42 Oilfield Review

From Reservoir Specificsto Stimulation Solutions

Ali O. Al-Qarni Saudi Aramco Udhailiyah, Saudi Arabia

Brian AultRoger Heckman Sam McClure Ultra Petroleum, Inc. Englewood, Colorado, USA

Stan Denoo Wayne RoweEnglewood, Colorado

David Fairhurst San Antonio, Texas, USA

Bruce KaiserHouston, Texas

Dale Logan Midland, Texas

Alan C. McNally Louis Dreyfus Natural Gas Inc. Midland, Texas

Mark A. Norville Milton R. Seim Kerns Oil and Gas Inc. San Antonio, Texas

Lee RamseyAl Khobar, Saudi Arabia

For help in preparation of this article, thanks to UsmanAhmed, Kamel Bennaceur, Leo Burdylo and Mo Cordes,Sugar Land, Texas, USA; Tom Bratton, Mike Donovan and Steve Neumann, Houston, Texas; Paul DeBonis,Englewood, Colorado, USA; and Joe Lima, Farmington, New Mexico, USA.AIT (Array Induction Imager Tool), CMR (CombinableMagnetic Resonance), CMR-Plus, CNL (CompensatedNeutron Log), DataFRAC, DESIGN-EXECUTE-EVALUATE, DSI (Dipole Shear Sonic Imager), ECS (Elemental CaptureSpectroscopy), ELAN (Elemental Log Analysis), FMI(Fullbore Formation MicroImager), FracCADE, Litho-Density,NODAL, Platform Express, PowerJet, PowerSTIM, QLA, RFT (Repeat Formation Tester) and UltraJet are marks of Schlumberger.

Comprehensive data acquisition, interpretation and modeling provide a thorough

understanding of basins and fields, a prerequisite for successful well completions.

With more complete information, teams of experts develop, refine and apply

improved models to design perforation, completion and development strategies

that enhance productivity.

Between two and three billion dollars US arespent annually to fracture more than 20,000wells worldwide.1 However, less than 1% of frac-turing treatments are optimally designed to max-imize production and recovery. In spite ofincreasing demand for stimulation services, theGas Technology Institute, formerly Gas ResearchInstitute, Chicago, Illinois, USA, reports that two-thirds of hydraulically fractured wells in theUnited States do not respond as expected andfail to meet operator objectives.2 The same istrue in other parts of the world. One reason citedfor this poor performance is lack of an optimiza-tion process. Operators, therefore, are continuallystriving to improve stimulation practices.

In the early 1980s, it seemed as though everywell needed a hydraulic fracturing treatment;reputations and résumés were built based onpounds of proppant pumped—many “records”were set. Later, the industry found that, as withmost things, there was a point of diminishingreturns, and optimization became a key word.

During the past two decades, there has beensome optimization of well stimulations, but notnearly enough. Even today, the tendency is to relyon fracturing treatments that have always beenperformed the same way in a particular area.This means that stimulation design utilizing com-prehensive data and detailed designs are not yetcommon practice.

In addition to enhancing oil production frommarginal reservoirs, well stimulation is becomingincreasingly important because of growing inter-est in natural gas, which often is found in lowerpermeability zones. Formations with low to mod-erate permeability may require hydraulic fracturingto produce at economic rates. Even for reservoirswith higher permeabilities, stimulation is an effec-tive way to improve production or acceleraterecovery, especially during periods of rising oil andgas prices or when project economics require a

quick return on investment. Stimulation technol-ogy is also applied as a preventive measure toavoid or delay productivity-related problems likesand production, movement of formation fines,scale deposition and organic deposits.3

These applications are especially importantoffshore where remedial well-intervention costsover the productive life of a well or reservoir canbe extremely high, often prohibitive. In manycases, stimulation is one of the main costs ofcompleting a well. Advances in three-dimensional(3D) simulation make reservoir characterizationand stimulation design more efficient, but acquir-ing input for these models continues to challengegeologists, petrophysicists and engineers whodesign drilling, completion or stimulation pro-grams. A vast amount of data and many variablesmust be considered, some of which are critical forpredicting potential production rates, reservesand recovery factors that are used to determinewell-completion or stimulation strategies.

The accuracy of petrophysical models and afew key reservoir characteristics, such as perme-ability, porosity, fluid saturation, magnitude anddirection of tectonic stresses, and other rockmechanical properties, dramatically impact fielddevelopment decisions. Many of these parame-ters, even those with significant influence on com-pletion and stimulation designs, are all too oftenbased on standard correlations, averages, esti-mates or even assumptions. Rather than rely onlimited data sets, rock catalogs, prior experienceand local practices, which may lead to inaccura-cies, miscalculations and completion inefficien-cies, optimized stimulation designs require themost complete and reliable data possible.

Advanced formation evaluation tools allow in-situ analysis and provide high-resolution, con-tinuous data acquisition across zones of interestto quantify important reservoir parameters and

1. During hydraulic fracturing treatments, fluid is injected at pressures above formation breakdown stresses to create a crack, extending in opposite directions from a well. These fracture wings—half-length—propagate in a preferred fracture plane (PFP) perpendicular to theleast rock stress. Held open by a proppant, theseconductive pathways increase effective well radius,allowing linear flow into the fracture and to the wellbore.Naturally occurring or resin-coated sand and high-strength bauxite or ceramic synthetics, sized by screeningaccording to standard U.S. mesh sieves, are used as proppants to maintain fracture conductivity.

2. Hydraulic Fracturing Survey of Industry Practices.Chicago, Illinois, USA: Gas Research Institute, 1995.

3. Economides MJ and Nolte KG: Reservoir Stimulation, 3rd ed. West Sussex, England: John Wiley & Sons Ltd, 2000.

Winter 2000/2001 43

improve predictive modeling. These direct mea-surements and knowledge from other sources,such as core, pressure and production data, andin-situ tests like DataFRAC minifracture treat-ments, provide correlations to complement andverify empirically derived values. Improved inter-pretation techniques and innovative formats forprocessing data are paving the way for bettermodels and simulations by supplying values thatwere previously unknown or difficult to determineaccurately, especially for low-permeability andheterogeneous reservoirs. In some cases, this typeof detailed data even helps identify productivezones that might otherwise be overlooked.

To select and apply the best stimulation tech-nologies and completion solutions, both operat-ing and service companies must use a full rangeof skills and expertise. Collaboration is essentialbecause operators bring reservoir knowledge andfield experience, and integrated service providerssupply the latest fit-for-purpose technology andexpertise derived from work in a variety of fieldsand basins. In addition to improved stimulationresults and enhanced well performance, pro-ducers desire cost-effective services delivered in a timely fashion. For this reason, established

data-processing services and a knowledge-management infrastructure are indispensable forreal-time data exchange between widely dis-persed office and field locations.

Comprehensive data acquisition coupled withWeb-based information technology improvesreservoir characterization to aid in stimulationoptimization. With proper knowledge manage-ment, experiences and lessons from previouswell completions are readily accessible. Infor-mation is distributed efficiently so multidisci-plinary teams are able to work together even atgreat distances, which means quicker turnaroundtimes for reservoir characterization. As a result,advances in formation evaluation and fracturingtechnology developed over the past 20 years canbe applied faster and more effectively than everbefore, often at reduced cost.

Stimulation optimization takes many forms,from slight modification of fracturing designs toapplication of new techniques or complete over-haul of field-development schemes. In Egypt forexample, development costs were reduced 42% inone field by changing from a 23-well infill-drillingprogram to 13 hydraulically fractured wells.

The potential exists to greatly improve comple-tion designs, optimize stimulation treatments andenhance production. Hydraulic fracturing of more

permeable formations, a proven technique inVenezuela and Prudhoe Bay, Alaska, USA, remainsuntried in other parts of the world. Refracturing tooptimize recovery is another stimulation applica-tion that is the subject of ongoing research.

This article focuses on stimulation optimiza-tion using the PowerSTIM stimulation and com-pletion process to develop field- or basin-specificmodels and apply customized well-completionsolutions. A proven engineering methodologyand unique Web-base workflow are the essenceof this initiative. Case histories illustrate howthis approach capitalizes on stimulation opportu-nities and improves financial results by fullyutilizing data captured while drilling, evaluating,completing and producing wells.

Finding the Right Solutions Optimized stimulations require a step change in delivery of formation evaluation, reservoircharacterization, and stimulation and completionservices. The PowerSTIM initiative provides aworkflow and tools for reengineering well com-pletions and stimulation treatments. It combinesopenhole and cased-hole reservoir characteriza-tion, drilling and measurements, completion andstimulation technologies to provide a fresh look atreservoirs. This methodology focuses on well pro-duction and field development, integrating petro-physical expertise and reservoir knowledge withdesign, execution and evaluation (above).

The PowerSTIM approach focuses on buildingfield- or basin-specific predictive models todeliver timely, customized completion recom-mendations. This approach helps teams ofexperts gather, process and evaluate as muchinformation as possible about a reservoir to opti-mize stimulation and completion designs.Experience and lessons learned are evaluatedand incorporated to close the optimization loop.

A thorough process internal to Schlumbergerand unique Web-based intranet tool combineanswers provided by wireline log data, well tests,core analysis and in-situ testing with stimulation

designs to maximize their benefits. PowerSTIMmethods generate more value than applying ser-vices and processing results separately. Takentogether, improved assessment of permeability,porosity, water saturation, mechanical rock prop-erties, stress profiles and net pay form the basisof specific solutions for a particular reservoir orfield development.

The PowerSTIM workflow can be divided intotwo stages. The first stage concentrates on a fewwells—three to five—in a field (next page, top).Through improved data acquisition, detailed anal-ysis and by working closely with the operatingcompany, operator and Schlumberger expertsdevelop a reservoir model that accurately predictskey parameters and forecasts production. Once aninitial model is established, emphasis shifts toidentifying technologies for improving well perfor-mance. Geologists, petrophysicists and reservoiror production engineers use this local model tomake completion and stimulation recommenda-tions for various stages in a field’s productive life.

In the second stage, geoscientists and engi-neers refine the reservoir model and completiondesigns to quickly deliver integrated stimulationsolutions for future wells (next page, bottom). Inmany cases, project teams deliver updatedmodels within hours of logging operations. This“in-time” approach makes the PowerSTIM

methodology an integral part of completion plan-ning rather than an afterthought. Knowledgegained from acquiring, interpreting and format-ting comprehensive data sets using the latestlogs, cores and in-situ tests, and state-of-the-artstimulation technology is key to the success ofthese projects.

One such technology is nuclear magnetic res-onance (NMR).4 Logging platforms such as theCMR Combinable Magnetic Resonance toolexcite hydrogen atoms in formations by settingup a magnetic moment, relaxing it and measuringthe time it takes atoms to realign. This NMRrelaxation time, T2, is dependent on pore size andporosity, which is related to permeability. The T2

distribution is used as an indication of porosityand permeability. Smaller pores have shorterrelaxation times; larger pores have longer relax-ation times. Laboratory analysis of core sampleshas identified a useful relaxation time for freeversus bound fluids. For typical sand-shalesequences, this cutoff is 33 msec. Pores withrelaxation time beyond this cutoff contain pro-ducible fluids.

Sonic tools like the DSI Dipole Shear SonicImager sonde excite formations with acousticwaves and measure resulting compressional and

44 Oilfield Review

> Stimulation optimization. The PowerSTIM methodology begins with a basic data set, incorporates geologic and reservoir considerations, and developsoptimized stimulation designs by acquiring and processing comprehensive data sets. Extensive post-stimulation evaluation provides feedback for continualimprovement of well completions, stimulation and field development. By comparison, standard fracturing services include only treatment design usingbasic, often limited, data and on-site execution with little post-job evaluation. Additional data acquisition, processing and interpretation are not included.

• Understand geological models• Evaluate input to geological models• Define structural and stratigraphic plays• Interpret depositional environment

• Design and analyze pressure buildup test• Analyze any flowing pressure and rate data• Make performance forecasts

• Recommend perforation interval• Design DataFRAC minifracture service• Optimize casing and tubulars• Perform reservoir simulation• Design artificial-lift or velocity string• Use appropriate new technology

• Pilot-test fracture fluids• Use local interpretation models and reservoir characterization expertise• Optimize fracture treatment

• Perform quality assurance and quality control• Ensure design criteria are met• Supervise implementation of pumping schedule• Analyze diagnostic tests

• Analyze log and core data• Analyze offset wells• Develop local interpretation models• Apply fit-for-purpose technology• Develop 3D fracture data set• Finalize prefracture reservoir characterization

• Analyze post-treatment data• Perform post-treatment production data analysis• Design and analyze post-treatment pressure transient test

Geologicconsiderations

Reservoirconsiderations

Welltesting

Well-completiondesign

PowerSTIM optimized stimulation solutions

Stimulationdesign

On-siteexecution

Post-jobevaluation

Standard fracturing services

DESIGN-EXECUTE-EVALUATE services

Winter 2000/2001 45

shear transit times.5 Transit times are convertedinto rock properties such as shear modulus,Young’s modulus and Poisson’s ratio.6 Theseinferred parameters could be further improved bycorrelation with direct measurements from coreand in-situ formation tests. Radial, or azimuthal,sonic data also can be measured to derive pre-ferred fracture plane (PFP) direction. These dataare useful for ensuring reservoir drainagethrough proper well placement. The FMI FullboreFormation MicroImager microresistivity tool isused to identify faults, natural fractures, sec-ondary porosity and borehole breakout. If presentand evident on FMI logs, borehole breakoutsoccur perpendicular to the principal stress direc-tion and help confirm DSI data.

These advanced sonic and NMR logging tech-nologies combined with core analysis, or in-situformation tests and well production testing pro-vide more accurate reservoir characterization forbetter fracture simulation and design. Designprograms like the 3D FracCADE model predicthydraulically induced fracture geometry (height,length and width) using formation parameterssuch as shear modulus, Young’s modulus,Poisson’s ratio, permeability, overburden stressand pressure. Several new tools have beendeveloped to help link formation evaluation, andpetrophysical and production analysis to stimula-tion and completion design.

For example, the ZoneAID program is a uniquezone-by-zone layering routine to identify and eval-uate individual zones in a layered formation. Thisanalysis tool is a critical link between formationevaluation data and the FracCADE program. TheFracVIZ program is a visualization tool to betterunderstand fracture geometry, fracture orientationand fracture barriers as well as their relationshipto reservoir size. Production-data analysis usingthe Production Data Fracture Interpretation Tool(PROFIT) program determines fracture half-lengthand conductivity, and effective permeability ofstimulated formations without shutting wells infor analysis.

The PSPLITR program uses production logdata to correctly allocate production to eachcompleted interval and ensure quantitative anal-ysis to reliably estimate production and evaluatefracture characteristics in commingled, multi-stage reservoirs. Well productivity is evaluatedby NODAL analysis, a technique that considersperforations, tubulars and surface facilities bytreating each pressure interface as a node withseveral variables.7 These tools and techniquescome together within the PowerSTIM environ-ment to generate innovative solutions.

> Reservoir characterization, the first stage of the PowerSTIM process.

> Optimizing stimulation and completion solutions, the second stage of the PowerSTIM process.

4. Allen D, Crary S, Freedman B, Andreani M, Klopf W,Badry R, Flaum C, Kenyon B, Kleinberg R, Gossenberg P,Horkowitz J, Logan D, Singer J and White J: “How toUse Borehole Nuclear Magnetic Resonance,” OilfieldReview 9, no. 2 (Summer 1997): 34-57.Allen D, Flaum C, Ramakrishnan TS, Bedford J, Castelijns K,Fairhurst D, Gubelin G, Heaton N, Cao Minh C, Norville MA,Seim MR, Pritchard T and Ramamoorthy R: “Trends inNMR Logging,” Oilfield Review 12, no. 3 (Autumn 2000):2-19.

5. Brie A, Endo T, Hoyle D, Codazzi D, Esmersoy C, Hsu K,Denoo S, Mueller MC, Plona T, Shenoy R and Sinha B:“New Directions in Sonic Logging,” Oilfield Review 10,no. 1 (Spring 1998): 40-55.

6. Shear modulus is an elastic material constant that is theratio of shear stress to shear strain. Young’s modulus isan elastic material constant that is the ratio of longitudinalstress to longitudinal strain. Poisson’s ratio is an elasticconstant that is the ratio of latitudinal to longitudinalstrain, or a measure of material compressibility perpen-dicular to applied stress, that can be expressed in termsof measured properties, including compressional- andshear-wave velocities.

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

• Existing data and local knowledge• Fracturing database• Petrophysical model library• Production database• Offset-well data• Client database• Basic data set

• Acquire new optimized data• Apply previous models• Calibrate data to core values• Revise current models• Optimize completion design• Apply innovative technology• Predict production performance • Perform post-job evaluation• Match production history• Refine models

• Acquire new optimized data• Apply refined models• Optimize completion design• Predict production• Verify models• Perform post-job evaluation• Refine models• Apply optimized stimulation and completion solutions

• Acquire comprehensive data set• Calibrate data to core values• Develop new models• Optimize completion design• Evaluate technical solutions• Predict production performance • Perform post-job evaluation• Match production history • Refine models

Well A Well B Well CPowerSTIM

stage 1

Field- or reservoir-specific model applied from stage 1

• Openhole logs• Petrophysical evaluation using interpretation models

• Geological, stratigraphic and structural evaluation

• Analyze well- test data

• Completion design• Completion optimization• Cementing evaluation

• Perforating design• Fracture stimulation design

• Predicted fracture properties• Production prediction

• Technical recommendation

• DataFRAC minifracture diagnostics

• Revise fracture stimulation design

• Evaluate fracture fluid flowback data

• Actual treatment parameters• Post-fracture pressure analysis • Match revised fracture properties• Analyze post-treatment production data• Complete evaluation by PowerSTIM team

• Recommendations for future development• Action plan• Update database• Full-cycle engineering and evaluation on every well• Revise models and database• Apply optimized solutions to next well

Completion Design

Summary and recommendationsExecution EvaluationReservoir

characterizationPowerSTIM stage 2

Step 1 Step 2 Step 3

Predicting Permeability Permeability impacts completion decisions anddictates optimal fracture designs. High-perme-ability formations may not need stimulation forenhanced productivity while low-permeability—tight—zones may require massive hydraulic frac-turing treatments (right). It is also important,however, to remember that stimulation of high-permeability formations is still a viable optionwhen sand production and formation fines move-ments are concerns.

Traditional methods of measuring or calculat-ing permeability do not always yield representa-tive values and may be costly, time-consuming orrisky. Core samples provide valuable informationto fill in gaps, but sample a statistically smallportion of the zone of interest. Pressure-builduptests and production-history matching supplyaverage permeability for perforated zones, but noinformation about adjacent formations or shales.In some cases, pay intervals may even have to bestimulated first just to flow and test.

To improve Lobo formation completions inSouth Texas, USA, Conoco looked to othermethods of acquiring reliable permeabilitydata.8 Accurate permeability measurementsfrom individual layers in sections with multiplepay zones are critical for predicting fracturegeometry, selecting treatment systems (prop-pants and fluids) and determining job executionparameters (pumping rates and pressures).Previously, the Lobo asset group obtained per-meability values from sidewall core and pres-sure-buildup data, relying on standard log dataand permeability correlations.

Several companies offer permeability loggingservices, so one option was to develop a methodfor estimating permeability from continuouswireline measurements. To assess this option,CMR logs were run in wells where permeabilitydata were also acquired by rotary sidewall cores.This project provided essential elements for anintegrated team of geoscientists and engineersto address stimulation optimization for a reser-voir with variable permeability by applying thePowerSTIM process.

Permeability, stress profile and net pay arekey reservoir parameters. The objectives of this evaluation were to calibrate log- and core-permeability data so a reliable local model couldbe built to predict permeability, quantify reservoirstress profiles and identify net pay, especiallypotential low-productivity zones. The modelneeded to be valid across the Lobo trend andavailable for real-time delivery. The final loggingprogram had to be more cost-effective than othermethods of acquiring permeability data.

The initial evaluation consisted of threewells. The logging program for the first wellincluded AIT Array Induction Imager Tool mea-surements, Litho-Density log, CNL CompensatedNeutron Log data and a gamma ray log for corre-lation. Additional logs were recommended toprovide permeability measurements, identify netpay and construct stress profiles for input to theFracCADE design program.

The CMR tool was used to determine pore-sizedistribution and relationship to permeability. TheECS Elemental Capture Spectroscopy tool providedclay typing and additional petrophysical analysis.Rotary cores taken with a Mechanical SidewallCore Tool (MSCT) system provided control for cali-brating CMR measurements. To be more represen-tative of in-situ conditions, core permeabilitieswere corrected for net overburden pressure.

The PowerSTIM team relied on ELANElemental Log Analysis output and DSI data toobtain mechanical rock properties and stressprofiles. To ensure good wellbore-to-formationcommunication, wells were perforated withdeep-penetrating PowerJet charges.9 Bottom-hole pressures were obtained while perforating.The best available methods were evaluated tomatch CMR permeability with core data.

Initially, permeability was computed using astandard Timur-Coates permeability equation with33-msec cutoff for T2 in sandstone, and equation

exponents based on experience in South Texas. Ingas zones, however, CMR porosity can be pes-simistic. Both CMR permeability predictions andconventional permeability-porosity relationshipsshowed mixed results when compared with corevalues (next page, top). The Lobo project team rec-ommended more core points to provide a betterpermeability baseline for correlation, but boreholeconditions prevented adequate data from beingobtained in the second well.

Additional rotary cores obtained in a thirdwell allowed testing and refinement of severalpermeability models. The initial Timur-Coatesequation did not work adequately in this welleither. Correlation between CMR-derived perme-abilities and core values was better in the lowerzone than in the top and middle zones (next page,bottom). After correction for net overburden, cor-relations using a Timur-Coates equation modifiedfor low-permeability reservoirs were encourag-ing, but still not acceptable.

46 Oilfield Review

30

Marginaleconomics

Fracturelength

Fractureconductivity

Naturalcompletion

25

20

15

10

5

00.0001

Core

por

osity

, %

Core permeability, mD

0.001 0.01 0.1 1.0 10.0 100.0 1000.0

Core point

> The impact of permeability on completion and stimulation decisions. Formations havedifferent optimal stimulation requirements. At the lowest permeabilities, economics aremarginal. For slightly higher permeabilities, fracture length becomes the crucial designparameter. At still higher permeabilities, fracture conductivity is the dominant characteristic.The highest permeability formations may not need a stimulation treatment at all.

8. Kerchner S, Kaiser B, Donovan M and Villereal R:“Development of a Continuous Permeability MeasurementUtilizing Wireline Logging Methods and the ResultingImpact on Completion Design and Post CompletionAnalysis,” paper SPE 63259, presented at the SPE AnnualTechnical Conference and Exhibition, Dallas, Texas, USA,October 1-4, 2000.

9. Behrmann L, Brooks JE, Farrant S, Fayard A,Venkitaraman A, Brown A, Michel C, Noordermeer A,Smith P and Underdown D: “Perforating Practices ThatOptimize Productivity,” Oilfield Review 12, no. 1 (Spring2000): 52-74.

Winter 2000/2001 47

Porosity

Neutron porosityvol/vol0.3 0.0

Density porosity

vol/vol0.3 0.0Total CMR porosity

Crossover CMR free fluid

vol/vol0.3 0.0

Free fluid CMR permeability, mD

Density porosity

vol/vol0.3 0.0CMR bound fluid

vol/vol0.3 0.0Pseudopermeability

equation-porosity relations0.002 20.0

StandardTimur-Coates equation0.002 20.0

Sidewall-cores corrected fornet overburden

. Initial log permeability versus Lobo formationsidewall core values. In track Three, permeabili-ties calculated from CMR Combinable MagneticResonance log data using a standard Timur-Coates equation with typical experience-basedSouth Texas exponents and a 33-msec cutoff forT2 (dashed purple curve) did not agree with side-wall core permeabilities (purple dots) in the firstwell of the Lobo reservoir characterization study.The correlation was acceptable in some zones,but not in others.

Porosity

Neutron porosity

vol/vol0.3 0.0Density porosity

vol/vol0.3 0.0Total CMR porosity

Crossover CMR free fluid

vol/vol0.3 0.0

Free fluid CMR permeability, mD

Density porosityvol/vol0.3 0.0

CMR bound fluidvol/vol0.3 0.0

Sidewall-cores corrected fornet overburden

Net-overburden correctedTimur-Coates equation0.002 20.0

StandardTimur-Coates equation0.002 20.0

Low-permeabilityTimur-Coates equation0.002 20.0

, Refining log permeability versus Lobo formationsidewall core values. More rotary sidewallcores were obtained in the third well of thisstudy to test and refine the permeability model(purple dots). The standard Timur-Coates equationdid not work in this well either (dashed purplecurve). Correcting CMR permeabilities for netoverburden improved the CMR permeabilityprediction in the lower zone, but not in the othertwo zones (green curve). A version of the Timur-Coates equation for low-permeability reservoirsmatched core permeability values in the topand middle lobes, but underestimated the lowestzone (black curve).

>

>

These new data were used to tailor the Timur-Coates equation specifically for the Lobo forma-tion. A modified CMR-permeability equation was developed using a Timur-Coates equationwith exponents provided by Schlumberger-DollResearch, Ridgefield, Connecticut, USA. Effectiveporosity from the ELAN output was used insteadof gas-corrected CMR density porosity because itwas corrected for gas, shale and invasion. Usinga 90-msec cutoff for T2 to take advantage of oil-base mud invasion characteristics in the Loboformation refined the ratio of free-fluid volume(FFV) to bound-fluid volume (BFV). Permeabilitiescalculated using this revised equation showacceptable agreement with core values through-out the well (above).

Conoco was concerned that, depending onlithology, the CMR tool would require ongoingcalibration to provide accurate reservoir per-meabilities. Laboratory measurements of T2,permeability and porosity were made on coresfrom several Lobo fields to determine if this wasthe case. To provide further validation, CMR data

from the first well were reprocessed using themodified permeability expression (below).Positive results gave Conoco confidence toremove sidewall cores from the logging programin the study area.

Net-pay calculations were also improved(next page). Standard QLA well-log analysisusing three criteria—porosity greater than 12%,water saturation less than 70% and shale vol-ume less than 50%—identified 42 ft [13 m] ofpay. An ELAN gas-resource profile using thesame criteria identified 50% more pay, or 65 ft[20 m]. The increase was a result of computingdifferent porosities and shale volumes based onadditional ECS and CMR data.

Other wells have been logged and com-pleted using CMR permeability as input. Fol-lowing reservoir characterization, Conoco andSchlumberger used more accurate and consistentpermeability, fluid saturation, net pay and stresspredictions from the Lobo-specific model todesign optimized fracture stimulations thatreduced well completion costs and significantlyincreased gas production.

48 Oilfield Review

Sidewall-cores corrected fornet overburden

Porosity

Neutron porosity

vol/vol0.3 0.0Density porosity

vol/vol0.3 0.0Total CMR porosity

Crossover CMR free fluid

Highporosity

Highporosity

Lowporosity

vol/vol0.3 0.0

Free fluid CMR permeability, mD

Density porosityvol/vol0.3 0.0

CMR bound fluidvol/vol0.3 0.0

StandardTimur-Coates equation0.002 20.0

Lobo-specificTimur-Coates equation0.002 20.0

Porosity and permeability do not correlate

> Optimizing log permeability versus Lobo formation sidewall core values. A Timur-Coates equation modified specifically for the Lobo formation using exponents providedby Schlumberger-Doll Research provided a better core-log correlation throughout thethird well of the Lobo permeability evaluation (red curve).

Porosity

Neutron porosityvol/vol0.3 0.0

Density porosity

vol/vol0.3 0.0Total CMR porosity

Crossover CMR Free fluid

vol/vol0.3 0.0

Free fluid

Density porosity

vol/vol0.3 0.0CMR bound fluid

vol/vol0.3 0.0

0.002 20.0

0.002 20.0

CMR permeability, mD

Sidewall-cores corrected fornet overburden

StandardTimur-Coates equation

Lobo-specificTimur-Coates equation

> Validating optimized log permeability versus Lobo formation sidewall core values. In Track 3,log data from the first well in this project were reprocessed using the Lobo-specific equationto validate the permeability model (red curve). Except for one core point near the bottom ofthe log section, the new expression provided a much better match between sidewall coreand CMR permeabilities (purple dots).

Winter 2000/2001 49

> Comparison of net pay analyses. Standard QLA well-log analysis using standard criteria of porosity greaterthan 12%, water saturation less than 70% and shale volume less than 50% identified 42 ft [13 m] of pay (top).With the same criteria, an ELAN gas-resource profile computed 50% more pay, or 65 ft [20 m], but with shale volumes obtained from the ECS Elemental Capture Spectroscopy tool for geochemical logging (bottom).

Permeability

Quartz

Clay-bound water

Calcite

Originalmodel shale

indicator

Chlorite

Illite

Porosity

Irreducible water

Net pay equals 65 ft using ELAN model with CMR and ECS datap.u.0.5 0

Gamma ray

API0 200 10 0.001

Permeability Effective porosity

mD 00.2 p.u.

Volume

100%0

Densityporosity

Neutronporosity

Effectiveporosity

Gas volume

Effective porosity

Net pay equals 42 ft using 12% porosity, 70% water-saturation and 50% shale-volume cutoff criteria

Gamma ray ResistivityPay

Neutron porosity

Density porosity

Shalevolume

Gasvolume

Apparent water

resistivity

Interactive Reservoir Characterization The PowerSTIM approach applies the best avail-able resources to understand wells, fields andbasins, and presents tailored recommendationsin a technically sound, easily understood format.A comprehensive log, or data montage, is built tocommunicate formation evaluation, well analy-sis, reservoir characterization and completion orstimulation designs along with predictions,results and post-treatment evaluations (right).This tangible hard copy embodies the valueinherent in an intangible methodology.

For the operator, this life-of-the-well montageis an invaluable reference for quick access towell information. The PowerSTIM montage con-tains detailed reservoir characterization andcompletion information in a continuous mode.The cover shows well location and relevant back-ground data. Inside, a section of mud log or open-hole log, core analysis and other test dataidentifies pertinent zones. A wellbore configura-tion shows the completion design and perforationrecord. Additional sections present stimulationdesigns, productivity analyses, production logsand actual production data. As a project pro-gresses, the PowerSTIM team updates the mon-tage, which can be used later to assess futurewell completions.

A Web-based intranet tool provides a frame-work for collaboration among various technicaldisciplines as well as between the operator,Schlumberger and third parties. Beginning withan entire montage presentation, team memberscan zoom in on any area of the montage for amore detailed view (next page, top). This tool ispart of Schlumberger’s internal Internet hub, aWeb-site accessible only by authorizedSchlumberger personnel. PowerSTIM teams canuse the intranet tool at any time, from any loca-tion with an Internet connection. For example,data uploaded from a field location in the MiddleEast by one team member can be accessed inminutes by a product center providing support orby another team member in an office in Houston.

The PowerSTIM intranet tool allows rapidintegration of knowledge from several disparatesources to facilitate faster, easier and betterteamwork. Project data uploaded by engineersare automatically incorporated into the PowerSTIMmontage, which is built in a fraction of the time itwould take each engineer just to handle and printvarious components separately. A montage iscompleted almost as soon as all data are col-lected. In fact, completion recommendationshave been delivered before logging trucks leavea location.

A stimulation-optimization project beginswhen the project manager, or coordinator, selects

experts from around the world who are notified bye-mail and assigned specific tasks. PowerSTIMteams should include petrophysicists, reservoirand production engineers and stimulation design-ers. This team should be involved as early as pos-sible in the drilling and completion designprocess. Optimization success is related directlyto project inception and timing. Projects that havea PowerSTIM team in place during the planningstages are usually extremely successful.

Once data are collected and analyzed, thePowerSTIM team designs a customized comple-tion based on improved reservoir characterization.Again, because of the intranet tool, these effortscan take place miles from the source of data.Completion designs and completion results canbe compiled and integrated into the montage, sothat the entire history of a stimulation treatmentis in one simple document. Running Monte Carloand other economic simulations for critical wellparameters also factors in risk.10

From initiation of a new project through post-completion, post-stimulation analysis, the intranettool essentially creates a virtual office for stimu-lation-optimization projects. Team members sepa-rated by hundreds or thousands of miles interactand exchange data efficiently and effectively todeliver completion and stimulation solutions “online” and “in time” to meet client needs. A fulland complete record is produced, just as if teammembers were working side by side.

PowerSTIM workflow using the intranet toolis an excellent example of how truly integratedsolutions are implemented by drawing on world-wide technical expertise, fit-for-purpose technol-ogy and knowledge-management systems withstate-of-the-art information technology. Quicklydelivering completion recommendations is a keycontribution of this workflow. Applications ofPowerSTIM methodology and tools for candidaterecognition and selection include: • optimize profitability in heterogeneous reservoirs

where conditions vary • improve performance of marginal fields • overcome completion problems or failures in

areas where others are succeeding • reengineer completions to maintain or exceed

past production at costs that are lower, thesame or higher, but economically justifiable.

Improving a Marginal Development Kerns Oil and Gas Inc., San Antonio, Texas, usedthe PowerSTIM methodology to optimize comple-tions in the Olmos and San Miguel tight-gasformations of South Texas, which include theCatarina S.W. and Dos Hermanos fields.11 Wellsin these fields are drilled and completed to pro-duce from either or both formations. Improvingstimulation results could justify additional devel-opment of this marginal area, but acquiring per-meability data to evaluate pay zones and designfracture treatments is difficult.

50 Oilfield Review

Well-completiondesign

Well testingReservoirconsiderations

Geologic considerations

Post-job evaluation

On-site job execution

Stimulation design

Stimulation design

Completion design

Reservoir characterization

Descriptiveheader

SummaryPost-job evaluation

On-site job execution

> PowerSTIM montage. The PowerSTIM montage is a large, paper report similar to a well log thatfully documents and completely displays relevant well data and interpretations from multiple sources.

Winter 2000/2001 51

The Olmos formation in this region consists ofindividual sands 5 to 50 ft [1.5 to 15 m] thick overan interval of several hundred feet. Many indi-vidual sands will not produce naturally andpermeability has to be assumed, leading to mis-calculations of fluid leakoff into the formation.Published permeabilities for laminated Olmossands range between 0.07 and 10 mD with avariation based on local experience of 0.04 mD toseveral millidarcies. The Olmos formation haslow compressive strength, and shales fracture aseasily as sands. Values for Young’s modulus inanother Olmos field range from 1.7 to 2.0 millionpsi [12 to 14 thousand MPa]. In the area whereKerns operates, hardness calculated from DSI logdata indicates that Young’s modulus is between1.2 and 4.8 million psi [8 to 33 thousand MPa].San Miguel sands vary from 0.1 to 33 mD.

To improve completion techniques and wellperformance, a PowerSTIM team studied severalwells and modified the completion techniques.First, high-performance UltraJet charges replacedstandard perforating charges to increase forma-tion penetration. Next, porosity and permeabilitymeasurements from CMR logs and mechanicalproperties from DSI logs were used to improvestimulation candidate selection. The team usedCMR-derived permeability to determine leakoffand design fracture stimulations. Accurate com-pressional and shear data to derive Young’s mod-ulus and Poisson’s ratio were obtained from DSIlogs. FMI microresistivity images helped addressfracturing fluid leakoff and well placement byidentifying fault planes and preferred fractureplane (PFP) orientation. Cost per well for perfo-rating increased $4,000 and stimulation costs

were $20,000 higher, but initial productionincreased from about 400 to 1000 Mscf/D[11,500 to 29,000 m3/d].

In the first well, CMR-derived permeabilitieswere compared with estimated well production(above). The zone of interest with 17% porosity and

Descriptiveheader

ReservoirCharacterization

CompletionDesign

Execution Fracturedesign

Summary

. PowerSTIM intranet tool. The entirePowerSTIM montage is built dynamicallyon the PC desktop to deliver integratedsolutions “in time” for ongoing projects.

> Nuclear magnetic resonance (NMR) permeability. The CMR CombinableMagnetic Resonance log confirmed a spontaneous potential interpretationof sand gradation fining upwards, with the most permeable zone—4 mD—at the bottom. If the well had been perforated in the top 6 ft [1.8 m] as origi-nally planned, initial production would have been low, and an unnecessarystimulation treatment would have been performed.

10. Bailey W, Couët B, Lamb F, Simpson G and Rose P:“Taking a Calculated Risk,” Oilfield Review 12, no. 3(Autumn 2000): 20-35.

11. Fairhurst DL, Marfice JP, Seim MR and Norville MA:“Completion and Fracture Modeling of Low-PermeabilityGas Sands in South Texas Enhanced by MagneticResonance and Sound Wave Technology,” paper SPE59770, presented at the SPE CERI Gas TechnologySymposium, Calgary, Alberta, Canada, April 3-5, 2000.

>

Bound fluid

MD1:120

ft

Gamma ray

API 1500

Timur-Coates

mD 100.001

CMR free fluid T2 cutoff=33 msec

ft3/ft3 00.3 30000.3

SP PermeabilityTotal CMR porosity T2 distribution

ft3/ft3 00.3

Density porosity

ft3/ft3 00.3

X150

X160

X170

X180

4 mD

Perforatedinterval

10-ohm-m resistivity calculates as 52% watersaturation. A RFT Repeat Formation Tester toolindicated 0.05-mD permeability at the top of thiszone; the bottom was not tested. A modifiedTimur-Coates permeability equation was used tocalculate 4-mD permeability from CMR data forthe lower 3 ft [0.9 m] of this section. The T2 dis-tribution supported an interpretation of sandgrains fining upward. The most permeable sectionwas in the 3 ft at the bottom of the sand. Neutronand density curves indicated significant porosity,but data were suspect because the section over-lies lignite and washed-out hole data. Late T2

decay times, however, were not due to hole con-ditions, so the lower section was tested.

This well was perforated over a 10-ft [3-m]interval indicated by gamma-ray data to be sand.Analysis using NODAL techniques for 4 mD and

3-ft net pay predicted production of 600 Mscf/D[17,000 m3/d]. Actual production was 700 Mscf/D[20,000 m3/d] without stimulation, a reasonablematch. The CMR permeability helped identify asection that might have been bypassed. The com-pletion interval and well productivity were antici-pated from the NMR-T2 distribution and resistivityprofile. Using permeability criteria, not densityporosity, resulted in an economical natural comple-tion, and a fracturing treatment was not needed.

Pore-size and permeability estimates fromCMR data were used for the second well, whichhas two thick Olmos sands located close together(below). The low-permeability sand was testedand stimulated because it met the standard 12%density porosity and 12% resistivity cutoffs, butresults were poor. This zone was plugged and theupper Olmos sand was completed successfully.

A CMR-based interpretation, which indicated thatthe lower sand has higher porosity and lower per-meability than the upper sand, would have rec-ommended that engineers intentionally bypassthe lower interval, saving stimulation and well-testing expense.

Permeability data are extremely important inall stimulation designs. Two FracCADE 3D frac-ture stimulation designs were compared for athird well (next page, top). The first design wasmodeled using permeability based on previouslocal knowledge, side-wall core estimates andpublished data for the region. The resulting frac-ture design length is less than required for opti-mal stimulation. To get the required fracturelength and width, the treatment size would haveto increased, which results in a more expensiveand, at the same time, less efficient stimulation.The possibility of a premature screenout is alsomuch higher with overdesigned jobs.12 The sec-ond design used permeability estimates from aCMR-Plus log to optimize the fracture designsand minimize potential job execution problems.

In a fourth well, DSI data were used in con-junction with CMR data. Young’s modulus fromsonic data varied from 2.5 to 4.5 million psi [17 to 31 thousand MPa] in sands and from 2.0 to 2.5 million psi [13 to 17 thousand MPa] inshales. Permeability varied from 0.003 to 0.5 mD in the sands. The fracture design for thiswell indicated a half-length of 800 ft [244 m].Interference was a concern because of an offsetwell about 1000 ft [305 m] to the northwest. Sonicdata show stress orientations even when boreholeovality, or breakout, is not apparent from FMIimages, so DSI logs were acquired to determinePFP direction. Determining fracture orientation canalso optimize well placement and gas recovery.

Directional data from DSI sonic logs and FMIimages were used to determine proper wellplacement and ensure effective reservoirdrainage. Most sands in this region have a NE-SW fracture orientation, but there is somevariation. The FMI data corroborated this direc-tion. The fracture direction was away from theoffset well, so pumping was initiated. No con-nectivity or interference was detected.

Another issue in these fields was inadequatebit performance. It was taking 20 days to drillwells, and the resulting poor borehole conditionswere adversely affecting formation evaluation and reservoir characterization. Overgauge andwashed-out holes caused log measurements to be unreliable, wasted cement and hindered zonalisolation. A synthetic rock-properties log wasdeveloped to select and run proper bits.13 A stableReed-Hycalog bit design cut drilling time to only

52 Oilfield Review

. CMR log and density log fortwo Olmos sands. An interpre-tation using density porosityalone indicated that the lowersand was more desirable.However, higher permeabilityin the upper zone correlateswith long relaxation times inthe NMR T2 distribution (top).Production results confirm thisCMR interpretation. The lowerzone was plugged after unsuc-cessful perforation and stimu-lation (bottom). The upper zonewas completed successfully.An interpretation based ondensity porosity alone indi-cated that the lower sand wasthe more desirable zone forproduction—an erroneousconclusion.

>

Bound fluid

Permeability

MD1:120

ft

Gamma ray

API0 150

CMR free fluid

ft3/ft30.3 0

Timur-Coates perm.

mD0.001 10

Total CMR porosity

ft3/ft30.3 0

Density porosity

ft3/ft30.3 0

T2 cutoff=33 msec

T2 distribution

30000.3

Resistivity less than 12 ohm-m Density porosity less than 12%

Stimulated and producing

X820

X830

X840

X850

0.1 mD

10 p.u.

Bound fluid

Permeability

MD1:120

ft

Gamma ray

API0 150

CMR free fluid

ft3/ft30.3 0

Timur-Coates perm.

mD0.001 10

Total CMR porosity

ft3/ft30.3 0

Density porosity

ft3/ft30.3 0

T2 cutoff=33 msec

30000.3

T2 distribution

Resistivity greater than 12 ohm-m Density porosity greater than 12%

Stimulated and plugged

X150

X160

X170

X180

0.01 mD

12 p.u.

Winter 2000/2001 53

10 days, improved log quality to help identify addi-tional pay and resulted in better cement jobs withless cement. This additional step ensured accuratedata for optimizing future well completions.

Well completions in this area are now more successful (right). Previously bypassed sandsthat once appeared marginal are adding signifi-cantly to total production. Data-acquisition costsincreased by $20,000 per well; perforating costsincreased $4,000; and stimulation charges rose$30,000. However, comprehensive data acquisi-tion and optimized completions have more thandoubled production rates. Average well productionincreased from about 1 MMscf/D [29,000 m3/d],and finding costs dropped by a factor of three,from $1.47 to $0.48/Mscf.

Solving Stimulation-Related Problems Fracturing the Morrow formation in southeastNew Mexico, USA, is problematic. Morrow gassands in this region are low-pressure and poten-tially water-sensitive, with permeabilities rang-ing over three orders of magnitude. The bestwells are completed naturally; those in lowerquality reservoir rock are fractured. To addresswater-sensitivity and avoid screenout while fracturing, a common practice is to pump low-viscosity foamed fluids with low proppant con-centrations that yield narrow fractures with lowconductivity. Operators in this area approachstimulation treatments cautiously and generallyaccept less than optimal results.

Most completions are planned using threegeneric guidelines: wells are completed withoutfracturing if zones produce at economical rates;fracture stimulation is a last resort if productionfalls below the economic limit; and aqueousfluid systems are avoided because of a water-sensitive formation. Although not universal,these approaches represent the thinking ofmany operators involved in Morrow develop-ment during the past 20 years.

Early attempts to fracture the Morrow withwater-base systems were marginally successful.Studies suggested that poor results were due towater-sensitive clays or capillary-pressureeffects that reduce reservoir permeability as aresult of exposure to fracturing fluids. Low reser-voir pressure exacerbates the latter. These issues

12. A screenout is caused by proppant bridging in the frac-ture near a wellbore, which halts fluid entry and fracturepropagation. If a screenout occurs early in a treatment,pumping pressure may become too high and the job maybe terminated before an optimal fracture can be created.

13. Besson A, Burr B, Dillard S, Drake E, Ivie B, Ivie C, Smith Rand Watson G: “On the Cutting Edge,” Oilfield Review 12,no. 3 (Autumn 2000): 36-57.

> More accurate fracture modeling. Permeability estimates used in the first FracCADEdesign (top) are higher than the CMR log-derived permeability that was used for thesecond design (bottom), resulting in a shorter fracture half-length—600 ft [183 m] versus 800 ft [244 m]. Other rock properties were held constant for both cases.

> Completion and field development optimization. During a 35-well drilling program, productionperformance was optimized through high-performance perforating, improved formation evalua-tion and log interpretation with advanced logging technologies, and proper fracture design anddrilling-bit selection.

35-Well Development Drilling Program (Groups of 5-Wells)

High-performanceperforating and improved

log interpretation

PowerSTIM reservoir characterizationand optimized fracture stimulations

Optimized bits

1998 1999 2000

Gas

rate

, Msc

f/D

3000

2500

2000

1500

1000

500

< 0.0 lbm/ft2 0.0-0.10.1-0.20.2-0.30.3-0.40.4-0.50.5-0.60.6-0.70.7-0.8> 0.8

X4800

X4900

X5000

X5100

X5200

X4800

X5000

Dept

h, ft

Dept

h, ft

X5100

X5200

X4900

3600 4400Stress, psi Fracture width at wellbore, in. Fracture half-length, ft

800-ft half-length

Greatest concentration of proppant

Greatest concentration of proppant

0.1 0 0.1 0 400 800 1200

3600 4400Stress, psi Fracture width at wellbore, in. Fracture half-length, ft

0.1 0 0.1 0 400 800 1200

< 0.0 lbm/ft2 0.0-0.10.1-0.20.2-0.30.3-0.40.4-0.50.5-0.60.6-0.70.7-0.8> 0.8

600-ft half-length

Proppantconcentration

Proppantconcentration

were addressed by pumping treatments ener-gized with nitrogen [N2] or carbon dioxide [CO2],and using methanol in fracturing fluids. However,stimulation results with these foamed systemshave been inconsistent.

In higher permeability zones where near-wellbore damage is bypassed, small fracturingtreatments using foams are successful, but inlower permeability zones where fracture lengthis critical to stimulation success, these systemsdo not provide consistent economical results.These treatments address water-sensitivity, butlow viscosity, high friction pressure and chemicalrequirements increase screenouts and cost. Earlyscreenout and low proppant concentrations leavewells producing at less than full potential.

Fracture-treatment designs that develop ade-quate hydraulic width and transport higher con-centrations and volumes of proppant were neededto maximize production, but this required reliablevalues for reservoir parameters and rock proper-ties. With a comprehensive understanding of thereservoir, stimulation specialists can design fluidsand proppants to create hydraulic fractures thatdeliver high conductivity and optimal results.

Louis Dreyfus Natural Gas Inc. drilled theETA-4 well in March 2000. Pressure data werenot available, but a bottomhole pressure of2000 psi [13.8 MPa] was measured in an offsetMorrow well. Wireline logs identified a homoge-neous, high-quality, 10-ft [3-m] Morrow zone withabout 14% porosity and 20% water saturation.Rotary sidewall cores verified the log interpreta-tion. A zone of this quality should produce natu-rally, but high permeability and low pressuremake the interval susceptible to drilling damage.Significant separation between resistivity curvesconfirmed deep invasion, so to overcome near-wellbore damage, the operator wanted to designa fracture stimulation in advance.14

Reservoir quality in the ETA-3 well, com-pleted two months earlier, was similar, but withhalf as much pay. This well was perforated andfracture stimulated down 5-in. casing with CO2

foam and high-strength, man-made ceramicproppant. Surface treating pressure was 5000 psi[34 MPa]; maximum proppant concentration was4 ppa; and there were indications of possiblescreenout near the end of the job. Post-stimula-tion production stabilized at 1.7 MMscf/D

[49,000 m3/d] and 500-psi [3.4-MPa] flowingtubing pressure (FTP) at surface.

Because reservoir quality was equivalent andpay was twice as thick, the operator expectedETA-4 to be an excellent well. However, produc-tion after perforating was only 500 Mscf/D[14,000 m3/d] with 220-psi [1.5-MPa] flowing cas-ing pressure (FCP), which was equivalent to anextremely damaged completion with a positive 45skin. The next step was to determine optimal frac-ture length using actual reservoir parameters (left).

To take full advantage of this quality reser-voir, the operator wanted to design a more con-ductive fracture using higher than conventionalproppant concentrations. However, because theoffset fracture treatment indicated a possiblescreenout at 4 ppa, this would not be easy.Premature screenout limits production rates afterfracturing and are common in the Morrow for-mation. Stimulation engineers consider near-wellbore tortuosity to be a factor that needs tobe addressed to minimize the likelihood of ascreenout (below). Properly oriented perforationsmitigate tortuosity effects. The maximum stressdirection, or preferred fracture plane (PFP), is per-pendicular to borehole breakout as indicated byFMI log data and oriented NW to SE in the ETA-4 well. This information was used to orient per-forating guns in the direction of maximumformation stress using a Wireline OrientedPerforating Tool (WOPT) system.

Higher proppant concentration—6 versus 4 ppa—to increase fracture width was possiblebecause oriented perforating reduces the risk ofa premature screenout caused by near-wellboretortuosity. At 6 ppa, the FracCADE programshows a fracture half-length of 300 ft [91 m] anda width of 0.15 in. [3.8 mm], twice as wide as

54 Oilfield Review

> Formation damage, or skin, effects and optimal fracture length. A NODALanalysis indicates that with zero skin, the ETA-4 well should produce 3.3 MMscf/D[94,500 m3/d] at 500-psi [3.4 MPa] flowing tubing pressure (top). An optimizedfracture would reduce skin to –4 and result in production above 5.5 MMscf/D[157,500 m3/d]. Maximum net present value (NPV) is achieved with a 200-ft[61-m] fracture half-length (bottom).

Pinch points

Preferred fractureplane (PFP) Maximum-stress

direction

Minimum-stressdirection

> Tortuosity and fracture initiation. When a frac-ture initiates randomly, there is a good chancethat it will propagate into the reservoir along aplane other than the maximum stress direction, orpreferred fracture plane (PFP). The fracture thenturns to follow the path of least resistance andalign with the PFP, causing pinch points that mayact as a choke and lead to premature screenout.

2000

1500

1000

500

01000

Pres

sure

, psi

Net

pre

sent

val

ue (N

PV),

$100

0

2000 3000 4000Gas rate, Mscf/D

5000 6000 70000

Skin = 0 Skin = 45 Skin = 4 Outflow

710

720

730

740

750

760

0 100Fracture half-length, ft

200 300 400 500

FracCADE output

NODAL output

Production time =1 year

Winter 2000/2001 55

a 4-ppa design (right). This treatment appears tobe overdesigned, but local experience suggeststhat to realize an effective 200-ft [60-m] conduc-tive fracture, a 300-foot design target is notunreasonable considering the potential for frac-ture conductivity damage after fracture closure and production begins.

Treatment pressures highlight the impact oforiented perforations on job execution (belowright). Pump rates for the two stimulation treat-ments are identical at 30 bbl/min [4.7 m3/min],but the conventional fracture stimulation showsa treating pressure of 5000 psi [34 MPa], whilepressures with oriented perforation rangebetween 3000 and 4000 psi [20 and 27 MPa].Another important indicator is pressure responseafter pumping stops. On the conventional job, ittakes 15 minutes for pressure to drop below3000 psi, suggesting that net pressure wasincreasing and this job was close to screenout.For the oriented fracture, pressure stabilizesalmost immediately, suggesting higher proppantconcentrations could be placed. Advances in fluidsystems and optimized fracture designs make itpossible to use either foam or water-base sys-tems to stimulate the Morrow formation.

Early production history for ETA-4 indicated a successful stimulation. Post-fracture produc-tion was 3.5 MMscf/D [1 million m3/d] with1280-psi [9-MPa] FTP compared with 500 Mscf/Dwith a flowing casing pressure of 220 psi beforestimulation. Because the goal was to bypassdrilling damage, skin is a good indicator of frac-turing success. Prestimulation production of 500 Mscf/D suggests a skin of 45, while a post-stimulation rate of 3.5 MMscf/D [99,000 m3/d]suggests that skin was reduced to -4.

Analysis shows that with 4 ppa maximumproppant concentration and 0.06-in. [1.5-mm]fracture width, the ETA-4 well would produce 2.2 MMscf/D [63,000 m3/d] at 1280-psi FTP. Iffracture width is increased to 0.15 in., productionrises to 3 MMscf/D [85,000 m3/d] with 1280-psiFTP. The well actually produced more, suggestingcreation of a slightly wider fracture. Orientedperforating allowed proppant concentration andfracture width to be increased, eliminated pre-mature screenout and the need to clean out wellsafter fracturing, and resulted in an additional 1.3 MMscf/D [34,000 m3/d]. The payout for incre-mental perforating costs was only three days.

14. Logan WD, Gordon JE, Mathis R, Castillo J and McNallyAC: “Improving the Success of Morrow Stimulations theOld-Fashioned Way,” paper SPE 67206, presented at theSPE Production Operations Symposium, Oklahoma City,Oklahoma, USA, March 24-27, 2001.

> Fracture designs for the ETA-4 well. Although fracture half-length and height are similar,fracture width using 4 ppa (top), as on offset well ETA-3, is half of that for 6 ppa (bottom).

7000

8000

5000

6000

3000

4000

1000

0

2000

35

40

25

30

15

20

5

0

10

84 87

Pres

sure

, psi

Pum

p ra

te, b

bl/m

in

90 93 97 100 103Pump time, min

106 109 113 116 119 122 125 129

Treating pressure—conventionalTreating pressure—oriented

Conventional—4 ppa

Oriented—6 ppa

Pump rateETA-4 well,oriented

Pump rateETA-3 well,conventional

> Comparison of conventional and oriented fracture treatments. The biggestdifference is in surface treating pressure. While proppant concentrationsteps from 1 to 4 ppa on the ETA-3 well and from 1 to 6 ppa on the ETA-4 well,treating pressures are significantly lower on ETA-4 than on ETA-3 as a resultof orienting the perforations with the maximum stress direction, or preferredfracture plane.

Dept

h, ft

Stress, 1000 psi Fracture width at wellbore, in. Fracture half-length, ft

Proppantconcentration

8 9 10 0.1 0 0 200 400 6000.1

< 0.0 lbm/ft2

0.0-0.10.1-0.2 0.2-0.3 0.3-0.4 0.4-0.5 0.5-0.6 0.6-0.7 0.7-0.8 > 0.8

X1900

X2000

X2100

X2200

X2300

X2400

Greatest concentrationof proppant

300-ft half-length

Dept

h, ft

0.1 0.10 0 200 400 600

< 0.0 lbm/ft2

0.0-0.1 0.1-0.2 0.2-0.3 0.3-0.4 0.4-0.5 0.5-0.6 0.6-0.7 0.7-0.8 > 0.8

X1900

X2000

X2100

X2200

X2300

X24008 9 10

Stress, 1000 psi Fracture width at wellbore, in. Fracture half-length, ft

Proppantconcentration

Greatest concentrationof proppant

400-ft half-length

Reengineering CompletionsUltra Petroleum, Inc., one of the most activeoperators in the prolific Jonah gas field ofSublette county, Wyoming, USA, wanted toreduce completion time and costs, and increaseproduction. This field is a fault trap within alarger gas accumulation in the Green River basin.Production comes from the Upper CretaceousLance formation, a thick sequence of stacked,interbedded channel sands, overbank silt-stones and floodplain shales from about 8900 to13,500 ft [2713 to 4111 m]. The Lance formationis gas-charged, but noncommercial throughoutmuch of the basin. Shallow overpressure makesgas production economical in the Jonah field.

Low permeability and multiple pay zonesacross large sections make it difficult to com-plete wells economically. The gross horizon is2800 to 3600 ft [853 to 1097 m] thick with morethan 100 separate sands that range from 2 to 30 ft [0.6 to 9 m] thick. Pay intervals consist ofboth individual 10- to 15-ft [3- to 4.5-m] zonesand stacked channel sequences more than 200 ftthick. Porosity is between 6 and 12%, and per-meability ranges from 0.001 to 0.5 mD. Each sandrequires stimulation to produce viable rates.

The completion interval is grouped intostages with 50 to 200 ft [15 to 61 m] of gross paythat are fractured separately. If there are toomany stages, costs increase significantly andreturn on investment is reduced. With too fewstages, some sands are not stimulated ade-quately and production is compromised. The fun-damental problems faced by engineers aredetermining which sands to complete and whichto skip; how many sands to include and how togroup them in stages; and how to isolate betweenstages after fracturing.

Historically, using limited-entry diversion overlonger intervals to control costs minimized thenumber of fracturing treatments. Stages rangedfrom 100 to 450 ft [30 to 137 m] with 20 to 28 per-forations per stage. After an interval was stimu-lated, the well was flowed to clean up andrecover fracturing fluids, and to test productivity.One to two weeks later, fractured intervals wereisolated with mechanical plugs—retrievable ordrillable—or sand plugs, and the next intervalwas perforated and fractured. This process con-tinued until the well was completed. Typically,three to six intervals were fractured per well overabout five weeks. These traditional completionsoften cost as much as drilling the well, take a lotof time and yield disappointing results.

The PowerSTIM methodology was applied tothis complex reservoir with impressive results

56 Oilfield Review

XX000

XX100

XX900

XX800

MDft

Inadequatelystimulated pay

intervals

Little to noproductioncontribution

Flow rateB/D

GRPass 2

API0 200

0 lbm/ft2 6

GRFracture width, in.

TotalScandium

TotalIridium

Cased-hole

Sandconcentration

Formation

Scandium

Iridium

Formation

Scandium ScandiumIridium

Iridium

> Diversion in original completions: radioactive tracers and production logs. With limited-entry tech-niques, some intervals were found to be unstimulated. In this example, six pay sands over 300 ft [91 m]of gross interval were fractured through 24 perforations. The lower zones are inadequately stimulated(left) and contribute little or no production. If an interval did not take fluid at the beginning of a treatment,perforation erosion in other sands eliminated the backpressure necessary for diversion.

3 of 4 sand layerstreated with

proppant fracture

No productioncontribution fromstimulated sand

layers

XX400

XX500

XX300

XX200

Flow rateB/D

GRPass 2

MDft

API0 200

0 lbm/ft2 6

GRFracture width, in.

TotalScandium

TotalIridium

Cased-hole

Sandconcentration

Formation

Scandium

Iridium

Formation

Scandium ScandiumIridium

Iridium

>Why sands at the top of perforated and fractured intervals fail to produce. Production logging quantifiedcompletion efficiencies throughout the field, and when combined with tracer surveys, provided clues tothe reason some zones did not produce. In original completions, tracer surveys indicated proppant inthe upper sands of most intervals where production logs often show no production. Analysis indicatedthat proppant placed during fracturing was overdisplaced in the formation during the setting of sandplugs for the next stage.

Winter 2000/2001 57

and far-reaching impact on completion success.Once again, the key was to develop a model toestimate permeability, rock properties and indi-vidual sand-layer productivity.

The first phase of the project incorporatedwork done by Amoco, now BP, and Schlumbergerto relate core and transient pressure tests to logresponses.15 After traditional completion tech-niques developed by other operators were ana-lyzed, several problems were apparent.Radioactive tracer surveys showed that manytargeted sands were not stimulated, and produc-tion logs indicated only about 60 to 70% of paysands were producing gas (previous page, top).

Sand plugs were difficult to place and oftenended up missing or too low, resulting in costlyprocedures to reset them. Production logsrevealed that there was no production from manyupper sands in a fractured interval, but radioac-tive tracers show proppant was placed in thesezones (previous page, bottom). Many wells hadthe same problem, indicating that proppant fromthe near-wellbore region may be displaced whensand plugs are set before fracturing subsequentstages. Tracer surveys also indicate fracturecontainment, but net-pressure plots show con-siderable fracture-height growth. Even withlimited-entry perforating for diversion, there may be connectivity between perforated andunperforated sands.16

The Jonah reservoir lacked complete charac-terization. Fracturing jobs rarely screened out andsand plugs for zonal isolation frequently end updisplaced into the formation, suggesting thepotential to optimize stimulation designs byincreasing treatment size. To make completionand stimulation decisions, the PowerSTIM teamneeded to evaluate key parameters, includingstress gradients for fracturing model geometryand proppant selection, Young’s modulus forfracture width, leakoff for fluid optimization, and reservoir pressure for staging strategy andfluid requirements.

The greatest challenge was deciding how toacquire additional data without compromisingprofitability. This was accomplished by carefullyplanning strategic logging programs, minifrac-ture treatments and pressure-transient analysis.Fracture gradient, Young’s modulus and leakoffparameters for fracture fluids were determinedfrom minifracture treatments using the DataFRACservice (above). Dipole sonic logs were used tobuild near-wellbore stress models, calibratestress profiles for sand-shale sequences anddetermine preferred fracture direction. Thesedata confirmed stress values from minifractures.

Fracturing models using more reliable stressmeasurement in sands and bounding layers,coupled with better Young’s modulus values,

provided improved fracture height and widthestimates. Cores were analyzed to understandfluid compatibility and verify rock mechanicalproperties. Early geologic interpretations assumedthere were natural fractures with high fluid leakoffthroughout the field, but image logs did not showsignificant natural fractures in the heart of thefield. Lower than expected fluid leakoff fromDataFRAC analysis and absence of natural frac-tures allowed pad volumes to be decreased, whichreduced costs.17

The geological complexity of this field requiresa method for completing multiple horizons in a single day without sand plugs or mechanicalisolation. This approach allows shorter intervalsto be stimulated, increases production efficiencyand improves development economics. Diversionwith limited-entry techniques and sand plugsresults in poor completion efficiencies. Mechan-ical isolation with bridge plugs or packers wascomplicated, costly and risky to retrieve by eitherconventional workover or coiled tubing.

To stimulate wells more effectively, theoperator decided to fracture smaller verticalintervals and perform multiple treatments in asingle day. The ideal diversion technique wouldallow cleanup of fracture intervals withoutneeding to wash out sand or retrieve packers.By designing a tip-screenout, net pressure

15. Christensen CJ, Cox DL, Lake EA, Dolan VB, Crisler JDand Lima JP: “Optimized Completion Techniques inJonah,” paper SPE 62853, prepared for presentation atthe SPE/AAPG Western Regional Meeting, Long Beach,California, USA, June 19-23, 2000.

16. Limited entry involves low shot densities—1 shot perfoot or less—across one or more zones with differentrock stresses and permeability to ensure uniform acid orproppant placement by creating backpressure and limit-ing pressure differentials between perforated intervals.

> Jonah field minifractures. The DataFRAC service is used to determine fluid leakoff coefficients,fracture-closure pressure and height growth, and Young’s modulus. A zone with good permeabilityand barriers away from adjacent intervals is selected to allow mechanical isolation. Pressure draw-down and buildup tests determine permeability, pressure and skin, or damage. A downhole memorygauge is used to acquire bottomhole pressures. A series of step-rate pump and flowback tests with2% potassium chloride water measures fracture-closure pressure. A small volume of fracturing fluid is pumped to determine its leakoff coefficient. Resulting net pressure is used to determine fractureheight and Young’s modulus.

The objective is to maximize stimulation efficiency andresults without mechanical isolation like drillable bridgeplugs and retrievable packers. Rubber ball sealers canbe used to seal open perforations and isolate intervalsonce they are stimulated so that the next interval can be treated. Because perforations must seal completely,hole diameter and uniformity are important.

17. The pad stage of a hydraulic fracturing treatment doesnot contain proppant, and is the volume of fluid thatcreates and propagates the fracture.

Determiningclosure values

Determiningleakoff coefficient

Redesignedfracture

Calculating Young’smodulus and

height growth

00 50 100 150 200 250

1

2

3

4

5

6

7

Trea

ting

and

botto

mho

le p

ress

ure

(BHP

) pre

ssur

es, 1

000

psi

Treatment time, min

8

9

10

0

15

30

45

60

75

90

105

Prop

pant

con

cent

ratio

n, p

paSl

urry

and

flow

back

rate

, bbl

/min120

135

150

Treating pressure, psi Calculated BHP, psi Slurry rate, bbl/min Proppant concentration, ppa Flowback rate, gal/min

generated during stimulation is used to divertsubsequent fracturing treatments into the nextinterval (below).18

After fracturing, wells are flowed to recoverat least one casing volume of fluid and allow thefracture to close. The next interval is then perfo-rated and fractured. This process is repeateduntil the entire producing horizon is stimulated.As many as 11 fracture stages have been per-formed in 36 hours, reducing the time required tocompletely stimulate a well from five weeks toless than four days, and increasing producing payto more than 90%.

In previous wells where limited entry wasused, an entire wellbore might have only 120 per-forations. With the new completion technique, a single fracture interval can have 120 per-forations. A well may have 1200 over the entire

interval to reduce the risk of leaving pay unstim-ulated. The new completion techniques increasethe maximum number of stages from 5 to 12intervals per well. Fracturing designs includedhigh proppant concentrations at the end of atreatment to maintain created width and maxi-mize net pressure after fractures close.

Tubulars were the final aspect to be analyzed.Initially, well designs used 4- or 5-in. casing toaccommodate high pump rates for fracturinglarge intervals. Because Jonah wells typicallyproduce water and condensate, unloading fluidsis critical to maintain production. After cleanup, 23⁄8- or 27⁄8-in. tubing was run into wells underpressure using a snubbing unit. With shortertreatment intervals and improved fluids, fracturesnow can be placed effectively at lower pump rates,making 3-in. tubing feasible for casing (above).This tubular configuration defers tubing installa-tion by several years and eliminates productionlimitations associated with small tubular sizes.

58 Oilfield Review

XX400

XX500

XX600

MDft

XX300

XX200

GRPass 2

API0 200 0 lbm/ft2 6

GR

Fracturewidth, in.

TotalScandium

TotalIridium

Cased-hole

Sandconcentration Formation

Scandium

Iridium

Formation

Scandium Scandium Iridium

Iridium

> Improved diversion. New completions, with 40% less gross interval perstage, allow pay sands to be treated more effectively. This radioactivetracer survey indicates successful placement of two separate fracturesless than 100 ft [30 m] apart without sand plugs or retrievable packersand drillable bridge plugs for positive isolation.

4 1/2

2 3/8

2 3/8 in 4 1/2

3 1/2

1 1/2

1 1/2 in 3 1/2

4.01.995 3.21 2.992 1.31 2.588

2850 675

1825 1525

290 1150

1287 1232 1267 1262 1200 1251

Tubular size, in. Equivalent insidediameter (ID), in. Rate, Mscf/D Bottomhole flowing

pressure (BHFP), psi

> Comparison of minimum gas rates to keep wellbore fluids unloaded. Based on a con-densate yield of 10 bbl/MMscf and a water yield of 3 bbl/MMscf with 700-psi [4.8-MPa]wellbore pressure, a well with 3-in. inside diameter (ID) casing continues to produce atabout half the rate of 4-in. ID casing before loading with fluid.

18. In standard fracturing, the fracture tip is the final areathat is packed with proppant. A tip-screenout designcauses proppant to pack, or bridge, near the end of thefracture early in a treatment. As additional proppant-laden fluid is pumped, the fracture balloons because it can no longer propagate deeper into the formation. This technique creates a wider, more conductive pathwayas proppant is packed back toward the wellbore.

Winter 2000/2001 59

With this background in mind, a specificPowerSTIM project was undertaken to furtherreduce completion time and costs in UltraPetroleum’s Jonah wells without jeopardizingproduction. Four reservoir rock types were identi-fied, and correlations were developed to computepermeability with CMR logs on key wells,Platform Express logs on infill wells and RSTmeasurements in wells where adverse hole con-ditions prevent openhole data acquisition.Similar correlations were developed to computerock mechanical properties—stress, Poisson’sratio and Young’s modulus. A unique zone-by-zone layering routine was developed to identifyand evaluate each of the hundreds of distinctlayers in the Lance formation. This routine even-tually evolved into the ZoneAID program.

A method was then developed to combineformation evaluation and stimulation designresults to predict production. This powerful toolallows a PowerSTIM team to quickly evaluatemultiple completion scenarios and determinewhich combination results in maximum produc-tion for the lowest cost. Currently, the time from

receiving log and well data to generating ratepredictions for all sands is only about four hours.Schlumberger delivers a PowerSTIM montage,including individual stimulation designs, produc-tion forecasts and economic evaluations for asmany as 17 fracturing stages, to Ultra Petroleumwithin forty-eight hours. Efforts are continuing toreduce this turnaround time even more with thehelp of the intranet tool.

However, the process does not stop here. Inperhaps the most important step, key wells areroutinely evaluated by running production logsafter three to six months to measure contribu-tions from each sand, make sure each sand wasadequately stimulated and assess productionpredictions. Production logs and production his-tories are evaluated using PSPLTR and PROFITsoftware to make sure fracture-model conductiv-ity and geometry are actually attained. By con-stantly evaluating and refining the optimizationprocess, the PowerSTIM methodology canachieve remarkable accuracy.

Like many of today’s lean and aggressiveenergy companies, Ultra Petroleum relies on newtechnology, innovative solutions and collabora-tive working relationships for technical servicesand support, innovative technologies and newintegrated solutions. New completion techniquesbased on extensive data collection reduced timeto first production and completion costs whileincreasing production and recovery factors.

Eliminating sand plugs and other positive iso-lation between fractured intervals, and extensiveflowback periods after each treatment savedmoney and almost four weeks of completion time(below). Reduced pad volumes, improved fluidand proppant selection, and optimized tubulardesign decreased costs. Overall completion costswere cut by 50%. Finding costs decreased from$0.45 to $0.23/Mscf.

When data are normalized for permeabilityand thickness, production from new completionsis 8% more than from original completions and30% higher than wells of other area operators,primarily from improved completion efficiencies.

SpudDrillingLoggingCementing

PerforatingFracturingFlowbackSetting plugs

TestingRetrieving plugsWell shut-inSales

August 1997

May 1998

February 1999

66 days

52 days

39 days

> Continual completion improvements. Over an 18-month period, time to first productionwas reduced by about 27 days, or four weeks, primarily by completing Jonah field wellswithout sand plugs or other forms of mechanical isolation.

Data also indicated an increase in estimated ulti-mate recovery (EUR) for new completions (above).

Completion Optimization Understanding reservoir characteristics across payzones in a well, an entire field and within a basinresults in optimized stimulation treatments andcompletion techniques that reduce costs, maximizeproduction and increase hydrocarbon recovery. ThePowerSTIM initiative uses an integrated approachto develop the required models for generatingtechnical solutions or well-completion strategiesthat are transportable from field to field and com-pany to company.

The positive impact and established record ofstimulation optimization in the mature onshorefield developments of North America are helpingto extend acceptance of the PowerSTIM method-ology into other areas both onshore and offshore,including regions like the Middle East.

A joint Saudi Aramco and Schlumberger pro-ject is now under way in the Hawiyah field, SaudiArabia, to eliminate sand production and maxi-mize production from the reservoir to meet thegas-deliverability target for this field. The projectinvolves stimulation optimization for a group of10 wells. Rather than undertake this effort inter-nally, Saudi Aramco chose to utilize thePowerSTIM approach and form a team of expertsto develop stimulation and completion solutions.

The PowerSTIM project manager is a SaudiAramco representative. A Schlumberger projectcoordinator heads joint technical and operational

teams. The technical team is made up of petro-phycists, geologists, reservoir engineers and stim-ulation engineers from each company who workwith the Saudi Aramco engineers that areassigned to designated wells. The operationalteam comprises Schlumberger field managersfrom wireline, well testing, cementing and stim-ulation, and coiled-tubing service segments whowork closely with Saudi Aramco site foremenand senior engineers.

The first stage of this project—model devel-opment—was completed in early 2001. A com-prehensive data set was developed to improvecompletion strategies and designs. Mechanicalrock properties, hydrocarbon resource profilesand sand-prediction models were either devel-oped or improved. In addition to optimal fracturedesign through integration of all available basin,field, reservoir and well data, this PowerSTIMproject is generating and documenting best prac-tices. Screenless completion guidelines wereadapted and distributed for use along with sand-control guidelines for well flowback.

Petrophysical models were initially applied tofour wells. In February 2001, the first well wasstimulated using recommendations based onreservoir and completion models developed bythe PowerSTIM team. Early results were extremelyencouraging. The project schedule calls for theremaining wells to be stimulated during the firsthalf of 2001.

Collaboration on this project proved to beextremely beneficial, particularly between SaudiAramco and Schlumberger, with interaction andworkflow continuing to improve. Staff from eachcompany appreciate the ability to contributeknowledge, experience and ideas to improvestimulation treatments and the well-completionprocess. Both companies benefit from thereduced engineering cycle time that is the resultof expediting the learning process, emphasizingadded value and targeting incremental produc-tion potential.

Stimulations based on estimates or on aver-age reservoir properties may result in hydraulicfractures of insufficient length and width withexcessive vertical height. Innovative methods toreliably establish the key parameters required forreservoir characterization, modeling and designprogram input overcome traditional limitationsinherent in acquiring these data.

Optimized stimulation treatments use continu-ous wireline measurements from advanced well-logging technology, core analysis, in-situ testing,and better data management, processing andinterpretation coupled with fit-for-purpose frac-turing technologies to ensure more contained,conductive conduits deeper into formations.

The PowerSTIM initiative is about full-cycleengineering, quality data and in-time delivery ofcustomized solutions. Procedures based on dataacquired throughout a field, region-wide experi-ence and application of careful reservoir evalua-tion are having beneficial effects on fielddevelopment. In addition, better evaluation offormation characteristics allows fracturing fluids,proppants and volumes to be optimized.Schlumberger is positioned to provide datameasurement, integration, formatting and pre-sentation, as well as interpretation expertise,technical design and evaluation, operations qualitycontrol and global support.

The way solutions are developed and commu-nicated within Schlumberger and externally tooperators is changing as the industry moves awayfrom static, or flat, documents and reports. Real-time information processing, data evaluation andlife-of-the well reports are becoming as importantas the answers and solutions they generate. Thelatest information technology (IT) systems andknowledge-management technologies are provid-ing us with methodologies and Web-based toolslike the PowerSTIM intranet tool and montage formaking informed decisions in a collaborative, vir-tual work environment. Through Internet meetingsand videoconferences, regional data houses andvisualization centers, we can work togetherwithout being in the same office. —MET

60 Oilfield Review

16,000

14,000

12,000

Estim

ated

ulti

mat

e re

cove

ry (E

UR),

MM

scf

Relative completion costs

10,000

8000

6000

4000

2000

00.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

Offset control wells

Optimizedcompletion

model

Previouscompletion

model

> Recovery versus relative completion costs. Compared with wells completed conventionally,the average Ultra Petroleum, Inc. well in the Jonah field completed using the new techniqueshas a significant increase in estimated ultimate recovery (EUR) per completion dollar spent.