Optimizing Frac Packs

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.

Bala Gadiyar
New Orleans, Louisiana, USA Tip screenout
Casing
Craig Meese
Greg Stimatz
Marathon Oil Company Dynamic fracture
Proppant
Houston, Texas, USA

Hugo Morales Washpipe

Houston, Texas
Perforation
Jose Piedras
Total E&P USA, Inc. Cement
Houston, Texas
Fracture inflation Screen
Jérôme Profinet
Total, Elf Petroleum Nigeria, Ltd.
Port Harcourt, Nigeria

Graham Watson Annular opening

Pau, France

For help in preparation of this article, thanks to Billy Greeson,

Houston, Texas, USA.
AFIV (annular-controlled FIV system), CoolFRAC, DataFRAC,
“External” proppant pack
DeepSTIM, FIV (Formation Isolation Valve), QUANTUM,
QUANTUM maX and SenTREE are marks of Schlumberger. Propped fracture
AllFRAC, AllPAC and Alternate Path are marks of Mobil Oil
Corporation, now ExxonMobil, that are licensed to
Schlumberger.
1. Ali S, Norman D, Wagner D, Ayoub J, Desroches J, Annular proppant pack
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.
> Frac packing. Tip-screenout (TSO) designs use fluids that leak off early in a treatment, causing
2. Acock A, Heitmann N, Hoover S, Malik BZ, Pitoni E, Riddles
C and Solares JR: “Screenless Methods to Control Sand,” proppant to pack off at the fracture tips (top). Pumping additional proppant-laden fluid, or slurry, causes
Oilfield Review 15, no. 1 (Spring 2003): 38–53. the biwing fractures to inflate as proppant packs back toward the well (middle). A TSO generates
enough 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 further
reduces near-well pressure drop (bottom).

18 Oilfield Review
As offshore oil and gas development moves into
harsher, more demanding deepwater environ-
ments, frac-packing utilization and methods
continue to expand and evolve based on specific
field experience and requirements. These
tip-screenout (TSO) fracturing treatments
performed in conjunction with gravel packing of
mechanical screens now represent almost 65%
of sand-control completions in the Gulf of
Mexico, USA. Since it was first applied in the
early 1990s, frac packing has become one of the
most widely used methods for completing wells
in weakly consolidated formations.
This combined stimulation and sand-control
technique has proved effective in a wide range
of formations with mobile solids, especially high-
permeability reservoirs.1 Frac packs consistently
yield sustained production increases compared
with viscous slurry packs or high-rate water packs.
Frac packing avoids many of the productivity
impairments common in conventional cased-hole
gravel packs by effectively bypassing formation
damage, or skin, and by creating an “external”
pack to stabilize perforations that are not aligned
with 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 cause
the proppant-laden slurry stages to dehydrate
early in a treatment. Proppants pack off near
the end, or tip, of dynamic fractures, causing
them to inflate like a balloon as additional slurry
is injected. Proppant then packs back toward
the well, which promotes grain-to-grain contact
and generates a wider, more conductive pathway
after the dynamic fracture closes.
In many ways, frac packing is a mature
technology. Service companies have comparable
pumping equipment, stimulation vessels,
downhole tools and laboratory support. They
also provide similar fluids, proppants, treatment
additives and damage-removal techniques.
Other well technologies—intelligent well com-
pletions, multiple-zone production monitoring
and control, real-time data and information
transfer, safety and quality control—have also
reached 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 for
improvement. Frac-pack productivity may be
lower than expected because of a combination of
factors, including perforation damage, failure to
achieve TSO, incomplete fracture or proppant-
pack coverage and high pressure drops through
downhole sand-exclusion screens and well-
completion equipment.
Autumn 2004
Propped fracture
100 Unaligned perforations
Ratio of fracture flow to total flow, %
Cased-hole perforated
80
completion, 4 spf, 90° phasing
60
40
Openhole completion
20
0
10 100 1,000 10,000
Formation permeability, mD
> Fracture flow versus inflow from unaligned perforations. Inflow is not limited
to propped fracture cross-sectional area. Computer modeling indicates that
perforations aligned away from the preferred fracture plane (PFP) contribute
almost 50% of the inflow from high-permeability formations, underscoring the
importance of TSO fracturing and creation of an external proppant pack. This
simulation compares a low perforation shot density of 4 shots per foot (spf) in
cased hole (red) with an openhole completion (green), that has infinite
perforation density.
Frac-pack optimization involves addressing experience and results from the ultradeepwater
all of these completion-design factors to reduce Aconcagua and Camden Hills fields in the
overall skin and improve well productivity, to Gulf of Mexico Canyon Express project,
maximize hydrocarbon recovery, and to help which contributed to a better understanding of
operators avoid future well interventions. The frac packing.
latter objective is critically important in deep-
water fields, particularly those with subsea Effective Perforations
wells, where remedial operations to remove Reservoir stimulation, or fracture conductivity,
damage or restimulate wells are extremely alone does not ensure an optimal frac pack. An
difficult, complex and costly. effective external proppant pack is also
Total, Marathon and Schlumberger refined required. A ring of proppant around the wellbore
existing completion practices and frac-packing stabilizes all the perforations and hydraulically
techniques in the Gulf of Mexico using field connects them with the propped fracture. This
experience and improved computer modeling further minimizes frac-pack skin and reduces
of fracturing and frac-packing processes. pressure drop across the completion interval to
Completion engineers now select optimal treat- help avoid formation failures and subsequent
ment fluids and adjust frac-pack designs to sand production. An external pack is also the
account for in-situ temperatures and fluid shear basis for screenless completions that control
during treatment execution. sand without mechanical screens and internal
This article discusses methods for cleaning gravel packs.2
perforations and for selecting treatment fluids Computer modeling indicates that unaligned
that achieve effective TSO fractures, including a perforations, those oriented away from the
reliable means of determining fracture-closure preferred fracture plane (PFP), contribute
pressure. It presents well-completion equipment as much as 50% of the inflow from high-
that ensures complete stimulation and gravel permeability formations (above). This under-
packing across long intervals, maximizes scores the importance of eliminating flow
internal flow area, and allows evaluation of restrictions in and around all of the perforations.
frac-packing efficiency. It also summarizes field
19
 40
XX,450
Small 3- to 5-ft
30 leakoff areas
above and
Dimensionless skin

below perforated

Depth, ft
XX,500 intervals
20

10 XX,550

0
0 20 40 60 80 100
Acid volume, gal/ft of perforated interval 0 50 100
Fracture length, ft
> Effective perforations. Comparing Gulf of Mexico well-completion skins
> Mitigating hydraulic fracture growth into shales.
with the volume of acid pumped to clean up perforation damage indicates
lower productivities when a hydrochloric acid [HCl] volume of less than Fracturing into a shale layer limits fluid leakoff and
20 gal/ft [0.24 m3/m] was used across the perforated interval. Optimized frac- may make it difficult to completely pack proppants
packing criteria recommend 40 to 50 gal/ft [0.5 to 0.6 m3/m] with effective around sand-control screens. Perforated intervals
diversion across the entire zone. can be reduced by 3 to 5 ft [0.9 to 1.5 m] near
shale interfaces to allow continued fluid leakoff
from dynamic fractures.

Explosive-jet perforating causes a crushed

zone of damage around perforation tunnels. This
damage can be addressed by pumping acid to
remove perforating damage and debris prior to
frac packing or by applying more effective perfo-
rating practices, such as dynamic underbalance
techniques.3 Analysis of Gulf of Mexico well com-
pletions indicates that skin factors were high
when acid volumes of less than 20 gal/ft 70
[0.24 m 3 /m] were used across perforated
intervals; pumping acid volumes of 40 to 50 gal/ft 60
[0.5 to 0.6 m3/m] with effective diversion across 50
the entire zone minimized completion skin
Gas rate, MMcf/D

(above left). 40
50 ppt 35 ppt
Careful consideration should also be given to
30
selecting perforated intervals to avoid unwanted
hydraulic fracture-height growth in shale layers 20
above and below productive intervals. Fractur-
ing into shale restricts fluid leakoff. Dynamic 10
fractures in shale remain open longer because
0
treating fluids do not leak off fast enough. This 1 2 3 4 5 6 7
also makes it difficult to obtain a complete Well
gravel pack around the top of sand-control > Improved frac-pack productivity. Production from frac-packed wells in the
screens. Reducing perforation intervals by 3 to Gulf of Mexico Matagorda Island area doubled after switching from a fluid
5 ft [0.9 to 1.5 m] at the top and bottom typically system with a 50-pounds per thousand (ppt) gallon polymer concentration in
allows enough leakoff from dynamic fractures to Wells 1 to 4 (red) to an optimized 35-ppt polymer fluid in Wells 5 to 7 (blue). The
productivity ratio in Well 7 would have been higher, but output was limited by
effectively complete the gravel-packing portion small production tubing.
of a treatment (above right).
After perforating, a successful TSO treatment
is essential to generate wide fractures and exter-
nal proppant packs, and to promote grain-to-grain

20 Oilfield Review
proppant contact from fracture tip to wellbore. 300 12,000
Achieving these interrelated objectives requires Injection calibration Frac pack
and DataFRAC tests

Bottomhole temperature (BHT), °F

selection of appropriate treatment fluids based 10,000

Bottomhole pressure (BHP), psi

on specific frac-packing criteria and analysis of 250
8,000
engineered injectivity-calibration tests. Acid Pressure
200 6,000
Fluid Selection
190°F Temperature
Treatment-fluid properties play an influential 4,000
role in generating hydraulic fracture geometry 150
and effectively placing proppant during any frac- 2,000
turing treatment, but are particularly important
100 0
during frac packing. Dynamic fracture width,
1,900 2,000 2,100 2,200 2,300
length, height and proppant-transport capability
Time, min
are determined primarily by fluid volume,
> Bottomhole temperature during fluid injection. Formation temperature is an
viscosity and leakoff coefficient. Optimal important consideration in the selection of fluids for frac packing. Field data
treatment-fluid characteristics are also impor- from downhole temperature gauges indicate that the near-wellbore region
tant for minimizing completion damage during cools to 190°F during pretreatment acid, step-rate and DataFRAC fracture data
posttreatment flowback and cleanup. determination testing. Low fluid efficiencies and high leakoff rates reduce
heat transfer from the reservoir and significantly reduce temperatures in
Initially, fluid-selection criteria for frac dynamic fractures. Therefore, frac-packing fluid selection and polymer
packing were based on conventional fracturing loadings should be based on actual in-situ temperatures.
treatments in low-permeability, consolidated reser-
voirs where fracture widths are narrow—high
fluid-shear rates—and fluid leakoff is low—less
formation cool down. This led to the use of frac-
packing fluids with high polymer concentrations
and higher efficiencies, or lower leakoff rates, even
in formations with higher permeabilities.
However, completion engineers soon found
that less-efficient frac-packing fluids with lower
polymer loadings and higher leakoff rates tend
to cause less formation and proppant-pack 1,000 10,000
damage, resulting in better well productivities
(previous page, bottom). Failing to consider 45 ppt
temperature changes and variations in shear 100
rate also resulted in unnecessarily high polymer
Shear rate, sec-1

40 ppt

loadings, which decreased the chance of achiev-
ing a tip screenout. Therefore, designers began
basing fluid selection and polymer loadings on
actual temperatures and shear rates in a fracture.4
Downhole temperatures decrease signifi-
cantly during pretreatment injectivity and
calibration tests and actual frac packing
because of rapid fluid leakoff into high-
permeability formations (top right). This cool
down subsequently increases apparent treatment-
fluid viscosity inside the dynamic fracture and
decreases leakoff into the formation.
In addition to changes in fluid properties
caused by temperature effects, treatment fluids
experience varying shear rates as fractures
extend, or propagate. Fluid velocities and shear
are high during fracture initiation, but decrease
by several orders of magnitude after TSO,
causing corresponding increases in apparent
viscosity (right).
Viscosity, cp
10 1,000
35 ppt
Fluid shear
1
Fracture extension Tip screenout
0.1 100
0 10 20 30 40
Time, min
> Fluid viscosity versus shear rate. Laboratory tests were conducted at fluid
shear rates (blue) typical of fracture extension and tip screenout during frac-
packing operations. A 35-ppt crosslinked hydroxypropyl guar (HPG) fracturing
fluid (green) exhibited adequate viscosity, while 40- and 45-ppt systems (red
and gold, respectively) had significantly higher viscosities. Changes in
temperature and shear can hinder TSO if polymer loadings are too high.
3. Bakker E, Veeken K, Behrmann L, Milton P, Stirton G, 4. Morales RH, Gadiyar BR, Bowman MD, Wallace C and
Salsman A, Walton I, Stutz L and Underdown D: “The Norman WD: “Fluid Characterization for Placing an
New Dynamics of Underbalanced Perforating,” Oilfield Effective Frac/Pack,” paper SPE 71658, presented at
Review 15, no. 4 (Winter 2003/2004): 54–67. the SPE Annual Technical Conference and Exhibition,
New Orleans, September 30–October 3, 2001.

Autumn 2004 21
 The viscosity of gelled polymer fluids must 35
break down completely after a treatment.
Frac-packing fluids that do not break quickly 30
can leave polymer residue in propped fractures

Fracture exposure time, %

25
and proppant packs, which impairs initial well
productivity. Chemical and breakers, such as 20
oxidizers, encapsulated oxidizers and enzymes,
15
added during various stages of a job, degrade
borate-crosslinked fluids over time. The type of 10
breaker and required concentration depend on
polymer loading, temperature and pump time. 5
The designed breaker loading for frac-packing
0
treatments is determined by exposure time Pad 1 2 3 4 5 6 7 8 9
inside the dynamic fracture (right). Stage
The initial stage, or pad fluid, which is > Fluid-retention time in dynamic fractures. A frac-packing treatment typically
pumped without proppant, leaks off faster involves gradually increasing proppant concentration in several stages.
because it is continuously creating and contact- Fracture exposure time for each stage expressed as a percentage of total
ing new dynamic-fracture surfaces. After a tip pumping time shows that the initial and later stages are exposed for less time
than the intermediate stages.
screenout, fracture-propagation rate decreases,
fluid efficiency increases, and the stages that
follow the pad remain in the open fracture
longer. Slurry stages pumped near the end of a
treatment schedule have the least exposure
time. When the first proppant-laden slurry stage
reaches the perforations, pad-fluid viscosity

Fracture-face contact
should start degrading, and then quickly break

and consolidation

Transition flow
down completely.

Radial flow
Linear flow
Intermediate slurry stages should remain
Receding fracture height and

stable for at least 30% of total pump time and

extending fracture length

then break. The final slurry stages should be sta-

ble for at least 20% of total pump time before
Receding fracture

Fracture
breaking. The total time that pad and slurry closure
stages remain stable should include workstring
length

traveltime. A reliable value for minimum in-situ

formation stress is critical in predicting fracture
dimensions and designing these TSO treatments.

Fracture-Closure Pressure
Minifracture injectivity tests performed before a Fracture- Instantaneous shut-in
extension
Bottomhole pressure

pressure (ISIP)
main treatment using the DataFRAC fracture pressure Net pressure
data determination service verify parameters
such as fracture-closure pressure, fluid leakoff Fracture-closure pressure
coefficient and fluid efficiency. Reliable values Rebound
for these parameters aid in calibration, optimiza- pressure

tion and final adjustment of treatment designs.

An incorrect fracture-closure pressure leads
to an incorrect closure time and net pressure,
and consequently, to a fluid efficiency, or leakoff
coefficient, that is too low or too high. As a
result, any adjustments to fluid and proppant
schedules during the main treatment could
result in failure to achieve a TSO and a less than
optimal stimulation. A reliable fracture-closure
pressure is also essential for generating accu-
rate real-time net-pressure plots, which are used
to predict fracture geometry and tip screenout.
Increasing Constant Constant Shut-in Constant Posttreatment
injection injection flowback injection rate pressure falloff
rate rate rate
Time
> Pretreatment injectivity or minifracture tests. Determining fracture-closure stress typically involves
injecting nondamaging, low-viscosity fluids to initiate a short, dynamic fracture. Step-rate tests
gradually increase the injection rate to identify the pressure required to propagate, or extend, fracture
length (bottom). Step-rate pressure data can be extrapolated to estimate the fracture-closure
pressure. Constant-rate injection followed by either constant-rate flowback or pressure decline after
shut-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).

22 Oilfield Review
 Equilibrium Step-Rate Flowback
7,600

Bottomhole pressure (BHP), psi

7,500
7,432-psi fracture closure
7,400

7,300

Equilibrium step-rate test DataFRAC test 7,200

3,500 28
Tubing treating pressure 7,100
3,000 24

Injection rate, bbl/min

7,000
2,500 20 4.5 4.6 4.7 4.8 4.9 5.0
Rate
Pressure, psi

2,000 16 Square root total time, min1/2

1,500 12
DataFRAC Flowback
1,000 Live- 8 7,800
annulus

Bottomhole pressure (BHP), psi

500 4
pressure
0 0 7,600
L1-E
250 260 270 280 290 300 310 L2-E 23% fluid efficiency
Time, min
7,432-psi fracture closure
7,400
L1-S
Fracture- Pressure L2-S
extension rate stabilizes 7,200

7,000
2.5 3.0 3.5 4.0 4.5 5.0
Square root total time, min1/2
> Equilibrium step-rate (ESR) test. In high-permeability formations, the ESR test provides a more reliable indication of fracture-closure pressure than other
methods. 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).

To achieve a TSO that ensures a wider
propped fracture, field engineers rely on net-
pressure plots to make real-time decisions about
continuing a treatment or ending the job early.
Matching net-pressure behavior using a com-
puter model also helps in estimating fracture
dimensions and adjusting treatment designs.
Step-rate tests using ungelled fluids and
DataFRAC analysis using the actual gelled
treatment fluids involve pumping fluid into a
formation to analyze pressure responses during
and after injection (previous page, bottom).
Pressure-decline analysis, the most commonly
used method, incorporates standardized plots to
identify the inflection point on a pressure-
decline curve that represents fracture closure.
Under some conditions, however, the
pressure response exhibits inflection points
associated with mechanisms other than fracture
closure—changes in flow regimes or gas
influx—that sometimes lead to erroneous esti-
mates. A more objective and reliable test was
needed to correctly and consistently determine
fracture-closure pressure and correctly and pressure stabilizes. Once this equilibrium
characterize hydraulic fracture behavior. pressure is reached, the well is shut-in and the
Ideally, fracture closure should be activated fracture closes.
by fluid flow and controlled primarily by the The ESR test provides a more reliable value
dynamic fracture opening and closing to provide for fracture-closure pressure, especially in high-
a unique pressure response. Performed permeability formations where treatment fluids
correctly, step-rate and flowback tests fit these leak off quickly and hydraulic fractures
criteria, as does the new equilibrium step- close quickly.
rate (ESR) test (above).5 The ESR test is similar The closure pressure is unique and easy to
to a conventional step-rate test with one exception. identify, which avoids ambiguities associated
This procedure decreases hydraulic fracture with other methods. Fluid efficiency can also be
surfaces to an equilibrium area where injection estimated from the decline slope. A field-derived
rate equals leakoff rate to provide a more correlation for deepwater wells provides a
reliable indication of fracture-closure pressure. reliable way to estimate in-situ stress.
Fluid is injected at increasing rates to create a In the past, a common misconception was
hydraulic fracture; then, rather than shutting in that higher frac-pack productivities resulted
the well, the fluid-injection rate is reduced to from placing larger volumes of proppant in a for-
the estimated fracture-propagation, or exten- mation. Skin data from frac-pack completions in
sion, rate and then held constant. the Gulf of Mexico plotted as a function of
The volume and pressure in the dynamic
5. Weng X, Pandey V and Nolte KG: “Equilibrium Test – A
fracture subsequently decrease until the fluid Method for Closure Pressure Determination,” paper
leakoff and injection rates reach equilibrium. At SPE/ISRM 78173, presented at the SPE/ISRM Rock
Mechanics Conference, Irving, Texas, October 20–23, 2002.
that point, fracture volume stops decreasing,

Autumn 2004 23
 40
Shunt
35 tubes
30
Basepipe
Dimensionless skin 25

20
15
10
5
0 Screen
-5
1 2,000 4,000 6,000
Volume of proppant placed, lbm/ft of perforated interval Nozzle

> Effective treatment designs. Comparing Gulf of Mexico frac-pack

completions indicates that placing higher volumes of proppant in the formation
does not correlate with reduced skin factors. Three wells had extremely high
skins, above 20 to more than 35, even though proppant volumes exceeded Casing
2,000 lbm/ft [3,000 kg/m] of perforated interval. In these cases, high skins were
attributed to not achieving a TSO. Shunt tubes

Perforations Screens

proppant volume per foot indicate that
increased proppant volumes do not necessarily
decrease skin if the treatment failed to achieve
a TSO (above). Another key aspect of ensuring
effective TSO treatments is complete and effec-
tive fracture coverage and proppant placement
across an entire productive interval.
Fracture Coverage and Proppant Placement
Frac-pack completion designs and downhole
equipment must address the complexities of
treating large reservoir sections and multiple
completion intervals, some with perforated
intervals longer than 150 ft [46 m] and with sig-
nificant permeability and stress contrasts. Even
carefully planned frac-packing treatments may
end in failure if proppant packs off, or bridges,
in the screen-casing annulus, restricting or
blocking annular flow. Proppant bridging may
cause early treatment termination, low fracture
conductivity and incomplete gravel packing
around screens.
Annular blockage near the top of a screen
assembly prevents fracture stimulation and
screen packing across deeper zones. Partial flow
restriction in the annulus increases frictional
pressure drop, restricts rate distribution and
limits fracture-height growth across the comple-
tion interval, especially when fracturing zones
with higher in-situ stresses. Annular voids below
a proppant bridge increase the likelihood of
subsequent screen failures caused by erosion
from produced fluids and sand.
For homogeneous intervals that are less than
60 ft [18 m] thick, fracture height typically
covers the entire zone. For longer intervals, the
probability of complete fracture coverage
decreases, and risk of proppant bridging
increases. Long intervals may be treated in
separate stages. This requires more downhole
equipment, such as stacked frac-packing
assemblies, and additional installation time, but
these factors are often outweighed by increased
treatment and completion effectiveness.
Alternate Path technology can also be
applied to frac pack longer intervals (right).
AllFRAC screens use hollow rectangular tubes, or
shunts, welded on the outside of screens to pro-
vide additional flow paths for slurry. Exit ports Annular
with carbide-strengthened nozzles on the shunt proppant bridge
tubes allow fluids and proppant to exit above and Annular
void Fracture
below annular restrictions, so that fracturing and
annular 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 proppant
bridges at the top of intervals or adjacent to
high-permeability zones with significant fluid
leakoff. If annular restrictions form, injection Nozzle
pressure increases and slurry diverts into the
shunt tubes, ensuring fracture coverage and
proppant packing around screens across the
entire completion interval.
> Effective treatment placement. AllFRAC
Alternate Path technology also allows screen
Alternate Path screens have shunt tubes with
manufacturers to maximize internal screen strategically located exit nozzles welded on the
diameters to reduce pressure drops through outside of screen assemblies (top and middle).
downhole assemblies and well-completion Large shunt tubes provide a flow path for slurry to
equipment. To accommodate higher injection bypass annular restrictions caused by the
formation of proppant bridges, allowing continued
rates, AllFRAC screens for frac packing have fracture stimulation of lower intervals and
shunt tubes with slightly larger cross sections complete packing of any annular voids around
than the AllPAC screens used for gravel packing. the screens that may form as a result of proppant
bridging in the screen-casing annulus (bottom).

24 Oilfield Review
 Fracture packing

11,500 50

Bottomhole pressure (BHP), psi

Proppant concentration, ppa

11,000 40
10,500

Rate, bbl/min
30
10,000
BHP Rate 20
9,500

9,000 10
Proppant
8,500 0
0 10 20 30 40 50 60
Time. min
Casing Screens Annular packing

6,000 50

Wellhead pressure (WHP), psi

Proppant concentration, ppa

40
5,000

Rate, bbl/min
WHP
30
4,000
20
3,000 Rate
10
Proppant
2,000 0
0 10 20 30 40 50 60
Time, min
> Optimized frac packing. Bottomhole treating-pressure response for an effective TSO treatment demonstrates a significant increase in net pressure
associated 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).

When these frac-pack design considerations 0 10 20 miles

are carefully addressed, operators should expect 0 15 30 km
desirable downhole and surface pressure
USA
signatures, indicating an effective TSO fracture
and complete packing of the screen-casing Canyon Express Project
annulus (above). Total and Marathon Oil Com-
pany successfully implemented optimized Gulf of Mexico
frac-packing techniques in two ultradeepwater
Gulf of Mexico gas fields. Destin Dome

Deepwater Proving Ground

Canyon Station
Total E&P USA, Inc. operates the Aconcagua
0
field 140 miles [225 km] southeast of New 1,00
Orleans in Mississippi Canyon Block 305. The 0
Camden Hills field operated by Marathon Oil
Viosca Knoll 3,00 Desoto Canyon
Company lies in adjacent Mississippi Canyon 0
h, f
t 5,00
Block 348. These two fields, along with the dept 7,00
0 King‘s Peak field
BP-operated King’s Peak field, comprise the ter Aconcagua field
Wa Mississippi
Canyon Express development (right).6 Canyon Camden Hills field

Camden Hills field–two wells operated by Marathon, 7,200-ft water depth

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 Expectations
from 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.
Aconcagua field–four wells operated by Total, 7,100-ft water depth
King’s Peak field–three wells operated by BP, 6,200-ft water depth
> 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 in
Gulf 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 to
the Canyon Station host platform located 56 miles [90 km] away in shallower outer continental shelf
water. Before agreeing to a shared gathering and processing system, Total, Marathon and BP
evaluated other options, such as spars and other stand-alone facilities. The complexity of subsea
operations and limited reserves made it uneconomic to develop these fields independently.

Autumn 2004 25
 Tubing hanger

4 1⁄2 -in. production

tubing

Methanol-injection
mandrel
TRC-DH-10-LO
safety valve
9 5⁄8 -in. liner top
Chemical-injection
mandrel
Packer-setting device
Packer-setting
mechanical override
Splice sub
9 5⁄8 -in. Production packer
liner Gauge carrier with
top three pressure and
temperature sensors
Cross-nipple for upper
zone isolation
Upper flow-control valve
Lower flow-control valve
Landing nipple
Wireline reentry guide
7-in. shroud
Landing nipple for
lower zone isolation
3 1⁄2-in. isolation tubing
QUANTUM isolation packer
Production-seal assembly
AFIV device
QUANTUM maX packer
Upper Mechanical FIV device
interval
2 7⁄8-in. tubing with
carbide blast rings
AllFRAC screens (Aconcagua field)
Weatherford prepacked screens
(Camden Hills field)
Shifting tool
Lower QUANTUM maX packer
interval Hydraulic/mechanical FIV device
AllFRAC screens (Aconcagua field)
Weatherford prepacked screens
(Camden Hills field)
Sump packer

> Typical Canyon Express well configuration. Aconcagua field wells

incorporated state-of-the-art Alternate Path sand-control screens because of
long completion intervals. Camden Hills field wells used Weatherford
prepacked screens without shunt tubes because of shorter completion
intervals. The sump packer, lower sand-control assembly, upper sand-control
assembly and isolation packer were installed in four separate runs. The upper
completion equipment above the production-seal assembly was installed in a
single operation. A SenTREE 7 subsea well control system installed in the
upper completion performed 24 functions, including automated flow control,
permanent monitoring and emergency well shut-in with a 15-second
disconnect at any water depth.

The four wells in the Aconcagua field, two
wells in the Camden Hills field and three wells
in the King’s Peak field are located in 6,200 to
7,200 ft [1,890 to 2,195 m] of water. These sub-
sea wells produce into a common multiphase
gathering system. Dual pipelines then transport
produced gas to Canyon Station, a processing
platform operated by Williams Energy Services,
about 56 miles [90 km] away in Main Pass
Block 261. Production is selectively controlled
using 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 multiple
sands separated by shale layers. Production
history from these types of reservoirs indicates
that gas output drops quickly once water
production begins.

26 Oilfield Review
 Completing and draining multiple sands and
having the capability of controlling water
production without conventional remedial rig
intervention were critical elements in the plan-
ning and design of these wells. Well-completion
equipment consisted of a sump, or bottom,
packer; AllFRAC screen assemblies and an isola-
tion packer with tubing or annular isolation
devices for each completion interval; and an
isolation packer and upper completion equip-
ment (previous page).
Total and Marathon completed a total of 13
perforated intervals in six subsea wells, four in
the Aconcagua field and two in the Camden Hills
field. Each well had at least two completion inter-
vals that were frac packed with the objective of
achieving completions skins of less than five.8
Well operations, which were suspended after
batch drilling all the wells, resumed by reenter-
ing the wells to drill out temporary cement plugs
and replace mud with less damaging completion
fluids. Surfactant pills were circulated to remove
remaining mud deposits. Casing brushes, 360°
scrapers, screen-type junk baskets and magnets
removed any other debris. The subsea riser and
cavities in the blowout preventer (BOP) stack
were cleaned with brush and jetting tools.
Following these cleanout procedures, the
wellbores were displaced with seawater followed
100
80
Number of points = 95
Average skin = 10.3
Dimensionless skin
60
40
20
0 DataFRAC analysis that included ESR tests
-20
1 100
Permeability-height (kh), mD-ft
> Frac-pack productivity. Gulf of Mexico pressure-buildup data indicate that
completion skin is higher than expected in many frac-pack completions,
leaving room for further optimization. In this evaluation, skins increased with
increasing permeability-height (kh), averaging 10.3 for these 95 treatments.
Autumn 2004
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 ultrasonic
cement bond log with a gamma ray correlation
log and casing collar locator on the Total wells.
Based on previous results, the Marathon well-
completion program did not include a cement
bond log. A lower sump packer run on wireline
and set below the deepest perforations provided
a reference depth for subsequent operations in
all the wells.
Productive intervals were perforated with a
400- to 600-psi [2.8- to 4.1-MPa] overbalance
using tubing-conveyed perforating (TCP)
techniques. The relatively simple TCP gun
assemblies consisted of a locating snap-latch,
gun sections with charges loaded at 12 shots per
foot (spf) and 120° or 60° phasing, a pressure-
disk to hold fluid in the tubing and a hydraulic
firing head with backup drop bar. One of the
Camden Hills field wells used 18 spf. The wells
were not flowed after perforating. This method
proved to be simple, reliable and relatively low-
risk compared to underbalanced perforating in
unconsolidated sands.
Perforated intervals in the Aconcagua field
ranged from 35 to 111 ft [11 to 34 m] in length
with a true vertical depth (TVD) of about
12,700 ft [3,870 m]. Average reservoir pressure
10,000 1,000,000
was 6,300 psi [43.4 MPa], and bottomhole static
temperature (BHST) was 128°F [53°C]. Two
zones had perforated interval lengths greater
than 100 ft [30 m] with high inclination angles
of 30° to 53°.
Productive zones in the Camden Hills field
were close to a water contact, so fracture-height
growth was a concern. Perforated-interval
lengths varied from 46 to 65 ft, [14 to 20 m] at a
TVD of about 14,000 ft [4,267 m]. Reservoir
pressure was 7,065 psi [48.7 MPa] and BHST was
155°F [68°C].
Because of high wellbore inclination angles
in the Aconcagua field, Total selected wire-
wrapped Alternate Path AllFRAC screens with
shunt tubes to obtain uniform fracture stimula-
tion and complete packing of the screen-casing
annulus across longer completion intervals.
Marathon chose prepacked Weatherford screens
with 20/40 mesh resin-coated proppant for the
Camden Hills field where shorter completion
intervals did not require the use of Alternate
Path technology.
After perforating, the sand-control screen
assembly for the lower completion interval,
including an FIV Formation Isolation Valve tool,
was run. Bottomhole temperature and pressure
gauges were run to evaluate treatment placement.
In some cases, the average skin for conven-
tional frac-packs in the Gulf of Mexico is greater
than 10 (left). Total and Marathon applied opti-
mized frac-packing techniques with the
objectives of reducing completion skin and
depleting these smaller fields more effectively
without future remedial intervention.
Prior to frac-packing operations, Total and
Marathon pumped 50 gal/ft [0.6 m3/m] of 10%
hydrochloric acid [HCl] to remove perforation
damage. Treatment fluids were selected based
on cool-down temperatures of 87 to 95°F [31 to
35°C] using the CoolFRAC optimized fracturing
service for high-permeability frac packs.
Before the main treatments, Total performed
using 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 and
Ferroni 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-Pack
and Intelligent Well Completion Systems,” paper OTC
15094, presented at the Offshore Technology Conference,
Houston, May 5–8, 2003.
27
 > Frac-packing Aconcagua field wells. The Schlumberger DeepSTIM I and DeepSTIM II stimulation
vessels performed frac-pack combined stimulation and gravel packing for Total while moored beside
the Transocean Discoverer Spirit drillship during the completion of subsea wells in Mississippi Canyon
Block 305.
thousand (ppt) crosslinked fluid. Marathon
eliminated DataFRAC tests from Camden Hills
field frac-pack procedures to avoid promoting
fracture growth into nearby water sands.
Frac-packing treatments were pumped from
the Schlumberger DeepSTIM I and DeepSTIM II
stimulation vessels, which were moored along-
side the Transocean Discoverer Spirit drillship
(above). The main frac-packing treatments were
performed at 20 to 30 bbl/min [3.2 to
4.8 m 3 /min] using 20/40 man-made ceramic
proppant and a 20-ppt crosslinked fluid.
After the operator encountered difficulty
achieving a complete annular pack in the first
Aconcagua well, perforation intervals were
reduced to limit fracture growth into bounding
shale layers and provide for continued fluid
leakoff. Subsequently, a 120-bbl [19-m3] slurry
stage with 8 to 10 pounds of proppant added
(ppa) per thousand gallons was included in the
pumping schedule to allow for controlled reduc-
tion in injection rates at the end of a treatment.
These steps ensured complete gravel packing on
subsequent jobs without pumping separate
treatments to top-off and pack the screen top.
28
In the Camden Hills field, Marathon and
Schlumberger specialists designed unconven-
tional frac packs using a less efficient
crosslinked fluid to control fracture-height
growth through excessive leakoff and to prevent
fracture propagation into wet sands. Design frac-
ture lengths were 20 to 30 ft [6.1 to 9.1 m] with a
1-in. propped fracture width. For the Aconcagua
field completions, Total increased the polymer
concentration slightly and designed fracture
lengths of 40 to 50 ft [12.2 to 15.2 m].
Following frac-packing operations on the
lower completion interval in each well, an isola-
tion plug was set in the lower frac-pack packer.
Sand or calcium carbonate pills were spotted on
top of this plug to facilitate cleaning out debris
after perforating the upper completion interval.
The operators perforated the upper completion
intervals, retrieved the packer isolation plug and
ran sand-control equipment with an FIV tool
before frac packing.
A seal assembly below the isolation packer
connected into a polished-bore receptacle (PBR)
at the top of the lower screen assembly. An AFIV
annular-controlled FIV system in the isolation
packer assembly provided flow isolation for the
upper and lower completion intervals.
Zonal Isolation and Fluid-Loss Control
Canyon Express well designs integrated an FIV
system with QUANTUM gravel-pack packer
equipment for lower completion intervals. The
sand-control equipment for upper intervals
included an AFIV system. These devices assure a
high level of well control without expensive and
risky well-intervention operations. The FIV and
AFIV systems have a proven record of providing
safe, reliable zonal isolation in various drillstem
testing and downhole tool applications.
These valves allow independent, two-way
isolation and control of each interval to prevent
fluid losses and gas influx during completion and
flowback operations. The FIV and AFIV valves
also facilitated pressure-integrity tests without
wireline or coiled tubing intervention before the
wells were opened for production.
In each of these Canyon Express wells, a
shifting tool below the internal washpipe passed
back through the screens as the gravel-packing
workstring tubing was retrieved. This tool
shifted a sleeve that closed the respective valves
to isolate each completion interval after
frac-packing treatments. The FIV and AFIV
valves could also be opened with a similar shift-
ing tool run on wireline or coiled tubing, and the
FIV ball valve could be milled through tubing as
a contingency.
Oilfield Review
 8,500 110 Fracture-closure pressures from conven-
8,400 tional injection and minifracture tests were too
Slope change ambiguous for critical ultradeepwater well
8,300
completions. More reliable ESR analysis ensured

Bottomhole pressure (BHP), psi

100
8,200 optimal treatment designs and execution in the
Middle temperature gauge
Aconcagua field.

Temperature, °F
8,100
Top temperature gauge Total performed Aconcagua field frac packs
8,000 90 in circulating position and tracked bottomhole
7,900 pressure (BHP) in real time by monitoring the
tubing-casing annulus during the jobs performed
7,800
Sharp pressure increase 80 in this field. Net-pressure gains of between 300
7,700 and 1,100 psi [2.1 and 7.6 MPa], indicating effec-
Bottom temperature gauge
7,600 tive TSO results, were observed while pumping
these optimized frac-packing treatments.
7,500 70
25 30 35 40 Marathon performed Camden Hills field frac
Time, min packs in squeeze position and did not monitor
> Effective stimulation and packing. Data from permanent bottomhole gauges BHP using a live annulus.
show pressure responses associated with changes in temperature that Calculated completion skins ranged from
are indicative of slurry diversion through Alternate Path shunt tubes during –1.5 to 4 with an average of 3.06 for 13 intervals
frac packing.
in the six wells in Aconcagua and Camden Hills
fields, much better than the previous average of
10.3 (below left). Production began just after
the sixth well was completed. Effective TSO
fracture treatments optimized production from
10
frac-pack completions in high-permeability,
unconsolidated formations. Each of the wells is
capable of producing more than the target rate
of 50 MMcf/D per well [1.4 million m3/d].
Dimensionless skin

Gulf of Mexico frac-packing trend Successful TSO treatments in the ultra-

0
-10
10
> Optimal frac-pack results. Optimized techniques resulted in an average
frac-pack skin of 3.06 for intervals completed in six subsea wells of the
Aconcagua and Camden Hills fields.
Installing the isolation packer assembly after
frac packing the upper interval mechanically
shifted the upper AFIV open. A series of specific
pressure cycles applied to the production tubing
hydraulically opened the lower FIV device. This
allowed the lower interval to be produced with-
out intervention after the production packer and
upper completion equipment were in place.
100 1,000 10,000 100,000
Permeability-height (kh), mD-ft
Reduced Completion Skins
During frac-packing operations, downhole
gauges recorded and transmitted pressure and
temperature data to surface. There was evidence
of treatment fluids bypassing localized bridges
through the shunts. There were also changes in
the pressure-curve slope associated with varia-
tions in temperature, which indicated diversion
through the shunts (top).
deepwater Aconcagua and Camden Hills gas
fields were achieved through improved fluid
designs and frac-packing procedures. Treat-
ments designed to achieve TSO fractures took
precedence over larger proppant volumes
because 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 rates
inside the hydraulic fracture. In some cases,
analysis of pretreatment injection-calibration
tests 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 to
ensure clean perforations. Alternate Path
screens with shunt tubes facilitated treatment
diversion in these multizone reservoirs with long
completion intervals.
As in many endeavors, stimulation and gravel-
packing improvements have evolved from a better
understanding of basic principles and the refin-
ing of existing technologies and practices. Tested
in this harsh ultradeepwater environment, these
optimized techniques can be applied in other
areas to ensure frac-packing success. —MET

Autumn 2004 29