SPE

Society of Petrolern1 Engineers of ... IME

SPE 12168

An Experimental Investigation of Fines Migration

in Porous Media
by G.A. Gabriel* and G.R. Inamdar, Exxon Production Research Co.
'Member SPE-AIME

Copyright 1983 Society of Petroleum Engineers of AIME

This paper was presented at the 58th Annual Technical Conference and Exhibition held in San Francisco, CA, October 5-8, 1983. The material is subject
to correction by the author. Permission to copy is restricted to an abstract of not more than 300 words. Write SPE, 6200 North Central Expressway,
Drawer 64706, Dallas, Texas 75206 USA. Telex 730989 SPEDAL.

ABSTRACT and on this basis have devised remedial treatments

by which damage can be prevented or inhibited.
Permeability decline in argillaceous sandstone
format ions has tendered a broad spectrum of As reported by Vel ey (1969),1 the factors con-
explanations and related remedial treatments. This tributing to the binding of clay particles in
decline has often been attributed to the mobiliza- compact, oriented aggregations are van der Waals
tion, migration, and plugging of fine particles in forces, recrystallization and chemical alteration,
the formation pore spaces. While previous inves- sorption of organic matter from oil, mutual
tigations have generally been directed toward sorption of ions between adjacent unit layers,
either a chemical or a mechanical characterization electrostatic attractions between positively
of the damage mechanism, this experimental study charged edges (at neutral or lower pH) and neg-
was undertaken to comprehensively analyze both the atively charged faces, and the thermodynamic drive
chemical and the mechanical interactions. - - to reduce interfaci al free energy by reduci ng
surface area. Opposing these considerations are
In linear core tests, the flow of a chemically the factors contributing to clay particle expansion
compatible, wetting fluid resulted in severe and dispersion such as hyd~ation of exchangeable
permeability loss when the fluid velocity exceeded cations, hydration of particle surfaces, repulSion
a critical value. The flow of a chemically of interacting atmospheres of exchangeable cations
incompatible, wetting fluid resulted in a total (double layer theory), desorption or chemical
loss of permeability which exhibited no dependency removal of sorbed binding matter, neutralization of
on fluid velocity or volumetric throughput. positive charges on particle edges, mechanical
Pretreatment of the core with a cationic polymer shear, and thermal (Brownian) motions.
(clay stabilizer) prevented permeability damage due
to chemical incompatibility. However, a cationic Of all these effects, Veley maintained that the
polymer pretreatment could not protect the core relative expansion and compression of the afore-
against damage caused by exceeding the critical mentioned cation atmosphere was critically impor-
velocity. tant in determining whether clay particles would
disperse. As noted by Reed (1972),2 the cation
lNTRODUCTION atmosphere develops as a result of the inherent
negative charge on almost all clay minerals due to
For several years fines migration has been recog- substitution of A1+ 3 for Si+ 4 in the tetrahedron
nized as a source of permeability damage and sheet and Mg+2 or Fe+2 for A1+3 in the octahedron
productivity decline in both consolidated and sheet. The negatively charged lattice is neutral-
unconsolidated formatiolls. Investigations of the ized by cation adsorption on the clay surfaces.
fines damage mechanism have basically originated However, the cations dissociate in solution
from two different perspectives -- chemical versus establishing an atmosphere of positively charged
mechanical mobilization, migration, and plugging. ions near the surface of the negatively charged
particle. This is the aforementioned electric
Chemical Mechanism: Laboratory work in the area of double layer and particles exhibiting this ionic
flnes mlgratlon induced by chemical interactions configuration repel one another.
has historically involved demonstrat ions of rapid
and drastic permeability decline resulting from Attraction of cations toward the particle (i .e .•
fresh water contact of clay-containing formations. compression of the double layer) is a function of
Several investigators have described the mechanism the particle's total charge density and the
by which clays expand, disperse, migrate, and plug, surrounding cations' total effective charge. The
diffusion forces are strongly dependent on the
References and illustrations at end of paper. concentration of ions in the bulk solution. Free
ion repulSion counteracts exchangeable cation
2 AN EXPERIMENTAL INVESTIGATION OF FINES MIGRATION IN POROUS MEDIA SPE 12168
repulsion and tends to compress the cation atmos- Prevention or control of permeability damage
phere. Consequent ly, the atmosphere wi 11 be induced by mechanical flow forces has not been
diffuse in distilled water and compressed in a specifically addressed by the industry. However,
solution having a high electrolytic concentration. alternative treatments to clay stabilizers have
The reduced dissociative tendency of higher valence been devised which are loosely based on a mechan-
cations also decreases the repulsive strength of ical repair mechanism. These treatments (e.g.,
the double layer and promotes compression of the asphaltene/resin adsorption,16 oil emulsion/
cation atmosphere. Counteracting double layer dispersion coating,17 etc.) are designed to change
repulsion and clay platelet dispersion can there- particle wettability or to adhere particles to pore
fore be accomplished in one of two ways: 1) by walls.
maintaining a bulk solution brine concentration
higher than the clay matrix brine concentration or RESEARCH OBJECTIVE
2) by decreasing the repulsive strength of the
cation cloud through exchange of interlayer cations The research program was designed to accomplish
with higher valence substituents. five specific objectives: 1) to develop a reliable
experimental scheme (equipment and procedures) to
Preventative treatments have been developed to study fines-related permeability decline, 2) to
address this second objective. that is, stabiliza- establish the existence of both a mechanical and a
tion of the cation atmosphere. The first stabil- chemi ca 1 damage mech an ism, 3) to de 1i neate th e
ization chemicals were inorganic salts such as characteristics of each damage mechanism, 4) to
aluminum chloride 2- 7 and zirconium oxychloride 1 ,8,9 analyze the interrelationship of mechanical and
which under controlled pH conditions underwent chemical fines damage, and 5) to assess the
hydration and/or hydrolysis to form large, multi- effectiveness of currently available preventative
valent, cationic polymers. These reportedly treatments.
adsorbed onto the anionic clay particles and
greatly compressed the double layer. The advantage It is important to note that even though the intro-
of multivalent polymeric cations over simple ductory literature review was outlined per the dual
divalent cations such as Ca+ 2 is their relative damage mechanism approach, such a distinction has
permanency. Though Ca+ 2 reduces the repulsive not been clearly identified much less characterized
strength of the double layer, it can be readily in the past. The primary motivation for this
exchanged with monovalent cations resulting in a distinction is the work of ~uecke (1979)14 and
clay structure which is highly susceptible to Gruesbeck and Collins (1982).1 5 These inves-
dispersion. The relative tenacity of the polymeric tigators demonstrated that it is not valid to
inorganic salts in clinging to particle surfaces simply equate fines migration with clay desta-
has been attributed to matching of their polymeric bilization because siliceous particulates can also
structure with the octahedral sh eet in some clays be mobilized. As a result, remedial treatments
or chemical bonding via silicon-oxygen or silicon- with conventional clay stabilization chemicals may
hydroxyl linkages. The more recently developed not be applicable to a general fines migration
organic cationic polymers 10 - 13 reportedly behave in phenomenon.
the same manner as the inorganic polymers but are
insensitive to pH. EQUIPMENT AND MATERIALS
Mechanical Mechanism: Fines migration damage Some of the most critical considerations of this
resulting from mechanical flow forces has not been experimental study involved equipment specifica-
as extensively studied as that caused by chemical tions. Equipment performance characteristics were
interactions. The majority of the work in this continually evaluated and re-evaluated to maximize
area is detailed in the investigations of Muecke confidence in the experimental results and to
(1979)14 and Gruesbeck and Collins (1982).15 As permit critical assessment of their reproduc-
identified in these studies, migrating particulates ibility. A discussion of the primary components of
span the compositional range from siliceous to the experimental system. represented schematically
argillaceous. Consequently, remedial treatments in Figure 1, will elucidate the importance of
must be capable of immobilizing both types of equipment selection.
particulates. In micromodel studies, Muecke iden-
tified the importance of particle wettability and Core Samples: Most of the flow tests were per-
surface/interfacial forces in determining particle formed on Berea sandstone cores whose mineralogical
mobility. It was concluded that particles will composition is characterized in Table 1. A few
move only if the phase that wets them is mobile. In tests were performed on unconsolidated 80/140 U.S.
addition, it was observed that particle bridges mesh (104-177 11m) sand packs "spiked" with a small
which formed at pore throat constrictions could be percentage of <400 U.S. mesh «37 11m) particles.
readily disrupted by pressure disturbances and/or All cores were 10.2 cm (4 in.) in length and 2.54
flow reversals. Permanent bridge formation was cm (1 in.) in diameter. The cores were confined in
greatly reduced in multiphase flow because local- a standard Hassler cell with a net overburden
ized pressure disturbances maintained a state of pressure in the 2.76-6.89 MPa (400-1000 psi) range.
fluid-particle agitation. As a follow-up to this
work, Gruesbeck and Collins identified a critical Berea sandstone core material was chosen due to the
velocity or minimum interstitial fluid velocity for relative homogeneity of its properties from lot to
fines entrainment. This velocity was determined to lot. This was critically important in that perme-
be highly dependent upon the properties of the ability damage due to particulate plugging is a
porous medium and the interstitial fluids. strong function of initial core permeability. Over
SPE 12168 G. A. GABRIEL and G. R. INAMDAR 3
the course of the two year research program, particles suspended in any carrier fluid which
quality control of the core permeability was main- could transmit collimated light across a 1-mm gap.
tained at 0.15±0.05 ~m2 (150±50 md). This not only allowed for more rapid sample
analysis but also eliminated some of the inherent
An attempt was made to e1 iminate some of the analytical inaccuracies of other equipment types.
inherent uncertainties regarding consolidated core For instance, some instruments require that
properties by constructing unconsolidated sand particles be transferred from the test carrier to a
packs comprised of a knuwn amount of fine par- special electrolytic medium, while others require
ticles. Several concerns surfaced with this 1 arger s amp 1e vo 1urnes or higher samp1 e concentra-
approach. Firstly, the packing process itself tions than were available from these tests.
generated an unknown quantity of additional fines
as a result of particle-particle abrasion. Fl uids: The majority of the flow tests were
Secondly, there was no guarantee that all of the performed with distilled water or KC1 brine
fine particulates initially implanted in the pack solutions at room temperature. In addition, tests
would be capable of being mobilized. Thirdly', the were conducted with CaC12 brines, NaC1 brines,
sand packs that were constructed contained 109-lOll cationic polymer-brine solutions, oil-water
fine particles. Analytical errors of only 1% would mixtures, and a light refined oil (Isopar M®). All
amount to 10 7-10 9 particles which was comparable to fluids were initially filtered to 0.2 ~m.
the total number of effluent particles during a
typical flow test. Therefore, it was concluded that The particle size analyzer indicated that the fil-
no further statistical certainty could be gained by tered (to 0.2 ~m) inlet fluids registered 200-400
using unconso1 idated sand packs as an alternative partic1es/cm 3 in the 1-10 ~m range with a mean
to consolidated Berea sandstone cores. diameter of approximately 2.5 ~m.
Pumps: In order to accurately characterize the Backpressure Regulator and Pressure Transducer:
~t of a given velocity on particle entrainment The performance characterlstlcs of backpressure
and plugging, it was imperative that f10wrates be regulators and pressure transducers necessitated
maintained constant during a given test. Past critical evaluation in the early stages of this
studies had indicated that pressure di sturbances -study. A series of flow tests was run to determine
caused by flowrate fluctuations could have a if the experimental results were influenced by the
significant effect on the construction and destruc- presence of a backpressure regulator and transducer
tion of particle bridges. A constant f10wrate was in the flow stream. Both of these pi eces of
accomp1 ished by means of four, high pressure, equipment have various crevices and corners upon
liquid chromatographic pumps. These provided a which particles may be deposited and later
f10wrate range of 0-1.9 cm 3/s which corresponded to released, thus skewing the resulting data.
a superfici~ velocity range of 0-0.37 cm/s for a
2.54-cm-diameter core. Velocities of this mag- A backpressure regulator was initially assumed to
nitude are typical in the near-we11bore region of a be necessary to obtain an accurate, stable measure-
perforated completion. ment of liquid permeability. It was determined
that complete removal of core gas without a
Piston Cylinders: Further pulse dampening was continuous backpressure could be ensured through a
achleved by employing large volume (2000-4000 cm 3) , series of air evacuating/C02 flooding/C02 evacu-
high pressure, piston cylinders to transfer fluid ating/C02 flooding operations. The liquid was
to the core. The system was arranged in such a way subsequently introduced under approximately 6.89
that the pumps injected a constant rate of filtered MPa (1000 psi) of pressure which was released as
oil to the cylinders which displaced a constant the test began. In tests for which a determination
rate of filtered fluid to the core. The fluids of an absolute value of permeability was not con-
were separated in the cylinder by a Tef10n® piston. sidered critical, simple core evacuation provided
This configuration also protected the pump heads the same qual itative results as the C02 flood and
from the corrosive water and/or brine. evacuation. Consequently, the particle-entraining
regulator was removed.
From early flow tests it was discovered that a
round trip of the piston up and down the cylinder The pressure transducer in the flow stream posed a
depOSited a very thin film of oil on the stainless more difficult problem. Fluid retention in the
steel wall. This film would become dispersed in the transducer following each run led to the formation
aqueous core fluid of a subsequent flow test and of bacteria colonies or corrosion products. If the
would be counted as fines by the particle analyzer. system experienced a pressure surge (e.g., trans-
Consequently, prior to filling the cylinders for ducer zeroing), these particulates would be
each test, they were flushed with a volatile, transmitted to the samp1 eli ne. To control the
organic solvent, dried, and rinsed with filtered problem of pressure and particle surges, zeroing
aqueous fl uid. frequency was minimized and/or effluent samples
were collected only after the particle pulse had
Particle Size Analyzer: An optical particle size subsided.
analyzer was employed for determining the particle
concentration and size distribution of core RESULTS AND DISCUSSION
influent and effluent samples. The analyzer,
operating on the principle of light blockage, was The following experimental results were observed
capable of detecting particles down to 1 ~m in from a series of linear flow tests on Berea
size. The advantage of the part i c u1 ar ins t rumen t sandstone cores (2.54-cm diameter, 10.2-cm length).
used in this study was its capability of analyzing Though a damage mechanism will also be proposed,
4 AN EXPERIMENTAL INVESTIGATION OF FINES MIGRATION IN POROUS MEDIA SPE 12168
the primary contribution of this study is the velocity was changed. However. the effluent
empirical characterization of the fines migration particle concentration quickly declined to a value
phenomenon. at or below the inlet background count (",400
particles/cm 3 ) within a couple of pore volumes of
One of the major results of this study was an continued flow. The particle surge was observed at
identification of the complexity of the fines both damaging and non-damaging velocities. and its
migration phenomenon. For this reason. perme- magnitude did not appear to be a function of flow
ability declines are presented as a range of values velocity or direction. The size distribution of
applicable only to the given initial conditions. In particles in the effluent stream was also indepen-
addition. only the trends in the effluent particle dent of flow conditions. remaining essentially
characteristics can be considered statistically unchanged (mean diameter of 2.0-2.5 ~m; 98% less
significa~t. Furthermore. extreme caution should than 10 ~m) with variations in velocity above and
be exercised in extrapolating the results of these below the critical value.
laboratory. linear. core flow tests on Berea
sandstone cores to wellbore configurations with The above resul ts suggest that particles are
different formation material. mobilized at all flow velocities of a chemically
compatible. wetting fluid. At non-damaging veloc-
Fines Migration - Mechanical ities. particles are either not mobilized in
sufficient concentrations to cause pore throat
Figure 2 illustrates the results of a series of bridging and constriction or else particle bridges
flow tests with 150 md (0.15 ~m2) Berea sandstone are not tight enough to effect a macroscopic change
cores initially saturated with 2% KC1. Following in permeability. At damaging velocities. particle
saturation. the cores were flowed with 2% KCl at a concentrations are high enough and/or particle
superficial velocity of less than or equal to 0.007 bridges tight enough to cause a loss of core
cm/s. At these flowing conditions. pressure drop permeability. Constant permeability decline at
across the core was constant for the determination flow velocities above the critical value with only
of initial permeability. When the velocity was an initial. effluent particle surge may be indica-
maintained at a value less than or equal to tive of constant pore wall scouring. particle
0.007 cm/s. no permeability decline was experienced mobilization. and bridge formation. or may indicate
for the duration of the test. usually 600 PV (pore initial particle release followed by continuous
volumes) of cumul ative flow. 1~ other words. the construction and compression of a pore throat
pressure drop across the core remained constant particle cake. The extent of damage characteri zed
throughout the flow period. When the test was by both mechanisms would be time dependent; thus.
conducted at a velocity above 0.007 cm/s. perme- both mechani sms coul d exp 1 a in why damage is more
ability declined throughout the flow period. If severe at velocities slightly in excess of critical
monitored at a damaging velocity. pressure drop for constant. cumulative. throughput volumes.
steadily increased and permeability correspondingly
decreased. If permeability was monitored at a The experimental results in Figure 6 demonstrate
non-damaging velocity after a flow period at a that flow-induced permeability decline can be
damaging velocity. its val ue stabil i zed but at a partially restored by flow reversal. Permeability
level less than the initial permeability. Higher damage did not recur as long as the velocity was
permeability core material was less susceptible to maintained below its critical value. These results
this mechanical flow damage. suggest that flow reversal may partially restore
permeability by loosening or breaking particle
These tests were conducted over a range of super- bridges. At non-damaging velocities. particles may
ficial velocities to determine the extent of settle out in the pore spaces or re-form loose
permeability decli~e. Cores were subjected to bridges at pore throat constrictio~s. At damaging
damaging velocities for 600 PV of cumulative velocities. additional particle entrainment along
throughput after which the velocity was decreased with bridge readjustment would explain both initial
to a non-damaging value to determine final perme- permeability increase and long-term permeability
ability. As illustrated in Figure 3. permeability decrease.
decline was 20-40% for superficial velocities
ranging from 0.007 cm/s to 0.4 cm/s. The higher end As reported in Figure 7. no critical velocity was
of the decline range tended to apply to velocities established for the flow of a chemically compat-
just in excess of the critical velocity while the ible. non-wetting fluid (oil) at connate water. At
lower end of the range applied to velocities a superficial flow velocities of 0.4 cm/s and flow
couple of orders of magnitude above critical. The volumes in excess of 5000 PV. no permeability loss
flow period at higher velocities was of course was experienced by the core. ihese results confirm
significantly shorter. and. as expected. the rate the behavior observed by Muecke (1979)14 in
of permeabil ity loss (monitored at the damaging micromodel studies. At all flow velocities of a
velocity) was sharper for the higher velocities. non-wetting fluid. particle confinement in the
connate water film precluded particle migration and
Figures 4 and 5 illustrate effluent particle plugging.
characteristics as a function of volume and
velocity for tests conducted with 2% KC1. Even Fines Migration - Chemical
though there were no visible particulates in the
effluent stream. the particle analyzer indicated Flow tests similar to the previous series were
that the effluent fines concentration surged from employed to characterize fines migration damage
0-100 particles/cm 3 (>1 ~m) to 104-10 6 particles/ resulting from chemical interactions. However.
cm 3 (>1 ~m) when flow was initiated and when fluid fresh water was used as the primary flow medium
instead of 2% KC1.
SPE 12168 G. A. GABRIEL and G. R. INAMDAR 5
The results of the fresh water flow tests were could be readily accomplished with small flow
highly dependent on the previous saturation and volumes of potassium chloride or sodium chloride
flow history of the core. The experimental solutions. From the standpoints of cation exchange
observations reported in Figure 8 demonstrate that capacity, chemical equi 1 ibri um, and double 1ayer
fresh water flow caused severe and abrupt perme- theory, these results are not surprising.
ab il ity damage (>95% los s) when the core was
previously contacted with either potassium chloride The severe permeability damage incurred by the
or sodium chloride. There was no apparent velocity contact of argi 11 aceous sandstone formations with
dependency under these conditions. Higher perme- fresh water has been previously studijd and
ability cores (600 md) were also susceptible to described by several investigators.1- 13 , 8 The
this severe damage, but the decline rate was not as above results impact this previous work by high-
rapid as with the lower permeability cores (150 lighting the marked differences between chemical
md). Severe permeability decline from chemical and mechanical damage. This was a critical
interaction was characterized by a visibly cloudy component for the interpretation of the following
discharge of fluid from the core. results on the effectiveness of remedial treat-
ments.
Figure 8 also illustrates that fresh water flow
below a superficial velocity of 0.007 cm/s caused Fines Migration - Damage Treatment
no permeability damage when the core was previously
contacted with either calcium chloride or fresh Virtually the same flow tests as those used to
water. At velocities in excess of 0.007 cm/s, study chemical incompatibility were employed to
20-40% permeability decline was observed after investigate the effectiveness of preventative
600 PV of fresh water flow following a calcium treatments. The one difference was that a volume
chloride or fresh water saturation. This was the of 2% KCl with a clay stabilizer was flowed through
same critical velocity that was identifi ed for KCl the core following KCl saturation and prior to the
as the flowing fluid (Figure 3). flow of fresh water. Only the cationic organic
polymers were tested in this study. No evaluation
Figure 9 illustrates that the severe permeability of the inorganic, polymeric clay stabilizers was
decline associated with fresh water flow following conducted. Treatment concentrations for the
KCl or NaCl contact could be halted by a change to organic clay stabilizers ranged from 0.5-2.0% (by
KCl flow. Further damage was then controlled by volume) in 2% KC1, and treatment volumes were
the critical velocity dependency. Permeability typically 2-10 PV. Over these ranges, treatment
steadily decl ined at a brine velocity above effectiveness was not observed to be a function of
0.007 cm/s, but decline was arrested at a velocity concentration and volume.
below 0.007 cm/s. Brine flow did not restore
permeabil ity. The results reported in Figures 12 and 13 illus-
trate that treatments of clay stabilizers prevented
As illustrated in Figure 10, flow reversal with the drastic permeability damage resulting from
brine following severe damage restored approx- fresh water contact of sensitized Berea sandstone
imately half of the lost permeability. Damage did cores. Clay stabilizers did not protect Berea
not recur as long as the brine velocity was cores against damage caused by exceeding the
maintained below the critical value of 0.007 cm/s. critical velocity of 0.007 cm/s with either a
If the velocity exceeded 0.007 cm/s, mechanical chemically compatible or a chemically incompatible
flow damage ensued. fl uid.
The final points of investigation in the area of Berea sandstone cores treated with a clay stabi-
chemical damage concerned core sensitization and lizer were not re-sensitized to fresh water contact
minimum sal inity criteria. As reported in even after a thousand pore volumes of flow with
Figure 11, Berea cores that were insensitive to fresh water, potassium chloride, sodium chloride,
fresh water flow (i.e .• those previously contacted or calcium chloride over a 0-0.4 cm/s range of
by only calcium chloride or fresh water) could be fluid velocities. Note, however, that any flow
readily sensitized by injection into the core of velocity in excess of 0.007 cm/s through a treated
less than a pore volume of potassium chloride or core always resulted in a 20-40% permeability
sodium chloride. Regarding minimum salinity decline indicative of mechanical flow damage.
requirements, Berea sandstone cores saturated with
2% KCl exhibited the characteristics of "fresh Core pretreatment with a cl ay stabil i zer is
water" sensitivity when the salinity of the flow apparently effective for negating the fresh water
medium dropped below approximately 0.1% KCl sensitivity of argillaceous sandstone cores. The
(1000 ppm). mechanism proposed by Veley (1969)1 and Reed
(1972)2 can expl ain this phenomenon. Calcium
This latter observation was the basis for one of chloride pretreatment is equally effective except
the major insights gained from this study, that calcium ions can be readily exchanged by potassium
being a clearer understanding of the subject of or sodium ions which re-sensitize the core to fresh
chemical incompatibility. Core sensitivity water attack. The polymeric cations of clay
seemingly cannot be defined by an absolute minimum stabilizers are nct readily exchangeable.
salinity criterion for the flow medium. Rather, the
relative change in fluid salinity is the critical Cl ay stabil izers are apparently not effective for
parameter. Consequently, if the salinity of the controlling permeability damage caused by exceeding
flow medium is comparable to the salinity of the the critical flow velocity. This was not unexpected
saturation fluid, no chemical incompatibility by virtue of the reported repair mechanism of clay
problems are anticipated. However, it was also stabilizers. For treated Berea cores contacted by
observed that sensitization of the core material brine or fresh water and for untreated Berea cores
 6 AN EXPERIMENTAL INVESTIGATION OF FINES MIGRATION IN POROUS MEDIA SPE 12168
contacted by brine, exceeding the critical flow perforated completion with wet production. In this
velocity resulted in 20-40% permeability decline environment, permeability restoration could likely
for 600 PV of fluid throughput. This damage is be accomp 1i shed through the app 1i cat i on of an
associated with the mechanical mobilization of fine appropriately diverted, matrix acidizing treatment.
particles. For other types of core material and
flow configurations, the extent of mechanical NOMENCLATURE
permeability damage will be a function of flow
area, direction, and velocity as well as core q = flowrate, cm 3/s
permeability and wettability. v = superficial velocity, cm/s
Vc = critical superficial velocity, cm/s
LABORATORY CONCLUSIONS (vc = 0.007 cm/s for 150 md Berea
sandstone cores.)
From this experimental investigation it can be con-
cluded that: ACKNOWLEDGMENTS
1) The permeability of porous media can be The authors thank Exxon Production Research Co. for
damaged by the migration and plugging of fine permission to publish this paper.
particulates native to the formation. These
particles can be mobilized by either a REFERENCES
mechanical or a chemical mechanism.
1. Veley, C. D.: "How Hydrolyzable Metal Ions
2) Mechanical particle mobilization occurs when a React with Clays to Control Formation Water
critical flow velocity is exceeded. For 0.15 Sensitivity," J. Pet. Tech. (Sept. 1969)
~m2 (150 md) Berea sandstone, the critical 1111-1118.
superficial velocity was 0.007 cm/s. The
extent of permeability decline is a function 2. Reed, M. G.: "Stabi lization of Formation
of flow area, direction, and velocity as well Clays with Hydroxy-Aluminum Solutions," J.
as core permeability and wettability. Pet. Tech. (July 1972) 860-864.
3) Chemical particle mobilization occurs when the 3. Coppel, C. P., Jennings, H. Y., and Reed, M.
native clays in the formation matrix are con- G.: "Field Results from Wells Treated with
tacted by a chemically incompatible fluid. Hydroxy-Aluminum," J. Pet. Tech. (Sept. 1973)
This results in a high concentration disp~r­ 1108-1112.
sion of particulates whose bridging leads to
sudden and virtually total permeability loss. 4. Blevins, T. R., Cotten, W. R., and Dugas, E.
J.: "OH-Al Treatments Sustain Acid Stimulated
4) A clay stabilizer (cationic polymer) pretreat- Production," World Oi 1 (Aug. 1973) 26-29.
ment prevented permeability damage due to
chemical incompatibility. A clay stabilizer 5. Haskin, C. A., Reed, M. G., and Coppel, C. P.:
pretreatment did not protect the core against "Here's When and How to Use Hydroxy-Aluminum
permeability damage caused by exceeding the Well Treatment," Oi 1 and Gas J. (Nov. 1975)
critical velocity. 217-220.
FIELD IMPLICATIONS 6. Reed, M. G. and Coppel, C. P.: "Sand Stabi-
lization with Hydroxy Aluminum Solutions,"
The results of this study suggest that the use of paper SPE 4186 presented at the SPE-AIME 43rd
clay stabilizers (cationic organic polymers) in the Annual California Regional Fall Meeting,
field should probably be limited to those situa- Bakersfield, Nov. 8-10, 1972.
tions in which a chemical incompatibility problem
has been specifically identified. In this applica- 7. Haskin, C. A.: "A Review of Hydroxy-Aluminum
tion, laboratory tests have indicated that clay Treatments," paper SPE 5692 presented at the
stabilizers can effectively protect the formation SPE-AIME Symposium on Formation Damage
against severe permeability damage caused by abrupt Control, Houston, Jan. 29-30, 1976.
salinity decreases. Once deposited, these stabi-
lizers are not readily exchanged by inorganic 8. Peters, F. W. and Stout, C. M.: "Clay
cations. However, it should be noted that the Stabilization during Fracturing Treatments
laboratory tests were performed at room tempera- with Hydrolyzable Zirconium Salts," J. Pet.
ture. Clay stabilizer performance characteristics Tech. (Feb. 1977) 187-194.
should be checked out at reservoir conditions prior
to treatment. 9. Thomas, R. L., Crowe, C. W., and Simpson,
B. E.: "Effect of Chemi cal Treatment upon
No preventative technique for controlling the Formation Clays Is Revealed by Improved SEM
mechanical mobilization of formation particulates Technique," paper SPE 6007 presented at the
was identified in this study. Damage from this SPE-AIME 51st Annual Fall Technical Conference
source is probably negligible in relatively high and Exhibition, New Orleans, Oct. 3-6, 1976.
permeability (>500 md) formations, dry hydrocarbon
reservoirs, and/or large flow area completions 10. Young, W. M., Mclaughlin, H. C., and
(open hole). The magnitude of the critical Borchardt, J. K.: "Clay Stabilization Agents
velocity observed in this study seems to confine - Their Effectiveness in High-Temperature
the problem of mechanical fines mobilization and Steam," J. Pet. Tech. (Dec. 1980) 2121-2131.
plugging to the very near-wellbore region of a
SPE 12168 G. A. GABRIEL and G. R. INAMDAR 7

11. Coulter, A. W., Copeland, C. T., and Harris- 15. Gruesbeck, C. and Collins, R. E.: "Entrain-
berger, W. H.: "A Laboratory Study of Clay ment and Deposition of Fine Particles in
Stabilizers," Soc. Pet. Eng. J. {Oct. 1979} Porous Media," Soc. Pet. Eng. J. {Dec. 1982}
267-269. 847-856.

12. McLaughl in, H. C., El phingstone, E. A., and 16. Clementz, D. M.: "Cl ay Stabil ization in Sand-
Hall, R. E.: "Aqueous Polymers for Treating stones through Adsorption of Petrol eum Heavy
Cl ays in Oil and Gas Producing Formations," Ends," J. Pet. Tech. {Sept. 1977} 1061-1066.
paper SPE 6008 presented at the SPE-AIME 51st
Annual Fall Technical Conference and Exhibi- 17. Barkman, J. H., Abrams, A., Darley, H. C. H.,
tion,· New Orleans, Oct. 3-6, 1976. and Hill, H•.J.: "An Oil Coating Process to
Stabilize Clays in Fresh Water Flooding
13. Copeland, C. 1., Coulter, A. W., and Harris- Operations," paper SPE 4786 presented at the
berger, W. H.: "Designed Appl ication of Cl ay SPE-AIME Symposium on Formation Damage
Stabilizer Improves Performance," paper SPE Control, New Orleans, Feb. 7-8, 1974.
6759 presented at the SPE-AIME 52nd Annual
Fall Technical Conference and Exhibition, 18. Khilar, K. C. and Fogler, H. S.: "Water
Denver, Oct. 9-12, 1977. Sensitivity of Sandstones," Soc. Pet. Eng. J.
{Feb. 1983} 55-64.
14. Muecke, T. W.: "Formation Fines and Factors
Controlling Their Movement in Porous Media,"
J. Pet. Tech. {Feb. 1979} 144-150.

TABLE 1 - COMPOSITION OF BEREA SANDSTONE CORES

Bulk Fines
Core
-- ( <5].lm)

Quartz 86% 51%

Feldspar 5 11
Dolomite 1 --
Siderite 1 1
III ite 4 15
Kaolinite 3 20
Chlorite -- 2
-- --
Total 100% 100%
 R.OATING PISTON
CYUNDERS
HIGH PRESSURE
PROPORTIONING
PUMPS

= 0 - 1.9 cm 3/s
q
-+-+--+-,v = 0 - 0.37 cm/s

1" DIAMETER CORE

HASSLER CEll

PARTICLE SIZE
ANALYZER

TRANSDUCER~ECORDER
SYSTEM

Fig. 1-Experimental apparatus.

100 r"-, I I
'-.
.....................

---------------
'--- ~~--------------------~-

-
< Vc
---------- V

----v>V c
-

o L -__________ L - I_ _ _ _ _ _ _ _ _ _L - I_ _ _ _ _ _ _ _- - '

0200400
CUMULATIVE PORE VOLUMES

Fig. 2-Permeability decline stabilizes at velocities below the critical value.

 _ 100 F=--=-=-=-=-=-=-=-"""=====r1~---------'
53
~
t::i,....111
C::

!§~
2
Ll.v Ll.v
'-----j b--
~~60
~Ll.V -------
:;;
lOO~-.~~,--------,--------,-----~
~ 1--___~v~<:...:cVc'----_ _ _ _ _+-_______v~_v_"c_ _ _ _ _--j

10'
I
I
I
I
I
I
I
I
I
I INPUT lEVEl
I
I
I
I IOl I~
I
I CUMULATIVE PORE VOLUMES
I
I
j--'-V C = 0.007 Fig. 4-Particle concentration surges with changes in the fluid velocity.

0.1 01 0.3 0.4

SUPERFICIAL VElOCITY, cm/s I
100 ' .
...........

Fig. 3-Srine flow above the critical velocity causes permeability damage.
............ --- -- ---__
1---- -
I __

----.J - - __

I
flOW
FORWARO flOW -
0.007 cm/s REVERSAL

f.----...r- v<
r~
Vc

"-
-
----V>Vc
r- t---.. t-- -
r-

FORWARO flOW OL-________~I__________~~I--------~

0.06 cm/s o ~ ~ IOl
CUMULATIVE PORE VOLUMES

Fig. 6-Flow reversal partially restores permeability.

I-
-- ............
i'-.t--..
-

1-1-1--

REVERSE flOW
006 cm/s
,....211

; I
I
I
I
2-10 I--- --............ - I

~
I

i
I
I-- ........... I
I
t- ~ I
I
~ 0 -r- I
I
I
I 2 4 6 7 8 9 10 211 I
I
SIZE, pm
1
I
I
I
i-- Vc
I
= 0.007
I
Fig. 5-Effluent particle size distribution is independent of flow velocity and direction. I
I

01 0.4
SUPERFICIAL VElOCITY, cm/s

Fig. 7-0il flow at connate water does not cause permeability damage.
 100====~----'-------~'-------~'-------~

80
I-
Z
LU
t..J
a:
~
goo
z
~
a:
~40
==
~ - - - CaCl 2 OR H20 SATURATION; H20 FLOW
:E
ffi ------ KCI OR NaCI SATURATION; H20 FLOW
£l..20

= 0.007
\,
OL-~~==~~~====~==-===-===-==~-===-==~·~~~
02 0.3 0.4
SUPERFICIAL VELOCITY, cm/s

Fig. a-Permeability damage from fresh water flow depends on initial saturation.

100 - - v-
A
I I
KCl OR NaCI \
SATURATION .
-
\ -
\ V< Vc
\ ---V> Vc
- \ -
\
\
-

- ----------
-
H2 O
flOW
" ' ........
........
- -
BRINE flOW

o I I
o 2 3 200 400 600
CUMULATIVE PORE VOLUMES

Fig. 9-Fresh water flow damage is arrested by brine flow at a controlled velocity.
 l00~--------~I----------~I~--------~

- - - - Hz 0 FLOW IANY vI
I- 80 t- -
as ___ BRINE FLOW
~
l:t. ,
COOL. -
~ I FLOW REVERSAL
~
5 ~~
i !r~"~._____
-----------------v-<~vc----~-----4_
~ i . . . . . . . . . . . __
~~ 20 ~, "

i,
--------
v> Vc --_.::

\..~-
DL-~--------~----------~----------~
1 I
0200400 600
CUMULATIVE PORE VOLUMES

Fig. lO-Flow reversal with brine partially restores fresh water flow damage.

l00~----------",----------------------~
' ..... v<v ,
..... __

V> ---
c _

vc -
I+-KCI OR NaCI INJECTION 1< 1 PVI

I~
I
-

i
! -
I
',eaozOR HzO SATURATION; HzO FLOW

D~---------L-I\~~~~=-~-=-~-=-~-~-=====-~-=-~~__~
0200400 600
CUMULATIVE PORE VOLUMES

Fig. ll-KCI or NaCI injection can readily sensitize cores to fresh water flow.
 100 KCI-OR NaCI~ CLAY T',. . . . . - - -v < vc
SATURATION STABILIZERJ ' ' ' __

-
I-

i:5 00 I- REATMEN ---- _ -

~
(..) v>v c - -__
H20 FLOW
-

I I

10 20 200 400
CUMULATIVE PORE VOLUMES

Fig. 12-Pretreatment with a clay stabilizer prevents only chemical incompatibility damage.

100 -l~ I I T

'"\
----
I,
1\
,
\,
'------- --- ---- _----

(1) KG! OR NaCI SATURATION

I (2) CLAY STABILIZER PRETREATMENT -
I
I
I
(3) H2 0 FLOW
I
,
I
201- I -
,
I

I
rvc = D.DDl
,
I
I I I
D~~~----~--------L---------~------~
0.1 Dl 0.3 0.4
SUPERFICIAL VELOCITY, cm/s

Fig. 13-Pretreatment with a clay stabilizer does not prevent mechanical flow
damage.