Novel Insights Into Microgel Systems
for Water Control
Céline Cozic, David Rousseau, and René Tabary, IFP
Summary
The performances of water-shutoff and conformance-control micro-
gels were investigated extensively in previous works (Chauveteau
et al. 2004; Rousseau et al. 2005; Zaitoun et al. 2007). However,
a broad range of potential application cases for microgels involves
harsh conditions such as high-salinity reservoir brines. This paper
studies how high salinity impacts microgel properties and treat-
ment efficiency.
In order to provide specifications for optimum microgel dissolu-
tion, viscosity measurements and filterability tests of microgel solu-
tions through calibrated membrane filters were first performed in an
extended domain of brine salinities. Then, coreflood experiments
in model granular packs were carried out with microgels diluted in
brines of various salinities and at various microgel concentrations.
These tests evaluate the injectivity and determine the permeability
reductions. After microgel injection, the packs were flushed with
brines of different salinities, and the permeability reductions were
determined under these new conditions. These additional tests were
aimed at reproducing the field conditions when setting a well back
in production after a microgel treatment.
Results showed that the hydrodynamic thicknesses of adsorbed
layers did not vary significantly when microgels were injected in
brines of moderate to high salinities. At very high salinity (200 g/L
NaCl with calcium), a shear-controlled overadsorption of the
microgels was observed. Results also showed that adsorbed layers
of microgels were very stable when exposed to brines of low to
very high salinity. These positive outcomes bring new encouraging
insights for the success of microgel applications.
Introduction
Modern oil-production operations require dedicating an increasing
amount of attention to water control, which involves conformance-
control and water-shutoff issues. Among the available methods for
conformance control or water shutoff (Seright et al. 2003) is injecting
a gelling system composed of a polymer and a crosslinker. This has
been used widely (Seright and Liang 1994; Fulleylove et al. 1996;
Zaitoun et al. 1999; Sydansk and Southwell 2000). However, since
gelling properties have been found to depend on many factors, such
as the gelling time and the final gel strength, the depth of the gel pen-
etration and, more generally, the efficiency of the treatments are quite
difficult to predict. More recent gel-based technologies include pre-
formed bulk gels (Seright 2004), partially preformed gels (Sydansk
et al. 2005), gels with polyethyleneimine (PEI) organic crosslinkers
(Al-Muntasheri et al. 2008a; Al-Muntasheri et al. 2008b), colloidal-
dispersion gels (CDGs) (Chang et al. 2006a; Chang et al. 2006b;
Muruaga et al. 2008), and preformed particle gels (PPGs) (Coste
et al. 2000; Bai et al. 2008), which are more dedicated to treatments
of fractures or very high permeability streaks.
Microgel technology presents original advantages. It is based
on preformed polymeric particles, the microgels, which are stable,
fully water-soluble, and size-controlled with a narrow size distribu-
tion. Microgels are nontoxic, satisfying the OSPAR environmental
tests. The use of microgels was shown to be an effective method
for in-depth permeability control and/or relative permeability
Copyright © 2009 Society of Petroleum Engineers
This paper (SPE 115974) was accepted for presentation at the SPE Annual Technical
Conference and Exhibition, Denver, 21–24 September 2008, and revised for publication.
Original manuscript received for review 07 July 2008. Revised manuscript received for
review 05 May 2009. Paper peer approved 19 May 2009.
590
modification for water-production control (Chauveteau et al. 2004;
Rousseau et al. 2005). Mechanical, chemical, and thermal stabil-
ity of microgels was found to be excellent because of microgels’
chemically crosslinked structure. The rheological behavior of the
microgels showed that they are intermediate between linear poly-
mers and hard spheres, depending on the crosslink density (Rous-
seau et al. 2005) because they are more rigid and/or less penetrable
than linear polymers. Because of their controlled size, softness, and
flexibility, microgels can be transported in reservoir rock without
mechanical trapping. The controlled permeability reduction to
water originates in the adsorption of the microgels as monolayers.
In addition, when they are injected in a multilayered reservoir,
microgels invade the low-permeability zones significantly less
because of the low viscosity of their solutions and because of
steric effects. This is unlike conventional polymer-based systems.
The depth of penetration in the highest-permeability layers is thus
largely improved. As microgels are readily available at industrial
scale, microgel technology was applied in 2005 for the treatment
of an underground gas-storage reservoir. The treatment was suc-
cessful, and the long-term results were very positive in terms of
water shutoff and sand control (Zaitoun et al. 2007).
In some well-treatment situations, the only water that can be
available on the field is high salinity with variable mono and
divalent ionic concentrations. When injecting gelling systems,
increasing the salt content of the mixing water results in increased
gelation time and reduced rate of viscosity buildup. In the case of
polyacrylamide-based gels, the cations shield the negative charges
available on the carboxylate groups leading to the polymer chains’
collapse, making the carboxylate groups no longer available for
the crosslinker molecules (Nasr-El-Din and Hawkins 1991). A
difference was also observed in the gel strength of gels prepared
in low- or in high-salinity waters. Furthermore, gels can swell or
shrink depending on the external environment, such as salinity
(Cohen 1996; Horkay et al. 2000; Molloy et al. 2000). In particular,
it is well known that under certain conditions, polyelectrolyte gels
may undergo a discontinuous volume change, the volume of gel
being influenced by osmotic pressure (Whittington and Naae 1992;
Molloy and Cowling 2000). Maitland (2000) has also pointed out
gel syneresis effects with chromium-based gels. In the presence
of reservoir brine with a high salinity, the use of preflushes and
buffers can minimize the impact of divalent ions on the polymer
but also could be hazardous. Concerning the more recent gel-based
technologies, the impact of variable brine ionic content on the
performances remains uncertain.
The purpose of the present paper is to complete and extend previ-
ous work regarding microgels in order to investigate their perform-
ances in conditions close to those of the field in terms of brine ionic
contents. A broad range of salinities, which can be met in the field,
are examined [20 to 215 g/L total dissolved solids (TDS)].
In the first section of this paper, the materials and methods
are described. Then the main results are presented and discussed,
beginning with the characteristics of the microgels in solution
(intrinsic viscosity, interaction properties, and rheology), continu-
ing with their filterability through calibrated membranes, and fin-
ishing with an evaluation of their performances in porous media.
The main outputs include (a) the good solubility of the microgels
in all the brines studied; (b) their easy in-depth propagation at
long distances, as well as their capacity to form a size-controlled
adsorbed monolayer in brines up to 108 g/L TDS; and (c) their
overadsorption when injected in the highest salinity brine (215 g/L
TDS), for which an interpretation is proposed.
November 2009 SPE Production & Operations
Experimental
Materials. Microgels are colloidal particles of acrylamide-based
polymeric gels with neutral organic covalent crosslinkers. They
are available with different sizes (from 0.3 to 2 ␮m), polymer-
chains chemistry, and crosslink density. The study presented here
is focused on one microgel grade with a size close to 2 ␮m and
an intermediate crosslink density.
Microgel powder was dissolved in four brines with different salt
concentrations. The lowest salinity brine, which will be referred
to as S20/04, was composed of 20 g/L NaCl and 0.4 g/L NaN3
(natrium azide) in distilled water. NaN3 was added to all brines
as bactericide and as stabilizer to prevent free-radical degradation
(concentration was fixed at 0.4 g/L, and the presence of NaN3 is
referred to as /04 in the following nomenclatures). A second brine,
referred to as SW40/04, was a synthetic seawater composed of 40
g/L TDS (24.79 g/L NaCl; 11.79 g/L MgCl2,6H2O; 0.8 g/L KCl;
1.6 g/L CaCl2,2H2O). The third brine, referred to as S100Ca/04,
was composed of 100 g/L NaCl and 8 g/L CaCl2. The last brine,
referred to as S200Ca/04, was composed of 200 g/L NaCl and
15 g/L CaCl2.
The brines were filtered through 0.22-␮m MilliporeTM cellulose
membranes to remove any fine particles in suspension. Before
injection into the cores, the brines were degassed under a vacuum
and kept under a helium cap to prevent bubble formation in the
porous medium.
A high-shear stirring method was used to prepare the microgel
solutions using an Ultraturrax® (IKA Labortechnik) mixing device.
The concentration of the solutions was fixed at 10 g/L, with a shear
rate of 24,000 s–1. The solution was stirred for 10 minutes at 25°C
in adapted brines. The solutions were then diluted before viscosity
measurements, filtration tests, and injections in porous media.
Characterization Methods. Viscosity Measurements. A low-
shear viscometer (LS30 from Contraves) having a Couette geom-
etry (0.5-mm gap) was used to measure the viscosities at various
shear rates (from 0.5 to 40 s–1) of both the microgel solutions and
the solvent to determine the relative viscosities accurately. For all
viscosity measurements, temperature was kept constant at 30°C.
Newtonian relative viscosities, ␩ro, of the solutions were deduced
from plots of relative viscosities vs. shear rates. Then the values
of both intrinsic viscosity [␩] and the Huggins constant KH of
the microgels were derived from plots of the Newtonian-specific
reduced viscosity ␩spr = (␩ro − 1) / C vs. microgel concentration C,
according to the Huggins equation (Russel et al. 1989; Chauveteau
and Sorbie 1991):
␩spr = [␩ ] + K H [␩ ]2 C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (1)
The parameters [␩] and KH characterize the internal density
of the microgel colloids and their interactions in solution, respec-
tively. [␩] is linked to the colloid density (Chauveteau and Sorbie
1991). The effective volume fraction, ␾, can be calculated from the
intrinsic viscosity: ␸ = 2 / 5 ⋅ C ⋅ [␩ ]. For interpenetrable polymeric
colloids, ␾ is an overlap parameter and can have values above
unity. For a neutral flexible polymer, KH increases from 0.3 in a
good solvent with a segment-segment net repulsion at very short
distances up to a value of 0.5 in a theta solvent. More generally,
for objects showing low internal permeability even in good solvent
conditions, KH value is expected to increase as internal permeabil-
ity decreases, reaching values as high as unity when the permeabil-
ity drops to zero (Chauveteau and Sorbie 1991). KH values higher
than unity indicate the presence of attractive interactions between
colloids. When the colloid size is smaller than the hydrodynamic
pore-throat size, the transport in porous media is expected to be
easy as long as the value of KH indicates that microgels are strictly
repulsive, since in this case the formation of multiple adsorbed lay-
ers is prevented. By contrast, when microgels are manufactured to
be partly attractive, a multilayer adsorption can occur, resulting in
a complete plugging when the adsorbed-layer thickness reaches the
pore-throat hydrodynamic radius. Such a property gives the pos-
sibility of completely plugging a water-producing layer, if desired
(Feng et al. 2003).
November 2009 SPE Production & Operations
It should be noted that a straightforward computation of the
experimental uncertainties cannot be carried out for intrinsic-viscos-
ity measurements. However, uncertainties can be estimated from
both experimental reproducibility from one microgel solution to
another and from sensitivity tests performed on the data-fitting
(Eq. 1) plots. In the present work, our investigations indicated that
uncertainties’ values could be estimated, for all the solutions studied,
to ±50 mL/g for the [␩] values and to ±0.1 for the KH values.
Filtration Tests. The solutions were injected through cali-
brated polycarbonate membrane filters with 8-␮m pore diameters
(Isopore®, Millipore, similar to Nuclepore® products) by using a
syringe pump (Perfusor, BBraun, Secura FT, 50 mL) at a flow
rate of 10 mL/h. This flow rate corresponded to a wall shear rate
of 32 s–1 in the pores, taking into account a porosity of 5% and
a membrane diameter of 47 mm. A microgel size between 2 and
2.5 ␮m corresponds to a jamming ratio between 3 and 4 for these
filtration tests. The jamming ratio represents the ratio between the
mean diameter of the microgels and the mean pore diameter.
For all tests, the inlet and outlet of the filter were connected to a
differential-pressure transducer. The filter was saturated previously
with the same brine as the one used for the microgel solution to be
tested. For each microgel solution, two filtration tests have been
carried out on two distinct samples of the solution, 24 hours and 7
days, respectively, after the solubilization. The concentration of the
microgel solutions was fixed to 2 g/L for all filtration tests.
Coreflood Experiments. The porous-media experiments were
aimed at evaluating and comparing the in-depth propagation as
well as the adsorption properties (hydrodynamic thicknesses of
adsorbed layers and adsorption densities) of microgels solubilized
in various brines. Thus, microgel solutions had to be injected
in model porous media having well-known, homogeneous, and
reproducible pore structure and surface properties. Silicon carbide
(SiC) particle packs are such a model of porous media (Médout-
Marère et al. 2000; Chauveteau et al. 2003). The granular media
in our experiments were 10-cm-long SiC packs. The 80-␮m grain
size led to porosities of ␾ = 41±1% and absolute permeabilities of
k = 3±0.5 ␮m2. The hydrodynamic pore radius, rh, was estimated
from a corrected capillary model (Chauveteau and Sorbie 1991):
rh = 1.15 ⋅ 8 k / ␸ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (2)
The estimated value of the hydrodynamic pore-throat radius
was then rh ≈ 9 ␮m (i.e., a pore-throat diameter of approximately
18 ␮m). The jamming ratio for microgels of size between 2 and
2.5 ␮m was, hence, between 7 and 9.
The coreflood experimental setup is shown in Fig. 1. Brine or
unfiltered microgel solutions were injected in the porous medium
by Labotron® or Pharmacia® pumps. On the microgels-injection
line, a Nylon grid with a mesh size of 10 ␮m was set upstream
of the porous media. This grid was aimed at providing a coarse
filtration in order to remove dust or other macroscopic impurities
from the injected solutions. On the brine-injection line, a 0.22-␮m
cellulose filter was set in order to prevent damage to the porous
media during the injections of brine. The pack holder was equipped
with pressure taps to measure pressure drops (⌬P) on different
sections of the 10-cm porous media: an overall section (noted 0 to
10 cm) and two 4-cm internal sections, located respectively from 1
to 5 cm and 5 to 9 cm. Downstream of the porous media, the setup
was equipped with a calibrated capillary in order to measure the
microgel-solution viscosity and, hence, the microgel concentration
in the effluents. In order to measure reference concentrations, the
setup also enabled bypassing the porous medium. Pressure drops
along the different sections of the porous medium and the cali-
brated capillary were measured with differential-pressure sensors
connected to a data-acquisition system composed of a data-logging
unit and a computer.
Both the porous medium and the capillary were immersed in a
thermoregulated bath adjusted at T = 30°C.
A standard procedure was implemented for all the coreflood
experiments. It was designed to determine microgel propagation,
microgel-induced permeability reduction, and microgel adsorption
591
 Fig. 1—Schematic of the experimental setup for the coreflood experiments.

TABLE 1—DESCRIPTION OF THE COREFLOOD EXPERIMENTAL PROCEDURE

Step # Description
3
1 Preliminary, injection of several pore volumes of brine at 2 cm /h
2 Brine injection at various flow rates and determination of the absolute permeability k
3
Microgel injection (C = 1 g/L) at 2 cm /h during 11.5 pore volumes (i.e., about 85 mg
3 st
of microgels injected) and measurement of the mobility reduction ("1 front")
3
Brine injection at different flow rates (from 2 to 50 cm /h) to determine permeability
4
reduction
3
Microgel injection (C = 1 g/L) at 2 cm /h during 15 pores volumes (i.e., about 110 mg
5 nd
of microgels injected) and measurement of the mobility reduction ("2 front")
3
Brine injection at various flow rates (from 2 to 50 cm /h) to determine permeability
6
reduction
3
Microgel injection (C = 2 g/L) at 2 cm /h during 11.5 pore volumes (i.e., about 170 mg
7 rd
of microgels injected) and measurement of the mobility reduction ("3 front")
3
Brine injection at various flow rates (from 2 to 50 cm /h) to determine permeability
8
reduction
3
Brine injections at various flow rates (from 2 to 50 cm /h) with all the available salinity
9
brines

for a given injection concentration (1 g/L); evaluate the effect of Four experiments were performed, which correspond to four
a higher concentration (2 g/L); and investigate the rheological distinct, but similar in terms of petrophysical properties, granular
behavior of the microgel adsorbed layers when exposed to various packs. For a given experiment, microgels solubilized in one of the
salinity brines. This procedure is described in Table 1. four brines were injected. The porous medium was then swept with
Mobility and permeability reduction, Rm and Rk, are defined as all of the four brines. Table 2 gives the brine salinity and the rela-
the ratios of pressure drops during and after microgel injection, tive viscosity ␩r determined at 30°C at the concentration C (g/L)
respectively, to pressure drops in presence of solvent (brine) before of the microgel solutions for the four experiments.
any microgel injection. Rm and Rk are also called resistance factor
(RF) and residual resistance factor (RRF), respectively. Results and Discussions
Characterization of the Solutions. Intrinsic-Viscosity Determi-
nation. Fig. 2 presents the reduced specific viscosities of solu-
tions vs. concentrations according to Huggins relation 24 hours
TABLE 2—LIST AND DESCRIPTION OF THE EXPERIMENTS
after solubilization. Values of intrinsic viscosities [␩] and Hug-
Relative Viscosity ηr, of the gins constant KH are summarized in the table shown alongside
Microgel Solutions the plot. Intrinsic-viscosity results show that microgels prepared
Dissolution
in S20/04 are more swelled than in the three other brines with
Exp. in Brine C = 1 g/L C = 2 g/L
[␩] = 510 ± 50 mL/g. The almost identical [␩] values for the three
1 S20/04 1.6 2.3 other solutions [[␩] close to 200 mL/g, taking into account the
2 SW40/04 1.3 1.8 experimental uncertainty of ± 50 mL/g] suggests that the com-
plete charge screening of the microgels is achieved from the ionic
3 S100Ca/04 1.2 1.6
strength of the SW40/04 brine. All these intrinsic-viscosity values
4 S200Ca/04 1.2 1.4 are significantly lower than the ones of corresponding linear poly-

592 November 2009 SPE Production & Operations

 800
Brine [η ] KH
solution (mL/g)
S20/04
700

Reduced Specific Viscosity (mL/g)

S20/04 510±50 0.3±0.1

600 SW40/04 210±50 1.9±0.1

S100Ca/04 190±50 2.3±0.1

500 S200Ca/04 180±50 2.8±0.1
SW40/04
400

300

200

100 S100Ca/04 S200Ca/04

0
0 0.5 1 1.5 2 2.5 3
C (g/L)

Fig. 2—Intrinsic viscosity and Huggins constant values for microgel solutions in the four brines investigated according to Hug-
gins relation.

mers of similar chemical composition [[␩] ≈ 3000 mL/g (Rousseau
et al. 2005)], as is expected from the chemically crosslinked nature
of the microgels. The Huggins constants derived from the slope
show increasing values with salinity: KH = 0.3 ± 0.1, 1.9 ± 0.1,
2.3 ± 0.1, and 2.8 ± 0.1 for the S20/04, SW40/04, S100Ca/04, and
the S200Ca/04 brines, respectively. KH value is higher than unity
for the three highest-salinity brines (SW40/04, S100Ca/04, and
S200Ca/04) and could suggest the existence of attractive interac-
tions between the colloids. Intrinsic viscosities remain stable with
respect to time over the duration of the porous-media experiments
(e.i., namely a few weeks).
Filtration Tests. For all filtration tests, viscosity measurements
were performed downstream of the filter in order to determine the
effluent concentration recovered. Results indicate that for all tests,
100% of injected material was quickly recovered. This means that
filterability was good for all microgel solutions. However, the pres-
sure-drop measurements over the filters reveal discrepancies. Fig. 3
presents the plots of mobility reduction (Rm) vs. injected volume
24 hours (left side) and 7 days (right side) after solubilization of
the microgels. Results presented concern solutions prepared with
brines of the lowest salinity (S20/04), the intermediate salinity
(SW40/04), and the highest salinity (S200Ca/04). It should be
emphasized that these results can lead only to qualitative interpre-
tations in terms of comparison of the various brines.
Results show that for all the solutions, filterability is better
7 days than 24 hours after solubilization. However, it has been
observed that the viscosities of all the solutions were identical after
24 hours and after 7 days. As a consequence, the filterability behav-
ior does not appear to be governed by the general swelling state of
the microgels, but rather by the presence of a limited amount of
large objects, which do not contribute to the viscosity of the solu-
tions. Indeed, for all the brines investigated, the full solubilization
of the microgels appears to be a rather slow mechanism, and the
poorer filterabilities after 24 hours are certainly only because of the
presence of some partially solubilized objects in the solutions.
Results also show that:
(a) Filterability is globally poorer in S20/04 than in the two
other brines investigated.
(b) Filterability is poorer in S200Ca/04 than in SW40/04.
Observation (a) is likely because of the fact that microgels in
S20/04 brine are more swelled than in the other brines, which
is consistent with the higher value of intrinsic viscosity. In our
experiments, a microgel size close to 2 to 2.5 ␮m for membrane
pore diameters of 8 ␮m corresponds to a jamming ratio, Jr, of
3 to 4. Theoretically, a minimum value of Jr to ensure a good
propagation is 3. Because of particle-size distribution, pore-throat
geometry, size distributions, concentration, and velocity effects,
higher values of Jr are requested in practice. In S20/04 conditions
where microgels present the most swelled conformation, their aver-
age hydrodynamic diameter is likely to be slightly higher than in
the two other brines, and Jr is probably too low. The increase in
mobility reduction can thus be explained by the formation in the

10 10
24h after solubilization 7 days after solubilization

S20/04
Rm/η r
Rm/η r

5 S200Ca/04 SW40/04 5 S20/04

S200Ca/04
SW40/04

0 0
0 10 20 30 40 50 0 10 20 30 40 50
Injected Volume (cm ) 3
Injected Volume (cm3 )

Fig. 3—Rm /␩r (mobility reduction over relative viscosity) vs. injected volume during filtration tests 24 hours (left) and 7 days
(right) after dissolution of microgels in three brines.

November 2009 SPE Production & Operations 593

 Pore Volumes Injected
0 1 2 3
10
9 100
8
Mobility Reduction (Rm) 7
6 0−10cm 1−5cm
5
4
3
2 20
5−9cm
1
0 0
0 10 20
Fig. 4—Mobility reduction on sections 0 to 10 cm, 1 to 5 cm, and 5 to 9 cm and effluent concentration during the first injection of mi-
crogels (first front) for Experiment 3 (S100Ca/04). The stabilization on the three sections indicates a good in-depth propagation.
pores of some bridges between two adsorbed microgels and a third
one transported by the flow.
Observation (b) is more consistent with the higher KH value
in S200Ca/04 (KH = 2.8 ± 0.1) than in SW40/04 (KH = 1.9 ± 0.1),
suggesting significant attractive interactions between microgels in
S200Ca/04. The filtration-test results can thus be interpreted by the
presence of some microgels aggregates in the S200Ca/04 solution,
which slow down the swelling process and deteriorate the filter-
ability only 24 hours after solubilization. After 7 days, microgels
seem to be satisfactorily swelled because the filterability of the
solution is significantly improved.
Finally, it appears that the optimum filterability that is observed
for the SW40/04 solution is because of both the charge screening
effects making the microgels less swelled than in S20/04 and the
KH value that is not high enough to entail significant microgel/
microgel attractive interactions.
Coreflood Experiments. Pressure Drops and Effluents Profiles.
Detailed results regarding the injection of the first front of micro-
gels prepared in S100Ca/04 (Experiment 3) are presented, as an
example, in Fig. 4. This figure shows the mobility reduction (Rm),
recorded on the three sections of the porous medium, and the
effluent concentration vs. injected volume (cm3) and pore volumes
injected.
For all sections, Rm increases almost linearly in a first step
with a certain delay because of adsorption and then levels out
at a plateau value, which is practically the same regardless of
distance from the inlet. The effluent concentration was calcu-
lated from the pressure-drop measurements in the capillary. It
appears that 100% of the injected concentration is recovered after
approximately 3.5 pore volumes injected. These results indicate
a very good in-depth propagation of the microgels, as well as a
stabilized adsorption.
All the experiments showed a complete absence of entry-face
plugging. In addition, the mobility-reduction profiles over the two
internal sections of the porous media were similar for all experi-
ments. It has therefore been decided, on the overall figures below,
to present results corresponding to the 1-to-5-cm section only.
Figs. 5 through 7 are regrouped results of Rm vs. injected
volume (and pore volumes injected) for the four experiments dur-
ing the first, second, and third microgel injections (first front at
1 g/L, second front at 1 g/L, and third front at 2 g/L), respectively.
Microgel concentrations at the outlet of the porous media were also
determined from the capillary viscosity measurements during all
594
4 5 6 7 8 9 10
80
Effluent Concentration (%)
60
40
Exp. 3 (S100Ca/04)
1 st front 1g/L
30 40 50 60 70
3
Injected Volume (cm )
experiments. For the sake of clarity, these results are not plotted.
They will be discussed when necessary.
For the solutions in S20/04, SW40/04, and S100Ca/04, results
of the three injections clearly show a very easy in-depth propaga-
tion of the microgels through the porous media. For these three
brines, Rm levels after one pore volume of injected solution, and
the outlet concentration reaches 100% referring to the injected con-
centration. Results show that stabilization of both Rm and effluent
concentrations is maintained over the duration of the experiments
(i.e., namely after about 27 and 15 pore volumes injected at 1 g/L
and 2 g/L, respectively).
Injections in S20/04 led to higher Rm values after stabilization
than those determined in SW40/04 and S100Ca/04. This behavior
mainly originates from the higher swelling state of the microgels
in the S20/04 solution. More precisely, since, at the concentrations
investigated, the viscosity in S20/04 is not significantly higher
than in the other brines, the increased Rm is mainly because of an
increased size of the swelled microgels in S20/04. This point will
be discussed further in the section Water Permeability Reduction
and Rheology of Adsorbed Layer. From the results shown in Fig.
5, Rm during injection in S20/04 seems to reach stabilization more
slowly than for the injections in SW40/04 and S100Ca/04. An
enlargement of the experimental curves clearly indicates that this
observation is actually incorrect, and that a slow increase of Rm
is also observed for the injections in SW40/04 and S100Ca/04
(enlargement not shown for clarity). The slow increases in Rm rep-
resent the rather slow overall deposition kinetics of the microgels
on the available surface. This particular deposition kinetics is likely
because of the fact that microgel deposition, even as a monolayer,
occurs in two steps: a first and quick step in which microgel
deposition can be treated as the deposition of solid particles and a
second, slower step in which additional microgels are deposited,
thus making the adsorbed layer denser, slightly thicker, and entail-
ing slow mobility-reduction increases. This process should take
place when the concentration is increased (i.e., when the osmotic
pressure exerted by the microgels becomes more significant). This
hypothesis of a second-step adsorption is consistent with the fact
that microgels are soft and deformable particles. However, quan-
titative determinations of the adsorption densities are difficult to
perform experimentally.
In the case of microgels injected in S200Ca/04, Rm curves are
dramatically different as with the injections in the three other
brines. After a first stage in which Rm seems to be stabilized and
effluent concentration reaches 100%, a second stage increase of Rm
November 2009 SPE Production & Operations
 Pore Volumes Injected
0 2 4 6 8 10
20
Section 1−5 cm
18
1 st front 1g/L
16

Mobility Reduction (Rm)

Exp. 4
14 Exp. 2
(SW40/04) (S200Ca/04)
12 Exp. 1 Exp. 3
(S20/04) (S100Ca/04)
10

0
0 10 20 30 40 50 60 70 80
3
Injected Volume (cm )

Fig. 5—Mobility reductions on the 1-to-5-cm section during the injection of the first front of microgels (C = 1 g/L) for the four
experiments. For Experiments 1, 2, and 3, the stabilized values of Rm indicate a good in-depth propagation, while an overadsorp-
tion of the microgels is observed for Experiment 4.

takes place, while effluent concentration is reduced to near 30 to
40%. During the second microgel injection, the increasing trend of
the Rm curve is carried on, and then stabilization occurs with the Rm
value being approximately three times higher than for the injection
in the other brines. During the third microgel injection, at 2 g/L,
another sharp increase followed by stabilization is observed with
an Rm value approximately three to four times higher than for the
other brines. The stabilizations of the mobility reductions coincide
with effluent concentrations reincreasing up to 100%.
The latter results suggest, first, an overadsorption of the micro-
gels injected in S200Ca/04. As will be discussed in the section
Water Permeability Reduction and Rheology of Adsorbed Layer,
permeability-reduction determinations performed after microgel
injections confirm this hypothesis. It appears therefore that a
threshold, in terms of brine ionic strength, exists above which
microgel overadsorption can occur. Interestingly, this threshold
does not correspond to the onset of the full charge screening of the
particles, the intrinsic viscosities being very similar in SW40/04,
S100Ca/04, and S200Ca/04. The threshold value is more likely
linked to a certain level of attractive microgel/microgel interac-
tions, as indicated by the continuously increasing KH values, with
increasing ionic strength. A critical KH value could then be associ-
ated to the overadsorption threshold. However, more experiments
would be required to draw a more precise picture of the onset of
microgel overadsorption.
The stabilization of the mobility-reduction curves for microgels
injected in S200Ca/04 could originate in the relative weakness
of microgel/microgel attractive interactions. Indeed, if microgel

Pore Volumes Injected

0 2 4 6 8 10 12 14
20

15
Mobility Reduction (Rm)

Exp. 1
Exp. 3 (S20/04) Exp. 4
(S100Ca/04) (S200Ca/04)
10

Exp. 2 Section 1−5 cm

(SW40/04) 2 nd front 1g/L
0
0 10 20 30 40 50 60 70 80 90 100 110
3
Injected Volume (cm )

Fig. 6—Mobility reductions on the 1-to-5-cm section during the injection of the second front of microgels (C = 1 g/L) for the
four experiments. For Experiments 1, 2, and 3, in-depth propagation remains good. For Experiment 4, microgels overadsorption
levels out.

November 2009 SPE Production & Operations 595


0 2
40
30
Mobility Reduction (Rm)
Exp. 2
(SW40/04) (S100Ca/04)
20
10
0
0 10 20
Fig. 7—Mobility reductions on the 1-to-5-cm section during the injection of the third front of microgels (C = 2 g/L) for the four ex-
periments. For Experiments 1, 2, and 3, in-depth propagation remains good. For Experiment 4, microgel overadsorption resumes
in a first step, then levels out in a second step.
multilayers are formed, the pore-throat diameters could be reduced
significantly, entailing an increase of the effective shear rate and
thus an increase of the hydrodynamic forces. When hydrody-
namic forces bring more energy than those involved in microgel/
microgel bonds, additional bonds can no longer be formed. As a
consequence, the overadsorption of microgels can be said to be
shear-rate limited. This hypothesis is consistent with the higher
sensitivity of Rm to the pressure pulses caused by the pump (the
periodic pulses in the plots) for the S200Ca/04 injections than for
the injections in the other brines.
Also regarding the injection in S200Ca/04 is that, generally
speaking, the initial Rm values during Injection Front n + 1 are
much lower than the final Rm values during Injection Front n (Figs. 6
and 7). These discrepancies cannot be explained by simple viscos-
ity changes between brine and microgel solutions, as the similar,
but much lower, discrepancies observed during the injections in
the other brines can. However, this behavior is well understood: It
also originates from the fact that the overadsorbed microgel layers
are very sensitive to shear rate. As will be discussed in the section
Water Permeability Reduction and Rheology of Adsorbed Layer
(rheological study of the adsorbed layers), it appears that some
overadsorbed microgels can be removed from the surface when
exposed to high shear rates. Accordingly, since high shear rates are
applied during the brine sweeps performed between each microgel
injection front, the permeability reduction-values are much lower
after these sweeps than before (hysteresis effect), entailing the
discrepancies observed between starting and final Rm values.
The fact that the onset of microgels overadsorption is delayed
(as it occurs after approximately eight pore volumes injected at
1 g/L) also probably originates in the relative weakness of micro-
gel/microgel attractive interactions. However, a satisfactory inter-
pretation of these results requires further investigations.
As general concluding remarks regarding adsorption and
microgel in-depth propagation, it appears that salinity can play
a significant role beyond a certain threshold. Nevertheless, in
our experimental conditions, salinity never compromises micro-
gel propagation, as a stabilization of the mobility reductions is
observed without any plugging tendency. Since it appears as being
shear-rate controllable, the microgel overadsorption at high salinity
could be exploited to obtain more significant permeability-reduc-
tion effects than with a monolayer adsorption.
Microgel Adsorption. Microgel adsorption was determined for
Experiments 1, 2, and 3 from the shift in the effluent-breakthrough
596
Pore Volumes Injected
4 6 8 10
Section 1−5 cm
3rd front 2g/L
Exp. 3 Exp. 1
(S20/04)
Exp. 4
(S200Ca/04)
30 40 50 60 70 80
3
Injected Volume (cm )
curves between the first and the second microgel injection. This
method assumes that maximum adsorption is reached after the
first injection. It was therefore not suitable for Experiment 4
(S200Ca/04). As will be shown in the next section, this condition
was also not exactly fulfilled for the experiments in SW40/04 and
S100Ca/04 brines because a slight increase of the adsorbed-layer
thicknesses is observed between the first and the second microgel
injection. As a consequence, the adsorption is slightly underesti-
mated for these two experiments.
The equation used to calculate Γ (␮g/g), the mass of adsorbed
microgels per unit mass of SiC, is
⌫ = ( ⌬V ⋅ C ) / (1000 ⋅ mSiC ) , . . . . . . . . . . . . . . . . . . . . . . . . . . (3)
where ⌬V (cm3) is the volume difference between the 50% con-
centration value determined from the effluent-analysis profile on
the first and second fronts, C (g/L) is the injected concentration,
and mSiC (g) is the mass of SiC in the granular pack.
Results are shown in Table 3. Adsorption values are all of the
same order of magnitude, but they increase when increasing the
ionic strength of the microgels dilution brine. These results can
be linked either to an increase of the adsorbed-layer thickness
with salinity or to an increase of the adsorbed-layer density. Their
interpretation will be provided in the next section, on the basis
of the determination of the adsorbed-layer thicknesses from the
permeability-reduction measurements.
A comparison of these adsorption values to the adsorption of
linear polyacrylamides would not be appropriate. Indeed, adsorp-
tion of linear polymer has been shown (Chauveteau et al. 1997)
to be significantly dependent on many parameters, including the
TABLE 3—ADSORPTION VALUES (Γ ) FOR EXP. 1 TO 3*
Exp. Γ (µg/g)
1 (SW20/04) 97
2 (SW40/04) 108
3 (S100Ca/04) 129
* Adsorption tends to increase with salinity. Since the hydrodynamic
thicknesses of the adsorbed layers do not significantly increase with salinity,
this result indicates that the adsorbed layers tend to be denser when
increasing salinity.
November 2009 SPE Production & Operations
content in ionic monomers of the chains. For example, in the Second, results related to the evolution of Rk and εh vs. shear rate
case of the sulfonated polyacrylamides, adsorption in Vosges will be presented for two specific experiments. Third, Rk and εh
sandstones with 20 g/L NaCl salinity ranges from 130 ␮g/g when values as a function of the amount of microgels injected and the
no sulfonates groups are present to only a few ␮g/g for 25-mol% brine salinities will be discussed from plots of synthetic results,
sulfonates. allowing comparison of the four experiments.
Water Permeability Reduction and Rheology of Adsorbed Examples of Rk and εh profiles on the three sections along the
Layer. As pointed out previously, the adsorption of microgels onto granular pack
•
(0 to 10 cm, 1 to 5 cm, and 5 to 9 cm) are plotted on
rock pore surfaces allows controlling the permeability reduction Fig. 8 vs. ␥ for the first front of Experiment 3 (microgels injected
to water near the wellbore where the microgels have been placed. in S100Ca/04 brine at a concentration of 1 g/L). Results show
Accordingly, microgel adsorption properties have been investi- similar Rk and εh values on the three sections. This indicates that
gated in detail in this work. For all experiments, after displacement the microgel-adsorption profile is homogeneous along the porous
of the nonadsorbed microgels by brine, as effluent concentration media because it is consistent with the good in-depth propagation
reached 0%, the pressure drops in the various sections of the pack evidenced from the mobility-reduction plots. Similar results, in
remained higher than before microgel injection. This indicates a terms of homogeneity of the adsorption profile, have been obtained
permeability reduction to water by microgel adsorption. Determi- for all the experiments reported in this paper. As a consequence,
nations of the permeability reductions, Rk, lead to estimations of it has been chosen, in the following discussions, to present only
the hydrodynamic thicknesses of the adsorbed layers, εh, according results concerning the 1-to-5-cm section of the porous media.
to the following capillary model relationship: Results plotted in Fig. 8 also show a slight shear-shinning behav-
ior of the adsorbed layer. This behavior, which has already been
ε h = rh (1 − Rk −1/ 4 ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (4) reported for microgels (Rousseau et al. 2005), is a consequence of
the limited deformability of the particles because of their covalent
In this work, the rheological behavior of the adsorbed layers has crosslinking. Because it was also already observed with microgels,
been investigated for all experiments after each microgel injection results show no hysteresis between the increasing and decreasing
through measurements of Rk and determination
•
of εh at different flow shear-rate stages. This result stands as a consequence of the strong
rates. Values of the wall shear rates ␥ were estimated from: adsorption of the microgels, which entails a very good mechanical
stability of the adsorbed layer. The adsorbed microgels induce a
• 4v permeability reduction of approximately 3, with a hydrodynamic
␥ = 2.5 , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (5)
rh thickness of the adsorbed layer of approximately 2 ␮m. The latter
value is consistent with the hydrodynamic size of the microgels
where v is the interstitial mean velocity ( v = Q / (S␸ ) , Q is the classically measured in solutions of moderate ionic strengths (typi-
flow rate, S is the cross section of the core, and 2.5 is a factor cally S20/04 brine). The fact that a 2-␮m-thick adsorbed layer is
that has been previously determined for sharp-edged granular obtained in higher salinity brine, in which the intrinsic viscosity
packs (Chauveteau and Sorbie 1991). Shear-rate values have been is lower than in S20/04, implying a value of the microgel size that
investigated between approximately 10 s–1 (i.e., the shear rate dur- should be smaller, is a striking result. Hydrodynamic layer thick-
ing the injection of the microgels) and approximately 200 s–1. In nesses as a function of the salinity will be discussed further, when
order to check for hysteresis effects, adsorbed-layer-rheology tests comparing results for the four experiments.
have been performed in two stages: first when increasing shear Similar results in terms of shear-rate dependence have been
rates and second when decreasing shear rates. All shear rates have observed for Experiments 1, 2, and 3. Fig. 9 shows, as a represen-
been calculated using initial values of the pore radii, not taking tative• example of these experiments, the plots of Rk and εh values
into account the presence of the adsorbed layers. Shear-rate values vs. ␥ on the 1-to-5-cm section after the three microgel injections
should then be considered as estimates, and detailed interpreta- for Experiment 3. The same shear-rate dependence is observed
tions based on absolute values of shear rates cannot be provided. after the three fronts. Fig. 9 also indicates that as a function of
Nevertheless, shear-rate dependencies of the measured Rk and εh the amount of microgels injected, adsorption increases, but tends
can be interpreted satisfactorily. to level out. This point will be discussed in more detail when
All the results discussed in this section will be presented in comparing results for all of the experiments.
terms of both permeability reductions and hydrodynamic thick- Microgels injected in the highest salinity brine (Exp. 4,
nesses of adsorbed layers. Such a presentation allows considering S200Ca/04) exhibit a very different behavior in terms of Rk and
simultaneously the effectiveness of the microgel treatment of the εh vs. shear rate. Results plotted in Fig. 10 show higher Rk and εh
porous media and the conformation of the adsorbed layers. Results values than for the three other experiments, with a marked increase
shown will first concern the homogeneity of the Rk and εh profile in the amount of microgels injected, indicating of an overadsorp-
along the porous media for a single representative experiment. tion of the microgels. The results also show a marked decrease of
Adsorbed Layer Thickness, εh ( μm)

5 3
st
Permeability Reduction (Rk)

1st front
front (1g/L) 1−5 cm
5−9 cm
Exp. 3
4
5−9 cm 2

3 st
1st front
front (1g/L)
0−10 cm Exp. 3
1
2 1−5 cm
0−10 cm

1 0
0 50 100 150 200 250 0 50 100 150 200 250
. .
Shear Rate, γ (s−1) Shear Rate, γ (s−1)

Fig. 8—First front of microgels during Experiment 3 (S100Ca/04) on sections 0 to 10 cm, 1 to 5 cm, and 5 to 9 cm. Rk and ␩h
are almost the same for the three sections of the granular pack. A slight shear-thinning behavior is observed. Results show no
hysteresis effects vs. shear rate.

November 2009 SPE Production & Operations 597

 Adsorbed Layer Thickness, εh ( μm)
5 3
Permeability Reduction (Rk) Section 1−5 cm 3rd front
Exp. 3 3rdfront (2 g/L)
(S100Ca/04)
4 (2 g/L)
2

3
2 nd front
1 1st front (1 g/L)
2 (1 g/L) Section 1−5 cm
1-5 cm
1st front 2nd front Exp. 3
(1 g/L) (1 g/L) (S100Ca/04)
(S100Ca/04)
1 0
0 50 100 150 200 250 0 50 100 150 200 250
.. ..
Shear Rate, γ (s−1) Shear Rate, γ (s−1)

Fig. 9—Evolution of Rk (left) and εh (right) vs. shear rate after the first, second, and third fronts of microgels for Experiment
3 (S100Ca/04) on the 1-to-5-cm section. Increase of Rk and εh vs. the amount of microgels injected tends to stabilize. A slight
shear-thinning behavior is observed. Results show no hysteresis effects vs. shear rate.

Rk and εh between the stages of increasing and decreasing shear experiment, values of εh ≈ 2.3 ␮m and Rk ≈ 3.5 are expected accord-
rates. This hysteresis effect indicates that the overadsorbed micro- ing to our previous investigations (Rousseau et al. 2005). Results
gel layer has a poorer mechanical stability than the adsorbed layers for Experiments 2 and 3 (injections in SW40/04 and S100Ca/04,
formed during injections to lower salinity. As from the stabilization respectively) show that Rk and εh increase from one injection
observed in the Rm curves, these results indicate again that overad- to another, but their values tend to level out at εh ≈ 2.3 ␮m and
sorbed microgels can be removed by hydrodynamic forces. This Rk ≈ 3.5 (i.e., the same values as for Experiment 1). Interpretation
interpretation is consistent with the hypothesis of a shear-limited of this result is not straightforward. Because the intrinsic viscosi-
overadsorption, originating from weak microgel/microgel attrac- ties measured in SW40/04 and S100Ca/04 are significantly lower
tive interactions for this salinity. than in S20/04, microgels hydrodynamic size must be smaller in
For the further discussions, it should be stressed that Rk and εh SW40/04 and S100Ca/04 than in S20/04. As a consequence, the
values measured in Experiment 4 depend on the maximum shear conformation of the adsorbed layers in Experiments 2 and 3 can-
rate applied. In particular, these values could have been lower if not be the same as in Experiment 1 (i.e., namely a monolayer of
higher shear rates had been applied. Nevertheless, for the four adsorbed particles). As previously stated, adsorption measurements
experiments performed, the shear-rate dependencies of Rk and εh indicate that adsorption is higher in Experiments 2 and 3 than in
during the stages of decreasing shear rates are similar. Results Experiment 1. As a result, the adsorbed layers are denser in the
regarding the variations of Rk and εh will be discussed later and latter three experiments than in Experiment 1. This also suggests
compared for the four experiments from values measured during that attractive microgel/microgel interactions play a role in the
the decreasing stage at a given shear rate of approximately 100 s–1, adsorption of the microgels for salinities higher than S20/04. In
corresponding to a flow rate of 20 cm3/h in the porous media. Experiments 2 and 3, these interactions likely allow more micro-
Fig. 11 summarizes the values of Rk and εh on the 1-to-5-cm gels to adsorb onto the surface than the amount corresponding to
section and at a given shear rate of approximately 100 s–1 for all a monolayer, but do not entail a shear-rate-limited overadsorption
experiments during the three microgel injections. Because the as observed in Experiment 4.
results will be discussed in terms of amount of microgels injected; Results for Experiment 4 show significantly higher Rk and
it is important to take into account the two injection concentra- εh values than for the other three experiments, with an increas-
tions and the volumes injected, the first, second, and third fronts ing trend vs. the amount of microgels injected. Final values of
correspond respectively to injections of 85, 110, and 170 mg of hydrodynamic adsorbed-layer thickness and permeability reduc-
microgels. The amount of microgels injected during the third front tion are εh ≈ 2.9 ␮m and Rk ≈ 6.6, respectively. However, these
is close to the total amount injected during the first two fronts. values depend on the maximum shear rate previously applied on
Results for Experiement 1 (injection in S20/04) show that Rk the adsorbed layer. Since stabilization of the Rm curves has been
and εh do not depend on the amount of microgels injected. For this observed in this experiment during the second and third front,
Adsorbed Layer Thickness, εh (μm)

17 4
Permeability Reduction (Rk)

Section 1−5 cm
1-5 cm 3 rd front
15 Exp. 4 (2 g/L)
(S200Ca/04)
(S200Ca/04)
13 3
rd
11 3 front
(2 g/L)
9 2
7 2 nd front
(1 g/L)
5 1 1 st front Section 1−5 cm
1-5 cm
st nd (1 g/L) Exp. 4
3 1 front 2 front (S200Ca/04)
(S200Ca/04)
(1 g/L) (1 g/L)
1 0
0 50 100 0 50 100 150 200 250
.150−1 200 250
.
Shear Rate, γ (s ) Shear Rate, γ (s−1 )

Fig. 10—Evolution of Rk (left) and εh (right) vs. shear rate after the first, second, and third fronts of microgels for Experiment 4
(S200Ca/04) on the 1-to-5-cm section (for the highest applied shear-rate value, the pressure sensor was saturated, and the Rk
and εh could not be calculated). Rk and εh strongly increase vs. the amount of microgels injected. A shear-thinning behavior, and
a strong hysteresis effect vs. shear rate are observed.

598 November 2009 SPE Production & Operations

 8 4

Adsorbed Layer Thickness, εh (μm)

Exp. 4 Exp. 4

Permeability Reduction (Rk)

7 (S200Ca/04) (S200Ca/04)
3
6
Exp. 1
5 (S20/04)
2
4
Exp. 1
3 (S20/04)
1 Exp. 2 Exp. 3
2 Exp. 2 Exp. 3 (SW40/04) (S100Ca/04)
(SW40/04) (S100Ca/04)
1 0
1st front 2nd front 3rd front 1st front 2nd front 3rd front

Fig. 11—Evolution of Rk (left) and εh (right) for the first, second, and third fronts (i.e., as a function of the amount of microgels
injected, for each experiment on the 1-to-5-cm section at a given shear rate of approximately 100 s–1. Rk and εh values are similar
for S20/04, SW40/04, and S100Ca/04, but are significantly higher for S200Ca/04.

Adsorbed Layer Thickness, εh (μm)

8 4
Permeability Reduction (Rk)

5
3
4

1 2
Exp. 1 Exp. 2 Exp. 3 Exp. 4 Exp. 1 Exp. 2 Exp. 3 Exp. 4
(S20/04) (SW40/04) (S100Ca/04) (S200Ca/04) (S20/04) (SW40/04) (S100Ca/04) (S200Ca/04)

Fig. 12—Evolution of Rk and εh for each experiment after injection of the three fronts of microgels on the 1-to-5-cm section at a
given shear rate of approximately 100 s–1. Values of Rk and εh are similar for Experiments 1, 2, and 3 and are significantly higher
for Experiment 4. An overadsorption of the microgels is suggested in the latter case.

Rk and εh do not seem to increase with the amount of microgels
injected, but rather with the injected concentration. This observa-
tion is consistent with an overadsorption originating from attractive
microgel/microgel interactions, as more microgel/microgel bonds
can be formed when concentration is increased.
Fig. 12 summarizes the values of Rk and εh after the third front
for all experiments. These plots clearly show that permeability
reduction by microgel adsorption is very stable vs. salinity when
microgels are injected in S20/04, SW40/04, and S100Ca/04. A
shear-controlled overadsorption occurs only at very high salin-
ity, when microgels are injected in the S200Ca/04 brine. As a
general result, it appears that microgels allow a precise control
of the permeability reduction in an extended scope of brines up
to a very high upper salinity limit. This result stands as a prom-
ising prospect for the robustness of future microgel-based well
treatments.
Effect of Salinity Changes on the Adsorbed Layer. This sec-
tion concerns the behavior of the microgel adsorbed layers when
exposed to the various brines after the injections of the microgels.
Fig. 13 plots the Rk and εh values measured after sweeping the
porous media with each brine for the four experiments. The larger
dots represent the initial salinities (i.e., the salinities of the injected
microgel solutions). Arrows indicate the order according to which
the various salinity sweeps have been performed.

4
Adsorbed Layer Thickness, εh (μm)

: starting points
points (brines
(brines used Exp. 2 Exp. 4
11
(S200Ca/04)
Permeability Reduction (Rk)

during microgel
microgel injections)
injections) (SW40/04)
Exp. 4
9 3
(S200Ca/04)

7
Exp. 1 2
(S20/04) Exp. 1
5 Exp. 3
(S20/04) Exp. 3
(S100Ca/04)
1 (S100Ca/04)
3 : starting points
points (brines
(brines used
used
Exp. 2 during microgel
during microgel injections)
injections)
(SW40/04)
1 0
S20/04 SW40/04 S200Ca/0 S20/04 SW40/04 S100Ca/04 S200Ca/04

Fig. 13—Evolution of Rk and εh for Experiments 1, 2, 3, and 4 during brine sweeps on the 1-to-5-cm section at a given shear rate
of approximately 100 s–1. The larger dots represent the initial salinity (salinity of the injection brine). Adsorbed layers of microgels
swell or flatten when exposed to different salinities. Hyteresis effect is observed for Experiment 4.

November 2009 SPE Production & Operations 599

 As a general result, microgel adsorbed layers tend to swell
when the salinity decreases and also tend to flatten when salinity
increases, as is expected from charge screening mechanisms.
For Experiments 1, 2, and 3, the microgel adsorbed layers are
only slightly sensitive to the variations of salinity. For these three
experiments, Rk and εh vary at the most from approximately 1.5
units and 0.5 ␮m, respectively, from the lower to the higher salin-
ity. In addition, almost no hysteresis is observed when increasing
and decreasing the salinity: The swelling and flattening of the
adsorbed layer appear to be reversible, that is to say controllable.
Interestingly, since the initial Rk and εh values are similar for
Experiments 1, 2, and 3, the results are consistent with the hypoth-
esis of denser adsorbed layers for Experiments 2 and 3 than for
Experiment 1. In the case of Experiment 1, the adsorbed microgels
are initially swelled and flatten when they are exposed to higher
salinities. In the case of Experiments 2 and 3, it clearly appears
that the adsorbed microgels are initially flattened since they swell
when the salinity is reduced. So, for the initial Rk and εh values
in Experiments 2 and 3 to be equivalent to those in Experiment 1,
microgel density on the surface has to be higher.
In Experiment 4, the sensitivity to salinity of the adsorbed layer
is significantly more marked: Rk and εh vary at the most from
approximately 5 units and 1 ␮m, respectively, from the lower to
the higher salinity. Hysteresis is also observed. The latter effect
likely originates from the shear sensitivity of the overadsorbed
layer. As the salinity is reduced, the adsorbed layer swells, entail-
ing a reduction of the pore-throat sizes. Thus, increased hydrody-
namic forces in the pore throats can remove microgels from the
adsorbed layer.
To summarize, results show that the robustness of the adsorbed
layers is very satisfactory when microgels are injected in S20/04,
SW40/04, and S100Ca/04 brines. When microgels are injected in
the very high salinity S200Ca/04 brine, because of the overadsorp-
tion, the absorbed layer is more sensitive to variation of salinity.
Conclusions
This work reports a lab-scale evaluation of the performances of
microgel technology in brine-salinity conditions, which can be
encountered in field applications. The salinities investigated were
aimed at being representative of both the dissolution brines, which
can be available on site, and the formation brines to which the
microgel-treated formations could be exposed after treatment.
Results of the first part of the study show that a satisfactory
dissolution of the microgels can be achieved in brines of salinity
up to 215 g/L TDS (200 g/L NaCl and 15 g/L CaCl2), and that their
viscosimetric properties evolve as expected for polyelectrolyte
colloids. Filterability of microgel solutions is also satisfactory in
all brines, although some discrepancies linked to the aging of the
solutions or their variable swelling states have been observed, in
particular for the highest salinities.
Microgels have then been injected in model porous media
in a large domain of salinities. Results show that no plugging is
observed for all the salinities investigated. Both mobility- and per-
meability-reduction values are stable when microgels are injected
in brines of moderate (20 g/L NaCl) to high salinity containing
calcium (100 g/L NaCl and 8 g/L CaCl2). At very high salin-
ity in presence of calcium (200 g/L NaCl and 15 g/L CaCl2), a
shear-controlled overpermeability reduction occurs. This latter
effect, of which the onset is likely because of increasing attractive
interactions between microgels when increasing salinity, could be
exploited for the treatment of very high permeability layers.
After microgel injection, the porous media were post-flushed
with brines of different salinities, and the permeability reductions
were determined under these new conditions. These tests were
aimed at reproducing the field conditions when setting a well back
in production after a microgel treatment. In the case of microgel
injections carried out in brines of moderate to high salinity (up
to 100 g/L NaCl and 8 g/L CaCl2), results show that the perme-
ability reductions are only moderately affected when varying the
salinity of the post-flush brines. In addition, these modifications
are reversible and are clearly controlled by the swelling and flat-
tening effects of the microgels also observed in bulk solutions. In
600
the case of microgel injection carried out in the very high salinity
brine (200 g/L NaCl and 15 g/L CaCl2), the permeability reduction
appears to be more sensitive to the post-flushes with variable salin-
ity, and hysteresis effects are observed. This behavior is consistent
with a shear-rate-controlled thickness of the overadsorbed layers
of microgels.
All these results highlight the excellent chemical stability of the
microgels compared to linear-polymer-based systems. They bring
encouraging insights for the success of forthcoming applications
of microgel technology.
Acknowledgments
The authors wish to thank IFP for permission to publish this work
and SEPPIC for manufacturing the microgels.
References
Al-Muntasheri, G.A, Nasr-El-Din, H.A. Peters, J.A., and Zitha, P.L.J.
2008a. Thermal decomposition and hydrolysis of polyacrylamide-co-
tert-butyl acrylate. European Polymer Journal 44 (4): 1225–1237. doi:
10.1016/j.eurpolymj.2008.01.022.
Al-Muntasheri, G.A., Nasr-El-Din, H.A., and Zitha, P.L.J. 2008b. Gelation
Kinetics and Performance Evaluation of an Organically Crosslinked
Gel at High Temperature and Pressure. SPE J. 13 (3): 337–345. SPE-
104071-PA. doi: 10.2118/104071-PA.
Bai, B., Huang, C.G., Liu, H., Seright, R.S., and Wang, Q.M. 2008. Case
Study on Preformed Particle Gel for In-Depth Fluid Diversion. Paper
SPE 113997 presented at the SPE/DOE Symposium on Improved Oil
Recovery, Tulsa, 19–23 April. doi: 10.2118/113997-MS.
Chang, H.L., Sui, X., Xiao, L, Guo, Z., Yao, Y., Xiao, Y., Chen, G., Song,
K., and Mack, J.C. 2006a. Successful Field Pilot of In-Depth Colloidal
Dispersion Gel (CGD) Technology in Daqing Oil Field. SPE Res Eval
& Eng 9 (6): 664-673. SPE-89460-PA. doi: 10.2118/89460-PA.
Chang, H.L., Zhang, Z.Q., Wang, Q.M., Xu, Z.S., Guo, Z.D., Sun, H.Q.,
Cao, X.L., and Qiao, Q. 2006b. Advances in Polymer Flooding and
Alkaline/Surfactant/Polymer Processes as Developed and Applied in
the People’s Republic of China. J. Pet Tech 58 (2): 84–89. SPE-89175-
MS. doi: 10.2118/89175-MS.
Chauveteau, G. and Sorbie, K.S. 1991. Mobility control by polymers. In
Basic Concepts in Enhanced Oil Recovery Processes, ed. M. Bavière,
Vol. 33, Chap. 2, 43–87. London: Critical Reports on Applied Chem-
istry, Elsevier Applied Science.
Chauveteau, G., Kuang Y., and Fleury, M. 2003. A Structural Model to Pre-
dict Transport Properties of Granular Porous Media. Paper SCA2003-
53 prepared for presentation at the International Symposium of the
Society of Core Analysis, Pau, France, 21–24 September.
Chauveteau, G., Tabary, R., and Kohler, N. 1997. Rheology of Polyacryl-
amide-Based Systems Under Near-Wellbore Conditions. Paper SPE
37299 presented at the International Symposium on Oilfield Chemistry,
Houston, 18–21 February. doi: 10.2118/37299-MS.
Chauveteau, G., Tabary, R., Blin, N., Renard, M., Rousseau, D., and Faber,
R. 2004. Disproportionate Permeability Reduction by Soft Preformed
Microgels. Paper SPE 89390 presented at the SPE/DOE Symposium of
Improved Oil Recovery, Tulsa, 17–21 April. doi: 10.2118/89390-MS.
Cohen Addad, J.P. ed. 1996. Physical Properties of Polymeric Gels. West
Sussex, UK: John Wiley & Sons Ltd.
Coste, J.-P., Liu, Y., Bai, B., Li, Y., Shen, P., Wang, Z., and Zhu, G. 2000. In-
Depth FLuid Diversion by Pre-Gelled Particles. Laboratory Study and
Pilot Testing. Paper SPE 59362 presented at the SPE/DOE Improved
Oil Recovery Symposium, Tulsa, 3–5 April. doi: 10.2118/59362-MS.
Feng, Y., Tabary, R., Renard, M., Le Bon, C., Omari, A., and Chauveteau,
G. 2003. Characteristics of Microgels Designed for Water Shutoff
and Profile Control. Paper SPE 80203 presented at the Interna-
tional Symposium on Oilfield Chemistry, Houston, 5–7 February. doi:
10.2118/80203-MS.
Fulleylove, R.J., Morgan, J.C., Sevens, D.G., and Trasher, D.R. 1996.
Water Shut-Off in Oil Production Wells—Lessons from 12 Treatments.
Paper SPE 36211 presented at the Abu Dhabi International Petroleum
Exhibition and Conference, Abu Dhabi, UAE, 13–16 October. doi:
10.2118/36211-MS.
Horkay, F., Tasaki, I., and Basser, P.J. 2000. Osmotic Swelling of Polyac-
rylate Hydrogels in Physiological Salt Solutions. Biomacromolecules
1 (1): 84–90. doi: 10.1021/bm9905031.
November 2009 SPE Production & Operations
Maitland, G.C. 2000. Oil and gas production. Current Opinion in Col-
loid & Interface Science 5 (5–6): 301–311. doi: 10.1016/S1359-
0294(00)00069-8.
Médout-Marère, V., El Ghzaoui, A., Charnay, C., Douillard, J. M., Chauve-
teau, G., and Partyka, S. 2000. Surface Heterogeneity of Passively
Oxidized Silicon Carbide Particles: Hydrophobic-Hydrophilic Parti-
tion. Journal of Colloid and Interface Science 223 (2): 205–214. doi:
10.1006/jcis.1999.6625.
Molloy, P.J. and Cowling, M.J. 2000. Volume and density changes in
polymer gels in seawater environments. Proc IMechE, Part L: Journal
of Materials: Design and Applications 214 (4): 223–228.
Molloy, P.J., Smith, M.J., and Cowling, M.J. 2000. The effects of salinity
and temperature on the behaviour of polyacrylamide gels. Materials &
Design 21 (3): 169–174. doi: 10.1016/S0261-3069(99)00068-0.
Muruaga, E., Flores, M., Norman, C., and Romero, J. 2008. Combining
Bulk Gels and Colloidal Dispersion Gels for Improved Volumetric
Sweep Efficiency in a Mature Waterflood. Paper SPE 113334 presented
at the SPE/DOE Symposium on Improved Oil Recovery, Tulsa, 20–23
April. doi: 10.2118/113334-MS.
Nasr-El-Din, H.A and Hawkins, B.F. 1991. Recovery of Waterflood Resid-
ual Oil Using Alkali Surfactant and Polymer Slugs in Radial Cores. Oil
& Gas Science and Technology—Revue de l’IFP 46 (2): 199–219.
Rousseau, D., Chauveteau, G., Renard, M., Tabary, R., Zaitoun, A., Mallo,
P., Braun, O., and Omari, A. 2005. Rheology and Transport in Porous
Media of New Water Shutoff/Conformance Control Microgels. Paper
SPE 93254 presented at the SPE International Symposium on Oilfield
Chemistry, Houston, 2–4 February. doi: 10.2118/93254-MS.
Russel, W.B., Saville, D.A., and Schowalter, W.R. 1989. Colloidal Disper-
sions. Cambridge, UK: Cambridge University Press.
Seright, R.S. 2004. Conformance Improvement Using Gels. Annual Tech-
nical Progress Report No. DOE/BC/15316-6, Contract No. DE-FC26-
01BC15316, New Mexico Institute of Mining and Technology/US
DOE, Socorro, New Mexico (September 2004).
Seright, R.S. and Liang, J. 1994. A Survey of Field Applications of Gel
Treatments for Water Shutoff. Paper SPE 26991 presented at the SPE
Latin American and Caribbean Petroleum Engineering Conference,
Buenos Aires, 27–29 April. doi: 10.2118/26991-MS.
Seright, R.S., Lane, R.H., and Sydansk, R.D. 2003. A Strategy for Attack-
ing Excess Water Production. SPE Prod & Fac 18 (3): 158–169. SPE-
84966-PA. doi: 10.2118/84966-PA.
Sydansk, R.D. and Southwell, G.P. 2000. More than 12 Years’ Experience
With a Successful Conformance-Control Polymer-Gel Technology.
November 2009 SPE Production & Operations
SPE Prod & Fac 15 (4): 270–278. SPE-66558-PA. doi: 10.2118/
66558-PA.
Sydansk, R.D., Xiong, Y., Al-Dhafeeri, A., Schrader, R.J., and Seright, R.S.
2005. Characterization of Partially Formed Polymer Gels for Application
to Fractured Production Wells for Water-Shutoff Purposes. SPE Prod &
Fac 20 (3): 240–249. SPE-89401-PA. doi: 10.2118/89401-PA.
Whittington, L.E. and Naae, D.G. 1992. Conformance gels—Formation by
contact with brine. J. Pet. Sci. Eng. 7 (1–2): 45–51. doi: 10.1016/0920-
4105(92)90007-N.
Zaitoun, A., Kohler, N., Bossie-Codreanu, D., and Denys, K. 1999. Water
Shutoff by Relative Permeability Modifiers: Lessons From Several
Field Applications. Paper SPE 56740 presented at the SPE Annular
Technical Conference and Exhibition, Houston, 3–6 October. doi:
10.2118/56740-MS.
Zaitoun, A., Tabary, R., Rousseau, D., Pichery, T., Nouyoux, S., Mallo,
P., and Braun, O. 2007. Using Microgels to Shut Off Water in a Gas
Storage Well. Paper SPE 106042 presented at the SPE International
Symposium on Oilfield Chemistry, Houston, 28 February–2 March.
doi: 10.2118/106042-MS.
Céline Cozic is a licensing project manager in material science
for CNRS. Previously, she was a research engineer in the reser-
voir engineering department at the exploration and produc-
tion business unit of IFP in a postdoctoral position. Her research
interests include promotion and development. She holds a
PhD degree in polymer physical chemistry from University of
Rouen. David Rousseau is a research engineer in the reservoir
engineering department at the exploration and production
business unit of IFP. His research activities include evaluation
and development of well treatments and EOR technologies
based on associative polymers and polymeric gels, as well as
formation damage control during produced-water reinjection.
He holds a PhD degree in soft matter physics from the University
of Strasbourg. René Tabary is a senior research engineer in
the reservoir engineering department at the exploration and
production business unit of IFP. He is responsible for evaluating
EOR polymer floods processes. He has worked on water shutoff
and chemical EOR since 1976. His activity relates to upstream
research, development of new methods and processes, field
applications, and on-site technical assistance. He is the project
leader for the JIP “STARGEL,” dedicated to new IFP microgel
technology for water shutoff and conformance control. He
holds a four-year technical degree in chemistry, mathematics,
and physics from the University of Paris.
601