pg 01 gastips-masttoc - netl.doe.gov · long-term cement integrity of hpht cement systems deep...

Imaging Deep Gas Prospects with Multi-component Seismic DataGas producers across the Gulf of Mexico are targeting deeper and deeper drilling objectives. Most opera-tors in the Gulf consider 30,000ft to be the deepest target depth that will be drilled for the next few years. Toimage geology at those depths, seismic reflection data must be acquired with offsets of at least 30,000ft.

TECHNOLOGY DEVELOPMENTS IN NATURAL GAS EXPLORATION, PRODUCTION AND PROCESSING

A Publication of Gas Technology Institute, the U.S. Department of Energy and Hart Energy Publishing, LP

GasTIPS®

Items of Interest 02 Editors’ Comments29 Publications, Events Calendar, Briefs and Contacts

Summer 2005 • Volume 11 • Number 3

GTI-05/0232

M-C Seismic 3

Integrated Seismic/Rock Physics Approach to Characterizing Fractured ReservoirsThe key to successful development of many low permeability reservoirs lies in reliably detecting, characterizing and mapping natural fractures. Fractures play a crucial role in controlling fluid transport intight reservoirs, and they can influence production, sometimes adversely, even in reservoirs with moder-ate to high permeability.

Rock Physics 6

Dual-Density Drilling Systems ReduceDeepwater Drilling Costs: Part I—Concepts and Riser Gas Lift MethodA study of dual-gradient deepwater drilling systems relying on riser gas lift and riser dilution concludedthat both concepts are feasible and warrant additional research. Editors note: This is the first in a twopart series. The second part will publish in GasTIPS fall issue.

Dual-densityDrilling 11

Laboratory Testing of an Active Drilling Vibration Monitoring & Control SystemThe deep, hard-rock drilling environment induces severe vibrations, which can cause reduced rates ofpenetration and premature failure of the equipment. An active vibration control system can prevent suchproblems and increase the rate of penetration.

DrillingVibrations 15

Long-term Cement Integrity of HPHT Cement SystemsDeep Trek, a U.S. Department of Energy program, seeks to improve HPHT well economics by improvingdrilling and completion technology. As part of the Deep Trek effort, CSI Technologies is attempting toimprove the economics of deep-well completions with the development of a “supercement” capable ofproviding long-term, HPHT sealing integrity.

HPHT CementSystems 20

Capturing Missed Reserves in Existing WellsThe motivation behind capturing the missed reserves is the geologic complexity of the reservoirs, suchas sequences of low-permeability sands and shales, existence of natural fractures and variable reservoirarchitecture caused by abrupt changes in stratigraphic facies.

Behind-pipePay Zones 24

A High-flying Alternative to Walking the LineMore than 2,000 years ago, natural gas was sent across Tibet via bamboo pipes. How did they maintain the integrity of that system? By walking the line. For as long as there have been natural gaspipelines, inspectors have had no choice but to walk the line to detect leaks and ensure public safety.Now the endless walk is over.

PipelineInspection 27

© 2005 Halliburton. All rights reserved.

Acidizing can be a superior route to sustained

production increases – with the right technology.

Carbonate 20/20 service is Halliburton’s complete

system for candidate selection, job design and

execution. And it’s generating long-lasting

production gains in all kinds of carbonate

formations and all types of wells. The key is

the focus on the rock properties, because the

reservoir drives all the decisions.

And Carbonate 20/20 service includes a full

suite of fluid systems designed for both fracture

and matrix acidizing to match every rock and

well condition and monitoring technologies to

measure results. Acidizing is an effective and

economical alternative to proppant fracturing –

with the Carbonate 20/20 system’s disciplined,

comprehensive approach.

To learn more about Halliburton’s energy

to help with Carbonate 20/20 system, write

[email protected] or visit

www.halliburton.com/stimulation.

Unleash the energy.™

Headline news on acidizing.

HALLIBURTON

Production Optimization

Carbonate 20/20

Carbonate 20/20difference

Carbonate 20/20difference

ConventionalAcidizing

Tota

l Pro

duct

ion

Time

Summer 2005 • GasTIPS 1

Managing EditorMonique A. BarbeeHart Energy Publishing, LP

Graphic DesignMelissa RitchieHart Energy Publishing, LP

EditorsGary SamesDOE-NETL

Kent PerryGas Technology Institute

Subscriber ServicesAmy [email protected] Energy Publishing, LP

PublisherHart Energy Publishing, LP

C O N T E N T SEditors’ Comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

M-C Seismic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

Rock Physics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

Dual-density Drilling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11

Drilling Vibrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

HPHT Cement Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

Behind-pipe Pay Zones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

Pipeline Inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Briefs and Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29

DISCLAIMER:LEGAL NOTICE: This publication was pre-

pared as an account of work sponsored

by either Gas Technology Institute (GTI) or

the U.S. Department of Energy, (DOE).

Neither GTI nor DOE, nor any person act-

ing on behalf of either of these:

1. makes any warranty or representa-

tion, express or implied with respect to the

accuracy, completeness or usefulness of

the information contained in this report

nor that the use of any information, appa-

ratus, method, or process disclosed in this

report may not infringe privately owned

rights; or

2. assumes any liability with respect

to the use of, or for damages resulting

from the use of, any information, appara-

tus, method or process disclosed in this

report.

Reference to trade names or specific

commercial products, commodities, or

services in this report does not represent

or constitute an endorsement, recommen-

dation, or favoring by GTI or DOE of the

specific commercial product, commodity

or service.

GasTIPS®

GasTIPS® (ISSN 1078-3954), published four times a year by Hart Energy Publishing, LP,reports on research supported by Gas Technology Institute, the U.S. Department of Energy,and others in the area of natural gas exploration, formation evaluation, drilling and com-pletion, stimulation, production and gas processing.

Subscriptions to GasTIPS are free of charge to qualified individuals in the domestic naturalgas industry and others within the Western Hemisphere. Other international subscriptionsare available for $149. Domestic subscriptions for parties outside the natural gas industryare available for $99. Back issues are $25. Send address changes and requests for backissues to Subscriber Services at Hart Energy Publishing, 4545 Post Oak Place, Suite 210,Houston, TX 77027, Fax: (713) 840-0923. Comments and suggestions should be directed toMonique Barbee, GasTIPS managing editor, at the same address.

GTISM and GasTIPS® are trademarked by Gas Technology Institute, Inc.© 2005 Hart Energy Publishing, LP

Unless otherwise noted, information in this publication may be freely used or quoted, pro-vided that acknowledgement is given to GasTIPS and its sources.

Publication OfficeHart Energy Publishing, LP4545 Post Oak Place, Suite 210Houston, TX 77027(713) 993-9320 • FAX:(713) 840-0923

POSTMASTER: Please send address changes to GasTIPS, c/o Hart Energy Publishing, 4545 Post Oak Place, Suite 210, Houston, TX 77027.

2 GasTIPS • Summer 2005

The issue of oil and gas industryresearch and development (R&D) is afocus of attention on multiple fronts

this summer.A Society of Petroleum Engineers (SPE)

forum held in Colorado from June 20-24 thatbrought together technologists representingoperators, service companies, academia and thegovernment with the objective of stimulatinginteraction and leveraging each other’s effortsrelated to exploration and production R&D.This forum is one of a number of effortsundertaken by the SPE R&D AdvisoryCommittee, created in 2002 to foster the con-tinuing development of petroleum engineeringtechnology through collaborations of acade-mia, industry and government worldwide. InOctober, at SPE’s Annual TechnicalConference and Exhibition in Dallas, an R&Dpanel discussion will focus on the perception ofa decline in operating company upstreamR&D, including a reduction in direct researchand reduced participation in collaborativeefforts. The committee also has sponsored aseries of eight articles in the Journal ofPetroleum Technology ( JPT) that focus onR&D in each of SPE’s technical disciplines.

The backdrop to these activities is the picturepainted by the 2005 update of the EnergyInformation Administration report Perform-ance Profiles of Major Energy Producers. Thisreport shows that 2003 expenditures onupstream R&D reported by 28 major U.S.energy companies continue to follow the down-ward trend exhibited during the past decade(see chart). Service company R&D has filled ina portion of the gap between today’s level ofupstream R&D investment and that of adecade ago. On the positive side, combinedR&D spending by three of the largest explo-ration and production (E&P) service compa-nies (Schlumberger, Halliburton and BakerHughes), as reported in their annual reports,hasgrown by 25% during the past 5 years, but this

has not filled in the gap.While the reporting rulesfor these statistics prevent aprecise quantification ofexactly how much money isspent on R&D, vs. engi-neering related to devel-oped technologies, it is clearthat total R&D spendingfocused on oil and gas E&Premains well below histori-cal levels for the industry.

Some believe we mayalready be seeing theeffects of decreased invest-ment. In an article in theJune 2005 issue of the JPTtitled “Technology’s Valuein the Upstream Oil andGas Industry,” AliDaneshy of the University of Houston andTom Bates of Lime Rock Partners suggest agradual increase in finding and developmentcosts since 1995 reflects the technology statusof the industry for the prior decade. Their fearis that the trend may indicate dramaticallyincreasing costs for the next decade, reflectingan increased aversion to risk by the industry,slower development of technology or both.

Another area of possible concern for thedomestic industry in the United States is thatwhat R&D the service companies are doing ispointed toward offshore and other high poten-tial areas rather than toward unconventionalresources. This is left to the independents whoproduce most of the oil and gas and drill mostof the wells in the United State but do notindividually own enough of the resource tojustify large R&D investments.

Don Green, professor of chemical andpetroleum engineering at the University ofKansas and one of the chairs of the recent SPEforum, said people are paying increased atten-tion to this trend.

“There is concern about both the decreasinglevels of E&P R&D spending on the part ofindustry and the government, and on the lim-ited degree of interaction between industryand the academic community,” Green said.“The industry representatives at the forumsupport more of both.”

One topic the forum attendees discussed atlength was how to better inform universityfaculty on the fundamental E&P problemsstill in need of solutions.These and other ideasdeveloped from the forum will help form thebasis for the October panel.

This issue of GasTIPS includes a range ofresearch topics from geophysical characteriza-tion of fractured reservoirs to novel drillingand completion technologies, to high altitudedetection of gas pipeline leaks. The array ofissues reflects the range of challenges facingthe industry. We hope you find this issue ofGasTIPS informative. ✣

Commentary

Industry R&D Spending in the Spotlight

0

100

200

300

400

500

600

700

800

900

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Oil

an

d G

as

Re

co

ve

ry R

&D

(m

illi

on

$)

0

0.5

1

1.5

2

2.5

R&

D P

erc

en

tag

e o

f T

ota

l E

&D

Exp

en

dit

ure

s

Oil and Gas Recovery R&D Spending

R&D Percentage of Total E&DExpenditures

Oil and gas research and development investment by major exploration and production companies. (Energy InformationAdministration, 2005)


M-C SEISMIC

Long-offset seismicsurveys are difficult toachieve using towed-

cable seismic technology inareas congested with pro-duction facilities, typical formany shallow-water blocksacross the northern GulfShelf. Ocean-bottom-cable(OBC) and ocean-bottom-sensor (OBS) technologiesare logical options for long-offset data acquisition insuch congested productionareas because ocean-floorsensors are immobile oncedeployed and can be posi-tioned close to platforms,well heads or other obstruc-tions that interfere withtowed-cable operations. Anexample illustrating thedeployment of ocean-floorsensors across an area ofnumerous production plat-forms and facilities in onestudy area of offshoreLouisiana is illustrated inFigure 1. A 6-mile (10-km)diameter circle is positioned on this map toillustrate the difficulty of towing a 6-mile cableacross the area in any azimuth direction. Incontrast to the difficulty of executing towed-cable operations, OBC lines AA, BB and CC(actual profiles used in one long-offset OBCdata-acquisition program) pass within a few

meters of several production platforms.An additional appeal of OBC seismic

technology is that 4-C seismic data can beacquired, allowing targeted reservoir intervalsto be imaged with P-SV wavefields, as well aswith P-P wavefields. Once 4-C seafloorreceivers are deployed, source boats towing

only air gun arrays canmaneuver along a receiverline to generate P-P and P-SV data from long-offsetdistances. For the multi-component data used inthis study, some fieldrecords were acquired withoffsets greater than 6 miles.However, data offsets werelimited to this distance dur-ing data processing.

A P-P image across onestudy area of offshoreLouisiana is shown inFigure 2. This north-southprofile is 45 miles (72 km)long, and a scale bar is posi-tioned on the image to rep-resent the dimension of thelongest source-receiver off-set used in processing thedata. This maximum-offsetbar can be compared withthe physical sizes of the saltstructures and rotated faultblocks along the profile toidentify where the seismicpropagation velocity can be

expected to change over lateral distances sim-ilar to the maximum offset and possibly affectdeep imaging.

Two horizons were interpreted along theprofile and neither is a structural horizon.Each is only a marker that indicates seismicreflection quality. Horizon 1, the shallower

By Bob A. Hardage, Bureau of Economic

Geology Imaging Deep Gas Prospects withMulti-component Seismic DataGas producers across the Gulf of Mexico are targeting deeper and deeper drilling objectives. Mostoperators in the Gulf consider 30,000ft to be the deepest target depth that will be drilled for the nextfew years. To image geology at those depths, seismic reflection data must be acquired with offsets ofat least 30,000ft.

Figure 1. Four-C ocean-bottom cable data acquisition of long-offset seismicprofiles across a congested production area.


M-C SEISMIC

of the two, marks the base ofcontinuous reflections. Thathorizon crosses geologic timelines and does not map struc-ture or indicate depth varia-tions of a fixed formation.Horizon 2, the deeper of thetwo, defines the base of dis-continuous but mappablereflections. It also crossesgeologic time lines and doesnot follow a fixed geologicstructure.

The profile shows there is alarge sediment accumulationin the northern one-third ofthe image space whereHorizon 2 drops down toabout 10 s. This sediment loadsqueezes the Jurassic saltsouthward, causing several saltstructures to punch upwardthrough overlying, youngerstrata shown by the salt-flowarrows. This salt movementcreates numerous echelonrotated fault blocks.The depthof shallower Horizon 1 is con-trolled to a great extent by thevertical depth to the tops ofthe various salt structuresalong the profile. The defini-tion of the base of continuousreflections (Horizon 1) in thenorth portion of the profile is subjective.Horizon 1 could be positioned at the northend of the seismic line deeper than where it isshown in Figure 2. The exact vertical positionof Horizon 1 in P-P image space is not toocritical because the surface is used as a data-quality indicator, not as a geologic horizon.

The 10-s image times of the deepest P-Preflections on this profile are considerablydeeper than the maximum P-P seismic imagetimes observed with “conventional” seismicdata in the area. Interpreters who have exam-ined these long-offset data acknowledge that

the data image deeper geology than doshorter-offset seismic reflection data acquiredto date over the northern Gulf Shelf.

In the north portion of this profile, a saltweld is near Horizon 1, where underlying salthas evacuated and flowed south. The classicalP-P seismic attributes of a salt weld arelabeled on the data display:

• events above the weld are usually discor-dant with events it;

• the signal-to-noise ratio is often low inthe data window that encompasses theweld; and

• events below the weldtend to be lower frequencythan those above it.

The P-SV image along thissame profile is displayed asFigure 3. The vertical axislabeled “warped P-SV time”means P-SV image time hasbeen adjusted to P-P equiva-lent image time, at least to afirst-order level of accuracy. Ifthe P-SV image in Figure 3 iscompared with its companionP-P image (Figure 2), P-SVHorizon 1 is at about the sameimage-time coordinates as P-PHorizon 1 (about 5 s) acrossthe profile. Locally, P-PHorizon 1 and P-SV Horizon1 differ. The important point isthat in a broad perspective, thetwo horizons are essentiallydepth equivalent. This obser-vation is a key principle. Manyexplorationists do not yetknow how deep P-SV data canimage. This data comparisonprovides critical informationsuggesting that P-SV dataprovide continuous, mappablereflections to the same depthsas P-P data. A second point toemphasize is that local differ-ences between P-P Horizon 1

and P-SV Horizon 1 are important only ifthese horizons are structural surfaces.Because the horizons are indicators of reflec-tion quality (specifically indicating the baseof deepest continuous reflections) and notstructure surfaces, local differences betweenP-P Horizon 1 and P-SV Horizon 1 are nottoo critical.

A different situation exists for Horizon 2.It is difficult to find any mappable P-SVevents at image times significantly below P-SV Horizon 1. Only a few short segments ofdeep P-SV events are labeled in Figure 3. In

Figure 2. Long-offset P-P image across study area. Horizons 1 and 2 aredata-quality horizons defined in the text.

Figure 3. Long-offset P-SV image across study area. This image should becompared with the P-P image of Figure 2 acquired along the same profile.

contrast, the P-P data contain a large popu-lation of deep Horizon 2 events (Figure 2).The lack of P-SV events near the super-deepdepths of P-P Horizon 2 does not reduce thevalue of P-SV data for evaluating drillingtargets down to depths of 30,000ft to33,000ft (9 km to 10 km) in this area of theGulf of Mexico basin.

The time-based horizons in Figures 2 and3 need to be converted to depth for thedepth-imaging capabilities of long-offset P-P and P-SV data to be better appreciated.The transformation from image time todepth was done using P-P rms migrationvelocities determined during seismic dataprocessing. Examples of P-P rms velocitiesdetermined across part of the study area areshown in Figure 4. A north-south velocityprofile is displayed to give a sense of thevelocity behavior associated with the selectednorth-south P-P and P-SV seismic images(Figures 2 and 3). The velocity layeringexhibits vertical oscillations and thicknesschanges in the image-time interval between3 s and 6 s where propagating wavefields firstencounter salt-related structures.

The offset scale bar on the velocity profileis helpful for recognizing locations wherelateral velocity variations occur over dis-tances of the same dimension as the positiveand negative-offset range used in processingthe 4-C OBC data. Lateral velocity changesof this physical scale will complicate deeperimaging. Below 6 s, the velocity layering isreasonably smooth and uniform. All velocitylayers drop deeper at the north end of theprofile (Figure 4) where the thickest sedi-ment accumulation is encountered andwhere little high-velocity salt is present.

The basic message this depth-convertedprocess provides is critical information forexplorationists operating in the Gulf ofMexico basin. The key finding is that con-tinuous, mappable reflections (Horizon 1)often extend to depths of 30,000ft to33,000ft for P-P and P-SV data. Long-off-set 4-C OBC data provide quality P-SV and

P-P reflection images of Gulf of Mexicogeology to challenging drilling depths. Thefact that good quality, continuous P-P reflec-tions extend down to 33,000-ft targets is notsurprising. The fact that equivalent-qualityP-SV reflections are obtained for these sametarget depths is important, new information.

ConclusionsOBC seismic technology allows long-offsetseismic data to be acquired across congestedproduction areas where long-offset towed-cable seismic technology is not feasible.Further, 4-C OBC seismic technology pro-vides P-P and P-SV data. Towed-cable tech-nology provides only P-P data.

Practical drilling targets across theLouisiana shelf are now limited to depths ofabout 33,000ft or less. Long-offset P-P andlong-offset P-SV data provide quality, con-tinuous reflections to these depths. The doc-umentation that P-SV images are of a qual-ity equal to that of P-P images at depths of30,000ft to 33,000ft is critical information

for operators wanting to use multicompo-nent seismic data for improved understand-ing of lithofacies distributions and overpres-sure conditions across deep prospects. Thestudy confirms that a fundamental require-ment for good imaging of deep targets isacquiring long-offset seismic data. Theseresearch findings should encourage operatorsin the Gulf of Mexico basin to integratelong-offset 4-C OBC seismic technologyinto their prospect evaluations, particularlyin areas where there are congested produc-tion facilities. ✧

AcknowledgementsThe seismic data shown in this article are multi-client data available from WesternGeco, who isthanked for allowing the data examples to beshown. Initial funding support for the researchwas provided by the Research Partnership toSecure Energy for America (Contract No.RPSEA-R-521A). Ongoing research support isprovided by the U. S. Department of Energy(Program No. DE-PS26-04NT42072).


M-C SEISMIC

Figure 4. P-P rms migration velocities across the study area used to convert Horizons 1and 2 to depth. Velocity values increase with time within each colored layer.


ROCK PHYSICS

M ost seismic methodsto map fracturesdepend on a few

anisotropic wave propagationsignatures that can arise fromaligned fractures. Commonamong these are shear wavesplitting, azimuthal and offsetvariations of seismic P-waveamplitude, and azimuthal varia-tion in P-wave normal moveoutvelocity. While seismic anis-otropy can be a powerful frac-ture diagnostic, a number of situations can lessen its usefulness or introduceinterpretation ambiguities, including:

• multiple fracture sets at different orien-tations that combine to lessen theanisotropy;

• the presence of nonfracture rockanisotropy as might occur with large dif-ferences in horizontal principal stresses;

• fracture occurrence in narrow bands orswarms that are not sampled well byazimuthal methods; and

• seismic acquisition geometries that pro-vide limited azimuthal coverage.

Laboratory and theoretical work in rockphysics indicates that a broader spectrum offracture seismic signatures can occur, includ-ing a decrease in P- and S-wave velocities, achange in Poisson’s ratio, an increase invelocity dispersion and wave attenuation, aswell as indirect images of structural features

that can control fracture occurrence.In this article, we summarize an interpre-

tation and integration strategy for detectingand characterizing natural fractures in rocks.

The goal of rock physics is to discover,understand and quantify the links betweengeophysical measurements and the underly-ing rock properties. In the case of fractures,our models help quantify links between seis-mic observables and fracture density, orien-tation, pore fluid and permeability.

The physical mechanism of the elastic frac-ture effect is simple: fractures, whether wet ordry, soften the rock in a way that depends onthe fracture direction. The fracture-inducedreduction in elastic stiffness translates directlyinto reduction of seismic P- and S-wavevelocities, reduction in impedance and achange in Poisson’s ratio. If the fractures arealigned, then these fracture signatures tend to

be directionally dependent,resulting in seismic anisotropy.

The historical basis for mod-ern seismic methods for fracturedetection was the classic stress-induced velocity anisotropyexperiment by Nur andSimmons (1969), illustrated inFigure 1. The experiment wasperformed on a sample of gran-ite under various levels of uniax-ial compressive stress. The plotsshow P-wave and S-wavevelocities vs. direction of wave

propagation relative to the direction of appliedpressure. The cracks appear to be initiallyisotropically distributed, because the velocitiesare independent of direction at zero stress. Asuniaxial stress is applied, crack anisotropy isinduced. The velocities vary with direction rel-ative to the stress-induced crack alignment. Itis these fracture-related velocity changes thatwe look for in the field.

Integrated strategy for fracture characterizationDifferent types of information can be used forfracture characterization. Outcrop studies givedirect observations of the fracture orientation,spatial density and sometimes their lengths.Outcrops also lend insight into geomechani-cal relations, such as how fractures are relatedto strain, bed thickness and facies brittleness.The challenge is to extrapolate this

By Gary Mavko, DianaSava, Juan-Mauricio

Florez and Tapan Mukerji,Rock Physics Laboratory,

Stanford University

Integrated Seismic/Rock PhysicsApproach to CharacterizingFractured ReservoirsThe key to successful development of many low permeability reservoirs lies in reliably detecting, characterizing and mapping natural fractures. Fractures play a crucial role in controlling fluid transportin tight reservoirs, and they can influence production, sometimes adversely, even in reservoirs withmoderate to high permeability.

Figure 1. Ultrasonic P- and S-wave velocities vs. the direction of wavepropagation relative to the direction of applied uniaxial compression.


ROCK PHYSICS

information to the reservoir at depth. Seismicdata, on the other hand, provide good cover-age at depth, but the measurements are indi-rectly related to fractures, and their resolutionis lower than the scale of the features in whichwe are interested. Nevertheless, seismic attrib-utes yield information about fracture density,orientation and sometimes the type of fluidsaturating the fractures.

Our approach is to integrate the geologic,well log and seismic information in theframework of an inverse problem, as definedby Tarantola (1982, 1987). The inverse prob-lem has three different elements:

• the model parameters, represented bythe fracture characteristics that wewish to map;

• the data parameters, such as seismic andlog measurements; and

• the rock physics theories that relate themodel parameters to the data.There alsooften is prior information about the sub-surface fracture parameters from geo-logic constraints.

A general way to express the prior geologicknowledge about the fractures is through apriori probability density functions (PDFs).

We can think of the prior modelas the geologist’s best predictionof where the fractures are likely tobe, before seeing the quantitativeseismic information. Expressingthe prior model as a PDF allowsus to describe multiple likely frac-ture hypotheses, while at thesame time quantifying our level ofconfidence in the predictions.The geophysical data, which areaffected by measurement errors,also can be described throughPDFs. Finally, the theoreticalrelations between the fracturecharacteristics and the seismicdata can incorporate PDFs toaccount for approximations in themodels and natural variability ofthe input rock properties.

Probability theory allows us to integratethe various types of information and, at thesame time, to estimate the uncertainty in ourpredictions. The solution is represented bythe a posteriori PDF for the fracture charac-teristics, from which we can obtain theexpected values for fracture parameters, aswell as maps of uncertainty.

The workflow we have developed consistsof the following steps:

• analysis of the geologic scenarios,including stratigraphic and structuralcontrols on fracture occurrence. Theresults of this step are a list of likely frac-ture scenarios to evaluate – Which inter-vals might be fractured? How manyfracture sets might exist? What pore flu-ids should be considered? – constraintson fracture occurrence, and a prior frac-ture model, expressed as a PDF;

• rock physics fracture modeling, toexplore the seismic signatures of theplausible fracture distributions at the site;

• analysis of field seismic data to identifypotential artifacts, and extract attributesthat will be useful for the fracture inter-pretation; and

• integration of the prior information, thetheoretical relations and the geophysicaldata to produce the final fracture inter-pretation as well as quantitative esti-mates of uncertainty.

Geologic model of fracture occur-rence in a fractured carbonateWith the cooperation of Marathon Oil, wereceived well logs, seismic, lithologicdescriptions and geologic insights into afractured tight carbonate field in EastTexas. The dominant structural style in thefield site is normal faulting and the currentstate of stress is extensional. The intervalsof interest consist of the interposition oflayers of limestone and calcareous shales ata variety of scales, which impose importantcontrols on the mechanism and styles offracture formation.

We have developed a model for the evolu-tion of conjugate normal faults in sedimen-tary sequences, such as this, having high brit-tle/ductile contrast. Slip along ductile (shaly)bedding surfaces creates stress perturbationsthat result in the formation of subverticalsplay joints. Slip along these subvertical splayjoints generates a new stress perturbationthat creates a new set of splay joints. Thenew splay joints constitute new low-frictionplanes that slip, forming a set of conjugatenormal faults.

This evolution of conjugate normal faultsresults in a particular shape called the tec-tonic bowtie, (Figure 2). The limits of thebowtie are defined by the parent (synthetic)normal fault and the main two opposite-dip-ping antithetic faults. The zones between theparent fault and the antithetic faults arecharacterized by high-density fractureswarms, parallel to the parent fault. Fractureswarms may occur associated with faultswith less than 161⁄2ft of offset, and are there-fore below the seismic resolution. These frac-ture swarms, however, have a significantimpact on rock impedance.

Integration of a multi-offset vertical seismic

Figure 2. Outcrop examples of tectonic bowties.


ROCK PHYSICS

profile (VSP) and a horizontal well,confirms the presence of subseismicfaults and fracture swarms withinthe limestone. The presence of smallfaults is indicated by changes in thestratigraphy along the well trajec-tory, as well as by small displace-ments of the top reflection observedin the VSP data. These faults alsoare associated with reduction in theP-to-P reflectivity (“dim spots”),most likely related to fractureswarms (Figure 3).

Rock physics analysis and fracture modelingThe reservoir is modeled with twotypes of fracture distributions: local-ized bowtie fracture swarms, withinwhich the fractures can be more orless randomly oriented; and a moreuniformly distributed set of alignedvertical joints that generate anazimuthally anisotropic medium. Loganalysis indicates three main facies:unfractured clean limestones; shalylimestones; and clean limestones.FMI data indicate that the fracturestend to occur in the cleaner (clay-free)limestone intervals, which have lowerporosities and higher velocities. The

goal of the rock physics analysis is to find theoptimal combination of seismic attributes fordistinguishing the gas-filled fractured zonesfrom the shaly and unfractured limestones inthe reservoir.

For each of the fracture distributions, westochastically modeled seismic interval andinterface properties such as interval veloci-ties, impedances, travel times, and P-to-Preflectivity as functions of angle of incidenceand azimuth. The modeling shows that allthese properties may be influenced by thepresence of the fractures, with gas-filledfractures generally more visible than brine-filled fractures.

The AVO gradient-intercept domain ispotentially useful for detecting gas-filledfractures in the reservoir. As the fracturedensity increases in the clean limestones, theP-to-P reflectivity from the fractured zonesdecreases and becomes indistinguishablefrom the P-to-P reflectivity from shaly lime-stones. However, shaliness moves the AVOgradient to smaller negative values, whilegas-filled fractures move the AVO gradientto larger negative values (Figure 4).Therefore, at this site, AVO gradient appearsto be a good fracture indicator, while P-waveimpedance, alone, is not.

The Integration Methodology

Prior information about fracture parametersAt our field site, the fracture occurrenceis controlled primarily by the existingfaults – both the localized intensely-damaged swarms associated with bed-scale normal faults and the more dis-tributed background subvertical frac-tures associated with larger faults.Additional factors controlling fractureoccurrence include bed thickness, bedcurvature and local tectonic stresses.

Figure 5 shows a seismic horizonslice (left), from which a major fault isinterpreted. A prior model of the mean

Figure 4. Joint probability distribution functions ofthe AVO gradient and intercept for the MonteCarlo simulations of the unfractured, clean lime-stones (blue), fractured limestones with randomlyoriented cracks (green) and shaly rocks (red). Thefractures are filled with gas.

Figure 5. Left panel: Amplitude map at the top of a fractured carbonate reservoir, with the inter-preted fault. Bin size is 200ft. Right panel: Map with the a priori spatial distribution of the meanfracture density at the top of the reservoir, based on the interpreted fault.

Figure 3. Fracture-induced amplitude dim spotsnear a large fault.


ROCK PHYSICS

fracture density is shown on the right,described as a decaying exponential with dis-tance from the fault. The maximum expectedfracture density near the fault is assumed tobe about 0.07, which corresponds to an uppervalue for the crack density expected for areservoir at about a 1-mile depth, based on aworld-wide compilation by Crampin (1994).At each location, the prior PDF of fracturedensity is expressed as a truncated exponen-tial, with the mean shown in Figure 5.

Rock-physics modeling andstochastic simulations:Theoretical PDF Next, we perform rock-physics forward mod-eling and stochastic simulations based on thewell-log data, under the chosen geologicalhypotheses. Using Monte Carlo simulations,we obtain a variety of realizations of plausiblefracture parameters and seismic attributes thatspan the intrinsic natural variability of therock properties. Based on these realizations,we can estimate the theoretical joint PDFdescribing the physical relations between thefracture parameters, such as fracture density,and the seismic attributes at our site. Figure 6shows an example for the theoretical jointPDF of the fracture density and the azimuthalreflectivity anisotropy under the hypothesis ofa single set of aligned vertical fractures. Therock physics theory predicts increasingazimuthal reflectivity anisotropy withincreasing fracture density.

Seismic dataThe data parameters are represented byvarious reflectivity attributes from a 3-D seismic dataset acquired over thereservoir. A pitfall is the artificial ampli-tude patterns associated with the acqui-sition footprint. Increasing the bin sizehelped remove the footprint, though atthe expense of resolution.

Azimuthal analysis of the P-to-Preflectivity involves partial stacking ofthe data within different ranges of

azimuth. There is a tradeoff between theazimuthal resolution, which requires smallranges of azimuth, and the signal-to-noiseratio that requires larger fold and implicitlylarger azimuthal bins. For a fixed azimuthalrange, we can increase the fold by increas-ing the bin size, at the expense of reducingthe spatial resolution.

Under the hypothesis of a nearly vertical setof fractures, as indicated by FMI data, thereflectivity at far offsets varies with azimuth.Figure 7 shows the observed azimuthal reflec-tivity anisotropy at far offsets (left), with theassociated standard deviations (right). Themean values as well as the standard deviationsare derived using a bootstrap method to takeinto account the measurement errors associatedwith the reflectivity. The uncertainty because ofmeasurement errors is assumed to be Gaussian.

Quantitative integration of priorgeological information withseismic data using rock physicsFinally, we quantitatively integrate the priorgeological information and seismic data usingthe theoretical PDF, given by the rock physicstheories. At each location, we derive the pos-terior PDF over the fracture parameters anddata by multiplying the prior joint PDF on thefracture parameters and the seismic data, withthe theoretical PDF, containing the rockphysics information. We then integrate the a

posteriori PDF over the seismic attributesspace to obtain the updated distribution offracture parameters. This posterior PDF rep-resents the updated measure of uncertaintyabout the fracture parameters after integratingthe prior geological information with the seis-mic data using rock physics theories.

From this a posteriori PDF of the fractureparameters, we can derive statistical quanti-ties, such as expected values of fracture den-sity, and confidence parameters, such as theprobability that the fracture density exceedscertain thresholds. These probability mapshelp us assess the uncertainty in our predic-tions and can help us make informed deci-sions about the reservoir management.

Figure 6. Theoretical joint PDF of crack den-sity and azimuthal reflectivity anisotropy,derived based on the rock physics theories.The uncertainty is because of natural variabil-ity of the rock properties.

Figure 7. Left panel: Map with the mean values of the azimuthal reflectivity anisotropy at far offsets at thetop of the reservoir. Right panel: Standard deviations associated with the mean reflectivity anisotropy.


ROCK PHYSICS

Figure 8 shows a map with the spatial dis-tribution of the posterior expected values forfracture density at the top of the reservoir,conditioned on the azimuthal anisotropy ofreflectivity at far offsets. We observe a rela-tively higher fracture density near the fault,as the prior geological information suggests.We can also observe the asymmetric distrib-ution of the expected crack density withrespect to the fault, with higher values offracture density in the hanging wall. Thisagrees with outcrop data. We also find otherzones of higher fracture density away fromthe fault. These zones also may correspondto possible subseismic faults.

3-D Seismic data and AVAZfor fracture characterizationOnce we observe an azimuthal variation in theseismic amplitudes, the challenge is to interpretit in terms of fracture density, orientation andfluid saturation. Rock physics modeling is anecessary step in interpreting the integratedfracture density map shown in Figure 8. Thepolarity of the amplitude variation withazimuth depends on gas saturation and fracturecompliance. As a result, the azimuthal variationmay be used to determine the saturation if thefracture orientation can be determined fromother information. On the other hand, the sat-uration-compliance behavior can be a serioussource of ambiguity. At our site, the rock physicsmodels indicate that the fracture strike is in theminimum amplitude azimuth. From this, wederive a map with the fracture orientation andrelative fracture density,. Using a bootstrapmethod, we also estimate the uncertainty in thefracture orientation and the azimuthalanisotropy in the reflectivity because of mea-surement errors (Figure 9).

SummaryWe have presented an interpretation and inte-gration strategy for detecting and characteriz-ing natural fractures in rocks. We used a for-malism based in probability theory to integrateprior geologic models of fracture occurrence,

geophysical (log and seismic) data andtheoretical rock physics models linkingfractures, background rock propertiesand observable seismic attributes. In afield study of a fractured tight lime-stone, we find agreement between themean fracture orientations derivedfrom the azimuthal variation of theseismic amplitudes at far offsets andthe fracture orientations derived fromthe FMI logs from a nearby well.Therealso is agreement between the meanfractures’ strike from azimuthal varia-tion of amplitude and the presentregional stress field. The mean fractureorientations are approximately parallelto the maximum horizontal stress inthe region. As expected from geome-chanical considerations, we infer higherfracture densities in proximity to a seis-mically interpreted fault, with higherfracture density in the hanging wall.We also infer apparent high-fracturezones away from the fault, which canonly be detected from the seismic.

A critical part of our work is care-ful rock physics modeling to providethe quantitative links between log,geologic and seismic information. ✧

AcknowledgementsThis work was supported by contract DE-AC26-99FT40692 from the U.S.Department of Energy, Federal EnergyTechnology Center and the Stanford Rock Physicsand Borehole Geophysics project.We especially thankMarathon Oil for their participation and support.

References1. Crampin, S., 1994, The Fracture Criticalityof Crustal Rocks, Geophys. J. Internat., 118,428-438.2. Hudson, J. A., 1980, Overall Properties of aCracked Solid, Math. Proc. Camb. Phil. Soc.,88, 371-384.3. Nelson, R.A., 1985, Geologic Analysis ofNaturally Fractured Reservoirs, Houston Gulf

Publishing Co., Book Division, 1985.4. Nur, A. and G. Simons, 1969, Stress-induced Velocity Anisotropy in Rocks: AnExperimental Study, Journal of GeophysicalResearch, Vol. 74, 6667.5. Schoenberg, M and Muir, F, 1989, ACalculus for Finely Layered AnisotropicMedia, Geophysics, 54, 581-589.6. Tarantola, A., Vallete, B., 1982, InverseProblems=Quest for Information, J. Geophys.,50, 150-170.7. Tarantola, A., 1987, Inverse ProblemTheory, Elsevier Science B. V.

Figure 8. Map of the expected values of fracturedensity derived from the a posteriori distribution,obtained by constraining the a priori geologicalinformation with the azimuthal anisotropy ofreflectivity at far offsets, in the hypothesis of a ver-tical set of aligned fractures.

Figure 9. Picked amplitude variation with azimuthat a fixed superbin, and the 100 cosine fitsobtained using a bootstrap method, taking intoaccount the measurement errors represented bythe error bars.


DUAL-DENSITY DRILLING

Development of theU.S. deepwater gasresources is limited

by the high capital costsinvolved in developing theseresources. Dual-gradientdrilling methods have beenproposed as a means to pro-vide simpler, safer, more eco-nomic well designs andtherefore increase the ulti-mate development and uti-lization of deepwater gasresources. Two dual-densitydrilling concepts – riser dilu-tion with a low density liquidand riser gas lift – wereinvestigated as potential means to imple-ment a dual-gradient system. The overallobjective was to establish whether furtherresearch concerning dual-density drillingsystems based on use of low-density fluids,either liquid or gas, is justified. The investi-gation was conducted by the Craft andHawkins Department of PetroleumEngineering at Louisiana State Universityand sponsored by the Research Partnershipto Secure Energy for America.

The investigation focused on providinginitial answers to several critical questionsabout the practical feasibility and potentialcommerciality of these two dual-density

drilling methods. The first is the probablecost-benefit relationship for each.The secondis whether effective well control methods canbe defined. The third question is the practi-cality of separating the low-density and high-density components of the mixed fluid thatreturns to the surface in the riser for reuse.

The cost reduction when using a riser gaslift approach was estimated to be at least9%, and most likely 17% to 24%, vs. esti-mated trouble-free costs for conventionaldrilling of three example wells selected torepresent future deepwater Gulf of Mexicooperations. In addition, a riser gas liftapproach also increased the feasibility of

drilling deep wells in deep-water that might otherwisebe impossible. Well controlwith a riser gas lift systemalso was found to be feasi-ble when using methodsgenerally analogous to con-ventional operations or touse a subsea mudlift pump.Riser dilution using liquidswas estimated to reducecosts vs. conventional oper-ations by 7% for the exam-ple studied. The practicalityof separating and reusingmud returning from theriser for a liquid dilution

system was only partially investigated.However, results from this work and thatdone by de Boer (2003) indicate mud sep-aration and reuse in a riser dilution systemis possible.

This article will focus on the concepts andrationale for dual-density drilling and theinvestigation of riser gas lift as one mecha-nism for achieving a dual-gradient system.Part 2 will focus on liquid dilution of fluids inthe riser as an alternative means for achievinga dual-gradient system as well as the expectedcosts savings from using dual-density insteadof conventional drilling methods.

The rationale for investigating these dual-

By John Rogers Smith,Louisiana State University;

and Mikolaj Stanislawek,ENSCO

Dual-Density Drilling SystemsReduce Deepwater DrillingCosts: Part I—Concepts andRiser Gas Lift MethodA study of dual-gradient deepwater drilling systems relying on riser gas lift and riser dilution concludedthat both concepts are feasible and warrant additional research. Editors note: This is the first in a two-part series. The second part will publish in GasTIPS fall issue.

Pressure (psi)

TV

D (

ft) [R

KB

]

9 10 11 12 13 14Mud Weight (ppg)

3,000 6,000 9,000 12,000 15,000

5,000

10,000

15,000

20,000

25,000

0

Constant Mud Weightfrom Surface

Casing Point

20”16”

13 5/8”11 3/4”

9 5/8”

7 5/8”

5 1/2”

3 1/2” forProduction

24000’ Well in

10000’ Water Depth

Conventional Designwith Returns to Rig

Figure 1. Conventional well design for 24,000-ft Gulf of Mexico well in10,000ft of water.



density drilling methods is the expectationthat development of deepwater natural gasreserves should contribute significantly tonew gas reserves in the lower 48 states dur-ing the next 15 years, according to recentGas Technology Institute (GTI) baselineprojections. However, the current economicsignificance of deepwater gas production isconstrained by the substantial capital costs ofdeepwater development. Effective dual-den-sity drilling systems could potentially removethat constraint.

Even though wells have been drilled inwater as deep as 10,000ft, the riser insidediameter can severely limit the number ofcasing strings that can be used and conse-quently the maximum practical well depthwhen conventional well designs are used.These limits become more severe withincreasing water depth. An example of aconventional well design for a 24,000-ftwell in 10,000ft of water is shown in Figure1. Use of a single mud density from the wellbottom to the surface results in wellborepressures that can only be contained over ashort distance of drilling without settingcasing and increasing the mud density. Anexample interval is labeled “constant mudweight from surface” in Figure 1. The resultin this case is that eight casing strings

ranging in size from 20-in. to 3.5-in. arerequired to reach and set casing at the totaldepth of 24,000ft.

Specific geologic conditions, such as longsalt intervals, or more costly well equipment,such as expandable tubulars, can offset thelimitations on maximum practical depth orcasing size at total depth (TD). Neverthe-less, some deepwater resources will be leftunexplored or undeveloped because the cur-rent well design technology is too limited orcostly to be economically feasible.

A simpler, potentially more cost-effec-tive well design would use a moderate den-sity fluid in the annulus of the riser and ahigher density fluid in the wellbore to pro-vide a more favorable pressure profile in thewell, specifically a pressure profile closer towhat naturally exists in the subsurface for-mations. The drilling system that wouldallow these two different fluid gradients inthe well has been called the dual-density ordual-gradient system. An example of howthe fluid gradients and casing points in thiskind of well design would correspond withformation pressure gradients is provided inFigure 2.

The water depth, well depth, pore pres-sures and fracture pressures in Figure 2 arethe same as in Figure 1. However, use of a

higher density fluid in the wellbore and alower fluid density in the riser give wellborepressures that only require three casingstrings to reach and set casing at TD. In thiscase, the casing at TD could easily be 9.625-in., which would allow economic well pro-duction rates that would be impossible withthe 31⁄2-in. casing in the conventional design.Ultimately, the dual-density well design hasthe advantages of fewer casing strings forlower well cost, larger mud weight marginsfor improved safety, a larger production cas-ing size for increased production and revenuerates, and reduced riser tension requirementsthat would allow longer risers to be usedwith existing tensioning systems.

Several industry projects have been con-ducted to address the potential of the dual-gradient concept. The subsea mudlift drillingproject led by Conoco and Hydril developedprototype seafloor pumps that were used in asuccessful offshore field trial. The Deep-vision concept also used seafloor pumps andwas researched by Baker-Hughes andTransocean. The concept of using hollowglass beads to reduce the density of riser flu-ids was researched by Maurer Technology,and a previous, preliminary feasibility studyof the riser gas lift concept was performed byLSU and Petrobras. Dual-Gradient Systems

Pressure (psi)

TV

D (

ft) [R

KB

]

9 10 11 12 13 14Mud Weight (ppg)

3,000 6,000 9,000 12,000 15,000

5,000

10,000

15,000

20,000

25,000

0

Constant Mud Weight from Seafloor

Casing Point20”

13 5/8”

9 5/8”24000’ Well in 10000’ Water Depth

Dual Density Design with Injection at Seafloor

Figure 2. Well design for 24,000-ft Gulf of Mexico well in 10,000ft of waterusing the dual-density concept.

Figure 3. Schematic of the riser gas lift method.(Source: Lopes, 1997)

has pilot-tested a patented liquid dilutionsystem. However, none of these systems hasreached commercial application, and theriser gas lift and liquid dilution methodshave received less attention than the pump-based systems even though they potentiallyrequire less rig modification and initialinvestment to put into service.

Project descriptionThe overall objective of this project was toestablish whether more comprehensiveresearch concerning dual-density drillingsystems based on use of low-density fluids,either liquid or gas, is justified. The projectwas intended to continue the research initi-ated by LSU and Petrobras on the riser gaslift method (Figure 3) and begin assessinginjection of unweighted liquid into the riseras another alternative. These methods areintended to offer alternative methods ofachieving a dual-gradient deepwater drillingsystem that utilize more standard equipmentthan the separate industry projects focusedon the use of seafloor pumps to achieve theadvantages of a dual-gradient method.

The focus of the project has been to eval-uate and develop the operational conceptsfor two dual-density methods that can beapplied using current riser-supported subseadrilling systems: riser gas lift and injection ofan unweighted liquid into the base of theriser to dilute the wellbore fluids entering theriser. The results are intended to provide afirst step toward answering critical questionsabout the practical feasibility and commer-ciality of these systems.

The primary business question is whatthe probable cost-benefit relationshipwould be for each of the two alternativeconcepts if applied to deepwater Gulf ofMexico development and exploratory wells.This question was addressed by applyingdual-density design concepts to three rep-resentative deepwater well designs and con-sidering the cost impacts of the differentdesigns and rig equipment requirements.

The principle technical question is whetheran effective well control method can bedefined for a system containing so manydifferent density fluids and flow paths. Thisquestion was addressed primarily by simu-lating well control operations for represen-tative situations using a transient multi-phase flow simulator. An additional con-cern is the practicality of separating themixed fluid that returns to the surface intolow-density and high-density flow streamsfor reuse. This concern was addressed bystudying lab formulations of candidate mud

systems as well as using a lab centrifuge anda hydroclone to separate a riser mud intolow-density and high-density flow streamsto be used as dilution fluid and wellborefluid, respectively.

Riser gas liftThe conceptual feasibility of riser gas lift isdependent on achieving a significantly lowereffective density in the riser than in the well.Figure 4 shows the result of one studydemonstrating the feasibility of achieving awellhead pressure equivalent to the seawater



0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

0 10 20 30 40 50

Gas rate, mmscf/d

Pre

ssu

re, p

si

1,500 gpm

150 gpm

250 gpm

500 gpm

1,000 gpm

2236 psi=seawater pressure

Figure 4. Wellhead pressure with riser gas lift using 16 ppg mud at 5000-ft water depth.

Surface choke

11,600

11,800

12,000

12,200

12,40012,600

12,800

13,000

13,200

13,400

1 2 3

BOP Shut In

4 5 6 7 8 9 10

Time, hr

Pre

ssu

re, p

si

0

0.01

0.02

0.03

0.040.05

0.06

0.07

0.08

0.09G

as in

flu

x, b

bl/s

BHPFormation influx rate

Kick circulationbegan

Kick removed

Figure 5. Simulation of bottomhole pressure and formation flow rate during well control.



hydrostatic pressure over a range of circulat-ing rates with 16 ppg mud in a 191⁄4-in.inside diameter riser in 5,000ft of water.

A primary concern was whether an effec-tive well control method could be defined fora system containing the many different den-sity fluids and different flow paths inherentwith a riser gas-lift system. The specific con-cerns addressed were kick detection, cessa-tion of formation feed-in and removal ofkick fluids with a constant bottomhole pres-sure method. These concerns were studiedusing OLGA2000™, a transient multiphasesimulator. The validity of using the simulatorto study this system was confirmed by com-parison to transient multiphase flow datafrom a test well.

Conventional kick detection methodsrelying on the pit gain and return flow ratewere concluded to be effective. However, aflow check to determine a kick is in progressis not possible. Two alternatives for stoppingformation flow were considered, a “load-up”method of reducing the nitrogen rate vs.closing a subsea blowout preventer (BOP).BOP closure was shown to be faster andmore reliable for stopping flow and minimiz-ing kick volume.

Figure 5 shows the results from an exam-ple simulation of well control with riser gaslift. In this case, the well was shut-in withthe BOP, and then the kick fluids were cir-culated out through the choke line and asurface choke using essentially conventionalmethods. Based on simulations like this ofseveral different alternatives, it was con-cluded that methods using a choke withreturns up the choke line were simpler andmore effective than alternatives with returnsup the riser and/or using nitrogen rateadjustments to control wellbore pressure.Bottomhole pressure was maintained rela-tively constant using only choke adjust-ments and no variation in the nitrogeninjection rate, which should indicate thatimplementation of successful well control inthe field would be relatively straightfor-

ward. However, the variation in bottomholepressure was larger than typically desired.The cause of this variation is two-fold. Thefirst cause is that there is always a gas phasein the choke line, which makes the effect ofa surface choke pressure adjustment lesspredictable than for a system that has liquidin the choke lines. Using a remote-con-trolled seafloor choke upstream of thechoke line could alleviate this. The othercause relates to the simulator rather than tothe real system. The simulator runs only ina batch mode, which prevents simple orquick correction of choke pressure andrequires rerunning the simulation each timea wrong adjustment in choke pressure ismade. Consequently, it is difficult and time-consuming to get simulations with exactlythe desired results, and those shown weredeemed acceptable for the purpose of show-ing system feasibility.

There are some differences vs. conven-tional well control. Nitrogen injection intothe base of the choke line lowers wellheadpressure, which alleviates the concerns relat-ing to excessive bottomhole pressure becauseof choke line friction that occurs in conven-tional well control. One additional item ofequipment vs. that in a conventional systemalso would be desirable. Use of a drillstringvalve that has been developed for use withthe subsea mudlift drilling system wouldprevent excessive bottomhole pressure whenthe well is on choke with the drillstring filledwith heavy wellbore mud.

Overall, well control with a riser gas-liftapproach to dual-density drilling should bepractical using essentially conventionalmethods.

ConclusionThis article has focused on the rationale fordual-density drilling and the different con-cepts for implementing a dual-gradient sys-tem. It also describes the investigation ofriser gas lift as one mechanism for achievinga dual-gradient system, and concludes that

riser gas lift is operationally feasible and wellcontrol operations will be essentially conven-tional. Part 2 of this series will focus on liq-uid dilution of fluids in the riser as an alter-native means for achieving a dual-gradientsystem as well as the expected costs savingsfrom using dual-density instead of conven-tional drilling methods.

Detailed reports on all parts of the projectto investigate the feasibility and economicsof these dual-density systems are available aspublications on the GTI Web site,www.gastechnology.org ✧

References1. de Boer, L.: Method and Apparatus forVarying the Density in Drilling Fluids inDeep Water Oil Drilling Applications, UnitedStates Patent No. 6,536,540, March 25, 2003.2. Forrest, N., Bailey, T., Hannegan, D.:“Subsea Equipment for Deep Water Drillingusing Dual-Gradient Mud System,”SPE/IADC Paper No. 67707 presented at the2001 SPE/IADC Drilling Conference,Amsterdam, Feb. 27-March 1, 2001.3. Lopes, C.: “Feasibility Study on the Red-uction of Hydrostatic Pressure in a DeepWater Riser using a Gas-lift Method,”Ph.D. Dissertation, Louisiana State Univers-ity, April 1997.4. Maurer Technology: “JIP to Develop Hollow-Sphere Dual-Gradient Drilling System,”http://www.maurertechnology.com/JIP/DGD/DGDProposal.pdf, September 2001.5. Schubert, J.J., Juvkam-Wold, H.C., Choe. J.:“Well Control Procedures for Dual-GradientDrilling as Compared to Conventional RiserDrilling,” SPE/IADC Paper No. 79880 pre-sented at the SPE/IADC Drilling Conference,Amsterdam, The Netherlands, Feb. 19-21, 2003.6. Smith, K.L., Gault, A.D., Witt, D.E.,Weddle, C.D: “Subsea MudLift Drilling Joint Industry Project: Delivering Dual-Gradient Drilling Technology to Industry,”SPE Paper No. 71357 presented at the 2001SPE Annual Technical Conference andExhibition, New Orleans, Sept.30–Oct. 3.


DRILLING VIBRATIONS

APS Technology isdeveloping a uniquesystem to monitor

and control drilling. This sys-tem has two primary elements:an active vibration damper(AVD) to minimize harmfulvibrations, whose hardness iscontinuously adjusted; and areal-time system to monitordrillstring vibration andrelated parameters. This mon-itor adjusts the damperaccording to local conditions.

The AVD is designed tohave several favorable effectson the time needed to drill awell. By keeping the bit inconstant contact with thewell bottom and maintainingthe actual weight-on-bit(WOB) at the optimumlevel, the instantaneous rateof penetration (ROP) isincreased. Additionally, byreducing the levels of vibra-tion throughout the bottom-hole assembly (BHA), the operating life ofall downhole components, such as bits,motors and measurement-while-drilling sys-tems, is increased, thereby reducing thenumber of trips required for a particular well.These advantages will apply in all wells, buttheir value increases disproportionately indeep drilling.

An earlier paper reported on the designand modeling of this system. After brieflyreviewing these, we present preliminary lab-oratory tests illustrating the ability of theAVD to adjust to a range of downhole con-ditions. Field test prototypes are beingdesigned and built, and will be field testedthis year.

The drilling environment,and especially hard-rockdrilling, induces severe vibra-tions into the drillstring. Theresults of the vibrations arepremature equipment failureand reduced ROP. The onlyway to control vibration withcurrent monitoring technol-ogy is to change the rotaryspeed or WOB. Thesechanges may move the drill-ing parameters away fromtheir optimum value and mayhave a negative effect ondrilling efficiency.

Shock subs are not a uni-versal solution, since they aredesigned for one set of condi-tions. When the drilling envi-ronment changes, shock subsbecome ineffective and mayresult in increased drillingvibrations.

Drillstrings develop vibra-tions when run at criticalrotary speeds. These vibra-

tions are difficult to control because of thestrings’ long length and large mass.Operating at a critical speed can impartsevere shock and vibration damage to thedrillstring, and fatigue drill collars and rotaryconnections. Vibrations also cause the drill-string to lift off the bottom, reducing ROP.The effect of axial, lateral or torsional

By Martin E. Cobern, APS Technology; and

Mark E. Wassell, APS Oilfield Services LP

Laboratory Testing of an ActiveDrilling Vibration Monitoring& Control SystemThe deep, hard-rock drilling environment induces severe vibrations, which can cause reduced rates ofpenetration and premature failure of the equipment. An active vibration control system can preventsuch problems and increase the rate of penetration.

0

1

2

3

4

5

6

7

8

9

10

0 5 10 15 20 25

Drillstring Frequency (Hz)

Dri

llstr

ing

Resp

on

se (

G)

Vertical Well (Light Damping)

Horizontal Well (Medium Damping)

Zone A

Zone B

Zone C

Figure 1. The frequency response of a typical drillstring.

0

1

2

3

4

5

6

7

8

9

10

0 5 10 15 20 25

Drillstring Frequency (Hz)

Dri

lls

trin

g R

es

po

ns

e (

G)

Vertical Well (Light Damping)Horizontal Well (Medium Damping)MR Damper (Optimum Damping)

Zone A

Zone B

Zone C

Zone A - With MR Damper

Figure 2. Typical drillstring response with an active vibration damper in use.


DRILLING VIBRATIONS

(stick-slip or bit whirl) vibra-tion upon drilling has beendocumented in the laboratoryand the field.

The natural frequencies ofthe drillstring often fall in therange excited by typical drillingspeeds: between 0.5 Hz and 10Hz, depending on the BHAand length of the drillstring.There are many sources thatexcite drillstring vibrations,including bits, motors, stabiliz-ers and drillstring imbalance.For example, a tri-cone bitimparts a primary excitationfrequency of three times therotary speed. If rotatingbetween 120 rpm and 180 rpm,the excitation frequency isbetween 6 Hz and 9 Hz.

Mud motors also are signif-icant sources of excitations onthe drillstring. The rotor of themud motor moves in an eccen-tric orbit that oscillates severaltimes per revolution. Depend-ing on the lobe configurationof the motor, excitations occurbetween 1 Hz and 30 Hz.Shocks from bit bounce andcollar impacts against theborehole result in higher fre-quency vibration.

The best situation for adrillstring is to operate belowits lowest critical speed. Bystaying below this first criticalspeed, drilling frequencies do not excite thedrillstring and the bit maintains contactwith the cutting surface of the borehole. InFigure 1, this safe range is Zone A. In thisexample, with a fundamental natural fre-quency of 6 Hz, Zone A extends to 4 Hz,corresponding to a rotary speed to 80 rpm orless for a tri-cone bit. Zone B is the resonantrange that results in high levels of vibration.

Shock and vibration damage and low ROPoccur in this zone. Zone C is above the firstcritical speed of the drillstring. Vibrationslevels are reduced compared with Zone Aand B; however, the bit does not maintaincontinuous contact with the drilling surface,since the natural frequency of the drillstringis lower than the excitations of the bit, pre-venting it from reacting to irregularities in

the bottom. This discontinu-ous contact with the drillingsurface of the borehole greatlyreduces the ROP.

Principles of operationThe AVD consists of electron-ics that monitor vibrations andother drilling parameters, anda spring-fluid damper thatcontrols the vibration. Thedamper properties are contin-uously modified to provideoptimal damping characteris-tics for the conditions. A keyinnovation in the AVD is theuse of magnetorheologicalfluid (MRF) as the means ofvarying the damping coeffi-cient of the AVD.

MRF is a “smart” fluidwhose viscous properties arechanged by passing a magneticfield through it. The fluid’scomponents have no movingparts, rapid response times andlow power requirements. Thedamping properties can beoptimized to detune the drill-string from resonant vibration.

MRF damping is being usedin such diverse applications assophisticated automotive sus-pensions and earthquake pro-tection systems for buildingsand bridges. The AVD modi-fies the properties of the BHAin two ways that combine to

increase ROP and reduce vibration. First, thedamper decouples the drillstring sectionbelow the damper from the one above it.Second, it optimizes the damping based uponthe excitation forces, significantly reducingthe vibration. The combination allows the bitto respond more quickly to discontinuities onthe cutting surface, while maintaining thedesired surface contact force.

0

100,000

200,000

300,000

400,000

500,000

600,000

0 1,000 2,000 3,000 4,000 5,000

Dampin (lb-sec/in.)

WO

B (

lb)

Max WOBMin WOB

Figure 3. Measured weight-on-bit: 30,000 WOB – 30,000-in.-lb spring rate.

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

0 1,000 2,000 3,000 4,000 5,000

Damping (lb-sec/in.)

Ac

ce

l -

g

Figure 4. Bit acceleration: 30,000 WOB – 30,000-in.-lb spring rate.

0

2

4

6

8

10

12

0 1,000 2,000 3,000 4,000 5,000

Damping (lb-sec/in.)

RO

P (

ft/h

r)

Figure 5. Rate of penetration: 30,000 WOB – 30,000-in.-lb spring rate.

Separating the bit from the rest of thedrillstring with a spring-damper assemblyreduces the effective mass that must respondto discontinuities of the drilled surface.Reducing the mass increases the first criticalspeed of the drillstring attached to the bit,while the adaptive damping reduces themagnitude of vibration at the resonance.This provides a wider Zone A, (Figure 2,which is based on a simple model of thedamper). For a tri-cone bit, Zone A nowcovers a range between 0 rpm and 220 rpm,a significant improvement compared withthe 0 rpm to 80 rpm in Figure 1.

The practical effects of these changes areshown in the following figures, which arebased on BHA models performed with APS’sWellDrillTM software. When the dampingcoefficient is optimized, for this case tobetween 200 lb-sec./in. and 300 lb-sec/in.,the bit remains in contact with the formation,the WOB remains constant (Figure 3), bitvibration is essentially eliminated (Figure 4),and the ROP increases (Figure 5). Note: WhenWOB goes to zero, the bit is off bottom.

Tool designAn overview of the AVD tool is shown inFigure 6. The tool has many features of aconventional shock sub, including a stack ofBelleville washers to support the weightapplied to the bit and bearings to absorb theaxial and torsional loads. The key differenceis that the damping coefficient is continuallyadjustable by varying the magnetic fieldapplied to the MRF. The details of the MRFdamper design are shown in Figure 7. Thisdrawing is of an earlier test piece, but theconfiguration is largely unchanged in theprototype tools. The MRF will be in the vol-ume between the housing (1) and mandrel(2). A series of coils wrapped in the groovesin the mandrel (in the latest design, the coilsare in the outer housing, not the mandrel)will create bucking fields, which will bestrongest in the gaps between the coils. TheMRF in these areas will become more viscous

as a function of the field strength, therebyvarying the damping of the motion of themandrel relative to the housing.

The MRF damper control algorithm uti-lizes displacement measurements taken in realtime during the drilling operation. Based onthis information, the damping properties arecontinuously modified throughout the drillingprocess. The intent is to reduce the motion ofthe bit relative to the well bottom and smoothout the vibrations above the damper. A hard-ening damper algorithm was developed as thesimplest and most robust method to controldamping (Figure 8). This method increasesdamping levels as the damper sections displacerelative to one another. For small displace-ments and low WOB, a low level of dampingis provided. As the deflection increases,because of higher WOB or larger vibrationlevels, the damping is increased. This methodwas shown analytically to provide properdamping levels throughout a range of condi-tions with minimal sensor data.

The new AVD tool prototype design uses

a Belleville spring stack with a compoundspring rate. This provides better isolation atvarious levels of vibration and WOB com-pared with shock subs having a linear springrate, which provides increased deflection forthe damping module at low levels of vibra-tion and improved support at high levels.

Test benchIn operation, the AVD will be supported andloaded by the entire drillstring above it.Considerable weight is applied from above,which will have resilience and damping.The damping will result from the intrinsicdamping in the drillstring, from thehydraulic damping of the drilling fluid andfrom contact with the borehole walls. At thebit, the driving force is the interaction of thebit and the irregular bottom of the hole. Thisinteraction will have a primary frequency,such as triple the rotation rate for a tri-conebit, but may have other harmonics as well ifthere is more than one high point on the wellbottom. In addition, the well bottom is not

Figure 6. Overview of active vibration damper sub.

Figure 7. Detail view of the adjustable damping element design.


DRILLING VIBRATIONS


DRILLING VIBRATIONS

completely rigid but can respond to the bitby flexing or being drilled away.

To simulate these conditions, we designedthe test bench (Figure 9). The prototype (5)is supported by linear bearings (4) on a largeload frame (6). At the “uphole” end, to theleft, a large pneumatic cylinder (1) applies aforce simulating the loading from the drill-string above the tool. The damping of thedrillstring motion is simulated by twohydraulic cylinders (2) configured to produceadjustable damping. To mimic the drivingforce of the bit’s interaction with the wellbottom, a lower assembly (7) is provided. In

this assembly, a cam (8) is rotated by a vari-able speed gear motor (9) at rates simulatingthe drillstring rotation rate. The cam, whichis supported by ball bearings, can have con-figurations that mimic a variety of degrees ofirregularity of the well bottom.

Test resultsEarly static testing established that the AVDcould provide adequate damping to supportabout 6,000 lbs of force. With the Bellevillesprings supporting the majority of the WOB,this should be adequate to provide the neces-sary damping in most drilling conditions.

In the next phase of testing, the full labora-tory prototype, including the Belleville springsand bearings, was mounted in the test benchand driven by the cam. A sample of the resultsis shown in Figure 10, which plots the dynamicstiffness of the AVD as a function of the cur-rent applied and the drive frequency. Thedynamic stiffness of the damper is a combina-tion of the stiffness of the springs and variabledamping applied by the AVD. This combina-tion is a function of the frequency of the dri-ving vibration. As the applied current increases,the dynamic stiffness of the AVD rises.

The AVD can instantaneously vary itsdynamic stiffness by a factor between 7 and10, depending upon the excitation frequency.This is more than adequate to obtain theresults described earlier in the article. Theability of the damper to reduce bit vibrationand bounce is shown in Figure 11, whichplots the maximum motion of the dampercollar (connected to the bit) relative to thecentral mandrel, (connected to the upperdrillstring. The driving displacement fromthe cam was 0.7. As the damping is increasedby increasing the applied voltage, the maxi-mum motion converges toward a level con-sistent with the bit’s remaining in constantcontact with the cam, which simulates theirregular bottom of the well.

ConclusionsAVD laboratory testing indicates it is amplycapable of providing the variable dampingnecessary to control bit bounce, maintainuniform WOB and increase drilling ROP.

Once some optimization issues have beenresolved, it is anticipated a prototype AVDtool will be tested, first in drilling laborato-ries and then in the field, later this year.

Based on the laboratory results, the AVDis likely to have a significant impact upondrilling operations, particularly in hard rockdrilling. As illustrated in Figure 11, a prop-erly damped drillstring will drill more effi-ciently, with fewer failures, than one withoutsuch damping. The modeling indicates theFigure 9. Active vibration test bench.

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

-6 -4 -2 0 2 4 6

Disp - in

Da

mp

ing

- lb

-se

c/in

Figure 8. Active vibration damper hardening algorithm.


DRILLING VIBRATIONS

potential for increasing the average ROP by10% to 50%, depending upon conditions.

Future effortsThe detection and active damping of axialvibrations represents the first step in thisprogram. There are other troublesome vibra-tion mechanisms, such as transverse vibra-tions and stick-slip torsional modes. It islikely that by modulating the WOB, theAVD will reduce the likelihood of the bit’sdigging into the borehole bottom and initi-ating sticking, followed by slipping when theWOB decreases. This effect will be evaluatedin the field trials.

Plans are underway to add torsional andlateral sensors, and damping to the AVDsystem. Initial designs have been prepared.Further studies will investigate whether thebenefit from these additional features justi-fies the additional costs involved. ✧

AcknowledgmentsThis effort has been partially funded by the DeepTrek program of the U.S. Department of EnergyNational Energy Technology Laboratory,Contract DE-FC26-02NT41664. The authorsare grateful for this generous support.

This article is based on paper AADE-05-NTCE-25, presented at the AmericanAssociation of Drilling Engineers NationalTechnical Conference, Houston, in April.

References1. Martin E. Cobern & Mark E. Wassell,Drilling Vibration Monitoring & ControlSystem, presented at the National GasTechnologies II Conference, Phoenix, Ariz., Feb.8-11, 2004.2. T.M. Warren, J.H. Oster, L.A. Sinor, andD.C.K. Chen, Shock Sub Performance Tests,International Association of DrillingContractors/Society of Petroleum EngineersPaper No. 39323, presented at the 1998IADC/SPE Drilling Conference, Dallas,March 3-6.

3. cf., e.g., M.W. Dykstra, D.C.-K. Chen, T.M.Warren & S.A. Zannoni, ExperimentalEvaluations of Drill Bit and Drill StringDynamics, Society of Petroleum EngineersPaper No. 28323, presented at the 61st SPEATCE, New Orleans, Sept. 25-28, 1994.4. cf., e.g., S.L. Chen, K. Blackwood and E.Lamine, Field Investigation of the Effects ofStick-Slip, Lateral and Whirl Vibrations onRoller Cone Bit Performance, Society ofPetroleum Engineers Paper No. 56439, presented

at the 58th SPE ATCE, Houston, Oct. 3-6, 1991.5. cf., e.g., Magnetic Ride Control, GM TechLinks, Vol. 4, No.1, pp.1-2, January 2002,http://service.gm.com/techlink/html_en/pdf/200201-en.pdf 6. B.F. Spencer Jr., S.J. Dyke, M.K. Sain andJ.D. Carlson, Phenomenological Model of aMagnetorheological Damper, Journal ofEngineering Mechanics, ASCE, Vol. 123, pp.230-238, 1997, http://www.nd.edu/~quake/papers/MRD.Journal.pdf

0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

0 0.5 1 1.5 2 2.5 3 3.5 4

Current - amps

Dy

na

mic

Sti

ffn

es

s -

lb

/in

0.5 Hz

1.0 Hz

1.5 Hz

2.0 Hz

Figure 10. Dynamic stiffness of the active vibration damper as a function of current and frequency.

Figure 11. Maximum relative motion of the damper components during testing.


HPHT CEMENT SYSTEMS

I ndustry growth in the United States findsdrillers moving into deeper and hotterfrontiers in search of oil and gas.

Remediation of failed cement jobs in deephot wells costs the industry more than $100million each year. While additives and well-designed placement techniques can increasechances for achieving long-term zonal isola-tion in high-pressure, high-temperature(HPHT) wells, a high incidence of gas pres-sure on deep hot wells because of inade-quate placement and mechanical failure ofcements, continues to be reported.

Although applications and methods mayvary, Portland cements have been usedworldwide and are recognized as the indus-try standard for well sealants. Temperatureand pressures encountered through produc-tion and intervention can induce highstresses in the wellbore, exceeding the resis-tance capabilities of conventional Portlandcement systems. Short-term fluid migrationof water and gas also affects the perfor-mance of wellbore cement, and mechanicalcement failure is exaggerated in wells withnarrow annuli because of higher stress onthe cement sheath.

With current completion technology, asupercement designed for long-term sealingintegrity in HPHT wells is needed to allevi-ate escalating costs associated with annularseal loss and associated remedial repair.Remedial procedures for restoring sealintegrity are expensive and multiple applica-tions often are required during the well’s life.Annular seal loss in HPHT wells can result

in well abandonment, and potential environ-mental and safety issues.

Deep Trek, a U.S. Department of Energy(DOE) program, seeks to improve HPHTwell economics by improving drilling andcompletion technology. As part of the DeepTrek effort, CSI Technologies is attemptingto improve the economics of deep-well com-pletions with the development of a superce-ment capable of providing long-term,HPHT sealing integrity. The material mayalso be applicable in less extreme environ-ments that still require high stress resistance,or where annular gas pressure is a knownproblem. With tensile and compressivestrength, permeability, expansive properties,and pipe and formation bonding sufficientfor long-term durability, the potential formechanical failures at temperatures exceed-ing 350°F and at pressures over 15,000psiwill be minimized.

Three years—Three phasesLast year, the CSI laboratory team beganworking on the first phase of the three-phase, 3-year Deep Trek project. Phase Ibegan with laboratory analysis of variousPortland and non-Portland materials, andadmixtures in conventional and unconven-tional tests. Identification of compositionsthat provide the mechanical propertiesrequired for withstanding extreme downholetemperatures and pressures resulted. InPhase II, the team’s work is continuing withscale-up testing and well trials to determineperformance on a field scale. Scale-up activ-

ities include manufacturing, mixing, place-ment mechanics and performance in a testwell. Demonstrations of the cement’s perfor-mance in three to six field applications in hotdeep wells and technology transfer activitieswill complete Phase III.

An industry advisory committee, comprisedof representatives from 12 companies, hasassisted in the technical development of thesupercement material and continues to providea combined expertise for project execution.Membership is voluntary and includesAnadarko, BHP Billiton, BP, ChevronTexaco,ConocoPhillips, the DOE, Dominion,McMoran, PDO, 3M, Shell and Unocal.

In Phase II, committee members willassist in monitoring and evaluating perfor-mance of the candidate supercement mater-ial(s) in actual wells. Additionally, commit-tee members contribute data to a deep hotwell database used for test design, specifica-tion development and as a technicalexchange medium.

Phase I—Project objectivesPhase I, conducted from October 2003through December 2004, concentrated onidentifying state-of-the-art sealant systems,especially those applicable to HPHT envi-ronments. A baseline series of Portlandcement designs commonly used in HPHTenvironments was identified and tested todetermine currently-available performance.Various candidate systems, designed andscreened at low and high temperatures, nar-rowed the field to a manageable number.

By Fred Sabins, LarryWatters and Kevin Edgley,

CSI Technologies; and Tim Grant, U.S.

Department of Energy

Long-term Cement Integrityof HPHT Cement SystemsDeep Trek, a U.S. Department of Energy program, seeks to improve HPHT well economics by improv-ing drilling and completion technology. As part of the Deep Trek effort, CSI Technologies is attemptingto improve the economics of deep-well completions with the development of a “supercement” capa-ble of providing long-term, HPHT sealing integrity.


HPHT CEMENT SYSTEMS

Finally, unconventional tests were utilized togain insight into the sealing capability of thecandidate systems. Thiswork is in preparation ofPhase II scale-up andfield trial work.

General projectexecutionDuring Phase I, 169 dif-ferent slurry compositionswere evaluated in morethan 1,100 laboratorytests. Low-temperaturescreening tests were aban-doned for some slurries,because they were engi-neered for reactions athigher temperatures andwould not attain strengthat lower temperatures.The “base” series of test-ing consisted of compres-sive strength, compre-ssive Young’s Modulus and tensile strength utiliz-ing the splitting tensilemethod (Figure 1).

Tests not routinely con-ducted in today’s industryinclude flexural strength,shearbond, anelastic strainand annular seal (Figure2). These unconventionalanalyses helped gaininsight into the interac-tions between a material’smechanical properties andwellbore sealing capabil-ity. Anelastic strain is ameasure of permanentdeformation in cement asa result of repeated low-stress load application.This permanent defor-mation represents cumu-lative damage, leading to

eventual failure at loads well below ultimatestress levels (Figure 3).

No combinations ofmechanical properties, andspecifically no single mechan-ical property, are sufficient topredict the ability of acementing material to pro-duce an effective and long-term annular seal in a well-bore environment. The pre-cise nature of the interactionof the various mechanicalproperties of cementitiousmaterials, in context of wellfailure, is yet unknown. Somecorrelation work has beendone, but it represents a pre-liminary evaluation of cementproperties required to form acompetent annular seal in awellbore environment.

Significant efforts wereexpended developing mono-lith production, test protocolsand equipment modificationsfor the project. Specializeddevices were designed andbuilt to measure deformationduring loading for compres-sive Young’s Modulus, as wellas for flexural test beams. TheSplitting Tensile Test (ASTMD3967-95A) was chosen fortensile strength testingbecause of the simpler testrequirements, sample produc-tions and test procedures. Acontrolled-rate press wasmodified to allow for cycle-testing determining anelasticstrain properties.

Annular seal testThe annular seal test is con-sidered the best available pre-dictor of a material’s sealing

performance in a well environment. Cementis placed in a geometrically representativetest fixture with loading applied from inter-nal pipe pressure. A cross section of the testfixture is shown in Figure 4.

The red outer tube represents the forma-tion or outer pipe and can be varied to simu-late formation of various strengths. The graymaterial is the cement sheath, and the bluepipe represents the wellbore tubing or casing.The yellow end piece conducts gas flow tothe cement sheath. In practice, the cement iscured in the fixture. Low-pressure gas(10psi) is connected to the open end of theyellow end piece, and a source of high-pres-sure water is connected to the inner diameterof the blue tubing.

Water pressure is applied in successive pres-sure cycles in increments of 1,000psi and thenreleased, starting at 1,000psi. When a maxi-mum pressure of 10,000psi is reached, fourmore cycles of 10,000psi are applied. Thissequence represents one full cycle; after which,the cycle is repeated starting at 1,000psi. Theflow of gas remains at zero while the annularseal is intact. When the cement fails, a flowpath opens through the cement and the gasflow rate is measured. In practice, the materialmechanical properties and the interactionsbetween the properties determine when theannular seal is compromised. Materials engi-neered for sealing performance at high tem-peratures may not set at low temperatures ormay exhibit mechanical properties differentfrom those obtained at higher temperatures.As Phase II progresses, the annular seal test isbeing redesigned to evaluate material proper-ties at elevated temperatures.

Energy analysisCement sheaths subject to loads imposed bywell conditions and intervention activitiesare subject to failure at some point. In thewell, a strong formation and heavy pipeessentially “back up” the cement, resulting inmore energy absorption before cement fail-ure. Laboratory studies conducted in the

Figure 2. Measurement ofthe shear bond between theinner pipe and cement.

Figure 1. Splitting tensile testmethod (ASTM D3967-95A).Diametrically-applied forceinduces tensile stress in aplane normal to the loadapplication.


HPHT CEMENT SYSTEMS

annular seal model indicate that cement fail-ure is related to the amount of energyimposed on the cement sheath during thelife of the well. Mechanical properties ofcement strength are important, but there areother cement and non-cement variables thataffect long-term integrity of the cementsheath. These variables include:

• cement properties- cement tensile strength- cement Young’s Modulus- anelastic strain - radius of cement sheath

• well properties- size and wall thickness of casings- hole size- formation Young’s Modulus

• loading data- anticipated intervention and produc-

tion loading profileThe correlation involves a dimensionless

energy application factor, representing theenergy applied to the cement sheath in a well;and an energy resistance factor, representingthe ability of the pipe/cement/formation system to resist the applied energy. The cor-relation derived to date is useful but not com-prehensive in terms of mechanical properties.Additional tests, using different mechanicalproperties and load rates, are planned.

Test compositionsCompositions tested throughout the projectinclude:

• baseline Portland cement recipes;• ceramicrete and its derivatives;• reduced water systems (highly dis-

persed);• high-concentration reactive and nonre-

active fibers;• Portland cement with unconventional

additives; and• non-Portland cements.Throughout this project, material selection

and chemistry efforts are based on the theorythat improvement in the crystalline structureresults in reducing porosity and permeability,thereby strengthening the mechanical structureof the cement and improving annular sealingperformance. As shown in the table on the nextpage, low values are desirable for anelastic strainand annular seal test results; high values aredesirable for the remainder of the reported tests.

Portland cements are comprised of a highlyheterogeneous matrix in which large particlesof various materials are present in a cementi-tious binder.These cements are relatively brit-tle, exhibit significant compressive strengthsand relatively low tensile strengths, and bondreasonably well to steel. Heat can degrade thecement during time, and the matrix is usually

porous and permeable because of the excesswater required to mix and pump the material.

In Portland systems, not much can be doneabout the mechanically inefficient matrixstructure, but decreasing the amount of waterpresent in the slurry can reduce shrinkageand improve strength, permeability andporosity. The following strategies were uti-lized to affect performance improvements.

Reduced water—Reducing the amount ofwater present in the slurry improved strengthand reduced porosity. This is straightforwardcement chemistry where strength is generallyproportional to the cement-to-water ratiofor a given cement composition. The slurriestested with reduced water were highly dis-persed to improve mixability. In commoncements, highly-dispersed slurries can causedifficulties because of settling. The slurriesinvestigated were not tested specifically forsettling, but appeared to produce reasonablesuspension properties for the solids due towater content reduction.

Fibers—Fibers are known to improve ten-sile strength, or more specifically, the tough-ness of cement. However, they can actuallycreate problems by bonding mechanically tothe matrix and providing fracture points whenthe cement is stressed. Reactive fibers, such asceramics and silicates, were used to provide

0

1,000

2,000

3,000

4,000

5,000

6,000

0:00:00 0:07:12 0:14:24 0:21:36 0:28:48 0:36:00 0:43:12 0:50:24 0:57:36 1:04:48 1:12:00

Time - Minutes

Load

- lb

s

0.094

0.095

0.096

0.097

0.098

0.099

0.1

Dis

plac

emen

t - in

Load - lbsDisplacement - in

Figure 3. Measure of permanent damage (measured in permanent strain) in cement sam-ple as a result of low-stress loading.

Figure 4. Measures annular seal integrityin a geometrically-representative model.Failure is related to a complex interactionof cement mechanical properties and ind-icated by measured gas flow.


HPHT CEMENT SYSTEMS

mechanical and chemical bonding prop-erties to the matrix. Various composi-tions and fiber lengths were investigated.Results incorporating ceramic fibers areencouraging, and fibers are applicable toa wide range of Portland and non-Portland systems.

Unconventional additives—These inc-lude additives commonly used incements, but in higher concentrationsthan typical, as well as new additives notutilized in the industry. Studies focusedon molybdenum and high concentrationsof magnesium oxide, both of which causesignificant expansion in the cementmatrix. If expansion occurs at the righttime during the hydration process, it canresult in a denser matrix, higher strength andlower permeability product.

Timing is critical during the hydrationprocess; if expansion occurs too early or toolate, it can lead to matrix disintegration.When cured under confined conditions,where dimensional expansion is restricted,the expansion is theoretically directed inwardto the matrix, thereby reducing porosity andpermeability. In practice, this method has ledto high shear bonds, although both compres-sive and tensile strengths of cored sampleshave been modest. Annular seal tests areplanned to determine the ultimate potentialof the material.

Non-Portland systems—Initial researchfocused on ceramicrete and other acid-basesystems, but produced insufficient results forinclusion in Phase II. Other systems includeda calcium-aluminum-silicate mix and a high-temperature resin.

A range of calcium aluminum silicate, cal-cium aluminum phosphate and magnesiumaluminum phosphate systems were screenedfor strength development at 350°F. Althoughresults were unexceptional, work in this areawas incomplete at the end of Phase 1.Testingwith a CAS system varying slurry density andcomponent ratios is continuing during PhaseII, and initial results are encouraging.

The high-temperature resin producedhigh tensile strengths, which were as muchas 10 times higher than those of conven-tional Portland cements. Typically, resinssoften with temperature, making them ofdubious value in high-temperature wells.The resin under consideration has been for-mulated to retain properties at elevated tem-peratures, thereby qualifying the material asa candidate for this project. The high elastic-ity exhibited by the material may be useful inattaining long-term sealing integrity; otherproperty testing is incomplete.

A full-scale mixing test yielded the followingdifficulties with handling the resin on a largescale in an oilfield cementing environment:

• vapors pose possible health risks;• surface mixing consists of mixing the

liquid resin with a liquid activator in pre-cise relative quantities;

• air and water entrainment causes a persis-tent entrainment issue in the mixing tub,which can cause problems if density mea-surements are required during mixing;

• resin chemistry is a highly exothermicreaction, creating high temperatures thatcan damage mixing equipment and posea health risk to personnel;

• resin can be weighted or lightened forplacement efficiencies. The limits of

weighting and the impact on materialperformance are unknown;

• resin does not mix with water, which maygenerate advantages or disadvantages.Issues include displacement mechanics,mixing requirements, and sweep effi-ciency for mud and spacer removal; and

• mixing on a field scale will require thedevelopment of specialized equipmentwith sealed containers, precise rate con-trol and dedicated pumps. The heat-activated nature of the material maymake it difficult to place the material inthe HPHT well environment. Cleanupprocedures and disposal of cleanupmaterials must be addressed.

ConclusionsThe first year of this Deep Trek-related pro-ject has generated new insights and theoriesinto the mechanisms by which cementmechanical properties translate to annularsealing performance in oil and gas wells. Testprotocol and analysis methods have beendeveloped to better understand cement per-formance, and a suite of materials has beenidentified for evaluation in scale-up field tri-als. Ultimate improvements in HPHT welleconomics will result with the successfuldevelopment of an HPHT supercement. ✧

Preliminary test results of candidate systems. Blue—Best properties (or nearly best). Red—Worst val-ues in systems studied. Low values—Desirable for anelastic strain and annular seal tests. High values—Desirable for remainder of tests.

System FormulaCompressive

Strength psi

TensileStrength

psi

AnelasticStrain

in/in/min

Shear Bondpsi

Annular Seal

Ml/min flow

Baseline 77 5,650 730 44% 471 1.0

Baseline 99 4,790 710 100% 256 2.4

MgO 128 3,190 280 In Proc 1,850+ In Proc

Moly 132 4,280 1,080 25% 310 2.3

Moly 133 6,680 1,370 18% 570 3+

Resin 120 In Proc 2,210 In Proc In Proc In Proc

Resin 121 In Proc 3,200 In Proc In Proc On Proc

Fiber 130 3,710 1,020 52% 299 0

Fiber 131 3,020 1,150 54% 186 0.6

Fiber 136 4,510 1,310 31% 1,021 In Proc

Ca Al Silicate 169 1,920 220 In Proc In Proc In Proc


BEHIND-PIPE PAY ZONES

The infill drilling approach is reasonableand has been successfully demonstratedin many basins. As a result of this

development and advances in technology, newand additional pay zones are identified andexploited. However, the transfer of this knowl-edge to already existing wells is frequentlyoverlooked. A research project funded byRPSEA investigated how to better define andeffectively stimulate behind-pipe pay zones inexisting wellbores in tight-gas sand sequences.The need in “existing wells” is driven by therequest of independents and reflects the pref-erence to optimize existing wells over capitalexpenditures for new development. In thisarticle, only stimulation aspects of behind-pipezones are discussed, however the entire reportis available from the Gas Technology Institute.

The San Juan Basin, roughly circular inshape (Figure 1), is an asymmetrical synclinein northwestern New Mexico and southwest-ern Colorado. The selected target for thisresearch is the Menefee Formation of the pro-lific Mesaverde Group in the San Juan Basin.

The Mesaverde is divided into a basal unit,the Point Lookout Sandstone, a relatively uni-form blanket-type, low-permeability sandstoneformation; the Menefee Formation, overlyingthe Point Lookout, a sequence of coals, sandsand shales; and the Cliff House formation,which is similar in reservoir characteristics tothe Point Lookout, but less extensive.

Production/development historyThe Blanco Mesaverde Pool in the San JuanBasin was discovered in 1927 and has grown inresponse to pipeline capacity, decreased well

spacing and price increases.Extensive development occur-redin the 1950s on 320-acre spacingwhen the Western gas marketbecame available. Since then, thespacing has been reduced twice, to160-acres per well in 1975 and to80-acre spacing in 1998.Cumulative production to date isgreater than 10 Tcf, with an esti-mate of 7 Tcf of additional provenreserves. At this time, more than4,900 completions yield a total of0.75 Bcf/d. It is expected thatwithin the next 20 years, about4,300 additional completions willoccur, many of which will be dualcompletions with the deeperDakota Sandstone.

Traditionally, the focus of mostoperators was to concentrateefforts on the clean, relativelythick and laterally extensive PointLookout and Cliff House Sandstones, ignor-ing the Menefee Formation between the two.This underdevelopment of the Menefee isbecause of the lenticular nature of the reser-voirs and difficulty of evaluating the reser-voirs using geophysical logs.

StimulationIn response to the low permeability of thesetypes of reservoirs, stimulation (particularlyhydraulic fracturing) is required to develop acommercial well. Typically, this stimulation isdone during the initial completion of a well.The addition of “behind-pipe” pay leads to a

variety of problems in designing an effectivestimulation treatment. Using the Menefee asour example, the lenticular and discontinuousnature of the embedded sands can lead tocomplex fracture geometries; exacerbated bythe reduced net effective stress because of par-tially-depleted layers above and below theMenefee.The mechanics of fracturing are sub-ject to maximum allowable pump rate andpressure for older wells, casing or tubing sizeand integrity, and avoiding re-fracing old per-forated zones in an attempt to stimulate onlythe new, “bypassed” pay zone.

To assess the Menefee contribution to over-

By Tom Engler, New Mexico TechCapturing Missed

Reserves in Existing WellsThe motivation behind capturing the missed reserves is the geologic complexity of the reservoirs,such as sequences of low-permeability sands and shales, existence of natural fractures and variablereservoir architecture caused by abrupt changes in stratigraphic facies.

Figure 1. The map of New Mexico portion of San Juan Basin shows a Mesaverde Group outcrop, BlancoMesaverde pool, Chacra line and Chacra pools. (RFD,FFO, U.S. BLM, 2001)



all well performance and theimpact stimulation plays on thisperformance, the approach was to:

• analyze the production perfor-mance of Menefee pay-adds.As mentioned previously, inmany cases the Menefee hasbeen bypassed as a potentialreservoir because of its poorreservoir characteristics;

• compare production perfor-mance from wells completedwith and without theMenefee, both original par-ent-type wells and daughterwells, in a selected region; and

• compare and discuss theeffectiveness of the variousstimulation treatments.

Menefee payadd refers to aremedial workover where theMenefee Formation is completed and stimu-lated within an existing wellbore. The successof these workovers can be evaluated by the dif-ference between the before and after produc-tion, or the incremental reserves developed byadding the Menefee. A subset of more than 40wells was identified as Menefee payadds, geo-graphically located across a wide area withinthe San Juan Basin. This remedial work beganin 2002.

Two examples are shown in Figure 2. Thefirst production curve represents the best well,resulting in first year incremental productionof 63 MMscf. The second production curveillustrates an unsuccessful payadd with poorproduction response, with after workover pro-duction less than the previous rate.

Overall, an estimated incremental ultimaterecovery of 1.7 Bscf of gas can be attributed tothe 40-well, Menefee program. Of the 40wells, production performance after theworkover of four wells was initially less thanthe prior rate. These wells have continued thispoor trend and therefore are deemed failures.The remaining wells are estimated to average46 MMscf of incremental gas per well, to be

retrieved during a 2-year span. The best wellwill incrementally produce 285 MMscf.

To further distinguish the Menefee contri-bution, production performance was com-pared among four types of wells; old wells(originally completed between the late 1950sand early 1960s) with Menefee, old wellswithout Menefee, wells completed in the1980s and those with recent (about 1998) welldevelopment. Several important factors thatplay a role in the variation of performancemust be considered:

• the change in completion effectivenesswith time. Tech-nological advances pro-mote the idea of improved hyd-raulicfracture performance;

• pressure depletion of the Cliffhouse andPoint Lookout Formations. A new wellmay add gas reserves from Menefee, butmay recover less reserves from theCliffhouse or Point Lookout because ofprevious pressure depletion. All producingzones are commingled in the wellbore;

• changing surface and/or market con-straints. Gas-gathering line pressureshave been reduced in the San Juan

Basin, resulting in longer produc-tion life and reserves. The addi-tion of compression will have thesame impact. In the past, a prora-tion system was in effect reflectingmarket conditions; and

• variations in flow and storagecapacity from well to wellbecause of the discontinuousnature of the formations.

The target area to be evaluatedwas on the east side of the basinwhere data for more than 75 newwell completions was collected andanalyzed along with 30 originaland 1980s wells, all within thesame geographic area.

Results from this analysis aredisplayed in Figure 3 as cumulativeproduction curves. The time scalehas been shifted so each series

begins at time zero, even though the actualdate of initial production is different. Thesecurves represent average production per wellfor each designated group. The comparisonbetween the original wells with and withoutthe Menefee has the advantage of minimizingthe impact of the four factors previouslydescribed. Inspection of the curves for theoriginal wells shows that adding the Menefeeincreases incremental per well cumulative pro-duction by 103 MMscf.

Figure 3 shows the average cumulative pro-duction for wells completed in the early 1980s.Note the identical trend of this group of wellsand the previous original, with Menefeegroup. This behavior implies:

• flow and storage capacity are reasonablyconsistent throughout the study area,not anunreasonable assumption since new wellsare between existing development wells;

• completion methods are similar and pro-vide no improvement in effectiveness;

• surface and market constraints are simi-lar for both groups of wells. This is pos-sibly true in this portion of the basin butnot scaleable to the entire basin; and

1

10

100

1,000

10,000

100,000

1,000,000

1 5 9 13 17 21 25 29 33

Time (mo)R

ate

(msc

f/m

on

th)

MenefeePay add

1

10

100

1,000

10,000

100,000

1,000,000

10,000,000

100,000,000

1 5 9 13 17 21 25 29 33

Time (months)

Rat

e (m

scf/

mo

nth

) MenefeePay add

Figure 2 Two examples of Menefee pay add wells. Success (a) and fail-ure (b). (Gas—red squares, oil—green diamonds, water—blue triangles)



• pressure depletion has beenminimal in the Cliffhouseand Point Lookout Form-ations. Recent pressure tran-sient tests on infill wells haveconfirmed the lack of pres-sure depletion.

A final comparison shown inFigure 3 is the recent wells drilledsince 1998. These wells are com-posed primarily of 160-acredevelopment wells, all of whichincluded the Menefee in theoriginal completions. Resultsindicate a substantial improvement of wellperformance during the first 6-year life ofthese wells. In this case, improved perfor-mance for new wells could reflect advancesin stimulation technology, more favorablemarket/surface conditions or increased con-tributions from the Menefee. A decreasedperformance for the new wells could resultfrom pressure depletion of the Cliff Houseand Point Lookout Formations. In bothinstances, variations of storage and flowcapacities are assumed constant.

A final area of interest is to compare stimu-lation methods and effectiveness. A cursoryreview of stimulation methods was undertakento identify whether any specific treatment wasmore effective than another treatment.

Early (1950s to 1960s) completions con-sisted of “block” perforating with two shots/ftor greater, and then hydraulically fracturingdown casing with slick water only or water andsand combined. By the early 1980s, wells wereselectively perforated with 20 to 40 holes overthe Mesaverde section. Stimulation consistedof a small near-wellbore, acid cleanup followedby 80,000 gal of water with 80,000 lb of 20/40sand, also pumped down the casing. The 1998wells were completed almost exclusively byselective perforating and fracture treating downcasing with slick water and 80,000 lb of 20/40sand, in two stages. The first stage treated thePoint Lookout only, while the second stagecombined the Menefee and Cliffhouse.

In summary this work does not reveal thespecific advances in fracture technology withtime such as cleaner frac fluids, improvedadditives or proppants. However, severalgeneral observations can be made:

• stimulation type was consistentthroughout any given time period.Only minor fluctuations in volumesand method occurred in the studygroup, and there is no difference instimulation between wells with or with-out the Menefee. Stimulation is not thereason for the incremental productionshown in Figure 3, but is a result of theMenefee pay included in the well;

• the slight variation of completions inthe early 1980s appears to have littleimpact on improving well performance.As shown in Figure 3, the cumulativeproduction curve is identical to the onebetween 1960 and 1983, with Menefeecurve; and

• well performance for the 1998 well grouphas been significantly better than its pre-decessors. This could be the result ofimproved stimulation effectiveness, inwhich case, acceleration of productionmay be dominant. Alternatively, betterquality of Menefee pay also would lead tothe addition of reserves.

The Menefee payadd wells were all selec-tively perforated and then hydraulicallyfractured down tubing with a 70-quality

nitrogen foam and about 70,000lb of 16/30 sand. Pump rateand allowable pressure restric-tions were encountered by treat-ing through tubing instead ofcasing. This requirement isbecause of the existing CliffHouse perforations above theMenefee, thus limiting the stim-ulation options available.

SummaryKey factors influencing perfor-mance are:

• variations in completion practices fromwell to well with time;

• pressure depletion of the Cliffhouse andPoint Lookout Formations;

• changing surface and/or market con-straints; and

• variations in flow and storage capacityfrom well to well.

Differentiation of the impact of each indi-vidual factor was not attempted; however, sev-eral general observations are apparent fromthis initial work:

• the Menefee adds incremental reservesto the production stream. This workdemonstrated better recovery (103MMscf ) when the Menefee is includedin the initial completion than whenadded later (46 MMscf ) as a remedialworkover;

• the payadd stimulation is constrained bythe existing Cliff House completionabove the Menefee target. Consequently,stimulation effectiveness is reduced;

• the latest stimulation practices appear tobe successful, resulting in improved wellperformance; and

• the best payadd candidate resulted in285 MMscf of incremental reserves.This potential confirms the need forimproved reservoir description toidentify the best targets to addMenefee and to avoid those areas thatdo not exhibit potential. ✧

0

200,000

400,000

600,000

800,000

1,000,000

1,200,000

0 5 10 15 20 25 30 35 40 45 50

Time, YEARS

Cu

mu

lati

ve P

rod

uct

ion

, MS

CF

Without Menefee (1960-2003)With Menefee (1960-2003)With Menefee (1982-2003)With Menefee (1998-2003)

103 mmscf

Figure 3. Average cumulative production for wells (with and with-out the Menefee contributing) as a function of date of completion.


PIPELINE INSPECTION

ITT Industries has introduced a break-through airborne technology that makes itpossible to inspect pipelines up to 100

times faster and provides 30 times more right-of-way (ROW) coverage than traditionalmethods.The Airborne Natural Gas EmissionLidar (ANGEL) Service detects and quanti-fies leaks with a high degree of accuracy. It alsosurveys, maps and images pipeline corridorsand surrounding areas then delivers detailedGIS-ready datasets and reports. And it does itall, day or night – more than 1,000 miles a day.

Because it’s airborne, the ANGEL Servicedoesn’t have to deal with obstacles that oftenslow down ground crews. Issues such as diffi-cult terrain, property access, inclementweather and insufficient field staff are a thingof the past. The ANGEL Service also can beused in known dangerous areas without hav-ing to put people in harms way. Perhaps bestof all, the ANGEL Service is a cost-effective,end-to-end, out-sourced service. By subscrib-ing to the ANGEL Service, natural gasproviders not only obtain necessary informa-tion required to meet today’s enhanced regu-latory compliance, but also detailed informa-tion that will benefit various departmentswithin their company including pipelineintegrity management (PIM); geomatics(GIS); environment, health and safety (EHS);and operations and maintenance (O&M).

While the technologies used by theANGEL Service (differential absorptionLIDAR, aerial mapping and surveying, anddifferential global positioning system (GPS) toname just a few) may be new to some in thenatural gas industry, they have been in use for

many years in different applicationsranging from cartography to environ-mental monitoring. ITT is knowl-edgeable in these technologies, as theyhave been developing innovativeremote sensing technologies fordefense, intelligence and aerospaceapplications for more than 50 years.ITT remote sensing technology isused everyday in weather and imagingsatellites, space telescopes and GPSsystems to collect information andimages of earth and space.

ANGEL Service operationsWhen a company subscribes to theANGEL Service, they’re buying aturnkey service from a Fortune 500company that is a worldwide leaderin the remote sensing sciences. ITThandles every detail of the logistics,including all aircraft operations,data collection and processing, andreport generation.

The ANGEL Service provides aninstrumented leak survey using a remotesensing system built on differential absorp-tion lidar (DIAL), an extremely sensitive gasdetection and quantification technology.Installed aboard a Cessna Grand Caravan208B aircraft, the ANGEL system collectsup to 180,000 laser measurements perminute while flying at 1,500ft above theground at 120 mph. That means it’s sensi-tive, thorough, efficient and accurate. TheANGEL system scans the pipeline corridor,

remotely sensing ethane and methane emis-sions. It then precisely images and maps theplume’s physical size and ground locationallowing the client to fully understand thesituation before responding.

To ensure targeting accuracy and pipelinecoverage, the ANGEL system features acomputer-guided laser pointing and trackingsub-system. It scans a 100-ft wide groundtrack along the pipeline ROW. This widecoverage area increases the likelihood of

By Steven Stearns, ITT Industries,

Space Systems DivisionA High-flying Alternative to Walking the LineMore than 2,000 years ago, natural gas was sent across Tibet via bamboo pipes. How did they maintain the integrity of that system? By walking the line. For as long as there have been natural gaspipelines, inspectors have had no choice but to walk the line to detect leaks and ensure public safety.Now the endless walk is over.

Figure 1. Conceptual view of the ANGEL aircraftcapturing survey data along pipeline right-of-way.


PIPELINE INSPECTION

detecting leaks, even those that might havemigrated away from the pipe.

At the same time, the ANGEL system cap-tures color high-resolution orthorectified digitalimagery of the entire ROW and surroundingareas. These GIS datasets can be overlaid withmaps and other database information. Infor-mation can be extracted from this imagery suchas high consequence area identification andthird-party encroachment. It can also be seam-lessly integrated into a client’s existing digitalgeomatics and integrity management systems.

Unparalleled high-level outputAfter the flight, the ANGEL ground teamprocesses and analyzes the datasets providingunparalleled high-level output and reports.

ITT provides certified findings in the form ofwritten reports plus GIS vector layers, orthorec-tified color imagery and a video log of the flight.These detailed reports offer ITT’s clients arange of capabilities.

These reports make it possible to respondquickly and effectively when leaks occur orpipeline integrity has been compromised. Inaddition, the data contained in reports pro-vides an overview of pipeline ROW condi-tion that makes it easier to plan maintenance.

If a client is researching the purchase or saleof a transmission system, the ANGEL

Service can produce better-informed deci-sions during due diligence. Records can bearchived and accessed as needed. Using digitalmaps and databases, viewers in remote loca-tions can call up an accurate model, a virtual“big picture” of the pipeline’s status. Finally,the digital record of the ROW makes it pos-sible to fly identical missions to compare pastdata with the current status of the pipeline.

With 100 times the speed and 30 times theROW coverage compared with traditionalmethods, the ANGEL Service is a leap forwardfor leak survey. The additionof current mapping imageryand videography makes for acomplete service that willaddress the growing needs ofmany diverse departmentswithin a company.

Factor in the ability toaccurately collect a rangeof valuable data, minimizetransmission downtime andensure public safety, do itall cost effectively via anend-to-end turnkey servicefrom ITT, and it’s no won-der the ANGEL Service isgaining wide acceptance inthe industry. ✧

AcknowledgementsThis work was partially funded by a cooperativedevelopment agreement from the U.S.Department of Energy’s National EnergyTechnology Laboratory.

For more information on the ANGELService, or to schedule a site visit, contact:Dan Brake, 800 Lee Road, PO Box 60488,Rochester, N.Y. 14606-0488. Tel: (585) 269-5121. E-mail: [email protected], Web site:www.ssd.itt.com/angel

Figure 3. The ANGEL system detects, quantifies,images and maps gas concentrations of methaneand ethane gas.

Figure 4. The conical scan display shows points color-codedto reflect the measured concentration wavelength values formethane and ethane.

Figure 2. Ten-time enlargement of natural gas plume. ITT ANGEL system hasdetected emissions of natural gas. The plume is integrated with color imageryand GIS coordinated to provide the precise location.

“STRIPPER WELL”TECHNOLOGIES HOLD PROMISETO BOOST DOMESTIC OIL ANDGAS PRODUCTION

Joint ventures in technology development by gov-ernment and industry have delivered six newdeployment-ready applications in 4 years to extendthe useful life of more than 650,000 stripper wellsthat deliver almost 15% of America's domestic oilproduction and almost 8% of natural gas produc-tion, according to a U.S. Department of Energy(DOE) review.

The technologies were developed by the StripperWell Consortium, an industry-directed group whoseresearch, development and demonstration efforts areco-funded by the DOE through the NationalEnergy Technology Laboratory’s Strategic Centerfor Natural Gas and Oil. The six new technologiesthat have been commercialized, or are near commer-cialization, generally serve the purposes of increasingproduction, raising efficiencies or lowering costs.The consortium has been active in bringing alongmore than 55 additional technologies, some ofwhich are approaching commercial readiness.www.netl.doe.gov/publications/press/2005/tl_strip-per_well.html

ENERGY DEPARTMENTMOVES FORWARD ON ALASKANATURAL GAS PIPELINE LOANGUARANTEE PROGRAM

The Department of Energy published a notice ofinquiry in the Federal Register seeking public com-ment on an $18 billion loan guarantee program toencourage construction of a pipeline that will bringAlaskan natural gas to the continental United States.The pipeline will provide access to Alaska's 35-Tcf of

proven natural gas reserves and would be a step for-ward in meeting America's growing energy needsand reducing our dependence on foreign sources ofenergy. It would also fulfill the Bush Administration'spolicy to bring Alaska's natural gas reserves to market.For more information visit www.fossil.energy.gov/news/techlines/2005/tl_pipeline_noi.html

DETECTION SYSTEM CAN HELPFIND PIPELINE LEAKS, DAMAGE

A new, lightweight device that uses natural gas todetect leaks in natural gas pipelines has been suc-cessfully tested on a transmission main owned andoperated by Dominion Transmission Inc. inMorgantown, W. Va.

The test was conducted by the U.S. Departmentof Energy’s National Energy TechnologyLaboratory (NETL) and West Virginia University,which has worked with NETL for the past 2 yearsto advance the detection system. The device is one ofa suite of technologies being developed by theEnergy Department’s Office of Fossil Energy toeffectively and efficiently monitor the 1.3 millionmiles of transmission and distribution pipelines thatcriss-cross the United States.

Known as the Portable Acoustic MonitorPackage, the device – 14-in. long, 18-in. tall andweighing 5 lb – uses a variety of tools to capture andrecord sound waves transmitted by natural gas.Computer software included in the package analyzesand interprets the recorded signals to detect possibleleaks. Natural gas compressor stations, which pumpgas from one station to another, have a distinctsound, as does natural gas flowing normally throughfittings and valves. Variations to these and otherbackground sounds alert operators that a leak mayexist. www.netl.doe.gov/publications/press/2005/tl_acoustic_monitor.html


BRIEFSPUBLICATIONS

EVENTS

HEAVY OIL POTENTIALKEY TO ALASKAN NORTHSLOPE OIL FUTURE

Alaska’s North Slope boasts a massive heavy oilresource that someday could underpin the survival ofone of the nation’s most critical oil-producingprovinces – and research funded by the U.S.Department of Energy may provide the key tounlocking this vast but, to date, largely intractable oilresource. North Slope operators have had some suc-cess producing the less-viscous crudes in the WestSak and Schrader Bluff formations by injecting slugsof water alternating with gas (WAG) into the reser-voirs; the gas acts as a solvent to reduce oil viscosity,while the water front helps sweep the reservoir, push-ing the crude to producing wells. Research by theUniversity of Houston has developed tools for mod-eling the optimum WAG flood design. The goal ofthe research was to focus on modeling tools thatwould determine the best solvent, injection schedule,and well architecture for a WAG process in NorthSlope shallow-sand viscous oil reservoirs.www.netl.doe.gov/scngo/index.html

Gas Technology Institute (GTI)1700 S. Mount Prospect RoadDes Plaines, IL 60018-1804Phone: (847) 768-0500; Fax: (847) 768-0501E-mail: [email protected] site: www.gastechnology.org

GTI E&P Research Center1700 S. Mount Prospect RoadDes Plaines, IL 60018-1804Phone: (847) 768-0500; Fax: (847) 768-0501E-mail: [email protected] site: www.gastechnology.org

GTI/CatoosaSM Test Facility, Inc.19319 N. E. 76th, Owasso, OK 74015Phone: Toll-free (877) 477-1910 Fax: (918) 274-1914E-mail: [email protected]

U.S. Department of Energy National Energy Technology Laboratory Web site: www.netl.doe.gov/scngo

3610 Collins Ferry RoadMorgantown, WV 26507-0880

626 Cochrans Mill RoadPittsburgh, PA 15236-0340

One W. Third St.Tulsa, OK 74103-3519

525 DuckeringFairbanks, AK 99775

Office of Fossil Energy1000 Independence Ave., S.W. Washington, DC 20585Web site: www.fe.doe.gov

▲

▲ ▲▲

CONTACT INFORMATION

AAPG EASTERN MEETINGSept. 18-20, Morgantown, W. Va. For more informationvisit: www.wvgs.wvnet.edu/www/esaapg05/

AAPG ROCKYMOUNTAIN MEETING

Sept. 24-26, Jackson, Wyo. For more informationvisit: www.spe.org

▲▲

Schlumberger coalbed methane (CBM) services combine global knowledge with extensivelocal experience earned working on thousands of CBM wells in the United States. Our dis-tinctive pairing of expertise and innovative technology helps producers improve operationalefficiency through a better understanding of the unique reservoir characteristics associatedwith CBM exploration and production.

Successful CBM evaluation begins with superior petrophysical characterization thatreduces the need for coring. The Schlumberger RST* Reservoir Saturation Tool and ECS*Elementary Capture Spectroscopy sonde can be applied in both open and cased holes toidentify remedial and bypassed coal rework opportunities. Our new-generation dipole sonictools help to evaluate coal cleating and determine directional permeability. Efficient decisionsfor either horizontal or vertical completions can be made quickly with the assistance ofSchlumberger Data & Consulting Services and ECLIPSE® CBM simulation software. Coalcleat permeability damage can be greatly reduced by using the LiteCRETE* CBM cementsystem and CoalFRAC* fracturing fluid.

To find out about Schlumberger CBM services for your U.S. operations, contact us at [email protected].

We have the technology.

More information on Schlumbergeranswers to your CBM concerns is at www.slb.com/oilfield.

05-OF-166

*Mark of SchlumbergerECLIPSE is a registered trademark of Schlumberger.

Looking for improved efficiency for your coalbed

methane operations?

Looking for improved efficiency for your coalbed

methane operations?

pg 01 gastips-masttoc - netl.doe.gov · long-term cement integrity of hpht cement systems deep...

Documents