18 Oilfield Review

Optimizing Frac Packs

Bala Gadiyar New Orleans, Louisiana, USA

Craig MeeseGreg Stimatz Marathon Oil Company Houston, Texas, USA

Hugo Morales Houston, Texas

Jose Piedras Total E&P USA, Inc. Houston, Texas

Jérôme Profinet Total, Elf Petroleum Nigeria, Ltd.Port Harcourt, Nigeria

Graham Watson Pau, France

For help in preparation of this article, thanks to Billy Greeson,Houston, Texas, USA. AFIV (annular-controlled FIV system), CoolFRAC, DataFRAC,DeepSTIM, FIV (Formation Isolation Valve), QUANTUM,QUANTUM maX and SenTREE are marks of Schlumberger. AllFRAC, AllPAC and Alternate Path are marks of Mobil OilCorporation, now ExxonMobil, that are licensed to Schlumberger.

Fracturing for sand control has evolved as applications expand to deeper,

more demanding reservoirs. A reliable test to establish fracture-closure

pressure along with improved fluid-selection criteria has helped engineers

reduce completion damage in ultradeepwater wells. These field-proven

techniques can also be applied in other areas to ensure successful tip-

screenout treatments and placement of highly conductive proppant packs.

1. Ali S, Norman D, Wagner D, Ayoub J, Desroches J,Morales H, Price P, Shepard D, Toffanin E, Troncoso J and White S: “Combined Stimulation and Sand Control,”Oilfield Review 14, no. 2 (Summer 2002): 30–47.

2. Acock A, Heitmann N, Hoover S, Malik BZ, Pitoni E, RiddlesC and Solares JR: “Screenless Methods to Control Sand,”Oilfield Review 15, no. 1 (Spring 2003): 38–53.

> Frac packing. Tip-screenout (TSO) designs use fluids that leak off early in a treatment, causingproppant to pack off at the fracture tips (top). Pumping additional proppant-laden fluid, or slurry, causesthe biwing fractures to inflate as proppant packs back toward the well (middle). A TSO generatesenough displacement in soft formations to create an annular opening around the wellbore that fills with proppant. This external pack prevents sand production from nonaligned perforations and furtherreduces near-well pressure drop (bottom).

Dynamic fracture

Tip screenout

Proppant

Casing

Cement

Perforation

Fracture inflation

Annular opening

Screen

Washpipe

Propped fracture“External” proppant pack

Annular proppant pack

Autumn 2004 19

As offshore oil and gas development moves intoharsher, more demanding deepwater environ-ments, frac-packing utilization and methodscontinue to expand and evolve based on specificfield experience and requirements. These tip-screenout (TSO) fracturing treatments performed in conjunction with gravel packing ofmechanical screens now represent almost 65%of sand-control completions in the Gulf of Mexico, USA. Since it was first applied in theearly 1990s, frac packing has become one of themost widely used methods for completing wellsin weakly consolidated formations.

This combined stimulation and sand-controltechnique has proved effective in a wide rangeof formations with mobile solids, especially high-permeability reservoirs.1 Frac packs consistentlyyield sustained production increases comparedwith viscous slurry packs or high-rate water packs.

Frac packing avoids many of the productivityimpairments common in conventional cased-holegravel packs by effectively bypassing formationdamage, or skin, and by creating an “external”pack to stabilize perforations that are not alignedwith the propped fracture (previous page).

A TSO design limits hydraulic fracture extension, or length, by using less-efficient stim-ulation fluids with high leakoff rates that causethe proppant-laden slurry stages to dehydrateearly in a treatment. Proppants pack off nearthe end, or tip, of dynamic fractures, causingthem to inflate like a balloon as additional slurryis injected. Proppant then packs back towardthe well, which promotes grain-to-grain contactand generates a wider, more conductive pathwayafter the dynamic fracture closes.

In many ways, frac packing is a mature technology. Service companies have comparablepumping equipment, stimulation vessels, downhole tools and laboratory support. Theyalso provide similar fluids, proppants, treatmentadditives and damage-removal techniques.Other well technologies—intelligent well com-pletions, multiple-zone production monitoringand control, real-time data and informationtransfer, safety and quality control—have alsoreached a relatively high level of maturity.

The average completion skin for frac packs is typically less than the skin for other sand-control methods, but there is room forimprovement. Frac-pack productivity may belower than expected because of a combination offactors, including perforation damage, failure toachieve TSO, incomplete fracture or proppant-pack coverage and high pressure drops throughdownhole sand-exclusion screens and well-completion equipment.

Frac-pack optimization involves addressingall of these completion-design factors to reduceoverall skin and improve well productivity, tomaximize hydrocarbon recovery, and to helpoperators avoid future well interventions. Thelatter objective is critically important in deep-water fields, particularly those with subseawells, where remedial operations to removedamage or restimulate wells are extremely difficult, complex and costly.

Total, Marathon and Schlumberger refinedexisting completion practices and frac-packingtechniques in the Gulf of Mexico using fieldexperience and improved computer modeling of fracturing and frac-packing processes. Completion engineers now select optimal treat-ment fluids and adjust frac-pack designs toaccount for in-situ temperatures and fluid shearduring treatment execution.

This article discusses methods for cleaningperforations and for selecting treatment fluidsthat achieve effective TSO fractures, including areliable means of determining fracture-closurepressure. It presents well-completion equipmentthat ensures complete stimulation and gravelpacking across long intervals, maximizes internal flow area, and allows evaluation of frac-packing efficiency. It also summarizes field

experience and results from the ultradeepwaterAconcagua and Camden Hills fields in the Gulf of Mexico Canyon Express project, which contributed to a better understanding offrac packing.

Effective Perforations Reservoir stimulation, or fracture conductivity,alone does not ensure an optimal frac pack. Aneffective external proppant pack is alsorequired. A ring of proppant around the wellborestabilizes all the perforations and hydraulicallyconnects them with the propped fracture. Thisfurther minimizes frac-pack skin and reducespressure drop across the completion interval tohelp avoid formation failures and subsequentsand production. An external pack is also thebasis for screenless completions that controlsand without mechanical screens and internalgravel packs.2

Computer modeling indicates that unalignedperforations, those oriented away from the preferred fracture plane (PFP), contribute as much as 50% of the inflow from high-permeability formations (above). This under-scores the importance of eliminating flowrestrictions in and around all of the perforations.

> Fracture flow versus inflow from unaligned perforations. Inflow is not limitedto propped fracture cross-sectional area. Computer modeling indicates thatperforations aligned away from the preferred fracture plane (PFP) contributealmost 50% of the inflow from high-permeability formations, underscoring theimportance of TSO fracturing and creation of an external proppant pack. Thissimulation compares a low perforation shot density of 4 shots per foot (spf) incased hole (red) with an openhole completion (green), that has infiniteperforation density.

Cased-hole perforatedcompletion, 4 spf, 90° phasing

Openhole completion

100

80

60

40

20

010 100 1,000 10,000

Ratio

of f

ract

ure

flow

to to

tal f

low

, %

Formation permeability, mD

Propped fracture

Unaligned perforations

Explosive-jet perforating causes a crushedzone of damage around perforation tunnels. Thisdamage can be addressed by pumping acid toremove perforating damage and debris prior tofrac packing or by applying more effective perfo-rating practices, such as dynamic underbalancetechniques.3 Analysis of Gulf of Mexico well com-pletions indicates that skin factors were highwhen acid volumes of less than 20 gal/ft[0.24 m3/m] were used across perforated intervals; pumping acid volumes of 40 to 50 gal/ft[0.5 to 0.6 m3/m] with effective diversion across the entire zone minimized completion skin(above left).

Careful consideration should also be given toselecting perforated intervals to avoid unwantedhydraulic fracture-height growth in shale layersabove and below productive intervals. Fractur-ing into shale restricts fluid leakoff. Dynamicfractures in shale remain open longer becausetreating fluids do not leak off fast enough. Thisalso makes it difficult to obtain a completegravel pack around the top of sand-controlscreens. Reducing perforation intervals by 3 to5 ft [0.9 to 1.5 m] at the top and bottom typicallyallows enough leakoff from dynamic fractures toeffectively complete the gravel-packing portionof a treatment (above right).

After perforating, a successful TSO treatmentis essential to generate wide fractures and exter-nal proppant packs, and to promote grain-to-grain

20 Oilfield Review

> Effective perforations. Comparing Gulf of Mexico well-completion skinswith the volume of acid pumped to clean up perforation damage indicateslower productivities when a hydrochloric acid [HCl] volume of less than20 gal/ft [0.24 m3/m] was used across the perforated interval. Optimized frac-packing criteria recommend 40 to 50 gal/ft [0.5 to 0.6 m3/m] with effectivediversion across the entire zone.

1000

40

30

20

10

Acid volume, gal/ft of perforated interval

Dim

ensi

onle

ss s

kin

0 20 40 60 80

>Mitigating hydraulic fracture growth into shales.Fracturing into a shale layer limits fluid leakoff andmay make it difficult to completely pack proppantsaround sand-control screens. Perforated intervalscan be reduced by 3 to 5 ft [0.9 to 1.5 m] nearshale interfaces to allow continued fluid leakofffrom dynamic fractures.

0 50Fracture length, ft

100

XX,550

XX,500

Dept

h, ft

XX,450

Small 3- to 5-ftleakoff areasabove andbelow perforatedintervals

> Improved frac-pack productivity. Production from frac-packed wells in theGulf of Mexico Matagorda Island area doubled after switching from a fluidsystem with a 50-pounds per thousand (ppt) gallon polymer concentration inWells 1 to 4 (red) to an optimized 35-ppt polymer fluid in Wells 5 to 7 (blue). Theproductivity ratio in Well 7 would have been higher, but output was limited bysmall production tubing.

70

60

50

40

30

Gas

rate

, MM

cf/D

20

10

01 2 3 4

Well5 6 7

50 ppt 35 ppt

Autumn 2004 21

proppant contact from fracture tip to wellbore.Achieving these interrelated objectives requiresselection of appropriate treatment fluids basedon specific frac-packing criteria and analysis ofengineered injectivity-calibration tests.

Fluid Selection Treatment-fluid properties play an influentialrole in generating hydraulic fracture geometryand effectively placing proppant during any frac-turing treatment, but are particularly importantduring frac packing. Dynamic fracture width,length, height and proppant-transport capabilityare determined primarily by fluid volume, viscosity and leakoff coefficient. Optimal treatment-fluid characteristics are also impor-tant for minimizing completion damage during posttreatment flowback and cleanup.

Initially, fluid-selection criteria for frac packing were based on conventional fracturingtreatments in low-permeability, consolidated reser-voirs where fracture widths are narrow—highfluid-shear rates—and fluid leakoff is low—lessformation cool down. This led to the use of frac-packing fluids with high polymer concentrationsand higher efficiencies, or lower leakoff rates, evenin formations with higher permeabilities.

However, completion engineers soon foundthat less-efficient frac-packing fluids with lowerpolymer loadings and higher leakoff rates tendto cause less formation and proppant-pack damage, resulting in better well productivities(previous page, bottom). Failing to considertemperature changes and variations in shearrate also resulted in unnecessarily high polymerloadings, which decreased the chance of achiev-ing a tip screenout. Therefore, designers beganbasing fluid selection and polymer loadings onactual temperatures and shear rates in a fracture.4

Downhole temperatures decrease signifi-cantly during pretreatment injectivity andcalibration tests and actual frac packingbecause of rapid fluid leakoff into high-permeability formations (top right). This cool down subsequently increases apparent treatment-fluid viscosity inside the dynamic fracture anddecreases leakoff into the formation.

In addition to changes in fluid propertiescaused by temperature effects, treatment fluidsexperience varying shear rates as fracturesextend, or propagate. Fluid velocities and shearare high during fracture initiation, but decreaseby several orders of magnitude after TSO, causing corresponding increases in apparent viscosity (right).

3. Bakker E, Veeken K, Behrmann L, Milton P, Stirton G,Salsman A, Walton I, Stutz L and Underdown D: “TheNew Dynamics of Underbalanced Perforating,” OilfieldReview 15, no. 4 (Winter 2003/2004): 54–67.

4. Morales RH, Gadiyar BR, Bowman MD, Wallace C andNorman WD: “Fluid Characterization for Placing an Effective Frac/Pack,” paper SPE 71658, presented at the SPE Annual Technical Conference and Exhibition,New Orleans, September 30–October 3, 2001.

> Bottomhole temperature during fluid injection. Formation temperature is animportant consideration in the selection of fluids for frac packing. Field datafrom downhole temperature gauges indicate that the near-wellbore regioncools to 190°F during pretreatment acid, step-rate and DataFRAC fracture datadetermination testing. Low fluid efficiencies and high leakoff rates reduceheat transfer from the reservoir and significantly reduce temperatures indynamic fractures. Therefore, frac-packing fluid selection and polymerloadings should be based on actual in-situ temperatures.

Frac pack

Pressure

300

250

200

150

1001,900 2,000 2,100 2,200 2,300

12,000

10,000

8,000

6,000

4,000

2,000

0

Time, min

Botto

mho

le te

mpe

ratu

re (B

HT),

°F

Botto

mho

le p

ress

ure

(BHP

), ps

i

Temperature

Injection calibrationand DataFRAC tests

190°F

Acid

> Fluid viscosity versus shear rate. Laboratory tests were conducted at fluidshear rates (blue) typical of fracture extension and tip screenout during frac-packing operations. A 35-ppt crosslinked hydroxypropyl guar (HPG) fracturingfluid (green) exhibited adequate viscosity, while 40- and 45-ppt systems (redand gold, respectively) had significantly higher viscosities. Changes intemperature and shear can hinder TSO if polymer loadings are too high.

Shea

r rat

e, s

ec-1

Visc

osity

, cp

0 10 20 30 40100

1,000

10,000

45 ppt

40 ppt

35 ppt

Fluid shear

Time, min

Fracture extension Tip screenout0.1

1

10

100

1,000

The viscosity of gelled polymer fluids mustbreak down completely after a treatment. Frac-packing fluids that do not break quicklycan leave polymer residue in propped fracturesand proppant packs, which impairs initial wellproductivity. Chemical and breakers, such asoxidizers, encapsulated oxidizers and enzymes,added during various stages of a job, degradeborate-crosslinked fluids over time. The type ofbreaker and required concentration depend onpolymer loading, temperature and pump time.The designed breaker loading for frac-packingtreatments is determined by exposure timeinside the dynamic fracture (right).

The initial stage, or pad fluid, which ispumped without proppant, leaks off fasterbecause it is continuously creating and contact-ing new dynamic-fracture surfaces. After a tipscreenout, fracture-propagation rate decreases,fluid efficiency increases, and the stages thatfollow the pad remain in the open fracturelonger. Slurry stages pumped near the end of atreatment schedule have the least exposuretime. When the first proppant-laden slurry stagereaches the perforations, pad-fluid viscosityshould start degrading, and then quickly breakdown completely.

Intermediate slurry stages should remainstable for at least 30% of total pump time andthen break. The final slurry stages should be sta-ble for at least 20% of total pump time beforebreaking. The total time that pad and slurrystages remain stable should include workstringtraveltime. A reliable value for minimum in-situformation stress is critical in predicting fracturedimensions and designing these TSO treatments.

Fracture-Closure Pressure Minifracture injectivity tests performed before amain treatment using the DataFRAC fracturedata determination service verify parameterssuch as fracture-closure pressure, fluid leakoffcoefficient and fluid efficiency. Reliable valuesfor these parameters aid in calibration, optimiza-tion and final adjustment of treatment designs.

An incorrect fracture-closure pressure leadsto an incorrect closure time and net pressure,and consequently, to a fluid efficiency, or leakoffcoefficient, that is too low or too high. As aresult, any adjustments to fluid and proppantschedules during the main treatment couldresult in failure to achieve a TSO and a less thanoptimal stimulation. A reliable fracture-closurepressure is also essential for generating accu-rate real-time net-pressure plots, which are usedto predict fracture geometry and tip screenout.

22 Oilfield Review

> Fluid-retention time in dynamic fractures. A frac-packing treatment typicallyinvolves gradually increasing proppant concentration in several stages.Fracture exposure time for each stage expressed as a percentage of totalpumping time shows that the initial and later stages are exposed for less timethan the intermediate stages.

35

30

25

20

15

Frac

ture

exp

osur

e tim

e, %

10

5

0Pad

Stage2 5 7 931 4 6 8

> Pretreatment injectivity or minifracture tests. Determining fracture-closure stress typically involvesinjecting nondamaging, low-viscosity fluids to initiate a short, dynamic fracture. Step-rate testsgradually increase the injection rate to identify the pressure required to propagate, or extend, fracturelength (bottom). Step-rate pressure data can be extrapolated to estimate the fracture-closurepressure. Constant-rate injection followed by either constant-rate flowback or pressure decline aftershut-in also helps determine fracture closure. During flowback and pressure decline, however,pressure responses often exhibit inflection points caused by events other than fracture closure,making it difficult obtain a reliable fracture-closure stress (top).

Botto

mho

le p

ress

ure

Time

Increasinginjection

rate

Constantinjection

rate

Constantflowback

rate

Shut-in Constantinjection rate

Net pressure

Fracture-closure pressure

Fracture-extensionpressure

Reboundpressure

Posttreatmentpressure falloff

Instantaneous shut-inpressure (ISIP)

Rece

ding

frac

ture

hei

ght a

ndex

tend

ing

fract

ure

leng

th

Rece

ding

frac

ture

leng

th

Frac

ture

-face

con

tact

and

cons

olid

atio

n

Line

ar fl

ow

Tran

sitio

n flo

w

Radi

al fl

ow

Fractureclosure

Autumn 2004 23

To achieve a TSO that ensures a widerpropped fracture, field engineers rely on net-pressure plots to make real-time decisions aboutcontinuing a treatment or ending the job early.Matching net-pressure behavior using a com-puter model also helps in estimating fracturedimensions and adjusting treatment designs.

Step-rate tests using ungelled fluids andDataFRAC analysis using the actual gelled treatment fluids involve pumping fluid into a formation to analyze pressure responses duringand after injection (previous page, bottom).Pressure-decline analysis, the most commonlyused method, incorporates standardized plots toidentify the inflection point on a pressure-decline curve that represents fracture closure.

Under some conditions, however, the pressure response exhibits inflection pointsassociated with mechanisms other than fractureclosure—changes in flow regimes or gasinflux—that sometimes lead to erroneous esti-mates. A more objective and reliable test wasneeded to correctly and consistently determine

fracture-closure pressure and correctly characterize hydraulic fracture behavior.

Ideally, fracture closure should be activatedby fluid flow and controlled primarily by thedynamic fracture opening and closing to providea unique pressure response. Performed correctly, step-rate and flowback tests fit thesecriteria, as does the new equilibrium step-rate (ESR) test (above).5 The ESR test is similarto a conventional step-rate test with one exception.

This procedure decreases hydraulic fracturesurfaces to an equilibrium area where injectionrate equals leakoff rate to provide a more reliable indication of fracture-closure pressure.Fluid is injected at increasing rates to create ahydraulic fracture; then, rather than shutting inthe well, the fluid-injection rate is reduced tothe estimated fracture-propagation, or exten-sion, rate and then held constant.

The volume and pressure in the dynamicfracture subsequently decrease until the fluidleakoff and injection rates reach equilibrium. Atthat point, fracture volume stops decreasing,

and pressure stabilizes. Once this equilibriumpressure is reached, the well is shut-in and thefracture closes.

The ESR test provides a more reliable valuefor fracture-closure pressure, especially in high-permeability formations where treatment fluidsleak off quickly and hydraulic fractures close quickly.

The closure pressure is unique and easy toidentify, which avoids ambiguities associatedwith other methods. Fluid efficiency can also beestimated from the decline slope. A field-derivedcorrelation for deepwater wells provides a reliable way to estimate in-situ stress.

In the past, a common misconception wasthat higher frac-pack productivities resultedfrom placing larger volumes of proppant in a for-mation. Skin data from frac-pack completions inthe Gulf of Mexico plotted as a function of

5. Weng X, Pandey V and Nolte KG: “Equilibrium Test – AMethod for Closure Pressure Determination,” paperSPE/ISRM 78173, presented at the SPE/ISRM RockMechanics Conference, Irving, Texas, October 20–23, 2002.

> Equilibrium step-rate (ESR) test. In high-permeability formations, the ESR test provides a more reliable indication of fracture-closure pressure than othermethods. Fluid is injected at increasing rates to initiate a hydraulic fracture and estimate the injection rate required for fracture propagation, or extension(left). The injection rate is then decreased and held constant at the estimated propagation rate. When the fluid leakoff and injection rates reach equilibrium,the pressure begins to stabilize and the well is shut-in. In contrast to conventional minifracture tests and DataFRAC pressure-decline analysis (lower right),fracture-closure pressure during an ESR test is unique and easy to identify (upper right).

3,500

3,000

2,500

2,000

1,500

1,000

500

0

28

24

20

16

12

8

4

0

Pres

sure

, psi

Inje

ctio

n ra

te, b

bl/m

in

250 260

Tubing treating pressure

Rate

Live-annuluspressure

270 280Time, min

Fracture-extension rate

Pressurestabilizes

290 300 310

7,600

7,500

7,400

7,300

7,200

7,100

7,000

Botto

mho

le p

ress

ure

(BHP

), ps

i

4.5 4.6 4.7 4.8 4.9Square root total time, min1/2

7,432-psi fracture closure

5.0

Equilibrium Step-Rate Flowback

7,800

7,000

7,200

7,400

7,600

Botto

mho

le p

ress

ure

(BHP

), ps

i

2.5 3.0

L2-S

L2-EL1-E

L1-S

3.5 4.0 4.5Square root total time, min1/2

DataFRAC Flowback

5.0

7,432-psi fracture closure23% fluid efficiency

DataFRAC testEquilibrium step-rate test

proppant volume per foot indicate thatincreased proppant volumes do not necessarilydecrease skin if the treatment failed to achievea TSO (above). Another key aspect of ensuringeffective TSO treatments is complete and effec-tive fracture coverage and proppant placementacross an entire productive interval.

Fracture Coverage and Proppant PlacementFrac-pack completion designs and downholeequipment must address the complexities oftreating large reservoir sections and multiplecompletion intervals, some with perforatedintervals longer than 150 ft [46 m] and with sig-nificant permeability and stress contrasts. Evencarefully planned frac-packing treatments mayend in failure if proppant packs off, or bridges,in the screen-casing annulus, restricting orblocking annular flow. Proppant bridging maycause early treatment termination, low fractureconductivity and incomplete gravel packingaround screens.

Annular blockage near the top of a screenassembly prevents fracture stimulation andscreen packing across deeper zones. Partial flowrestriction in the annulus increases frictionalpressure drop, restricts rate distribution andlimits fracture-height growth across the comple-tion interval, especially when fracturing zoneswith higher in-situ stresses. Annular voids belowa proppant bridge increase the likelihood of subsequent screen failures caused by erosionfrom produced fluids and sand.

For homogeneous intervals that are less than60 ft [18 m] thick, fracture height typically covers the entire zone. For longer intervals, the

probability of complete fracture coveragedecreases, and risk of proppant bridgingincreases. Long intervals may be treated in separate stages. This requires more downholeequipment, such as stacked frac-packing assemblies, and additional installation time, butthese factors are often outweighed by increasedtreatment and completion effectiveness.

Alternate Path technology can also beapplied to frac pack longer intervals (right).AllFRAC screens use hollow rectangular tubes, orshunts, welded on the outside of screens to pro-vide additional flow paths for slurry. Exit portswith carbide-strengthened nozzles on the shunttubes allow fluids and proppant to exit above andbelow annular restrictions, so that fracturing andannular packing can continue even after restric-tions form in the screen-casing annulus.

Shunt tubes provide conduits for slurry to bypass collapsed hole and external zonal-isolation packers as well as annular proppantbridges at the top of intervals or adjacent tohigh-permeability zones with significant fluidleakoff. If annular restrictions form, injectionpressure increases and slurry diverts into theshunt tubes, ensuring fracture coverage andproppant packing around screens across theentire completion interval.

Alternate Path technology also allows screenmanufacturers to maximize internal screendiameters to reduce pressure drops throughdownhole assemblies and well-completionequipment. To accommodate higher injectionrates, AllFRAC screens for frac packing haveshunt tubes with slightly larger cross sectionsthan the AllPAC screens used for gravel packing.

24 Oilfield Review

> Effective treatment placement. AllFRACAlternate Path screens have shunt tubes withstrategically located exit nozzles welded on theoutside of screen assemblies (top and middle).Large shunt tubes provide a flow path for slurry tobypass annular restrictions caused by theformation of proppant bridges, allowing continuedfracture stimulation of lower intervals andcomplete packing of any annular voids aroundthe screens that may form as a result of proppantbridging in the screen-casing annulus (bottom).

Screen

Basepipe

Fracture

Shunttubes

Nozzle

Casing

Shunt tubes

Perforations Screens

Annularproppant bridge

Annularvoid

Nozzle

> Effective treatment designs. Comparing Gulf of Mexico frac-packcompletions indicates that placing higher volumes of proppant in the formationdoes not correlate with reduced skin factors. Three wells had extremely highskins, above 20 to more than 35, even though proppant volumes exceeded2,000 lbm/ft [3,000 kg/m] of perforated interval. In these cases, high skins wereattributed to not achieving a TSO.

6,0004,0002,000-5

40

35

30

25

20

15

10

5

0

Volume of proppant placed, lbm/ft of perforated interval

Dim

ensi

onle

ss s

kin

1

Autumn 2004 25

When these frac-pack design considerationsare carefully addressed, operators should expectdesirable downhole and surface pressure signatures, indicating an effective TSO fractureand complete packing of the screen-casingannulus (above). Total and Marathon Oil Com-pany successfully implemented optimizedfrac-packing techniques in two ultradeepwaterGulf of Mexico gas fields.

Deepwater Proving Ground Total E&P USA, Inc. operates the Aconcaguafield 140 miles [225 km] southeast of NewOrleans in Mississippi Canyon Block 305. TheCamden Hills field operated by Marathon OilCompany lies in adjacent Mississippi CanyonBlock 348. These two fields, along with the BP-operated King’s Peak field, comprise theCanyon Express development (right).6

6. Carré G, Pradié E, Christie A, Delabroy L, Greeson B, Watson G, Fett D, Piedras J, Jenkins R, Schmidt D, Kolstad E, Stimatz G and Taylor G: “High Expectationsfrom Deepwater Wells,” Oilfield Review 14, no. 4 (Winter 2003/2003): 36–51. “Canyon Express Setting Records in Subsea Development,”supplement to Hart’s E&P and Hart’s Oil and Gas Investor,April 2003. “Camden Hills: A World Record Achieved Through Innovative Solutions,” supplement to Hart’s E&P,April 2003.

> Optimized frac packing. Bottomhole treating-pressure response for an effective TSO treatment demonstrates a significant increase in net pressureassociated with proppant packing in a wide fracture rather than continued length propagation and height growth (top). As in conventional gravel packing, a spike in the surface pressure is indicative of a completely packed screen-casing annulus (bottom).

Botto

mho

le p

ress

ure

(BHP

), ps

i

Time. min

Time, min

Wel

lhea

d pr

essu

re (W

HP),

psi

Rate

, bbl

/min

Prop

pant

con

cent

ratio

n, p

pa

WHP

BHP Rate

Proppant

Rate

, bbl

/min

Prop

pant

con

cent

ratio

n, p

pa

11,500

11,000

10,500

10,000

9,500

9,000

8,5000 10 20

0 10 20 30 40 50 60

6,000

5,000

4,000

3,000

2,000

50

40

30

20

10

0

30 40 50 60

20

10

0

30

40

50

Rate

Proppant

Fracture packing

Annular packingCasing Screens

> Canyon Express subsea development. The Total Aconcagua and Marathon Camden Hills fields along with the BP King’s Peak field are located about 140 miles [225 km] southeast of New Orleans inGulf of Mexico Mississippi Canyon Blocks 305 and 348. The subsea infrastructure is tied to a subsea,multiphase gathering system. The subsea wells in all three fields produce gas through dual pipelines tothe Canyon Station host platform located 56 miles [90 km] away in shallower outer continental shelfwater. Before agreeing to a shared gathering and processing system, Total, Marathon and BPevaluated other options, such as spars and other stand-alone facilities. The complexity of subseaoperations and limited reserves made it uneconomic to develop these fields independently.

Camden Hills field–two wells operated by Marathon, 7,200-ft water depthAconcagua field–four wells operated by Total, 7,100-ft water depthKing’s Peak field–three wells operated by BP, 6,200-ft water depth

Canyon Station

Destin Dome

Water depth, ft

1,000

3,000

5,000

7,000

20 miles100

0 15 30 km

Aconcagua fieldKing‘s Peak field

Camden Hills field

Desoto Canyon

MississippiCanyon

Gulf of Mexico

Canyon Express Project

USA

Viosca Knoll

The four wells in the Aconcagua field, twowells in the Camden Hills field and three wellsin the King’s Peak field are located in 6,200 to7,200 ft [1,890 to 2,195 m] of water. These sub-sea wells produce into a common multiphasegathering system. Dual pipelines then transportproduced gas to Canyon Station, a processing

platform operated by Williams Energy Services,about 56 miles [90 km] away in Main PassBlock 261. Production is selectively controlledusing an intelligent well completion system.7

These fields comprise a series of high-permeability, unconsolidated sands overlying

water-bearing sands in some areas. In addition,most of these reservoirs consist of multiplesands separated by shale layers. Production history from these types of reservoirs indicatesthat gas output drops quickly once water production begins.

26 Oilfield Review

> Typical Canyon Express well configuration. Aconcagua field wellsincorporated state-of-the-art Alternate Path sand-control screens because oflong completion intervals. Camden Hills field wells used Weatherfordprepacked screens without shunt tubes because of shorter completionintervals. The sump packer, lower sand-control assembly, upper sand-controlassembly and isolation packer were installed in four separate runs. The uppercompletion equipment above the production-seal assembly was installed in asingle operation. A SenTREE 7 subsea well control system installed in theupper completion performed 24 functions, including automated flow control,permanent monitoring and emergency well shut-in with a 15-seconddisconnect at any water depth.

Tubing hanger

Methanol-injectionmandrelTRC-DH-10-LOsafety valve

Chemical-injectionmandrelPacker-setting device

Splice subProduction packer

Upper flow-control valveLower flow-control valveLanding nippleWireline reentry guide7-in. shroudLanding nipple forlower zone isolation3 1⁄2-in. isolation tubingQUANTUM isolation packerProduction-seal assembly

AFIV device

QUANTUM maX packerMechanical FIV device2 7⁄8-in. tubing withcarbide blast ringsAllFRAC screens (Aconcagua field)Weatherford prepacked screens(Camden Hills field)Shifting tool

Gauge carrier withthree pressure andtemperature sensorsCross-nipple for upperzone isolation

Packer-settingmechanical override

9 5⁄8 -in. liner top

4 1⁄2 -in. productiontubing

QUANTUM maX packerHydraulic/mechanical FIV deviceAllFRAC screens (Aconcagua field)Weatherford prepacked screens(Camden Hills field)Sump packer

Upperinterval

9 5⁄8 -in.linertop

Lowerinterval

Autumn 2004 27

Completing and draining multiple sands andhaving the capability of controlling water production without conventional remedial rigintervention were critical elements in the plan-ning and design of these wells. Well-completionequipment consisted of a sump, or bottom,packer; AllFRAC screen assemblies and an isola-tion packer with tubing or annular isolationdevices for each completion interval; and an isolation packer and upper completion equip-ment (previous page).

Total and Marathon completed a total of 13perforated intervals in six subsea wells, four inthe Aconcagua field and two in the Camden Hillsfield. Each well had at least two completion inter-vals that were frac packed with the objective ofachieving completions skins of less than five.8

Well operations, which were suspended afterbatch drilling all the wells, resumed by reenter-ing the wells to drill out temporary cement plugsand replace mud with less damaging completionfluids. Surfactant pills were circulated to removeremaining mud deposits. Casing brushes, 360°scrapers, screen-type junk baskets and magnetsremoved any other debris. The subsea riser andcavities in the blowout preventer (BOP) stackwere cleaned with brush and jetting tools.

Following these cleanout procedures, thewellbores were displaced with seawater followed

by filtered calcium chloride [CaCl2] brine to provide a 300- to 700-psi [2.1- to 4.8-MPa] hydro-static overbalance. Total ran an ultrasoniccement bond log with a gamma ray correlationlog and casing collar locator on the Total wells.Based on previous results, the Marathon well-completion program did not include a cementbond log. A lower sump packer run on wirelineand set below the deepest perforations provideda reference depth for subsequent operations inall the wells.

Productive intervals were perforated with a400- to 600-psi [2.8- to 4.1-MPa] overbalanceusing tubing-conveyed perforating (TCP) techniques. The relatively simple TCP gunassemblies consisted of a locating snap-latch,gun sections with charges loaded at 12 shots perfoot (spf) and 120° or 60° phasing, a pressure-disk to hold fluid in the tubing and a hydraulicfiring head with backup drop bar. One of theCamden Hills field wells used 18 spf. The wellswere not flowed after perforating. This methodproved to be simple, reliable and relatively low-risk compared to underbalanced perforating inunconsolidated sands.

Perforated intervals in the Aconcagua fieldranged from 35 to 111 ft [11 to 34 m] in lengthwith a true vertical depth (TVD) of about12,700 ft [3,870 m]. Average reservoir pressure

was 6,300 psi [43.4 MPa], and bottomhole statictemperature (BHST) was 128°F [53°C]. Twozones had perforated interval lengths greaterthan 100 ft [30 m] with high inclination anglesof 30° to 53°.

Productive zones in the Camden Hills fieldwere close to a water contact, so fracture-heightgrowth was a concern. Perforated-intervallengths varied from 46 to 65 ft, [14 to 20 m] at aTVD of about 14,000 ft [4,267 m]. Reservoir pressure was 7,065 psi [48.7 MPa] and BHST was155°F [68°C].

Because of high wellbore inclination anglesin the Aconcagua field, Total selected wire-wrapped Alternate Path AllFRAC screens withshunt tubes to obtain uniform fracture stimula-tion and complete packing of the screen-casingannulus across longer completion intervals.Marathon chose prepacked Weatherford screenswith 20/40 mesh resin-coated proppant for theCamden Hills field where shorter completionintervals did not require the use of AlternatePath technology.

After perforating, the sand-control screenassembly for the lower completion interval,including an FIV Formation Isolation Valve tool,was run. Bottomhole temperature and pressuregauges were run to evaluate treatment placement.

In some cases, the average skin for conven-tional frac-packs in the Gulf of Mexico is greaterthan 10 (left). Total and Marathon applied opti-mized frac-packing techniques with theobjectives of reducing completion skin anddepleting these smaller fields more effectivelywithout future remedial intervention.

Prior to frac-packing operations, Total andMarathon pumped 50 gal/ft [0.6 m3/m] of 10%hydrochloric acid [HCl] to remove perforationdamage. Treatment fluids were selected basedon cool-down temperatures of 87 to 95°F [31 to35°C] using the CoolFRAC optimized fracturingservice for high-permeability frac packs.

Before the main treatments, Total performedDataFRAC analysis that included ESR testsusing a linear gel to accurately determine fracture-closure pressure. Subsequent injection-calibration treatments used a 20-pound per

7. de Reals BT, Lomenech H, Nogueira AC, Stearns JP andFerroni L: “Canyon Express Deepwater Flowline System:Design and Installation,” paper OTC 15096, presented at the Offshore Technology Conference, Houston, May 5–8, 2003.

8. Piedras JM, Stimatz GP, Jackson Nielsen VB and Watson GM: “Canyon Express: Design and Experience on High-Rate Deepwater Gas Producers Using Frac-Packand Intelligent Well Completion Systems,” paper OTC15094, presented at the Offshore Technology Conference,Houston, May 5–8, 2003.

> Frac-pack productivity. Gulf of Mexico pressure-buildup data indicate thatcompletion skin is higher than expected in many frac-pack completions,leaving room for further optimization. In this evaluation, skins increased withincreasing permeability-height (kh), averaging 10.3 for these 95 treatments.

1,000,00010,000100-20

0

20

40

60

80

100

Permeability-height (kh), mD-ft

Number of points = 95Average skin = 10.3

Dim

ensi

onle

ss s

kin

1

thousand (ppt) crosslinked fluid. Marathoneliminated DataFRAC tests from Camden Hillsfield frac-pack procedures to avoid promotingfracture growth into nearby water sands.

Frac-packing treatments were pumped fromthe Schlumberger DeepSTIM I and DeepSTIM IIstimulation vessels, which were moored along-side the Transocean Discoverer Spirit drillship(above). The main frac-packing treatments wereperformed at 20 to 30 bbl/min [3.2 to4.8 m3/min] using 20/40 man-made ceramicproppant and a 20-ppt crosslinked fluid.

After the operator encountered difficultyachieving a complete annular pack in the firstAconcagua well, perforation intervals werereduced to limit fracture growth into boundingshale layers and provide for continued fluidleakoff. Subsequently, a 120-bbl [19-m3] slurrystage with 8 to 10 pounds of proppant added(ppa) per thousand gallons was included in thepumping schedule to allow for controlled reduc-tion in injection rates at the end of a treatment.These steps ensured complete gravel packing onsubsequent jobs without pumping separatetreatments to top-off and pack the screen top.

In the Camden Hills field, Marathon andSchlumberger specialists designed unconven-tional frac packs using a less efficientcrosslinked fluid to control fracture-heightgrowth through excessive leakoff and to preventfracture propagation into wet sands. Design frac-ture lengths were 20 to 30 ft [6.1 to 9.1 m] with a1-in. propped fracture width. For the Aconcaguafield completions, Total increased the polymerconcentration slightly and designed fracturelengths of 40 to 50 ft [12.2 to 15.2 m].

Following frac-packing operations on thelower completion interval in each well, an isola-tion plug was set in the lower frac-pack packer.Sand or calcium carbonate pills were spotted ontop of this plug to facilitate cleaning out debrisafter perforating the upper completion interval.The operators perforated the upper completionintervals, retrieved the packer isolation plug andran sand-control equipment with an FIV toolbefore frac packing.

A seal assembly below the isolation packerconnected into a polished-bore receptacle (PBR)at the top of the lower screen assembly. An AFIVannular-controlled FIV system in the isolationpacker assembly provided flow isolation for theupper and lower completion intervals.

Zonal Isolation and Fluid-Loss Control Canyon Express well designs integrated an FIVsystem with QUANTUM gravel-pack packerequipment for lower completion intervals. Thesand-control equipment for upper intervalsincluded an AFIV system. These devices assure ahigh level of well control without expensive andrisky well-intervention operations. The FIV andAFIV systems have a proven record of providingsafe, reliable zonal isolation in various drillstemtesting and downhole tool applications.

These valves allow independent, two-way isolation and control of each interval to preventfluid losses and gas influx during completion andflowback operations. The FIV and AFIV valvesalso facilitated pressure-integrity tests withoutwireline or coiled tubing intervention before thewells were opened for production.

In each of these Canyon Express wells, ashifting tool below the internal washpipe passedback through the screens as the gravel-packingworkstring tubing was retrieved. This toolshifted a sleeve that closed the respective valvesto isolate each completion interval after frac-packing treatments. The FIV and AFIVvalves could also be opened with a similar shift-ing tool run on wireline or coiled tubing, and theFIV ball valve could be milled through tubing asa contingency.

28 Oilfield Review

> Frac-packing Aconcagua field wells. The Schlumberger DeepSTIM I and DeepSTIM II stimulationvessels performed frac-pack combined stimulation and gravel packing for Total while moored besidethe Transocean Discoverer Spirit drillship during the completion of subsea wells in Mississippi CanyonBlock 305.

Autumn 2004 29

Installing the isolation packer assembly afterfrac packing the upper interval mechanicallyshifted the upper AFIV open. A series of specificpressure cycles applied to the production tubinghydraulically opened the lower FIV device. Thisallowed the lower interval to be produced with-out intervention after the production packer andupper completion equipment were in place.

Reduced Completion Skins During frac-packing operations, downholegauges recorded and transmitted pressure andtemperature data to surface. There was evidenceof treatment fluids bypassing localized bridgesthrough the shunts. There were also changes inthe pressure-curve slope associated with varia-tions in temperature, which indicated diversionthrough the shunts (top).

Fracture-closure pressures from conven-tional injection and minifracture tests were tooambiguous for critical ultradeepwater well completions. More reliable ESR analysis ensuredoptimal treatment designs and execution in theAconcagua field.

Total performed Aconcagua field frac packsin circulating position and tracked bottomholepressure (BHP) in real time by monitoring thetubing-casing annulus during the jobs performedin this field. Net-pressure gains of between 300and 1,100 psi [2.1 and 7.6 MPa], indicating effec-tive TSO results, were observed while pumpingthese optimized frac-packing treatments.Marathon performed Camden Hills field fracpacks in squeeze position and did not monitorBHP using a live annulus.

Calculated completion skins ranged from–1.5 to 4 with an average of 3.06 for 13 intervalsin the six wells in Aconcagua and Camden Hillsfields, much better than the previous average of10.3 (below left). Production began just afterthe sixth well was completed. Effective TSO fracture treatments optimized production fromfrac-pack completions in high-permeability,unconsolidated formations. Each of the wells iscapable of producing more than the target rateof 50 MMcf/D per well [1.4 million m3/d].

Successful TSO treatments in the ultra-deepwater Aconcagua and Camden Hills gasfields were achieved through improved fluiddesigns and frac-packing procedures. Treat-ments designed to achieve TSO fractures tookprecedence over larger proppant volumesbecause placing more proppant does not neces-sarily impact frac-pack productivity significantly.

Treatment fluids were selected based on in-situ cool-down temperatures and shear ratesinside the hydraulic fracture. In some cases,analysis of pretreatment injection-calibrationtests using the new ESR method helped deter-mine more accurate fracture-closure pressures.

Prior to executing frac-pack treatments, a sufficient volume of acid was pumped toensure clean perforations. Alternate Pathscreens with shunt tubes facilitated treatmentdiversion in these multizone reservoirs with longcompletion intervals.

As in many endeavors, stimulation and gravel-packing improvements have evolved from a betterunderstanding of basic principles and the refin-ing of existing technologies and practices. Testedin this harsh ultradeepwater environment, theseoptimized techniques can be applied in otherareas to ensure frac-packing success. —MET

> Effective stimulation and packing. Data from permanent bottomhole gaugesshow pressure responses associated with changes in temperature that are indicative of slurry diversion through Alternate Path shunt tubes duringfrac packing.

8,500

Botto

mho

le p

ress

ure

(BHP

), ps

i

Tem

pera

ture

, °F

Time, min

8,400

8,300

8,200

8,100

8,000

7,900

7,800

7,700

7,600

7,500

110

100

90

80

7025 30 35 40

Slope change

Sharp pressure increase

Middle temperature gauge

Top temperature gauge

Bottom temperature gauge

> Optimal frac-pack results. Optimized techniques resulted in an averagefrac-pack skin of 3.06 for intervals completed in six subsea wells of theAconcagua and Camden Hills fields.

Dim

ensi

onle

ss s

kin

Permeability-height (kh), mD-ft10 100 1,000 10,000 100,000

10

0

-10

Gulf of Mexico frac-packing trend