Oilfield Review - Schlumberger/media/Files/resources/oilfield_review/ors15/... · Oilfield Review...

September 2015

Borehole Imaging

Sand Control

Advanced Fixed Cutter Bit

CO2 Engineering

Oilfield Review

15-OR-0003

Oilfield Review AppsOilfield Review communicates advances in finding and producing hydrocarbons to oilfield professionals. Articles from the journal are augmented on the apps with animations and videos, which help explain concepts and theories beyond the capabilities of static images. The apps also offer access to several years of archived issues in a compact format that retains the high-quality images and content you’ve come to expect from the print version of Oilfield Review.

To download and install the app on your Android† or iPad‡ device, search for “Schlumberger Oilfield Review” in the App Store‡ or Google Play† online store or scan the QR code below, which will take you directly to the device-specific source.

†Android and Google Play are marks of Google Inc. ‡App Store and iPad are marks of Apple Inc., registered in the US and other countries.

Anyone working in the upstream oil and gas industry under-stands the importance of technology for finding new reserves and producing what’s been discovered. Without continuous innovation in E&P technologies, the world’s thirst for oil and gas would have long ago outstripped their availability. Doomsday practitioners who predict the “end of oil” consistently fail to appreciate the ability of science and technology to extend the era of hydrocarbons. This issue of Oilfield Review, like every issue, bears witness to essential technology breakthroughs, from those for drilling to comple-tion to those for defining reservoirs and challenges beyond.

Finding and producing oil and gas is fraught with obsta-cles and risk, and pioneers in the early days were as chal-lenged to innovate as we are today. Upstream innovators have come from all walks of life—from academia, from machine shops and sometimes from a life that had previ-ously nothing to do with oil and gas. By ingenuity, perse-verance and circumstance, larger-than-life characters have transformed wild and improbable ideas into viable busi-nesses, many of which still bear their names.

The recently published Groundbreakers: The Story of Oilfield Technology and the People Who Made It Happen tells the story of these innovators, from the earliest days to the present day. In the four years that Mark Mau and I took to write this book, we interviewed more than 120 sci-entists and engineers, many of whom are familiar names, and perused the huge literature provided by the industry’s professional societies. This new book covers all players in the industry, and because of our editorial independence, shows bias to none.

Whoever is plying upstream technology, whether an oil company, the service industry or a startup, faces chal-lenges that are compelling and unique to the oil field. By instinct, the industry is wary of novelty—and for good rea-son. The natural risks of operating remotely at the bottom of deep oceans, releasing pressure deep in the earth and producing highly combustible substances to the surface warrant a highly conservative and safety-conscious atti-tude. In addition, incremental technology improvements often necessitate an adjustment to the bigger engineering picture. Money may be thrown at new ideas, but progress can remain slow.

In the end, research and development budgets and return on investment calculations provide only part of the story. All the money in the world is worthless without the right idea. And ideas come from people. Luckily for the industry and the world, exceptional individuals have stepped up, and such individuals have been determined to find a better way, whatever it takes.

Groundbreakers: The Birth and Growth of an Industry

1

From its inception, Oilfield Review has sought to honor these inventors and innovators and communicate the most significant advances in upstream technology. And as the Schlumberger portfolio and presence have grown, so too has the range of topics covered by the journal. Explaining the new requires patience and careful exposition. I am proud that Oilfield Review continues this vital tradition in the same spirit in which we began it more than 25 years ago.

Henry EdmundsonDirector, R9 Energy Consultants LimitedCambridge, England

Henry Edmundson worked more than 45 years for Schlumberger, was founding editor of the Oilfield Review and now runs his own energy consulting business.

www.slb.com/oilfieldreview

Schlumberger

Oilfield Review1 Groundbreakers: The Birth and Growth of an Industry

Editorial contributed by Henry Edmundson, Director, R9 Energy Consultants Limited

4 Imaging: Getting the Picture Downhole

Downhole image logs help geologists identify and analyze reservoir features such as fractures, folds and faults. Stratigraphic features, including paleotransport direction and the presence of bioturbation, clasts and scours, can also be seen in image logs. Acquiring quality images in oil-base mud systems has been more difficult because oil and mudcake often render conductivity-based imaging tools ineffective. A newly introduced imaging tool provides photorealistic quality images even in oil-base mud environments.

Oilfield Review SEPTEMBER 15Imaging Fig 19ORSEPT 15 IMG 19

22 Sand Screen Selection

Sand control equipment is typically selected based on time-honored methods and laboratory tests. Recent research suggests a more efficient, more accurate way may exist for operators to select an optimal sand control strategy.

Executive EditorCharlie Cosad

Senior EditorsTony SmithsonMatt VarhaugRick von Flatern

EditorsIrene FærgestadRichard Nolen-Hoeksema

Contributing EditorsDavid AllanGinger Oppenheimer

Design/ProductionHerring DesignMike Messinger

Illustration Chris LockwoodMike MessingerGeorge Stewart

PrintingRR Donnelley—Wetmore PlantCurtis Weeks

Oilfield Review is published and printed in the USA.

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

© 2015 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.

About Oilfield ReviewOilfield 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.

On the cover:

A bit is being readied for its trip into a test well. Advances in materials and manufacturing have led to the develop-ment of a new type of cutter. The conical diamond element is being incorporated into a variety of fixed cutter bits. The bit (inset) employs the new conical diamond element cutters along with polycrystalline diamond compact cutters.

2

September 2015Volume 27Number 2

ISSN 0923-1730

3

30 A New Approach to Fixed Cutter Bits

A new drillbit design incorporates conically shaped diamond cutting elements across the bit face for greater resistance to impact and wear. This bit, which has been tested in more than 1,000 wells around the world, has helped operators increase run lengths and sustain high rates of penetration through notoriously difficult formations.

36 Carbon Dioxide—Challenges and Opportunities

Its function in climate change has caused carbon dioxide to be a topic of significant public interest and scientific investiga-tion and a focus of hydrocarbon operators. Ongoing projects in the oil field reflect several priorities—managing carbon dioxide’s corrosive effects, using it to recover more oil after waterflood and storing it in underground formations.

CO2 solid CO2 liquid

CO2 gas

CO2 supercriticalfluid

Oilfield Review SPRING 15CO2 Fig 1ORSPRNG 15 CO2 1

Hani Elshahawi Shell Exploration and Production Houston, Texas, USA

Gretchen M. Gillis Aramco Services Company Houston, Texas

Roland Hamp Woodside Energy Ltd. Perth, Australia

Dilip M. Kale ONGC Energy Centre Delhi, India

George King Apache Corporation Houston, Texas

Michael Oristaglio Yale Climate & Energy Institute New Haven, Connecticut, USA

Advisory Panel

Editorial correspondenceOilfield Review 5599 San FelipeHouston, TX 77056United States(1) 713-513-3760E-mail: [email protected]

Distribution inquiriesMatt VarhaugOilfield Review 5599 San FelipeHouston, TX 77056United States(1) 713-513-2634E-mail: [email protected]

Oilfield Review onlineAll Oilfield Review issues and the complete defining series are availableat www.slb.com/oilfieldreview. Sign up there for an email alert to find out when new issues are available online and in the app.

51 Contributors

53 Coming in Oilfield Review

54 The Defining Series: Artificial Lift

The series provides E&P professionals with concise, authoritative summaries of a wide range of industry topics.

56 September 2015 Article Summaries

4 Oilfield Review

Imaging: Getting the Picture Downhole

Geologists and petrophysicists use image logs to visualize rocks in situ and determine

structural geometry and formation properties. Image data help them analyze reservoir

properties such as heterogeneity, sedimentary conditions and structural features,

including fractures, folds and faults. Engineers have found that acquiring images

in oil-base mud systems is difficult because the insulating properties of oil often

renders conductivity-based imaging tools ineffective, especially for fracture analysis.

That limitation has been addressed with a newly introduced imaging tool for

oil-base mud systems.

Janice BrownFort Worth, Texas, USA

Bob DavisOklahoma City, Oklahoma, USA

Kiran Gawankar Southwestern EnergyThe Woodlands, Texas

Anish KumarBingjian LiCamron K. Miller Houston, Texas

Robert LarongaPeter SchlichtClamart, France

Oilfield Review 27, no. 2 (September 2015).Copyright © 2015 Schlumberger.adnVISION, FMI, FMI-HD, Formation MicroScanner, MicroScope HD, OBMI, OBMI2, Quanta Geo, Sonic Scanner, SonicScope and UBI are marks of Schlumberger.

A picture is worth a thousand words because visualizing an object or concept is a powerful means of assimilating large amounts of informa-tion. Geologists and petrophysicists may use imaging tools to visualize downhole formations. These tools provide information that can be cru-cial for determining rock and formation proper-ties, especially when physical core samples are not available. Wireline logging tools that can image the borehole are based on dipmeter tools, which were originally designed to determine for-mation geometry and structural properties.

The evolution of imaging tools is part of a long history of petrophysical tool development. The first wireline logs were euphemistically referred to as electrical coring; some of the early logging units displayed “Electrical Coring” below the Schlumberger name (Figure 1). And yet, early wireline logs offered far too little information to substitute for coring. Service providers advancing the science of well logging have developed tools that probe the structure and mineralogy of for-mations almost to the level available from studies performed on cores.1 Images that represent the

Figure 1. Electrical coring. As evidenced by this 1932 photograph from the California, USA, oil fields, the originators of the wireline logging industry envisioned the concept of electrical coring.


1. For more on coring services: Andersen MA, Duncan B and McLin R: “Core Truth in Formation Evaluation,” Oilfield Review 25, no. 2 (Summer 2013): 16–25.

September 2015 55September 2015

6 Oilfield Review

electrical properties of the borehole may also provide geologists with core-like visualizations of downhole conditions.

Cores, however, are preferred by geologists studying downhole rock properties although the length of coring intervals is often limited by costs, and physical changes in the cores can occur while bringing the samples to the surface. From a cost and coverage standpoint, modern logging tools can sometimes provide details of

the reservoir that might otherwise be unavailable from physical cores. Although images cannot replace cores, they can provide qualitative and quantitative visual information when core is absent; from a visual perspective, they are per-haps the closest devices available for meeting that original electrical coring vision.

The first imaging devices—introduced in the 1980s—were developed from tools designed to acquire dipmeter measurements.2 Dipmeter tools

use a combination of electrical and mechanical sensors to acquire data from which the magni-tude and direction of formation dip can be deter-mined. Geologists use dip information to help them understand the subsurface geometry of geo-logic structures; the information may then be used to project structural geometry away from the borehole out into the formation.

Continuous improvements and changes in hardware, measurement physics, processing power,

Figure 2. LWD azimuthal imaging. The bulk density image from an adnVISION tool (Track 3) provides information about the borehole circumference in a horizontal well. Density data are also presented as curves (Tracks 3, 4 and 5) and are displayed according to the quadrant from which the data were acquired (Tracks 3 and 5). Bulk density and neutron porosity data may be

affected by hole conditions as can be observed around X50 ft, where the caliper indicates a washout (Track 1, blue shading). Because this well is horizontal, the tool’s azimuthal outputs are referenced to up, down, left and right. In a vertical well, the references are north, south, east and west.


in.8 10

0 150

X10

X20

X30

X40

X50

0

0.2 2,000 1.9 2.9g/cm3

g/cm3

1.9 2.9g/cm3

0 capture units 50

1.9 2.9g/cm3

1.9 2.9g/cm3

1.9 2.9g/cm3

ohm.m

0.02 200ohm.m

500rpm

40 –15%gAPI

Gamma Ray

Collar Rotation

Depth, ft

Deviation

16-in. Phase Shift Bulk Density UpperNeutron Porosity

(Thermal) Up Density

Bulk Density

Neutron Density

Image-Derived Density

Sigma

22-in. Phase Shift

28-in. Phase Shift

34-in. Phase Shift

40-in. Phase Shift

Ultrasonic Caliper

QualityFlagsClay

Sandstone

Water

Pyrite

–0.8 0.2g/cm3

Density Correction

1.7 2.7

Density Image

0 90degreein.8 10

Density Caliper

Washout

Mudcake16-in. Attenuation

22-in. Attenuation

28-in. Attenuation

34-in. Attenuation

40-in. Attenuation

Resistivity

1.9 2.9g/cm3

Right Density

1.9 2.9g/cm3

Left Density

1.9 2.9g/cm3

Bottom Density

1.9 2.9g/cm3

Average Density

Quadrant BulkDensity Data

September 2015 7

data transmission and tool technologies eventually led to imaging tools that provided much more than formation dip. Imaging tools acquire high-resolu-tion conductivity (or the reciprocal resistivity) data from a very shallow depth of investigation and provide an image of a portion of the wellbore cir-cumference. These images are representative of features intersecting the borehole.

By interpreting information generated by both computer processing and manual correlations, geologists can identify geologic features. Before imaging tools were available, geologists used dip-meter data primarily for guidance in planning the next well location. They were able to determine the direction needed to move up or down struc-ture, the location of faults and the presence of structural anomalies. Modern image logs provide an opportunity to better understand reservoir geo-logic characteristics and visualize the well within the context of the reservoir.

Advancing beyond dipmeter tools, imaging tools now allow interpreters to identify structural features such as faults, folds, angular unconformities and bed-ding geometry and infer paleotransport direction of sands and conditions that existed during deposition. Geologists can also use image logs to detect fractures and define their properties—a crucial element in characterizing tight reservoirs. They then incorpo-rate fracture properties in completion designs and use the information for field optimization.

The ability to detect small features such as fractures is not easily performed in wells drilled with oil-base mud (OBM) systems.3 The mud and mudcake add a layer of electrical insulation in the wellbore that usually renders traditional con-ductivity-based imaging tools ineffective. Imaging

tools designed for use in OBM systems have not delivered the level of resolution that tools designed for water-base mud (WBM) systems are able to provide—determining quantitative properties of fractures has been especially difficult. The Quanta Geo photorealistic reservoir geology service, which acquires images that are representative of the borehole wall in the challenging environment of OBM systems, was recently introduced to address this situation.

This article reviews the evolution of imaging ser-vices—from dipmeter tools to the latest generation imaging devices. Case studies demonstrate the use of image logs in OBM wells for stratigraphic analysis of wells drilled in deepwater Gulf of Mexico environ-ments and for analyzing fractures in wells drilled in unconventional reservoirs.

Painting a Wellbore PictureBefore computers became readily available, relatively high-resolution dipmeter data were acquired from downhole, and the information was presented on photographic film. Analysts read and interpreted these data manually—a tedious process. The introduction of computer-ized logging units and digital data processing enabled higher sample-rate data to be acquired than was previously possible. Modern logging tools acquire more information than most humans can assimilate, integrate and process. Computer processing has become indispensable for delivering information in a usable format.

The ability of logging-while-drilling (LWD) tools to make azimuthal measurements from around the circumference of the borehole has also changed the way many analysts visualize

downhole data. In a similar manner to that used by conventional wireline logging devices, LWD tools acquire data linearly via tool movement along the well; however, azimuthal tools also acquire data from the full circumference of the wellbore as the tool rotates. Azimuthal data are then presented as an image of the borehole, “painting a picture” of the inside of the wellbore. Because the tool orientation is measured simul-taneously, the images can be aligned with the geometry of the wellbore. However, the resolu-tion of these data is insufficient for detecting small details (Figure 2).

Many LWD tools can provide azimuthal data presented in the form of wellbore images; such tools include azimuthal gamma ray devices, the MicroScope HD high-definition imaging-while-drilling tool and the adnVISION azimuthal den-sity neutron service.4 Image interpretation of data from azimuthal tools has become crucial for adjusting wellbore trajectory—up, down, left or right—in real time in many horizontal drilling operations (Figure 3).

Figure 3. Well placement using image data. Azimuthal log data in the shapes of smiles and frowns help well placement engineers determine bit corrections while drilling. When a wellbore crosses a bedding plane, the azimuthal logging tool response indicates whether the wellbore is exiting an ascending or descending geologic layer. When the wellbore cuts an ascending layer (left ), the first contact with the formation is at the bottom

of the hole; when the bit exits the layer, the last contact will be at the top of the hole. The image data appear as a frown. Conversely, measurements from a wellbore that exits a descending bedding plane (right ) appear as a smile. Based on these interpretations, drilling engineers may guide the bit up or down to ensure that the wellbore remains in or reconnects with a target zone.


Bed Dipping Toward Kickoff Point

Top

Top

Botto

m

Top

Top

Botto

m

Bed Dipping Away from Kickoff Point

2. For more on legacy imaging tool and image interpretation: Wong SA, Startzman RA and Kuo T-B: “A New Approach to the Interpretation of Wellbore Images,” paper SPE 19579, presented at the 64th SPE Annual Technical Conference and Exhibition, San Antonio, Texas, USA, October 8–11, 1989.

3. For more on early logging services used for imaging in OBM systems: Cheung P, Hayman A, Laronga R, Cook G, Flournoy G, Goetz P, Marshall M, Hansen S, Lamb M, Li B, Larsen M, Orgren M and Redden J: “A Clear Picture in Oil-Base Muds,” Oilfield Review 13, no. 4 (Winter 2001/2002): 2–27.

4. For more on LWD azimuthal imaging tools and using azimuthal data for structural steering: Amer A, Chinellato F, Collins S, Denichou J-M, Dubourg I, Griffiths R, Koepsell R, Lyngra S, Marza P, Murray D and Roberts I: “Structural Steering—A Path to Productivity,” Oilfield Review 25, no. 1 (Spring 2013): 14–31.

8 Oilfield Review

Wireline logging tools were the first to acquire data that could be displayed as images from the circumference of a wellbore although few wireline tools have the azimuthal acquisition capabilities commonly found in LWD tools. An exception is the UBI ultrasonic borehole imager tool, which uses a rotating assembly to map the full circumference of the borehole from ultrasonic reflections of the borehole wall. Because the UBI tool depends on the quality of the reflections from the borehole, it works best in hard formations.

Older generation devices that have multiple pads, such as the HDT high-resolution dipmeter tool, acquired data from four regions inside the borehole. By correlating the data acquired from around the wellbore, bedding or feature dip mag-

nitude and direction could be determined manu-ally or by computer. Successive generations of dipmeter tools increased the number of sensors and pads, ultimately giving way to tool designs that had sufficient sensor density to provide imaging capabilities (Figure 4).

The FMS Formation MicroScanner tool was one of the first successful borehole imaging ser-vices. Equipped with four pads, the original tool had 27 sensors on two of the pads, which acquired data every 2.5 mm [0.1 in.].5 The other two pads had only two button sensors each. This design permitted basic imaging of the borehole; how-ever, covering the inside of the wellbore required multiple passes and manual depth matching. An updated FMS tool had two rows of eight sensor

buttons on each of its four pads, which covered more of the borehole in a single pass.

The FMI-HD high-definition formation micro-imager is the latest generation Schlumberger tool for assessing structure and stratigraphy of rocks in WBM systems and some OBM systems.6 This tool is equipped with 192 pad-mounted sen-sors, or button electrodes, and samples every 2.5 mm (Figure 5). The button electrodes are arranged in parallel rows across the face of each pad, and each pad has a hinged flap extension that has its own parallel rows of sensors. When the pads, which are mounted on caliper arms, are extended, the flaps open and increase the cir-cumferential coverage of the borehole. In an 8-in. borehole, the tool covers 80% of the circumfer-ence. The design results in a 5-mm [0.2-in.] reso-lution; any feature 5 mm or larger can be directly measured although much smaller features, including fractures, can be imaged if there is suf-ficient electrical contrast with the background.

For interpreters to visualize these data, the measurements are converted from conductivity values into images. These images are created from the electrical measurements, which are converted to pixels. Before image logs existed, however, dipmeter interpretation relied on tad-poles computed from wellbore data.

Answers in the TadpolesLog analysts still use tadpoles from dipmeter logs to describe downhole structural geometry and stratigraphy. Tadpoles represent information computed from raw dipmeter data; they provide two main quantities: dip direction and dip magni-tude (Figure 6). Each tadpole consists of a head and a tail. The head of the tadpole is plotted on a graph scaled from 0° to 90°, and the position of the head on the scale indicates the magnitude of the dip. The tail points in the downward direc-tion, or dip, of the formation or feature, and the display is based on a compass dial. True north is at the top followed clockwise by east, south, west and back to north through a full 360° cycle. By reading the dip magnitude from the location of the head and the direction from the tail, inter-preters infer formation or feature geometry.

Tadpoles are computed from data acquired as the tool traverses the borehole during logging; if bedding planes with contrasting resistivities are

Figure 4. Imaging evolution. The original dipmeter tool from 1945 had three pads; each pad had a single sensor button (top left ). As successive generations of tools were developed, engineers added pads and increased the number of sensor buttons on each pad. The FMS tool (bottom left ), introduced in the mid-1980s, was one of the first wireline tools to provide image logs. Developers found that multiple parallel rows of buttons in the original design were not necessary, and the original FMS tool was modified in 1988 to have only two rows of sensors on each of its four pads for a total of 64 sensors (bottom middle). The FMI fullbore formation microimager (bottom right ), introduced in 1991, has four pads that have four flaps and a total of 192 sensors. The wellbore schematic below each tool shows coverage by the pads in an 8-in. borehole.


1945

3 pads3 sensors

Pad position in the borehole

Pad position in the borehole

2 pads29 sensors

2 pads2 sensors

4 pads64 sensors

8 pads192 sensors

4 pads4 sensors

4 pads8 sensors

1986 1988 1991

1968 1982

5. An FMS tool with four pads for imaging was a forerunner of the FMI-HD tool. For more: Bourke L, Delfiner P, Trouiller J-C, Fett T, Grace M, Luthi S, Serra O and Standen E: “Using Formation MicroScanner Images,” The Technical Review 37, no. 1 (January 1989): 16–40.

6. For more on the FMI service: Adams J, Bourke L and Buck S: “Integrating Formation MicroScanner Images and Cores with Case Studies,” Oilfield Review 2, no. 1 (January 1990): 52–65.

September 2015 9

encountered, the tool will detect those events along the borehole (Figure 7). Should all pads detect an event at the same depth, the relative dip is 0°. If the sensors encounter a dipping bed or feature crossing the wellbore, the sensors detect it at various points inside the borehole. The magnitude of dip is determined by comput-ing the displacement of these events. A structural dip of just 1° will cause approximately 5 mm of displacement across an 8-in. borehole, which is within the resolution range of the tool.

The position of one pad is referenced with respect to true north, which determines the ori-entation of the tool. This also defines the posi-tion of the other pads and sensors. The orientation of the pads in the borehole along with the displacement between conductive or

Figure 5. Latest generation imaging tool for WBM systems. The FMI-HD tool, which has four pads and four flaps, has a total of 192 button sensors. The caliper arms extend, and the flaps rotate to provide an acquisition surface that is twice as wide as that of tools that have only four pads. The close spacing and fixed distances between sensor buttons result in high-resolution data; fixed spacing provides a systematic method for speed correction. The tool generates a continuous stream of high-resolution data (inset ) from its 192 buttons from which images are generated.


X,376

X,377

X,378

X,379

Figure 6. Tadpole plots from dipmeter data. A single tadpole computed from dipmeter tool data indicates a variety of reservoir geometric properties. The location of the head of the tadpole on the scale indicates the magnitude of formation dip. The tail of the tadpole points in the downward direction. This example tadpole indicates formation dip of 27° down to the west. Tadpoles have evolved over the years to include color coding, quality indicators and modifications that represent fractures or other features.


N

SEW

Tadpole

0° 10° 20° 30° 40° 50° 60° 70° 80° 90°

Figure 7. Detecting bed boundaries and formation dip. As a dipmeter tool is pulled through the wellbore (left ), sensors on the pads intersect the bedding plane or feature at various points along the borehole wall. By correlating the points and determining the tool’s cardinal coordinates (middle left ), the bedding plane’s geometry can be computed. When the data from along the inner surface of the borehole are unwrapped (middle) and presented in 2D (middle right ), a dipping bedding plane will form a sinusoid, which gives an

indication of the direction and magnitude of the formation dip. Analysts use images from the inner surface of the borehole wall to visualize formation geometry and identify features such as fractures and unconformities. The down dip direction in the image appears to be to the west, although most image data are presented using apparent dip. Based on the tadpoles computed from these data (right ), which include rotation for wellbore and tool drift, the true dip is down to the south.


Displacementbetweenpad responses

Low-angledip

High-angledip

BoreholePad response

4 dipmeterpads

Resistive thin bed

12

2 3

0°

0° 90°

90°

NN

N

N

NS S S

W

W WE E E

180° 270°

Tadpoles

Low-anglesouth dip

High-anglesouth dip

360°

41

3 4

10 Oilfield Review

resistive events are used to define the depth, direction and dip magnitude of a bedding plane or a feature. The direction and magnitude are then presented as an apparent dip, which is

related to the tool orientation. This apparent dip can also be corrected for the angle and incli-nation of the wellbore, also referred to as bore-hole drift (Figure 8). Sensors measure the

position of the tool with respect to true north and determine tool deviation from vertical. When the contribution of the tool and well posi-tion are rotating out, the true formation dip from horizontal can be displayed as a tadpole.

A single tadpole is not sufficient for determin-ing formation geometry. In the past, dipmeter interpretation, which is both an art and a sci-ence, was a process whereby analysts identified trends or patterns in the tadpoles from whence downhole structures could then be described. The three primary patterns are often referred to by the colors red, blue and green (Figure 9). A red pattern is increasing dip magnitude with depth, a blue pattern is decreasing dip magni-tude with depth and a green pattern is uniform, or unchanging, dip with depth. The azimuth of the dips should be constant or changing slowly across the section or feature. Patterns can result from a variety of features, but interpreters used the patterns primarily as guides for selecting the direction of offset well locations or defining depo-sitional direction. Dipmeter interpretations are often used to explain why a well encountered unexpected or missing formations sections, for instance, as a result of crossing a fault.

Geologists interpreting dipmeter data and images today have gone far beyond recognizing red, blue and green patterns. From images, they are able to interpret downhole structure and stratigraphy.

Evolution: Tadpoles to ImagesTraditional tadpole pattern recognition involved taking a 2D concept and constructing a 3D vision of the reservoir. This macro view of the downhole environment was used to describe formation geometry, but the view inside the wellbore can show the interpreter much more about rock and formation characteristics. This task is accom-plished using borehole image data.

The conversion of tool measurements to images is analogous to the processes used in mod-ern digital photography. One type of digital camera in use today is the charge-coupled device (CCD).7 The heart of the camera is a densely packed array of sensors. Incoming photons strike a portion of the sensor surface and are converted to electrons (Figure 10). An analog-to-digital converter accu-mulates the charge information from these elec-trons and transmits it for further processing and eventual storage. The more densely packed the sensors are on the array, the greater the number of pixels and the higher the resolution.

7. Charge-coupled device sensors were invented by Willard Boyle and George Smith at AT&T Bell Laboratories, New Jersey, USA, in 1969.

Figure 8. Correction for borehole drift and formation geometry. Apparent dip (AD) is the computed angle of the formation bedding plane or feature as it crosses the borehole. True dip (TD) is AD corrected for the geometry of the well and tool drift; these rotated data reflect deviation from horizontal. Some stratigraphic features such as the paleodepositional direction can be more easily seen in data that have the structural dip deleted (not shown) because the resulting data may be representative of conditions that existed at the time of deposition.


BoreholeAD angle

TD angle

AD: Apparent dip noncompensated

TD: True dip compensated for borehole drift

Figure 9. Pattern recognition dipmeter interpretation. Green patterns (Track 2) represent the general structural dip of a formation and are usually more consistent in low-energy depositional environments such as shales, as indicated in the correlation curve (Track 1) than in high-energy deposition typical of sandstones. Abrupt changes in structural dip can occur when a well crosses an unconformity or a fault. Red patterns are increasing dip with depth and may be indicative of approaching faults, drape over structures and channels. Blue patterns are decreasing dip with depth and may indicate bedding, paleodepositional direction, deformation below faults and unconformities. The borehole geometry can also be represented by tadpoles (Track 3). This well is drifting about 2° from vertical toward the ENE.


Correlation Curve True dip angle and direction, degree

10 10020 30 40 50 60 70 80 90

Boreholedrift,

degree

Structural dip

Structural dip

Structural dip

Crossbedding andfracturing

Current patterns,unconformities andforeset beds

Slope patterns: faults, bars, reefs, channels and downdip thickening

September 2015 11

Unlike in film photography, in digital photog-raphy, there is no “physical” image—photo-graphs are reconstructed from digital data that represent light falling onto the sensors. Similar to the process in which digital cameras convert signal data to pixels and collect pixels into images, the high-resolution conductivity data from the sensor buttons of imaging tools are con-verted to pixels and then displayed together as an image (Figure 11). The image is not an actual picture but a representation of the changes in conductivity along the inside of the wellbore.

Imaging ProcessData acquired during logging have little resem-blance to the final image product. The buttons produce a continuous stream of parallel conduc-tivity measurements, which are transmitted uphole and recorded. The 192 buttons of the FMI-HD tool—each of which has a 5 mm diame-ter—acquire a measurement with each 2.5 mm of tool movement. The tool’s horizontal and verti-cal sensor spacing, along with high sampling fre-quency, enable the tool to measure features as small as 5 mm, but it can resolve much smaller

Figure 10. Creating digital images. A charge-coupled device (CCD) camera consists of an array of sensors. Light (photons) strikes the surface of the CCD (left ), and the sensors detect the photons and convert them to electrons. Electrons are measured and converted to a voltage. The analog voltage measurement is sent to a processor, where the measurement is converted to digital data for storage. A CCD sensor does not create an image as film cameras do; the image is recreated from stored data at each pixel location. This process is similar to the process used for conversion of conductivity (or resistivity) data to pixels for creating image logs.


Gain

Photon-to-electron

conversion

Photonsstrikearray

Charge-Coupled Device

Electron-to-voltageconversion

Analog-to-digital converter

Figure 11. Data from the FMI-HD tool converted to an image log. The 192 buttons located on the FMI-HD tool’s four pads and four flaps generate a stream of conductivity data (left ). These data are processed, the values are assigned a scaled color, and an image is produced (right ). The geologist analyzing the images can modify the color scale and range to enhance features. The cardinal location of Pad 1 can be identified from the green curve at the far right.


XX5.75

XX6.00

XX6.25

XX7.25

XX6.50

XX7.50

XX6.75

XX7.75

XX7.00

Depth,ft

Raw Button Conductivity Data

Image Data

Borehole Position, Pad 1

North0 360degree

West South East North

12 Oilfield Review

events. Measurements such as tool position, the Earth’s magnetic field properties, caliper data and low-resolution sensor information are sam-pled every 3.8 cm [1.5 in.].

Raw data must be reviewed for quality, and corrections are applied as necessary during pro-cessing. A crucial step in the QC process is speed correction, in which the objective is to position each measurement at the correct depth in the borehole relative to all the other measurements. Speed correction attempts to overcome nonuni-form tool movement and ensure data integrity. Even slight changes in tool movement during acquisition of high-resolution data can affect image quality.

Speed correction is often a two-step process. Accelerometers in imaging tools detect incremental tool movements; offsets for these small variations are applied as a first-level correction. Because the sensor buttons are arranged in parallel rows with a fixed spacing, changes in resistivity at boundary crossings can be compared. If the same event is found to be displaced between rows, the data can be shifted to adjust for the offset. Software-based tool movement detection methods help to further refine the initial speed correction. Combining the meth-ods produces a robust correction; however, when extreme tool movement irregularities occur, espe-cially those of the stick-and-release variety, data may not be recoverable.

The next step in processing is to harmonize the button responses (Figure 12). Raw button responses are not calibrated, but button-to-but-ton normalization can be used to ensure a reason-able image is generated. In this step, gains and offsets are computed for each button over a slid-ing window—typically 5 to 30 m [15 to 100 ft]—to give all of the buttons a comparable response. These normalized responses are then assigned a color or gray scale value and presented as an image of the borehole from 0° to 360°, with the left edge at 0° and the right edge at 360° representing true north. The center of the image at 180° repre-sents south. For horizontal wells, the top of the well is on the left and right (0° and 360°) edges and the bottom of the well is in the center of the image (at 180°).

Data are usually presented in both static and dynamically enhanced modes—the latter can increase the visible range of usable images (Figure 13). The static image helps the inter-preter maintain the image context—to recognize whether one is interpreting a conductive or resis-tive bed—and to recognize major bed boundaries by their association with significant resistivity changes. The dynamic image allows the inter-

Figure 12. Processing raw data. Streams of raw data from the button sensors (top) are depth shifted, offset and equalized (bottom). This processing produces more consistent data and better image quality than would be available from the raw data.


Conductivity Curves

Histograms

Freq

uenc

y

Original data Data after offsetequalization

Data after offset and gain equalization

Figure 13. From borehole conductivity to borehole images. After offsets and normalization are applied, the processed data are assigned a color or gray scale based on the measured conductivity (or resistivity). In this scheme, conductive features are represented by dark colors and resistive features are represented by light colors. Because the resistivity range of the tool is large, the data are usually presented in static mode and a dynamically adjusted mode. For static imaging, the peak value (green shading) corresponds to a color or gray shade. For dynamic scaling, the computer samples the data outside the peak value (blue shading) and uses the information to create an enhanced image. The color or gray scales may also be reversed to highlight resistive or conductive features. The various modes allow analysts to see details and features that might otherwise be masked.


Histogram of image data

Sampling windowfor dynamic imaging

Sample window for static imaging

Minimum0

1

Maximum

Minimum Maximum

Colo

r or g

ray

scal

e ou

tput

September 2015 13

preter to see the maximum detail of formation texture and is useful for identifying sedimentary structures, textures associated with complex porosity systems and both natural and drilling-induced fractures. Static equalization alone may be used if the contrast range is small.

Images may be presented in a variety of colors or in gray scale. A typical color scheme, referred to as heated, uses a yellow-to-brown gradient that is scaled from light to dark extremes. The actual color is arbitrary but may help to highlight fea-tures. Comparing color image logs to physical cores can be disconcerting because the actual rock will not have as much physical contrast as the visual contrast typical of image logs. For that reason, some analysts prefer gray scale images for comparing images to core.

Computer interpretation software is often used first to analyze the data and generate tad-poles. The processed image data are then dis-played on a workstation, where image analysts, usually geologists, observe and identify features such as structural dip, faults, fractures, crossbed-ding planes and unconformities (Figure 14).

The process of image interpretation has been described as observation, interpretation and implication. The analyst’s first task is to review the data in search of recognizable or observable features. After features have been identified, the analyst interprets them by making manual picks. Because features striking the wellbore at an angle present themselves as sinusoids, the ana-lyst can define points along a feature and allow

the image workstation software to fit a sinusoid and compute formation dip or feature geometry. The software can also correct the data for bore-hole drift. Stratigraphic features may be more meaningful if the borehole dip is further cor-rected by subtracting the structural dip compo-nent, restoring the geometry to that of the apparent depositional orientation.

The final task of the image analyst is to assess the interpretation for implications. Analysis of the structural geometry may be used to help plan the next well, determine the lateral landing point or establish field development alterna-tives. Stratigraphic interpretation may include identifying depositional implications and apply-ing that information to understanding the nature of the rocks. Identification of both natural and induced fractures can be used in determining fracture properties, confirming in situ stress rela-tionships and designing effective stimulation and completion programs.

In addition to wireline imaging tools, other technologies are available for imaging boreholes. These include resistivity-based LWD logging tools and acoustic imaging tools run on wireline.

LWD ImagingAlthough many LWD tools provide image logs, the combination of accurate tool movement, high-resolution accelerometer data and high data transmission rates have given wireline imaging tools an advantage over LWD imaging services.

However, the inability of LWD images to resolve small features has been addressed by a novel processing approach developed by Schlumberger researchers.

The MicroScope HD service has 1-cm [0.4-in.] buttons and can sample every 5 mm. Although this tool design can provide high-resolution mea-surements, design alone is not sufficient to resolve small features because tool movement cannot be controlled to the level needed in the drilling environment. Complicating the depth control issue is the fact that LWD data are time based rather than depth based, and pipe move-ment at the drilling floor is indexed to tool move-ment downhole. The large separation between the point of acquisition and the depth reference affects resolution quality.

To overcome tool movement issues, high-reso-lution data are acquired with the MicroScope HD tool along with magnetometer-based tool orienta-tion data as the tool rotates.8 Since each tool has a fixed sensor spacing, data from the borehole circumference can be viewed as strips that have a constant and known thickness. The time-based measurements are converted to a depth-indexed image using high-resolution axial and azimuthal

8. For more on the LWD imaging technique: Allouche M, Chow S, Dubourg I, Ortenzi L and van Os R: “High-Resolution Images and Formation Evaluation in Slim Holes from a New Logging-While-Drilling Azimuthal Laterolog Device,” paper SPE 131513, presented at the SPE EUROPEC/EAGE Annual Conference and Exhibition, Barcelona, Spain, June 14–17, 2010.

Figure 14. A case for images. Structural dip can be identified in the raw data (left, Track 1). These data have been computer processed to generate tadpoles (Track 2). The red tadpoles are generated from data covering 1-ft intervals; the blue tadpoles are output from 2 ft of data. Computed results that have lower confidence are shown as open circle tadpoles; the lines represent the computed sinusoids. In the image display (center), the computer-generated tadpole is not related to a specific feature but indicates trends derived from the raw data. A fracture crossing the wellbore is more easily visualized from the image log than on the raw data,

and it has been marked by the analyst (right, purple) using image workstation software. The analyst traces the fracture and allows the software to compute its true dip magnitude and direction (purple tadpole). The fracture crosses the wellbore at an angle magnitude of around 87°; the true dip direction is down to the NE and its strike direction is NW–SE. The analyst can also trace features such as bedding planes (green lines and tadpoles) and faults and compare those with computer-generated results (blue and red). The formation dip is about 10°, dipping to the NNE, as indicated by both manual and computer-generated results.


X,388

X,389

Image Data TadpolesDepth,ft Image Data TadpolesDepth,

ft

X,388

X,389

X,387

Depth,ft Raw Data Tadpoles

X,388

X,389

14 Oilfield Review

sensor positions. As the tool is advanced up or down the well, overlapping strips of data are acquired. These strips of data have features that are a fixed distance apart, which allows the image strips to be merged, correlated to axial tool movement and continually adjusted for depth (Figure 15).

Using image data from the MicroScope HD tool, analysts have been able to detect small fea-tures. Datasets from this service are quite large but are transmitted continuously during the drill-ing operation or retrieved from the tool when it

returns to the surface. As with many other imaging tools, this tool requires a conductive mud system.

Oil-Base Mud ImagingWell operators use OBM systems because they facilitate improved drilling performance.9 Since the 1990s, most deepwater wells have been drilled with OBM systems that use nonconductive fluids; such systems preclude the use of logging tools that function only in conductive fluids.10

Conductivity-based imaging tools rely on detecting small changes in conductivity along the

surface of a borehole wall. However, OBM and mudcake behave similar to electrical insulators, obstructing current flow. Therefore, acquiring wellbore images in OBM systems may not be fea-sible using tools designed for acquiring data in WBM; modifications to the FMI-HD tool, however, have enabled acquisition of images in some OBM environments.11

Early attempts to acquire dipmeter data in OBM wells were often met with frustration. Blades and scratchers were first used to remove mud and mudcake from the borehole wall to pro-

Figure 15. High-resolution LWD imaging. Tool movement for LWD tools is referenced to changes in drillpipe depth measured at the surface. Downhole LWD data are time based. Depth is derived by associating the time of acquisition to the depth measured at the surface. Depth accuracy available from this system of measurement is insufficient for resolving fine details and features. Schlumberger engineers developed a method that ties the fixed spacing of the tool sensors to the data and correlates depth to data.

Overlapping data, viewed as strips (top left ), are aligned and adjusted to match tool movement and then merged. Examples of noncorrelated and correlated data (top right ) demonstrate the enhanced image resolution. MicroScope HD image data can now be used to define structural features and fractures (bottom). A log analyst has marked the almost vertical resistive fractures (cyan) and faults (magenta and blue) crossing this horizontal wellbore along with the bed boundaries (green) cut by the wellbore.


498

2,951

2,950

2,949

2,948

2,947

2,946

2,945

2,944

499

Time, 100 s

Dept

h, ft

500 501 502 503 504 505

NoncorrelatedImages

CorrelatedImages

Dept

h

Mergedstrip

+ =

Correlation depthSurface-measureddepth

Dyna

mic

Imag

e

Top

Top

Bottom

Resistive Fracture

Fault

Bed Boundary

Microfault

Imagestrip

Imagestrip

9. For more on drilling with OBM systems: Bloys B, Davis N, Smolen B, Bailey L, Houwen O, Reid P, Sherwood J, Fraser L and Hodder M: “Designing and Managing Drilling Fluid,” Oilfield Review 6, no. 2 (April 1994): 33–43.

10. Chen Y-H, Omeragic D, Habashy T, Bloemenkamp R, Zhang T, Cheung P and Laronga R: “Inversion-Based Workflow for Quantitative Interpretation of the New-Generation Oil-Based Mud Resistivity Imager,” Transactions of the SPWLA 55th Annual Logging Symposium, Abu Dhabi, UAE (May 18–22, 2014): paper LL.

11. For more on the FMI-HD service in OBM: Laronga R, Lozada GT, Perez FM, Cheung P, Hansen SM, Rosas AM: “A High-Definition Approach To Formation Imaging In Wells Drilled With Nonconductive Muds,” Transactions of the SPWLA 52nd Annual Logging Symposium, Colorado Springs, Colorado, USA, May 14–18, 2011, paper FFF.

12. Cheung P, Pittman D, Hayman A, Laronga R, Vessereau P, Ounadjela A, Desport O, Hansen S, Kear R, Lamb M, Borbas T and Wendt B: “Field Test Results of a New Oil-Base Mud Formation Imager Tool,” Transactions of the SPWLA 42nd Annual Logging Symposium, Houston (June 17–20, 2001): paper XX.

13. Bourke LT and Prosser DJ: “An Independent Comparison of Borehole Imaging Tools and Their Geological Interpretability,” Transactions of the SPWLA 51st Annual Logging Symposium, Perth, Western Australia, Australia (June 19–23, 2010): paper GGG.

14. Bloemenkamp R, Zhang T, Comparon L, Laronga R, Yang S, Marpaung S, Guinois EM, Valley G, Vessereau P, Shalaby E, Li B, Kumar A, Kear R and Yang Y: “Design and Field Testing of a New High-Definition Microresistivity Imaging Tool Engineered for Oil-Based Mud,” Transactions of the SPWLA 55th Annual Logging Symposium, Abu Dhabi, UAE (May 18–22, 2014): paper KK.

September 2015 15

vide an electrical path to ground, but these meth-ods did not prove feasible for imaging. The OBMI oil-base microimager tool was the first successful tool for imaging in OBM systems.12 This tool has four pads located 90° from each other; each pad has five pairs of sensors spaced 1 cm apart (Figure 16). This spacing provides approximately 1-cm vertical and horizontal resolution, and pad coverage is approximately 32% of an 8-in. well-bore. The OBMI tool delivers images of large fea-tures but is unable to detect fine details.

The OBMI2 integrated dual oil-base microim-agers features two stacked OBMI sondes oriented 45° apart. This design doubles the circumferen-tial borehole coverage. In general, the OBMI, OBMI2 and other OBM imaging devices do not image small features as well as their WBM coun-terparts do.

Not only can spatial resolution be a problem, the measurement technology used in most OBM imaging tools can introduce artifacts such as shadow beds on the shoulders of high-contrast environments, or the images may be affected by the orientation of the bedding planes. Mud-filled cracks and drilling-induced features often distort the image and mask formation geology. One inde-pendent study found that OBM imaging tools resolved an order of magnitude fewer sedimen-tary features compared with those resolved by tools run in WBM environments.13

Realizing the need for a high-resolution imag-ing solution, Schlumberger researchers began

developing a tool that could produce images in OBM systems comparable to those available from wells drilled with WBM. In 2014, the Quanta Geo service was introduced (Figure 17).14 In design-ing the new tool, engineers used button elec-trodes that function in a different manner compared to those used in WBM imagers.

To determine the formation’s electrical con-ductivity, imaging tools in WBM inject current directly into the formation from the button elec-trodes. Because OBM and mudcake act similar to electrical insulators, current is impeded from going into the formation and from returning to the tool. To overcome this dilemma, the button electrodes of the Quanta Geo service establish capacitive contact with the formation by sending current at much higher frequencies—in the MHz range—than the current used in WBM imagers. Imagers designed for WBM operate with currents in the kHz range.

When OBM is used for drilling, rather than acting as a true insulator, the fluid and the mud-cake actually behave like a lossy dielectric. A dielectric is a material that acts as a poor con-ductor of electric current and impedes current flow. Although a dielectric has properties similar to those of an insulator, it differs in that the impedance—defined as the resistance to flow of an AC composed of resistive and reactive compo-nents—of a dielectric decreases inversely with increased frequency.

Figure 16. Sonde and pad of the OBMI service. The OBMI tool (left ) has four pads. Each pad (right ) has two rows of sensor buttons. Current is emitted from the sensors and returns to the electrodes above and below the sensors. Because of the limited borehole coverage of this design, the OBMI2 tool—consisting of two tools stacked and offset by 45°—was developed (not shown).


Figure 17. The Quanta Geo photorealistic reservoir geology service. This tool has two sonde sections; each section has four pads oriented at 90°, and the two tool sections are offset by 45°. Each spring-mounted pad is fully independent and can swivel ±15° around the tool axis. The mechanical design allows the tool to be logged in an up or down direction. Azimuthal coverage in an 8-in. borehole is 98%.


Pad positionin the borehole

16 Oilfield Review

Conversely, formations act like resistors, and the resistance remains fairly constant over a large resistivity and frequency range—up to a point. There is an upper limit to the frequency of the current at which the contribution from the permittivity of the formation becomes non-negligible. Permittivity is a measure of how an electric field affects, and is affected by, a dielec-tric medium. Above the critical frequency, per-mittivity of the formation combines with the

dielectric properties of the mud system. Below the frequency upper limit, the permittivity of the formation is negligible and frequency-related changes in impedance measured by the tool arise from the mud and mudcake proper-ties. Design engineers correct for the contribu-tion to the impedance measurement from the mud and mudcake by using the phase difference between the signals passing through the forma-tion and the signals passing through the mud

and mudcake at two frequencies. During pro-cessing, the analyst can determine which fre-quency provides the optimal response.

The Quanta Geo service outputs an imped-ance measurement seen by the electrodes at the two separate frequencies rather than the conduc-tivity normally measured by WBM-based tools. A consequence of using this technique is that the measured impedance is not directly proportional to the formation resistivity. Computing an invaded zone resistivity (Rxo) from measured data, which is usually available from WBM tool measurements, is not an option.

The Quanta Geo sonde has four pads oriented at 90° and a second set of four pads located below the first set offset by 45°. The fully independent pads are spring mounted and can swivel ±15° around the tool axis as well as longitudinally; this mechanical design allows the tool to be logged in an up or down direction. Azimuthal coverage in an 8-in. borehole is 98%. The tool operates across a resistivity range of 0.2 to 20,000 ohm.m.

Each pad has a horizontal row of button electrodes bounded above and below by guard rings and return electrodes (Figure 18). High-frequency current emitted from each of the 192 buttons capacitively connects to the forma-tion and returns back to the tool. Using two return electrodes provides a symmetrical tool response. The current flowing from each button is mea-sured, and the impedance is computed. This impedance contains both the amplitude ratio

Figure 18. Quanta Geo Pad design. Each of the eight identical pads for the Quanta Geo service has a row of button electrodes surrounded by a guard electrode (left ). Return electrodes are above and below the button electrodes. Two high-frequency alternating currents are forced to flow through the mud and mudcake into the formation; the currents (right , white arrows) return to the upper and lower electrodes, providing a symmetrical response. Current is prevented from returning directly to the tool by the guard electrode. The current flowing from each button is measured and the impedance is computed. This impedance contains both the amplitude ratio between the voltage and current and the phase shift for the two AC frequencies. Tool design provides a vertical resolution of 6 mm [0.24 in.] and a horizontal resolution of 3 mm [0.12 in.].


Measurement currentflow path

Nonconductive mud

Guard electrode

Button electrodes

Return electrode

Figure 19. High-quality images. The Quanta Geo service provides high-quality images in wells drilled with OBM. Visible bedding planes in a whole core (left ) can be easily seen in the dynamic image (Track 1) but not so clearly seen in the static image (Track 3). A fault crossed by the wellbore (right ) is visible in both the dynamic image (Track 1) and the whole core taken across this section.


Quanta GeoDynamic Image

Quanta GeoStatic Image

Quanta GeoDynamic Image

Quanta GeoStatic Image

Image Orientation, degree

Whole Core Whole Core


0 150gAPI

High-Resolution

Gamma Ray

0 150gAPI

High-Resolution

Gamma Ray

East ENorth NSouth SSouth SWest W

East ENorth NSouth SSouth SWest WImage Orientation, degree Image Orientation, degreeResistiveConductive ResistiveConductive

ResistiveConductive ResistiveConductive

September 2015 17

between the voltage and current and a phase shift for the two frequencies. The measured impedance is a mix of the formation and the mud responses. Electrode spacing and tool design pro-vide a vertical resolution of 6 mm [0.24 in.] and a horizontal resolution of 3 mm [0.12 in.] Two sam-pling interval rates are available—5 mm [0.2 in.] and 2.5 mm [0.1 in.]

Processed image data from the Quanta Geo service produces photorealistic images never before possible in OBM systems (Figure 19). Log analysts use these high-resolution data to define structural features such as faults and unconfor-mities. Stratigraphic features such as crossbed-ding and foreset beds can be identified; depositional characteristics such as bioturba-tion, clasts and scours can be recognized in the

images. The high-quality images allow interpret-ers to identify natural and drilling-induced frac-tures and quantitatively determine their physical properties. The image quality for data acquired with the Quanta Geo tool is comparable to that of the images available from the FMI-HD service (Figure 20).

Deepwater ApplicationIn the Gulf of Mexico, deepwater exploration offers the potential for significant discoveries. In the search for new sources of oil, operators rou-tinely drill to 30,000 ft [9,100 m] and beyond.15 The cost of drilling a single well is high, and the number of wells drilled into a structure is inten-tionally kept to a minimum. Because of the extreme depths and possible subsalt placement of target reservoirs, the structure and reservoir

architecture may not be as well understood as it is for shallower horizons. For proper placement of the limited number of wells that are drilled to develop these reservoirs, geologists must have a clear understanding of the subsurface geometry. Geologists start with seismic data to develop res-ervoir models, but for fine-tuning the models, dip-meter and image data are crucial.

Image acquisition in deepwater wells must be performed almost universally in OBM drilling sys-tems, and the ability to acquire high-quality images has been challenging. This difficulty is often compounded by the presence of low forma-tion resistivity and little resistivity contrast between beds; the formation signals are small, and the system has little tolerance for measure-ment error or noise.

15. Bloemenkamp et al, reference 14.

Figure 20. Photorealistic examples. Static (left, Track 1) and dynamic (Track 2) images are presented from data acquired with an FMI-HD tool in a well drilled with WBM. Dynamic (right, Track 1) and static (Track 2) images from a well drilled with OBM using data from a Quanta Geo service. These images can be used to identify stratigraphic features and structural dip.


FMI-HD Tool in WBM

Static Images Static ImagesDynamic Images Dynamic Images

Quanta Geo Tool in OBM

X,358

X,359

X,360

X,361

X,362

X,363

X,280

X,281

X,282

X,283

X,284

X,285

18 Oilfield Review

Older generation OBM imaging tools provided reasonable success for structural analysis although structural dip determination can be difficult in shales that have been altered during lithification and burial. Sedimentological interpretation is usually beyond the limits of these tools.

When high-resolution data are available, geol-ogists can identify textural features from bore-hole images to help them understand the internal structure of thick sediment sections and to define orientable features indicative of stratigraphy. Conventional cores can provide this information; however, because of their prohibitive cost in rig time, acquiring cores in a large number of wells

or over extended openhole intervals is impracti-cal in deepwater projects. The Quanta Geo ser-vice was developed, in part, to address the need for a tool capable of producing photorealistic images in these challenging environments.

To test the imaging capabilities of the Quanta Geo service, a deepwater Gulf of Mexico operator ran the tool in a 97/8-in. wellbore. The well was drilled with a synthetic OBM typically used in the region. The logging toolstring included a dipole sonic tool for determining for-mation mechanical properties. Images were acquired logging down while running in the hole and logging up while pulling out of the hole. The

tool achieved an 80% circumferential coverage of the wellbore.

Thick channel sands are common drilling targets in deepwater exploration. Characterizing these sands, and correctly understanding the stratigraphy, can be illustrated by looking at the information gleaned from the Quanta Geo ser-vice. Interpreters were able to determine that a sequence started with low-energy channel fill followed by a rapid, high-energy influx of mate-rial. Geologists further discovered that what appeared to be a massive sand sequence from standard log interpretation was actually a series of approximately 50 individual depositional

Figure 21. Gulf of Mexico deepwater exploration well. Analysts initially interpreted the sands encountered in a deepwater Gulf of Mexico well as massive channel sands. However, based on interpretation from image data (Track 2), there may have been as many as 50 individual sand bodies. This particular 7-ft [2.1-m] sand interval appears to be uniform; however, the rapid change in dip direction between the bottom and top (Track 3) and the distorted bedding planes (Track 4) are indicative of

a slump fold. The 3D view (right ) clarifies this; the magenta planes show the bedding orientation. The net sand thickness may be much less than it appears in conventional logs and may be disconnected from the rest of the channel sand complex, which will have implications for further field development. This type of information is crucial for development of deepwater reservoirs because operators limit the number of wells drilled. (Adapted from Bloemenkamp et al, reference 14.)


79

80

81

82

83

84

85

86

87

Depth,ft


Interpreted Feature

Bedding PlanesTrue DipGamma Ray

Caliper 1Caliper 2Caliper 3Caliper 4

gAPI0 0 0N NE S W

90 90degree180 270 360


0N NE S W

90 180 270 360

8 in. 18

150

ResistiveConductive ResistiveConductive

September 2015 19

events (Figure 21). In addition, the paleotrans-port direction was inferred from the images. Structural features—unconformities, faults and high-stress regions—were also clearly visual-ized in the image data. Understanding original depositional conditions and structural geometry aided in planning the optimal program for drill-ing and development within the field.

The Quanta Geo service depth of investigation is an order of magnitude shallower than than that of legacy OBM imaging tools. In an OBM environ-ment, mud filtrate usually flushes free formation fluids from permeable sands in the shallow region from which the Quanta Geo service acquires data. The OBM filtrate filling pores in this shal-low region has a high resistivity value. The rela-tive resistivity measured by the Quanta Geo service in shales relates primarily to the conductive and immovable claybound water. Because shales, which have little intrinsic permeability, are not invaded, the relative resistivity computed from

16. Nelson R: Geologic Analysis of Naturally Fractured Reservoirs 2nd ed. Woburn, Massachusetts, USA: Gulf Professional Publishing, 2001.

17. For more on fractures and hydraulic stimulations: Gale JFW, Reed RM and Holder J: “Natural Fractures in the Barnett Shale and Their Importance for Hydraulic Fracture Treatments,” AAPG Bulletin 91, no. 4 (April 2007): 603–622.

18. Nelson, reference 16.19. For more on image logs used to analyze drilling-induced

fractures and in situ stress direction: Aadnøy BS and Bell JS: “Classification of Drilling-Induced Fractures and Their Relationship to In-Situ Stress Directions,” The Log Analyst 39, no. 6 (November 1998): 27–42.

the Quanta Geo data should be unaffected by invasion and comparable to the shale resistivities measured from other sources. By comparing rel-ative resistivities from this shallow depth of investigation, geologists can obtain an accurate net sand count.

The orientation and geometry of drilling-induced fractures were also identified. Such frac-tures are helpful for establishing the maximum horizontal stress direction, especially in combi-nation with mechanical properties determined from advanced acoustic measurements. Natural fractures, which could rarely be visualized in images from older generation OBM tools, were numerous and easily identified.

Fracture CharacterizationNaturally fractured reservoirs make up a signifi-cant portion of global oil and gas reserves.16 The presence of fractures and fracture networks adds complexity to reservoir analysis and reser-voir characterization—a complexity that is absent in reservoirs in which the matrix pore space dominates.17 Operators must understand the nature and characteristics of fractures and fracture networks in reservoirs that must be hydraulically stimulated to produce commer-cially. These fracture systems will greatly affect well performance and field development. As such, completion programs and stimulation designs must include the effects of natural frac-tures and fracture networks.

A common fracture description system labels fractures as open, healed and partially healed. Open fractures generally increase reservoir per-meability and offer conduits to fluid flow. During drilling operations, open fractures fill with drill-ing fluid or seal with mudcake.18 Healed frac-tures, also referred to as mineral-filled and closed fractures, are common (Figure 22). After they form, fractures can fill over time with a sec-ondary cementing material, which is often quartz, carbonate or a combination of minerals. Unlike open fractures, healed fractures can impede reservoir fluid flow. However, fracture stimulation programs often reactivate the frac-ture network along healed surfaces. In some cases, drilling alone can reactivate healed frac-tures. Partially healed fractures exhibit varying degrees of open and closed properties.

For identifying and characterizing natural fractures in situ, analysts may use well logs, which are usually integrated with other techniques to develop a macroscopic view of the reservoir. Fractures are often inferred from logging tool responses rather than measured. For instance, analysts may use sonic log data to identify anoma-

lous responses, such as cycle skipping, which may indicate the presence of fractures.

Drilling-induced fractures are frequently observed in newly drilled wells. These fractures result from wellbore failure during drilling and can be caused by a high mud weight that breaks down the formation. Drilling-induced fractures can usually be distinguished from natural frac-tures in well logs because they appear as mostly parallel but incongruent pairs on opposing sides of the wellbore in vertical wells (Figure 23).19 These fractures, which are indicative of the

Figure 22. Healed fracture. This fracture is filled with mineralized material. Although healed fractures such as this one may not contribute to the intrinsic permeability of a reservoir section, these types of fractures may be reactivated during hydraulic stimulations.


2 cm

Healedfracture

Figure 23. Drilling-induced fractures. Mechanical failure of the borehole wall is evidenced by drilling-induced fractures. These types of fractures are usually parallel features on image logs. Drilling-induced fractures do not contribute to production although they are useful indicators of the direction of maximum principal stress. Drilling engineers can use this information when developing well profiles and may change drilling mud properties to avoid future occurrences.


X,546

X,547

X,548

X,549

X,550

X,551

X,552

X,553

X,554


0North

Resistive Conductive

NorthEast South West90 180 270 360

Depth,ft

Drilling-inducedfracture

20 Oilfield Review

stress profile because they are associated with the direction of maximum principal stress, do not contribute to production.

Another indicator of stress direction is bore-hole breakout, which is characterized by an oval borehole observed in caliper logs. The presence of breakout along one axis is usually an indica-tion of the direction of minimum principal stress. Breakout in one direction combined with the presence of fractures in the adjacent axes are indications of drilling-induced fractures; analysts can infer principal stress orientation based on these features.

Many tools and methods have been developed to detect natural fractures downhole. Some basic methods commonly deployed use seismic, ultra-sonic, sonic, optical and electrical systems.

Geologists can use seismic data to detect frac-ture swarms but not individual fractures. Ultrasonic tools, such as the UBI service, produce full circumferential images of the borehole wall; however, image quality from reflected ultrasonic

pulses is highly dependent on the geomechanical properties and quality of the borehole surface. The best results are achieved in hard formations that have few drilling-induced effects. Dispersion curves from elastic shear waves are also used to characterize fracture systems.20 Tools such as the Sonic Scanner acoustic scanning platform can acquire these measurements on wireline. LWD options include the SonicScope multipole sonic-while-drilling service. Optical methods include downhole cameras and televiewers; mud-filled environments are difficult to image using optical devices, however. The most common and effec-tive method for fracture evaluation involves high-resolution electrical measurements. The FMI-HD and Quanta Geo services are examples of wireline logging tools and the MicroScope HD tool is an LWD example.

Until recently, fracture characterization using imaging logs for wells drilled with OBM posed problems for analysts. The spatial resolu-tion for tools such as the OBMI and OBMI2 ser-

vices is about 1 cm, which is sufficient for structural dip determination. However, the pho-torealistic images provided by the Quanta Geo service redefines OBM imaging both in quality and resolution (Figure 24). This was demon-strated recently in a horizontal well drilled in an unconventional reservoir that has vertical and subvertical fractures.

Finding Fractures in Unconventional PlaysThe advances in technology that enable oil and gas operators to exploit unconventional resources such as organic shales, coalbed methane and tight rocks include horizontal drilling and hydraulic fracture stimulation. The presence of natural fractures, and the activation of those fractures using hydraulic stimulation, is one of the key components for success. When operators lack a thorough understanding of the fracture networks in place, drilling operations, comple-tion designs and stimulation programs may not be optimal.21

For wells in which the presence of fracture networks is key to success, completion and stimu-lation designs that properly leverage the fracture systems can mean the difference between com-mercial success and failure. Many of these wells are drilled using OBM systems, which makes frac-ture characterization difficult because an OBM-filled open fracture will have a resistivity signature similar to that of a mineral-filled healed fracture. The Quanta Geo service identifies fractures and may, in some cases, be able to differentiate open fractures from healed fractures.

Southwestern Energy drilled a vertical evalu-ation well in an unconventional resource play in the Northeast US. The zone of interest was drilled using OBM and had an 81/2-in. borehole. The well was used for data acquisition to understand and characterize the reservoir and would then serve as a pilot hole for a lateral well. A UBI tool was run in addition to the Quanta Geo service, and the images from the two sources were compared.

The increased circumferential coverage of the borehole and its enhanced resolution allow the Quanta Geo service to overcome limitations inherent in older generation OBM imaging tools. Analysts can usually detect high-angle fractures intersecting the borehole that might not be obvi-ous using data from other tools.

Data from the Quanta Geo service can also be processed to evaluate tool standoff, a measure of the degree of pad contact with the borehole wall. The standoff image, generated using an advanced inversion processing technique, is then used to correct the image for standoff effects.22 These data can also be used to generate a sensor stand-

Figure 24. Images from a naturally fractured zone. In this image from a well drilled in the Northeast US, the gamma ray log (Track 1) is indicative of a shale. The dynamically generated image (Track 2) has been interpreted by a geologist. The fractures on the image appear to be dipping to the south; however, these data have not been corrected for borehole drift and tool position. The high-angle fractures are actually dipping to the NNW (Track 3), as indicated by the modified tadpoles, which have been corrected to give true dip. Their strike is ENE–WSW, which the stereonet plot clearly indicates; stereonet plots stack interpreted data to simplify trend identification. The uninterpreted image (Track 4) is presented for reference.


True DipDepth, ft 0 degree 900

N NE S

X,590

X,600

X,610

W90 180 270 360 0

N NE S W90 180 270 360

Image Orientation, degree Image Orientation, degree

Conductive Resistive Conductive Resistive

Classification

W E

N

S

Conductive fractureResistive fracture

W 270°

0°30°

60°

90°

120°

150°180°

210°

240°

300°

330°

E

N

S

Gamma Ray

Bit Size

gAPI0 150

in.6 16

Stereonet plot

September 2015 21

off image, which is a quality indicator of the image generated by the tool and can reflect the presence of hole rugosity. Because the measure-ment comes directly from the borehole wall, the image resolves both geologic and drilling-induced features. Another application of the standoff image is the identification of open fractures.

Tools such as the UBI service are often used to determine fracture status. Images derived from the UBI service result from variations in the acoustic reflectivity of the inner surface of the wellbore. These images are sensitive to minor changes in the surface. Open fractures filled with fluid can be distinguished; however, healed frac-

tures filled with a material that has an acoustic impedance similar to that of the surrounding for-mation are invisible to the tool.23 This application has been used to infer open or closed fractures.

Log analysts compared acoustic reflection images with those from the Quanta Geo service, including the standoff image (Figure 25). The interpreters were able to identify fractures in the dynamic and static images. The inverted standoff images clearly identified the open frac-tures, but healed fractures were not resolved. By inference, fractures observed only in the standoff image are considered open; those not seen in the standoff images are considered closed or very small open fractures. For geolo-gists, the ability to characterize the state of fractures in downhole conditions from Quanta Geo data has great implications for well completion designs and field development.

The Future of ImagingElectrical coring was a vision of the early develop-ers of wireline logging tools. The pictures painted by the latest generation of photorealistic imaging

tools in some ways approach that vision. Logging tools will never completely replace conventional coring because cores provide information that extends beyond visual analysis. However, new techniques and technologies are giving geologists insights into downhole conditions in both WBM and OBM wells never before possible.

The answers from these technologies help guide developers of completion programs to focus on the sweet spots in individual wells and also provide insight into reservoir properties on a scale previously unattainable in OBM-drilled wells. When combined with information from other petrophysical measurements and surface and subsurface seismic data, these new approaches will enable operators to effectively evaluate their resources, optimize development programs and, in some cases, move marginal plays into the realm of commerciality. From a financial standpoint, the resulting picture will be worth more than mere words alone. —TS

Figure 25. Differentiating open and healed fractures. Sinusoids indicate fractures in the static image (Track 1) and the dynamic image (Track 2) from the Quanta Geo service. Modified tadpoles and the stereonet plot (Track 3) indicate the high angle of the fractures and their NNW–SSE strike direction. Determining the fracture status—open or closed—from these images alone is not possible. The UBI image (Track 5) shows many superficial

drilling artifacts—the effects of drilling and backreaming—and vertical scratches from previous logging runs. The fracture at about 977 ft is visible on all images. Two fractures at about 982 ft do not appear in the UBI image or the inverted standoff image (Track 4). These fractures are likely closed. By inference, the standoff image may be a useful indicator of the status of fractures—open and closed.


975

980

True Dip

ClassificationQuanta Geo Static Images Quanta Geo Dynamic Images Quanta Geo Inverted Standoff Static Images

Acoustic Dynamic Amplitude Images

0 degree 90


0N NE S W

90 180 270 360


0N NE S W

90 180 270 360


0N NE S W

90 180 270 360


Dept

h, ft

0N NE S W

90 180 270 360

W E

N

S

Healed fractureOpen fracturePartial fracture

W 270°

0°30°

60°

90°

120°

150°180°

210°

240°

300°

330°

E

N

S

Stereonet plot

ResistiveConductive ResistiveConductive Low Standoff High Standoff

20. For more on sonic data used for fracture detection: Haldorsen JBU, Johnson DL, Plona T, Sinha B, Valero H-P and Winkler K: “Borehole Acoustic Waves,” Oilfield Review 18, no. 1 (Spring 2006): 34–43.

21. For more on geosteering and horizontal drilling: Amer et al, reference 4.

22. For more on using standoff images for fracture characterization: Chen et al, reference 10.

23. For more on using images from the UBI service for fracture characterization: Ellis D, Engelman B, Fruchter J, Shipp B, Jensen R, Lewis R, Scott H and Trent S: “Environmental Applications of Oilfield Technology,” Oilfield Review 8, no. 3 (Autumn 1996): 44–57.

22 Oilfield Review

Sand Screen Selection

Many formations produce sand that may hinder production or damage completion

and surface equipment. For decades, the industry has chosen sand control screens

to address this threat based on traditional practices. Research suggests a new

methodology that uses numerical simulation for selecting screen size and type may

improve outcomes.

Jamie Stuart AndrewsStatoilStavanger, Norway

Joseph A. AyoubRajesh A. ChanpuraMehmet ParlarSugar Land, Texas, USA

Somnath MondalShell International E&PHouston, Texas

Mukul M. SharmaThe University of Texas at AustinAustin, Texas

Oilfield Review 27, no. 2 (September 2015).Copyright © 2015 Schlumberger.

Adopting methods first used in water wells, early 20th century oil and gas operators concerned with potential sand production from unconsoli-dated formations completed wells using pipe that had slotted or round openings. The openings, placed across the production interval, were sized to prevent sand from entering the wellbore while minimally constricting fluid flow.

In time, the oil and gas industry developed sand retention methods that incorporated screens, resin- or plastic-coated particles and gravel packs. Some companies have, in recent years, sought to distinguish between sand man-agement and sand retention, in which the former uses techniques such as orientation of the well-bore and perforations, monitoring and control of

1. Tronvoll J, Dusseault MB, Sanfilippo F and Santarelli FJ: “The Tools of Sand Management,” paper SPE 71673, presented at the SPE Annual Technical Conference and Exhibition, New Orleans, September 30–October 3, 2001.

2. Coberly CJ: “Selection of Screen Openings for Unconsolidated Sands,” Drilling and Production Practice (January 1, 1937): 189–201.

The notation refers to the percent of sand particles by mass within the formation that are larger than that value. A d10 designation means 10% of the sand particles in a formation are larger than that value; thus, 10% of the sand particles in a formation are larger than a d10 sand particle size.

3. Coberly, reference 2.

September 2015 2323

well pressures, fluid rates and sand influx to limit sand production.1 Sand retention, or sand con-trol, refers to the use of screens and other tools to reduce the risks of sand production without restricting oil and gas productivity.

Early sand control efforts centered on the assumption that choosing the optimal sand screen was based on a relationship between screen opening and a single point in grain size distributions. Experiments performed under ideal prepack test conditions using spheres of a single diameter led early researcher C.J. Coberly to conclude that negligible particle production occurs through rectangular slots of widths that are twice the particle diameter or through circu-lar openings that have diameters three times that particle diameter (Figure 1).2

In formation sand samples, particles have a size distribution, which forced Coberly to pick a characteristic diameter, d, within that size distri-bution based on physical experiments using for-mation sand samples. Sizing the slot width to twice d10 (2d10) to allow negligible transient sand production is known as the Coberly rule. In response to Coberly’s work, H.D. Wilson wrote that for sand samples from the US Gulf Coast, for example, proper retention of sand required sizing the slots to no larger than d10.3 Industry experts have concluded that the differences in those con-clusions are related to what constitutes a negli-gible amount of produced sand and to the attempt to characterize the entire particle size distribu-tion using a single parameter.

Other aspects of selecting a slot or screen size based on traditional practices involve taking rep-resentative sand samples and characterizing those samples. Most representative samples are obtained through conventional cores retrieved from known depths.

To characterize formations, laboratory tech-nicians determine the particle size distributions (PSDs), typically by sieve or laser analysis or both. In recent years, the use of laser particle size analysis (LPSA) has become common in some companies because such analysis can bet-ter provide the details of the finer portion of the particle size distribution than can sieve analysis. In addition, laser analysis is less labor intensive than sieve analysis and thus typically lower in cost, which allows operators to economically ana-lyze many samples.

Using the most representative sample avail-able, engineers typically determine proper screen openings based on the coarsest 10% of a particle size distribution, or d10. Screens that have slot widths determined by this process are designed to allow some amount of sand to pass

while the coarsest particles are retained by size exclusion or bridging. In the process of reten-tion, fine particles are retained by the pore space of the coarse grains and even finer parti-cles retained between the pore space of the fine particles; this process repeats until sand pro-duction ceases.

This article describes the process by which engineers match optimal wire wrap and metal mesh stand-alone screen (SAS) size and type to target formations in openhole completions. In addition, this article discusses a technique that allows engineers to use the entire sand size dis-tribution when selecting a screen and to quickly narrow the range of screen sizes and types to optimize sand control. This process often results in sand control decisions more suited to the well at hand than is possible using past practices that use only one design parameter, such as d10, and reduces the number of laboratory tests that must

be performed to determine the optimal choice for the target formation. A case history from offshore West Africa demonstrates the potential for the methods discussed.

How Choices Are MadeBefore the drill bit breaks ground, operators must make various decisions that will impact how the completion is finally configured. Engineers must then decide whether to case, cement and perforate the production interval or to use an openhole completion.

Openhole completions, typically less costly than cased hole completions, may be completed using gravel packs or stand-alone screens if the formation is expected to produce sand. Stand-alone screen types include wire wrap screens (WWSs) and metal mesh screens (MMSs). To cre-ate a WWS, manufacturers wrap wire around a perforated base pipe. The wire is either placed

Figure 1. Traditional method for sizing screens. Tests performed in the 1930s, using sands of a single grain size on screens that had rectangular slots, resulted in curves that are nearly linear functions. A stable particle bridge formed across slots whose width was about twice that of the grain diameter (red), and all sand passed through the screen that had a slot size about three times the grain diameter (blue). (Adapted from Coberly, reference 2).

0.35

0.30

0.25

0.20

0.15

0.10

0.05

0.000.10 0.090.080.070.060.05

Screen opening, in.0.040.030.020.010.00

Slot

wid

th, i

n.

Sand p

asse

s thr

ough

Sand bridges

24 Oilfield Review

around the pipe during manufacturing or manu-factured as an individual jacket that is later welded to a base pipe. Mesh screens include one or more layers of woven stainless steel or mesh wire wrapped around a base pipe. The mesh, which acts as a filter, is covered by a protective shroud (Figure 2). Although uncommon, opera-tors have included shrouds on WWSs in side-tracked wells that have challenging casing exits.

Even when widespread agreement exists that SASs are appropriate, recommendations for screen type and opening size often vary widely. Early efforts at screen sizing were based on a single point (d10) on the PSD and some amount of sand production that was assumed to be accept-able, as described earlier.4

In the 1990s, a mathematical model was developed to optimize sizing of slots in sand con-trol devices. This model was based on a fractal description of the entire PSD given in terms of the number of particles rather than particle mass.5 A series of laboratory tests were performed to establish a database of wire wrap screen behavior results using sands from the North Sea and the Haltenbanken Area offshore Norway. From these experiments and the number-based particle size distributions, four slot widths were defined for each sand type tested: d22, d2, d+ and d++ (Figure 3). The designation d22 was the largest slot size at which severe plugging occurred and d++ was the smallest slot size at which continuous sand production occurred. The d2 and d+ slot widths were defined as the small-est hole size that did not allow plugging and the largest slot size that did not allow continuous sand production, respectively.6 The ideal slot size was stipulated to be between d2 and d+.

Completion engineers often use these crite-ria to constrain screen size options before per-forming sand retention tests (SRTs) in the laboratory to determine a final screen size. Two types of SRTs are available: slurry tests and pre-pack tests. Slurry tests are designed to replicate

4. Coberly, reference 2. 5. Markestad P, Christie O, Espendal Aa and Rørvik O:

“Selection of Screen Width to Prevent Plugging and Sand Production,” paper SPE 31087, presented at the SPE Formation Damage Control Symposium, Lafayette, Louisiana, USA, February 14–15, 1996.

6. Markestad et al, reference 5.7. Chanpura RA, Hodge RM, Andrews JS, Toffanin EP,

Moen T and Parlar M: “A Review of Screen Selection for Standalone Applications and a New Methodology,” SPE Drilling & Completion 26, no. 1 (March 2011): 84–95.

8. Chanpura et al, reference 7. 9. Mondal S, Sharma MM, Chanpura RA, Parlar M and

Ayoub JA: “Numerical Simulations of Sand-Screen Performance in Standalone Applications,” SPE Drilling & Completion 26, no. 4 (December 2011): 472–483.

Figure 2. Wire wrap and metal mesh screens. Both wire wrap screens (left) and metal mesh screens (right) are constructed around a perforated base pipe. Wire wrap screens include a screen that can be slipped over the base pipe and welded into place. The metal mesh screens, made of woven metal layers that may include sintered metal, are placed between the base pipe and the perforated protective shroud.

Wire wrap screen Metal meshscreen

Perforatedbase pipe

Protectiveshroud

Figure 3. Slot width ranges for sand screen design. Mathematical modeling and laboratory results led scientists to define four slot widths for each target sand based on sand grain diameter (d). The lower and upper limits of width sizes are defined by d22 and d++. The optimum size range that will neither plug nor produce sand is bounded by d2 and d+ (green).

May plug

d– – d– d+ d++

Optimum zone May producesand

Will notproduce sand

Will alwaysproduce sand

Will notplug

Will alwaysplug

September 2015 25

gradual failure of the rock surrounding the borehole (Figure 4). During slurry tests, a low-concentration slurry is pumped at a constant rate to form a sandpack around the screen. The mechanism of sand retention, therefore, is dic-tated only by particle size exclusion.

To perform prepack tests, which represent complete hole collapse, technicians place a sandpack on the screen and pump clean solids-free liquid through the pack. Because a sand-pack is already in place, sand retention during a prepack test is achieved through both size exclusion and bridging.

Recent research has shown that current SRT setup and interpretation methods tend to favor one screen type or other. The traditional criteria used to choose between a gravel pack or an SAS are overly conservative and often lead analysts to opt for a gravel pack. Numerous experiments indicate that, contrary to accepted wisdom, screen plugging is rarely a problem in clean sand formations; when plugging is a threat as a result of other factors such as contaminated fluids, the risk can be mitigated through proper hole prepa-ration procedures.7

To address the variability and inconsistency inherent in screen selection and to better under-stand the physics of sand control, scientists recently used a numerical simulation approach to evaluate sand screen performance. The effort was part of a larger plan to produce a systematic screen selection process.

Screen sizing practices that relied on accepted standards were based on PSDs that did not use the results of sand retention tests. Despite the limitations of these standards, which are based on a few parameters of the formation sand size distribution and implicit assumptions about acceptable levels of sand production, most experts continue to use such standards not only to narrow screen size options but also to perform SRTs to confirm final screen selection.

In general, three results from SRTs are of interest: sand production correlated to the screen’s sand retention efficiency, pressure development correlated to screen plugging ten-dency and size distribution of produced particles with which to evaluate the risk of screen erosion. However, because it has now been established that screen plugging is rarely a problem in clean formation sand of any PSD, the main criteria for screen selection become transient sand produc-tion and PSD of produced particles. Engineers can determine both criteria using models devel-oped in the last five years for specific screen and PSD combinations without having to conduct actual SRTs.8

Model AlternativeA team from academia and industry reviewed recent screen testing advancements, interpreta-tion and modeling for SAS applications. Based on its findings, the team has proposed a screen selection method based on laboratory test–veri-fied numerical and analytical models.

The primary purpose of this method is to elim-inate or reduce the number of physical SRTs that must be performed when selecting a screen size and type for a given application and to better

understand the science of sand retention. The study used numerical SRT simulations that matched experimental data in an effort to aid the team in understanding and relating PSD-screen combinations and to correlating sand production with formation sand PSD until sand production stops or is limited to fines.9

The team first studied WWSs, which have a simpler geometry than that of other screen types, and performed simulations using the discrete element method (DEM). This numerical model describes mechanical behaviors, such as mass,

Figure 4. Two types of sand retention tests. Slurry tests (top) are designed to simulate gradual failure of the formation surrounding the borehole. Technicians pump a low-concentration sand slurry through a screen coupon, then measure the weight of solids produced through the screen and the pressure buildup across the screen versus the amount of sand contacting the screen. Laboratories design prepack tests (bottom) to simulate a complete borehole collapse by placing a sand sample directly on the screen. A liquid is then flowed through the sand and screen. Technicians then create a confining stress on the sample that forces the sand into full contact with the screen. The test measures the amount of sand that passes through the screen—measured by weight—and the pressure drop across the screen.

Pressuremonitoringpoints

Sand-fluid inlet

Screen coupon

Outlet for sample capture

Pressuremonitoringpoints

Sand-fluid inlet

Outlet for sample capture

Weave sample

Sandpack

Retaining mesh

Piston

26 Oilfield Review

velocity, force and angular momentum, of assem-blies of spheres (Figure 5).10 The study simulated prepack experiments by first generating a pack-ing of polydisperse granular spheres over a WWS geometry and then flowing a fluid through the pack. The research team could then compute the

mass of sand produced per unit area of screen for various screen sizes and PSDs.

To accurately represent the physics of the problem, the model was tested and validated using a range of various parameters. The team found that friction and shear forces are necessary

to form stable particle bridges, whereas the most critical parameter affecting the number of sand particles produced is the ratio of the slot width to particle diameter. Similarly, high fluid viscosities and low pressure gradients facilitate particle bridging; increased fluid pressure increases par-ticle production when pressure gradients are up to about 2.3 MPa/m [100 psi/ft]. At higher gradi-ents, however, there is no such dependence.

When the results from the DEM model were plotted, the team observed a power-law relation. This relationship was confirmed by plotting the experimental data, which revealed excellent agreement and consistent trends between model and experimental results. Based on this newly established relation, the team developed the Mondal-Sharma (M-S) method, which uses the number and size of the produced solids to esti-mate the mass of sand produced (Figure 6). When comparing the estimated mass of sand pro-duced using the M-S method with the mass of sand produced in experiments, a good match was found. The M-S method, which uses DEM simula-tion results to develop a simple correlation, can be used to estimate the mass of sand produced without performing DEM simulations for every possible sand and screen combination.11

The research team next extended the applica-tion of the M-S method to include plain square mesh (PSM) screens, achieving much the same outcomes. Some conclusions from WWS and PSM simulations included the following:• Simulations are able to estimate the mass of

sand produced for a given PSD and screen size.• Simulations results strongly agree with those

from carefully controlled prepack experiments.• Simulations show that the mass of sand pro-

duced per unit screen area and for unit open flow area is larger for single layer PSMs than for slot geometry of the same rating and corre-sponding standard open flow area.

• Simulations show that the ratio of wire thick-ness to opening size seems to be a key factor contributing to the increased mass of sand pro-duction from single layer PSMs.12

Researchers then turned their attention to analytical solutions and Monte Carlo simulations to predict sand production through WWSs and PSM screens under slurry test conditions. Their results showed that the analytical solution and the numerical simulation were in excellent agreement. The team showed that its proposed methods were able to estimate both mass and size distribution of the produced solid in a slurry-type SRT, taking into account the full PSD of for-mation sand. Simulations also showed that, with the exception of a mobile fines problem, sand

Figure 5. Simulations of sand retention tests using the discrete element method (DEM). Using the DEM, scientists track information such as mass, velocity, force and momentum about each particle within the computational domain, or simulation box (left). Researchers used the DEM and a molecular dynamics simulator to model performance in a prepack experiment by generating a packing of polydisperse granular spheres (multicolored balls) over a wire wrap screen geometry (blue layer) and then flowing liquid through the pack. The individual size and number of particles per size were obtained from the measured particle size distribution of the formation sand sample used for the corresponding experiment. Discrete element method simulations were then used to calculate the mass of sand produced per unit area of screen for various screen sizes and particle size distributions. Near the end of the polydisperse simulation, which required 24 hours on a 48-processor network cluster, sand particles (right, green, purple, brown, blue and white) bridge across the screen openings (pink).

Simulation Box Simulation End

Figure 7. Particle size distribution (PSD) of retained and produced formation sand. The PSD of the first layer of sand retained on the screen (red) has the expected distribution of the sand particles of widths greater than the slot size. The PSD of the second layer (green) is approaching that of the formation sand (blue). Because the sizes of the particles retained on the second layer are dictated by the pore sizes of the first layer, the retained particles will eventually be of the same PSD and permeability as the formation sand. (Adapted from Chanpura et al, reference 13.)

100

90

80

70

60

50

20

10

40

30

0

Particle size, microns1,000

Cum

ulat

ive

% b

y w

eigh

t

100 10

September 2015 27

production becomes negligible once the slot opening has been covered by particles larger than the opening (Figure 7).

As in the case of modeling prepack-type SRTs, the proposed methods can be used to estimate sand production in slurry-type SRTs for various screen sizes, thereby enabling screen size selec-

tion based on an acceptable level of sand produc-tion. Final screen selection may then be confirmed through a slurry-type SRT.13 Results showed that more than 90% of the total sand pro-duction by mass occurs during the formation of the first layer of particles on the screen and that the PSD of the retained sand approaches that of

the formation sand after a few layers of sand accumulate on the screen.

Results also revealed that the mass of sand produced during the formation of the first layer of particles on the screen is independent of the shape of the PSD for grains smaller than the aperture-pore size and is governed by the shape

10. Cundall PA and Strack ODL: “A Discrete Numerical Model for Granular Assemblies,” Géotechnique 29, no. 1 (March 1, 1979): 47–65.

11. Mondal et al, reference 9.

Figure 6. Determining mass of sand produced and entire formation particle size distribution. The Mondal-Sharma (M-S) method uses a correlation between the number of particles of diameter Dp produced through a screen slot opening of width, W. The number of particles of each diameter produced through the screen are counted and plotted against Dp/W from every simulation (top). In this case, formation PSDs A and B were distributed into five bin sizes each (bottom, dashed lines) to generate the number-based size distributions (D1A to D5A and D1B to D5B) used to populate the simulation box (bottom). (Adapted from Mondal et al, reference 9.)

1 × 1012

1 × 1011

1 × 1010

1 × 109

1 × 108

1 × 107

1 × 106

1Dp /W

y = A5x – b 5

R 2 = 0.8998

0.01 0.1

Num

ber o

f par

ticle

s pr

oduc

ed/m

2 for 1

0% o

pen

flow

are

a

100

80

60

40

20

01

Diameter, microns1,000 100 10

Cum

ulat

ive

by w

eigh

t,%

D 5A D 4A D 3A D 2A D 1A

PSD A PSD B

12. Mondal S, Sharma MM, Hodge RM, Chanpura RA, Parlar M and Ayoub JA: “A New Method for the Design and Selection of Premium/Woven Sand Screens,” SPE Drilling & Completion 27, no. 3 (September 2012): 406–415.

13. Chanpura RA, Fidan S, Mondal S, Andrews JS, Martin F, Hodge RM, Ayoub JA, Parlar M and Sharma MM: “Advancements in Screen Testing, Interpretation and Modeling for Standalone Screen Applications,” paper SPE 143731, presented at the SPE European Formation Damage Conference, Noordwijk, The Netherlands, June 7–10, 2011.

28 Oilfield Review

of the PSD of grains greater than the aperture-pore size. In addition, researchers found that sand production through the filter layer of a PSM screen of a given pore size is greater than that of a WWS of the same slot size (Figure 8).14

The Mythology of Screen SelectionThe team’s work has cast doubt on, or added qualifications to, numerous widely held industry beliefs about WWSs and PSMs. These axioms, upon which many traditional screen selection methodologies for SASs have been based, include the contention that formation sand plugs screens.

However, research has shown that following SRTs, when only trapped particles remained on the screens, final screen permeability was in the range of 5% to 100% of original screen permea-bility; the final value, then, of even the low-per-meability SAS screens, which have an original screen permeability of about 300 D, would be a minimum 15 D. The screen permeability is thus

significantly higher than most formations and thus too great to cause plugging; plugging is commonly quantified by a pressure differential created across the screen. Instead, plugging more likely occurs as the result of poorly condi-tioned mud or filtercake mixed with formation sand, mixed coarse and fine formation sands from a variety of zones or clay and shale mixed with formation sand.15

PSD and PoSD When SRTs are performed in the laboratory using formation sand, the sand PSD is often not needed. However, PSD is required if there is a large spread in formation PSD along the well, or if the SRT is performed using a sample that was gener-ated based on specified PSD or if a model is used to estimate sand production for a given sand PSD–screen combination. Particle size distribu-tion of formation sand is typically determined through dry sieve analysis or laser particle size analysis (LPSA).16

Dry sieve analyses determine PSD through a mechanical separation of particles by filtering them from top to bottom through a series of pro-gressively finer sieves. The measured weight of the sand captured in each sieve is used to calcu-late cumulative percentage mass of each, which is then plotted against sieve size on a semiloga-rithmic scale.

Laser particle size analyses determine PSD by measuring how light is scattered as a laser beam is passed through a sand sample. The angle of scatter is inversely proportional to the particle size.17 To ensure that the sand samples are deliv-ered to the measurement device in the correct concentration and in a stable state, LPSA is per-formed on samples whose dispersion is controlled by dry or, when necessary, fluid dispersants.

Sand control experts have long used dry sieve and LPSA nearly indiscriminately, and persistent differences in the results obtained from the two methods have been well docu-mented. Recent research indicates these incon-sistencies may be caused by the aspherical shape of the particles, sampling practices for LPSA, fluids used and various light blocking lev-els used in the LPSA. Based on these observa-tions, PSD determined by dry sieve analysis is recommended for both slurry-type SRT testing and sand production prediction using the above

Figure 8. Comparison of sand production through plain square mesh (PSM) and wire wrap screens (WWSs). The mass of sand produced for seven PSD values through a 175-micron single-layer PSM is greater than that produced through a 175-micron WWS per unit screen area (top) and per unit open flow area (OFA) (bottom). (Adapted from Chanpura et al, reference 15.)

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0

Formation sample

Mas

s of

san

d pr

oduc

ed/s

cree

n ar

ea, k

g/m

2

175-4175-3175-2175-1 175-5 175-6 175-7

2.5

0.0

0.5

1.0

1.5

2.0

Formation sample

Mas

s of

san

d pr

oduc

ed/O

FA, k

g/m

2

175-4175-3175-2175-1 175-5 175-6 175-7

Single-layer PSM WWS

Single-layer PSM WWS

14. Chanpura RA, Mondal S, Andrews JS, Mathisen A-M, Ayoub JA, Parlar M and Sharma MM: “New Analytical and Statistical Approach for Estimating and Analyzing Sand Production Through Plain Square-Mesh Screens During a Sand-Retention Test,” SPE Drilling & Completion 28, no. 2 (June 2013): 135–147.

15. This research revealed that numerous assumptions regarding sand production and screen characteristics were unfounded. For more on the team’s discussion of traditional assumptions: Chanpura RA, Mondal S, Sharma MM, Andrews JS, Mathisen A-M, Martin F, Marpaung F, Ayoub JA and Parlar M: “Unraveling the Myths Associated with Selecting Standalone Screens and a New Methodology for Sand-Control Applications,” SPE Drilling & Completion 28, no. 3 (September 2013): 227–236.

16. Zhang K, Chanpura RA, Mondal S, Wu C-H, Sharma MM, Ayoub JA and Parlar M: “Particle Size Distribution Measurement Techniques and Their Relevance or Irrelevance to Sand Control Design,” paper SPE 168152, presented at the SPE International Symposium and Exhibition on Formation Damage Control, Lafayette, Louisiana, February 26–28, 2014.

17. Ballard T and Beare S: “Particle Size Analysis for Sand Control Applications,” paper SPE 165119, presented at the SPE European Formation Damage Conference and Exhibition, Noordwijk, The Netherlands, June 5–7, 2013.

18. Zhang et al, reference 16. 19. Mondal S, Wu C-H, Sharma MM, Chanpura RA, Parlar M

and Ayoub JA: “Characterizing, Designing, and Selecting Metal Mesh Screens for Standalone Screen Applications,” paper SPE 170935, presented at the SPE Annual Technical Conference and Exhibition, Amsterdam, October 27–29, 2014.

20. Agunloye E and Utunedi E: “Optimizing Sand Control Design Using Sand Screen Retention Testing,” paper SPE 172488, presented at the SPE Nigeria Annual International Conference and Exhibition, Lagos, Nigeria, August 5–7, 2014.

September 2015 29

models. However, errors or differences attrib-uted to particle shape differences may still occur.18 These differences may be minimized by characterizing the particle shape and aspect.

Recent investigations of mesh screens have highlighted the need to account for screen com-plexity from layered design when modeling sand production. Using microcomputed tomography (microCT) images, researchers constructed 3D images of two metal mesh screen types: PSM and plain Dutch weave (PDW) (Figure 9). These 3D images of virtual screens were validated by com-parison with the microCT images.

The team then conducted DEM simulations that were validated by experiments of prepack SRTs through multilayer PSMs and PDWs. Analyses of microCT scan meshes indicated that mesh screen layers overlap significantly and thus impact retention efficiency. The group developed a method to calculate the retention pore size dis-tribution (PoSD) and effective pore size for a given overlap of PSM samples. The calculated PoSD can be used in the analytical model to improve sand production prediction in a slurry-type SRT.

As a consequence of this work, the perfor-mance of nominal size MMSs can be simulated using any reservoir sand size distribution. To date, because the team has been able to charac-terize PSMs, operators are able to evaluate a large number of PSMs in a short time and thus reduce the number of SRTs that must be run to choose the optimal screen size for a given reser-voir.19 In time, this work will be expanded to include additional screen types.

By the Numbers Engineers use SRTs to choose the optimal screen from a range of screens selected based on a rela-tionship between screen openings and grain sizes. Although SRT results can be impacted sig-nificantly by relatively small changes to test con-ditions, when performed properly, the SRT is widely considered a reliable method for finalizing screen choice. The drawback to this process, however, lies in the dubious traditional practices used to narrow the range of screen choices and in misinterpretation of pressure developments in standard SRT experiments. This process often forces operators to choose to perform many time-consuming SRTs before qualifying a screen as optimal for long horizontal sections that have varying sand PSD.

By replacing traditional methods with numer-ical and analytical models, operators may reduce and eventually eliminate the dependence on SRTs. In addition, because traditional screen selection methodology tends to be conservative, a software-based approach may allow operators to opt for SASs over gravel packs, which are typi-cally more expensive.

When working offshore West Africa required sand control for a nonuniform unconsolidated formation, a major operator based its screen selection process on traditional d10 preselection criteria and on SRTs for finalizing its selection. The completions team also compared the results of the laboratory tests to numerical models.

The targeted reservoir is the second sand in the offshore field; wells in the first sand of the field were completed using sand control devices selected based solely on traditional methods.

However, the first formation produced is made up of highly uniform, well-sorted reservoir sands that have very low levels of fines content. By con-trast, the targeted sand in the second reservoir is much less uniform, poorly sorted and has higher fines content. In the face of these adverse sand control indicators, the operator opted to perform as rigorous a selection process as possible and to check selections based on traditional and SRT methods against those using simulations and mathematical models.

In comparing results, the operator concluded that selections based on the results of SRTs and those based on the mathematical models matched closely. The operator added that although models require laboratory data for proper calibration, they held significant potential for aiding screen size selection without the need for continued laboratory testing when applied in regions for which extensive SRT data existed.20

The quantity of and interaction between the variables that engineers must consider in choos-ing a sand control strategy can be daunting. For decades, engineers have relied on the experience of their predecessors to help them sort the data and arrive at decisions. Today, however, because of the growth of computing power and capacity, operators may avail themselves of more accurate and less compromising methods for sand control selection. Based on physics and mathematics, these new methods promise not only a quicker, less costly path through the selection process, but one that provides engineers with the cer-tainty that they have chosen an optimal sand con-trol strategy for any given formation. —RvF

Figure 9. High-resolution microCT scans of PSM. A 3D PSM screen image (left) can be reconstructed from a microCT scan using a commercially available computer-aided design format that is able to preserve and reproduce minute detail (center and right). (Adapted from Mondal et al, reference 19.)

30 Oilfield Review

A New Approach to Fixed Cutter Bits

The goal of drillers everywhere is to drill as quickly as possible from casing shoe to

casing point without compromising borehole quality. The bit, which must withstand

variations in lithology, formation compressive strength and numerous other factors,

is central to achieving this goal. A new bit, which has conical diamond cutting

elements arrayed across its face, is attaining extended run lengths and increased

penetration rates through challenging formations. This bit also delivers higher build

rates and a balanced steering response in directional drilling applications.

Michael AzarWiley LongAllen WhiteHouston, Texas, USA

Chance CopelandMidland, Texas

Ryan HemptonCimarex Energy CompanyMidland, Texas

Mikhail PakMoscow, Russia

Oilfield Review 27, no. 2 (September 2015). Copyright © 2015 Schlumberger.For help in preparation of this article, thanks to Diane Jordan, Houston.IDEAS, ONYX 360, StingBlade and Stinger are marks of Schlumberger.

1. Ortiz B, Casallas C and Parra H: “Improved Bit Stability Reduces Downhole Harmonics (Vibrations),” paper IADC/SPE 36413, presented at the IADC/SPE Asia Pacific Drilling Technology Conference and Exhibition, Kuala Lumpur, September 9–11, 1996.

2. Bit whirl occurs when a bit’s axis of rotation is not in line with the bit’s physical center. Instead, one of the cutters becomes an instantaneous center of rotation, forcing the bit to rotate about this contact point rather than about the bit center. As the bit rotates about this contact point, friction builds between the wellbore wall and bit, and torque in the drillstring increases, which can force the bit to move in reverse relative to the surface rotation of the drillstring, or laterally, creating high-impact loads on the bit and BHA.

For more on bit whirl: Centala P, Challa V, Durairajan B, Meehan R, Paez L, Partin U, Segal S, Wu S, Garrett I, Teggart B and Tetley N: “Bit Design—Top to Bottom,” Oilfield Review 23, no. 2 (Summer 2011): 4–17.

Brett JF, Warren TM and Behr SM: “Bit Whirl: A New Theory of PDC Bit Failure,” paper SPE 19571, presented at the 64th SPE Annual Technical Conference and Exhibition, San Antonio, Texas, USA, October 8–11, 1989.

3. Allamon JP, McKown T, Hill D, Brooks BA, Bayoud BB and Winters WJ: “Diamond Bit Handling and Operation,” paper SPE/IADC 16144, presented at the SPE/IADC Drilling Conference, New Orleans, March 15–18, 1987.

September 2015 3131

The price of drilling into certain formations is paid in terms of shock and vibration to downhole tools, slow penetration rates and damaged bits. Hard or abrasive sandstones, interbedded sands and shales, conglomerates, carbonates containing chert and clays containing pyrite are particularly tough on drill bits. Encounters with such forma-tions may compel drillers to trip out of the hole to exchange their damaged bit for something harder.

Diamonds, one of the hardest materials in the world, have been used in drilling applications since about 1910, when they were first used in cor-ing bits. By the early 1920s, they were incorporated into fullbore drill bits. In the 1970s, synthetic dia-monds, bonded onto tungsten carbide, led to the development of fixed cutter polycrystalline dia-mond compact (PDC) bits. Further advances in materials science and manufacturing have led to a new generation of fixed cutter PDC bits, which continue to evolve to meet the challenge of drilling in variable lithologies and along complex trajecto-ries. However, even a PDC cutter is subject to chip-ping and impact damage that can slow progress or force the driller to trip for a new bit.

Although rate of penetration (ROP) typically increases following a bit change, the time spent tripping out of the hole and back to bottom is flat, or nonproductive, time not spent on drilling, which adversely impacts efficiency and drilling costs. The most obvious way to increase drilling efficiency and reduce costs is to drill from the cas-ing shoe to the next casing point as quickly as pos-sible using just one bit. When this ideal is not attained, operators must choose between staying on bottom and enduring lower penetration rates or tripping for a new bit to increase ROP. Each choice exacts a penalty in terms of rig time. Often, bit selection requires a compromise that balances impact and wear resistance against ROP.

Within local basins, bit selection is typically driven by operator experience in drilling through a particular formation. Carbonate formations can be characterized by a range of lithologies—some of which are easier to drill than others—from soft marls and limestones to hard and brittle dolo-mites. Evaporites also present a variety of chal-lenges, including cutter overload in hard anhydrites, inhibited drilling efficiency in lami-nated gypsum, and washouts in soluble salts. Clastics may reduce ROP when the cuttings stick to the bit and obstruct the bit’s junk slots and waterways. Sandstones and siltstones often cause abrasive wear. Some plays lie beneath basalts, which can be especially hard and abrasive.

Formation depth also plays a role in bit selec-tion because formation compressive strength tends to increase with depth. Some formations are notori-

ously hard, having compressive strengths that range from 207 to 380 MPa [30,000 to 55,000 psi] and, depending on thickness, may require several days and several bits to drill through.

Matching the right bit to a formation might not be so difficult but for the fact that most formations are not homogeneous. Frequently, multiple or mixed lithologies lie between the bit and the next casing point. And it is the abrupt transition from one rock type to another that can lead to bit dam-age and durability problems. Drilling from one lithology to another—or from one compressive strength regime to another—can produce high instantaneous impact forces, cyclic lateral forces and vibrations that can cause accelerated bit wear and failure. When retrieved to surface, a PDC bit that has failed because of vibration will have chips, fractures and gross cutter breakage—all attribut-able to severe impacts on the diamond table of the PDC cutting structure.1

At the surface, lithologic changes may regis-ter as fluctuations in rotary torque or ROP, but such indicators only hint at what is happening downhole. A hard, abrasive sandstone, for exam-ple, may cause accelerated cutter wear or dam-age. Some otherwise soft shale formations offer deceptively hard drilling, owing to the presence of calcite concretions or pyrite nodules, which are significantly harder than the shale mass itself. Calcite [CaCO3] concretions are formed through solution deposition and may range from a few cm to 30 cm [1 in. to 1 ft] in diameter. These concretions can have compressive strengths in excess of 260 MPa [38,000 psi], whereas the sur-rounding shale may have a compressive strength

of around 34 MPa [5,000 psi]. Similarly, small nodules of pyrite [FeS2], often found in shale, can also be troublesome.

Drilling into a formation characterized by mixed lithologies can create intense cutter loading and cyclic lateral forces that cause bit whirl, which in turn, leads to impact damage of PDC cutters.2 The formation characteristics, bit design and the required bit performance will determine if modifi-cation of operating parameters will support fur-ther drilling or warrant a trip for a new bit.3

In hard formations, the driller must increase weight on bit (WOB) to overcome the formation shear strength needed to fail the rock and main-tain an acceptable ROP. However, higher WOB sig-nificantly increases cutter loading, which can lead to microchipping of the diamond table in PDC cut-ters. The bit dulls as the cutter wear flat area increases, which increases frictional heating at the interface between cutter and rock, potentially weakening the diamond cutting element.

Not only is transition drilling a problem; the capability to drill through a curve is a significant challenge for plays in which economics of pro-duction depend on lateral drilling. Building angle generates considerable torque at the bit and can create toolface control difficulties for some PDC bits, making it difficult to maintain trajectory.

To address these challenges, bit engineers developed a fixed cutter bit that employs a unique type of cutting element. The Stinger conical dia-mond element (CDE) provides a significantly thicker layer of diamond than do conventional PDC cutters (Figure 1).

Figure 1. Conical diamond element (CDE). The Stinger CDE (left ), manufactured under high temperature and pressure, has a layer of synthetic diamond that is substantially thicker than that of a conventional PDC cutter (right ). The polycrystalline diamond material of the conical cutter is engineered to provide a level of impact strength and resistance to abrasive wear that is higher than that of the conventional PDC cutter (center ).

Rela

tive

scal

e

Wear resistance

Impact strength

Diamond thickness

CDE cross section

Diamond

PDC cross section0

0.25

0.50

0.75

1.00

1.25

CDE PDC

Diamond

Substrate

Substrate

32 Oilfield Review

The StingBlade conical diamond element bit incorporates an array of Stinger cutting elements across the bit face. Depending on the application, these cutting elements may be positioned any-where from bit center to gauge. This cutter array has enabled operators to improve ROP and drill significantly longer intervals than was possible using conventional PDC bits. In some wells, StingBlade bits were able to drill continuously from shoe to casing point in a single run through formations in which this was previously impossi-ble. In addition, the new bit design has provided improved toolface control in challenging direc-tional applications. The following discussion

focuses on the StingBlade bit, its design and its performance in drilling some of the toughest for-mations around the world.

Bit DesignThe Stinger conical diamond element was initially introduced as a stand-alone cutting element placed at the bit center to improve ROP and enhance dynamic stability for PDC bits (Figure 2). In this center position, the conical element frac-tured and crushed the rock as the PDC cutters sheared the rock.4 The design team at Smith Bits recognized the potential for increased drilling effi-ciency using multiple Stinger elements to fail the rock through a combination of shearing and plow-ing actions. Bit design engineers used finite ele-ment analysis (FEA) to experiment with CDE cutter placement and to model the resulting changes in drilling performance.

The conical elements were placed at various positions across the bit face. This design process yielded a stronger overall cutting structure com-pared with that of conventional fixed cutter PDC bits. As they experimented with Stinger element placement across the bit face, the design engi-neers recognized the potential for improving design configurations and the benefits of using specific configurations to address specific drill-ing challenges (Figure 3).

Testing the HypothesisDesign engineers conducted a series of laboratory tests to evaluate Stinger conical diamond element

performance and durability. One test compared impact strength relative to that of a conventional polycrystalline diamond cutter element. Both ele-ments were dropped onto a hardened steel block with an impact force of 80,000 N [18,000 lbf]. This experiment simulated typical transitional drilling conditions when a PDC bit drilling at an ROP of 18 m/h [60 ft/h] exits soft shale and penetrates hard limestone. On first impact with the steel block, the sharp edge of the conventional PDC cutter was severely damaged (Figure 4). By con-trast, the conical element survived more than 100 impacts at 80,000 N without damage. Greater impact resistance of the CDE, which has a thicker diamond layer, translates into extended run lengths and improved penetration rates in impact-prone settings.

In a separate test, a vertical turret lathe was used to measure wear resistance. The conical ele-ment was lowered onto a rotating test bed of granite having a compressive strength of 207 MPa [30,000 psi]. After force was applied to the CDE, depth of cut and amount of wear were measured. Compared with a standard PDC cutter, the coni-cal element exhibited greater wear resistance and cutting efficiency. For example, under an applied force of 5,300 N [1,200 lbf], a 0.5-mm [0.02-in.] depth of cut by the CDE resulted in a 70% increase in cutting efficiency; at 1.3-mm [0.05-in.] depth of cut, the CDE cutter was 35% more efficient.5 The results also showed that the conical element dissipated frictional heat more efficiently than did conventional PDC cutters.

To investigate the conical element’s capabil-ity to induce rock failure, bit design engineers turned to FEA modeling, which allowed them to evaluate the Stinger element’s performance within the controlled environment of a virtual downhole setting. The FEA modeling demonstrated that the conical diamond element exerts concen-trated point loading to fail high–compressive strength formations. By creating a high-stress concentration at the contact point, the CDE increases fracture generation at the rock face while requiring significantly less applied force compared to that of standard PDC cutters.6

Through FEA modeling, engineers investi-gated the effects of conical elements on bit and BHA stability by comparing the forces sustained by conventional PDC cutters with those of CDE cutters. Among the most destructive products of those forces are lateral and axial vibration. In addition to damaging downhole equipment, these vibrations create undesirable drillstring harmon-ics and divert mechanical energy from the drill-ing system, resulting in lower ROPs. The modeling

Figure 2. Centrally placed conical diamond element (CDE). By removing cutting structures from the bit center (left ), space is created for the CDE (right ). During drilling, this space leaves a small column of rock, which is easily crushed by the CDE.

Figure 3. Variations in cutter placement. Depending on the application, cutter placement on the StingBlade bit may vary from single and double rows of CDE cutters (left ) to alternating rows of PDC and CDE cutters (right ).

September 2015 33

4. Bruton G, Crockett R, Taylor M, DenBoer D, Lund J, Fleming C, Ford R, Garcia G and White A: “PDC Bit Technology for the 21st Century,” Oilfield Review 26, no. 2 (Summer 2014): 48–57.

5. Azar M, White A, Velvaluri S, Beheiry K and Johny MM: “Middle East Hard/Abrasive Formation Challenge: Reducing PDC Cutter Volume at Bit Center Increases ROP/Drilling Efficiency,” paper SPE/IADC 166755, presented at the SPE/IADC Middle East Drilling Technology Conference and Exhibition, Dubai, October 7–9, 2013.

6. German V, Pak M and Azar M: “Conical Diamond Element Bit Sets New Performance Benchmarks Drilling Extremely Hard Carbonate/Chert Formations, Perm Region Russia,” paper SPE/IADC 173144, presented at the SPE/IADC Drilling Conference and Exhibition, London, March 17–19, 2015.

showed that the balanced profile of the conical diamond element subjects the cutter to less lat-eral force, which provides greater stability for longer bit runs while mitigating shock and vibra-tion effects to prolong the life of LWD and steer-ing components in the BHA (Figure 5).

The design process also led bit engineers to surmise that the plowing action of the conical ele-

ment might produce less torque than the shearing action of conventional PDC cutters. To confirm this hypothesis, the engineers subjected the bit to extensive testing, starting with FEA modeling, fol-lowed by evaluations in their rock mechanics laboratory. Next, downhole testing was carried out at a wellsite on the grounds of the Schlumberger Cameron Test and Training Facility in Texas, USA. This test compared the directional response of a StingBlade bit to that of a conven-tional PDC bit as each drilled a curve section through interbedded limestone, shale and sand-stone that had compressive strengths ranging from 69 to 103 MPa [10,000 to 15,000 psi]. The bit tests were conducted from identical kickoff points in adjacent wells on the same pad, using the same rig, motor type and directional driller. The StingBlade bit attained 23% higher build rates. It also exhibited better toolface control, requiring less intervention by the directional driller to stay

Figure 4. Impact testing. A technician prepares cutters for testing (left ). Still images from a motion picture indicate that the conventional PDC cutter (center, gray rounded element) failed on the first impact; the conical diamond element survived 100 impacts without damage (right ).

Figure 5. Stability as a function of resultant lateral force. The FEA modeling shows how resultant lateral force, applied by combining weight on bit with torque, is distributed at the cutter element. When applied to a conventional PDC cutter (left ), the forces (dashed orange lines) are spread along the leading edge of the cutter. The forces concentrate more symmetrically at the tip of the conical element (right ). Balancing this distribution of resultant lateral forces is key to reducing lateral shocks and vibrations induced at the drill bit.

Leadingedge

34 Oilfield Review

on target (Figure 6). The higher build rates deliv-ered by the StingBlade bit enabled it to land the curve 20 m [65 ft] sooner than the standard PDC bit.

Drilling the Curve in Variable LithologiesIn Lea County, New Mexico, USA, Cimarex Energy is targeting the Delaware basin Avalon shale

play.7 There, wells are typically drilled vertically to the Bone Spring Limestone, then kicked off with a bent sub and motor. The directional driller builds angle to 90° at 12°/100 ft [12°/30 m], to land the wellbore in the Avalon shale, after which the well is extended horizontally. The Avalon shale contains numerous stringers of interbed-

7. Hempton R, Copeland C, Cox G, Faught J, Blackmon W, Prewitt E, McDonough S and White A: “Innovative Conical Diamond Element Bits Drill Back-to-Back Tight Curves in One Run, Improving Economics While Reducing Risk in Avalon Shale Play, New Mexico,” paper SPE 175534, presented at the SPE Liquids-Rich Basins Conference—North America, Midland, Texas, September 2–3, 2015.

8. For more on the ONYX 360 rolling cutter: Bruton et al, reference 4.

ded carbonates and is characterized by uncon-fined compressive strengths ranging from 9,000 to 30,000 psi [62 to 207 MPa].

The highly variable lithology creates chal-lenges for directional drillers in the form of bit whirl and axial, lateral and torsional vibrations. These problems cause the bent motor assembly to deviate from its intended course, thus forcing the directional driller to reorient the toolface and adjust the trajectory to get back on target. Each toolface adjustment creates additional time not spent drilling in the desired direction, result-ing in a longer curve section and increased potential for missing the target.

In general, standard fixed cutter bits can be affected by variable formations, as evidenced by erratic toolface control and difficulty in drilling tight curves. Consequently, operators in this area typically rely on roller cone bits to drill the curve and have lately turned to a premium roller cone hybrid bit. These bits produce consistent torque responses for better steering control; however, they also drill at lower ROPs than do PDC bits.

Although the operator had success with the roller cone hybrid, the bit did not consistently drill the entire curve in a single run. A review of bit records for nine wells drilled by Cimarex within five miles of the target wellsite showed completion of the curved section using one bit in only 55% of the wells and an average ROP of 20.8 ft/h [6.34 m/h].

Based on bit performance and wear analysis in offset wells, Smith Bit engineers evaluated key areas along the bit face to determine where CDE placement would prove most effective. Using the IDEAS integrated drillbit design platform, they developed a fixed cutter bit having an alternating CDE and PDC cutter configuration. With this design, the conical diamond elements score the rock, creating two adjacent troughs. A PDC cut-ter, which trails behind the pair of CDEs, then shears away the unconfined rock ridge between the troughs (Figure 7). This arrangement requires lower force than is needed using tradi-tional PDC cutting structures, providing more efficient rock removal with less reactive torque.

Cimarex engineers selected an 83/4-in. StingBlade bit to drill the curve interval in its next two Avalon shale wells. Each bit drilled the curve in just one run with no significant toolface

Figure 6. Toolface angle. Changes in mud motor toolface orientation served as a gauge of downhole torque in field tests comparing PDC bit performance with that of a CDE bit. Despite changes in drilling parameters, such as weight on bit, the StingBlade bit experienced low fluctuations in reactive torque while building angle through formations of varying compressive strength.

3,025

3,0751800 90–90

Toolface angle, degree

PDC bit StingBlade bit

–180

Dept

h, ft

Figure 7. Alternating CDE and PDC cutting elements. Using FEA modeling of cutting action, bit engineers created a StingBlade bit design to plow, then shear the rock (left ). Stinger cutting elements create troughs separated by a small ridge (right). This ridge of stress-relieved rock is then easily sheared by the PDC cutter.

CDE1

CDE2PDC

September 2015 35

control problems. The configuration of the coni-cal diamond elements also helped protect the PDC cutters; when pulled from the hole, the bits were graded in very good condition (Figure 8). Furthermore, protection of the PDC cutters con-tributed to an improvement in ROP. Compared with bit performance from the previous nine wells, the StingBlade bits were able to complete the curve interval at an ROP that was 36% faster than the average roller cone hybrid one-run bit.

Broader HorizonsAdvances in bit design software, materials sci-ence and manufacturing enable bit engineers to not only test their ideas in the laboratory but to also see their designs come to fruition within days of conception. As a result, the variety of StingBlade bit designs is expanding rapidly to address a num-ber of challenges. Already, Stinger cutting ele-ments are being mounted on steel or composite bit bodies of various blade configurations, fre-

quently in conjunction with conventional PDC cutters or with ONYX 360 rolling cutters.8

Variations on the original design now include several types of StingBlade bits in a range of bit diameters (Figure 9). Although early StingBlade bit designs addressed specialized applications, its versatility is allowing the Stinger conical dia-mond element to quickly expand into more rou-tine applications. —MV

Figure 8. Conical diamond element bits after a full curve run. Bits pulled from wells are assessed using industry standard dull grading criteria. Increasingly, these assessments are supplemented with digital photographs. The first bit (left ) displayed slight chipping on one Stinger element in the trailing position on the nose of Blade 3 and on one PDC cutter on the nose of Blade 4 (circled). The bit pulled from a second well (right ) shows a delaminated and worn PDC cutter in the cone of Blade 3 and a chipped and worn CDE cutter on the shoulder of Blade 5 (circled).

1

2

34

4

5

6

6

5

3

2

1

Figure 9. StingBlade bit variations. Of the dozens of configurations designed for different drilling applications, five examples are shown. Designed for drilling hard carbonates with high concentrations of chert, this bit (A) uses Stinger elements to help support PDC cutter loading in applications that have potential for impact damage. Designed for highly interbedded formations, this bit (B) has alternating PDC and Stinger elements on the leading position of each blade to reduce torque variation and improve toolface control for curve intervals. A third variation (C), for hard, abrasive formations, uses Stinger elements to help support PDC cutter loading; ONYX 360 rolling cutters are strategically placed for wear resistance. Another design utilizes Stinger elements only (D); this bit is intended for granites or other extremely hard, abrasive igneous rocks. The Stinger cutting elements provide high concentrated point loading to fail the rock. Utilized in soft formations with hard stringers, the three-bladed bit (E) tends to drill faster than conventional five-bladed bits; the Stinger elements protect the PDC cutters from impact damage while the bit is transitioning through hard stringers.

A B C D E

Rolling cutterPDC

CDE

36 Oilfield Review

Carbon Dioxide—Challenges and Opportunities

In the early days of the oil and gas industry, companies noted that carbon dioxide had

corrosive effects on well internals; operators later found opportunities to use the

compound to their advantage. Projects now underway in the oil field reflect several

priorities—managing carbon dioxide’s corrosive effects, using it to recover more oil

after waterflood and storing it in underground formations. Because of its role in

climate change, carbon dioxide has emerged as a topic of significant public interest

and scientific investigation as well as the focus of hydrocarbon producers.

Mehdi AnsarizadehCalgary, Alberta, Canada

Kevin DoddsAustralian National Low Emissions Coal Research and DevelopmentCanberra, Australian Capital Territory, Australia

Omer GurpinarLawrence J. PekotDenver, Colorado, USA

Ülker KalfaSecaeddin S‚ahinSerkan UysalTurkish Petroleum CorporationAnkara, Turkey

T.S. RamakrishnanCambridge, Massachusetts, USA

Norm SacutaPetroleum Technical Research CentreRegina, Saskatchewan, Canada

Steve WhittakerCommonwealth Scientific and Industrial Research OrganisationPerth, Western Australia, Australia

Oilfield Review 27, no. 2 (September 2015).Copyright © 2015 Schlumberger.

Carbon dioxide is in the news. Whether because of the link to climate change and its consequences or for the concept of long-term storage, carbon diox-ide has captured the interest of the public and the global scientific community.1 The oil and gas indus-try has a long history of addressing the effects of

this compound, ranging from studies of carbon dioxide–methane hydrates in the 1940s to current studies on corrosion.2 Although anthropogenic—human-generated—carbon dioxide plays a nega-tive role in climate change, its role is positive in enhanced oil recovery (EOR).

Figure 1. Carbon dioxide phases. Phase boundary lines (blue) define the areas in which each CO2 phase exists. At the triple point, all three phases—solid, liquid and gaseous CO2—coexist in thermodynamic equilibrium. Along the solid-gas line below the triple point, CO2 sublimes—converts directly—from a solid to a gas without going through a liquid phase. The marked sublimation point corresponds to 0.101 MPa [14.7 psi] of CO2 vapor. Along the solid-liquid line above the triple point, solid CO2 melts to a liquid. Along the liquid-gas line above the triple point, liquid CO2 evaporates to a gas. At the critical point, the liquid and gaseous states of CO2 are indistinguishable, and phase boundaries no longer exist. These attributes at the critical point and at higher temperature and pressure characterize the area in which CO2 is a supercritical fluid (green). (Adapted with permission from Bassam Z. Shakhashiri, University of Wisconsin–Madison, USA.)

CO2 solid CO2 liquid

CO2 gas

Pres

sure

, MPa

Temperature, °C

Sublimationpoint

Triple point,–56.6°C, 0.519 MPa

Critical point,31.1°C, 7.37 MPa

CO2 supercriticalfluid

0.0001

0.001

0.01

0.1

1.0

10

–120–140 –100 –80 –60 –40 –20 0 20 40 60 80 100

100

Liquid and gas

Solid

and g

as

Solid

and

liquid


1. Zimmer C: “Ocean Life Faces Mass Extinction, Broad Study Says,” The New York Times (January 15, 2015), http://www.nytimes.com/2015/01/16/science/earth/study-raises-alarm-for-health-of-ocean-life.html (accessed January 15, 2015).

Fountain H: “Turning Carbon Dioxide Into Rock, and Burying It,” The New York Times (February 9, 2015), http://www.nytimes.com/2015/02/10/science/burying-a-mountain-of-co2.html (accessed June 1, 2015).

Cannell M, Filas J, Harries J, Jenkins G, Parry M, Rutter P, Sonneland L and Walker J: “Global Warming and the E&P Industry,” Oilfield Review 13, no. 3 (Autumn 2001): 44–59.

2. Unruh CH and Katz DL: “Gas Hydrates of Carbon Dioxide–Methane Mixtures,” Journal of Petroleum Technology 1, no. 4 (April 1949): 83–86.

Choi Y-S, Young D, Nešic S and Gray LGS: “Wellbore Integrity and Corrosion of Carbon Steel in CO2 Geologic Storage Environments: A Literature Review,” International Journal of Greenhouse Gas Control, 16S (January 2013): S70–S77.

3. “Global Ecology—Understanding the Global Carbon Cycle,” Woods Hole Research Center, http://whrc.org/global/carbon/ (accessed January 15, 2015).

Falkowski P, Scholes RJ, Boyle E, Canadell J, Canfield D, Elser J, Gruber N, Hibbard K, Högberg P, Linder S, Mackenzie FT, Moore B III, Pedersen T, Rosenthal Y, Seitzinger S, Smetacek V and Steffen W: “The Global Carbon Cycle: A Test of Our Knowledge of Earth as a System,” Science 290, no. 5490 (October 13, 2000): 291–296.

4. Riebeck H: “The Carbon Cycle,” NASA Earth Observatory, http://earthobservatory.nasa.gov/Features/CarbonCycle/ (accessed January 15, 2015).

5. Riebeck, reference 4.

September 2015 3737

Carbon, one of the two constituents of carbon dioxide [CO2], is an essential element on Earth. The mass of carbon on Earth is 5.37 × 1016 kg [11.83 × 1016 lbm], which is distributed among several reservoirs: the Earth’s atmosphere; plants; animals; soil; minerals; the shallow and deep ocean; and coal, oil and gas.3 The movement of carbon between these reservoirs—the carbon cycle—maintains a balance between carbon in the atmosphere and in the ocean and rocks.4 This cycle has two components: a slow cycle that takes 100 to 200 million years to move carbon between the oceans, soil, rock and the atmosphere and a fast cycle that takes 50 to 100 years to move car-bon through the biosphere.

Historically, the carbon cycles have resulted in a nearly constant level of carbon in the atmo-sphere, but that is changing. Current data point to deforestation and combustion of fossil fuels as

prime causes for changes in the fast carbon cycle.5 Plants, trees and microscopic marine plants are important components of the fast car-bon cycle. During decay, burning and consump-tion of these life forms, carbon, present as CO2, is released and accrues in the atmosphere. Similarly, much of the CO2 from anthropogenic activities also accumulates in the atmosphere. Plants and the oceans absorb about 55% of this anthropogenic CO2, but the rest stays airborne. Scientists attribute persistent changes in the composition of the atmosphere, such as the increasing CO2 content, to be an important driver of climate change.

The role of CO2 in climate change is signifi-cant, but the compound has a different function in the oil and gas industry. Carbon dioxide can be captured and stored in depleted reservoirs,

which helps arrest atmospheric accumulation. For EOR, CO2 enables increased yield from oil fields after primary recovery and waterflood. This article discusses and illustrates these aspects of CO2, its effect on climate change and its role in the oil and gas industry. Examples from oil and gas fields in Canada, Algeria and Turkey demon-strate the use and storage of carbon dioxide.

Carbon Dioxide CharacteristicsCarbon dioxide, a molecule that consists of two oxygen atoms covalently bonded to a single carbon atom, has a molecular weight of about 44 g/mol. Depending on temperature and pressure, CO2 can exist as a solid, liquid or gas (Figure 1). At tem-peratures and pressures at or above the critical point, CO2 is a supercritical fluid, which has some properties of a gas and some of a liquid. As a super-critical fluid, CO2 develops miscibility—the ability

Oilfield Review SPRING 15CO2 Fig OpenerORSPRNG 15 CO2 Opener

PhotosynthesisPhotosynthesis

Plantrespiration

Plantrespiration

Anthropogeniccarbon

Anthropogeniccarbon

Soil carbonSoil carbon

Ocean sediment carbonOcean sediment carbon

DecompositionDecomposition DecompositionDecompositionPhotosynthesisPhotosynthesis

CO2CO2

CO2CO2

CO2CO2

CO2CO2 CO2CO2O2COCCOCOCO2OO22

Year

Tem

pera

ture

cha

nge,

°F

1890

0

1

2

1920 1950 1980 2010

38 Oilfield Review

to mix homogeneously—with crude oil and improves oil recovery.

Carbon dioxide is stable in the atmosphere. Its concentration in the atmosphere depends on competing processes within the carbon cycles that consume or release CO2. Photosynthesis is one chemical reaction that involves CO2. During photosynthesis, plants, algae, ocean plankton and certain types of bacteria use light energy to convert CO2 and water to oxygen, carbohydrates and water.6 Each year, more than 10% of the atmo-spheric CO2 is reduced to carbohydrates by pho-tosynthesis. Plants, algae and plankton use carbohydrates for growth whereas animals, including humans, use it as an energy source.

Carbon dioxide may be produced in several ways. Natural sources of CO2 production include plant and animal respiration and decay, fires and volcanic release. Anthropogenic sources include fossil fuel combustion and certain manufacturing activities, including cement and ammonia produc-tion, natural gas processing and petrochemical manufacturing. Humans release CO2 indirectly through deforestation.7

Carbon dioxide may undergo several reac-tions of interest in the oil field. For example, dis-solved in water, CO2 forms carbonic acid [H2CO3] and other H2CO3 analogs.8 The CO2 may also react with the minerals of the reservoir; in carbonate reservoirs, the reaction can be relatively rapid

while in silicate reservoirs, the reactions are gen-erally much slower. These reactions may result in some of the CO2 being mineralized and perma-nently trapped.9

Another important set of reactions involving CO2 is associated with corrosion. Carbon dioxide may be corrosive or noncorrosive depending on the materials employed, temperature at the con-tact surface, water vapor concentration and CO2 partial pressure. The most likely metal to corrode is carbon steel in storage environments and cas-ing and tubular steel in wells. At a moderate pres-sure of 1.0 MPa [145 psi], the corrosion rate of X65 pipeline steel is independent of temperature from 50°C to 130°C [120°F to 270°F].10 Increasing water concentration, on the other hand, causes a significant increase in corrosion for steel. For example, at a pressure of 8 MPa [1,160 psi] and a temperature of 40°C [104°F], increasing the water concentration in supercritical CO2 from 1,000 to 10,000 parts per million (ppm) causes the corrosion rate of steel to increase by 87%.11 Similarly, for carbon steel in aqueous CO2 solu-tions at 25°C [77°F], increasing the CO2 partial pressure from 0.1 MPa [14.5 psi] to 1 MPa pro-duces a corrosion rate increase of about 450%.12

Another potential area of concern for oilfield operators is the effect of CO2 on cement in wells.13 Carbon dioxide saturated with water deteriorates the cement used in wells. This dete-rioration can occur in cement that is adjacent to the well casing either in the annulus between the casing and rock or at the interface between the casing and a cement well plug (Figure 2).14 Therefore, the cement used in CO2 injection wells must be able to resist the damaging effects of CO2 because operational periods can last from 25 to 100 years and mandated safety periods that last much longer. For wells to reach these time objectives intact, using additives that make the cement more resistant to harm from CO2 may be advantageous. Reaction of CO2 with wellbore cement is slow in a well in which good construc-tion practices and appropriate materials were used; in these cases, CO2 should not pose a prob-lem. Many old, abandoned wells—completed and shut in using practices and cement accept-able at the time—are not suitable to use in long-term CO2 storage systems. Leakage from abandoned wells has been identified as a signifi-cant risk in geologic storage of CO2.

Greenhouse Gas Effect and Climate ChangeThe greenhouse gas (GHG) effect is the process by which atmospheric insulation, imparted by certain gases, keeps the Earth warmer than it

Figure 2. CO2 migration. Wells offer several potential pathways for gas migration. The pathways consist of the following: between the casing and the annular cement (A), between the cement plug and casing (B), through the cement plug and annular cement pore space as a result of deterioration (C), through the casing as a result of corrosion (D), through fractures in the annular cement (E) and between the annular cement and rock (F). (Adapted with permission from Michael Celia, Princeton University, New Jersey, USA.)


Rock formationAnnular cementWell casing

Cement well plug

A

B

C

D

E

F

September 2015 39

would be without them (Figure 3).15 Although the concept of climate change associated with green-house gases may seem recent, the idea of the GHG effect dates back to the 19th century.16 Scientists then were intrigued by the possibility that lower levels of CO2 might explain the ice ages. In 1896, Swedish scientist Svante Arrhenius calculated that industrial emissions from coal combustion might someday cause an increase in the Earth’s surface temperature.17 More recently, in the 1960s and 1970s, Russia considered ways to warm its large areas of icy tundra and convert them to fertile farmland through human-engi-neered climate change. These and subsequent attempts to alter the climate are encapsulated by the term climate engineering, also known as geo-engineering.18 Geoengineering is the intentional manipulation of planetary-scale processes to affect Earth’s climate system—for example, to cool down the Earth’s atmosphere or remove CO2 from the Earth’s atmosphere.

The GHG effect comprises a natural and an enhanced component. Warming of the Earth’s sur-face associated with indigenous gases is the natu-ral GHG effect. The main GHGs are, in order of abundance, water vapor [H2O], CO2, methane [CH4], nitrous oxide [N2O], ozone [O3] and other minor components. These gases in the atmosphere heat the Earth’s surface by absorbing and reradiat-ing some of the infrared radiation coming from the surface. In addition to the natural GHG effect, an enhanced effect occurs when human activities increase the level of greenhouse gases—primarily CO2 but also CH4, N2O, sulfur hexafluoride [SF6]

and other fluorinated hydrocarbons.19 Increased concentrations of these gases add to the atmo-sphere’s insulating qualities, thereby increasing Earth’s surface temperatures.

In 2012, CO2 accounted for 82% of all GHG emissions in the US.20 Electrical power genera-tion was responsible for 32% of the CO2 emitted,

and transportation contributed 28%. Industry accounted for 20% and the remaining 20% was attributable to emissions from residences, com-mercial buildings and agricultural activity. The end result of the enhanced GHG effect and increased concentrations of CO2 is that the Earth’s surface is getting warmer.

6. Whitmarsh J and Govindjee: “The Photosynthetic Process,” in Singhal GS, Renger G, Sopory SK, Irrgang K-D and Govindjee (eds): Concepts in Photobiology: Photosynthesis and Photomorphogenesis. Dordrecht, The Netherlands: Springer Science+Business (1999): 11–51.

7. “Overview of Greenhouse Gases: Carbon Dioxide Emissions,” US Environmental Protection Agency, http://www.epa.gov/climatechange/ghgemissions/gases/co2.html (accessed January 10, 2015).

8. When CO2 dissolves in water, it forms H2CO3* and H2CO3°. The former is the concentration of H2CO3 present and is typically about 0.3% of the CO2. The latter is another state entirely and could be called liquid CO2. For more on H2CO3 states: Langmuir D: Aqueous Environmental Geochemistry. Upper Saddle River, New Jersey, USA: Prentice Hall, 1997.

9. Cardoso SSS and Andres JTH: “Geochemistry of Silicate-Rich Rocks Can Curtail Spreading of Carbon Dioxide in Subsurface Aquifers,” Nature Communications 5, article 5743 (December 11, 2014).

10. Sim S, Cole IS, Choi Y-S and Birbilis N: “A Review of the Protection Strategies Against Internal Corrosion for the Safe Transport of Supercritical CO2 via Steel Pipelines for CCS Purposes,” International Journal of Greenhouse Gas Control 29 (October 2014): 185–199.

11. Sim S, Bocher F, Cole IS, Chen X-B and Birbilis N: “Investigating the Effect of Water Content in Supercritical CO2 as Relevant to the Corrosion of Carbon Capture and Storage Pipelines,” Corrosion 70, no. 2 (February 2014): 185–195.

Figure 3. Greenhouse gas effect. Most solar radiation (left ) is absorbed by and warms the Earth’s surface while some radiation is reflected by the Earth and the atmosphere back into space. Some solar radiation that reaches Earth’s surface is emitted as infrared radiation (right ), some of which passes directly back through the Earth’s atmosphere into space. Greenhouse gas molecules absorb and reemit infrared radiation in all directions, including back toward the Earth’s surface. The net effect is a warming of the Earth’s surface and lower atmosphere. (Adapted from the US EPA, reference 15.)


Solarradiation

Earth’s surface

Atmosphere Infraredradiation

12. DeBerry DW and Clark WS: “Corrosion Due to Use of CO2 for Enhanced Oil Recovery,” US Department of Energy, Report DOE/MC/08442-1, September 1979.

13. Rutqvist J: “The Geomechanics of CO2 Storage in Deep Sedimentary Formations,” Geotechnical and Geological Engineering 30, no. 3 (June 2012): 525–551.

14. Ramakrishnan TS: “Climate Initiative and CO2 Sequestration,” presented at the Fourth Annual Conference on Carbon Capture and Sequestration, Alexandria, Virginia, USA, May 2–5, 2005.

Gasda SE, Bachu S and Celia MA: “Spatial Characterization of the Location of Potentially Leaky Wells Penetrating a Deep Saline Aquifer in a Mature Sedimentary Basin,” Environmental Geology 46, no. 6–7 (October 2004): 707–720.

Celia MA, Bachu S, Nordbotten J, Gasda S and Kavetski D: “Implications of Abandoned Wells for Site Selection,” presented at the International Symposium on Site Characterization for CO2 Geological Storage, Berkeley, California, USA, March 20–22, 2006.

Ide ST, Friedmann SJ and Herzog HJ: “CO2 Leakage Through Existing Wells: Current Technology and Regulations,” in Proceedings of the 8th International Conference on Greenhouse Gas Technologies. Kidlington, Oxford, England: Elsevier Ltd. (2006): 2531–2536.

15. Cannell et al, reference 1. US Environmental Protection Agency (US EPA): “Climate

Change Indicators in the United States, 2nd ed.,” Washington, DC: US EPA, Report 430-R-12-004, December 2012.

16. Weart SR: The Discovery of Global Warming. Cambridge, Massachusetts, USA: Harvard University Press, 2008.

17. Svante Arrhenius was a recipient of the 1903 Nobel Prize in chemistry for his electrolytic theory of dissociation. He also proposed what came to be known as the Arrhenius equation, which shows the temperature dependence of reaction rate constants.

18. Keith DW: “Geoengineering the Climate: History and Prospect,” Annual Review of Energy and the Environment 25 (November 2000): 245–284.

Caldeira K, Bala G and Cao L: “The Science of Geoengineering,” Annual Review of Earth and Planetary Sciences 41 (May 2013): 231–256.

19. Sulfur hexafluoride [SF6] is used for high-density plasma etching and as a dielectric. According to the Intergovernmental Panel on Climate Change, SF6 is the most potent greenhouse gas and has a global warming potential of 23,900 times that of CO2 based on a 100-year timeframe. For more on the warming potential of SF6 in relation to the other greenhouse gases: “Direct Global Warming Potentials,” Intergovernmental Panel on Climate Change, http://www.ipcc.ch/publications_ and_data/ar4/wg1/en/ch2s2-10-2.html (accessed May 10, 2015).

20. “Overview of Greenhouse Gases: Carbon Dioxide Emissions,” reference 7.

40 Oilfield Review

Global surface temperatures measured since 1900 have risen 0.79°C [1.4°F].21 During the same period, the CO2 level in the atmosphere rose from 296 ppm in 1900 to 392 ppm in 2010.22

The CO2 level in 1900 was typical of levels for the 400,000 years before 1950, in which the CO2 level was never above 300 ppm (Figure 4).23 The effect of rising CO2 on the Earth’s surface tem-

21. Five independent government agencies or their predecessors have measured surface temperatures annually since 1900. These agencies are the US National Aeronautics and Space Administration Goddard Institute for Space Studies, the US National Oceanic and Atmospheric Administration National Centers for Environmental Information, the Met Office Hadley Centre and the University of East Anglia Climatic Research Unit in England and the Japan Meteorological Agency in Japan. Data from all these agencies show nearly identical long-term trends and variations. “Despite Subtle Differences, Global Temperature Records in Close Agreement,” NASA Goddard Institute for Space Studies (January 13, 2011), http://www.giss.nasa.gov/research/news/20110113/ (accessed May 17, 2015).

22. US EPA Office of Air and Radiation: “Climate Change Science Facts,” Washington, DC: US EPA, Report 430-F-10-002, April 2010.

23. “Global Temperature,” NASA Global Climate Change: Vital Signs of the Planet, http://climate.nasa.gov/vital-signs/global-temperature/ (accessed May 29, 2015).

“Trends in Atmospheric Carbon Dioxide: Recent Monthly Average Mauna Loa CO2,” NOAA Earth System Research Laboratory Global Monitoring Division, http://www.esrl.noaa.gov/gmd/ccgg/trends/ (accessed May 17, 2015).

24. Although surface temperatures appear to be directly correlated with CO2 levels, this has not been proven from first principles. This correlation is based on evidence-based fact.

25. Archer D, Eby M, Brovkin V, Ridgwell A, Cao L, Mikolajewicz U, Caldeira K, Matsumoto K, Munhoven G, Montenegro A and Tokos K: “Atmospheric Lifetime of Fossil Fuel Carbon Dioxide,” Annual Review of Earth and Planetary Sciences 37 (2009): 117–134.

Doney SC, Fabry VJ, Feely RA and Kleypas JA:

Figure 4. Increase in CO2 in the atmosphere. The levels of atmospheric CO2 can be measured from the distant past by analyzing ancient air bubbles trapped in polar ice. Data from 400,000 years before 1950 and to the present show a cyclic pattern dipping lower than 200 ppm during cold cycles and rising to nearly 300 ppm during warmer periods. Starting in the mid-1950s (inset ), atmospheric CO2 level rose above 300 ppm and continues to rise to the current level—slightly more than 400 ppm. (Adapted from “Global Temperature” and “Trends in Atmospheric Carbon Dioxide: Recent Monthly Average Mauna Loa CO2,” reference 23.)


–50,000–100,000–200,000–250,000–300,000–350,000

Years before 1950–400,000

200

250

350

450

300

400

CO2, p

pm

–150,000 0

400

1960 1970Year

CO2, p

pm

1980 1990 2000 2010 2015

320

340

360

380

peratures is well documented.24 Less well docu-mented is the effect of increased atmospheric CO2 on global oceans.

Researchers believe that the effects of increased CO2 can be observed in global oceans.25 Oceans absorb about one-third of the CO2 added to the atmosphere. Ocean absorption of CO2 is not benign—CO2 has caused a significant increase in ocean acidification. The weather sta-tion at Mauna Loa, Hawaii, USA, measured atmo-spheric CO2 and ocean pH from 1990 to 2010. During that 20-year period, atmospheric CO2 rose from 352 ppm in 1990 to 387 ppm in 2010. Concomitant with this rise in CO2, the ocean pH decreased from 8.12 to 8.08, indicating an increase in ocean acidification.26

Acidification of oceans and the warming effects from climate change are leading to the extinction of some types of ocean animal life.27 Coral reefs are sensitive to both acidification and warming. The net effect of acidification is an increase in the hydronium ion [H3O+] concentra-tion and a corresponding decrease in the carbon-ate ion [CO3

–2], which results in less coral being formed than in healthy ocean waters.28 Coral reefs are also sensitive to increases in temperature as small as 1°C to 2°C [1.8°F to 3.6°F] over times that are too short for the corals to adapt. These changes in temperature affect many of the microscopic and higher marine life forms that live in a symbiotic relationship with coral.

“Ocean Acidification: The Other CO2 Problem,” Annual Review of Marine Science 1 (January 2009): 169–192.

26. Since pH is on a logarithmic scale, this 0.5% decrease in pH represents a 9.6% increase in the acidity as measured by the hydronium [H3O+] ion concentration.

27. Zimmer, reference 1. McCauley DJ, Pinsky ML, Palumbi SR, Estes JA,

Joyce FH and Warner RR: “Marine Defaunation: Animal Loss in the Global Ocean,” Science 347, no. 6219 (January 16, 2015): 1255641-1–1255641-7.

Urban MC: “Accelerating Extinction Risk from Climate Change,” Science 348, no. 6234 (May 1, 2015): 571–573.

28. Coral reefs are composed of calcium carbonate [CaCO3] and small amounts of other minerals. Coral formation is dependent on the concentrations of the calcium ion and the carbonate ion—a decrease in the carbonate ion means less coral is formed.

29. Lake LW, Schmidt RL and Venuto PB: “A Niche for Enhanced Oil Recovery in the 1990s,” Oilfield Review 4, no. 1 (January 1992): 55–61.

Al-Mjeni R, Arora S, Cherukupalli P, van Wunnik J, Edwards J, Felber BJ, Gurpinar O, Hirasaki G, Miller CA, Jackson C, Kristensen MR, Lim F and Ramamoorthy R: “Has the Time Come for EOR?,” Oilfield Review 22, no. 4 (Winter 2010/2011): 16–35.

30. International Energy Agency Greenhouse Gas R&D Programme (IEA GHG): “CO2 Storage in Depleted Oilfields: Global Application Criteria for Carbon Dioxide Enhanced Oil Recovery,” Stoke Orchard, Cheltenham, England: IEA GHG, Technical Report 2009-12, December 2009.

31. The 10 basins in the world that have oil recoverable using CO2 are the following: In the Middle East, the Mesopotamian Foredeep, Greater Ghawar Uplift, Zagros Fold Belt and the Rub Al Khali basins; in Russia, the

West Siberian and Volga Ural basins; in South America and Mexico, the Maracaibo and Villahermosa Uplift basins; in the US, the Permian basin; and in Europe, the North Sea Graben basin.

32. In miscible conditions, two or more fluids mix in all proportions and form a single homogeneous phase. In immiscible conditions, two fluids are incapable of forming molecularly distributed mixtures or attaining homogeneity.

For more on the evolution of CO2 flooding: Holm LW: “Evolution of the Carbon Dioxide Flooding Processes,” Journal of Petroleum Technology 39, no. 11 (November 1987): 1337–1342.

33. Holm LW and O’Brien LJ: “Carbon Dioxide Test at the Mead-Strawn Field,” Journal of Petroleum Technology 23, no. 4 (April 1971): 431–442.

Hill B, Hovorka S and Melzer S: “Geologic Carbon Storage Through Enhanced Oil Recovery,” Energy Procedia 37 (2013): 6808–6830.

34. Martin FD and Taber JJ: “Carbon Dioxide Flooding,” Journal of Petroleum Technology 44, no. 4 (April 1992): 396–400.

35. Mohan H, Carolus M and Biglarbigi K: “The Potential for Additional Carbon Dioxide Flooding Projects in the United States,” paper SPE 113975, presented at the SPE/DOE Improved Oil Recovery Symposium, Tulsa, April 19–23, 2008.

36. IEA GHG, reference 30.37. Mohan et al, reference 35.38. Benson S: “Status and Opportunities in CO2 Capture,

Storage and Utilization,” presented at the American Physical Society Workshop on Energy Research and Applications for Physics Students and Postdocs, San Antonio, Texas, USA, March 1, 2015.

September 2015 41

Carbon Dioxide FloodingThe oil and gas industry injects CO2 into oil and gas fields for two primary purposes—rejuvenating producing fields and storing in depleted or unused reservoirs; these processes contribute to the global effort to minimize climate change. Carbon dioxide can be used in EOR to recover additional oil following primary production and waterflood.29 In addition, CO2 can be captured from a variety of sources and stored underground.

The amount of technically recoverable oil worldwide has been estimated at 450 billion bbl [72 billion m3].30 Oil that could theoretically be recovered using CO2 is concentrated in ten large basins worldwide—in the Middle East, Russia, South America, Mexico, the US and Europe.31 Oil recovery on this scale using CO2 EOR would require large sources of CO2 close to the reser-voir; proximity to a CO2 source is a challenge for most basins. Because the volume of potential recoverable oil worldwide is large, operators have been able to use a variety of EOR techniques to recover additional oil from reservoirs for decades, despite the difficulty of matching CO2 sources with sinks.

In the 1950s, researchers looking at CO2 flooding found that the compound was miscible in oil at pressures above 700 psi [5 MPa].32 Building on this and subsequent findings about CO2 behavior in oil, operators conducted the early successful field test of miscible CO2 flooding at the Mead-Strawn field near Abilene, Texas, USA, in 1964.33 Test results showed a 35% increase in incremental oil recovery using CO2 over the results of conventional waterflooding. Since that field test, many successful operations using mis-cible CO2 flooding have been conducted.

Carbon dioxide flooding for EOR can be grouped into two broad categories—miscible and immiscible. The process that is ultimately employed by the oilfield operator will depend on reservoir conditions and characteristics of the oil. Miscible CO2 flooding is the most common application although immiscible flooding may be applied in some situations because of oil density or reservoir pressure.34

Several factors make miscible CO2 flooding an effective method for additional oil recovery. Carbon dioxide is soluble in crude oils, swells net oil volume and reduces oil viscosity even before it achieves miscibility. As the point of complete mis-cibility is approached, the CO2 phase and the oil phase start to flow together homogeneously as a result of reduced interfacial tension and the increase in volume of the combined oil-solvent phase relative to the water phase.

At constant temperature, the lowest pressure at which liquids achieve miscibility is defined as the minimum miscibility pressure (MMP). Miscible CO2 flooding is applicable in many reser-voirs and is most effective when the reservoir has a pressure greater than the MMP. Typically this occurs at a depth greater than 760 m [2,500 ft].35 Additionally, the oil should have greater than 22 degree API gravity [less than 0.92 specific gravity] and less than 10 cP [10 mPa.s] viscosity. For best results, the reservoir needs to have oil saturation greater than 20% of the pore volume.

Ideally, a typical miscible flood injects CO2 at one end of the desired zone and recovers oil driven to producer wells (Figure 5).36 Although miscible floods account for the majority of CO2 EOR projects, some systems may benefit from immiscible flooding. Immiscible CO2 EOR proj-ects depend on a reduction in oil viscosity accompanied by oil swelling to achieve addi-tional oil recovery. Projects that would benefit from immiscible CO2 EOR have low-gravity crude oil and reservoir pressures less than the MMP.

A dependable source for CO2 is a prerequisite for both miscible and immiscible CO2 flooding. Natural and industrial sources of CO2 are avail-able.37 In 2008, the US produced about 3 bil-lion ft3/d [80 million m3/d] of CO2, primarily from natural sources in New Mexico, Colorado and Mississippi. Approximately 75% of the naturally produced CO2 in the US is sent by pipeline to the

Permian Basin, Texas, from the McElmo Dome, which is located near the border between Utah and Colorado and has one of the world’s largest accumulations of naturally occurring CO2.

Industrial sources of CO2 within the US include natural gas processing plants in Texas, Oklahoma, Wyoming and Michigan, an ammonia plant in Oklahoma, a coal gasification plant in North Dakota and power plants that have carbon capture capability. In Europe, significant indus-trial CO2 sources are located in the UK, the Netherlands, Belgium, France and Germany. However, none of these sources have yet been used for CO2 EOR operations.

Carbon Dioxide StorageMore than 80% of the world’s energy comes from the combustion of fossil fuels, and a rapid transi-tion to low-carbon energy sources will likely be difficult and expensive.38 One method of mitigat-ing the effects of CO2 on climate change is carbon capture and storage (CCS). Because about 7,400 industrial sources worldwide have CO2 emissions greater than 100 thousand metric tons/yr [110 thousand tonUS/yr], CCS and other strate-gies will be necessary over a 50-year period just to arrest the increase of CO2 in the atmosphere. To actually reduce CO2 will require an even greater effort (See “Taming Carbon Dioxide Emissions,” page 42).

Figure 5. Miscible CO2 flooding. Purchased and recycled CO2 are injected into a formation (top left ); water is also injected and acts as a driver. Some of the CO2 dissolves in the oil and is stored in the formation (bottom left ). The remainder of the CO2 causes vaporization of the lighter oil fractions into the CO2 phase (bottom center ) while the CO2 condenses into the oil phase. Driven by the water flood, the oil, and any residual CO2, reaches the production well (bottom right ) and both are pumped to the surface. At the surface (top right ), the oil and CO2 are separated, and the CO2 is recycled back to the injection point. (Adapted with permission from the IEA GHG, reference 30.)

Drivewater Water Miscible

zone

Immobile oil

Zone of efficient sweepIndustrial plant Producer

well

Immobile oil

Oilbank

OilrecoveryCO2

StoredCO2

DissolvedCO2

CO2


Injected CO2Purchased CO2 Recycled CO2

(continued on page 44)

42 Oilfield Review

Data about the state of the Earth suggest that CO2 must be brought under control to arrest the deleterious effects of climate change. Although not the only culprit, of all potential factors, CO2 carries the most weight in influ-encing undesired changes in the atmosphere, surface temperatures and oceans. A variety of proposals have been suggested to bring CO2 under control. Two such proposals illustrate the magnitude of the CO2 emissions challenge and show workable paths for halting emissions

growth and reducing the absolute level of those emissions over the next 15 to 35 years. Climate change caused by CO2 and other fac-tors can be arrested and reduced, but con-crete action must be taken now to accomplish the task.

Stabilization WedgesThe stabilization wedge concept, introduced in 2004 and refined in 2007, shows how CO2

emissions could be brought under control.1 A

plot of carbon emission rate versus time helps explain the concept and its application to con-trolling CO2 levels in the atmosphere (Figure S1).

To frame CO2 emissions stabilization and how it might be achieved, the stabilization tri-angle is divided into eight wedges. Each wedge represents the amount required from an area of focus, or major effort, to accomplish this objective. Carbon dioxide capture and storage is recognized as one area of focus (Figure S2). The strategies in each area can

Taming Carbon Dioxide Emissions

50 years

Year

Carb

on e

mis

sion

rate

, bill

ion

met

ric to

ns/y

r

19500

8

16

2000 2050 2100

Historicalemissions

Extrapolatedemissions path

Reduced emissions path

Constant emissions path

Eight carbon wedges

Double the fuel efficiency of 2 billion cars from 13 to 25 km/l[30 to 60 mi/galUS].

Decrease the distance traveled by hydrocarbon-fueled cars by half.

Use best energy efficiency practices in all residential andcommercial buildings.

Produce coal-based electricity that has twice that of today’s efficiency.

Replace 1,400 coal-fired electric plants with naturalgas–fired plants.

Capture and store emissions for 800 coal-based electric plants.

Capture the carbon from 180 coal-to-synfuel plants and store the captured CO2.

Double the current global nuclear capacity to replacecoal-based electricity.

Increase electricity generated by wind by 10 times the currentrate to be achieved by a total of 2 million windmills.

Increase electricity generated from solar radiation to 100 times the present capacity.

Use 40,000 km2 [15,000 mi2] of solar panels to producehydrogen for fuel cell cars.

Increase ethanol production from biomass by a factor of 12using farms that have an area equal to one-sixth of theworld’s croplands.

Eliminate tropical deforestation.

Adopt the practice of conservation tillage in all agriculturalsoils worldwide.

Produce hydrogen from coal at six times that of today’s rate and store captured CO2.

Efficiency

Fuel switching

Carbon captureand storage

Nuclear power

Wind power

Solar power

Biomass

Natural sinks

Carbon Wedge Strategy

Figure S1. Stabilization triangle and wedges. Historical carbon emission rates (red line) are shown up to 2007, at which time the emission rate was about 8 billion metric tons/yr [9 billion tonUS/yr]. If extrapolated along the current path (black dashed line) for 50 years, the emission rate could reach about 16 billion metric tons/yr [18 billion tonUS/yr], and the atmospheric CO2 level could be nearly 850 ppm in 2057. The triangle (green) formed by the constant emissions path (orange line)—defined here as a 50-year span—and the extrapolated emissions is called the stabilization triangle. The area of this triangle quantifies the amount of carbon that must be removed to stabilize atmospheric CO2 at close to current levels. The stabilization triangle can be divided into eight carbon wedges. Within each wedge, the carbon emission rate grows from zero to 1 billion metric tons/yr [1.1 billion tonUS/yr] of carbon after 50 years and, consequently, its area represents 25 billion metric tons [28 billion tonUS] of carbon emissions. Therefore, the area of the stabilization triangle represents 200 billion metric tons [220 billion tonUS] of carbon that will not be released to the atmosphere over the 50-year span. Once emissions have stabilized, industry and the public must begin to employ technologies that reduce emissions (right, blue line). (Adapted with permission from the Carbon Mitigation Initiative, Princeton University, New Jersey.)

Figure S2. Carbon stabilization wedges and strategies for lowering CO2 emissions to Earth’s atmosphere.

September 2015 43

1. Pacala S and Socolow R: “Stabilization Wedges: Solving the Climate Problem for the Next 50 Years with Current Technologies,” Science 305, no. 5686 (August 13, 2004): 968–972.

“Stabilization Wedges Introduction,” Princeton University Carbon Mitigation Initiative, http://cmi.princeton.edu/wedges/intro.php (accessed March 24, 2015).

2. International Energy Agency (IEA): “Energy and Climate Change: World Energy Outlook Special Report,” Paris: IEA, 2015.

3. IEA, reference 2.4. Energy efficiency is using less energy to provide the

same service.5. Renewable energy sources include solar, wind,

hydropower and biomass.

eventually reduce global carbon emissions by 1 billion metric tons/yr [1.1 billion tonUS/yr] by 2057.

Bridge ScenarioA major climate meeting, the 21st Conference of the Parties (COP21) to the United Nations Framework Convention on Climate Change, will take place in Paris in December 2015. In advance of the COP21 conference, countries have pledged to make intended nationally determined contributions (INDCs) for reduc-ing energy-related greenhouse gas (GHG) emissions toward the objective of slowing the pace of climate change.2 Under this INDC sce-nario, if countries adhere to their pledges, the growth of the carbon emission rate is pro-jected to slow down but not stop (Figure S3).

An alternative path to taming and eventu-ally reducing CO2 emissions has been devel-oped by the International Energy Agency (IEA).3 This concept, called the bridge sce-nario, seeks a more aggressive approach to battling carbon emissions than does the INDC scenario. Implicit in the bridge scenario is the recognition that global economic output and energy-related GHG emissions are indepen-dent phenomena. The bridge scenario calls for implementing five policy measures:• Increase energy efficiency.4

• Reduce the use of inefficient coal-fired power plants.

• Increase investment in renewable energy.5

• Phase out subsidies for fossil-fuel consumption.

• Reduce upstream methane emissions.Adoption of these measures is a start

toward achieving a maximum surface temper-ature rise of 2°C [3.6°F] from current levels. However, other measures will be required to achieve the goal.

If these five strategies are fully employed immediately worldwide, under the bridge sce-nario, the carbon emission rate will peak in 2018 followed by a steady reduction. When compared with the INDC scenario, the bridge scenario promotes a reduction of 1.3 billion metric tons/yr [1.4 billion tonUS] from the calculated 2030 carbon emission rate to the 2010 emission rate of 8.9 billion metric tons/yr [9.8 billion tonUS/yr].

Figure S3. Bridge scenario. Two bounding curves define the emissions reduction envelope of the bridge scenario. The upper curve (black) is the trend line for the carbon emission rate if the global community honors its INDC pledges. The bottom curve (red) represents the emission rate reduction possible under the bridge scenario. The largest contributor to reduced carbon emissions in 2030 is energy efficiency (light orange), which contributes a 49% reduction. Increased investments in renewable energy sources (green) provide a 17% reduction. Upstream methane reduction (light blue), fossil fuel subsidy reform (purple) and reducing inefficient coal use (brown) make up the remaining 34% reduction. (Adapted from the IEA, reference 2.)

Carb

on e

mis

sion

rate

, bill

ion

met

ric to

ns/y

r

Year

Energy efficiency increaseSubsidy reformRenewables investment

Inefficient coal use reductionUpstream methane reduction

INDC scenario

Bridge scenario

10.5

10.0

9.5

9.0

2015 2020 2025 2030

44 Oilfield Review

39. The potential for reuse in the oil and gas industry is limited. Most of the CO2 in the reuse category will go to CO2 flooding.

40. Benson SM: “Carbon Capture and Sequestration (CCS) 101,” presented at the Stanford University Global Climate and Energy Project Symposium, Stanford, California, September 28–29, 2010.

Alvi A, Berlin EH, Kirksey J, Black B, Larssen D, Carney M, Chabora E, Finley RJ, Leetaru HE, Marsteller S, McDonald S, Senel O and Smith V: “CO2 Sequestration—One Response to Emissions,” Oilfield Review 24, no. 4 (Winter 2012/2013): 36–48.

41. Benson, reference 38.42. Capillary action, or capillarity, is the phenomenon in

which surface tension draws fluid into the interstices of a material.

43. Zoback MD and Gorelick SM: “Earthquake Triggering and Large-Scale Geologic Storage of Carbon Dioxide,” Proceedings of the National Academy of Sciences 109, no. 26 (June 26, 2012): 10164–10168.

44. Hunter S: “The Tiltmeter: Tilting at Great Depths to Find Oil,” Lawrence Livermore National Laboratory Science and Technology Review (October 1997): 14–15.

Granda J, Arnaud A, Payàs B and Lecampion B: “Case Studies for Monitoring of CO2 Storage Sites, Based on Ground Deformation Monitoring with Radar Satellites,” paper C01, presented at the Third EAGE CO2 Geological Storage Workshop: Understanding the Behaviour of CO2 in Geologic Storage Reservoirs, Edinburgh, Scotland, March 26–27, 2012.

45. Benson, reference 38.46. Wright I, Ringrose P, Mathieson A and Eiken O: “An

Overview of Active Large-Scale CO2 Storage Projects,” paper SPE 127096, presented at the SPE International

Carbon capture and storage technologies can be applied to much of the 60% of the CO2 emis-sions that come from stationary sources such as power plants, cement plants and refineries. The remaining 40% is released to the atmosphere and comes from other stationary sources that emit CO2 such as residences, commercial buildings, small cement kilns, small steel plants and com-bustion of biomass.

Implementing CCS requires four sequential steps: CO2 capture, compression, transport via pipeline or marine transport and storage or reuse.39 For power plants, options for CO2 capture include precombustion, such as coal gasification, postcom-bustion and oxygen combustion. Each option has its advantages and disadvantages, and no single option fits every situation.40 The next step—con-version of CO2 to a liquid state—is a mature tech-nology and requires CO2 compression to 7.6 MPa [1,100 psi] or higher. Pipeline transport is also a mature technology; 3,000 mi [4,800 km] of pipe-lines are in place in the US alone. This pipeline network continues to slowly increase. Storage, the last step in CCS, is more complex.

Depleted oil and gas reservoirs and deep saline formations either onshore or offshore are options for geologic storage of CO2. To store CO2, the gas is injected into reservoirs that are at depths of 1 km [0.6 mi] or greater to ensure the CO2 remains in a dense liquid or supercritical fluid state. Several mechanisms trap and keep CO2 immobile.41 The primary trapping mecha-nism is usually a seal of low-permeability rock above the storage area, similar to that for natural oil and gas accumulations. Secondary mecha-nisms include solubility trapping, a mechanism by which a portion of the CO2 dissolves in water, and residual gas trapping, a mechanism by which the CO2 is trapped by capillarity.42 In some forma-tions, CO2 can be eventually trapped by its reac-tion with the rock and conversion to solid minerals. These secondary trapping mechanisms tend to become more effective over time, yield-ing, in most cases, a more secure storage site.

An ideal storage site is close to stationary sources of CO2, has the capacity to contain the projected volume of material over a long period of time, is able to sustain a high injection rate and has a low-permeability barrier to act as a caprock or seal. In addition, the storage site must be at an appropriate depth for CO2 to be liquid and have good mechanical strength to withstand injection pressures.

The injection of high volumes of fluid under high pressure into fault zones near a CCS site may create problems; doing so may cause faults to slip and generate microseismic activity.43 Passive seis-

Figure 6. Tiltmeter. A tiltmeter, which measures surface deformation, is similar to a sensitive carpenter’s level. The tiltmeter can measure a tilt of about 1 × 10−9 radians [57 × 10−9 degrees], which is equivalent to lifting one end of a 4,000 km [2,500 mi] long beam just 0.64 cm [0.25 in.] from level. To measure deformation of this size, the tiltmeter uses a sensor that has a glass case that contains a gas bubble, conductive liquid and several electrodes (inset ). When the tiltmeter case tilts to one side, the gas bubble changes position, and the resistance between the electrodes changes. This resistance change is calibrated to give the degree of deformation. (Photograph courtesy of Steven Hunter, Lawrence Livermore National Laboratories, California, USA.)


CaseElectrode

Liquid conductor

Gas bubble

Conference on CO2 Capture, Storage and Utilization, San Diego, California, November 2–4, 2009.

For more on the Sleipner project: Bennaceur K, Gupta N, Monea M, Ramakrishman TS, Randen T, Sakurai S and Whittaker S: “CO2 Capture and Storage—A Solution Within,” Oilfield Review 16, no. 3 (Autumn 2004): 44–61.

47. Fairley P: “A Coal Plant That Buries Its Greenhouse Gases,” MIT Technology Review 118, no. 1 (January/February 2015): 84–87.

48. IEA GHG, reference 30.49. Barnhart WD and Coulthard C: “Weyburn CO2 Miscible

Flood Conceptual Design and Risk Assessment,” paper 95-120, presented at the Sixth Petroleum Conference of the South Saskatchewan Section of the Society of Canadian Institute of Mining, Metallurgy and Petroleum, Regina, Saskatchewan, Canada, October 16–18, 1995.

Brown K, Jazrawi W, Moberg R and Wilson M: “Role of Enhanced Oil Recovery in Carbon Sequestration—The Weyburn Monitoring Project, a Case Study,” presented at the First National Conference on Carbon Sequestration, Washington, DC, May 14–17, 2001.

50. Protti G: “Win-Win: Enhanced Oil Recovery and CO2 Storage at EnCana’s Weyburn Oilfield,” paper WPC-18-0986, presented at the 18th World Petroleum Congress, Johannesburg, South Africa, September 25–29, 2005.

Bennaceur et al, reference 46.51. For more on the Great Plains Synfuels Plant: Dakota

Gasification Company, http://www.dakotagas.com/index.html (accessed May 12, 2015).

52. Fairley, reference 47.

September 2015 45


Project area

Williston basin

S a s k a t c h e w a n

N

Weyburn

Regina

U N I T E D S T A T E S

C A N A D A


N o r t h D a k o t a

M a n i t o b aS a s k a t c h e w a n

Bismarck

Estevan

Weyburn

Beulah

0 km

0 mi 100

100

Oil field

CO2 pipeline tap

CO2 pipeline

mic methods may be used to monitor and record any injection-induced seismicity. Increased forma-tion pressure may cause some degree of uplift, which can be monitored using interferometric syn-thetic aperture radar (InSAR) satellite imagery or tiltmeters (Figure 6).44

The practice of carbon capture and storage continues to expand worldwide. Eight operating industrial-scale projects have come on stream in the past 40 years. These projects represent storage of about 14 million metric tons [15 million tonUS] of CO2 annually.45 Eight additional projects are under construction, representing 13 million metric tons [14 million tonUS] of CO2 annual storage capacity.

At the Sleipner project in the North Sea, CO2 is produced with natural gas, separated offshore and then injected into a disposal interval. About 1 million metric tons [1.1 million tonUS] of the produced CO2 has been injected annually into the Utsira formation over a 15-year period.46 The Boundary Dam Power Plant CCS project is located at Estevan, Saskatchewan, Canada. SaskPower, the owner and operator of the proj-ect, invested more than US$ 1 billion to equip one of its generators for carbon capture. SaskPower sells the captured CO2 to Cenovus Energy Inc. who uses it for EOR to boost output from matur-ing wells nearby.47 SaskPower also operates its own injection well at the power plant site.

Although CCS is a major initiative in several oil fields, CO2 is also being used for EOR in many production environments. A project in Canada is a good example of combined EOR and CCS.

Miscible CO2 Flooding and StorageCenovus Energy Inc. has embraced a long-term commitment to use CO2 for miscible flooding and to store excess amounts underground at the Weyburn-Midale field in Canada. This project is at the forefront of combined CO2-EOR and geo-logic CO2 storage.48

The Weyburn-Midale field, discovered in 1954, is located in southeast Saskatchewan (Figure 7).49 The operation covers 180 km2 [70 mi2] and is one of the largest medium-sour oil reservoirs in Canada. Original oil in place (OOIP) was estimated at 1.4 bil-lion bbl [220 million m3]. Following initial produc-tion over a 9- to 10-year period, the operator started waterflooding in 1964 followed by horizontal drilling in the 1990s. Although these measures helped pro-duction, the operator opted to use CO2 EOR to reverse the long-term production decline and to demonstrate large-scale geologic storage of CO2.50

The Dakota Gasification Company operates a synfuel plant in Beulah, North Dakota, that gener-ates natural gas from coal.51 The byproduct CO2

produced at the Beulah plant is compressed to 2,200 psi [15 MPa] and transported 210 mi [340 km] via pipeline to the Weyburn-Midale field. Deliveries of CO2 from the Beulah plant vary, rang-ing from 6,000 to 8,500 metric tons/d [6,600 to 9,400 tonUS/d]. The nearby Boundary Dam CCS project supplies an additional 2,300 metric tons/d

[2,500 tonUS/d] to the Weyburn-Midale field.52 Storage of CO2 was initiated in September 2000 in a limited area of the field. This early phase of the operation had 16 vertical and 13 horizontal injec-tion wells. A study of this injection area is ongoing and will address all of the technical aspects of long-term geologic storage (Figure 8).

Figure 7. Weyburn-Midale project. Located near Weyburn, Saskatchewan, Canada, this project uses CO2 for enhanced oil recovery and stores it in underground formations. The Weyburn-Midale reservoir is located principally in North Dakota, USA, and Saskatchewan; the fields extend west into Montana, USA, and east into Manitoba, Canada. Most of the CO2 used at the Weyburn-Midale project is piped (red) from a coal gasification plant near Beulah, North Dakota; the remainder comes from the nearby Boundary Dam Project at Estevan, Saskatchewan.

Figure 8. Weyburn-Midale research project. The Petroleum Technology Research Centre (PTRC) conducted research at the Weyburn-Midale project. The project area encompasses a 100,000-km3 [24,000-mi3] volume (dashed lines) and is part of the Williston basin (red line). The oil reservoirs and the levels above and below them in the area earmarked for CO2 storage were characterized before the initial injection. During the injection of CO2, measurements were made in a smaller area (gray rectangle) of the Weyburn-Midale field. In this field, PTRC scientists carried out an array of monitoring and verification research that included soil, gas and water sampling, subsurface monitoring, seismic monitoring, sampling from wells and risk assessment; these studies are ongoing. (Adapted with permission from PTRC.)

46 Oilfield Review

Miscible CO2 EOR and geologic storage have been successful at the Weyburn-Midale field, giv-ing the field new life and potentially extending its operational period by more than 25 years. Currently, the field produces about 26,000 bbl/d [4,100 m3/d] of light crude oil (Figure 9). Carbon dioxide injection has tripled oil production from the estimated lowest production rate for the field, about 8,000 bbl/d [1,200 m3/d] in 1988.

To date, about 24 million metric tons [26 mil-lion tonUS] of CO2 have been stored, and about 55 million metric tons [61 million tonUS] will be stored underground over the life of the project.

Carbon Dioxide Storage at In SalahThe In Salah Gas (ISG) project, a joint venture between Sonatrach, BP and Statoil, is currently executing a phased development of eight gas fields in the Ahnet-Timimoun basin in the Algerian cen-tral Sahara desert (Figure 10). These fields com-prise an area of 25,000 km2 [9,600 mi2] and have estimated recoverable gas reserves of 0.23 trillion m3 [8.1 trillion ft3].53 The gas from these fields contains 1% to 10% CO2, which is removed at the Krechba central processing facility (CPF). Carbon dioxide and any residual hydrogen sulfide [H2S] in the pro-duced gas are removed by monoethanol amine (MEA) absorption. The cleaned up gas from the Krechba CPF contains 0.3% or less CO2 and is trans-ported by pipeline to export terminals. The ISG proj-ect started production in 2004 and is currently producing 9 billion m3/yr [320 billion ft3/yr] of gas for export.

Carbon dioxide recovered from the produced gas was injected about 1,900 m [6,200 ft] into the water-filled downdip flank of the Krechba gas field. The three CO2 injection wells have horizontal sec-tions measuring up to 1.8 km [1.1 mi] in length (Figure 11). The joint venture conducted extensive monitoring of CO2 storage using a variety of tech-niques such as surface and soil gas monitoring, downhole gas measurements and tracer chemical tagging. Geophysical and InSAR satellite monitor-ing were also conducted to check for ground deformation and microseismicity (Figure 12).54

53. “In Salah Southern Fields Development Project, Algeria,” Hydrocarbons Technology, http://www.hydrocarbons-technology.com/projects/in-salah-southern-fields-development-project/ (accessed December 12, 2014).

54. Mathieson A, Midgley J, Dodds K, Wright I, Ringrose P and Saoul N: “CO2 Sequestration Monitoring and Verification Technologies Applied at Krechba, Algeria,” The Leading Edge 29, no. 2 (February 2010): 216–222.

Shi J-Q, Sinayuc C, Durucan S and Korre A: “Assessment of Carbon Dioxide Plume Behaviour Within the Storage Reservoir and the Lower Caprock Around the KB-502 Injection Well at In Salah,” International Journal of Greenhouse Gas Control 7 (March 2012): 115–126.

Stork AL, Verdon JP and Kendall J-M: “The Microseismic Response at the In Salah Carbon Capture and Storage (CCS) Site,” International Journal of Greenhouse Gas Control 32 (January 2015): 159–171.


19550

10,000

20,000

30,000

40,000

50,000

1970

Prod

uctio

n, b

bl/d

1985Year

2000 2015Year

Primary and waterflood Horizontal infill wells

CO2 EORVertical infill wells

Figure 9. Weyburn-Midale production. This facility (inset ) has produced oil since 1955. It and others at the Weyburn-Midale field have been used for production during four distinct phases (bottom). In the first phase, primary production and waterflood produced a total of 3.3 million bbl [0.52 million m3] of oil. The second and third phases, which used vertical and horizontal infill wells, produced a total of 5.9 million bbl [0.94 million m3] of oil. The last phase, CO2 EOR, has produced 9.4 million bbl [1.5 million m3] to date. (Adapted with permission from Cenovus Energy Inc.)

Figure 10. In Salah Gas (ISG) project. The In Salah CO2 storage project in central Algeria consists of several gas fields and a central processing facility (CPF, inset ) at Krechba, where CO2 and other impurities are removed from the produced natural gas. The cleaned up gas is sent by pipeline to a distribution station at Hassi R’Mel, Algeria, for further shipment to export terminals and markets in Europe.


Krechba CPF

300

km0

0

300

mi

S P A I N

PORT

UGAL

T U N I S I A

L I B Y A

A L G E R I A

M A L IMAU

RITA

NIA

Hassi R’Mel

AGas fieldExport terminal

ISG pipelineGas export pipeline

MOROCCO

September 2015 47

Figure 11. Carbon dioxide injection at Krechba. At the In Salah project, the Krechba central processing facility (CPF, inset ) consists of several sections—power generation (right ), CO2 removal and dehydration (center ) and CO2 injection. Recovery and injection of the CO2 removed from the natural gas are straightforward. The producing gas reservoir is about 20 m [66 ft] thick and lies about 1,900 m [6,200 ft] deep below a 950 m [3,100 ft] thick caprock formation of Carboniferous mudstones. A 900 m [3,000 ft] thick layer of Cretaceous sandstone and mudstone lies above the mudstone section. Produced gas from the reservoir is treated at the CPF to remove CO2, H2S and other impurities. The treated CO2 is then reinjected into water-saturated rock of the same reservoir from which the gas is produced.

Water

Gas

CO2 injection

Gasproduction


Gasd i

Cretaceous sandstonesand mudstonesCarboniferous mudstones

Gas

Water

Pipeline

Well casing

C

Figure 12. Monitoring ground deformation. Ground deformation is monitored at the In Salah project using interferometric synthetic aperture radar (InSAR). This technique uses a dedicated satellite (left ) to collect infrared radar images of the ground elevation using side-beam radar. Measurement of the vertical and horizontal displacements requires two passes of the satellite. Displacements are determined by comparing the wave phase changes of the radar signal between the two passes. The deformation and CO2 plume spreads at each well (marked by a cross) were estimated from the InSAR data for 2005, 2007 and 2009 (right ). The color intensity indicates the degree of vertical deformation (scale, far right ) while the size of the colored area around each well infers the horizontal spread of the CO2 plume.


KB-503

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

KB-502

2005

2007

2009

KB-14

KB-11

KB-12

KB-501

+ +

+

++

+

KB-503KB-502

KB-14

KB-11

KB-12

KB-501

+ +

+

++

+

KB-503KB-502

KB-14

KB-11

KB-12

KB-501

+ +

+

++

+

Nor

th, k

mN

orth

, km

Nor

th, k

m

0 5 10 15 20

East, km

20

15

10

5

–5

–10

–15

–20

0

Verti

cal d

efor

mat

ion,

mm

20

15

10

5

–5

–10

–15

–20

0

Verti

cal d

efor

mat

ion,

mm

20

15

10

5

–5

–10

–15

–20

0Ve

rtica

l def

orm

atio

n, m

m

Flight path

Displacement toward satellite

02.83 cm

Second passFirst pass

Flight path

Displacement toward satellite

02.83 cm

Second passFirst pass

48 Oilfield Review

Figure 13. Bati Raman project. The Bati Raman field and associated CO2 EOR project in Turkey are located about 720 km [450 mi] southeast of Ankara. The Bati Raman project has two CO2 injection stations—AP2 (inset ) and 3TP2 (not shown). (Photograph used with permission from the Turkish Petroleum Corporation.)


Ankara

Istanbul

Bati Raman

CYPRUS

BULGARIAR U S S I A

G E O R G I A

LEBANON

S Y R I A I R A Q

T U R K E Y

M e d i t e r r a n e a n S e a

B l a c kS e a

ARMENIA

IRAN

55. Ringrose PS, Mathieson AS, Wright IW, Selama F, Hansen O, Bissell R, Saoula N and Midgley J: “The In Salah CO2 Storage Project: Lessons Learned and Knowledge Transfer,” Energy Procedia 37 (2013): 6226–6236.

56. Sahin S, Kalfa U and Celebioglu D: “Bati Raman Field Immiscible CO2 Application—Status Quo and Future Plans,” SPE Reservoir Evaluation & Engineering 11, no. 4 (August 2008): 778–791.

Since 2004, approximately 3.5 million metric tons [3.9 million tonUS] of CO2 have been separated from the produced gas and reinjected into the Krechba reservoir.

Important lessons were learned about CO2 storage during the design, startup and operation of the ISG project, including the need for detailed geologic and geomechanical character-ization of the reservoir and the overburden.55 These data helped the operator develop the injection strategy and ensured the long-term integrity of the storage facility. The operator also realized the importance of flexibility in the design and control of the capture, compression and injection well systems.

Immiscible CO2 FloodingThe Bati Raman field, in southeast Turkey, is one of the largest oil fields in that country (Figure 13). Owned and operated by the Turkish Petroleum Company (TP), the field was discovered in 1961

and produces from a Garzan limestone—a het-erogeneous carbonate from the Cretaceous period.56 The heavy crude produced at the Bati Raman field has 11 degree API gravity [0.99 spe-cific gravity], high viscosity and low solution–gas content. The OOIP was estimated to be 1.85 bil-lion bbl [300 million m3]. From 1965 to 1970, the number of producing wells increased from almost 20 to more than 130.

During the primary production period from 1961 to 1986, reservoir pressure decreased from about 1,800 psi [12 MPa] to as low as 400 psi [2.8 MPa] in some parts of the field. Similarly, crude production declined from a peak rate of 9,000 bbl/d [1,400 m3/d] in 1969 to 1,600 bbl/d [250 m3/d] in 1986. During the primary produc-tion period, recovery was estimated to be less than 2% of OOIP.

Following the primary recovery period, the operator studied several processes for EOR and chose immiscible CO2 flooding primarily because of the proximity of the Dodan gas field.

The Dodan field is 55 mi [89 km] from the Bati Raman field and produces gas that is mostly CO2 and has 3,000 to 4,000 ppm H2S. The wellhead pressure at the Dodan field is about 1,050 psi [7.2 MPa]. After it is cleaned up, the CO2 from the Dodan field is sent to the Bati Raman field via pipeline (Figure 14).

Before implementing full-scale CO2 flooding at the Bati Raman field, TP performed a pilot test using 17 CO2 injection wells in the western part of the field. The original plan was cyclic injection of CO2 followed by water. After studying the pilot test results, TP engineers converted the initial

September 2015 49

injection plan to CO2 flooding. The operator made several observations based on the pilot test: CO2 injection helped produce a considerable amount of oil, and diffusion of CO2 into the oil was effec-tive for displacing oil in the fractured carbonate reservoir. After evaluating the results from the pilot test, TP engineers gradually extended the CO2 flooding to the rest of the field. Currently, 95% of the production wells in the Bati Raman field are influenced by CO2 flooding.

In 2012, the CO2 injection project was 25 years old, far beyond what was envisioned during the initial field design. More than 6% of the OOIP has now been recovered, a significant increase over the less than 2% recovered during primary field production. Primary recovery was 32 million bbl [5.1 million m3] while total field production, including that from primary, secondary and EOR recovery, was 114 million bbl [18 million m3] as of the end of 2014 (Figure 15).

Figure 14. Dodan and Bati Raman process flow. At the Dodan gas plant (left ), produced gas is stripped of H2S and water, compressed and sent by pipeline to the Bati Raman process facility, where it is injected (right ). At the Bati Raman facility, oil and CO2 in the produced stream are separated—the oil

goes to refining and the gas to cleanup. The cleaned up CO2 is compressed and sent to the injection wells. (Adapted with permission from the Turkish Petroleum Corporation.)


Pipeline

Production well

Flash tank

Compressor

Recycle compressors

Flar

e

Rege

nera

tion

Reboiler

Reboiler

Reboiler Tank

Vent

Flash tank

Separator

Separator

To refinery

H 2S a

bsor

ber

H 2O a

bsor

ber

Oil and CO2 production

H2O

H 2O a

bsor

ber

CO2

CO2

CO2

Injection wells

Dodan Gas Plant Bati Raman Recycle Station

Tank

Figure 15. Bati Raman production history. The Bati Raman field started producing in 1961, and as additional wells were brought on stream, production ramped up and peaked in 1970 at almost 10,000 bbl/d [1,600 m3/d]. Following this peak, production declined because of decreasing reservoir pressure. Water flooding started in 1975 and slowed the rate of production decline but did not reverse it. In 1986, primary production reached a low of about 2,000 bbl/d [300 m3/d], and CO2 injection for EOR was initiated. After CO2 EOR was introduced, production peaked around 1992 at about 15,000 bbl/d [2,400 m3/d]. Production decreased until 2004 when the practice of integrated reservoir management was implemented and arrested the decline. Production has held steady at about 7,500 bbl/d [1,200 m3/d] since that time. (Adapted with permission from the Turkish Petroleum Corporation.)


0

15,000

2,500

5,000

7,500

10,000

12,500

Prod

uctio

n, b

bl/d

1965 1970

Year

Primary Production Immiscible CO2 Injection

1975 1980 1985 1990 1995 2000 2005 2010 20141961

50 Oilfield Review

Carbon Dioxide and the FutureScientists have had an interest in CO2 for more than a century. The business sector, government regulators and the public have joined scientists in the quest to slow the atmospheric accumula-tion of CO2. The oil industry is tackling this chal-lenge in part by injecting CO2 underground, both for EOR and for long-term storage purposes.

In addition to EOR and storage, the industry can also take advantage of new, less-expensive sorbents for CO2 capture, which will also help reduce CO2 emissions into the atmosphere. Current technology depends on absorbents, including aqueous MEA, to remove CO2 from streams such as power plant flue gas.57 The MEA solution is corrosive, degrades into toxic byprod-ucts and requires a large amount of energy to clean it up for reuse.

The unfavorable characteristics of current sorbents such as MEA have led researchers to develop new ones—both solid and liquid. One of these new sorbents is solid, microporous carbon that is synthesized from asphalt (Figure 16).58 This sorbent is inexpensive, has high surface area CO2 uptake and excellent properties for revers-ibly capturing CO2. Another new sorbent is a liq-uid carbonate enclosed within polymer microcapsules with shells of highly permeable silicone.59 These microcapsules are reported to have rapid CO2 uptake and release.

Although new sorbent technology can help, it is only part of the solution for emissions. The reduction and mitigation of GHG emissions will require simultaneous implementation of several technologies and significant governmental action on a worldwide basis. These technologies range from efficiency improvements to alternate energy sources to conservation soil tillage. Likewise, gov-ernments can help, for example, by reducing sub-sidies for inefficient hydrocarbon use and intelligent mandates on fuel efficiency.

One area in which the oil and gas industry can play an important role is geologic storage in CCS.60 A technical challenge related to geologic storage is risk associated with faulty CO2 confinement. The oil and gas industry has the technical tools to assess the potential and risk for CO2 migration

away from storage sites. Although extra costs associated with large-scale geologic storage of CO2 will be incurred, these costs are fundamen-tally no different than additional costs already borne by the public for cleaner water and air.

The consequences of climate change are potentially enormous. In the last decade, as evi-dence for the effects of climate change mounts, moving beyond maintaining the status quo and doing business as usual has become important. Although it can help to reduce the problem, the oil and gas industry can offer only some solutions. Worldwide, industries, governments and the pub-lic must be educated and ready to support a vigor-ous effort to arrest climate change. —DA

Figure 16. Carbon dioxide capture. Asphalt (left ) is carbonized by treating it with potassium hydroxide [KOH] at 700°C [1,300°F]. This process yields treated asphalt (center ) that has a surface area of nearly 2,800 m2/g. At a pressure of 30 bar [440 psi] and at room temperature, the treated asphalt (right ) can absorb 93% of its weight in CO2. The CO2 can be desorbed using a simple pressure swing absorption process. (Adapted with permission from Jalilov et al, reference 58.)


50 nm

KOH

700°C

CO2

Treatedasphalt

Asphalt

57. Vericella JJ, Baker SE, Stolaroff JK, Duoss EB, Hardin JO IV, Lewicki J, Glogowski E, Floyd WC, Valdez CA, Smith WL, Satcher JH Jr, Bourcier WL, Spadaccini CM, Lewis JA and Aines RD: “Encapsulated Liquid Sorbents for Carbon Dioxide Capture,” Nature Communications 6, article 6124 (February 5, 2015).

58. Jalilov AS, Ruan G, Hwang C-C, Schipper DE, Tour JJ, Li Y, Fei H, Samuel ELG and Tour JM: “Asphalt-Derived High Surface Area Activated Porous Carbons for Carbon Dioxide Capture,” ACS Applied Materials and Interfaces 7, no. 2 (January 21, 2015): 1376–1382.

59. Vericella et al, reference 57.60. Bryant S: “Geologic CO2 Storage—Can the Oil and Gas

Industry Help Save the Planet?,” Journal of Petroleum Technology 59, no. 9 (September 2007): 98–105.

Contributors

September 2015 51

Jamie Stuart Andrews is a Sandface Completion Specialist for Statoil in the Drilling and Well Engineering group in Stavanger, where he focuses on production-related rock mechanics, including sand control, sand prediction, perforating and injection and fracturing challenges. He began his oil industry career in 1993 at Shell International as a production technol-ogist then joined Statoil in 1995 and spent two years researching formation damage and sand control prob-lems. He spent the next several years in various asset teams concentrating on production optimization, well performance and well completions. In 2004, he became lead advisor in the company’s central technol-ogy unit in Stavanger and started his present position in 2009. Jaime earned a PhD degree in theoretical chemistry at the University of Cambridge, England.

Mehdi Ansarizadeh is a Principal Reservoir Engineer for PetroTechnical Services at Schlumberger Canada in Calgary. He leads the local engineering team and provides technical support to enhanced oil recovery (EOR) and unconventional resources projects. Before joining Schlumberger, he worked as a reservoir simula-tion engineer for the National Iranian Oil Company. During his 27 years of experience in the oil and gas industry, he has simulated complex reservoir systems that have various recovery mechanisms, from simple expansion drives to EOR processes such as miscible CO2 and alkaline surfactant polymer floods. Mehdi holds a BSc degree in reservoir engineering from the Abadan Institute of Technology, Iran, and an MEng degree in chemical and petroleum engineering from the University of Calgary.

Joseph A. Ayoub is the Schlumberger Production and Completion Engineering Discipline Manager in Sugar Land, Texas, USA. He holds more than twelve patents and has published more than 35 papers, pre-dominantly in the areas of well testing, hydraulic fracturing and frac pack. His involvement in develop-ing the pressure derivative method of well testing was instrumental in bringing the technique to the Gulf of Mexico in the early 1990s. More recently, he has led the formation of industry consortia to investi-gate technical challenges in the areas of stimulation and sand control. He has served on numerous SPE committees and was an SPE Distinguished Lecturer in 1998/1999 on Improving Productivity of Sand Control Completions and in 2009/2010 on Realizing Full Potential of Hydraulic Fracturing. He was named Schlumberger advisor in 1999 and SPE Distinguished Member in 2005 and served as SPE technical director drilling and completions from 2010 to 2013. Joseph has an engineering degree and a DEA degree from Ecole Centrale Paris.

Michael Azar is a Director of Design Engineering and an Advisor for Smith Bits, a Schlumberger company, in Houston. He began his career with Smith International in 1985, specializing in the design and development of PDC and diamond bits. Over the past 30 years, Michael held multiple positions within the company, including assignments in Europe and North Africa. He currently

manages the research and development of new drillbit technologies for the oil and gas industry. Michael received his BS degree in mechanical engineering from The State University of New York at Buffalo, USA.

Janice Brown, based in Fort Worth, Texas, has been a Senior Borehole Geologist for Schlumberger since 2000. She began her career in 1978 as a geologist at Sun Oil in Dallas, working the Anadarko, Arkoma, and Eastern Permian basin exploration plays. She also pro-vided software support for geologic cross section appli-cations for Schlumberger and Mobil in Dallas. Janice obtained a BS degree in geology from The University of Texas at Dallas.

Rajesh A. Chanpura is a Product Champion for screens and inflow control devices for Schlumberger Sand Management Services in Sugar Land, Texas. He is responsible for defining key product requirements and introducing to the field new products developed under the screens and inflow control devices portfolio. Rajesh joined Schlumberger in 2002. Following his involvement in the development of the Schlumberger gravel packing simulator, he developed an in-house methodology and answer product for screen selection in openhole completions. He holds an undergraduate degree in construction engineering from the University of Mumbai, an MS degree in civil engineering from the Indian Institute of Technology, Mumbai, and a PhD degree in civil engineering from the Georgia Institute of Technology, Atlanta, USA.

Chance Copeland is the Smith Bits Product Engineering Manager for the Permian basin in Midland, Texas. Prior to his more than five years of work in the Permian basin, he was a bits product engi-neer and has also assumed technical sales responsibil-ities. Chance earned a BS degree in mechanical engineering from Texas Tech University in Lubbock.

Bob Davis recently retired as geology discipline career manager for Schlumberger, where he managed 600 geologists worldwide; he was based in Oklahoma City, Oklahoma, USA. He served as division geologist for Schlumberger Wireline & Testing in New Orleans prior to his most recent position. He joined Schlumberger in 1977 and has had various assignments, including field engineer and field service manager in Louisiana, USA, and sales engineer and district manager in Pennsylvania, USA. Bob focused on geologic applica-tions and dipmeter interpretation and helped develop models for stratigraphic interpretations in both hard and soft rock basins. He authored and coauthored sev-eral papers on dipmeter field studies and borehole image interpretation and applications. He has a BS degree in aerospace engineering from the University of Oklahoma, Norman.

Kevin Dodds is the General Manager for Research for Australian National Low Emissions Coal Research and Development, based in Canberra, Australian Capital Territory, Australia. He is responsible for managing the carbon capture and storage (CCS) research program in support of existing and prospective demonstration projects in Australia. He has been a technical project manager with Schlumberger and the Commonwealth Scientific and Industrial Research Organisation

(CSIRO), where he led the monitoring and verification program for the Otway CCS project. Before his current position, he spent eight years with BP in Houston, con-tributing to global CCS programs through commission-ing projects for the CO2 Capture Project in the US Department of Energy regional projects and working on Alberta, Canada, government regulatory commit-tees. Kevin received a BSc (Hons) degree in upper atmospheric physics from the University of New England at Armidale, New South Wales, Australia, and an MSc degree in applied geophysics from the University of Birmingham, England.

Henry Edmundson worked more than 45 years for Schlumberger, was founding editor of the Oilfield Review and now runs his own energy consulting busi-ness. He is based in Cambridge, England.

Kiran Gawankar is a Senior Staff Petrophysicist with the Appalachian Division at Southwestern Energy in The Woodlands, Texas; he’s been with the company since 2013. He began his career as a wireline field engineer with Halliburton in Mumbai and was later team leader for the company’s reservoir evaluation group in Oklahoma City, Oklahoma. He next served as team leader in the Shale Study Group with NUTECH Energy Alliance in Houston. Kiran obtained a bache-lor’s degree in mechanical engineering from the University of Pune, India.

Omer Gurpinar is the Schlumberger Technical Director of Enhanced Oil Recovery (EOR) in Denver. He leads the development of technologies and services to help improve recovery factors in oil fields. He has more than 35 years of industry experience around the world in various aspects of EOR, naturally fractured reservoirs, critical fluids and field optimization. Since joining Schlumberger, he has served as the vice presi-dent and technical director in various segments and played a key role in building the Schlumberger E&P consulting organization. Before working for Schlumberger, he was the vice president for reservoir simulation with INTERA and then chief reservoir engi-neer for Scientific Software-Intercomp. Omer started his professional career with the Turkish Petroleum Corporation in Ankara, Turkey. He earned BSc and MSc degrees in petroleum engineering from Middle East Technical University, Ankara, and holds several key industry patents relevant to reservoir and field optimization and EOR.

Ryan Hempton is a Drilling and Completions Engineer for Cimarex Energy Company in the Permian basin in Midland, Texas. During his six years in the industry, he has also worked as an MWD field engineer, directional driller and project manager at Schlumberger. In his current position, he focuses on improving the efficiency of drilling and development of Delaware Basin shale plays. Ryan holds a BS degree in mechanical engineering from Kettering University, Flint, Michigan, USA.

52 Oilfield Review

Ülker Kalfa, who works at the Turkish Petroleum Corporation (TP) headquarters in Ankara, Turkey, is Manager for the Bati Raman EOR Project. She over-sees drilling and production and injection monitoring and surveillance. She held a research assistantship at Middle East Technical University in Ankara before she joined TP. Prior to taking her position on the Bati Raman team in 2002, she worked as a petroleum engi-neer for the TP district office in Batman, Turkey, and as an information technology staff member for the pro-duction department in Ankara. Ülker has BS and MS degrees in petroleum engineering from Middle East Technical University, Ankara.

Anish Kumar is a Geology Domain Champion for Schlumberger North America Offshore Wireline. He started his career with Schlumberger in 2001 as an interpretation development geologist in New Orleans; his primary focus was on borehole geology and deepwa-ter deposition. Anish received an MSc degree in geology from University of Roorkee, India, and a PhD degree in geology from Texas Tech University in Lubbock.

Robert Laronga is the Schlumberger Wireline Headquarters Geologist in Clamart, France, a position he has held since 2010. He began his career with Schlumberger as a wireline field engineer and worked in West Texas, the Gulf of Mexico and aboard the drill-ship of the Ocean Drilling Program. Rob was next the OBMI* field-test engineer and then the OBMI product champion in Clamart. He then served as the sales and marketing manager for Wireline in the North Sea before becoming the Wireline marketing manager for continental Europe, based in Bucharest. Rob received a BA degree in archaeology and geology from Cornell University, Ithaca, New York, USA.

Bingjian Li is Principal Geologist for Unconventional Measurements for the Schlumberger USA Land GeoMarket* in Houston. Responsible for geology solu-tion product development, his main focus is on shale reservoirs using geology measurements, including those from borehole images and sidewall cores. He previously worked as a geologist in Southeast Asia, the Middle East and Canada. Prior to joining Schlumberger, he also worked in China. For the past three years, he has served as committee member and abstract reviewer for the SPE annual conferences and has published or coauthored more than 20 papers; he is a technical reviewer for the Journal of Unconventional Oil and Gas Resources. Bingjian obtained a BSc degree in petroleum geology from Northeastern Petroleum University, Daqing, China, and a PhD degree in reservoir geology and sedimentol-ogy from the University of Aberdeen.

Wiley Long is the Product Champion for StingBlade* conical diamond element bits for Smith Bits in Houston. During his 10 years with Smith Bits, he has served as a bits product engineer in South Texas and in Russia, an engineering manager for Russia and Central Asia and the bits global drilling optimization manager. Wiley earned a BS degree in mechanical engineering from Texas A&M University, College Station.

Camron K. Miller is a Principal Geologist with Schlumberger Production Management in Houston. He has directed the reservoir characterization of potential North American assets since early 2015. He began his

career with Schlumberger as a wireline field engineer in 2004 and has held various positions in the company in Oklahoma and Texas. His areas of expertise are borehole geology, reservoir characterization and the exploration and appraisal of unconventional resources. Camron has a BS degree in geology from The College of Wooster, Ohio, USA, and an MS degree in geology from the University of Akron, Ohio.

Somnath Mondal is a Production Technologist for the Unconventional Gas and Tight Oil–Completions and Stimulation R&D Team for Shell International E&P in Houston. He performs research on comple-tions and stimulations to increase the financial and environmental performance of tight gas and oil devel-opments. His responsibilities include proposing field trials and designing, running and analyzing experi-ments and computer simulations in collaboration with internal and external partners. His current areas of focus are understanding hydraulic fracturing effectiveness, refracturing strategy and novel prop-pant and fluid designs. He began his career with Shell E&P Company in 2010 as an intern in the deep-water technology deployment group. Somnath, who has coauthored numerous journal and conference papers, received a BSc degree in chemical engineer-ing from the Birla Institute of Technology, Pilani, India, and MS and PhD degrees in petroleum engi-neering from The University of Texas at Austin.

Mikhail Pak is the Smith Bits Engineering Support Manager for Russia and Central Asia; he is based in Moscow. Previously, he worked in various locations in Russia and the Caspian region focusing on product development and technical support for Smith Bits. Mikhail obtained a master’s degree in mechanical engineering from Yeungnam University in Gyeongsan, South Korea.

Mehmet Parlar is a technical advisor at Schlumberger Sand Management Services in Sugar Land, Texas. He has 26 years of industry experience; 7 of those years are in product development and 19 years in sand con-trol, all with Schlumberger. Contributor to more than 60 technical papers and the author of a chapter in the SPE Frac Packing Handbook, Mehmet also holds 27 US patents. He is an SPE Distinguished Member and SPE Distinguished Author and has served in a variety of roles—including organizing committee member, technical committee member, session chair, discussion leader, moderator and speaker—for many SPE events. He was an SPE Distinguished Lecturer in 2007/2008 on Current Practices and Challenges in Gravel Packing Open Holes with Reactive Shales and in 2011/2012 on State of the Art in Openhole Sand Control Completions: Advancements & Gap. He has organized numerous internal and external training courses on sand control. Mehmet holds a BS degree in petroleum engineering from Istanbul Technical University, Turkey, and MS and PhD degrees in petro-leum engineering from the University of Southern California in Los Angeles.

Lawrence J. Pekot is a recently retired Schlumberger senior project manager. He joined the company in 2004 as a Schlumberger Eureka Technical Career principal and held positions in unconventional reservoir consult-ing and project management in Data & Consulting Services and as a Europe and Africa Area technical

manager for the Carbon Services Segment. Before join-ing Schlumberger, he was vice president with Advanced Resources International, Inc., an unconven-tional resources consulting firm in Washington, DC. He also worked as senior reservoir engineer for Phillips Petroleum Company, developing new North Sea reser-voirs. He left Phillips in 1991 to begin a career in con-sulting and has since authored or coauthored more than 25 publications. Lawrence earned BSc degrees in both civil engineering and in geological science from The Pennsylvania State University, State College.

T.S. Ramakrishnan, a Scientific Advisor at Schlumberger-Doll Research Center in Cambridge, Massachusetts, USA, is the Research Director for the Enhanced and Unconventional Recovery department. He has authored more than 100 papers in carbonate petrophysics, pressure transient and formation testing, fluid mechanics, induction logging, nuclear magnetic resonance and applied mathematics. Rama, who holds 50 patents, is an SPE Distinguished Member and received the 2009 SPE Formation Evaluation Award. In addition to the Henri Doll Award for Innovation and other awards from Schlumberger, he was the recipient of the 1980 Acharya P.C. Ray award, an SPWLA best paper award, a 2012 Charles W. Pierce Distinguished Alumni Award from the Illinois Institute of Technology in Chicago (IIT Chicago) and a 2013 Distinguished Alumni Award from the Indian Institute of Technology Delhi (IIT Delhi), New Delhi, India. He has a BTech degree from IIT Delhi and a PhD degree from IIT Chicago, both in chemical engineering.

Norm Sacuta is the Communications Manager at the Petroleum Technology Research Centre (PTRC) in Regina, Saskatchewan, Canada. He is responsible for communications and public outreach for the International Energy Agency Greenhouse Gas R&D Programme with the Weyburn-Midale CO2 Monitoring and Storage Project. Norm currently manages com-munications for the PTRC enhanced oil recovery research programs. He came to the PTRC in 2008 after eight years working in communications for Natural Resources Canada in oil sands environmental research at the CANMET Energy Technology Centre in Devon, Alberta, Canada. Norm received an MA degree in English from the University of Alberta, Canada, and an MFA in creative writing from the University of British Columbia in Vancouver, Canada. Norm has many years of experience in journalism and creative writing and has published the poetry collec-tion Garments of the Known.

Secaeddin Sahin is Project Coordinator of the Bati Raman field immiscible CO2 injection project and manages the steam injection pilot project in the same field. He joined Turkish Petroleum Corporation (TP) in 1986 and has more than 25 years of experience in oil and gas production operations. At TP, he has held posi-tions in oil and gas production operations, enhanced oil recovery (EOR) and underground gas storage (UGS) operations. He served as project manager of the first UGS project in Turkey. Currently based in Ankara, Turkey, his focus is on heavy oil production, carbon capture and storage, CO2-EOR and UGS. Secaeddin obtained a BSc degree in petroleum engineering from Istanbul Technical University, Turkey.

An asterisk (*) denotes a mark of Schlumberger.

September 2015 53

Peter Schlicht is the Technical Director, Exploration and Development and Geology Advisor to Schlumberger Wireline Headquarters in Clamart, France. He is responsible for supporting borehole imaging and coring activities. His primary focus is on field support, including hardware and software devel-opment. Peter began his career in Schlumberger as a wireline field engineer in 2001 in Hobbs, New Mexico, USA. He has served as geology domain champion for Sub-Saharan, West Africa, based in Lagos, Nigeria, and Luanda, Angola. He supported operations in explora-tion activities in Nigeria, Ghana, Gabon, Congo, Angola, Mozambique and Tanzania, including forma-tion evaluation activities in most major deepwater projects in West and East Africa. Prior to his current assignment, he was geology research program manager at the Schlumberger Brazil Research and Geoengineering Center in Rio de Janeiro. He holds a German diploma degree in geology-paleontology from the University of Cologne, Germany.

Mukul M. Sharma is a Professor and holds the “Tex” Moncrief Chair in the Department of Petroleum and Geosystems Engineering at The University of Texas at Austin, where he has taught for the past 30 years. He served as chair of the department from 2001 to 2005. His areas of expertise are hydraulic fracturing, injec-tion water management, formation damage, improved oil recovery and petrophysics; he has published more

than 300 journal and conference articles and holds 15 patents. He is the recipient of the 2009 Lucas Gold Medal, the highest technical award from the SPE, the 2004 SPE Faculty Distinguished Achievement Award, the 2002 Lester C. Uren Award and the 1998 SPE Formation Evaluation Award. He served as an SPE Distinguished Lecturer in 2002, has served on the editorial boards of many journals and has taught and consulted for more than 50 companies worldwide. He cofounded two private exploration and production companies and a consulting company. Mukul earned a bachelor of technology degree in chemical engi-neering from the Indian Institute of Technology, Kanpur, and MS and PhD degrees in chemical and petroleum engineering from the University of Southern California, Los Angeles.

Serkan Uysal is a Chief Engineer in the Production Department of Turkish Petroleum Corporation (TP) in Ankara, Turkey. He has more than 21 years of experi-ence in the upstream oil and gas industry and has worked on several projects in Turkey, Azerbaijan and Kazakhstan. His work has focused on production and development of oil fields; water, CO2 and steam-injection enhanced oil recovery (EOR); and construction of oil and gas processing facilities. He currently works on the EOR project in the Bati Raman field, in which immisci-ble CO2 injection has been ongoing for almost 30 years. Serkan has a BS degree in petroleum engineering from Middle East Technical University, Ankara.

Allen White is the Operations Manager for Smith Bits in Jakarta. He was previously product champion for novel cutting structures for Smith Bits in Houston. During his nine years in the industry, he has also worked as a product champion for PDC cutters, as an engineering manager for design engineering and as a design engineer. Allen obtained BS and MS degrees in mechanical engineering from Texas Tech University in Lubbock.

Steve Whittaker is based in Perth, Western Australia, Australia, where he is the Research Group Leader—Reservoir Dynamics with the Commonwealth Scientific and Industrial Research Organisation (CSIRO). Previously, he was principal manager for geologic stor-age of CO2 at the Global CCS Institute in Canberra, Australian Capital Territory, Australia, and chief tech-nology manager at the Petroleum Technology Research Centre in Regina, Saskatchewan, Canada, where he managed a program studying storage and monitoring of CO2 injected for enhanced oil recovery at Weyburn, Saskatchewan. Steve is a petroleum geologist and has worked in petroleum-related fields for more than 15 years. He received BSc and PhD degrees in geology from the University of Saskatchewan.

Developing the Vaca Muerta Unconventional Resource Play. The Vaca Muerta Formation in the Neuquén basin of Argentina is an unconventional resource play with an estimated 103 billion m3 [661 billion bbl] of original oil in place and in excess of 28 billion m3 [1,000 Tcf] of natural gas. In developing the play—one of the first successful shale plays out-side North America—the Argentine energy company YPF is deploying an integrated approach that includes a dynamic unconventional fracture model to optimize stimulation programs and maximize field production.

Coming in Oilfield Review

Artificial Lift in Unconventional Wells. Typically, extremely low-permeability horizontal shale wells are characterized by rapid decline in production rates, fluid composition changes and a need for artificial lift soon after initial production. Installing rod pump systems in unconventional resource wells requires engineers to adapt technologies to operating environments that dif-fer from those found in conventional vertical wells for which they were originally intended.

Mud Removal. Proper removal of drilling fluids is essential in achieving effective zonal isolation and ultimately primary cementing success. Typically, mud displacement solutions are selected based on prior experience, but customized spacers and engineered solutions that enhance cleaning performance and effi-ciency are now available.

Improving Exploration Success. Seismic imaging cannot resolve many potential exploration targets lying beneath shallow rock layers. The overburden behaves as a defective lens, distorting seismic imag-ing of deeper geologic structures. As a result, targets appear indistinct, distorted, mislocated or, in extreme cases, completely obscured. New developments in multimeasurement marine seismic acquisition and imaging are making it possible to compensate for the distortions, sharpen images of deeper targets and reduce the uncertainty of seismic information about drilling prospects.

Oilfield Review54

THE DEFINING SERIES

Because they lack sufficient reservoir pressure to produce fluids to the sur-face, the majority of the world’s oil and gas wells are unable to produce at economic rates without assistance. This condition may be the result of pres-sure depletion over time or be caused by low original reservoir pressure.

To compensate for the lack of natural energy in these formations, opera-tors equip the wells with artificial lift (AL) systems. Artificial lift well can-didates are those completed in formations that have economically viable reserves and sufficient permeability for the fluids to move to the wellbore but do not have sufficient reservoir drive to lift those fluids to the surface. Secondary recovery efforts, such as waterfloods designed to capture remain-ing reserves in pressure-depleted reservoirs, often result in increased fluid volumes that can be lifted to the surface only through AL methods.

When choosing a specific AL system, engineers must consider—in addi-tion to surface conditions based on location—a host of parameters, includ-ing reservoir characteristics, production properties, fluid types and operational considerations. Choice of an optimal AL system may be influ-enced by subsurface conditions, expected production rates, fluid composi-tion, well geometry, reservoir depths, completion configuration and surface facilities. In addition, operators must consider the potential return on their investment by balancing the value of increased production against the cost of hardware for and installation and maintenance of an AL system.

Artificial lift systems are deployed predominantly to extend well life. But these systems may also help shorten the time from first production to well abandonment. For example, operators may gain an economic advan-tage by accelerating recovery rates, a process that more quickly drains the reservoir, thus saving expenses in situations characterized by high operat-ing costs.

After an operator has established that an AL system is advisable, pro-duction engineers choose the type best suited to the situation. For example, electric submersible pumps and gas lift systems are often chosen to boost production in offshore wells because such systems have small footprints, are able to handle high production volumes and may be deployed at signifi-cant depths below the wellhead. On the other hand, sucker beam pumps, which require a significant amount of surface space but are reliable, easily serviced and one of the least expensive of the AL options, are often the optimal solution for land-based, marginally economic wells.

Artificial lift systems fall into two basic types: pumping and gas lift. Pumping systems include electric submersible pumps, beam pumps, pro-gressing cavity pumps, plunger lifts and hydraulic pumps.

Electric Submersible PumpsPerhaps the most versatile AL systems are electric submersible pumps (ESPs). These pumps comprise a series of centrifugal pump stages con-tained within a protective housing. A submersible electric motor, which drives the pump, is deployed at the bottom of the production tubing and is connected to surface controls and electric power by an armored cable strapped to the outside of the tubing.

An ESP derives its versatility from a wide range of power output drives and from variable speed drives that allow operators to increase or decrease volumes being lifted in response to changing well conditions. Additionally, modern ESPs are able to lift fluids with high gas/oil ratios (GORs), can be designed using materials and configurations able to withstand corrosive flu-ids and abrasives and can operate in extreme temperatures.

Beam PumpsA beam pump system is composed of a prime mover, a beam pump, a sucker rod string and two valves (Figure 1). The gas- or electric-driven prime mover

Artificial Lift

Oilfield Review 27, no. 2 (September 2015).

Copyright © 2015 Schlumberger.

For help in preparation of the article, thanks to Kyle Hodenfield, Houston, Texas, USA.

Rick von FlaternSenior Editor

Figure 1. Beam pumps. A traveling ball valve at the end of the rod string is pushed off seat as it travels downward through the fluid column. When the traveling valve achieves maximum downward reach and the beam is at its lowest point, the beam begins its upward movement, and the rods are pulled upward, which forces the ball of the traveling valve to be forced back on to its seat; As a result the fluid column (green) is captured above it. As the fluid is pulled toward the surface, the pressure in the tubing decreases, which causes a standing ball valve at the end of the tubing to open. Formation fluid (green arrows) flows through this lower valve and fills the wellbore. When the traveling valve begins its descent, the pressure of the fluid column forces the stationary valve ball to fall back onto the valve seat, and the cycle is repeated. The standing and traveling ball valves are often contained within an insert pump so the entire assembly can be retrieved with the rod string.

Insert pump

Rods

Tubingpump

Tubing

Casing

Prime mover Beam pumping unit

Perforations

Producedfluids

Plunger

Fullbore barrel

Traveling valve

Standing valve

turns a crank arm, which causes a beam to reciprocate. The resulting up and down movement lifts and lowers a rod string attached to one end of the beam. The motion of the rod string opens and closes traveling and standing ball valves to capture fluid or allow fluid to flow into the wellbore. In some configurations, the valves are part of an integrated assembly called an insert pump, which can be retrieved using the rods while leaving the production tubing in place.

Beam pump equipment and parameters—valves, prime mover, rod and tubing diameter, and stroke length—are determined according to reservoir fluid composition, depth to the fluid top and reservoir productivity. The sys-tems are typically equipped with timers that turn the pumps off to allow fluid time to flow through the formation and into the wellbore. The timer then restarts the pump for a period calculated to produce the fluid that has accumulated in the well.

Progressing Cavity PumpsThe progressing cavity pump consists of a rotor placed inside a stator. The rotor is a screw that has deep round threads and extremely long pitch—the distance between thread tops. The stator has a longer pitch and one more thread than the rotor. When the rotor turns inside the stator, the thread and pitch differences create a cavity within the pump barrel that is filled by formation fluid. The rotor is turned by a rod string connected to a motor at the surface or by an electric-drive motor located downhole at the pump mov-ing the fluid uphole.

PlungersPlunger lift systems, the simplest form of artificial lift, consist of a piston, or plunger, that has only small clearance through the production tubing and is allowed to fall to the bottom of the well. They are used primarily in high GOR wells to lift liquids out of the well to allow the gas to be recovered. A valve on the surface is closed, which causes natural pressure from the res-ervoir to build in the casing annulus. At a preset pressure level, the valve on the surface opens and pressure from the annulus enters the tubing below the plunger, which forces it upward. The plunger pushes the fluid column above it to the surface. When it reaches the surface, the plunger enters the lubricator, a short section of pipe, which extends above the wellhead. Because the plunger is no longer in the flow path, the gas that provided the lifting energy can pass beneath it and along the flowline. When the pressure at the wellhead has dropped to a predetermined level, the surface valve closes, the plunger falls from the lubricator to the bottom of the well, and the cycle is repeated.

Hydraulic PumpsIn some situations, operators may install a hydraulic pumping system that pumps a fluid, called a power fluid, from the surface through tubing to a subsurface pump. The subsurface pumps, which may be jets, reciprocating pistons or rotating turbines, force the formation fluids and the power fluid up a second tubing string to the surface.

Hydraulic pumping systems offer two specific advantages. Because the subsurface pump is free floating, it can be circulated out of the hole for repair with little intervention cost. And the power fluid, which is typically refined oil, mixes with the produced fluid; the resulting fluid column exerts a lighter hydrostatic pressure than does the formation fluid alone, reduces the resistance to flow and lessens the work required of the downhole pump. As a consequence, hydraulic pumps are frequently chosen for use in heavy oil operations.

Gas Lift SystemsAs an alternative or in addition to pump solutions, gas lift systems aid flow to the surface by reducing the density of formation fluids in the wellbore. Gas lift systems consist of valves installed at various depths along the tubing string, which open in response to pressure exerted on them by the rising fluid column. When the valve opens, injected gas mixes with and lightens the fluid column, reducing the hydrostatic and thus the bottomhole pres-sure. The lower hydrostatic pressure reduces the drawdown pressure and allows formation fluid to enter the wellbore. The less dense fluid column may then be lifted to the surface by reservoir pressure alone.

Optimally, gas lift systems use continuous gas injection at a rate that ensures a steady flow of fluids to the surface. However, if drawdown pres-sures are insufficient, intermittent injection schemes using gas lift valves may be implemented to allow formation fluids time to enter the wellbore; the gas then lifts slugs of fluid to the surface. Although effective, slug pro-duction can cause fluid handling problems at the surface and surges down-hole that may initiate sand production.

Gas valve locations and injection rates are based on the needs of the indi-vidual well. Gas lift valves may be set in side pocket mandrels—receptacles that are included as part of the completion design. Because valves are placed in the mandrels using running and setting tools carried downhole via slick-line, when well conditions change, operators can retrieve and change the gas lift valves without pulling the tubing from the well. Technicians can adjust the valves to open at pressures that meet the needs created by the new conditions and replace the valves in the well with minimal intervention costs.

Indispensable TechnologyThe great majority of the world’s approximately one million active oil wells use some form of artificial lift (Figure 2). Such demand has resulted not just in innovation in AL technologies but in an AL discipline. Operators are able to design the best AL system for each well and field and to adjust those systems to meet changing well and reservoir conditions.

Today, AL systems include technologically advanced downhole pumps that can be monitored and controlled remotely in real time and are able to pump thousands of barrels of fluid per day even in wells that have signifi-cant amounts of solids. The efficiency of modern beam pumps is such that hundreds of thousands of stripper wells in the US are able to remain profit-able while producing less than 2.4 m3 [15 bbl] of oil or 2.5 m3 [990 Mcf] of gas per day. Since most oil wells and fields eventually rely on AL to continue production, these advances ensure that operators are able to provide a con-tinuous stream of oil and gas to an energy-hungry world.

Figure 2. Artificial lift choices. By some estimates, the four most common artificial lift types are currently deployed in more than 800,000 wells worldwide with capacity to lift fluids ranging from negligible amounts to 60,000 bbl/d. System capacity may be limited by depth, wellbore trajectory or the ability of the formation to deliver liquids to the well. As a consequence, most systems are most efficient when operating in the middle range of their volume capabilities and at less than maximum depth.

16,0008,60018,00015,000

Beam pumpProgressing cavity pumpGas liftElectric submersible pump

Artificial lift type Maximum depth(TVD), ft

Typical operatingvolume, bbl/d

Number of wells worldwide

0 to 1,0000 to 5,000

100 to 80,000400 to 60,000

600,00031,50048,300146,700

September 2015 55

56 Oilfield Review

Imaging: Getting the Downhole Picture

Downhole image logs help geologists identify and analyze reservoir fea-tures such as fractures, folds and faults. Stratigraphic features can also be seen in image logs. A long-standing gap in imaging technology left the technique significantly less effective in wells drilled using oil-base mud (OBM) systems than in those using water-base mud (WBM) systems. This situation was recently addressed with the introduction of the Quanta Geo* photorealistic res-ervoir geology service.

Imaging tools evolved from dipmeter tools, which are wireline conveyed devices designed to determine formation geometry and structural properties. In the 1980s, Schlumberger introduced the forma-tion MicroScanner* service—one of the industry’s first borehole imaging tools. Imaging tools provide data that allow geologists to visual-ize structural and stratigraphic features. Because of the electri-cal insulating properties of OBM, obtaining useful images in wells drilled using OBM was challenging, and the results were often less infor-mative than in those drilled with WBM systems. In 2014, engineers at Schlumberger introduced the Quanta Geo service, which provides photorealistic images of wellbores drilled with OBM.

The ability to image boreholes in OBM systems is significant because almost every well drilled in deep-water exploration uses OBM, which improves drilling performance and reduces overall costs. Because oper-ating costs are high in deepwater environments, operators drill only a limited number of wells to extreme depths, which limits traditional subsurface mapping methods that integrate surface seismic with down-hole data. This limitation can be problematic for geologists modeling formations such as deep structures located beneath salt. Interpretations of image data from wells drilled using OBM systems help geologists fill the knowledge gap and aid them in opti-mizing well placement. Page 4.

Sand Screen Selection

A successful sand control strategy is one that minimizes sand produc-tion while allowing formation fluid to flow freely into the wellbore. The sand control method of choice is typically screens placed across the producing formation. Traditionally, the type and size of the screens selected for the pur-pose are based on time-honored methods and laboratory tests.

Traditional selection methods center on choosing the optimal sand screen based on a relation-ship between screen opening and a single point in grain size distributions. But recent research suggests that it may be more effi-cient for operators to select sand screens based on numerical and analytical models.

A team from Schlumberger, academia and industry have arrived at a process by which engineers match optimal wire wrap and metal mesh stand-alone screen size and type to target formations in open-hole completions. In addition, they have developed a technique that allows engineers to use the entire sand size distribution when select-ing a screen and to quickly narrow the range of screen sizes and types to optimize sand control.

This new process often results in sand control decisions better suited to the well at hand than is possible using past practices that use only one design parameter. The process also reduces the number of labora-tory tests that must be performed to determine the optimal choice for a target formation. Page 22.

A New Approach to Fixed Cutter Bits

The price of drilling into certain formations is tool-jarring shock and vibration, slow rates of penetration and damaged bits. The usual solu-tion is to trip for a new bit.

Drillers have long used diamond-tipped bits to improve rates of penetration in challenging or pun-ishing environments. First used in coring bits around 1910, diamonds were incorporated into drilling bits by the early 1920s. In the 1970s, synthetic diamonds, bonded on tungsten carbide, led to the devel-opment of fixed cutter polycrystal-line diamond compact bits.

Fixed cutter bit durability has improved with the development of advanced materials and design. However, as horizontal drilling becomes the norm, and the success of many projects hinges on produc-tion from laterals, drillers need a bit that not only stands up to harsh environments but also helps them drill long intervals with good rates of penetration.

To address these demands, designers developed the StingBlade* conical diamond ele-ment bit. Incorporating a unique cutting element across the bit face, the new bit has been used to drill from casing shoe to casing point in a single run, including in formations in which historically such performances have been impossible. The StingBlade bit has also demonstrated an ability to sub-stantially reduce the time required to drill curved sections, providing improved toolface control that allowed the driller to keep within targeted zones. Page 30.

Carbon Dioxide—Challenges and Opportunities

For reasons both positive and nega-tive, carbon dioxide has long held the attention of oil industry experts. This byproduct of many gas wells and industrial processes can severely corrode critical metal parts but can also serve as a key ingredi-ent in enhanced oil recovery.

Today, increased levels of car-bon dioxide [CO2] in the Earth’s atmosphere, much of it the result of burning hydrocarbons, is con-sidered the leading contributor to global climate change. The irony is that many solutions to address-ing rising CO2 levels depend on oil industry techniques. Carbon diox-ide, captured from various large-scale industries, is being pumped down wells and stored deep under-ground; the long-term integrity of both wells and storage areas relies on oil industry expertise. Operators are reinjecting captured CO2 into production zones for enhanced oil recovery projects.

At the Sleipner Field in the North Sea, CO2 is produced with natural gas. The gases are sepa-rated offshore and about 1 million metric tons [1.1 million tonUS] of CO2 per year is reinjected into a disposal interval in the field. Cenovus Energy is using CO2 cap-tured from a synfuel manufacturing plant in North Dakota, USA, to extend the life of its 60-year-old Weyburn-Midale field in Canada while simultaneously storing mil-lions of tons of CO2 underground. Similar projects around the world are resulting in both the produc-tion of millions of incremental bar-rels of oil and the sequestration of millions of tons of CO2. Page 36.

September 2015 Article Summaries

An asterisk (*) denotes a mark of Schlumberger.

E&P Defining SeriesThe Oilfield Review Defining Series provides E&P professionals with concise, authoritative, up-to-date summaries of a wide range of industry topics. The two-page format of each article quickly informs and educates.

Defining Series articles are created by the Oilfield Review team in collaboration with technical experts fromacross the industry.

Exclusively online and available immediately: Defining Electrical Submersible Pumps. To see this and all archivedDefining Series articles, visit http://www.slb.com and search for Defining Series.

Oilfield Review - Schlumberger/media/Files/resources/oilfield_review/ors15/... · Oilfield Review...

Documents

Transcript of Oilfield Review - Schlumberger/media/Files/resources/oilfield_review/ors15/... · Oilfield Review...