SPE 84263
Optimization of Frac/Pack Completions Based on Field Experience
R.H. Morales, Schlumberger; J. Profinet, J. Piedras, TOTAL E&P, USA Inc.; B. Gadiyar, S. Harris, Schlumberger
Copyright 2003, Society of Petroleum Engineers Inc.
This paper was prepared for presentation at the SPE Annual Technical Conference and
Exhibition held in Denver, Colorado, U.S.A., 5 – 8 October 2003.
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers is
prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300
words; illustrations may not be copied. The abstract must contain conspicuous
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
Abstract
As the frac/pack technology continues to evolve, there are a
number of topics that need more understanding and
improvement. Some of these are: The estimation of correct
closure pressure, which in turn leads to evaluation of the
correct fluid efficiency, and ultimately leads to the design of
the slurry schedule. In addition, a proper fluid selection
criterion is critical to the placement of highly conductive
propped fracture wings via tip screenout.
This paper describes the long strides that have been taken to
address these areas. Special attention is given to skin data
from build-up tests showing skin increasing with increase in
kh. In essence the paper shows effective ways to lower the
skin, and consequently increase production. This goal is
accomplished by adjusting the design of the frac/pack slurry
schedule to observation from field data and to numerical
modeling of the fracturing process, and also, by considering
the issues stated above. As a consequence, the paper shows
that important steps to minimize the skin are to:
• Design the fracturing fluids to suit the character of
the fractures in soft formations where cooldown,
fluid velocities, and fluid exposure time
are important.
• Use equilibrium method to fine-tune the closure
pressure (and fluid efficiency) evaluation.
• Clean and pack the perforations not aligned with
the fracture.
• Use diversion techniques to pack the
entire reservoir.
• Evaluate the packing efficiency with bottomhole
gauge data (temperature and pressure) and/or
radioactive tracers.
When these steps are meticulously followed, fracture
performances with lower skins can be obtained as
demonstrated in the case history.
Introduction
After more than one decade, the frac-packing technology has
gradually established as one of the most widely used
completion techniques for high permeability formations.1-3 In
many fronts it is becoming a mature industry where, across the
borders, service companies have comparable state of the art
vessels, downhole tools, equipment, and fluids technology;
and in other fronts, such as, intelligent completions, multiple
zone production diagnosis, and damage repair, technology is
continuously improving. Even with all these advancements,
there is still room for improvements to achieve maximum well
potential. For example, average production skins are higher
than desired (Figure 1). Like in the sports arena, excellence
comes through refining and perfecting the basic game
principles; the core of this paper is to refine the use of existing
technology with the aim of optimizing production.
With this goal in mind, we present principles for refining the
estimation of closure pressures, fluid efficiency, and
perforation clean up. We also summarized some lessons
learned that could prevent costly interventions.
In-situ stresses
The estimation of the correct in-situ stress is important
because it leads to the correct evaluation of other parameters
that are critical to the design of the slurry schedule for a tip-
screenout (TSO) treatment and prediction of fracture
geometry. An inflection point on the pressure decline plot of a
calibration test is a common method used to measure in-situ
stress. However, in many cases the pressure decline plot
shows inflection points not related to in-situ stresses but to
other mechanisms like change in flow regime or gas kicks that
could lead to the wrong estimate. In this section, we present
application of a new method that facilitates the closure pick,
known as the Equilibrium Step Rate Test (ESR). In addition,
we present a field-derived correlation for deepwater wells that
provides a reliable way to estimate in-situ stress.
Equilibrium Step Rate Test. This test is a modification of
the equilibrium test presented in SPE paper 78173.4 The
objective is the same – to facilitate the closure pick by
decreasing the fracture area to an equilibrium area in which
the injection rate equals the leakoff rate. That area is attained
when the treating pressures become stable after the initial rate
2 SPE 84263
has been decreased to the propagation rate. In the original
paper the propagation rate is assumed to be a fraction (0.2) of
the initial injection rate, while the equilibrium step-rate-test
allows estimating the actual propagation rate on the fly whilst
stepping up the rate.
The following example shows better depiction of closure
pressure in the plot of the square root of time vs. pressure
decline of an ESR compared to that of a calibration test. In
fact, the first inflection point (Point A) in the ESR decline plot
(Figure 2) corresponds to the closure pressure. That closure
(Point B) can scarcely be seen in the pressure decline plot of
the calibration treatment (Figure 3). Linear gel was used for
ESR, whereas crosslinked fluid was pumped during the
calibration test. Using an incorrect closure pressure from the
calibration treatment would have resulted in an incorrect fluid
efficiency. This is just one example of scores of tests in which
the ESR has helped correctly define the closure. The
evaluation of the fracturing fluid efficiency (which is a direct
function of the closure pressure) from the subsequent
calibration treatment becomes more reliable.
Figure 4 shows plot of the ESR rates and pressures. The
figure shows that the propagation rate is estimated from the
inflection point in the stepping up pressure plot (Point C). The
figure also shows that at the end the test instead of shutting
down pumps, as is conventionally done in a step-rate test, the
rate is decreased to the propagation rate and pumping is
continued until the pressure becomes stable (Point D).
Stress Prediction Correlation. For many logistical reasons it
is important to have an estimate of the closure pressure prior
to pumping the injection tests. As explained in detail in
Appendix A, the following correlation uses the classical
Hubbert and Willis equation for the estimation of closure
pressure for reservoirs with conditions similar to those
encountered in the Gulf of Mexico where there is no
significant tectonic activity, the overburden is the principal
stress, and the ratio between the maximum and minimum
horizontal stresses is close to one.
σ h min = A(TVD) + B( Pp ) − C (Wdepth ) Eq. (1)
Where TVD is true vertical depth (ft), Pp is pore pressure (psi)
and Wdepth is water depth (ft). The values of the constants A,
B, and C are 0.342, 0.658, and 0.190, respectively. The
correlation was derived from a database of 30 fracturing
injection tests in reservoirs with water depths raging from
1300 ft to 7200 ft. The error between the predicted and
measured stresses was minimized using the Poisson’s ratio as
the calibration variable. The Poisson’s ratio resulting in the
least error was 0.255 and the average error between the
predicted and measured stresses was 2.8% (Figure 5).
Fluids
Fluid plays a major role in the success of frac-pack treatment.
It facilitates: 1) creating the fracture geometry, 2) proppant
transport as well as placement, and 3) most importantly
initiating tip-screenout (TSO). The most commonly used fluid
in the Gulf of Mexico is borate crosslinked guar. This is a
non-Newtonian fluid that exhibits shear-thinning behavior.
During treatment execution, the temperature in the wellbore
and fracture will cool down significantly due to leak-off. The
cool down temperature increases the fluid apparent viscosity.
Simultaneously, as the fracture propagates, the fluid
continuously experiences change in shear rate and
consequently the apparent viscosity changes too. At the
initiation of the fracture, the shear rate could be above 100 sec-
1
but after the TSO event, it could drastically decrease by
orders of magnitude. These changes could hinder TSO as
typically observed with higher than required polymer loading.
5
Thus the fluid selection (polymer loading) should be based
on cool down temperature and shear rate. 5
Production contribution of perforations
Previous modeling6 indicates that the production from
perforations not aligned with the fracture plane can be
significant in high permeability reservoirs (Figure 6). This
figure demonstrates two limiting cases – low perforation
density (4 spf) and infinite perforation density (open hole)
having a skin of zero. This phenomenon could explain why
Gulf of Mexico data shows that skin increases for high
kh reservoirs.
Necessity of Acid Cleanup. On most occasions, the
perforation tunnels incur damage during perforating.
Therefore, it is important to pump and effectively divert an
acid cleanup treatment prior to frac-pack with the objective of
removing perforation damage. To demonstrate the effect of
acid, skin data from build-up tests was plotted as a function of
acid volume pumped as shown in Figure 7. It is evident that
the skin was high when less than 20 gal/ft of acid was
pumped. Consequently, to minimize skin it is recommended to
pump 40 to 50 gal of acid per foot of perforated interval with
some means of diversion in long interval wells.
Proppant Placement
The perception in the industry is that wells produce better
when more proppant is placed behind perforations. To
investigate this, skin data as a function of proppant volume per
foot (Figure 8) was plotted. This figure indicates that skin did
not decrease with increase in proppant. In fact in four cases,
skin was very high even though the proppant exceeded 2000
lbs per foot of perforated interval. This is attributed to not
achieving a TSO.
Case History
The application of above techniques is demonstrated in the
economic success of the ultra-deepwater Aconcagua and
Camden Hills gas fields where the goal was to achieve
completions with low skins of less than 5.7 These fields are
located in Gulf of Mexico, Mississippi Canyon blocks 305 and
348, at water depth of 7000 ft to 7200 ft. There were total of 6
wells with 13 perforated zones completions. All these zones
were frac-packed.
The zones of interest in Camden Hills were close to water
contact.8 Hence fracture height growth was a concern. The
SPE 84263
perforated interval height varied from 46 to 65 ft, and TVD
was approximately 14,000 ft. The reservoir pressure was 7065
psi and Bottom Hole Static Temperature (BHST) was 155O F.
In the Aconcagua field, the perforated interval height varied
from 35 to 111 ft, and TVD was 12,700 ft. The average
reservoir pressure was 6300 psi and BHST was 128O F.
In both fields, the formations were over-balance perforated
(450 to 600 psi) with Tubing Conveyed Perforating (TCP)
guns. The shot density was 18 shot per foot (spf) at 120/60O
phasing. The wells were not flowed after perforating. This
method was very simple, reliable, and relatively low-risk
compared to under-balanced perforating in
unconsolidated sands.
In the case of Aconcagua field, two of the zones had long
perforated intervals (> 100 ft) with high deviation angles (30-
53O). In addition, most of these reservoirs consisted of
multiple sands separated by shale layers. Beacuse of the
inclination and long intervals it was decided to use shunt tubes
to allow uniform annular packing and bypass possible
localized bridging. Bottom hole temperature and pressure
gages were used for post job evaluation of fracture
height coverage.
Frac-pack fluid selection was based on cool down temperature
of 87 to 95O F. In the case of Camden Hills, instead of
pumping a gravel pack or high-rate-water-pack as is typically
done to prevent fracture propagation into wet sand, we
designed unconventional frac-packs using an inefficient
crosslinked fluid to control height growth through excessive
leak-off. The designed fracture lengths were 20-30 ft and 1
inch propped width. For the Aconcagua completions, we
slightly increased the gel concentration and designed fracture
lengths were 40-50 ft. During pumping operations,
bottomhole pressure was monitored real time via live annulus.
The sequence of steps during job execution were:
• Pumped 50 gal/ft of 10% HCl to
cleanup perforations.
• ESR test using a linear gel.
• Calibration treatment with crosslinked gel.
• Frac-pack with rates ranging from 20-30 bpm.
• Annular pack was achieved by controlled rate slow
down procedure.
Results. While pumping the frac-pack treatments, we
observed net pressure gains from 300-1100 psi. The
temperature gages showed that slurry covered the entire
interval. The closure pressure estimated from ESR test
assured the correct closure estimation that in some cases was
ambiguous from calibration test. The leak-off coefficient and
proppant placement ranged from 0.01-0.032 ft/min1/2 and 740-
1540 lb/ft, respectively. The skins calculated from injection
tests ranged from –1.5 to 4.
There was evidence of bypassing localized bridges through the
shunts as illustrated in Figure 9 where changes in pressure
3
slope associated with changes in temperature slope indicate
diversion through the shunts.
Lessons Learned and recommendations. Key lessons
learned are listed below:
• Achieving the annular pack was challenging in the
first Aconcagua well possibly due to fractures
growing into shales that remained open due to lack of
leak-off. This resulted in pumping a slurry top-off
treatment and a stuck cross over tool. To prevent this
from happening again, in addition to the designed
fracturing slurry volume, we allocated 120 bbl (8-10
ppa) of slurry volume for the slow down stage. This
proved to be successful.
• The overbalanced perforating worked well since it
controlled fluid loss and prevented the spotting of
damaging loss circulation pill.
• The closure pressure estimated from ESR test assured
the correct closure estimation that in some cases was
ambiguous in the calibration test.
• Designing the fracturing fluid based on cool down
temperature and fracture shear rates was successful in
achieving TSO.
• Using extremely low polymer concentration cross-
linked fluid achieved the goal to control height
growth via excessive leak-off.
• Shunt tubes demonstrated effectiveness in bypassing
localized bridging.
Conclusions
Based on the content of this paper following conclusions
are made:
• Production of high permeability and unconsolidated
formations can be optimized only if TSO can be
achieved during frac-pack completion process.
• Fluid should be designed based on cooldown
temperature and shear rate in the fracture.
• Prior to executing frac-pack, sufficient quantity of
acid should be pumped to clean perforations.
• ESR is a simple and reliable technique to determine
closure pressure.
• Shunts help in diversion in multi-zone reservoirs
and/or long interval completions.
• Proppant placement volume does not impact
production as long as TSO is achieved.
Acknowledgment
The authors would like to thank the management of
Schlumberger and Total E&P, USA Inc. for granting the
permission to publish this paper. We extend our appreciation
to Tricia O'Flaherty Lespreance for helping with
data compilation.
References
1) Tiner, R.L., Ely, J.W., and Schraufnagel, R.: “Frac
Packs – State of the Art,” Paper SPE 36456 Presented
at SPE ATCE, Denver, CO, October 6-9, 1996.
2) Smith, M.B. and Hannah, R.R.: “High-Perm
Fracturing: The Evolution of a Technology,” Paper
4 SPE 84263

SPE 27984 Presented at University of Tulsa ν (σ ov − p p )

Centennial Petroleum Engineering Symposium, σ h min = + pp Eq.
Tulsa, OK, August 29-31, 1994. 1 −ν
3) Stewart, B.R., Mullen, M.E., Brown, J.E., and (A2)
Norman, W.D.: “Step-Rate, Calibration Injection and
Treating Pressure Anomalies in Soft Rock High Where ν is the Poisson’s ratio, pp is the pore pressure; and the
Permeability formations: An Explanation Based on overburden stress, σov is function of the water depth (Wdepth),
Bottom Hole Pressure and Production Results,” Paper true vertical depth (TVD), water gradient (gw =0.444 psi/ft),
SPE 29444 Presented at Production Operations and overburden gradient (gf = 1 psi/ft),
Symposium, Oklahoma City, OK, April 2-4, 1995.
4) Weng, X., Pandey, V., and Nolte, K.G.: “Equilibrium σ ov = Wdepth g w + (TVD − Wdepth ) g f Eq.
Test - A Method for Closure Pressure
Determination,” Paper SPE 78173 Presented at (A3)
SPE/ISRM Rock Mechanics Conference, Irving, TX,
October 20-23, 2002. With the exception of ν, all the parameters described above
5) Morales, R.H., Gadiyar, B.R., Bowman, M.D., were measured from downhole gauges or were known by
Wallace, C, and Norman, W.D.: “Fluid other means. The error between the σmin predicted by
Characterization for Placing an Effective Frac/Pack,” Hubbert and Willis equation and the σmin measured was
Paper SPE 71658 Presented at SPE ATCE, New minimized using ν as the calibration parameter. The ν
Orleans, LA, September 30 – October 3, 2001. resulting in the least error (2.8 %) was 0.255 (Figure 5). Thus,
6) Guinot, F., Zhao, J., James, S., and d'Huteau, E.: the constants A, B, and C (0.342, 0.658, and 0.190) were
“Screenless Completions: The Development, obtained by using a ν of 0.255 in Eq. A2.
Application and Field Validation of a Simplified
Model for Improved Reliability of Fracturing for
Sand Control Treatments,” Paper SPE 68934
Presented at SPE European Formation Damage
Conference, The Hague, The Netherlands, May 21-
22, 2001.
7) Piedras, J., Stimatz, G.P., Jackson Nelson, V.B., and
Watson, G.M.: “Canyon Express: Design and
Experience on High-Rate Deepwater Gas Producers
Using Frac-Pack and Inteligent Well Completion
Systems,” Paper OTC 15094 Presented at 2003 OTC,
Houston, TX, May 5-8, 2003.
8) Hart’s E&P Supplement: Camden Hills: A World
Record Achieved Through Innovative Solutions,
April 2003.

Appendix A
A set of data consisting of 30 in-situ minifrac tests were used
to derive the following correlation,

σ h min = A(TVD) + B( Pp ) − C (Wdepth ) Eq. (A1)

The TVD depths for the reservoirs ranged between 10000 ft to

20000 ft and the water depths ranged from 1300 ft to 7200 ft.
Furthermore, the basis for the correlation was the Hubbert and
Willis equation:
SPE 84263 5

100
Number of Points: 95
80 Ave. Skin: 10.30

60
Skin
40
20

0
-20
1 100 10000 1000000

kh (md)

Figure 1: Gulf of Mexico Build-up data showing skin vs. kh

7150
7100 ESR Test
7050
A
BHP (psi)-

7000
6950
6900
6850
6800
6750
5.6 5.8 6 6.2 6.4
1/2
Sqrt (Total Time) (min )

Figure 2: Equilibrium Step Rate test pressure decline

6 SPE 84263

7150
7100
Calibration
7050 B
BHP (psi)-
7000
6950
6900
6850
6800
6750
2 2.5 3 3.5 4
1/2
Sqrt (Total Time) (min )
Figure 3: Calibration pressure decline

4500 40
4000 35
3500 D
30
Pressure (psi)
(psi)

3000
Rate (bpm)
25
2500
20
2000
(bpm)

TP
C 15
1500
1000 10
500 R ate AP 5
0 0
260 280 300 320 340
Tim e (m in)

Figure 4: Field Example showing equilibrium step rate test and calibration test
SPE 84263 7

5.0

4.5

Average Error (%)

4.0

3.5

3.0
Error = 2.8%
2.5 ν = 0.255

2.0
0.15 0.2 0.25 0.3 0.35

Poisson's Ratio

Figure 5: Figure showing the error as function of Poisson’ s ratio for derivation of Eq. 1

100
90
FracFlow/TotalFlow (%)

80
4 SPF 90o Phase
70 Zero Skin
60
50
40
30 OH: Zero Skin
20
10
0
10 100 1000 10000
Permeability (md)

Figure 6: Contribution to production from the fracture compared to flow from non-aligned perforations
8 SPE 84263

40

30

Skin 20

10

0
0 20 40 60 80 100

Acid Volume (gal/ft)

Figure 7: Effect of pre-fracturing acid treatment on skin

40
35
30
25
20
Skin

15
10
5
0
-5
0 2000 4000 6000
Proppant Placement (lb/ft)

Figure 8: Effect of proppant placement on skin

SPE 84263 9

8500 110
8400
8300
Slope change 100
8200

Temperature (oF)
BHP (psi) 8100
Middle
8000 90
Upper
7900
7800
Lower 80
7700
Shap
Sharp Pressure
pressure
7600 increase
increase
7500 70
25 30 35 40
Time (min)

Figure 9: Bottom hole gages illustrating shunt action