Oilfield Review

Summer 2013

Sourceless Density
Routine Core Analysis
Multistage Stimulation
Hydraulic Fracture Design Software

13-OR-0003

Core Analysis: Combining Expertise for Insight into the Reservoir

Managing an asset to optimize oil or gas production
requires knowing reservoir rock and uid properties. For
example, planning well locations relies on predictions of
porosity and other rock properties from seismic surveys.
Stimulation and completion designs depend on knowledge
of geomechanical strength and permeability from logging
and core measurements. Reservoir simulation demands
data on a wide range of properties of the formation rocks
and uids on many scales to engineer eld production.
Evaluation of the rocks and the uids within them is vital
for reservoir development and management throughout the
life of a eld.
While many reservoir properties can be evaluated
remotely with seismic or logging studies, the most detailed
and accurate measurements of rocks and uids come
from laboratory evaluations of core samples. The new
Schlumberger Reservoir Laboratories organization focuses
on the interrelationships of rock and uid properties to
help operators understand oil and gas assets.
This organization comprises more than 25 laboratories
globally and employs standardized procedures and equipment to support analysis of core data. This comprehensive
integration of rock and uid analysis services helps customers reduce risk in making reservoir development decisions. Combining both services in one organization
expands the expertise at hand for execution of the experiments and interpretation of the results.
One of the most obvious areas of rock-uid interaction
is in enhanced oil recovery (EOR) studies. To forecast
improvements from miscible gas and chemical displacements, engineers require rock and uid properties at reservoir conditions. EOR laboratory core oods use live
reservoir uids under reservoir conditions.
Schlumberger has offered advanced uids expertise,
geomechanics and unconventional resource evaluation for
years. In 2012, the company added several commercial
conventional core analysis laboratories and established a
laboratory hub in Houston. These laboratories provide
routine and special core analysis capabilities, with a
special emphasis on EOR evaluations for miscible gas
and chemical ooding.
Throughout its history, Schlumberger has used its
breadth and depth of knowledge of the subsurface to
develop logging tools, logging-while-drilling techniques,
fracturing techniques and other services that require
understanding of rock-uid interactions. The new commercial core analysis capabilities build on this long history of
rock studies.
As a company, we have always answered questions about
reservoirs through petrophysical analysis. Now, we extend

this tradition to routine core analysis, answering basic

questions about a formation: Does the formation contain
uids? Are hydrocarbons present? Can they ow through
the formation? (See Core Truth in Formation Evaluation,
page 16). The answers come from measurements of porosity, saturation and permeability.
These properties are a part of any petrophysical study
of a reservoir. In addition, core measurements provide a
means to calibrate log interpretations of electrical properties and nuclear magnetic resonance responses to obtain
downhole estimates of porosity and saturations. The
petrography and sedimentology of the reservoir core can
also be evaluated in detail.
Reservoir engineering relies on dynamic ow simulations,
and core analysis is fundamental to this activity. Multiphase
propertiesrelative permeability and capillary pressure
come from measurements made in a special core analysis
laboratory. Other ow studies conducted in these laboratories are designed to evaluate EOR processes and assess formation damage caused by a variety of sources.
Core analysis is often referred to as the ground truth of
rock properties. In the laboratory, we can measure properties more precisely and accurately than through remote
sensing. But it is also necessary to understand that laboratory measurements may not be reective of eld conditions. Field conditions can be simulated to a greater or
lesser degree, but some alterations to the rock caused by
drilling and retrieval are difcult to reverse. To get the full
picture, it is necessary to integrate the information from
all sourceslaboratory and eld. With the wide range of
expertise found in Schlumberger, we can provide the necessary perspective.
The future of core analysis is bright. We are commercializing our digital core analysis effort with services
ranging from whole core to nanoscale imaging and ow
simulation. Additional innovative services will be introduced in coming years.
Mark A. Andersen
Core Physics Domain Head
Schlumberger Reservoir Laboratories
Houston, Texas, USA
Mark A. Andersen, Schlumberger Domain Head for Core Physics in Houston,
joined the company in 2000. He spent 11 years as an Oilfield Review editor
and executive editor before returning to his roots in core analysis to help build
a new business for Schlumberger. He began his career in 1981 as a researcher in rock properties at Amoco Research Center in Tulsa. He subsequently
spent several years in Stavanger, where he managed the Amoco Norway
external research program and wrote Petroleum Research in North Sea Chalk.
Mark is the author of many technical papers, including 23 articles for Oilfield
Review. He earned a BS degree in engineering physics from the University
of Oklahoma at Norman, USA, and MS and PhD degrees in physics from The
Johns Hopkins University in Baltimore, Maryland, USA.

Schlumberger

Oilfield Review
www.slb.com/oilfieldreview

Executive Editor
Lisa Stewart
Senior Editors
Tony Smithson
Matt Varhaug
Rick von Flatern

Core Analysis: Combined Expertise for Insight into the Reservoir

Editorial contributed by Mark A. Andersen, Core Physics Domain Head,

Schlumberger Reservoir Laboratories

Editor
Richard Nolen-Hoeksema
Contributing Editors
Ginger Oppenheimer
Rana Rottenberg
Design/Production
Herring Design
Mike Messinger
Illustration
Chris Lockwood
Tom McNeff
Mike Messinger
George Stewart

Formation Density from a Cloud, While Drilling

A recently introduced formation density tool uses a pulsed

neutron generator to induce gamma rays in a formation and
compute bulk density. The LWD tool that houses the new
measurement system is the rst to offer a compact logging
suite comparable to a triple-combo service, but without the
use of radioisotopic sources.

Printing
RR DonnelleyWetmore Plant
Curtis Weeks

16 Core Truth in Formation Evaluation

Oil and gas companies obtain physical samples of subsurface
formations through coring. Careful testing of these samples
allows operators to determine if the rock contains uid-lled
pores, if those pores contain hydrocarbons and if those
hydrocarbons are producible. Routine core analysis helps
operators answer these questions and more.

On the cover:
Core analysis is an essential building
block of formation evaluation. Most
E&P companies rely on the specialized
equipment and expertise of a core
analysis laboratory to evaluate their
core samples. Here, a core specialist
removes a core plug from a solvent
distillation and extraction device used
to clean the core and measure the
volume of any uids contained therein.
A computed tomography scan of a
core (inset) shows changes in density
indicative of variations in mineralogy
or porosity.

About Oilfield Review

Oilfield Review, a Schlumberger journal,
communicates technical advances in
finding and producing hydrocarbons to
customers, employees and other oilfield
professionals. Contributors to articles
include industry professionals and experts
from around the world; those listed with
only geographic location are employees
of Schlumberger or its affiliates.

Oilfield Review is published quarterly and

printed in the USA.
Visit www.slb.com/oilfieldreview for
electronic copies of articles in English,
Spanish, Chinese and Russian.

2013 Schlumberger. All rights reserved.

Reproductions without permission are
strictly prohibited.
For a comprehensive dictionary of oilfield
terms, see the Schlumberger Oilfield
Glossary at www.glossary.oilfield.slb.com.

Summer 2013
Volume 25
Number 2
ISSN 0923-1730

Advisory Panel

26 Multistage Stimulation in Liquid-Rich

Unconventional Formations

Hani Elshahawi
Shell Exploration and Production
Houston, Texas, USA

To optimize the economics of producing oil from liquid-rich

shales, service companies are rening the completion technology that has made possible the protable exploitation of
these tight formations. Operators are now able to take
advantage of new completion tools and systems, which are
designed to signicantly improve the efciency and effectiveness of stimulating low-permeability formations.

Gretchen M. Gillis
Aramco Services Company
Houston, Texas
Roland Hamp
Woodside Energy Ltd.
Perth, Australia
Dilip M. Kale
ONGC Energy Centre
Delhi, India

GG

Stage 15
Stage 14

GG

Stage 12

Good

GG

Stage 11

GG

Good

Good

Good

Stage 13

Good

GG

Good

GG

Good

Andrew Lodge
Premier Oil plc
London, England

Stage 10
Stage 9

Good

Stage 8

GG

Good
Good

Taking advantage of the combination of horizontal drilling

and hydraulic fracturing technologies, operators are able to
access ultralow-permeability reservoirs that contain oil and
gas. A systematic, engineered completion design approach
using a comprehensive workow management software
system is helping make hydrocarbon extraction from unconventional reservoirs more effective.

Good

34 Stimulation Design for Unconventional Resources

Good

George King
Apache Corporation
Houston, Texas

Stage 7

od

Stage 6

47 Contributors
49 New Books and Coming in Oilfield Review
51 Defining Hydraulic Fracturing:
Elements of Hydraulic Fracturing
This is the tenth in a series of introductory articles describing basic concepts of the E&P industry.

Editorial correspondence
Oilfield Review
5599 San Felipe
Houston, TX 77056
United States
(1) 713-513-1194
Fax: (1) 713-513-2057
E-mail: editorOilfieldReview@slb.com

Subscriptions
Customer subscriptions can be obtained
through any Schlumberger sales office.
Paid subscriptions are available from
Oilfield Review Services
Pear Tree Cottage, Kelsall Road
Ashton Hayes, Chester CH3 8BH
United Kingdom
E-mail: subscriptions@oilfieldreview.com

Distribution inquiries
Matt Varhaug
Oilfield Review
5599 San Felipe
Houston, TX 77056
United States
(1) 713-513-2634
E-mail: DistributionOR@slb.com

Formation Density from a Cloud, While Drilling

Environmental, health and security concerns have encouraged service companies

to search for alternatives to the traditional logging sources relied on for formation
density measurements. Scientists recently developed a reliable LWD measurement
that uses a pulsed neutron generator similar to those that have been deployed in
wireline logging tools for decades.

Franoise Allioli
Valentin Cretoiu
Marie-Laure Mauborgne
Clamart, France
Mike Evans
Sugar Land, Texas, USA
Roger Grifths
Petaling Jaya, Malaysia
Fabien Haranger
Christian Stoller
Princeton, New Jersey, USA
Doug Murray
Abu Dhabi, UAE
Nicole Reichel
Stavanger, Norway
Oileld Review Summer 2013: 25, no. 2.
Copyright 2013 Schlumberger.
For help in preparation of this article, thanks to Doug Aitken,
Sugar Land, Texas.
EcoScope and NeoScope are marks of Schlumberger.

Formation density logs rst appeared in the mid1950s. Henri Doll, a Schlumberger research scientist who is credited with the development of
the density measurement and many other petrophysical measurements in use today, received a
patent for the concept in 1951. The formation
density tool he helped design uses a radioisotopic
source that emits gamma rays and then counts
the gamma rays that return to the tool after passing through the formation. Recently, a new technique has been introduced that eliminates the
traditional gamma ray source in logging-whiledrilling (LWD) applications.
Density tools were originally referred to as
gamma-gamma density (GGD) devices because
gamma rays were emitted from a logging source
and then returning gamma rays that passed
through the formation were counted by the tool.1
The hardware and the electronics used in counting those returning gamma rays have undergone
evolutionary changes over the past half century,
yet the source has remained a fundamental
requirement for formation density logging.
Traditional wireline and LWD formation density tools use a cesium [137Cs] gamma ray source.2
To gain a statistically precise measurement, a
63-gigabequerel (GBq) or higher source strength
is normally used.3 Density tools are not the only
tools that use sources for petrophysical measurements. Traditional thermal neutron porosity measurements rely on americium beryllium
[241AmBe] sources to generate the neutrons used
in the measurement.

Service companies go to great lengths to minimize the risks associated with the use of sources;
these devices must be handled carefully to avoid
health, security and environmental concerns.4 In
a number of locations throughout the world, the
use of traditional source material is being discouraged or even banned. In response, service
companies have sought to develop alternatives to
tools that require sources.5 Increasingly, pulsed
neutron generators (PNGs) are replacing
241AmBe neutron sources in both LWD and wireline applications.6
PNGs produce high-energy, fast neutrons
using a charged particle accelerator. Inelastic
collisions between these fast neutrons and the
nuclei of a variety of atoms found in formation
uids and minerals can put those nuclei in an
excited state. Typically, the nuclei return to
ground state by emitting one or more gamma
rays. These gamma rays form a cloud that can act
as a distributed source in the formation. The
gamma rays undergo attenuation as they travel
through the formation. As in the case of a radioisotopic source, the attenuation of these gamma
rays depends mainly on the electron density of
the materials making up the formation.
Scientists have developed a technique that
takes advantage of the distributed gamma ray
cloud to compute formation density, although
they rst had to develop a method that accurately
modeled gamma ray transport from the formation to one or more detectors on a tool. The resultant bulk density measurement is similar to that

Oileld Review

from a GGD tool, but it comes from the neutroninduced gamma rays. The density derived from
this technique is referred to as a sourceless neutron gamma density (SNGD) measurement.7
This article presents the SNGD measurement
theory and discusses some of the advantages of a
sourceless LWD density tool. Field results validate this new technique.

As Low as Reasonably Achievable

Traditional sources used for petrophysical analysis are protected and isolated while being transported to and from drilling rigs and are stored in
shields that protect personnel from exposure.
Pressure vessels that house the radioactive elements are made from materials designed to protect sources from mechanical damage and

corrosion in the harsh wellbore environment.

While inserting a source into a logging tool, workers follow strict safety practices to eliminate
potential for exposure. When the tool is lowered
below the rig oor, the potential for human exposure goes with it. Sources must be handled carefully, but when established safety precautions are
followed, there is little risk of exposure.

1. In this article, a source refers to a radioisotopic device

used in petrophysical logging tools that emits ionizing
radiation.
2. The radioisotope 137Cs has a half-life of 31.17 years and
emits gamma rays with an average energy level of
662 keV.
3. A becquerel (Bq) is the activity of a quantity of
radioactive material in which one nucleus decays per
second. Prior to the adoption of Bq as a standard SI unit
of measurement, radioactivity was expressed in curies
(Ci), which was the radioactivity of 1 g of the radium
isotope 226Ra. 1 GBq = 0.027027 Ci.

4. Evans M, Allioli F, Cretoiu V, Haranger F, Laporte N,

Mauborgne M-L, Nicoletti L, Reichel N, Stoller C,
Tarrius M and Grifths R: Sourceless Neutron-Gamma
Density (SNGD): A Radioisotope-Free Bulk Density
Measurement: Physics, Principles, Environmental Effects,
and Applications, paper SPE 159334, presented at the
SPE Annual Technical Conference and Exhibition,
San Antonio, Texas, USA, October 810, 2012.
5. Reichel N, Evans M, Allioli F, Mauborgne M-L, Nicoletti L,
Haranger F, Laporte N, Stoller C, Cretoiu V, El Hehiawy E
and Rabrei R: Neutron-Gamma Density (NGD):
Principles, Field Test Results and Log Quality Control

of a Radioisotope-Free Bulk Density Measurement,

Transactions of the SPWLA 53rd Annual Logging
Symposium, Cartagena, Colombia, June 1620, 2012,
paper GGG.
6. For more on pulsed neutron generators: Adolph B,
Stoller C, Archer M, Codazzi D, el-Halawani T, Perciot P,
Weller G, Evans M, Grant J, Grifths R, Hartman D,
Sirkin G, Ichikawa M, Scott G, Tribe I and White D:
No More Waiting: Formation Evaluation While Drilling,
Oileld Review 17, no. 3 (Autumn 2005): 421.
7. The term sourceless indicates that this measurement
does not use radioisotopic sources.

Summer 2013

Pulsed Neutron Generator

Controls

n
Ion source

High-voltage
supply

On-off
switch
Main
power

Target

n n
p+

n n
p+ p+

Deuterium

Tritium

Helium

Neutron

2H

3H

4He

n
p+

Kinetic
energy
E (17.6 MeV)

> Pulsed neutron generator (PNG). PNGs are self-contained particle

accelerators that produce neutrons using a fusion reaction. A high
voltage potential accelerates ionized deuterium and tritium isotopes of
hydrogen toward a target doped with tritium (top). The fusion reaction
(bottom) results in the production of a 4He nucleus and a neutron. The
reaction energy is transferred into the kinetic energy of the two particles
and is converted into heat when the particles are stopped in matter. The
neutrons leave the reaction with very high speed, having kinetic energy of
approximately 14 MeV of the total 17.6 MeV released. When the main power
is disconnected, the PNG produces no neutrons.

In the early days of the nuclear age, which

coincided with the development of many of the
tools used in petrophysical analysis, radiation
safety practices focused on time, distance and
shielding: Minimize exposure time, keep maximum reasonable distance from radiation sources
and maintain barriers (shielding) between people
and material. These principles are still applied
today for working with traditional sources, and
exposure limits have been established to ensure
the safety and health of workers who routinely
handle these materials. Workers are also closely
monitored to determine exposure levels.
Observations of the long-term effects of radiation on humans resulting from surface detonation of nuclear devices, however, led scientists to
develop a new methodology for dealing with
human exposure. As low as reasonably achievable
(ALARA) has emerged as the standard for regulators. The goal of ALARA is to eliminate exposure
whenever and wherever possible, which has
driven service companies to investigate alternatives to traditional sources such as 137Cs and
241AmBe. A PNG is one example of an alternative
to traditional sources.8
A PNG is a miniature particle generator.
Deuterium [2H] and tritium [3H] are accelerated
into a tritium-doped target, and high-energy
(approximately 14 MeV) neutrons are released
(above). When not electrically energized, PNGs
do not emit external radiation. Scientists and
engineers developed the rst PNGs in the 1950s.
These devices have since been adopted for many

downhole applications, including neutron porosity tools, cased hole formation evaluation tools
and capture and inelastic spectroscopy services.
PNGs have emerged as a viable alternative to
241AmBe sources. For LWD operations, turbine
generators have been developed to supply the
downhole electrical power needed to operate
PNGs. This advance has allowed design engineers
to incorporate PNGs in services such as the
EcoScope multifunction logging-while-drilling service and the NeoScope tool.9 Attempts to replace
137Cs sources used in GGD tools used for formation
density, considered by many geoscientists to be
one of the most critical parameters for the quantitative determination of formation porosity, have
not met with similar success until recently.
Scientists have been unable to replace
137Cs-dependent measurements for a number of
reasons. For example, there is no comparable
electronic gamma ray generator, and replacing
other sources was deemed a higher priority. The
half-life of 241AmBe is 432 years, much longer
than the approximately 30-year half-life of 137Cs.
The activity of an 241AmBe source is higher and
also more difcult to shield.10 If an LWD logging
tool becomes stuck in a well, operators must
ensure that the source will remain in place,
intact and isolated for hundreds or even thousands of years. The shorter half-life of 137Cs and
its lower radiotoxicity do not remove the risk,
but, compared to 241AmBe, there is a reduced
potential for long-term consequences.11

As a way to mitigate risk associated with

sources, some operators have opted to
use PNG-based wireline and LWD neutron porosity tools exclusively rather than tools with a traditional source. Additionally, the prospect that
some countries may mandate the elimination of
traditional sources entirely is a concern to both
operators and service companies.
Another reason for the delay in replacing
density sources is that bulk density resulting
from the GGD measurement is a fairly straightforward petrophysical parameter that has been
accepted by the interpretation community for
decades. Replacing GGD tools with SNGD tools
adds a greater level of complexity and introduces some differences in measurement physics.12 As a consequence, scientists have invested
considerable time and resources in understanding the physics involved in using induced gamma
rays for density measurements. In 2005, scientists and engineers at Schlumberger introduced
the algorithms needed to compute an SNGD
measurement. They were able to demonstrate
that a sourceless density measurement that replicated traditional formation density measurements could be produced. Seven years later,
they launched the rst commercial PNG-based
LWD gamma density tool in the oil and gas
industry. This tool delivers a high-quality bulk
density measurement comparable to that of traditional GGD tools. Because the technique uses
a PNG in place of a traditional source, the tool
complies with ALARA objectives.13

241AmBe

8. For more on radioactive sources used in logging tools:

Aitken JD, Adolph R, Evans M, Wijeyesekera N,
McGowan R and Mackay D: Radiation Sources in
Drilling Tools: Comprehensive Risk Analysis in the
Design, Development and Operation of LWD Tools,
paper SPE 73896, presented at the SPE International
Conference on Health, Safety and Environment in Oil
and Gas Exploration and Production, Kuala Lumpur,
March 2022, 2002.
9. Japan Oil, Gas and Metals National Corporation
(JOGMEC), formerly Japan National Oil Corporation
(JNOC), and Schlumberger collaborated on a research
project to develop LWD technology that reduces the
need for traditional chemical sources. Designed around
the pulsed neutron generator (PNG), NeoScope and
EcoScope services use technology that resulted from
this collaboration. The PNG and the comprehensive
suite of measurements in a single collar are key
components of the NeoScope and EcoScope services
that deliver game-changing LWD technology.
10. Sources that emit gamma rays can be shielded using
lead, although lead is not an effective shield for
neutrons. Shields for neutron sources generally
contain polyethylene.
11. Aitken et al, reference 8.
12. In some regions, operators consider the anhydrite
measurement a validation of proper tool calibration. This
valuea density of 2.98 g/cm3is outside the quoted
formation density range of the SNGD measurement.
13. The PNG used in the NeoScope tool contains a small
amount1.6 Ciof tritium, a radioisotope of hydrogen.
The half-life of tritium is 12.3 years. Tritium is also used
in conjunction with phosphorous in luminous watch dials
and exit signs in buildings.

Oileld Review

PNG-Based Measurements

Other Measurements

Neutron-gamma density

Array resistivity

Azimuthal gamma ray

Neutron porosity

Dual ultrasonic caliper

Near-bit inclination

Spectroscopy

Annular pressure while

drilling

Three-axis shock and

vibration

Sigma

Temperature

Near epithermal detector

Pulsed neutron generator
Neutron flux
detector

Short-spacing gamma ray detector

Far thermal neutron detectors

Long-spacing
gamma ray detector

Near thermal
neutron detectors

> NeoScope LWD logging tool and its capabilities. Engineers designed the NeoScope tool (bottom) with several collocated petrophysical measurements
on a single 7.6-m [25-ft] collar. The table (top) summarizes the tools capabilities.

Summer 2013

the sensors are simultaneously measuring the

same formation volume under identical static
and dynamic conditions.
The NeoScope service measures neutronbased petrophysical properties, along with bulk
density. Most wireline and historical neutron
porosity data come from tools that use 241AmBe
sources; the NeoScope service provides a comparable thermal neutron measurement. Formation
hydrogen index (HI), the basis of neutron porosity computation, is also an output of the tool. The
neutron count rates in near and far helium-3
detectors are used to determine HI and thermal
neutron porosity. Compared with traditional
thermal neutron porosity, this PNG-based HI is
less sensitive to environmental conditions.

Sigmaanother output available from the

NeoScope toolis the macroscopic thermal
neutron capture cross section of the formation.
Sigma is a measurement of the formations ability to capture, or absorb, thermal neutrons, and
the measurement can provide resistivity-independent uid saturation in the presence of
saline formation water. High-energy, fast neutrons are emitted by the tool, slowed by collisions with the nuclei of elements in the
formationprimarily hydrogenand then
absorbed by receptive atoms and molecules.
After these neutrons are absorbed, capture
gamma rays are generated, which are counted
by the detectors. The rate at which thermal neu-

1.0

Plan View
0.8

Fraction of response

More Than Just Density

The scientists who developed the SNGD model
worked with engineers to include this new design
concept in the NeoScope sourceless formation
evaluation while drilling service. Six petrophysical measurements are incorporated in the
NeoScope platformSNGD, neutron porosity,
elemental capture spectroscopy, sigma, resistivity and azimuthal natural gamma rayand they
are collocated on a single, relatively short collar
(above). The NeoScope LWD tool is generally
located close to the bit, giving well placement
engineers early and precise geosteering data.
Near-bit positioning allows the tool to make measurements when drilling uid invasion is still
minimal, which further simplies data interpretation and modeling. This is especially important
for sigma measurements. The NeoScope tool also
contains sensors to measure hole size, annular
pressure and temperature, near-bit borehole
inclination and triaxial shock and vibration.
In addition to collocated measurements close
to the bit, the NeoScope tool design has other
benets; the SNGD measurement has a greater
depth of investigation (DOI) than traditional
GGD tools have and is less dependent on wellbore
wall contact for accurate measurements. Even a
small standoff for the GGD tools may result in
compromised measurements, and hole rugosity
has always been problematic for traditional density tools (right).
The SNGD measurement is collocated with
the other neutron-based measurements and
resistivity measurements. Conventional logging
strings often have separate tools for each measurement. Collocating the sensors reduces the
effects of irregular tool movement that can
cause misalignment of depth reference points.
Collocation also simplies interpretation because

Borehole

0.6

Azimuthal
density

0.4

Depth of
investigation

GGD data
SNGD data
0.2

SNGD
measurement volume

10

12

Depth into formation, in.

> Greater DOI of the SNGD measurement. Traditional GGD measurements, such as from LWD azimuthal
density tools, read only a few inches into the formation (left, red) and have a narrow measurement
aperture (right). Hole rugosity may negatively impact the quality of the measurement. Although the
SNGD (green) has a greater DOI, which results in a measurement that is less sensitive to rugosity and
standoff, it does not have an azimuthal component.

Compton Scattering
Gamma ray
Formation

Detectors

Nuclear source

Scattered
gamma ray

Incident gamma ray

e
> Compton scattering of gamma rays. For traditional density tools (left), gamma rays are emitted by
a source and then interact with the formation in three main ways. Compton scattering (right) is the
primary interaction related to bulk density measurements. Pair production and photoelectric effect
(not shown) are the other two interactions. For most well logging situations, the amount of Compton
scattering is related to the electron density of the atoms that make up the minerals and uids in the
formation. Electron density is directly related to bulk density. The formation bulk density is computed
from the number of gamma rays that make their way from the source, through the formation and back
to the detectors. Higher density results in fewer returning gamma rays compared with measurements
in lower density formations.

Electronic source
High energy
Traditional source
10 6

Neutron energy
leaving source
Neutron energy, eV

Intermediate energy
10 4

Inelastic
region
10 2

Epithermal energy

Capture
gamma ray
emitted

10 0

Average
thermal
energy
0.025 eV
102

Neutrons with thermal energy

200

400

Time, s

> Life of a neutron. Both electronic and traditional sources emit high-energy,
fast neutrons. Neutrons from the PNG electronic source used in the
NeoScope tool have an initial kinetic energy of about 14 MeV but in a few
microseconds reach thermal energy level (approximately 0.025 eV). During
those rst few microseconds, before neutron kinetic energy falls below
about 1 MeV, the neutrons experience inelastic collisions that produce
gamma rays. These are the gamma rays used for SNGD processing. After
several microseconds, the neutrons reach thermal energy level and are
eventually captured. The capturing atoms generate gamma rays to return to
ground state.

trons are captured depends on the capture cross

sectionsigmaof the element absorbing them.
The capture cross section of chlorine, which
is the strongest neutron absorber of common elements encountered in well logging, is higher than
that of oil or gas. If the porosity and formation
water salinity are known, the water saturation
can be determined from sigma. Because the measurement is acquired near the bit, it is possible to
determine sigma in the absence of mud ltrate
invasion. This establishes a reliable baseline for
comparison with future cased hole sigma logs.
An added benet of water saturation computed from sigma data occurs when logging in
high-angle wells. When high-angle and horizontal wells cross or approach bedding planes with
resistivity contrasts, the resistivity measurements often exhibit anomalous readings.
Because sigma data are not similarly affected at
bed boundaries, saturation measurements computed from sigma may be more accurate than
traditional computations that are based on
Archies equation.
Missing from the SNGD measurement is
the photoelectric factor (PEF) measurement.
Conventional density tools include this lithology
indicator for inferring the rock matrixa crucial
input for computing density porosity. Although the
PEF measurement is not available with the new
technique, the NeoScope tool provides neutron capture spectroscopy, which delivers formation elemental composition information. These data offer
petrophysicists a more reliable and accurate lithology determination than do PEF measurements.
The primary drivers for development of a
sourceless density tool have been environmental
and security concerns. In some areas of the world,
regulations prevent drillers from reentering a reservoir in which a traditional source has been left
behind in a stuck drilling assembly. Because PNGs
are inactive and cannot produce neutrons when
circulation ceases, operators are often permitted
to drill sidetrack wells very near a wellbore in
which a sourceless tool has been lost.14
The radioisotope-free nature of the NeoScope
service is also attractive in unconventional plays
because many of these are located near population
centers, where the public may be wary of traditional sources. There are no traditional sources
with the NeoScope service, completely eliminating
their transportation and handling at the wellsite.
The NeoScope service provides real-time natural
gamma ray images to steer the well, triple combo
data for petrophysical analysis and spectroscopic
lithology information to accurately evaluate reservoir quality, but avoids raising public concern
regarding the presence of radioactive sources.

Oileld Review

Its Not Simple

The physics of formation density measurements
with GGD tools is relatively straightforward. As the
137Cs in a typical logging source decays, it emits
about 5.0 1010 gamma rays/s (GR/s). These GRs
interact with the electrons of atoms in the formation in a variety of ways but primarily by Compton
scattering (previous page, top).15 These interactions result in most of the GRs being absorbed by
the formation, but a few travel back to detectors in
the tool located a xed distance from the source.
Formation density measurements are computed
from the number of gamma rays traveling from the
source to the detectors.
From the original pool of GRs emitted by the
source, a small fraction of the scattered gamma
raysa few hundred to more than 10,000 GR/s
will make it to the detectors. High-density rocks
with little porosity result in fewer GRs returning
to the tool than occurs in porous rocks lled with
water, oil or gas.
Gamma ray output can also vary from source
to source. To compensate for differences in individual sources and detector efciencies, each
tool is calibrated to a xed reference so the tool
delivers the correct downhole density value.
As previously noted, engineers have successfully developed tools that replace the 241AmBe
source with PNG-based tools for both neutron
porosity and capture spectroscopy. The pursuit of
a high-quality, radioisotope-free density measurement has been more elusive because of the
lack of electronic gamma ray emitters analogous
to PNGs to replace 137Cs. To overcome this hurdle, Schlumberger scientists adapted some of the
principles used for neutron-based measurements, such as spectroscopy and sigma, to
develop the SNGD measurement.
PNGs generate high-energy neutrons in short
bursts. Neutrons leave the tool and interact with
the various elements of the formation rocks and
uids. The interactions that have the greatest
effect are predominantly elastic collisions with
hydrogen nuclei (previous page, bottom). With
successive collisions, the initial high-energy neutrons slow down and reach thermal energy level.16
Thermal neutron porosity tools count the number
of thermal neutrons that arrive back at the tool;
from this count rate, the traditional thermal neutron porosity is computed.17
Not all the collisions are elastic. Immediately
after the initial burst of neutrons from the PNG,
but before the neutrons reach thermal level,
inelastic collisions occur between the fast neutrons and atomic nuclei in the formation (above
right). Inelastic collisions cause some atomic
nuclei to become excited and emit one or more

Summer 2013

Inelastic Neutron Scattering

Excited
nucleus

Inelastic
gamma rays

Neutron Capture
Excited nucleus
Slow
neutron

Capture
gamma ray

> Neutron interactions. The neutron interactions relevant to petrophysical

logging can be separated into three categories: Inelastic scatter (top),
elastic scatter (not shown) and capture (bottom). Inelastic gamma rays are
generated by the interaction of a fast neutrontypically with energy greater
than 1 MeVwith a nucleus. The interaction lifts the nucleus into an excited
state, the neutron emerges with less energy and one or more gamma rays are
emitted. Also counted among the inelastic gamma rays are those following
a high-energy nuclear reaction, such as a reaction in which the neutron
knocks out a particlesuch as an alpha particle, a proton or a second
neutronfrom the nucleus. In elastic scattering, the neutron bounces off the
nucleus without pushing it into an excited state. The only energy loss is from
the kinetic energy imparted to the nucleus on which the scattering occurs.
Elastic scattering from hydrogen, the essential mechanism underlying the
neutron porosity measurement, is a result of the collision between particles
of equal massneutron and protonwhich causes maximum energy loss.
The neutron capture reaction, in which a neutron can be absorbed by a
nucleus, dominates at low neutron energy. This leaves the absorbing nucleus
in an excited state and the resulting deexcitation is accompanied by the
emission of gamma rays.
14. In 1999, the US Nuclear Regulatory Commission (NRC)
modied existing regulations to exempt PNGs from well
abandonment procedures applied to radioisotopic
sources. For more: NRC: Regulatory Analysis of Energy
Compensation Sources for Well Logging and Other
Regulatory ClaricationsChanges to 10 CRF Part 39,
Ofce of Nuclear Materials Safety and Safeguards
(December 1999), http://pbadupws.nrc.gov/docs/
ML0036/ML003690515.pdf (accessed April 29, 2013).
15. Compton scattering occurs when a gamma ray collides
with an electron, transferring part of its energy to the
electron, while itself being scattered. The gamma ray
continues at a reduced energy. The degree of Compton

scattering depends on the electron density of the target

material. As the electron density increases, there is
more attenuation of gamma ray energy.
16. PNGs emit fast neutrons with a kinetic energy level of
about 14 MeV. Thermal neutrons have a kinetic energy
of about 0.025 eV at room temperature.
17. Weller G, Grifths R, Stoller C, Allioli F, Berheide M,
Evans M, Labous L, Dion D and Perciot P: A New
Integrated LWD Platform Brings Next-Generation
Formation Evaluation Services, Transactions of the
SPWLA 46th Annual Logging Symposium, New Orleans,
June 2629, 2005, paper H.

Inelastic
gamma ray
source volume

PNG
Neutron
detector
Inelastic
scattering

Inelastic
gamma ray
scattering
volume

Gamma ray
detector

> Inelastic gamma ray cloud. The PNG generates

neutrons that move away from the source and
collide inelastically with atoms in the formation
(blue shading). These collisions cause a cloud
of inelastic gamma rays to form (green shading).
Some of these gamma rays will travel back to the
tool and be counted by the detectors.

10

Long-spacing
detector response

Inelastic count rate, counts/s

GRs as they return to ground state. Scientists are

able to use the energy spectrum of inelastic GRs to
identify elements such as carbon, oxygen, silicon,
calcium, iron and sulfur. Engineers use the volumetric yields of these elements to compute lithology, and this is the basis of neutron spectroscopy
measurements. The energy spectrum of inelastic
gamma rays is also the basis of carbon/oxygen ratio
tools, which are used to identify hydrocarbonbearing zones in cased holes.
During the short period of inelastic collisions,
a GR cloud forms (below). This articially generated cloud emits around 108 GR/s, about two
orders of magnitude lower than the number emitted by a typical 137Cs source. Scientists have
determined, however, that there are sufcient
GRs produced to function in a manner similar to
that of a traditional source. The GR cloud is
short-lived because the neutrons that create it
collide with other nuclei, rapidly slow to thermal
level and are subsequently captured.
The number of gamma rays that result from
inelastic collisions and reach the detectors from
the GR cloud is inuenced by three factors: the

Gamma ray
transport

Neutron
transport

Formation density, g/cm 3

> Nuclear transport and long-spacing detector response. The response

of the long-spacing gamma ray detector (black) is largely determined
by neutron (blue) and gamma ray transport (red). Neutron transport is
related to the interactions of neutrons with atomic nuclei in the formation.
Inelastic gamma rays are produced during inelastic scattering of fast
neutrons. Elastic scattering, which occurs primarily when neutrons
collide with hydrogen nuclei, reduces the energy of the fast neutrons
below the threshold for producing inelastic gamma rays. Thus, with
increased formation density (lower porosity), there are fewer hydrogen
nuclei available for elastic scattering and, as a result, there are more fast
neutrons available for the production of inelastic gamma rays. Gamma ray
transport and the number of inelastic gamma ray counts decrease with
increased formation density because the higher electron density provides
more opportunity for gamma ray interactions and energy reduction.

fast neutron transport from the PNG to the

point where inelastic GRs are produced within
the formation, the subsequent transport of GRs
from their origin back to the detectors in the
tool and the electron density of the formation.
The GRs generated in the formation by inelastic
interactions move rapidly through the formation, interacting in a manner similar to GRs generated by a radioisotopic source, and they are
attenuated by collisions with electrons within
the formation primarily through Compton scattering (above). Properly characterized, the
counts at the detector are used to compute electron density, which in turn is used to compute
the formation bulk density.18
If only inelastic GRs were present, the characterization would be more easily performed; however, another major source of GRs complicates the
measurement. Fast neutrons eventually become
thermal neutrons and are captured by atoms in
the formation. Nuclei that capture thermal neutrons emit GRs to return to a stable energy state
in a manner similar to the emission of GRs resulting from inelastic collisions. The population density of thermal neutrons available for capture is
directly related to the number of hydrogen atoms

in the formation. In a typical downhole environment, the element with the highest probability of
absorbing thermal neutrons is chlorine [Cl],
whose number density is related to the salinity of
the formation uids. The SNGD measurement is
based only on GRs generated by the inelastic collisions. To correctly compute the bulk density
value, the contributions from capture GRs resulting from neutron capture must be quantied and
removed from the measurement.19
Engineers must also account for the variability of the initial source strength. The output of a
traditional source may vary, depending on age
and activity level of the radionuclide, but the output is fairly constant and its change over time is
predictable. Calibration of GGD tools accounts
for variability between sources and detector efciencies by correcting to a known reference. The
output of a PNG is not as predictable and may
vary over short periods of time and even between
bursts. A control loop in the NeoScope tool
adjusts the PNG to maintain a constant average
output, and the tool includes a detector at the
18. Reichel et al, reference 5.
19. Epithermal neutrons have an energy range between
about 0.02 eV and 10 keV at room temperature.

Oileld Review

NeoScope Calibration Facility

1

NeoScope
tool

Mud
channel

Aluminum
calibration sleeve

Water

Calibration
sleeve

Detectors

> NeoScope calibration device. A special calibration facility was developed specically for the NeoScope tool. Four
measurements are performed in a water-lled tank using a calibration sleeve and a simulated mud channel. With the PNG turned
on, responses are measured in four congurations: sleeve raised, mud channel lled with air (1); sleeve raised, mud channel
lled with water (2); sleeve lowered, mud channel lled with water (3); and sleeve lowered, mud channel lled with air (4). These
four measurements allow calibration gains and offsets to be computed and provide quality checks for tool verication.

PNG to determine the neutron output and compensate for variations.

To provide the specied 0.025-g/cm3 accuracy
for the density measurement, the SNGD model uses
a combination of responses from multiple detectors
and requires a complex and demanding calibration.
This calibration consists of correlating the count
rates measured by each of the tools detectors to

those measured in the same environment with the

reference tool. For this purpose, engineers have
designed a new calibration tank that allows measurements over a wide range of count rates (above).
The uncertainties found in downhole log measurements arise from the primary measurement,
applied corrections and conversion of measured
parameters to formation properties. To mitigate

PNG
Neutron monitor

these uncertainties, the NeoScope service includes a

quality control system that begins with general tool
system hardware and moves to specic sensor functions, individual sensor measurements and integrated measurements that may involve multiple
individual sensor responses (below). The last step of
the process is quality control of the nal integrated
answers that may use multiple measurements.

Long-spacing gamma ray detector

Source output correction
(neutron monitor)
Neutron transport correction
(near epithermal and far thermal detectors)

Near epithermal detector

Near thermal detector

Sigma input

Short-spacing
gamma ray detector

Spectroscopy input

Far thermal detector

Neutron porosity input

Long-spacing
gamma ray detector

Neutron-gamma density input

Fast neutron correction

(short- and long-spacing gamma ray detectors)
Sigma correction
SNGD output

> Multi-input, multioutput measurements. The nuclear portion of the NeoScope tool (left) uses a single PNG to generate
neutrons, but the responses from multiple detectors are integrated to produce specic measurements. For example, sigma
data are derived from near thermal, short-spacing gamma ray and long-spacing gamma ray detectors. SNGD data, the most
complex measurement from the NeoScope tool, are primarily computed using counts from the long-spacing gamma ray detector,
but inputs from the neutron monitor, near epithermal detector, short- and long-spacing gamma ray detectors and far thermal
detectors are required to provide an accurate nal answer. The owchart (right) traces the corrections applied to arrive at the
nal density output.

Summer 2013

11

Density range
Precision at ROP 61 m/h [200 ft/h]

SNGD

GGD

1.7 to 2.9 g/cm3

1.7 to 3.05 g/cm3

g/cm3

0.006 g/cm3

Clean sandstone,
limestone and dolomite

0.025 g/cm3

0.015 g/cm3

Shale

0.045 g/cm3

0.015 g/cm3

Salt

Not applicable

0.015 g/cm3

Anhydrite

Not applicable

0.015 g/cm3

Axial resolution

89 cm [35 in.]

36 cm [14 in.]

Depth of investigation

25 cm [10 in.]

10.2 cm [4 in.]

Image capability

No

Yes

0.018

Accuracy

> Specications for SNGD and GGD tools.

Individual quality control considerations

that may impact accuracy include sensor and
hardware functionality, density values within
the 1.7- to 2.9-g/cm3 range of SNGD and tool
standoff. In addition, environmental quality
controls include borehole size, deviation, ROP
and formation shaliness, all of which may
impact measurement accuracy (above). The
indicators are combined into a measurement
quality control ag. A green ag suggests that
the measurement is accurate and within specied limits. A yellow ag indicates that the measurement is likely to be within its specied

range but may require further interpretation,

and a red ag means that the measurement is
outside specied accuracy parameters. These
quality ag values are crucial for comparing the
accuracy of GGD and SNGD measurements.
Field Testing and Beyond
Field tests for the SNGD measurements consisted of comparing them with GGD measurements using a modied tool that allowed
engineers to acquire both measurements simultaneously from the same well using the same

3.0

Data within tolerance

Data at limit of tolerance

GGD data, g/cm3

2.8

2.6

2.4

2.2

2.0
2.0

2.2

2.4

2.6

2.8

3.0

SNGD data, g/cm3

> Crossplot comparison. Density data from a GGD tool were compared with
data from an SNGD tool; the data are color-coded by their quality ag value.
There is good agreement between the two when SNGD data are within
tolerance. The data align well along the ideal axis and are agged as green.
Invasion effects start to occur in the lower density range at approximately
2.3 g/cm3. The spread of the data points around the ideal line is attributed to
differences in the axial resolution of the two measurements while crossing
various layers at high deviations.

12

bottomhole assembly. Objectives for eld testing included logging in the following:
clean sandstone, limestone and dolomite
formations
anhydrite
shale
gas and light hydrocarbon reservoirs
large boreholes
deviated and vertical wells.
Scientists compared the GGD measurement,
considered the benchmark, with SNGD results
and accounted for the differences and limitations
of both measurements. Test acceptance criteria
were based on a systematic evaluation of both
measurements, and nal analysis was based on a
set of numerical interpretation criteria.20
The maximum acceptable error when two
independent measurements are compared is the
sum of their individual accuracies. In this case,
the acceptable error for the two measurements is
0.040 g/cm3 in clean formations and 0.060 g/cm3
in shales.21 The data from the combined tools
were plotted, which allowed engineers to quantify any deviation from perfect agreement.
Additionally, scientists had to account for
conditions in each well that might impact GGDto-SNGD comparisons. These conditions included
ltrate invasion, the presence of gas or light
hydrocarbons that may change with time and
various drilling conditions, such as mud weight,
uid variations and changes in ROP. If a large discrepancy between the two measurements could
be explained by environmental effects, the test
was considered acceptable. All tests were performed in 8 1/2-in. boreholes.
In a eld test of the NeoScope service, the
operator drilled a well with an average inclination of 60 through a sandstone reservoir using
1.26-g/cm3 [10.5-lbm/galUS] water-base mud
(WBM). The caliper log indicated the borehole
was in gauge, and no GGD data corrections were
required. Additionally, the GGD data indicated
no major azimuthal effects. Sigma was within a
range that indicated minimal correction to the
SNGD. In the hydrocarbon-bearing section of the
formation, the resistivity log indicated some invasion (next page). Because of the difference in
their DOIs, the SNGD and GGD outputs were
slightly different in this zone. By contrast, these
measurements were almost identical in a noninvaded water-bearing section of the formation.
The SNGD data were within accuracy limits
throughout the well (left).
20. Reichel et al, reference 5.
21. Theys P: Log Data Acquisition and Quality Control.
Paris: Editions Technip, 2nd edition, 1999.

Oileld Review

Resistivity
Quadrant Bulk
Density Data

40-in. Attenuation
34-in. Attenuation

Average Density

28-in. Attenuation
22-in. Attenuation

Mudcake
0.02

Deviation

Density Caliper
in.

in.

Collar
Rotation

10
0

Gamma Ray
0

gAPI

150

RPM 500

Depth, ft

ohm.m

0.8
200 1.7

2.7
1.9

28-in. Phase Shift

Image-Derived Density 1.9

22-in. Phase Shift

1.9

ohm.m

cu

g/cm3

g/cm3

50

2.9

g/cm3

1.9

g/cm3

g/cm3

2.9 40

g/cm3

2.9

g/cm3

2.9

Right Density

2.9

Bulk Density Upper

2.9

Left Density
1.9

Bulk Density

g/cm3

Bottom Density

2.9

Neutron Porosity
(Thermal)

2,000 1.9

1.9

0.2

Neutron Density

34-in. Phase Shift

16-in. Phase Shift

0.2

g/cm3

Sigma

40-in. Phase Shift

10 0 degree 90

Ultrasonic Caliper
8

Density Correction

16-in. Attenuation

Washout

Density Image

1.9

15 1.9

g/cm3

Water
2.9

Up Density
g/cm3

Pyrite

2.9

Sandstone
Clay

Quality
Flags

X10

X20

X30

X40

X50

X60

X70

> Density comparison in an invaded oil zone. The interval from X10 to X40 ft is an oil-bearing sandstone with mud ltrate invasion. The invasion is indicated
by separation in the resistivity curves (Track 2, blue shading). The sandstone below X60 ft (red shading) is water lled, and the lack of separation indicates
little to no invasion. The NeoScope toolalong with a conventional GGD LWD toolwas run in this well. The density image (Track 3) indicates a fairly
homogeneous reservoir, as does the lithology computed from spectroscopy data (Track 6). Quadrant density data (Track 5) overlie each other through the
two sections, as would be expected with the high-quality wellbore conditions. There is excellent agreement between the traditional density (Track 4, red)
and the NeoScope density (black), although there is a slight difference between the two datasets in the oil-bearing interval because of the invasion. These
data overlie the thermal neutron porosity data (blue) in clean, water- or oil-lled rocks. (Adapted from Reichel et al, reference 5.)

Summer 2013

13

Resistivity
40-in. Attenuation

Quadrant Bulk
Density Data

Density Image

34-in. Attenuation

Average Density

28-in. Attenuation
22-in. Attenuation

Mudcake
0.02

Deviation

Density Caliper
in.

10 0 degree 90

Ultrasonic Caliper
8

in.

10
0

Gamma Ray
0

gAPI

Collar
Rotation

150

cu

Density Correction
0.8

50

Image-Derived Density 1.9

34-in. Phase Shift

1.9

22-in. Phase Shift

0.2

2.7

40-in. Phase Shift

28-in. Phase Shift

RPM 500

Depth, ft

200 0

ohm.m

g/cm3

Sigma

16-in. Attenuation

Washout

1.7

g/cm3

2.9

Bulk Density Bottom 1.9

1.9

g/cm3

2.9

16-in. Phase Shift

Bulk Density Upper

ohm.m

g/cm3

2,000 1.9

2.9 40

g/cm3

1.9

g/cm3

1.9

g/cm3

g/cm3

2.9

g/cm3

2.9

Right Density

2.9

Neutron Porosity
(Thermal)

2.9

Left Density

2.9

Bulk Density

g/cm3

Bottom Density

0.2

Neutron Density

1.9

1.9

15 1.9

g/cm3

Carbonate
2.9

Up Density
g/cm3

2.9

Sandstone
Clay

Quality
Flags

X10

X20

X30

> Comparison of washout effects on density. Density data were acquired using a NeoScope tool and a conventional GGD LWD tool across a predominantly
water-lled carbonate section (Track 6, lithology) of a test well. Caliper data (Track 1) from the NeoScope tool (black) and the traditional density tool (red)
indicate an enlarged borehole (blue shading) above and below X12 ft. Resistivity data are presented in Track 2. Track 3 contains density image data from
the traditional tool, along with azimuthal density from the bottom (red dashed) and upper (green) quadrants, an image-derived density (black) and sigma data
(purple). The bulk density data from the conventional tool (Track 4, red) are affected by hole conditions from X10 to X18 ft, but the NeoScope tool provides
good density data (black). The differences in the quadrant data from the traditional GGD tool (Track 5) demonstrate the effects of the enlarged borehole.
The left quadrant (blue) and the upper quadrant (green) data are invalid, as is the average computed density (red). The bottom quadrant (pink) and the right
quadrant (dark red) data are closer to the NeoScope density in Track 4. While the NeoScope density has a greater DOI and is less affected by washouts or
hole rugosity, the yellow quality ag (Track 7) indicates the measurements are approaching the limits. (Adapted from Reichel et al, reference 5.)

In another eld test conducted in a limestone formation at the Schlumberger test facility in Cameron, Texas, USA, engineers drilled a
well with an average inclination of 25 using
1.13-g/cm3 [9.4-lbm/galUS] WBM (above). The
caliper log indicated hole enlargement in the
top section of the log. In zones where the SNGD
quality control ag was yellow, there were signicant differences between the SNGD and
GGD data. The density correction on GGD data

14

was generally between 0.1 and 0.15 g/cm3, which

is not usually indicative of compromised data
quality resulting from hole rugosity, although
the quadrant density data clearly showed effects
of the enlarged borehole.
Analysis of these two logs highlighted the value
of the greater DOI of the SNGD measurement. The
SNGD data were borehole corrected and, because
of the NeoScope tools greater DOI, were less inuenced by variations in the near-borehole environ-

ment. The SNGD curve tracks the thermal neutron

porosity curve in clean formations as expected.
The SNGD data appear more reliable than the traditional GGD measurement.
A Middle East operator tested the new SNGD
design in four environments.22 The NeoScope tool
was run in a high-angle, high gas saturation reservoir drilled with nonaqueous mud, a high gas
saturation reservoir drilled with WBM, an oilsaturated carbonate reservoir drilled with high-

Oileld Review

salinity WBM and an oil-saturated carbonate

reservoir drilled with low-salinity WBM. To validate the measurements, traditional GGD tools
were run for comparison.
The rst test was in an 8 1/2-in. wellbore, in
which the high-angle well approached 90 deviation at TD. The nonaqueous mud system was
barite-saturated, which invalidated PEF measurements from the GGD tool. The reservoir section was predominantly limestone and the
formation density ranged from around 1.95 to
2.7 g/cm3. A comparison of the data from the GGD
tool with those from the NeoScope SNGD measurement shows excellent agreement (right).
One benet of the NeoScope tool is the availability of neutron capture spectroscopy data.
Although the PEF measurement from the traditional tool was affected by barite in the mud system, lithology could still be determined using
spectroscopy data from the NeoScope tool. The
majority of the interval was limestone, although
some dolomite was observed.
The second example was a vertical well drilled
with WBM through a gas-lled carbonate reservoir
in the same eld as the previous well. Comparison
of GGD with SNGD data again showed good agreement across a wide range of values.
A third example was drilled with high-salinity
WBM through an oil-saturated carbonate reservoir. In this highly deviated well, the porosity
data from the GGD and SNGD measurements
compared favorably, well within statistical precision limits of the measurements. As is typical of
liquid-lled reservoirs, the neutron porosity data
values were similar to porosities computed from
formation density data.
A fourth well was a high-angle well drilled
with low-salinity polymer-base WBM. As with
the other three wells, there was excellent agreement between the SNGD data and conventional
GGD measurements.
Petrophysical analysis of data from these four
wells demonstrated that in a variety of wells with
a wide range of density values, SNGD data from
the NeoScope tool compare favorably with data
from conventional density tools. In addition to
the SNGD data, the neutron porosity and resistivity measurements provide a sourceless triplecombo logging option for LWD applications.
Sigma and spectroscopy data are added benets
that petrophysicists can use to better characterize and understand reservoirs.
22. Atfeh M, Al Daghar KA, Al Marzouqi K, Akinsanmi MO,
Murray D and Dua R: Neutron Porosity and Formation
Density Acquisition Without Chemical Sources in Large
Carbonate Reservoirs in the Middle EastA Case
Study, Transactions of the SPWLA 54th Annual Logging
Symposium, New Orleans, June 2226, 2013, paper KKK.

Summer 2013

Density Correction
0.8

g/cm3

0.2

Neutron Porosity (Corrected)

Resistivity
Bit Size
8

in.

10

34-in. Phase Shift

Ultrasonic Caliper
8

in.

10

Gamma Ray
0

gAPI

40

100

Depth,
ft
0

Sigma
cu

1.9

g/cm3

28-in. Phase Shift

Neutron Density

22-in. Phase Shift

g/cm3

1.9

16-in. Phase Shift

50 0.2

ohm.m

15

Bulk Density

40-in. Phase Shift

2,000 40

Lithology
2.9

2.9 1.9

Neutron Porosity (Thermal)

%

Density Image
g/cm3

Dolomite
2.75

Bulk Density
15 1.95

g/cm3

Calcite
Sandstone

2.95

Clay

X,300

X,400

X,500

X,600

> Density comparison in a barite-weighted mud system. Barite in drilling mud can render PEF
measurements invalid. PEF is important for inferring lithology, which is used for porosity calculations.
In this high-angle Middle East carbonate reservoir, the spectroscopy data from the NeoScope tool
provide mineralogy information (Track 6) that would not have been available from traditional density
tools. For example, the data show dolomite mixed with calcite from X,350 to X,420 ft. In the high-density
carbonate intervals, such as from X,400 to X,520, the NeoScope density data (Track 4, black) compare
favorably with traditional bulk density (red). Traditional thermal neutron porosity (blue) is presented
along with a density-corrected thermal neutron porosity (green). The NeoScope tool does not provide
azimuthal density or density images as are available from the traditional LWD GGD tool (Track 5).
Sigma data (Track 2) may be used to determine changes in hydrocarbon saturation or uid contacts
over time. Track 3 presents resistivity data. (Adapted from Atfeh et al, reference 22.)

The Pulse of Things to Come?

It has been a long time coming, but the introduction of SNGD technology may revolutionize
LWD porosity logging. Replacing sources with
PNGs has the potential to eliminate exposure
risks and reduce costs associated with source
storage, transportation and record keeping.

Introducing similar measurements for wireline applications is the next obvious step.
Unfortunately, modeling borehole effects on the
measurement for wireline tools has been beyond
the reach of current research and software. It
may take some time, but if traditional sources
can be replaced in wireline tools, the ALARA
standardas low as reasonably achievablewill
be reached in the oil and gas industry.
TS

15

Core Truth in Formation Evaluation

The nature of subsurface exploration forces oil and gas companies to investigate
each reservoir remotely, primarily through well logs, seismic surveys and well tests.
Through analysis of rock samples obtained downhole, core laboratories provide a
wealth of information about lithology, porosity, permeability, uid saturation and other
properties to help operators better characterize the complex nature of the reservoir.

Mark A. Andersen
Brent Duncan
Ryan McLin
Houston, Texas, USA
Oileld Review Summer 2013: 25, no. 2.
Copyright 2013 Schlumberger.
For help in preparation of this article, thanks to Angela
Dippold Beeson, David Harrison, Mario Roberto Rojas and
Leslie Zhang, Houston; Carlos Chaparro and Adriano Lobo,
Ecopetrol, Bogot, Colombia; Alyssa Charsky, Michael
Herron and Josephine Mawutor Ndinyah, Cambridge,
Massachusetts, USA; William W. Clopine, ConocoPhillips
Company, Houston; Rudolf Hartmann, BCHI Labortechnik
AG, Flawil, Switzerland; Thaer Gheneim Herrera, Bogot,
Colombia; Wendy Hinton, Himanshu Kumar and David R.
Spain, BP, Houston; Upul Samarasingha, Salt Lake City,
Utah, USA; Tony Smithson, Northport, Alabama, USA; and
Elias Yabrudy, Coretest Systems, Morgan Hill, California, USA.
Techlog, TerraTek and XL-Rock are marks of Schlumberger.
PHI-220 Helium Porosimeter is a mark of Coretest Systems,
Inc.
LECO is a mark of the LECO Corporation.

Cores provide essential data for the exploration,

evaluation and production of oil and gas reservoirs. These rock samples allow geoscientists to
examine rsthand the depositional sequences
penetrated by a drill bit. They offer direct evidence of the presence, distribution and deliverability of hydrocarbons and can reveal variations
in reservoir traits that might not be detected
through downhole logging measurements alone.
Through measurement and analysis of porosity,
permeability and uid saturation from core samples, operators are better able to characterize

Whole Core
Segment

pore systems in the rock and accurately model

reservoir behavior to optimize production.
Core analysis is vital for determining rock
matrix properties and is an important resource for
formation characterization. The process known as
routine core analysis helps geoscientists evaluate
porosity, permeability, uid saturation, grain density, lithology and texture. Routine core analysis
laboratories (RCALs) frequently provide a variety
of additional services such as core gamma logging
for correlating core depth with wellbore logging
depth, core computed tomography (CT) scans for

Full Diameter
Core Analysis

Core Plug
Analysis

2.5 to
3 in.

3 ft

1 ft

1 ft

1 ft

1 ft

1 ft

1 ft

> Divided cores. At the wellsite, whole cores are typically cut into smaller
segments for ease of shipping. At the laboratory, the whole core segments
may be cut and subsampled.

16

Oileld Review

characterizing rock heterogeneity and core photographs for documenting and describing the core.
When operators need to understand complex
reservoir behaviors, they turn to special core analysis for detailed measurements of specic properties. Special core analysis laboratories (SCALs) are
typically equipped to measure capillary pressure,
relative permeability, electrical properties, formation damage, nuclear magnetic resonance (NMR)
relaxation time, recovery factor, wettability and
other parameters used for calibrating logs. SCAL
services are also used to characterize reservoirs for
enhanced oil recovery (EOR) and for studying multiphase ow and rock-uid interactions. Only a few
samples are selected for these extensive tests, some
of which require weeks to complete.

Summer 2013

For years, Schlumberger has maintained a

number of core analysis laboratories to support
research into wireline tool response, drilling
uid chemistry, formation damage, EOR or completion technology. However, these facilities did
not provide core analysis on a commercial scale.
Until recently, the companys commercial core
analysis services were centered in Salt Lake City,
Utah, USA, where the TerraTek rock mechanics
and core analysis facility is known for its focus
on geomechanics and unconventional reservoirs.
The 2012 inauguration of Schlumberger Reservoir
Laboratories has opened the way for integrating rock measurement technologies with uids
expertise to help customers better understand
reservoir behavior. Schlumberger now offers rock
and uid analysis through 27 laboratories around

the globe. Several companies offer similar analyses of conventional cores. This article focuses on
routine analysis of conventional sandstone and
carbonate cores carried out by specialists at the
Schlumberger Reservoir Laboratory in Houston.
Sample Sizes
Cores come in a variety of lengths and diameters
(previous page). The information extracted from
a core depends in part on the size and quantity of
the core, which control the types of analyses that
may be performed. To meet customer needs, the
core analysis laboratories must be exible
enough to process the various types of core sent
from the wellsite, be they bottomhole cores or
sidewall cores.

17

Bottomhole cores, also referred to as whole

cores, or conventional cores, are obtained during
the drilling process using a special coring bit
(below). The cores typically range in diameter
from 4.45 to 13.3 cm [1.75 to 5.25 in.] and are
generally drilled in 10-m [30-ft] increments,
which correspond to the length of the core barrel
or its liner. Whereas a conventional bit is designed
to grind away the rock at the bit face, the doughnut-shaped coring bit creates a cylinder of rock
that passes through the center of the bit and is
retained in a protective core barrel.
When the core barrel is full, the driller pulls
the assembly out of the hole, and a wellsite coring
specialist lays the barrel liner on the pipe rack.
The liner, with core inside, is then scribed with
depth markings and orientation lines. For ease of
shipping, the metal liner is usually cut into 1-m
[3-ft] segments and sealed at each end. To prevent shifting during transit, the wellsite corehandling team may inject epoxy or foam into the
liner to stabilize the core.
For sidewall cores (SWCs), the process is far
less involved. SWCs are obtained by a wireline
sampling device, which is typically run in the
hole near the conclusion of an openhole wireline
logging job after the operator consults the logs to

identify zones that merit sampling. The SWC

device can extract up to 90 samples from the side
of the wellbore at selected depths. Once on the
surface, sidewall cores are retrieved from the
tool, sealed in individual bottles and shipped to
the laboratory for analysis.
Percussion-type sampling devices obtain
SWCs measuring from about 2.86 to 4.45 cm
[1.125 to 1.75 in.] in length by 1.75 to 2.54 cm
[0.688 to 1 in.] in diameter. Percussion sampling
devices are known as core guns because they use
small explosive charges to propel individual core
barrels, called bullets, into the formation. The
core barrels are attached to the gun with strong
cables that are used to pull the core bullet from
the borehole wall as the gun is reeled uphole. By
contrast, rotary cores are cut from the formation
using a miniature, horizontally oriented coring
bit. The XL-Rock large-volume rotary sidewall
coring tool can drill cores 6.4 cm [2.5 in.] long by
3.8 cm [1.5 in.] OD from the side of the borehole.
This device produces core samples that have more
than three times the volume of percussion SWCs.
A third type of rock sample is the core plug.
The core plug is extracted from segments of
whole core. These plugs are taken as a representative subsample of the whole core and are useful

> Coring bit. This polycrystalline diamond compact (PDC) bit employs a xed
cutter design that leaves the center of the borehole untouched. The bit
creates a cylindrical core of the formation that passes through the middle of
the bit, to be retained within the bottomhole assembly.

18

in analyzing intervals of relatively homogeneous

core. Core plugs in conventional reservoirs are
routinely taken at 0.3-m [1-ft] intervals along the
length of the core and measure about 6.4 cm long
by 2.54 or 3.8 cm in diameter. Variations in lithology may require smaller sampling intervals, but if
the core is highly heterogeneous, as seen in vugular or fractured carbonates or thinly laminated
sand-shale intervals, the operator may elect to
analyze the whole core rather than plugs.
Initial Processing
The basic workow for conventional core analysis
moves from receiving to preliminary imaging,
then to preparation and analysis. Each process
involves several steps. Whole cores typically
require more initial processing than sidewall
cores. Although routine core analysis provides a
standard set of measurements, not all cores go
through the entire workow described here.
At the laboratory, cores are received and inventoried. Whole cores are run through a core gamma
ray logger, which measures gamma rays that are
naturally emitted from the cores. By comparing
core gamma ray measurements to LWD or wireline
gamma ray logs, geoscientists can correlate core
depth to log depth and identify intervals from
which core may have been lost or damaged.
The core gamma ray logging device uses a
conveyor to move the coreeither exposed or
sealed within the linerpast a gamma ray detector. The detector scans the core along its length
from bottom to top, which replicates the logging
sequence used in obtaining wireline logs.
Next, the core is run through a computed
tomography scanner to obtain a CT image. The
CT device obtains a 3D image of the whole core,
taking a series of closely spaced scans that can be
sliced at any point or orientation to create a virtual slab of the core. The CT scanner permits a
quick reconnaissance across the core. When
zones of interest are identied, they may be
scanned again for detailed examination (next
page). CT scanning is especially useful for detection and evaluation of internal features such as
bedding planes, vugs, nodules, fossils and frac1. For more on CT scans in oileld applications: Kayser A,
Knackstedt M and Ziauddin M: A Closer Look at Pore
Geometry, Oileld Review 18, no. 1 (Spring 2006): 413.
2. Passey QR, Dahlberg KE, Sullivan KB, Yin H, Brackett RA,
Xiao YH and Guzmn-Garcia AG: Digital Core Imaging In
Thinly Bedded Reservoirs, in Dahlberg KE (ed):
Petrophysical Evaluation of Hydrocarbon Pore-Thickness
in Thinly Bedded Clastic Reservoirs. Tulsa: The American
Association of Petroleum Geologists, AAPG Archie
Series, no. 1 (June 30, 2006): 90107.
3. Perarnau A: Use of Core Photo Data in Petrophysical
Analysis, Transactions of the SPWLA 52nd Annual
Logging Symposium, Colorado Springs, Colorado, USA,
May 1418, 2011, paper Z.

Oileld Review

tures.1 Operators occasionally specify that sidewall cores be scanned as well. CT scans are
noninvasive, require no further preparation of
the core and can be performed rapidly on exposed
cores or on cores retained inside the core barrel.
After performing initial scanning, the core
analyst frees the core from the barrel to prepare
it for further testing. The analyst uses a band saw
or radial saw, equipped with a diamond-impregnated blade, to cut the core lengthwiseparallel
to the core axisinto slabs. In most cases, the
core is cut off center rather than sliced down the
middle. The thickness of the slab dictates the
maximum size of any plugs subsequently sampled
from the core. The at face of the thinner slab is
polished to remove saw marks in preparation for
slab photography.
In some cases, portions of the core are not
slabbed. When a core exhibits substantial, largescale heterogeneitytypical of vuggy carbonates
or severely fractured or conglomeratic rockthen
sections of the core may be set aside without slabbing to permit analysis of the full diameter core.
The core slabs are photographed using a
35-mm digital camera linked to a dedicated computer, which can digitize, display and transmit
images to the client. Photographs can often
resolve individual layers within thin beds measuring just tenths of an inch.
Digital photography helps highlight important geologic and petrophysical characteristics.
These high-resolution color images provide an
important visual record of lithology, bedding
characteristics, contacts, fractures, fossils,
porosity, vugs and sedimentologic variations that
may be studied in detaillong after the core has
been subjected to further testing. Subsequent
manipulation and analysis of core images often
yield valuable information not readily apparent
from the original photographs. In some cases,
they can be used to reconcile discrepancies
between core and log analysis, detecting formation laminations that are too thin to be resolved
by the logging tool.
Upon client request, a 360 axial wrap of the
core is imaged. This is accomplished using a
digital camera and a table with rollers that
rotate the full diameter core longitudinally as it
is photographed.
Photographs are taken in white and in ultraviolet (UV) light. Images shot in plain white
light show the cores under natural lighting conditions. The UV light may highlight certain types
of minerals but, more importantly, it can
enhance the contrast between nonreservoir and
oil-bearing zones. Oil-bearing reservoir rocks
frequently exhibit strong ultraviolet uores-

Summer 2013

> Whole core CT scans. Pore characteristics are brought into sharp focus in a virtual slice of core (top,
foreground) as the sample passes through the CT scanner (top, background). Color coding on the scan
helps distinguish regions of differing density or mineralogy. By contrast, grayscale images are used to
highlight core damage. A core obtained through a friable formation at the Casabe eld in Colombia
was scanned prior to being removed from the core liner (bottom). Stacks of cross sections revealed
areas where portions of the core were damaged. The white exterior ring is the core liner; a layer of
caked drilling mud inside the liner surrounds the core. By avoiding fractured intervals (bottom left), the
core analyst was able to select undamaged sections (bottom right) from which to extract plugs. (CT
images courtesy of Carlos Chaparro and Adriano Lobo, Ecopetrol, Bogot, Colombia.)

cence. Typically, an oil emits uorescence, the

brightness and color of which are affected by
its composition. However, some oils do not
uoresce. Furthermore, if ushing has removed
some of the oil as the core was brought to the
surface, or if the core was not well preserved,
the pay interval may not uoresce uniformly.2
It is difcult to assess uorescence using the
naked eye; however, digital color photography
records numerical input, which some operators
use for subsequent computer analysis.3 Each photograph is made up of pixels, and each pixel may

be assigned one of more than 16 million shades.

Geoscientists lter or manipulate these colors to
highlight important features. Statistical analysis
of color data can help geologists differentiate
between lithologies or establish porosity or permeability cutoffs. The computer can count how
many pixels fall within a specied color range to
determine net sand or net uorescence in thinly
laminated zones.
Although core facilities typically accommodate
a wide range of samples, core plugs are most frequently used for routine core analysis. Plugs provide

19

Solvent

Boiling Point

Solubility

Methylene chloride

40.1C [104.25F]

Oil and limited water

Hexane

49.7C to 68.7C [121.5F to 155.7F]

Oil

Chloroform/methanol azeotrope 53.8C [128.8F]

Oil, water and salt

Acetone

56.5C [133.7F]

Oil, water and salt

Methanol

64.7C [148.5F]

Water and salt

Tetrahydrofuran

65.0C [149.0F]

Oil, water and salt

Cyclohexane

81.4C [178.5F]

Oil

Ethylene chloride

83.5C [182.3F]

Toluene

110.6C [231.1F]

Tetrachloroethylene

121.0C [249.8F]

Oil

Xylene

138.0C to 144.4C [280.4F to 291.9F]

Oil

Naphtha

160.0C [320.0F]

Oil

100C

Oil and limited water

Oil

> Common core-cleaning solvents, ordered by boiling point temperature. The

choice of solvent typically depends on wettability interactions between the
crude oil and the minerals contained within the rock. Complete extraction of
some oils may require mixtures or series of solvents. The red line represents
the boiling point of water. (Table adapted from API, reference 6.)

a reliable characterization of the core when the

pore system is relatively homogeneous.4
The core analyst, sometimes working in concert with the operators geologist, drills sample
plugs from a full diameter core. Most laboratories
use a mill or a drill press with a diamond bit to
drill the core plugs. The analyst cuts the plug to a

standard length then applies a precision nish

using an end face grinder. The result is a right cylinder, typically 38 mm [1.5 in.] diameter by 64 mm
[2.5 in.] long, with a at face at each end. By creating plugs of standard shape and size, the analyst
obtains samples with the same cross-sectional
area and length; thus, each plug has essentially

Distilled solvent

Solvent vapor

Condenser
pill point
point
Spill

tracctor
trac
Extractor

SSiph
h
hon
Siphon

Coree
samples
samp
p

Cor
rre
Core
sam
mplee
m
samples

Distillation
flask

Return lilliquid
R
id
solvent
Heating
mantle

Solvent vapor

> Soxhlet distillation extractor. Solvent in the distillation ask (left) is gently
heated until it vaporizes. The solvent vapors rise from the ask and cool when
they reach the condenser. The cooled liquid solvent drips onto the core to
permeate the sample. The solvent condensate carries away the hydrocarbons
and brine from the sample. When distilled solvent in the extractor reaches its
spill point, the used solvent siphons back into the ask to be redistilled (right).
This process is repeated continuously and can be sustained as long as needed.
The hydrocarbons from the core are retained and concentrated in the distillation,
or boiling, ask. Some Soxhlet devices can accommodate multiple core plugs.

20

the same bulk volume. A standard size plug also

reduces the potential for measurement errors
stemming from irregularly shaped samples.
Core Cleaning and Fluid Extraction
In addition to rock matrix, core samples contain
formation uids. If the core is taken from a productive zone, these formation uids will typically
contain a mixture of hydrocarbons and saltwater,
or brine. At the laboratory, these uids, which
would otherwise interfere with routine core analysis measurements of porosity and permeability,
must be completely removed from the pore
spaces of the rock.
Core cleaning and uid extraction are combined in a delicate process, which must be thorough enough to remove heavy fractions of crude
oil yet gentle enough to prevent damage to the
mineral constituents of the rock. This process
must avoid creation of additional pore space
resulting from dehydration of clays and hydrous
minerals, such as gypsum, or from erosion caused
by high ow rates as solvent passes through the
sample.5 Several techniques have been developed
for removing residual formation uids; the most
widely used involve distillation extraction or continuous solvent extraction.
Unslabbed cores, plugs and SWCs are carefully cleaned using a specialized closed-loop system that employs either a Soxhlet cleaning
treatment or a Dean-Stark uid extraction process. In the Soxhlet process, the sample is
allowed to soak in the solvent; in the Dean-Stark
method, solvent vapors and liquids ow through
the sample. Both techniques rely on heat to drive
the water from the core sample and on solvent to
extract any hydrocarbons (above left).
Soxhlet extraction uses a distillation process to
clean the core. The Soxhlet apparatus consists of a
heating mantle with a thermostatic controller,
boiling ask, extractor and condenser (left). The
solvent is gently boiled, and the distilled solvent
collects in the extractor, in which one or more core
samples soak. The heated solvent is continually
distilled, condensed and reuxed. The cleanliness
of the sample is determined from the color of the
solvent that siphons periodically from the extractor; the process is repeated until the extract
remains clear after an extended soak cycle. This
method uses one or more solvents to dissolve and
extract oil and brines from the core sample.
After repeated cycles, the extract should become
clear as no more oil is carried out of the rock. However,
the fact that one solvent is clear may not necessarily
mean that the oil has been completely removed from
the sample.6 Sequentially stronger solvents may be
required to clean the sample.

Oileld Review

Another distillation method, Dean-Stark

extraction, is an industry standard method for
determining uid saturation (right). The core
analyst rst weighs the sample on an analytical
balance before placing it into a thimble in the
Dean-Stark apparatus above its heating ask.
The ask is heated to raise the solvent temperature to its boiling point, and the sample becomes
enveloped in solvent vapors as they rise from the
ask. Water in the sample is vaporized by the solvent and rises with the solvent vapors to the condenser. There, the vaporized water and solvent
cool and condense before falling into a calibrated
receiving tube.
The water, denser than solvent, settles to the
bottom of the receiving tube. When the solvent
condensate overows the tube, it drips down onto
the sample. The condensate mixes with oil in the
rock, and this mixture drips back into the ask
below, where the solvent is again heated and the
vaporization-condensation cycle continues. Once
the volume of water in the receiving tube reaches
a constant value, with no more water produced
from the sample, the Dean-Stark distillation is
complete. Because the sample may not be completely cleaned of oil and salts, a Soxhlet cleaning often follows the Dean-Stark method before
the sample is placed in the oven to dry.
The core analyst weighs the sample after
Dean-Stark extraction and Soxhlet cleaning
and periodically during drying (right).7 The difference between sample weights before and
after cleaning is attributed to the weight of the
extracted fluids. The calibrated receiving tube
measures the extracted water volume, which is
converted to weight by using the density of distilled water. The remaining weight difference
is a result of any oil that has been extracted.
Typically, an oil density value is assumed for
determining the oil volume based on weight.
4. Almon WR: Overview of Routine Core Analysis, in
Morton-Thompson D and Woods AM (eds): Development
Geology Reference Manual, Part 5Laboratory
Methods. Tulsa: The American Association of Petroleum
Geologists, AAPG Methods in Exploration Series, no. 10
(October 1, 1993): 201203.
5. Macini P and Mesini E: Petrophysics and Reservoir
Characteristics, in Macini P and Mesini E (eds):
Petroleum EngineeringUpstream, Encyclopaedia of Life
Support Systems (EOLSS) 2008, developed under the
auspices of the United Nations Educational, Scientic
and Cultural Organization, EOLSS Publishers, Oxford,
England, http://www.eolss.net (accessed July 16, 2013).
6. American Petroleum Institute (API): Recommended
Practices for Core Analysis. Washington, DC: API
Exploration and Production Department, Recommended
Practice 40, Second Edition, February 1998.
7. The core is dried in an oven until its weight is constant
over a specied time interval, which implies that all the
water has evaporated. Normally, a convection or vacuum
oven is used to dry the samples. However, if the cores
contain gypsum or hydratable clays, then they are dried
in an oven equipped with a water vapor injection system
to regulate relative humidity.

Summer 2013

Desiccant

Condenser

Water
trap

Adapter

Thimble
basket
support
Extraction
thimble

Core
Distillation
flask
Heating
mantle

> Dean-Stark apparatus. A core analyst inserts a core plug into the sample chamber (photograph). The
typical setup (left) consists of an electric heating element, or heating mantle, a boiling ask with
extractor chamber, a sample thimble or support screen, a water trap or calibrated receiving tube, and
a condenser. The Dean-Stark method results in a quantitative measure of water volume extracted from
a core, and therefore each sample is cleaned individually in a separate apparatus.

> Weighing core plugs. Precision weighing of every sample at each stage in the cleaning and
extraction process is required because small differences in weight affect grain density calculations
and subsequent determination of other important reservoir parameters such as uid saturation.

21

Later, analysts measure the pore volume of the

core sample; the difference between the pore
volume and the sum of the water and oil volumes is the gas volume. These fluid volumes
are converted to saturations by dividing by the
pore volume.
Laboratories sometimes use other cleaning
and extraction techniques to accommodate different types of rock. Analysts developed a technique for cores containing very ne clays with
delicate mineral structures. The cores are
cleaned with a series of mutually miscible solvents, injected in sequence such that each solvent displaces a specic pore uid and each
solvent in the sequence is displaced by the next.
During ow-through cleaning, solvent may be
either injected continuously or periodically
halted to allow it to permeate the core.
For quick processing of cores, laboratory
workers may use a rapid extractor that injects
heated solvent into the sample. Multiple cores
can be analyzed in a single operation; each sample is placed into a separate pressure vessel, then
the rapid extractor heats and pumps solvent into
the samples at high pressure. The displaced uids are collected separately for each sample.
Key Measurements
Porosity and permeability are essential measurements for understanding how a reservoir will produce. Porosity, a measure of reservoir storage
capacity, can be determined by measuring grain
volume, pore volume and bulk volume (below).
Only two of the three volumes are required to
determine porosity, and pore volume is measured
under simulated overburden stress conditions.

= Vp /Vb ,
= (Vb Vg )/Vb ,
= Vp /(Vp +Vg ),
where

= porosity
Vp = pore volume
Vb = bulk volume
Vg = grain volume

> Porosity relationships. Porosity is dened as the

ratio of pore volume to bulk volume. Because bulk
volume is the sum of grain volume and pore
volume, measuring any two of these volumes
allows for calculation of the third, with
subsequent calculation of porosity.

22

Valve

Pressure
transducer

Valve

Pressure
transducer
V1

Pressure
transducer
Valve

V2

Valve
Vent

He
Core plug
Reference
cell

Sample
cell

Helium
tank

> Boyles law porosimeter. A porosimeter (top) measures the pressure difference between a reference
chamber and a sample chamber to determine pore and grain volumes. The basic system diagram
(bottom) shows the inner workings of a porosimeter, with its reference chamber of xed, known
internal volume and a sample chamber. The device also has valves to admit a gas under pressure to
each chamber, transducers to measure pressure and requisite plumbing to permit communication
between a pressurized gas container and the two chambers. Calibration, operation of valves and
calculation of results are completely automated. (Photograph courtesy of Coretest Systems, Inc.)

Over the years, scientists have developed various methods for measuring these core volumes;
most are based on physical measurements of
weight, length, volume or pressure. Some of these
measurements are obtained directly from the
sample; others rely on the displacement of uids.
Direct measurements may be taken to determine bulk volume. The core analyst may simply
use a digital caliper or micrometer to measure
the core plug length and diameter. A minimum of
ve measurements is recommended. The crosssectional area of the core plug is calculated from
the average diameter, then multiplied by the
average length to yield bulk volume.8 In some
laboratories, digitally calipered core measurement data are automatically logged into a computer, which calculates the geometric bulk
volume, shape factor, effective ow area and caliper bulk factor.
Other techniques are based on Archimedes
principle of uid displacement: A solid submerged completely in uid displaces an amount
of uid equal to its volume. Displacement can be
measured volumetrically or gravimetrically.
The volumetric approach for nding bulk volume uses a small amount of mercury in a porosimeter.9 First, the empty sample chamber of the
porosimeter is lled with mercury to determine
its volume. The mercury is then drained from the
chamber and the core plug is inserted. The cham-

ber is again lled with mercury. The volume of

mercury that lled the empty chamber minus the
volume needed to ll the chamber while it held
the sample equals the samples bulk volume.
The gravimetric approach uses a beaker of
mercury placed on a laboratory balance. After
the beaker and mercury are weighed, a cleaned,
dried core plug of known weight is submerged in
the mercury. The weight gain from the sample
submersion, divided by the density of mercury,
gives the bulk volume. Today, many laboratories
prefer not to use mercury and instead apply
Archimedes principle using other uids such as
brine, rened oil or toluene.10
After determining bulk volume, the analyst
measures the grain and pore volume of the samples. The most rapid and widely used device for
measuring grain volume and pore volumeand
hence, determining porosityis the automated
porosimeter (above). This device employs
Boyles law to calculate porosity based on the
pressure decrease measured when a known
amount of uid is vented into an expansion
chamber containing a core. In this case, the
uid is helium gas.11
To measure pore volume, the analyst places a
cleaned and dried core sample in a core holder
tted with an elastomer sleeve. When air pressure is applied to the outside of the sleeve, it conforms to the shape of the core. The core holder is

Oileld Review

Pi Vi = Pf (Vi + Vl + Vp ) ,
where
Pi = initial pressure
Pf = final pressure in the system
Vi = initial volume in reference chamber
Vl = volume of connecting lines
Vp = pore volume of sample

> Pore volume calculation. Following Boyles law,

the pore volume can be calculated using the
difference between initial and nal pressures in
the porosimeter.

used in place of the porosimeter sample chamber. The reference chamber is initially isolated
from the core in the holder and lled with helium
to a specied pressure. The valve to the sample
chamber is then opened to permit the helium
pressure to equilibrate between the reference
chamber and the pore volume of the conned
sample. Porosity is calculated using the bulk and
pore volume measurements (above). The process
for measuring grain volume is similar, except that
the sample is not conned, but is placed, with no
sleeve, directly into the sample chamber.
Permeability, the measure of a rocks capability to transmit uids, is another key reservoir
characteristic. In the laboratory, analysts determine permeability by owing a uid of known
viscosity at a set rate through a core of known
length and diameter then measuring the resulting pressure drop across the core. For routine
core analysis, the uid may be air, but is more
often nitrogen or helium, depending on the type
8. API, reference 6.
9. Mercury is used because it is a nearly perfect
nonwetting uid and does not enter the rock pores
under normal pressure.
10. API, reference 6.
11. Helium is used because it is an inert gas that does not
readily adsorb onto mineral surfaces of the core and
tends to exhibit ideal gas behavior at moderate
pressures and temperatures. Furthermore, the small size
of the helium atom enables it to rapidly enter the
micropore system of the core, penetrating very small
pores approaching 0.2 nm.
For more on porosity analysis: Cone MP and Kersey DG:
Porosity, in Morton-Thompson D and Woods AM (eds):
Development Geology Reference Manual, Part 5
Laboratory Methods. Tulsa: The American Association of
Petroleum Geologists, AAPG Methods in Exploration
Series, no. 10 (October 1, 1993): 204209.
12. API, reference 6.
13. Klinkenberg LJ: The Permeability of Porous Media to
Liquids and Gases, Drilling and Production Practice,
(1941): 200213.
Rushing JA, Newsham KE, Lasswell PM, Cox JC and
Blasingame TA: Klinkenberg-Corrected Permeability
Measurements in Tight Gas Sands: Steady-State Versus
Unsteady-State Techniques, paper SPE 89867,
presented at the SPE Annual Technical Conference and
Exhibition, Houston, September 2629, 2004.

Summer 2013

of permeameter used. The analyst loads a clean,

dry core into a specially designed core holder,
where it is enclosed in a gas-tight elastomer
sleeve (below). The permeameter forces pressurized gas through the inlet port and into the core.
The pressure differential and ow rate are
metered at the outlet port. This conguration is
used in a steady-state gas permeameter.
In an alternative method for determining
permeability, analysts charge a chamber to a
high gas pressure and then open a valve, allowing the gas to pass through the core plug as the
pressure declines. If the laboratory uses this
unsteady-state, or pressure-transient permeameter, analysts can use the time rate of change of
pressure and efuent ow rate to solve for the
plug permeability.
Analysts apply corrections to compensate for
the differences between laboratory and downhole conditions.12 They account for stress differences by applying conning stress to one or more
representative plug samples; some permeameters are capable of imposing conning pressures
to 70 MPa [10,000 psi]. Often, analysts use several conning stresses to determine the stress
effect on permeability then apply a correction

factor for the reservoir conning stress to the

other routine permeability measurements.
Gas ow in pores differs from liquid ow
because the ow boundary condition at the pore
walls is different for gases and liquids. Liquids
experience greater ow resistance, or drag, at
the pore wall than gas does. This gas slip effect
can be corrected by increasing, in steps, the
mean gas pressure in the plug, which increases
the drag at the pore wall. The Klinkenberg correction is an extrapolation of these measurements to innite gas pressure, at which point gas
is assumed to behave like a liquid.13
Analysts apply an additional correction for
high gas ow rates through tortuous ow paths.
The Forchheimer correction accounts for effects
produced when gas accelerates as it passes
through small pore throats and decelerates upon
entering the pores. In many automated unsteadystate permeameters, both the Klinkenberg and
Forchheimer corrections are automatically
solved during analysis.
To flowmeter
Outlet port
Metal cap

Rubber disk
Elastomer sleeve
Core

High air pressure

port (sealing)

Inlet port
Low air pressure (flow)

> Hassler chamber for measuring permeability to gas. A core sample is placed in an elastomer sleeve.
The caps at either end of the device are tted with axial ports to admit gas. The permeameter
(photograph) forces gas through the inlet port at the bottom, and the gas passes through the core
before exiting to a owmeter. Permeability is calculated using the Darcy equation. (Photograph
courtesy of Coretest Systems, Inc.)

23

100

100
Carbonate

DRFT-IR, wt %

DRFT-IR, wt %

Clay

50

50
DRIFTS, wt %

50

100

100

2.5
DRIFTS, wt %

5.0

Kerogen
Total organic carbon 1.2, wt %

DRFT-IR, wt %

100

5.0
Quartz

50

50
DRIFTS, wt %

50
DRIFTS, wt %

100

2.5

> Conrmation of DRIFTS measurements. Results from the DRIFTS measurement in the vertical
evaluation well compare favorably with companion DRFT-IR mineralogy for clay, carbonate and quartz
content. The kerogen content from DRIFTS was compared with a LECO TOC measurement. DRIFTS
measures wt % of the kerogen, which includes other elements than carbon; therefore, the industry
uses a factor of 1.2 to correlate between these TOC and kerogen measurements. The above plots show
good agreement between DRIFTS and the other measurements.

Upon completion of its analyses, the laboratory transmits a report, along with digital copies
of photographs and scanning data, to the client.
Depending on client instructions, the core may
be kept in storage, returned to the client, or
archived at a core library for future reference.
Petrographic Measurements
Routine core analysis helps operators evaluate
reservoir lithology, bedding features, residual uids, porosity and permeability, but these provide
only a portion of the information that can be
extracted from a core. Complementary petrographic tests furnish additional analytical results
and visual records of the core.
Scanning electron microscopy allows
inspection of core surface topographies with
magnifications capable of resolving features at
the nanometer scale. A scanning electron
microscope scans the surface of a sample with
a finely focused electron beam to produce an
image based on beam-specimen interactions.

24

Electron detectors receive data on specimen

surface topography while backscatter detectors resolve compositional variations across
the sample surface.
Analysts use a color cathodoluminescence
detector to examine variations in mineral composition, including phase and trace element distribution. This detector permits visualization of
chemical overprinting and overgrowths, growth
zonation and internal healed fractures. These
images provide insight into processes involving
the growth of mineral crystals as well as their
replacement, deformation and provenance.
Petrologic applications include investigations of
cementation and diagenesis of sedimentary rocks,
provenance of clastic rock materials and examination of the internal structures within fossils.
Diffuse reectance infrared Fourier transform spectroscopy (DRIFTS) provides a t-forpurpose technique for measuring mineralogy and
organic contentdata that are proving instrumental in supporting completion designs in mud-

stone reservoirs. Scientists can analyze core,

cuttings or outcrop samples. DRIFTS analysis is
fast; a 50-second scan determines mineralogy
and organic content. A sample preparation procedure that is proprietary to Schlumberger
enables it to be used as a quantitative measure of
rock constituents. The process requires a small
sample of only 5 g [0.18 ozm] for analysis.
The device scans the sample with infrared
light at multiple wavelengths. The light scatters
as it passes through the rock. The reected infrared energy is analyzed based on regression analysis of the frequency and amplitude of the spectra
to determine the lithology, mineralogy and
organic content of each sample.
To obtain an initial calibration of mineralogy
and kerogen measurements, analysts run dual
range Fourier transform infrared (DRFT-IR)
spectroscopy and X-ray uorescence (XRF) on
representative samples to verify mineralogy,
and they run a LECO total organic carbon (TOC)
scan to verify organic matter content and kerogen amount. After the full study of DRFT-IR,
XRF and TOC is complete, DRIFTS data can be
acquired on a rapid basis to give operators
timely information for completion decisions.
An operator in the western US performed
DRIFTS analyses on Mancos Shale wells to evaluate their utility for identication of mineralogy
and kerogen content in unconventional plays.
The rst study focused on a vertical well that was
cored and logged. This well provided an extensive
database for evaluating the Mancos Shale.
The laboratory used crushed samples from
cores to determine mineralogy and organic content. In addition to DRIFTS data, the evaluation
included mineralogy from DRFT-IR, XRF and
TOC measurements. These traditional analytical
methods utilized material from the same crushed
samples that were subjected to DRIFTS analysis.
Results from these evaluations showed good
agreement between the DRIFTS results and the
other three more accurate analytical methods
(above left). The results gave the operator condence to apply the DRIFTS method to other wells
drilled in the Mancos formation.
The second well in this study was a horizontal
production well. Near the heel of the well, high
gamma ray measurements and low mud log gas
readings suggested that the well might be out of
the primary target zone. The gamma ray response
decreased as the well moved upward in the stratigraphic section, indicating better reservoir quality in the targeted portion of the well; this was
consistent with mud log gas readings toward the
toe of the well.

Oileld Review

5,720
ft

True Vertical Depth

85 5,800
TG units

Total Gas

300 0
gAPI

Gamma Ray

Z,500

Z,250

Z,000

Y,750

Y,500

Y,250

Y,000

X,750

X,500

X,250

X,000

W,750

W,500

W,250

W,000

V,750

V,250

V,500

0
ft

Depth

30

40

50

60

70

80

10
8
6
4
2

Average, Rate, bbl/min

Maximum Pressure, 1,000 psi

14

13

12

11

10

40

50

15

Total Quartz

50

Total Clay

DRIFTS Composition, %

100

20

30
10
Total Kerogen

125 150
100
75

Tracer Concentration,
parts per billion (ppb)

50

Total Carbonate

25

Sand Pumped, 1,000 lbm

Stage

> Response in a bentonite zone. The gamma ray log (Track 3 from top) reads high from V,100 to V,400 ft measured depth, interpreted as a bentonite zone (red
box). The well trajectory (Track 1) was changed to drill upward through this zone. The high gamma ray signature, combined with low mud gas readings
(Track 2), is a typical indicator of poor reservoir quality. All zones of this horizontal well were stimulated. In the bentonite zone, the injection rate required to
fracture was higher at the same pressure (Track 4). A tracer survey (Track 5) shows the presence of a chemical tracer throughout the 15-stage interval,
indicating that the proppant was successfully placed at each stage. The DRIFTS analysis (Track 6) indicates that kerogen is present throughout this zone,
which gave petrophysicists condence that the bentonite zone (yellow box) was a thin zone in a productive Mancos formation, despite high gamma ray and
low mud gas readings.

In the toe of the wellbore, the DRIFTS data

recorded values above the operators minimum
cutoff for kerogen content along with acceptable
values of clay, carbonate and quartz, in agreement with gamma ray and mud gas data (above).
From the heel of the well, other data showed
unacceptably high clay content, supporting the
earlier gamma ray interpretation that the well
had drilled out of zone. However, the DRIFTS
data showed anomalously high kaolinite values at
the heel of the well, typical of thin bentonite layers known to exist in the Mancos Shale. The heel
also exhibited higher kerogen content (5.6% compared with a range from 3.1% to 4.3% in the toe of
the well). These observations, not available
through MWD logs obtained in this well, gave the
operator condence that the bentonite zone was
a thin anomaly, not representative of the Mancos
formation surrounding it.

Summer 2013

Postfracturing data indicate that the well was

successfully stimulated with adequate proppant
placement, although higher injection rates were
needed to successfully break down the clay-rich
bentonite zone. The operator is incorporating
these data in completion optimization studies for
future wells.
Operators combine other data with these
important ground truth measurements in their
formation evaluation programs. Core analysis, in
its many forms, will continue to inform operator
decisions to drill ahead, abandon or complete
their wells.
In some cases, routine analysis and petrology
provide an operator with all the core information
needed. More commonly, additional analyses are
obtained from this valuable asset. These include
evaluations of multiphase saturation and ow
properties, such as capillary pressure and relative

permeability; log-tuning measurements, such as

electrical properties for determining porosity and
saturation from logs; ow assurance studies; geomechanical measurements or enhanced oil recovery evaluations. These measurements add
tremendous value to reservoir evaluation, and
they all begin with routine core analysis. MV

25

Multistage Stimulation in Liquid-Rich

Unconventional Formations

Production of liquids from shale formations, pioneered in North America, has grown
exponentially in the past decade. The economics of these plays, however, remain
sensitive to prices and demand, thus operators and service companies must continually
develop more-efcient methods of recovering these once-overlooked hydrocarbons.

Isaac Aviles
Jason Baihly
Sugar Land, Texas, USA
Guang Hua Liu
CNPC-Dagang Oileld Company
Tianjin, Peoples Republic of China
Oileld Review Summer 2013: 25, no. 2.
Copyright 2013 Schlumberger.
For help in preparation of this article, thanks to
Amy Simpson, Houston.
Copperhead, DiamondBack, Falcon, KickStart, nZone,
PowerDrive Archer and Spear are marks of Schlumberger.
1. For more on kerogen and oil shales: Allix P, Burnham A,
Fowler T, Herron M, Kleinberg R and Symington B:
Coaxing Oil from Shale, Oileld Review 22, no. 4
(Winter 2010/2011): 415.
2. Baihly J, Altman R and Aviles I: Has the Economic Stage
Count Been Reached in the Bakken Shale?, paper
SPE 159683, presented at the SPE Hydrocarbon,
Economics and Evaluation Symposium, Calgary,
September 2425, 2012.
3. Jabbari H and Zeng Z: Hydraulic Fracturing Design for
Horizontal Wells in the Bakken Formation, paper
ARMA 12-128, presented at the 46th US Rock Mechanics/
Geomechanics Symposium, Chicago, June 2427, 2012.
4. Baihly et al, reference 2.
5. Martin R, Baihly J, Malpani R, Lindsay G and Atwood WK:
Understanding Production from the Eagle FordAustin
Chalk System, paper SPE 145117, presented at the
SPE Annual Technical Conference and Exhibition,
Denver, October 30November 2, 2011.
6. Martin et al, reference 5.
7. For more on PowerDrive Archer technology: Felczak E,
Torre A, Godwin ND, Mantle K, Naganathan S,
Hawkins R, Li K, Jones S and Slayden F: The Best of
Both WorldsA Hybrid Rotary Steerable System,
Oileld Review 23, no. 4 (Winter 2011/2012): 3644.
For more on the Spear bit: Centala P, Challa V,
Durairajan B, Meehan R, Paez L, Partin U, Segal S, Wu S,
Garrett I, Teggart B and Tetley N: Bit DesignTop to
Bottom, Oileld Review 23, no. 2 (Summer 2011): 417.
8. Baihly et al, reference 2.

During the past decade, oil companies have

performed thousands of hydraulic stimulations on intervals along horizontal wellbores drilled through ultralow-permeability
formations. Operators are using these techniques to exploit organic-rich shales, which
were traditionally viewed only as source rock for
conventional reservoirs.
These extremely tight sedimentary formations differ signicantly from oil shales, which
are sedimentary rocks containing kerogenpartially degraded organic materialthat has not
yet matured enough to generate hydrocarbons.1
In contrast, as a result of the pressure and heat of
the burial process, the kerogen in gas- and liquidrich shales has matured sufciently to generate
signicant amounts of gas and oil, which remain
trapped within the shale.

Numerous liquid-rich shale formations exist

in North America. Among these are the Bakken
and the Eagle Ford formations. Unlike in other
unconventional plays around the world, operators and service companies have years of experience working in these two extensive plays. The
formations are familiar to petrophysicists and
engineers and have been the proving grounds for
much of the technology now used to exploit
unconventional liquid-rich reservoirs.
Covering an area of 780,000 km2 [300,000 mi2],
the Bakken formation lies within the Williston
basin of North Dakota, South Dakota and
Montana in the US and in parts of Manitoba and
Saskatchewan in Canada (below).2 Operators
rst produced oil and gas from this formation in
the early 1960s through conventional vertical
wells. In the 1980s, production increased when

Saskatchewan
CANADA

Manitoba

Regina

Bakken Formation

USA
Montana

North Dakota
Billings

Bismark
Williston Basin

Wyoming

South Dakota

> The Bakken. The Bakken formation (pink) covers an area of more than 780,000 km2 across the states of
Montana and North Dakota in the US and parts of the Canadian provinces of Manitoba and Saskatchewan.

26

Oileld Review

operators began drilling horizontal wells.3 When

operators combined the complementary technologies of horizontal drilling and hydraulic stimulation to maximize the amount of formation
exposed to the wellbore, North Dakota production from Bakken elds rose dramatically, from
16,000 m3 [100,000 bbl] per day in 2005 to
96,000 m3 [600,000 bbl] per day in 2012.4
These increased production rates in North
Dakota led geoscientists to consider using the
same techniques to produce oil from source rock
in other existing plays, including the Eagle Ford
Shale, in Texas, USA, which is the source rock for
the massive hydrocarbon accumulation that has
produced from the Austin Chalk for 80 years.
That trend overlies the Eagle Ford Shale across
large swaths of south Texas (right).5 The Eagle
Ford play, which stretches from central Texas
southwest into Mexico, is 160 km [100 mi] long
and averages 100 km [60 mi] wide.6
In an effort to help operators optimally
exploit unconventional plays, service companies
have rened certain critical technologies. Today,
operators are able to drill long horizontal wells
and place them accurately within formation
sweet spots. Production and completion engineers have also sought to improve methods for
stimulating the numerous potentially productive
intervals pierced by these wells (see Stimulation
Design for Unconventional Resources, page 34).
Renements to directional drilling assemblies such as the PowerDrive Archer rotary steerable system have led to more efcient drilling
through higher build rates and improved rates of
penetration. In addition, engineers have designed
bits specically for use in shale formations. The
Spear steel bit from Smith Bits, a Schlumberger
company, is designed to meet the demands
unique to rotary steerable systems drilling in
shale formations.7
In ultralow-permeability formations, operators nearly always use multistage stimulation
(MSS) techniques to access commercial volumes of oil, condensate and dry gas. These
methods enable engineers to stimulate multiple intervals along horizontal sections.
Typically, completion engineers isolate individual intervals and, either through perforating or by opening sliding sleeves, expose the
zone to be treated. The well is then hydraulically stimulated. Engineers repeat this
sequence, moving upward along the wellbore
until all targeted zones have been stimulated.
This article examines various MSS methods.
Case histories from the US and China illustrate
their use and advantages.

Summer 2013

USA
Texas

Texas

MEXICO

Eagle Ford
Shale Formation

Oil window
Wet gascondensate window
Dry gas window

lf of
Gu

Mexico

km 100

mi

100

> The Eagle Ford. The Eagle Ford Shale formation, which is the source rock
for the Austin Chalk play, covers a large swath across southern Texas and
runs parallel to and north of the Gulf of Mexico coastline. The burial process
of the Eagle Ford formation has resulted in a trend of oil (green), wet gas and
condensate (yellow), and dry gas (blue) from the northwest to southeast.

Balls, Seats and Valves

As the industry improved its ability to drill horizontally, wellbore lengths increased. So too did
the number of intervals that had to be isolated
and treated. In 2007, the average treatment number, or stage count, in Bakken wells was three.
By the end of 2011, that number was nearly 30,
and some wells had more than 40 stages in a
single lateral.8
While the economies of scale seemed to dictate treating as many intervals as possible per
wellbore, operators sought to improve well economics further by reducing the time required to
stimulate all the stages in a given well. Exploiting
a liquid-rich shale play is drilling intensive, and
despite the advantages of hydraulically treating
longer wells, the drainage area of each wellbore
in these tight formations is limited. With more
than 200 rigs working in the Bakken at the end of
2011, there was great economic incentive to move
the rigs off one well and on to the next as quickly
as possible.

Traditionally, stimulating multiple zones in a

conventional vertical well involved perforating
the lowest zone, retrieving the perforating guns
and pumping the treatment. The operator owed
the well back to drain extra proppant and carrying uids and to force closure of the propped
fracture. The completion engineer then set a
bridge plug to isolate the interval from those
above, pulled out of the hole to pick up perforating guns and repeated the process. Once all zones
were treated, drillers milled out or retrieved the
plugs and brought the well on line. Often, the
well had to be completed with multiple strings of
tubing or isolation valves to prevent crossow
between zones having different pressures. While
this was a time-consuming process, it was not
economically prohibitive in a vertical well with
only two or three stages.
However, when dozens of intervals in each
wellbore needed treatment, operators sought

27

Perforations

Heel

Toe
Frac plugs

> Plugging and perforating procedure. In a typical cemented and cased well plug and perf scenario,
the deepest interval at the toe of the well is perforated and treated first. A plug is then set above the
perforation cluster. The next stage is treated, a plug is set, perforations are added and the process is
repeated until all intervals are stimulated. The driller mills the plugs using coiled tubing or a
conventional drillstring. The operator then commingles production for all intervals.

to reduce the time required between reaching

total depth and initial production. In response,
service companies developed more-efficient
treatment methods that relied on external
packers, balls and seats, or plugs, to isolate
and treat intervals. They also developed valves
that, in some circumstances, could be substituted for perforations.
Today, most horizontal wells are completed
so that each interval can be isolated and perforated in one interventionusing pumpdown
wireline or coiled tubing conveyanceand then
treated. A final intervention may be required to
mill out isolation plugs. Because the intervals
are in one zone and equally pressured, the well
is ready to produce.

Typically, service company completion specialists use plugs or ball-and-seat systems to isolate each stage. When the company opts for a
plug, it is placed on wireline and pumped down
the hole, or less commonly, it is run and set on
coiled tubing. The assembly includes perforating
guns. Once the plug is set above the topmost perforation cluster of the previous stage, the completion team pulls the guns into position. Each
cluster of the next stage is then perforated, and
the tools, along with the spent guns, are retrieved.
Next, the team stimulates the open interval,
and this plug and perf procedure is repeated
(above). When all intervals have been treated,
the driller mills out the plugs, and production
from all intervals is commingled.

Closed Position

Heel

Valve seat

Frac sleeve

Toe

Making Good Ideas Better

In the mid-1960s, water depth of more than about
60m [200ft] was considered by the E&P industry
to be a very deep working environment. But

Open Position

Flow port

Ball

Frac sleeve

> Ball and seat. A valve device is run into the hole in the closed position (top). When the ball (bottom,
red) lands in a valve seat in a frac sleeve (green), pressure applied at the surface causes the sleeve to
slide downward and open a flow port, which exposes the interval to be treated. The ball seals against
the valve seat to isolate the previously treated stage below it. This process is repeated for each
stimulation stage.

28

65263schD6R1.indd 28

In other completion designs, a valve containing a ball seat and sliding sleeve is run into the
hole as part of the completion. External packers
isolate each interval. The ball seat is designed to
capture a ball of a specific size that is pumped
into the well. The diameters of the seats become
successively larger from the bottom to the top of
the completion. When the ball lands in the seat,
continued pumping causes pressure to build
against the ball and seat (below left). At a specified pressure, the ball-and-seat assembly moves
downward, which opens a sleeve in the valve to
expose the formation between the external packers. The interval is then treated. The next larger
size ball is then run, isolating the treated zone.
Completion specialists repeat this ball drop
stimulation treatment sequence for all intervals beginning at the toe and moving toward the
heel of the well. The method offers an advantage
over the use of plugs because, as long as the ball
seats do not represent a significant flow restriction, the balls may be flowed back to the surface, obviating the need for and risk of milling.
Additionally, the operation is continuous, thus
less time consuming.
For cemented completions, engineers may perform similar operations using specially designed
valves run as part of the completion string. When
the ball is pumped downhole, it lands and creates
a seal in the seat of the deepest exposed valve,
which results in a closed system. Pressuring the
well causes the sliding sleeve to open, allowing the
interval to be treated directly through the cement.
As a result, the operator does not need to perforate
the casing and cement first.
Despite the success of these systems, operators still seek hedges against narrow profit margins and unpredictable commodity prices that
govern the economics of unconventional plays. In
an effort to cushion profit margins, service companies are working with operators to refine MSS
practices and tools and to shave costs and risks
from well completion operations while simultaneously increasing production rates and ultimate
recovery from these wells.

9. Raulins GM: Well Servicing by Pump Down

Techniques, paper OTC 1016, presented at the First
Annual Offshore Technology Conference, Houston, May
1821, 1969.
10. Stegent N and Howell M: Continuous Multistage
Fracture-Stimulation Completion Process in a Cemented
Wellbore, paper SPE 125365, presented at the SPE
Eastern Regional Meeting, Charleston, West Virginia,
USA, September 2325, 2009.

Oilfield Review

8/19/13 6:18 PM

operators were already contemplating the implications of servicing subsea wells completed with
wellheads on the seaoor in water as deep as
3,650 m [12,000 ft]. To address the challenges of
deep water, engineers developed several technologies, including pumpdown for interventions
traditionally performed using slickline.9
Pumpdown systems convey tools downhole
using uid pressure. When uid is pumped
against mandrels equipped with swab cups, tools
move up or down the tubing. Because this system
requires the uid to circulate, designers created
a crossover port that allowed circulation between
the tubing and the annulus.
Today, completion engineers apply this
method to push plugs and perforating guns
attached to electric line to depth in horizontal
wells. Service technicians set the plug above the
shallowest perforation cluster of the previous
fracture stage, detach the perforation guns from
the plug assembly and move up the hole to create
the next perforation cluster. After the guns have
been red, they are retrieved to the surface, and
the interval stage is stimulated. The process is
repeated for each stage. Once stimulation operations are complete, the driller must mill out each
plug before putting the well on production.
In this form of MSS, the last stepmilling
is often the most difcult and time-consuming
portion of the operation in high-angle wells
because weight on bit is limited. Engineers have
developed plugs of varying design and material
that are able to withstand stimulation pressures
while at the same time are more easily ground
into cuttings than are traditional cast-iron bridge
plugs; these cuttings are small enough to be circulated out of the hole.
The aluminum Copperhead drillable and owthrough fracture plug and the DiamondBack
composite fracture plug are examples of these
new plugs (above right). The former is rated to
103.4 MPa [15,000 psi] and 205C [400F] and is
designed to withstand multiple pressure and
temperature cycles. The latter may be used when
downhole conditions are less extreme and is
rated for pressures up to 68.9 MPa [10,000 psi]
and temperatures up to 177C [350F].
Both plugs are signicantly easier to mill than
are standard cast-iron plugs. Researchers have
also developed a mill specically for drilling out
the Copperhead plug. The new mill reduces milling time and creates smaller cuttings. Because
the DiamondBack plug is constructed of a composite material that is considerably softer than
metal plugs, it is easily and quickly milled using
standard mills.

Summer 2013

Copperhead Plug

DiamondBack Plug

Aluminum slip pads

with cast-iron facing
Shear
ring

Aluminum slip pads

with cast-iron facing

Sealing aluminum
backup ring
Element backup
Shear
ring

Aluminum slip pads

with cast-iron facing

Aluminum slip pads

with cast-iron facing
Pumpdown ring

> Bridge plugs. The Copperhead plug body (left) includes slip pads constructed
of aluminum with cast-iron facing, which helps prevent cracks in the slips
when deployed in hard casing. The Copperhead plug includes a shear ring
embedded in the slips (not shown) to help ensure that presetting does not
occur and an element backup that enhances seal effectiveness while the
bridge plug is exposed to multiple pressure changes. Because the plug body
and slips are composed mainly of aluminum rather than cast iron, when the
plug is milled, its cuttings are more easily owed from the well. The
DiamondBack plug (right) consists of composite material. Like the Copperhead
plug, the DiamondBack plug includes an internal shear ring to prevent
presetting. It also features rigid slips and a pumpdown ring to minimize uid
use. Because the plug is composed of a composite material, it can be quickly
milled with a standard mill.

In addition, both plugs are designed to prevent

premature setting, which can be a problem with
plugs that are landed and set on slips intended to
wedge against the casing wall. If plugs of this
design are conveyed downhole at excessive speeds,
the slips may overtake the lower mandrel on the
plug, causing them to expand against the casing
wall and the tool to set. To help prevent this problem, both plugs use shear rings to hold the slips in
place until at least half the set-down weight is
applied to the assembly. This signicantly reduces
the chances that the plugs will be set prematurely
even when they are run or pumped into the well at
relatively high speeds.
When developing the pumpdown concept,
designers incorporated a circulation path for
fluids exiting the tubing and returning to the

surface via the casing annulus. This process is

not possible in cemented horizontal wells
because until the well is perforated, the well is
a closed system. Therefore, during plug and
perforation operations using the pumpdown
technique in a cemented horizontal wellbore,
the first set of perforating gunsthose at the
toe of the wellmust be conveyed on coiled
tubing, wireline tractor or drillpipe. Service
industry experts have tried several methods to
avoid this costly step, including overdisplacing
the cement to leave a flow path open through
the casing shoe. For numerous reasons, including the inability to get a pressure test of the
casing and cement, most operators deemed
this wet shoe solution unacceptable.10

29

Fracture ports

Piston

Rupture discs

> First stage valve. The KickStart rupture disc valve eliminates an intervention
during MSS operations by facilitating circulation at the toe of horizontal
wells. The valve is part of the casing string and is cemented in place along
with the casing. After the casing is pressure tested, the well is pressured to
some value higher than the test pressure to rupture the discs and open the
valve. The fracture ports are designed to ensure that at least one opening is
within 3 of the minimum stress direction of the formation to be stimulated.

Schlumberger researchers addressed the

problem of opening this closed system without
mechanical intervention through the KickStart
pressure-activated rupture disc valve for
cemented multistage fracturing. The valve is
run as part of the casing string one or two joints
above the oat shoe. Its internal diameter is
nearly equal to that of a 4 1/2- or 51/2-in. casing,
which allows it to accommodate standard
cement wiper plugs and requires no change to
cementing procedures. The valve includes two
discs, but only one must rupture for the stimulation to be successful.
Once the disc ruptures, a helical port pattern
made up of seven 6-in. long ports with 15 phasing is open to the cement sheath through which
the interval is stimulated (above). Following
numerous iterations in the laboratory, mathematical modeling and nite element analysis, designers arrived at this conguration, which minimized
fracture initiation pressure through the cement

30

while promoting a single vertical fracture in the

cement. Experts have expressed doubts about
the efcacy of fracture stimulations performed
through ports instead of perforations, but the
KickStart valve arrangement allays those concerns by ensuring that at least one of the fracturing slots is no more than 3 from one of the
minimum stress points on the hoop stress envelope of the wellbore. The total area of all the
ports is 69 cm2 [10.7 in2], which is the equivalent
of six 0.6-m [2-ft] long perforation clusters with
19 shots per meter [6 shots per ft].11
After the casing string is cemented in place
and the casing has been pressure tested, the
driller increases pump pressure to a predetermined level, which ruptures the discs in the
KickStart valve. This exposes the cement in the
annulus through which the formation is stimulated. The nal step of the treatment is to pump a
ush uid, which can also be used as the pumpdown uid for the next plug and perforation gun
assembly. Then, the remaining intervals may be

plugged and perforated using standard pumpdown practices.

Cabot Oil & Gas Corporation, a long-time
Eagle Ford Shale operator, has implemented
numerous innovations in the play, including
reducing well spacing without sacricing wellbore length. In one campaign, the completion
engineer used the rupture disc valve to stimulate
the toe section of the formation in more than a
dozen wells. Typically, the wells in the Buckhorn
area of the shale play are drilled in 5,500-ft
[1,676-m] laterals and are stimulated in 14 to
20 stages.
The operator tested its casing to 10,000 psi
[68.9 MPa] with the KickStart valve discs set to
rupture when the pressure was 10,600 to
10,800 psi [73 to 74.5 MPa]. Cabot engineers routinely pumped the rst treatment consisting of
more than 250,000 lbm [113,000 kg] of proppant,
at 65 bbl/min [10 m3/min] through the valve
ports. When engineers compared the results of
the section treated through the valve with those
treated through perforations, they found that the
pump pressures, rates and volumes all compared
favorably. They also concluded that the KickStart
valve saved the operator more than US$ 100,000
per well by eliminating the coiled tubing intervention to perforate the interval at the toe of
the well.12
Savings per well is critical for operators
producing from low-permeability formations
because these plays are typically exploited
using many wells that produce at rates near
their economic limit. To make such a strategy
work requires each well to be drilled, completed
and produced efciently. The plug and perforate
procedure with the rst stage performed using
the KickStart rupture disc valves helps operators reach that goal.13
Ball and Seat
In the past decade, operators have come to view
openhole completions in horizontal wells as substantially more cost-efcient than cemented
completions. These systems use hydraulically set
or swellable packers to isolate each interval.
Sliding sleeve valves, run as part of the completion tubulars between packers, are opened by
hydraulic pressure applied to a seal created by a
ball that is dropped from the surface to land in a
mated seat. These seats increase in size from the
smallest in the toe to the largest in the heel of the
well, thus the smallest ball passes through each
seat to the toe and the largest stops in the rst
seat near the heel (next page).

Oileld Review

The industry has embraced these openhole

systems because they may result in certain
advantages over plug and perforate cased hole
completions:
less time-consuming and less expensive completion operations
production from the open hole as well as the
fractures
a simpler connection between the wellbore and
the fractures
wellbore fractures that generate higher early
production.14
These systems also have potential disadvantages. Unlike cased wells that are stimulated
through valves or perforations, openhole stimulations are conned only by packers, which may
leave large sections of formation exposed
between them. As a consequence, the operator
has little control of the fracture location or
number of fractures created in a stage. In addition, as the ball seats decrease in size with well
depth, friction pressures increase, which may
result in higher overall fracture initiation and
extension pressures.15
There may also be problems with the interaction between the ball and seat. When the ball
lands and pressure is applied, the sleeve slides
downward, exposing the annulus for treatment.
The ball and seat then become the barrier isolating the lower intervals of the hole that have been
previously treated. Both these actuator and seal
functions are critical. If the ball fails to create a
seal, the sliding sleeve may not move and the
interval cannot be treated. At the same time, previously treated zones below the seat may be
exposed a second time to stimulation uids and
pressure, which can damage production from
that zone as well.
After all the intervals have been successfully
treated, the balls must ow off their seats and not
impair production. Operators had long assumed
that the balls oated off their seats even though
not all balls were accounted for in a ball catcher
at the surface. The widely accepted explanation
for this seeming discrepancy has been that some
balls ow back to a highly deviated point in the
well where they churn in the ow and smash
against each other until they break into pieces
that are small enough to ow out of the well.16
However, some operators have become sufciently concerned about ball material left in the
well that they routinely mill the seats to make
certain the ow path is clear. One operator
found that after milling the ball-and-seat
sleeves in 10 wells, estimated ultimate recovery

Summer 2013

nesses in the layer bonding that may cause them

to fail under pressure. If they land on the seat in
certain positions relative to the layering, they
may delaminate and break apart.
Schlumberger engineers have incorporated
several solutions to address these concerns in the
Falcon multistage stimulation method for uncemented wells. While testing various ball materials, the engineers also tested seat designs and
discovered that spherical seats far outperformed
typical cone-shaped seats. They also found that a
magnesium alloy was superior to phenolic or
composite ball material.
The lightweight magnesium used in the
Falcon system balls minimizes ball extrusion; in
addition, the balls are temperature insensitive,
ow back intact and do not break on contact with
the seat or during stimulation. They are rated to
68.9 MPa differential pressure and are easily
milled. In one conguration of the Falcon system,
the toe valves have multiple smaller balls that
land in a single seat. These balls are able to easily
pass through the upper seats to reach the lower
sections of the well, but the total ow-through
area remains large enough even at the lowest

increased signicantly; the experiment was

expanded to more than 300 wells.17 But elimination of a coiled tubing intervention to mill out
plugs was one of the original drivers for adopting the ball-and-seat technology, and service
companies have sought to eliminate the possibility of balls staying in place through numerous
methods, including retrievable seats and valves;
like most methods, however, that alternative
also requires a coiled tubing intervention.18
One of the issues with balls not seating or not
oating off the seat after treatment is in the
material used for the ballspredominantly phenolic, composite or metal alloy. These balls must
be light enough to ow out of the well but strong
enough to land in the seat at a high velocity without being deformed or damaged. Some industry
experts believe these lowcompressive strength
balls are breaking before they have a chance to
work. Under pressure, the balls may extrude,
causing them to become stuck in their respective
seats or one of the next seats uphole as they are
owed up the well.
Additionally, some types of these balls are
constructed in layers and have inherent weak-

Openhole
packer

3-in.
frac sleeve

2 1/2-in.
frac sleeve

2-in.
frac sleeve

1 1/2-in.
frac sleeve

> Typical ball-and-seat conguration. Ball-and-seat MSS systems use frac valves, or sleeves, with
seats that decrease in size from heel to toe. This allows the lower valves to be activated by balls (red)
small enough to pass through the upper valves. In long horizontal sections, this can become a problem
as friction pressures increase with wellbore length and decreasing seat diameters.

11. Baihly et al, reference 2.

12. Baihly et al, reference 2.
13. Arguijo AL, Morford L, Baihly J and Aviles I:
Streamlined Completions Process: An Eagle Ford Shale
Case Study, paper SPE 162658, presented at the
SPE Canadian Unconventional Resources Conference,
Calgary, October 30November 1, 2012.
14. Daneshy A: Hydraulic Fracturing of Horizontal Wells:
Issues and Insights, paper SPE 140134, presented at
the SPE Hydraulic Fracturing Technology Conference,
The Woodlands, Texas, USA, January 2426, 2011.

15. Daneshy, reference 14.

16. Baihly et al, reference 2.
17. Wozniak G: Frac Sleeves: Is Milling Them out Worth the
Trouble?, paper SPE 138322, presented at the SPE Tight
Gas Completions Conference, San Antonio, Texas,
November 23, 2010.
18. Grifn J, Barraez R and Campbell S: To Mill or Not to
Mill: A Fully Retrievable Multistage Fracturing System,
paper SPE 163936, presented at the SPE/ICoTA Coiled
Tubing and Well Intervention Conference and Exhibition,
The Woodlands, Texas, March 2627, 2013.

31

point of the well to eliminate the effects of friction pressure on fracture initiation (below). The
material and design of the seats allow them to be
easily and quickly milled.
No Limits
Engineers at Schlumberger have recently developed a variation on ball-activated systems that
may be used in cemented wells. The technique
uses balls or darts to activate sliding sleeves that
provide stage isolation. Because this technique
does not require seats of decreasing diameter to
get the balls to TD, the technique can be used to
stimulate a nearly unlimited number of stages in
a single continuous operation.
The nZone multistage stimulation system
includes a control line connected to sequential
valves that make up part of the completion. To
initiate the stimulation operation, a dart, which
is pumped from the surface, lands on a C-ring
an incomplete circlein the lowermost valve.
The completion engineer then applies pressure
against the dart, which opens the sliding sleeve
and pressurizes the control line. This pressure is
transferred to a piston in the valve immediately
above it, which closes the C-ring, creating an
O-ring with a reduced ID (next page, top).
The rst stage of the stimulation is pumped,
and during the ush stage, another dart is
released. This dart lands on the now-compressed
C-ring, isolating Stage 2 from Stage 1. The resulting increase in pressure forces the sleeve to slide
for Stage 2 and the control line to become pressurized and close the next C-ring, which is then
ready to catch the next isolation dart. Stage 2 is
treated, and during the ush stage, another dart

Single Ball Seat

is pumped. When fracturing operations for all

stages are complete, the well can be produced.
The darts can remain in the well, but to obtain
full wellbore access for future interventions, the
darts must be milled out. Alternatively, the operator can deploy dissolvable darts.
Recently, in an effort to increase production
and reduce completion costs per well, the operator of the Dagang eld in the Huanghua depression of eastern China, which had previously
drilled only vertical wells, changed to horizontal
wells. The rst commercial discovery in this eld
was made in 1963 in the Tertiary Guantao group.
By 1996, this oil-bearing play had expanded to
564 km2 [218 mi2] with proven original oil in
place of 790 million metric tons (t) [5.8 billion
bbl]. This same trend has a proven gas-bearing
area of 104.5 km2 [40.3 mi2] with original gas in
place of 31 billion m3 [1.1 Tcf]. Additionally, the
depression has estimated proven gas condensate
reserves of 7.34 million t [54 million bbl].
Currently, there are 23 oil and gas elds in

19. Hua LG, Kai CH, Fould J, Lee JS, Long WH, Guo ZX,
Aviles I and Baihly J: An Efcient Horizontal Cased Hole
Multistage Stimulation Well: China Case Story, paper
SPE 153339, presented at the SPE Oil and Gas India
Conference and Exhibition, Mumbai, March 2830, 2012.
20. Baihly et al, reference 2.
21. Martin et al, reference 5.

Multiple Ball Seat

1,314

3,359

1,267

2,523

0.875-in. ball seat

~1,300 psi

1,687

1,220

851

1,174

821

Static pressure, psi

Static pressure, psi

~3,400 psi

15

the depression, including 15 oil- and gasproducing elds in 24 development areas in the
Dagang eld. Annual production from this eld
is 4.3 million t [31.4 million bbl] of oil and
380 million m3 [13 Bcf] of gas.
This eld has been traditionally produced
through cased and cemented vertical wells.
Because many wells of this type are required to
produce these relatively low-permeability formations, the economics may be considered marginal despite the large production volumes. The
operator recently set an oil production target
rate for the eld of more than 6,000 t/yr
[44,000 bbl/yr] oil equivalent. Completing wells
quickly and achieving incremental production
gains in each well are the keys to reaching the
operators objective. To do so, engineers must
properly identify and complete as many pay
zones per well as possible using appropriate
technology, including horizontal drilling.
Additionally, the operator calculated that vertical wells in the target formation would produce
an average of 15 m3/d [94 bbl/d] of oil, while
horizontal wells would produce an average
45 m3/d [283 bbl/d] using traditional completion techniques. To increase the return on horizontal wells, and after assessing the plug and
perforate methodology, engineers opted for an
nZone completion that included a rupture disc
valve placed at the toe of the well to expose the
formation for treatment of the rst stage.

Four 0.875-in. ball seats

1,127
1,081

1,657

1,034

2,492

988

3,328

941

> Multiple ball seats. By replacing a single ball-and-seat conguration (bottom left) with multiple seats (photograph, top), the Falcon multiple seat valve
(bottom right) enables the system to use balls small enough to reach and activate the lower valves. The smaller ball size also reduces friction pressure and
pump horsepower requirements as well as wear on the ball seat.

32

Oileld Review

Hydraulic control line to

upper valves and surface

Fracture 3.75-in. internal

ports
diameter

Dart in
sliding sleeve seat

Shifted sleeve

Hydraulic control line

to lower valves

3.25-in. internal
diameter

Open C-ring

Closed circular ring

> Unlimited numbers of stages. Using an nZone valve, operators ready the stage below the valve for
treatment when a ball or dart lands in the seat of the sliding sleeve. Pressure increases in a hydraulic
control line that connects numerous valves. When a lower nZone valve opens, stimulation uids are
pumped into the formation (yellow arrows). Pressure on the hydraulic line shifts a sleeve downward,
causing a C-ring to move into the smaller inner diameter of the valve and form a smaller diameter
circular seat that is ready to receive the next dart or ball to begin the process again. Because the
seats are not in descending size, the process can be repeated for as many stages as are required to
stimulate the entire well.

The horizontal section of the well was completed as a 51/2-in. monobore casing cemented in
an 81/2-in. hole and treated via a four-stage stimulation. The disc valve at the shoe was opened at
3,500 psi [24 MPa] above the casing test pressure, which allowed engineers to test the casing
as part of the cementing operation. After the disc
valve ruptured, which manifested as a sudden
pressure drop observed at the surface, engineers
rst performed a minifracture to determine for-

Stage 1

mation parameters and conrm injectivity into

the rst zone; they followed that with the rst
stimulation stage.
Completion engineers launched a ball from the
surface during ush to isolate the rst stage and
begin Stage 2. When the ball landed in the rst
seat, the pressure increased, and engineers shut
down the pumps. When pumping resumed, a sudden drop in pressure indicated the valve had
opened and the formation was fractured using less

Stage 3

Stage 4

Pressure

Stage 2

8:24

9:36

10:48

12:00

Pressure
Pump rate
Proppant concentration
Fluid volume
Sand weight

13:12

14:24

15:36

16:48

18:00

19:12

Time

> Stimulation treatment. Following the opening of the rupture disc valve to begin the MSS operation in
the Dagang eld in China, fracturing operations started with the treatment of the rst zone. After a full
ush on the rst stage, the rst ball was released into the well. This operation took about 1.5 h per
stage. Once the ball landed on the rst seat (second stage) at about 10:48, the pressure increased
quickly (A), and all pumps were shut down. Pumping resumed, and a sudden drop in pressure indicated
that the valve had opened (B). The pump rate was increased further (C), and the Stage 2 fracture was
initiated. These steps were repeated until all four stages were treated. (Adapted from Hua et al,
reference 19.)

Summer 2013

than 4,800 psi [33 MPa] pressure as measured at

the surface. Engineers attribute this low fracture
pressure to the helical port design of the Falcon
fracture valves. These steps were repeated until all
four stages were stimulated, during which fracture
initiation pressures from Stage 1 to Stage 4 were
5,100, 4,800, 5,800 and 5,500 psi [35, 33, 40 and
38 MPa], respectively (below left). That pressures
were different at each stage is a strong indication
that all four stages were treated.
Unlike most other wells in the area, the treated
well was able to ow back immediately and without articial lift. Production was 8 to 10 times
greater than that of a vertical offset well and was
expected to be triple that of an unstimulated horizontal well. After ve weeks, because ow rates
were higher than those in other wells in the eld,
the operator was able to produce the well using a
less expensive electric submersible pump instead
of a rod pump. Payout from the well in which the
operator used the nZone system was calculated at
two and one-half months in contrast to four
months for the unstimulated horizontal well and
eight months for vertical wells. The operator plans
several more wells using MSS technology.19
Not One Size Fits All
As MSS technology rapidly transitions from
emerging to mature status, the industry remains
uncertain about how best to apply it. Exploiting
liquid-bearing shales and other ultralowpermeability formations is a relatively recent
endeavor, and long-term data are nonexistent.
For example, while engineers have doubled
the length of laterals in the Bakken Shale in the
past decade, the stimulation stage count has
increased 10-fold. At the same time, as lateral
lengths have increased, operators have generally
decreased stage spacing and the amount of proppant and uid pumped per stage. And although
data seem to indicate a limit to the rate of return
on investment from more stages per wellabout
37 stages in the Bakkenlong-term economic
analysis of these plays is currently impossible;
these wells have not been producing for enough
time to generate sufcient data for meaningful
decline curve analysis.20
Similarly, the industry is still learning how to get
the most from the shales. For example, the industry
does not yet fully understand the storage mechanisms of the Eagle Ford Shale and the factors that
differentiate a good producing area from a mediocre
one.21 Only data gathered over time will answer the
economic and reservoir questions of unconventional resources, even as technologies emerge to
take advantage of that knowledge.
RvF

33

Babatunde Ajayi
Seneca Resources Corporation
Pittsburgh, Pennsylvania, USA
Iroh Isaac Aso
Ira J. Jay Terry, Jr.
Kirby Walker
Kevin Wutherich
Canonsburg, Pennsylvania

Stimulation Design for

Unconventional Resources
The oil and gas industry has undergone a renaissance brought on by the development
of ultralow-permeability reservoirs, made possible through horizontal drilling and

Jacob Caplan
Dewey W. Gerdom
PDC Mountaineer LLC
Bridgeport, West Virginia, USA

hydraulic fracturing. Recent innovations in systematic engineering design are improv-

Brian D. Clark
Utpal Ganguly
Houston, Texas, USA

monitoring and calibration and production evaluation.

Xianwen Li
Yonggao Xu
Hua Yang
PetroChina Changqing Oileld Company
Xian, Shaanxi, Peoples Republic of China
Hai Liu
Beijing, Peoples Republic of China
Yin Luo
Chengdu, Sichuan, Peoples Republic of China
George Waters
Oklahoma City, Oklahoma, USA
Oileld Review Summer 2013: 25, no. 2.
Copyright 2013 Schlumberger.
For help in preparation of this article, thanks to Paul
A. Babasick, Houston; John P. McGinnis and Barry L.
McMahan, Seneca Resources Corporation, Houston; and
Michael Yang, Beijing.
Mangrove, Petrel, RST, Sonic Scanner, StimMAP LIVE and
UFM are marks of Schlumberger.
INTERSECT is a joint mark of Schlumberger, Chevron
and Total.
1. For more on current horizontal drilling technology:
Felczak E, Torre A, Godwin ND, Mantle K, Naganathan S,
Hawkins R, Li K, Jones S and Slayden F: The Best of
Both WorldsA Hybrid Rotary Steerable System,
Oileld Review 23, no. 4 (Winter 2011/2012): 3644.
For more on steering horizontal wells: Amer A,
Chinellato F, Collins S, Denichou J-M, Dubourg I,
Grifths R, Koepsell R, Lyngra S, Marza P, Murray D and
Roberts I: Structural SteeringA Path to Productivity,
Oileld Review 25, no. 1 (Spring 2013): 1431.
2. Miller C, Waters G and Rylander E: Evaluation of
Production Log Data from Horizontal Wells Drilled in
Organic Shales, paper SPE 144326, presented at the
SPE North American Unconventional Gas Conference
and Exhibition, The Woodlands, Texas, USA,
June 1416, 2011.

34

ing stimulation effectiveness and well production. Completion engineers are able to
perform the entire design loop, from reservoir characterization to stimulation plan,

The ability to efciently exploit ultralow-permeability plays has invigorated the oil and gas industry around the globe. The transition from vertical
to horizontal wells was spurred by development
of revolutionary techniques for drilling and completion. Eventually, completion and stimulation
design for horizontal wells evolved into a standard templatethe geometric method, whereby
engineers divide the horizontal wellbore length
evenly into the number of planned intervals, or
stages, designated for fracture treatment. To promote fracture growth from multiple starting
points, engineers then design stages typically
with two to eight perforation clusters distributed
uniformly along the stage length.
The geometric approach for fracture design
ignores the vertical and horizontal heterogeneity
of unconventional reservoirs. Vertical wells may
penetrate a stack of highly variable sandstone and
shale strata. Horizontal wells may wander through
heterogeneous portions of a reservoir, or even
completely out of a reservoir, depending on how
closely the driller was able to follow the target
zone. Geologic heterogeneity along wellbores
causes wide variability of rock properties that, in
turn, directly affect where fracturing stages will
encounter producible reservoir rock. Consequently,
the geometric placement of stages often results in
poor well performance, leading completion engineers to use manual, time-intensive methods of
picking stage and perforation locations based on
subtle well log characteristics.

Increasingly, directional wells are being

drilled and steered based on logging-while-drilling (LWD) data.1 Engineers can use these
measurements to characterize small-scale heterogeneities that horizontal wells encounter as
they penetrate stratied formations. However,
even with the addition of LWD data to help in
planning stimulation programs, well performance has been difcult to predict.
Recently, Schlumberger engineers analyzed
production logs from more than 100 horizontal
shale gas wells in six US shale basins to identify
factors that inuence the effectiveness of hydraulic fracture completions.2 The analysis indicated
that perforation efciencythe percentage of
perforation clusters that contribute to productionwas about 70%. Nearly a third of the clusters contributed nothing to production. The
investigators looked deeper into the data to
explain this inefciency.
The data showed that increasing the number
of fracture stages and decreasing the distance
between stages and between perforation clusters
correlated with a rise in production rate from a
well. Stimulation design is a compromise between
the extremes of a single customized fracture
stage and of multiple stages to cover a wide variety of rocks. Increasing the number of perforation clusters and stages is not a guarantee for
success. The analysis suggested that focused
staging is important: Fracture stages should target rocks with similar petrophysical and geomechanical properties.

Oileld Review

Summer 2013

35

Reservoir Quality (RQ)

Organic content

Completion Quality (CQ)

Mineralogymainly clay, carbonate and silica

Thermal maturity
Effective porosity

Mechanical propertiesYoungs modulus,

Poissons ratio and tensile strength

Intrinsic permeability
Fluid saturationsoil, gas,
condensate and water
Organic shale thickness
Hydrocarbons in place

Natural fracturespresence, density, orientation

and condition (open, closed or cemented)
In situ stressvariations between intervals
accounting for mechanical properties anisotropy

> Reservoir quality and completion quality factors.

Because it was apparent that not every stage

contributed equally to well productivity, the
investigators also examined the contribution of
perforation clusters within fracture stages. They
determined that, like fracture stages, not every
cluster contributed equally to production, and
they concluded that the optimal number of perforation clusters per stage ranged from two to ve.
The analysis suggested that strategic placement
of clusters within productive and fracturable geologic units was more important than the number
of clusters.
The study results led to fundamental design
questions:
Is there an optimal number of treatment stages?
Is there an optimal location for each treatment
stage along a wellbore?
Is there an optimal place for perforation clusters within stages?
To address these questions, Schlumberger completion engineers developed the Mangrove reservoir

stimulation design software for engineering,

modeling and designing hydraulic stimulations.
The software facilitates a systematic strategy for
designing multistage stimulations centered on
single wells embedded within the context of a 3D
earth model of the reservoir. Completion and
stimulation design is based on multidisciplinary
reservoir characterization that is combined with
microseismic information for model calibration
and integrated with production forecasting for
design evaluation.3
This article describes the Mangrove software
and outlines case studies that demonstrate how
the software helps operators improve well productivity. Two examples from the eastern US
show side by side comparisons of well productivities that result from conventional and engineered completions in the Marcellus Shale. An
example from the Ordos basin of China illustrates improvements to production from lowpermeability sandstones.

3. Cipolla C, Weng X, Onda H, Nadaraja T, Ganguly U and

Malpani R: New Algorithms and Integrated Workow for
Tight Gas and Shale Completions, paper SPE 146872,
presented at the SPE Annual Technical Conference and
Exhibition, Denver, October 30November 2, 2011.
Cipolla C, Lewis R, Maxwell S and Mack M: Appraising
Unconventional Resource Plays: Separating Reservoir
Quality from Completion Effectiveness, paper
IPTC 14677, presented at the International Petroleum
Technology Conference, Bangkok, Thailand,
February 79, 2012.
4. Fabric refers to the spacing, arrangement, distribution,
size, shape and orientation of the constituents of rocks
such as minerals, grains, porosity, layering, bed
boundaries, lithology contacts and fractures.
5. For more on fracture staging algorithms: Cipolla et al
(2011), reference 3.
6. For more on conventional hydraulic fracture models:
Brady B, Elbel J, Mack M, Morales H, Nolte K and Poe B:
Cracking Rock: Progress in Fracture Treatment Design,
Oileld Review 4, no. 4 (October 1992): 417.
7. Jeffrey RG, Zhang X and Thiercelin M: Hydraulic
Fracture Offsetting in Naturally Fractured Reservoirs:
Quantifying a Long-Recognized Process, paper SPE
119351, presented at the SPE Hydraulic Fracturing
Technology Conference, The Woodlands, Texas,
January 1921, 2009.

Surez-Rivera R, Deenadayalu C, Chertov M,

Hartanto RN, Gathogo P and Kunjir R: Improving
Horizontal Completions on Heterogeneous Tight Shales,
paper CSUG/SPE 146998, presented at the Canadian
Unconventional Resources Conference, Calgary,
November 1517, 2011.
Surez-Rivera R, Burghardt J, Stanchits S, Edelman E and
Surdi A: Understanding the Effect of Rock Fabric on
Fracture Complexity for Improving Completion Design and
Well Performance, paper IPTC 17018, presented at the
International Petroleum Technology Conference, Beijing,
March 2628, 2013.
8. For more on the wiremesh model: Xu W, Thiercelin M,
Ganguly U, Weng X, Gu H, Onda H, Sun J and Le Calvez J:
Wiremesh: A Novel Shale Fracturing Simulator, paper
SPE 132218, presented at the CPS/SPE International Oil
and Gas Conference and Exhibition in China, Beijing,
June 810, 2010.
9. For more on the UFM model: Weng X, Kresse O, Cohen C,
Wu R and Gu H: Modeling of Hydraulic FractureNetwork Propagation in a Naturally Fractured
Formation, SPE Production & Operations 26, no. 4
(November 2011): 368380.
Kresse O, Cohen C, Weng X, Wu R and Gu H: Numerical
Modeling of Hydraulic Fracturing in Naturally Fractured
Formations, paper ARMA 11-363, presented at the
45th US Rock Mechanics/Geomechanics Symposium,
San Francisco, June 2629, 2011.

36

Engineered Stimulations
While Mangrove software provides a specic
engineering workow intended for predictive
model building and evaluation of hydraulic fracture treatment in unconventional reservoirs, it
also continues to support workows and modeling necessary for conventional reservoirs. The
Mangrove system is capable of accommodating
reservoir heterogeneity, rock fabric, physical
properties and geomechanical properties at a
ne level of detail without compromising computational efciency.4
Input to the workow comes from geologic,
core, well log, seismic, production log and engineering data. Geologists, geophysicists and
engineers compile, synthesize and interpret
these data and summarize them in a common
3D earth model. This integration and display are
performed within the Petrel E&P software platform. The earth model forms the basis for geologic, discrete fracture network (DFN) and
geomechanical models that are input to the
completion advisor as well as to a number of
hydraulic fracture models and to production
and forecasting simulators accessible within
the Mangrove workow.
Engineers use the Mangrove completion advisor to assign levels of reservoir quality and completion quality to the reservoir rock (above left).
Reservoir quality (RQ) is a prediction of how
prone the rock is to yield hydrocarbon.
Completion quality (CQ) is a prediction of how
effectively the rock may be stimulated using
hydraulic fractures. The RQ and CQ parameters
typically receive binary scores of good or bad
based on cutoff criteria for a reservoir. They are
then combined into composite scores that grade
the intervals from best to worst for placing fracturing stages and perforation clusters within
each stage. The best locations have good RQ and
CQ grades, meaning the rock should be productive and fracturable (next page).5 The completion
advisor also allows similar quality rocks to be
grouped in the same stage, leading to the most
effective multistage treatment. The completion
advisor is able to accommodate user-provided
operational constraints, such as the maximum
stage interval or minimum and maximum perforation interval, and structural constraints such
as fault locations and distances of perforation
clusters from these faults.
After deciding where to locate stages and perforation clusters, engineers design the stimulation treatments using hydraulic fracture (HF)
simulators. In situations in which the geology is

Oileld Review

relatively simple, conventional HF simulators are

adequate. These time-tested 2D and pseudo-3D
models treat HFs as planes propagating away
from the well in the direction of the maximum
principal compressive stress.6 Engineers have the
option to use these models in the Mangrove workow and determine which model is best suited
for a given reservoir.
Conventional models are not comprehensive
enough for heterogeneous and naturally fractured reservoirs. Hydraulic fracture growth is
complex, and its characterization requires 3D
models that incorporate interactions of HFs with
natural fractures while also considering the
impact of HFs on local principal stresses.7 To
address complex situations, the Mangrove system
provides two fracture models: the wiremesh
hydraulic fracturing model and the UFM unconventional fracture modeling simulator.
The wiremesh model provides a mathematical equivalent representation of the hydraulic
fracture network.8 The wiremesh approach is
relatively fast and suitable for environments that
lack signicant reservoir characterization data.
To improve well productivity, completion designers are able to iterate and parameterize the input
values to obtain an optimal stimulation design for
fracture length, height, surface area and proppant distribution.
The UFM model is the rst commercially
available complex hydraulic fracture model to
incorporate fracture-to-fracture interactions.9
The model accounts for the effects of natural
fractures and geomechanical properties on
hydraulic fracture growth and predicts dendriticmultiple branchinghydraulic fracture
propagation as well as uid ow and proppant
transport. Hydraulic fracture growth is governed
by the rock fabric and geomechanical properties
of the reservoir, the preexisting fracture network
and prevailing in situ stress magnitudes and
anisotropy. As the HF network develops, it perturbs the stress eld as each fracture surface
becomes pressurized, opened and propped.
Engineers may use the UFM simulator for HF network design to optimize well productivity.
Regardless of the HF model engineers use to
prepare their initial design, the result must be
calibrated during HF stimulations. The Mangrove
workow is able to incorporate results obtained
from monitoring microseismicity induced by
propagating HFs to calibrate the predicted
model. Geophysicists process the microseismicity
data to locate seismic emissions from small slip
events associated with the development of the

Summer 2013

Segments of Similar Lithology

Well segments

Austin Chalk

Upper Eagle Ford Shale

Lower Eagle Ford Shale

Buda Limestone

Stages of Similar Rock Quality and Stress Gradient

Hydraulic fracturing stages

Austin Chalk
Rock Quality
Good RQ and good CQ
Bad RQ and bad CQ
Bad RQ and good CQ
Good RQ and bad CQ
Rock quality

Eagle Ford Shale

Stress gradient
Stress gradient
Low

High

Buda Limestone

> Dividing horizontal laterals into segments and stages. This horizontal well (top center) targets a
reservoir zone near the boundary horizon (purple) between the upper and lower Eagle Ford Shale,
which was deposited above the Buda Limestone and below the Austin Chalk. The other horizons are
the top surfaces of the Buda Limestone (blue) and the upper Eagle Ford Shale (brown). Engineers
divided the lateral into segments based on location within the reservoir, the wellbore trajectory and
rock properties. Each segment contains similar lithology along its length. Engineers further subdivided
the segments into stages (bottom center) based on similar minimum horizontal stress gradients,
reservoir quality (RQ) and completion quality (CQ) along the length of each stage. Each stage is then a
candidate for hydraulic stimulation. A color-coded rock quality index, shown above the well, combines
RQ and CQ and indicates the best intervals for stimulation. The relative magnitude of the far eld
minimum horizontal stress gradient, shown along the bottom of the well, indicates the relative
pressure levels at which the reservoir interval will fracture. [Adapted from Cipolla et al (2011),
reference 3.]

37

Perforation cluster

Geometric Placement of Fracture Stages and Perforation Clusters

Fracture stage

Rock Quality
Good RQ and good CQ
Bad RQ and bad CQ
Bad RQ and good CQ
Good RQ and bad CQ
Rock quality

Stress gradient
Stress gradient
Low

High

Engineered Placement of Fracture Stages and Perforation Clusters

Rock quality

Stress gradient

> Comparing hydraulic fracture designs for a horizontal well in the Eagle Ford Shale. In a geometric
design (top), fracture stages (inset, four disks of the same color) and perforation clusters (individual
disks) were distributed uniformly along the length of the lateral. In the engineered design from the
Mangrove workow (bottom), engineers determined the location and length of each fracture stage
and the placement of each perforation cluster from analysis of the composite rock quality scores and
minimum horizontal stress gradients. The optimal design is for all perforation clusters (PCs) to break
down and initiate fractures at more or less the same pressure. The composite RQ and CQ rock quality
index is shown along the top of the well. The relative magnitude of the far eld minimum horizontal
stress gradient is shown along the bottom of the well. [Adapted from Cipolla et al (2011), reference 3.]

HFs.10 Often, to increase the precision and accuracy of the event locations, geophysicists adjust
their geologic and velocity models. These adjustments, in turn, are used to update the geomechanical and the DFN models for the HF models.
Before and after completion of HF stimulations, production engineers run reservoir ow
models to predict the resulting production performance. These models couple mechanical
deformation and pore volume changes. The

38

fracture models simulate rock deformation, the

creation of conductive fractures and channels in
the reservoir and the placement of proppant into
them. The reservoir simulators predict the ow of
uids from the reservoir into and through the
higher conductivity pathways created by HFs that
have been propped open. Within the Mangrove
workow, these calculations may be performed
using the INTERSECT reservoir simulator, which
allows unstructured gridding for a range of grid

densities. Fine gridding in the vicinity of the wellbore and HF network captures ne-scale details.
Coarse gridding is usually sufcient far from the
wellbore and HF network.11
The Mangrove workow provides analysis
from data entry to model updates. In this process,
geologic and engineering eld data are input for
building models of the reservoir. Engineers use
the models to estimate RQ and CQ (above).
Engineers input the completion design into 2D or

Oileld Review

3D HF simulators for evaluating the fracture

stimulations that will be pumped and then feed
the stimulation design into reservoir simulators
to forecast production.
The system is able to incorporate microseismicity monitoring to calibrate steps in the
Mangrove workow. Such calibration comes from
locating microseismic events precisely and comparing these locations with predicted HF growth.
The location of microseismic events may help the
system estimate the effective stimulated reservoir volume, which may then be used to adjust the
completion and stimulation strategies of subsequent fracture stages or make adjustments even
while stimulation is occurring in some stages.
In addition, to obtain precise microseismic
event locations, geophysicists conduct seismic
velocity inversion, and in the process, adjust the
starting model of the geologic and mechanical
properties within the reservoir zone. The adjusted
model may be used to update predictions of
hydraulic fracture growth and forecasts of reservoir production.
The Mangrove workow centers around completion and stimulation design for single wells
within the 3D context of a larger reservoir model.
The focus on single wells reduces model size,
enables faster calculations and gives completion
engineers exibility to make quick decisions and
adjustments to stimulation programs.
The Mangrove software may be run on a single platform, which eliminates the need to
migrate data from one software application to
another and to address problems of software
interfaces and interoperability.
10. For more on hydraulic fracture monitoring: 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,
Oileld Review 17, no. 4 (Winter 2005/2006): 4257.
Burch DN, Daniels J, Gillard M, Underhill W, Exler VA,
Favoretti L, Le Calvez J, Lecerf B, Potapenko D,
Maschio L, Morales JA, Samuelson M and Weimann MI:
Live Hydraulic Fracture Monitoring and Diversion,
Oileld Review 21, no. 3 (Autumn 2009): 1831.
11. For more on the INTERSECT simulator: Edwards DA,
Gunasekera D, Morris J, Shaw G, Shaw K, Walsh D,
Fjerstad PA, Kikani J, Franco J, Hoang V and
Quettier L: Reservoir Simulation: Keeping Pace with
Oileld Complexity, Oileld Review 23, no. 4
(Winter 2011/2012): 415.
12. Walker K, Wutherich K, Terry J, Shreves J and Caplan J:
Improving Production in the Marcellus Shale Using an
Engineered Completion Design: A Case Study, paper
SPE 159666, presented at the SPE Annual Technical
Conference and Exhibition, San Antonio, Texas,
October 810, 2012.
Gerdom D, Caplan J, Terry IJ Jr, Wutherich K, Wigger E
and Walker K: Geomechanics Key in Marcellus Wells,
The American Oil & Gas Reporter 56, no. 3
(March 2013): 8491.

Summer 2013

A software-mediated systematic approach to

planning, engineering and executing stimulations
has proved to be more effective than conventionally planned stimulations. PDC Mountaineer LLC
and Schlumberger obtained favorable results with
engineered completions in the Marcellus Shale.
Comparing Completion Methods
PDC Mountaineer LLC (PDCM) focuses primarily
on natural gas production from the Marcellus
Shale formation. In the companys efforts to
develop a Marcellus Shale eld near Bridgeport
in Harrison County, West Virginia, USA, its rst
three horizontal wells were only marginally economic. Consequently, PDCM wanted to determine how to improve production.
The company started each of these rst
wells by drilling and logging a vertical pilot well.
Engineers used these data to determine the target reservoir zone and the landing point for the
horizontal well, or lateral. PDCM then drilled
the laterals using data derived from mud logs
and logging-while-drilling (LWD) gamma ray for
guidance to stay within the target zone. The laterals were completed using designs based on a
geometric methodstages and perforation
clusters distributed uniformlyfollowed by
slight manual adjustments to the design to move
perforation clusters within each stage to zones
that were estimated to have lower minimum
horizontal stress.12
PDCM and Schlumberger engineers analyzed
the data from the rst three wells and concluded
that the completion designs paid little attention to
specic conditions in each welllithology, reservoir quality, mechanical properties and in situ
stresses. Furthermore, examination of stimulation-induced microseismicity monitored during
the treatments showed a relationship between the
locations of perforation clusters, predicted minimum in situ horizontal stress and microseismic
activity; the highest microseismic activity concentrated near perforations in rocks of low stress, and
lower activity occurred elsewhere. Fractures
started and grew by taking paths of least resistance. Areas near the geometrically located perforation clusters were effectively stimulated only
when the clusters happened to be located in easily
fractured rock. Otherwise, areas tended to be
understimulated because the perforation clusters
were not strategically located.
The analysis indicated that optimal stimulations would result if the completions were engineered so each stage and each perforation cluster
contributed to the overall production in proportion to their number. Horizontal wells would be

divided into segments of similar lithology that did

not include discontinuitiesprimarily faults,
fractures and highly laminated intervals. The segments would then be divided into stages and perforated in rocks of similar minimum horizontal
stress. During each fracture stage, all perforations would initiate fractures at roughly the same
pumping pressure, the fractures would extend
and propagate together, and eventually, production would ow from the fractures in proportion
to the stimulated reservoir volume they contact.
To test this procedure, the PDCM team
selected three new well locations, near the original three wells, which had similar reservoir and
completion quality. Except for the engineered
design for distributing the staging and perforation locations along the laterals, the new wells
would be completed in the same way as the earlier wells.
The wells were drilled in the direction of the
regional minimum in situ principal horizontal
stress to facilitate opening of hydraulic fractures
emanating perpendicularly from the wells. The
lateral wells cut across rocks of variable lithology
and, consequently, mechanical properties, which
dictate how the regional stress eld is transmitted through the rock to the local borehole wall.
After drilling the wells and before designing
the completions, the team collected the following well information: wellbore directional surveys, gamma ray logs, petrophysical and
mechanical properties for evaluating RQ and
CQ, planned fracture uid types and properties,
pumping rates, number of stages, number of
perforation clusters per stage and perforation
diameter, density and phasing. The completion
design called for slickwater to be pumped at
80 bbl/min [13 m3/min] through ve perforation
clusters in each stage.
Engineers assembled this information in the
Mangrove workow software and constructed
3D earth models of each well. Based on data
from the 3D earth models, engineers were able
to segment the wells into lengths of similar
lithology; each segment was subdivided into
stages, such that each stage length contained
rock of similar reservoir quality and was capable
of accepting the planned pumping rate. The
team selected perforation locations based on
completion quality. The perforation locations
were adjusted until the models showed that
fractures initiated at each perforation cluster
within a stage at the same pressure within a tolerance of 0.01 psi/ft [0.23 kPa/m] for the

39

Measured
Depth,
ft

Gamma
Ray
0

Minimum
Stress
Gradient

Poissons
Ratio

gAPI 460 0.67 psi/ft 1.01 0.17

Youngs
Modulus

0.44 2.33 MMpsi 3.93 0

higher production. During the rst 30 days, the

engineered completions resulted in 106% higher initial cumulative production per foot of stimulated
wellbore length than the original three wells.
Based on these positive results, PDC
Mountaineer now performs engineered completion designs for all its horizontal wells. The company has determined that the time and effort
spent on the design are more than offset by the
savings from operational effectiveness during completions and revenue from increased production.14
Perforating Low-Stress Intervals
Seneca Resources Corporation and Schlumberger
conducted another test of engineered completion
design. Seneca Resources produces natural gas
from Marcellus Shale reservoirs in Pennsylvania

Calcite
Volume
%

100 0

Quartz
Volume
% 100 0

Kerogen
Volume

Effective
Porosity

25 0

15

and New York, USA. The company sought to

increase production by maximizing reservoir contact through hydraulic fracture stimulations from
horizontal wells.
Seneca Resources had been stimulating wells
in the Marcellus Shale but results were highly
variable, even from apparently identical wells.
However, the Marcellus Shale comprises many
thin laminations, each distinct from its neighbor
in terms of physical and mechanical properties.
As horizontal wells cut through the formation,
they intercept these varied laminations. The
company teamed with Schlumberger to conduct a
controlled pilot study to test the effectiveness of
engineered completions compared with what had
been standard practice for the companygeometric completions.

Stimulation
Stages

Measured
Depth, ft

minimum in situ stress gradient.13 When the

team was satised with the completion plans,
the wells were stimulated (below).
Completion engineers conducted each fracture treatment according to the intended proppant schedule. Compared with the treatments in
the original three horizontal wells, the engineered completions were pumped at 10.3% higher
average pumping rates and 5.7% lower average
treating pressures. In addition, the treatments
succeeded in placing 30% more of the designed
proppant load per lateral and experienced no
screenouts (next page, top left).
The team compared the rst 30 days of production from each well, which revealed a second measure of success. Compared with the original wells,
the engineered completions resulted in signicantly

Perforation
Cluster

Stage 14

X1,500

5-ft Moving Average

Minimum
Stress Gradient

Smoothed Minimum
Stress Gradient

0.67

Segment 1
X1,000

Original

psi/ft

1.01 0.67

psi/ft

1.01

Stage 13
Stage 12

Segment 2

X3,850

Stage 11
X2,000

X2,500

X3,900

Stage 10
Stage 9

Segment 3
X3,950

Stage 8
X3,000

Stage 7
Stage 6

X4,000

X3,500
Stage 5
X4,000

Stage 4

Segment 4

X4,050

Stage 3
X4,500
Stage 2
X5,000

Stage 1

Stress gradient
Low

High

> Segments, stages and clusters. Stresses typically change from one lithology to another. To prevent a fracture stage from crossing a lithology barrier,
engineers divide the well into segments of similar lithology. Stimulation stages (left, Track 9, green and light blue) should be contained within a segment,
and their lengths should be within prescribed minimum and maximum values. Engineers position the perforation clusters (Track 9, short horizontal lines to
the left and right of the fracture stages) based on preset design criteria: the number of clusters per stage, the minimum and maximum distance between
clusters and a minimum horizontal stress gradient (Track 2) tolerance of 0.01 psi/ft [0.23 kPa/m]. During completion design and modeling, these criteria may
need to be relaxed to account for the minimum horizontal stress variation. A close-up of the red box (right) from Track 2 shows the stress gradient ranges
from high (blue) to low (red). The original stress gradient logs were recorded every half foot (inset, Track 1) and smoothed using a 5-ft [1.5-m] moving
average algorithm (inset, Track 2) to account for imprecision during the perforating operation. (Adapted from Walker et al, reference 12.)
13. The rate of these stress variations within a few borehole
diameters of the wellbore, away from the immediate
inuence of the borehole, is the wellbore-parallel stress
gradient and, for wells drilled parallel to the minimum
in situ principal stress direction, is equivalent to the
minimum stress gradient.
14. Walker et al, reference 12.

40

15. Wutherich K, Walker K, Aso I, Ajayi B and Cannon T:

Evaluating an Engineered Completion Design in the
Marcellus Shale Using Microseismic Monitoring, paper
SPE 159681, presented at the SPE Annual Technical
Conference and Exhibition, San Antonio, Texas,
October 810, 2012.

16. Waters G and Zhao R: Measuring the Impact of

Geomechanical Heterogeneity in Organic Shales on
Hydraulic Fracture Initiation and Propagation,
paper CSUG/SPE 147597, presented at the Canadian
Unconventional Resources Conference, Calgary,
November 1517, 2011.

Oileld Review

Design Summary
Design
Pumping
Rate,
bbl/min

Well

Completion
Method

Fluid

Lateral
Length, ft

Stages

Well 1

Nonengineered

Slickwater

3,375

14

241

1,670

80

Well 2

Nonengineered

Slickwater

2,312

330

1,220

80

Well 3

Nonengineered

Slickwater

Average

Perforation
Clusters
per Stage

Design
Proppant
per Lateral,
lbm/ft

Average
Stage
Length, ft

2,140

306

1,320

80

2,609

9.3

292

1,400

80

Well 4

Engineered

Slickwater

4,500

12

375

1,080

80

Well 5

Engineered

Slickwater

3,950

12

329

4.5

1,230

80

Well 6

Engineered

Slickwater

3,925

12

327

4.5

1,240

80

4,125

12

344

4.7

1,180

80

Average

Completion Summary

Well

Average
Average
Treating Treatment
Pressure,
Rate,
psi
bbl/min

Percentage
of Proppant
Placed
Versus Design

Gross,
Mcf

Normalized
by Lateral
Length,
Mcf/ft

Normalized
by Number
of Stages,
Mcf/ft

7,749

78.1

1,783

107.0%

63,194

18.7

4,514

903

Well 2

7,557

76.3

672

55.0%

42,396

18.3

6,057

1,211

Well 3

7,716

66.3

855

65.0%

65,039

30.4

9,291

1,858

Average

7,674

73.6

1,103

75.7%

56,876

21.8

6,094

1,219

Well 4

7,308

79.2

1,002

92.8%

212,631

47.3

17,719

3,544

Well 5

7,105

81.9

1,251

101.7%

162,652

41.2

13,554

3,012

Well 6

7,298

82.3

1,245

100.5%

180,436

46.0

15,036

3,341

Average

7,237

81.1

1,166

98.3%

185,240

44.9

15,437

3,308

437

7.6

63

22.7%

128,363

23.1

9,343

2,089

5.7%

10.3%

5.7%

30.0%

226%

106%

153%

171%

> Summary of completion design and results. Data from six horizontal wells drilled into the Marcellus
Shale illustrate the results of nonengineered and engineered completion methods (top). Wells 1 to 3
were drilled and completed conventionally. Wells 4 to 6, which were drilled near Wells 1 to 3, were
completed using an engineered design method that species stage and perforation cluster placement.
The engineered completions were more effective than the nonengineered completions (bottom); the
success of the engineered completions is measured by lower treating pressures, higher pumping
rates, more efcient proppant placement and higher cumulative production after 30 days compared
with those in the nonengineered completions. (Adapted from Walker et al, reference 12.)

The company drilled three horizontal wells

into the same Marcellus Shale reservoir zone
from the same drilling pad. The laterals were
drilled parallel to one another, 800 ft [240 m]
apart and aligned to the northwest, in the direction of the regional minimum in situ principal
horizontal compressive stress (above right).
Well A, the base case, was completed using the
standard geometric method.15
Wells B and C were completed using the engineered approach. The RST reservoir saturation
tool and Sonic Scanner acoustic scanning tool
were run along each lateral after casing had been
set to determine the extent of variation between
lithologic and mechanical properties and the
resolved stresses in the three wells.16 These measurements were compiled and interpreted using
the Mangrove workow software to produce an
engineered completion strategy for each well.

Summer 2013

305
1,000
N

Well C
Well B
Well A

Normalized
by Number of
Perforation
Clusters,
Mcf/cluster

Well 1

Average
difference
Percent average
difference

m
ft

Monitor well

30-Day Cumulative Production

Placed
Proppant
per Lateral,
lbm/ft

0
0

Although completion strategies were customized to optimize production from each well,
engineers kept a number of completion variablesuid, proppant type and size and pumping ow ratethe same and also kept the
number of stages, number of perforation clusters
per stage and amounts of proppant per length of
lateral similar for both wells. Nonetheless, some
variability existed across the three wells. By their
nature and because they are intended to account
for the rock and stress heterogeneity along the
wellbore, engineered completion designs inevitably result in variable stage lengths, perforation
cluster spacings and pumping schedules.
To accommodate these variations and maintain the spirit of consistency, the company staggered the timing of the well stimulations using a

> Well plan. From a single pad, Seneca

Resources drilled horizontal Wells A, B and C and
drilled a vertical monitor well for recording
stimulation-induced microseismicity. Well A was
completed following a geometric design and
Wells B and C were completed according to
engineered completion designs. The disks on
each well, which represent perforation clusters,
are grouped into fracture stages with adjacent
stages differentiated by color. (Adapted from
Wutherich et al, reference 15.)

zipper-fracture method, whereby plug and perforation operations followed by stimulation of stages
were rotated from one well to the next. As Well A
was being stimulated, Well C was undergoing plugging and perforating. Then stimulation moved to
Well B, while plugging and perforating moved to
Well A. This process continued until stimulation of
all stages in all the wells was complete.
The stimulation engineering team analyzed
pilot study results by comparing treatment,
microseismicity and initial flowback data from
the geometrically designed completion in
Well A to similar data from the engineered
completions in Wells B and C. Because all of
the perforation clusters were engineered to be
located in wellbore intervals of relatively low
minimum principal stress, the average fracture
breakdown and treatment pressures were 7%

41

Design Summary
Average
Stage
Length, ft

Perforation
Clusters
per Stage

Design
Proppant
per Lateral,
lbm/ft

Design
Pumping
Rate,
bbl/min

Well

Completion
Method

Fluid

Proppant
Size

Lateral
Length, ft

Well A

Geometric

Slickwater

40/70

5,312

18

295

1,650

90

Well B

Engineered

Slickwater

40/70

4,528

20

226

3.7

1,585

90

Well C

Engineered

Slickwater

40/70

4,998

20

250

3.9

1,675

90

Stages

Completion Summary

Flowback Results

Well

Average
Breakdown
Pressure,
psi

Average
Treating
Pressure,
psi

Average
Treatment
Rate,
bbl/min

Placed
Proppant
per Lateral,
lbm/ft

Percentage
of Proppant
Placed
Versus Design

Maximum
Flow,
Mcf/d/1,000 ft

Tubing
Pressure,
psi

Choke, in.

Well A

5,572

7,277

69.7

1,122

68%

450

1,500

5/8

600

1,800

5/8

640

1,800

5/8

Well B
Well C

5,160

7,095

81.1

1,353

83%

Difference
Percent
difference

412

182

11.4

231

15%

170

300

7%

3%

16%

21%

22%

38%

20%

> Summary of completion design and results. Of three horizontal wells drilled into the Marcellus Shale,
Well A, the reference case, was completed following a geometric design (top). Wells B and C were
completed according to engineered completion designs, which were more effective than the
geometric completion. Their relative success is measured by lower breakdown and treating
pressures, higher pumping rates, more effective proppant placement and higher owback rates than
those of Well A (bottom). (Adapted from Wutherich et al, reference 15.)

and 3% lower and the average treatment rate

and amount of proppant placed were 16% and
22% higher in Wells B and C, respectively, than
in Well A. The treatment comparison indicated
that the engineered completions were more
effective than the geometric completion (above).
Initial gas owback rates from Wells B and C
were 33% and 40% higher than the rates from
Well A on the same 5/8-in. choke size. In addition,
fracture-water owback recovery from Wells B
and C was twice that from Well A. These owback
data suggest that the wells stimulated by engineered completions were making better reservoir
contact, leading to better production, than was
the geometrically completed well.
During the pilot study, the team placed a vertical monitor well between Wells A and B; the well
was instrumented with geophones for monitoring
microseismicity induced by the stimulations in the
three wells. The StimMAP LIVE real-time microseismic monitoring service recorded and analyzed
microseismicity. When compared with perforation
cluster locations, microseismic event locations
from the StimMAP LIVE service revealed that as

42

much as 35% of the perforation clusters in Well A,

with the geometric completion, were not contributing to the reservoir volume targeted for stimulation. In contrast, microseismicity from the
engineered completions and stimulations in
Wells B and C showed improvement in the percentage of perforation clusters that contributed
to the stimulated reservoir volumeonly 20% of
the perforation clusters made little to no contribution (next page). The microseismicity comparison indicated that the engineered completions
resulted in better placement of perforation clusters than did the geometric completion.
The Mangrove workow software not only produced the designs that led to these positive
results but also reduced completion design time
from several hours to about one hour per well.
Moreover, the software rationalized data handling and procedural operations, which led to
fewer inaccuracies and improved perforation
placement. Seneca Resources continues to use
computer-aided completion design and microseismic analysis on other wells in its elds.17

Stimulation of Tight Oil Sandstone

Conventional reservoirs are also candidates for
the application of the systematic, engineering
approach to reservoir stimulation. The
PetroChina Changqing Oileld Company conducted a pilot study using the engineered
approach for designing reservoir stimulation in a
conventional clastic reservoir.
The Ordos basin, in north-central China, is a
gentle monocline that dips stratigraphically
about 1 from east to west. Its ll, which consists
of sediments deposited during the Paleozoic,
Mesozoic and Cenozoic eras, thickens in the dip
direction with an average thickness of 4 to 5 km
[2.5 to 3.1 mi]. The Paleozoic sediments are
marine deposits that yield primarily natural gas,
while the Mesozoic sediments have a continental
origin and yield oil.18
17. Wutherich et al, reference 15.
18. For more on the Ordos basin: Yang Y, Li W and Ma L:
Tectonic and Stratigraphic Controls of Hydrocarbon
Systems in the Ordos Basin: A Multicycle Cratonic
Basin in Central China, AAPG Bulletin 89, no. 2
(February 2005): 255269.

Oileld Review

Well A
Event count

100

Event count

35

B
200

Event count

Event count

100

Well B
40

Stress gradient
Low

High

Event count

Event count

40

Stress gradient
Low

High

B
40

Stress gradient
Low

High

Event count

Event count

250

Stress gradient
Low

High

> Microseismicity comparison. Microseismicity resulting from four fracture stages in Well A (top) and Well B (bottom) indicate improved stimulations from
the engineered completions in Well B over the stimulations from the geometric completions in Well A. In each panel, the data show results from a fracture
stage; the disks along the colored well trace represent stimulated perforation clusters and the dots are induced microseismic event locations. To show
correlation, the disks and dots have the same color. Above the well trace, the height and color of the orange-to-green bars indicate the number of
microseismic events along each wellbore interval. Below the well trace on Well B, the minimum horizontal stress gradient is plotted; the amplitude and
color of the pink-to-blue shading specify the closure stress gradient level. The company placed perforation clusters based on engineering design principles
at locations with relatively low stress gradients. There is a better one-to-one correspondence between microseismicity and perforation locations in Well B
than in Well A, indicating improved perforation performance results from an engineered completion design. (Adapted from Wutherich et al, reference 15.)

Summer 2013

43

300 m

250 m

300 m

Ordos Basin

Beijing

500 m

MW1

C H I N A

500 m

MW3

250 m

Xian
Shanghai

MW2

HW2
HW1

0
0

250
750

500 m
1,500 ft

Basin
Gas field
Oil field

uth
So

Ch

ina

Se a

> Ordos basin, north-central China. A completions team conducted a pilot study to test engineered completion
designs from Mangrove software. The eld test area (white box) is in southwest Ordos basin. The well layout (inset)
consists of two parallel horizontal production wells (HWs) and three vertical monitoring wells (MWs, blue circles)
constucted for recording microseismicity. The Chang 7 member of the Yanchang Formation was the target horizon.
(Adapted from Liu et al, reference 20.)

The Yanchang Formation is a thick sequence of

lake and delta sediments deposited during the
Late Triassic period. The formation consists of 10
lithologic members, named Chang 1 to Chang 10
from top to bottom. The members are stacks of
alternating mudstone, siltstone and sandstone layers that result in vertical heterogeneity. The reservoirs in the Yanchang Formation are naturally
fractured, low-permeability sandstones in which
porosity is typically about 10% and permeability is
generally 0.1 to 10 mD. The natural fractures occur
in two sets that tend to dip steeply and generally
strike in the ENE and NNW directions.19
To produce oil from these low-permeability
reservoirs, an operator must stimulate the production intervals through multistage hydraulic
fracturing. Historically, most production wells
have been vertical, and after HF stimulation,
their initial production rates have varied from
5 to 8 m3/d [30 to 50 bbl/d]. In the few horizontal
wells, the initial production rates after HF stimulation have averaged 32 m3/d [200 bbl/d]. While
still considered economic, these production rates
are only marginally acceptable. To improve the
production outcomes from its stimulation programs, the company partnered with Schlumberger

44

in a pilot project to test the Mangrove workow in

horizontal wells in a tight oil reservoir zone in the
southwest Ordos basin.20
The company drilled two 1,500-m [4,920-ft]
parallel horizontal wells in the Chang 7 member
of the Yanchang Formation. The wells, 600 m
[1,970 ft] apart, were drilled in the N15W direction, which is parallel to the minimum in situ
principal horizontal stress direction in the Ordos
basin. The company drilled three vertical wells
500 m [1,640 ft] apart between and along a line
parallel to the horizontal wells; these vertical
wells were added for microseismicity monitoring
(MSM) during the fracture stimulations of the
horizontal wells (above).
The pilot study team constructed 3D geologic,
geomechanical and DFN models from the pilot
study well log data and from core descriptions
and geologic studies in the surrounding area
(next page). These models were calibrated using
data from the three monitoring wells and integrated using the Mangrove system to form the
bases for modeling reservoir quality, completion
quality, stimulation staging and perforation
placement, hydraulic fracture stimulation design
and production performance forecasting.

Optimal stimulation design requires that

each stage and its perforation clusters be placed
in wellbore intervals that have a high likelihood
of producing economic amounts of hydrocarbon
and breaking down by fracturing in response to
increased pressure during stimulation. These
wellbore intervals possess good RQ and good CQ.
The team used the Mangrove completion advisor
to select 18 stages per well.
In conjunction with the completion advisor,
the team used the UFM simulator to predict HF
propagation, growth and interaction with natural
fractures (NFs) in the reservoir. Depending on
the in situ stress direction and anisotropy in relation to the reservoir NF system, hydraulic fractures may take advantage of the NFs to produce
19. For more on the Yanchang Formation: Lianbo Z and
Xiang-Yang L: Fractures in Sandstone Reservoirs with
Ultra-Low Permeability: A Case Study of the Upper
Triassic Yanchang Formation in the Ordos Basin, China,
AAPG Bulletin 93, no. 4 (April 2009): 461477.
20. Liu H, Luo Y, Li X, Xu Y, Yang K, Mu L, Zhao W and Zhou S:
Advanced Completion and Fracturing Techniques in
Tight Oil Reservoirs in Ordos Basin: A Workow to
Maximize Well Potential, paper SPE 158268, presented
at the SPE Annual Technical Conference and Exhibition,
San Antonio, Texas, October 810, 2012.
Yang H, Xu YG, Yang KW, Zhou SX, Liu H and Luo Y:
Optimized Treatment Design Shows Promise, E&P 86,
no. 2 (February 2013): 4650.
21. Weng et al, reference 9.

Oileld Review

Total Vertical
Depth, m

Horizon Surface

MW1

MW2

MW3

Neutron
Porosity
(left)

Neutron
Porosity
(left)

Neutron
Porosity
(left)

% 100

% 100

Bulk Density
(right)
Resistivity
Gamma Ray

Bulk Density
(right)
Resistivity
Gamma Ray

7 ohm.m 617 1 g/cm3 2.85 0

2 ohm.m 2,200 1 g/cm3 2.85 0

gAPI 200

gAPI 200

MW1

% 100

Bulk Density
(right)
Resistivity
Gamma Ray
3 ohm.m 1,700 1 g/cm3 2.85 0

MW2

MW3

HW2 HW1

7
6
5
4

gAPI 200

X,100

2
1
X,200

6
X,300

5
X,400

X,500

X,600

MW1

X,700

MW2

MW3
X,800

N
HW2
HW1

> Model building for Ordos basin wells. Because there were no seismic or geologic data for the location, model building started after well logs were
acquired from the three vertical monitor wells (left, MWs). Logs for each well display resistivity (Track 1), neutron porosity and bulk density (Track 2) and
gamma ray (Track 3). Geoscientists began model building by extracting geologic horizon surfaces based on well-to-well correlations between the
monitoring wells. Engineers used the surfaces for well placement guidance (top right) and for 3D model development (middle right) by upscaling
petrophysical properties derived from well log data and lling in between the wells while honoring the horizon surfaces. Geologists created a simple
discrete fracture network (DFN) model (bottom right) based on geologic studies and core descriptions. The DFN contained two dominant steeply dipping
fracture sets, characterized by average strike orientations of N75E (cyan) and N15W (purple) and average fracture spacing of 15 m [49 ft]. The DFN was
calibrated later and modied based on microseismicity data. (Adapted from Liu et al, reference 20.)

complex HF networks and, consequently, high

fracture surface area to make contact with the
reservoir. The production of complex HF networks is more likely when the in situ stress
anisotropy is low.21
During the UFM modeling, the team was also
concerned about determining how existing HFs
affected the behavior of subsequent HFs. After an
HF is created and lled with proppant, the immediate vicinity of the HF changes forever. The HF
imposes a compressive stress component, or
stress shadow, that acts outward from the HF

Summer 2013

plane in the minimum principal stress direction

It alters the local stress magnitude and anisotropy near the fracture and affects adjacent fractures through mechanical interactions. To
properly space HF stimulation staging, engineers
must include such stress shadow effects when
calculating CQ.
After selecting the stage and perforation locations, the team began to execute its design.
During stimulation operations, the team
employed the StimMAP LIVE real-time microseismic monitoring (MSM) service. After each
stage, the team used MSM results to recalibrate

the 3D models, UFM model and stimulation

design. For the next stage, the engineers wanted
to maximize the HF surface area and proppantlled volume to obtain the best production from
the stimulated reservoir interval. MSM data suggested that the HFs being created tended to be
long and contained within the targeted Chang 7
reservoir interval.
While monitoring the rst ve to six stages,
the team observed considerable overlap of microseismicity from neighboring stages, indicating
suboptimal stage spacing. The team decided to

45

Measured Depth, m

Rock Quality
Good RQ and good CQ
Bad RQ and bad CQ
Bad RQ and good CQ
Good RQ and bad CQ
Composite

Bad

Bad

BB

Bad

GG

Good

CQ

Bad

RQ

Good

0.30

Initial
Stimulation
Stages

Updated
Stimulation
Stages

Perforation
Cluster

Perforation
Cluster

Stage 18

Stage 17
Stage 12

Stage 15
GG

Good

Good

Stage 16

X,600

Stage 13

GG

psi/m

Good

250 0

Good

X,400

gAPI

Good

X,200

Minimum Stress
Gradient

Gamma Ray
0

Low

Good

Stress Gradient
High

Stage 14

Stage 11

GG

Stage 12

Good

GG

Stage 11

GG

Good

Good

X,800

Good

Stage 13

Good

GG

Good

Stage 9

Stage 9

Good
Y,000

Stage 10

Stage 10

Good

GG

Stage 8

Stage 8

Stage 7

Stage 7

Stage 6

Stage 6

Stage 5

Stage 5

Stage 4

Stage 4

Stage 3

Stage 3

Stage 2

Stage 2

Stage 1

Stage 1

GG
GG
Good
GG

Good

Y,200

Good

Good

Good

GG
BB

Good
Bad

Y,600

Bad

Good

Y,400

> Completion advisor results. Engineers used the Mangrove completion advisor to compile and analyze
petrophysical data to select fracture stages and perforation cluster locations for wells in the Ordos
basin. Gamma ray (Track 1) and the minimum horizontal stress gradient (Track 2) were key parameters
for the design. For the stress gradient prole, blue is high and red is low. Reservoir quality (Track 3),
completion quality (Track 4) and composite (RQ plus CQ) quality scores (Track 5) provide color-coded
quality indicators for stage and cluster selection. Initially, engineers proposed 18 stimulation stages
(Track 6). After 5 stages were stimulated, engineers recalibrated the stimulation program using
microseismic monitoring data and, as a result, reduced the number of stages to 13 (Track 7). The blue
spikes (Tracks 6 and 7, left and right of stimulation stages) indicate proposed perforation cluster
locations. (Adapted from Liu et al, reference 20.)

46

increase the spacing of stages and reduce the

number of stages from 18 to 13 per well (left).
After all 26 stages were stimulated in both
horizontal wells, the operator put the wells into
production. Initial production rates were
103.2 m3/d [649.1 bbl/d] and 124.5 m3/d
[783.1 bbl/d], a three- to fourfold improvement
over the average production rate of 32 m3/d from
previous horizontal wells. After three months, the
production rates from these wells stabilized and
were 50% higher than the previous best production
from any horizontal well in the formation.
Stimulation by Design
Unconventional reservoirs provide special challenges because they are heterogeneous reservoirs
composed of highly stratied sediments. Staying
within a reservoir zone during horizontal drilling is
difcult. Consequently, the wellbore intersects
variable lithologies, which exhibit dissimilar petrophysical and mechanical properties.
Unconventional reservoirs are also usually
anisotropic and naturally fractured. Shales possess layering caused by the horizontal alignment of
nely laminated sediments and platy clay minerals. This layering causes rock properties, such as
permeability, elastic moduli and electrical resistivity, to be anisotropic.22 These properties may vary
more from layer to layer than within layers.
Natural fractures may cut across this layering and
superimpose additional anisotropy on the shales.
Both anisotropy and natural fractures complicate
the propagation of hydraulic fractures.23
Recent advances in multistage stimulation technology are making it possible to stimulate and
develop unconventional hydrocarbon resources
more successfully (see Multistage Stimulation in
Liquid-Rich Unconventional Formations, page 26).
Parallel advances in the Mangrove stimulation
design software are making it possible to design
completions that are more effective. Integration of
the two technologies promises a positive future for
unconventional resource development. RCNH
22. For a discussion of permeability anisotropy: Ayan C,
Colley N, Cowan G, Ezekwe E, Wannell M, Goode P,
Halford F, Joseph J, Mongini A, Obondoko G and Pop J:
Measuring Permeability Anisotropy: The Latest
Approach, Oileld Review 6, no. 4 (October 1994): 2435.
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, Oileld Review 6,
no. 4 (October 1994): 3647.
For more on the anisotropy of electrical properties:
Anderson B, Bryant I, Lling M, Spies B and Helbig K:
Oileld Anisotropy: Its Origins and Electrical
Characteristics, Oileld Review 6, no. 4
(October 1994): 4856.
23. Wu R, Kresse O, Weng X, Cohen C and Gu H: Modeling
of Interaction of Hydraulic Fractures in Complex
Fracture Networks, paper SPE 152052, presented at the
SPE Hydraulic Fracturing Technology Conference,
The Woodlands, Texas, February 68, 2012.

Oileld Review

Contributors
Babatunde Ajayi is a Senior Completions Engineer for
Seneca Resources Corporation in Pittsburgh,
Pennsylvania, USA. Previously, he was a eld engineer
and then a production stimulation and completions
engineer for Schlumberger and a supply chain management program intern for Halliburton. He has a BS
degree (Hons) in chemical engineering from the
Bogazii University in Istanbul, Turkey, and an MS
degree in petroleum engineering from Texas A&M
University, College Station, USA.
Franoise Allioli is a Principal Physicist and
Development Physics Team Leader for LWD tools with
Schlumberger. Since joining the company at the
Schlumberger Riboud Product Center in Clamart,
France, in 1995, she has worked on both wireline and
LWD tool projects, including nuclear and resistivity
tools. She has coauthored and presented papers at SPE,
SPWLA and other conferences. Franoise holds a PhD
degree in nuclear physics from Universit Paris Diderot.
Mark A. Andersen, Schlumberger Domain Head for
Core Physics in Houston, joined the company in 2000.
He spent 11 years as an Oileld Review editor and
executive editor before returning to his roots in core
analysis to help build a new business for Schlumberger.
He began his career in 1981 as a researcher in rock
properties at Amoco Research Center in Tulsa. He subsequently spent several years in Stavanger, where he
managed the Amoco Norway external research program and wrote Petroleum Research in North Sea
Chalk. Mark is the author of many technical papers,
including 23 articles for Oileld Review. He earned a
BS degree in engineering physics from the University
of Oklahoma at Norman, USA, and MS and PhD
degrees in physics from The Johns Hopkins University
in Baltimore, Maryland, USA.
Iroh Isaac Aso, based in Canonsburg, Pennsylvania,
has been a Production and Stimulation Engineer with
Schlumberger since 2009. Some of his responsibilities
include completions optimization and production
performance analysis of horizontal shale wells and
hydraulic fracture design and evaluation. Prior to
his career with Schlumberger, Iroh was a graduate
research assistant at the Integrated Core
Characterization Center, University of Oklahoma,
Norman. He obtained a BS degree in electrical and
electronics engineering from the University of Lagos,
Nigeria, and an MS degree in petroleum engineering
from the University of Oklahoma, Norman.
Isaac Aviles is the Schlumberger Global Portfolio
Manager for Multistage Stimulation and has been
with the company since 2000; he joined the multistage stimulation team in 2010 and began his current
assignment in 2013. Before moving to Sugar Land,
Texas, he held eld operations positions in Alaska,
USA; Argentina and Colombia. He was also a eld service manager in Mexico and spent three years performing technical support of worldwide eld
operations for fracturing and stimulation. Isaac
received a BS degree in chemical engineering from
Universidad de las Amricas, Puebla, Mexico, and an
MBA degree in nance from Rice University, Houston.

Summer 2013

Jason Baihly, based in Sugar Land, Texas, is the

Schlumberger Subsurface Manager for the Eagleville
Joint Venture Project in Texas for which he formulates
and executes plans to optimize production of horizontal Eagle Ford oil wells. Previously, he was a multistage
stimulation product line manager, was instrumental in
merging Smith and Schlumberger R&D teams for a
new business unit and created new R&D teams to
focus on multistage stimulation hardware solutions.
Prior to this role, and as the horizontal optimization
and completion team leader within the Schlumberger
Petro-Technical Services group, he led and was
exposed to many integrated shale projects. Jason is a
Principal Engineer with more than a dozen years of
industry experience. He has a bachelors degree in
civil engineering from the South Dakota School of
Mines and Technology in Rapid City, USA, and a masters degree in management in the oil and gas industry
from Heriot-Watt University in Edinburgh, Scotland.
Jacob Caplan is a Senior Completions Engineer for
PDC Mountaineer LLC in Bridgeport, West Virginia,
USA. He joined the company in 2010 and has focused
on completions in the Marcellus Shale; he has also
worked in the Utica Shale. He began his career with
Schlumberger, where he was a eld engineer in the Well
Services segment in south Texas and the Texas
Panhandle and then was engaged in starting the
Haynesville Shale play as the eld service manager for
the stimulation department in Bossier City, Louisiana,
USA. He also served as a sales engineer in Dallas. Jacob
holds a BS degree in petroleum and natural gas engineering from West Virginia University, Morgantown.
Brian D. Clark, a Senior Production and Completions
Engineer with Schlumberger in Houston, has worked
on a variety of projects in shale reservoirs. He started
his career with Schlumberger in 1994 as a eld engineer; he was involved in cementing and fracturing
operations and performed laboratory work. Brian
earned a BS degree in mechanical engineering from
Clarkson University, Potsdam, New York, USA.
Valentin Cretoiu, based at the Schlumberger Riboud
Product Center, Clamart, France, is the Project
Manager for the EcoScope 675* Measurement
Improvements and NeoScope* projects with a focus on
NeoScope commercialization. Valentin joined
Schlumberger as an electronics software engineer in
1999 in Houston. In 2001, he moved to Clamart as an
electrical engineer working on nuclear acquisition and
control design. He has been project manager for various Wireline and Drilling & Measurements tools,
recently focusing on the development of advanced
measurements for the EcoScope* tool and the
NeoScope service. He received an MS degree in electrical engineering from the Institut National des
Sciences Appliques de Lyon, France.
Brent Duncan has been the Schlumberger Routine
Core Analysis Supervisor in Houston since 2011. He
began his career in the oil industry in 2001 as a hydrogeologist with Eagon & Associates, Inc. and has since
worked as a hydrogeologist and soil technician, laboratory supervisor, geologist and project manager. Brent
obtained a BS degree in geology from Olivet Nazarene
University, Bourbonnais, Illinois, USA.

Mike Evans is a Scientic Advisor for Schlumberger

working with the EcoScope nuclear group, in Sugar
Land, Texas. He joined Schlumberger in 1981 as a development engineer and worked on several wireline
nuclear logging tools. In 1986, Mike joined the original
Schlumberger LWD project development team and has
been involved in the design of several nuclear tools.
During his career, Mike has coauthored dozens of
papers and holds 13 patents. He has a BS degree in
physics, an MS degree in computer science and a
PhD degree in physics, all from Texas A&M University,
College Station.
Utpal Ganguly is Global Stimulation Software Portfolio
Manager for Schlumberger. Based in Houston, he directs
the companys stimulation software development,
denes technology strategy and conducts product marketing. He began his career with Schlumberger as a Well
Services eld engineer. With more than 18 years of
experience in the oil eld, he has held various roles in
eld operations, engineering and marketing. Utpal
received an MS degree in petroleum engineering from
the Norwegian Institute of Technology, Trondheim, and
an MS degree in computer science from the University
of Tulsa. He currently focuses on novel applications of
software technologies enabling unconventional reservoir development and production.
Dewey Gerdom is Vice President of Eastern
Operations at Petroleum Development Corporation
(PDC) and Chief Executive Ofcer of PDC
Mountaineer LLC in Bridgeport, West Virginia. With 26
years of experience with both major and independent
producers, he previously served as vice president of
exploration, acquisitions and divestures at PDC. He
joined the company in 2000 as project manager and
was responsible for all exploration and land activities
in the Rocky Mountain region in the US. Prior to that,
he was the chief operating ofcer of Fruehauf
Production Company. He also served as an asset manager at Chevron USA and Gulf Oil Company. Dewey
holds a BS degree in business from the University of
Wyoming, Laramie, USA.
Roger Grifths is the Petrophysics Domain Head for
Schlumberger Drilling & Measurements in Petaling
Jaya, Malaysia. He joined the company in 1987 as a
wireline eld engineer and has worked in Asia, the
Middle East, Europe, Africa and North America in
eld, management, engineering and technical positions. He is a Technical Advisor in petrophysics and
well placement, has written two books, coauthored
numerous technical papers and holds several patents
related to petrophysics and well placement. Roger
obtained a degree (Hons) in mechanical engineering
from The University of Melbourne, Victoria, Australia.
Fabien Haranger is a Senior Physicist who began his
career with Schlumberger in 2006 as a physicist working on technical support for the nuclear detectors and
generators in the Drilling & Measurements (D&M)
EcoScope, arcVISION* and SlimPulse* tools. Based in
Princeton, New Jersey, USA, his current project is to
develop new nuclear detectors for D&M and Wireline
tools. Fabien has a degree in engineering from the
Ecole Nationale Suprieure dIngnieurs de Caen,
France, an MS degree in physics from Universit ParisSud and a PhD degree in physics from Universit de
Caen Basse-Normandie, France.

47

Xianwen Li is the Vice President of the Petroleum

Technology Institute of PetroChina Changqing Oileld
Company, based in Xian, Shaanxi, Peoples Republic of
China. He is also a senior technical expert recognized by
the China National Petroleum Corporation. He has been
engaged in research and management work in oil and
gas eld development since 1985. His major interest is
in production enhancement technologies. Xianwen
received a masters degree in petroleum production
engineering from Daqing Petroleum Institute and a PhD
degree in materials science from Xian Jiaotong
University, both in the Peoples Republic of China.
Guang Hua Liu is Cell Lead of the technical design
cell for the Oil Development Segment of the CNPCDagang Oileld Company in Tianjin, Peoples Republic
of China. His responsibilities include well design, procedure and the introduction of new techniques. He
began his career in 1999 as a junior engineer for the
Dagang Oileld Company and became an engineer and
then senior engineer with the company. Guang Hua
earned a BS degree in petroleum engineering from the
Jianghan Petroleum Institute, Jingzhou, Peoples
Republic of China.
Hai Liu is currently Well Services General Manager for
Schlumberger covering China, Japan, Korea and
Taiwan; he is based in Beijing. In his 18-year career, he
has worked in China, the Middle East and Texas and
focused on production optimization and stimulation.
He moved to a management role in early 2012. His
main areas of expertise are in reservoir production
and stimulation. He is recognized as Principal
Engineer in Schlumberger Eureka Technical Careers
and received Performed by Schlumberger Bronze,
Silver and Chairman awards for his technical contributions. Before joining Schlumberger, he worked for the
Production Technology Research Institute of
PetroChina Xinjiang Oil Company from 1995 to 2000.
Hai holds a BS degree in petroleum geology from Xian
Petroleum Institute, Peoples Republic of China.
Yin Luo is Senior Technical Sales Engineer for
Schlumberger China. He joined Schlumberger in 2006
after earning a masters degree in computer science
from Peking University in Beijing. He has worked in
Nigeria, Texas and Beijing and is currently based in
Chengdu, Sichuan, Peoples Republic of China. His primary areas of expertise are in reservoir stimulation
and software architecture. Yin is the recipient of two
Performed by Schlumberger Bronze awards for his
technical contributions.
Marie-Laure Mauborgne is a Senior Physicist working
on nuclear measurements for Drilling & Measurements
at the Schlumberger Riboud Product Center in
Clamart, France. Her career with Schlumberger began
in 2006, when she worked on the EcoScope tool as a
postdoctoral physicist. Marie-Laure earned a BS
degree in nuclear engineering and an advanced degree
in nuclear physics from Ecole Nationale Suprieure
dIngnieurs de Caen, France, and a PhD degree in
nuclear physics from the University of Caen at the
Institut de Recherche sur les Lois Fondamentales de
lUnivers, Saclay, France.
Ryan McLin, who is a Senior Geologist, Petrology
Manager and Diffuse Reectance Infrared Fourier
Transform Spectroscopy Product Champion, works at
the Schlumberger Reservoir Laboratory in Houston. In
his early career, which centered on the environmental
sciences, he worked as a soils laboratory technician. In
2005, he moved to support the oileld industry, when

48

he began as a geologist at TerraTek* Services in Salt

Lake City, Utah, USA. Ryan has BS and MS degrees in
geology from New Mexico Institute of Mining and
Technology, Socorro, USA.
Doug Murray is LWD Domain Champion and
Petrophysics Advisor for Schlumberger in Abu Dhabi,
UAE. Since joining Schlumberger in 1982, he has held
various positions in the eld and in management, engineering and formation evaluation. His career includes
assignments in Canada, Algeria, Nigeria, Saudi Arabia,
Trinidad and Tobago, Argentina, Japan and China.
Doug holds a BS degree in electrical engineering from
Lakehead University, Thunder Bay, Ontario, Canada,
and an MA degree in management from the University
of Hull, Yorkshire, England. He is a member of the
SPWLA, the SPE and SEG.
Nicole Reichel is the LWD Petrophysics Associate
Domain Champion and Senior Petrophysicist for
Schlumberger in Stavanger, Norway. She joined
Schlumberger in 2007 and worked as an LWD eld
engineer in Canada and Oman prior to her post at the
Schlumberger Riboud Product Center in Clamart,
France, where she was part of the LWD petrophysics
interpretation group. In this function, she participated
in the development of interpretation and answer products for LWD acquisition and provided global petrophysics eld support. Nicole teaches courses in LWD
petrophysics and has served as an SPWLA
Distinguished Lecturer. She obtained BSc and MSc
degrees in applied geosciences at the
Montanuniversitt in Leoben, Austria.
Christian Stoller is a Schlumberger Scientic Advisor
and Physics Mtier Manager for the Princeton
Technology Center, New Jersey, and for the Houston
Formation Evaluation Center. Before joining
Schlumberger, he worked on a variety of phenomena in
nuclear, atomic and applied physics at the Nuclear
Physics Laboratories of Eidgenssische Technische
Hochschule (ETH) Zurich, Switzerland, and at
Stanford University, California, USA. During more than
20 years with Schlumberger, he has been involved in
the design and testing of most of the Schlumberger
nuclear wireline tools for downhole applications. In
2005, he moved to the Princeton Technology Center,
where he helped with the development of photomultipliers and detectors for downhole applications and the
development and manufacturing of pulsed neutron
generators. Most recently, he worked with the team
that developed the rst LWD tool that incorporates a
pulsed neutron generator. Chris received MS and PhD
degrees in physics from ETH Zurich.
Ira J. Jay Terry, Jr. joined Schlumberger in 1987 as a
Wireline eld engineer in Shreveport, Louisiana, and
has held eld, sales and technical support positions in a
variety of North American locations. Currently in
Canonsburg, Pennsylvania, he has worked for the past
18 years as Senior Sales Engineer in the Charleston,
West Virginia, and Pittsburgh, Pennsylvania, ofces,
amassing extensive technical expertise in the Marcellus,
Utica and other Appalachian region reservoirs. Jay holds
a BS degree in geological engineering from Louisiana
State University, Baton Rouge, and a BS degree in geology from Centenary College of Louisiana, Shreveport.
Kirby Walker has served in numerous operational,
management and technical capacities for
Schlumberger in south Texas, Alaska and the
Appalachian basin, USA; Venezuela; and Russia in his
11-year career. He has concentrated on production
engineering and stimulation techniques. Previously, as

stimulation domain manager in Canonsburg,

Pennsylvania, Kirby managed a group of completion
engineers focused on designing and analyzing shale
completions for the Northeast US. He is Chair of the
SPE Pittsburgh Petroleum Section and has a BS
degree in petroleum and natural gas engineering from
Pennsylvania State University, University Park.
George Waters is an Unconventional Completions
Technical Manager for Schlumberger in Oklahoma
City, Oklahoma. He is responsible for the geomechanical assessment and completion design of organic shale
reservoirs under exploration and for initial reservoir
development outside of North America. He has performed appraisals of hydraulic fracture dimensions
and producibility via laboratory and eld measurements on multiple reservoirs in more than 20 countries
on six continents. George has been involved in the
stimulation optimization of organic shales since 2000
and has extensive experience in shale basins throughout North America. He earned a BS degree in petroleum engineering from West Virginia University,
Morgantown, an MS degree in environmental engineering from Oklahoma State University, Stillwater, and an
MS degree in petroleum engineering from Institut
Franais du Petrole, France. George was a 20092010
SPE Distinguished Lecturer on completion of organic
shale reservoirs.
Kevin Wutherich, based in Canonsburg, Pennsylvania,
is the Stimulation Domain Expert for Schlumberger in
the Northeast US. He is primarily engaged in optimizing completion practices in the Marcellus and Utica
shales with a focus on integrating expertise from multiple domains. Previously, he was the stimulation
domain manager for Schlumberger in Europe, where
he engaged in numerous high-prole projects, including the rst shale gas developments in Europe. Kevin
holds a bachelors degree in chemical engineering
from the University of Waterloo, Ontario, Canada.
Yonggao Xu recently became Deputy General Manager
and Chief Engineer for Shaanxi Yanchang Petroleum
(Group) Corporation Ltd. in Xian, Shaanxi, Peoples
Republic of China. Previously, he served as deputy
chief engineer of PetroChina Changqing Oileld
Company. Yonggao received a bachelors degree in
petroleum production engineering from Jianghan
Petroleum University, Jingzhou, Hubei, and a masters
degree in reservoir and production engineering from
Southwest Petroleum University, Chengdu, Sichuan,
both in the Peoples Republic of China.
Hua Yang is Deputy General Manager for PetroChina
Changqing Oileld Company in Xian, Shaanxi,
Peoples Republic of China, and in charge of oil and
gas exploration and appraisal for reservoirs in the
Ordos basin, Peoples Republic of China. He is also the
Director of the National Engineering Laboratory for
Low Permeability Oil and Gas Field Exploration and
Development. He graduated from the Southwest
Petroleum University in Nanchong, Sichuan, Peoples
Republic of China. Hua has also been recognized as
professoriate senior engineer by the Peoples Republic
of China for his contribution to the discovery of the
Jiyuan, Huaqing, Yulin and Sulige oil and gas reservoirs and holds a rst prize and second prize for
National Science and Technology Progress in the
Peoples Republic of China.
An asterisk (*) is used to denote a mark of Schlumberger.

Oileld Review

NEW BOOKS

Coming in Oileld Review

diamondoids chemistry, their molecular structure and physical properties.

Periic O: Book Reviews, Journal of
Nanophotonics 7, no. 1 (April 4, 2013):
079899-1079899-2.

Diamondoid Molecules:
With Applications in
Biomedicine, Materials
Science, Nanotechnology &
Petroleum Science
G. Ali Mansoori, Patricia Lopes
Barros de Araujo and Elmo Silvano
de Araujo
World Scientic Publishing Company
27 Warren Street, Suite 401402
Hackensack, New Jersey 07601 USA
2012. 424 pages. US$ 128.00

As someone who has worked in

nanoscience for many years, I was
gratied to see a book that links the
dots in the different areas of science
and technology. Diamondoid
Molecules opens a door for newcomers, even in industry and government,
and for seasoned researchers who
want to explore new opportunities in
other areas of the research. I strongly
recommend the book to researchers
and students in all the relevant elds.
Zhang G: Book Review, Physics Today 66, no. 3
(March 2013): 5960.

ISBN: 978-981-4291-60-6

Diamondoids are stable, saturated

ringed hydrocarbons with applications
in elds ranging from chemistry to
geology. This book looks at the history
and fundamentals of diamondoid
science and technology and examines
the importance of the diamondoid
molecule to the global scientic
community. The authors explore both
theoretical and experimental work
with diamondoids.
Contents:
Molecular Structure and Chemistry
of Diamondoids
Diamondoids in Petroleum and
Other Fossil Fuels
Physical Properties of Diamondoids
Diamondoids as Nanoscale
Building Blocks
Properties of Diamondoids Through
Quantum Calculations
Biomedical Applications of
Diamondoids
Diamondoids in Materials Science
Glossary, Index
Diamondoid Molecules is a
thorough presentation of current and
future applications of diamondoids.
. . . Every chapter is excellently
illustrated with plenty of clearly
designed gures and graphs and
well-organized tables. . . . [T]he book is
an excellent introduction to theoretical and technological aspects behind

Summer 2013

Maverick Genius: The

Pioneering Odyssey of
Freeman Dyson
Phillip F. Schewe
Thomas Dunne Books, an imprint of
St. Martins Press
175 Fifth Avenue
New York, New York 10010 USA
2013. 352 pages. US$ 27.99
ISBN: 978-0-312-64235-8

Freeman Dyson, one of the foremost

physicists of our time, has played many
roles: mathematician, astronomer,
biologist and engineer, among others.
This biography explores his life and the
impact his work has had on the world.
Considered a genius by many, Dyson
colleague and friend of J. Robert
Oppenheimer, George F. Kennan and
Richard Feynmanalso won the
Templeton Prize for his writing on
science and religion.
Contents:
Killing Time: Dyson Bombs Berlin
(19231945)
Life Is a Blur: Dyson as
Mathematician (19451947)
Ecumenical Councils: Dyson as
Seminarian (19471948)
The Secret Signature of Things:
Dyson as Artist (19481949)
Recessional: Dyson as Professor
(19491953)

Nuclear Opera: Dyson and the

Cold War (19541956)
Intrinsically Safe: Dyson as
Engineer (19561957)
Space Travelers Manifesto:
Dyson as Rocketeer (19571959)
Civilized Behavior: Dyson Searches
for Extraterrestrial Intelligence
(Early 1960s)
Nuclear Manifesto: Dyson as
Diplomat (Early 1960s)
On the Oregon Trail: Dyson as
Pentagon Consultant (1960s1970s)
Success in Life: Dyson as
Astronomer (Mid 1960s to
Mid 1970s)
Science and Sublime: Dyson as
Essayist (19761985)
Nuclear Slavery: Dyson as
Abolitionist (1980s)
The Arc of Life: Dyson as Biologist
(1980s and 1990s)
God and Man at Princeton:
Dyson as Preacher (19852000)
Splintering the Species: Dyson as
Heretic (19902010)
Long-Term Thinking: Dyson as
Storyteller (Recent Years)
A Many-Colored Glass
Notes, Index
A fascinating account of an
iconoclastic scientic polymath and
the lively collection of scientists who
were his friends.
Book Review, Kirkus (December 16, 2012),
https://www.kirkusreviews.com/book-reviews/
phillip-f-schewe/maverick-genius/ (accessed
April 2, 2012).

Discussion of Dysons opposition

to the Superconducting Super Collider
and to the Hubble Space Telescope . . .
would . . . have brought a sharper,
more critical focus on Dysons
contrarian personality, and made the
end of Schewes book read less like an
extended attering magazine portrait
of a prophet.
Still, Maverick Genius provides a
vivid and enjoyable sketch of one of
the most prominent scientic rebels of
our day.
Crease RP: Rebel Without a Pause, Nature 494,
no. 7437 (February 21, 2013): 311.

Despite some strange digressions,

Dr. Schewe knows the physics, and he
gained access to colleagues, family
everyone except Freeman Dyson, who
politely declined, saying, Maybe in
50 years youll be able to tell whether
I did anything important.
Hoffman J: Growing Wings and Rising Oceans,
The New York Times (February 25, 2013),
http://www.nytimes.com/2013/02/26/science/
growing-wings-and-rising-oceans.html?_r=0
(accessed April 2, 2013).

Mudrocks. Until the late 20th century, mudrocksincluding shales

were largely ignored except for their
source rock potential and capacity to
form seals, trapping hydrocarbon
beneath them. Research is now
revealing that mudrocks are as interesting and complicated as sandstones and carbonates. They form
under similar hydrodynamic ow
regimes as sandstones and carbonates, and their characteristics are
predictable using the principles of
sequence stratigraphy. Consequently,
techniques used to characterize
sandstone and carbonate reservoirs
may be adapted for characterizing
mudrock resources.
Underreaming. To enlarge the wellbore diameter below a casing shoe
or other restriction that might limit
the maximum diameter of a reaming
tool, operators employ underreamers. These tools have been used for
decades with varying degrees of
success. Advanced underreamer
designs, coupled with innovative
activation and deactivation techniques, are helping operators drill
deep or extended-reach wells with
greater efciency.
Measuring Casing Corrosion.
The bane of all things metal is corrosion. In the upstream industry, tubulars, designed to protect the well
and the environment, are continuously exposed to naturally occurring
elements and chemicals that aggressively promote corrosion. Operators
can use monitoring to determine the
extent and location of corrosion in
time to manage it. By implementing
advanced downhole corrosion monitoring techniques and tools, companies are protecting their investments
and the environment.
Geomagnetic Referencing.
Accurate positioning of well trajectories is required to optimize hydrocarbon recovery, determine where
each well is relative to the reservoir
and avoid collisions with other
wells. Advances in geomagnetic
referencing now allow companies
to use real-time data from accelerometers, magnetometers and gyroscopes to land horizontal wells
according to plan. This article
examines the science of well guidance, focusing on modern magnetic
surveying techniques.

49

Crowd-sourced science has rarely

been so thrilling. As Deborah R. Coen
reveals, the rumbustious history of
seismology began with roving scientists gathering locals accounts of
shocks, shudders and thumps. . . .
Coen argues for a hybridized
disaster science, factoring in such
responses from human seismographs
with geology and instrumental data.

The Great Fossil Enigma:

The Search for the
Conodont Animal

The Earthquake Observers:

Disaster Science from Lisbon
to Richter

Simon J. Knell
Indiana University Press
Ofce of Scholarly Publishing
Herman B Wells Library 350
1320 East 10th Street
Bloomington, Indiana 47405 USA
2012. 440 pages. US$ 45.00

Deborah R. Coen
University of Chicago Press
1427 East 60th Street
Chicago, Illinois 60637 USA
2012. 360 pages. US$ 35.00

ISBN: 978-0-253-00604-2

Eyewitness accounts of earthquakes

once knitted together the stories of
seismic devastation. The author
explores this citizen science, which was
abandoned with the introduction of the
Richter scale, through commentary
from observers such as Charles Darwin,
Mark Twain, Ernst Mach and John Muir
as well as from ordinary citizens, and
discusses how it developed into a eld
of scientic research. The book tells the
history of this dialogue between scientists and citizens, which has since been
revived in the 21st century.

Through an exploration of 150 years

of scientic thinking, studies,
misconceptions and misunderstanding
of the conodont, the author reveals an
emerging consensus among scientists
about this creatures place in the
fossil records.
Contents:
The Road to El Dorado
A Beacon in the Blackness
The Animal with Three Heads
Another Fine Mess
Outlaws
Spring
Diary of a Fossil Fruit Fly
Fears of Civil War
The Promised Land
The Witness
The Beast of Bear Gulch
The Invention of Life
El Dorado
Over the Mountains of the Moon
Afterword: The Progress of Tiny
Things
Notes, Index
. . . Knell uses the history of
conodont research to show how the
ideas and actions of scientists are
inuenced not merely by the clinical
interpretation of the evidence but also
by their imagination. . . . If you want
[an] entertaining and interesting
account of the discovery of knowledge
through the analytical, political, and
idiosyncratic activities of researchers,
The Great Fossil Engima will serve
you well.
Donoghue P: Fascinating Little Whatzits,
Science 340, no. 6134 (May 17, 2013): 813.

Books in Brief, Nature 491, no. 525

(November 21, 2012), http://www.nature.com/
nature/journal/v491/n7425/full/491525a.html
(accessed July 1, 2013).

The author . . . blends the condensed-matter topics in a manner that

blurs the distinction between soft and
hard parts. Thats a novel idea.
Potentially, this book will modernize
the undergraduate condensed-matter
physics course by introducing more of
the soft-matter component.
. . . Since the book covers such a
wide variety of topics, the level of
detail suffers a bit.
Overall, though, [the book]
succeeds at covering many fundamental concepts of solid-state and softmatter physics and at combining them
in an approachable manner.
Smalyukh I: An Introduction that Blends Hard
and Soft Condensed Matter, Physics Today 66,
no. 5 (May 2013): 49.

ISBN: 978-0-226-11181-0

Fundamentals of Condensed
Matter and Crystalline Physics
David L. Sidebottom
Cambridge University Press
32 Avenue of the Americas
New York, New York 10013 USA
2012. 418 pages. US$ 75.00
ISBN: 978-1-107-01710-8

Contents:
The Human Seismograph
The Planet in the Village: Comrie,
Scotland, 17881897
News of the Apocalypse
The Tongues of Seismology:
Switzerland, 18551912
Geographies of Hazard
The Moment of Danger
Fault Lines and Borderlands:
Imperial Austria, 18801914
What Is the Earth?
The Youngest Land: California,
18531906
A True Measure of Violence:
California, 19061935
Conclusion, Notes, Bibliography,
Index
. . . The author demonstrates how
the approach, and even the goals, of
earthquake science are intertwined
with and inuenced by their historical
and political context. The book is well
written, the documentation meticulous, and the depth of research
impressive. . . . The chronology of
attempts to recruit amateur earthquake observers that Coen assembles
to make her case is fascinating, and
on that basis alone the book is worth
reading.

Intended for undergraduates, this

textbook combines solid-state physics
with condensed matter physics. The
author takes a holistic approach to
condensed matter physics by integrating
the crystalline and amorphous states and
relating areas of condensed matter
such as electronic properties of solids
and statistical mechanics of hard and
soft materialsto one another.
Contents:
Part I. Structure: Crystal Structure;
Amorphous Structure; Bonds and
Cohesion; Magnetic Structure
Part II. Scattering: Scattering
Theory; Scattering by Crystals;
Scattering by Amorphous Matter;
Self-Similar Structures and Liquid
Crystals
Part III. Dynamics: Liquid
Dynamics; Crystal Vibrations;
Thermal Properties; Electrons: The
Free Electron Model; Electrons:
Band Theory; Bulk Dynamics and
Response
Part IV. Transitions: Introduction to
Phase Transitions; Percolation
Theory; Mean Field Theory and
Renormalization; Superconductivity
Appendices, Index

Beroza GC: Did You Feel It?, Science 340,

no. 6130 (April 19, 2013): 274275.

50

Oileld Review

DEFINING HYDRAULIC FRACTURING

Elements of Hydraulic Fracturing

P breakdown

Richard Nolen-Hoeksema
Editor

P reopening

The Physics of Fracturing

The size and orientation of a fracture, and the magnitude of the pressure needed to create it, are dictated by the formations in situ stress
field. This stress field may be defined by three principal compressive
stresses, which are oriented perpendicular to each other (below). The
magnitudes and orientations of these three principal stresses are determined by the tectonic regime in the region and by depth, pore pressure
and rock properties, which determine how stress is transmitted and distributed among formations.
In situ stresses control the orientation and propagation direction of
hydraulic fractures. Hydraulic fractures are tensile fractures, and they
open in the direction of least resistance. If the maximum principal com-

Fracture

Hmax
Fracture

Hmin

> In situ stresses and hydraulic fracture propagation.

The three principal compressive stresses (red arrows)
are a vertical stress (V) and a maximum and
minimum horizontal stress (Hmax and Hmin). Hydraulic
fractures open in the direction of the least principal
stress and propagate in the plane of the greatest and
intermediate stresses.
Oileld Review Summer 2013: 25, no. 2.
Copyright 2013 Schlumberger.
For help in preparation of this article, thanks to Jerome Maniere, Mexico City.

Summer 2013

Pumping rate

A wells ability to produce hydrocarbons or receive injection uids is limited

by the reservoirs natural permeability and near-wellbore changes resulting
from drilling or other operations. Hydraulic fracturing, also known as
hydraulic stimulation, improves hydrocarbon ow by creating fractures in
the formation that connect the reservoir and wellbore.
A hydraulic fracture is a pressure-induced fracture caused by injecting
uid into a target rock formation. Fluid is pumped into the formation at
pressures that exceed the fracture pressurethe pressure at which rocks
break. To access a zone for stimulation, engineers perforate the casing
across the interval and use retrievable plugs to isolate the interval from
other open zones. This interval is then pressurized to the formation breakdown pressure, or fracture initiation pressure, the point at which the rock
breaks and a fracture is created.

Bottomhole pressure, P

P closure

Reopening

After
closure
Breakdown
P initial
Time

> Fracture pressures. During a stimulation treatment,

engineers pump uid into the targeted stimulation
zone at a prescribed rate (blue polygons), and
pressure (red line) builds to a peak at the breakdown
pressure, then it drops, indicating the rock around
the well has failed. Pumping stops and pressure
decreases to below the closure pressure. During a
second pumping cycle, the fracture opens again at
its reopening pressure, which is higher than the
closure pressure. After pumping, the fracture closes
and the pressure subsides. The initial pore pressure
is the ambient pressure in the reservoir zone.

pressive stress is the overburden stress, then the fractures are vertical,
propagating parallel to the maximum horizontal stress when the fracturing pressure exceeds the minimum horizontal stress.
The three principal stresses increase with depth. The rate of increase
with depth denes the vertical gradient. The principal vertical stress,
commonly called the overburden stress, is caused by the weight of rock
overlying a measurement point. Its vertical gradient is known as the lithostatic gradient. The minimum and maximum horizontal stresses are the
other two principal stresses. Their vertical gradients, which vary widely by
basin and lithology, are controlled by local and regional stresses, mainly
through tectonics.
The weight of the uid above a measurement point in normally pressured basins creates in situ pore pressure. The vertical gradient of pore
pressure is the hydrostatic gradient. However, pore pressures within a
basin may be less than or greater than normal pressures and are designated
as underpressured or overpressured, respectively.
Beyond Fracture Initiation
At the surface, a sudden drop in pressure indicates fracture initiation, as
the uid ows into the fractured formation. To break the rock in the target
interval, the fracture initiation pressure must exceed the sum of the minimum principal stress plus the tensile strength of the rock. To nd the fracture closure pressure, engineers allow the pressure to subside until it
indicates that the fracture has closed again (above). Engineers nd the
fracture reopening pressure by pressurizing the zone until a leveling of
pressure indicates the fracture has reopened. The closure and reopening
pressures are controlled by the minimum principal compressive stress.

51

DEFINING HYDRAULIC FRACTURING

Therefore, induced downhole pressures must exceed the minimum principal stress to extend fracture length.
After performing fracture initiation, engineers pressurize the zone for
the planned stimulation treatment. During this treatment, the zone is pressurized to the fracture propagation pressure, which is greater than the
fracture closure pressure. Their difference is the net pressure, which represents the sum of the frictional pressure drop and the fracture-tip resistance
to propagation.

X,200

Y,000

Depth, ft

X,600

Y,400

Y,800

Keeping Fractures Open

The net pressure drives fracture growth and forces the walls of the fracture
apart, creating a width sufcient to allow the entry of the fracturing slurry
composed of uid and proppantsolids that hold the fracture open after
pumping stops.
Once the pumping is halted, the pressures inside a fracture subside as
the uids either ow back into the well or leak away into the reservoir
rock. This drop in pressure allows the fracture to close again. To ensure
that fractures stay open, engineers inject additional materials, depending
on lithology. In sandstone or shale formations, they inject proppant
sand or specially engineered particlesto hold fractures open (below).
In carbonate formations, they pump acid into the fractures to etch the
formation, creating articial roughness.
The stimulation treatment ends when the engineers have completed
their planned pumping schedule or when a sudden rise in pressure indicates that a screenout has taken place. A screenout is a blockage caused by
bridgingaccumulation, clumping or lodgingof the proppant across the
fracture width that restricts uid ow into the hydraulic fracture.
Controlling Hydraulic Stimulation
Stimulation engineers maintain a constant rate of fluid injection. The
volume injected includes the additional volume created during fracturing and the fluid loss to the formation from leakoff through the permeable wall of the fracture. However, the rate of fluid loss at the growing
fracture tip is extremely high. Therefore, it is not possible to initiate a
fracture with proppant in the fracturing fluid because the high fluid loss
would cause the proppant at the fracture tip to reach the consistency of
a dry solid, causing bridging and screenout conditions. Consequently,
some volume of clean fluida padmust be pumped before any proppant is pumped.

1 mm

1 mm

1 mm

> Proppant. Several proppant types, including high-strength bauxite

(left), resin-coated silica (middle) and lightweight ceramic (right), are
pumped into fractures to maintain open fractures for enhanced
hydrocarbon production.

52

Z,200

400

800

1,200

1,600

Horizontal

2,000

2,400

2,800

3,200

3,600

, ft

departure

> Microseismic monitoring of multiple-stage hydraulic stimulation.

Analysis of microseismic data provides operators with information about
the effectiveness of hydraulic stimulation treatments. In this example,
five fracturing stages were pumped into the treating well (red line) while
being monitored from a second well (green line with location of
geophones shown as green disks). Microseismic events during stages 1
through 5 are indicated by the yellow, blue, red, cyan and magenta dots,
respectively. Real-time microseismic monitoring may allow completion
engineers to adjust operations during execution to improve the
effectiveness of the treatment.

When designing a hydraulic fracture treatment, engineers must establish the leakoff rate and volume of the pad in relation to the timing of
slurry and proppant injection so that when the fracture reaches its
designed length, height and width, the rst particle of proppant reaches
the fracture tip. To design a hydraulic fracturing job, engineers must
understand how pumping rate and stimulation uid properties affect
hydraulic fracture geometry and propagation within the in situ stress eld
to achieve a targeted propped fracture length.
Operators design stimulation treatments to control fracture propagation
and to ensure that the hydraulic fracture stays within the reservoir and does
not grow into the adjacent formation. To reduce this risk, operators monitor
fracture growth. As fracturing uid forces the rock to crack and fractures
grow, small fragments of rock break, causing tiny seismic emissions, called
microseisms. Geophysicists are able to locate these microseisms in the subsurface (above). Laboratory and eld data have shown that these microseisms track growing fractures. Armed with the knowledge of the direction
of fracture growth, engineers may be able to take action to steer the fracture into preferred zones or to halt the treatment before the fracture grows
out of the intended zone.
The propagation of hydraulic fractures obeys the laws of physics. In
situ stresses control the pressure and direction of fracture initiation and
growth. Engineers carefully monitor the stimulation process to ensure it
goes safely and as planned.

Oileld Review

SCHLUMBERGER OILFIELD REVIEW

SUMMER 2013

VOLUME 25 NUMBER 2