Successful Implementation of Horizontal Multistage Fracturing to

Spring 2013

THE SAUDI ARAMCO JOURNAL OF TECHNOLOGYA quarterly publication of the Saudi Arabian Oil CompanyJournal of Technology

Saudi Aramco

Successful Implementation of Horizontal Multistage Fracturing to EnhanceGas Production in Heterogeneous and Tight Gas Condensate Reservoirs: Case Studies see page 2

Selecting Optimal Fracture Fluids, Breaker System, and Proppant Type forSuccessful Hydraulic Fracturing and Enhanced Gas Production – Case Studiessee page 22

62884araD1R2_ASC027 3/17/13 3:46 PM Page 1

creo

On the Cover

A photograph of a reservoir core and thin section and Scanning

Electron Microscope (SEM) photomicrographs of dolomitic and

anhydritic limestone core. The grains appear to have undergone a

minor to moderate amount of compaction as evidenced by the

numerous point and long grain contacts and fewer concavo-convex

grain contacts and stylolites. The dolostone are poorly sorted and the

original fabric of the remaining dolostone has been partially to almost

completely obscured by the dolomitization process. Cementation by

calcite and anhydrite is the main cause of reduction of the primary

pore volume. Later dissolution of grain and dolomitization generated

secondary pores that make up much of the total porosity. Placing a long horizontal wellbore toward the minimum stress

direction plays a major role in the success and effectiveness offracturing — to enhance and sustain productivity from tight gasreservoirs. The Gas Reservoir Management Team has been successfullyexploiting nonassociated gas reservoirs and meeting the Kingdom’s gasdemand by using this process. Pictured here discussing the mosteffective drilling and completion plans for nonassociated gas wells(from left to right) is Ali Habbtar, Adnan Al-Kanaan, Dr. Zillur Rahimand Dr. Hamoud Al-Anazi from the Gas Reservoir ManagementDepartment.

The Saudi Aramco Journal of Technology ispublished quarterly by the Saudi Arabian OilCompany, Dhahran, Saudi Arabia, to providethe company’s scientific and engineeringcommunities a forum for the exchange ofideas through the presentation of technicalinformation aimed at advancing knowledgein the hydrocarbon industry.

Complete issues of the Journal in PDF formatare available on the Internet at:http://www.saudiaramco.com(click on “publications”).

SUBSCRIPTIONS

Send individual subscription orders, addresschanges (see page 81) and related questions to:

Saudi Aramco Public Relations DepartmentJOT DistributionBox 5000Dhahran 31311, Saudi ArabiaFax: +966/3-873-6478Website: www.saudiaramco.com

EDITORIAL ADVISORS

Zuhair A. Al-HussainVice President, Southern Area Oil Operations

Abdulaziz M. JudaimiVice President, Chemicals

Ziyad M. ShihaVice President, Power Systems

Abdullah M. Al-GhamdiGeneral Manager, Northern Area Gas Operations

Salahaddin H. DardeerManager, Riyadh Refinery

EDITORIAL ADVISORS (CONTINUED)

Sami A. Al-KhursaniProgram Director, Technology

Ashraf A. GhazzawiManager, Lab Research and Development Center

Samer S. AlAshgarManager, EXPEC ARC

CONTRIBUTIONS

Relevant articles are welcome. Submissionguidelines are printed on the last page.Please address all manuscript and editorial correspondence to:

EDITOR

William E. BradshawThe Saudi Aramco Journal of TechnologyRoom 2240 East Administration BuildingDhahran 31311, Saudi ArabiaTel: +966/3-873-5803E-mail: [email protected]

Unsolicited articles will be returned onlywhen accompanied by a self-addressedenvelope.

Khalid A. Al-FalihPresident & CEO, Saudi Aramco

Mohammed Al-QahtaniVice President, Saudi Aramco Affairs

Essam Z. TawfiqGeneral Manager, Public Affairs

PRODUCTION COORDINATION

Robert M. Arndt, ASC

DESIGN

Pixel Creative Group, Houston, Texas, U.S.A.

ISSN 1319-2388.

© COPYRIGHT 2013 ARAMCO SERVICES COMPANYALL R IGHTS RESERVED

No articles, including art and illustrations, inthe Saudi Aramco Journal of Technology,except those from copyrighted sources, maybe reproduced or printed without thewritten permission of Saudi Aramco. Pleasesubmit requests for permission to reproduceitems to the editor.

The Saudi Aramco Journal of Technologygratefully acknowledges the assistance,contribution and cooperation of numerousoperating organizations throughout the company.

ATTENTION! MORE SAUDI ARAMCOJOURNAL OF TECHNOLOGY ARTICLESAVAILABLE ON THE INTERNET.

Additional articles that were submitted forpublication in the Saudi Aramco Journal ofTechnology are being made available online. Youcan read them at this link on the Saudi AramcoInternet Web site: www.saudiaramco.com/jot.html

62884araD1R2_ASC027 3/17/13 3:46 PM Page 2

creo

Spring 2013

THE SAUDI ARAMCO JOURNAL OF TECHNOLOGYA quarterly publication of the Saudi Arabian Oil CompanyJournal of Technology

Saudi Aramco

Contents

Successful Implementation of Horizontal Multistage Fracturing to Enhance Gas Production in Heterogeneous and Tight Gas Condensate Reservoirs: Case Studies 2Dr. Hamoud A. Al-Anazi, Dana M. Abdulbaqi, Ali H. Habbtar andAdnan A. Al-Kanaan

Evaluation of Nonreactive Aqueous Spacer Fluids for Oil-based Mud Displacement in Open Hole Horizontal Wells 10Peter I. Osode, Msalli Al-Otaibi, Khalid H. Bin Moqbil, Khaled Kilany and Eddy Azizi

Selecting Optimal Fracture Fluids, Breaker System and Proppant Type for Successful Hydraulic Fracturing and Enhanced Gas Production – Case Studies 22Dr. Zillur Rahim, Dr. Hamoud A. Al-Anazi and Adnan A. Al-Kanaan

Assessment of Multistage Stimulation Technologies as Deployed in the Tight Gas Fields of Saudi Arabia 30Mohammed A. Al-Ghazal, Saad M. Al-Driweesh, Fadel A. Al-Ghurairi,Abdulaziz M. Al-Sagr and Mustafa R. Al-Zaid

An Iterative Solution to Compute Critical Velocity and Rate Required to Unload Condensate-Rich Saudi Arabian Gas Fields and Maintain Field Potential and Optimal Production 39Hamza Al-Jamaan, Dr. Zillur Rahim, Bandar H. Al-Malki and Adnan A. Al-Kanaan

Microbial Community Structure in a Seawater Flooding System in Saudi Arabia 46Mohammed A. Al-Moniee, Dr. Indranil Chatterjee, Dr. Gerrit Voordouw, Dr. Peter F. Sanders and Dr. Tony Y. Rizk

Comprehensive Diagnostic and Water Shut-off in Open and Cased Hole Carbonate Horizontal Wells 52Nawawi A. Ahmad, Hussein S. Al-Shabebi, Dr. Murat Zeybek andShauket Malik

Black Oil, Heavy Oil and Tar in One Oil Column Understood by Simple Asphaltene Nanoscience 59Douglas J. Seifert, Dr. Murat Zeybek, Dr. Chengli Dong, Dr. Julian Y. Zuo and Dr. Oliver C. Mullins

Cementing Abnormally Over-pressurized Formations in Saudi Arabia 68Abdulla F. Al-Dossary and Scott S. Jennings

62884araD2R1_ASC026 3/15/13 11:23 PM Page 1

creo

ABSTRACT INTRODUCTION

The increase in energy demand has led operators to exploit allhydrocarbon resources, including tight gas reservoirs. Accord-ingly, service companies have developed several technologiesfor well completion and stimulation to enable the operators totarget tight gas reservoirs and ensure enhancing and sustaininggas productivity in the most effective and economical manner.In Saudi Arabia, most of the conventional gas wells have beendrilled in the maximum horizontal stress direction to avoid anypotential wellbore instability during drilling operations. Thistechnique has been successfully implemented and has long pro-vided the target sustained gas production from conventionalgas reservoirs1, 2. In the early stage of deploying multistagefracturing (MSF) in tight reservoirs that were to be acid or hydraulically fractured, though, it was found in such wells thatthe fracture grew along the wellbore in the direction of thewell azimuth and resulted in longitudinal fractures; this causedthe overlapping of two adjacent induced fractures, and therebycommunication between the stages, which meant only two tothree stages of fracture treatments could be performed, whilethe remaining stages ended with high rate stimulation. Depend-ing on the length of the wellbore reservoir contact, reservoirdevelopment and stress barriers, more than four fracture treat-ments in such wells can become redundant or even cause premature screen-out in proppant fracture treatments3-6.

Wells drilled in the direction of minimum horizontal stressare potentially more favorable candidates for fracturing fromthe perspective of reservoir development and optimal production.In such situations, hydraulic fractures grow transverse to thewellbore axis, allowing multiple fractures to be placed withoutthe possibility of fracture overlapping or communication betweenstages. Yet a few wells drilled in the minimum horizontal stressdirection encountered several drilling related problems such asstuck pipe, hole breakouts causing ovality or formation break-downs. A comprehensive study is essential to investigate thefeasibility of drilling wells in the minimum horizontal stress direction to overcome such wellbore instability issues. Correctmud weight prediction is one key factor during the drillingstage that helps keep the wellbore stable for the good boreholegeometry needed to run the MSF assembly without complica-tion. Multiple transverse hydraulic fractures can be created in

The heterogeneity and tightness of retrograde carbonate reservoirs are the main challenges to maintaining gas well productivities. The degree of heterogeneity changes over thefield and within well drainage areas, where permeability candecrease from a few millidarcies (md) to less than 0.2 md.Thorough studies conducted to exploit these tight reservoirsnot only have focused on well performance, but also have beenextended to assure enhancing and sustaining gas productivitythrough practical applications of new technologies. The mainobjective of this article is to assess the performance of multistagefracturing (MSF) in horizontal wells that were drilled conven-tionally and did not meet gas deliverability expectations. Thisarticle provides a detailed analysis of well performances, exploitation approaches and successful implementation of newcompletion technologies, such as horizontal MSF, to revive lowproducing gas wells due to reservoir tightness. Placing the hori-zontal wellbore in reference to the stress directions plays a major role in the success and effectiveness of fracturing in enhancing and sustaining productivity.

Several wells had been drilled in tight reservoirs, but couldnot achieve or sustain the target gas rate. Recently, two ofthese wells were geometrically sidetracked, targeting the devel-opment intervals based on logs of the original hole, and completed with MSF toward the minimum stress direction.Open hole logs showed a low porosity development similar tothat of the vertical holes; however, after conducting multiplestages of fracturing, both wells produced a sustainable rate ofmore than 25 million standard cubic feet per day (MMscfd),which prompted connecting them to gas plants. Placing thesesidetracks in the minimum stress direction helped to createtransverse fractures that connected to sweet spots and sustainedgas production. This article provides thorough guidelines forselecting optimal candidates for MSF, based on reservoir heterogeneity, and for the proper design and execution of fracturing. It also addresses various components that contributedto the success of both wells, such as reservoir development,workover pre-planning, geomechanical studies, drilling opera-tions and real-time support, completion operations optimiza-tion and best practices, and performance evaluation of otherproducers in the field.

2 SPRING 2013 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

Successful Implementation of HorizontalMultistage Fracturing to Enhance GasProduction in Heterogeneous and Tight GasCondensate Reservoirs: Case StudiesAuthors: Dr. Hamoud A. Al-Anazi, Dana M. Abdulbaqi, Ali H. Habbtar and Adnan A. Al-Kanaan

62884araD2R1_ASC026 3/15/13 11:23 PM Page 2

a good wellbore geometry to maximize the reservoir contactarea, so as to increase and sustain productivity from the lowquality tight reservoir. The key objectives of the geomechanicalstudy are to define a safe mud weight program for the horizontalsection of the planned wells by conducting a wellbore stabilitystudy, and to determine a real-time strategy to mitigate and/ormanage wellbore instability problems as they arise. Develop-ment of the comprehensive mechanical earth model should allowoptimization in drilling trajectory, running the completion, per-foration interval selection and fracture design3, 6, 7.

This article provides thorough guidelines for the selection of optimal candidates for MSF, based on reservoir heterogeneity,and for the proper MSF design and job execution. It also ad-dresses various components that contributed to the success ofboth wells, such as reservoir development, workover pre-plan-ning, geomechanical studies, drilling operations and real-timesupport, completion operations optimization and best practices,and performance evaluation of other producers in the field.

RESERVOIR DESCRIPTION

The major nonassociated gas reservoirs in a major Saudi Ara-bian gas field (Field-A) are present in the upper Permian-LateTriassic formation, which is divided into four depositional cycles.Three reservoirs (A, B and C) are gas bearing, while Reservoir-D is anhydrite. Reservoir-B represents a third order compositecycle that commenced with a sea level rise following a longtime of exposure and nondeposition at the Permo-Triassicboundary. Reservoir-B comprises two high frequency sequences,initiated with the deposition of an open marine thrombolyticlime mudstone, that shallow upwards into lagoonal and periti-dal facies. Reservoir-B is represented by three reservoir faciescomposed of oolitic peloidal grainstone, mud-lean ooliticpeloidal packstone and horizontally burrowed shallow subtidaldolostone. The oolitic peloidal grainstone is the most common,with moldic porosity in the calcareous upper part of the reser-voir. The porosity of the grainstone is enhanced where the rockis dolomitized to include moldic and inter-crystalline porosity.The moldic porosity associated with the ooid grainstone repre-sents the main reservoir rock.

The reservoir is highly heterogeneous and exhibits anomalousfluid and stress characteristics. The formation has limited pre-served primary porosity development, with reservoir qualityrelated to the digenetic process of dolomitization, selective dissolution of limestone and cementation (anhydrite). Litho-logical studies show that the reservoir is composed of dolomiteintermingled with limestone and intermittent anhydritestringers within the tighter section of the reservoir. The threetypes of porosity observed in the reservoir are inter-particle, inter-crystalline and moldic. Natural fractures have also beenobserved in some cores. Therefore, it is fair to say that thereservoir is structurally complex and heterogeneous. The bestreservoir development is typically noticed in the dolomitizedgrainstone with high inter-particle porosity. Reservoir-B in

particular is a large heterogeneous and compartmentalizedreservoir with multiple gas-water contacts, faulting and varia-tion in flow capacity. Regionally, the entire field is divided intoseveral sections based on reservoir characteristics, porosity de-velopment and varying production rates. As such, area specific development methodologies need to be established to optimizegas exploitation from each area8.

PETROGRAPHY

Petrographic evaluations of several core samples, Fig. 1, fromvarious wells indicated a composition of limestone and dolo-stone: calcite, dolomite and anhydrite are commoncementing/replacement minerals in many samples. Scanningelectron microscope (SEM) and X-ray diffraction (XRD)analysis conducted on these samples confirmed the observedmineralogy. The allochems in the lime grainstones are moder-ately sorted, and average grain size ranges from 330 to 383microns (medium sand size). Some of the micritic grains havebeen replaced by dolomite, Fig. 2. Grains appear to have undergone a minor to moderate amount of compaction, as evidenced by the numerous point and long grain contacts andthe fewer concavo-convex grain contacts and stylolites. On the

SAUDI ARAMCO JOURNAL OF TECHNOLOGY SPRING 2013 3

Fig. 1. Photographs of reservoir cores recovered from pay zone at various depths.

62884araD2R1_ASC026 3/15/13 11:23 PM Page 3

other hand, the dolostone is poorly sorted, and average grainsize ranges from 438 to 882 microns (upper medium to coarsesand size). The original fabric of the remaining dolostone hasbeen partially to almost completely obscured by the dolomiti-zation process. The average crystal size in the dolostone rangesfrom 14 to 25 microns (finely crystalline). Cementation by calcite and anhydrite is the main cause of the reduction of primary pore volume. Later dissolution of grains and dolomiti-zation generated the secondary pores that make up much ofthe total porosity. Porosity and permeability data from conven-tional core analysis was integrated and cross-plotted by lithology.The average porosity and permeability in limestone is 12.5%and 0.196 md, while that in dolostone is 16.6% and 5.88 md.The low values in limestone are due to the pore-filling calcitecement that left few primary pores. The secondary pores arepoorly connected due to the extensive calcite cementation5.

RESERVOIR HETEROGENEITY

Reservoir-B is a naturally fractured gas carbonate reservoirthat covers most of the field. It is the largest in size comparedto the other carbonate and sandstone reservoirs in the field.The reservoir is part of the carbonate formation and belongsto the Triassic age. The reservoir quality varies regionally according to the ratio of anhydrite to carbonate components,and the matrix porosity and permeability, as illustrated in thecross section of wells drilled in the field, Fig. 3. The fracturedensity increases from the central area, where the fractures arethin, dispersed and mostly short in length (< 1 ft)9. Therefore,the reservoir performance varies widely among offset wells inthe same field1, 5.

Analysis of reservoir data indicates the presence of signifi-cant areal and vertical pressure compartmentalization. Seismicdata shows variability in reservoir characteristics, which isusually thin up to 20 ft true vertical depth (TVD). Due to thethin nature of the reservoir, seismic impedance inversion is notprecise and many times cannot be correlated with log porosityand reservoir performance. In many places, multiple contami-nations of the data make it impossible to arrive at a correct interpretation. Changing dip and structures also pose majorchallenges for correct interpretation. Another challenge is thepresence of multiple gas-water contacts, as observed in formation


analysis logs and well tests. Therefore, well placement is criti-cal to avoid wet zones and mitigate water encroachments3, 10.

Reservoir heterogeneity necessitates the use of effectivedrilling and completion fluids that reduce induced formationdamage if the wells are to achieve their expected potential11, 12.Pressure compartmentalization has a major impact on produc-tion performance due to the potential drop in the bottom-holeflowing pressure below the dew point pressure, which wouldtrigger the onset of condensate banking13. Several techniqueshave been deployed to address this onset, such as solvent treat-ment to remove the condensate banking around the wellboreregion, but production has been enhanced only up to severalmonths14. More effective treatments, such as wettability alteration, have been extensively tested and approved in thelab, and are now undergoing field trials on candidate gaswells15-17.

BEST PRACTICES TO EXPLOIT TIGHT GAS RESERVOIRS

Tight gas reservoir development requires good reservoir char-acterization based on sufficient data from core analysis, offsetwell logs, reservoir parameters and production performance.The following steps are a prerequisite for effective develop-ment of a tight gas reservoir:

• Identify the bottom-hole location based on seismic andoffset well data.

• Drill a vertical pilot hole across all layers of the targetreservoir.

• Run open hole logs (density/neutron/resistivity/gammaray/caliper).

• Take pressure points and samples to assess fluidgradients, fluid type and mobility.

• Drill a geometric horizontal hole in the minimum stressdirection targeting the most developed sections observedin the pilot hole. The geomechanical study parametersmust be determined prior to drilling the sidetrack.

• Maintain the recommended mud weight and inclination.

• Run open hole logs to assess the reservoir developmentacross the geometric lateral.

Fig. 2. Thin section and SEM photomicrographs of dolomitic and anhydriticlimestone core.

Fig. 3. Porosity development profiles indicate the heterogeneity of Reservoir-Bwithin the field.

62884araD2R1_ASC026 3/15/13 11:23 PM Page 4


• Design the MSF according to porosity developmentpackages and ensure there is enough spacing betweenstages to avoid communication.

• Place packers across the gauged hole section away fromthe washout zones.

• Flow back the well completed with MSFs for cleanupbefore conducting the fracturing treatment.

• Drill the geometric horizontal lateral with nondamagingfluids (no barite) to achieve the needed fracturing fluidinjectivity and avoid the necessity for any wellboreintervention to open the ports mechanically.

CASE STUDIES

Well-A

Well-A was drilled in 2007 as an open hole Reservoir-C hori-zontal development well. Due to the poor reservoir qualityseen in the well’s motherbore, it was suspended with a 7”bridge plug and three cement plugs. In December 2011, planswere made to sidetrack the well as a horizontal gas produceracross Reservoir-B in the minimum horizontal stress directionas part of a strategy to exploit that area’s tight gas reservoirs.The well was sidetracked from inside the 95⁄8” casing using amechanical whipstock. After milling the window, an 83⁄8” di-rectional hole section was drilled across Reservoir-A, buildingfrom around a 3° inclination to a 89° inclination at the 7”liner point inside Reservoir-B, with 103 to 106 pounds per cu-bic ft (pcf) of potassium chloride (KCl) polymer mud. Therewas no major problem in drilling this hole section, with a rateof penetration (ROP) averaging at 8.3 ft/hr. After running andcementing the 7” liner, the 57⁄8” section was drilled using adownhole motor for better ROP (due to continuous rotationwithout having to slide for directional control). Potassium (K)formate mud type was used as it is nondamaging to the reser-voir, and its lubricity helped reduce torque and drag whiledrilling this hole in the minimum stress direction. A highermud weight of 103 pcf mud was chosen, as recommended bythe geomechanical studies, to mitigate wellbore instability is-sues due to the well azimuth’s being drilled towards the mini-mum horizontal stress direction. With this mud weight,Reservoir-B was overbalanced by ~700 psi. Proper sizedCaCO3 chips were added to the K formate mud system to helpcreate a bridging action across the permeable reservoir sec-tions, thereby minimizing the chance of differential sticking.

Nevertheless, the string got mechanically stuck momentarilywhile moving across the reservoir, but it was freed after spot-ting an acid pill and jarring. While drilling at 15,793 ft meas-ured depth, the downhole motor drive shaft broke, leaving thebit sub and 57⁄8” bit at the bottom of the well. After runninglogs across the open hole section, Fig. 4, the decision wasmade to call total depth to avoid risky fishing operations andto not jeopardize the hole. Therefore, a total of 3,566 ft of

57⁄8 ” lateral was drilled compared to the 5,400 ft originallyplanned. The open hole logs showed development in only twozones in Reservoir-B with an average porosity of 6%. The wellwas completed with two-stage MSF equipment to enhance thewell productivity. Three mechanical packers were installed be-tween the stages to reduce the potential of communication dur-ing pumping of the fracturing fluids. The lower frac-port wasopened with a rig on location since it is a pressure actuated port.

The fracturing treatment was designed to create a fracturein each stage with a gelled pad, after which alternating stagesof acid systems and additional pads were pumped. A polymer-free acid system was used as the diverter system to assist inmaximizing the fracture half-length in each stage. Table 1 liststhe fracturing treatment components and volumes for eachstage. Prior to performing the fracturing, the well was flowedback for cleanup and achieved a flush rate of 11 MMscfd at1,723 psi flowing wellhead pressure (FWHP), followed with agradual decline. This rate confirmed the intersection of thewellbore with natural fractures as a result of drilling in theminimum stress direction18, 19. Both stages were pumped suc-cessfully with positive indication of isolation between the twostages. After conducting the acid fracturing treatments, thewell was flowed back for cleanup at a gas rate of 43 MMscfdwith 2,942 psi FWHP. The productivity index of this well isshown in Fig. 5, which indicates the effectiveness of the two-stage fracturing treatment in enhancing well productivity fromthis tight heterogeneous reservoir. Based on these encouragingresults, drilling in the minimum horizontal stress direction andcompletion with the MSF assemblies were followed in otherwells designed for the exploitation of the tight gas.

Fig. 4. Schematic of original and sidetracked wellbores of Well-A with porositydevelopment profiles.

Component Volume, bbl

15% HCI Spearhead Acid 48

Acid Fracturing Pad 870

28% Emulsifi ed Acid 821

Acid Diverting System-1 197

Acid Diverting System-2 197

Table 1. Fracturing treatment components and volumes for each stage

62884araD2R1_ASC026 3/15/13 11:23 PM Page 5

Well-B

Well-B was initially drilled and completed in 2004 as a verticalproducer across a deep clastic reservoir; it was permanentlyplugged due to poor development and performance after test-ing. The open hole logs across Reservoir-B showed moderatedevelopment in the B-1 layer. Therefore, a workover wasplanned to drill a geometric sidetrack in the B-1 layer in theminimum horizontal stress direction and to complete it with aMSF assembly as part of the initiative to exploit Reservoir-B.The sidetrack operation was carefully planned based on thelessons learned from Well-A to avoid getting stuck, and drillingachieved the planned reservoir contact. Due to reservoir het-erogeneity, the well encountered 650 ft net reservoir contact inthis Reservoir-B geometric lateral. Three distinct porosity loopswere identified at 440 ft, 160 ft and 50 ft, with the bottom,middle and upper sections having 5%, 12% and 7% averageporosity, respectively, Fig. 6.

To perform the acid fracturing treatment, initially a 1¾”coiled tube (CT) was run in hole to clean/displace the wellborewith 240 bbl of treated water from a depth of 15,568 ft to thesurface. Then two 10 bbl pills of 26% hydrochloric acid werepumped to try to achieve the minimum injection rate of 5 bar-rels per minute (bpm) required to displace the balls needed toactivate the ports in the completion. But the maximum injec-tion rate achieved was only 0.6 bpm at 5,800 psi, which implied that it was not possible to pump the scheduled acidmatrix stimulation treatment in this first stage. Therefore, acidfracturing of the first stage was canceled due to the poor injec-tivity caused by either reservoir tightness or plugging of thefrac-port. The activation of the second stage port by pumpinga ball was also not possible at this low rate, so the port had tobe opened mechanically. A 2” CT fitted with a 3” activatortool was run to open the second stage frac-port at 14,356 ft.The 2¾” frac-port ball seat was tagged at 14,363 ft, and 4,000lb of slack-off force was applied on it. To open the port,treated water was then pumped at 1.3 bpm until the surfacepumping pressure stabilized at 4,900 psi, indicating that the

port was opened and continuous injection into the formationwas taking place. The CT was then pulled up to 10,000 ft, andthe well was opened to flow over a 90-minute period, duringwhich it produced gas at a rate of 10 MMscfd at 3,880 psiFWHP with 75% basic sediments and water. Next, an injectivitytest was performed by pumping 250 bbl of treated waterthrough the CT and CT/tubing annulus into the formation, at a stabilized injection rate and pressure of 4 bpm and 6,400psi, respectively. The rate was deemed adequate to displace the second stage ball and isolate the first open port, a required step before proceeding with the scheduled acid fracturing treatment.

Operations to acid fracture the second and third stages ofthe MSF completion system in Reservoir-B were successfullycompleted. The second frac-port, set at a depth of 14,494 ft,was successfully opened by dropping a 3” ball. A mini fall-off(MFO) was performed at a maximum pumping rate, treatingpressure and bottom-hole pressure (BHP) of 20 bpm, 9,900 psiand 13,800 psi, respectively, followed by a step rate test (SRT)at a maximum rate, treating pressure and BHP of 25 bpm,11,000 psi and 14,400 psi, respectively. Then the second stagemain fracturing treatment was performed by displacing a totalvolume of 2,094 bbl at a maximum rate, treating pressure andBHP of 60 bpm, 11,600 psi and 13,200 psi, respectively, intothe formation. The third frac-port, set at a depth of 13,312 ft,was successfully opened by dropping a 3¼” ball. The MFOwas performed by displacing a total volume of 50 bbl of


Fig. 7. Productivity index of Well-B after the two-stage fracturing treatment.

Fig. 5. Productivity index of Well-A after the two-stage fracturing treatment.

Fig. 6. Schematic of original and sidetracked wellbores of Well-B with porositydevelopment profiles.

62884araD2R1_ASC026 3/15/13 11:23 PM Page 6


treated water at a maximum rate, treating pressure and BHP of15 bpm, 8,400 psi and 12,700 psi, respectively, followed by aSRT at a maximum pumping rate, treating pressure and BHPof 20 bpm, 10,000 psi and 14,000 psi, respectively. Then thethird stage main fracturing treatment was successfully per-formed by displacing a total volume of 1,020 bbl at a maxi-mum rate, treating pressure and BHP of 59 bpm, 11,880 psiand 13,900 psi, respectively, into the formation. The well wasflowed back for cleanup. The productivity index profile isshown in Fig. 7. The gas rate reached 36 MMscfd at 2,560 psiFWHP. The successful implementation of sidetracking andfracturing converted Well-B from a suspended well drilled in2004 into a strong producer that will be connected to the nearestnonassociated gas plant.

CONCLUSIONS AND RECOMMENDATIONS

The strategy of drilling a pilot hole to help in placing the hori-zontal hole to target the best porosity development in a hetero-geneous reservoir was very practical. Placing these sidetracksin the minimum stress direction and using MSF completionshelped to create transverse fractures that connected to sweetspots and sustained gas production. The sidetracks alsoopened the possibility to intersect with the natural fracturesthat exist parallel to the maximum stress direction. Geome-chanical studies helped control wellbore instability by predict-ing the proper mud weight needed to drill the horizontallateral in the minimum stress direction. The application of theMSF completion proved successful in enhancing gas productiv-ity from these tight reservoirs.

Based on these case studies, it is recommended to considerthe following in the exploitation of tight gas reservoirs:

• Drilling horizontal wells in the minimum stress directionis a prerequisite for successful MSF in tight gas reservoirdevelopment.

• Geomechanical studies are essential to ensure problem-free drilling and placing of the horizontal wellbore inthe direction of the minimum stress.

• Mud weight windows for hole breakouts and loss ofcirculation need to be predicted from the geomechanicalstudy as a function of well deviation and azimuth.

• A real-time geomechanical model has proven to beeffective in predicting the proper mud weight windowand preventing wellbore instability and drilling relatedproblems.

• Sufficient spacing between frac-ports in the MSFcompletion plays a major role in achieving desiredfracturing pressure and eliminating communicationthrough packers, which must be placed across thegauged hole section away from the washout zones.

• Wells completed with MSF need to be flowed back forcleanup before conducting the fracturing treatment.

ACKNOWLEDGMENTS

The authors would like to thank Saudi Aramco managementfor the permission to present and publish this article. We appreciate the help of all personnel from the Gas ReservoirManagement and Gas Production Engineering Departmentsfor their assistance.

This article was presented at the Abu Dhabi InternationalPetroleum Exhibition and Conference (ADIPEC), Abu Dhabi,U.A.E., November 11-14, 2012.

REFERENCES

1. Al-Qahtani, M.Y. and Rahim, Z.: “Optimization of AcidFracturing Program in the Khuff Gas Condensate Reservoirof South Ghawar Field in Saudi Arabia by ManagingUncertainties Using State-of-the-Art Technology,” SPEpaper 71688, presented at the SPE Annual TechnicalConference and Exhibition, New Orleans, Louisiana,September 30-October 3, 2001.

2. Rahim, Z. and Petrick, M.: “Sustained Gas Productionfrom Acid Fracture Treatments in the Khuff Carbonates,Saudi Arabia: Will Proppant Fracturing Make RatesBetter? Field Example and Analysis,” SPE paper 90902,presented at the SPE Annual Technical Conference andExhibition, Houston, Texas, September 26-29, 2004.

3. Rahim, Z., Al-Anazi, H.A., Al-Kanaan, A.A. and AbdulAziz, A.: “Successful Exploitation of Khuff-B LowPermeability Gas Condensate Reservoir through OptimizedDevelopment Strategy,” SPE paper 136953, presented atthe SPE/DGS Saudi Arabia Section Technical Symposiumand Exhibition, al-Khobar, Saudi Arabia, April 4-7, 2010.

4. Rahim, Z., Al-Anazi, H.A., Al-Malki, B. and Al-Kanaan,A.A.: “Optimized Stimulation Strategies Enhance AramcoGas Production,” Oil and Gas Journal, October 4, 2010,pp. 66-74.

5. Al-Anazi, H.A., Al-Baqawi, A.M., Ahmad Azly, A.A. andAl-Kanaan, A.A.: “Effective Strategies in Development ofHeterogeneous Gas-Condensate Carbonate Reservoirs,”SPE paper 136399, presented at the SPE Russian Oil andGas Conference and Exhibition, Moscow, Russia, October26-28, 2010.

6. Ahmed, M., Rahim, Z., Al-Anazi, H.A., Al-Kanaan, A.A.and Mohiuddin, M.: “Development of Low PermeabilityReservoir Utilizing Multistage Fracture Completion in theMinimum Stress Direction,” SPE paper 160848, presentedat the SPE Saudi Arabia Section Annual TechnicalSymposium and Exhibition, al-Khobar, Saudi Arabia, April8-11, 2012.

7. Rahim, Z., Al-Qahtani, M.Y., Bartko, K.A., Goodman, H.,Hilarides, W.K. and Norman, W.D.: “The Role ofGeomechanical Earth Modeling in the Unconsolidated pre-

62884araD2R1_ASC026 3/15/13 11:23 PM Page 7

Khuff Field Completion Design for Saudi Arabian GasWells,” SPE paper 84258, presented at the SPE AnnualTechnical Conference and Exhibition, Denver, Colorado,October 5-8, 2003.

8. Clerke, E.A.: “Electrofacies and Geological Facies forPetrophysical Rock Typing: Khuff C,” SPE paper 126086,presented at the SPE Saudi Arabia Section TechnicalSymposium, al-Khobar, Saudi Arabia, May 9-11, 2009.

9. Ameen, M.S., Buhidma, I.M. and Rahim, Z.: “TheFunction of Fractures and In-Situ Stresses in the KhuffReservoir Performance, Onshore Fields, Saudi Arabia,”AAPG Bulletin, Vol. 94, No. 1, January 2010, pp. 27-60.

10. Al-Shehri, D.A., Rabaa, A.S., Duenas, J.J. and Ramanathan, V.: “Commingled Production Experiences of Multilayered Gas-Condensate Reservoir in Saudi Arabia,” SPE paper 97073, presented at the SPE Annual Technical Conference and Exhibition, Dallas, Texas, October 9-12, 2005.

11. Al-Anazi, H.A., Bataweel, M.A. and Al-Ansari, A.A.: “Formation Damage Induced by Formate Drilling Fluids in Gas Bearing Reservoirs: Lab and Field Studies,” SPE/IADC paper 119445, presented at the SPE/IADC Drilling Conference and Exhibition, Amsterdam, The Netherlands, March 17-19, 2009.

12. Al-Anazi, H.A., Okasha, T.M., Haas, M.D., Ginest, N.H. and Al-Faifi, M.G.: “Impact of Completion Fluids on Productivity in Gas/Condensate Reservoirs,” SPE paper 94256, presented at the SPE Production and Operations Symposium, Oklahoma City, Oklahoma, April 17-19, 2005.

13. Al-Anazi, H.A., Solares, J.R. and Al-Faifi, M.G.: “The Impact of Condensate Blockage and Completion Fluids on Gas Productivity in Gas Condensate Reservoirs,” SPE paper 93210, presented at the Asia Pacific Oil and Gas Conference and Exhibition, Jakarta, Indonesia, April 5-7, 2005.

14. Garzon, F.O., Al-Anazi, H.A., Leal, J.A. and Al-Faifi, M.G.: “Laboratory and Field Trial Results of Condensate Banking Removal in Retrograde Gas Reservoirs: Case History,” SPE paper 102558, presented at the SPE AnnualTechnical Conference and Exhibition, San Antonio, Texas, September 24-27, 2006.

15. Al-Anazi, H.A., Xiao, J.J., Eidan, A.A., Buhidma, I.M., Ahmed, M.S., Al-Faifi, M.G., et al.: “Gas Productivity Enhancement by Wettability Alteration of Gas Condensate Reservoirs,” SPE paper 107493, presented at the 7th SPE European Formation Damage Conference, Scheveningen, The Netherlands, May 30-June 1, 2007.

16. Xie, X., Liu, Y., Sharma, M. and Weiss, W.W.: “Wettability Alteration to Increase Deliverability of Gas

Production Wells,” SPE paper 117353, presented at the SPE Eastern Regional/AAPG Eastern Section Joint Meeting, Pittsburgh, Pennsylvania, October 11-15, 2008.

17. Ahmadi, M., Sharma, M.M., Pope, G.A., Torres, D.E., McCulley, C.A. and Linnemeyer, H.: “Chemical Treatment to Mitigate Condensate and Water Blocking in Gas Wells in Carbonate Reservoirs,” SPE Production & Operations, Vol. 26, No. 1, February 2011, pp. 67-74.

18. Demarchos, A.S., Porcu, M.M. and Economides, M.J.: “Transverse Multifractured Horizontal Wells: A Recipe for Success,” SPE paper 102262, presented at the SPE Annual Technical Conference and Exhibition, San Antonio, Texas, September 24-27, 2006.

19. Bahrami, H., Rezaee, M.R. and Asadi, M.S.: “Stress Anisotropy, Long-Term Reservoir Flow Regimes and Production Performance in Tight Gas Reservoirs,” SPE paper 136532, presented at the SPE Eastern Regional Meeting, Morgantown, West Virginia, October 12-14, 2010.


BIOGRAPHIES

Dr. Hamoud A. Al-Anazi is theGeneral Supervisor of the NorthGhawar Gas Reservoir ManagementDivision in the Gas ReservoirManagement Department (GRMD).He oversees all work related to thedevelopment and management of huge

gas fields like Ain-Dar, Shedgum and ‘Uthmaniyah.Hamoud also heads the technical committee that isresponsible for all new technology assessments andapprovals for GRMD. He joined Saudi Aramco in 1994 asa Research Scientist in the Research & Development Centerand moved to the Exploration and Petroleum EngineeringCenter – Advanced Research Center (EXPEC ARC) in2006. After completing a one-year assignment with theSouthern Area Reservoir Management Department,Hamoud joined the Gas Reservoir Management Divisionand was assigned to supervise the SDGM/UTMN Unit andmore recently the HWYH Unit. With his team hesuccessfully implemented the deepening strategy of keywells that resulted in a new discovery of the Unayzahreservoir in UTMN field and the addition of Jauf reservesin the HWYH gas field.

Hamoud’s areas of interests include studies of formationdamage, stimulation and fracturing, fluid flow in porousmedia and gas condensate reservoirs. He has publishedmore than 50 technical papers at local/internationalconferences and in refereed journals. Hamoud is an activemember of the Society of Petroleum Engineers (SPE) wherehe serves on several committees for SPE technicalconferences. He is also teaching courses at King FahdUniversity of Petroleum and Minerals (KFUPM), Dhahran,Saudi Arabia, as part of the Part-Time Teaching Program.

62884araD2R1_ASC026 3/15/13 11:23 PM Page 8


In 1994, Hamoud received his B.S. degree in ChemicalEngineering from KFUPM, and in 1999 and 2003,respectively, he received his M.S. and Ph.D. degrees inPetroleum Engineering, both from the University of Texasat Austin, Austin, TX.

Dana M. Abdulbaqi has been aPetroleum Engineer with SaudiAramco since 2004. She has hadseveral assignments with variouspetroleum engineering anddevelopment departments, includingthe Production and Facilities

Development Department and Reservoir ManagementDepartment.

Dana is an active member of the Society of PetroleumEngineers (SPE) from which she obtained her PetroleumEngineering Certification. She is also a member of theInternational Association for Energy Economics as well asthe Saudi Council of Engineers. In addition to herinvolvement in these professional societies, in 2012, sheestablished and chaired Qudwa (www.qudwa.org), which isan affinity group that aspires to encourage dialogue andopen discussion by providing opportunities for its membersto interact via networking, skill building, and knowledgesharing and mentoring with special consideration to genderdifferences.

Dana received her B.S. degree in Architecture fromVirginia Tech, Blacksburg, VA. She completed an M.S.degree in Petroleum Engineering from Texas A&MUniversity, College Station, TX, and is currently pursuing aPh.D. degree in Mineral and Energy Economics at theColorado School of Mines, Golden, CO.

Ali H. Habbtar is a Supervisor of theHWYH Unit in the Gas ReservoirManagement Department and isresponsible for the management of allreservoirs feeding the Hawiyah GasPlant. He has over 10 years of industryexperience in reservoir engineering and

well productivity enhancement through stimulation. As a member of the Society of Petroleum Engineers (SPE),Ali has published numerous SPE papers. He is the chairmanof the upcoming 2013 SPE Saudi Arabia TechnicalSymposium.

Ali received his B.S. degree in Petroleum Engineeringfrom Pennsylvania State University, University Park, PA,and an M.B.A. from the Instituto de Estudios Superiores dela Empresa (IESE Business School), Barcelona, Spain.

Adnan A. Al-Kanaan is the Managerof the Gas Reservoir ManagementDepartment (GRMD) where heoversees three gas reservoir manage-ment divisions. Reporting to the ChiefPetroleum Engineer, Adnan is directlyresponsible for making strategic

decisions to enhance and sustain gas delivery to theKingdom to meet its ever increasing energy demand. Heoversees the operating and business plans of GRMD, newtechnologies and initiatives, unconventional gas develop-ment programs, and the overall work, planning anddecisions made by his more than 70 engineers andtechnologists.

Adnan has 15 years of diversified experience in oil andgas reservoir management, full field development, reservesassessment, production engineering, mentoring youngprofessionals and effectively managing large groups ofprofessionals. He is a key player in promoting and guidingthe Kingdom’s unconventional gas program. Adnan alsoinitiated and oversees the Tight Gas Technical Team toassess and produce the Kingdom’s vast and challengingtight gas reserves in the most economical way.

Prior to the inception of GRMD, he was the GeneralSupervisor for the Gas Reservoir Management Divisionunder the Southern Reservoir Management Department for3 years, heading one of the most challenging programs inoptimizing and managing nonassociated gas fields in SaudiAramco.

Adnan started his career at the Saudi Shell Petro-chemical Company as a Senior Process Engineer. He thenjoined Saudi Aramco in 1997 and was an integral part ofthe technical team responsible for the on-time initiation ofthe two major Hawiyah and Haradh Gas Plants thatcurrently process more than 6 billion cubic feet (bcf) of gasper day. Adnan also directly managed the Karan and Wasitfields — two major offshore gas increment projects — withan expected total production capacity of 4.3 bcf of gas perday.

He actively participates in the Society of PetroleumEngineers’ (SPE) forums and conferences and has been thekeynote speaker and panelist for many such programs.Adnan’s areas of interest include reservoir engineering, welltest analysis, simulation modeling, reservoir charac-terization, hydraulic fracturing, reservoir developmentplanning and reservoir management.

He will be chairing the 2013 International PetroleumTechnical Conference to be held in Beijing, China.

Adnan received his B.S. degree in Chemical Engineeringfrom King Fahd University of Petroleum and Minerals(KFUPM), Dhahran, Saudi Arabia.

62884araD2R1_ASC026 3/15/13 11:23 PM Page 9

ABSTRACT immediately after well completion to avoid long-term mud andsolids aggregation in the wellbore. Residual mud cakes afterwellbore displacement with solids-free OBM DIFs are rela-tively thinner and easier to remove at low drawdown pressuresduring the initial production phase1, 2. Nevertheless, in manyother conditions, wellbore cleanup with reactive treatment fluids is required for filter cake dissolution and removal.

An effective cleanup treatment delivers optimum life cycleproductivity by allowing access to the entire pay zone at a min-imum drawdown pressure across the reservoir, and therefore,lowers the risk of early water breakthrough and fines migration3.Uniform placement of conventional breaker fluids for completetreatment of the horizontal wellbore, however, is difficult toachieve, especially in high permeability sandstone reservoirs,because of rapid fluid reaction and leak off at the first point ofcontact. Alternative systems, such as delayed reaction breaker(DRB) fluids, have provided only limited respite due to therapid cake solubility associated with complete hydrolysis of esters for in-situ generated organic acid at high bottom-holetemperatures4. Other DRB fluids with ethylene diamine tetra-acetic acid (EDTA) or its derivatives have indicated risks of reprecipitation when used in a divalent salt environment, whilethe inclusion of hydroxyl ethyl cellulose as a delay mechanismin DRB fluids shields calcium carbonate (CaCO3) particlesfrom the reactive fluid component and reduces the productivityperformance5. Dual-purpose delayed cleanup fluids that arebased on reversible invert emulsion DIF systems are complicatedand rely on a delicate pH control to be effective6, 7. CurrentDRB fluids are also deemed suboptimal for cleanup in extendedreach horizontal or multilateral wells when a noneffective mechanical isolation device is utilized with a wash pipe in thecompletion bore8.

Nonreactive Cleanup Fluids

The ideal cleanup solution for a high risk, high permeability/fractured reservoir is an extended delay breaker fluid systemthat is benign at the surface but provides homogeneous treat-ment of OBM DIF mud cake without causing severe wellborefluid losses during completion. The absence of such an idealfluid has prompted the use of nonreactive aqueous fluids witha properly designed displacement process to facilitate wellbore

Reactive mud cake breaker fluids in long open hole horizontalwells located across high permeability sandstone reservoirshave had limited success because they often induce massivefluid losses. The fluid losses are controlled with special pills,polymers and brine or water, causing well impairment that isdifficult to remove when oil-based mud (OBM) drill-in fluids(DIFs) are used. This situation has resulted in a drive for an alternative cleanup fluid system that is focused on preventingexcessive fluid leak off, maximizing the OBM displacement efficiency and allowing partial dispersion of the mud cake forease of its removal during initial well production. The two-stage spacer cleanup fluid is composed of a nonreactive fluidsystem, which includes a viscous pill with nonionic surfactants,a gel pill, a completion brine and a solvent.

Extensive laboratory testing was conducted at simulatedreservoir conditions to evaluate the effectiveness of the OBMdisplacement fluid system. The study included dynamic high-pressure/high temperature (HP/HT) filter press tests and core-flood tests, in addition to wettability alteration, interfacialtension and fluid compatibility tests.

The spacer fluid parameters were optimized based on well-bore fluid hydraulic simulation and laboratory test results,which indicated minimal fluid leak off and a low risk of emul-sion formation damage. Three well trials then were conductedin a sandstone reservoir drilled with OBM in a major offshorefield. All three trial wells (one single lateral and two dual later-als) treated with the displacement fluid system have demon-strated improvement in production performance. This articlewill discuss in detail the spacer fluids’ optimization process,the laboratory work conducted and the successful field treat-ments performed.

INTRODUCTION

Oil-based mud (OBM) drill-in fluids (DIFs) are favored fordrilling extended horizontal wells located in reservoirs withwater sensitive shale sections since they provide superior inhi-bition, greater lubricity, reduced mechanical friction and improved wellbore stability relative to water-based mud(WBM) DIFs. Ideally, removal of OBM cake should be done


Evaluation of Nonreactive Aqueous SpacerFluids for Oil-based Mud Displacement inOpen Hole Horizontal Wells

Authors: Peter I. Osode, Msalli Al-Otaibi, Khalid H. Bin Moqbil, Khaled A. Kilany and Eddy Azizi

62884araD3R1_ASC026 3/15/13 11:25 PM Page 1

OBM clean out and create a uniform mud cake “pinhole”prior to gradual liftoff of the residual cake during an earlyflow back/production kickoff operation9, 10. This technique issupported by previous formation damage studies, which indi-cate that DIF design optimization for filter cake removal viadrawdown can deliver up to 95% inflow performance for gasand oil reservoirs with minimum permeabilities of 1-2 mD and0.5-1 D, respectively11.

OBM DIFs generally utilize CaCO3 solids as a density andbridging material. OBM filter cake and solids removal in openhole/sand screen completion wells demands the use of cleanupfluids that can disperse the oily particles and thereby enhancethe residual DIF solids clean out from the wellbore. The poten-tial success of nonreactive fluids in achieving wellbore cleanout is predicated on the premise that only a limited filter cakeremoval, albeit uniformly across the wellbore, is required foroptimum well production performance. One well productivityassessment model estimates that less than 5% filter cake re-moval is required in a high permeability sandstone reservoirwith a slotted liner completion12-14. The solids-free, post-cleanup displacement brine fluids will also reduce the risk ofdamage in wells that are suspended with low solids, oil-basedDIFs/completion fluids in the wellbore long before the well iscleaned up and brought onstream.

Nonaqueous treatment fluids will not produce the desiredwettability changes in the near-wellbore area, whereas conven-tional aqueous surfactant cleanup fluids may cause damage,which will hamper oil production if an emulsion block formsin the wellbore due to water saturation15-17. With the advent ofmicroemulsion technology, nonreactive aqueous treatment flu-ids can be customized to achieve a relatively more effectivewell cleanup. Microemulsions are thermodynamically stabilized

multicomponent fluids composed of oil, water and surfactantblends, which solubilize the oil component of the OBM withlimited mechanical agitation18-24. Since acid-free micro-emul-sion fluids are incapable of dissolving OBM solid particles, it iscritical that dispersed residual filter cake solids are able to flowthrough the sand screen completion apertures when used instand-alone screen completions. Additionally, the mechanicalaspect of the displacement process must be optimal for maxi-mum removal of fluid solids in the wellbore, with final brinereturns having a solids/sediments content < 1% or fluid claritybelow 300 nephelometric turbidity units (NTUs)25.

Reservoir OBM DIFs and Spacer Fluids Design Options

The predominant development oil reservoir in the field selectedfor the cleanup fluid trials is relatively heterogeneous with awide variation of permeability (0.25 to 6 D) across the targetpay zone section, located at a shallow total vertical subseadepth of <5,500 ft. The reservoir is a thick sequence of uncon-solidated sandstone with siltstone, shale and limestone in-terbeds. Formation fluid is composed of medium light crudeand relatively saline formation water with a maximum bot-tom-hole static temperature (BHST) of ~160 °F. The well later-als were drilled with a relatively low density, invert emulsionOBM (75 pcf to 80 pcf, 70/30 oil/water ratio (OWR)) andcompleted as open hole horizontal wells with 5½” inflow con-trol devices (ICDs)/sand screens and production equalizers in-stalled in the 8½” lateral section (4½” ICDs/sand screens andproduction equalizers were used in the 61⁄8” laterals forslim/sidetracked wells). The CaCO3 loading required toachieve the desired mud weight was approximately 120 lb/bbl,Table 1. Previous laboratory investigation of field muds for


Additive Unit Conc. Property Unit Value

Mineral Oil bbl 0.52 Density lb/ft3 ~75

Emulsifi er gal 1.5 Plastic Viscosity cp 18-20

Lime lb 6.0 Yield Point lb/100 ft2 20-25

Filtration Control lb 6.0-8.0 10 sec. Gel lb/100 ft2 4-6

Water bbl 0.22 10 min. Gel lb/100 ft2 8-12

Organophilic Clay/Viscosifi er lb 6.0-8.0 Filtrate, HP/HT ml/30 min 1-2

Organic Surfactant gal 0.5 Electric Stability volts >800

CaCl2 (78%) lb 41 Chlorides mg/l ±350,000

CaCO3 (fi ne) lb 90 Excess Lime lb/bbl 4-6

CaCO3 (medium) lb 30.0 Oil/Water Ratio 70/30

Table 1. Composition and properties of OBM DIF

62884araD3R1_ASC026 3/15/13 11:25 PM Page 2

their role in DIF induced formation damage had detected per-meability reductions of 25% to 65% after mud exposure tocore samples, with higher alterations recorded for higher per-meability cores. Improvements observed after physical mudcake removal and core spinning down suggested that mud cakewas the primary barrier to flow, while higher density (~90 pcf)muds caused additional alteration in permeability26.

Following traditional practice, the newly drilled wells werecirculated using a solids-free version of the same OBM formu-lated with a higher density base brine (~90 ppb CaCl2) to facil-itate the installation of the sand screen/completion linerassembly on the bottom. Some of the wells were subsequentlyleft untreated for weeks and brought onstream only after pro-duction hookup facilities were installed. With the rig on-site,other wells were treated with breaker fluids, which resulted insevere losses and difficult well control situations. When there isa high risk of severe losses with breaker fluids, nonreactiveaqueous spacer fluids are recommended to displace the DIFsfrom the well. A combination of chemical and mechanical actions by the spacer fluid system is required to achieve mini-mum damage in extended horizontal wells during cleanup27, 28.Criteria that effective spacer fluids must achieve in a water-based spacer and completion formulation are:

• Effective displacement of the OBM.

• No excessive losses during different displacement stages.

• Thinning and weakening of the mud cake by solubi-lization of the oil from the OBM and filter cake into thespacer fluid, and wettability reversal (to water-wet) forbetter mud cake dispersion and easier lift-off duringproduction.

The aqueous spacer fluids train options considered included:

• Dispersant base oil, viscous push/gel pill, wash/surfactant pill (3-spacer fluids train).

• Viscous push pill, viscous push/gel pill, brine spacer,surfactant/solvent wash pill (4-spacer fluids train).

• Dispersant base oil, viscous push/gel pill, brine spacer,wash/surfactant pill, solvent pill (5-spacer fluids train).

Following a decision to test an acid-free microemulsionspacer fluid (MSF) system, the 4-spacer fluids train system con-taining a surfactant/solvent wash pill was selected. The compo-sition and properties of the spacer train are given in Table 2.The proposed nonionic surfactants used in the above spacersystem were reported to be insensitive to temperature andsalinity.

Fluids Hydraulics and Spacer Displacement Modeling

Wellbore fluids displacement efficiency is essentially deter-mined by the hydrodynamic properties of the OBM and thecleanup fluids, in addition to the chemical interaction of theDIFs, completion fluids and formation fluids. Wellbore fluidhydraulics analysis software was used to evaluate the fluid-


fluid displacement behavior at expected downhole conditionsand determine optimum cleanup fluid performance. The soft-ware applied the well geometry, fluids density and rheologydata to generate different fluids flow/interface profiles at spe-cific pump rates. Previous industry experience had identifiedthe need for contrasts between the mechanical properties of thefluid being displaced and those of the displacement fluid to enhance the wellbore fluid’s clean out29, 30.

A base case model was developed using a spacer fluid sys-tem, i.e., a base oil, a weighted/viscous spacer (push pill) and alow weight cleaning/wash pill, which was a blend of brine andsurfactants, Figs. 1a and 1b. Two sets of simulations were con-ducted to optimize the spacer train design parameters, such asdensity, rheology, fluid volume and contact times. This was required to determine which spacer train displacement processdemonstrated the most displacement efficiency. The two sets ofsimulations also tested the sensitivity of the wellbore fluid dis-placement performance to the physical properties (density andrheology) of the key spacer (push pill) and the volume/contacttime of the component spacers. Table 3 describes the variedparameters for the different case scenarios.

The simulation results reflected displacement performance

Spacer-1

(76 pcf Weighted/Viscous Surfactant Spacer):

Mix Water + 22 ppb Viscosifi er Additive + 88 ppb Barite + 2.75 gals/bbl Surfactant Additive + 0.36 gal/bbl Co-Surfactant Additive + Defoamer

Spacer-2

(75 pcf Gel Spacer):

Mix Water + 0.04 ppb Specialty Additive + 0.8 gal/bbl Gel Additive + Defoamer (as needed)

Spacer-3 (75 pcf Brine Spacer)

Spacer-4

(62 pcf Solvent/Brine Wash Fluid):

Mix 75 pcf Brine + 40% by vol. Solvent Additive

Table 2. Spacer fluids formulation

Simulation Case (Viscous Push Pill) Density Rheology (PV/YP)

Base Case 90 pcf 25 cp/60 lb/100 ft2

Sensitivity Cases

Case-1 90 pcf 42 cp/96 lb/100 ft2

Case-2 80 pcf 34 cp/52 lb/100 ft2

Table 3. Fluids displacement simulation variables

62884araD3R1_ASC026 3/15/13 11:25 PM Page 3


for each scenario in terms of “fluid concentrations” and “riskof mud left on the wall” snapshots. The absence of visible improvement with higher rheology spacers (Sensitivity Case 1)and the significantly poorer mud removal observed at lowerdensity (r= 80 pcf) (Sensitivity Case 2) indicated that the den-sity difference is a more dominant factor than the rheology dif-ference, Figs. 2a and 2b. The second set of simulation resultsalso showed that increasing the volume of the high densitypush pill relative to that of the wash/cleaning pill gave im-provement in the cleanup. It was noted that the key spacerfluid/push pill was unable to remove bulk mud from the nar-row side of the open hole section in all cases at a poor pipestandoff of �50%. These simulation results were instrumentalin altering the push pill density to 90 pcf, which led to im-proved performance in subsequent spacer fluid applications.

EXPERIMENTAL STUDIES

HP/HT Filter Press and Rheology Tests

A fluid loss performance test carried out with a HP/HT filterpress on the field OBM DIFs indicated a minimal fluid loss atstatic conditions with a 35-micron ceramic disc at 140 °F (to-tal filtrate volume ~5.0 ml after 60 minutes), Fig. 3. Table 4shows the rheology for the laboratory OBM, field OBM andkey spacers, with the field mud showing higher rheological valuesdue to the additional solids accumulated during the drilling

Fig. 1a. Base case flow profile (push pill displacement). Fig. 1b. Base case flow profile (wash pill displacement).

Fig. 2a. Sensitivity Case-2 flow profile (Push pill displacement). Fig. 2b. Sensitivity Case-2 flow profile (Wash pill displacement).

Fig. 3. OBM DIF filtrate vs. square root of time.

OBM/Spacer Fluid RPM Readings PV cp

YP lb/100

ft2

600 300 200 100

Field OBM 119 74 55 32 45 29

Lab OBM 97 60 52 30 37 23

Push Pill 114 78 63 48 36 42

Gel Pill 73 58 51 43 15 43

Table 4. OBM and spacer fluids rheology

62884araD3R1_ASC026 3/15/13 11:25 PM Page 4

process. The push pill designed in this work showed a favor-able yield point (YP) in contrast with the conditioned DIF(similar to the lab DIF) and field OBM before commencementof the cleanup operation. The YP value of the key displacingfluid (push pill) was approximately 1.5 times the YP for thedisplaced OBM (laboratory and field), as recommended by Javora and Adkins30.

The dispersion effect of the surfactant/solvent wash pill onthe OBM was evaluated by measuring the change in the emul-sion stability and rheology of the OBM when it was mixedwith different volumes of the wash pill. This change in emul-sion stability and rheology was measured using an electricalstability meter and a viscometer, respectively. Figure 4 showsthe increased reduction in electrical stability achieved by in-creasing the mixing ratio of the surfactant spacer with theOBM. At around 12 wt% of wash pill added to the OBM, areduction of 90% in emulsion stability was measured. This re-duction is an indication of how well the wash pill was dispers-ing the OBM and reversing the wettability to more water-wet.A complete dispersion of the mud components in the wash pillwas accomplished at a concentration of 20 wt%.

Figure 5 shows the change in viscometer reading that wascaused by the addition of 10% vol/vol of the wash pill to theOBM at speeds ranging from 100 rpm to 600 rpm. The

microemulsion surfactant wash fluid reduced the OBM rheol-ogy by 30% to 60%. Measurement of the rheology of theOBM and spacer fluid mixtures was required to determine thefluid’s behavior at the mixing zone/interface during wellboredisplacements. The test also enabled performance comparisonof different surfactants or surfactant concentrations on specificOBM DIFs.

Compatibility/Wettability and Interfacial Tension Tests

A bottle test was performed to confirm the ability of the sur-factant/wash pill to water-wet the OBM particles. Tests thatsimulated the OBM/surfactant solution interaction were pre-pared with an OBM/solution ratio of 10/90 that was left tosoak overnight at ~120 °F. Visual observation of solid particledispersion, with none of the particles sticking on the glass,gave an indication of the cleaning effectiveness. Mud particleswere fully dispersed and water wetted for the mixed solution,Photos 1a and 1b. See-through cell tests were also carried outto assess the compatibility of the solvent additive with theOBM DIF base oil by observing the mixed fluids at differentratios of 25/75, 50/50 and 75/25, Photos 2a and 2b. Similarcompatibility tests were carried out between the solvent andthe base brine, Photos 3a and 3b. No precipitation or emulsiondroplets were observed for the different fluids at bottom-holeconditions, i.e., a circulating pressure of 1,000 psi and a tem-perature of 120 °F. A Winsor Type III middle-phase micro-emulsion was also confirmed after mixing the OBM with a surfactant/solvent wash pill, Photo 4.

An inter-facial tension (IFT) test was conducted on the sur-factant based wash pill/OBM fluid system, using the spinningdrop method for measuring ultra-low IFTs to determine the ef-fectiveness of the surfactant solutions in solubilizing the oil inthe aqueous surfactant based solution and in water wetting the


Photos 1a and 1b: OBM DIF sample before and after surfactants at 120 °F.

Fig. 5. Conventional/Microemulsion surfactant effect on OBM rheology.

Fig. 4. Surfactant effect on OBM electrical stability.

Photos 2a and 2b: Compatibility test of solvent pill with OBM base oil at 120 °Fand 1,000 psi.

62884araD3R1_ASC026 3/15/13 11:25 PM Page 5

creo


OBM filter cake. This test followed from the established factthat cleaning of oil and oily dirt from solid surfaces with sur-factant solutions is largely dependent on ultra-low IFTs (<< 1µN/m = 1 dyne/cm) between the immiscible fluids. Table 5shows two different surfactant/solvent solutions that gave rela-tively low IFTs with the OBM at 70 °C (158 °F), i.e., 0.160and 0.078 dynes/cm as against the ~48 dynes/cm expected fora typical water/oil fluid interface. Also, the surfactant/solventsolution was completely haze-free, indicating salinity toleranceat the test temperatures.

Performance of Cleanup Flush/Circulation Treatment

To study the ability of the spacer train to thin and weaken thefilter cake while maintaining minimum fluid losses during thewellbore clean out, a filter press test was conducted on thecleanup spacers using a synthetic ceramic disc of the perme-ability range, 35.0 µm, (equivalent to 10 Darcies) and OBMDIFs at expected reservoir conditions. OBM filter cake wasprepared by circulating the mud for 30 minutes at an expectedoverpressure of 500 psi and a bottom-hole circulating tempera-ture of 140 °F, followed by 3 hours of static conditions. Thespacer fluids were circulated sequentially, one after the other,

on top of the filter cake, with dynamic conditions at 350 psiand 140 °F. Filtrate volume was monitored during the circula-tion of each spacer, and the total fluid leak off (TFL) after thecirculation treatment was recorded. The thickness and weightof the mud cake were also recorded before and after thecleanup flush treatment, and the percent filter cake reduction(FCR) was computed.

It was observed that the solvent wash pill altered the wetta-bility of the mud cake and OBM particles, changing from oil-wet to water-wet after circulation treatment. Also, it wasshown that the wash pill thinned the mud cake and reduced itsweight, Photos 5a and 5b. The results showed a maximumTFL < 30 ml (~20% of treatment fluid) and a FCR of ~10% to20% with optimized spacer fluid formulations after repeatedtests at expected operating conditions, Table 6.

Coreflood Tests

Coreflood tests were conducted to determine the return perme-ability using different spacer trains in a dynamic fluid loss in-strument with two test cells. The tests were conducted at athird-party laboratory facility using these procedures:

• Base Permeability Measurement: Cores were loaded intothe test cells, and the flow of mineral oil was initiated inthe production direction to obtain initial corepermeability at 150 °F.

• Dynamic Fluid Loss Measurement: Mud was loadedinto the system, and the pump was started at apredetermined shear rate that matched the wellboreflow conditions. Differential pressure across the coreswas 350 psi while system temperature was maintainedat 150 °F, with fluid loss lines opened for 4 hours.

• Static Fluid Loss Measurement (pump shutdown): Themud differential pressure across the core was reduced to

Photos 3a and 3b: Compatibility test of solvent pill with 67 pcf NaCl completionbrine at 120 °F and 1,000 psi.

Photos 5a and 5b. OBM sample mud cake and after cleanup flush with solventspacer at 120 °F.

Photo 4. Confirmation of Winsor Type III microemulsion using surfactantsolution with field OBM sample.

Fluid Interface IFT Measurement

Water: OBM 48

Solvent/Wash Pill-A: OBM 0.160

Solvent/Wash Pill-B: OBM 0.078

Table 5. Results of IFT tests at 70 °C

62884araD3R1_ASC026 3/15/13 11:25 PM Page 6

250 psi while the system temperature was increased to150 °F, with fluid loss lines opened for 2 hours.

• Cleanup Flush/Circulation Treatment: Two differentcleanup spacer fluids trains were circulated with thedifferential pressure across the two cores maintained at350 psi.

• Final Permeability Measurement: Mineral oil was againinitiated in the production direction at the same bottom-hole conditions used for the base permeabilitymeasurement above.

The proportional retained permeability computed for thetwo spacer fluids trains enabled the selection of the superiorsurfactant/solvent wash formulation with acceptable retainedpermeability (> 70%), Fig. 6. The selected spacer train wascom posed of nonreactive components, i.e., nonionic surfac-tant, gel pill, sodium chloride (NaCl) completion brine and solvent pill.

FIELD APPLICATION AND CASE HISTORIES

Test Well-1

The well was originally drilled and completed as a deviatedcased hole/perforated completion across the target reservoir (7”casing was cemented from total depth to the surface) in 1984.The well was subsequently sidetracked using a 75 pcf dieseloil-based DIF and thereafter completed with a 4½” sand screenand ICDs on the bottom after sidetracking and cementing a4½” casing off the bottom inside a 7” open hole in July 2009.

The two-stage cleanup wash with a 4-spacer fluids train wascarried out as planned in August 2009, Table 7. The post-com-pletion production test indicated a production increase of 10%(5% water cut) compared to offset wells in the area. Well per-formance was better, with a 60% higher production rate com-

pared to an offset well that had experienced severe fluid lossesduring breaker fluid treatments at a similar well completionstage, with those losses controlled using killing fluid, Table 8.

Test Well-2

The dual horizontal well was drilled with 75 pcf to 80 pcf min-eral oil-based DIFs and completed with a 5½” ICD/sandscreen in the lower lateral and a 4½” ICD/sand screen in theupper lateral in July 2009. The 3,440 ft lower lateral wastreated with 200 bbl of a reactive microemulsion/mesophasefluid system due to the unavailability of the spacer fluid addi-tives. The treatment fluid was formulated with NaCl brine/10% acetic acid and nonionic surfactant additive (displacedand spotted in open hole with 125 bbl of 70 pcf NaCl brine).

The 3,300 ft upper lateral cleanup was carried out usingacid-free MSFs in two stages with NaCl brine as the displace-ment fluid in July 2009, Table 7. The initial displacement ratewas limited at <1.2 bpm with maximum pressure at 700 psiduring treatment of the upper lateral to avoid prematurepacker setting. The post-completion production test indicateda 157% (0% water cut) production rate when compared withthe offset well. Well performance was better than that of theoffset wells that had encountered severe fluid losses while be-ing treated with breaker fluids during completion, Table 8.

Test Well-3

The last test well had a hole configuration and completion design similar to that of test Well-2, but both laterals werecleaned out with the microemulsion fluid system in August2009. A two-stage cleanup wash with a 4-spacer fluids systemwas carried out prior to completion brine displacement andcirculation in both laterals. For the 61⁄8” upper lateral (~2,540ft), initial displacement was maintained at <1 bpm with maxi-mum pressure at 800 psi to avoid premature packer setting.Similarly, the initial displacement was kept below 5 bpm forthe lower lateral (~3,180 ft), Fig. 7.

Brine samples were collected on the surface after the first-stage and second-stage cleanup followed by displacement brineto assess the performance of the well cleanup operation. Exten-sive analysis of the brine returns after more than 200% holevolume displacement indicated adequate removal of the solidsor sediments contained in the wellbore (less than 0.3% solids


Spacer W1, g W2, g W3, g FCR% Ti, mm Tf, mm Reduced Cake Thickness, %

Solvent Pill 45.99* 56.36 54.451 18.41 9.57 8.89 7.11

Surfactant 52.581 57.97 57.491 8.89 9.28 9.04 2.56

Cleanup Flushes 53.743 62.281 60.520 20.63 10.03 9.38 6.48

*10 micron ceramic disk

Table 6. Results of the filter press tests

Fig. 6. Retained permeability vs. pore volume of cleanup treatment fluid.

62884araD3R1_ASC026 3/15/13 11:25 PM Page 7


Test Well-1 Test Well-2 Test Well-3

Upper Lateral Upper Lateral Lower Lateral

Stage-1

Weighted Spacer

Gel Spacer

Brine Spacer

Solvent Pill

60 bbl

60 bbl

60 bbl

45 bbl

60 bbl

60 bbl

60 bbl

35 bbl

60 bbl

60 bbl

60 bbl

35 bbl

60 bbl

60 bbl

60 bbl

40 bbl

Displacement Brine (75 pcf) NA 350 bbl 390 bbl 380 bbl

Gel Spacer* NA 70 bbl 70 bbl 140 bbl

Stage-2

Weighted Spacer

Gel Spacer

Brine Spacer

Solvent Pill

30 bbl

30 bbl

30 bbl

35 bbl

30 bbl

30 bbl

53 bbl

45 bbl

30 bbl

30 bbl

40 bbl

40 bbl

30 bbl

30 bbl

67 bbl

35 bbl

Displacement Brine**(2-3 hole volumes until clean returns)

75 pcf CaCl2 Brine 75 pcf NaCl Brine 75 pcf NaCl Brine 75 pcf NaCl Brine

*Spotted in open hole prior to stinging out of the sand screen PBR**Displacement after setting production packer

Table 7. OBM spacer fluids pump sequence and volumes

Test #1 Offset #1 Offset #2

Date

*Prod Rate %

Water Cut %

Feb. 2010

110

5.0

Aug. 2007

41

59.1

Jan. 2002

100

36.3

Test #2 Offset #3

Date

*Prod Rate %

Water Cut %

June 2010

157

0

Feb. 2008

100

4.3

Test #3

Same offset well with Test #2 well above

Date

*Prod Rate %

Water Cut %

June 2010

145

0

*Compared to offset wells with acid cleanup and severe losses

Table 8. Well production performance of test well and offset well

62884araD3R1_ASC026 3/15/13 11:25 PM Page 8


content was recorded for the test Well-3 upper lateral), Photos6a and 6b. The post-completion production test indicated aproduction rate of 145% (0% water cut) compared to thesame offset wells used for the test Well-2 assessment. The wellperformance was appraised as better than that of the offsetwells that had breaker fluids treatment while encountering severe losses at completion, Table 8.

CONCLUSIONS

1. Reactive mud cake breaker fluids are incapable of effec-tively removing OBM filter cake in long open hole horizon-tal wells located across high permeability sandstone reser-voirs without inducing severe fluid losses and emulsion induced formation damage as a result of the OBM, comple-tion and formation fluids mixing together.

2. A two-stage circulation treatment with acid-free MSFs has been proven effective in facilitating open hole sandstone wellbore cleanup by altering the wettability of the oily filter cake and mud particles without completely removing the filter cake and so inducing fluid losses that need to be con-trolled with more damaging materials.

3. It is recommended to evaluate the probability and potential risk of severe losses with breaker fluid application to the filter cake by reviewing the completion and cleanup fluid performance in offset wells prior to using the acid-free MSFs.

4. The surfactant/solvent fluids were effective in dispersing and water-wetting the OBM DIFs. The OBM base oil and formation brine were found to be compatible with the sur-factant/solvent pills as no precipitation or emulsion was ob-served at bottom-hole conditions. The generation of a WinsorType III middle-phase microemulsion was confirmed.

ACKNOWLEDGMENTS

The authors would like to thank Saudi Aramco managementfor the permission to present and publish this article. Wewould also like to thank all the members of the formationdamage and stimulation laboratory for their support towardsthe success of the laboratory work and field trials. We also acknowledge the technical support of members of the DrillingFluids and Cement Unit and Saleh M. Ammari at the time ofthe project.


REFERENCES

1. Giard-Blanchard, C., Audibert-Hayet, A. and Dalmazzone,C.: “Development and Application of Surfactant-BasedSystems for Treatment of Wells Drilled with OBM,” SPEpaper 68960, presented at the SPE European FormationDamage Conference, The Hague, The Netherlands, May21-22, 2001.

2. Chambers, M., Hebert, D.B. and Shuchart, C.E.:“Successful Application of Oil-based Drilling Fluids inSubsea Horizontal, Gravel-Packed Wells in West Africa,”SPE paper 58743, presented at the SPE InternationalSymposium on Formation Damage Control, Lafayette,Louisiana, February 23-24, 2000.

3. Brady, M.E., Bradbury, A.J., Sehgal, G., Brand, F., Ali,S.A., Bennett, C.L., et al.: “Filter Cake Cleanup in OpenHole Gravel-Packed Completions: A Necessity or Myth?”SPE paper 63232, presented at the SPE Annual TechnicalConference and Exhibition, Dallas, Texas, October 1-4,2000.

4. Bin Moqbil, K.H., Al-Otaibi, M.A., Al-Faifi, M.G., Al-Khudair, W.S. and Al-Aamri, A.D.: “Cleanup of Oil-BasedMud Filter Cake Using an In-Situ Acid Generator Systemby a Single Treatment,” SPE paper 126065, presented atthe SPE Saudi Arabia Section Technical Symposium andExhibition, al-Khobar, Saudi Arabia, May 9-11, 2009.

5. Kim, L.S., Ravitz, R., Patel, A., Martens, H., Luyster, M.and Kuck, M.: “Specially Formulated Delayed-BreakerSystem for Extended Reach Injectors in a Viscous OilWaterflood Application,” SPE paper 144136, presented atthe SPE European Formation Damage Conference,Noordwijk, The Netherlands, June 7-10, 2011.

6. Luyster, M., Patel, A. and Ali, S.: “Development of aDelayed-Chelating Cleanup Technique for Open HoleGravel Pack Horizontal Completion Using a ReversibleInvert Emulsion Drilling System,” SPE paper 98242,presented at the International Symposium and Exhibitionon Formation Damage Control, Lafayette, Louisiana,February 15-17, 2006.

Fig. 7. Pump and displacement brine data for lower lateral in test Well-3.

Photos 6a and 6b. Displacement brine returns after first-stage treatment and aftersecond-stage treatment.

62884araD3R1_ASC026 3/15/13 11:25 PM Page 9


7. Ali, S., Luyster, M. and Patel, A.: “Dual Purpose ReversibleReservoir Drill-in Fluid Provides the Perfect Solution forDrilling and Completion Efficiency of a Reservoir,” SPEpaper 104110, presented at the SPE/IADC Indian DrillingTechnology Conference and Exhibition, Mumbai, India,October 16-18, 2006.

8. Infra, M., Coronado, M.P., Woudwijk, R., Al-Mumen,A.A. and Al-Baggal, Z.A.: “New Inflow Control DeviceProvides Solid-Liner Functionality Throughout Installationand Fluid Loss Control During Completion,” SPE paper134576, presented at the SPE Annual Technical Conferenceand Exhibition, Florence, Italy, September 19-22, 2010.

9. Abiodun, A., Nwabueze, V., Opusunju, A. and Sibigem, F.:“Successful Application of Mud Cake Pop-Off Techniquein Horizontal Well Cleanup – Case Histories,” SPE paper82277, presented at the SPE European Formation DamageConference, The Hague, The Netherlands, May 13-14,2003.

10. Ding, Y., Longeron, D., Renard, G. and Audibert-Hayet, A.: “Modeling of Near-Wellbore Damage Removal by Natural Cleanup in Horizontal Open Hole Completed Wells,” SPE paper 68951, presented at the SPE European Formation Damage Conference, The Hague, The Netherlands, May 21-22, 2001.

11. Lohne, A., Han, L., van der Zwaag, C., van Velzen, H., Mathisen, A.M., Twyman, A., et al.: “Formation Damageand Well Productivity Simulation,” SPE paper 12224, presented at the European Formation Damage Conference,The Hague, The Netherlands, May 27-29, 2009.

12. Davis, E.R., Beardmore, D., Burton, R., Hedges, J., Hodge, R., Martens, H., et al.: “Laboratory Testing and Well Productivity Assessment of Drill-in Fluid Systems in Order to Determine the Optimum Mud System for Alaskan Heavy Oil Multilateral Field Developments,” SPE paper 96830, presented at the SPE Annual Technical Conference and Exhibition, Dallas, Texas, October 9-12, 2005.

13. Burton, B. and Hodge, R.: “Comparison of Inflow Performance and Reliability of Open Hole Gravel Packs and Open Hole Stand-alone Screen Completion,” SPE paper 135294, presented at the SPE Annual Technical Conference and Exhibition, Florence, Italy, September 19-22, 2010.

14. Brown, S.V. and Smith, P.S.: “Mud Cake Cleanup to Enhance Productivity of High-Angle Wells,” SPE paper 27350, presented at the SPE Formation Damage Control Symposium, Lafayette, Louisiana, February 7-10, 1994.

15. Goode, D.L. and Stacy, A.L.: “Aqueous-Based Fluids for Perforating and Oil Phase Mud Removal,” SPE paper 12901, presented at the SPE Rocky Mountain Regional Meeting, Casper, Wyoming, May 21-23, 1984.

16. Berry, S.L. and Beal, B.B.: “Laboratory Development and

Application of a Synthetic Oil/Surfactant System for Cleanup of OB and SBM Filter Cakes,” SPE paper 97857,presented at the International Symposium and Exhibition on Formation Damage Control, Lafayette, Louisiana, February 15-17, 2006.

17. Hutton, A., Vickers, S., Davidson, M., Wharton, I., Hatch, A., Simmonds, R., et al.: “Design and Application of Invert Emulsion Drilling and Aqueous Completion Fluids for Long Horizontal Multilateral Wells,” SPE paper 121905, presented at the European Formation Damage Conference, Scheveningen, The Netherlands, May 27-29, 2009.

18. Lavoix, F., Leschi, P., Aubry, E., Quintero, L., Le Prat, X. and Jones, T.: “First Application of Novel MicroemulsionTechnology for Sand Control Remediation Operations – A Successful Case History from the Rosa Field, a Deep Water Development Project in Angola,” SPE paper 107341, presented at the European Formation Damage Conference, Scheveningen, The Netherlands, May 27-29, 2007.

19. Berry, S.L.: “Optimization of Synthetic-Based and Oil-Based Mud Displacements with an Emulsion-Based Displacement Spacer System,” SPE paper 95273, presented at the SPE Annual Technical Conference and Exhibition, Dallas, Texas, October 9-12, 2005.

20. Bustin, B., Phillips, J., Al-Otabi, M., BinMoqbil, K.H., Abou Zeid, S., Christian, C.F., et al.: “Improved WellboreCleanup – Successful Case Histories in Saudi Arabia fromDevelopment to Field Implementation,” SPE paper 120801, presented at the SPE Saudi Arabia Section Technical Symposium and Exhibition, al-Khobar, Saudi Arabia, May 10-12, 2008.

21. Ruwaily, A.A., Phillips, J.E., Ben Saad, Z.R. and Christian, C.F.: “Microwash Treatment Case History,” SPE paper 119591, presented at the SPE Middle East Oil and Gas Show and Conference, Manama, Bahrain, March15-18, 2009.

22. Otaibi, M.A., BinMoqbil, K.H., Al-Rabba, A.S. and Abitrabi, A.N.: “Single-Stage Chemical Treatment for Oil-Based Mud Cake Cleanup: Lab Studies and Field Case,” SPE paper 127795, presented at the SPE International Symposium and Exhibition on Formation Damage Control, Lafayette, Louisiana, February 10-12, 2010.

23. van Zanten, R. and Ezzat, D.: “Surfactant Nano-technology Offers New Method for Removing Oil-Based Residue to Achieve Fast, Effective Wellbore Cleaning and Remediation,” SPE paper 127884, presented at the SPE International Symposium and Exhibition on Formation Damage Control, Lafayette, Louisiana, February 10-12, 2010.

24. Quintero, L., Jones, T.A. and Pietrangeli, G.: “Phase Boundaries of Microemulsion Systems Help to Increase

62884araD3R1_ASC026 3/15/13 11:25 PM Page 10


BIOGRAPHIES

Peter I. Osode is a Petroleum EngineerSpecialist with the Formation Damageand Stimulation Unit in SaudiAramco’s Advanced Technical ServicesDivision.

He has over 30 years of diverseupstream industry experience spanning

wellsite petroleum engineering operations, productiontechnology (well and reservoir management, productionoptimization and production chemistry) and drilling andcompletion fluids management. Peter started his careerwith Baroid/Halliburton as a Technical Sales Engineerbefore moving to Shell Petroleum Development Companyin Nigeria and Shell International’s affiliate-PetroleumDevelopment Oman (PDO) in Oman. He has participatedin a number of Shell Global E&P Well PerformanceImprovement projects and was the subject matter expert ondrilling fluids performance assessment process prior tojoining Saudi Aramco in 2009.

Peter received his B.S. degree with honors in PetroleumEngineering from the University of Ibadan, Ibadan,Nigeria.

He is an active member of the Society of PetroleumEngineers (SPE) International and has authored a numberof published technical papers. Peter is currently involved information damage evaluation of reservoir drilling andcompletion fluids.

Msalli Al-Otaibi joined Saudi Aramcoin 2005 and began working with theFormation Damage and Stimulationunit of the Exploration and PetroleumEngineering Center Advanced ResearchCenter (EXPEC ARC) as a PetroleumEngineer. His work experience includes

formation damage evaluation and prevention strategies forexploration drilling, reservoir development and waterinjection projects in addition to impaired well diagnosisand remedial treatments.

Msalli was a principal member of the focused teamtasked with promoting innovation in Saudi Aramcothrough the development and launching of the firstInnovation Tournament (InTo) in 2010. He has been anactive member in the Society of Petroleum Engineers (SPE)by publishing seven technical papers and leading the YoungProfessionals (YP) and Students Outreach committee of theSPE-Saudi Arabia Section (SAS) for 2010/2011. Also,Msalli served as the 2010/2011 SPE-SAS representative onthe North Africa and Middle East (MENA) YP committee.

He received his B.S. degree in Chemical Engineeringfrom Louisiana State University, Baton Rouge, LA, in2005. In 2011, Msalli received his M.S. degree in ChemicalEngineering from KFUPM. He is currently pursuing hisPh.D. degree in Petroleum Engineering at the ColoradoSchool of Mines, Golden, CO.

Productivity,” SPE paper 144209, presented at the SPE European Formation Damage Conference, Noordwijk, The Netherlands, June 7-10, 2011.

25. Berg, E., Sedberg, S., Kaarigstad, H., Omland, T.H. and Svanes, K.: “Displacement of Drilling Fluids and Cased-Hole Cleaning: What is Sufficient Cleaning,” SPE paper 99104, presented at the IADC/SPE Drilling Conference, Miami, Florida, February 21-23, 2006.

26. Shahri, A.M., Kilany, K., Hembling, D., Lauritzen, J.E., Gottumukkala, V., Ogunyemi, O., et al.: “Best CleanupPractices for an Offshore Sandstone Reservoir with ICD Completions in Horizontal Wells,” SPE paper 120651, presented at the SPE Middle East Oil and Gas Show and Conference, Manama, Bahrain, March 15-18, 2009.

27. Al-Yami, A.S. and Nasr-El-Din, H.A.: “Completion FluidsChallenges in Maximum Reservoir Contact Wells,” SPE paper 121638, presented at the SPE International Symposium on Oil Field Chemistry, The Woodlands, Texas, April 20-22, 2009.

28. Davison, J.M., Jones, M., Shuchart, C.E. and Gerard, C.: “Oil-Based Muds for Reservoir Drilling: Their Performance and Cleanup Characteristics,” SPE paper 58798, presented at the SPE International Symposium and Exhibition on Formation Damage Control, Lafayette,Louisiana, February 23-24, 2000.

29. Ladva, H.K.J., Brady, M.E., Sehgal, P., Kelkar, S., Cerasi, P., Daccord, G., et al.: “Use of Oil-Based Reservoir Drilling Fluids in Open Hole Horizontal Gravel Packed Completions: Damage Mechanisms and How to Avoid Them,” SPE paper 68959, presented at the SPE EuropeanFormation Damage Conference, The Hague, The Netherlands, May 21-22, 2001.

30. Javora, P.A. and Adkins, M.: “Optimizing the Displacement Design – Mud to Brine,” SPE paper 144212, presented at the SPE European Formation Damage Conference, Noordwijk, The Netherlands, June 7-10, 2011.

62884araD3R1_ASC026 3/15/13 11:25 PM Page 11

Khalid H. Bin Moqbil started hispetroleum engineering career in SaudiAramco’s Exploration and PetroleumEngineering Center – AdvancedResearch Center (EXPEC ARC) in2005. His area of interests includestudies in formation damage related

aspects of reservoir drilling, completion and wellstimulation fluids in addition to water injection studies.

Khalid is currently working with the Gas ReservoirManagement Department where he is involved with gasproduction optimization and reservoir managementprojects.

In 2005, Khalid received his B.S. degree in ChemicalEngineering and in 2011, he received his M.S. degree inPetroleum Engineering along with a graduate certificate inSmart Oil Field Completions, all from the University ofSouthern California, Los Angeles, California.

He is an active member of the Society of PetroleumEngineers (SPE) and has authored and coauthored severalSPE technical papers.

Khaled A. Kilany has over 25 years ofindustry experience while working as aReservoir and Production Engineer. Hestarted his career in the oil fields as aProduction Engineer working from1986 to 1990, and then Khaledswitched to reservoir engineering,

working as a Reservoir Simulation and ReservoirManagement Specialist in several international companiesin Egypt, Canada and the Gulf area, including AGIP inEgypt, the Kuwait Oil Company and Shell International inCanada and Oman prior to coming to Saudi Aramco.

Since joining Saudi Aramco in August 2005, Khaled hasworked as a Senior Reservoir Engineer with the NorthernArea Reservoir Management Department where he wasinvolved in introducing innovative completion equipmentand production optimization techniques in Safaniya field.Khaled’s experience here includes his participation inseveral reserve assessment studies, short- and long-termproduction forecasts, waterflood management and full fielddevelopment plans. He currently leads a sub-team of theManifa Incremental Project Team that is tasked with thelargest ongoing offshore incremental development projectin the company.

In 1982 Khaled received his B.S. degree in PetroleumEngineering from Cairo University, Giza, Egypt.

Eddy Azizi has over 17 years ofexperience that consolidates his currentposition as Senior Production Engineerwithin the multidisciplinary NorthernArea Production Engineering team inSaudi Aramco. He has worked in bothoffshore and onshore environments at

both Shell International and Saudi Aramco. Eddy startedhis career in the oil field as a Process Engineer for 2 years,and then worked as a Well Site Drilling/CompletionEngineer for 2 years and one year as a Well ServicesSupervisor in the field. He later worked as a ProductionTechnologist and/or Production Engineer for the next 12years with involvement in several field developmentassessment studies/plan, short- and long-term productionforecasts, sand management, production systemmodleing/nodal analysis and ESP operations andunconventional oil production systems.

Eddy has been involved in a number of new productionoptimization initiatives, which has resulted in improvedstimulation fluid placement, zero flaring, and completionintegrity management in addition to reduced coil tubingutilization in Safaniya while he currently leads the WellIntegrity team working on the Qatif field.

Eddy received his B.S. degree (First class honors) inChemical Engineering from London University, London,U.K., in 1995.


62884araD3R1_ASC026 3/15/13 11:25 PM Page 12


ABSTRACT enhance productivity. The unconventional gas reservoirs willnot produce unless a long, conductive fracture is in place. Asuccessful treatment can improve well production and sustain-ability; however, fracturing is never optimum, and the industryhas been continuously working at different fronts to obtain im-proved results. Some of these fronts include a complete changein the fluid system so that the gel damage is reduced and thefluids pumped can be flowed back efficiently. The gel breakingagent is therefore an important ingredient, and the rightamount and type of that agent are some of the key factors thathelp minimize proppant conductivity damage from gel residue.This article focuses more on the gel and polymer systems andthe breaking agents.

On the other hand, the proppant system and pump schedulehave also been revised. The change is intended to providelonger and more conductive fractures and to reduce chances ofpremature screen-outs.

This article also deals with many important issues and illus-trates the gradual progression in the application of high-endtechnology, thereby improving overall hydraulic fracturingtreatments to achieve sustained production rates. Many exam-ples in this article confirm the successes achieved through theidentification of problems and a follow-up with remedial actions.

FRACTURE FLUIDS CHEMISTRY

Water-based fracturing fluids are the most common types andare widely used. The viscosity is obtained by mixing 20-70 lbof guar polymer, or its derivative, per 1,000 gal of water. Thismixture is known as the base gel and typically provides 30-50cP viscosity at surface conditions1-3.

Developed around 1968, cross-linked agents were added tolinear gels, resulting in a complex, high-viscosity fracturingfluid that provides higher proppant transport performancethan do linear gels. Cross-linking also reduces the need forfluid thickener and extends the viscous life of the fluid. Thefracturing fluid remains viscous until a breaking agent is intro-duced to break the cross-linker, and eventually, the polymer.Although cross-linkers make the fluid more expensive, theycan improve hydraulic fracturing performance considerably.

When a gel is cross-linked, the viscosity can increase on the order of 100 times or more. The base gel and polymer

Hydraulic fracturing is required to commercially produce low tomoderate permeability gas reservoirs. The selection of fracturingfluids, additives and proppant types is a major component whendesigning and implementing a hydraulic fracturing treatment.A viscous, unbroken fracture fluid that remains after the treat-ment compounds the effects of fracture face skin and causes severe deterioration to proppant conductivity. With the ad-vancement of technology, many novel fracture fluid systemsare now available in the industry with reduced polymer con-centration to preserve reservoir and proppant integrity. Theadvantages of these fluids include less formation damage,lower cost and reduced treatment pressure. Subsequent to thefracture operation, an aggressive breaker treatment is oftennecessary to effectively clean up the fracture and restore prop-pant conductivity. Proppant conductivity plays a tremendousrole in the post-fracture production enhancement, and anydamage left from the fluids can impair well potential consider-ably. Similarly, the correct choice of proppant, based on the rockstrength, reservoir fluid properties, expected production rate,pressure and temperature, is important. Proppant type andscheduling determine the ultimate propped fracture geometrythat controls the gas flow from the reservoir to the wellbore.

The application of new technologies in combination withbetter job design is ongoing to obtain improved results in the deepsandstone reservoirs of Saudi Arabia. In the process of opti-mization, fluids along with their gel type, polymer concentrationand additives have been modified and changed to provide betterresults. Similarly, proppant size, type and scheduling have beenoptimized. Different types of aqueous-based fracturing fluidswith various polymer loadings, as well as hybrid systems andviscoelastic surfactant (VES) fluids for deep and high tempera-ture reservoirs are currently in use. Several case studies providedin this article demonstrate how the critical fracturing parametershave progressed with time, been customized and can now bemade to fit the reservoir conditions to make a noticeable impacton well productivity and recovery.

INTRODUCTION

The primary purpose of hydraulic fracturing treatment is to

Selecting Optimal Fracture Fluids, BreakerSystem and Proppant Type for SuccessfulHydraulic Fracturing and Enhanced GasProduction – Case StudiesAuthors: Dr. Zillur Rahim, Dr. Hamoud A. Al-Anazi and Adnan A. Al-Kanaan

62884araD4R1_ASC026 3/15/13 11:27 PM Page 1


cross-linker are the two most critical elements to consider in afracturing fluid.

The most frequent base gel and cross-linker used in the in-dustry are hydroxyethyl cellulose (HEC) and hydroxypropylguar (HPG), respectively. Many different products are nowavailable to address fracturing of unconventional reservoirs,and the fluids often selected are highly dependent on the per-meability of the formation, the reservoir fluid and the reservoirpressure of the candidate well.

Table 1 provides some of the components of fracture fluidscommonly used and their respective functions during the treat-ment. The chemical components of some of the additives aregiven in Table 2.

As the fracturing fluid is pumped at high rates, it creates aninduced fracture. The proppant is transported along with thefluid. A certain portion of the fluid leaks off into the formationwhile a cake of concentrated guar polymer is formed at thefracture face. This cake can be like an elastic membrane, and ifnot removed, can severely damage the reservoir flow capacity.

Many major aspects need to be considered when selectingthe most appropriate fluid to ensure good results after the fracture

treatment. Most importantly, the fluid must be compatiblewith the reservoir pressure-volume-temperature (PVT) proper-ties so that there remains absolutely no chance of creating for-mation damage. Other aspects to consider include how to:

• Create sufficient width to the fracture.

• Provide enough viscosity to transport proppant at thedesigned concentration.

• Resist pressure and shear degradation as the treatmentprogresses.

• Provide lower friction loss to reduce injection pressure.

• Include sufficient additives to control vital fluidproperties such as pH level and viscosity.

• Provide adequate and effective breaker systems to breakthe polymer gel once treatment is complete and thefracture has closed.

• Control fluid loss and provide optimum fracturegeometry.

• Provide higher regained permeability of the proppantpack.

Use of viscoelastic surfactant (VES) fluids, a polymer-free,low surface tension system, is a good option in fracturingtreatment4. These fluids use surfactants with inorganic salts tocreate an ordered structure resulting in increased viscosity andelasticity5. The initial two shortcomings inherent with otherfluids, i.e., low viscosity at high temperatures due to thermalthinning and the lack of internal breaking mechanisms, aresupposedly overcome by the new fluid formulation. These flu-ids are insensitive to salinity, are compatible with N2 and CO2,and do not require clay control agents5. The fluid does notform filter cake; therefore, there is no plugging of the forma-tion, and the post-treatment retained permeability is high5. Ba-sically, VES uses surfactants in combination with inorganic salts tocreate ordered structures, which eventually results in increasedviscosity and elasticity. The fluids tend to be shear degradablebut can transport proppant with lower loading and withoutthe comparable viscosity requirements of conventional fluids.

SHEAR AND TEMPERATURE TOLERANCE OF FRACTURE FLUIDS

One of the most critical aspects of selecting a fracture fluid isto ensure that the viscous characteristics are maintained untilthe treatment is over and the fracture has closed. Many frac-ture fluid systems are affected by pressure and temperature,Fig. 16. Traditionally, complex fracture fluid systems with highgel and polymer loading and a high concentration of additiveshave been used for treatments in complex reservoirs, such as ahigh-pressure/high temperature environment. A number offluid additives are used, resulting in a complex chemistry thatmust be kept in a tight range to ensure quality fluid perform-ance. Some current new fluids have been tested in the laboratory

Frac Fluid Additives Functions

Breaker Breaks gel and polymer after treatment and fracture closure.

Cross-linked Agent Maintains fl uid viscosity as temperature increases.

Base Gel Thickens the water in order to suspend the sand.

Iron Control Agent Prevents precipitation of metal oxides.

KCl Creates a brine carrier fl uid.

Proppant Is main fracturing component providing conductivity.

SurfactantMinimizes formation damage, leaves no residue, reduces friction.

Table 1. Main fracture fluid additives and their functions

Frac Fluid Additives Chemical Components

Gelling Agent Hydroxyethyl Cellulose/ Hydroxypropyl Guar

Proppant Quartz Sand, Ceramics

Cross-linker Borate Salt

Breaker Ammonium Persulfate

Surfactant Isopropanol

Table 2. Fracture fluid chemistry

62884araD4R1_ASC026 3/15/13 11:27 PM Page 2

first and then used in the actual fracturing treatments of a fewSaudi Arabian gas wells. Although there is no single, all-pur-pose fluid fit for any application, and every well must be evalu-ated for selecting the most appropriate fluid characteristics, thenewer fluids with fewer additives and a dual cross-linker (bo-rate and zirconate) system have proven good results. Typically,the borate cross-linkers are shear tolerant but are affected bytemperature. In contrast, the zirconate cross-linkers are tem-perature resistant but shear degradable. The laboratory test re-sults for different fluid specifications are illustrated in Fig. 2.The dual cross-linker system therefore is considered to be anappropriate type of fluid to use.

GEL DAMAGE AND TRAPPED FLUIDS

The gel and the cross-linked fluids pumped during fracturetreatment cannot be fully recovered during production. In dif-ferent field studies, it has been found that 60% to 80% of thepumped fluids can be recovered over a long period of time.The amount of fluids recovered decreases and the recoverytime increases, in low permeability, tight reservoirs. Palmer de-scribed a “check valve” effect where the width of the fracturedecreases after treatment and does not allow larger size poly-mers to flow back7. Also, during injection, the hydraulic gradi-ent is higher and carries the polymers farther away. During theflow back, the hydraulic gradient is much lower and does notgenerate sufficient force for the fluids to be produced back,therefore, the need for polymer breakers to reduce the injectedfluid viscosity is as low as possible.

In conventional reservoirs, the gel damage is compensatedfor by the reservoir permeability and increased apparent well-bore radius due to hydraulic fracturing. In tight formations,and also in naturally fractured reservoirs (therefore in coalbedmethane), the effect of the gel damage is more severe.

BREAKERS

Using the appropriate type and volume of breakers is of primeimportance. Though a gel is needed to create the fracture andcarry the proppants, it also has to degrade and be producedback to leave a clean, high conductivity, propped fracture behind.

Breakers are usually mixed with the fracturing fluid duringpumping. Most breakers are typically acids, oxidizers or en-zymes. The breakers include ammonium persulfate, ammoniumsulfate, copper compounds and glycol. Types include time-release, shear-release or temperature dependent breakers4, 8.

Residue-after-break tests have shown that enzyme breakersleave fewer residues than oxidative breakers used at the sametemperature. Polymer degradation by enzymes continues for amuch longer time, so has a cleaner effect on the gel residue after a fracture treatment.

Breaker concentration is important for proper cleanup ofthe fracture. Improved well performance, indicated by higherflow rate and sustainability, has been observed when usinghigher than normal breaker concentrations. The results are related to achieving and maintaining higher proppant conduc-tivity as the magnitude of gel damage is reduced.

Basically, two types of breakers are used. The enzymes ofthe first one are mixed with the fracturing fluids at variousconcentrations as the job is being pumped. Introduction of thesecond one is delayed through encapsulation; the capsulesbreak and release their ingredients under certain temperatureand stress conditions, which typically happen post-treatmentwhen the fracture closes. Figure 3 presents some laboratory ex-periments on a certain breaker showing the effect of gel break-ing and the loss of viscosity as functions of concentration.

USE OF FRACTURE FLUIDS AND BREAKERS IN SAUDIARABIAN WELLS

Since the inception of hydraulic fracturing, many different types offracturing fluids have been used in the sandstone gas formations

Fig. 1. Effects of pressure and temperature on fracture fluid viscosity.

Fig. 2. Effect of shear degradation on fracture fluid types4.


62884araD4R1_ASC026 3/15/13 11:27 PM Page 3

in Saudi Arabia9. The fluid volume size, gel loading and addi-tives are customized to fit the needs of a particular field andreservoir. The fluid quality and type have also advanced duringthis time, and new fluids systems are progressively being used.

Figure 4 shows a comparison chart of the average percent-age of basic fracture fluid additives (x-axis) used in 2011 com-pared to 2008 in a few of the Saudi Arabian wells. Other thandifferences in some of the fluid chemistry, it is noticeable howthe quantity of some of the ingredients has increased over time.The pH control and bactericide are used to maintain the in-tegrity of the fluids and provide compatibility with the forma-tion. The cross-linker concentration was increased to providebetter proppant carrying capacity and generate a larger frac-ture width. The breaking agent in particular has increased bymore than 60%, indicating the importance of ensuring a cleanfracture after the treatment and a quick flow back of the de-graded gel. Figure 5, which shows the breaker-to-gel ratio usedin the treatment of about 100 wells analyzed since 2000, illus-trates the trend toward using increased breaker concentrations.This change in the fracturing program is due to the fact that ahigher concentration of breaking agent is conducive in achiev-ing cleaner fractures, thereby leading to higher productivitywells. The field results confirmed the benefits of using a higherbreaker amount, so the trend continues. The gel loading didnot change, Fig. 6, showing that the proppant transport andfracture dimensions were being achieved as per expectation. In

fact, attempts have been made to decrease the gel loading with-out compromising fracture quality so as to incur less damageto the proppant and formation.

Figure 7 presents the use of different breaker types and theirrespective quantities as a function of the total gel volume. Thechoice and use of both oxidative and encapsulated breakers,along with their specific activation characteristics, are impor-tant to cover the range of temperature between the cooleddown fracture during the treatment and the reservoir tempera-ture. Therefore, the proper mix of low temperature and hightemperature (LT and HT) breaking agents ensures that thebreaking of gel initiates when the fracture closes and is rela-tively cool, and continues for a prolonged period as the frac-ture eventually attains reservoir temperature.

EXAMPLE WELLS

The effects of breaking down the gel are seen in results fromtwo recent vertical wells where additional breakers werepumped after it was realized that the post-treatment productionrates were not up to the expectation based on open hole log dataand rates from some of the offset wells. The inflow performance

Fig. 3. Effect of gel concentration on fracture fluid viscosity6.

Fig. 4. Change of additive quantity between 2008 and 2011.

Fig. 5. Breaker-to-gel ratio in gas wells between 2000 and 2012.

Fig. 6. Normalized gel loading showing a constant trend.

Fig. 7. Breaker-to-gel ratio for different breaker types.


62884araD4R1_ASC026 3/15/13 11:27 PM Page 4

cern, appropriate measures should be taken, as was done withthe two wells illustrated here. The production rate before andafter the breaker treatments for Well-B, Fig. 10, shows that thewell stabilized at 18 MMscfd at a high flowing wellhead pres-sure after the treatment. The reservoir lithology, porosity andsaturation of the two wells are comparable.

ADDITIONAL BREAKER TREATMENT

The polymer removal breaker treatment used in the two wellsconsisted of a high temperature, high concentration, water sol-uble breaker system. The purpose was to efficiently break anyresidual gel in the proppant pack and also leak off into the for-mation and clean up the matrix during flow back. Amongother components, a bactericide and pH buffer were added,along with a corrosion inhibitor to preserve fluid integrity andmake it compatible with the reservoir as well as with the tubu-lars. Nitrogen was added to enhance flow back. In addition,methanol was pumped to treat any water blockage in the fracture and formation. Removal of water blockage should beconsidered because it can hinder gas flow significantly10.

PROPPANT OPTIMIZATION

Good proppant selection is an integral part of successful hy-draulic fracturing. Among the different types of proppantsavailable, the major ones used in Saudi Aramco are the light-weight ceramics and the intermediate/high density ceramics,some of which are resin coated proppants (RCP). RCP is rou-tinely used as a tail-end in the pumping treatment to preventproppant flow back, and this process has been working verywell. The main criteria of proppant selection depend on theconductivity requirement at downhole conditions. The evalua-tion is usually done based on the contrast between the flow capacity of the fracture and the reservoir, known as the dimen-sionless fracture conductivity, FCD

=kmLf

kf Wf————— .

Selection criteria are also based on reservoir pressure andtemperature, embedment, multiphase flow and crushing. Othervery important aspects to take into account while selecting theproppant are the flow convergence effects, particularly intransverse fractures, non-Darcy flow, gel damage, and nonopti-mal proppant concentration and placement, as well as reducedconductivity due to fines migration and pressure cycling.

Maintaining a high conductivity fracture has always provento be a preferred option since it overcomes many of the abovementioned problems that can reduce gas production rate. Aproppant type that shows high conductivity at higher stress inthe laboratory, however, can fall short in the field, failing tomaintain that level of conductivity due to non-Darcy effects orflow convergence11. The non-Darcy flow permeability, whichis the effective permeability, kF, can be computed from thelaminar flow equation by relating Darcy permeability, kD, withthe flow turbulence expressed by Reynold’s number, NRE, usingthe equation: kF

= kD1+NRe——————— . Therefore, the higher the Reynold’s

curves from Well-A and Well-B presented in Figs. 8 and 9, re-spectively, clearly show the improved rates from both wells,where the increase of absolute open flow ratio ranged from25% to over 100%. The measured rate and pressure are plot-ted on the graph. The improvement varies, depending on theinitial treatment schedule and what was pumped in terms ofgel loading and breaker quality. The optimum procedure is totake into consideration all damage and cleanup possibilities soas to optimize the fluids pumped during the treatment. Thatway, added intervention in the well is avoided, saving time andadditional expenditure. Consequently, post-frac productionanalysis must be conducted on all wells, and if there is a con-

Fig. 8. Before and after breaker treatment IPR, Well-A.

Fig. 9. Before and after breaker treatment IPR, Well-B.

Fig. 10. Flow rate and pressure before and after breaker treatment, Well-B.


62884araD4R1_ASC026 3/15/13 11:27 PM Page 5

the post-fracture expected rate is not achieved, although frac-ture treatment was pumped as designed. When this happens,well performance has been compromised because of gel residuein the fracture and suboptimal cleanup. The following aresome of the key points drawn from this article.

• Efficient polymer breaker treatment contributes tohigher well productivity. Low polymer loading and/oran adequate amount of breaker is necessary for acomplete post-fracture cleanup.

• Methanol or similar surface tension reducing agentsshould be routinely used to minimize water blockageeffects. Wells have shown significant improvement aftersuch treatments are conducted.

• Additional pressure drop due to non-Darcy flow or flowconvergence phenomenon occurring in transversefracture geometry may be significant and will contributeto low well productivity.

• The gas production rate is also adversely affected by thepresence of condensate.

• The use of high strength and high conductivityproppant, without risking embedment or crushing, isessential to maintain good connectivity between the welland the formation.

• Higher proppant loading is equally effective to providesufficient fracture width, which is directly related to theultimate conductivity of the created fracture.

ACKNOWLEDGMENTS

The authors would like to thank Saudi Aramco managementfor the permission to present and publish this article.

This article was presented at the SPE Unconventional GasConference and Exhibition, Muskat, Oman, January 28-30,2013.

REFERENCES

1. Economides, M.J. and Nolte, K.G.: Reservoir Stimulation,3rd edition, New York: John Wiley and Sons, 2000, p. 818.

2. Gall, B.L. and Raible, C.J.: “Molecular Size Studies ofDegraded Fracturing Fluid Polymers,” SPE paper 13566,presented at the SPE Oil Field and Geothermal ChemistrySymposium, Phoenix, Arizona, April 9-11, 1985.

3. Langedijk, R.A., Al-Naabi S., Al-Lawati H., Pongratz, R.,Elia, M.P. and Abdulrab, T.: “Optimization of HydraulicFracturing in a Deep, Multilayered, Gas-CondensateReservoir,” SPE paper 63109, presented at the SPE AnnualTechnical Conference and Exhibition, Dallas, Texas,October 1-4, 2000.

4. Courtesy of Schlumberger.

5. Gupta, S.: “Unconventional Fracturing Fluids: What,

number (high flow rate), the lower the effective permeabilitywill be. In the proppant fracturing jobs performed in SaudiArabian gas wells, ensuring an effective conductivity of morethan 3,000 md-ft has become the norm. Even though in thetight reservoirs this number seems to be high, the higher con-ductivity helps maintain a long-term rate in reservoirs wherecondensate dropout becomes a challenge.

Some examples of non-Darcy-related rate loss as a functionof some specific reservoir properties in Saudi Arabian gas fieldsare provided in Figs. 11 and 12. The rate loss is pronounced inhigh permeability wells due to their high flow rates, and thisnumber can be significant. For a fracture conductivity of 1,000md-ft, the rate loss in a 5 md reservoir can be as much as 35%,whereas there will be no loss in a 0.1 md reservoir. Even in a 1md reservoir, the loss will be negligible; therefore, the selectionof proppant type and concentration should be based on reser-voir flow capacity. If proppant crushing and embedment con-ditions are met, high permeability proppants are alwayspreferred, pumped at high concentration so as to achieve significant propped fracture width at fracture closure.

CONCLUSIONS

Hydraulic fracturing is a necessary technique to improve gasproduction from tight or conventional reservoirs. Many times

Fig. 11. Normalized rate loss as a function of fracture conductivity for differentreservoir permeabilities.

Fig. 12. Normalized gas rate illustrating both laminar and non-Darcy effects fordifferent reservoir permeabilities.


62884araD4R1_ASC026 3/15/13 11:27 PM Page 6

Where, and Why,” Baker Hughes, Tomball TechnologyCenter, Tomball, Texas.

6. England, K.W. and Parris, M.D.: “The UnexpectedRheological Behavior of Borate Cross-linked Fluid,” SPEpaper 140400, presented at the SPE Hydraulic FracturingTechnology Conference, The Woodlands, Texas, January24-26, 2011.

7. Palmer, I.D., Frayar, R.T., Tumino, K.A. and Puri, R.:“Comparison between Gel-Fracture and Water-FractureStimulations in Black Warrior Basin,” SPE paper 23415,presented at the Coalbed Methane Symposium, Tuscaloosa,Alabama, May 13-16, 1991.

8. Courtesy of Halliburton.

9. “2009-2011 Gas Program,” Saudi Aramco Gas ReservoirManagement Division internal documentation.

10. Holditch, S.A.: “Factors Affecting Water Blocking and Gas Flow from Hydraulically Fractured Gas Wells,” Journal of Petroleum Technology, Vol. 31, No. 12, December 1979, pp. 1,515-1,524.

11. Gidley, J.L.: “A Method for Correcting Dimensionless Fracture Conductivity for non-Darcy Flow Effects,” SPE Production Engineering, Vol. 6, No. 4, November 1991, pp. 391-394.

BIOGRAPHIES

Dr. Zillur Rahim is a PetroleumEngineering Consultant with SaudiAramco’s Gas Reservoir ManagementDepartment (GRMD). He heads theteam responsible for stimulationdesign, application and assessment forGRMD. Rahim’s expertise includes

well stimulation, pressure transient test analysis, gas fielddevelopment, planning, production enhancement, andreservoir management. Prior to joining Saudi Aramco, heworked as a Senior Reservoir Engineer with Holditch &Associates, Inc., and later with Schlumberger ReservoirTechnologies in College Station, TX, where he used toconsult on reservoir engineering, well stimulation, reservoirsimulation, and tight gas qualification for national andinternational companies. Rahim is an Instructor ofpetroleum engineering industry courses and has trainedengineers from the U.S. and overseas. He developedanalytical and numerical models to history match andforecast production and pressure behavior in gas reservoirs.Rahim developed 3D hydraulic fracture propagation andproppant transport simulators and numerical models tocompute acid reaction, penetration, and fractureconductivity during matrix acid and acid fracturingtreatments.

Rahim has authored 65 Society of Petroleum Engineers(SPE) papers and numerous in-house technical documents.He is a member of SPE and a technical editor for theJournal of Petroleum Science and Engineering (JPSE).Rahim is a registered Professional Engineer in the State ofTexas and a mentor for Saudi Aramco’s TechnologistDevelopment Program (TDP). He is an instructor of theReservoir Stimulation and Hydraulic Fracturing course forthe Upstream Professional Development Center (UPDC) ofSaudi Aramco. Rahim is a member of GRMD’s technicalcommittee responsible for the assessment and approval ofnew technologies.

Rahim received his B.S. degree from the InstitutAlgerien du Petrole, Boumerdes, Algeria, and his M.S. andPh.D. degrees from Texas A&M University, CollegeStation, TX, all in Petroleum Engineering.


62884araD4R1_ASC026 3/15/13 11:27 PM Page 7

Adnan A. Al-Kanaan is the Manager ofthe Gas Reservoir ManagementDepartment (GRMD) where heoversees three gas reservoir manage-ment divisions. Reporting to the ChiefPetroleum Engineer, Adnan is directlyresponsible for making strategic

decisions to enhance and sustain gas delivery to theKingdom to meet its ever increasing energy demand. Heoversees the operating and business plans of GRMD, newtechnologies and initiatives, unconventional gas develop-ment programs, and the overall work, planning anddecisions made by his more than 70 engineers andtechnologists.


Prior to the inception of GRMD, he was the GeneralSupervisor for the Gas Reservoir Management Divisionunder the Southern Reservoir Management Department for3 years, heading one of the most challenging programs inoptimizing and managing nonassociated gas fields in SaudiAramco.

Adnan started his career at the Saudi Shell PetrochemicalCompany as a Senior Process Engineer. He then joinedSaudi Aramco in 1997 and was an integral part of thetechnical team responsible for the on-time initiation of thetwo major Hawiyah and Haradh Gas Plants that currentlyprocess more than 6 billion cubic feet (bcf) of gas per day.Adnan also directly managed the Karan and Wasit fields —two major offshore gas increment projects — with anexpected total production capacity of 4.3 bcf of gas per day.

He actively participates in the Society of PetroleumEngineers’ (SPE) forums and conferences and has been thekeynote speaker and panelist for many such programs.Adnan’s areas of interest include reservoir engineering, welltest analysis, simulation modeling, reservoir charac-terization, hydraulic fracturing, reservoir developmentplanning and reservoir management.



Dr. Hamoud A. Al-Anaziis the GeneralSupervisor of the North Ghawar GasReservoir Management Division in theGas Reservoir ManagementDepartment (GRMD). He oversees allwork related to the development andmanagement of huge gas fields like

Ain-Dar, Shedgum and ‘Uthmaniyah. Hamoud also headsthe technical committee that is responsible for all newtechnology assessments and approvals for GRMD. Hejoined Saudi Aramco in 1994 as a Research Scientist in theResearch & Development Center and moved to theExploration and Petroleum Engineering Center – AdvancedResearch Center (EXPEC ARC) in 2006. After completinga one-year assignment with the Southern Area ReservoirManagement Department, Hamoud joined the GasReservoir Management Division and was assigned tosupervise the SDGM/UTMN Unit and more recently theHWYH Unit. With his team he successfully implementedthe deepening strategy of key wells that resulted in a newdiscovery of the Unayzah reservoir in UTMN field and theaddition of Jauf reserves in the HWYH gas field.

Hamoud’s areas of interests include studies of formationdamage, stimulation and fracturing, fluid flow in porousmedia and gas condensate reservoirs. He has publishedmore than 50 technical papers at local/internationalconferences and in refereed journals. Hamoud is an activemember of the Society of Petroleum Engineers (SPE) wherehe serves on several committees for SPE technicalconferences. He is also teaching courses at King FahdUniversity of Petroleum and Minerals (KFUPM), Dhahran,Saudi Arabia, as part of the Part-Time Teaching Program.

In 1994, Hamoud received his B.S. degree in ChemicalEngineering from KFUPM, and in 1999 and 2003,respectively, he received his M.S. and Ph.D. degrees inPetroleum Engineering, both from the University of Texasat Austin, Austin, TX.


62884araD4R1_ASC026 3/15/13 11:27 PM Page 8


ABSTRACT Hydraulic fracturing is required in tight multilayered reser-voirs for increased oil and gas recovery. Effective wellborecompartmentalization by means of open hole packers, espe-cially in low and nonuniform permeability reservoirs, is key tosuccessful multistage stimulation operations. It is, therefore,important to describe and compare the modes of operation ofstimulation systems and the effects of the various downholeconditions on the main open hole packer designs available toour industry today.

Since the beginning of 2007, a total of 40 wells in the tightgas fields of Saudi Aramco’s Southern Area have been com-pleted with open hole multistage stimulation systems. Targetformations have spanned the Khuff B and C carbonates andthe pre-Khuff (Unayzah) sandstone reservoir. Hole sizes haveincluded both 8⅜” and 5⅞”, and the number of stages per wellhas been as high as seven. Figures 1 to 6 show more detailsabout the 40 wells where open hole multistage stimulation sys-tems were run by the various technology suppliers. Out of the 40wells:

The increasing demand for oil and gas resources to supportworldwide development plans means that the petroleum indus-try is always actively engaged in exploring new frontiers indrilling and production, including tight multilayered reservoirs.It is becoming evident more than ever that producing the mostoil and gas out of drilled reservoirs is an absolute necessity. Accordingly, completion techniques have presented themselvesas a crucial well construction parameter, one that is key to optimally producing wells.

Several completion techniques have been exhaustively trialtested in Saudi Aramco to determine the most successful com-pletion mode for each reservoir. Among those various tech-niques, open hole multistage stimulation has demonstratedsuperior performance in minimizing skin damage and maxi-mizing reservoir contact through efficient propagation of fracture networks within the rock matrix.

Overall, the production results from wells completed usingopen hole multistage stimulation systems — as deployed in thetight gas fields of Saudi Arabia — have been very positive. Var-ious open hole multistage completion systems were run overapproximately 40 wells. While production results varied wherethis new technology was utilized, the majority of the wellshave met or exceeded the pre-stimulation expectations for gasproduction.

This study highlights these systems and discusses their impacton wells during the fracturing operation and the final stabi-lized production. This study will also present some case studiesin multistage fracturing operations and investigate the opera-tional impact of such operations on productivity enhancement.With correct implementation, the findings from this studyshould increase the probability of having a more successfulmultistage stimulation job from a productivity standpoint.

INTRODUCTION

While the well trajectory planning, reservoir characterizationand completion design are important determinants of well pro-ductivity, open hole multistage stimulation completion hasdemonstrated that it can have significant effects on long-termstabilized production and reservoir draining efficiency.

Assessment of Multistage StimulationTechnologies as Deployed in the Tight GasFields of Saudi Arabia

Authors: Mohammed A. Al-Ghazal, Saad M. Al-Driweesh, Fadel A. Al-Ghurairi, Abdulaziz M. Al-Sagr andMustafa R. Al-Zaid

Fig. 1. Number of wells completed per supplier.

Fig. 2. Number of MSS wells completed per year per supplier.

62884araD5R1_ASC026 3/15/13 11:29 PM Page 1


• 35 wells (87.5%) have exceeded the target productiongas rates.

• 32 wells (80%) have exceeded the target rates with

enough flowing wellhead pressure (FWHP) to connectdirectly to the trunk line.

• Two wells (13%) have exceeded the target rates, butcould not be connected to the trunk line due toinsufficient FWHP.

• Five wells (12.5%) could not meet the expected post-stimulation target rates.

DEFINITION OF AN UNBALANCED AND BALANCEDSYSTEM

An unbalanced open hole multistage completion system designmeans that the lowest stage in the completion is open at thebottom to allow fracturing out of the toe stage, Fig. 7. This isopposed to a balanced system where the first stage stimulationzone is between two packers, Fig. 8.

COMPARISON OF OPEN HOLE PACKERS

Inflatable Packers

Often referred to as external casing packers (ECPs), thesepackers are normally constructed of base pipe similar to thecompletion casing/liner/tubing, Fig. 9. The construction is suchthat the packer element is mechanically fixed to the outside diameter (OD) of the base pipe at both ends, leaving an annu-lus between the pipe’s OD and the element’s inside diameter(ID). The base pipe would normally have a valve system thatwould open at a predetermined pressure to allow tubing fluid tofill the annulus and “inflate” the element. The valve systemwould then trip closed at another predetermined higher pressureto lock the fluid inside the element and retain the post-inflationelement dimensions and seal against the wellbore. Inherent de-sign limitations of these packers (i.e., their very low differentialpressure capabilities) have discounted their use in open holemultistage system applications.

Fig. 3. Percentage success in reaching target depth per supplier.

Fig. 4. Percentage success in opening frac sleeves per supplier.

Fig. 5. Percentage success in packer zonal isolation per supplier.

Fig. 6. Percentage success in ball seat milling operations.

Fig. 7. Unbalanced system with the hydraulic fracture sleeve (stage 1) at thebottom of the lower completion.

Fig. 8. Balanced system with the hydraulic fracture sleeve (stage 1) between twopackers.

62884araD5R1_ASC026 3/15/13 11:29 PM Page 2

Swell Packers

Often referred to as swellable element packers (SEPs) and/orreactive element packers (REPs), these packers are also con-structed of a base pipe similar to the completion liner/tubing,Figs. 10a and 10b. With this application, specific rubber ismolded, thermally cured and glued to the base pipe. Sometimesbackup rings are integrated into the design to keep the rubber

element in place. Swell packers are typically between 10 ft and32 ft long, and they feature various grades and types of rubber.Upfront design and planning is required so that the properpacker is selected for each operation (job specific design).

Three factors dictate the downhole performance of swellpackers:

• Bottom-hole temperature (the most determinant factor,as temperature variations could be crucial).

• Wellbore fluid type (completion, stimulation andproduction fluids).

• Ratio of base pipe OD to wellbore ID.

Significant pre-job data should be collected for each well-bore section. Once the necessary information is gathered, it ispossible to estimate the packer dimensions (base pipe OD andelement thickness) as well as the swell period required toachieve the desired pressure rating.

As soon as the element comes into contact with its corre-sponding fluid (water or hydrocarbon), it begins to swell.Therefore, to avoid premature swelling, retardant chemicalsare normally mixed in the rubber recipe or otherwise appliedto the element OD, depending on the swell packer suppliercompany.

The swell process is a function of time, temperature andfluid type, so these crucial factors must be carefully observedduring job design and execution.

In swell packers such as that provided by Supplier C, the re-tardant chemical is applied to the outside of the SEP. This typi-cally creates a huge risk when running in the open hole, as it ispossible that the retardant chemical could be removed orscraped off, and premature swelling could occur. The optimumswell packer to use is one where the retardant chemical ismixed in the rubber, so the possibility of its removal and pre-mature swelling does not occur. Another disadvantage of Sup-plier C’s swell packer is the 32 ft length with a 5.60” OD,which makes deployment a major issue when running severalstages in the well and heightens the risk of not reaching the tar-get depth due to mechanical or differential sticking issues. Theshorter the length and the smaller the OD, the better, when selecting swell packers from the deployment standpoint.

The time to swell could range from hours to weeks depend-ing on well conditions, element design and swell packer sup-plier company.

MECHANICAL PACKERS

Hydraulically set mechanical open hole packers use rubberpack off elements, which are compressed when set to form aseal between the completion and the open hole, Fig. 11. A suc-cessful packer design used in the Saudi Aramco Southern Areagas fields is an open hole packer that features more than onerubber element.

The setting mechanism of this packer is characterized by adynamic setting mode that uses the fracture surface pumping

Fig. 9. Inflatable packer element.

Fig. 10a. Swellable packer element.

Fig. 10b. Swellable packer element.


62884araD5R1_ASC026 3/15/13 11:29 PM Page 3

pressure to continuously adjust the pack off force on each element to maintain sealing. When the packer is subject to apressure higher than its initial setting pressure, the ratchet willmove further and pack off the element — this not only copeswith borehole changes, but also increases the differential pres-sure rating due to the additional pack off force delivered to theelement with the increased hydrostatic pressure.

The open hole mechanical packer is approximately 5 ftlong, which makes it readily adaptable to high doglegs andbuild rates, facilitating easier reach to target depth.

COMPARISON OF THE EXTERNAL PRESSURESLEEVE/PORT TOOLS

The hydraulic fracture sleeve (HFS) provided by Supplier B(HFS-B; see yellow column in Fig. 12) has encountered prob-lems with opening on some operations to date. On one well, ittook over two days of pressure cycling using coiled tubing(CT) with jetting acid to finally shift it open. The port was setto 4,500 psi and finally opened at 7,474 psi, Fig. 13. Well-Bwas a similar case, and the sleeve took even longer to open, re-quiring an application of 8,000 psi, Fig. 14. Finally, on a thirdwell, the P-sleeve was cycled for three days, first to 7,100 psithrough the wellhead and second to 12,100 psi through a treesaver, and it still did not open.

Fig. 13. Treatment chart for first well: CT pressure cycling attempts. The port was sat to 4,500 psi and finally opened at 7,474 psi.

Fig. 14. Treatment chart for Well-B: Frac pumps attempted five times to open the HFS-B by bullheading. On the fifth attempt, it was opened at 8,000 psi and 4 bpm.

Fig. 11. Mechanical open hole packer.

Fig. 12. Summary graph showing the history of the hydraulic frac-port openingson all MSS operations.


62884araD5R1_ASC026 3/15/13 11:29 PM Page 4

On all Supplier A jobs, the HFS has opened immediately asplanned, except on one well where barite mud was used for thefirst time. It was concluded that barite should never be usedagain due to the potential problems of mud particulates plug-ging the wellbore pores and preventing injectivity into thereservoir rock. Despite the mud issues, the HFS-A on that onewell still opened after a short period of pressure cycling.

COMPARISON OF UNBALANCED AND BALANCEDLOWER COMPLETIONS

In two key wells, large pressure drops were seen when pump-ing into the first stage. In the first well, a drop was seen during

the first injection step rate test (SRT), from 10,100 psi to8,400 psi, Fig. 15. In the second well, a similar drop in surfacepressure was observed following spotting acid/mutual acidduring the main treatment; here there was a drop of 5,254 psisurface pressure from an initial 10,800 psi, Fig. 16.

When pumping commenced into the second stage for bothwells, it was very clear that there was communication betweenzones and that the packers were likely no longer holding pres-sure. As seen in Figs. 17 and 18, there was an immediate pres-sure decline to 0 psi surface pressure when the pumping wasstopped.

For the two initial gas well completion operations in 2007,the open hole multistage systems were all in a balanced config-

Fig. 15. Pressure drop seen on first well during injection test within water.

Fig. 16. Pressure drop seen on second well during main treatment.


62884araD5R1_ASC026 3/15/13 11:29 PM Page 5

uration. Due to the aforementioned mechanical and/or differ-ential sticking issues during deployment, where the completionwas unable to reach the target depth, it was decided that anunbalanced system was preferred. The theory was that if thelower multistage completion was unable to reach target depth,then the toe section of the well could still be treated.

One important consideration is that the open hole swellpackers or mechanical packers offer near negligible anchoringcapability. Testing performed in open hole conditions has

shown that it is possible to piston the packer uphole with cer-tain overpull, depending on various downhole conditions.Given the open hole diameter and the high pressures involvedduring the stimulation treatments, the upward forces createdthat are acting on the lowermost packer are very high: up to400,000 lb upward force on the lower packer, Figs. 19 and 20.With an unbalanced system and with high forces acting on thebottom packer, the completion will undergo a rapid upwardspistoning effect, and all of the lower completion will stroke a

Fig. 17. Communication between Stages 1 and 2 on first well.

Fig. 18. Communication between Stages 1 and 2 on second well.


62884araD5R1_ASC026 3/15/13 11:29 PM Page 6

significant distance uphole. This phenomenon is well provenand simply related to the forces resulting from the pressures.

For example, with an unbalanced system inside an 8⅜”hole, the piston area trying to push the packers up the hole is37.742”, resulting in an upwards force of 377,400 lb with10,000 psi differential pressure applied. In this case, the tubingshrinkage increases the force by approximately another 50,000lb; therefore the total upwards force is ~420,000 lb.

For a balanced system inside an 8⅜” hole, the piston areatrying to push the packers apart is 55.092”, so with 10,000 psidifferential pressure applied, there is over 550,000 lb of forcetrying to part the tubing. This is counteracted somewhat bytubing shrinkage due to the temperature drop, reducing theforce down to ~500,000 lb.

In a 5⅞” hole, the numbers are 27.112” for a balanced sys-tem and 17.492” for an unbalanced system, equaling 270,000lb (220,000 lb with shrinkage) and 175,000 lb (225,000 lbwith shrinkage), respectively.

Therefore, as shown in Table 1, the upward movement ofthe lowest open hole packer can be as much as 15 ft. When thepressure is released, the completion will slide back towards itsoriginal position. With every pressure cycle on this lowestzone, the upward force will be created again, resulting in com-pression of the liner. The implication is that with a set packersliding along the open hole rock face several times during thepressure cycles, it would be highly likely that the packer sealswould be damaged and thereby reduce its sealing ability, re-sulting in communication between zones. Initially, this wouldbe between Zones 1 and 2, but subsequent zones would alsobe involved as with movement of the entire completion, allopen hole packer seals would very likely be damaged.

CARBONATE AND SANDSTONE COMPLETION CONFIGURATIONS

For all formation types, a balanced system would be the pre-ferred method of running the multistage fracturing completion.This is simply because the first stage is in a balanced condition,and the forces created during the fracturing treatment areequally applied, in opposing directions, to each packer. Forcarbonate formations, the need for a balanced system is greatlyincreased because with an unbalanced system the potential riskof the acid treatment eroding away the formation around theopen hole packer is higher than in sandstone formations.

ADDITIONAL RECOMMENDATIONS FOR FUTURE MULTISTAGE STIMULATION OPERATIONS

Due to improved operational running procedures and the useof centralization of the liner, almost all of the recent systemshave reached target depth without issue1.

The recommendation for forthcoming wells has been tostandardize operations based on balanced systems. The ideabehind the design is to run a balanced multistage stimulationcompletion with a single joint above the circulation valve as-sembly (with float collars and guide shoe). Above the lowest

Well NameOpen Hole

SizeCompletion

Size ODCompletion

Size ID

Open Hole Annulus Area

(sq. in)

Bottom-hole Pressure (psi)

Reservoir Pressure (psi)

Well-A 8½” 5½” 4.7” 37.74” 16,200 6,600

Well-B 5½” 4½” 3.813” 17.49” 13,500 5,200

Differential Pressure (psi)

Force Created on Lowest

Packer (lbsf)

Shrinkage due to Temperature

Difference (lbsf)

Resulting Upwards Force Created (lbsf)

Friction Forces based on T&D Analysis (lbsf)

Completion Length (ft)

Resulting Liner Movement (ft) Uphole from

TD

9,600 362,304 50,000 412,304 80,000 3,760 8 ft

8,300 145,164 50,000 195,164 60,000 5,284 12 ft

3 8

7 8

Table 1. Liner upward movement resulting from applied forces on the lower packer when treating Stage 1 for an unbalanced system

Fig. 19. Area (shaded in yellow) where the force is applied to the lowermost openhole packer.

Fig. 20. Diagram showing the calculated forces acting on the lower open holepacker during the Stage 1 treatment in an unbalanced system.


62884araD5R1_ASC026 3/15/13 11:29 PM Page 7

packer will be the first stage hydraulic. In this way, allforces/movement will be balanced, and the open hole anchorpacker will be well set with a reduced chance of erosion from theacid treatment. Second, as an added precaution to eliminate anycommunication seen between Stage 1 and Stage 2, it is recom-mended to place an extra open hole packer between the stages.

CONCLUSIONS

This investigation is part of a more detailed report currentlybeing complied of evaluations performed on the multistagestimulation fracturing and completion efficiencies. High rateand high-pressure acid fracturing treatments have pushed thecompletion equipment to its limits, and there is much still to belearned on the interaction with carbonate formations. The welldirection and resulting fracture orientation is certainly a majorinfluence on the fluid placement. This investigation focusedsolely on the completion equipment set-up and configurations.

The completion that was run on the second well by SupplierB was an unbalanced system, and the completion was only an-chored at the top of the liner by the liner hanger. The lowestsingle sealing packer therefore began to slide immediatelywhen pressure was applied to it. A major pressure drop of ap-proximately 1,700 psi was seen immediately during the SRTwhen pumping water at approximately 8,300 psi differentialpressure (surface pressure less than reservoir pressure).

With 8,300 psi differential pressure and 145,164 lb of forceapplied to the lowest packer in the 5⅞” open hole, with 5,284ft of liner in total, the upwards movement can be as much as12 ft. This would potentially damage the packer seal andtherefore allow communication between zones.

The completion used on the first well saw a pressure dropoccur following three days of pumping, which included an acidtreatment designed to dissolve some of the barite mud away.Stage 1 of the completion system had been pressure cycledmany times up to its maximum differential of 9,600 psi by thetime the pressure drop was observed.

As a result of the deployment issues that led to a failure toreach the target depths and the port opening problems, Sup-plier C was placed on hold from future operations in February2011 and has not resumed multistage stimulation operations.

As a result of the hydraulic P-sleeve problems as well as CTmill-out problems, Supplier B was placed on hold from futureoperations in March 2011 and has yet to resume multistagestimulation operations.

For future wells, it is recommended to run a balanced sys-tem with an open hole packer at the bottom of the first stageprior to the hydraulic fracturing sleeve.

ACKNOWLEDGMENTS

The authors would like to thank Saudi Aramco managementfor the permission to present and publish this article. Furtherthanks are provided to the Saudi Aramco Multistage Fracturing

Team and the field crew for their outstanding work.This article was presented at the International Petroleum

Technology Conference, Beijing, China, March 26-28, 2013.

REFERENCE

1. Al-Jubran, H.H., Wilson, S. and Johnston, B.: “SuccessfulDeployment of Multistage Fracturing Systems inMultilayered Tight Gas Carbonate Formations in SaudiArabia,” SPE paper 130894, presented at the SPE DeepGas Conference and Exhibition, Manama, Bahrain,January 24-26, 2010.


BIOGRAPHIES

Mohammed A. Al-Ghazal is aProduction Engineer at Saudi Aramco.He is part of a team that is responsiblefor gas production optimization in theSouthern Area gas reserves of SaudiArabia. During Mohammed’s careerwith Saudi Aramco, he has led and

participated in several upstream projects, including pressurecontrol valve optimization, cathodic protection systemperformance, venturi meter calibration, new stimulationtechnologies, innovative wireline technology applications,upgrading fracturing strategies, petroleum computer-basedapplications enhancement and safety managementprocesses development.

In 2011, Mohammed assumed the position of GasProduction HSE Advisor in addition to his productionengineering duties. During his tenure as HSE Advisor, hefounded the People-Oriented HSE culture, which hasbrought impressive benefits to Saudi Arabia gas fields,resulting in improved operational performance.

In early 2012, Mohammed went on assignment with theSouthern Area Well Completion Operations Department,where he worked as a foreman leading a well completionsite in remote areas.

As a Production Engineer, Mohammed played a criticalrole in the first successful application of several high-endtechnologies to present new possibilities in the Kingdom’sgas reservoirs.

Mohammed’s areas of interest include formationdamage investigation and mitigation, coiled tubingapplications, wireline operations, matrix acidizing,hydraulic fracturing and organizational HSE performance.

In 2010, Mohammed received his B.S. degree withhonors in Petroleum Engineering from King FahdUniversity of Petroleum and Minerals (KFUPM), Dhahran,Saudi Arabia.

He has also authored and coauthored several Society ofPetroleum Engineers (SPE) papers and technical journalarticles as well as numerous in-house technical reports.Additionally, Mohammed served as a member on theindustry and student advisory board in the PetroleumEngineering Department of KFUPM from 2009 to 2011.

62884araD5R1_ASC026 3/15/13 11:29 PM Page 8

Abdulaziz M. Al-Sagr is a Supervisorin the Southern Area ProductionEngineering Department (SAPED). Hehas been very involved in the gasdevelopment program in the SouthernArea to meet the growing global gasdemand. Abdulaziz’s experience covers

several aspects of production optimization, including acidstimulation, coiled tubing applications and fishingoperations.

In 1995, he received his B.S. degree in ChemicalEngineering from King Fahd University of Petroleum andMinerals (KFUPM), Dhahran, Saudi Arabia.

Mustafa R. Al-Zaid is a GasProduction Engineer at Saudi Aramcoworking for the Southern AreaProduction Engineering Department(SAPED).

In 2010, he received his B.S. degreein Petroleum Engineering from the

University of Adelaide, Adelaide, Australia.

Saad M. Al-Driweesh is a GeneralSupervisor in the Southern AreaProduction Engineering Department(SAPED), where he is involved in gasproduction engineering, wellcompletion and fracturing andstimulation activities. His main interest

is in the field of production engineering, includingproduction optimization, fracturing and stimulation, andnew well completion applications. Saad has 24 years ofexperience in areas related to gas and oil productionengineering.

In 1988, he received his B.S. degree in PetroleumEngineering from King Fahd University of Petroleum andMinerals (KFUPM), Dhahran, Saudi Arabia.

Fadel A. Al-Ghurairi is a PetroleumEngineering Consultant and TechnicalSupport Unit Supervisor working ongas fields. He has 24 of years ofexperience in production and reservoirengineering. In the last 12 years, Fadelhas specialized in stimulation and

fracturing of deep gas wells.In 1988, he received his B.S. degree in Petroleum

Engineering from King Fahd University of Petroleum andMinerals (KFUPM), Dhahran, Saudi Arabia.


62884araD5R1_ASC026 3/15/13 11:29 PM Page 9


ABSTRACT INTRODUCTION

Liquid loading in gas reservoirs is a very important aspect toconsider when the goal is maintaining the production rate of afield. Many gas reservoirs produce some amount of liquid inassociation with the gas, either a hydrocarbon phase known ascondensate or an aqueous phase known as formation brine. Ifthis liquid accumulates in the wellbore, it will impair well pro-ductivity. The productivity can be restored if proper remedialaction is taken on the well. Figure 1 illustrates how the liquidloading can drastically decrease the well rate until a properwell intervention is implemented.

Liquid loading mainly occurs in low energy formations(with low reservoir pressure) and in tight gas regions. Thisproblem can also occur in moderate to high permeability reser-voirs with a high condensate to gas ratio (CGR).

For some wells, the liquid exists as a mist of droplets in theproduced gas. If the gas flow velocity in the production tubingis high enough, the gas will carry the droplets up the wellboreto be co-produced with the gas. The minimum gas velocity sat-isfying this condition is the Qcrit

1, which is a function of rateand is therefore related to the flowing wellhead pressure(FWHP)2. As the FWHP increases, both gas rate and velocitydecrease. If a well’s production rate falls below the Qcrit, liquidstarts accumulating in the wellbore, which not only decreasesthe production capacity of the well, but also adds to the back

Saudi Arabian nonassociated gas reservoirs produce variousamounts of condensate depending upon field and reservoirconditions. In most cases, these wells are hydraulically frac-tured, and at the initial stage after such stimulation treatment,each well needs to unload a large quantity of the pumped fluidto ensure the well’s full potential. If the liquid starts accumu-lating in the wellbore during production, the well productivitywill gradually decrease, and the well eventually may stop pro-ducing. If the gas flow velocity in the production string is highenough, the gas will continue flowing and carry the liquiddroplets up the wellbore to the surface. The minimum velocityand critical gas rate (Qcrit) are therefore the determining fac-tors to ensuring the stable field production rate and maintain-ing the production plateau while producing a well or severalwells from a condensate-rich field.

An analytical model has been developed to iteratively com-pute the critical velocity (Vcrit) and Qcrit for a given flowingwellhead pressure (FWHP), tubing diameter, and many otherreservoir and completion properties. If the FWHP is set and acertain production rate is expected of a well, the program automatically computes the pressure drop due to friction, dynamic hydrostatic head and bottom-hole pressure (BHP). Simultaneously, both the Vcrit and Qcrit required to unload thefluids are calculated. If the Qcrit is above the expected produc-tion rate, a different wellbore completion is automatically selected, and computation is continued until the Qcrit is lowerthan the expected rate of the well. If this is not possible, theprogram will display an appropriate message.

Several wells were analyzed from a condensate gas reservoirin a field that has to maintain certain production potential fora given number of years. The analyses show that the wells thatare producing without intervention are those that are con-firmed by this model to be flowing above the Qcrit, and thatthe wells that were intermittently producing and ultimatelycould not sustain production were producing at rates belowthe critical values. Using this iterative model, those rates can beautomatically adjusted for intermittent producers or a newcompletion string will be suggested to bring them back intoproduction.

An Iterative Solution to Compute CriticalVelocity and Rate Required to UnloadCondensate-Rich Saudi Arabian Gas Fields andMaintain Field Potential and Optimal ProductionAuthors: Hamza Al-Jamaan, Dr. Zillur Rahim, Bandar H. Al-Malki and Adnan A. Al-Kanaan

Fig. 1. Well intervention is essential to restore production.

62884araD6R1_ASC026 3/15/13 11:33 PM Page 1


pressure on the reservoir and can eventually completely kill thewell. All gas wells go through a natural decline that can bemodeled accurately. It is only during the liquid loading that thedecline curve deviates. The curve can then be restored to itsoriginal position by proper intervention on the well.

Early detection of liquid loading is essential to overcomeproductivity decreases. Fluctuations in the daily gas rates andcasing pressures are characteristic of liquid accumulation ingas wells. Determination of a fluid gradient in the tubing andregular bottom-hole pressure (BHP) surveys can also indicateliquid loading.

For proper reservoir management, it is imperative that eachwell be closely followed to ensure that no fluid is building upin the wellbore. This requires a good understanding and moni-toring of the field, reservoir conditions, hydrocarbon proper-ties, and the production and facility requirements andconstraints. Intense reservoir management and engineeringmust be conducted so that a remedy can be quickly consideredif liquid buildup starts impairing well productivity.

One of the easier methods used to overcome liquid loadingis to produce the problem well intermittently. This involvessustaining the natural flow of the well by alternatively shuttingin and opening the well. During the shut-in period, energygathers near the wellbore and then helps to unload the liquidas the well is opened. The downside of this approach is thatthe production from the well may be lost for several days orweeks depending on how quickly the near wellbore pressurebuilds up. This solution is also temporary as, with the deple-tion of the reservoir, the well will eventually stop producing.

The iterative software model developed in this study is anexcellent reservoir management tool that accurately computesthe Qcrit of a gas well, taking into consideration all the impor-tant reservoir and well properties. The model then provides remedial actions for wells that flow below the Qcrit.

These remedial actions can include changing the tubing sizethrough a workover or decreasing the FWHP. A viable artifi-cial lift method is also sometimes used through the implemen-tation of a free piston or plunger to lift fluids to the surfaceusing the energy stored in the gas — the installation ofplungers reduces the problem faced with the intermittent pro-duction strategy. When liquid accumulation is considered andacted upon, the intervention will restore well productivity andmaintain the overall field production rate.

DESCRIPTION OF THE ITERATIVE MODEL

The purpose of this software application is to calculate theQcrit for any given gas well utilizing the “Turner DropletModel” to ensure stable flow conditions. From certain inputvalues for a specific gas well, the program will calculate theQcrit and will test whether the well is flowing below or abovethe Qcrit, which determines whether it is a candidate for inter-vention or not. The program will also test and plot how theQcrit for that specific well will vary with changes in tubing

diameter and BHP to find the optimal conditions for flowingthe well without causing any liquid holdup.

Benefits

The iterative program has several benefits that make signifi-cant contributions to the management of a gas condensatereservoir. The program is a quick guide to the stable flow con-ditions needed for gas wells to avoid possible accumulation ofliquids in the tubing. This model then provides proactive solu-tions to maintain continuous gas production. The model alsorecognizes and predicts liquid loading that can happen in thefuture, and simultaneously provides practical remedial actionto be taken at the outset to overcome later production impair-ment. By preventing liquid loading, it enhances the productionlife of a gas condensate reservoir and ensures the most efficientreserves exploitation.

Signs of Liquid Loading

Liquid loading is not easily identified. Even when a well is liq-uid loaded, it may continue to produce for a long time. It fol-lows that if liquid loading is recognized and reduced at anearly stage, higher producing rates can be achieved and main-tained. Symptoms indicating liquid loading include the follow-ing2:

• Pressure Gradient: Pressure surveys reveal a heaviergradient.

• Variance from the Decline Curve: Typically gas wellswill follow an exponential-type curve decline; however,liquid loading generally leads to a deviation from thecurve with a lower than predicted production rate.

• Liquid Slugging: Liquid production does not arrive tothe surface in a steady continuous flow, but instead inslugs of fluid. This is readily observed throughproduction monitoring.

As shown in Fig. 2, the flow regime that is desirable in gaswells is the “mist flow,” where there is a continuous gas phasewith evenly dispersed liquid droplets. When a gas well flowsbelow the critical gas flow rate, the flow regime changes to“slug flow,” where the liquid starts accumulating in the well-bore.

Procedure

To predict liquid loading in gas wells, the Turner DropletModel is used to calculate the critical gas velocity3 using thefollowing equation:

VCrit-T = 1.92 [σ (�i_�g)]1⁄4

––––––––––––– ————————————

�g1⁄2

where VCrit-T is the Turner critical velocity in feet/second, is thegas-liquid interfacial tension in dyne/cm (dynes/centimeter), �i

62884araD6R1_ASC026 3/15/13 11:33 PM Page 2

is the liquid density in lb/ft3 (pounds/cubic foot), and �g is thegas density in lb/ft3. In this equation, the gas density is approx-imated at the bottom-hole conditions, and the BHP is calcu-lated from the FWHP using Guo’s analytical method4. Thecritical flow rate is subsequently computed and converted tostandard conditions using the following formula: Qcrit =Vcrit-T x ATubing.

At standard conditions of 60 °F and 14.6 psi, the molar vol-ume is 379.48 ft3/lb-mol. To perform this computation andconduct a sensitivity analysis, the data provided in Table 1 isinput in the program.

Based on this input, the program will calculate the criticalgas flow rate and simultaneously assess whether the well isflowing above or below that critical rate. The program willalso output a plot showing how the critical gas flow rate varieswith tubing diameter and flowing bottom-hole pressure(FBHP). If the well is flowing below the Qcrit, several interven-tions are automatically presented for consideration to fix theproblem. These include:

• Reduction of FBHP, subject to constraints imposed byreservoir engineering.

• Use of minimum operating FWHP input by the user.

Reducing the FBHP will affect the situation in two ways: itwill decrease the density of the gas in the tubing and will in-crease the production rate from the formation into the tubing.

The gas well production rate is defined as Q = C (Pr2 –Pwf

2 )n, where C is a constant that includes drainage radius, radius ofthe wellbore, formation thickness, reservoir permeability, reser-voir temperature, gas compressibility, etc., and n accounts fornon-ideal gas behavior. It is assumed that the C and n values of

the well do not change when the FBHP is reduced. The vari-able Pr in this equation is the average reservoir pressure (psia),Pwf is the FBHP (psia), Q is the gas rate (Mscfd), the value of nranges from 0.5 to 1, and C is defined by Mscfd/psia2.

The user can also evaluate the effect of replacing the tubingwith the next smaller size (velocity string concept) and the effect of reducing the gas-liquid interfacial tension (soap-sticksconcept). The following sections present a few examples wheregas condensate wells were analyzed using the iterative program.

EXAMPLE WELL-1

Figure 3 presents the well parameters and reservoir conditionsthat were input for the well. They show that the well is flowing ata low rate of 1 million standard cubic ft per day (MMscfd), andthe condensate yield is 130 bbl/MMscf. From a reservoir pressureof 4,000 psi and a reservoir temperature of 240 °F, the programcomputed a critical gas rate of 4.18 MMscfd for this well.

The result box provided at the bottom of Fig. 3 shows thatthe current well production rate (1 MMscfd) is lower than thecritical gas flow rate, and therefore the well is loading up withliquids. An intervention box thereby appeared to suggest a re-duction in the tubing size to overcome the slug flow. At thecurrent well flowing conductions and based on the inflow per-formance curve that the program automatically computes, theBHP is also low. Therefore, the only possible solution for get-ting the well to produce above the critical gas flow rate is toreduce the tubing size, which is also aligned with the velocitystring concept that reduces the flow area of a well by insertingan external string in the wellbore.

After clicking the “show intervention” button marked ingreen, Fig. 3, a plot of Qcrit vs. tubing internal diameter appeared, Fig. 4.

Fig. 2. Flow regime at different conditions.

Parameters Units

Gas Rate MMscfd

FWHP psia

Condensate-to-Gas Ratio (CGR) bbl/MMscfd

Average Wellbore Gradient psi/ft

Reservoir Depth ft

Gas Gravity (Air=1)

Liquid Density kg/m3

Gas-Liquid Interfacial Tension dynes/cm

Tubing Diameter inches

Average Reservoir Pressure psia

Bottom-hole Temperature °F

Minimum Operating FWHP psi

Table 1. Model input parameters


62884araD6R1_ASC026 3/15/13 11:33 PM Page 3

The iterative software application in this example has com-puted the following:

• The critical gas flow rate for the initial conditions.

• A plot showing the effect on the critical gas flow rate ofreducing the tubing internal diameter and FBHP.

If intervention is needed, the program will calculate the optimal gas rate that can be achieved by reducing the FBHP,based on a minimum operating FWHP input by the user(1,000 psia), and will inform the user whether the test was successful or not. In all cases, if a well is flowing below thecritical gas flow rate, then intervention is always required to

prevent slug flow. Reducing the tubing size will result in alower critical gas flow rate. If the FBHP is reduced, the criticalgas flow rate also decreases; this is because the gas density inthe tubing will decrease as a result of the FBHP reduction andwill thereby increase the production rate from the formationinto the tubing. In the case of Well-1, the FBHP was alreadylow, and further reduction of the FBHP was not possible dueto the constraint imposed by the engineer (a minimum operat-ing FWHP of 1,000 psia). The only possible solution in thiscase was to reduce the tubing size.

Fig. 3. Well-1 input variables.

Fig. 4. Critical gas flow rate vs. tubing internal diameter plot for Well-1.

Fig. 5. Well-2 input variables.

Fig. 6. Critical gas flow rate vs. tubing internal diameter plot for Well-2.


62884araD6R1_ASC026 3/15/13 11:33 PM Page 4

EXAMPLE WELL-2

The input variables in Fig. 5 show that, for the given reservoirproperties, the Qcrit for Well-2 is 5.75 MMscfd. This well wascurrently producing at 10 MMscfd, which is above the criticalgas rate, so no intervention was required.

It is also worth noting that plots of Qcrit vs. tubing internaldiameter and FBHP, Fig. 6, were still generated so users canbetter understand how these critical rates change throughoutthe life of the well as reservoir pressure depletes. Also, in case atubing replacement is required for this well due to corrosion ordamage, an assessment of the effect of the new tubing size onthe production rate can be quickly performed.

The critical gas rate for Well-2 is higher than that for Well-1. That is because certain factors, such as tubing diameter andBHP, have a significant impact on the Vcrit calculation com-pared to gas gravity, interfacial tension and bottom-hole tem-perature. Well-2 has a much higher BHP than Well-1, resultingin higher gas density and a higher Vcrit. The Qcrit vs. the FBHPplot is provided in Fig. 7.

Example Sensitivity Runs

Qcrit is a function of many parameters, such as reservoir andwell configuration.

A sensitivity example presented in Fig. 8 shows the impactof FWHP and tubing size on the Qcrit. A decrease in FWHPand tubing size means a lower flow rate is required to keep awell continuously unloaded. The figure also illustrates that anincrease in CGR increases the Qcrit and that the proportion depends on hydrocarbon properties and well configuration.Figure 9 shows the Qcrit as a function of hydrocarbon densityfor a well with CGR = 100 bbl/MMscfd.

Figure 10 illustrates the effects of CGR and reservoir pres-sure on gas well performance. The CGR value has been variedbetween 10 and 500 bbl/MMscfd for reservoir pressures be-tween 5,000 psi and 8,000 psi, respectively. The FWHP washeld constant at 3,000 psi. Figures 11 and 12 illustrate gas ratesand changing gas density as a function of CGR and reservoir

Fig. 7. Critical gas flow rate vs. FBHP.

Fig. 8. Critical gas flow rate as a function of FWHP and CGR.

Fig. 9. Critical gas flow rate as a function of hydrocarbon density.

Fig. 10. Effects of CGR on well performance.

Fig. 11. Declining gas rate with higher CGR and lower reservoir pressure.


62884araD6R1_ASC026 3/15/13 11:33 PM Page 5

pressure. It is noted that in low-pressure reservoirs (depletedreservoirs) some of the high CGR wells will not produce. A remedial plan therefore needs to be considered in advance toovercome such situations.

CONCLUSION

Liquid loading is a complex phenomenon, and accurately model-ing the process is very difficult due to the various flow regimes andthe dynamics of fluid flow and its interaction among reservoir,wellbore and surface hydraulics. Most models are based onsteady-state flow solutions and therefore cannot necessarilycapture the full process that occurs throughout the life of a well.

Liquid loading is currently one of the major challengesfaced in high CGR fields, and several wells have been shut-indue to the inability to unload fluids accumulated in their well-bores. If the Qcrit is calculated and predicted earlier, then stepscan be taken to maintain the well rate above the Qcrit to avoidliquid loading. The software application developed in thisstudy detects the loading process and automatically generates asolution so that well intervention can be planned in advance.This application was initially developed and coded in visualBASIC and was then transferred into an easier and more user-friendly interface to better conduct the runs and sensitivityanalysis. Several wells have been analyzed using this model,which has greatly helped in improving good reservoir manage-ment practices.

ACKNOWLEDGMENTS


This article was presented at the SPE Kuwait InternationalPetroleum Conference and Exhibition, Kuwait City, December10-12, 2012.

REFERENCES

1. Hearn, W.: “Gas Well Deliquification ApplicationOverview,” SPE paper 138672, presented at the

International Petroleum Exhibition and Conference, AbuDhabi, U.A.E., November 1-4, 2010.

2. Lea, J.F. and Nickens, H.V.: “Solving Gas Well Liquid-Loading Problems,” Journal of Petroleum Technology, Vol.56, No. 4, April 2004, pp. 30-36.

3. Turner, R.G., Hubbard, M.G. and Dukler, A.E.: “Analysisand Prediction of Minimum Flow Rate for the ContinuousRemoval of Liquids from Gas Wells,” Journal of PetroleumTechnology, Vol. 21, No. 11, November 1969, pp. 1,475-1,482.

4. Guo, B.: “Use of Wellhead Pressure Data to Establish WellInflow Performance Relationship,” SPE paper 72372,presented at the SPE Eastern Regional Meeting, Canton,Ohio, October 17-19, 2001.

Fig. 12. Gas density as a function of reservoir pressure.

BIOGRAPHIES

Hamza Al-Jamaan is a PetroleumEngineer with the Gas ReservoirManagement Department at SaudiAramco. His interests include generalreservoir engineering, fielddevelopment and productionoptimization. Currently, Hamza is

pursuing his M.S. and Ph.D. degrees in PetroleumEngineering at Stanford University, Stanford, CA. Hiscurrent research involves the characterization andpetrophysics of shale gas.

He received a dual B.S. degree with honors in PetroleumEngineering and Economics from the University of Texas atAustin, Austin, TX.

Dr. Zillur Rahim is a PetroleumEngineering Consultant with SaudiAramco’s Gas Reservoir ManagementDepartment (GRMD). He heads theteam responsible for stimulationdesign, application and assessment forGRMD. Rahim’s expertise includes

well stimulation, pressure transient test analysis, gas fielddevelopment, planning, production enhancement, andreservoir management. Prior to joining Saudi Aramco, heworked as a Senior Reservoir Engineer with Holditch &Associates, Inc., and later with Schlumberger ReservoirTechnologies in College Station, TX, where he used toconsult on reservoir engineering, well stimulation, reservoirsimulation, and tight gas qualification for national andinternational companies. Rahim is an Instructor ofpetroleum engineering industry courses and has trainedengineers from the U.S. and overseas. He developedanalytical and numerical models to history match andforecast production and pressure behavior in gas reservoirs.Rahim developed 3D hydraulic fracture propagation andproppant transport simulators and numerical models tocompute acid reaction, penetration, and fracture


62884araD6R1_ASC026 3/15/13 11:33 PM Page 6

conductivity during matrix acid and acid fracturingtreatments.

Rahim has authored 65 Society of Petroleum Engineers(SPE) papers and numerous in-house technical documents.He is a member of SPE and a technical editor for theJournal of Petroleum Science and Engineering (JPSE).Rahim is a registered Professional Engineer in the State ofTexas and a mentor for Saudi Aramco’s TechnologistDevelopment Program (TDP). He is an instructor of theReservoir Stimulation and Hydraulic Fracturing course forthe Upstream Professional Development Center (UPDC) ofSaudi Aramco. Rahim is a member of GRMD’s technicalcommittee responsible for the assessment and approval ofnew technologies.

Rahim received his B.S. degree from the Institut Algeriendu Petrole, Boumerdes, Algeria, and his M.S. and Ph.D.degrees from Texas A&M University, College Station, TX,all in Petroleum Engineering.

Bandar H. Al-Malki joined SaudiAramco in 1998 as a ProductionEngineer, working in the company’sgas fields. He is currently the GeneralSupervisor of the Gas ReservoirManagement Division. This rolerequires him to monitor the

production capacity of the plants, while optimizing theproductivity of the wells and preventing wasted time andresources.

Bandar received his B.S. degree in PetroleumEngineering from King Fahd University of Petroleum andMinerals (KFUPM), Dhahran, Saudi Arabia. In 2004, heearned his M.S. degree in Petroleum Engineering from theImperial College, London, U.K., focusing on gascondensate reservoirs.


Adnan A. Al-Kanaan is the Managerof the Gas Reservoir ManagementDepartment (GRMD) where heoversees three gas reservoir manage-ment divisions. Reporting to the ChiefPetroleum Engineer, Adnan is directlyresponsible for making strategic

decisions to enhance and sustain gas delivery to theKingdom to meet its ever increasing energy demand. He over-sees the operating and business plans of GRMD, newtechnologies and initiatives, unconventional gas developmentprograms, and the overall work, planning and decisionsmade by his more than 70 engineers and technologists.


Prior to the inception of GRMD, he was the GeneralSupervisor for the Gas Reservoir Management Division underthe Southern Reservoir Management Department for 3 years,heading one of the most challenging programs in optimizingand managing nonassociated gas fields in Saudi Aramco.

Adnan started his career at the Saudi Shell Petro-chemical Company as a Senior Process Engineer. He thenjoined Saudi Aramco in 1997 and was an integral part ofthe technical team responsible for the on-time initiation ofthe two major Hawiyah and Haradh Gas Plants thatcurrently process more than 6 billion cubic feet (bcf) of gasper day. Adnan also directly managed the Karan and Wasitfields — two major offshore gas increment projects — with anexpected total production capacity of 4.3 bcf of gas per day.

He actively participates in the Society of PetroleumEngineers’ (SPE) forums and conferences and has been thekeynote speaker and panelist for many such programs.Adnan’s areas of interest include reservoir engineering, welltest analysis, simulation modeling, reservoir characterization,hydraulic fracturing, reservoir development planning andreservoir management.



62884araD6R1_ASC026 3/15/13 11:33 PM Page 7

ABSTRACT (~55,000 mg/l); microbial content present throughout the sys-tem differs from one location to another due to exploitation ofthe biocide batch treatment further downstream of the system.Moreover, produced water re-injection may enrich microbialcontent and may allow different microbial species to live onthe higher concentration of organics in some parts of the system.

The waterflooding system in Saudi Arabia was subjected toa microbial community structure review. Conventional micro-bial investigations were conducted to assess the microbial ac-tivity in the system. The progress in molecular biology andDNA sequencing technologies has opened endless possibilitiesto analyze microbial communities and identify the types of mi-croorganisms responsible for relevant microbial activities, suchas souring and corrosion. Pyrosequencing, a massive parallelDNA sequencing technology, was used here to characterize thecommunity composition of the seawater injection system.

MATERIALS AND METHODS

Sample Description and Preparation

Two samples of scraping solids (Table 1, Samples 5 and 6)were collected from a water pump station in the system andanalyzed using environmental scanning electron microscopy(ESEM) coupled with an energy dispersive X-ray spectroscopy(EDS) analyzer to assess the presence of sulfate-reducing bacte-ria (SRB) in the system.

Four seawater samples from the system (Table 1, Samples 1to 4) were collected from various locations in the field and an-alyzed using the 16S pyrosequencing method to assess the mi-crobial community composition in the system. The seawatersamples were filtered on-site at the field locations where thesamples were collected, using 0.2 µm filters. The filters werefixed and dried. Each of the six samples (the scraping solidsand the water samples) were transferred into 2 ml Eppendorftubes and sent to the University of Calgary for DNA extractionand pyrosequencing analysis. Upon arrival at the laboratory,the tubes were strongly vortexed with seawater filtrate to sus-pend the biomass. This was repeated two to three times to en-sure that the greatest amount of biomass was recovered. Next,the tubes were centrifuged at 17,000 x g for 10 minutes. Thesupernatant was discarded and the cell pellet was frozen at -70°C until DNA extraction.

A pyrosequencing survey of planktonic seawater and sessilepipeline solids samples from a seawater injection system inSaudi Arabia indicates the presence of distinct microbial com-munities. The pipeline surface had a microbial community consisting of the anaerobic heterotrophs Roseovarius, Ruege-ria, Colwellia, Lutibacter and Psychrobacter, which fermentrefractory organic carbon to intermediates (e.g., lactate andH2) and are then used by sulfate-reducing bacteria (SRB) of thegenus Desulfovibrio to reduce sulfate to sulfide. All of thesemicrobes were present in a much smaller fraction in the seawa-ter, e.g., Desulfovibrio was present in a 100-fold smaller frac-tion in the planktonic seawater population than in the pipelinesolids. The presence of sulfur in the pipeline solids, as deter-mined by X-ray powder diffraction (XRD), and of high num-bers of cultivatable SRB (108/g) also indicated the potential forsignificant microbially influenced corrosion (MIC) risk, bio-fouling and water quality deterioration. The data suggests thatmeasures to control SRB should be continued and possibly adjusted to decrease the risk of operational problems causedby SRB growth and activity.

INTRODUCTION

In water injection, or waterflooding, either aquifer water ordeoxygenated and filtered seawater is injected at strategicpoints along the periphery of the oil reservoir, displacing theoil and “pushing” it towards oil supply wells in the center ofthe formation. The technique increases crude oil recovery substantially and allows for greater returns from the field.

Nonpotable water from underground aquifers locatedabove the oil reservoirs is usually used in injection programs tomaintain reservoir pressure. Oil companies also have convertedsome of their water injection facilities to use treated seawaterin waterflooding to conserve the aquifers for future use.

The seawater injection system studied uses water from aseawater treatment plant in Saudi Arabia that treats millions ofgallons of seawater per day from the Gulf region and ships itover very long distances (hundreds of kilometers) throughmassive transfer lines. Given the size and complexity of the injection system and the high salinity of the water it uses

Microbial Community Structure in a SeawaterFlooding System in Saudi Arabia

Authors: Mohammed A. Al-Moniee, Dr. Indranil Chatterjee, Dr. Gerrit Voordouw, Dr. Peter F. Sanders andDr. Tony Y. Rizk


62884araD7R1_ASC026 3/15/13 11:36 PM Page 1

DNA Extraction Technique

The cell pellets stored at -70 °C were taken out and thawed toroom temperature. The cell pellets were re-suspended in 280 µlof 0.15 M NaCl and 0.1 M ethylene diamine tetra-acetic acid(pH 8). Genomic DNA was isolated using a procedure outlinedin Marmur1. In brief, the cell pellets were treated withlysozyme (to weaken the bacterial cell wall), followed by treat-ment with 25% sodium dodecyl sulfate and then with threerounds of freeze-thaw cycles (-70 °C to 68 °C).

Treatment with DNase-free RNase and recombinant Pro-teinase K (Roche Diagnostics, GmbH) was done to removeRNA and protein contaminants, respectively. DNA was furtherpurified by precipitation with a DNA precipitation mix(sodium acetate + ethanol) and by washing with 70% ethanol.DNA was re-suspended in buffer EB (10 mM Tris-Cl, pH 8.5;Qiagen QIAquick kit).

Community Structure Analysis by Pyrosequencing

DNA samples were amplified through a two-step polymerasechain reaction (PCR) amplification. The first PCR (25 cycles)was performed with 16S primers 926Fw (AAACTYAAAK-GAATTGRCGG) and 1392R (ACGGGCGGTGTGTRC).Agarose gel analysis confirmed the presence of the desiredPCR product at approximately 500 bp, Fig. 1.

Using this as the template, a second round of PCR (10 cycles)was performed using the FLX Titanium Amplicon primers454T_RA_X and 454T-FB. These have the sequences for 16Sprimers 926Fw and 1392R as their 3 ft ends. Primer 454T_RA_Xhas a 25 nucleotide A-adaptor (CGTATCGCCTCC-CTCGCGCCATCAG) and a 10 nucleotide multiplex identifierbarcode sequence X, whereas primer 454T-FB has a 25 nu-cleotide B-adaptor sequence (CTATGCGCCTTGCCAGCC-CGCTCAG). Following the second PCR amplification, thePCR product was checked on a 0.7% agarose gel, Fig. 2, andpurified with a QIAquick PCR purification kit (Qiagen).

The second PCR product concentration, Table 1, was thendetermined by a Qubit fluorometer (Invitrogen), using aQuant-iT™ dsDNA high sensitivity (HS) assay kit (Invitrogen).

PCR products (typically 20 µl of 5 ng/µl) were sent for pyrose-quencing analyses. Pyrosequencing was performed with aGenome Sequencer FLX instrument, using a GSFLX TitaniumSeries kit XLR70 (Roche Diagnostics Corporation).

RESULTS AND DISCUSSION

Initial Bacterial Assessment

Initial bacterial assessment of the scraping solids samples andfour seawater samples from different locations in the field con-firmed the presence of SRB. The scraping solids contained ahigh concentration of SRB, in the range of 108/g of scrapingsolids. The ESEM and EDS results, Fig. 3, showed that themain elements in the samples were sulfur, oxygen, sodium andiron. The samples were rich in FeS (mackinawite) and NaCl.The X-ray powder diffraction (XRD) method was also used todetermine the phase identification and quantification of thescraping solids (one scraping solid and scraping filter). The re-sults showed that the major phases are 55% magnetite [Fe3O4]and 22% akaganeite [FeO(OH)] for the scraping solid, and44% mascagnite [(NH4)2S] and 42% mackinawite [FeS] forthe scraping filter. A high amount of Fe (41%) as well as thepresence of sulfur (8%) was detected through an X-ray fluorescence (XRF) elemental analysis on the scraping solid

Fig. 1. Agarose (0.7%) gel analysis of first PCR amplification product. M = lHindIII molecular marker. Samples 1 to 6 are as described in Table 1. -vecnt =negative control without added DNA.

Sample Description PCR Product (ng/ml)

Sample #1 Injection well 45.8

Sample #2 Water 2 51

Sample #3 Water 3 41.9

Sample #4 Water 4 25.7

Sample #5 Scraping solids-1 48.9

Sample #6 Scraping solids-2 48.8

Table 1. Concentration of second PCR amplification products using Quant-iTdsDNA HS assay kit

Fig. 2. Agarose (0.7%) gel analysis of second PCR amplification product. M = lHindIII molecular marker. Samples 1 to 6 are as described in Table 1. -vecnt =negative control without added DNA.


62884araD7R1_ASC026 3/15/13 11:36 PM Page 2

and indicated sulfate reduction in agreement with the highnumbers of SRB.

Pyrosequencing Data

The pyrosequencing data were analyzed by Phoenix-2, a bioin-formatics pipeline developed in-house2. Sequence reads weresubjected to stringent systematic checks to remove low qualityreads and minimize sequencing errors that can be introducedduring the pyrosequencing process3. Eliminated sequences in-cluded those that: (1) did not perfectly match the adaptor andprimer sequences, (2) had ambiguous bases, (3) had an atypi-cal length of 1 SD away from mean length after removingadaptor and primer sequences, (4) had an average qualityscore below 25, and (5) contained homopolymer lengthsgreater than 8 bp.

The remaining high quality sequences were comparedagainst the nonredundant SSU reference data set of SILVA1024

using the Tera-BLAST algorithm on a TimeLogic decypher sys-tem from Active Motif, Inc., consisting of 12 boards. TheTera-BLAST results were used to screen for problematic,chimerical and eukaryotic sequences. Sequences having a bestalignment covering less than 70% or having a best BLASTsearch hit an e-value greater than e-50 were excluded as prob-lematic sequences.

Putative chimeras were identified by using a two-stage approach. The sequences having a best alignment covering lessthan 90% of the trimmed read length, with greater than 90%sequence identity to the best BLAST match, were identified aspotential chimeras. The potential chimeras were excluded fromfurther analysis if they were also identified as chimeras at mini-mum 80% bootstrap support in chimera.slayer implemented in

the Mothur software package5. The filtered sequences, afterpassing the quality control process for problematic, chimericaland eukaryotic sequence removal, were clustered into opera-tional taxonomic units (OTUs) at 3% distance by using thecomplete linkage algorithm in Mothur.

A taxonomic consensus of each representative sequencefrom each OTU was derived from the recurring species within5% of the best bit score from a BLAST search against theSILVA database. Of the good reads generated by pyrosequenc-ing, 36,138 were assigned taxonomic identifiers, which wereidentified at the genus level.

Microbial Communities in the Injection System

Both planktonic seawater samples and sessile pipe sampleswere analyzed. Table 2 and Fig. 4 show the percentage of readsin each sample. This gives an indication of the microbial diver-sity in the samples. Bar diagrams, Fig. 4, were constructed us-ing the average percentage of reads of each sample. It indicatesthe main microbial population (genus level) in the seawatersamples. Some interesting differences among the samples areapparent. For example, sample 2 has a high fraction of Po-laromonas, which was not found in any other sample. In theabsence of information on what these samples represent, wecannot make suggestions on why these microorganisms arefound at these particular sites.

The betaproteobacterium Delftia was found to be the mostprominent genus in the seawater samples (39%). Delftiatsu-ruhatensis, a terephthalate-assimilating bacterium, has beenisolated from activated sludge from a domestic wastewatertreatment plant in Japan6. In another recent study, Delftia spp.were isolated as a novel peptidoglycan-degrading bacterium insamples from mesotrophic lake water in Denmark7.

In addition to Delftia, the alphaproteobacterium Sphin-gomonas as well as the gammaproteobacteria Pseudomonas,Pseudoalteromonas and Sedimenticola were also dominant inseawater samples. Among these, Sedimenticola has been docu-mented as an anaerobic selenate-respiring bacterium isolatedfrom estuarine sediment8. The well-known sulfate-reducingdeltaproteobacterium Desulfovibrio was present with an aver-age fraction of 2.37%.

Fig. 3. ESEM image at 1-2 µ and the corresponding EDS X-ray spot analysisspectrum.

Fig. 4. Graphical representation of genus level survey of 16S sequences in theseawater injection system. The average for all six samples (Table 2) is shown.


62884araD7R1_ASC026 3/15/13 11:36 PM Page 3

Sessile Microbial Community

The data obtained allowed comparison of the planktonic com-munity (Table 2, A1, the average for seawater samples 1 to 4)and the sessile community present on the pipeline wall (Table 2,A2, the average for scrapings samples 5 and 6). The ratioR=A2/A1 was calculated for each entry in Table 2 and indicatedthe tendency of a given microbe to attach to the pipeline wall.

The sessile community was dominated (in order of decreas-ing R) by Petrobacter, Desulfovibrio, Achromobacter, Roseo-varius, Colwellia, Sedimenticola, Thiomicrospira, Lutibacter,Ruegeria and Psychrobacter. Of these, Petrobacter and Achro-mobacter are potentially anaerobic heterotrophic bacteria, ca-pable of degrading organic carbon in seawater. Roseovariusand the related Ruegeria, Colwellia, Lutibacter and Psy-chrobacter are commonly isolated from seawater, with Colwellia being capable of Fe-III reduction.

Collectively, these bacteria may form a biofilm on thepipeline wall, anaerobically degrading organic carbon in sea-water. Degradation products (e.g., lactate or H2) are then used

by SRB of the genus Desulfovibrio to reduce sulfate to sulfide.Sulfide may be reoxidized by Thiomicrospira if traces of oxy-gen remain in the seawater.

CONCLUSIONS

Planktonic seawater and sessile pipeline solids samples (Sam-ples 1 to 4 and Samples 5 and 6, respectively) from the seawa-ter injection system in Saudi Arabia harbor a diverse microbialcommunity, which shows very significant differences, Table 2.The pipeline surface has a microbial community with a 100-fold higher fraction of SRB of the genus Desulfovibrio, whichmay contribute to microbially influenced corrosion and bio-fouling. Therefore, treatment to limit pipeline damage, as cur-rently being undertaken, must continue or must be adjusted toprevent further proliferation of SRB.

ACKNOWLEDGMENTS

This work was supported by an NSERC Industrial Research

Number of Reads (n) 24,485 14,156 10,329

Saudi Aramco Sample Numbers SA-1 to SA-6 SA-1 to SA-4 SA-5 to SA-6

Sample Type All Seawater Srapings Ratio

Average Reads (%) A1 (%) A2 (%) A2/A1

Class Genus

Betaproteobacteria Delftia 39.58 53.50 11.72 0.22

Alphaproteobacteria Sphingomonas 16.24 22.05 4.62 0.21

Gammaproteobacteria Pseudomonas 6.28 0.24 18.95 78.48

Gammaproteobacteria Pseudoalteromonas 4.89 7.33 0.02 0.00

Gammaproteobacteria Sedimenticola 2.56 0.18 7.32 39.78

Betaproteobacteria Polaromonas 2.50 3.75 0.00 0.00

Gammaproteobacteria Colwellia 2.47 0.15 7.11 47.40

Deltaproteobacteria Desulfovibrio 2.37 0.07 6.98 98.70

Betaproteobacteria Petrobacter 2.17 0.05 6.43 142.96

Flavobacteria Lutibacter 1.20 0.24 3.13 13.04

Gammaproteobacteria Thiomicrospira 1.15 0.19 3.07 16.57

Gammaproteobacteria Acinetobacter 1.03 0.93 1.25 1.35

Alphaproteobacteria Roseovarius 0.99 0.03 2.91 90.28

Deltaproteobacteria Desulfurivibrio 0.81 1.08 0.26 0.24

Alphaproteobacteria Bradyrhizobium 0.71 1.00 0.14 0.13

Actinobacteria Microbacterium 0.69 0.91 0.24 0.26

Gammaproteobacteria Psychrobacter 0.64 0.23 1.46 6.34

Betaproteobacteria Achromobacter 0.60 0.02 1.77 92.05

Betaproteobacteria Hylemonella 0.59 0.88 0.00 0.00

Alphaproteobacteria Ruegeria 0.57 0.18 1.33 7.35

Table 2. Genus level survey of 16S sequences in samples of seawater and scrapings listed in Table 1. The number of pyrosequencing reads (n) and the average fraction(%) of these for each genus are indicated for the 20 most prevalent genera. The list is ranked in order of most to least prevalent genus (average for all samples).Averages for seawater (A1) and scrapings (A2) samples are also provided, as well as the ratio R=A2/A1, which indicates prevalence in pipeline scrapings


62884araD7R1_ASC026 3/15/13 11:36 PM Page 4

Chair Award to GV, which was also supported by BakerHughes Inc., Commercial Microbiology Ltd. (Intertek), theComputer Modeling Group Ltd., ConocoPhillips Company,YPF SA, Aramco Services, Shell Canada Ltd., Suncor EnergyDevelopments Inc. and Yara International ASA, as well as bythe Alberta Innovates-Energy and Environment Solutions. Thework was also supported by funding from Genome Canada,Genome Alberta, the Government of Alberta and Genome BC.We thank Xiaoli Dong and Christoph Sensen for the bioinfor-matics analyses.

REFERENCES

1. Marmur, J.: “A Procedure for the Isolation ofDeoxyribonucleic Acid from Microorganisms,” Journal ofMolecular Biology, Vol. 3, No. 2, April 1961, pp. 208-218.

2. Park, H.S., Chatterjee, I., Dong, X., Wang, S.H., Sensen,C.W., Caffrey, S.M., et al.: “Effect of Sodium BisulfiteInjection on the Microbial Community Composition in aBrackish Water Transporting Pipeline,” Applied andEnvironmental Microbiology, Vol. 77, No. 19, October 1,2011, pp. 6,908-6917.

3. Huse, S.M., Huber, J.A., Morrison, H.G., Sogin, M.L. andWelch, D.M.: “Accuracy and Quality of Massively ParallelDNA Pyrosequencing,” Genome Biology, Vol. 8, No. 7,July 20, 2007.

4. Pruesse, E., Quast, C., Knittel, K., Fuchs, B.M., Ludwig,W., Peplies, J., et al.: “SILVA: A Comprehensive OnlineResource for Quality Checked and Aligned RibosomalRNA Sequence Data Compatible with ARB,” NucleicAcids Research, Vol. 35, No. 21, October 17, 2007, pp.7,188-7,196.

5. Schloss, P.D., Westcott, S.L., Thomas, R., Hall, J.R.,Hartmann, M., Hollister, E.B., et al.: “Introducing Mothur:Open Source, Platform Independent, CommunitySupported Software for Describing and ComparingMicrobial Communities,” Applied EnvironmentalMicrobiology, Vol. 75, No. 23, October 2, 2009, pp.7,537-7,541.

6. Shigematsu, T., Yumihara, K., Ueda, Y., Numaguchi, M.,Morimura, S. and Kida, K.: “Delftiatsuruhatensis sp. nov.,a Terephthalate-assimilating Bacterium Isolated fromActivated Sludge,” International Journal of SystematicEvolutionary Microbiology, Vol. 53, September 2003, pp.1,479-1483.

7. Jørgensen, N.O.G., Brandt, K.K., Nybroe, O. and Hansen,M.: “Delftialacustris sp. nov., a Peptidoglycan-degradingBacterium from Fresh Water, and Emended Description ofDelftiatsuruhatensis as a Peptidoglycan-degradingBacterium,” International Journal of SystematicEvolutionary Microbiology, Vol. 59, 2009, pp. 2,195-2,199.

8. Narasingarao, P. and Häggblom, M.M.: “Sedimenticolaselenatireducens, gen. nov., sp. nov., an Anaerobic Selenate-respiring Bacterium Isolated from Estuarine Sediment,”Systematic and Applied Microbiology, Vol. 29, January 20,2006, pp. 382-388.

BIOGRAPHIES

Mohammed A. Al-Moniee joinedSaudi Aramco’s PetroleumMicrobiology Unit of the Research &Development Center (R&DC) in 1998.He is currently working as a SeniorLab Scientist with the MaterialPerformance Group of the Technical

Services Division, R&DC. In June 2005, Mohammedundertook an internship program with the BiotechnologyDepartment at the Institute Francias du Petrol (IFP),France, working on bio-denitrogenation of diesel oil. Hehas over 15 years of professional and field experience inthe areas of microbial corrosion, bactericides and microbialsensing, biofouling and bioprocessing for oil upgrading.Mohammed has handled various projects covering SaudiAramco’s oil fields. In particular, he has worked onbacterial monitoring and control in the seawater injectionsystem and oil pipeline system.

In 1997, Mohammed received his B.S. degree inChemistry from the University of Toledo, Toledo, OH, andin 2012, he received his M.S. degree in ProjectManagement (Oil and Gas Specialty) from the University ofLiverpool, Liverpool, U.K.

Mohammed has authored or coauthored numerousjournal and international conference publications in hisareas of expertise. He is an active member of the AmericanChemical Society (ACS) and the Saudi ArabianInternational Chemical Science Chapter of ACS.

Dr. Indranil Chatterjee is the SeniorResearch Microbiologist at the PuneTechnology Center, India, for the OilField Chemical Division of Nalco (AnEcolab Company). He acquiredexperience in various microbiologicaland molecular techniques in addition

to projects dealing with global genomic analysis(transcriptomics and proteomics). In addition, Indranil wasalso involved in pharmaceutical industrial projects withBayer Vital, GmbH and Wyeth Pharma, GmbH.

Following his 6 years of research experience withMedical Microbiology, he joined the Petroleum Micro-biology Research Group (PMGR) at the University ofCalgary, Calgary, Alberta, Canada. Here, Indranil wasassigned to a project funded by Genome Canada/GenomeAlberta, working as a senior postdoctoral fellow. Duringthis time, he was responsible for conducting research intothe composition of microbial communities within variedhydrocarbon resource environments using modernmetagenomic tools and evaluating biotechnologies toimprove oil production. Indranil was involved in several


62884araD7R1_ASC026 3/15/13 11:36 PM Page 5

projects with oil and gas companies before joining theNalco Technology Center in 2011.

Indranil received his B.Pharm. degree from theUniversity of Pune, Pune, India, and his M.S. degree inMolecular Genetics from the University of Leicester,Leicester, U.K. Following this, he successfully completed hisPh.D. degree with the dissertation “Senescence ofStaphylococci: Metabolic and Environmental FactorsDetermining Bacterial Survival and Persistence” at theInstitute of Medical Microbiology and Hygiene, Universityof Saarland-Hospital, Homburg, Germany. Indranilfollowed this with an additional 3 years of postdoctoralexperience in medical and infectious microbiology.

He has published in several peer-reviewed journals inthe areas of both medical microbiology and petroleummicrobiology.

Dr. Gerrit Voordouw has been aProfessor of Microbiology in theDepartment of Biological Sciences atthe University of Calgary since 1986and has held the NSERC IndustrialResearch Chair in PetroleumMicrobiology since 2007. As an

Industrial Research Chair holder, he works closely withmajor energy companies to coordinate the researchactivities in his lab focused on sulfur cycle management,corrosion control and improved production. Gerrit servedas a member of the Technical Advisory Committee to theSaudi Aramco Research & Development Center (R&DC)from 2009 to 2011.

In addition to researching practical aspects of petroleummicrobiology, he is project leader of a 4-year GenomeCanada funded project, aimed at characterizing themicrobial communities in hydrocarbon resourceenvironments through state-of-the-art DNA sequencingtechnologies. This project started in 2009 and involves 12co-investigators, as well as participation by other industryprofessionals.

Gerrit received his B.S. and M.S. degrees in Chemistryfrom the University of Utrecht, Utrecht, The Netherlands,in 1970 and 1972, respectively, and a Ph.D. degree inPhysical Biochemistry from the University of Calgary,Calgary, Alberta, Canada, in 1975.

Dr. Peter F. Sanders is a ResearchScience Consultant in Saudi Aramco’sResearch & Development Center(R&DC). He worked for 12 years as aSenior Microbiologist and ResearchManager for Oil Plus Ltd., an oil fieldconsultancy company in the U.K.,

working on solving microbiological problems for most ofthe major oil field operators all over the world. Prior tothat, Peter was a Research Fellow at Aberdeen University,Scotland, and ran a small oil field microbiology company

He joined Saudi Aramco in 2001, and has been workingon new technologies to predict, monitor, assess and controlmicrobial corrosion, biofouling and contaminationproblems in water injection, oil production, andtransportation and utilities systems. Peter has also beenstudying downhole microbial growth and microbiology inextreme environments to develop biotechnology-basedprocesses. He has also consulted widely within SaudiAramco to address operational problems caused bymicrobial growth in oil field systems.

He received his B.S., M.S. and Ph.D. degrees inMicrobiology from Exeter University, Exeter, U.K.

Dr. Tony Y. Rizk joined SaudiAramco’s Research & DevelopmentCenter (R&DC) in July 2006 and iscurrently a Science Specialist.Throughout his career in the oil andgas industry for well over two decades,Tony has initiated and managed a

number of research and deployment projects. He alsopioneered the development of new technologies that havebeen successfully implemented in the oil and gas industry.

Tony assumed a number of roles while at the R&DC,and he has been handling the Biotechnology TechnicalServices activities for the last two years. His work hasinvolved microbially induced corrosion, encapsulation fordownhole slow release, MEOR methodologies, reservoirsouring and control mechanisms, nitrate corrosion,corrosion inhibitor selection, and corrosion evaluationunder high shear stress and hydrotesting.

Tony has chaired a number of international and regionalconferences, including the Energy Institute ReservoirMicrobiology Forum, London, U.K. (2007-2008), the SaudiAramco Technical Exchange Forum (2009), TechnicalChairman of the Middle East Corrosion Conference(2011), Session Chairman of the Society of PetroleumEngineers (SPE) conference on MIC at Calgary, Canada(2009), and Session Chairman of both Chemindex andLabtech in Bahrain (2010) and Qatar (2011), respectively.He is also currently the Technical Chairman of the 15thMiddle East Corrosion Conference and Exhibition to beheld in Bahrain in 2014.

Tony received his B.S. in Industrial Engineering andgraduated with a Ph.D. in Corrosion Science fromManchester University, Manchester, U.K., in 1992.


62884araD7R1_ASC026 3/15/13 11:36 PM Page 6

ABSTRACT For many reasons, water production can increase earlierthan expected and impair the performance of the well. In some cases, horizontal wells have died suddenly3. Water shut-off (WSO) and remedial work is crucial to revive them and to reduce water cut to improve well performance. Successful shut-offs require an understanding of the water entry mecha-nism, the reservoir heterogeneities and the wellbore operations.Accurate diagnostics and successful remedial actions can lead to significant improvements in well performance. In onecase, this process led to the significant reduction of gas entry4.Because all wellbore and reservoir parameters and hetero-geneities are unique, each case requires a customized workflow.The feasibility of an intervention depends on the specific conditions and the environment in each case. For illustrationpurposes, this article presents several field examples, includingopen hole and cased hole completions, with well history and performance documented before and after the remedialwork.

INTEGRATED PRODUCTION LOGGING TOOL

As described in several publications, the production loggingtool, used in these case studies provides continuous multiphasevelocity distribution measurements and holdup data that arethen used to identify water entries, establish water profiles andanalyze complex horizontal flow behavior.

The vertical-axis orientation of the sensors enables themeasurement of mixed and segregated flow regimes, includingdirect independent measurement of gas velocity in a multi-phase horizontal well. All measurements are taken simultane-ously at the same depth level.

The tool runs are decentralized in highly deviated and hori-zontal wells to ensure proper sensor placement across the verti-cal axis. Caliper and tool orientation measurements enablereal-time calculation of the sensors’ positions.

Each spinner responds to the velocity of the fluid passingthrough it, which enables the calculation of the multiphase ve-locity profile. Each of the six electrical probes and six opticalprobes reads the localized water and gas holdup, which en-ables the calculation of the multiphase holdup profile. The cor-responding holdup and velocity profiles permit the calculationof the multiphase flow rate profile using dedicated algorithms.

Increases in water production can significantly reduce well performance and the life of a well, leading to decreased oilproduction. To mitigate this situation, water management iscrucial. Water influx can occur through several mechanismsand approach from several directions. Accurate diagnostic in-formation is important for the design of successful shut-offsand effective results. One option is to isolate the water produc-ing zone with a rigless water shut-off (WSO) technique, whichis less costly than the use of workover rigs for interventions.

This article presents case histories of five horizontal wellsdrilled in carbonate formations and producing excess water;three were completed in open hole and two were cased. A mul-tiphase production logging (MPL) tool, equipped with fiveminiaturized spinners for phase velocity measurement, and sixelectrical and six optical probes for holdup data, provided im-portant diagnostic data for the decision making on remedialactions. Using the tool data, the operator pinpointed the waterentries and performed shut-off operations based on the sourceof the entries and water flow profiles. Subsequent productiontest results showed that the water cut was reduced in all thewells. Examples from open and cased hole completions areshown, utilizing a number of different shut-off techniques. Inaddition, oil production was considerably increased in many ofthe wells. These results demonstrate that accurate diagnosticinformation and an integrated approach are keys to successfulrigless WSOs.

INTRODUCTION

Most horizontal wells are drilled to improve oil productionand to minimize water production. In addition, the drilling ofhorizontal sidetracks is increasing to further maximize oil re-covery. The monitoring and management of these wells arechallenging operations because their completions and interven-tions are complex, and it is difficult to obtain accurate diag-nostics in the complex flow regimes occurring in theirundulating deviations. It has been shown1, 2 that the use of anintegrated compact production logging tool with multiplemini-spinners can provide accurate information on water entries and flow profiles.

Comprehensive Diagnostic and WaterShut-off in Open and Cased HoleCarbonate Horizontal Wells

Authors: Nawawi A. Ahmad, Hussein S. Al-Shabebi, Dr. Murat Zeybek and Shauket Malik


62884araD8R1_ASC026 3/15/13 11:38 PM Page 1

WATER MANAGEMENT WITH SHUT-OFFS

The production of oil reservoirs is affected by water produc-tion, which can originate from either aquifers or water injec-tion. In fact, water production is a direct consequence ofhydrocarbon depletion in all fields5. The production of water,its handling at the surface and the re-injection process com-prise the “water cycle,” which must be effectively managed.

Water control services are one of the fastest and least costlyroutes to reduce operating costs and simultaneously improvehydrocarbon production. High water production can have ad-verse effects on reservoir performance, which can result in pro-duction losses; it can add to oil production cost with increasedlifting, separation and disposal costs, and it can lead to scaling,corrosion and degradation in the wellbore, tubing, flow linesand processing facilities. Water management is crucial to re-duce water production, optimize oil production and either in-crease well life or revive dead wells. The detection of waterentry intervals and the establishment of production profiles areneeded to gain the understanding of reservoir dynamics andwell performance that are necessary to achieve successful water isolation.

Whether options for effective WSO are feasible depends on several reservoir and well parameters and diagnostic results. The solutions can include recompletion, mechanicalisolation, chemical isolation and sidetracks. The option dis-cussed in this article involves isolating the water producingzone through a mechanical means that (except in one instance)does not require a workover rig, which is easier and cheaper toimplement.

FIELD EXAMPLES

All the wells presented in this article were drilled in the Juras-sic formation of a giant oil field. The formation is thick andhas high permeability. The formation is divided from top tobottom into lithostratigraphic zones 1 to 4; zones 2 and 3 aredivided into subzones A and B. The best reservoir quality is inzone 2, described as having been formed in a high energy, shal-low marine environment. The oil is of relatively light quality,and the formation water has a very high salinity, above 200Kppm total dissolved solids. The field has been under peripheralwater injection for a long time to maintain pressure and im-prove production. The multiphase production logging (MPL)tool, Fig. 1, was run in all these examples to determine the

multiphase flow profile and the water entry intervals. To date,no MPL runs have been made on these wells since the WSOjob. Performance of the wells is also provided before and afterthe WSO job (blue curve as water cut, green curve as oil pro-duction). For the operations presented in this article, pre-jobpreparation, including assembly of drilling history, productionhistory, open hole logs and well details, was carried out to en-sure effective data acquisition. During the job, data was trans-mitted and observations were communicated to the well sitefor real-time decisions to ensure the objectives of reduced water production were accomplished.

Well-A: Shut-off Job Using Inflatable Packer (Rigless Operation)

Background: Well-A, as illustrated in Figs. 2 and 3, was drilledunderneath a gas cap and completed as a horizontal cased holeoil producer in zones 2 and 3. Due to lost circulation encoun-tered at X265, the well was completed with a 4½” perforatedliner to avoid gas cusping. These lost circulation zones havetight intervals above and below, as illustrated by the blue arrows in track 9 of Fig. 2. The well has seven perforations selectively placed in zones 2 and 3.

Logging Job: The MPL tool was deployed using 2” coiled tub-ing (CT), with about 87% coverage of the completed interval;greater coverage was not attempted because of indications thatdeeper logging may cause the tool to get stuck. Subsequently,the relevant intervals were covered so that an appropriateWSO decision could be made. The logging was done undershut-in and natural flowing conditions.

Logging Results: Figure 2 shows the results, with the well

Fig. 1. MPL tool.


Fig. 2. Results from integration of production log and open hole formationevaluation data in Well-A.

62884araD8R1_ASC026 3/15/13 11:38 PM Page 2

sketch/perforated interval, flow profile and open hole data dis-played in tracks 2, 6 and 9, respectively. The MPL data analy-sis showed that all of the water production was identified ascoming from the second perforation (perfo-2) and the fourthperforation (perfo-4), and most of the produced oil was identi-fied as coming from the first perforation (perfo-1), as shown intrack 6, Fig. 2. Perforations 2 and 4 were in communication(indicated by downward water cross flow) during shut-in, asshown in track 8, Fig. 2. It was also observed that perforations3, 5, 6 and 7 were not contributing to oil production.

Shut-off Job: The MPL data showed that all of the water pro-duction was identified from below X600, and the open holelog data showed that there are tight intervals above X600.Therefore, a WSO job was performed by setting an inflatablepacker at X560 using CT, as illustrated in track 10, Fig. 2.Consequently, the producing perforated interval after the WSOjob is now distinctly above the tight zones at X560.

Shut-off Result: The production history is shown in Fig. 3,with the WSO event indicated by an orange dashed line. Afterthe WSO job, compared with values recorded during the MPLrun, the water production dropped sharply — by about 45%— and oil production nearly doubled. This is considered to bea successful WSO job and was done at relatively low cost.

Well-B: Shut-off Job through a Five-stage Cementing Job(Rigless Operation)

Background: Well-B, as illustrated in Figs. 4 and 5, was drilledand completed as a horizontal open hole oil producer overzones 2 and 3. Open hole logs showed several tight/low poros-ity intervals; the relevant one for this example is at X160(shown by a blue arrow in track 7, Fig. 4).

Logging Job: The MPL job was done using 2” CT, with about99% coverage of the completed interval. The logging was doneunder shut-in and natural flowing conditions.

Logging Results: Figure 4 shows the results, with the wellsketch, flow profile and open hole data displayed in tracks 2, 6and 7, respectively. The MPL data analysis showed that the

flow zones can be divided into three major units: X030 toX160, X160 to X400, and below X400. All of the water pro-duction was identified as coming from below X160, and mostof the produced oil (about 78%) was identified as comingfrom X030-X160 (track 6, Fig. 4).

Shut-off Job: A WSO job was accomplished through a five-stage cementing job, performed over the horizontal sectionfrom X150 to X570 (track 8, Fig. 4). Consequently, the pro-ducing open hole interval after the WSO job is now above thetight zone at X160.

Shut-off Result: The production history is shown in Fig. 5 withthe WSO event indicated by an orange dashed line. After theWSO job, compared with values recorded during the MPLrun, the water production dropped sharply by about 48%, andoil production has remained constant over a 4-month period.This is also considered a successful WSO job, as oil productionremained constant despite reducing the open hole length/com-pleted interval by about 75%.Fig. 3. Production history before and after WSO in Well-A.

Fig. 4. Results from integration of production log and open hole formationevaluation data in Well-B.

Fig. 5. Production history before and after WSO in Well-B.


62884araD8R1_ASC026 3/15/13 11:38 PM Page 3

Well-C: Shut-off Job Using Mechanical Plug and Cement(Rigless Operation)

Background: Well-C, as illustrated in Figs. 6 and 7, was drilledand completed as a horizontal cased hole oil producer in zone2A. Open hole logs showed several tight/low porosity inter-vals; the relevant one here is at X810 (shown by a blue arrowin track 7, Fig. 6). The well was completed with a 4½” equal-izer string and inflow control devices (ICDs), with seven equal-izer segments separated using six mechanical open holepackers. During the MPL job, the well was dead.

Logging Job: The MPL tool was deployed using 2” CT and afriction reducer to extend the reached depth and avoid the tool’sdamage, with 98% coverage of the completed interval. The log-ging was only done under shut-in conditions because the well wasdead and no attempt to date had been made to revive the well.

Logging Results: Figure 6 shows the results, with the wellsketch/equalizer intervals (marked with numbers for easy reference), shut-in profile and open hole data displayed in

tracks 2, 6 and 7, respectively. The MPL tool data analysisshowed that a strong upward cross flow of water was identi-fied as coming from the seventh equalizer segment (blank pipewith bull plug), and it was concluded that the bull plug wasbroken. This upward water cross flow was the reason prevent-ing the well from naturally flowing.

Shut-off Job: The MPL data showed that all of the watermovement (cross flow) during shut-in was coming from theseventh equalizer segment (below X880). Because of the tightinterval at X810, the shut-off job was done by setting a me-chanical plug at X560 (as illustrated in yellow in track 8, Fig.6) and pumping cement through it. Consequently, the produc-ing equalizer interval after the WSO job is now above the tightzone at X810.

Shut-off Result: The production history is shown in Fig. 7,with the WSO event indicated by an orange dashed line. Afterthe WSO job, the well was revived and flowed naturally with15% water cut, producing thousands of barrels of oil per day.This is a remarkably successful WSO job; a dead well was re-vived to produce oil at a high rate and relatively low water cut.

Well-D: Shut-off Job Using Blank Pipe and Equalizer String(Workover Rig Operation)

Background: Well-D, as illustrated in Figs. 8 and 9, was drilledand completed as a horizontal open hole oil producer overzone 2A. Open hole logs showed several tight/low porosity in-tervals; the relevant one here is at X250, shown by a blue ar-row in track 11, Fig. 8. The well was logged three times usingMPL tools over a four-year period, from 2005 to 2008.

Logging Job: The logging summary of each job is as follows:

Fig. 6. Results from integration of production log and open hole formationevaluation data in Well-C.

Fig. 7. Production history before and after WSO in Well-C. Fig. 8. Results from integration of production log and open hole formationevaluation data in Well-D.


62884araD8R1_ASC026 3/15/13 11:38 PM Page 4

• First run in 2005: The job was done with a 2” CT, with61% coverage over the objective interval, in both shut-in and flowing conditions.

• Second run in 2007: The job was done with 2⅜” CT,with 89% coverage over the objective interval, only inshut-in condition because of operational issues.

• Third run in 2008: The job was done with 2⅜” CT,with 99% coverage over the objective interval, to avoidthe tool’s damage. The relevant intervals were covered,enabling appropriate production and reservoir man-agement decisions. The well was logged under shut-inand natural flowing conditions.

Logging Results: Figure 8 shows the results of all three MPLjobs, with the well sketch, flow profile from 2005, shut-in profile from 2007, flow profile from 2008, shut-in profile from2008 and open hole data displayed in tracks 2, 5, 7, 9, 10 and11, respectively.

The logging result summary of each job is as follows:

• First run in 2005: The MPL showed no water, whichwere also in agreement with the test results. The MPLdata showed major oil entry (82% of total oil) atintervals between X250 and X330 (track 5, Fig. 8). Thiswas attributed to the presence of conductive fracturesover this interval. No cross flow was observed duringthe shut-in and flowing surveys.

• Second run in 2007: The MPL showed a strongdownward oil cross flow during the shut-in survey (noflowing survey was done). It was discovered that the zoneswith conductive fractures (between X250 and X330)were responsible for this cross flow (track 7, Fig. 8).

• Third run in 2008: The MPL showed a strongdownward oil and water cross flow during both shut-inand flowing surveys. During the flowing survey, theshallower fracture at X260 was bringing all the water tothe wellbore, as shown in track 9, Fig. 8. During theshut-in survey, it was also discovered that the zoneswith conductive fractures (between X250 and X330)were responsible for this cross flow (track 10, Fig. 8).

Shut-off Job: From the latest 2008 MPL data, the water

production was identified as coming from the shallow fractureat X260. The WSO job was done by installing an equalizerstring with bull plug (as illustrated in track 12, Fig. 8). Blankpipe was installed against the conductive fractures (betweenX250 and X330) to eliminate the cross flow and the major wa-ter production from this highly conductive interval.

Shut-off Result: The production history, Fig. 9, with the WSOevent is indicated by an orange dashed line. After the WSOjob, the water production dropped sharply to almost a dry oilwell, and oil production increased about three times. This isconsidered to be a successful WSO job, even though it wasdone at a high cost.

Well-E: Shut-off Job Using Inflatable Cement Retainer andCement Plug (Rigless Operation)

Background: Well-E, as illustrated in Figs. 10 and 11, wasdrilled and completed as a horizontal open hole oil producerover zones 2, 3 and 4. Open hole logs showed several tight/lowporosity intervals; the relevant one here is at X240, indicatedby a blue arrow on track 7, Fig. 10.

Logging Job: The MPL job was done using a 1¾” CT, with

Fig. 9. Production history before and after WSO in Well-D.

Fig. 10. Results from integration of production log and open hole formationevaluation data in Well-E.

Fig. 11. Production history before and after WSO in Well-E.


62884araD8R1_ASC026 3/15/13 11:38 PM Page 5

80% coverage over the objective interval. The limited coveragewas because of CT lockup; however, the relevant intervalswere covered, enabling appropriate production and reservoirmanagement decisions. The logging was done under shut-inand natural flowing conditions.

Logging Results: Figure 10 shows the results, with the wellsketch, flowing profile and open hole data displayed in tracks2, 6 and 7, respectively. The MPL tool data indicated that allof the water production was emanating from the fracturedzone intersecting the wellbore at X425, which also contributedabout 42% of the total oil. It was observed two years after ob-taining the MPL data, that the well was dead, possibly due toexcessive water production from the fractured zone at X425.

Shut-off Job: The WSO job was done by installing two inflat-able cement retainers, squeezed with cement, at X236 andX200; the top of the cement was at X183 (illustrated in track8, Fig. 10). Consequently, the producing open hole interval after the WSO job is now above the tight zone at X240.

Shut-off Result: The production history, Fig. 11, with the WSOevent indicated by an orange dashed line. After the WSO job,the water production dropped sharply by about 80%, and oilproduction increased by about one-third. This is considered tobe a successful WSO job.

CONCLUSIONS

Integration of MPL results, open hole data and other static anddynamic data is essential for a successful shut-off job (and otherproduction and reservoir management decisions). The presentedresults demonstrate that successful WSO is achievable in hori-zontals wells, even though there is a potential for water coningdue to the homogeneous character and high permeability ofthe reservoirs. It was also observed that reservoir barriers/lowpermeability intervals above the shut-off interval play an im-portant role in preventing water coning after the WSO job.

These field examples showed that increased water produc-tion can significantly reduce oil production and impair wellperformance. In one example, water production had causedthe horizontal well to become a dead well. As demonstrated inthat example, the execution of a successful WSO job can revivesuch a well and make it flow naturally at a high rate and atlow water cut.

Accurate production logging diagnostic input and a me-thodical shut-off design can lead to significant improvement inwell performance and increased well life. Although the riglessshut-off technique is generally desired because it is a fast andcost-effective intervention, the shut-off solution may requiremore expensive options, such as using a workover rig to installequalizer strings and ICDs and/or to sidetrack the well. Thesuccess of WSO depends on accurate problem diagnostics,careful job design and excellence in execution.

RECOMMENDATIONS

The following guidelines and recommendations will improvethe potential for WSO success:1. Production logging data should be recent when planning the

shut-off design and execution, as the reservoir dynamics canrapidly change, especially in mature fields.

2. Ensure there is a prominent reservoir barrier/low permeabil-ity zone above the shut-off interval, as shown by open hole log and/or image data.

3. Numerical simulation within an integrated petroleum engi-neering study will help assess more quantitatively the effec-tiveness of the shut-off job and the added value (cost, rate, etc.).

ACKNOWLEDGMENTS

The authors would like to thank Saudi Aramco managementand Schlumberger for their permission to present and publishthis article and to thank Mohammad M. Al-Mulhim for pro-viding relevant data.


REFERENCES

1. Baldauff, J., Runge, T., Cadenhead, J., Faur, M., Marcus, R., Mas, C., et al.: “Profiling and QuantifyingComplex Multiphase Flow,” Oilfield Review, Vol. 16, No. 3, October 1, 2004, pp. 4-13.

2. Al-Muthana, A.S., Ma, S.M., Zeybek, M. and Malik, S.:“Comprehensive Reservoir Characterization withMultiphase Production Logging,” SPE paper 120813,presented at the SPE Saudi Arabia Section TechnicalSymposium, al-Khobar, Saudi Arabia, May 10-12, 2008.

3. Nawawi, A., Bawazir, M., Zeybek, M. and Malik, S.:“Pinpointing Water Entries in Dead Horizontal Wells,”IPTC paper 15375, presented at the InternationalPetroleum Technology Conference, Bangkok, Thailand,February 7-9, 2012.

4. Al-Behair, A., Malik, S., Zeybek, M., Al-Hajari, A. andLyngra, S.: “Real Time Diagnostics of Gas Entries andRemedial Shut-off in Barefoot Horizontal Wells,” IPTCpaper 11745, presented at the International PetroleumTechnology Conference, Dubai, U.A.E., December 4-6,2007.

5. Bailey, B., Crabtree, M., Tyrie, J., Kuchuk, F., Romano, C.,Roodhart, L.; “Water Control,” Oilfield Review, Vol. 12,No. 1, 2000, pp. 30-51.


62884araD8R1_ASC026 3/15/13 11:38 PM Page 6


BIOGRAPHIES

Nawawi A. Ahmad is a PetroleumEngineer Specialist and is currently theLead Engineer for day-to-dayevaluation of production logs for allfields in Saudi Aramco. He started hisoil field career in 1989 with Shell inSoutheast Asia as a Well Site

Petroleum Engineer, Operational Petrophysicist and FieldStudy Petrophysicist in new and mature oil and gas fields.Nawawi then worked as a Senior Petrophysicist and fieldstudy leader for Petroleum development Oman in theMiddle East. His last position before joining Saudi Aramcowas as a division head of one of the petrophysic units in aShell operating company in Southeast Asia.

Nawawi received his B.Eng. degree in Mining andPetroleum Engineering from Strathclyde University,Glasgow, U.K., in 1989 and an M.B.A. from BruneiUniversity, Brunei, in 2005.

He has been a member of the Society of PetroleumEngineers (SPE) since 1989.

Hussain S. Al-Shabibi joinedSchlumberger Oilfield Services in 2006as a Borehole Production Engineer inthe Petro-Technical Services (PTS)segment. He has 6 years of experiencein job planning, real-time monitoringand post-acquisition data processing

and interpretation related to production logging in verticaland horizontal wells. Hussain also assists the company inthe marketing and support of integrated solutions.

In 2006, he received his B.S. degree in PetroleumEngineering from King Fahd University of Petroleum andMinerals (KFUPM), Dhahran, Saudi Arabia.

Hussain has been a member of the Society of PetroleumEngineers (SPE) since 2003.

Dr. Murat Zeybek is a SchlumbergerReservoir Engineering Advisor andReservoir and Production DomainChampion for the Middle East area.He works on analysis/interpretation ofwireline formation testers, pressuretransient analysis, numerical modeling

of fluid flow, water control, production logging andreservoir monitoring.

He is a technical review committee member for theSociety of Petroleum Engineers (SPE) journal ReservoirEvaluation and Engineering. Murat also served as acommittee member for the SPE Annual TechnicalConference and Exhibition, 1999-2001. He has been adiscussion leader and a committee member in a number ofSPE Applied Technology Workshops (ATWs), including atechnical committee member for the SPE Saudi TechnicalSymposium, and he is a global mentor in Schlumberger.

Murat received his B.S. degree from the TechnicalUniversity of Istanbul, Istanbul, Turkey, and his M.S.degree in 1985 and Ph.D. degree in 1991, both from theUniversity of Southern California, Los Angeles, CA, all inPetroleum Engineering.

Shauket Malik is currently working asa Senior Geoscientist in Saudi Arabiawith Schlumberger where he has beenfor over 20 years. He started his careerin Iraq as a Log Analyst (open hole)and then worked in Angola as a LogAnalyst (open and cased hole). Shauket

was transferred to Saudi Arabia, where he led the DataManagement group and then worked as a Log Analyst(open and cased hole) until 1999. In 2000, he wastransferred to Reservoir Domain and then to ProductionDomain, where currently he is performing vertical andhorizontal production analysis.

Sauket received his B.S. degree in Physics and a M.S.degree in Applied Mathematics (fluid mechanics anddynamics), both from Punjab University, Chandigarh,India.

Shauket is the author and coauthor of numerous paperson production domain.

62884araD8R1_ASC026 3/15/13 11:38 PM Page 7

ABSTRACT oil-water contact (OWC) can preclude any aquifer support andany effectiveness of water injection in the aquifer. In spite ofthe overriding impact of viscosity gradients in black oil, heavyoil and tar, there has been very little understanding of the ori-gin and distribution of these gradients. The reason for thisglaring deficiency in petroleum science and engineering is sim-ple to understand. The viscosity gradients in black oil/heavy oilsystems are dominated by asphaltene gradients, and until re-cently, there has been no proper theoretical framework for understanding the distribution of asphaltene gradients in oilreservoirs. For example, the ubiquitous use of the cubic equa-tion of state (EoS) in reservoir models traces back to the Vander Waals Equation, which was developed to treat gas-liquidequilibria and has no provisions for handling colloidal solids,such as the asphaltenes. The reason for the inability to treat as-phaltenes in thermodynamic models, so as to give asphaltenegradients, is quite clear; there has been a long-standing, orders-of-magnitude debate in the asphaltene science literature aboutthe size of asphaltene molecules1. If the size is unknown, thenthe effects of gravity are indeterminate, thereby precluding themodeling or prediction of gradients. In short, this deficiencyhas now been resolved: the molecular and colloidal sizes of asphaltenes in crude oil and in laboratory solvents have beencodified in the Yen-Mullins model2. Indeed, with this resolu-tion, the Flory-Huggins-Zuo Equation of State (FHZ EoS) hasbeen developed3 and proven to give accurate asphaltene gradi-ents in heavy oils4, black oils5 and condensates6.

In this article, a brief review of the new asphaltene formal-ism is given, showing that the formalism is extremely simplefor low gas-oil ratio (GOR) fluids. This simple formalism isthen applied to a double plunging anticlinal oil field (4-wayclosure) that has black oil in the crest, mobile heavy oil in theflank and a tar mat at the OWC. (For this work, mobile heavyoil is defined to have a viscosity less than ~1,000 cP; in manyfields such oil is produced conventionally.) It is shown that thesimple precepts herein properly account for detailed observa-tions; chemical analysis of the oils and tar show that the simplemodel captures the primary features of the data. Indeed, thetreatment of such important properties, such as the viscosity ofa large volume of oil over great distances, with a simple, effec-tive model might be called stunning. Certain unresolved issuesare discussed within the context of this new foundation of asphaltene science.

A Jurassic oil field in Saudi Arabia is characterized by black oilin the crest with mobile heavy oil underneath, all of it under-lain by a tar mat at the oil-water contact (OWC). The viscosi-ties in the black oil section of the column are fairly similar andare quite manageable from a production standpoint. In con-trast, the mobile heavy oil section of the column contains alarge continuous increase in asphaltene content with increasingdepth, extending to the tar mat. The tar shows very high as-phaltene content, but it is no longer monotonically increasingwith depth. Because viscosity depends exponentially on asphal-tene content in these oils, the observed viscosity varies fromseveral to ~1,000 centipoise (cP) in the mobile heavy oil andincreases to far greater viscosities in the tar mat. Both the ex-cessive viscosity of the heavy oil and the existence of the tarmat present major, distinct challenges in oil production. Con-ventional pressure-volume-temperature modeling of this oilcolumn grossly fails to account for these observations. Indeed,the very large height of this oil column poses a stringent chal-lenge for any corresponding fluid model. A simple new formal-ism used to characterize the asphaltene nanoscience in crudeoils, the Yen-Mullins model, has enabled development of theindustry’s first predictive equation of state (EoS) for asphaltenegradients: the Flory-Huggins-Zuo (FHZ) EoS. For a low gas-oil ratio (GOR) such as those in this field, the FHZ EoS re-duces to the simple gravity term. Robust application of theFHZ EoS, employing the Yen-Mullins model, accounts for themajor property variations in the oil column and by extension,the tar mat as well. Moreover, as these crude oils are largelyequilibrated throughout the field, reservoir connectivity is indi-cated in this field. This novel asphaltene science is dramaticallyimproving the understanding of important constraints on oilproduction.

INTRODUCTION

Huge viscosity gradients in oil columns have an enormous impact on production. Oil flow rate depends inversely on viscosity. Water sweep efficiency is greatly reduced when theviscosity ratio between oil and water exceeds ~5 centipoise(cP), causing water fingering instead of sweep. Tar mats at the

Black Oil, Heavy Oil and Tar in One OilColumn Understood by Simple AsphalteneNanoscience

Authors: Douglas J. Seifert, Dr. Murat Zeybek, Dr. Chengli Dong, Dr. Julian Y. Zuo and Dr. Oliver C. Mullins


62884araD9R1_ASC026 3/15/13 11:41 PM Page 1

ASPHALTENE NANOSCIENCE

The Yen-Mullins Model

After a lengthy literature debate, the centroid and distributionof asphaltene molecular weights and sizes has largely been re-solved by many different experimental methods and by manydifferent groups around the world7. In addition, there is nowextensive consensus on the nanocolloidal picture of asphaltenes.Most importantly, the fact that there are now two nanocol-loidal species of asphaltenes has a major bearing on asphalteneand viscosity gradients in oil reservoirs. The dominant molecu-lar and colloidal structures are represented in a model withprototypical structures, now called the Yen-Mullins model8. Aschematic of the model showing the nominal sizes of mole-cules, nanoaggregates and clusters is shown in Fig. 1.

Generally, different fields are seen to exhibit these sizeswithin 10% variability. It is not currently known whetherthere are actual size differences in the asphaltene nanoaggre-gates, varying from one oil to the next, or whether apparentdifferences are actually from errors in measurements. It is im-portant to note, however, that asphaltene molecular propertiesfrom many different crude oils are seen to be rather uniformand not dependent on the specific crude oil7.

The salient components of this nanoscience model are asfollows: asphaltene molecular weights are ~750 g/mole with arange of 500 g/mole to 1,000 g/mole. The predominant molec-ular architecture has a large central ring system with periph-eral groups (Fig. 1, Left). At low asphaltene concentrations,asphaltene molecules are not aggregated, and asphaltenes aredispersed as molecules; this applies to condensates6. At higherconcentrations, such as in black oils, asphaltene molecules self-assemble into nanoaggregates (of roughly six molecules) with asingle, central, disordered stack of aromatic groups (Fig. 1,Center). At yet higher asphaltene concentrations, for example,found in mobile heavy oil, asphaltene nanoaggregates self-as-semble into clusters of roughly eight nanoaggregates (Fig. 1,Right). These structures figure prominently when determiningthe direct effect of gravity on asphaltene gradients.

The FHZ EoS

With the size known for these distinct asphaltene species, afirst principles model can be developed for describing asphal-tene gradients. The Flory-Huggins equation has been used ex-tensively to describe asphaltene solubility and asphaltene phasebehavior9. Adding the gravity term to the Flory-Huggins equa-tion enables the calculation of asphaltene gradients in reser-voirs. This modification yields the powerful FHZ EoS10:

(1)

where OD(hi) is the optical density or oil color (typicallymeasured by downhole fluid analysis) at height hi in the oilcolumn, fa(hi) is the asphaltene concentration at height hi, va

is the molar volume of the asphaltene species of interest (eithermolecule, nanoaggregate or cluster, cf. Fig. 1), v is the molarvolume of the crude oil, g is the earth’s gravitational accelera-tion, Δρ is the density contrast between the asphaltene and theliquid crude oil, δa is the solubility parameter of the asphal-tene, δ is the solubility parameter of the crude oil, k is Boltz-mann’s constant, and T is temperature. The color of the crudeoil scales linearly with asphaltene content, as has been shownin numerous case studies.

The first term in the argument of the exponential is thegravity term. For low GOR black oils and heavy oils, the grav-ity term dominates. This gravity term contains Archimedes’buoyancy, which has had two millennia of validation, va Δρg.The asphaltenes are negatively buoyant (more dense) than theliquid crude oil. Newton’s force (F=ma) is mass times accelera-tion. With Archimedes’ buoyancy, it is not the total mass of theasphaltene species that matters but rather the effective buoyantmass, va Δρ (volume times density = mass). This buoyant massis multiplied by g to obtain the gravitational force on the as-phaltene particle. Of course, with larger asphaltene species(with larger volume va), the force is greater. In effect, the en-ergy required to lift an asphaltene particle off the base of theoil column to some height, h, equals the gravitational force, va Δρg, multiplied by h.

If gravity were the only determinant for the asphaltene dis-tribution, then all asphaltenes would be at the base of the oilcolumn; however, as Boltzmann showed over 100 years ago,available thermal energy can lift particles to higher energystates. In a gravitational field, this amounts to thermal energylifting particles off the floor to some higher height. The Boltz-mann distribution describes the population distribution ofground (E=0) and excited (ΔE) states in the very simple form:exp{-ΔE/kT}. This applies to all systems. Most importantly, theBoltzmann distribution represents an equilibrated state. Having particles in an excited state is not a transient condition;it is an equilibrium condition that will not change with time.

One system that clearly shows the Boltzmann distribution isthe earth’s atmosphere. If gravity were the only determinantfor the distribution of air molecules, then all air molecules

Fig. 1. The Yen-Mullins model of asphaltene science showing the predominantmolecular and colloidal structures of asphaltenes1. Left: At low asphaltene concen-trations such as in condensates, asphaltenes are dispersed as molecules. Center: Atlarger asphaltene concentrations such as in black oils, asphaltene molecules self-assemble, forming nanoaggregates with about six molecules per nanoaggregate.Right: At even higher asphaltene concentrations such as in (mobile) heavy oils,asphaltene nanoaggregates self-assemble, forming asphaltene clusters with abouteight nanoaggregates.


62884araD9R1_ASC026 3/15/13 11:41 PM Page 2

would be pulled to the surface of the earth and everyonewould suffocate. Thermal energy lifts air molecules to eleva-tions above the earth’s surface. Because air molecules are small(two heavy atoms in N2 and O2), the available thermal energylifts the air molecules to a great height. Here, the air moleculesare suspended in a vacuum, so the Boltzmann distribution issimply exp{-mgh/kT}, where m is the weighted molar mass ofthe air molecules, 80% N2 and 20% O2. This is what is plot-ted in Fig. 2 with T=298° Kelvin. Such a simple prediction,Fig. 2, closely matches observations.

For asphaltenes, one replaces m with va Δρ, thereby usingArchimedes’ buoyancy (essentially because the liquid is incom-pressible, so buoyancy is used), and the rest of the Boltzmanndistribution expression remains the same as for the atmosphere.For low GOR crude oils, the asphaltene gradient is predomi-nantly just given by the gravity term with all variables definedabove.

(2)

Asphaltene molecules contain ~70 heavy atoms, nanoaggre-gates contain ~400 heavy atoms, and clusters contain 3,000carbon atoms. Consequently, the gravitation gradient of asphaltenes depends critically on the particular asphaltenespecies. For a fixed thermal energy (temperature), asphaltenemolecules are suspended to a considerable height (but muchless than air molecules, which have only two heavy atoms),nanoaggregates are suspended less high, and clusters with their~3,000 heavy atoms reach the least height. Figure 3 shows thegradients for asphaltenes, presuming molecules, nanoaggregatesand clusters in a crude oil of 0.90 g/cc liquid phase density andT=393° Kelvin.

In Eq. 1, the second and third terms in the argument of theexponential incorporate the effects of entropy. This term tendsto be small, so it can largely be ignored. The effect of entropy

is to randomize or equally disperse the asphaltenes.The last term in the argument of the exponential of Eq. 1 is

the solubility term. In chemistry “like dissolves like,” and thischemical heuristic is formalized in the solubility term. For ex-ample, water and alcohol are mutually soluble since both haveOH groups. In contrast, oil with its CH groups is dissimilar towater with its OH groups, so oil and water are not mutuallysoluble. Here, given an interest in gradients, it is the variationof the solubility term with height in the oil column that is im-portant in establishing asphaltene gradients. The asphaltenesolubility parameter is determined by asphaltene chemicalproperties and is invariant, aside from a slight temperature dependence10. If the composition of the liquid oil does notchange in an oil column, then there is no variation of the solu-bility parameter or solubility term in Eq. 1 vs. height in the oilcolumn, so the gravity term still dominates.

The primary factor that determines whether or not there is avariation of the liquid oil solubility parameter (for equilibratedoil columns) is the solution gas content. Solution gas is a color-less gas, where asphaltenes are a dark brown solid — they arechemically very different and don’t dissolve in each other. Asphaltene does not partition to gas, making gas colorless.

Asphaltene does not dissolve well in crude oil with high so-lution gas. If there is a significant solution gas variation in anoil column, then there will be a large variation of the liquid oilsolubility parameter with height, and this can dominate cre-ation of an asphaltene gradient. Crude oils with low solutiongas have largely homogeneous solution gas. For these crude

Fig. 2. Calculated atmospheric pressure from the equation exp{-mgh/kT} using theweighted average of the molecular mass of air molecules (and 298° Kelvin) closelymatches observations. The prediction for Mount Everest is slightly high because ofthe assumption of constant room temperature. Virtually the same equation appliesto mobile heavy oil gradients, substituting the negative buoyancy of asphalteneparticles for mass2.

Fig. 3. The asphaltene gradient from the gravity term alone for the threeasphaltene species in the Yen-Mullins model from Fig. 1. The large clusters (5.0nm) show a rapid decline of % asphaltene with height, while the intermediatenanoaggregates (2.0 nm) and the small molecules (1.5 nm) show a very gradualdecline. For low GOR crude oils, the gravity term tends to dominate theasphaltene gradient, while for large GOR crude oils, the solubility term in theFHZ EoS can dominate the asphaltene gradient (cf. Eq. 1).


62884araD9R1_ASC026 3/15/13 11:41 PM Page 3

oils, the gravity term dominates. For crude oils with high solu-tion gas (>700 scf/bbl), there is a significant solution gas varia-tion, and the solubility parameter then becomes dominant,creating the asphaltene gradient. The GOR variation is largelytraceable to compressibility. Crude oils with high solution gasare compressible. As the hydrostatic head pressure of the oilcolumn increases density at the base of the column, the lightcomponents get “squeezed out” of the base, creating a solutiongas variation. Crude oils with low solution gas are incompress-ible. For these oils, the hydrostatic head pressure does not in-crease the oil density at the base of the column; therefore, thereis no density gradient to drive a compositional gradient.

BLACK OIL, HEAVY OIL AND TAR IN A SINGLE RESERVOIR

Mobile Heavy Oil

A large anticlinal structure contains a black oil reservoir of lowGOR11. The asphaltenes underwent some instability, formingthe mobile heavy oil section of the oil column and a tar mat atthe OWC. Here, the focus is on the mobile heavy oil and thetar mat in the field. Small fractions of the asphaltenes in theblack oil were destabilized, possibly by a gas or condensatecharge. The destabilized asphaltenes formed clusters, whichthen accumulated at the base of the oil column. In a local sec-tion of the field spanning roughly 8 kilometers, the asphaltenesare in clusters and are equilibrated, Fig. 4, in total agreementwith the reservoir scenario just discussed11.

Figure 4 also shows that the simple gravity term of the FHZEoS accounts for the huge increase in asphaltene content at aheight of 120 ft. Such a large height in the oil column and thecorresponding sixfold increase in the asphaltene content fromtop to bottom represent a stringent test of any model. Thegravity term has only one tightly constrained parameter, thesize of the asphaltene cluster. The fitted data gives a size of 5.2nm, which is a very close match to the nominal 5.0 nm clustersize, as previously shown in Fig. 1. Moreover, traditional mod-eling finds almost no asphaltene gradient because of the lack ofany GOR gradient. That is, traditional fluid modeling of themobile heavy oil fails miserably and here it is all but useless.

Asphaltene data from eight wells around the entire circum-ference of the field is shown in Fig. 5 (and includes the datafrom Fig. 4). The fit is very good, indicating that the simpleBoltzmann distribution of asphaltene clusters accounts for thehuge increase in asphaltene content in the height of the mobileheavy oil section for the entire circumference of the anticline.The FHZ EoS with the Yen-Mullins model represents a dra-matic improvement in the understanding of mobile heavy oilcolumns. Moreover, the measured size of the asphaltene clusterclosely matches that found in an Ecuador heavy oil column(5.0 nm)4 and in a Gulf of Mexico heavy oil column (5.2 nm)12.

Figure 5 also provides dramatic confirmation that asphalteneclusters are in thermodynamic equilibrium, as given by the

FHZ EoS. This indicates that this reservoir is in flow commu-nication — that is, it is a connected reservoir13. Gross differ-ences in asphaltene concentration in crude oil vs. height atdifferent reservoir locations could trigger convection, whichwould then rapidly smooth out these differences. In addition, it

Fig. 4. A local section of a large anticline with fluid data from three wells. Top:The asphaltene content vs. height agrees exactly with a simple equilibrium modelwith only one tightly constrained parameter, the size of the asphaltene cluster, heredetermined to be 5.2 nm, closely matching the nominal 5.0 nm cluster size in Fig.1. Bottom: The viscosity matches a simple Pal-Rhodes model, showing thatviscosity is largely exponentially dependent on asphaltene content.

Fig. 5. Data from eight wells shows that the mobile heavy oil column around theentire circumference of the field matches the simple gravity term of the FHZ EoSwith one tightly constrained parameter, the asphaltene cluster size (here 5.2 nm vs.the nominal 5.0 nm in Fig. 1). Moreover, the large height of the column yields afactor-of-6 variation of asphaltene content. This field represents an extreme test ofour simple model for mobile heavy oil – and represents the best data set there is(to the knowledge of the authors) to test thermodynamic modeling of mobileheavy oil.


62884araD9R1_ASC026 3/15/13 11:41 PM Page 4

is plausible that distal parts of the field underwent similargravitation accumulations of asphaltene to arrive at currentobservations of substantial uniformity around the flank. As-phaltene migration through reservoirs is a subject of currentresearch, and the consequence of this migration is seen repeatedly.

Above the mobile heavy oil section, there is less data. Theasphaltene content of the highest samples here is only a fewpercent. It is known that the oil in the crest, at a much greaterheight in the column, is black oil. At the asphaltene concentra-tion of a few percent (in this oil) is found the point of transi-tion from asphaltene cluster to asphaltene nanoaggregate. Atconcentrations lower than a few percent asphaltene, the as-phaltenes are dispersed as nanoaggregates. We saw in Fig. 3that the gradient of nanoaggregates is not so great. Even atmuch greater heights in this oil column, the oil remains blackoil. If asphaltenes were still within clusters even at low concen-trations, then the huge reduction of asphaltene concentrationwith height would continue until there would be almost no as-phaltenes, as previously shown in Fig. 3. In other words, if thehuge gradient of asphaltene concentration with height, whichholds for clusters, continued throughout the entire height ofthe oil column, then there would be a condensate (no as-phaltenes) practically on top of the mobile heavy oil section.This is not correct, and is resolved by postulating asphaltenesas being present as nanoaggregates at lower concentrations —thereby yielding much smaller gradients (cf. Fig. 3).

A critical component of the model of gravitation accumula-tion of asphaltenes is that the ratios of other saturates, aromat-ics, resins and asphaltenes (SARA) components are notchanging or are changing at a rate an order of magnitudeslower than the asphaltenes.

There is significant scatter in the SARA data, which is notthat unusual. Nevertheless, the trends are clear; the primaryvariation in the mobile heavy oil samples is their asphaltenecontent. The variations of ratios of other SARA fractions arefive to 10 times smaller. Indeed, if any other fraction were toassociate with asphaltenes, one would expect that to be resins.But, clearly, bulk resins are not accompanying asphaltenes.This limits an age old model showing strong asphaltene resinassociation. Figure 6 shows that bulk resins do not associatewith asphaltenes. Indeed, very similar results were obtained ina lab centrifugation experiment of live black oil.

Figure 7 shows the results from centrifugation of live blackoil14. This oil had a GOR of 800 scf/bbl, so both the solubilityterm and the gravity term contribute to establishing the asphal-tene and resin gradients. It took one month without seal loss toachieve equilibrium in this spin. The asphaltene gradient is~10x, while the resin gradient is 25% relative. Therefore, bulkresins are not migrating with the asphaltenes. Analysis of thecentrifugation results did conclude that a fraction of the heavi-est resins do associate with the asphaltenes. The picture thatemerges is that there is a molecular continuum going fromresins to asphaltenes. The criterion of n-heptane insolubility todefine the asphaltenes captures most but not all of the crude

oil fraction that self-assembles into aggregates (cf. Fig. 1)14.The field data presented in Figs. 5 and 6 is consistent with thecentrifugation data of Fig. 7. The asphaltenes by far dominatethe fraction of crude oil that self-assembles. Moreover, mobileheavy oils, such as those found in this study, have large asphal-tene fractions that are all in asphaltene clusters. These clustersequilibrate in the gravitational field, yielding large gradients(cf. Fig. 5).

Tar Mat

At the base of the mobile heavy oil section, Fig. 5 indicates thata tar mat was found. Several wells were drilled to intersect thistar mat for characterization. The organics were extracted fromcore sections at different depths in the tar mat and characterizedin terms of SARA fractions. Figure 8 shows an example of the

Fig. 6. For the mobile heavy oils plotted in Fig. 5, the primary variation is theasphaltene content. The variation of the other SARA fractions is a factor of 5 to10 smaller. This data shows consistency with the finding of a simple gravitationalequilibration of asphaltene clusters through the height and circumference of thefield.

Fig. 7. Live black oil centrifugation shows a similar result to that found in Fig.614. A giant asphaltene gradient (10x) was formed by centrifuging a live black oilwith moderate GOR so both the gravity term and the solubility term contribute tothe asphaltene gradient. Due to the lower asphaltene fraction in this black oil, theasphaltenes are present as nanoaggregates.


62884araD9R1_ASC026 3/15/13 11:41 PM Page 5

asphaltene content in the extracted tar vs. depth for two sepa-rate wells at the same depth scale.

Figure 8 also shows that there is a nearly random variationof tar with height in each of the two “tar” wells. The as-phaltenes are not equilibrated vs. height, even in a single well,which is a huge contrast to the heavy oil sections where the as-phaltene content is (or appears to be) largely equilibrated overthe circumference of the mobile heavy oil flank. Figure 8 alsoshows that there is no correlation of asphaltene concentrationlaterally for these two wells. The asphaltene content showslarge increases and decreases over very short vertical distances.

The mobile heavy oil section was shown to be characterizedby a simple gravitational accumulation and equilibration of as-phaltene vs. depth. Figure 8, on the other hand, shows that theasphaltene content of the tar is not even monotonic with depthand does not even approximate any equilibration. It is impor-tant to check whether the tar is simply an accumulation of as-phaltene in oil or whether other SARA fractions show largevariations in the tar as well.

According to Fig. 8, there is a huge variation of asphaltenecontent in the tar. Since the asphaltene content shows largevariations, the other SARA fractions must also show varia-tions; the sum of all SARA fractions must add to 1. Therefore,it is the ratio of the other SARA fractions that is of interest.Figure 9 shows the ratios of asphaltenes to paraffins, aromat-ics to paraffins and resins to paraffins. By far the largestchange is in the asphaltene-to-paraffin ratio. That is, the tar isprimarily an addition of a variable amount of asphaltene to anoil with fixed ratios of paraffins (or saturates), aromatics andresins.

Figure 9 also shows that the tar is dominated by changes inasphaltene content. Indeed, the variation of the asphaltenecontent is enormous, in one well changing from ~30% to65%. This picture is consistent with the origin of tar in thisfield as being due to the gravitational accumulation of asphal-tene at the base of the oil column, and it is consistent with thesame conclusion drawn for the origin of the mobile heavy oilcolumn immediately above the tar column. The primary differ-

ences between the tar and the mobile heavy oil is that: (1) themobile heavy oils have asphaltene content less than ~30% (cf.Fig. 8), while the tar has asphaltene content greater than~30%, and (2) the mobile heavy oil is vertically and laterallyequilibrated, while the tar is not equilibrated even over shortvertical distances, let alone large lateral distances. Two factorsplay an important role in equilibration: distance and viscosity.

Figure 10 shows the viscosity as a function of asphaltene con-tent in an oil phase of fixed composition15. This viscosity profileis not that of the oil and tar presented in this article, but never-theless shows the dependence of viscosity on asphaltene content.

Figure 10 provides a plausible reason why the tar is notequilibrated, while the mobile heavy oil directly above the taris equilibrated. (“Equilibrated” here means that the asphaltenecontent is varying monotonically vs. depth according to Eq. 2.)By showing that the viscosity is high at 30% asphaltene content, and that every 5% increase in asphaltene content is

Fig. 8. Asphaltene content vs. depth for tar wells below the mobile heavy oilsection in two wells (cf. Fig. 5). The asphaltene content does not varymonotonically, even in a single well. In addition, there is no lateral correlation ofasphaltene content, in contrast to the mobile heavy oil sections. In the tar mat,there are large increases and decreases of asphaltene within very small intervals ofheight.

Fig. 9. The SARA fractions are divided by paraffins vs. asphaltene content forsamples from two “tar” wells (saturates = paraffins). By far the largest variation isin the asphaltene/paraffin ratio; the aromatic/paraffin ratio and the resin/paraffinratio exhibit much smaller changes. Consequently, the tar can largely be describedas having large, variable asphaltene content in an oil of fixed composition.

Fig. 10. Viscosity is shown to depend exponentially (or more) on asphaltenecontent for several different carbonaceous systems15. For the range of %asphaltene relevant to the mobile heavy oil and tar sections of the hydrocarboncolumn, the viscosity in this figure increases by a factor of 100,000,000. Note thehydrocarbon system is not the crude oil and tar column from this field, but thedependence of viscosity on asphaltene is similar.


62884araD9R1_ASC026 3/15/13 11:41 PM Page 6

associated with another huge increase in viscosity, Fig. 10 indi-cates that the viscosity in sections of the tar mat is extraordi-narily high, precluding equilibration.

Plausible Geoscenarios Matching Field Observation

This Jurassic reservoir initially contained black oil. A subse-quent charge of a lighter hydrocarbon could have occurred because, in a normal burial sequence, the kerogen generateslighter hydrocarbons with longer times and greater tempera-tures. The lighter hydrocarbon often goes to the top of thereservoir without good mixing16. This lighter hydrocarbon (itcould even be gas) can diffuse into the oil column, causing instability of the asphaltene17, 18. If the instability is not toogreat, the asphaltenes can migrate great distances in the reser-voir, in some cases going to the base of the reservoir. High con-centrations of asphaltenes at or near the OWC can thereforeoccur. One can imagine separate destabilizing events yieldingpulses of asphaltenes, all snowing down towards the OWC. Athigh asphaltene concentrations, the viscosity increases, and ifthe viscosity increase is also associated with a permeability re-striction in the reservoir, then low viscosity tar can becometrapped or “perched” below the high viscosity tar. At somehigh asphaltene concentrations, there might also be a phasetransition, yielding a phase very rich in asphaltenes that mightblock pore throats. This is under investigation. If this occurs, itrepresents a second mechanism that can cause lower viscositytar to be trapped underneath higher viscosity tar. For asphal-tene concentrations below 30%, the viscosity is sufficientlylow that diffusion enables equilibration of the asphaltene inthe mobile heavy oil section.

CONCLUSIONS

Traditional EoS modeling of heavy oils has failed miserablydue to: (1) the previous lack of knowledge about asphaltenecolloidal sizes, and (2) the lack of a proper model to treat col-loidal solids in crude oil. The Yen-Mullins model of asphaltenenanoscience specifies the size of three distinct species of as-phaltenes: molecules, nanoaggregates and clusters. Thisnanoscience model enables accounting for the effects of grav-ity, which has been incorporated into the FHZ EoS for asphal-tene gradients. Moreover, for mobile heavy oils, only thegravity term contributes significantly to asphaltene gradients.In a field in Saudi Arabia, a mobile heavy oil rim has been fitto the model using a simple exponential equation (the Boltz-mann distribution). Moreover, the asphaltene content varies bya factor of six within this height. The simple Boltzmann distri-bution of asphaltene clusters accounts for this entire volume ofmobile heavy oil. SARA analysis of the crude oil confirms thatthe mobile heavy oil column simply has added asphaltene intoa crude oil of fixed composition. A tar mat below the mobileheavy oil does not show a monotonic increase of asphaltenestowards the base. This is linked to the extraordinarily high

viscosities within the tar mat. SARA analysis of the tar estab-lishes that, similar to the mobile heavy oil, there is variable asphaltene added to a crude oil of fixed composition.

Gravitational accumulation of asphaltenes at the low points ofthe reservoir is consistent with all observations. The applicationof new asphaltene science to heavy oils is seen to greatly improvethe understanding and prediction of reservoir observations.

ACKNOWLEDGMENTS



REFERENCES

1. Mullins, O.C.: “The Modified Yen Model,” Energy &Fuels, Vol. 24, No. 4, January 19, 2010, pp. 2,179-2,207.

2. Mullins, O.C., Sabbah, H., Eyssautier, J., Pomerantz, A.E.,Barré, L., Andrews, A.B., et al.: “Advances in AsphalteneScience and the Yen-Mullins Model,” Energy & Fuels, Vol.26, No. 7, April 18, 2012.

3. Freed, D., Mullins, O.C. and Zuo, J.Y.: “TheoreticalTreatment of Asphaltene Gradients in the Presence of GORGradients,” Energy & Fuels, Vol. 24, No. 7, June 3, 2010,pp. 3,942-3,949.

4. Pastor, W., Garcia, G., Zuo, J.Y., Hulme, R., Goddyn, X.and Mullins, O.C.: “Measurement and EoS Modeling ofLarge Compositional Gradients in Heavy Oils,” SPWLApaper, presented at the 53rd Annual Logging Symposium,Cartagena, Colombia, June 16-20, 2012.

5. Betancourt, S.S., Dubost, F.X., Mullins, O.C., Cribbs,M.E., Creek, J.L. and Mathews, S.G.: “PredictingDownhole Fluid Analysis Logs to Investigate ReservoirConnectivity,” IPTC paper 11488, presented at theInternational Petroleum Technology Conference, Dubai,U.A.E., December 4-6, 2007.

6. Elshahawi, H., Shyamalan, R., Zuo, J.Y., Mullins, O.C.,Dong, C. and Zhang, D.: “Advanced Reservoir EvaluationUsing Downhole Fluid Analysis and Asphaltene Flory-Huggins-Zuo Equation of State,” paper prepared for the53rd Annual Logging Symposium, Cartagena, Colombia,June 16-20, 2012.

7. Mullins, O.C., Sheu, E.Y., Hammami, A. and Marshall,A.G., eds.: Asphaltenes, Heavy Oils and Petroleomics, NewYork: Springer, 2007.

8. Sabbah, H., Morrow, A.L., Pomerantz, A.E. and Zare,R.N.: “Evidence for Island Structures as the DominantArchitecture of Asphaltenes,” Energy & Fuels, Vol. 25,No. 4, March 8, 2011, pp. 1,597-1,604.


62884araD9R1_ASC026 3/15/13 11:41 PM Page 7

9. Buckley, J.S., Wang, J. and Creek, J.L.: “Solubility of theLeast-Soluble Asphaltenes,” in Asphaltenes, Heavy Oilsand Petroleomics, eds. O.C. Mullins, E.Y. Sheu, A.Hammami and A.G. Marshall, New York: Springer, 2007.

10. Zuo, J.Y., Mullins, O.C., Freed, D. and Zhang, D.: “A Simple Relation between Solubility Parameters and Densities of Live Reservoir Fluids,” Journal of Chemical and Engineering Data, Vol. 55, No. 9, May 4, 2010, pp. 2,964-2,969.

11. Mullins, O.C., Seifert, D.J., Zuo, J.Y., Zeybek, M., Zhang, D. and Pomerantz, A.E.: “Asphaltene Gradients and Tar Mat Formation in Oil Reservoirs,” WHOC12-182 paper, presented at the World Heavy Oil Conference,Aberdeen, Scotland, September 10-13, 2012.

12. Nagarajan, N.R., Dong, C., Mullins, O.C. and Honarpour, M.M.: “Challenges of Heavy Oil Fluid Sampling and Characterization,” SPE paper 158450, presented at the SPE Annual Technical Conference and Exhibition, San Antonio, Texas, October 8-10, 2012.

13. Pfeiffer, T., Reza, Z., Schechter, D.S., McCain, W.D. and Mullins, O.C.: “Fluid Composition Equilibrium; A Proxy for Reservoir Connectivity,” SPE paper 145703, presentedat the SPE Offshore Europe Oil and Gas Conference and Exhibition, Aberdeen, Scotland, September 6-8, 2011.

14. Indo, K., Ratulowski, J., Dindoruk, B., Gao, J., Zuo, J.Y. and Mullins, O.C.: “Asphaltene Nanoaggregates Measured in a Live Crude Oil by Centrifugation,” Energy& Fuels, Vol. 23, No. 9, August 7, 2009, pp. 4,460-4,469.

15. Lin, M.S., Lumsford, K.M., Glover, C.J., Davison, R.R. and Bullin, J.A.: “The Effects of Asphaltenes on the Chemical and Physical Characteristics of Asphalt,” in Asphaltenes: Fundamentals and Applications, eds. E.Y. Sheu and O.C. Mullins, New York: Plenum Press, 1995, pp. 155-76.

16. Stainforth, J.G.: “New Insights into Reservoir Filling and Mixing Processes,” in Understanding Petroleum Reservoirs: Toward an Integrated Reservoir Engineering and Geochemical Approach, eds. J.M. Cubit, W.A. England and S. Larter, Special Publication, London: Geological Society, 2004.

17. Elshahawi, H., Latifzai, A.S., Dong, C., Zuo, J.Y. and Mullins, O.C.: “Understanding Reservoir Architecture Using Downhole Fluid Analysis and Asphaltene Science,” SPWLA-FF paper, presented at the 52nd Annual Logging Symposium, Colorado Springs, Colorado, May 14-18, 2011.

18. Zuo, J.Y., Elshahawi, H., Dong, C., Latifzai, A.S., Zhang,D. and Mullins, O.C.: “DFA Asphaltene Gradients for Assessing Connectivity in Reservoirs under Active Gas Charging,” SPE paper 145438, presented at the SPE Annual Technical Conference and Exhibition, Denver, Colorado, October 30-November 2, 2011.

BIOGRAPHIES

Douglas J. Seifert is a PetrophysicalConsultant with Saudi Aramco, wherehe works as the PetrophysicsProfessional Development Advisor inthe Upstream ProfessionalDevelopment Center (UPDC). Dougspecializes in real-time petrophysical

applications and fluid analysis. Before joining SaudiAramco in 2001, he was the Western Hemisphere RegionalPetrophysicist for Pathfinder Energy Services in Houston,TX, and the Eastern Hemisphere Regional Petrophysicist inStavanger, Norway. Doug also worked as the SeniorPetrophysicist for Mærsk Olie og Gas in Denmark; forHalliburton Energy Services in various operational,research and technical support functions; and for Texaco intheir Technical Services and Production Operations.

Doug is the President of the Saudi Petrophysical Society,the Saudi Arabian Chapter of the Society of Petrophysicistsand Well Log Analysts (SPWLA), and he also serves on theSPWLA Technology Committee.

He received a B.S. degree in Statistics and a M.S. degreein Geology, both from the University of Akron, Akron,OH.

Dr. Oliver C. Mullins is a ScienceAdvisor to Executive Management inSchlumberger. He is the primaryoriginator of downhole fluid analysisfor formation evaluation. For this, hehas won several awards, including theSociety of Petroleum Engineers (SPE)

Distinguished Membership Award and the Society ofPetrophysicists and Well Log Analysts (SPWLA)Distinguished Technical Achievement Award; Oliver alsohas been a Distinguished Lecturer four times for theSPWLA and SPE.

He authored the book The Physics of Reservoir Fluids:Discovery through Downhole Fluid Analysis, which wontwo Awards of Excellence. Oliver has also co-edited threebooks and coauthored nine chapters on asphaltenes. Hehas coauthored >190 publications and has ~3,100literature citations. Oliver has co-invented 80 allowed U.S.patents. He is a fellow of two professional societies and isAdjunct Professor of Petroleum Engineering at Texas A&MUniversity. Oliver also leads an active research group inpetroleum science.

Dr. Murat Zeybek is a SchlumbergerReservoir Engineering Advisor andReservoir and Production DomainChampion for the Middle East area.He works on analysis/interpretation ofwireline formation testers, pressuretransient analysis, numerical modeling

of fluid flow, water control, production logging andreservoir monitoring.

He is a technical review committee member for theSociety of Petroleum Engineers (SPE) journal Reservoir


62884araD9R1_ASC026 3/15/13 11:41 PM Page 8

Evaluation and Engineering. Murat also served as acommittee member for the SPE Annual TechnicalConference and Exhibition, 1999-2001. He has been adiscussion leader and a committee member in a number ofSPE Applied Technology Workshops (ATWs), including atechnical committee member for the SPE Saudi TechnicalSymposium, and he is a global mentor in Schlumberger.

Murat received his B.S. degree from the TechnicalUniversity of Istanbul, Istanbul, Turkey, and his M.S.degree in 1985 and Ph.D. degree in 1991, both from theUniversity of Southern California, Los Angeles, CA, all inPetroleum Engineering.

Dr. Chengli Dong is a Senior FluidProperties Specialist in the ShellFEAST team (Fluid Evaluation andSampling Technologies), andpreviously he was a SchlumbergerReservoir Domain Champion. Chenglihas been a key contributor on the

development of downhole fluid analysis (DFA) as well asDFA applications in reservoir characterization. Heconducted extensive spectroscopic studies on live crude oilsand gases, and led the development of interpretationalgorithms on the DFA tools. In addition, Chengli hasextensive field experience in design, implementation andanalysis of formation testing jobs.

He has published more than 50 technical papers, and heco-invented 10 granted U.S. patents and nine patentapplications.

Chengli received his B.S. degree in Chemistry fromBeijing University, Beijing, China, and his Ph.D. degree inPetroleum Engineering from the University of Texas atAustin, Austin, Texas.

Dr. Julian Y. Zuo is currently aScientific Advisor in ReservoirEngineering at the SchlumbergerHouston Pressure and SamplingCenter. He has been working in the oiland gas industry since 1989. Recently,Julian has been leading the effort to

develop and apply the industry’s first simple Flory-Huggins-Zuo equation of state (EOS) for predicting compositionaland asphaltene gradients to address a variety of major oilfield concerns such as reservoir connectivity, tar matformation, asphaltene instability, flow assurance,nonequilibrium with late gas charging, etc.

He has coauthored more than 140 technical papers inpeer-reviewed journals, conferences and workshops.

Julian received his Ph.D. degree in Chemical Engineeringfrom the China University of Petroleum, Beijing, China.


62884araD9R1_ASC026 3/15/13 11:41 PM Page 9

ABSTRACT through wall cake is accredited to the failure of cement tomake a good bond to the formation2. There are several factorsthat directly affect this phenomenon, including the type of ce-ment used, chemical additives, mud and cement density, tem-perature, pressure, mud cake film, centralization, movement ofcasing string and reciprocation while pumping the cementslurry and cement filtrate3.

Most of the theories have been developed to address prob-lems associated with cement channeling that leads to severeloss in the hydrostatic pressure of the cement column duringgelation (i.e., when cement passes from liquid to a solid state).Most of the suggested solutions in the past have focused onone property of the cement but neglected the change in otherproperties, assuming that changes that may be directly respon-sible for the gas migration occur only for some of the physicalor chemical properties. Casing centralization, use of ascratcher to clean mud cakes and use of fluid spacers are someof the solutions implemented to help improve zonal isolation4.

The cement is capable of transmitting pressure as long as itis in the liquid state until it attains gel strength — enough toform an effective seal5. Upon cement placement, cement suffersa gradual drop in hydrostatic pressure, following a downwardgradient until it reaches that of water. Hydrostatic pressurewill further drop and become less than that of water duringgelation due to dehydration within the cement matrix and fluidloss. Excessive dehydration rates and fluid loss will cause highshrinkage that might form a passage for formation fluids totransfer from a high-pressure zone into a low-pressure zoneand into the well through filter cake or casing leak6. Cooke7

studied the pressure behavior during the first six hours afterpumping cement and observed that cement loses pressure at 39psi/ft. Also, he found that application of annular pressure canmake up for the drop, maintaining the pressure required tooverbalance the formation until the cement develops enoughcompressive strength for effective zonal isolation.

Gas or fluid migration will not take place if the cement isable to develop gel strength between 500 lbf/100 ft2 within 15minutes after the start of transition time8. It would be impossi-ble for gas to migrate at 500 lbf/100 ft2, especially if the ce-ment has low permeability, zero free water, high gel strength,low viscosity and a short transition time. In such situations,gas will enter into the cement matrix and create channels

Cementing is one of the most important and crucial issues inoil fields, especially for high-pressure and gas bearing forma-tions. It is difficult to achieve a good zonal isolation in suchformation types because pressure is abnormal and formationfluid contains corrosive fluids and gases. A common problemassociated with highly over-pressurized zones is cross flow after cementing. Fluid flow from an over-pressured zone to alow-pressure, high permeability zone can lead to deteriorationof the existing production hardware. Workover operations thatattempt to repair cement voids, including perforation, squeez-ing and use of casing patches or scab liners, are not recom-mended as they do not provide long-lasting results.

One onshore field in Saudi Arabia has experienced a persist-ent problem related to cementing at high-pressure zones. Re-cently, communication between Formation-A (an abnormallyover-pressurized zone) and Formation-B (a low-pressure zone)is occurring with increasing frequency due to long-term seawa-ter injection, which has resulted in production interruption inseveral wells. This article addresses the problems by investigat-ing field practices that include drilling, cementing and comple-tion. It also reviews the field reports and cased hole logs forthe affected wells. Three-month and six-month studies wereconducted to evaluate the effects of Formation-A water on ce-ment, where the cement was exposed to Formation-A waterunder downhole conditions. Tests for mechanical properties,including permeability, a thermogravimetric analysis (TGA)and tests using energy dispersive X-ray fluorescence (EDXRF)are presented, in addition to discussions of some of the prelim-inary findings.

INTRODUCTION

Cement channeling is viewed as one of the major completionissues in the petroleum industry. Several attempts have beenmade by cementing companies and individual researchers totackle this problem; however, so far there is no reputable im-provement. Fluid migration in cement happens in the course ofspotting cement or afterwards. The main cause of gas channel-ing is believed to be the inability of cement to maintain enoughpressure on the formation before it sets1. Fluid migration


Cementing Abnormally Over-pressurizedFormations in Saudi Arabia

Authors: Abdulla F. Al-Dossary and Scott S. Jennings

62884araD10R1_ASC026 3/15/13 11:42 PM Page 1

within the cement. Sometimes it overcomes the tensile strengthof the cement structure, breaks the cement matrix and travelsthrough the micro-fractures. It is assumed that the hydrostaticpressure of the cement column will decrease when the gas bub-bles are already inside and that the gas will try to expand untilthe pressure difference is large enough to overcome the cementtensile strength and in turn break the cement9, 10.

On the other hand, water does not migrate in the samemanner as gas since it is not compressible11. There is no wayfor liquid, i.e., water or oil, to travel up or down the cementcolumn unless there are channels big enough within the cementfor it to flow through. Such channels possibly can form aftergas migration, when the channels get wider and wider due tothe high-pressure/high temperature (HP/HT) environment12.During cementing at an over-pressurized zone, the formationmight be underbalanced before the cement becomes strong inthe sense that it resists fluid movement. If this happens, forma-tion fluid will displace or squeeze cement into the formationsabove or below the high permeability zone, eventually result-ing in a non-cemented pipe.

Improper drilling practices also contribute to poor cement-ing to some extent. For instance, drilling with mud that leadsto uncontrolled fluid loss leaves excessive filter cake that is dif-ficult to remove. It has become evident that filter cake gives upat 2 psi13. Also, high mud weight along with high circulationrate while drilling through high permeability zones or a low-pressure, highly porous formation encourages fractures andwash outs to develop, which are difficult to cement. Hole con-ditioning practices are vital to a successful primary cementingjob. The hole should be clear of fill and filter cake, and ingauge before cementing; therefore, a clean out trip with a holeopener is required to further clean the hole by removing anyremaining filter cake and gelled mud. Other means, like use oflow viscosity mud and high circulation rate, will help effec-tively remove wall cake and mud pockets14. Mud buckets,which emerge when mud remains static a long time in the holewith formation cuttings inside, provide a route for fluid afterthey dehydrate.

It is very challenging to have an effectively cemented pipe inhighly deviated and horizontal holes. A couple of factors thatplay an important role in cementing such wells include central-izing, mud displacement efficiency and hole cleaning. It is well-known from past research that fluid tends to flow more in awide side offering least resistance than in a low side that re-stricts flow15. To overcome this problem, the number of cen-tralizers needs to be selected in a way that improves standoffwithout increasing drag, which might present additional prob-lems. Also, the design or shape of the centralizers should beoptimized in a way that helps provide a uniform flow regimearound the pipe and improve the displacement efficiency ratio.A spacer volume that provides a four-minute contact time withthe hole and the use of low viscosity mud at a circulation rateof 3 barrels per minute (bpm) will improve filter cake removalefficiency according to field and lab results. Moreover, the

spacer should be compatible with cement, as well as lighterthan the cement and heavier than the mud in the hole, to im-prove displacement efficiency and avoid mud channeling andcement contamination.

During the life of the well, the cement sheath is vulnerableto failure when different events take place, such as stimulation,well testing, communication testing, casing pressure tests andcement squeeze jobs, which generate thermal and cyclicstresses as a result of changes in hydrostatic pressure and tem-perature16. Mechanical stresses generated by tubular run inholes also contribute to cement fracture in the long term. Ce-ment contracts and expands frequently in response to tempera-ture changes, and if this movement exceeds the cement tensilestrength, cement will fracture. Radial expansion of the cement-casing interface, due to high-pressure induced stresses, willradically compress the cement and induce tensile tangentialstress that can cause a crack. When that happens, the tensilestrength of the cracked section will drop to zero, and the distri-bution of stress in the cement will be changed. This change willhelp the cracks creep outward and eventually reach the casingformation interface. If the crack occurs across the long axialdistance, a channel will form through which liquid can flow17.

Cement deterioration can accelerate in the presence of cor-rosive CO2 gas. The effect of CO2 is much worse in HP/HTformations. In such an environment, cement degradation dueto carbonation will occur in a short time. Three differentchemical reactions occur when cement comes in contact withCO2:

• Formation of Carbonic Acid (H2CO3): It lowers pH. Itseffect depends on temperature, partial CO pressure andother ions dissolved in the water.

• Carbonation of Cement or Cement Hydrates: It causesan increase in density, which leads to the increasedhardness and decreased permeability of the cementsheath. As a result, CO2 diffusion will decrease andvolume will increase by up to 6%. In such cases, crackswill develop.

• Dissolution of CaCO3: This phenomenon happens inthe presence of water containing CO2 for a long periodof time. Effects of this reaction include an increase inpermeability and porosity and a loss of mechanicalintegrity. This dissolution process will lead to poorformation isolation.

It is still in dispute whether or not carbonation is detrimen-tal to cement integrity. Some researchers showed that the me-chanical properties of cement will suffer degradation due CO2

exposure, leading to fluid migration. On the other hand, somestudies conducted on 20- to 30-year old cement samples fromCO2 wells showed that they maintained their integrity despitecarbonation. Cement mainly consists of tricalcium silicate(C3S) and dicalcium silicate (C2S). When cement reacts withwater, calcium silicate hydrate (C-S-H) and calcium hydroxide(Ca2) evolves. During exposure to CO2 dissolved in water,


62884araD10R1_ASC026 3/15/13 11:42 PM Page 2

calcium carbonate (CaCO3) will form. This product is harmfulto cement sheaths at high concentrations because it cracks ce-ment. There are two solutions to minimize the carbonation ef-fect and prolong the life of cement18-20:

• Reduce the cement permeability so that it withstandswell operations with low dehydration volume shrinkage.

• Optimize the cement design so that dehydrationproducts have fewer materials that are reactive withCO2.

Cement mechanical properties, including compressivestrength, yield stress, permeability, Young’s modulus and Pois-son’s ratio, should be taken into account when designing a cement system to guarantee that the cement will survive longerwhen exposed to cyclic loads. Physical properties like thickeningtime, fluid loss and viscosity also must be considered carefully tohelp reduce transition time and achieve the required compres-sive strength as quickly as possible. Cement that is mechanically,thermally and chemically stable will be able to survive HP/HTand corrosive environments.

COMMUNICATION PROBLEM BETWEEN FORMATIONSA AND B

A communication problem between Formations A and Bemerged recently in several newly drilled and sidetrackedwells. Three wells showed a recurrence of Formation-A casingleak. This problem is a big concern, and quick intervention isneeded before it escalates and becomes a major issue. The reason why the leak occurred has not been identified yet; however, there are three possible explanations for how it devel-oped. First, Formation-A water made its own way behind thecement, through the mud cake and into the well, since Forma-tion-A pressure is higher than the pressure of the productivezone across Formation-B. Second, Formation-A gas transferredthrough cement channels and reacted with the casing, whichmeans the casing got corroded and holes developed, paving theway for Formation-A water to enter the wellbore and eventu-ally kill the well. Third, water influx attacked the cement andcreated a severe contamination because the cement hydrostaticpressure was not enough to overbalance the Formation-A highpressure, allowing communication to take place during weighton cement (WOC).

Most of the wells with a casing leak problem across Forma-tion-A were drilled in early 2006 to increase the oil produc-tion. Basically, these wells were completed as either vertical orhorizontal open hole wells with 7” liners across Formations Aand B. All wells were completed with 7” downhole packersand 4½” tubing. Soon after the first completion, these wellsstarted producing Formation-A water, which was an indicationof communication between Formations A and B.

It is important to note that Formation-A has a higher pres-sure than that of Formation-B, which resulted from the poorcement behind the pipe and the erosion of anhydrite between


the two formations. As an undesirable consequence, Forma-tion-B injectors are feeding Formation-A along with Forma-tion-B. In addition, high injection volumes and velocities haveeroded the Formation-B anhydrite cap rock and established acommunication between the reservoirs. Formation-A pressureis higher only in the central area. Formation-B pressure at theflanks is higher than Formation-A pressure due to peripheralinjection. Formation-B pressure declines at the center becauseof oil production; however, because Formation-A does nothave any production, Formation-A pressure builds up continu-ously in the center.

FIELD PRACTICES

A survey was made of the field practices implemented in wellswhere the communication problem arose, including drilling,hole conditioning and cementing. In addition, the cement bondlog (CBL) was reviewed. Two wells were chosen for this study:Well-A and Well-C. Well-A is a horizontal well, while Well-C isvertical.

DRILLING

Well-A

This well was drilled and completed as a Formation-B horizon-tal open hole producer in mid-2007. In this well, a 8½” curvedsection (0° to 81°) was drilled from two formations above For-mation-A all the way down to a 2 ft true vertical depth (TVD)inside Formation-B with full circulation, Fig. 1. Mud weightwas 64 pounds per cubic foot (PCF) at the start until Forma-tion-A was hit, at which point the well started flowing at 40barrel per hour (BPH). The well was then shut-in until pressurestabilized. The stabilized shut-in pressure was 450 psi. Themud weight was increased to 84 PCF to kill Formation-A. Af-ter that, the rest of the hole was drilled to a 2 ft TVD belowthe top of Formation-B. The hole was swept with a Hi/LowVis pill to effectively clean the well by improving cutting liftingefficiency. In addition, a wiper trip was performed from thebottom up to the 9⅝” casing shoe to boost the hole cleaningefficiency before running the 7” liner.

Well-C

This well was drilled and completed as a Formation-B verticalopen hole producer in early 2006. In this well, an 8½” openhole was drilled from two formations above Formation-A allthe way down to a 2 ft TVD inside Formation-B with full cir-culation, Fig. 2. Mud weight was 64 PCF at the start, untilFormation-A was hit, at which point the well started flowingat 25 BPH with H2S traces. The mud weight was raised to 87PCF to kill Formation-A. After that, the rest of the hole wasdrilled to Formation-B. The hole was swept with a Hi/Low Vispill to effectively clean the well by improving cutting lifting

62884araD10R1_ASC026 3/15/13 11:42 PM Page 3


efficiency. In addition, a wiper trip was performed from thebottom up to the 9⅝” casing shoe to boost hole cleaning effi-ciency before running the 7” liner.

The wells were placed on production in mid-2006 and early2007, respectively. They both produced oil with zero water cutfor six months before being declared dead due to communica-tion between Formations A and B, which was confirmed bywater sampling and a production logging tool log as well.

HOLE PREPARATION AND CEMENTING

In both wells, a 7” liner was run consisting of a float shoe,float collar, landing collar, 7” casing joints and a mechanicalhanger along with a top packer and tie-back receptacle. Uponreaching the bottom, the casing was rotated and reciprocated,in addition to circulating the well at the highest possible rate,to remove mud cake. Then the mechanical liner hanger wasset. After that, water spacer was pumped ahead of the cementto remove any residual impurities and prevent any potential cement contamination from contact with mud.

In Well-A, the 7” liner was centralized as follows:

• Every joint from the bottom until an inclination of 44°and every second joint above that to the kickoff pointwere centralized with a spiral centralizer.

• Every other joint was centralized with a collapsiblecentralizer to the 9⅝” casing shoe, and after that everythird joint was centralized inside the casing to the 7”liner hanger using a bow rigid centralizer.

In Well-C, the 7” liner was centralized as follows:

• The first five joints and then every second joint to the9⅝” casing shoe were centralized with collapsiblecentralizers.

• Every third joint was centralized inside the casing to the

Fig. 2. The sketch for Well-C.Fig. 1. The sketch for Well-A.

Class G + 0.6% (Dispersant) + 0.3% (Fluid loss) + 0.05 gps (Retarder) + 0.005 gps (Defoamer)

Slurry Weight 101 PCF

Thickening Time 5 - 5.5 hours

Table 1. Lead cement recipe

Class G + 1.2% (Dispersant) + 0.4% (Fluid loss) + 0.22% (Retarder) + 0.01 gps (Defoamer)


Thickening Time 4 - 5 hours

Table 2. Tail cement recipe

62884araD10R1_ASC026 3/15/13 11:42 PM Page 4

7” liner hanger using a bow rigid centralizer.

After centralization, the wells were cemented in two stagesusing the cement recipes shown in Tables 1 and 2.

During cementing, no lost circulation was encountered in either well. At the end, the excess cement was reversed out,and the liner top packer was tested with water up to 2,000 psi,with no leak detected.

A CBL log was run across the entire 7” liner and showedpoor cement across Formations A and B, which confirmed thatFormation-A water was dumping into Formation-B, Figs. 3and 4.

EFFECT OF FORMATION-A WATER ON CEMENT

The cement was placed in a harsh environment where the pres-sure reaches 4,000 psi and CO2 and H2S gases exist. It is still

debatable whether Formation-A sour conditions contributed tothe poor cement behind the liner that led to the communica-tion problems or not. The effect of Formation-A water shouldnot be overlooked when rooting out the problem. But giventhe results of tests to date, as described below, this study as-sumes that there is no effect of Formation-A water on cement.

EXPERIMENTAL WORK AND EQUIPMENT

A three month study was conducted to find out the degree towhich Formation-A water contributed to the communicationproblem. In this study, cement was exposed to Formation-A


Fig. 3. CBL for Well-A. Fig. 4. CBL for Well-C.

62884araD10R1_ASC026 3/15/13 11:43 PM Page 5


water for three months under molded downhole conditions,Fig. 5. Formation-A water contains 4.5% CO2 gas and 1.28ppm H2S gas. The same cement used in the wells was used toprepare the cement samples. Some of the cement samples werecured in raw water at 215 °F before being exposed to Forma-tion-A water at 215 °F and 4,000 psi, Fig. 6. Parallel to that,other cement samples were cured in raw water at the sameconditions. Upon completion of the curing process, all cementsamples were tested for mechanical properties, namely, perme-ability, compressive strength, Poisson’s ratio and Young’s modulus. In addition, thermogravimetric analysis (TGA) andenergy dispersive X-ray fluorescence (EDXRF) tests were conducted.

A well was drilled to collect the Formation-A water samplesneeded in this project. After hitting Formation-A, the well wasflowed with a test packer isolating the zone until clean waterreached the surface. A total of 40 gallons of water was col-lected.

In total, 18 samples were prepared using the same cementrecipe used in the field. Cement samples were then poured indifferent cubical and cylindrical molds. These molds wereplaced in the curing chambers at 215 °F for two days. Afterthe curing period, the cement samples were removed and theweight was recorded. Each test specimen was assigned a num-ber. Four samples were tested for mechanical properties, in-cluding permeability, and subjected to TGA and EDXRF testsafter the initial curing. The remaining samples were dividedinto two groups. The first set was cured under sour conditionsin Hastelloy metal autoclaves for three months, while the sec-ond set was cured in raw water in autoclaves for the same pe-riod of time. Samples cured for six months in sour conditionsin Formation-A water are shown in Fig. 7. At the end, the ce-ment samples were taken out of the autoclaves and tested formechanical properties, including permeability, and subjected toTGA and EDXRF tests.

Permeability Test

The permeability test is conducted using permeability equip-ment. It consists of a core holder in which the cement sample isplaced, a fluid cylinder for fluid injection, a beaker to collectfluid, if any, a pump for injection purposes and a computer tocollect data. The sample is placed in the core holder after beingcleaned and trimmed. Then brine is injected into the cementsample at 700 psi differential pressure and an injection rate of2 cc/min. At the end, the amount of water collected is meas-ured, Tables 3 and 4.

Young’s Modulus and Poisson’s Ratio Test

In the test conducted to calculate Poisson’s ratio, Young’s mod-ulus and peak strength, axial stress is applied to a test specimenuntil the cement starts to break or fracture. The cement samplesare cut into 3” length x 1.5” outer diameter size using the trimming machine. Then the sample surfaces are finished orground using a surface grinding machine. The degree of

Fig. 5. Some cement samples after being exposed to Formation-A water for threemonths.

Fig. 6. Some cement samples before being exposed to Formation-A water.

Fig. 7. Some cement samples after being exposed to Formation-A water for sixmonths.

Initial Curing 3 Months Raw Water Curing 3 Months Formation-A Water Curing

Sample-13 Sample-14 Sample-19 Sample-110 Sample-115 Sample-116

Permeability CC/min 0 0 0 0 0 0

Table 3. Permeability tests results after short-term exposure to Formation-A water

Class G + 1.2% (Dispersant) + 0.4% (Fluid loss) + 0.22% (Retarder) + 0.01 gps (Defoamer)


Thickening Time 4 - 5 hours

Table 4. Permeability tests results for long-term test

62884araD10R1_ASC026 3/15/13 11:43 PM Page 6

parallelism of the surfaces of the sample is then measured. Toensure the load is applied evenly over the surface, the acceptedtolerance should be equal to or less than 2/1,000”. The sample isthen placed inside the Tri-Axial equipment, which consists of acore holder, a piston, a vessel, a control panel, a camera and acomputer for data acquisition. At first, a plastic jacket is used toprotect the plug while applying the confining pressure to avoidfluid entry into the plug. After that, the core is placed into thecore holder. Three voltage linear differential transducer wires areconnected to the core holder. Two wires are used to measurethe axial distance change while the third one is for changemeasurement in radial distance. Next, confining pressure is ap-plied at 700 psi, and axial load ranging from 5 to 15 MPA isapplied to the piston at a temperature of 150 °C. Then Young’smodulus, Poisson’s ratio and peak strength are calculated,Table 5.

EDXRF Test

In this test, samples are tested to determine the elemental com-positions that make up the cement system. The cement sampleis crushed and milled until it becomes a powder. Then thepowder is mixed with 0.5 grams of a chemical binder. Themixture is poured into a pellet mold before being pressed at 15psi by the X-Press machine. The pellet is then placed inside aspectrometer that consists of a 400 watt X-ray tube, a computercontrolled high voltage generator for the X-ray tube, liquid N2,a cooled Si(Li) detector, a multichannel analyzer and a com-puter for data acquisition. The EDXRF analyzes the samplefor elemental composition after entering the weights of thesample and binder, Table 6.

TGA Test

This test is conducted to measure the thermal stability andcomposition of the cement as a function of time. The effect isquantified by the weight loss that elements suffer due to heat.First, the cement sample is crushed and milled until it becomesa powder. Then a pellet is produced with 50 mg of cement andplaced into the TGA test apparatus. Then the temperature israised at a rate of 2 °C/min from room temperature until

1,000 °C is reached. Data, including the amount of weight lossand remaining mass percentage, are calculated by the com-puter, Table 7.

RESULTS AND DISCUSSION

Short-term Test

All samples were examined physically upon their removal fromthe CO2 autoclave. All samples were inspected and were found


6 Months Raw Water Curing 6 Months Formation-A Water Curing

Sample-117 Sample-118 Sample-125 Sample-126

Compressive Strength, psi 10,944.9 11,548.6 12,656.7 13,093.8

Dynamic E, psi 1.809E+06 1.816E+06 2.005E+06 1.859E+06

Static E 3.084E+06 2.998E+06 2.904E+06 3.142E+06

Dynamic Y, psi 0.260 0.276 0.241 0.268

Static Y, psi 0.242 0.288 0.189 0.172

Table 5. Mechanical properties for long-term test

CompoundsApproximate Weight Percentages

111 112 129 130

CaO 59.85 60.06 56.96 57.90

SiO2 19.65 19.71 17.58 17.93

Fe2O3 4.80 4.71 4.24 4.32

Al2O3 2.40 2.30 2.20 2.45

SO3 1.88 1.89 2.39 2.38

MgO 1.70 1.68 1.66 1.88

K2O 0.08 0.12 0.15 0.09

TiO2 0.22 0.20 0.20 0.21

Mn2O3 0.04 0.04 0.04 0.04

SrO <0.05 <0.05 0.07 0.06

Table 6. Chemical composition for cement after long-term exposure Formation-Awater and raw water

Initial Curing Short-termWater Curing

Short-term CO2 Curing

Sample# 116 115 111 112 117 118

Mass loss % 13.06 12.98 12.66 13.15 16.42 16.09

Residual Mass % (150-1,000 ºC)

74.21 73.57 76.57 74.12 77.59 77.57

LOL % (20-150 ºC) 25.8 26.4 23.43 25.88 22.41 22.43

Table 7. TGA results after initial setting, water curing and Formation-A water curing

62884araD10R1_ASC026 3/15/13 11:43 PM Page 7


to be intact. All samples were found to have turned to a blackcolor due to their reaction with the H2S gas. Mechanical prop-erties, including permeability, Young’s modulus and Poisson’sratio, were all calculated before and after Formation-A waterexposure, Table 8. According to the permeability test, the ce-ment stayed solid for 15 minutes during brine injection at apressure of 700 psi, indicating that it is impermeable. Also, re-sults showed a slight change in the rest of the mechanical prop-erties. For example, Static E increased from 2.322E+06 to2.400E+06 psi, while Dynamic E increased from 2.930E+06 to3.001E+06 psi. Tests also showed that Static Y increased afterexposure from 0.125 to 0.29, and that Dynamic Y increasedfrom 0.282 to 0.290. All results pertaining to the tests of me-chanical properties, including permeability, for all samples arein Tables 3 and 4.

The TGA analysis showed that the cement lost approxi-mately 13% of mass due to moisture evaporation between 20°C to 150 °C. The cement sample suffered a further weightloss of 13% as the temperature rose to 1,000 °C due to the de-cay of some elements. The sample mass decreased by 26% intotal during the test. EDXRF results showed that the cementsamples after initial curing mainly consisted of CaO (60%)and SiO2 (19%) by weight, Fig. 8. After curing in Formation-Aconditions, less than 1% change in mass occurred.

These findings showed that Formation-A water did not sig-nificantly harm the cement integrity even in the presence ofhigh pressure for the three month test period. This is mostlikely due to the small amount of CO2 gas present in the curing

water. The picture will be clearer after the end of the six monthtest period.

Long-term Test

According to the permeability test, the cement stayed solid for15 minutes during brine injection at a pressure of 700 psi, indi-cating that it is impermeable. Also, results showed a slightchange in the rest of the mechanical properties. For example,dynamic Young’s Modulus (E) increased from 1.089 E+06 to2.005 E+06 psi while Static E increased from 2.998E+06 to3.142E+06 psi. In regard with Poisson’s ratio (Y), tests showedthat Static Y decreased from 0.288 to 0.189 and Dynamic Ydecreased from 0.276 to 0.241, Table 5.

The TGA analysis showed that the cement lost approxi-mately 4.61% of mass due to moisture evaporation between 0 °C to 150 °C. The cement sample suffered further weightloss of 16.77% as the temperature rose to 1,000 °C due to thedecay of some elements, Table 9. The sample mass decreasedby 21.38% in total during the test. EDXRF results showedthat the cement samples after six months of curing mainly con-sisted of CaO (60% to 57%) and SiO2 (19.5% to 17.5%) byweight, respectively, Tables 10 and 11. In addition, the weightof these two elements decreased by 2% to 3% due to an en-countered error while taking the WOC. No major change inmass has been observed. Moreover, the cement color changedfrom gray to black owing to the reaction with H2S gas.

These findings showed that Formation-A water did not harmcement integrity even in the presence of high pressure. This isdue to the small amount of CO2 gas present in the curing water.

Initial Curing 3 Months Water Curing 3 Months Formation-A Water Curing

Sample-11 Sample-12 Sample-17 Sample-18 Sample-113 Sample-114

Compressive Strength, psi 8,609.1 10,024.8 11,587.7 12,030.6 11,279.9 12,270.0

Dynamic E, psi 2.949E+06 2.930E+06 2.994E+06 2.933E+06 3.025E+06 3.001E+06

Static E 7.153E+05 2.322E+06 2.177E+06 2.120E+06 1.958E+06 2.400E+06

Dynamic Y, psi 0.282 0.281 0.275 0.276 0.174 0.219

Static Y, psi 0.125 0.125 0.298 0.275 0.290 0.258

Table 8. Mechanical properties after short-term exposure to Formation-A water

Fig. 8. TGA chart after initial curing.

Long-term Raw Water Curing

Long-term CO2 Curing

Sample# 111 112 1,129 130

Mass Loss % 16.32 16.11 15.85 16.77

Residual Mass % (150-1,000 °C) 79.09 79.49 79.59 78.62

LOL % (20-150 °C) 20.91 20.51 20.41 21.38

Table 9. TGA results after long-term water curing and Formation-A water curing

62884araD10R1_ASC026 3/15/13 11:43 PM Page 8


After an extensive review of the field practices, it is clearthat the dominant factor contributing to communication be-tween Formations A and B is the loss of hydrostatic pressure ofthe cement column in addition to high Formation-A pressure.No deficiencies were found in field cementing practices, includ-ing mixing and pumping the cement, conditioning the holeprior to the cement job, mud cake removal, and mud displace-ment and casing centralization.

A batch mixer was used in all cement jobs as it gives an ac-curate density reading of the cement slurry. The number ofcentralizers used in the horizontal wells was selected to obtain70% standoff across critical open hole sections. According tofield findings, this degree of concentricity is fair enough for

good zonal isolation21. This supports the conclusion that cen-tralization was not poor since the problem also occurred invertical wells where the standoff is as high as 95%. Liner rota-tion and reciprocation within 60 ft stroke, in addition to circu-lation at a rate of 4 bpm, helped clean filter cake and provideuniform cement distribution around the casing. Conditioningmud to reduce its viscosity improves mud displacement effi-ciency by enhancing fluid mobility. In addition, liner rotationand reciprocation increases the mud’s ability to erode and remove bypassed mud by reducing casing-to-mud and well-bore-to-mud drag forces. The presence of a spiral centralizerimproved the flow regime of cement across the horizontal sec-tions. A compatible viscous spacer was used to separate the cement and drilling fluid. The spacer helps avoid prematuresetting of cement, cement channeling and cement contamina-tion. The volume of the spacer was calculated to give a contacttime of 10 minutes, which is consistent with widely used ce-menting practices. The spacer density was higher than mudand lighter than cement. This best practice in cementing helpseffectively displace mud and avoid mud bypassing cement.

The results of the survey of field practices were surprisingsince they showed that all practices were perfect. Therefore, itwas advisable to go back to the literature and examine theproblem more deeply by focusing on the effect of a loss of hy-draulic pressure while waiting on the cement to set and by ig-noring the other factors after it was confirmed that they werenot linked to the problem completely. During the second lookat the literature, an interesting experiment conducted in thefield by Cooke7 to study the behavior of cement hydraulicpressure during the first six hours after cement placement wasfound. The results of this experiment showed that cement pres-sure decreases at 39 psi/ft during the first six hours afterpumping the cement. These results are supported by the experi-ment Levine2 conducted three years earlier that showed thatcement is able to transfer pressure during gelation time untilthe cement gets set, after which the cement is not able to trans-mit pressure.

Such a finding was utilized along with field data to plotpressure vs. depth charts to study the behavior of cement hy-drostatic pressure while pumping cement and six hours later.The red line up to the intersection point indicates the pressureof the mud column, while the rest of it shows the pressure ofthe cement and mud columns six hours following cementplacement. In contrast, the blue line shows the pressure of thecement and mud columns right after cement placement. As illustrated in Fig. 9, the hydrostatic pressure at the top of theFormation-A pressure was 4,570 psi before it decreased to 700psi below the Formation-A pressure, creating an underbal-anced situation during which Formation-A water displaced ce-ment into permeable zones above and below, leaving the lineruncemented and allowing communication to take place whilewaiting on the cement to set. As a result, communication wasestablished between these two zones.

Formation-A in this area has high reservoir pressure.

Elements

Approximate Weight Percentages

8948 1-5 (Initial Curing)

8948 1-6 (Initial Curing)

CaO 60.12 60.30

SiO2 19.85 19.76

Fe2O3 4.52 4.54

Al2O3 2.68 2.74

SO3 1.89 1.88

MgO 1.87 1.93

K2O 0.43 0.46

TiO2 0.20 0.20

Mn2O3 0.04 0.05

SrO 0.04 0.04

Table 10. Chemical composition for cement after initial setting

CompoundsApproximate Weight Percentages

Short-term Forma-tion-A Water

Short-term Water Curing

CaO 58.89 59.32 60.92 60.88

SiO2 18.08 18.32 19.15 19.10

Fe2O3 4.37 4.38 4.59 4.57

Al2O3 2.5 2.54 2.52 2.51

SO3 2.53 2.49 1.94 1.94

MgO 1.9 1.86 1.76 1.79

K2O 0.19 0.17 0.06 0.05

TiO2 0.19 0.20 0.21 0.21

Mn2O3 0.04 0.04 0.04 0.04

SrO 0.06 0.06 0.04 0.04

Table 11. Chemical composition for cement after short-term Formation-A watercuring and raw water curing

62884araD10R1_ASC026 3/15/13 11:43 PM Page 9

creo


Therefore, it was easy for the cement column to be underbal-anced against Formation-A before it was able to develop therequired static gel strength of 500 lbf/100 ft2. When the under-balance occurred, the inflow of water from Formation-A con-taminated the cement column in the annulus. Actual reductionin hydrostatic pressure experienced by a cement column is dependent on the development of its gel strength and reductionin the slurry volume. To illustrate the occurrence of water flowfrom Formation-A during the primary cementing job in WellsA and C, the pressure loss profile calculated from Cooke’s7

data was used. As shown in Figs. 9 and 10, the loss in the hydrostatic pressure likely caused the cement column to be underbalanced against Formation-A.

Figures 9 and 10 also demonstrate that in the first six hoursfollowing the cement placement, the hydrostatic pressure ofthe cement column dropped by 700 psi, creating an underbal-anced situation and allowing for communication between for-mations. Without doubt, the main factor that caused poorprimary cementing across Formation-A behind the 7” liner isloss of hydrostatic pressure in the cement column after it wasspotted in place in the annulus. In addition, setting a 7” linertop packer had isolated the hydrostatic pressure from acting

downward onto the annulus and formations below. This fur-ther encouraged the flow of influx from Formation-A into theannulus. Use of 3,000 ft liner lap also contributed to the lossof hydrostatic pressure, since the amount of loss in pressure ishigher there compared with a short cement column.

CONCLUSIONS

1. The root cause of the communication problem was found clearly to be the loss of hydrostatic pressure before the cement attained enough compressive strength.

2. Cementing practices, including setting the liner top packer and use of long liner laps, further encouraged water influx to attack and contaminate the cement.

3. CBL logs showed poor cement and water channeling, con-firming the occurrence of communication.

4. Lab results showed that Formation-A water is not detrimentalto cement during this period of time.

5. Solutions, including use of a short cement column, elimina-tion of the liner to packer, application of annular pressure and use of a zonal isolation packer between Formations A and B, will help avoid cement contamination due to water influx during WOC.

6. The CBL should be run immediately after the cement job sothat corrective measures can be taken in a timely manner.

7. Field practices showed no deficiencies except those previouslyhighlighted.

REFERENCES

1. Cheung, P.R. and Beirute, R.M.: “Gas Flow in Cements,”Journal of Petroleum Technology, Vol. 37, No. 6, June1985, pp. 1,041-1,048.

2. Levine, D.C., Thomas, E.W., Bezner, H.P. and Tolle, G.C.:“Annular Gas Flow after Cementing: A Look at PracticalSolutions,” SPE paper 8255, presented at the SPE AnnualTechnical Conference and Exhibition, Las Vegas, Nevada,September 23-26, 1979.

3. Bonett, A. and Pafitis, D.: “Getting to the Root of GasMigration,” Oil Field Review, Vol. 8, No. 1, March 1,1996, pp. 36-49.

4. Hartog, J.J., Davies, D.R. and Stewart, R.B.: “AnIntegrated Approach for Successful PrimaryCementations,” SPE paper 9599, presented at the SPEMiddle East Technical Conference, Manama, Bahrain,March 9-12, 1981.

5. Soran, T.U., Chukwu, G.A. and Hatzignatiou, D.G.: “GasChanneling and Micro-Fractures in Cemented Annulus,”SPE paper 26068, presented at the SPE Western RegionalMeeting, Anchorage, Alaska, May 26-28, 1993.

6. Robert, B. and Art, T.: “Expansive and ShrinkageCharacteristics of Cement under Actual Well Conditions,”

Fig. 9. Behavior of cement column pressure after 6 hours from cement placement(Well-A).

Fig. 10. Behavior of cement column pressure after 6 hours from cement placement(Well-C).

62884araD10R1_ASC026 3/15/13 11:43 PM Page 10


Symposium, San Francisco, California, June 29 - July 2, 2008.

17. Hibbeler, J.C., DiLullo, G. and Thay, M.: “Cost-Effective Gas Control – A Case Study of Surfactant Cement,” SPE paper 25323, presented at the SPE Asia Pacific Oil and Gas Conference, Singapore, February 8-10, 1993.

18. Santara, A., Reddy, B.R., Liang, F. and Fitzgerald, R.: “Reaction of CO2 with Portland Cement at Downhole Conditions and the Role of Pozzolanic Supplements,” SPEpaper 121103, presented at the SPE International Symposium on Oil Field Chemistry, The Woodlands, Texas, April 20-22, 2009.

19. Santara, A., Reddy, B.R. and Antia, M.: “Designing Cement Slurries for Preventing Formation Fluid Influx after Placement,” SPE paper 106006, presented at the SPE International Symposium on Oil Field Chemistry, Houston, Texas, February 28 - March 2, 2007.

20. Moroni, N., Santara, A., Ravi, K. and Hunter, W.: “Holistic Design of Cement Systems to Survive CO2

Environment,” SPE paper 124733, presented at the SPE Annual Technical Conference and Exhibition, New Orleans, Louisiana, October 4-7, 2009.

21. Osazuwa P., Mathias, A. and Herve, F.: “New Centralizers Improve Horizontal Well Cementing by 100% Over Conventional Centralizers in the Niger Delta Basin,” SPE paper 67197, presented at the SPE Production and Operations Symposium, Oklahoma City, Oklahoma, March 24-27, 2001.

Journal of Petroleum Technology, Vol. 25, No. 8, August 1973, pp. 905-909.

7. Cooke Jr., C.E., Kluck, M.P. and Medrano, R.: “FieldMeasurements of Annular Pressure and Temperatureduring Primary Cementing,” Journal of PetroleumTechnology, Vol. 35, No. 8, August 1983, pp. 1,429-1,438.

8. Sabins, F.L., Tinsley, J.M. and Sutton, D.L.: “TransitionTime of Cement Slurries between the Fluid and Set State,”SPE paper 9285, presented at the SPE Annual TechnicalConference and Exhibition, Dallas, Texas, September 21-24, 1980.

9. Myrick, B.D.: “Field Evaluation of an ImpermeableCement System for Controlling Gas Migration,” SPE paper11983, presented at the SPE Annual Technical Conferenceand Exhibition, San Francisco, California, October 5-8,1983.

10. Jones, R.R. and Carpenter, R.B.: “New Latex, ExpandingThixotropic Cement Systems Improve Job Performance and Reduce Costs,” SPE paper 21010, presented at the SPE International Symposium on Oil Field Chemistry, Anaheim, California, February 20-22, 1991.

11. Dean, G.D. and Brennen, M.A.: “A Unique Laboratory Gas Flow Model Reveals Insight to Predict Gas Migrationin Cement,” SPE paper 24049, presented at the SPE Western Regional Meeting, Bakersfield, California, March30 - April 1, 1992.

12. Ramirez, H.B., Santara, A., Martinez, C. and Ramos, X.: “Water-Gas Migration Control and Mechanical Properties Comparison with a Quick-Setting Slurry Design (QSSD) to be Applied in a Production Cementing Job for Ecuador,” SPE paper 123085, presented at the SPE Latin American and Caribbean Petroleum Engineering Conference, Cartagena, Colombia, May 31 - June 3, 2009.

13. Farias, A.C., Suzart, W.P., Ribeiro, D., Santos, P.R. and Santos, R.: “High Static Gel Strength Cement Slurries to Hold Gas Migration-Laboratory Surveys,” paper presented at the 7th Well Engineering Seminar in El Salvador, October 16-18, 2007.

14. Calloni, G., Antona, P.D. and Moroni, N.: “A New Rheological Approach Helps Formulation of Gas Impermeable Cement Slurries,” Cement and Concrete Research, Vol. 29, No. 4, April 1999, pp. 523-526.

15. Moran, L. and Savery, M.: “Fluid Measurements through Eccentric Annuli: Unique Results Uncovered,” SPE paper 109563, presented at the SPE Annual Technical Conference and Exhibition, Anaheim, California, November 11-14, 2007.

16. Inverson B., Darbe, R. and McMechan, D.: “Evaluation of Mechanical Properties of Cement,” ARMA paper 08-293, presented at the 42nd U.S. Rock Mechanics

62884araD10R1_ASC026 3/15/13 11:43 PM Page 11


BIOGRAPHIES

Abdulla F. Al-Dossary joined SaudiAramco in December 2005. He beganhis career as a Workover Engineerworking with the WorkoverDepartment. In April 2012, Abdullawent to work with the Northern AreaOil Drilling Department as a Drilling

Engineer.He received his B.S. degree in Mechanical Engineering

from King Fahd University of Petroleum and Minerals(KFUPM), Dhahran, Saudi Arabia, in 2005. In 2011,Abdulla received his M.S. degree in Petroleum Engineering,also from KFUPM.

He has published and presented four Society ofPetroleum Engineers (SPE) papers.

Scott S. Jennings is the Group Leaderfor cementing at Saudi Aramco’sExploration and PetroleumEngineering Center – AdvancedResearch Center (EXPEC ARC). Hehas 32 years of experience incementing. Prior to joining Saudi

Aramco in 1987, Scott assumed duties that includedstimulation, cementing and sand control with HalliburtonCo. in East Texas and the Middle East Region. His areas ofinterest are developing standards and test equipment, wellconstruction, gas migration prevention and long-termcement durability. Scott is the Saudi Aramco votingmember of the American Petroleum Institute Subcommittee10 and a long-term member of the Society of PetroleumEngineers (SPE).

In 1980, he received a B.S. degree in Chemistry fromAngelo State University, San Angelo, TX.

62884araD10R1_ASC026 3/15/13 11:43 PM Page 12


62884araD11R1_ASC026 3/15/13 11:43 PM Page 1


SUBSCRIPTION ORDER FORM

To begin receiving the Saudi Aramco Journal of Technology at no charge, please complete this form.

Please print clearly.

Name ______________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

Title __________________________________________________________________________________________________________________________________________

Organization _______________________________________________________________________________________________________________________________

Address _____________________________________________________________________________________________________________________________________

City __________________________________________________________________________________________________________________________________________

State/Province ______________________________________________________________________________________________________________________________

Postal code _________________________________________________________________________________________________________________________________

Country _____________________________________________________________________________________________________________________________________

E-mail address _____________________________________________________________________________________________________________________________

Number of copies _________________________________________________________________________________________________________________________

TO ORDER

By phone/e-mail:

Saudi Aramco Public Relations Department JOT Distribution+966/[email protected]

By fax:

Saudi Aramco Public Relations Department JOT Distribution+966/3-873-6478

By mail:

Saudi Aramco Public Relations Department JOT Distribution Box 5000Dhahran 31311Saudi Arabia

Current issues, select back issues and multiple copies of some issues are available upon request.

The Saudi Aramco Journal of Technology is published by the Saudi Aramco Public Relations Department, Saudi Arabian Oil Company, Dhahran, Saudi Arabia.

62884araD11R1_ASC026 3/15/13 11:43 PM Page 2


GUIDELINES FOR SUBMITTING AN ARTICLE TO THE SAUDI ARAMCO JOURNAL OF TECHNOLOGY

These guidelines are designed to simplify and help standardizesubmissions. They need not be followed rigorously. If youhave additional questions, please feel free to contact us atPublic Relations. Our address, fax and phone numbers arelisted on page 81.

Length

Varies, but an average of 2,500-3,500 words, plusillustrations/photos and captions. Maximum length should be 5,000 words. Articles in excess will be shortened.

What to send

Send text in Microsoft Word 6.0/95 or higher (do not submit UNIX files) via e-mail or on disc, plus one hard copy.Send illustrations/photos and captions separately but concurrently, both as e-mail or as hard copy (moreinformation follows under Format).

Procedure

Notification of acceptance is usually within three weeks. Thearticle will be edited for style and clarity and returned to theauthor for review. All articles are subject to the company’snormal review. No paper can be published without a signatureat the manager level or above.

Format

No single article need include all of the following parts. Thetype of article and subject covered will determine which partsto include.

Working title

Abstract

Usually 100-150 words to summarize the main points.

Introduction

Different from the abstract in that it “sets the stage” for thecontent of the article, rather than telling the reader what it is about.

Main body

May incorporate subtitles, artwork, photos, etc.

Conclusion/summary

Assessment of results or restatement of points in introduction.

Endnotes/references/bibliography

Use only when essential. Use author/date citation method inthe main body. Numbered footnotes or endnotes will beconverted. Include complete publication information.Standard is The Associated Press Stylebook, 46th ed.

Acknowledgments

Use to thank those who helped make the article possible.

Illustrations/tables/photos and explanatory text

Submit these separately. Do not place in the text. Positioningin the text may be indicated with placeholders. Initialsubmission may include copies of originals; however,publication will require the originals. When possible, submitboth electronic versions, printouts and/or slides. Color ispreferable.

File formats

Illustration files with .EPS extensions work best. Otheracceptable extensions are .TIFF, .JPEG and .PICT.

Permission(s) to reprint, if appropriate

Previously published articles are acceptable but can bepublished only with written permission from the copyrightholder.

Author(s)/contributor(s)

Please include a brief biographical statement.

Submission/Acceptance Procedures

Papers are submitted on a competitive basis and are evaluatedby an editorial review board comprised of various departmentmanagers and subject matter experts. Following initialselection, authors whose papers have been accepted forpublication will be notified by e-mail.

Papers submitted for a particular issue but not accepted forthat issue will be carried forward as submissions forsubsequent issues, unless the author specifically requests inwriting that there be no further consideration. Paperspreviously published or presented may be submitted.

Submit articles to:

EditorThe Saudi Aramco Journal of TechnologyRoom 2240 East Administration BuildingDhahran 31311, Saudi ArabiaTel: +966/3-873-5803Fax: +966/3-873-6478E-mail: [email protected]

Submission deadlines

Issue Abstract submission deadline Release date

Fall 2013 June 1, 2013 September 30, 2013Winter 2013 September 1, 2013 December 31, 2013Spring 2014 December 1, 2013 March 31, 2014Summer 2014 March 1, 2014 June 30, 2014

62884araD11R1_ASC026 3/15/13 11:43 PM Page 3

H2S Early Notification System for Production Pipelines: A Pilot TestGeorge J. Hirezi, Faisal T. Al-Khelaiwi and Mohammed N. Al-Khamis

ABSTRACT

The produced fluid of an oil field located in the Eastern Province of Saudi Arabia contains relatively high levels of H2S. A pilottest was conducted by Saudi Aramco to install a wireless gas detection system along an oil pipeline in this field. The pilot testobjectives included:

• Determining the communication availability and reliability of the remote wireless sensors in areas where extending hardwired and fiber optic networks proved impractical and expensive.

• Evaluating the usefulness of this system for early notification of toxic gas releases or pipe leaks in and around critical geographical areas by alerting the console operator via email and Short Message Service (SMS).

Intelligent Field Infrastructure Adoption: Approach and Best PracticesSoloman Almadi and Tofig Al-Dhubaib

ABSTRACT

The drive to implement the latest optimal intelligent field infrastructure (IFI) is a continuous goal for oil and gas operators. Thisrequires the right balance between technology, business drivers and evolving implementation requirements. A successfulintelligent field implementation relies on a robust real-time field to desktop data acquisition and delivery system designed withclearly defined data acquisition requirements. The data acquisition requirements definition should include data type, acquisitionfrequency, resolution, integrity, quality and reliability.

Real-Time Estimation of Well Drainage Parameters Mohammad S. Al-Kadem, Faisal T. Al-Khelaiwi and Meshal A. Al-Amri

ABSTRACT

The well drainage pressure and radius are key parameters of real-time well and reservoir performance optimization, well testdesign and location identification for new wells. Currently, the primary method of estimating the well drainage radius is builduptests and a subsequent well test analysis. Such buildup tests are conducted using wireline run quartz gauges for an extended wellshut-in period, resulting in deferred production and risky operations.

Solar Power Integration Challenges: Intermittency and Voltage Regulation Issues Mahmoud B. Zayan

ABSTRACT

Grid-connected solar energy generation is expected to multiply over the coming decades. Solar power generation brings manybenefits, such as reduced greenhouse gases and pollutant emissions, diversity of fuel supplies and displacement of costly fossilfuel generation. Consequently, achieving higher penetration levels of solar energy in the market depends primarily on theviability and reliability of the integrated system. A considerable barrier to the sustainability of solar power generation is theconstrained ability to control voltage as a result of weather related intermittency and the heavy reliance on inverters and otherpower electronic devices to interface with the grid. To overcome those barriers, distribution networks will have to be designeddifferently, and innovative smart grid technologies will have to be developed so as to optimize contributions from solar resourceswhile preserving the integrity of the grid.

Additional Content Available Online at: www.saudiaramco.com/jot

62884araD1R2_ASC027 3/17/13 3:46 PM Page 3

creo

62884araD1R2_ASC027 3/17/13 3:46 PM Page 4

Documents

Successful Implementation of Horizontal Multistage Fracturing to