Spring 1999

Shear-Wave Seismic Data

Cementing Technology

Production Services

Artificial-Lift Methods

Oilfield Review

MORAMultimedia Oilfield Review Archive

Point your browser to www.connect.slb.com to find:• technical presentations, forums and

publications, including Oilfield Review• detailed service information,

applications and case studies • log interpretation charts and

nonexclusive seismic maps• an oilfield dictionary, mnemonics

database and other oilfield links • location and telephone directory• DataLink* secure data transfer• InterACT* real-time data transmission

from wellsites• SuperVision* project monitoring.

Connect Schlumberger is the Internet communications resource for Schlumbergerclients. Updated daily, this password-protectedinterface is growing rapidly, providing timelyand interactive access to the information, dataand tools needed by the E&P industry.

You can also visit the publicSchlumberger Web site atwww.slb.com.* Mark of Schlumberger

Did you know that Oilfield Review can be found on-line?

MP-6066

Name

Company

Address

City State Postal Code

Country

Telephone

Fax E-mail

Method of Payment: ❏ MasterCard ❏ Visa ❏ Check

Account Number-Please include all digits

VAT number (EEC countries only)

Please print name as it appears on credit card

Expiration Date

Order Date Signature

Statement Address (if different from above)

MORA or Oilfield Review Services Order Form

Viewgraph and Slide sets of Oilfield Review Illustrations

Oilfield Review Back Issues and Articles

Multimedia Oilfield Review Archive

Platform: ❏ Macintosh ❏ Windows ❏ UNIX

Computer Model

Quantity Ordered @$25.00 each = Total $

Back-issue price:$15. Call for article reprint price.

Issue date Price Quantity Subtotal

1.

2.

3.

4.

Call for added cost of delivery. Total $Specify: ❏ Courier ❏ Air mail

ORDER TOTAL

Article title No. of sets Subtotal(call for prices)

1.

2.

3.

4.

Specify: ❏ Viewgraphs: $16 each ❏ Slides: $14 each Total $

Oilfield Review ServicesBarbour Square, High Street, Tattenhall,Chester CH3 9RF England☎ (44) 1829-770569 • Fax: (44) 1829-771354 E-mail: orservices@t-e-s.co.uk

Call or Copy and FAX this form.

$

MultimediaMORA—the Multimedia Oilfield Review Archive—providesquick and easy access to issues published from 1992 to 1998.MORA preserves the look of the printed magazine in a cross-platform format. Full-color articles can be printed or exploredon-screen using key-word searches and cross-document links.

Color 35-mm slides and viewgraphsThe ideal way to prepare high-quality sales presentations andlectures—Oilfield Review Services offers color 35-mm slidesand viewgraphs of illustrations from Oilfield Review and TheTechnical Review. The cost of a slide or viewgraph set rangesfrom $120 to $450. Please call for more information.

Article reprints and back issuesReprints of Oilfield Review articles can be supplied. Minimumorder is 500. Back issues of the Oilfield Review and TheTechnical Review can also be ordered, subject to availability.

Exhibition postersSpecial display materials such as posters or large transparencies,incorporating illustrations from Oilfield Review, can be ordered.

Oilfield Review Services also creates tailor-made multimediapresentations for exhibitions, conferences and seminars, and canprovide staff and equipment to run the presentations.

Browse Oilfield Review Services online at www.t-e-s.co.uk

Multimedia, graphics andreprints fromOilfieldReview

Oil from upper sands onlyOil from lower sands only

Commingled oils A B C D E F G H I J K L

% u

pper

san

d pr

oduc

tion,

ste

amflo

od

0

10

20

30

40

50

60

70

80

90

100

Well

Before steamfloodAfter steamflood

1.0

0.5

1.52.0

2.5

A

B

C

DE

F

G

H

I

J

K

LM

N

O

P

Lower sands(3 wells)

Upper sands(2 wells)Producing

well

Productionwater Steam

Injectionwells

Steamfloodedupper sands

Waterfloodedlower sands

0 ft 300

A

B

C

D

E

F

G

H

I

JK

L

Available from Corporate Express/Ross-Martin

Available from Oilfield Review Services:

Corporate Express Document and Print Management3635 Willow Bend, Suite 300Houston, Texas 77054☎ 1(713)660-7226 • Fax 1(713)660-0362E-mail: rmartin1@idt.net

Oilfield Review

Your electronic journey into the pages of

Oilfield Review begins here.

This archive provides quick and easy access

to Oilfield Review articles published from 1992

to 1998. Select a button to explore articles by

topic, issue or search by keyword.

MORA

About Topics Issues Help Search

NOW A

VAIL

ABLE: 1

992-

1998

MULT

IMED

IA O

ILFI

ELD R

EVIE

W A

RCHIVE

CD

SEE

INSI

DE BACK C

OVER

Properly applying artificial lift is a multifaceted processthat relies on understanding hydraulic, mechanical andelectrical basics. Decision-making depends on experienceand technology—both old and new—to help select thebest lift methods. Even when oil prices are low, installingartificial lift or switching to a different lift system canincrease oil output and economic return. Therefore, it isimportant to minimize well interventions and deferredproduction, reduce installation costs and operatingexpenses, and decrease failure frequencies.

What is the right artificial-lift system—rod, hydraulic or electric submersible pump, or gas lift? There is alwaysdebate about efficiency. However, relative performance isjust one measure. The best system is not one that is mostefficient or has the least failures, but one that maximizesprofit and asset value. Good designs consider the overallimpact on existing or new facilities, bottomhole pressuredifferential and reliability. Published comparisons of depthversus rate and costs for artificial-lift methods are severalyears old. New guidelines are needed to show lift applica-tions, operating parameters, failure rates and costs.

Another consideration is artificial-lift maintenance cost.Direct costs are often obvious, but associated downtimemust also be considered. In offshore or remote locations,workover expenses and deferred production may be significant. Some interventions require $50,000 to morethan $100,000 in special fluids that may damage zones andlimit productivity. Improved downhole shut-in and well-control safety barriers reduce costs and formation damage.

Other developments, such as electric submersible pumpsystems deployed with cable inside coiled tubing, alsoreduce well intervention costs and deferred production insome areas. Combining a second system, like gas lift or jetand progressing cavity pumps, with submersible systemsextends artificial-lift applicability beyond the range of asingle system and can maintain production when primarylift fails. Through-tubing wireline, coiled tubing and cable-retrievable pumps also reduce expenses and downtime.

Can trends that indicate lifting problems be recognized?With more wells and a smaller workforce, it is difficult torely solely on discrete data sampling. Lift methods need tobe monitored continuously to optimize well performance.Monitoring systems that provide real-time data to helpoperators make decisions are essential. These systems usesurface and downhole measurements to determine if problems exist. In high-cost areas, using these data toreduce or eliminate failures justifies monitoring systeminvestments. Once there is confidence in this approach,the industry will move toward closed-loop automation anduse computers to make real-time operational decisions.

An Operator’s Perspective on Artificial Lift

To make key decisions, both failures and well activitiesneed to be tracked. ARCO has used a single database tocollect information about all types of wells from gas orwater injectors to flowing, rod or submersible pump andgas-lift producers. A few wells usually contribute a largeportion of failure costs. By identifying these wells and reasons for the failures, Arco reduced expenses.

Historical data are valuable in selecting lift methods forcurrent operations or future projects. Therefore, the dilemma with any tracking system is how much data areneeded. Minimal, but meaningful data are required. If toomuch detail is provided, it burdens those collecting data.Databases can now be structured so service providers can help input data. Several companies are developing uniform, consistent platforms. It remains to be seen ifagreement can be reached on the ideal level of detail.

Well drilling and construction have changed artificial lift. Horizontal, high-angle and multilateral wells add newchallenges by complicating flow regimes. Slimhole wellsrequire smaller equipment, but are expected to produce atthe same rates as wells with larger casing. Performance ofsmaller equipment has improved, but there are practicallimits based on well output and equipment size. Althoughcapital costs may be reduced, there is risk of lower thanexpected production and possibly increased failure rates.

Those who develop, apply and manage artificial-lift technology must address these limitations and understandtechnical issues. Wells should be designed around lift systems and the data needed to evaluate equipment andreservoir performance. Since many new developmentsaddress specific problems, operators must understandartificial-lift fundamentals and document conditions thatestablish lift-system applications (see “Artificial Lift forHigh-Volume Production,” page 48). Finally, to tap localknowledge and experience, companies and operatingareas must share data efficiently.

John C. PattersonChief Production Engineer ARCO Exploration & Production TechnologyPlano, Texas, USA

John Patterson is responsible for production equipment performance, artificial-lift solutions and cost control for worldwide operations. He has held positionsas production superintendent and supervisor, and project director in Texas and Alaska, USA. A recognized artificial-lift authority, John developed andupdated many ARCO production engineering practices. He earned a BS degreein petroleum engineering from Texas A&M University in College Station andholds six US patents for artificial-lift and facility technologies.

Oilfield Review is published quarterly by Schlumberger to communicatetechnical advances in finding and producing hydrocarbons to oilfieldprofessionals. Oilfield Review is distributed by Schlumberger to itsemployees and clients.

Contributors listed with only geographic location are employees ofSchlumberger or its affiliates.

© 1999 Schlumberger. All rights reserved. No part of this publicationmay be reproduced, stored in a retrieval system or transmitted in anyform or by any means, electronic, mechanical, photocopying, recordingor otherwise without the prior written permission of the publisher.

Address editorial correspondence to:

Oilfield Review225 Schlumberger Drive Sugar Land, Texas 77478 USA

(1) 281-285-8424Fax: (1) 281-285-8519E-mail: OilfieldReview@sugar-land.oilfield.slb.com

Address distribution inquiries to:

Mark E. Teel(1) 281-285-8434Fax: (1) 281-285-8519E-mail: teel@sugar-land.oilfield.slb.com

Oilfield Review subscriptions are available from:

Oilfield Review ServicesBarbour Square, High StreetTattenhall, Chester CH3 9RF England

(44) 1829-770569Fax: (44) 1829-771354E-mail: orservices@t-e-s.co.uk

Annual subscriptions, including postage, are 160.00 US dollars, subject to exchange rate fluctuations.

Executive EditorDenny O’BrienSenior Production EditorMark E. TeelSenior EditorLisa StewartEditorsRussel C. HertzogGretchen M. GillisDavid E. Bergt

Contributing EditorsRana Rottenberg IllustrationTom McNeffMike MessingerGeorge StewartDesignHerring DesignPrintingWetmore Printing Company, USA

Advisory PanelTerry AdamsAzerbaijan International Operating Co., Baku

Syed A. AliChevron Production Co.New Orleans, Louisiana, USA

Antongiulio AlborghettiAgip S.p.AMilan, Italy

Svend Aage AndersenMaersk Oil Qatar ASDoha, State of Qatar

Michael FetkovichPhillips Petroleum Co.Bartlesville, Oklahoma, USA

George KingAmocoTulsa, Oklahoma

David Patrick MurphyShell E&P CompanyHouston, Texas, USA

Richard WoodhouseIndependent consultantSurrey, England

Spring 1999Volume 11Number 1

Schlumberger

48 Artificial Lift for High-Volume Production

Subsurface equipment complexity, surface facility requirements and the energyneeded to bring fluids to surface make high-rate lift expensive to install andoperate. Selecting the right system is important because a single installationmay produce as much as some small fields. Choosing suitable methods is evenmore critical when evaluated in terms of failure, downtime and interventioncosts. This article reviews artificial-lift selection, design and optimization as well as technological advances in gas lift and electric submersible pumps.

16 Concrete Developments in Cementing Technology

The quality and integrity of primary cement jobs often determine how long wells will produce without requiring repair. Slurry properties, including rheology, fluid loss, pumpability and thickening time, traditionally clashed with mechanical properties of hardened cement, such as compressive strength, porosity and permeability. New technology optimizes both slurry and set-cement properties simultaneously while reducing cost and risk, even under difficult operating conditions and in harsh environments.

2 Shear Waves Shine Brightly

Some reservoirs are impossible to see in conventional compressional-waveseismic images, but are clearly imaged on surveys that blend informationfrom shear and compressional waves. A new, seafloor acquisition system captures data from both wave types and establishes the viability of this technique offshore. Case studies show that, in addition to better imaging,combining the two types of wave data provides better interpretations of reservoir lithology variations and helps identify fluid contacts and properties.

30 Keeping Producing Wells Healthy

From first completion, the process of monitoring, diagnosing and interveningcan ensure well viability and productivity. Experienced personnel and the latest technology have been committed to identify and solve production-relatedproblems economically. We discuss developments in production logging andinterpretation through several field examples. Remedial solutions, includingthrough-tubing casing-patch technology, perforating techniques and innovativecombinations of production logging and completion services, are also reviewed.

64 Contributors

66 Coming in Oilfield Review

Oilfield Review Services and MORA Order Form (inside back cover)

Oilfield Review

1

2 Oilfield Review

Shear Waves Shine Brightly

Jack CaldwellHouston, Texas, USA

Phil ChristieFolke EngelmarkSteve McHugoHüseyin ÖzdemirGatwick, England

Pål KristiansenOslo, Norway

Mark MacLeodChevron UK Ltd.Aberdeen, Scotland

For help in preparation of this article, thanks to LorraineClark, John Kingston, Scott Leaney and Tony Probert, Geco-Prakla, Gatwick, England and Lars Sonneland, Geco-Prakla, Stavanger, Norway; and to the Alba fieldpartnership for permission to publish.MultiWave Array, Nessie and Seismos are marks ofSchlumberger.1. Jolly RN: “Investigation of Shear Waves,” Geophysics 21,

no. 4 (October 1956): 905-938.

Imagine a tool that sees through fog and mirrors as well

as it does through glass, and in the dark. This sums up

the improvements that converted shear waves bring to

hydrocarbon reservoir imaging and characterization.

Scientists have known for almost a hundredyears that combining the information from com-pressional and shear waves can provide valuableinsight into earth properties. Resource explo-ration methods have concentrated on compres-sional (P) waves alone, but long ago when theonly seismic sources were earthquakes, shear (S)waves were accorded their due respect. Oftenmore energetic than P waves by orders of magni-tude, shear waves were recognized as the powerbehind an earthquake’s devastation.

In the 70 or so years that the hydrocarbonexploration industry has been applying seismicwaves, it might seem that adequate results havebeen achieved with P waves alone. Many reser-voirs have been discovered and delineated ascompressional-wave surveys have evolved fromtwo- to three-dimensional (2D to 3D) and recentlyadded a fourth dimension with 4D time-lapseseismic techniques.

However, there are several instances inwhich compressional waves from a standard sur-vey do not adequately image a reservoir ordescribe its properties. Gas, even in smallamounts, disrupts P-wave transmission andobscures underlying targets from compressionalsurface seismic view (next page, top left). Whenthe gas is shallow, it can cloud the entire subsur-face. Some reservoirs do not present sufficientimpedance contrast to the overburden, and so donot reflect P waves strongly enough to producean interpretable image (next page, top right). Inareas where the overburden itself is a high-impedance material, such as salt or hard volcanicrock, imaging the underlying reservoir is difficultbecause so little P-wave energy returns to sur-face after transmission twice—down then up—through high-impedance rocks. Waves thatreverberate in the water column, or that reflectmultiple times at the sea-earth interface, canalso distort the seismic image. Bad weather orsurface obstructions such as platforms, pipelinesand other infrastructure can make acquisition ofconventional towed marine surveys impossible,leaving gaps in subsurface coverage.

In addition to these imaging shortcomings,compressional waves often fail to resolve manyreservoir property questions. In many cases, P waves do identify the target, but do not clearlydelineate its extent. This is a common problem instratigraphic traps, where reservoirs pinch outand are replaced laterally by other lithology.Compressional waves may detect lateral varia-tions in reservoir properties, but be unable to dis-tinguish between changes in lithology andchanges in fluid content or pressure. Bright spots

Spring 1999 3

and other amplitude anomalies can be seen in P-wave surveys, but they can sometimes point tohard, tight rock instead of hydrocarbon accumula-tions unless additional information is supplied.Compressional waves can also be less sensitiveto aligned fractures or rock textures that imposeazimuthal variations of velocity, or other types ofanisotropy, in the reservoir or the overlying strata.

There are several ways of addressing thesechallenges. Advances in acquisition, processingand interpretation of P-wave data are all con-tributing to better images and reservoir charac-terization, but none is meeting the challenge theway that shear waves can.

Shear ExcitementExploration and production geophysicists havealready experimented with shear waves in bothsurface and borehole seismic contexts. From theearliest attempts, dating back to the 1950s, geo-physicists concluded that using shear waves forseismic exploration was not practical.1 Thoseexperiments relied on generation of direct shearwaves that reflected back as shear waves, andwere restricted to land surveys. Generation ofdirect shear waves requires a special, orientedsource, one that excites predominantly sidewaysshaking, or shear motion, as opposed to compres-sional motion, which is easily excited by readilyavailable volume injection—explosive or implo-sive—sources. Shear-wave recording requiresmultiple, orthogonally oriented receivers to regis-ter each component of the motion. This has giventhe name “multicomponent” to methods involvingboth compressional and shear waves.

Since the early shear experiments, improve-ments in acquisition technology have led to somesuccessful land shear-wave surveys, but notmany such surveys are run. They are time-con-suming to set up, because each geophone mustbe properly oriented and planted firmly in theground. Vibroseis trucks with special shakingplates are used as sources.

At first glance, shear waves might seem lessuseful than compressional waves for explorationand production (E&P) purposes. After all, shearwaves are known to propagate only in solids, so

they can neither be excited by conventionalmarine seismic sources nor travel through water.In addition, shear waves are almost insensitiveto a rock’s fluid content: shear-wave velocity andreflectivity remain virtually unchanged whetherthe formation contains gas, oil or water.

However, the combination of shear and com-pressional waves has the potential to revealmore about the subsurface than either wave typein isolation. With information on a formation’s P- and S-wave velocities, an interpreter canbegin to pinpoint lithology more readily than withonly pure P-wave data. The velocity ratio Vp / Vs

has a known range for many rock types. Fluidcontent of a formation is more reliably inter-preted with the combination of wave types. A lateral change in P-wave reflection amplitude

along an interface is more likely to indicate fluid-content change than lithology change if the cor-responding S-wave reflection amplitudes areconstant. If the S-wave amplitudes also change,the variation is more likely to mean the rock prop-erties are changing.

While generating direct shear waves doesrequire a special source, generating reflectedshear waves does not. Shear waves are reflectedwhenever P waves impinge on a solid interfaceat any angle other than 90°. The resultant shearwaves are called converted waves, and can occuras reflected or transmitted waves (below). Areflection in which a P wave converts to an Swave is called a P-S reflection, to distinguish itfrom the more familiar P to P reflection, which islabeled P-P.

Streamer Data

Two-

way

tim

e, s

ec

1.8

2.0

2.2

Gas saturation, %0 100

P wave

S waveN

orm

alize

d ve

loci

ty

> A conventional, P-wave reflection seismicimage from the Far East. Subsurface structure is unclear because shallow gas disrupts P-wavepropagation. In a later figure (page 7), shearwaves reveal the structure. Compressional-waveand shear-wave velocities plotted as a functionof gas saturation, show that even the slightestamount of gas causes P-wave velocities todecrease, resulting in troublesome imaging conditions (inset).

Conventional P-Wave Image

> A P-wave reflection image of a prolific NorthSea reservoir. The low impedance contrast atthe reservoir top makes this structure nearlyinvisible to P waves. In a later figure (page 9),this structure will be revealed by the S-waveimage of the reservoir.

IncidentP wave

ReflectedS wave

ReflectedP wave

TransmittedP wave

TransmittedS wave

Vp Vs Vp

> Upon reflection, P waves converting to S waves.

Two-

way

tim

e, m

sec

400

500

600

700

800

400

500

600

700

800

600

700

800

900

1000

600

700

800

900

1000

Offset, ft CDP Offset, ft CDP0 667 1667 2667 1 10 0 667 1667 1 8

P-PPrestack

P-PStack

P-SPrestack

P-SStack

> Synthetic seismograms for the oil reservoir top (green), oil-water contact (blue), and reservoir base (red). Traces modeled fordifferent source-receiver offsets (track 1) show the vertical-incidence P-P trace has no amplitude at the top of the reservoir, butlarge amplitude at the oil-water contact. Once stacked (track 2), the P-P reflection does show some amplitude at the reservoirtop, because traces from far offsets contribute amplitude. Stacked traces are all the same, but repeated for clarity and displayedat different common depth point (CDP) numbers. Stacked amplitudes tend to cancel at the reservoir base. Converted-wave P-Straces computed for each offset (track 3) exhibit large amplitude at the reservoir top for all offsets. After stacking, P-S amplitudesremain high at the reservoir top and base (track 4).

4 Oilfield Review

Dept

h, ft

2000

3000

2500

3500

10,000 11,000 2.00 2.50 4500 7000 0 0.5 22,000 24,000 11,000 13,500

P-wave velocity Density S-wave velocity Poisson’s ratio P-wave impedance S-wave impedanceft/sec g/cm3 ft/sec ft/sec.g/cm3 ft/sec.g/cm3

Shale

Shale

Oil sand

Brine sand

> Logs from an acoustically invisible oil sand reservoir encased in shales. The increase in compressional velocity (track 1) at the top of the reservoir (green) is offset by a decrease in density (track 2) so that acoustic impedance (track 5), or density timesP-wave velocity, of the oil-filled sandstone is the same as that of the overlying shale. The lack of P-wave, or acoustic, impedancecontrast across the interface induces no P-wave reflection. The large contrast in S-wave velocity (track 3) gives rise to a largeS-wave, or shear, impedance (track 6), so S waves will reflect. Poisson’s ratio (track 4) is a function of Vp /Vs ratio, and is sometimes used to interpret rock and fluid properties. Red represents the bottom of the reservoir sand and blue represents the oil-water contact (OWC).

Spring 1999 5

Forward modeling can predict how P and Swaves will reflect at an interface. Compressionaland shear sonic-log measurements and densitylogs are combined to produce an impedancemodel, which can be ray-traced and convolvedwith a basic seismic pulse to create a syntheticseismogram. An idealized example, with ficti-tious but physically realistic log values, showshow the top of a reservoir can be invisible to aconventional P-wave survey but clearly visible toS waves (previous page, top). In this case, theacoustic, or P-wave, impedance—density timesP-wave velocity—of the oil-filled sandstone isthe same as that of the overlying shale. Thus theacoustic impedance contrast across the inter-face is zero. The shear impedance, or densitytimes S-wave velocity, is sharply higher in thereservoir. The P-wave impedance does jump atthe oil-water contact (OWC) deeper in the reser-voir, where, as expected, the change in fluidshardly affects shear impedance.

Synthetic seismograms show how P-P and P-S reflections will react to such a reservoir (previous page, bottom). The vertical-incidenceP-P trace shows no signal at the top of the reser-voir, but there are reflections off the OWC andthe bottom sand-shale interface. At higherangles of incidence, or source-receiver offsetsgreater than about 1500 ft, the top of the reser-voir does support significant P-P reflections. Butwhen reflections from all offsets are stacked, asoccurs in conventional processing, the P-P reflec-tion from the reservoir top diminishes, while theP-S stack shows a clear reservoir reflection.

Shear Waves at SeaWhile it’s true that shear waves will be reflectedand can image otherwise invisible reservoirs,before that can happen, the reflected energymust be recorded by multicomponent receivers.This is already known to be a challenge on land.Knowing that shear waves cannot propagate inwater, how can this task possibly be accom-plished at sea?

The solution lies in recording the shear waveswhile they are still in the solid earth, at theseafloor. Academic institutions initiated seafloorrecording with ocean-bottom seismometers forearthquake and other analyses decades ago.Recently, the recording of P and S waves at theseafloor has been resurrected by the oil and gasindustry, and two methods of multicomponentrecording have been attempted. The first requiresplacing individual recording devices on theseafloor with a remotely operated vehicle (ROV).This method achieves acceptable results, with

high-quality converted-wave data, but mobiliza-tion and operation of the ROV make these typesof surveys expensive.

The second, more tractable, method makesuse of instrumented cables packed withreceivers, similar to the streamers that are towedin conventional marine surveys, but designed tooperate on the seafloor. Geco-Prakla has devel-oped a new marine acquisition technique thataccomplishes this with a rugged multicompo-nent, instrumented cable called the Nessie 4CMultiWave Array system. The cable is laid on theseafloor by the recording vessel, and anothervessel activates sources (above).

The Nessie 4C cable had its origins in a cableof Russian design that contained hydrophonesand geophones and was dragged on the seafloorby a vessel that simultaneously fired sources.Geco-Prakla engineers devised an innovativeway to place the sensors—each containing onehydrophone and three orthogonally oriented geo-phones—inside the cable, distributing theirweight for optimal coupling to the seafloor(below). Each cable houses hundreds of four-component sensors. The four-component designhas given a new acronym, 4C, to marine multi-component acquisition. With four components to

0 to 1000 m

Recordingvessel

3 to 6 km

Sourcevessel

Sourcevessel

600 to 700 m

> Acquisition with the seabed multicomponent system. The cables are laid on theseafloor by the recording vessel, and another vessel activates sources. The source vessel may traverse the cables at any angle, adding flexibility to survey geometry. aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaP P

P P PS

MultiWave Array system

Z

Y

X

Hydrophone

P-wavesource

> The four components of the Nessie 4C MultiWave Array system. Each sensor stationwithin the cable comprises one hydrophoneand three orthogonally oriented geophonesto record both pressure and particle velocity.Compressional waves are recorded primarilyon the hydrophone and vertical geophonecomponent (Z), while shear waves arerecorded mostly on the transverse (Y) andradial (X) geophones.

each sensor package, full particle-motion vectorrecording of all P and S wavefronts is achieved,along with the pressure wavefront familiar fromtowed streamers. The system is different fromordinary ocean-bottom cables (OBCs) that have ahydrophone and a vertically oriented geophonestrapped to the outside, which is incapable ofrecording the full particle-motion vector and haspoor coupling quality.

The new cable is positioned on the seaflooreither by letting the cable drape into place fromthe sea surface, or by dragging the cable from onebottom location to the next. Environmental author-ities have expressed concern that a cable posi-tioned in this way may adversely agitate the softsediments of the ocean-bottom habitat. However,video recordings of Nessie 4C MultiWave Arraycable being deployed for some Gulf of Mexico sur-veys show that the cable disrupts the sedimentsless than does the native sea life.

Geco-Prakla has acquired about 50 marinemulticomponent surveys in different seabed envi-ronments and different water depths. Data qual-ity is high in all environments, even in soft,unconsolidated seafloor sediments. The cableshave operated in water depths reaching 800 m[2624 ft]. Even an irregular seafloor does not

cause problems; sensors are grouped, so oneimperfectly coupled geophone does not affectthe overall response of the group.

The two-vessel seafloor multicomponenttechnique allows acquisition of true zero-offsetdata, unlike the towed streamer technique thatacquires data at receivers that are offset fromthe source location. With two vessels, seismiclines can be shot across, in the direction of, or atany desired angle to the cable lines.

The seafloor position of the receivers shieldsthem from rough weather that can delay acquisi-tion. Compared to towed-streamer surveys,seabed sensor surveys encounter less time wait-ing on weather (above).

Protection from the weather also lets theseafloor cable acquire data of greater bandwidththan from streamers towed near the sea surface.Isolation from surface conditions reduces noiseand improves signal acquisition (below).

6 Oilfield Review

Perc

enta

ge o

f tot

al ti

me

0

20

40

60

80

Oct Nov Dec Jan

Towed streamer

Seabed cable

Feb Mar Apr

> Less time spent waiting on weatherwith the seabed acquisition system.

0 10 20 29 39 49 59 68 78 88 98 107 117 0 10 20 29 39 49 59 68 78 88 98 107

Frequency, Hz Frequency, Hz

3D Towed Streamer 2D MultiWave Array Z Geophone

> Greater bandwidth recorded with seabed system (right) compared to towed streamers (left). Acquisition of the seabed data took place under severe weather conditions that would have precluded any towed-streamer acquisition. [Data courtesy of Saga Petroleum.]

Spring 1999 7

Seeing Through GasIn 1996 the Nessie 4C seafloor cable system wasused for the first time to acquire 2D surveys inthe North Sea for imaging reservoirs shadowedby overlying gas. Since then, about 30 such sur-veys have been acquired by Geco-Prakla, and thegeographic scope has expanded to include theGulf of Mexico and several areas in the Far East.

The power of the technique becomes evidentby comparing seismic lines shot with conven-tional towed-streamer P-P techniques to thoseacquired with the seabed multicomponent sys-tem. The first example replays the earlier imageof a reservoir in the Far East obscured by gas(above). The towed-streamer results are so per-turbed by the gas that no clear image of eitherlayering or faults can be achieved. The multicom-ponent recordings of converted waves clearlyimage the layering and faults.

As important as the detection and delin-eation of faults is the confirmation that there areno faults. The second example shows two por-tions of a P-P reflection line shot over a fault inthe South China Sea (right). Only one image isover the faulted zone, but gas in the subsurfaceclouds the images so that both sections look dis-rupted enough to contain a fault somewherealthough it is not clear where. The line producedfrom the P-S reflections shows the first part ofthe line to be fault-free, at least on the scale ofthe seismic survey, with the fault revealed andimaged well in the second section.

Conventional P-P Image Seabed P-S Image

< Comparison of P-P and P-S images of a FarEast reservoir. The P-P image (left), repeatedfrom the first figure on page 3, shows littledetail compared to the P-S image (right)achieved with the Nessie 4C cable.

> Two portions of a P-P seabed sensor reflection line (top) shot over one fault in the South China Sea.Both portions are disrupted by gas, so pinpointing the fault location is impossible. The P-S images(bottom) from the same seabed line reveal a fault only on the portion at the right.

Seabed P-P Images

Seabed P-S Images

The Invisible ReservoirEvery operating company probably has an exam-ple of a reservoir that is invisible on traditionalseismic sections. In the case of Chevron, 1984appraisal drilling into a North Sea Cretaceoustarget encountered another, shallower, oil reser-voir that had not been detected on seismic sec-tions. Thus the Alba field was discovered(below). The reservoir is a poorly consolidatedEocene turbidite sand channel about 9 km [5.6miles] long, 1.5 to 3 km [0.9 to 1.8 miles] wideand up to 100 m [330 ft] thick.2 How could such areservoir go unsuspected? Sonic logs wereacquired, examined, and found to hold the rea-son. The compressional sonic logs showedbehavior similar to the synthetic log example pre-sented earlier: no P-wave impedance contrast atthe top of the reservoir, but a sizable contrast atthe oil-water contact. Shear sonic logs showed ahigh contrast in shear-wave impedance at the topof the reservoir.

Revisiting the image of the reservoir derivedfrom 1989 towed-streamer data, some indicationof reflected energy can be imagined, but no clearimage presents itself (below right). A few verticalappraisal wells delineated the general limits ofthe field, but its detailed 3D shape remaineduncertain. New wells encountered intrareservoirshales that could cause significant drilling andproduction problems, but the shales were mostlyinvisible in the seismic images. Developmentplans called for horizontal production wells to bedrilled as close as possible to the reservoir top,but the shape of the top of the reservoir was any-one’s guess. To position the horizontal wells opti-mally, without losing a single foot of pay,required a better seismic image of the reservoir.

8 Oilfield Review

> Conventional towed-streamer P-P reflection image of the Alba reservoir (repeatedfrom the second figure, page 3). Some broken reflections are recorded, but there isno clear image.

> Indications that converted waves can image the Alba reservoir. Modeling usingsonic and density logs and original fluid-saturation information (middle) shows thatP waves (top) will reflect off the oil-water contact, while converted waves (bottom)illuminate the top and base of the reservoir.

Aberdeen

Areaof survey

S C O

T L A

N D

Brae

Forties

Albafield

> The Alba field, anoil-filled unconsolidatedEocene sand in the North Sea,discovered while drilling to adeeper target.

Conventional P-P Image

NW SEOWC

A8 A16

Top of sand Base of sand

Oil-Water Contact: Initial Saturation

Converted P-S Image

Streamer Data

Two-

way

tim

e, s

ec

1.8

2.0

2.2

Spring 1999 9

A feasibility study was conducted to deter-mine if the desired reservoir description could beobtained from a 3D survey with ocean-bottomcables. Three objectives were defined for such asurvey. First, the reservoir should be delineatedusing converted waves (previous page, top).Second, long-offset P waves should contributereflection amplitude variation with offset (AVO)information for distinguishing lithologic contactsfrom fluid contacts. Third, reflected P waves fromthe new 1998 survey should be suitable for com-parison with the 1989-vintage data to mapchanges in the oil-water contact brought on bythe four years of production since January 1994.

Prior to acquiring the 3D seabed survey,Chevron acquired 2D seabed test lines with twocontractors. The P-wave seabed results obtainedwith the Geco-Prakla Nessie 4C MultiWave Arrayseabed system were found to be most compara-ble to the 1989 streamer data.

Chevron commissioned a 67-km2 [26-sq mile]3D multicomponent survey, which was recordedin 14 swaths parallel to the 1989 survey in orderto facilitate time-lapse comparison for fluid-contact monitoring. Acquisition took 8 weeks inrough weather during a period of intense fieldactivity. The seismic acquisition crew workedaround divers, construction, platforms, umbilicalsand pipelines—an environment in which astreamer survey would be extremely difficult(right). The 3D converted-wave processing wassteered by information obtained from the 2D test

line and from initial analysis of the 3D recordeddata.3 A four-month time limit was set on the pro-cessing to ensure that interpretable resultswould be ready in time to guide the next drillingdecisions, and after 31⁄2 months the 3D data setwas ready.

For the first time, the detailed shape of theAlba reservoir was revealed. The converted-wave survey imaged the reservoir with astonish-ing clarity compared to the compressional-wavesurvey (left).4 The new images have enabled amore confident interpretation of the reservoirsand and identification of some of the larger

> Surface activity and infrastructure preventing conventional towed-streamer surveys of the Alba reservoir.

> Sighting the Alba reservoir with converted waves. The top, base, lateral extension andeven compartmentalization of the channel sand are imaged in this seabed line that cutsacross the reservoir axis. These features are not seen in the compressional-wave imagefrom streamer data.

2. Newton SK and Flanagan KP: “The Alba Field: Evolution of the Depositional Model,” in Parker JR (ed): PetroleumGeology of Northwest Europe: Proceedings of the 4thConference, vol. 1. London, England: The GeologicalSociety, 1993.

3. McHugo S, Probert A, Hadley MJ and MacLeod MK:“Processing of 3D Multicomponent Data from the AlbaField,” presented at the 61st European Association ofGeoscientists and Engineers Conference and Exhibition,Helsinki, Finland, June 7-11, 1999.

4. MacLeod MK, Hanson RA, Hadley MJ, Reynolds KJ,Lumley D, McHugo S and Probert A: “The Alba Field OBC Seismic Survey,” presented at the 61st EuropeanAssociation of Geoscientists and Engineers Conferenceand Exhibition, Helsinki, Finland, June 7-11, 1999.

Pipeline

Anchor chain

Umbilical

Receiver line

Albaproductionplatform

Accommodationsplatform

Seabed P-S Image

Alba field

Two-

way

tim

e, s

ec

3.4

3.8

4.0

intrareservoir shales. Interpreters feel there ismore information in the shear-wave data than inthe long-offset compressional AVO dataacquired for lithologic interpretation. The newshear-wave information is improving theChevron team’s understanding of the deposi-tional model and the importance of injected sandfeatures. Interpretation and visualization of theconverted shear-wave cube now play a leadingrole in the understanding of the geometry of theAlba reservoir (left).

The ongoing drilling program in the Alba fielddemands a precise representation of reservoirfluid distribution. Since 1993, the field has pro-duced 130 million barrels [20.6 million m3] of oil.Currently, 15 horizontal wells produce 83,000barrels [13,190 m3] of oil per day. As more wellsare drilled close to existing production and injec-tion wells, understanding water movementbecomes critical. Two recently drilled wells haveencountered unpredicted fluid profiles: one wellfound a water-wet sand above a sand with pri-mary oil saturation; in another, a partially oil-sat-urated sand was found below the originaloil-water contact.

Modeling studies showed that a water satu-ration change of 40% or more at the Alba OWCwould cause a decrease in P-wave reflectionamplitude large enough to be seen in time-lapseseismic surveys (left). The seabottom cablesrecorded excellent P-wave data for this purpose.Comparison with the streamer data acquiredbefore production shows changes in the locationand reflection strength of the oil-water contactdue to oil production and water injection (nextpage, top). The observed fluid-contact changesare compatible with predicted changes seen onsynthetic seismic lines generated for the Albasand with fluid saturations estimated from reser-voir simulations.

10 Oilfield Review

< Modeled response of P-P reflections at theAlba OWC before and after production. Beforeproduction (top), the reservoir sand has a flat OWC, and modeled P waves image that contact. After production (bottom), the modeledP-wave image shows changes in the strengthand location of the OWC. The seismic responseto two different shapes and saturation changes (middle) is modeled.

Streamer P-P Data

P-S Data

> Three-dimensional renderings ofAlba P-P and P-S reflections greaterthan a given amplitude threshold.The cube of P-P reflections showsroughly uniform reflection amplitudethrough the 3D volume, while thecube of P-S reflection data clearlymaps the sand body.

Before Production

After Production

Oil-Water Contact: Saturation, September 1998

100 m

OWC

Spring 1999 11

The Alba case may be unique in that it is thefirst reservoir to have been brought to light sodramatically by P-S converted waves, but similarreservoirs are probably still eluding detection.The conditions that gave rise to the low-impedance-contrast deposits of the Alba field arewidespread in the North Sea, and probably else-where in the world (below left). Feasibility stud-ies will show in advance where multicomponentsurveys could have the same impact in pinpoint-ing reservoir location, shape and fluid content.

The Shear Process Seismic sections generated by converted wavesresemble traditional P-wave sections closelyenough to allow most Oilfield Review readers tofollow the similarities and differences describedin the text and figure captions. One might thinkthat S waves are similar enough to P waves toallow data processing developed for the latter tobe applied to the former. In fact, the first 2D mul-ticomponent seabottom cable surveys were pro-cessed with commercial P-wave algorithms thathad been tricked into thinking the shear arrivalswere compressional waves.

Except for a few key differences, converted-wave processing does follow a scheme similarto that of P-wave surveys. However, those dif-ferences require sufficient attention to detail torender each multicomponent processing job dif-ferent from the last.

The three major differences between pro-cessing converted wave surveys and those usingpurely compressional waves are asymmetricreflection at conversion, difference in geometriesand conditions of source and receiver, and thepartitioning of energy into orthogonally polarizedcomponents. The first, asymmetric reflection,arises because shear waves have slower velocitythan compressional waves. This means that the

1989 Streamer1.8

2.1

Two-

way

tim

e, s

ec

1998 Seabed1.8

2.1

Two-

way

tim

e, s

ecOWC

Top of sand

> Time-lapse comparison between 1989 streamer data (top) and 1998 P-Pdata (bottom) acquired with the seabed cable. The observed fluid-contactchanges, including the lack of a clear OWC in the 1998 data, are compatiblewith changes predicted from production simulations.

211

32 30 31205 206 207

213

6 7 9

11 12 13 14 15

19 20 21

28 29 30 1 2 3

38

8 9

17 18

5606

5605

5604

5504 5505A

B

CE F44434241

47 E F4948

53 P Q

M

G

110

113

Aberdeen

Edinburgh

Bergen

Stavanger

NORWAY

GERMANY

UK

DENMARK

THE NETHERLANDS

< North Sea Eocene sediments. The red areaindicates where sediments similar to those ofthe Alba field were deposited, and whereseabed seismic technology may reveal otherlow-impedance-contrast discoveries.

angle at which an S wave is reflected is alsosmaller, so S waves are reflected up moresharply, with near-vertical raypaths (above).

Compressional waves, incident upon a planarinterface, reflect at the angle of incidence.Processing takes advantage of this symmetry inseveral ways. Signal-to-noise ratio is increasedby stacking traces from several source-receiveroffsets centered on one common midpoint (CMP)located halfway between source and receiverand directly above the common depth point(CDP). Another advantage of symmetric reflectionis source-receiver reciprocity. That is, to a firstapproximation, a trace recorded for a givensource-receiver pair is the same as one for whichthe source and receiver have exchanged posi-tions. Many conventional processing steps, suchas normal moveout (NMO) and dip moveout(DMO) corrections, as well as the wavefieldextrapolation processes of trace interpolation,redatuming and migration, rely on reciprocity.

Converted-wave traces may also be stacked,but because of asymmetric reflection, the reflec-tion point, called the common conversion point(CCP), is not halfway between source andreceiver. The conversion point is always dis-placed from the source-receiver midpoint in thedirection of the receiver, whatever the sourceposition. To bin, or put traces at their correct CCP,a special converted-wave correction, called theCCP binning correction, is achieved by a space-and time-variant grouping of traces from a givensource-receiver pair. Then an NMO correctionmust be applied to achieve a proper stack oftraces. However, the difference in velocitybetween the incident P wave and the convertedS wave, along with the different angles of inci-dence and reflection, causes the NMO correctionin converted-wave processing to be nonhyper-bolic. A common occurrence in P-wave process-ing is to simultaneously process traces atcommon offsets, whether receivers are located“ahead” or “behind” the current source position,and to apply the same NMO velocity to all traces

from a given offset. This must be modified forconverted waves, again for reasons of asymmet-ric reflection, because the velocities computedfor negative and positive offsets are different.

The second major difference with convertedwaves is that the source and receiver are at dis-parate levels and in dissimilar materials. Thesource is at the sea surface but the receiver isdeeper, at the seafloor. This is called a differencein datum. The datum is the arbitrary referenceplane on which sources and receivers areassumed to lie to minimize near-surface effects.Most processing algorithms such as NMO, DMOand migration assume that the source andreceivers are at the same datum. When thesource and receivers are at different datums, atime shift and time-variant spatial shift are intro-duced. A correction is required to bring thesource and receivers to the same datum plane.Mean sea level is a natural choice of datum, andis often selected for converted-wave surveys. Forshallow-water acquisition, the spatial shift issmall, so the correction is simply a static shift.For deeper water, redatuming techniques requireextrapolation using the seismic wave equation tocorrect the spatial position.

Differences in near-surface conditions at thesource and receiver locations require further cor-rections in processing. The thin layer of soft sed-iments on which the cables lie has an extremelylow shear velocity. Variation in the thickness orvelocity of this layer from one receiver to the nextcan cause abrupt changes in the traveltime of theshear wave coming up through this layer. Theseeffects are commonly corrected for in land pro-cessing and are known as statics. The traveltimedifferences occur over short distances relative tothe seismic wavelength, they can vary signifi-cantly from trace to trace, and have to beresolved before stack. In the North Sea, shear-wave statics of up to 150 msec one-way timehave been observed. Fortunately, because thesource location can be controlled, statics affectonly the receivers—only the shear-wave portionof the travel path and not the compressional.

The third major difference with convertedwaves is that energy is partitioned into orthogo-nally polarized components, and that is why mul-ticomponent sensors are needed to fully recordthe converted wavefield. When a P wave strikesan interface at nonnormal incidence, the particlemotion of the reflected S wave is polarized in theplane containing the raypaths of the incident Pwave and reflected S wave.5 The three receivergeophones are typically oriented such that onecomponent (Z) is aligned vertically, the second (X)is parallel to the cable, and the third (Y) is trans-verse to the cable in the horizontal plane. In 2Dsurveys, if nothing disrupts the S wave as it trav-els to the receiver, its particle motion will berecorded mainly on the X receiver, and to aninconsiderable extent (but one that increaseswith offset) on the Z component. In practice,especially in 3D surveys, S-wave arrivals will berecorded on more than one geophone. Thisoccurs because the receiver components are ori-ented at some angle to the source-receiver lineor because rock properties cause the particlemotion to change along the raypath between theconversion point and the receiver.

When S waves are recorded on more thanone geophone, the recorded data must be math-ematically rotated into the radial and transversecoordinate system aligned with the source andreceiver before processing can proceed. The sig-nals recorded on the radial component are takento represent the converted-wave arrivals. Anyresidual transverse-component signal, called off-axis energy, indicates the presence of velocityanisotropy. Anisotropy, caused by alignment ofsmall-scale features in rock layers, forces waveparticle motion to adhere to the principal align-ments of the rock, such as horizontal layering orvertical fractures.

Velocity anisotropy in the vertical plane,called TIV anisotropy, occurs when velocity varieswith angle from the vertical. Horizontal layeringis a common cause of TIV anisotropy. It has beenobserved in many compressional-wave surveys,but it sometimes can be neglected in traditionaltowed-streamer data processing. These surveysare not greatly affected by velocity anisotropybecause they are generally restricted to P-waveray angles of 30 to 40°, and most of the effect ofanisotropy occurs between 40° and horizontal. Inaddition, since particle motion is close to thedirection of wave propagation, only one directionis involved. TIV anisotropy becomes most notice-able in towed-streamer surveys at long source-receiver offsets if horizontal P-wave velocitiessurpass vertical velocities by more than about10%—a common occurrence. Compressional-

12 Oilfield Review

IncidentP wave

ReflectedS wave

ReflectedP wave

> Near-vertical raypathsof converted-wavereflections. In this configuration, shearwaves arriving at the surface will berecorded on the radial,or X-component, geophone.

Spring 1999 13

wave velocity anisotropy in some shales canreach more than 30%, and that of shear wavescan be of the same order of magnitude.

Seabed multicomponent converted-wave sur-veys are more likely to be affected by TIVanisotropy, because these surveys use raypaths atlarge angles to the vertical. In addition, horizontalparticle motion of the vertically traveling S wavebrings an additional direction into the picture andincreases the sensitivity of the shear wave tovelocity variation with azimuth, or angle, in thehorizontal plane. This is called TIH anisotropy, andcan be caused by vertical fractures.

The influence of TIH anisotropy shows up asshear-wave energy recorded on both the X and Ycomponents, even when the X component isaligned with the source-receiver axis. The effectscan be pervasive, complicating computation ofconversion point location, stacking, correction ofreflection times for dip (dip moveout, or DMO)and migration, which positions reflections attheir correct location in time and space.

In most converted-wave surveys run to date,as with P-wave surveys, anisotropy has beenconsidered a second-order problem and not thegoal of the survey. However, detection and quan-tification of anisotropy can yield important reser-voir characterization information, pointing todominant stress directions, small-scale layering,fracturing or other internal alignments that may affect imaging, drilling or production.6

Multicomponent surface and borehole seismicsurveys on land have already been used to estimate fracture directions by measuringanisotropy. Multicomponent seafloor surveys aresure to follow.

Converted-wave survey processing mustincorporate all these aspects of asymmetricreflection, source-receiver differences and multi-component polarization into the processingchain, and special multicomponent algorithmshave been developed for this purpose. In addi-tion, algorithms have been devised for special-purpose processing, which can include:angle-dependent summation of hydrophone and

vertical geophone signals to produce better P-P data than from hydrophones alone; velocityinversion for lithologic interpretation; andanisotropic prestack depth migration for imagingcomplex structures.

Geco-Prakla engineers have placed con-verted-wave processing steps into the commer-cial Seismos data-processing software. TheSeismos system allows faster, more efficientdata transfer from field to office, leaving lessroom for error between acquisition, processingand interpretation. It also enables concurrentprocessing in which the effort is shared betweenfield, processing center and client office. Dataquality control and preliminary processing allowdelivery of stacked data as soon as acquisition is complete. However, since each survey isacquired in a different environment and has a different objective, final processing requires individual attention.

In Addition to ImagesIn most cases so far, the objective of the multi-component seabed survey has been to produce aseismic image where conventional technologyhas failed. However, converted waves can domore: in addition to supplying images of subsur-face layers, converted waves give informationabout the rocks and fluids within the layers.

Compressional waves have been used ashydrocarbon indicators, but with mixed success.In some regions, bright spots, or reflections ofanomalously high amplitude, have pointed to oiland gas accumulations. Unfortunately, the bright-spot technique is unreliable. High amplitudesassociated with tight zones or hard rock can havethe same appearance as those from oil and gas,leading to some disastrously expensive dry holes.

One way to characterize the cause of a high-amplitude reflection is to examine the behaviorof the reflection amplitudes as the source-

receiver offset changes. Amplitude variation withoffset analysis can be a powerful tool for distin-guishing fluid-content changes from lithologychanges.7 The method requires acquisition ofcompressional-wave data at a variety of offsets,the range of which can be determined by a feasi-bility study during the survey planning stage.Then follows careful processing to achieve highsignal-to-noise level without stacking, to pre-serve true relative amplitudes. The observed AVOsignatures are then compared to modeledresponses to infer pore-fluid type.

Of course, AVO analysis is an indirect methodto obtain shear information from P-P reflectiondata, since the AVO signal is often parameterizedin terms of the Poisson’s ratio contrast betweenthe caprock and the reservoir. At first glance, itmight seem that there should be a direct linkbetween AVO and P-S reflectivity: when P-Simaging works, so should AVO and when AVOworks, so should P-S imaging. However, the linkcan be more subtle and not quite so direct.Because Poisson’s ratio is a function of the Vp / Vs

ratio, the AVO response is to this ratio, while P-Sreflectivity responds to contrasts in shearimpedance. The Alba field is one example of areservoir top displaying greater effect on P-Sreflectivity than on AVO.

Another way to discriminate fluid changesfrom lithologic changes is to take advantage ofthe additional information contained in convertedwaves. An example from the North Sea showshow the comparison between P-wave and S-wave response at a reflector can help identifyfluid content of the formation. Nessie 4CMultiWave Array seabed cables were used toacquire P-P and P-S images over a speculatedtarget formed by dipping beds abutting a poten-tially sealing fault (below). Compressional-wavereflection amplitudes are high as the layersapproach their termination with the fault, but the

P-P Reflection Amplitude P-S Reflection Amplitude

> Compressional-wave (left) and converted-wave (right) images of target dippingbeds. Analysis of amplitude variations can help map fluid content in the dipping layer.

5. Geophysicists sometimes call the reflected wave an SVwave, because for horizontal reflectors, the polarization isin the vertical plane containing the source and receiver.Another kind of reflected shear wave, called the SHwave, with particle motion polarized in the horizontalplane, can exist. The SH reflection is generated by incident SH waves—the kind excited by shear-wave vibrator sources—or by shear-wave splitting caused byanisotropy or other features with a preferred orientation.

6. Armstrong P, Ireson D, Chmela B, Dodds K, Esmersoy C,Miller D, Hornby B, Sayers C, Schoenberg M, Leaney Sand Lynn H: “The Promise of Elastic Anisotropy,” Oilfield Review 6, no. 4 (October 1994): 36-47.

7. Chiburis E, Franck C, Leaney S, McHugo S and Skidmore C:“Hydrocarbon Detection with AVO,” Oilfield Review 5,no. 1 (January 1993): 42-50.

high amplitudes could have a number of causes.The objective of the multicomponent survey wasto see if converted-wave reflections would com-plement the P-wave data and help map fluid con-tent in the dipping layer.

To the unaided eye, the color displays of P-Pand P-S reflection amplitude look similar (below).Analysis of the recorded amplitudes by reflectiontracking software shows a subtle difference inthe trend of P-P compared to P-S reflection ampli-tudes. At the updip end near the fault on theright, the normalized P and S amplitudes are rel-atively constant and equal. Downdip, somewherebetween traces 1500 and 1600, the amplitudetrends diverge. The P-S amplitude remainsroughly constant, while P-P amplitude increases.

Interpreters associate the divergence ofthese amplitude trends with a change in fluidcontent. A change in the fluid content of a rocklayer will produce a change in P-wave velocity ofthat layer, and a corresponding change in theamplitude of a P-P reflection off the top of thelayer. The same change in fluid content will havenegligible effect on the S-wave velocity of thelayer, so will not appreciably change the P-Sreflection amplitude. The nature of the fluids con-

tained in this North Sea example might be pre-dicted through forward modeling, but cannot beconfirmed until drilled.

All the Way to InversionThe new proficiency in acquiring multicomponentdata offshore is just one of the ways the E&Pindustry is working to extract more informationfrom shear waves. Even more value can bederived from multicomponent surveys whenborehole seismic and log inputs are integrated.At the earliest stages of survey feasibility, bore-hole seismic and log data can help model seismicwave response and plan acquisition. Three-component vertical seismic profiles (VSPs) areroutinely separated into up and down compres-sional- and shear-wave sections. Because thesesurveys record data at a common depth scale,questions of P- and S-event correlation, waveletpolarity and conversion strength are resolved.Compressional and shear velocities can be ana-lyzed and AVO effects may be calibrated. Shearsonic measurements complemented by multi-component VSPs are also used to detect andquantify velocity anisotropy in the vicinity of theborehole. With multicomponent surface seismic

data, this important information might beextended from the borehole to quantifyanisotropy at the reservoir scale.

One means of combining borehole informa-tion with compressional surface seismic datainvolves inverting the seismic reflectivity data toobtain acoustic impedance. Borehole-measuredacoustic impedance is introduced as a constraintto the inversion process. Through calibration ofacoustic impedance with logged porosity values,the resulting acoustic impedance sections maybe interpreted as porosity sections.8

This technique could also be applied to shear-wave surface seismic data. The method has beentested in the Middle East in a pilot survey over a portion of the Unayzah reservoir. The objectiveof the experiment was to demonstrate the feasibility of predicting lithology—specificallydistinguishing sand from shale—using multi-component seismic data. Studies of the availablewell data in the area indicate a good correlationbetween high sand/shale ratio and low Vp / Vs

ratio. After calibration with borehole sonic andgamma ray logs, the Vp / Vs ratios mapped fromsurface seismic data might be used as reservoirquality indicators.

14 Oilfield Review

P-S Image

p

54

3

2

10

1100 1200 1300 1400 1500 1600 1700 1800

P-P Image

P-S

P-P

P-P and P-S Amplitudes

Trace Number

> Displays of P-P (top) and P-S (middle) reflection-amplitude images from dipping layers, flattened on the target horizon.Reflection amplitude plotted versus trace number (bottom) shows relatively equal P and S amplitudes at the updip edge on the right that diverge, probably due to changes in fluid content, downdip to the left.

Spring 1999 15

The land survey consisted of three multi-component 2D surface seismic lines and five multicomponent VSPs, all acquired with com-pressional and shear vibrator sources. A numberof wells along the 2D lines penetrated the targetreservoir, and full suites of logs, including shearsonic logs and shear check shots, provided ade-quate well control. The seismic recordings wereprocessed to produce seismic sections contain-ing P-P and SH-SH reflections, and then invertedfor impedance, where the shear impedance cor-responded to that of the SH modes. Each trace inthe P-impedance section was divided by the cor-responding trace in the S-impedance section;density was cancelled out; and a Vp / Vs trace wasderived. The average Vp / Vs over the reservoirthickness can then be compared with log data indry and producing wells to see where the seis-mic-based Vp / Vs indicates there may be dry orproductive zones elsewhere on the seismic line.Interpreters are evaluating the results and willpresent findings in the near future.9

Shear AdvancesThe arrival in 1996 of multicomponent surfaceseismic technology in the offshore environmenthas, in a few short years, solved many of theproblems for which it was designed. Aspromised, it successfully images subsurface lay-ers previously obscured by shallow gas, some-thing no other surface seismic technique hasaccomplished. It can also acquire seismic data inareas of surface obstructions where streamerscannot be towed. Multicomponent methods canimage low-impedance reservoirs, register time-lapse data for monitoring fluid-contact displace-

ment, resolve changes in saturation and reservoirpressure, indicate fluid content and characterizelithology and other reservoir quality parametersbetter than methods using P waves alone.

The seabed multicomponent technique doeshave some limitations, however. Several servicecompanies operate seabottom multicomponentcable systems, each with different advantagesand disadvantages. Although there is no inherentlimit to the water depth at which the technologycan be applied, cable handling, strength specifi-cations and positioning capabilities drive practi-cal constraints. Surveys have been acquired inwater depths slightly exceeding 2000 m [6560 ft].As water depth increases, handling the longercables can require special winches or capstansfor tension reduction. Deep water also adds diffi-culty to tracking the position of the cable: tem-perature and salinity variations alter soundvelocity in water, affecting pinging systems usedfor pinpointing cable location.

To date, Geco-Prakla surveys have beenacquired with one or two Nessie 4C cablesdeployed on the seabed. In the future, acquisitioncould become more efficient with more cablesactive. The time, equipment and expertise neces-sary to deliver a high-quality seabed cable surveycommand a premium price, and currently off-shore 4C surveys cost from 1.5 to 4 times that ofa towed-streamer survey. As more such surveysare run, acquisition experience will continue togrow and the cost will decline.10

As more types of surveys are acquired andprocessed, expertise in processing will grow.Time-based processing applied today givesresults in time, not depth. Since S waves travelabout half as fast as P waves, reflections on S-wave time sections appear about twice as“deep” as reflections on P-wave sections. Forquality control of converted-wave processing, itis useful to adjust the S-wave section with atime-varying correction to tie and compare majorreflectors to the P section. To further complicatematters, what may appear as a positive event ona P section may show up as a negative event onan S section, and some events on S sections maynot even appear on P sections. For detailed inter-pretation purposes, more careful correlationtechniques may be necessary before reflectorsimaged with S waves can be compared with P-wave images.

Ultimately, converted-wave surveys, like P-wave surveys acquired to image complex struc-tures, may need depth-based processing. Thisyields images on a scale of depth instead of time.Depth-based processing requires construction ofa velocity model of the subsurface, and researchis being done to find efficient ways to do this forS waves. Advances in model-based processingare expected to help. Application of depth-basedprocessing will also be able to accommodateanisotropic velocity models, crucial to the pro-cessing of converted waves. When these pro-cessing methods are ready, there may be furtherapplication of converted-wave surveys, such asclearer imaging below salt.

To support the evolution of multicomponentmethods from their current level to one of morewidespread use requires advancing interpreta-tion techniques to get the most out of the data.The key to new interpretation techniques will beintegration of data from many sources. Oneapproach under development is to use 4C seabeddata to enhance the value of conventional 3Dtowed streamer data. In a case study from theDanish sector of the North Sea, a technique hasbeen tested in which Vp / Vs ratios were com-puted from converted-wave sections. Theseratios were used to calibrate the lithology effecton the 3D-survey AVO response so that any resid-ual AVO effect could be attributed to pore fluids.11

Integration of borehole seismic and log infor-mation will also contribute to improved con-verted-wave interpretation. This integration canprovide valuable input for forward models thatare key to lithology and fluid characterization. Inaddition to velocity and density, two propertiesparticularly important for converted-wave appli-cations are anisotropy and attenuation. Theseproperties are difficult to measure accuratelyfrom surface seismic data alone, but are rela-tively easy to measure with borehole seismicdata. Schlumberger geophysicists are investigat-ing ways to include this information to constructmodels for more accurate simulation.12

As more multicomponent surveys are run,more applications will be found. If there is aproblem that P waves can’t solve, technology isavailable to test the feasibility of solving it withshear waves. Some geophysicists believe theimpact of multicomponent methods will be asgreat as that resulting from the move from 2D to3D seismic surveys in the 1980s. What is certainis that we don’t yet know the limits of what theE&P industry can gain from shear waves. —LS

8. Ariffin T, Solomon G, Ujang S, Bée M, Jenkins S, Corbett C,Dorn G, Withers R, Özdemir H and Pearse C: “SeismicTools for Reservoir Management,” Oilfield Review 7, no. 4 (Winter 1995): 4-17.

9. Macrides CG, Kelamis PG, Marschall R, Potter G andGunaratnam K: “Lithology Estimation of a PermianClastic Reservoir Using Multicomponent Seismic Data,” presented at the 61st European Association ofGeoscientists and Engineers Conference and Exhibition,Helsinki, Finland, June 7-11, 1999.

10. Caldwell J: “Marine Multicomponent Seismic-Acquisition Techniques,” paper OTC 10981, presented atthe Offshore Technology Conference, Houston, Texas,USA, May 3-6, 1999.

11. Sønneland L, Veire HH, Hansen JO, Hutton G, Nickel M,Reymond B, Signer C and Tjøstheim B: “ReservoirCharacterization Using 4C Seismic and Calibrated 3D AVO,” presented at 60th European Association ofGeoscientists and Engineers Conference and Exhibition,Leipzig, Germany, June 8-12, 1998.

12. Leaney S, Cao D and Tcherkashnev S: “CalibratingAnisotropic, Anelastic Models for Converted WaveSimulation,” presented at the 61st European Associationof Geoscientists and Engineers Conference andExhibition, Helsinki, Finland, June 7-11, 1999.

Oilfield Review

Concrete Developments in Cementing Technology

Jean Marc BoisnaultDominique GuillotMontrouge, France

Abderrahim BourahlaTimothy TirliaAnadarko Algeria CompanyHassi Messaoud, Algeria

Trevor DahlPanCanadian Petroleum Ltd.Calgary, Alberta, Canada

Chris HolmesA.M. RaiturkarPetroleum Development OmanMuscat, Sultanate of Oman

Pierre MaroyClamart, France

Charles MoffettHunt Petroleum CorporationJena, Louisiana, USA

Genaro Pérez MejíaIgnacio Ramírez MartínezPetróleos MexicanosVillahermosa, Mexico

Philippe RevilHouston, Texas, USA

Robert RoemerAberdeen, Scotland

Perhaps the most difficult borehole fluid to handle, cement is critical to the

performance and life of a well. Optimal slurry properties for placement of

standard oilfield cements typically do not coincide with optimal mechanical

properties of set cement necessary for long-term zonal isolation. New

technology optimizes both slurry and set-cement properties simultaneously.

Since a flawlessly cemented wellbore protectsthe conduit that links reservoir fluids to the sur-face where they are used, high-quality oilfieldcement is an essential ingredient in any success-ful well. The quality and integrity of a cement jobcan determine how long a well remains stableand productive without requiring repair. In addi-tion to promoting ongoing operational safety andsuccess, today’s cements must also be designedwith cost savings and challenging operating envi-ronments in mind. Environmental protection is agreater concern than ever, especially protectionof shallow aquifers during and after drilling. Agood primary cement job is essential becauseremedial cementing (squeezing) is difficult toaccomplish and provides only temporary, localzonal isolation—it is preferable to do the job cor-rectly the first time. Overcoming the trade-offbetween cement slurry properties, includingrheology, fluid loss, pumpability and thickeningtime, and mechanical properties of set cement,such as compressive strength, porosity and per-meability, is a major challenge.

Traditional Cementing ApproachesThere are several fundamental purposes for plac-ing cement in oil and gas wells. Cement is usedto support the casing. In addition, it hydraulicallyisolates the various formations the well pene-trates, thereby protecting aquifers and prevent-ing fluid flow from high-pressure to low-pressureformations, which might result in a loss of hydro-carbon production or excessive water production.Cement guards against fluid broaching to the sur-face, which could lead to a catastrophic blowout.Cement also protects the casing from corrosionby chemically aggressive brines.

In the past, the least expensive material andtechnology—typically displacing drilling fluids by pumping Portland cement behind casing—were acceptable in all but the most difficultcases. Portland cement mixes easily with waterto produce a slurry that is readily pumpable andcan be placed anywhere within hydrostatic pres-sure constraints of a wellbore. Prepared at therecommended water-to-cement ratio, Portlandcement fulfills the most important objective,hydraulically isolating the formations. Further-more, Portland cement is readily available world-wide and is inexpensive.

For help in preparation of this article, thanks to AndrewAcock and Kevin England, Dowell, Houston, Texas, USA;Tyler Bittner, Walter Chmilowski and Mike Roy, Dowell,Calgary, Alberta, Canada; Leo Burdylo, Oilfield Services,Sugar Land, Texas; Erling Prado-Velarde, Dowell,Villahermosa, Mexico; Tarek Ramadan, Dowell, Muscat,Sultanate of Oman; and Eugene Toukam, Dowell, HassiMessaoud, Algeria.CemCADE, CemCRETE, DeepCRETE, DensCRETE, DESC(Design and Evaluation Services for Clients), FLAC (fluid-loss additives for cement), GASBLOK, LiteCRETE,SqueezeCRETE, USI (UltraSonic Imager) and VariableDensity are marks of Schlumberger. Ping-Pong is a mark of Parker Brothers, Inc.

16

Spring 1999 17

The usual method for placing a slurry in a wellduring primary cementing operations consists ofpumping a series of fluids down the casing whilethe fluid already in the well—the drilling mud—flows out the casing-formation annulus to sur-face. The first fluid pumped is usually a preflushor spacer, or both, that separates the drilling fluidfrom the cement slurry. The spacer must be com-patible with both the drilling fluid and the slurry,yet keep those fluids apart to preclude contami-nation of the slurry by drilling fluid. Such con-tamination degrades the quality of the setcement. This is followed by as many as fourslurries. The preflush-spacer-cement series mustdisplace from the annulus all fluids ahead of it toprevent development of mud channels within thecement sheath.1 Such channels allow formationfluid migration. The presence of mud can alsonegatively affect set-cement properties, for

example by inducing shrinkage cracks, reducingcompressive strength or increasing permeability.A mechanical plug is then launched into the cas-ing and displaced to the bottom of the well byanother fluid, typically the drilling fluid needed todrill the next section of hole. At the end of theoperation, the cement occupies the annularspace between the casing and the penetratedformation from the bottom of the hole up to thedesired level.

During the cementing operation, the criticalgoal is to maintain the pressure in the annulusbetween the pore and fracture pressures of thepenetrated formations at all times and all depthsthroughout the openhole interval.2 If the annularpressure becomes lower than the formation porepressure, fluids can flow into the annulus andlead to a potentially catastrophic situation, ablowout. At the other extreme, if the annularpressure becomes higher than the formation frac-ture pressure, then annular fluids can split thesurrounding rock, damaging the borehole andescaping into the formation.

The first factor affecting annular pressureduring drilling or cementing operations is thedensity of the fluids, which exert hydrostaticpressure on the exposed formations. The secondfactor, fluid rheology, governs the frictional pres-sures during placement. Though density is aparameter that can be controlled easily duringthe design and operation phases, the actualrheology of a fluid is more difficult to control ormodify. Once these properties have beendesigned properly for a given operation, such aswith CemCADE cementing design and evaluationsoftware or other simulators, it is important thatthey be maintained within reasonable toler-ances during the entire placement operation.The cement slurry must be stable—solid parti-cles that are denser than the water in which theyare suspended must not separate from the liquidduring either static or dynamic conditions. The

1. Bonett A and Pafitis D: “Getting to the Root of GasMigration,” Oilfield Review 8, no. 1 (Spring 1996): 36-49.

2. Aldred W, Cook J, Bern P, Carpenter B, Hutchinson M,Lovell J, Rezmer-Cooper I and Leder P: “Using DownholeAnnular Pressure Measurements to Improve DrillingPerformance,” Oilfield Review 10, no. 4 (Winter 1998): 40-55.

slurry must not lose excessive interstitial waterto the formation when the pressure in the annu-lus is higher than in the formation. Excessivefluid loss from a slurry can increase the viscosity,which might result in incomplete placement ofthe slurry and bridging of the annulus, and canalso lead to volume reduction in the cement, pro-ducing channels or other defects.3 Finally, theslurry should not thicken or set prematurelyduring placement.

Performance of conventional cement slurriesultimately is a function of many variables, includ-ing the amount and types of solids, water, chem-ical additives, temperature and pressure.Weighting agents increase density; extendersdecrease it. Dispersants control rheology bybreaking larger particles into smaller ones, whichcan reduce viscosity. Stability is either intrinsic tothe design, or improved by using free water con-trol or solid-suspending agents (antisettlingagents). Fluid-loss control is achieved by addingFLAC fluid-loss additives for cement. Retarders oraccelerators control thickening time. Clearly,chemical additives define the performance ofPortland cement slurries.

Once in place, the cement slurry should setquickly and develop adequate strength to mini-mize the time spent waiting on cement (WOC) sothat the operator can proceed with the nextphase of well construction as soon as possible.

Limitations of Conventional Cementing TechnologyGood slurry and set-cement properties are mutu-ally exclusive in many conventional cementingsituations. For example, standard high-densitycements, while necessary for well control inhigh-pressure drilling, are difficult to pump andprone to sedimentation as weighting agents set-tle out of suspension. Low-density slurries withproportionately higher liquid volumes developcompressive strength slowly and attain low finalcompressive strengths, limiting their value whencementing production casing. Although chemicaladditives are crucial to successful cementingoperations, the ultimate performance of conven-tional cement systems is dominated by thewater-to-cement ratio.

The optimal water-to-cement ratio is about44% by weight for a low-viscosity, stable slurryof API (American Petroleum Institute) Class Gcement, one of the most commonly used Portlandcements in the oil field. This gives a density ofaround 15.8 lbm/gal [1900 kg/m3]. Higher densi-ties can be reached by either decreasing thewater-to-solid ratio or increasing the density ofthe solid blend at a given water-to-solid ratio.When the water-to-cement ratio is close to theoptimal value, the better choice is to reduce theamount of water; but this quickly leads tounpumpable or unmixable slurries. At that point,the only option is to add weighting materials tothe cement, normally high-density minerals suchas barite, hematite (the most common weightingagent) or ilmenite. The densities of these miner-als are 35 to 43 lbm/gal [4200 to 5200 kg/m3],whereas the density of Portland cement is about27 lbm/gal [3200 kg/m3].

To achieve lower densities, the methods arereversed: either increase the water-to-solid ratioor add lightweight aggregates. Another possibleoption is foaming the slurry with gas—usuallynitrogen or air. When the water-to-cement ratioapproaches the optimal value, the simplestapproach is to add more water to the slurry, butthis jeopardizes its stability, reduces the strengthof the set cement and increases porosity and per-meability. To rectify stability problems, intersti-tial water can be viscosified using colloidal clays(bentonite or attapulgite), sodium silicates orhydrosoluble polymers. However, these cementsystems exhibit higher porosity and permeabilityonce set, which often precludes their use in crit-ical casing strings. Another technique consists ofblending Portland cement with lighter solid mate-rials such as diatomaceous earth, perlite, fly ash,fumed silica, blast furnace slag or hollow micro-spheres. This method works only in relativelynarrow density ranges where the water-to-solidratio is maintained above a given threshold forthe slurry to be mixable and pumpable.

A further problem with standard cementsystems is that remediation of unsatisfactory pri-mary cement jobs is difficult. Squeeze cementing,even when performed satisfactorily, merely pro-vides a temporary patch. Conventional cementsare difficult to place in small defects, such aspartially plugged perforations and damaged cas-ing, because of their relatively large particlesizes and poor injectability.

18 Oilfield Review

> Particle-size optimization. A slurry made from particles of a single size (left) contains larger water-filled spaces than a slurry made from an optimized blendof several particle sizes (right). The smallest particles fill the spaces between larger particles and function much like lubricating ball bearings.

3. Bonett A and Pafitis D, reference 1: 38.4. For more on primary cementing: Fraser L, Stanger B,

Griffin T, Jabri M, Sones G, Steelman M and Valkó P:“Seamless Fluids Programs: A Key to Better Well Con-struction,” Oilfield Review 8, no. 1 (Spring 1996): 42-56.

Spring 1999 19

Operators seek cementing materials that notonly are easier to place the first time, but alsooffer the best long-term performance. Cementsthat achieve compressive strength earlierreduce waiting time and increase efficiency.Because drilling a well is typically the culmina-tion of months or even years of intensive effort,including the acquisition and interpretation ofseismic data and planning well construction, itis critical to achieve 100% cementing successat the outset.4

Concrete ImprovementTypically, cements are weighted without consid-eration for the particle sizes of the ingredients(primarily cement and weighting agents). As therequired density increases, conventional addi-tives alone quickly lead to either an unpumpableor unmixable slurry if the solid-to-liquid ratio istoo high, or to a system that does not containenough cement to develop a reasonable strength.A new system, CemCRETE technology, is con-crete-based slurry technology to optimize slurryperformance during placement while ensuring ahigh set-cement quality. By adjusting the parti-cle-size distribution (PSD) of the different solids,this technique uses more solid particles in agiven slurry volume while keeping slurry rheologyreasonably low. This allows slurries with densi-ties as high as 24 lbm/gal [2900 kg/m3] to be

used to cement critical casing strings in wellswith high pressure gradients.

Because many traditional cement slurrieshave single-size particles, they can be visual-ized as a box full of Ping-Pong balls (previouspage). Between each ball, there are large air-filled voids. In a real slurry, the void spacewould be filled with water rather than air. In ahigh-performance slurry with engineered PSDoptimization, particles of three or more differentsizes are carefully selected. A box of Ping-Pongballs with green peas and grains of sand fillingthe voids is crudely analogous to a trimodal PSDCemCRETE system.

By adjusting the PSD of the solids in theblend, CemCRETE technology increases thesolids per unit volume of slurry above that of Portland cement slurries. This increasescompressive strength and reduces porosity andpermeability by achieving a higher packingvolume fraction (PVF) independent of slurrydensity. Packing volume fraction is defined as the ratio of the sum of the absolute volumesof all particles in the dry blend divided by the bulk volume of the dry-blend components. HigherPVF values generally indicate better set-cementproperties. For example, hexagonal packing ofidentical spheres results in a PVF of 0.74, butrandom packing of the same spheres achieves a PVF of 0.64. The packing volume fraction of

an optimized dry blend is increased by using a trimodal PSD, which in turn decreases set-cement permeability (below left).

Because the remaining fluid content is usedmore efficiently, CemCRETE technology usuallyrequires lower concentrations of most chemicaladditives compared to traditional approaches.Gas-migration technology is more easily appliedbecause of the lower water-to-solid ratio andbecause of the lower permeability and porosityof the cement slurry during the transition fromliquid to solid as the cement sets. The 35 to 45%porosity, or water content, of the new high-performance slurries is significantly lower thanthe average 55 to 75% porosity for standardslurries (below right).

In contrast to conventional Portland cement,state-of-the-art cements contain a specific blendof particles engineered for each specific slurrydensity. The PVF of the optimized blends com-monly exceeds 0.80. The high solids contentresults in stable systems that disconnect theslurry density from rheology, require few addi-tives and are easy to mix and place in operationsthat are as simple as ordinary jobs yet require nospecialized equipment. These systems exhibitlow porosity and permeability once set, even for slurry densities as low as 10 lbm/gal [1200 kg/m3]. More simply stated, physics suc-ceeds where chemistry often fails.

Extendedlightweight

cement

15.8-lbm/galClass Gcement

CemCRETEcement

0

0.05

0.10

0.15

0.20

Perm

eabi

lity,

mD

> Set-cement permeability. Permeabilities to gasof set conventional and extended lightweightcements can be as high as 0.20 mD. (“Extended”lightweight cements have high porosities, typi-cally 75%, because the slurry density is loweredby increasing the water-to-cement ratio.) Thegranulometric optimization of CemCRETE blendsresults in set-cement permeability below 0.05 mD.

12.5-lbm/galExtended

lightweightcement

15.8-lbm/galClass Gcement

10- to 24-lbm/gal

CemCRETEcement

0

12

10

8

6

4

2

14

77%porosity

59%porosity

40%porosity

Mix

wat

er n

eede

d, g

al/s

ack

> Slurry porosity. High water content, or porosity,of a cement slurry improves its pumpability, butcan lead to sedimentation in the slurry andhigher permeability and lower compressivestrength once the cement sets. Conventionalslurry porosities range from 55 to 75% or more,whereas CemCRETE slurry porosities are typi-cally 35 to 45%. Sized particles in the optimizedblend ensure high strength in the set cement andgood slurry rheology despite low water content.

20 Oilfield Review

DensCRETEcement

Conventionalcement

DensCRETEcement

Conventionalcement

DensCRETEcement

Topsectiondensity

Middlesectiondensity

Bottomsectiondensity

17.8lbm/gal

18.1lbm/gal

17.5lbm/gal

18.7lbm/gal

19.3lbm/gal

19.6lbm/gal

18.7lbm/gal

20.7lbm/gal

20.9lbm/gal

18lbm/gal

18lbm/gal

19.5lbm/gal

19.5lbm/gal

21lbm/gal

21.2lbm/gal

> Sedimentation and segregation. In the BP settling test, a column of set cement cured under controlled pressure andtemperature is cut into sections and the density of each cylindrical section is measured. High-density conventionalcements tend to show greater vertical density variation because the weighting agent tends to settle out of suspensionas the cement sets. DensCRETE cements, or high-density CemCRETE cements, show little variation in density from topto bottom because the network of particles and associated reduced water content inhibit sedimentation or segregationof the heaviest particles. Each column represents a different cement type and density, with density variation measuredin the top, middle (where the designed density is most likely to be found) and bottom sections of the column.

0

50

100

150

200

250

300

350

400

450

500

30 40 50 60Solid volume fraction, %

Conventional slurryCemCRETE slurry

Conventional slurry (20% fluid loss)CemCRETE slurry (20% fluid loss)

Plas

tic v

isco

sity

, cp

> Fluid-loss effects. As slurries lose fluids to permeable formations, plasticviscosity tends to increase. Compared with optimized slurries, conventionalPortland cement slurries tend to suffer greater increases in plastic viscosityper unit of fluid loss. The bottom two curves show the difference in viscositybetween an optimized blend and a standard blend. The top two curves showthe increase in viscosity after both slurries have lost 20% of their fluid. Optimized blends suffer less viscosification per unit of fluid loss.

030 40 50 60

250

200

150

100

50

Solid volume fraction, %

Monomodal silica suspension in 0.15 M NaCl, PVF 0.5

Trimodal silica suspension in 0.15 M NaCl, PVF 0.8

Plas

tic v

isco

sity

, cp

> Plastic viscosity of silica suspensions. A dry blend consisting of amonomodal particle-size distribution produces a high-viscosity slurryeven at a relatively low solids content. The blend with the trimodalparticle-size distribution, typical of CemCRETE technology, achievesbetter slurry properties and contains more solids per unit volume.

Spring 1999 21

The rheology of CemCRETE slurry is decou-pled from its density (previous page, top left).These water-reduced slurries have constant vis-cosities even at high densities, low gel strengthsand are easy to place. Low water content dimin-ishes sedimentation (previous page, bottom), orseparation of liquid and solids during cementing,yielding higher compressive strength and lowerpermeability (previous page, top right). The spe-cially engineered particle sizes allow easy mixingand pumping because the smallest particles actlike ball bearings to provide lubricity for thelarger solids in the slurry. The compressivestrength of set CemCRETE slurries, whether ofhigh or low density, develops faster and reacheshigher levels than conventional cements (right)because of the low water content.

CemCRETE technology benefits not only pri-mary cementing applications, but also remedia-tion. Particle-size optimization inhibits prematuredehydration of the slurry and the associated fric-tion-pressure increase that commonly preventsany remedial slurry from achieving deep penetra-tion. Water-reduced primary cements have alower incidence of costly remediation thanPortland cements.

Additional benefits are that CemCRETE tech-nology does not require specialized equipment orpersonnel, and while never desirable, mixingerrors are better tolerated in the new slurriesthan in Portland cement. Optimized dry blendsmay be mixed with fresh water, seawater or saltwater. Optimized slurries can include conven-

tional defoamers, accelerators, dispersants,retarders, fluid-loss control additives, right-angleset (RAS) additives and GASBLOK gas migrationcontrol cement technology. In fact, the combina-tion of specialized gas-migration control addi-tives, low bulk shrinkage and rapid strengthdevelopment of optimized cements is breakingnew ground in gas-migration control. Clearly, asexemplified in the case histories that follow,advanced cementing technology can be tailoredto specific needs by changing components of the dry blend.

Specialized ApplicationsThere are four broad applications of CemCRETEtechnology, encompassing low-density, high-density, remedial and deep-water cementingsituations. LiteCRETE slurry systems have lowdensities and are ideal for cementing weak for-mations or eliminating a casing string or a riskymultiple-stage operation (below). LiteCRETE slur-ries of 9.7 to 13 lbm/gal [1166 to 1563 kg/m3]perform comparably to ordinary 15.8-lbm/gal[1900 kg/m3] slurries. Optimized lightweightcement develops compressive strength earlier

6000

5000

4000

3000

2000

1000

00 2 4

Time, hr

8 16 24

18-lbm/gal DensCRETE slurry 12-lbm/gal LiteCRETE slurry15.8-lbm/gal conventional slurry

Com

pres

sive

stre

ngth

, psi

> Compressive-strength development. CemCRETE slurries, both low-density LiteCRETEcement and high-density DensCRETE cement, develop compressive strength earlier and reach higher levels than conventional cement slurries. Rapid compressive-strengthdevelopment reduces waiting-on-cement time and speeds well construction.

Conventional cement LiteCRETE cement

Weak zone

LiteCRETE Stage-Operation Replacement

Conventional cement LiteCRETE cement

Zone 2

Zone 3

Zone 1

LiteCRETE Cementing Production Liner

Tailslurry1

2

LiteCRETE Cement Plugs

Conventional cement LiteCRETE cement

Fillerslurry

> New approaches to common problems. LiteCRETE cement (left) can replace stage-cementing operations, saving rig time and avoiding a complex, moreexpensive operation. Here, the two-stage cementing operation on the left has a weak zone that is eliminated in the single-stage LiteCRETE operation on theright. For cementing production liners (center) or casing across a weak or depleted zone, high-quality cement is placed across the primary pay zone as atail slurry at the bottom of the well. Shallower formations, isolated with lower-quality filler slurry, cannot be completed without additional cementing work.LiteCRETE cement can be placed throughout the entire annulus so that any zone may be completed without additional cementing work, such as blocksqueezes. Placing a higher density cement plug in a lightweight fluid (right) can lead to instability as the fluids intermix. Cement placement is improved bymatching low fluid densities with LiteCRETE slurries, which prevents fluid contamination and degradation of set-cement properties.

than conventional cement, reducing WOC time.In addition, this type of slurry is more stable thanlow-density Portland cement slurries because ofits low water content. It is strong enough to beperforated cleanly and withstands fracturing andstimulation treatments (above left).5

DensCRETE technology offers better rheol-ogy at high density, adjustable density at thewellsite and improved well control duringcementing (above right). High-density, water-reduced cement is useful for whipstock plugsand high-pressure cementing operations, for sit-uations where the fracture and pore pressuremargin is narrow, and for grouting (injection of cement to consolidate seabed sediments orinjection of high-strength cement betweenpipes such as the legs of offshore platforms). A

high-performance, high-density slurry of 17 to24 lbm/gal [2040 to 2900 kg/m3] has a lowerequivalent circulating density than that of a con-ventional high-density cement slurry, allowingplacement even when the window between porepressure and fracture pressure is tight and con-ventional high-density slurries are inadequate.Slurry density can be adjusted by as much as 1 lbm/gal at the last minute on location withoutperturbing other slurry properties. DensCRETEslurries usually develop compressive strengthswell in excess of 5000 psi [34.5 MPa] and canreach 20,000 psi [138 MPa] in especiallydemanding applications.

For remediation of faulty cement jobs and for water control, SqueezeCRETE technologyoffers a new solution for wellbore repairs, such

as casing leaks, liner top leaks, old partiallyplugged perforations, channels behind casing,leaking stage tools, fractures or even squeezinga gravel pack (below). A SqueezeCRETE slurrysystem applies the new technology at themicroscale for injection into very small gaps orfractures in primary cements and casing.Optimized slurries with specially engineeredparticle-size distributions penetrate deeply notonly because of the small particle sizes of theblend, but also because their improved resis-tance to dehydration reduces viscosification dur-ing placement. The improved injectability thatresults from fine-sized particles is key to successin remediation. In addition to high injectability,SqueezeCRETE cement has high compressivestrength and low permeability. Strength makes

22 Oilfield Review

Squeeze throughgravel pack

Microannulussqueeze

Casing leakrepair

SqueezeCRETE Applications

Old perforationsqueeze

Top of linersqueeze

Repair of channelbehind casing

> Remediation success. Perhaps the most versatile application of CemCRETE technology, SqueezeCRETE slurries penetrate more effectively than othercement slurries. SqueezeCRETE slurries repair small microannuli and leaks in casing, channels in cement and liner tops. They can also isolate old, partiallyplugged perforations and even be placed through gravel packs.

> Clean perforating. While conventional cements can shatterduring perforating, CemCRETE cement remains intact afterperforating. The perforation diameter is 0.4 in.

DensCRETE Applications< Cementing high-pressure formations. In high-pressurewells with narrow pore-fracturepressure windows, the frictionpressure increase in a tight annulus during cementing canfracture the formation (left), leading to improper zonalisolation. DensCRETE slurrieshave lower viscosity, allowingslurry placement throughout the annulus. In deviated holes, standard high-density slurries are prone to sedimentation ashematite particles settle on the low side of the wellbore and do not contribute to the totalhydrostatic pressure (right). This instability can lead to serious well control problems.

Spring 1999 23

SqueezeCRETE cement an appropriate materialto plug wells upon abandonment, although it ismore commonly applied to remediate wellboreproblems that cannot be repaired with typicalcementing materials.

SqueezeCRETE technology succeeds wherestandard gels used for water-control applica-tions might fail, including remediation of cross-flow behind casing and as a tail behindconventional gel treatments. When water cross-flow behind the casing is diagnosed, the paththrough the primary cement sheath might not yet be large enough to place ordinary squeezeslurries. On the other hand, the path may alreadybe so large that a standard gel used for water-control applications cannot perform correctly or withstand the differential pressure once thewell returns to production. The advanced slurryexperiences a lower viscosity increase for thesame volume of fluid loss than conventionalsqueeze cements. Its enhanced fluid-loss controlproperties, commonly better than those ofdrilling fluids, greatly improve slurry penetrationproperties: it can penetrate 120 micron slotsmore than 10 times farther than well-dispersedsqueeze slurries (top right).

Engineered slurry for squeeze applications isplaced after deep penetration through the chan-nel and set like ordinary primary cement. In thismanner, SqueezeCRETE technology restores theintegrity of the cement sheath and provides com-petent zonal isolation.

An alternative to foamed cement, DeepCRETEtechnology, has been developed for deepwaterwells. Foamed cement—cement plus nitrogen or air—requires specialized equipment and acementing team trained in its use (as well asavailability of nitrogen when air is not used),which might be logistically challenging andcostly on some offshore rigs and platforms.DeepCRETE cement develops strength faster,even at temperatures as low as 39°F [4°C], soWOC time is reduced when rig costs are calcu-lated by the minute, such as in deepwater areas.No specialized equipment clutters up limitedfloor space. LiteCRETE slurry systems can alsosubstitute for foamed cement.

Traditionally, cement jobs were planned byidentifying the application of the cement and thetotal hydrostatic limitations on the placedcement column. The liquid slurry density wasinferred from the physical properties necessaryfor the set cement. A major change precipitatedby new cementing technology is that the initialplanning step is to decide the slurry density first

and then the slurry porosity. From that, the spe-cific gravity of the dry blend is calculated and ablend designed according to the job parameters.6

CemCRETE technology results in cementproperties that ensure long-lasting zonal isola-tion. Its strong resistance to corrosion from acidstimulations and formation fluids is enhanced byits low permeability (above left). Its mechanicalintegrity is high, even in workover, perforatingand other specialized applications (above right).

Oilfield cement must withstand corrosionand CemCRETE cements provide good sulfateresistance when designed for that purpose.

SqueezeCRETE slurry Standard microcement slurry

Injection point Injection point

> Improved penetration of remedial cement. Squeeze cementing materials were injected through the valve on the left side of the 120-micron slots shown in the photographs. As indicated by the blackarrows below the slots, the SqueezeCRETE slurry (left) achieved deeper penetration into the narrowslot than the conventional microcement slurry (right), which lost more water earlier, viscosified andplugged the left side of the slot. Improved penetration reflects lower fluid loss and reduced viscosifi-cation, allowing SqueezeCRETE slurry to better repair tiny wellbore defects.

0

5

10

15

20

25

30

35

40

Extendedlightweight

cement

15.8-lbm/galClass Gcement

CemCRETEcement

Cem

ent s

olub

ility

in m

ud a

cid

afte

r 4 h

r, w

t %

> Resistance to acid attack. Better zonal isolationis inherent in all CemCRETE systems because oftheir improved resistance to aggressive, corro-sive fluids, as demonstrated in laboratory testson cement solubility by acid or brine. This prop-erty makes LiteCRETE systems particularly valu-able for geothermal applications or when acidstimulation is planned, since low density andresistance to corrosive fluids are of paramountimportance in those situations.

1.00

1.05

1.10

1.15

1.20

1.25

1.30

12-lbm/galCemCRETE

cement

15.8-lbm/galClass Gcement

18-lbm/galCemCRETE

cementT/

E

> Cement integrity. The mechanical integrity ofcement, or its ability to withstand stresses fromperforating, hydraulic fracturing and other opera-tions, is critical for long-term zonal isolation. Theratio of the tensile strength (T) and Young’s modu-lus (E) is one indicator of the relative performanceof different cements. The higher T/E of CemCRETEcements reflects their superior integrity.

5. For more on high-performance, lightweight cement slurries: Moulin E, Revil P and Jain B: “Using ConcreteTechnology to Improve the Performance of LightweightCements,” paper SPE/IADC 39276, presented at theSPE/IADC Middle East Drilling Technology Conference,Bahrain, November 23-25, 1997.Revil P and Jain B: “A New Approach to Designing High-Performance Lightweight Cement Slurries for ImprovedZonal Isolation in Challenging Situations,” paperIADC/SPE 47830, presented at the IADC/SPE Asia PacificDrilling Technology Conference, Jakarta, Indonesia,September 7-9, 1998.Sumartha I and Martinez R. JA: “Application of a New Technique for Lightweight Cement Formulation toImprove Cement Placement in Campeche Bay Area,”paper SPE 39889, presented at the SPE InternationalPetroleum Conference and Exhibition, Villahermosa,Mexico, March 3-5, 1998.

6. Moulin E et al, reference 5.

Also, their low permeability inhibits water per-colation into the cement, slowing corrosion (seebottom left figure, page 19 ). Destructive events,such as repeated freeze-thaw cycling, tectonicactivity, production-induced subsidence andthermal expansion during production and testsprior to abandonment of wells, can impactcement integrity.

Protection of shallow aquifers is an ongoingconcern, so regulatory requirements for cementperformance, such as in well abandonments, arebecoming stricter in many areas. Recently,prudent operators have recognized that surfacecasing should be cemented as carefully asproduction liners. New high-performance oilfieldcements have greater reliability than traditionalcements, even in extreme conditions, so using thebest technology available might help operatorsmeet stricter environmental protection standards.

During 1998, more than 250 CemCRETE jobswere carried out in 20 countries (left). LiteCRETE,DensCRETE and SqueezeCRETE technologieshave been used in most cases, althoughDeepCRETE technology, introduced at the end of1998, is also gaining popularity.

Elimination of Stage-Cementing OperationsIn the Hassi Berkine field in the Ghadames basinof Algeria, Anadarko Algeria Company usesLiteCRETE cement to avoid stage-cementingoperations and better protect the supply of freshwater coming from the overpressured Albiansandstone. The Albian aquifer overlies oil-producing Cambrian sandstones and underliessalty Senonian carbonate and evaporite rocks.Additional geologic complications include theweakness of certain formations below the Albianthat are prone to lost circulation during drillingand the potential for flowing salt. The previousapproach had been to set a stage tool below the Albian, cement the lower zones, and thenisolate the Albian in the second stage of cement-ing operations.

Stage cementing resulted in higher costs thana single-stage operation and suboptimal zonalisolation that often required remedial cementing.After careful consideration of the risks andrewards of different approaches, Anadarko chosea solution proposed by Dowell engineers—single-stage cementing using a LiteCRETE slurry.Key factors that make this preferable to conven-tional cementing include rapid setting time, highcompressive strength, low set-cement porosityand permeability that result in better zonal isola-tion and superior resistance to corrosive forma-tion fluids (left).

24 Oilfield Review

257 CemCRETE jobs worldwide in 1998

> Locations of CemCRETE operations. The size of each circle is proportional to the number of jobs inthe area. During 1998, more than 250 cementing operations using CemCRETE technology demonstratedthe versatility of optimized cement blends in a variety of critical casing operations. Stage-operationreplacement has been the most significant application to date.

Fresh water

Typical Casing Program LiteCRETE Casing Program

Low fracturegradient

Fresh water

Low fracturegradient

The stage tool created aweakness in the 9 5/8-in.casing, requiring 7-in.casing to surface

9 5/8-in. intermediate casingcemented in two stages to cover freshwater zone withlow-permeability cement

9 5/8-in. casingcemented in onestage with LiteCRETE slurry

7-in. full production string

7-in. production linerreplaces the full stringdue to the eliminationof the stage tool

> Elimination of stage-cementing operations. In the Hassi Berkine field, Alge-ria, LiteCRETE technology meets multiple operational challenges: protectionof freshwater supplies, high strength with low density and reduced cost andrisk. Senonian carbonate and evaporite rocks must be isolated from underly-ing Albian sandstone, a freshwater aquifer. Oil production comes from deeperCambrian sandstones. By eliminating stage-cementing operations, a 7-in. pro-duction string to the surface can be replaced by a 7-in. production liner.

Spring 1999 25

The cost savings associated with the single-stage operation and decreased need for remedialcementing were also compelling. A typical sin-gle-stage operation in this area can save almosta full day of rig time and decrease costs of fluidcontamination that might occur during the firststage of cementing. Additional savings stemfrom the low incidence of remedial work, whichtypically requires two days of rig time as well asadditional cementing costs. The elimination ofthe stage tool removes a known weak point fromthe 95⁄8-in. casing string, making it possible toreplace a full 7-in. production casing to surfacewith a 7-in. production liner, saving on tubularand cementing costs as well as rig time (above).7

In the United Arab Emirates, Abu DhabiCompany for Onshore Oil Operations (ADCO) has performed similar successful single-stageLiteCRETE cementing operations.8

Ongoing collaboration between engineersfrom Dowell and Schlumberger Wireline &Testing has improved interpretation of bond logsof lightweight cementing systems. In the past,acoustic properties were incorrectly related tocompressive strengths of cement, resulting in afalse expectation of similar log responsesbetween 15.8-lbm/gal Portland and LiteCRETEsystems. The new systems have compressivestrengths as high as 15.8-lbm/gal Portland

cements, but their acoustic impedances arebetween 15.8-lbm/gal cements and ordinarylightweight cements. LiteCRETE systems displaya lower acoustic impedance contrast withdrilling fluids, producing a different logresponse, so log interpretation for these systemsis not as straightforward.

Fluid compensated CBL amplitude (CBLF)

0 MV 50Transit time (TT)

400 µsec 200Transit time (Sliding Gate) (TTSL)

0 MV 50

CCL (CCLU)

-35 5

Predicted Amplitude for 100% BI fromDowell cement data (DCD PA 100 BI)

0 MV 50

Predicted Amplitude for 80% BI fromDowell cement data (DCD PA 80 BI)

0 MV 50

10 20in.Bit size (BS)

10 20in.

Caliper 1(DCD CALI1)

0 100API

Gamma Ray(GR)

Gas fromLHF2 to USGI

Liquid fromUSGI to USBI

Bonded fromUSBI to LHF2

0.00000.30001.90002.09092.28182.47272.68362.85453.04543.23643.42733.61823.80914.0000

Cement map withimpedance

classification

Min Amplitude

Sonic_VDL_Curve (VDL)

Max

200 µsec 1200

< Evaluation of LiteCRETE cement using bondlogs. The USI UltraSonic Imager (USI) log,cement bond log (CBL) and Variable Density log(VDL) from a well in Algeria give informationrelated to the presence of a 10.85-lbm/gal [1.33-kg/l] LiteCRETE cement behind 95⁄8-in. casing. In the first track (from left to right), thegreen gamma ray curve shows minor lithologyvariation with depth; the black curve indicatesbit size and the red curve hole size (as uploadedfrom the CemCADE software). The bond index isdenoted from 100% to 0% in track 2, with yellowindicating cement behind the casing. Thecement map in the third track is a circumferen-tial representation of the material presentbehind the casing. The cement map was generated by rescaling USI UltraSonic Imagerdata from the default (0 to 8 MRayl) to a scaleof 0 to 4 MRayl to better fit the lower acousticimpedance of LiteCRETE cement, which averages 3 to 4 MRayl. Dark areas, equivalent to 4 MRayl here, indicate excellent cement bondto the casing. The fourth track displays classiccement bond log information, including ampli-tude (solid purple), transit time (blue and red dotted) and casing collar locations (black). Additionally, the orange and green solid linesrepresent the expected amplitude for 100% and 80% bond (as predicted by the CemCADEsimulator). The amplitude values are higherfor LiteCRETE cement than for standard, heavier cements, which typically have greaterattenuation. Finally, the Variable Density cementbond log (VDL) in track 5 provides informationabout the quality of the cement-formation bondby displaying a color-coded traveltime trace atevery depth. The relatively low color contrast(low amplitudes) at early times indicates weakcasing arrivals, which is to be expected for agood bond between the casing and a relativelylow acoustic impedance cement. (A high acoustic impedance cement under the same circumstances would give lower amplitudes and weaker casing arrivals, if any.) The highercolor contrast (high amplitudes) at later times represents arrivals from the formation, whosevelocity varies with lithology, and correlatesroughly with lithology indicated in the gamma ray log.

7. For more on single-stage cementing operations inAlgeria: Toukam E: “New Cement Improves Costs,Operations In Northern Africa,” Petroleum EngineerInternational 72, no. 3 (March 1999): 23-29.

8. Mukhalalaty T, Al Suwaidi A and Shaheen M: “IncreasingWell Life Cycle by Eliminating the Multistage Cementerand Utilizing a Light-Weight, High-Performance Slurry,”paper 53283, presented at the SPE Middle East Oil Show,Bahrain, February 20-23, 1999.

Whipstock Plugs and Liner CementingIn Mexico, Petróleos Mexicanos (PEMEX) hasused LiteCRETE cement for whipstock plugs andliner cementing. PEMEX initially used thelightweight optimized blend for whipstock plugsto kick off deviated wells past irretrievable fish.The success ratio of kickoff plugs has beenimproved greatly by using the new technology ina low-density environment. The matched densi-ties of the drilling fluids and cement slurries pre-vented swapping and mixing of fluids duringplacement and ensured development of therequired compressive strength.

In a field with a low fracture gradient in theVillahermosa region, CemCRETE technologyproved to be the best answer for cementing deep (4500- to 5000-m) [14,760- to 16,400-ft],depleted, fractured, dolomitic Mesozoic carbon-ate reservoirs. Lightweight cement is employedbecause the reservoirs have a low fracturegradient.9 In one deviated well, PEMEX elected tokick off in order to reach a better part of thereservoir (above left). A special 15-lbm/gal opti-mized whipstock plug material designed forPEMEX reached a compressive strength of 3750 psi [26 MPa] within eight hours and a finalcompressive strength of 4203 psi [29 MPa] in12.5 hours, allowing the sidetrack to be com-pleted successfully.

Liner cementing has also been improvedthrough the use of new cementing technology.Because of the low formation pressure and sus-ceptibility to fracturing, a low-density slurry wascritical to success. In one case, an 11.1-lbm/gal[1330-kg/m3] LiteCRETE slurry was used tocement a 5-in. production liner from 13,399 ft to 15,095 ft [4084 to 4600 m]. The cementdeveloped a compressive strength of 1200 psi[8273 kPa] after eight hours. A cement bond logconfirmed a good seal between the liner cementand formation.

The overall cost of using LiteCRETE technol-ogy, including service, products and rig time, islower than the cost of using traditional technol-ogy. PEMEX reduced the cost of rig time duringcementing by 30% because new lightweight slur-ries develop compressive strength rapidly. Byusing optimized cement for kickoff plugs, PEMEXsaved 45% of the total operation cost comparedwith the use of conventional cement, which com-monly entailed repeating the cement plug. Also,remedial squeeze operations have not been nec-essary. Conventional jobs commonly required oneor two squeezes.

Cementing Shallow, Low-Pressure WellsHunt Petroleum Corporation has used LiteCRETEcement to complete five wells in the Olla field,LaSalle Parish, Louisiana, USA. Shallow Wilcoxoil wells, with total depths of 3500 ft [1067 m]and bottomhole static temperatures of 129°F[54°C], have low bottomhole pressures and lowfracture gradients, so getting a column of cementhigh enough in the annulus has proven difficult.In the past, as many as three block squeezes perwell were performed to remediate poor primarycement jobs in 51⁄2-in. casing (left).

The Wilcox reservoir in Olla field has a strongwaterdrive. Productive zones are completed byperforating the top of the productive intervalabove the oil-water contact. Offset wells com-monly produce high volumes of water at watercuts greater than 95%. The wells completedwith LiteCRETE cement produce at water cutsless than 85% water, but, more importantly, thetotal volume of water produced is significantlyreduced. Hunt Petroleum interprets the reducedwater production as verification of proper isola-tion of the producing zone from nearby zonesthat contain 100% water. The additional waterproduction in the offset wells has beenattributed to water channeling from nearbywater zones; radioactive tracer injection logshave verified this. None of the wells in whichHunt Petroleum used LiteCRETE cement hasrequired remedial work.

Besides reducing the need for remedial work,Hunt Petroleum has lowered total well costs onWilcox completions by avoiding the mechanicalrisks associated with squeezing operations. Suchrisks include the possibility of setting the cementretainer incorrectly, drilling a hole in the casingwhen drilling out the cement retainer, splittingcasing during the squeeze, cementing the work-string if cement sets up early, or fracturing into awater zone. Because LiteCRETE cement columnsextend higher in the annulus, upper zones of theWilcox may be completed without additionalcementing to cover these zones, which generallyare not covered during conventional operations.

26 Oilfield Review

Cement plug

2878 m

4150 m4160 m

Optimized CemCRETE Plug for Sidetracking

> High-performance lightweight slurry. Optimized,low-density blends are used for whipstock plugsand liner cements in depleted reservoirs with low fracture gradients. In this example, PEMEXdecided to sidetrack to reach a better part of the reservoir. By using CemCRETE technology,PEMEX has improved its success ratio for kickoff plugs and minimized WOC time.

8 5/8-in., 24-lbm/ft surfacecasing at 1712 ft

Weak zone

5 1/2-in., 15.5-lbm/ft production casingat 3150 ft

Cementing Low-Pressure Zones

> Cementing in a low-pressure gradient. Theuse of conventional cements in the Olla fieldtypically required two or three block squeezesafter each primary cementing operation. UsingLiteCRETE slurry systems on five wells improvedzonal isolation without block squeezes. It alsomakes it possible to complete shallower zoneswithout additional cementing work. In thisexample, the LiteCRETE slurry column could be placed high enough in the annulus to coverthe weak zone.

9. Pérez Mejía G, Ramírez Martínez I and Prado-Velarde E:“Optimización de los Tapones de Desvío y Liners,Utilizando un Sistema de Cemento de Baja Densidad yAlta Resistencia a la Compresión (LBDARC), en laRegión Sur de Pemex, México,” presented at the XICongreso Latinoamericano de Perforación, BuenosAires, Argentina, October 25-29, 1998.

10. In the North Sea, a LiteCRETE blend remained on a sup-ply boat for several days in bad weather. Nevertheless,the blend did not segregate during its rough journey tothe wellsite.

Spring 1999 27

Cementing High-Pressure WellsHigh-pressure wells benefit from the use ofreduced-water cements. Petroleum DevelopmentOman (PDO) first adopted DensCRETE technologyto address numerous challenges in fields such asthe Al Noor and Sarmad fields of southern Oman.While adjustments to the mud system and casingprogram can reduce the cost and risk of drillingoperations, the use of new cementing technologywas the most important factor in improving oper-ations for PDO.

In the southern Oman fields, PDO produces oilfrom stringers of tight Cambrian Athel silicilyteembedded in salt. The Athel reservoir, which isalso a world-class hydrocarbon source rock, is upto 400 m [1312 ft] thick and contains 80 to 90%microcrystalline silica, with an average porosityof 22% and permeability below 0.05 mD. Highdrawdown pressures are applied to produce oilfrom such a tight reservoir, so it is crucial tomechanically isolate the individual stringers ofreservoir rock.

Drilling and completing such wells suc-cessfully are challenging. At depths of 3500 to4800 m [11,483 to 15,748 ft] and temperatures of90°C [194°F], pressure control dictates a high-density slurry. Segregation of the weightingagent, hematite, from conventional dry blendsduring transport across graded roads led to diffi-culty mixing and pumping slurries and up tothree hours of lost time to clean pluggedcementing lines. Displacing heavy muds withhigh rheologies was inefficient. There was a nar-row window between the formation pore pres-sure of 16.2 lbm/gal [1941 kg/m3] and formationfracture pressure of 20.4 lbm/gal [2444 kg/m3],as well as a low differential pressure betweenthe 17-lbm/gal [2037-kg/m3] mud system, 18.3-lbm/gal [2193-kg/m3] spacer and 19.6-lbm/gal[2348-kg/m3] cement. There was little leeway toadjust densities and displacement rates.

Contamination of fluids by salt-saturated mudled to instability. Bulk shrinkage of set cementoften resulted in microannuli. In at least onewell, a microannulus was not detectable with acement bond log, but was discovered when pres-sure in the annulus rose. Finally, when comparedwith conventional cements, CemCRETE slurriesset faster at the top of the liner, which reducesthe risk of fluid migration. In one well, a gas kickoccurred 14 hours after conventionally cementingthe liner and it took four days to control the welland avoid a blowout.

Before approval for the initial use ofDensCRETE cement by PDO, numerous tests byPDO and by Dowell in Oman and at theSchlumberger-Riboud Product Center in Franceconfirmed that the advanced technology wouldsurpass critical performance requirements. Inaddition to exceeding the performance of tradi-tional cements in 8-hour compressive strength,24-hour compressive strength, stability andshrinkage, DensCRETE cement offered greaterability to optimize slurry rheology and density(below). A yard trial in early 1998 also demon-strated that the DensCRETE blend would not seg-regate during transport, remained mixable aftertransport and passed relevant API tests, such asrheology, compressive strength and fluid loss.10

The first DensCRETE operations in Omanwere performed during the second quarter of1998 on the Sarmad-1 well, placing cement plugsat 4100 m [13,451 ft] and 4300 m [14,108 ft] with 21.5-lbm/gal [2576-kg/m3] slurry and a 7-in.liner at 3850 m [12,631 ft] with 19.5-lbm/gal[2337-kg/m3] slurry (above). Because the wellencountered a fault and fluid losses occurred justabove total depth, PDO decided to set plugsabove the fault and then cement the liner usingDensCRETE cement for both operations. Theplugged interval exceeded 200 m [656 ft] in thick-ness, so the plug was set in two stages.

To date, seven DensCRETE cement jobs havebeen performed in the area for PDO, includingthree liner jobs and four plugs for abandonmentof high-pressure wells. The slurry is less sensi-tive to salt-saturated mud contamination thanordinary cement. As the optimized high-densitycement sets, it is less prone to forming amicroannulus because it suffers less bulk shrink-age. Even in long liners, no density gradient isobserved in the set cement column in the annu-lus. The column is uniform and stable, even asthe cement is setting, so the risk of a blowout isreduced. The top of DensCRETE plugs is closer tothe theoretical top than that of conventionalplugs because the rheology of optimized high-density slurries allows more efficient removal ofdrilling fluids.

Compressive strength at 8 hr

Initial set 50 psi

Compressive strength at 24 hr

Shrinkage

Stability of set cement(BP settling test)

Tolerance to density variations

Separation of heavy particlesfrom blend during transport

0 kPa

After 20 hours

18,275 kPa [2651 psi]

1.5% after 24 hours

0.35 kPa/m [0.297 lbm/gal]top to bottom

Low

High risk

18,616 kPa [2700 psi]

Properties Conventional slurry DensCRETE slurry

After 4 hours

24,132 kPa [3500 psi]

0% after 24 hours

0.20 kPa/m [0.169 lbm/gal]top to bottom

High

Very low risk

< Laboratory testing. In testsconducted before the firstuse of high-performanceheavyweight slurry byPetroleum DevelopmentOman (PDO), DensCRETEslurries outperformed conventional heavyweightslurries. This superior performance carried overto field applications.

DensCRETE Plug and Liner Cementing

7-in. liner3850 m

4100 m

4300 m

Cement plug(21.5-lbm/gal)

Liner cement(19.5-lbm/gal)

> High-pressure cementing. In the deep, high-pressure Sarmad-1 well, PDO set 21.5-lbm/galDensCRETE cement plugs to counter fault-related fluid losses near total depth and then cemented the liner using a 19.5-lbm/galDensCRETE slurry.

> Water-control diagnosticplots. Log-log plots of theactual water-oil ratio (WOR)and its derivative (WOR’)versus time help differentiatebetween water-control problems, including waterconing and channeling, dur-ing production. Systematicflow model numerical simulations produced characteristic curves. Thesecurves are used to diagnoseproblems and then decidethe appropriate remedy. Thetheoretical representation ofbottomwater coning (upperleft) is similar to the actualfield example below it. Simulated multilayer channeling (upper right)also mimics actual multilayerchanneling observed in thefield (lower right).

The WOC time for conventional cements todevelop adequate compressive strength underthe conditions in the southern Oman fields is atleast 28 hours. DensCRETE cement achieves highcompressive strength in as few as 15 hours (theworst case to date has been 26 hours) and ulti-mately develops higher compressive strengththan standard high-density cement (right). Thedecrease in WOC time has proven especiallyimportant in drilling exploration wells, and therehas been a decreased need to repeat plugs orremediate liner cements. Thus, PDO plans tocontinue to use DensCRETE cement for high-pressure cementing operations.11

Water-Control ApplicationsSqueezeCRETE technology has been used inAlberta, Canada, for numerous squeeze jobs. Inthe Halkirk field northeast of Calgary, an oil welloperated by PanCanadian Petroleum Ltd. pro-duced 35 m3 [220 bbl] of oil nearly water-freefrom the Upper Manville ”I“ Glauconitic forma-tion upon its initial production in 1995. Within a year, however, water production increased from1 m3 [6 bbl] per day to more than 20 m3 [126 bbl]per day. By late 1998, the well was completelywatered out. From knowledge of reservoir geol-ogy and performance, the water influx wasattributed to layer breakthrough.

The attractive economics for remedial workprompted action. On the basis of known wellboreintegrity, a bridge plug was set above existingperforations at 1266.5 m [4154 ft] and the zoneabove it was reperforated, but water productioncontinued. After reviewing geological, reservoirand completion data, the water influx wasascribed to poor cement behind the bridge plug.Because of a large drawdown and close water

proximity, the Dowell DESC Design andEvaluation Services for Clients engineer wasasked to verify that water coning was occurring.

Water-control diagnostic plots, which displayraw historical production data versus time on alog-log scale, help identify water sources, suchas differentiating bottomwater coning frommultilayer channeling (below). Systematic flowmodel numerical simulations were performed to

28 Oilfield Review

0:00 1:45 3:30 5:15 7:00 8:45 10:30 12:15 14:00 15:45 17:30 19:15

Elapsed time

196

176

156

137

117

98

78

58

39

19

0

Compressive strength

Temperature

Com

pres

sive

stre

ngth

, psi

Tem

pera

ture

, °C

5200

4680

4160

3640

3120

2600

2080

1560

1040

520

0

> Rapid development of compressive strength. High-density optimized slurries develop compressive strength sooner than their conventional counterparts. In this examplefrom the Al Shomou-4 well, the 22-lbm/gal DensCRETE slurry achieved a strength of5000 psi in only 17 hr.

0.00011 10 100 1000

10

1

0.1

0.01

0.001

Time, days

0.0011 10 100 1000 10,000

10

100

1000

1

0.1

0.01

Time, days

0.00110 100 1000 10,000

10

1

0.1

0.01

Time, days

0.00011 10 100 1000 10,000

10

100

1000

10,000

1

0.1

0.01

0.001

Time, days

WORWOR'

WORWOR'

WORWOR'

WORWOR'

WOR

or W

OR'

WOR

or W

OR'

WOR

or W

OR'

WOR

or W

OR'

Spring 1999 29

produce characteristic curves for different typesof water production. On the basis of the water-control analysis for the Halkirk well, the diagno-sis was a high-permeability layer with waterbreakthrough (right). This problem was compli-cated by a microannulus that allowed water flowbehind the casing.

Because of low oil prices and the fact that themature Halkirk field is undergoing waterflooding,workover costs must be minimized to achieveacceptable economic results. Considerable effortis made to mitigate the risk and impact of unsuc-cessful treatments. Therefore, procedures with ahigh probability of success are favored. In thiswell, a conventional cement squeeze wasdeemed too risky. The SqueezeCRETE treatmentwas predicted to have a much higher probabilityof success, so the economics for that treatmentwere acceptable.

SqueezeCRETE slurry was placed across theperforations from 1263 to 1265.25 m [4144 to4151 ft] as a balanced plug, and a hesitationsqueeze was performed.12 After 24 hours, thecement was drilled out and successfully pres-sure- and swab-tested. Following reperforation,the zone is producing 28 m3 [176 bbl] of oil, 3100 m3 [110 Mcf] of gas and 0.32 m3 [2 bbl] of water per day, reversing the water cut from 99.5% before to only 1.1% after thesqueeze operation.

In another well in southern Alberta,PanCanadian wanted to shut off old perforationsand complete a deeper interval. Because theslurry feed rate into the old perforations was lessthan 20 L/min [5.3 gal/min], ordinary slurrieswould not be effective. After acid was spottedacross the perforations to increase the injectionrate, only minor improvement occurred.SqueezeCRETE slurry was then batch mixed and1.2 m3 [8 bbl] placed across the perforations, fol-lowed by a hesitation squeeze. After 48 hours,the cement was drilled out and the perforationswere successfully pressure tested and swabtested. The lower interval was subsequently per-forated and completed. Without a highlyinjectable remedial system like SqueezeCRETE

slurry, the operator might have risked impairingthe additional completion by using a casing patchto shut off the abandoned perforations.

SqueezeCRETE cement has the potential toaddress stringent well plugging requirements assome of the many shallow gas wells in westernCanada are abandoned. Its high injectability andlow permeability can repair gas leaks better thantraditional cementing materials.

Successful water-shutoff jobs have beenperformed using engineered squeeze cementselsewhere. In one case in the North Sea, oil pro-duction increased from 2000 to 4000 bbl per day[317 to 635 m3/d] while water productiondecreased from 7000 to 1500 bbl per day [1112 to 238 m3/d]. This sharp reduction inwater production made gas lift unnecessaryafter production resumed.

Also in the North Sea, BP Amoco plc success-fully abandoned a reservoir section in a well fromits Bruce platform using a single optimizedcement plug. After remedial completion effortsand other attempts to isolate and abandon thereservoir failed, SqueezeCRETE slurry waspumped through coiled tubing across the perfora-tions and then squeezed. BP Amoco plc was thenable to sidetrack an adjacent wellbore to reachthe reservoir.

Merely pumping a superior slurry does notalways effect the desired repair. Sound comple-tion engineering concepts, proper design andexecution are critical ingredients for successfulwell remediation.

Present and Future Value of Optimized CementsCemCRETE technology has proven its value onseveral fronts. Its early development of compres-sive strength saves rig time because drillingoperations can resume sooner. The reliability ofthe technology decreases the need for remedialblock squeezes or repetition of plugs. Repairs offaulty cement and casing are more effective thanever before. The risk and expense of stage-cementing operations are avoided with a single-stage operation using CemCRETE technology.Well designs can be optimized to avoid costlycasing strings.

The lower porosity and permeability of set cements using CemCRETE technology willallow safer abandonment of wells and isolationof aquifers from hydrocarbon zones. Low-permeability cements are more resistant tocorrosive brines and there is less bulk shrinkageas the cement sets, resulting in better zonal iso-lation over time. Studies are nearing conclusionon the enhanced durability of the new systemsover conventional cements when perforating.

The successful application of CemCRETEtechnology in 257 wells during 1998 provides afoundation for expansion of this versatile tech-nology from specialized initial applications tomainstream cementing operations. —GMG

0.001

0.01

0.1

1

10

100

1,000

10,000

10 1000100 10,000

Cumulative production time, days

High-permeability layerwater breakthrough

Initial wellborefluid cleanup

Wellbore waterholdup

Post-treatment WOR

WORWOR'

WOR

or W

OR'

> Halkirk water-control diagnostic plot. The PanCanadian Halkirk well produces oil from several layers. A log-log plot of the actual water-oil ratio (WOR) and its derivative (WOR’) versus cumulativeproduction time illustrates water breakthrough from a high-permeability layer. Increasing hydro-static pressure from wellbore water holdup significantly reduced oil production, and the water cut reached 99.5%. Successful shutoff of the water layer restored the previous oil production andreduced the water cut to 1.1%.

11. For more on the use of high-density slurries in well construction: Adamson K, Birch G, Gao E, Hand S,Macdonald C, Mack D and Quadri A: “High-Pressure,High-Temperature Well Construction,” Oilfield Review 10,no. 2 (Summer 1998): 36-49.

12. In a hesitation squeeze, a portion of the slurry is pumped,then pumping stops to expose the slurry to differentialpressure against the zone of interest in stages over a period from several minutes to several hours. This pressure, higher than necessary for fluid movement, isapplied to force filtrate from the cement slurry, leavingonly solid material in the area requiring repair. This pro-cedure is repeated until all the slurry has been pumped.The dehydrated cement remaining in the zone forms aseal with a higher compressive strength and lower per-meability than the original slurry design.

Oilfield Review

Ramez AkhnoukhJames LeightonRennes, France

Yann BignoPatrick Bouroumeau-FuseauEddie QuinTotal Oil MarineAberdeen, Scotland

Gérard CatalaLisa SilipignoClamart, France

Jim HemingwayJack HorkowitzSugar Land, Texas, USA

Xavier HervéColin WhittakerAberdeen, Scotland

Koji KusakaMontrouge, France

Dan MarkelAndy MartinRosharon, Texas

For help in preparation of this article, thanks to MikeBrandie, Schlumberger Wireline & Testing, Aberdeen,Scotland; Alison Goligher, Schlumberger Wireline &Testing, Montrouge, France; Steve Grayson, SchlumbergerWireline & Testing, Bakersfield, California, USA; NeilHewitt, GeoQuest, Anchorage, Alaska, USA; TimothyMcLaughlin, Schlumberger Wireline & Testing, Lafayette,Louisiana, USA; Brad Musgrove, Dowell, Anchorage,Alaska; Daniel Palmer, Schlumberger Wireline & Testing,Prudhoe Bay, Alaska; Marc Pearcy, Schlumberger Wireline& Testing, Oklahoma City, Oklahoma, USA; Darryl Radke,Lee Tool, Red Deer, Alberta, Canada; George Spencer,Schlumberger Perforating & Testing, Rosharon, Texas, USA; David Turner, Schlumberger Wireline & Testing,Duncan, Oklahoma; and TeS Technical Editing Services,Chester, England.CMR (Combinable Magnetic Resonance Tool), CPLT(Combinable Production Logging Tool), CQG (Crystal QuartzGauge), Enerjet, FIV (Formation Isolation Valve), FloView,FloView Plus, FMI (Fullbore Formation MicroImager),GeoFrame, Gradiomanometer, HSD (High Shot Density),HTX, MAXIS (Multitask Acquisition and Imaging System),MAXPRO, PatchFlex, PLT (Production Logging Tool), PLFlagship, PS PLATFORM (Production Services Platform),PosiSet, PowerJet, PVL (Phase Velocity Log), RST(Reservoir Saturation Tool), Sapphire, SpectroLith, TPHL(three phase fluid holdup log), UNIGAGE and WFL (WaterFlow Log) are marks of Schlumberger. Unix is a mark ofUNIX System Laboratories.

To survive in the Exploration and Production (E&P)industry’s turbulent economic environment, oper-ators must produce hydrocarbons at the lowestpossible cost. The most cost-effective way toreverse declining profits is to capitalize on moneyalready invested. This means that wells alreadydrilled and producing are the most viable sourcesof future earnings. Keeping these wells produc-ing economically at optimal rates throughouttheir lifetimes is a top priority.

Production problems come from manysources. For example, in the wellbore they mayarise from the tubing or completion hardwareand perforations—both inside casing and in the

annulus behind casing. Typical problems includetubing and casing leaks, packer failures, flowbehind casing through cement channels, and fluidcrossflow inside casing between permeablelayers. Production logging can identify theseproblems. Others—such as water coning andcusping, changing oil-water contacts, poor arealsweep, high-permeability-induced breakthroughor fracture conduits—are associated with thereservoir itself. Many of these problems can be detected with cased-hole formation moni-toring tools.

30

Keeping Producing Wells Healthy

Monitoring, diagnosing problems and intervening

in a timely manner help operators maintain

well productivity. New technology and

experience are providing the most

economic solutions.

Spring 1999 31

Although complex production problems oftenrequire detailed analysis, the solutions may besimple and inexpensive. Repairs for leaks fre-quently involve mechanical shutoff such as cas-ing patches, casing liners and straddle packers.Other approaches use shutoff fluids, such ascement, plastic or resins, either in the formationor the annulus between the formation and cas-ing (see ”Concrete Developments in CementingTechnology,“ page 16 ). Crossflow and floodingbreakthrough problems are often more difficult.Simple solutions may involve a plug or casingpatch, while a long-term approach might bereentry drilling with a horizontal section.1

Assiduous water coning and cusping problemsusually can be solved with lateral drainholes ormultiple completions.2

This article looks at how production servicescan provide sustained optimized production,while lowering operating and intervention costs.First, we discuss how the changing productionenvironment has challenged the traditionalmethods of well diagnosis and monitoring. Then,we review the recent developments in productionlogging measurements and interpretation, illus-trated by several field examples. We also surveyremedial solutions to production problems,including a new through-tubing casing patchcalled PatchFlex technology, new perforatingtechniques that have been developed in responseto specific client requests, and innovative appli-cations of combined production logging and com-pletion services.

The ChallengeThe production logging environment has becomeextremely challenging. Mature reservoirs—oftenunder waterflood to maintain reservoir pres-sure—experience increased water breakthrough,usually through high-permeability layers. Asfields become depleted and reservoir pressuredrops, three-phase flow is encountered more fre-quently. Shutting off unwanted water or gas andidentifying nonperforming zones for stimulationtreatment are becoming key diagnostic activitiesfor production logging. It has become necessaryto precisely locate hydrocarbon and water entrypoints and to accurately measure the flow ratesof each phase. The increasingly complex environ-ment is driving the need for new production log-ging sensors and interpretation methods toprovide data for confident remedial interventions.

To tap the full potential of existing reservoirsand make marginal fields more productive, newwells are becoming more complex. Many havehigh-angle or horizontal wellbores that exhibitcomplicated downhole fluid behavior. The natureof multiphase flow itself is a major challenge tothe development of new sensor technology.Before 1995, conventional production loggingtools developed for vertical wells could not pro-vide flow profiles, holdup or effective well lengthin horizontal wells, and were also subject to thesame limitations found in vertical wells.

In conventional wells with deviations of morethan 60º, fluid density from the Gradiomanometerspecific gravity profile suffers significant loss of accuracy. In horizontal wells, fluid velocitymade with spinners and capacitance holdupmeasurements are significantly affected bystratified flow regimes. Small changes in welldeviation can cause large changes in fluid veloc-ity and holdup, particularly at lower flow rates.The measurement of either the phase velocity orholdup in isolation can cause a misdiagnosis offluid entry. Furthermore, horizontal wells produc-ing water-free oil at the surface can have waterpooled in sumps and gas in traps downhole thatperturb the response of conventional productionlogging tools.

1. Hill D, Neme E, Ehlig-Economides C and Mollinedo M:“Reentry Drilling Gives New Life to Aging Fields,” OilfieldReview 8, no. 3 (Autumn 1996): 4-17.

2. Bosworth S, Saad El-Sayed H, Ismail G, Ohmer H, StrackeM, West C and Retnanto A: “Key Issues in MultilateralTechnology,” Oilfield Review 10, no. 4 (Winter 1998): 14-28.

A Quantum Leap in ResolutionIn recent years, there has been a significantimprovement in the understanding of multiphaseflow in producing oil wells (left).3 The flowregime is specific to the number of flowingphases (two-phase oil and water, or gas and oil;three-phase oil, gas and water) and the flow rateof each phase and the wellbore geometry, espe-cially wellbore deviation. An understanding ofthe impact of the oil, water and gas flow regimesin vertical, deviated and horizontal wells hasdriven the development of improved productionlogging technology for the most challengingdownhole flow conditions.4 Conventional pro-duction logging sensors such as spinners orGradiomanometer tools sense only the combinedor average physical effect of fluid and flow prop-erties in the wellbore. The interpretation of theresponse from spinners and nuclear tool mea-surements is sensitive to the geometry and dis-tribution of fluid flow. These sensors providelarge volume or global measurements in thewellbore, and in order to interpret their netresponse, a global fluid mechanics interpretationmodel is needed.5 However, these models arestill in their infancy, so computer programsdesigned to solve a system of equations are usedto derive flow rates using global measurements.

As the logging environment becomes morecomplex, understanding fluid flow becomesmuch more difficult. Phase segregation acrossthe wellbore occurs, and traditional global fluidmeasurements react differently to the separatephases traveling at different velocities. Con-ventional interpretation models, approximationsand assumptions about the nature of the flowregime start to break down. For example, therecan be complete phase separation in horizontalwells, and segregated fluid phases traveling atdifferent velocities will defeat conventional fluidflow models (left). Therefore, a new approach isneeded both in making measurements and inusing models to interpret these data.

32 Oilfield Review

Wavy stratified flow (SW)

Slug flow (SL)

Annular flow (A)Stratified flow (SS)

Plug flow (PL)

Dispersed bubble flow (DB)

0.01 0.1 1 10 100

0.01

0.001

0.1

1

1090 degrees

Superficial gas velocity, m/sec

Sup

erfic

ial l

iqui

d ve

loci

ty, m

/sec

0.01

0.001

0.1

1

1091 degrees

0.01 0.1 1 10 100

0.01

0.001

0.1

1

1089 degrees

0.01 0.1 1 10 100

DB

PL

SS

SW

SL

A

DBPL

SS SW

SL

A

DB

PL

SS

SW

SL A

> Flow regimes. In oil-water flow regimes, the flow patterns dependon the fraction of the pipe cross section occupied by water—calledwater holdup—and the wellbore deviation. In vertical and deviatedwells, the flows are essentially bubbly, in which one phase flows asa dispersed phase, and the other is continuous. Phase inversion fre-quently occurs in oil wells, where the flow changes from being con-tinuous water to continuous oil as more oil enters the pipe.

< Gas-liquid flow regimes. The flow pattern pic-tures (top) and crossplots (bottom) show theprincipal flow regimes for wellbores with devia-tions of 89º (uphill flow), 90º (horizontal flow) and91º (downhill flow). When gas is present, thedominant regime is slug flow (SL) and plug flow(PL) in uphill flows and stratified flow (SS) andwavy stratified flow (SW) in downhill flowsshown by the largest area on the crossplots.

Pure water

Pure oil

Pure water

Mixed

0 10 20 30 5040 8060 70 10090Water holdup, %

90

80

70

60

50

40

30

20

10

0

Dev

iatio

n, d

eg

Pure water

Pure oil

Spring 1999 33

New measurement techniques have beendeveloped using small sensors that can quicklydetect changes in the local fluid-flow regime.These approaches combine new local measure-ments with other existing global measurementsfor a better understanding of fluid dynamics. Forexample, the introduction of small electric probesfor mapping water and hydrocarbon holdup in awellbore has made production logging in highlydeviated and near-horizontal wells more quanti-tative.6 The small size allows sensors to providedata on a physical scale measured in fractions ofa bubble diameter, which then eliminates theneed for a mixing law to interpret their response.

New Production Logging TechnologyResearch into the dynamics of multiphase fluidflow in wells has sparked exciting new develop-ments in production logging technology.7 The PLFlagship advanced well flow diagnosis service—developed jointly by British Petroleum ExplorationOperating Co. Ltd. and Schlumberger CambridgeResearch, Cambridge, England—is the first suiteof production logging sensors for diagnosing fluidproduction problems in high-angle and horizontalwells.8 Applications include differentiation ofwater and hydrocarbon entries, distribution ofthe flow regime, individual phase holdup mea-surements and determination of the flow rates ofeach phase. The measurements include pressureand temperature with the CPLT CombinableProduction Logging Tool string, total fluid velocitywith a spinner in the CPLT tool, three-phaseholdup and fluid movement with the RSTReservoir Saturation Tool, flow regime and waterholdup with electrical impedance probes in theFloView tools, and fluid phase-velocity measure-ments with the PVL Phase Velocity Log tool.Many of the concepts developed and tested aspart of the PL Flagship service have now beenembodied in the new generation of compact inte-grated-sensor production logging technology,called the PS PLATFORM Production Servicestool. This tool is being used to provide detailedoil, gas and water flow-rate measurements alongwith wellsite real-time analysis.

The PS PLATFORM tool is the latest multi-phase production logging tool developed bySchlumberger. It can be operated with anytelemetry logging cable (including mono-conductor and coaxial cables) or slicklinerecorder operations in vertical and deviated wells(see “New Technology for Production Services,”next page). In the latter case, the logging dataare memorized downhole and retrieved at sur-face with a portable computer. Major tool-lengthreductions are measured using a high level ofsensor integration within the Flow-CaliperImaging tool, which can provide important mea-surements of flow velocity and holdup only 16 in.[41 cm] from the bottom of the tool string (right). Combined with a depth resolution of 6 in. [15 cm], this tool can accurately pinpointfluid entries in the lowest levels of the borehole.An array of locally interpretable measurements,when combined with global measurements,allows analysis of complex, multiphase fluidbehavior. The advantage of using many dis-tributed small sensors is that the sensor arraycan be arranged to map fluid characteristicsacross a single plane in the wellbore. These sen-sors can then be collocated with other globalsensors, such as a spinner, to obtain simultane-ous measurements at the same depth. Thisapproach greatly enhances the depth resolutionof fluid entries. This new production logging platform has been successfully conveyed in high-angle wells, without the assistance ofcoiled tubing.

The Multi-Capacitance Flowmeter (MCFM)tool developed by Baker Atlas uses pairs of rigidsteel plates, which when extended form a capac-itor array, to measure holdup across the well-bore.9 Fluid velocity is determined by correlatingmeasurements from adjacent pairs. These mea-surements are combined to determine the flowprofile. The MCFM tool is run with the AtlasPulsed Neutron Holdup Indicator (PNHI) tool thatuses carbon-oxygen and count-rate measure-ments to determine gas, oil and water holdups.The PNHI tool is also capable of making oxygen-activation measurements to help identify waterentry points.

> Flow-Caliper Imaging tool. The Flow-CaliperImaging tool is compact, measuring only 5.2 ft[1.6 m] long. The caliper arms close to encasethe spinner with its foldable blades to allowsafe conveyance through tight restrictions inboth cased and open holes. The diameter of thefolded tool is 111⁄16 in. [4.29 cm] when the ends ofthe caliper arms are equipped with protectiveskids, and 21⁄8 in. [5.08 cm] when equipped withrollers for easier conveyance. The spinner andcaliper settings can be changed easily at thewellsite to adapt the tool for a range of tubularand hole sizes encountered during the survey.

(continued on next page)

3. Catala G, Théron B, Conort G and Ferguson J: “FluidFlow Fundamentals,” Oilfield Review 8, no. 4 (Winter1996): 61-64.

4. Rounce J, Lenn C and Catala G: “Pinpointing Fluid Entriesin Producing Wells,” paper SPE 53249, presented at the1999 SPE Middle East Oil Show, Bahrain, February 20-23,1999. Also see Bamforth S, Besson C, Stephenson K,Whittaker C, Brown G, Catala G, Rouault G and Théron B:“Revitalizing Production Logging,” Oilfield Review 8, no. 4(Winter 1996): 44-60.

5. In this article global measurements are those that involverelatively large volumes of the wellbore. Global modelstypically require mixing laws to determine the bulk fluideffect on the sensor response. For example, holdup istraditionally derived from fluid density measurementsmade with pressure sensors in the Gradiomanometer tool 21 in. [53 cm] apart.

6. Halford FR, MacKay S, Barnett S and Petler JS: “AProduction Logging Measurement of Distributed LocalPhase Holdup,” paper SPE 35556, presented at the SPEEuropean Production Operations Conference andExhibition, Stavanger, Norway, April 16-17, 1996.

7. Bamforth et al, reference 4. 8. Lenn C, Bamforth S and Jariwala H: “Flow Diagnosis in

an Extended Reach Well at the BP Wytch Farm OilfieldUsing a New Toolstring Combination Incorporating NovelProduction Logging Technology,” paper SPE 36580, pre-sented at the SPE Annual Technical Conference andExhibition, Denver, Colorado, USA, October 6-9, 1996.

9. von Flatern R: “Spinning a New Line,” Offshore Engineer(January 1999): 36-37.

Production Log Interpretation Traditional production logs require considerablemathematical manipulation to convert rawcurves of pressure, temperature, density, localholdup and spinner revolutions into a flow profile.Two- and three-phase flow interpretations areespecially difficult. For example, the spinner dataand cable velocity are combined and transformedinto an apparent fluid velocity. This apparent fluidvelocity is then corrected for variations in veloc-ity across the pipe. The correction factor used isa function of spinner blade size, completion inter-nal diameter and Reynolds number. The Reynoldsnumber is a function of fluid velocity, completioninternal diameter, true fluid density and fluid vis-cosity. The true fluid density is a function ofapparent fluid density from the Gradiomanometerlog together with the well deviation, the surfaceroughness of the completion, the Reynolds num-ber and the true fluid velocity.10 The downholedensities of oil, water and gas may be known orestimated from well-known local correlations.11

Bubbles of oil and gas will rise faster than waterbecause of their buoyancy. The oil slip velocity,which is the difference between the average oiland water velocities, is a function of oil density,water density, water holdup, flow regime andpipe deviation. The gas slip velocity, which is thedifference between the average gas and liquidphases, is a function of gas holdup, flow regimeand pipe deviation. Even when the above compu-tations have been completed, the task of con-verting downhole flow rates into surface flowrates remains.

It is not surprising that after the first manda-tory hand calculation of a production log flowrate, the production engineer quickly turns tocomputers and software packages. Since the mid-dle 1980s, Schlumberger has used the ProductionLog Quick Look (PLQL) program for wellsite inter-pretation. The PLQL module features tight inte-gration between data acquisition and subsequentinterpretation, and in skilled hands, the interpre-tation begins as soon as the last logging pass iscompleted. Very often, the interpretation is fin-ished before the tools are pulled back to the sur-face. PLQL software is now being replaced withSingle Pass Rate Interpretation (SPRInt) software,which is more user friendly and supports all thenew production imaging tools.

34 Oilfield Review

New Technology for Production Services

Traditional production logging measurements of pressure, temperature, density and spinnerrate have been redesigned to deliver theimproved accuracy and resolution required in today’s complex production logging environ-ments. The new, compact integrated technology,called the PS PLATFORM tool string, combinesreliability with modern sensor design. Electricprobe water holdup sensors have been collo-cated with the spinner and integrated with an X-Y caliper for pinpointing fluid entries deepin the wellbore.

The length of the tool string can varydepending on the required sensor configuration.The basic tool string containing three tools can provide three-phase diagnosis in verticaland deviated wells (next page). This config-uration is 18.5 ft [5.6 m] long. Starting from the bottom, the tool string consists of the Flow-Caliper Imaging tool, the Gradiomanometer tooland a Basic Measurement tool.

The Flow-Caliper Imaging tool—The Flow-Caliper Imaging tool is a high-resolution direc-tional fullbore spinner flowmeter equipped withFloView electrical probes—mounted on each of the four self-centering linked X- and Y-caliperarms—that deliver water holdup images andhydrocarbon measurements. The electrodes pro-vide a measure of the conductivity of the bore-hole fluid, thus discriminating between water(conductive) and hydrocarbon (nonconductive)components. The probes are capable of measur-ing droplets of fluid as small as 0.04 in. [1 mm]in diameter. The ratio of probe conduction timeto total time is a measure of local water holdupat the probe site. The transition from conduc-tive to nonconductive fluid provides a measureof bubble count. An internal relative-bearingsensor allows recording of the spatial orienta-tion of the four probes in the pipe. The orienta-tion is used to construct a holdup map of thewellbore derived from the local probe measure-

ments. Laboratory flow-loop tests and fieldexperience show that the presence of the tooldoes not seriously affect the flow regime.

A fullbore high-resolution directional spinneris positioned within the caliper in the plane of the electrical probes permitting collateralflow and holdup measurements within 16 in. [40 cm] of the end of the tool—an advantagewhen logging wells with reduced sumps. Thespinner has a low threshold and high resolution,yielding accurate flow-rate measurements aslow as 2 ft/min [0.6 m/min]. The four-armcaliper can determine the diameter and eccen-tricity of the wellbore.

Gradiomanometer tool—The Gradiomanom-eter tool provides a global or wellbore-averagedmeasurement of the fluid density using a solid-state differential pressure measurement. Thepressure sensors have been designed to delivera fluid density resolution of 0.002 g/cm3. Thepressure ports face away from the outside of the tool, minimizing the risk of damage to thesensor from fluid jetting from perforations or the formation.

Basic Measurement tool—The BasicMeasurement tool consists of a sensor packagecontaining several other traditional measure-ments needed in production logging and thedata-acquisition electronics package. It con-tains either a power supply and wireline teleme-try system for real-time operations or downholebattery and recorder for memory operations.The wireline telemetry system is capable oftransmitting up to 150 kbps on monocable andup to 200 kbps on coaxial cable. Recorder elec-tronics and the associated lithium battery allow

1. Treseder RS, Baboian R and Munger CG: NACE CorrosionEngineer’s Reference Book, 2nd ed. Houston, Texas, USA:National Association of Corrosion Engineers, 1991.

10. Catala et al, reference 3.11. McCain WD: The Properties of Petroleum Fluids. Tulsa,

Oklahoma, USA: PennWell Books, 1973. See also: FluidConversions in Production Log Interpretation. Houston,Texas, USA: Schlumberger, 1974.

Spring 1999 35

more than 100 hours of continuous data record-ing in memory mode. Logging sessions can be scheduled to allow much longer, sophisti-cated logging programs that include many daysof stabilization and buildup times for downholeformation testing.

The sensor package in the Basic Measure-ment tool contains a gamma ray and casingcollar locator used for accurate, precise depthcontrol. It also contains a temperature gaugeand a CQG Crystal Quartz Gauge pressuresensor that delivers high-resolution andextremely accurate pressure measurementswith sufficiently fast temperature stabilization,required when logging flow profiles with high-temperature gradients. If necessary, a Sapphirepressure gauge sensor with improved dynamicresponse and stability, but similar pressureresolution, can be used instead of the CQGsensor. A UNIGAGE well test quartz recordercarrier is an optional tool that provides a sec-ond high-accuracy pressure measurement tradi-tionally required for validation and redundancyinsurance on key high-profile wells.

This PS PLATFORM system is also combin-able with other tools for production loggingservices such as the RST Reservoir SaturationTool service useful for the WFL Water Flow Logtechnique to determine water-velocity informa-tion, the PVL Phase Velocity Log technique tomeasure both oil and water velocity, and car-bon-oxygen (C/O) measurements for evaluatingformation saturation behind casing. Also, the PS PLATFORM tool is combinable with the Slim Cement Mapping Tool (SCMT) service forthrough-tubing cement evaluation. Tool stringsare subjected to a harsh environment, so pack-aging must meet logging-while-drilling (LWD)specifications for logging where tools are usuallyconveyed by coiled tubing in memory mode. Allthe tools in the PS PLATFORM string are corro-sion resistant and exceed National Associationof Corrosion Engineers requirements.1

Telemetry

Gamma ray

Battery

Recordermode

Telemetry mode

Memoryelectronics

Pressure,temperature,CCL

Power supply

FloView electric probes

Directionalfullbore spinner

IndependentX-Y calipers

Gradiomanometer tool

Flow-Caliper Imaging tool

Basic Measurement Tool

18.5 ft

< Basic PS PLATFORM tool string. The basic tool string consists of the Flow-Caliper Imaging tool that providesfluid velocity with the directional fullbore spinner, hole size and geometry (eccentricity) with the X- and Y-caliperarms, water holdup and hydrocarbon bubble-count measurements with the electrical impedance probes; the Gradiomanometer tool that provides fluid density measurements; and a Basic Measurement tool that contains a common sensor package with tempera-ture, pressure, gamma ray and casingcollar locator (CCL) sensors. This tool can be equipped with either a memory for slickline operations or telemetry module for monoconductor or multicon-ductor cable wireline operations.

Measurement Integration Modern production logging interpretation, partic-ularly in complex multiphase flow, has beenmade less complicated by integration of localand global measurements—especially criticalholdup and velocity measurements made at thesame point in the wellbore—provided by theFlow-Caliper Imaging tool. New and moretractable algorithms, fewer parameters, and the fact that the measurement methodologyrequires only a single logging pass, all lead to the ability to perform fluid-flow interpretationduring logging.

With the new measurements, solving formultiphase flow in vertical or deviated wells isrelatively straightforward. Spinner rates arerelated directly—through calibrated spinnerpitch, efficiency and threshold parameters—tofluid mixture velocity. Wellbore size (from caliper)and mixture velocity (from spinner) are used todetermine mixture flow rates and to make frictioncorrections to bulk fluid density measured with a

Gradiomanometer tool. Local electric-probe-based water holdup measurements combinedwith corrected fluid density enable oil, water andgas holdup determination. The bubble velocityand, therefore, hydrocarbon flow rate can beestimated from the average bubble size and thebubble count rate also measured with the electricprobes. The average bubble size can be derivedfrom an empirical model—where bubble sizedepends on the water holdup (below). Finally, oiland gas slip velocities can be estimated, enablingthe flow rate of each phase to be calculated.

The SPRInt software eliminates the dataredundancy of a conventional spinner calibrationplot. Instead, just one pass of the spinner isneeded, together with well-known, accurate spin-ner threshold and spinner pitch. The eliminationof multiple spinner passes allows real-time inter-pretation during the logging pass. However, theSPRInt program can do even more. Based on anew bubble-size correlation, a SPRInt algorithmcan compute the discontinuous-phase flow rate

from local probe holdup measurements. Thiseliminates the need for unreliable empirical cor-relations used for slip velocity. The SPRInt pro-gram has been designed for the local probemeasurements provided by the PS PLATFORMtool. This means that the display of the computedflow rates is closely integrated with the imageplots of local holdup.

An example from the North Sea shows thevalue of the new measurements leading tounambiguous pinpointing of unwanted fluidentries. A 31º deviated well with 5-in. [13-cm]casing was producing oil, water and a vastamount of unwanted gas. The objective of pro-duction logging was to obtain a flow profile andlocate the source of undesired water and gas.The PS PLATFORM tool string, consisting of theFlow-Caliper Imaging tool, the UNIGAGE welltest quartz recorder, the Gradiomanometer tooland a Basic Measurement tool equipped with atelemetry module and a Sapphire pressure gaugesensor, was used.

The logs clearly show the first productioncoming from the lowest perforations only a fewfeet above total depth at X645 ft (next page). Thechange in water holdup can easily be seen intrack 2 and the bubble count in track 4. Thesecorrelate well with the changes in fluid density intrack 3 and spinner rotation rate in track 5,although the bubble count and holdup curvesgive much sharper indications of hydrocarbonentries than the spinner or density curves.Additional evidence of production can be seen inthe second and third sets of perforations at X440and X320 ft. The real-time interpretation showsthat the lowest interval was producing oil at 150 B/D [24 m3/d] and water at 800 B/D [130 m3/d]. The second entry is producing oil at 500 B/D [80 m3/d] and water at 2200 B/D [350 m3/d]. The top set of perforations at X320 ftis producing a large amount of water-free gasand oil at 2000 B/D [300 m3/d].

36 Oilfield Review

> Estimating bubble size from water holdup. The average bubble size can be determined from a phenomenological model (solid curves), wherebubble size depends on the water holdup. The minimum bubble sizeoccurs when the water holdup is 100% and depends on the hydrocarbondensity. The maximum bubble size expands asymptotically to the size ofthe casing pipe. Field and laboratory data (circles) are compared with themodel predictions. These data were inferred from the bubble-count ratemeasurements and assume the bubble velocity was the same as that ofthe fluid mixture. The bubble velocity and therefore flow rate can be esti-mated from the average bubble size and the electrode bubble-count rate.

Water holdup, %

Bub

ble

diam

eter

, mm

0 20 40 60 80 100

0

5

10

15

Flow loop

Model

North Sea well

Middle East well

Spring 1999 37

The bubble flow-rate curve in track 7, derivedfrom the electric probe bubble-count rate, agreeswell with independent total oil and gas flow-rateestimates from spinner and holdup measure-ments. It is interesting to see estimates of bub-ble size grow from 1.3 mm at the bottom entry to12 mm at the top of the logged interval. This indi-cates that initially the bubbles are close to theirminimum size—where they first enter and watercut is nearly 97%—and then start to coalesceand increase in size at each additional hydrocar-bon entry. These relative sizes are in agreementwith laboratory studies and observations made inother wells worldwide.

To control this water entry, the operator had several options. One alternative was plug-ging the well at X400 ft to reduce water produc-tion. However, other wells in the area werere-evaluated, which indicated that the bottom ofthe well could be cleaned and reperforated. Thiscourse of action was chosen, and increased oilproduction by 500 B/D.

Detecting Difficult Flow RegimesIn horizontal wells, phase separation can be totaland fluid phases traveling at very different veloc-ities can be difficult to detect. PVL logs, WFL logsand TPHL three-phase fluid holdup logs can help.Recent developments in pulsed-neutron cased-hole spectroscopy, called RSTPro technology,have improved borehole fluid analysis. Thesedevelopments include enhanced spectral calibra-tions, new spectral standards and better signalprocessing techniques.12 Other improvementsinclude MAXIS Multitask Acquisition andImaging System software and hardwareupgrades resulting in refined operational and logquality control. Today, TPHL measurements madewith the RSTPro tool are wellbore deviation inde-pendent. The velocity of gas slugs can now bemeasured from inelastic spectra to derive a gasflow rate. The RSTPro tools can be combinedwith the PS PLATFORM service using monocableor coaxial cable telemetry systems. RSTPro toolsare rated for high-temperature operations—175ºC [350ºF] and above, with special flaskedpressure housings.

0.75

1.00

Bubble

0.00

0.50

0.25

holdup

0.75

1.00

0.00

0.50

0.25

rps 600

5000 bubbles/sec

°C 112102

Temperature

Bubble count

Spinner

psi 14501200

150 mm

150

Bubble size (nominal)

Bubble size (apparent)

PressureWater holdup

% 1000

10 g/cm

Mixture density

10

Gradiomanometer

g/cm 3

API

ft/hr 50000

63.52000 in.

in. 63.5

Caliper Y

Caliper X

Cable speed

GR

Depthft B/D 25,0000

Water

Gas

Oil

Flow rates

B/D 25,0000

Bubble rate

X600

X500

X400

X300

Water andoil entry

Water andoil entry

Oil and gas entry

AmplifiedCCL

Watercount

3

> Pinpointing oil production with production logs. The gamma ray and perforations (black bars) areshown in the depth column on the left. The X-Y caliper data and an amplified casing collar locator (CCL)are in shown in track 1. The caliper and CCL indicate scaling or casing corrosion, as well as possibledamage to the perforations due to the casing guns at X640 ft and X315 ft. The water holdup imagemapped from the electric probes is shown in track 2. Solid blue indicates 100% water and solid whiteindicates 100% hydrocarbon. The top of the deviated wellbore is aligned with the middle of the colortrack, as indicated by the lighter hydrocarbon shading along the middle of the display. Track 3 containsaverage water holdup computed from the probe measurements (green), raw fluid density measured fromthe Gradiomanometer tool (red) and friction-corrected fluid density (purple). A bubble-count map fromthe electric probes is shown in track 4, and the average bubble count (green) is shown in track 5 alongwith fluid temperature (purple) and spinner rotation rate (red). Apparent bubble size (solid green) andnominal bubble size (dashed green) are computed and displayed in track 6 with the downhole pressuremeasured with the Sapphire gauge (red). The real-time computed oil (green), gas (pink) and water (blue)flow rates are shown in the composite display in track 7 along with the bubble-flow rate (dashed green).

12. Harness P, Shotts N, Hemingway J, Rose D and van derSluis R: “Accurate Oil Saturation Determination andMonitoring in a Heavy Oil Reservoir,” paper SPE 46245,presented at the SPE Western Regional Meeting,Bakersfield, California, USA, May 10-13, 1998.

An example from Alaska shows how TPHLmeasurements can help distinguish difficult flowregimes in highly deviated and undulating well-bores.13 Both the FloView Plus fluid imaging andRSTPro tools were deployed to identify fluid entry(above). The interval logged contained two longhorizontal segments at different depths con-nected by a 250-ft [76-m] uphill segment. Thewater holdup answers from TPHL and FloViewPlus logs are similar over most of the intervallogged. Above XX500 ft, the TPHL measurementsdetect a low water flow in the uphill and first hor-izontal segment, but the FloView Plus tool is

unable to measure the low water flow due to thehigh water velocity and low holdup. The highestpoint in the second horizontal segment occurs atXX560 ft, and since little fluid is produced beyondthis point, the water holdup is high in thisextended, slightly down-dipping section of thewellbore. In the wellbore from XX300 ft to XX560ft, the produced fluids are flowing down the uphillsection. Here, as expected, the oil and watertravel faster than the low-density gas and conse-quently have a lower relative holdup. The TPHLmeasurements are not sensitive to fluid velocityor droplet size, and can detect the high-velocitywater even at low holdups. Above XW800 ft, thewell becomes more vertical, slowing the oil flowand increasing its holdup.

Post-Logging AnalysisComplementing the quick-look analysis at thewellsite are other, more powerful software pack-ages at regional computing centers. Post-log pro-cessing often provides better understanding inthe most difficult-to-understand environments.Of comparable vintage to PLQL software is theProduction Log Global (PLGLOB) software. Thiswas the first production log software to use theglobal approach of tool response models, over-determined information sets and flow-rate solu-tions that minimized reconstruction errors. As inthe case of PLQL software, the outdated humaninterface, lack of support for horizontal wells,and poor integration with newer imaging toolshas meant that the PLGLOB program is alsobeing retired. The replacement is the BorFlowprogram.

The BorFlow program is one of the growingsuite of GeoFrame programs. Embodying theglobal approach of PLGLOB analysis, this pro-gram includes a graphical user interface,improved analysis for horizontal wells, and sup-port for the new generation of imaging tools interms of graphical display and interpretationmodels. To improve user access, the BorFlowprogram comes in a standard Unix version and anemulation version that runs stand-alone on high-end PCs.

Formation MonitoringAnother development in pulsed-neutron spec-troscopy called SpectroLith processing hasimproved the ability to perform saturation moni-toring behind casing using the RSTPro tool. In anexample from West Texas, USA, the RSTPro toolwas run through a 25 p.u.-section of the SanAndres formation to establish a baseline oilsaturation for carbon dioxide [CO2] flood monitor-ing. The reservoir section is primarily dolomite,with varying proportions of anhydrite, gypsumand eolian silts composed of quartz, orthoclasefeldspar and illite clay. Reservoir production was initially through a natural bottomwater driveand limited gas-cap expansion. Waterfloodingbegan nearly 30 years ago for secondary recov-ery, and a tertiary CO2 flooding program has justbeen initiated.

The operator drilled a new production-moni-toring well and acquired a full suite of openholelogs including the FMI Fullbore FormationMicroImager log and the CMR CombinableMagnetic Resonance log. Full core was alsotaken. The goal of the RSTPro monitoring pro-gram was to detect flood fronts and oil banking inorder to optimize sweep efficiency and minimizerecycling of water or CO2. Zones of high perme-

38 Oilfield Review

> Three-phase RST and FloView Plus logging measurements. Track 1 contains water holdup from RST(solid blue) and FloView Plus measurements (dotted blue). Some stationary measurements (circles)taken with both tools are plotted with the continuous logging results. Tracks 2 and 3 show oil (green)and gas (orange) holdup measurements. Track 4 is the two-phase water and hydrocarbon holdup fromthe FloView Plus electric probe measurements plotted along with the well trajectory of the horizontalsections of the wellbore. Track 5 shows the TPHL three-phase holdup log plotted against the wellboretrajectory. Track 6 is a flow-rate profile computed from the TPHL holdup data and velocity measure-ments from the other tools in the PLT Production Logging Tool string. Nearly continuous perforationsare along both horizontal sections.

%Water holdup Water holdup Water flow profile

Hydrocarbon holdup

X720 X685True vertical depth

ft

Oil holdup Oil flow profile

Gas holdup Gas flow profile

X720 X685True vertical depth

ft B/D0 35,000Total flow rate

FloView RST

Depthft

FloView

GR

API

0

0 200

100

RST

FloView

RST Water

RST

Holdup

0 100

100RST

0RST

%

%%

%

0 100%

0 100%

0 100%

0 100

RST

Oil Gas

0 100

Firsthorizontalsection

Secondhorizontalsection

XX000

XW800

XY000

XX500

Spring 1999 39

ability can create local zones of easy production,eventually leading to waterflood breakthrough,also known as fingering. If such zones aredetected, plugs can be used to make the produc-tion more uniform over the thick reservoir, mini-mizing fingering.

Carbon-oxygen ratio (C/O) measurements area sensitive indication of oil and water concentra-tions in formations, and are frequently used tomonitor oil saturation and detect the changingposition of oil-water contacts. However, C/Ointerpretation is strongly influenced by the car-bon contribution from limestone or dolomite, andthe effective porosity of the formation. As theamount of carbonate in the rock varies, the rela-tive carbon and oxygen elemental concentrationsof the formation change, and can result in errorsin the interpreted oil saturation. Using theRSTPro SpectroLith interpretation and the ele-mental gamma rays from calcium, silicon, sulfur,iron and other elements, a reliable estimate ofthe formation lithology can be made (right).

This interpretation also determines the for-mation evaporite content correctly from the cal-cium and sulfur yield, which is quite important inthis case since gypsum is the dominant evaporitemineral (not anhydrite). Determining porosityaccurately in gypsum-bearing carbonates can bequite problematical from standard interpretationmethods in which porosity is usually overesti-mated due to the water of hydration associatedwith gypsum. Therefore, this quantitative litho-logic knowledge helps to determine porosity andoil saturation from C/O logging, leading toimprovements in the accuracy of cased-hole oilsaturation monitoring and enhanced recoveryprogram planning.

Remedial OperationsAfter production logging is completed and a wellproblem has been identified, analysis of differentoptions is performed with the help of the produc-tion technologist. Sometimes additional opinionsare needed. For example, assistance fromSchlumberger Holditch-Reservoir Technologiesmay be enlisted for extensive reservoir analysis.Depending on the operator’s long-term strategy forthe well, courses of action may include workoveroperations such as pumping cement or a coiledtubing operation performed by Schlumberger; aMechanical Plug Back Tool (MPBT), through-tubingplugback performed by Schlumberger Wireline &Testing wireline or slickline; or Camco gas-lift opti-mization (see “Artificial Lift for High-VolumeProduction,” page 48 ).

There are often multiple options for improvingwell productivity. The cost of well intervention ishigh, so minimizing unnecessary or unsuccessfulintervention is vital. Some may be long-termsolutions, while others may provide temporaryimprovement and permit or even require furtherintervention at a later date. The final choice maydepend on something as simple as time con-straints or the availability of a workover rig. Eachwell has a different set of problems leading to awide variety of possible integrated solutions. Thenext examples of remedial solutions demonstratethe wide scope of technology that can beemployed in innovative and economic ways tohelp maximize production.

Production logging detects unwantedinjection water—The Dunbar field is 84 miles[134 km] east of the Shetland Islands in the UKsector of the North Sea. The field is under jointownership by ELF Exploration UK and Total OilMarine. The field consists of several accumu-lations of high structural and geological com-plexity. Recently drilled wells have multipledrains, including horizontal and multilateraldrains. Production logging of new and existingdevelopment wells plays a major role in fieldmanagement.14 The reservoir drive uses a combi-nation of natural depletion and water injection.Several aspects of water production are of cru-cial interest. Scaling caused by mixing of injectedseawater and formation water is a potentialproblem and can have an immediate effect onwell productivity.

One well, drilled into the Brent reservoir, wasflowing a mixture of volatile hydrocarbons whenit started to produce water. The primary loggingprogram objective was to determine the produc-tion profile under flowing and shut-in conditions.This would help identify where the water wasentering from the large number of perforatedintervals over the 394-ft [120-m] long producing

13. Morris F and Hemingway J: “Continuous Oil, Gas andWater Holdup Using Pulsed-Neutron SpectroscopyTechniques,” Transactions of the SPWLA 40th AnnualLogging Symposium, Oslo, Norway, May 30-June 3, 1999,paper O.

14. Hervé X, Bouroumeau-Fuseau E, Quin E and Bigno Y:“Advanced Production Logging Applications in DunbarField,” World Oil 220, no. 1 (January 1999): 45-50.

> RST SpectroLith processing. Formation lithology in track 1 (shown as dry weight fraction) isderived from the RST spectrally measured dry-weight elemental concentrations in tracks 2 to 8.Aluminum is not measured, but is calculated from silicon, calcium and iron. Knowing formationlithology is critical for determining porosity and making correct oil-saturation interpretations withRST logs. Many of the formation capture gamma rays measured by the RST tool come from therocks themselves, thus providing high-quality quantitative carbonate, evaporite, clay and silt discrimination. In this example, silt is composed of quartz, feldspar and mica.

Formation

Quartz, feldsparand mica

Carbonate

Clay

Anhydrite andgypsum

Al 0 20

Si0 20

Ca0 20

Fe

Elemental concentrations, dry weight %

Depthft

X700

X750

X800

0 20S

0 20Ti

0 20Gd

0 20

section. Secondary objectives were to quantifythe extent of any crossflow under shut-in condi-tions and assess scale deposits in the liner. Scalecan cause diameter changes, which can reduceproduction and affect flow-rate computations.

Interpretation of the flowing measurementsshows zones of significant water production atZones A and B in the wellbore (below). Theseintervals also coincide with diameter reductionsshown by the caliper tool. Significant correlatedincreases in the gamma ray response confirm thepresence of scale deposits in the liner. In compar-ing flowing and shut-in conditions, the FloViewdata clearly show the wellbore oil-water contact(OWC) moves from Zone C, where produced watersettles during shut-in conditions on top of the

completion mud in the wellbore sump, to Zone B,148 ft [45 m] higher during flowing conditions.Without the information in the well-flowing logs,the analyst might incorrectly locate the wellboreOWC. A crossflow from the lower into the upper-most perforations can be seen in the shut-inFloView data, even though the flow rate from thelower perforations in Zone C is small, only 45 B/D[7 m3/d]. Scale buildup on the lowest perfora-tions, also confirmed by the gamma ray andcaliper logs, could account for reduced productionin this zone.

The PS PLATFORM logs confirmed many ofcomplex characteristics of the Dunbar field thatuntil now had not been possible with traditionalproduction logging measurements. By running

the PS PLATFORM tool string in another well, theoperator, Total Oil Marine, was able to determinethat water was coming from a particular layer. Asimple solution was employed. After a plug wasset above the same layer in the supporting injec-tor well, water production stabilized and oilrecovery increased.15

Through-tubing water shutoff techniques—Water shutoff techniques that can be deployedthrough tubing offer cost-effective ways ofreducing water cut from a well, while increasingtotal oil production. Through-tubing water shut-off methods can be classified into three main cat-egories: pumping directly through the tubing withgel treatments, electric wireline deployment ofinflatable sleeves or mechanical bridge plugs,

40 Oilfield Review

> Detecting water entry in an oil producer. Several passes of the production taken while the well is flowing (left) are compared with thosetaken while the well is shut-in (right). Perforations are shown between tracks 3 and 4 in both the flowing and shut-in logs. In the well-flowinglogs, the location of oil-water contact in the wellbore at Zone B is obvious in the FloView holdup logs in track 2. The density logs in track 1are more ambiguous. In Zones A and B, significant water production can be seen in the oil (green) and water (blue) flow-rate logs in track 4.The gamma ray in track 6 indicates that scale buildup is occurring in the liner. In the shut-in logs, crossflow can be seen in the fluid velocitylog in track 3 and the oil and water flow-rate logs in track 4. The crossflow, due to differential pressure support, enters this well through thelower perforations at Zones B and C, and is taken by the reservoir through the upper perforations at A.

Depthm

Water holdup

%0 120

Densityg/cm3

0.2 1.2

Vapparentm/min

–10 100

Flow rate

Per

fora

tions

Per

fora

tionsWell flowing Well shut-in

B/D–1k 15k

Bubble countBubbles/sec0 600

Gamma rayAPI

0 1000

Densityg/cm3

0.5 1.2

Water holdup

%–20 120

Vapparentm/min

–2 4

Flow rateB/D

–200 500

Bubble countBubbles/sec0 10

X740

X760

X780

X800

X820

X840

CX860

B

A

Zone

Cro

ssflo

w

Spring 1999 41

and finally coiled tubing deployment of inflatableplugs and cement squeezes. Mechanical bridgeplugs, such as those done with the MPBT plugs,have been successfully set on wireline in a vari-ety of well conditions including crossflows,hydrogen sulfide [H2S], CO2 and dry gas. Plugshave been set in gravel-pack screens, perfora-tions and in openhole. Almost any application orwell problem that once required pulling tubingand setting a cast-iron plug can be reliably reme-died using the Schlumberger PosiSet plug. Theseplugs, which come in sizes from 111⁄16 in. for 41⁄2-and 51⁄2-in. casing sizes to sizes large enough for95⁄8-in. casing, can be accurately placed with aCasing Collar Locator (CCL) log correlation.

An operator in an offshore California, USAlocation had a well originally producing 9000BOPD [1430 m3/d] when water productionincreased to 26% at the expense of oil produc-tion. Production logs showed that most of thewater was coming from the lowest 100 ft [30 m]of a 260-ft [79-m] perforated zone. The traditionalsolution would have involved moving an expen-sive workover rig in for remedial work. A fasterand lower-cost solution was proposed by the pro-duction services team. A PosiSet plug was runthrough tubing and set 70 ft [21 m] above the bot-tom perforation (right). An additional 30 ft [9 m]of cement was placed on top of the plug toensure integrity. When the well was put back onstream, production had increased to 11,500BOPD [1830 m3/d] with only a 4% water cut.

When the PosiSet plug was set, a nonexplo-sive deployment system using a hydraulicallypowered tensile rod was pulled through the plug,compressing the seal against the casing wall.The upper and lower arms have opposed carbide-tipped locking anchors and provide strong metal-to-metal mooring of the tool. The seals andlocking anchors are self-centering, allowing theplug to be set in deviated wells. Once anchored,the smallest PosiSet plug can withstand a mini-mum of 25,000 lbf [111 kN] without moving, andstrength increases to 90,000 lbf [400 kN] forlarger plugs. The entire operation can be donequickly because the PosiSet plug seals as soonas it is set. Depending on its size, the PosiSetplug has a differential pressure rating from 500to 1500 psi [3.5 to 10.3 MPa], and a cement plugis usually built up on top to further improve dif-ferential pressure rating. Typically a 10-ft [3-m]cement plug will increase this rating to 3000 psi[20.6 Mpa] for a 95⁄8-in. casing, increasing to amuch higher rating for smaller casing sizes.

15. Quin E and Hervé X: “High Tech for Deep Probes,” TOMTOM, no. 65, produced by The Corporate Communica-tions Division of Total Oil Marine, Aberdeen, Scotland(December 1998): 7.

31/2-in.tubing

7-in.casing

Packer

Wat

er p

rodu

cing

Oil

prod

ucin

g

Before PosiSet plug9000 BOPD at 26% water cut

Running inPosiSet plug

After PosiSet plug 11,500 BOPDat 4% water cut

Perforations

Setting tool

Anchors

PosiSet plug

Anchors

Cement plug

> Stopping water entry with a PosiSet plug. The PosiSet plug was run through tubing and set in casing within a set of perforations. Oil production increased from 9000 B/D [1430 m3/d] with a 26% water cut to 11,500 B/D [1830 m3/d] with a 4% water cut.

PatchFlex technology—Repairs often needto be made without blocking entry into lowersections of the wellbore. Schlumberger is thesole provider of services based on Drillflex prod-ucts, which include many rigless remedial solu-tions such as the new PatchFlex through-tubing,inflatable casing-patch technology (above).Conveyed on electric wireline and based on theuse of in-situ polymerization, PatchFlex technol-ogy forms a hard and impermeable lining, flushwith the well casing with minimal internal diam-eter loss. PatchFlex service can be used to sealperforations for water shutoff, gas shutoff,selective injection profile modification and tub-ing or casing repair.

BP Exploration and Production uses PatchFlextechnology to control unwanted water productionin their Forties Delta field in the North Sea.16 TheForties field is one of the first and largest fieldsdiscovered in the North Sea. First oil was deliv-ered in 1975, and a maximum oil production rateof 500,000 B/D [80,000 m3/d] was achievedbetween 1978 and 1980. Oil in place was esti-

mated to be 4.2 billion bbl [0.67 billion m3] withan estimated 60% recovery. To date, almost 57%of the original oil in place has been produced.

In the late 1980s, efforts began to reversedeclining production rates. Artificial gas-lift facil-ities were installed on the four main platforms,and since then over 50 wells have had theirproduction liners replaced, or have been drilledto new infill locations. Nevertheless, by 1997 theoil production rate was down to 86,000 B/D[13,700 m3/d] with an associated water produc-tion rate of 270,000 B/D [43,000 m3/d] and aver-age water cut of 76%, incurring high waterseparation and disposal costs.

In one example, a Forties Delta well wassidetracked in 1995 to replace the old productionliner, and the 56º deviated sidetrack encountereda thick oil sand. Two sections totaling 33 ft [10 m]along the lower E-sand were perforated, and one21-ft [6.5-m] section along the upper H-sand wasalso perforated. When the well came on-line,water cut increased rapidly. The original produc-tion logs showed that one-third of the productionwas from the lower E-sand with about 50%

water cut. The upper H-sand had a water cut of87%. The logs also indicated that H-sand pro-duction was dominated by flow through a cementchannel behind the liner, even though earlycement bond logs had indicated good cementover the entire perforated interval. The originalproduction logs also indicated approximately 500 B/D [80 m3/d] crossflow from the H-sand into the E-sand with a pressure differential of 50 psi [345 kPa].

Two years later, a cement squeeze was usedto repair the cement channel causing the excesswater production from the H-sand, but after aninitial decline in water production the water cutincreased again. New production logs run afterone year showed that the H-sands were againproducing two-thirds of the total flow, but now ata 100% water cut. The lower E-sand water cut

42 Oilfield Review

Cable

Cable head

CCL

Electronics

Pump

MechanicaldeflateInflatable settingelement (ISE)Core(resins and fibers)

Permanent sleeve

External seals

Outer skin Run-in Anchor Progressive inflation

Polymerization Deflate ISE Extract ISE

A B C D E F

> PatchFlex setting sequence. When the patch sleeve is in place (A), a surface-controlled module activates a pump in the downhole running tool andbegins expansion of the inflatable setting element (ISE). The lower end inflates first (B) acting as an anchor for the rest of the patch. The inflation—controlled by progressive expansion rings that break—progresses upward (C) preventing any well fluid from becoming trapped behind the permanentsleeve and creating a hydraulic lock. Heating begins when the PatchFlex patch is fully inflated (D). Electric current from the surface is supplied through the electric wireline multiconductor cable to the heating elements in the inflatable setting element to heat the resin to the polymerization temperature. Heat locks the resins in place and gives the composite sleeve the required mechanical properties. The resin-curing temperature is monitored to verify the polymerization process and controlled to ensure that all parts of the permanent sleeve are fully hardened. After polymerization is complete and thesleeve hardened, the pump is reversed to deflate the inflatable setting element (E). When the pressure has been bled-off, the inflatable setting element is pulled through the permanent sleeve patch and recovered at the surface with the running tool (F).

16. Saltel JL, Leighton J, Morrison J, Welch R and Pilla J:“Water Shut-Off Using an Inflatable Composite SleevePolymerised In-Situ. A Case History on Forties Delta,”paper SPE 50620, presented at the 1998 SPE EuropeanPetroleum Conference, The Hague, The Netherlands,October 20-22, 1998.

Spring 1999 43

had also increased to 67%. The primary remedialobjective on this well was to reduce the watercut by isolating the H-sands and producing onlythe relatively water-free E-sand. However, timingwas an issue because of a 6-month drilling pro-gram planned for the platform. There was only ashort time available for well repair before theprogram started; then oil could be producedwhile drilling continued.

Two options existed: a repeat attempt with acement squeeze or use of the new compositepatch technique. The cement squeeze had theadvantage of potentially sealing the channelbehind the liner permanently, permitting subse-quent access to the potential H-sand oil produc-tion. However, the failed first attempt at acement squeeze indicated that the squeezewould probably not withstand the drawdownpressures needed to maximize production. Also,the length of time required for a cementsqueeze would have delayed the start of thedrilling program.

The new PatchFlex composite patch tech-nique has the advantages of ready availabilityin a variety of sizes to fit the perforation inter-val, capability of handling crossflow pressures,and, as a quickly deployed wireline service,induces no delays in the drilling program. It isalso less costly than a cement-squeeze opera-tion. The composite liner is millable usingthrough-tubing techniques if a subsequentsqueeze operation is desired, or it can be reper-forated at a later date to allow access to the H-sands. The only disadvantage of the compos-ite liner is a slight reduction in wellbore diam-eter (below). However, other mechanical patchremediations take up much more of the avail-able casing inner diameter.

The Forties Alliance team reviewed the pro-posed well intervention project and selected aPatchFlex patch solution. The total set lengthwould be long enough to overlap and isolate theentire 21-ft [6.4-m] water-producing H-sand per-forations. Although the technology was new,numerous PatchFlex trials had been made in testwells and no problems setting and centralizingthe patch were expected in this highly deviatedwell. Since the polymerization process is heatsensitive, and earlier PatchFlex field tests weredone only to 85ºC [185ºF], the well was cooled to80ºC [176ºF]—by pumping 500 bbl of inhibitedseawater before running in—preventing prema-ture hardening of the polymer seal. The wateralso helped clean the area where the PatchFlexexternal seal would be set.

A 16-arm caliper run before the job showedthat the liner was approximately to gauge, andno cement nodules or obstruction existed in thesetting zone. The PatchFlex patch, control andmonitoring electronics, CCL tool for depth corre-lation, and cable head were made up in a 4-in.[10-cm] riser, in which a full set of electrical testswas carried out before the assembly was low-ered into the well. The assembly was run in thehole without difficulty at 4000 to 6000 ft/hr [1220to 1830 m/hr], and positioned over the H-sandperforations using the CCL log. After a final sys-tem electronics check, the downhole inflationpump was started.

A 4-gallon [15-liter] volume of borehole fluidwas pumped to inflate the 29-ft [8.8-m] inflat-able setting element (ISE). A differential pres-sure of about 150 psi [1.03 MPa] is required tobreak each of 900 rings that control the pro-gressive expansion. The inflation process is themost critical part of the PatchFlex installation.The inflation process was monitored in fourways: by total pumped volume, by increasedcable tension caused by shortening of thePatchFlex patch, by counting the number of pro-gressive expansion rings broken with a down-hole geophone that can hear rings breaking, andfinally and most reliably, monitoring differentialinflation pressure (above).

Differential pressure reached a plateau of150 psi after 2 minutes and remained nearlyconstant for 12 minutes. Pressure then rose to260 psi [1.8 MPa]—the level of the pressure-relief safety valve. A total of 843 progressiveexpansion rings were counted during the 12- to15-minute inflation period, and the cable wasslacked off from the surface about 2 ft [0.6 m] tomaintain downhole tension, confirming the prop-erly inflated PatchFlex seal.

> PatchFlex cross section. A cross section of aPatchFlex patch set inside a recovered joint ofperforated 7-in. casing shows the sleeve will notexpand through the holes.

> Inflation pressure. The most reliable way of monitoring the inflation of thePatchFlex system is with differential pressure. The pressure (red curve) shouldincrease steadily while the anchor zone is being inflated until it reaches the 150psi [1.03 MPa] required to break the first progressive expansion ring. The differ-ential pressure fluctuates slightly within a 30-psi [210-kPa] range as each pro-gressive expansion ring is broken until the sleeve is fully inflated. At this point thepressure should increase sharply to reach the limit of the pressure-relief valve.

0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0

50

100

150

200

250

300

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Time, min

Infla

tion

pres

sure

, MP

a

Infla

tion

pres

sure

, psi

Safety-relief valve limit

Next, electric power was sent down theseven-conductor wireline cable to heater wires inthe sleeve, and the temperature was monitoredto keep it above 145ºC [293ºF] until the thermo-setting resin in the composite seal was success-fully polymerized. The final step, after thePatchFlex patch was fully hardened, was runningthe pump in reverse at a negative differentialpressure of 70 psi [500 kPa] to separate and col-lapse the inflatable setting element away fromthe permanent composite patch. The ISE wasremoved with a 460-lbf [2.1-kN] overpull withoutany difficulty. No anomalies in the ISE werefound when examined at the surface.

Production results showed that the patch was successful. Initial oil production increased1300 B/D [210 m3/d] and the total water cutreduced to 60%, leading to an additional250,000 bbl [40,000 m3] of oil in the first year(below). The pay-off time from the additional oilproduced was less than one month. From anenvironmental perspective, the patch reducedthe oil-in-water discharge from the platform, andthe savings from the cleanup alone paid back thecost of the repair in less than a year. BP, satisfiedwith these results, continued using thePatchFlex system to treat other wells in theForties field with similar success.

Organizing Production ServicesThe new Schlumberger production maximization,or MAXPRO initiative, has its origins in the GulfCoast of North America. There, in 1998, produc-tion services operations organized into teams offield engineers, sales, technical and manage-ment experts dedicated to identifying areaswhere, in response to customer needs, improvedservice could help maximize production. The firstMAXPRO team, after researching and evaluatingproduction history in the Gulf of Mexico, made aproposal to Chevron on how to optimize produc-tion from one of their largest fields. A ProductionEnhancement Group (PEG) identified under-performing wells, and the MAXPRO teamprovided efficient, cost-effective services toimprove production.

Today, MAXPRO services along withFormation Evaluation and Well Testing make upSchlumberger Wireline & Testing (W&T). Thesegroups work together with other Schlumbergercompanies to offer better solutions to customersworldwide (bottom left). The ability of MAXPROservices to help customers is based on three keycomponents: people, technology and a respon-sive, closely linked support network.

People—First and most important are theexperts who provide the production services andinterface with operators. These experts knowhow to find the most effective way to addresswell intervention, completion design and execu-tion issues. MAXPRO field engineers are skilledin the latest tool technology, operation, measure-ment capabilities and limitations. Their trainingand experience provide efficient and cost-effec-tive production logging and cased-hole services.They interpret logs at the wellsite, recognize pro-duction problems and implement well interven-tion solutions—such as perforating, plugs andcasing patches. Every strategic MAXPRO locationhas a key specialist or production technologist—usually trained as a log analyst or reservoir engi-neer—who applies the latest technologies toimprove and optimize well productivity.

Technology—Technology is the cornerstoneof production services. Schlumberger is currentlydeveloping more technology for improving pro-duction operations than at any other time in itshistory. Listening to operator needs and concernscoupled with years of research and developmenthas led to a new generation of production mea-surements, tools, analysis software and remedialsolutions, aimed not only at difficult new wellssuch as highly deviated and horizontal boreholes,but also at traditional vertical wells where pro-duction has been undergoing gradual declineover decades.

44 Oilfield Review

>World map showing MAXPRO locations. Engineering and manufacturing centers arelocated in Texas, USA; Red Deer, Canada; and Paris, France. Major field MAXPRO loca-tions are found in Aberdeen, Scotland; Al-Khobar, Saudi Arabia; US Gulf Coast region,Texas; Bergen, Norway; Cañadón Seco, Argentina; Duri, Indonesia; Hassi Messaoud,Algeria; Macaé, Brazil, Lake Maracaibo, Venezuela; and Port Harcourt, Nigeria.

x

x

x

xxx

xxxxxx

x xxx

xx

xxx

xxxx xxxxxx

xxxxx

xxx

xxxx xxxx

x xxxx

xx

x xx

xxx

x

x

x

x

xx xxx x x x xxxx

xx

xxx x

xx

x xx

xx x x

x xxxxx

xxxx xx

xx x x

0Jan-97 Apr-97 Jul-97 Oct-97 Jan-98 Apr-98 Jul-98

500

100

90

80

70

60

50

40

30

20

10

0

1000

1500

2000

2500

3000

3500

Oil rate - well testWater cut - wellhead sampleWater cut - well test

Oil

flow

rat

e, B

/D

Wat

er c

ut, %

Cement squeeze Composite patch

>Water control for production enhancement. After the PatchFlex patch was installed,oil production increased 1300 B/D [210 m3/d] leading to an additional 250,000 bbl[40,000 m3] of oil in the first year. At the same time, water production dropped by 6000 B/D [1000 m3/d] (water cut dropped from 87% to 60%) and stayed low.

Red Deer

Texas Gulf Coast

Lake Maracaibo

Macaé

Cañadón Seco

Port HarcourtDuri

BergenAberdeen

Paris

Al-KhobarHassi Messaoud

Spring 1999 45

Linking Solutions to Problems—The MAXPROorganization is committed to providing cost-effective services to improve well productivityeverywhere. Responsive engineering and manu-facturing centers in Canada, France and the USAteam with remote field locations to provide directsupport and rapid, custom, fit-for-purpose solu-tions to specific needs. Intranet communicationsand electronic bulletin boards, monitored bySchlumberger technical experts worldwide, helpaddress new and difficult problems—sometimeseven at the wellsite. The expertise of one personat the wellsite or in the interpretation center isimmediately multiplied through these instantchannels of communication.

Rapid Response ProjectsRapid response teams at two key MAXPROSolution Centers help provide communicationsand engineering coordination with field locationsand clients to develop special fit-for-purposesolutions to specific production problems. Thisresults in a faster solution development cycle—helping to quickly cope with the ever-changingneeds in production services.

Lee Tool, in Red Deer, Alberta, Canada, a newdivision of Schlumberger Canada Limited, isalready providing many engineered solutions forproducing wells—particularly in slickline andmemory logging services. These projects includeutilizing the standard Schlumberger neutronsource to develop a slimhole thermal andepithermal neutron tool for production logging,adapting high-precision pressure sensors for thememory Gradiomanometer specific gravity profiletool and improving production logging tools for the increasingly important high-pressure,high-temperature (HPHT) environment. Specialhostile environment pressure housings were also built for RST operations in extreme H2Sservice conditions.

The Schlumberger Perforating & Testing (SPT)center in Rosharon, Texas, USA has a long historyof developing perforating and completion sys-tems. Over 35 rapid response projects have beendeveloped for MAXPRO production services. Oneof the earliest projects was the development of aFIV Formation Isolation Valve tool.

Completion of the BP Andrew field requiredunderbalanced perforating of the 3600-ft [1100-m] long horizontal section, without exposing theformation to damaging kill fluids, to achieve asufficiently high productivity index for the eco-nomic viability of the project. This was achievedby using the FIV tool to isolate the reservoir afterperforating, allowing long gun strings to be

retrieved without killing the well.17 This devicesaves at least two days and $300K per well, andhas led to an entire range of products. For exam-ple, an operator needed to control pressure whenretrieving perforating guns in gas wells. This ledto the design of the Surface controlled FormationIsolation Valve (SFIV) tool. A single control lineallows the valve to be opened or closed from thesurface and eliminates a trip into the well tooperate the valve mechanically.

A special gun system was developed for anoperator in Norway using heavy-wall casing tocounter the effects of subsidence. The challengewas to make holes in the heavy casing to fracturethe wells. A high shot density (33⁄8-in., 6 spf) gun system was modified in two weeks to pro-duce hole sizes of 0.4 to 0.54 in. in the 65⁄8-in., 66-lbm/ft heavy-wall casing. In South America,an operator needed a gun system to perforatethrough the damaged zone in a high compressivestrength 19-kpsi [130.9-MPa] reservoir rock.These high-performance gun systems for hard-rock formations took longer—nearly 9 months—to develop. The new gun system penetrates 33%deeper than previous systems in high compres-sive strength rock.

In another project, several customized solu-tions have been designed to address the chal-lenge of minimizing sand production in weak, butconsolidated, reservoir formations. Sand produc-tion is affected by the perforated tunnel stability.The stability of the tunnel is a function of the var-ious forces acting on the tunnel, the tunnelgeometry and the position of each tunnel relativeto others. Rapid response projects for severaloperators have been focused on controlling theseparameters. For a North Sea operator, gun phas-ing was optimized with deep-penetratingcharges to produce large shot-to-shot separa-tion—maximizing tunnel stability. In another pro-ject, an upwards shooting oriented gun systemwas developed to reduce stresses on perfora-tions for an operator in The Netherlands.

Expendable Spiral Enerjet—A large-scalerapid response project at SPT was the develop-ment of a fully expendable spiral gun system(right). The request for this system came fromPEMEX in Mexico because a reliable low-cost,deep-penetrating, through-tubing gun systemwas needed to shoot in as many directions aspossible for maximum productivity. A system wasdeveloped and field tested within 3 months of theinitial request, and is now commercially available.

17. Edmonds P: “Linking Solutions to Problems,” OilfieldReview 8, no. 4 (Winter 1996): 4-17.

> Expendable spiral Enerjet gun. The wireline-conveyed, through-tubing Enerjet expendablespiral gun can be set up with either 45º or 60ºphasing. The charges and strip break up intosmall pieces leaving only the cable head, CCL,weights and gun head for recovery. Gun dia-meter pulling out of hole after firing is thesame as that running in, allowing passagethrough tight restrictions.

The new wireline-conveyed, through-tubingEnerjet gun system with charges mounted on anexpendable spiral carrier is available in threesizes: 111⁄16, 21⁄8, and 21⁄2 in. The gun is used toshoot past formation damage with the new, deeppenetrating PowerJet charges. These chargesincorporate new liner designs and manufacturingmethods to provide increased penetration overearlier designs, but are less damaging to the for-mation. Deeper penetration helps increase wellproductivity. Other charge options are available,including HTX high-temperature explosivecharges and big-hole charges for fracturing andgravel packing.18

Controlling the amount and size of the debrisproduced from a gun is important, especially withexpendable systems. This new expendable spiralEnerjet gun system produces less debris volumefrom the charge case, caps, carrier strip andexplosives. The expendable spiral Enerjet gunuses a patented strip, which breaks up into smallpieces after detonation to reduce the amount andsize of the debris. The charges and caps alsobreak into small pieces. Therefore, the volume ofdebris in the wellbore is 25% less than that leftwith previous expendable Enerjet gun systems.

Double Your MoneyIt is particularly important to combine productionservices whenever possible to save rig time andreduce costs. For example, Shell was one of thefirst operators to adopt a novel fit-for-purposeproduction service developed in the North SeaBrent field by running memory production loggingtools below perforating guns.19 Production log-ging data can be acquired before and after firingguns, without having to pull out of the hole (left).This saves many hours of rig time—especiallywhen the technique is extended to coiled tubingoperations—and provides the capability to eval-uate the effectiveness of the perforating programbefore the guns are removed from the borehole.The technique is especially helpful in identifyingand understanding crossflow after perforatingnew zones in a producing well, performing leakinvestigations on recently set casing patches,and determining production and injection flow inthe wellbore before and after perforating.

46 Oilfield Review

Flow rate–2K B/D 8K

Gamma ray0

X200

A

B

C

D

E

F

G

L

Depthft

X250

X300

X350

X400

X450

X500

X550

X600

X650

API 150

Fluid density0.2 g/cm3 1.2

Pressure4.6K psi 5.1K

Flow rate–2K B/D 12K

Fluid density0.2 g/cm3 1.4

Pressure

Well shut-in Well flowing

4.6K psi 5.3K

Cro

ssflo

w

> Detecting crossflow in an oil producer. The gamma ray, in track 1, clearlyidentifies many sand zones (yellow) separated by shale beds (grey). Theold perforations (black bars) in the B-, C-, D- and E-sands shown in thedepth column between tracks 1 and 2 were producing oil with a high water cut. The new perforations (red bar) are in the middle of the L-sand.Production logs with the well shut-in are shown in track 2. These show anunexpectedly high 2000 B/D [300 m3/d] upward crossflow from the L-sandinto the B-, C- and D-sands. Production logs with the well flowing, shown in track 3, confirm that most of the total production is coming from the new perforations.

Spring 1999 47

In another example of two-for-one productionservices, a high-shot density job was run inDuncan, Oklahoma, USA using a 27⁄8-in. [7.3-cm]HSD High Shot Density gun mounted above aspecially built gauge hanger. The gauge hangerpermitted the UNIGAGE recorder to attachbeneath the guns for pressure transient analysisafter the perforations. The guns were shot atnearly 14,000 ft [4300 m] with a full column offluid and an additional 1000 psi [7 MPa] back-pressure applied at the surface. After the gunswere fired, the pressure increased to 2000 psi [14 MPa] at the surface. Pressure transient anal-ysis of the logged UNIGAGE data shows the skinfactor to be low from the perforating process at1200 psi [8.3 MPa] underbalanced pressure. Twoadditional zones were perforated in the samemanner. Testing impulse gauges were used tomonitor one-hour buildup pressures.

Finally, a fit-for-purpose solution wasdesigned by Dowell, Schlumberger Wireline &Testing and GeoQuest to help BP Exploration loga number of wells under difficult weather condi-tions in Alaska, USA (above). On one well, the PSPLATFORM and RSTPro tool combination was runin a well using coiled tubing to a total depth of14,930 ft [4550 m]. During well construction, holeconditions prevented openhole logging and thewell was completed with only basic loginformation. The well produced at 6200 B/D [986 m3/d] with a 99.5% water cut.

Remedial action was needed. Unfortunately,mast height and wellhead conditions imposedspace restrictions and limited logging toollengths. The new, compact PS PLATFORM toolstring was combined with the RSTPro tool toprovide both production logging and formationevaluation through casing. Only one trip into the hole was required for both shut-in and flow-ing production profiles. In addition, the RSTProsaturation interpretation identified lenses of oilin the formation missed during the initial com-pletion. GeoQuest and Wireline & Testing per-sonnel working in Anchorage, Alaska processedthe logs and advised on remedial work. Within

hours of rigging down, the team was back run-ning in the well with the solution—whichincluded setting a plug and reperforating in thenewly found oil zones. The results were highlyprofitable for BP—oil production increased 2921 B/D [464 m3/d] and water production wasreduced by 28%.

The concept of maximizing production for thelife of the well is a driving force that is reinvigo-rating today’s production services. Since opera-tors are drilling increasingly complex wells andat greater depths, new direct measurement,real-time production logging promises to be apowerful tool to describe complex fluid flowregimes in these wellbores. Combined withcased-hole monitoring techniques, new produc-tion logs are becoming indispensable for diag-nosing well-flow problems at the source andconfidently determining remedial solutions.MAXPRO teams are taking advantage of themany technological advances made in the pastfew years to provide proactive solutions to wellintervention, completion design and execution.The MAXPRO focus is designed to meet theneeds of operators and reshape the manage-ment of healthy wells. —RCH

18. Baird T, Fields T, Drummond R, Mathison D, Langseth B,Martin A and Silipigno L: “High-Pressure, High-Temperature Well Logging, Perforating and Testing,”Oilfield Review 10, no. 2 (Summer 1998): 50-67.

19. Davies J, van Dillewijn J, Hervé X and Kusaka K:“Spinners Run While Perforating,” paper SPE 38549,presented at the 1997 SPE Offshore Europe Conference,Aberdeen, Scotland, September 9-12, 1997.

< PS PLATFORM service,coiled tubing and teamwork.Despite extreme weather conditions and limited craneheight, the compact PS PLATFORM tool combined with the RST tool was able tolog, evaluate producing inter-vals and find zones of missedoil reserves in the reservoirbehind casing. The logs helpincrease production nearly3000 B/D [500 m3/d].

Oilfield Review48

Electric submersible pumpHydraulic pumps

Jet Other

Gas liftProgressing cavity pumpRod pumpSpecificCondition

Number of wells

Well depth

Production rate

Well inclination

Dogleg severity

Temperature

Safety barriers

Casing size

Single1 to 20

Greater than 10,000 B/DLess than 2500 ft2500 to 7500 ft

Greater than 7500 ft

Greater than 20Less than 1000 B/D1000 to 10,000 B/D

5 1/2 in.7 in.

9 5/8 in. and largerVertical

DeviatedHorizontal

Less than 3° per 100 ft3 to 10° per 100 ft

Greater than 10° per 100 ftLess than 250°F

250 to 350°FGreater than 350°F

0

4 1/2 in.

1

221231111

1221

1

1

3122

112

1

331231331

1231

1

1

3123

112

3

111111111

1111

2

2

1211

121

1

111121121

1111

2

1

1211

121

12

RequiredNot required

SimpleDual or multiple zones

Greater than 1000 psi100 to 1000 psi

Less than 100 psi

VariablePrimary

Secondary waterfloodTertiary

LowModerate

HighLess than 100 cp

100 to 500 cpGreater than 500 cp

Yes

Stable

1

1111112

1112

1

3113

111

31

1111112

1112

1

3

3112

111

1

1231121

1132

1

1

1111

123

1

1111132

2112

1

2

2112

111

Vapor/liquid ratio

Gas/oil ratio

Treatment

Surface InfrastructureLocation

Electrical power

Fuel

Contaminants

NoLess than 10 ppm

500 to 2000 scf/STBGreater than 2000 scf/STB

Less than 0.10.1 to 1.0

10 to 100 ppmGreater than 100 ppmLess than 500 scf/STB

ScaleParaffin

AsphalteneScale inhibitor

Corrosion inhibitorSolvent

Acid

OnshoreOffshoreRemote

Sensitive environmentUtility

Greater than 1.0

1

112

13221

2221

2

1

2312

231

1

232

12121

1112

2

1

2212

111

1

111

11221

1221

1

1

1121

112

1

222

11111

2222

2

1

1212

231

SCADA1

Space restrictions

Well service

1. Supervisory Control and Data Acquisition

GenerationNatural Gas

YesNo

Workover rigPulling unit

Diesel or gasolineYesNo

Snubbing unitWireline

Coiled tubing unit

2

33

3

11

31

2

33

3

11

21

2

221111111

1221

1

1

3221

112

2

2111122

1122

1

3

3113

112

1

112

12221

2221

3

1

2212

231

1

11

1

11

21

2

221111111

1221

1

1

3221

112

2

2111122

1122

1

3

3113

111

1

112

12221

2221

2

1

2212

231

1

11

1

11

21

1

11

1

11

21

2

23

2

11

11

Reservoir access

Completion

Flowing pressure

Water cut

Fluid viscosity

Corrosive fluid

Sand and abrasives

Stability

Recovery

Produced-Fluid Properties

1 Good to excellent 2 Fair to good 3 Not recommended or poor

Production, Reservoir and Well Constraints

Artificial Lift for High-Volume Production

Roy FleshmanBartlesville, Oklahoma, USA

Harryson Obren Lekic Houston, Texas, USA

For help in preparation of this article, thanks to Rick Baileyand Duane Russell, Reda, Bartlesville, Oklahoma, USA;James Garner, Camco Products & Services, Houston, Texas,USA; Peter Schrenkel, Reda, Dallas, Texas; and Dave Bergt,Schlumberger Oilfield Services, Sugar Land, Texas. NODAL is a mark of Schlumberger. AGH (Advanced GasHandler), CDPS (Cable Deployed Pumping System) and HOTLINE are marks of Reda. Camco EOR (EngineeringOptimization Resources) is a mark of Camco Products &Services. Camco Products & Services and Reda areSchlumberger companies.

Rod pumps bring oil to surface in many fields, but for better flow rates more than

100,000 wells use subsurface electric pumps or inject external gas to lighten the

fluid column. Specialized approaches are needed to optimize existing gas-lift or

submersible systems and to design new installations for more complex applications.

Less than a fourth of producing oil wells flow nat-urally. When a reservoir lacks sufficient energyfor oil, gas and water to flow from wells atdesired rates, supplemental production methodscan help. Gas and water injection for pressuresupport or secondary recovery maintain well pro-ductivity, but artificial lift is needed when reser-voir drives do not sustain acceptable rates orcause fluids to flow at all in some cases. Lift pro-cesses transfer energy downhole or decreasefluid density in wellbores to reduce the hydro-static load on formations, so that available reser-voir energy causes inflow, and commercialhydrocarbon volumes can be boosted or displacedto surface. Artificial lift also improves recovery byreducing the bottomhole pressure at which wellsbecome uneconomic and are abandoned.

Because reservoir pressure declines andmore water is produced late in field life, artificiallift is generally associated with mature oil andgas developments. However, driven by activity indeep water and areas that require constructionof complex wells, the mature state of hydrocar-bon exploitation worldwide has increaseddemand for high lifting rates to produce oilquickly and efficiently at low cost. Offshore andin difficult international regions, artificial-lifttechniques accelerate cash flow, generate profitssooner and help operators realize better returns,even in wells that flow naturally.

Rod pump, gas lift and electric submersiblepumps are the most common artificial-liftsystems, but hydraulic and progressing cavitypumps are also used. Each is suited to certainlifting requirements and operational objectives,

1. Brown KE: The Technology of Artificial Lift Methods, vol.2A. Tulsa, Oklahoma, USA: PennWell Books, Inc., 1980.

Spring 1999 49

but there is overlap between systems dependingon subsurface conditions, fluid types, requiredrates, well inclination angles, depths, comple-tion configurations, lift-system hardware andsurface facilities.

Lift optimization to get the most fluid from awell or field at the lowest cost offers opportuni-ties for substantial production gains in new wellsor mature fields. When selecting and designinglift systems, engineers must consider reservoirand well parameters, but field developmentstrategies should be factored in as well.Artificial-lift selection is specialized and oftentedious, but guidelines provide the relative appli-cability of each method (previous page).1

Artificial-lift technology is well established,but new developments continue to play a role insolving problems and meeting production chal-lenges. Recent improvements reduce lifting coststhrough system components that resist hostileenvironments, optimize power usage and improvereliability. Alternative means of deploying liftsystems allow profitable production from previ-ously uneconomic wells or fields. Traditional arti-ficial-lift limits are expanded by using more thanone lift method in the same well, such as gas liftor jet pumps combined with electric submersiblepumps and progressing cavity pumps driven byelectric submersible motors. This article reviewsbasic lift systems, discusses high-volume artifi-cial lift and presents selection, design and opti-mization strategies along with new gas-lift andsubmersible technology.

< Artificial-lift selection. Making artificial-liftdecisions is primarily a process of choosing thelift methods most applicable to expected sur-face, reservoir, production, fluid and operationalconditions. This table provides applicabilityvalues and selection criteria or conditions for the basic forms of artificial lift. To choose amethod that meets production requirements,select the range that applies—good to excellent(1), fair to good (2) and not recommended or poor(3)—for key criteria, tally these values and weighthe results.

Basic System Descriptions The four basic subsurface artificial-lift groupsinclude rod or progressing cavity displacementpumps; jet, piston, turbine or plunger hydraulicpumps; gas lift; and electric submersible cen-trifugal pumps.2

Rod pumps combine a cylinder (barrel) andpiston (plunger) with valves to transfer well fluidsinto the tubing and displace them to surface.These pumps are connected to surface by a metalrod string inside the tubing and operated byreciprocating surface beam units, or pumpingjacks, that are powered by a prime mover—elec-tric or gas motors—(below). There are two typesof linear-displacement rod pumps. Tubing pumpshave a fullbore barrel with standing valve and areattached to the end of the tubing. A plunger, ortraveling valve, is run into this barrel on the rods.Tubing must be pulled to repair or replace tubingpumps. Smaller insert pumps consist of a barrel,intake valve, plunger and discharge valve com-

bined in an integral assembly run inside tubingon rods. Insert pumps can be retrieved andrepaired or replaced without disturbing the pro-duction tubing by just pulling the rods.

Fluids are pulled into pump barrels by close-fitting plungers with check valves to displacefluid into the tubing. Standing, or intake, valvesconsist of a stationary ball-and-seat. The dis-charge, or traveling valve, moves during eachreciprocating pump cycle. Rod pumps are simple,familiar to most operators and used widely.However, rod pump capacity, or volumetric effi-ciency, is limited in wells with high gas/liquidratios, small tubing diameters or deep producingintervals. Other disadvantages are a large surfacefootprint (space requirement), high capital invest-ment and potential wellhead leaks or spills.

Progressing cavity pumps are based on rotaryfluid displacement. This spiral system consists ofa rotor turning eccentrically inside a stationarystator (next page, top left). The rotor is a small-diameter screw with deep round threads andextremely long pitch—distance between threadpeaks. The stator has one more thread and longerpitch than the rotor, which forms cavities thatprogress in a rotating motion to create almostpulsation-free linear flow. Like rod pumps, therotor is generally turned by rods connected to asurface motor. New rodless installations use sub-surface electric motors and a speed-reducinggearbox to turn the rotor.

In most cases, progressing cavity pumps areflexible, reliable, resistant to abrasive solids andvolumetrically efficient. Use of small motorsresults in efficient power usage and low liftingcosts. Compared to rod pumps, progressingcavity pumps last longer and have fewer rod ortubing failures because of slower operatingspeeds. Capital costs are typically less than otherartificial-lift methods. Progressing cavity pumpsproduce up to 1700 B/D [270 m3/d] and are usedto depths of about 4000 ft [1220 m]. Elastomercomponents limit operating temperatures tobetween 212 and 302°F [100 and 150°C] and maynot be compatible with some chemicals or hydro-gen sulfide.

Hydraulic systems transfer energy downholeby pressurizing a special power fluid, usuallylight refined or produced oil, that flows throughwell tubing to a subsurface pump, which trans-mits this potential energy to produced fluids (next page, bottom left). Common pumps consistof jets, also known as venturi and orifice nozzles,reciprocating pistons, or less widely used rotat-ing turbines. A free-floating feature allowspumps to be circulated in and out of wellshydraulically, eliminating slickline or rig opera-tions to replace pumps or pull tubing. Hydraulicpumps are used at depths from 1000 to 18,000 ft

[305 to 5486 m] and produce rates from 100 to10,000 B/D [16 to 1590 m3/d] or more. Manyhydraulic installations produce 150 to 300 B/D[24 to 48 m3/d] from deeper than 12,000 ft[3658 m]. Heavy, viscous crudes are often easierto produce after mixing with lighter power fluids.Because pumps can be circulated out, systemscan be modified for changing conditions.

Gas lift uses additional high-pressure gas tosupplement formation gas. Produced fluids arelifted by reducing fluid density in wellbores tolighten the hydrostatic column, or backpressure,load on formations. Primary criteria for thismethod are gas availability and compressioncosts. Most gas-lift wells produce by continuousinjection, which is the only lift method that fullyutilizes formation gas energy (next page, topright). External gas, injected into special gas-liftvalves at specific design depths, mixes with pro-duced fluids and decreases the pressure gradientfrom the point of injection to surface. Bottomholepressure is reduced to provide a differential, orpressure drawdown, for required flow rates. Ifdrawdown is insufficient, instantaneous high-volume injection, or intermittent gas lift, can beused to displace slugs of liquid to surface. Theon-off nature of this option causes surface gas-handling problems as well as surges downholethat may result in sand production.

Gas lift is flexible and adjustable. Slickline-retrievable gas-lift valves can be pulled andreplaced without disturbing tubing if designs orsystem performance need to be changed. Costsvary depending on gas source and pressure, butcan be high if additional surface compressorsand processing facilities are needed. Gas-liftinstallations handle abrasive materials like sandand can be used in low-productivity, high gas/oilratio (GOR) wells or deviated wellbores. Naturalgas shortages limit or prevent gas-lift use.Freezing and gas hydrates are problematic, as isslickline valve retrieval in high-angle wells.Scale, corrosion and paraffin increase systemfriction or backpressure and reduce lift efficiency.Tubing size and long flowlines also limit systempressure and restrict efficiency. The main disad-vantage of gas lift is difficulty depleting low-pressure, low-productivity wells completely. Insome gas-lift wells, a change in lift method maybe required before abandonment.

Electric submersible systems use multiplecentrifugal pump stages mounted in series withina housing, mated closely to a submersible elec-tric motor on the end of tubing and connected tosurface controls and electric power by an armor-protected cable (next page, bottom right).

50 Oilfield Review

Insert pump

Rods

Tubingpump

Tubing

Casing

Prime mover Beam pumping unit

Perforations

Producedfluids

Plunger

Fullbore barrel

Traveling valve

Standing valve

> Reciprocating displacement rod pumps.2. Bradley HB (ed): Petroleum Engineering Handbook, First

Printing. Richardson, Texas, USA: Society of PetroleumEngineers, 1987.

Spring 1999 51

Electric motor

Rods

Tubing

Casing

Stator

Rotor

> Progressing cavity displacement pumps.

Power-fluidstorage

High-pressurepump

Perforations

Tubing

Power fluid

Casing

Jet, piston or turbinedownhole pump

Producedfluids

> Hydraulic-lift pumping systems.

Perforations

Produced oil and gas,and injection gas

Producedoil and gas

Injection gas

Gas-liftedproduced

fluids

Flowingproducedfluids

Side-pocketmandrels

Gas-liftvalves

> Injection gas lift.

Electric drives and controllers protectsystems by shutting off power if normaloperating limits are not maintained. A variable-speed driveadjusts pump outputby varying motor speed.

Producedfluids

Gas separators segregate some free gas fromproduced fluids into the tubing-casing annulusby fluid reversal or rotary centrifuge beforegas enters the pump.

Submersible motors are two-pole,three-phase induction motors.

Pump intakes allow fluids to enter thepump and may be part of the gas separator.

Electric transformersconvert source voltageto required downhole

motor voltage.

Motor protectors connect pumps to motors;isolate motors from well fluids; serve as a

motor-oil reserve and equalize pressurebetween wellbore and motor; and allow

expansion or contraction of motor oil.

Pump housings contain multistage rotatingimpellers and stationary diffusers. The

number of centrifugal stages determinesrate, pressure and required power.

Power cables supply electricity to submersiblemotors through armor-protected, insulatedconductors. Cables are round except for a flatsection along pumps and motor protectorswhere space is limited.

Downhole monitoring tools incorporatepressure and temperature sensing instruments

that send signals through the power cable toa surface readout unit.

Perforations

Gas

> Electric submersible centrifugal pump systems.

Submersible systems have a wide perfor-mance range and are one of the more versatilelift methods. Standard surface electric drivespower outputs from 100 to 30,000 B/D [16 to4770 m3/d] and variable-speed drives add pump-rate flexibility. High GOR fluids can be handled,but large gas volumes can lock up and destroypumps. Corrosive fluids are handled by using spe-cial materials and coatings. Modified equipmentand procedures allow sand and abrasive particlesto be pumped without adverse effects. Operatingsubmersible pumps at temperatures above 350°F[177°C] requires special high-temperature motorsand cables.

Historically, electric submersible pumps wereused in high-water, low-oil producers that per-form like water wells. A submersible pump canoperate in high-angle and horizontal wells, butshould be placed in a straight or vertical section.Subsurface submersible equipment may be sev-eral hundred feet long, so bending reduces runlife by causing internal wear on motor and pumpbearings. Wells deeper than 12,000 ft can be pro-duced efficiently with electric submersible sys-tems and these pumps can be used in casing assmall as 4.5-in. outside diameter (OD). At 20 to70% efficiency, electric submersible pumps areperhaps the most efficient and economical liftmethod on a cost-per-barrel basis, but depth andhigh GOR restrict capacity and efficiency.

Another disadvantage is the need for expen-sive rig interventions to pull tubing for pumprepairs or replacement. In addition, individualinstallations have limited production ranges dic-tated by the number of pump stages. Alternativedeployment methods and variable-speed surfacedrives address these limitations.

Current ApplicationsBecause hydrocarbon developments worldwideare in various stages of maturity, producing wellscan be grouped into categories (below). At oneend of this spectrum, which includes subseacompletions and wells requiring advanced con-struction methods or new equipment technolo-gies, there is a limited but growing number ofcomplex, high-cost wells that produce at highrates. Sizable installation and operating costs,combined with technology or equipment con-straints, limit use of artificial lift in these wells.

In general, this sector is not very active, but isundoubtedly the direction of future hydrocarbondevelopment. Offshore, because of reliability andflexibility, robust gas-lift and electric submersiblesystems are now used almost exclusively whenartificial lift is required. Exploitation of deep-water reserves requires improved technology.Alternative deployment methods and combinedlift systems for subsea wells in conjunction withpermanent downhole monitoring allow efficient,economic artificial lift and process control.

At the other extreme, stripper and develop-ment, or harvest, wells produce limited rates andvolumes. Incremental production due to artificiallift is small. Rod, progressing cavity or hydraulic

pumps are often applied in these wells. Althoughwell numbers are high, activity in this sector islimited to low-cost new installations and systemsalvage or replacement.

Between these categories are many medium-volume wells, often in secondary recovery fields,that produce significant rates and oil volumes.Incremental gains in these wells representimportant potential production. These wells drivea majority of engineering and technologicaldevelopments, generate cash flow, and representthe most active and high-value artificial-lift sec-tor. Medium- to high-volume lift methods, likegas lift or electric submersible pumps, areapplied in these wells. Ease of installation andoperational simplicity make these two systemspreferred and popular among operators.

Selection of artificial-lift methods and systemdesigns are best accomplished by studying fieldsas a whole, including reservoirs, wells, surfacefacility infrastructure and overall project eco-nomics. Service companies play an importantrole by providing installation, operation, trouble-shooting and optimization services in addition toartificial-lift technology, equipment and designsfor specific applications.

52 Oilfield Review

Number of wells

Prod

uctio

n ra

te

Subsea

High-rate wells

Stripper wells

Harvest fields and infill wells

Complex completions

Remote environments

Secondary recovery fields

> Artificial-lift applications. Across the spectrumof producing wells, artificial lift is applicable from simple, low-cost stripper wells where low-volume sucker-rod, progressing cavity andhydraulic lift are used most often to high-costsubsea developments. In between, there arelarge numbers of development, infill and sec-ondary recovery wells that produce significantvolumes of gas and oil, primarily by gas lift and electric submersible pumps. Increasingly,artificial-lift methods are being combined toovercome single-system limitations in thesecomplex, high-volume wells.

>

Spring 1999 53

System Evaluation and Selection Various approaches are used to develop oil andgas assets, add value or simply to reduce thecosts associated with potential prospects, newfields and late-life strategies for existing devel-opments. Choosing the best methods involveshydraulic, mechanical and electrical engineeringconsiderations. Ideally, artificial-lift evaluationsincorporate production system parameters fromreservoir boundaries to process plants.

Equipment requirements, the size and com-plexity of production systems and the powerrequired to lift well fluids make high-volume arti-ficial lift expensive to install and operate.Selecting the most suitable methods and equip-ment is important, because one artificial-liftinstallation may produce more oil than the pro-duction of some small mature fields. Selectingthe right system or combination of methods iseven more critical when evaluated in terms offailure, downtime and intervention costs.

Engineering teams review technical, eco-nomic and risk factors, generate options andmake recommendations. The best approach is aniterative total systems evaluation, whetherapplied a short time after discovery when morereservoir information is known, following initialdevelopment at a stage before further drilling orwhen reviewing late-life strategies (right).Artificial-lift strategies should maximize optionsthat are available over the life of a field.

Initial evaluation might indicate an artificial-lift method like electric submersible pump toobtain higher production rates, but later analysismay reveal that gas lift is best. Conversely, gaslift might be considered suitable initially becauseof poor submersible pump economics and equip-ment performance, but a review might show sub-mersible systems to be the right approach aslong as proper design, installation and operationare carried out. In some cases, electric sub-mersible pumps are installed and operated, butwhen sand, scale or emulsion problems developand actual production is reevaluated, gas lift orprogressing cavity pumps might be better.

For example, a field in North Africa withdeclining pressure and increasing water ratesappeared to be a candidate for electric sub-mersible pumps. The reservoir has a strongwaterdrive, and pressure declines about 100 psi[690 kPa] per year. No water injection is planned.The wells flow to field-gathering manifolds thatconnect with a pipeline linked to a distant pro-cessing plant. Increased water rates led to ces-sation of natural flow in some wells, indicatingthat artificial lift or pressure support was needed.This field appeared to be a candidate for electricsubmersible pump installations.

Three members of the Camco EOR Engineer-ing Optimization Resources group conducted anartificial-lift evaluation. Flowing gradient surveyshelped select the best vertical-flow correlation.Field flowline network and export pipeline pres-sures and rates were recorded to select a hori-zontal flow correlation. Water rates at whichnatural flow ceased were predicted and matchedby NODAL techniques and well performance

models.3 Reservoir pressure and water produc-tion forecasts were used to project when thefield would require artificial lift to produce highwater-cut wells.

Electric submersible pump evaluation deter-mined rates that could be achieved given reser-voir and well limitations. Pump designs weregenerated and production benefits were quanti-fied. Also estimated were the expected pump runlife and power requirements for developing thefield with submersible technology. Gas lift wasevaluated for a range of well conditions over thelife of the field. Injection pressure, gas-lift rateand tubing size were calculated to maximize production under existing processing facility con-straints. Compressor requirements were deter-mined from solution gas and lift-gas usage.Pipeline pressure and capacity with lift gasadded to the production stream were analyzed.Gas-lift designs were generated and 20 to 40%production increases were estimated.

Team evaluationof artificial-lift

methods

Commercialanalysis

Riskanalysis

Generateoptions

Makerecommendations

Installartificial lift

Evaluate resultsand review options

Technical inputfrom equipmentproviders and

service companies

Installationinfrastructure

Environmentalconsiderations

Reservoirconsiderations

Drillingconsiderations

Interventionconsiderations

Top-side processconsiderations

Wellboreconsiderations

System dataconsiderations

Technologypresent and future

Safetyconsiderations

> Artificial-lift evaluation. Because there are many strategies for developing oilfields, artificial-lift alternatives need to be identified and evaluated based ontechnical, commercial, risk and overall system factors. Engineering teams recom-mend development strategies and artificial-lift methods from the options generatedby these evaluations. When additional reservoir, well and facility information orperformance data are available, perhaps after initial field development or laterduring mature stages of production, these techniques are used to cycle throughthe process loop again to assess performance, investigate late-life strategies orreevaluate and change artificial-lift methods.

3. Bartz S, Mach JM, Saeedi J, Haskell J, Manrique J,Mukherjee H, Olsen T, Opsal S, Proano E, SemmelbeckM, Spalding G and Spath J: “Let’s Get the Most Out ofExisting Wells,” Oilfield Review 9, no. 4 (Winter 1997): 2-21.

Reservoir constraints like water and gas con-ing, sand production and gas breakout at perfo-rations were identified. Gas-lift and submersiblepump performance were compared, and down-time was estimated based on electric pump runlife and required gas-lift valve changes.Operational capability and safety issues wereidentified and costs were estimated. A compari-son clearly identified gas lift as the optimal arti-ficial-lift method. Intervention and pumpreplacements made electric submersible pumpsuneconomic even though production could beincreased by 30 to 40% initially. The EOR teamrecommended gas-lift implementation to avoidlost production. This evaluation was completedin one month.

Another example shows the complexity of liftselection. Petrobras-operated Ceara offshoreproduction area in Brazil consists of nine plat-forms producing four fields—Atum, Curima,Espada and Xareu. Production from these fieldswas 10,550 B/D [1680 m3/d]. As a result of lowreservoir pressures, all but 6 of 63 wells requireartificial lift. Because of poor pump performanceand scale-related failures, a proposal was madeto switch from electric submersible pumps to gaslift in all four fields.

The objective was to reduce expenses andincrease production by recompleting wells withsingle producing intervals to dual-zone produc-ers. In general, commingling zones was restrictedby wide pressure differentials. Gas lift was pro-posed as a solution that allowed dual zones to beproduced. However, large capital expenditureswere needed to convert from current electric sub-mersible systems.

Dual submersible pump systems also allowsimultaneous production of two isolated zonesand are an alternative to dual gas-lift comple-tions. Using dual submersible pumps in eachwell achieves the same production rates, andinvestment is limited to new completions foronly 26 proposed dual wells. Forecast oil-rateincreases of 725 m3/d [4560 B/D] to 1190 m3/d[7485 B/D] from dual zones can be achieved withgas lift or electric submersible pumps. Existingsubmersible installations and optional solutionswere reviewed, and gas-lift viability as a replace-ment for submersible pumps was assessed.

Based on a flow correlation verified by flow-ing pressure-gradient surveys in a trial gas-liftwell and NODAL analysis, field economics over-all are not affected by switching from electricsubmersible pump to gas lift in existing wells.Scale problems are not alleviated. Gas-lift valvesmust be placed at depths where scale builds up,so chemical injection is still required to ensurethat wells remain productive and serviceable.Gas lift does not draw reservoir pressure downas much as submersible pumps, which results in lower production rates. However, this loss of output is balanced by less downtime for gas-lift completions.

Several wells produce at low rates, particu-larly in the Xareu field, and flow would not be sta-ble under continuous gas lift. In the Cearaproduction area, gas lift is not the ideal artificial-lift method for every well, since some wellswould not continue to produce. Intermittent gaslift or progressing cavity pumps may be neededfor low-rate wells. A subsea completion with lowreservoir pressure that was fitted with gas-liftvalves would not flow naturally. A stand-alonecompressor was proposed to get this well on line.

Significant efficiency and oil output could begained by addressing submersible pump perfor-mance. A better chemical inhibition program wasneeded to reduce failures due to scale andimprove pump operations. Increasing run lifefrom 16 months to 24 months reduces the num-ber of workovers. Better designs could eliminateinefficiency, and installing two submersible pumpsystems per completion would cut workover fre-quency by half. When one pump fails, the othercan be used without pulling tubing and comple-tion equipment. Increasing submersible pump runlife and improving efficiency reduce expenses.After evaluation, submersible systems stillappear to be best, but reliability and life-cyclecosts need to improve.

The Camco EOR group recommended thatsubmersible pumps be retained as a primary arti-ficial-lift method and that alternatives for reduc-ing cost and increasing production be reviewed.One option was to use more than one artificial-lift method. By using redesigned submersiblepumps and better operating practices in theAtum, Xareu and Espada fields, and convertingthe Curima field to gas lift, production andexpense targets were achieved at reduced capitalexpenditure for facilities. This approach addressesthe field with highest lifting costs due to short sub-mersible pump run life and allows installation ofdual gas-lift completions in the Curima field,which has the best production potential.

Expertise is required to select, install andoperate high-rate artificial lift. Aside from techni-cal evaluation, system designs must be depend-able to realize optimal value in the face ofprobable commercial and risk scenarios. To mini-mize technical and financial risks, and addressspecialized applications, outsourcing and results-based contracts are becoming standard practiceamong operators for procuring equipment andimplementing artificial lift through systemsdesign and engineering services (see ”Artificial-Lift and Field Optimization,“ page 61).

54 Oilfield Review

Nitrogen-filledbellows

Casing ports

Tubing ports

> Subsurface gas-lift valves. The choice ofvalve and operating principle depends on wellparameters and well intervention costs. Con-ventional gas-lift valves and mandrels are runas part of the production tubing string. Retriev-able valves in side-pocket mandrels that areoffset from the centerline of the tubing areused offshore and in remote locations whererig interventions are expensive. Closing forcefor pressure-operated valves is provided by a spring, nitrogen-charged bellows, or both.Using a surface test rack, valves are preset toopen at the required operating pressure for awell. Smaller miniature values are available forlimited clearance and slimhole applications.

Spring 1999 55

High-Rate Gas-Lift SystemsAlthough trailing electric submersible pump useworldwide, gas lift—generally the most eco-nomical artificial-lift method if a cost-effectivegas supply is available—is common in NorthAmerica, the US Gulf Coast and offshore. Unlikesubmersible systems, gas lift does not addenergy, or lifting power. Reservoir pressure, sup-plemented by gas injected into tubing valves atspecific depths to lighten the fluid column, stilldrives fluid inflow and outflow. There are manytypes of gas-lift valves that use a variety ofoperating principles (previous page). Productionengineers choose the valves that fit well andfield conditions.

In gas-lift systems, downhole equipment andsurface facilities are closely related. Becausewell parameters and conditions like reservoirpressure are dynamic, producing operations

change over time. By using sophisticated soft-ware to link wellbore, surface facilities and pre-dicted reservoir response in a single model,integrated engineering teams can balance sur-face and subsurface considerations. Reservoirparameters are productivity, changes in perfor-mance with time and specific problems like sandor water influx. Well factors include tubing andcasing size as well as depth, completion configu-ration—packers, perforations and sand-controlscreens—type of gas-lift valve, wellborehydraulics and fluid-flow regimes. Surface facili-ties involve compressors, separators, manifolds,field flowlines and export pipelines (below).

Compressor discharge pressure impactsinjection valve setup and operation, and is thefirst gas-lift design consideration. Available pres-sure at the wellhead establishes gas injectiondepth, which determines lift efficiency. The

deeper gas is injected, the higher the productionrate. The cost of injecting deeper is related toadditional compression, plant upgrades andoperating expenses, as well as factors related toother surface facilities, like separator perfor-mance and pressures. There are, however, solu-tions that balance compression cost with theproduction rates that can be achieved.

It is important to ensure dependable gas pres-sure and volumes through mechanical reliabilityand operating procedures. Trained operators andproperly installed and maintained compressionequipment are crucial to gas injection. In somefields, gas lift is limited by existing infrastructure.Like gas-lift valves, compressors can also bechanged. Skid-mounted, portable compressionfacilities can be modified for use in other loca-tions or applications to improve outflow andminimize costs.

Oil storage

Productionmanifold

Compressorstation

Gas exportpipeline

Producedgas

Injection gasmanifold

Meteringand control

Producingwells

Producedfluids

Gas and oilseparator

Waterdisposal

well

Oil exportpipeline

Producedwater

Producedoil

Wellheadtubing and

casingpressure

Injectiongas

> Gas-lift networks and facilities. On the surface,gas-lift infrastructure includes compressors,separators, manifolds, field flowlines and exportpipelines, which are closely related to subsur-face equipment operation and performance.Changes in facility or reservoir performance influ-ence both systems. Often, there is not enough gasto lift every well at maximum efficiency. Produc-tion can be enhanced by optimizing gas injectionwithin existing field networks. If gas lift is limitedby existing surface infrastructure, skid-mounted,portable compression facilities can be used toimprove field output. [Adapted from Book 6 of theAmerican Petroleum Institute (API) Vocational TrainingSeries: Gas Lift. Dallas, Texas, USA: API, 1984.]

Surface gas compressors and subsurfacevalves need to operate in a stable manner, butchanges in facility or reservoir performance influ-ence both systems. Most of the time there is notenough gas to lift every well at maximum effi-ciency. Required injection rates often cannot beachieved because of gas source, equipment,pressure, economic or other limitations. Productionoutput can be enhanced by effective and efficientinjection gas distribution within existing fieldnetworks.

Many criteria are considered when choosingthe best injection rate. For example, wells withhigh productivity or low injectivity need more gas volume or higher gas-injection pressure.

Sensitivity analyses determine how wells affecteach other and define injection rates that resultin optimal production. Gas-lift valve port, or ori-fice, size can be calculated and adjusted forrequired gas injection. Subsurface gauges supplydata for subsequent evaluations. Surface andfinancial constraints often restrict gas throughputand need to be addressed using an integratedsystems approach.4 In such cases, field outputinstead of single-well rates are optimized. Forthis purpose, field-wide models are built basedon production system data such as compressors,separators, flowlines and chokes. Along withwell performance curves, data are gathered intoa field-wide NODAL analysis program.

Maximizing gas-lift performance one well at a time was standard in the past. Today, ongoingproduction optimization and management on asystem-wide basis, which includes compressors,increase revenue, enhance profitability and pro-vide long-term value more effectively. This sys-tems approach is made possible by improvementsand advances in computing, downhole monitor-ing, data-collection and information technologies.

Camco, now a Schlumberger company, manu-factures surface and subsurface flow-controldevices, side-pocket tubing mandrels and gas-liftvalves, latches, running and kickover tools forgas-lift systems (left). New technology, like elec-tric gas-lift valves, are also being developed.Conventional valves have one port size with thecapability to open and close. Simple orificevalves have no open-close mechanism. Electricvalves allow port size to be adjusted remotelyfrom surface over a range of fully open to closed.This provides better control when unloading flu-ids during well startup, real-time gas-lift opti-mization and the option of changing gas injectionpoints without well intervention. This flexibilitywill help meet future oil and gas exploitationchallenges by reducing gas-lift costs for deep-water and subsea wells.

In future optimization efforts, valves will berun with gauges to read casing and tubing pres-sure. Combined with information currently avail-able, such as well tests and surface pressuremeasurements, these readings will validate mod-els and forecasts, and be used to establish opti-mum gas-injection rates. Based on requiredrates, port size will then be adjusted remotely.The resulting effect on casing and tubing pres-sure is monitored and then used as feedback forthe next generation of closed-loop automatedcontrol systems.

High-Rate Electric Submersible SystemsWith liquid-lifting capacities up to 30,000 B/D[4770 m3/d], depending on electric power limita-tions, oilfield submersible pumps are used pri-marily for medium- and high-volume production.Within this artificial-lift sector are several typesof applications and configurations, includingstandard installations, booster or injection ser-vice, bottom intake or discharge, shroudedinstallations, offshore platforms and surfacehorizontal systems (next page). Design andinstallation of submersible systems combinehydraulic, mechanical and electrical componentsin a complex subsurface environment, so relia-bility is a key to success. If run life is short,retrieving an electric submersible pump thatfails prematurely is expensive and detrimental toproject economics.

56 Oilfield Review

Side-pocketmandrel

Gas-liftvalve

Slicklinekickover

tool

> Retrievable gas-lift valves. Slickline-retrievable valves can be installed or removed withoutpulling tubing. Kickover tools are designed to selectively locate side-pocket mandrels.

Spring 1999 57

Well stimulation or chemical injection areoften required, so it is important to ensure com-patibility between chemicals and downholeequipment. Treatment fluids can damage coat-ings and elastomer components like cable, motor,pump and motor-protector seals. Improveddesigns and advanced construction materials,including new metal alloys and elastomers forhanding corrosive fluids and harsh subsurfaceconditions such as extreme temperatures or high-ratio gas producers, are continuing to be devel-oped. These new technologies, coupled withalternative methods of deploying electric sub-mersible pumps, are expanding the range ofapplications for this versatile artificial-lift form.

High temperatures—For many years, electricsubmersible pumps were used in the Wilmingtonoil field, which consists of about 600 wells drilledfrom man-made islands in the harbor of LongBeach, California, USA, near San Diego. A subsetof these wells includes low-rate, high-oil-cut pro-ducers with 9 5⁄8-in. casing. The THUMS LongBeach Company operation had problems withmotors that failed prematurely in about 20 ofthese installations.5 Pumps were subjected totemperatures above 400˚F [205˚C] because oflimited oil rates and low water production thatdid not cool motors adequately. Pumps ran foronly 30 to 60 days.

An advanced design, HOTLINE motor serieswith capability to run continuously at up to 550˚F[228˚C] was developed by Reda, also aSchlumberger company. High-temperature ther-moplastic motor-winding insulation initiallydeveloped and patented for geothermal andsteamflood wells was applied. This successfulnew technology resulted in average runs inexcess of 1000 days and annual savings of over$200,000 per well, including fewer well interven-tions, less equipment repair and reduced

> Submersible pump configurations. In booster service, a standard pump, protector and motor unit are used to lift fluid from flowlines or other sources and simultaneously provide injection for waterflood, pipeline or other applications. In bottom-intake configurations, fluid enters the pump through a stingerin a permanent packer. Pump and motor are inverted from conventional installations. This setup is used when casing clearance limits production becauseof tubing friction loss or pump diameter interference. A bottom-intake configuration pump and motor can also be suspended by small diameter, high-strength cables, conventional tubing or coiled tubing to improve output. The bottom-discharge pump is used to inject water from shallow aquifers intodeeper producing zones. A shrouded configuration directs fluid past the motor for cooling or allows free gas to separate from fluids ahead of the intake and allows pumps to be set below perforations or producing zones. To save space on platforms and in other surface operations, submersiblepump surface units are used for mixing mud, washing down and fire protection, sump and water supply pumps and off-loading oil from storage.

4. Lekic O and Watt GW: “‘System Approach’ OptimizesGas Lift,” The American Oil & Gas Reporter 41, no. 6(June 1998): 124-128.

5. The THUMS (Texaco, Humble—now Exxon, Unocal,Mobil and Shell) Long Beach Company name wasderived from the first letter of the five companies thatjoined together to bid successfully for the rights todevelop and produce oil from under the city and harborof Long Beach, California.

Bottom Intake

Bottom Discharge Shrouded

Booster Service

Pump

Intake

Cable

Motor

Pump

Intake

Motor protector

Motor

Cable

Motor

Motor protector

Pump discharge

Pump

Packer

Motor

Motor protector

Intake

Pump

Packer

Cable

Cable

Motor protector

deferred-production cost. A life-cycle cost com-parison before and after introduction of this tech-nology shows a significantly expanded range ofsubmersible applications (above). These motorsare used in conventional submersible pumps forbetter reliability, even at low temperatures.6

High gas volumes—Like any artificial-liftmethod, submersible pumps reduce flowing bot-tomhole pressure to obtain better inflow. In gassywells, however, more vapor evolves from crudeoil at lower pressures. At higher vapor/liquid

ratios, pump performance begins to deteriorate.If a critical vapor-liquid limit is reached, pumpoperation becomes unstable, surging, cavitatingor stopping as gas blocks liquid flow insidepumps. Centrifugal force does not acceleratelow-density vapor. In fact, gas tends to lagbehind liquids and separate further, accumulat-ing in low-pressure, low-velocity areas of pumpimpellers and diffusers. Vapor restricts flowthrough these components, causing poor lift per-formance. Depending on fluid types, well charac-

teristics and hydraulic design of individualpumps, vapor can completely block flow into andthrough submersible pumps. Catastrophic fail-ures result if pumps are not protected from thisgas-lock condition (left).

The traditional solutions to gas problemswere shrouded, or tailpipe configurations, androtary separators to remove vapor ahead of pumpinlets. Production rate could also be limited, sothat inlet pressure is high enough to avoid detri-mental vapor/liquid ratios inside pumps. None ofthese solutions are optimal. Gas separatorsintroduce other limitations and mechanical com-plications while robbing the system of energy inthe form of gas, which lightens fluid density inthe tubing just as it does in gas-lift installations.However, keeping inlet pressure high limits pro-duction and may make artificial lift uneconomic.

Field experience shows that, depending onhydraulic design and fluid characteristics, cen-trifugal pumps tolerate vapor concentrations ofonly about 10 to 20% at moderate inlet pres-sures. New multiphase fluid-conditioning deviceslike the AGH Advanced Gas Handler componentprovide a way to produce at higher rates andlower flowing bottomhole pressures with greaterreliability and less wasted energy. The AGHmodule homogenizes liquid and gas entering the pump to reduce separation and accumulationin the first few pump stages, allowing sub-mersible systems to tolerate vapor concentra-tions greater than 50%.7

Field testing performed by Intevep S.A., theresearch branch of Petroléos de Venezuela S. A.(PDVSA), confirmed that the AGH component canallow stable pump operation with 48% vaporentering the pump. In Lake Maracaibo, Venezuela,where AGH technology was directly applicable inproduction operations previously restricted bygas-lift limitations or gas interference withsubmersible pumps, Intevep estimated thevalue of this capability to be at least 75,000 B/D[11,920 m3/d] of incremental oil productionfrom 250 wells.8 The AGH module can be usedalone or with a traditional rotary gas separator.

Alternative deployment—Techniques for run-ning electric submersible pumps in subsea com-pletions and on cable or coiled tubing expandartificial-lift applications and increase productionflexibility for offshore systems, remote locationswith limited rig availability and high-costworkover areas (next page).

A cable-deployed submersible-lift alternativelike the Reda CDPS Cable Deployed PumpingSystem technology reduces intervention costs byeliminating dependence on workover rigs. Thesystem is lowered into wells with a power cable

58 Oilfield Review

HOTLINE insulation

455°F[235°C]

Standard insulation

Subm

ersib

le sy

stem

life

500°F[260°C]

> High-temperature electric submersible pump performance. Through advanced materials, sub-mersible technology has developed to the point that operating temperatures greater than 400°F are possible, but many factors, including design and installation, must be addressed. Critical among these are cable type, equipment sizing, cable bands and elastomers. Each component ofthe HOTLINE production system is modified to operate up to 550°F [288°C], which expands sub-mersible lift use to steamfloods, geothermal applications and wells with poor cooling conditions.

Gasinterference

Gas lock

Pump impellers

AGHAdvanced

GasHandler

Rotarygas

separator

> Submersible pump performance in high gas/liquid ratio wells. When gas is present in producedfluids, lift efficiency deteriorates as gas takes up space inside pump stages and interferes with per-formance. Eventually, the system becomes gas-locked and stops producing or fails catastrophically.Traditionally, high gas volumes are handled by rotary gas separators that segregate gas into thetubing-casing annulus. Research and dynamic modeling prove that fluids can be homogenized byreducing pressure differentials within the pump. The AGH Advanced Gas Handler module conditionsfluid to behave as a single phase before it enters pumps.

Spring 1999 59

banded to a torque-balanced tension cable, andseated in a profile landing nipple of 5- or 7-in. tub-ing. A customized drawworks, which can be trans-ported by helicopter to offshore platforms andremote or environmentally sensitive areas withlimited access, is used to run and pull submersiblesystems.9 The CDPS lift system is inverted. Unliketypical submersible installations, running themotor on top and pump on bottom allows largerdiameter, higher volume pumps to be used sincethere is no need for a flat motor cable and guard torun beside the pump and motor protector. Blowoutpreventers are available to seal around the cables.

The economics that make coiled tubingattractive for other oilfield applications in high-cost workover areas also apply for artificial-liftsystem deployment. Electric submersible pumpson coiled tubing pump fluids conventionallythrough the coiled tubing or can be inverted toproduce fluids up the annulus. Power cables maybe installed inside coiled tubing or banded to theoutside. Internal cables are protected frommechanical damage, chemicals and well fluids.

In the Middle East, ARCO Qatar used coiledtubing with internal power cables to deploy sub-mersible pumps and produce fluids up the annu-lus inside 7-in. production tubing.10 In Brunei,Shell converted a well from gas lift to high-rateelectric submersible pump with a riglessworkover made possible by a coiled tubingdeployed system.11 Offshore, coiled tubingexpands submersible pump applications whenthrough-tubing installation is feasible, reducingthe need for conventional rig workovers and min-imizing downtime as well as deferred production.This unique, flexible technique has potential insmall or marginal offshore fields where no gas-lift infrastructure exists.

Subsea completions—Flow from subsea wellsis driven by reservoir pressure supplemented withgas injection when necessary. However, if wellsare far from host platforms, gas lift is inefficientbecause of long flowlines. Well-to-platformdistance is limited by the capacity of gas-liftinjection and reservoir pressure, which declinesas fields are depleted and water cut increases, to drive outflow. Distances greater than 8 miles[13 km] are considered uneconomic.12

Compared with gas lift, submersible pumpsare not as adversely affected by well-to-platformdistances and offer increased flow rates. Subseasubmersible installations were not feasible untilrecent advances in wet mateable connections.These connectors allow seafloor electrical tie-insand eliminate the need for dry connections to bemade at the surface. Offshore reservoirs that areuneconomic to operate by conventional means

Electricsubmersible

pump system

Electricsubmersiblepump system

Coiled tubingdeployed

Cable deployed

Packer

Cable bands

Armoredpowercable

Perforations

Torque-balancedwire rope

> Alternative deployment techniques. Cable-deployed and coiled tubing submersible pump systemsare designed to reduce expenses and production downtime associated with remote or high-costwells and offshore platforms where space and rig availability are limited.

6. Fuller D, Fickes B and Dowdy R: “Electric SubmersiblePumping Systems Applied in High-TemperatureEnvironments,” presented at the SPE ElectricalSubmergible Pump Workshop, Houston, Texas, USA,April 27-29, 1994.

7. Kallas, P and Way K: “An Electrical SubmergiblePumping System for High GOR Wells,” presented at theSPE Electrical Submergible Pump Workshop, Houston,Texas, USA, April 26-28, 1995.

8. Castro M, Pessoa R and Kallas P: “Successful Test ofNew ESP Technology for Lake Maracaibo Gassy OilWells,” presented at the SPE Electrical SubmergiblePump Workshop, Houston, Texas, USA, April 28-30, 1999.

9. Toubar M, Bahaa H and Guindi R: “Cable DeployedPumping System Case Study,” presented at the SPEElectrical Submergible Pump Workshop, Houston, Texas,USA, April 28–30, 1999.

10. Patterson JC, Stamey RC, Penny R and Dwiggins JL:“Coiled Tubing and Electrical Submersible PumpTechnology Improve Field Evaluation Cost,” presented at the SPE Electrical Submergible Pump Workshop,Houston, Texas, USA, April 30-May 2, 1997.

11. Pastor G, Knoppe R and Shepler R: “South China SeaGas Lifted Well Conversion Utilizing Coil Tubing ElectricalSubmersible Pumping Systems,” presented at the SPEElectrical Submergible Pump Workshop, Houston, Texas,USA, April 28-30, 1999.

12. Al-Mashgari A, Breit S, Christmas D, Leslie D and Smith J: “Subsea Electrical Submersible Pumps at Large Step-Out Distances,” paper SPE 38537, presentedat the 1997 SPE Offshore Europe Conference, Aberdeen,Scotland, September 9-12, 1997.

can now be produced with submersible lift.Operating satellite wells at greater distancesmeans that fewer platforms are needed; hostplatforms can be in shallow water; and marginalfields can be produced without platforms, whichreduces initial costs and operating expenses.

Electric submersible pumps offer many bene-fits over other artificial-lift methods in subseaapplications. The capability to operate wells far-ther from host platforms is the most important,but other benefits include improved well perfor-mance, reduced capital costs and lead times,improved energy efficiency and less environmen-tal impact. The first subsea submersible pumpwas installed in Brazil for Petrobras in 1994.13 Tomaximize recovery over a five-year production

contract, the BP Amoco plc Liuhua field in theSouth China Sea employs 24 of 29 subsea sub-mersible pumps in operation today, all installedby Reda (above).14

For electric submersible pump systems, Redamanufactures and supplies multistage centrifugalpumps, motors, protectors, gas-handling equip-ment, power cables, surface variable-speed drives(VSD) and controllers, and other accessories. In future submersible systems, vital operatingstatistics from fields or wells can be gathered bysurface instrumentation and reliable permanentdownhole gauges to be transmitted by SupervisoryControl and Data Acquisition (SCADA) systems tooffices where data are processed.

Pump inlet and outlet pressures, well andmotor temperatures, insulation resistance, sys-tem vibration and power supply can be inter-preted using software to make decisions, identifyor prevent pump problems and premature fail-ures, monitor performance and evaluate operat-ing options. Then, before actions are taken,NODAL analysis is used to simulate new systemconditions and validate motor frequency. Ifresults look good, executable commands aretransmitted to the well or field. Advanced vari-able-speed drives will be able to change operat-ing speeds automatically based on downholemeasurements and estimated torque to avoidelectric current fluctuations in motors.

Design, Installation and OperationArtificial-lift methods work well if systems aredesigned and installed properly. Changing reser-voir and well conditions need to be anticipatedso that proper equipment is selected andinstalled to ensure flexibility. Availability of datais important to achieve good designs that workeffectively in the field. In gas-lift design forexample, well data, completion diagrams, welldeviation, gas-lift equipment, surface productionsystem information, and reservoir and fluid char-acteristics are basic requirements. Good pres-sure-volume-temperature (PVT) data with flowingpressure and temperature surveys improvedesigns. The less uncertainty, the more economi-cal the design.

This principle also applies to other artificial-lift designs. In electric submersible designs, over-sized or undersized pumps and motors, whichcause inefficient energy consumption and shorterpump life, are often the result of limited or poordata. Variable speed drives can avoid these prob-lems, but may add to project capital expenditures.

Good data may have been available in thepast, but those designing artificial-lift systemsdid not always have access to this informationdue to inadequate communication within operat-ing companies or with pump manufacturing andservice companies. Reorganized and realignedbusiness units focus information and experiencelocally rather than company-wide. This trendrequires more openness between operators andservice provides to share nonconfidential infor-mation. Companies and operating areas need toshare knowledge and data efficiently to benefitfully from isolated pockets of industry expertiseand experience.

60 Oilfield Review

Power cable

9 5/8-in. casing

Packer

7-in. liner oropen hole

Bypass tubing

4 1/2-in. tubing

Y-tool bypass

Electric submersible pump

30-in. casing

Subsea wellhead

Subsurface safety valve

13 3/8-in. casing

Wet-mateable connector

> Subsea electric submersible installations. Under certain conditions, submer-sible pumping systems may offer advantages over other forms of artificial lift in subsea applications, including improved well performance, reduced capitalcost and lead time, improved energy efficiency, reduced environmental impactand more efficient operations at longer distances from host platforms.

Spring 1999 61

Once installed, artificial-lift systems must beoperated and managed. In gas-lift systems, sta-ble gas-injection pressure and rate are importantto prevent gas from being injected into multiplevalves or short circuiting above the operating-valve design depth. Effective monitoring providesearly indication of submersible pump problems,so preventive steps can be taken or future wellinterventions can be scheduled. If artificial liftfails, these data can be used in failure analysisand contribute to a process of continuous improve-ment. Teamwork among production, reservoir,completion and artificial-lift engineers, relateddisciplines, equipment providers and service sup-pliers is a key to production optimization.

Artificial-Lift and Field Optimization Maximizing field value is an important, but diffi-cult and often neglected task. Optimizing produc-tion well by well is one way to improve fieldoutput, but this approach is limited by constraintsfrom other wells and facilities. Another approachis to look at entire production systems—wells,reservoirs over time and surface networks. In thisway, constraints can be identified and elimi-nated. On an individual-well basis, optimizationis carried out using single-rate and multirate welltest results. When a group of wells is addressed,more involved methods from spreadsheets tofield models may be needed.

The value of production optimization may bedifficult to quantify and varies from case to case.Incremental production above baseline declinecurves through focused production managementand continuous optimization is the objective (right). The area under production curvesbetween optimized and baseline rates representscumulative incremental production and ulti-mately additional reserve recovery, particularlywhen ultimate abandonment pressure can bereduced. Added value can be significant, espe-cially in large fields. Experience shows that 3 to25% incremental production can be achievedwith production optimization. This percentagevaries, depending on the degree of optimizationthat has already been achieved and the quality orage of the original production system.

A modest 1% improvement in productionrates may deliver millions of dollars in addedvalue. Three to 25% increases equate to tens ofmillions of dollars per year in added revenue.Moreover, value is delivered not just fromincreased production, but also by better gas orpower usage, reduced operating costs and lowercapital expenditures. For example, after existingwells are optimized, fewer new or infill develop-ment wells may be required. Whatever the levelof production performance—from basic dataacquisition, system control and communicationto the actual optimization process—more isachieved with a systemized plan implementedand followed in a disciplined, structured approach.

When optimization is considered, often thefirst thought is in relation to gas-lift oil fields.Today, however, the approach and tools toachieve optimization allow all producing sys-tems—natural flow, gas lift, electric submersiblepump and gas wells—to be considered.Moreover, this process lends itself to performingshort studies to assess commercial and technicalimpacts of alternative development scenariosand provide important data for decision-makingand field management. Before optimizationbegins or strategic, economic and design choicesare made, it is necessary to evaluate productionsystems. This includes topside compressors,flowlines, manifolds and separators; wellboresubmersible pump or gas-lift design and opera-tion, fluid hydraulics and completion designs;reservoir productivity and changes with time,sand or water problems; and operating environ-ments from geographic location to type of instal-lation and export method.

Computer models aid in production systemoptimization. It is essential to have simulationsthat match reality by adjusting well and surfaceparameters—formation damage, tubing, flowlinecompressors, separators, manifolds, pipelinesand a flow correlation—in models. Often simula-tions that match measured, or known, cases areused as a predictive tool. Therefore, regularlyscheduled well tests are an important componentof modeling and optimization. As predictive tools,models are used to perform “what-if” scenariosand sensitivity analyses on different parametersto evaluate options. Continuous monitoring ofcompressor pressures, gas-injection rates orelectrical amp charts in submersible systems isneeded. These data are used to update modelsregularly and match actual well tests so that fieldconditions are represented accurately.

By studying oil and gas operations as com-plete systems, the most economic developmentstrategies are identified. Surface equipment andfacilities, well completion configurations, reser-voirs and operating environments are all takeninto consideration. Over the productive life of afield, optimization includes well modeling andmonitoring, liaison between field and officepersonnel, reconciling model predictions withmeasured data, updating recommendations peri-odically, training, data management and regularreporting of actual performance against targets.How far this process is taken depends on exist-ing conditions and limitations. In some cases,drilling new wells might cost less than optimiza-tion work. Therefore, a comprehensive study isneeded before making decisions.

Economic limit

Well life

Naturalflow

Gaslift

Electricsubmersible

pump

Prod

uctio

n ra

te

> Unlocking value: area under the curve. Incremental production value may be difficult to definebecause it varies from case to case, but output above a baseline decline is the target of artificial-liftoptimization. The difference between initial output and enhanced production is cumulative additionalproduction, or reserve recovery. Artificial lift increases the area under decline curves by improvingproduction rate, extending well life and reducing ultimate abandonment pressure. Changing from oneartificial-lift method to another may be necessary to further reduce flowing bottomhole pressure andmaximize reserve recovery.

13. Mendonca JE, Hodge RC, Izetti R, Nicholson A, DwigginsJL, Morrison D, Cia M and Alfano PP: “First Installationof an Electrical Submersible Pump in a Subsea Well,”presented at the SPE Electrical Submergible PumpWorkshop, Houston, Texas, USA, April 26-28, 1995.

14. Baillie AR and Chen Jing Hue: “Liuhua 11-1 FieldDevelopment: An Innovative Application of Technology,”presented at the SPE Electrical Submergible PumpWorkshop, Houston, Texas, USA, April 29-May 1, 1992.

Forties field is a decade-long example ofongoing artificial-lift optimization in a harsh off-shore environment where both gas lift and elec-tric submersible pumps are utilized. This NorthSea development consists of four main platformsproduced predominantly by gas lift and a smallerplatform lifted exclusively by electric sub-mersible pumps. Within the main field platforms,submersible systems have been used strategi-cally for tasks ranging from starting up platformsto proving new technologies. Submersible pumpoperations began in the late 1980s and gas liftwas initiated in the early 1990s. Incrementalgains from gas-lift optimization continue toincrease, and electric submersible pump reliabil-ity as well as run life have increased steadilywith improvements in operating techniques. Avalue-pricing arrangement led to $50 million inproject savings over five years.

Initially, a gas-lift group focused on supportingmore than 40 gas-lifted wells through studies,designs, monitoring, performance analysis, train-ing and trouble-shooting problems. Over time, astructured management process evolved thatincluded gas lift, reservoir surveillance and pro-duction engineers and encompassed all aspects ofproviding gas lift to the fields. Another team con-centrated on a systems approach to electric sub-mersible pump installation and operation with thegoal of improving run life and establishing anagreement between operating and service compa-nies that shared the financial risk of pump failuresas well as the benefits of prolonged production.

Team members are involved directly in analyz-ing and selecting artificial-lift methods bestsuited to meet short- and long-term field devel-opment goals. This approach brings new tech-nologies forward to address a variety of issuesfrom reservoir constraints to cost reduction.Combining gas-lift and electric submersible pump

expertise helps to better define and manage pro-duction. Today, an integrated team is modelingthe field, and networking all the wells and asso-ciated infrastructure. This will allow strategic andeconomic decisions to be made that take intoaccount a variety of constraints from platformelectricity generation, gas compression, flow-lines, separators, gas availability and water han-dling to subsurface pump performance, motorpower, pump stages, pressure drawdown limitsand well geometry.15 This asset is in decline, butsubstantial recoverable oil remains.

The next technological step is for optimiza-tion to be performed in real time with automatedclosed-loop systems. Automation can be appliedat different levels, from semi-automatic—stillinvolving field personnel to gather data or adjustvalves and engineers to make decisions—to fullyautomatic computerized systems. Automationcan be done using simple proportional-integral-differential (PID) or complicated fuzzy-logic con-trol systems.

Combining Systems Downhole There is a trend toward artificial-lift method com-binations to yield higher rates at lower cost,under better operating conditions and with moreproduction flexibility than could be expected fromjust one method. These approaches overcomerestrictions and limitations of individual methodssuch as tubing sizes, operating depth, high waterrates and corrosive conditions. Combined lift sys-tems are also more adaptable to changing oper-ational conditions, resulting from reservoirpressure depletion, gas injection for pressuremaintenance and secondary recovery water-floods. Combined lift methods reduce equipmentrequirements and power consumption, and yieldbeneficial results in terms of costs, investmentsand asset value.

For example, combining gas lift and electricsubmersible pumps in the same well offers manyopportunities to enhance production, optimizeflow rates and ensure uninterrupted operation.Gas lift in a combined installation provides abackup in case electric submersible pumps failand can be used prior to pump startup to unloadwells or stabilize wells that produce excessivegas or sand (below). Applications that operateboth systems concurrently include using a sub-mersible system to extend the life of an existinggas-lift installation. The submersible pump actsas a bottomhole pressure booster to increase theflowing pressure at the gas-injection depth. Fromthe standpoint of design, electric submersiblepumps develop greater flowing bottomhole pres-sure differentials than gas lift for a given rate.Simultaneous gas lift and submersible pumpoperation allows smaller pumps and motors to beused. Cost-savings can be utilized to install sys-tems with advanced materials and designs thathandle harsh conditions and extend pump andmotor life.16

62 Oilfield Review

Gas-lift valves

Packer

Injection gas

Electricsubmersible

pump system

Producedliquids

Perforations

> Combining gas lift and electric submersiblepumps. Combined artificial-lift systems performbetter in terms of improved production rates and reduced initial investments or operatingexpenses than can be expected from using onlyone method. Combining submersible pumps withgas lift allows smaller pumps or fewer stages tobe used and wells can continue to produce evenif submersible equipment fails.

15. Lekic O: “Enhancing Production,” Hart’s Oil and Gas World 90, no. 3 (March, 1998): 38-41.

16. Divine DL, Eads PT, Lea JF and Winkler HW: “Com-bination Gas Lift/Electrical Submersible Pump SystemIncreases Flexibility,” World Oil 211, no. 4 (October 1990): 77-82.Borja H and Castano R: “Production by CombinedArtificial Lift Systems and Its Application in TwoColombian Fields,” presented at the 1999 SPE LatinAmerican and Caribbean Petroleum EngineeringConference, Caracas, Venezuela, April 21-23, 1999; andSPE Electrical Submergible Pump Workshop, Houston,Texas, USA, April 28-April 30, 1999.Kahali KK, Deuri B and De SK: “Electrical SubmersiblePump-Gas Lift Combination—A Successful Trial for theFirst Time in ONGC, India,” presented at the SPEElectrical Submergible Pump Workshop, Houston, Texas, USA, April 28-April 30, 1999.

17. Chachula RC and Mann JS: “Selecting the AppropriateRodless Progressing Cavity (PC) Lift System for a HighlyDeviated Wellbore,” presented at the 1999 SPE LatinAmerican and Carribean Petroleum EngineeringConference, Caracas, Venezuela, April 21-23, 1999.Mann J, Ali I and Keller M: “Wireline RetrievableProgressing Cavity Electric Submergible PumpingSystem Updated Field Case Study,” presented at the SPEElectrical Submergible Pump Workshop, Houston, Texas,USA, April 29-May 1, 1998.Haworth CG: “Updated Field Case Studies onApplications & Performance of Bottom DriveProgressing Cavity Pumps,” presented at theSPE Electrical Submergible Pump Workshop,Houston, Texas, USA, April 30-May 2, 1997.

18. Carvalho PM, Podio AL and Sepehrnoori K: “Perform-ance and Design of an Electrical Submersible-Jet PumpSystem for Artificial Lift,” presented at the SPE ElectricalSubmergible Pump Workshop, Houston, Texas, USA,April 28-April 30, 1999.

Spring 1999 63

Progressing cavity pumps are popular for pro-ducing fluids with high-solids content, aromaticcondensates and tight emulsions as well asheavy crudes, especially in high-angle wells. Innonvertical wells, however, conventional sur-face-driven systems experience rod failure andwear-induced tubing leaks. Various rodless sys-tems are being used to solve these problems.One alternative is a bottomdrive configurationthat uses a power cable, submersible motor, pro-tector and flexible gearbox to drive progressingcavity pumps. This eliminates rod breaks, tubingwear and wellhead leaks, which reduces down-time and repair costs. The primary cause of pro-gressing cavity system failure is pump wear.Harsh subsurface conditions reduce pump perfor-mance and efficiency, but electric submersiblepump motors and drive components are usuallyunaffected and can be rerun.17

Deployment alternatives include conventionaltubing or coiled tubing. Using slickline or coiledtubing to replace pumps without pulling the driveassembly offers order-of-magnitude cost-savingsand makes combined systems attractive in high-cost areas if pumps fail frequently (left). Slickline-retrievable, bottomdrive progressing cavitypumps were evaluated originally for the AlaskanNorth Slope of the USA, where conventionalworkovers cost $200,000, but slickline operationscost $20,000. Wireline-retrievable progressingcavity lift systems were recently used in highlydeviated wells in Southeast Asia with sand, scaleor heavy-oil problems and small-diameter tubu-lars. When a failure occurred, the operator wasable to retrieve and replace the pump.

In subsea applications, artificial lift mustoperate effectively with multiphase gas-liquidmixtures since it not practical to use the tubing-casing annulus for downhole separation or ventproduced gas from the casing to an extra flowlinefor each well. Prototype testing proved thathydraulic jet pumps can be operated in combina-tion with electric submersible pumps to allowproduction of high gas/liquid ratio wells in deepwater. A rotary gas separator (RGS) to reduce the

volume of free gas that enters the pump intake,increases pump performance. Placing a jet pumpin the tubing above the electric submersiblepump discharge allows gas segregated into theannulus by rotary separation to be compressedand injected back into the liquid flow stream thatis pumped to surface by the submersible pump(above). The power fluid is the liquid—producedfluids less free gas—pumped by the electric sub-mersible pump. The jet-pump intake fluid is thefree gas that was separated upstream of theelectric submersible pump intake.18

Gas lift and electric submersible pumps havebeen used for decades, but new developmentsare still being introduced. Separating oil and gasdownhole, subsurface dewatering and disposal,and horizontal electric submersible pump sys-tems for surface oil and gas operations are justsome of the future applications for artificial-lifttechnology. Combining production processesdownhole to provide environmentally friendlysolutions that improve profitability blurs the dis-tinctions between various artificial-lift methods,and between subsurface equipment and surfacefacility functions. —MET

Bolt-on head

Progressing cavity pump

Casing

Intake

Cable

Gearbox and flex drive

Motor protector

Motor

Tubing

Slickline fishing head

Tubing-Conveyed

Slickline-Retrievable

> Subsurface-motor-driven progressing cavitypumps. Using progressing cavity pumps is away to lift heavy oil and high-solids content fluids. Rodless, bottomdrive progressing cavitypumps eliminate rod failures, tubing wear, rod torque and back spinning, and surfacewellhead leaks, which are major problems in conventional surface-driven systems.

Discharge

Producedliquids

Suction

Casing

Tubing

Sliding sleeve

Flowline

PlatformPower cable

Electricsubmersiblepump

Jet pump

Perforations

Subseawellhead

Gas

Gas

Producedfluids

> Combining submersible and jet-pump lift systems. Using a jet pump above the discharge of electricsubmersible pumps allows gas segregated into the tubing-casing annulus by a rotary gas separatorto be compressed and injected back into the liquid stream being boosted to surface by the submersiblepump. Prototype testing proved that this combination of artificial-lift methods can be used offshore,especially in deep water, where individual flowlines to vent annulus gas are complex and expensiveto install.

Ramez Akhnoukh, PatchFlex* Product Champion forSchlumberger Wireline & Testing in Rennes, France,provides support for the PatchFlex introduction out-side North America. He joined the company as a fieldengineer in 1987 with assignments in Italy, Algeria,Gabon, Congo and Libya. From 1993 to 1994, he wasan operations engineer in Egypt and became accountmanager there in 1994. He assumed his current postin 1999. Ramez holds a BS degree in electrical engi-neering from Ein Shams University in Cairo, Egypt.

Yann Bigno is a reservoir engineer for Total OilMarine, Aberdeen, Scotland, where he oversees reservoir management of the Dunbar field. Previously,from 1993 to 1996, he was with TOTAL ExplorationProduction in Paris, France, working on reservoirstudies for TOTAL subsidiaries. He has also workedfor Elf Aquitaine and Schlumberger in various loca-tions (1991 to 1993). Yann is a graduate of Ecole desMines, Nancy, France, and has an MS degree in petroleum engineering from Heriot Watt University,Edinburgh, Scotland.

Jean Marc Boisnault, Dowell WCS TechnologyImplementation Manager since 1998, is based in Montrouge, France. He joined the company in 1983 as a field engineer in Sarjah, United Arab Emirates.After assignments in Mexico, Guatemala and Brazil,he became station manager in El Tigre, Venezuela. In 1989, he became country manager in Gabon andCameroon, and four years later was country managerin Colombia. Jean Marc is a civil engineer with a master in management degree earned at Institutd’Administration des Entreprises, Paris, France.

Abderrahim Bourahla is with Sonatrach–AnadarkoGroupement Association in Hassi Messaoud, Algeria.Since 1997, he has served as Groupement DrillingManager reporting to the Groupement General Manager. His primary functions are to representGroupement in all drilling-related activities, to prepare the drilling department’s short- and long-range objectives and to ensure execution of thedrilling plan. He joined the Sonatrach HydrocarbonBranch, Drilling Division in 1983 as a drilling supervisor in the Hassi Messaoud field, Algeria, andfour years later, was promoted to drilling engineer for development and exploration wells in this field. In 1990, he became drilling superintendent and subsequently, development drilling manager. Hereceived a degree in drilling engineering from theAlgerian Petroleum Institute in Boumerdes.

Patrick Bouroumeau-Fuseau heads drilling and well servicing of the Alwyn platforms. He joinedTOTAL in 1977 and has held various positions indrilling and well servicing in TOTAL subsidiariesthroughout his career. He has a BS degree in petroleum engineering from the Institut Français du Pétrole in Rueil-Malmaison, France.

Jack Caldwell holds a BS degree in mathematicsfrom Davidson College, North Carolina, USA, and aPhD degree in geophysics from Cornell University,Ithaca, New York, USA. He began as a research geophysicist at the Texaco research lab in Bellaire,Houston, Texas, USA. In 1980, he joined Marathon Oil Company in their Littleton Research Center inColorado, USA. He joined Schlumberger in 1987, andwas seismic product development manager for USAonshore. In 1988, he became geophysical manager,responsible for Wireline & Testing borehole seismicbusiness in North America. Four years later, he transferred to Geco-Prakla in Houston as market

development manager for North and South America.After a stint in Calgary as Canada chief geophysicist,Jack returned to Houston, first as marine chief geophysicist for North and South America, and thenas reservoir characterization and monitoring manager for this area. In his current position asReservoir Solutions Manager, he has been responsiblefor developing multicomponent (4C) marine business in North and South America. A member of many committees of the Society of Exploration Geophysicists, he served as secretary-treasurer and as chairman of the research committee.

Gérard Catala heads the production services productline at Schlumberger in Clamart, France. Among hisprojects are the design and implementation of newproduction logging tools. Previously, he headed interpretation software products. He has been with the company for 25 years, and based at Clamartsince 1984. He has held field assignments in Italy,Yugoslavia, Tunisia, Nigeria, Australia and the FarEast. Gérard earned an engineering degree in industrial physics from Institut National des SciencesAppliquées, Toulouse, France.

Phil Christie took a BA degree in theoretical physicsfrom the University of Oxford, England in 1972 andthen went to Africa as a Schlumberger logging engineer. After three years in Angola, Nigeria, Gabonand Niger, he returned to the England to take a PhDdegree in seismology at the University of Cambridge.Following postdoctoral work in high-resolution seismic methods, he returned to Schlumberger in1981 as unit geophysicist for borehole seismic techniques within the European region. In 1985, he set up the Schlumberger borehole seismic engineering department in Clamart, France, wherethe present generation of Schlumberger openhole vertical seismic profile tools were designed. In 1987,Phil transferred to Schlumberger-Doll Research,Ridgefield, Connecticut, USA, to develop new applications in sonics, ultrasonics and borehole seismic measurements. Three years later, he established the seismic department at SchlumbergerCambridge Research, dedicated to new surface andborehole seismic applications. From 1996 to 1997, Philwas seconded to the BP Atlantic Margin Explorationgroup in Dyce, Scotland, where his projects includedthe joint reservoir monitoring experiment in Foinaven sponsored by BP, Shell and Geco-Prakla. Heis currently working with Geco-Prakla in Gatwickwhere he manages the Corporate Geophysics group,supporting applications such as multicomponent andtime-lapse seismic techniques to improve reservoirdefinition and to aid lithology and fluid prediction.

Trevor Dahl is a production engineer for PanCanadian Resources Ltd, Calgary, Alberta,Canada. Responsible for production technical issues, he is a member of a multidisciplinary teammanaging an oil and gas area. He joined PanCanadianPetroleum in 1996, working first in the Wilson Creekand Westerose area and then the Halkirk andLeahurst areas (1997 to 1999). Trevor has a BS degree in electrical engineering from the University of Saskatchewan, Saskatoon, Canada. He also won an award in the Myron Zucker worldwide IEEE design competition.

Folke Engelmark joined Geco-Prakla, Stavanger, Norway in 1996 as geoscience manager in the Reservoir Characterization and Monitoring department and is currently domain leader of interpretation workflow and rock physics in theReservoir Geophysics group at Geco-Prakla inGatwick, England. This group focuses on time-lapse(4D) seismic techniques for reservoir monitoring andmulticomponent (4C) data for state-of-the-art characterization of lithologies, pore fluids, abnormalpore pressures, fracture trends and anisotropy. Hebegan as a research geoscientist for the CaledonianResearch Project in Sweden (1974 to 1977) and spentthe next three years as a mining geophysicist and field geologist for LKAB Prospecting AB, Sweden. From1983 to 1990, he was project manager of the offshoreGulf of Mexico log analysis project for Geophysical Development Corporation in Houston, Texas. In 1990,he joined Pecten International Co. in Houston as anexploration geophysicist. He has also worked forPennzoil Exploration and Production Co., as E&Pgeophysicist and petrophysicist. Before joiningSchlumberger, he was principal geophysicist withSmedvig Technology in Stavanger. Folke has a degreein geology and geophysics from the University ofGothenburg, Sweden, and an MS degree in geophysicsfrom the Colorado School of Mines in Golden.

Roy Fleshman manages technology engineering marketing services at Reda Pump Company inBartlesville, Oklahoma, USA. He provides worldwide technical and marketing support, productdevelopment and introduction, intellectual propertyand technical training related to electric submersiblepump products and associated services to Reda fieldengineers and clients. He began with Reda Pump inMidland, Texas, as an application engineer in 1985.His next post was in Singapore (1987 to 1990) as atechnical sales and field engineer in the Asia Pacificarea. For the next two years, he was a technical application engineer in Bartlesville. From 1992 to1994, he was manager of application engineering.Prior to his current assignment he managed applications and special projects engineering. Royearned a BS degree in petroleum engineering at Oklahoma State University in Stillwater.

Dominique Guillot, Cementing Technology Specialistfor Dowell worldwide, is based in Clamart, France. Hismain responsibility is providing support to specialistswho work in Technology Application Centers inAberdeen, Scotland, Houston, Texas, and Singapore.He joined Dowell in 1981 in Saint-Etienne, France, as development engineer and section head (1981 to1984); and then as section head and product teammanager on projects related to well cementing (1984to 1989). In 1990, he became a cementing specialist in Houston, Texas, working on introduction of newtechnology. The following year he returned to Saint-Etienne as cementing engineering specialist, to workon cement mixing and cement job evaluation. From1994 to 1996, he was section head of Process and Software and field support in the Clamart ProductCenter. Before assuming his current position, he was acementing engineering specialist in Clamart, servingas the link between the Product Center and the engineering centers of North American-based clients.Dominique is a civil engineer, who trained at theEcole Nationale des Ponts et Chaussées in Paris andearned a Thèse de Docteur Ingénieur from the Centrede Géologie de l’Ingénieur de l’Ecole des Mines deParis et de l’Ecole Nationale des Ponts et Chaussées.

Contributors

64 Oilfield Review

An asterisk (*) is used to denote a mark of Schlumberger.

Harryson, who is an engineer with the Camco Engineering & Optimization Resources (EOR) groupin Houston, Texas, is responsible for well-performanceanalysis, and gas-lift design and analysis. He has beenwith the company since 1998. Harryson obtained a BS degree in mechanical engineering from BandungInstitute of Technology in Indonesia.

Jim Hemingway received a BS degree in chemistryfrom Emporia State University in Kansas, USA in 1978 and a BS degree in chemical engineering fromTexas A&M University, College Station, in 1979. Thefollowing year he joined Schlumberger as a field engineer and later became a sales engineer in Kansas.Since 1984, he has had several assignments as a loganalyst, data services manager and interpretationdevelopment engineer in Kansas, Texas and California.In 1998, he joined the Formation Evaluation department at the Schlumberger Sugar Land ProductCenter to work on the RSTPro* tool and three-phaseholdup interpretation techniques.

Xavier Hervé, the UK MAXPRO* marketing manager, has more than 20 years of experience with Schlumberger, originally in surface well testing beforespecializing in data acquisition services. Experiencedin testing operations worldwide, he was fully trained inproduction logging services. He began his career as afield engineer in the Middle East in 1979. Six yearslater he became a field analyst engineer in the FarEast. From 1987 to 1992, he was a field engineer inWest Africa. He spent the next three years as locationmanager for testing in Libya. Prior to his current position he was Multiline Production Services Manager in Aberdeen, Scotland. His degree is fromUniversité de Bretagne Occidentale, Brest, France.

Chris Holmes earned a degree in mechanical engineering from the University of Nottingham in England before joining the Shell group in 1989. Hebegan with Shell as a wellsite drilling engineer andgained experience offshore in Brunei prior to becoming a high-pressure, high-temperature (HPHT)drilling supervisor in Norway. He moved to Oman in1994 where he spent another two years in the fieldbefore moving to the office as a well engineer. For thepast two years, Chris has been an HPHT well engineer,involved in drilling high-pressure, intrasalt explorationwells in the south of Oman.

Jack Horkowitz is lead petrophysicist and productchampion for elemental capture spectroscopy loggingat the Schlumberger Sugar Land Product Center inTexas. He is currently developing open- and cased-hole applications for enhanced formation evaluationin carbonate and clastic reservoirs through the integration of geochemical spectroscopy measure-ments from ECS* Elemental Capture Spectrometerand RSTPro tools with nuclear magnetic resonanceand conventional wireline measurements. Before joining Schlumberger in 1995, he was a senior petrophysicist in the Enhanced Oil Recovery divisionof Pennzoil in Houston, Texas. He also spent five yearsas a petrophysicist with BP Exploration in Houston,involved in seismic calibration and lithology and fluidprediction for shelf and deep-water Gulf of Mexicoprospects. Jack holds MS and PhD degrees in geologyfrom the University of South Carolina, Columbia, USA.

Pål Kristiansen, Manager Seabed Acquisition Systems, is based at Geco-Prakla in Oslo, Norway. He isthe project manager for the Foinaven Active ReservoirMonitoring project. After joining Geco in 1983, heworked in seismic processing programming until 1987,

when he became involved in borehole seismic studies.Since rejoining Geco-Prakla in 1995, he has workedwith multicomponent seismic techniques and reser-voir monitoring. Pål was graduated from the Universityof Oslo in 1976 with a degree in physics.

Koji Kusaka, MAXPRO Product Champion for Schlumberger Wireline & Testing in Montrouge,France, leads the worldwide MAXPRO initiative, particularly in marketing, new technology develop-ment and training. After joining Schlumberger in 1983,he worked as a wireline logging engineer for sevenyears in Libya, Italy, Taiwan and Malaysia. He thenbecame a reservoir engineer in Indonesia andVenezuela. Before taking his current post, he was customer support engineer in Aberdeen, Scotland. Koji obtained a BS degree in geology from HiroshimaUniversity in Japan, and a diploma in reservoir management jointly awarded by Institut Français duPétrole in Rueil-Malmaison, France; Delft TechnicalUniversity, The Netherlands; and Imperial College,London, England.

James Leighton, Marketing and Operations Director,Drillflex, is based in Rennes, Brittany, France. Since1994, he has been responsible for marketing and fieldintroduction of the PatchFlex range of products basedon in-situ polymerization. He began his career in 1980with Halliburton in the Middle East. From 1987 to1994, he was division manager with Petrometalic inFrance. James earned a BA degree (hons) in politicalscience and an MBA degree from University of Nantesin France. He also holds three international patents onin-situ polymerization technology and has writtenmany papers on this subject.

Obren Lekic was graduated from the University ofLeeds, England, with a degree (hons) in mining engineering in 1988. The following year he joined BPExploration in Aberdeen, Scotland, as a petroleumengineer, working on production technology and wellengineering. In 1993, he moved to Camco Products &Services, based in Aberdeen, where he created a groupto provide engineering support along with productsales, predominantly gas lift. During this time, heworked with most operating companies in the NorthSea in their gas-lift fields as well as operations locatedaround the world. Since 1997, Obren has been based in Houston, Texas as the manager of the Camco Engineering & Optimization Resources (EOR) group,responsible for all aspects of business development.This team provides a range of engineering servicesincluding artificial-lift optimization, production management and alternative-lift studies.

Mark MacLeod, Earth Science Technical Manager atChevron UK Ltd in Aberdeen, Scotland, has beenresponsible for technical oversight of the earth scienceactivities in the Chevron Europe Business units inLondon, Oslo and Aberdeen since 1998. He initiatedand led a group effort on the Alba field ocean-bottomcable survey. He began his career in 1981 withChevron Geosciences in Houston, Texas, and thenspent two years in exploration with Chevron USA inSan Francisco and Concord, California, USA. From1983 to 1985, he conducted geophysics research atChevron Oilfield Research in La Habra, California.After eight years in various postings (technical staff,exploration and development), he began working onseismic integration in geostatistical reservoir models.In 1996, he transferred to Aberdeen to work on reservoir characterization and technical geophysicalstudies. Mark has a BS degree in earth science fromLaSalle College, Philadelphia, Pennsylvania, USA andan MS degree in exploration geophysics from StanfordUniversity in California.

Dan Markel, a senior engineer at the SchlumbergerPerforating & Testing Center in Rosharon, Texas, hasbeen in the Wireline Perforating MAXPRO group since1996. His main responsibilities are wireline perforatingrapid response projects, specializing in through-tubingguns and sustaining projects. He joined Schlumbergerin 1988 in the Through-Tubing Perforating group, thenjoined the Gun Systems group. He has three patentsrelating to perforating products. Dan received a BSdegree in mechanical engineering from University ofWest Virginia in Morgantown, USA and an MS degree,also in mechanical engineering, from the University ofHouston, Texas.

Pierre Maroy, Engineering Advisor and Quality Manager at Dowell Engineering in Clamart, France,also manages environmental product registration forthe North Sea and universities. After joining TOTAL in 1969, he was involved in land management for 15years, management of R&D for seven years, and ofmajor hazards and environment for one year. Pierre joined Dowell in 1991 as head of chemistry engineering in Saint Etienne, France, and subsequently moved to Clamart. He assumed his current post in 1998. Pierre is a graduate of EcolePolytechnique in Paris, France, with degrees inphysics and mathematics. He earned a State Doctoratedegree in chemistry from the University of Paris andhas an MS degree in economics from Dauphine University in Paris. Holder of 36 patents, includingseveral for CemCRETE* and SqueezeCRETE* systems, he is known for originating the “CRETE”slurry concept.

Andy Martin is a staff engineer in the marketinggroup at Schlumberger Perforating & Testing Center,Rosharon, Texas. His duties include oil company presentations and tours of the facility, training, marketresearch, as well as general marketing documentationpreparation. After joining Schlumberger Wireline &Testing as a field engineer in 1979, he had various fieldassignments, mainly in the Middle East. In 1990, hemoved to Montrouge, France as a staff engineer atWireline & Testing headquarters before moving to Livingston, Scotland as Production Services Tutor atthe British Training Centre. In 1993, Andy joined theOilfield Review editorial staff, writing on a variety oftopics including magnetic resonance, corrosion, paleo-magnetics, permeability and cased hole evaluation. Hehas an MA degree in engineering science from KebleCollege, University of Oxford, England.

Steve McHugo is domain leader for reservoir monitor-ing in the Reservoir Geophysics group in Gatwick, England. This group develops applications that flowfrom multicomponent and time-lapse seismic studiesto improve reservoir definition and aid lithology andfluid prediction. After obtaining a degree in appliedphysics from Middlesex Polytechnic in England, Stevejoined Geophysical Services Inc. in 1975 as processinggeophysicist. Between 1975 and 1986, he processed avariety of land and shallow-water 2D and 3D seismicdata from the Middle East and Europe. In 1986, hejoined the Geco-Prakla Land Special Projects andTesting group. In 1989, he became supervisor of thestratigraphic processing group and was involved inpreparing seismic data for stratigraphic analysis,developing strategies and tools for amplitude variationwith offset analysis and amplitude inversion. In 1993,this group became the inversion group, responsible for2D and 3D stratigraphic and structural inversion projects in Europe, Africa and the Middle East. In1996, he joined the marketing department as productchampion for reservoir geophysics with emphasis onmulticomponent processing and time-lapse studies.

Spring 1999 65

Charles Moffett, District Engineer with Hunt Petroleum Corp., is based in Jena, Louisiana, USA.Since joining Hunt in 1981, he has worked in drillingand production and is currently supervising produc-tion operations in the Goodpine District. Charles holdsa BS degree in petroleum engineering from LouisianaState University in Baton Rouge, and has 29 years ofexperience in the oil industry. He is a registered Petroleum Engineer in Texas.

Hüseyin Özdemir is a member of the Seismic Solutions group at Geco-Prakla, in Gatwick, England.Hüseyin gained his BS degree in geophysics and geology from the University of Istanbul, Turkey, andworked for the Mineral Research Institute of Turkeyas a field geophysicist. He then completed an MSdegree in applied geophysics at the University of Birmingham, England, in 1972, and a PhD degree inseismic data processing at Imperial College, London,England, in 1977. He subsequently was a researchassociate at the Technical University of Istanbul, andbecame an associate professor of applied geophysicsbefore joining Schlumberger in 1985. His assignmentswith Schlumberger included division geophysicist inKuwait and Abu Dhabi (1985 to 1990), and positions in the reservoir characterization group, Stavanger,Norway (1990 to 1993), and in the engineering andproduct development group in Gatwick, England.

Genaro Pérez Mejia, Engineering Submanager atPetróleos Mexicanos in Villahermosa, Mexico, helpsidentify, evaluate and supply new technology in thedrilling and workover management of wells in thesouthern Mexico region. In his 28 years with PetróleosMexicanos, he served as technical counselor duringthe blowout of several offshore wells. He has a degreein petroleum engineering from Instituto PolitecnicoNaçional Mexico, D. F.

Eddie Quin, who is with Total Oil Marine plc. inAberdeen, Scotland, is well superintendent for theAlwyn, Dunbar and Caister platforms. His primaryresponsibility is well interventions. Eddie has beenwith Total Oil Marine for nine years and spent the previous 12 years in the service sector with Halliburton, Baker and several small, independentwell-operation agencies.

M. Raiturkar received a BS degree in chemical engineering from the University of Bombay, India in1978. After working as process engineer for threeyears in Bombay, he joined the Production Technologydepartment of Petroleum Development Oman, Muscat(Shell OPCO). In 1986, he was transferred to the Production Chemistry department and since then hasbeen involved in completion fluids, drilling fluids,stimulation and cementing.

Ignacio Ramirez Martinez, a petroleum engineerwith Petróleos Mexicanos in Villahermosa, Mexico, is responsible for seeking and applying new technologies. He joined the company in 1984 and forthe next five years was a field engineer working onspecial operations in oil wells. Then he became budgetengineer for special projects in Comalcalco, Mexico.Before assuming his current post, he was operationssection chief in Comalcalco, providing hardware andsoftware support (1992 to 1997). He holds a degree inpetroleum engineering from Instituto PolitecnicoNaçional Mexico, D. F., and an MS degree in oil engineering from Universidad Naçional Autonoma de Mexico, D.F.

Philippe Revil, former Dowell CemCRETE ProductChampion for the Americas, is based in Houston,Texas. He was responsible for technical support andassistance in the introduction and use of CemCRETEtechnology in South and North America. He joined thecompany in 1992 as a field engineer with postings inSaudi Arabia, Indonesia and Ecuador. He also workedas a sales engineer in Venezuela and became DESC*

Design and Evaluation Services for Clients engineer inthe Shell Urdaneta Project in Maracaibo, Venezuela.In 1996, he became CemCRETE Product Champion forSouth America. He then served two years as worldwideCemCRETE Product Champion based in Clamart,France. Philippe has an Engineering Diploma inapplied physics and physics instrumentation from theEcole Nationale Supérieure de Physique de Grenoble(INPG), France, and studied communications systemsat McGill University in Montreal, Quebec, Canada.

Robert Roemer, CemCRETE Project Manager inAberdeen, Scotland since 1998, coordinates all phasesof engineering, marketing, field application and revenues related to CemCRETE products. He beganwith the Dowell division of Dow Chemical in SouthLouisiana in 1980, later serving as district engineer inPoint Noire, Congo. From 1985 to 1988, he was fieldservice manager for Dowell Schlumberger in Alice,Texas. He spent the next two years as Dowell RegionalLaboratory Manager in Houston, Texas, where he wasresponsible for expert technical support to the Southwestern USA. In 1990, he moved to St. Etienne, France, charged with the development, commercialization and support of new chemical technologies. Before his current position, Robert was Northern Area Business Manager, wellbore construction services in Calgary, Alberta, Canada,where he provided engineering, marketing and operations support for Dowell drilling and cementingfluids. Author of many publications, he received a BSdegree in geology from the University of Georgia atWest Georgia, Carrollton, USA.

Lisa Silipigno, PS PLATFORM* and RSTPro ProductChampion, is based in Clamart, France. She is responsible for product development and marketing of the physical and software enhancements to theseMAXPRO measurement services. Previously she wasSchlumberger customer support manager for the Belle Chasse district in Louisiana, responsible for job planning, technical support, and providing solutionsfor Schlumberger clients. She joined SchlumbergerWireline & Testing in 1994 as a junior field engineer,and attained general field engineer status in 1996. She became field service manager for Belle ChasseProduction Services in 1997. Lisa earned a BS degreein civil engineering systems from the University ofPennsylvania in Philadelphia.

Timothy Tirlia is currently drilling superintendentassigned by Anadarko Algeria Corporation toSonatrach–Anadarko Groupement Association inHassi Messaoud, Algeria. Here he represents Groupement in all activities required to drill, evaluateand complete wells. He joined Anadarko PetroleumCorporation in 1996 as a staff drilling engineer forAnadarko Algeria Corporation in Hassi Messaoud andwas promoted to his current position in 1998. Beforejoining Anadarko he was with Chevron (1989 to 1996),serving as a drilling supervisor throughout the UnitedStates. From 1994 to 1996, he was a floating drillingengineer at Chevron Drilling Technology Center inHouston, Texas. He has also worked for Tenneco OilCompany/Gas Pipeline (1981 to 1989). Timothy has adegree in petroleum engineering from Marietta College in Ohio, USA, and an MBA degree from theUniversity of St. Thomas in Houston, Texas.

Colin Whittaker is a production log interpretationexpert in charge of the cased-hole interpretationgroup in GeoQuest Aberdeen, Scotland, where he hasbeen based for the past two years. After earning adegree in electrical engineering from Imperial College, London, England, he worked for several yearsfor British Railways. He then spent more than adecade as a wireline field engineer in the Middle andFar East. Colin also was involved in production loggingtechniques for horizontal wells at Schlumberger Cambridge Research in England.

Coming in Oilfield Review

Model Validation. Reservoir simulation must honor data with different scales from many sources. Validated simulations helpimprove hydrocarbon field development by constraining multidisciplinary interpretations, and by integrating past and predicted reservoirperformance. Model-validation advances, such as using independent data to confirm interpretations, improve simulation reliability and extract more value from hard-won data.

Managing Drilling Risk. A new approachfocuses on avoiding stuck pipe, reducing drill-string failures, optimizing drilling efficiency, and monitoring and controlling wellbore stability to cut costs and avoid surprises. Technology is used to classify potential hazards beforehand, and to help detect and solve problems while drilling. Through teamwork, strategies and contingencies can be formulated to deal with these risks.

Real-Time Measurement Corrections. A new generation of wireline technology is revolutionizing fundamental formation evaluation. For efficiency and to quickly provide high-quality data at the wellsite, Platform Express* tools integrate a suite of logging measurements in a shorter, lighter package to help unravel formation mysteries. We review various applications and the underlying innovations that support them.

Production Management. Maximizing the economic potential of any asset requires a focused effort. By outsourcing oilfield production management, operators can realize the most from available infrastructure,resources and services. Armed with new technology, production analysts and operatingpersonnel are developing techniques thatincrease the longevity, productivity, profitabilityand reserve recovery of oil and gas fields.

66 Oilfield Review

MORAMultimedia Oilfield Review Archive

Point your browser to www.connect.slb.com to find:• technical presentations, forums and

publications, including Oilfield Review• detailed service information,

applications and case studies • log interpretation charts and

nonexclusive seismic maps• an oilfield dictionary, mnemonics

database and other oilfield links • location and telephone directory• DataLink* secure data transfer• InterACT* real-time data transmission

from wellsites• SuperVision* project monitoring.

Connect Schlumberger is the Internet communications resource for Schlumbergerclients. Updated daily, this password-protectedinterface is growing rapidly, providing timelyand interactive access to the information, dataand tools needed by the E&P industry.

You can also visit the publicSchlumberger Web site atwww.slb.com.* Mark of Schlumberger

Did you know that Oilfield Review can be found on-line?

MP-6066

Name

Company

Address

City State Postal Code

Country

Telephone

Fax E-mail

Method of Payment: ❏ MasterCard ❏ Visa ❏ Check

Account Number-Please include all digits

VAT number (EEC countries only)

Please print name as it appears on credit card

Expiration Date

Order Date Signature

Statement Address (if different from above)

MORA or Oilfield Review Services Order Form

Viewgraph and Slide sets of Oilfield Review Illustrations

Oilfield Review Back Issues and Articles

Multimedia Oilfield Review Archive

Platform: ❏ Macintosh ❏ Windows ❏ UNIX

Computer Model

Quantity Ordered @$25.00 each = Total $

Back-issue price:$15. Call for article reprint price.

Issue date Price Quantity Subtotal

1.

2.

3.

4.

Call for added cost of delivery. Total $Specify: ❏ Courier ❏ Air mail

ORDER TOTAL

Article title No. of sets Subtotal(call for prices)

1.

2.

3.

4.

Specify: ❏ Viewgraphs: $16 each ❏ Slides: $14 each Total $

Oilfield Review ServicesBarbour Square, High Street, Tattenhall,Chester CH3 9RF England☎ (44) 1829-770569 • Fax: (44) 1829-771354 E-mail: orservices@t-e-s.co.uk

Call or Copy and FAX this form.

$

MultimediaMORA—the Multimedia Oilfield Review Archive—providesquick and easy access to issues published from 1992 to 1998.MORA preserves the look of the printed magazine in a cross-platform format. Full-color articles can be printed or exploredon-screen using key-word searches and cross-document links.

Color 35-mm slides and viewgraphsThe ideal way to prepare high-quality sales presentations andlectures—Oilfield Review Services offers color 35-mm slidesand viewgraphs of illustrations from Oilfield Review and TheTechnical Review. The cost of a slide or viewgraph set rangesfrom $120 to $450. Please call for more information.

Article reprints and back issuesReprints of Oilfield Review articles can be supplied. Minimumorder is 500. Back issues of the Oilfield Review and TheTechnical Review can also be ordered, subject to availability.

Exhibition postersSpecial display materials such as posters or large transparencies,incorporating illustrations from Oilfield Review, can be ordered.

Oilfield Review Services also creates tailor-made multimediapresentations for exhibitions, conferences and seminars, and canprovide staff and equipment to run the presentations.

Browse Oilfield Review Services online at www.t-e-s.co.uk

Multimedia, graphics andreprints fromOilfieldReview

Oil from upper sands onlyOil from lower sands only

Commingled oils A B C D E F G H I J K L

% u

pper

san

d pr

oduc

tion,

ste

amflo

od

0

10

20

30

40

50

60

70

80

90

100

Well

Before steamfloodAfter steamflood

1.0

0.5

1.52.0

2.5

A

B

C

DE

F

G

H

I

J

K

LM

N

O

P

Lower sands(3 wells)

Upper sands(2 wells)Producing

well

Productionwater Steam

Injectionwells

Steamfloodedupper sands

Waterfloodedlower sands

0 ft 300

A

B

C

D

E

F

G

H

I

JK

L

Available from Corporate Express/Ross-Martin

Available from Oilfield Review Services:

Corporate Express Document and Print Management3635 Willow Bend, Suite 300Houston, Texas 77054☎ 1(713)660-7226 • Fax 1(713)660-0362E-mail: rmartin1@idt.net

Oilfield Review

Your electronic journey into the pages of

Oilfield Review begins here.

This archive provides quick and easy access

to Oilfield Review articles published from 1992

to 1998. Select a button to explore articles by

topic, issue or search by keyword.

MORA

About Topics Issues Help Search

NOW A

VAIL

ABLE: 1

992-

1998

MULT

IMED

IA O

ILFI

ELD R

EVIE

W A

RCHIVE

CD

SEE

INSI

DE BACK C

OVER

Spring 1999

Shear-Wave Seismic Data

Cementing Technology

Production Services

Artificial-Lift Methods

Oilfield Review