Polymer Reduction Leads to Increased
Success: A Comparative Study
Natasha Kostenuk, SPE, Trican Well Service, and Pierre Gagnon, SPE, ConocoPhillips Canada
Summary
Recent advances in guar and crosslinker technologies have re-
sulted in the development of high-viscosity crosslinked borate-
fracturing fluids without increasing polymer loadings. These low
polymer (LP) borate fracturing fluids are being used successfully
in various formations previously believed to be too hot and or too
deep for LP fracturing fluids.
Historically, polymer loadings of 3.6 to 4.2 kg/m3 (30 to 35
lbm/1,000 gal) were commonly pumped in the Western Canadian
Sedimentary basin (WCSB) for formations deeper than 2500 m
and bottomhole temperatures greater than 80°C. These same
formations are now fracture stimulated using the LP fluids
with loadings as low as 1.8 kg/m3 (15 lbm/1,000 gal) with excep-
tional results.
This paper demonstrates that LP fracture fluids can be used in
place of fluids requiring higher polymer loadings with minimal
changes to the overall design of the fracture treatment. The new
fluid can be pumped on-the-fly at conventional pump rates and
proppant concentrations because of the fluid’s improved shear and
temperature stability.
The advantages of using a reduced-polymer fracturing fluid
include increased production, lower treatment costs, and lower
frictional pressure loss.
This paper illustrates these advantages as it compares the LP
fracture fluid with HP fracture fluids in more than 200 wells in
the WCSB. The formations where LP fluids were used have
depths of up to 3250 m and reservoir temperatures reaching
over 100°C.
Introduction
Low-permeability gas wells in the WCSB often require hydraulic
fracturing to be ecomonic (such fracturing treatments are consid-
ered to be a critical step in the completion process). The choice of
fracturing fluid is a major component and concern when designing
optimal fracture stimulations. The primary roles of the fracturing
fluid are to initiate and extend a hydraulic fracture and to distribute
proppant. The majority of fracture treatments use crosslinked wa-
ter-based fluids. Fluid viscosity is attained by adding some type of
guar polymer to the water. Many types of guars can be and have
been used, including derivatized guars, xanthan gum, and poly-
acrylamides. For the LP fluids discussed in this paper, a nonderiva-
tized guar crosslinked with borate was used. For the purpose of this
paper, LP fluids are described as those with polymer loadings of
1.8 to 2.4 kg/m3 (15 to 20 lbm/1,000 gal). Conversely, HP fluids
are those containing polymer loadings of 3.0 to 4.2 kg/m3 (25 to
35 lbm/1,000 gal).
Recent advances in guar processing have made the LP sys-
tems achievable. Historically, guar has been known to contain 8
to 12% residue (Gidley et al. 2001), but newer, more-processed
guars available today contain 1 to 5% residue. This enhanced guar
has a higher molecular weight than previous guars (Dawson et al.
2004), thereby increasing the yield of the polymer that, in turn,
increases viscosity. The shear and temperature stability of the
Copyright © 2008 Society of Petroleum Engineers
This paper (SPE 100467) was accepted for presentation at the 2006 SPE Gas Technology
Symposium held in Calgary, 15–17 May 2006 and revised for publication. Original manu-
script received for review 30 June 2006. Revised manuscript received 3 October 2007.
Paper peer approved 08 October 2007.
March 2008 SPE Drilling & Completion
resulting fluid is superior to other guars, leading to a wide range
of applications.
Why Reduce Polymer Loadings?
As stated previously, polymer is required to create enough viscos-
ity for the fracture fluid to initiate and to extend the fracture, as
well as properly place the proppant along the fracture. Unfortu-
nately, polymer is also known to be damaging to the proppant pack
(Kim and Losacano 1985; Roodhart et al. 1988). This damage
adversely affects proppant-pack conductivity and, therefore, pro-
duction potential. There can also be damage from filter-cake
buildup on the formation face, which can also reduce the fracture’s
effectiveness (Volk et al. 1983).
Reducing the amount of polymer in the fracturing fluid is an
obvious way to reduce polymer damage within the proppant pack.
This has been demonstrated in regain-permeability tests conducted
on samples of 1.8-kg/m3 (15 lbm/1,000 gal) polymer-loaded fluid
and 3.6-kg/m3 (30 lbm/1,000 gal) polymer-loaded fluids. The tests
were performed by an independent laboratory as part of a larger
study. The regain-permeability tests were performed using 2 lbm/
ft3 of 20/40 Northern White Sand between Ohio sandstone at 2,000
psi closure stress and at 50°C. The baseline regain conductivity test
was performed using 2% KCl water. As Fig. 1 indicates, the LP
fluid exhibits improved regain proppant-pack conductivity. The
1.8-kg/m3 polymer-loaded fluid resulted in 58% regained conduc-
tivity compared with 35% for the 3.6-kg/m3 loading.
The costs associated with fracture treatments are also reduced
when using LP fluids. Obviously, the less polymer and breaker
used, the lower the cost, but there is also a reduction in horse-
power costs. The LP fluids exhibit lower frictional pressure
losses that, in turn, reduce average pumping pressures during
the treatment.
Comparison of HP and LP Fracture Treatments
A minimum amount of polymer is required to achieve an effective
crosslinked fluid. Also, a minimum viscosity is required to meet
the design parameters and fracture geometry. This study demon-
strates that polymer loadings can be lowered without adversely
affecting the fracture treatment.
Tables 1 and 2 compare the viscosities of linear and
crosslinked fluids of different polymer loadings. A common con-
cern when reducing polymer loadings is that the reduction of vis-
cosity will cause the fracture treatments to prematurely screen out.
Field data demonstrate the opposite. When using the LP fluid, 22%
of the treatments screened out compared to 18% of the treatments
using HP fluids. LP treatments placed 88% of the proppant de-
signed, as compared to 75% for the HP treatments. Although the
number of screenouts is similar, the LP treatments placed more
proppant before screening out.
Using LP fluids reduces the average pumping pressures of the
fracture treatments. The average rates of the LP and HP treatments
were 2.9 and 3.2 m3/min, respectively; and the average pumping
pressures were 30.2 and 34.1 MPa. This equates to an average
reduction of 3.9 MPa, respectively or 11.4% reduction in treating
pressure. Lower pumping pressures translate to fewer kilowatts
required for treatments. The average kilowatts used per treatment
with LP was 1,427 and 1,635 kW for HP fluids, a 208 kW (or
12.7%) reduction.
The savings from reducing the amount of polymer used in
fracturing fluids are significant. The average cost per tonne of
55
 Fig. 1—Regain-permeability test results.
proppant placed decreased from CAD 3,150 (HP loadings) to
$2,645 (LP). This is a savings of $495/tonne placed, or a 15.7%
reduction (All dollar figures are shown in Canadian dollars).
An important trend to note is that the majority of HP treatments
were performed in the years before the introduction and subse-
quent acceptance of the LP fluids. While the use of LP systems
increased, so did the average treatment size. The increase in treat-
ment size is solely attributable to lower-quality formations recently
becoming economical with higher energy prices. This is important
for the study because the average treatment size does increase from
HP treatments to LP treatments.
Because of the large number of treatments performed in this
study, the treatment schedules are somewhat varied. A typical
treatment schedule can be seen in Table 3.
There were more than 230 treatments initially examined for this
study. Four treatments were excluded for mechanical failures that
prematurely ended the treatment. Of the remaining treatments, 165
were completed with LP fluid and 61 were treated using HP fluids.
The average depths of the wells treated with LP and HP fluids
were 2,628 and 2,513 m, respectively. The average reservoir tem-
peratures for the LP sampling was 87°C; whereas the HP sampling
averaged 83°C.
Study Area
The subject wells for this study are located in an area paralleling
the Rocky Mountains from southern Alberta to northeast British
Columbia as seen in Fig. 2. All of the sample wells are located in
the WCSB.
56
Comparison of Production Rates: HP vs. LP
Of the initial data set of 226 treatments, only those zones that were
tested after the fracture stimulation were used for the production
comparison. Only formations were used that had a conventional
post-frac single-point flow and buildup test performed. A pressure-
transient analysis was performed for each example. Therefore for-
mation-flow capacity (kh), reservoir temperature, and reservoir
pressure were recorded. Zones with abnormally high kh values
(greater than 30 md-ft) were excluded from the study to avoid
skewing results.
There were 73 LP treatments tested, with an average kh value
of 4.7 md-ft. This sampling had an average reservoir pressure of 21
305 kPa and an average final test rate of 1,140 Mcf/D. There were
36 HP treatments tested. The average kh for the HP sampling was
4.1 md-ft, with an average reservoir pressure of 19 884 kPa and an
average final test rate of 803 Mcf/D. This resulted in an average
increase in production of 337 Mcf/D or 42%, when using LP
fluids. The average amount of proppant placed in the LP treat-
ments was 25.3 tonnes, compared with an average of 19.1 tonnes
placed using the HP fluids. The average polymer loading was 1.92
kg/m3 (16 lbm/1,000 gal) for the LP wells and 3.56 kg/m3 (29.7
lbm/1,000 gal) for the HP wells.
As a further comparison of fluid effectiveness, longer term
production was collected and analyzed. A 1-month flow rate was
used and is defined as the flow rate after 1 full calendar month of
March 2008 SPE Drilling & Completion
production. The overall sampling was reduced because some of the
zones were commingled and some wells had not been tied in at the
time of writing.
There were 32 zones treated with LP fluid on production for at
least 1 month. Their average kh value was 6.0 md-ft, with a cor-
responding average reservoir pressure of 19 126 kPa.
There were 22 zones treated with HP fluids on production for
at least 1 month. Their average kh value was 5.5 md-ft, with a
corresponding average reservoir pressure of 19 424 kPa.
The average 1-month flow rate for the zones treated with LP
fluids was 1537 Mcf/D. Zones treated using HP fluids averaged
740 Mcf/D. This translates into an increase of 797 Mcf/D or 108%,
when using LP fluids. The average amount of proppant placed in
the LP treatments was 26.6 tonnes compared with 19.3 tonnes
placed for HP treatments. Fig. 2—Areas of application.
The 2-, 3-, and 6-month flow rates were also compared. Again,
these are the flow rates after 2, 3, and 6 full calendar months on
production. The number of treatments in the study set was again An interesting point to note was that the average 1-month flow
reduced because some of the wells have not been on production the rates were significantly higher than the final test rates when using
full 6 months. LP fluids compared with HP fluids.
The average 2-, 3-, and 6-month flow rates for zones treated
with LP fluids were 1,232, 1,094, and 868 Mcf/D. Zones treated
with HP fluids averaged 575, 500, and 373 Mcf/D. Increases of Gething Treatments in Townships 58 to 61,
657, 594, and 495 Mcf/D or 114, 119, and 133%, respectively Range 20 to 26 W5
were ibtained with the use of LP fluids. Fig. 3 shows the produc- There were 18 treatments that were performed on the Gething
tion comparision. formation in the area of township 58 to 61 and range 20 to 26W5.
The average amount of proppant placed in the LP treatments The Gething formation is a medium-to-fine grained sandstone
was approximately 26 tonnes, compared with 19 tonnes for ranging from 2 to 20 m in gross thickness, with net pay ranging
HP treatments. from 1 to 8 m, porosities ranging from 7 to 15%-and in-situ per-

Fig. 3—Production comparison of all wells.

March 2008 SPE Drilling & Completion 57


Fig. 4—Gething-study map.
meabilities ranging 0.1 to 5 md on the basis of post-frac pressure
transient analysis (PTA).
Fig. 4 indicates the wells used for the study. Some of the wells
on the map had 2 Gething intervals that were fracture stimulated.
Eight of these treatments used LP fluids and 10 used HP fluids.
The initial flow rates as well as 1-, 2-, 3-, and 6-month flow rates
were compared. In all cases, LP fluids performed better.
Initial flow rates were taken as the final test rate on the post-
frac single-point flow test. HP treatments tested averaged a kh of
11.3 md-ft, average reservoir pressure of 21 950 kPa and an av-
erage final test rate of 1128 Mcf/D. Treatments using LP fluids
averaged 12.4 md-ft, 24 691-kPa reservoir pressure, and 1,541
Mcf/D for a test rate. This is an increase of 413 Mcf/D (or 37%)
when using LP fluids.
The average 1-, 2-, 3-, and 6-month flow rates of the zones
treated with LP fluids were 2,025, 1,760, 1,701, and 912 Mcf/D.
Zones treated with HP fluids averaged 1,391, 1,160, 1,015, and
743 Mcf/D. These are increases of 634, 600, 686, and 169 Mcf/D
or 46, 52, 68, and 23%, respectively. The following chart
shows this production comparison. Fig. 5 shows this produc-
tion comparison.
The average amount of proppant placed was higher for the LP
treatments (27.8 tonnes) compared with 21.8 tonnes placed for HP
Fig. 5—Production comparison of Gething wells.
58
treatments. The average polymer loading was 2.2 kg/m3 (18.3
lbm/1,000 gal) for the LP wells and 3.6 kg/m3 (30 lbm/1,000 gal)
for the HP wells.
Cutbank Treatments Townships 11–15, Range
26–28 W4
There were 13 treatments performed on the Cutbank formation in
the area of townships 11 to 15 and range 26 to 28W4. The Cutbank
formation is fine-grained sandstone ranging from 1 to 10 m in
gross thickness with net pay ranging from 1 to 6 m, porosities
ranging from 8 to 14%, and in-situ permeabilities ranging from 0.2
to 10 md. Fig. 6 indicates the wells in the study.
Five of these treatments used the LP fluid. The average kh was
8.4 md-ft, and the average reservoir pressure was 12 740 kPa for
this sampling. Eight of these treatments were performed using the
HP fluids. The average kh was 7.9 md-ft and the average reservoir
pressure was 14 213 kPa for this sampling. Again, initial flow rates
as well as 1-, 2-, 3-, 4-, 6-, and 12-month flow rates were com-
pared. In all cases, the formations treated with LP fluids performed
better than those using the HP fluids.
The zones treated with HP fluid had an average final test rate
of 942 Mcf/D compared with an initial flow rate of 1,069 Mcf/D
for zones treated with LP fluid, which is an increase of 127 Mcf/D,
or 14%.
The average 1-, 2-, 3-, 4-, 6-, and 12-month flow rates of the
zones treated with LP were 915, 610, 509, 456, 389, and 320
Mcf/D, respectively. This compares to the zones treated with HP
averaging 649, 547, 472, 398, 364, and 324 Mcf/D. Therefore,
increases of 266, 63, 37, 58, 25, and -4Mcf/D or 41, 12, 8, 15, 7,
and −1%, respectively, were observed.
The average amount of proppant placed was higher in the LP
treatments (21.5 tonnes) compared with 18 tonnes placed on av-
erage when using HP fluids. The average polymer loading was
1.92 kg/m3 (16 lbm/1,000 gal) for the LP wells and 3.23 kg/m3
(26.7 lbm/1,000 gal) for the HP wells. Fig. 7 shows this produc-
tion comparison.
What Does the Future Hold?
It is possible to lower the polymer loadings further than that of this
study. Treatments have been successful with polymer loadings as
low as 1.6 kg/m3 (13.3 lbm/1,000 gal). There are practical limita-
tions to pumping gel loadings lower than 1.6 kg/m3 in conven-
tional treatments that include viscosity levels below what is needed
March 2008 SPE Drilling & Completion

Fig. 6—Cutbank-study Map.
to place proppant successfully. There are treatments using non-
crosslinked polymer fluids such as slickwater that are used for
treatments that require low viscosity and high rates. There are also
nonpolymer viscoelastic surfactants available, but they not suitable
for all applications because of the higher costs of using them.
Fig. 7—Production comparison of Cutbank Wells.
March 2008 SPE Drilling & Completion
Summary
Reducing the amount of polymer in fracturing fluid reduces the
amount of polymer damage.
• Successful placement of proppant during fracture treatments
does not appear to be affected adversely by using lower poly-
mer loadings.
• Cost per tonne of proppant placed is reduced when using LP
fluids as compared with HP fluids.
• Cleanup appears to be enhanced when using LP fracturing
fluids. An increase (cleanup) from initial to 1 month flow rate was
more pronounced on wells stimulated with the LP fluids than on
wells stimulated with HP fluids.
• LP fluids appear to increase production potential for similar
fracture-treatment sizes.
Acknowledgments
The authors would like to thank Burlington Resources Canada for
allowing us to share their data. We would also like to thank Kari
Charlton from Burlington for her help compiling the data. The
following Trican Well Service personnel also need to be acknowl-
edged for their contribution to the paper: Greg Henderson, Con-
tessa Griffith, and Jeff Peterson.
References
Dawson, J.C., Cramer, D.D., and Le, H.V. 2004. Reduced Polymer Based
Fracturing Fluid: Is Less Really More? Paper SPE 90851 presented at
the SPE Annual Technical Conference and Exhibition, Houston, 26–29
September. DOI: 10.2118/90851-MS.
Gidley, J.L., Holditch, S.A., Nierode, D.E., and Weatch, R.W. 1989. Re-
cent Advances in Hydraulic Fracturing. Monograph Series, SPE, Ri-
chardson, Texas, 12.
Kim, C.M. and Losacano, J.A. 1985. Fracture Conductivity Damage Due
to Crosslinked Gel Residue and Closure Stress on Propped 20/40 Mesh
Sand. Paper SPE 14436 presented at the SPE Annual Technical Con-
ference and Exhibition, Las Vegas, Nevada, 22–26 September. DOI:
10.2118/14436-MS.
McDonald, C.E. 2005. The Application of a Reduced Polymer Borate
System. Paper SPE 94202 presented at the SPE Production Operations
Symposium, Oklahoma City, Oklahoma, 16–19 April. DOI: 10.2118/
94202-MS.
Roodhart, L.P., Kulper, T.O.H., and Davis, D.R. 1988. Proppant Pack and
Formation Impairment During Gas-Well Hydraulic Fracturing. SPEPE
3 (4): 438–444. SPE-15629-PA. DOI: 10.2118/15629-PA.
Volk, L.J., Gall, B.L., Raible, C.J., and Carroll, H.B. 1983. A Method for
Evaluation of Formation Damage Due to Fracturing Fluids. Paper SPE
11638 presented at the SPE/DOE Low Permeability Gas Reservoirs
Symposium, Denver, 14–16 March. DOI: 10.2118/11638-MS.
59
 Natasha Kostenuk has been working with Trican Well Ser-
SI Metric Conversion Factors vice for 8 years and is currently a technical specialist in Frac-
turing. Kostenuk graduated from the University of Calgary with
°F (°F−32)/1.8 ⳱ °C
a mechanical engineering degree and petroleum minor.
ft × 3.048* E−01 ⳱ m Pierre Gagnon is currently the manager of completions and
psi × 6.894 757 E+00 ⳱ kPa well servicing for ConocoPhillips Canada. Gagnon has 25
lbm/ft2 × 4.882 428 E+00 ⳱ kg/m2 years of industry experience in the Western Canadian Sedi-
mentary Basin. He has been with ConocoPhillips and its pre-
m3/h × 2.271 247 E−01 ⳱ U.S. gal/min decessor companies for 20 years in the completions field.
ton (metric) × 1.0* E+00 ⳱ Mg Before this, he was employed by a major pumping service
*Conversion factor is exact.
company for 5 years performing both fracture stimulations
and cementing.

60 March 2008 SPE Drilling & Completion