are used by exploration
and development personnel to make predictions about the presence, nature
and size of subsurface hydrocarbon accumulations.
Unifying Characteristics
The challenges usually encountered in geophysics are posed in the form of
an inverse problem, such that a set of measurements and known physical
laws will permit a geophysicist to determine the Earths structure and characteristics that are consistent with those measurements. In geophysics, the
answers are almost always nonunique, meaning there is more than one possible solution that satisfies the measurements. Geophysicists attempt to
resolve this ambiguity by integrating complementary data acquired from
dissimilar methods or by adding supplemental knowledge such as wellbore
measurements to determine which solution is correct.
In addition to nonuniqueness, all geophysical methods exhibit a
decrease in the resolving power with distance from the measuring equipment. This concept is analogous to the difficulty of distinguishing objects
by sight at increasing distances. The characteristic is more pronounced
for some measurement methods than for others, but the result is that the
deeper the subterranean structures, the less precise are the images of
such structures.
Seismic vessel
Airgun array
Sea surface
Streamer with hydrophone sensor arrays, 6 to 12 m deep
Seabed
Sedimentary layers
Figure 1. Marine seismic acquisition. An airgun array (top) produces pulses

of seismic energy (green) that penetrate the subsurface and are reflected
back (red) from the seabed and interfaces between rock layers. These
reflections are detected by hydrophone arrays towed behind the seismic
vessel. Geophysicists invert the recorded data to construct a 3D image of
the subterranean layers. Modern seismic acquisition methods (bottom)
illuminate 3D swaths of the Earth from a variety of angles. In one common
geometry, four ships cruise in line abreast approximately 1.2 km [0.75 mi]
apart. Each ship tows an airgun source array (red rectangles) a short
distance behind. The outermost ships also tow streamers (black lines)
typically 10 km [6 mi] in length, which record the reflections from below the
seabed and within a rock volume (tan) beneath and between the two sets
of streamers.
Seismic Surveying
In the oil field, the dominant geophysical data acquisition method is the
seismic survey, whose history dates from the early 1920s. Seismic surveying
employs a sourcetypically an airgun or a vibrating truckto generate
vibrations, or seismic waves, that propagate into the Earth. The seismic
waves are refracted and reflected by subterranean strata and structures
(Figure 1). Some of the energy returns to the surface, where it is recorded
by sensors such as hydrophones or geophones. The distances between
source and sensor can exceed 15 km [9.3 mi].
Geophysicists process the survey data to form an image and to estimate
the physical characteristics of the subsurface. This requires two steps:
develop a 3D velocity dataset, or volume, to produce a smooth estimate of
the spatially varying velocity with which the seismic waves propagate in the
Eartha process called tomographythen, with the help of this velocity
dataset, locate the subsurface layers from which the seismic waves were
reflected, a process called migration.
The resulting 3D representation of the Earth is called a structural
image, or volume. The reflecting surfaces are interpreted as the interfaces
between rock layers, some of which may have been folded, cracked, faulted
or eroded over geologic time. It can be sliced vertically to obtain a cross
section or horizontally to map the depths of the rock layers beneath the
survey area. The operator can use these interpretations to help determine
suitable drilling targets. Modern seismic surveys routinely produce detailed
3D images of these reflecting surfaces to depths of 10 km [6 mi].
Additional information about the characteristics of the rocks can be
extracted from seismic data. For example, by studying the size, or amplitude, of the reflections and how the amplitude changes with the angle at
which the seismic waves hit the reflectors, geophysicists may be able to
determine whether the pores within the rocks contain gas, oil or water. This
step, known as amplitude versus offset (AVO), often has a higher level of
uncertainty than does structural imaging.
Although most seismic work uses active sources designed to create seismic waves, the detection of weak seismic waves generated during hydraulic
fracturing is of increasing interest. These faint signals are used to determine the locations of microseismic events, which can indicate the position
and extent of the hydraulic fractures.

62
Oilfield Review
Magnetic Surveying
Magnetic surveying is another type of subsurface prospecting. Unlike EM
methods, which rely on fields that fluctuate rapidly in time, magnetic surveying depends on the permanent magnetic properties of rocks, whose strength
and orientation are fixed at the time of their deposition and may be in contrast with those of the surrounding rock. Measuring these subtle anomalies
can help geophysicists map subsurface formations over large areas.
The advantage of magnetic surveying is that data can be collected from
aircraft or satellites as well as from land or by ship. Consequently, magnetic
surveys can inexpensively cover large geographic areas as well as sites that
are otherwise difficult to access. Because the strongest anomalies are produced by volcanic or metamorphic formations, magnetic surveys are widely
used for mineral exploration.
Gravimetry Surveying
Gravity measurements have been applied in the oil field since the 1920s.
The technique is based on recording spatial variations in the Earths gravitational field, caused by differences in the density of rocks below the survey
location. The size of these variations is typically less than 1/100,000th of
Earths gravitational fields nominal value of about 9.81 m/s2 [32.2 ft/s2].
Detecting such small variations requires extremely sensitive instruments and the application of multiple corrections. For example, the
Bouguer correction accounts for variations in gravity caused by local
topography and corrects for the influence of latitude and measurement
altitude that might otherwise mask the signal. Because the low density of
salt generates a large gravity anomaly, the most common oilfield application of gravity surveying is to help delineate salt domes. Gravity data are
May 2016
Air (resistive)
CSEM transmitter
Natural-source
magnetotelluric fields
Seawater (conductive)
Electric and magnetic field recorders

Oil and gas (resistive)
Seafloor
(variable conductivity)
Salt, carbonates
and volcanics
(resistive)
Distance
Resistivity, ohm.m
10
Depth
Electromagnetic Methods
To reduce the interpretational uncertainty remaining after seismic surveying, geophysicists can choose from several techniques. The most common
are electromagnetic (EM) methods, which leverage the fact that some
important subterranean formations have strong EM signatures. For example, rocks saturated with hydrocarbons often have much higher electrical
resistivity than those containing water and is the basis of wireline resistivity
logging. Salt deposits have both a high seismic velocity and a high electrical
resistivity. Their high velocity makes seismic imaging beneath them problematic, but their high electrical resistivity makes them easy to detect using
EM surveys.
Geophysicists have two distinct methods for acquiring information
about the electrical characteristics of rocks at depth. They can use either a
high-powered EM source or fluctuations in the Earths magnetic field
induced by the solar wind as a natural EM source. In both cases, the
response of the Earth is detected via an array of receivers deployed on, or
near, the surface. The first technique is called controlled-source EM
(CSEM) and was developed in the 1980s. It is most commonly used in
marine settings, where anthropogenic noise, for example, radio signals or
power line noise, is less problematic than on land. The second EM technique, magnetotellurics (MT), was introduced in the 1950s. Some modern
systems can acquire CSEM as well as MT data when the controlled source is
not active (Figure 2).
Because of the frequency of the EM signal and the acquisition geometry,
MT surveys are best suited for basin-scale studies, while CSEM surveys are
more appropriate for detailed reservoir-scale targets and high-resistivity
anomalies. Consequently, the CSEM method is typically used to investigate
potential hydrocarbon reservoirs previously suggested by seismic images.
Joint imaging base salt

Seismic base salt
1 km
1 km
Figure 2. Marine magnetotellurics (MT) and controlled-source EM (CSEM)

acquisition. For marine MT studies (top), electric and magnetic recorders on
the seafloor make time series measurements of the Earths varying magnetic
field and the induced electric field, which can be interpreted to infer deep
geologic structures. Towing a CSEM transmitter, which has receivers behind
it, close to the seafloor allows geophysicists to map shallow structures,
including thin, resistive features such as hydrocarbon reservoirs. Joint
evaluation of multiple geophysical measurements (bottom) enables
geophysicists to obtain a consistent interpretation of the base of salt. The
best interpretation based solely on seismic data showed a thick section of
salt to the right of middle, whose base is indicated by the white line. Adding
MT resistivity data (colors) provides significant new information. Combining
seismic and MT data improves the previous interpretations of the base of salt
and gives interpreters greater confidence in their result (yellow dashed line).
frequently acquired using aircraft and satellites; taking measurements by

ship is also common.
Variations and Value
Geophysical methods are applied in various ways. For example, seismic
receivers are sometimes deployed in boreholes to generate detailed images
of small portions of the Earth. In addition, certain niche techniques, such
as the hyperspectral imaging, spontaneous potential and electrokinetic
seismoelectric and electroseismicmethods, are available but are not
widely used. Of all geophysical techniques, seismic surveying is by far the
most widespread. Because of this dominance, seismics and geophysics
are often used interchangeably in the oil industry, although for purists, this
is wrong. Nevertheless, the integrated use of complementary geophysical
methods provides critical information about the subsurface. This information is used by exploration and development personnel to make decisions
about where and how to drill.
63
LOOKING BACK
Looking Back on Wireline

On the occasion of 90 years following the July 1926 official
registration of Socit de Prospection lectrique (abbreviated
to SPE and soon nicknamed La Pros), this Looking Back article
examines the origins of the first logs and a contributed article
by Philippe Theys (Birth of La Pros, page 57) discusses those
first engineers who established an industry.
In the late 1920s, an alternative to the time-consuming process of
obtaining and analyzing rock cores was born in the wine region of
Alsace, France. This new technique became known as carottage
lectrique, or electrical coring, and eventually developed into the
wireline logging recognized today.1
Oil had been produced near the small village of Pechelbronn, France,
since the 1740s. By the 1920s, the well count was at 3,000 and increasing.
The Pechelbronn Oil Company had opened a new refinery that could
handle 80,000 metric tons, approximately 11,000 bbl, of oil annually and
needed to know that adequate reserves were available to feed the refinery.
In early 1927, company personnel discussed with Conrad Schlumberger
the idea of making resistivity measurements in the borehole to help
company geologists better understand the oil-bearing formations.
Marcel Schlumberger had already tested the technique in 1921,
taking resistivity measurements over about a meter [several feet] at
the bottom of a 760-m [2,500-ft] hole in Molires-sur-Cze in southern France. The results were inconclusive, but the feasibility of a
downhole resistivity measurement had been proved. The geophysical
community remained skeptical, however.
On September 5, 1927, Henri-Georges Doll, Conrads son-in-law, and
two colleagues, Roger Jost and Charles Scheibli, conducted the first
electrical logging operation in a 500-m [1,600-ft] Pechelbronn well
named Diefenbach 2905. The team logged an interval of 140 m [460 ft],
starting from a depth of 279 m [915 ft]. They rigged up a hand-operated winch that lowered into the hole three insulated wirescables
of the type used for lighting fixturestied together with friction tape.
The longest of the wires injected current into the well and formation;
the return was at the surface. The other two wires, shorter and of
slightly different lengths, measured the resulting potential field and
provided the resistivity readings. Measurements were made at 1-m
[3 ft] intervals; the entire operation took five hours. The result was a
resistivity log that distinguished between the many layers of sand and
shale pierced by the borehole.
Resistivity logging continued throughout 1928 at Pechelbronn,
and the resulting correlations of resistivity from one well to the next
revolutionized the understanding of the stratigraphy of the field.
Soon, the Pechelbronn Oil Company was able to raise the annual
capacity of its new refinery to 100,000 metric tons. By 1929, the new
logging technique had gone globalSchlumberger logging crews
were engaged by Shell for their explorations in Venezuela, the US
and the Dutch East Indies, and by the Soviet Union for the oil fields
of Grozny, Chechnya, and Baku, Azerbaijan.
64
Meanwhile, Conrad was playing with new ideas for his well logging. Realizing that oil was infinitely resistive to electricity, he
postulated that if a zone was oil bearing and reasonably thick, then
increasing the spacing between the wires measuring potential would
allow the logging tool to see deeper into the formation and record a
higher resistivity; the phenomenon was not observed in Pechelbronn
because the oil zones were too thin. In May 1930, he asked his field
engineers to try the idea and report back. Within a month, Marcel
Jabiol, the companys solitary engineer in northern Sumatra, sent
a telegram back confirming Conrads hypothesis. Electrical logging
could locate oil from the borehole.
Pulley used in Pechelbronn in 1928.

(Illustration courtesy of and
adapted from Mau and Edmundson,
reference 1.)
Harnessing the Potential

Combining another measurement technique with electrical logging,
however, was necessary to fully optimize the electrical logging
methodnamely, spontaneous potential (SP). Natural electrical
potentials in the subsurface had been discovered about a century
earlier in Cornwall, England, by the British geologist, natural philosopher and inventor Robert Were Fox. Potentials in the borehole are
caused by electrochemical interactions between the borehole fluid
and adjacent sand and shale formations.
Conrad Schlumberger secured a French patent on the SP technique in 1929, claiming it could be used to locate permeable strata,
but found no practical application. A year later, Doll observed
natural potentials while logging in the Oklahoma Seminole oil field.
While the battery was disconnected, he noticed the potentiometer
needle vibrating back and forth as the electrodes were lowered into
the well. Experiments followed at Pechelbronn, and by 1930, they
concluded that the SP method could differentiate permeable beds
such as sand and limestone from impermeable formations such as
shale. The combination of SP and resistivity curves turned out to be
of much greater value than the resistivity log alone in detecting production possibilities. Electrical logging established a reliable method
of achieving stratigraphic correlation, providing a way to distinguish
between shale and porous rock and between hydrocarbon- and
water-bearing rock.
In the decades that followed, electrical logging became increasingly sophisticated and paved the way for numerous other well logging techniques. It became an industry on its own, taking the key
position among formation evaluation techniques.
1. Mau M and Edmundson H: Groundbreakers: The Story of Oilfield Technology and the
People Who Made it Happen. Peterborough, England: Fast-Print Publishing, 2015.
Grow your Knowledge.

For experienced professionals, newcomers and those simply
interested in learning more about our industry, the Defining
Series provides summaries of a wide range of industry
topics, efficiently communicating basic principles and
underlying science.
Added to the series and appearing in this issue are subsea
infrastructure and geophysics. See the entire series online
to grow your knowledge.
http://www.slb.com/oilfieldreview
Oilfield Review
Oilfield Review

May2016 Oilfield Review

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

May2016 Oilfield Review

Uploaded by

Copyright:

Available Formats

Oilfield Review

Marine Seismic Imaging

Schlumberger and Cameron: A Meeting of Minds and Technologies

as a consequence, they are recognized as innovators in

The integration of Schlumberger and

is a registered trademark of Special Metals Corporation.

Viewing Complex Structures

The Defining Series

Marine Imaging in Three

About Oilfield Review

Hydraulic Fracturing Insights

Oilfield Review online

CorrosionThe Longest War

Oilfield Review app

Marine Imaging in Three Dimensions:

Single Channel Interpolation

Figure 1. Streamer element. An element of the IsoMetrix streamer system

the wavefield (blue) can be measured at each streamer location (black

engineers who develop static and dynamic

crossline direction may be aliased, or inadequately sampled.

By adding 3C accelerometers, the marine

seismic data using streamers spaced farther

hydrophone-only streamers; towing deep often

Geologic dip of surface of interest

Main Survey Area

Entire Survey Area

as high as 50 occur in the area along a WE

Estimated Survey Duration:

Bathymetry from Multibeam Echo Sounder

Bathymetry from IsoMetrix Technology

After acquisition, the data were preprocessed

The data proved to be of high quality. For

Dip shooting direction, inline stack

Strike shooting direction, crossline stack

direction, geophysicists judged the datasets to be

The WesternGeco seismic vessel Western

spatial bandwidth in all directions. In the spatial

were preprocessed and then simultaneously spatially dealiased and receiver-deghosted in 3D by

10. Gabrielsen RH, Frseth RB, Jensen LN, Kalheim JE and

12. For more on petroleum systems: Al-Hajeri MM,

International Meeting and Exhibition, Las Vegas,

fluvial and floodplain geology and reveal that

results as seismic waves travel through the Earth

Figure 10. Floodplain channels. This seismic

Formation, composed predominantly of shale. The

sandstone but are below the detection

3km [1.9mi] long, spaced 75m apart and towed

The geophysicists wanted to know whether the

15. zbek et al, reference 5.

Using Multimeasurement Towed Streamer Data: North

of the difference between the horizontal and vertical

Build initial velocity model

Control test for FWI: Insert known channel

One iteration of common image point (CIP)

Low-frequency FWI iterations using

Intermediate-frequency FWI iterations using

One iteration of CIP tomography to refine

Multiparameter FWI to refine

Final high-frequency FWI iterations to enhance

Figure 11. Workflow for full waveform inversion.

higher frequency bands until there is an

dataset acquired using IsoMetrix technology can be

Before FWI Processing

Depth slice, 158 m