SPE 151774

Productivity Impairment of Gas Wells Due to Fines Migration

A. Zeinijahromi, The University of Adelaide; A. Vaz, North Fluminense State University; P. Bedrikovetsky,
The University of Adelaide

Copyright 2012, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE International Symposium and Exhibition on Formation Damage Control held in Lafayette, Louisiana, USA, 15–17 February 2012.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to
reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract

Well productivity decline have been widely observed for gas wells producing the reservoir fines. The phenomenon has been
explained by the lifting, migration and subsequent plugging of the pores by the fine particles, finally resulting in permeability
decrease. It has been observed in numerous core flood tests and field cases. The new basic equations for the detachment of fine
particles, their migration and size exclusion, causing the rock permeability decline during gas production, have been derived.
The analytical model, developed for the regime of steady state gas production with a gradual accumulation of strained
particles, exhibits the linear skin factor growth vs the amount of produced reservoir fines. The modeling results are in a good
agreement with the well production history. The model predicts well productivity decline due to fines production based on the
short term production data.

1. Introduction

Extremely low-permeable gas reservoirs currently form a majority in the current exploitation market for unconventional gas
production including shale gas reservoirs, tight gas sands and gas producing coal beds. Well productivity decline is the major
problem during exploitation of low-permeable unconventional gas fields.
Oil well productivity decline under fines production is a well known phenomenon in the low consolidated and high
clay content reservoirs, as well as in the heavy oil and high rate gas fields (Mungan, 1965, Lever and Dawe, 1984, Tiab et al,
2004, Civan, 2007). This decline is due to detachment of the in-situ particles and clay fines by drag and lifting forces from the
moving fluid when the mobilized fines plug thin pores causing the decrease in permeability.
The previous publications report transportation of the fine particles by a wetting fluid - water in the case of gas-water
flow (Miranda and Underdown, 1993, Bennion et al., 1996, 2000, 2005). Although gas phase does not transport the fines, their
migration has been observed in many gas reservoirs prior to water movement (Watson et al, 2001, Paveley et al, 2002). High
pressure draw downs with the exploitation of low-permeability unconventional gas reservoirs may cause mobilization of fines
with consequent decline in well productivity. In low-permeable reservoirs, fines may be formed by booklets of Kaolinite clays,
hairy illite clays, silt particles, uncrystalline silica, quartz, feldspar, etc. (Byrne, 2009, 2010). Significant decline in gas well
productivity due to fines migration and straining has been reported in the literature (Byrne 2010).Well productivity prediction,
well stimulation project and design in unconventional gas reservoirs are based on reliable mathematical models.
The reliable prediction of productivity index decline is based on mathematical modeling. Kinetics of particle capture
by a rock from the flowing suspension is described by the filtration equation (Herzig et al., 1970, Vafai, 2000)

∂σ
= λ cU
∂t (1)

where cv and σ are the concentrations of suspended and retained particles, U is the flow velocity and λ is the filtration
coefficient.
Various mathematical models of fines detachment produce different expressions for particle detachment rate that is
assumed to be proportional to the retained concentration and to the detaching factors such as drag force, difference between
equilibrium and current velocities, difference between the equilibrium and current suspension concentrations, etc. (Ju et al.,
2 SPE 151774

2007, Tufenkij, 2007, Civan, 2010, Massoudieh and Ginn, 2010). The shortcoming of the models describing kinetics of
particle detachment is the asymptotical stabilization of the retained concentration and permeability as time tends to infinity,
whereas fines release due to an abrupt increase in pressure gradient or decrease in salinity happens almost instantly (Miranda
and Underdown, 1993, Khilar and Fogler, 1998). The corefloods with sharp rate increase show an immediate permeability
response (Ochi and Vernoux, 1998).
It was long recognized that particle detachment occurs if a particle retained on the internal filter cake is not in the
mechanical equilibrium (Schechter, 1992, Bergendahl and Grasso, 2000, Civan, 2007, Bradford and Torkzaban, 2008,
Bradford et al., 2009). A particle located on the internal cake surface is under the action of electrostatic, drag, lifting and
gravitational forces. Some authors consider a force balance between the drag force, acting on the particle from the by-passing
fluid, and the friction force by introduction of an empirical Coulomb coefficient (Civan, 2007), while the others describe
mechanical equilibrium of a retained particle as the balance of all forces moments/torques (Jiao and Sharma, 1994, Freitas and
Sharma, 2001). Mathematically, these two approaches are equivalent. The current mathematical models describing the
detachment kinetics of a particle do not consider the mechanical equilibrium of a retained particle. Therefore, the expression
for the detachment rate is not affected by the mechanical equilibrium of a single particle.
A recently developed deep bed filtration model with a migrating layer of the fine particles attached in the secondary
energy minimum (Yuan and Shapiro, 2010b) also does not consider the forces acting on the retained particles.
Since these forces depend on micro parameters like particle and pore sizes, stochastically distributed in natural rocks,
the detailed modeling studies on micro (pore) scale have been carried out (Payatakes et al., 1973, 1974). These include:
population balance models (Sharma and Yortsos, 1987a,b,c), random walk equations (Shapiro, 2007, Yuan and Shapiro,
2010a), stochastic models (Lin et al., 2009) and direct pore scale simulation (Roussel et al., 2007, Bradford et al., 2009). The
population balance and random walk models, as well as large scale phenomenological models, use the kinetics detachment rate
term with an empirical coefficient and also do not consider the forces, acting on a single particle.
The modified particle detachment model uses the maximum (critical) retention function instead of kinetics
expressions describing the detachment rate: if the retention concentration does not exceed its maximum value, the particle
capture occurs according to the classical deep bed filtration model; otherwise, the maximum retention concentration value,
which depends on the flow velocity and brine ionic strength, holds (Bedrikovetsky et al., 2011a). The maximum retention
function is determined by the condition of mechanical equilibrium of a particle on the matrix or deposit surface, which is
described by the torque balance of electrostatic, drag, lifting and gravitational forces. Yet, this model describes just one
particle capture mechanism (attachment) whereas the permeability decrease with fines migration is explained by both
attachment and size exclusion.
In the present work, the new basic equations for one phase gas flow towards a well, accounting for two simultaneous
fines capture mechanisms such as detachment and straining, are derived. The filtration equation (1) describes the particle
retention rate by size exclusion, whereas the maximum retention function is used to model the attached fine particles
mobilization. Non-Darcy gas flow at high pressure gradients and varying connate water saturation are accounted for. The
developed analytical model for quasi steady state gas production with fines captures long term production with stabilized
suspended fines concentration near to a well along with deposit accumulation and skin growth proportionally to the amount of
produced fines. The results derived from the developed analytical model are in a good agreement with the field data for the gas
well impairment.
The structure of the paper is as follows. Initially, we briefly describe non-Darcy inertial gas flow towards a well. A
mathematical model for fines mobilization and straining resulting in the permeability decline is presented in the third section.
Then, the system of governing equations for one-phase flow towards a well with fines lifting, migration and size exclusion by
the rock is derived. This is followed by the analytical model corresponding to the steady state flow towards a well (Section 5)
and by the exact formulae for the productivity index vs time, which are derived in Section 6. Finally, the analytical model is
tuned to the field data on well productivity decline (Section 7).

2. High speed gas flow under the presence of connate water

The particularities of high velocity gas flow in low-permeable formations near a wellbore follow the non-Darcy law of flow
accounting for the flow turbulence, gas compressibility, high connate water saturation and also the variability of the connate
water saturation due to high pressure gradient variation from the wellbore to the extent of the drainage area. Let us derive the
basic flow equations accounting for the above factors.
The mass balance equation for axi symmetric gas flow around the well is

∂ (φ (1 − swi ) ρ ( p ) ) ∂
r
∂t
+
∂r
( r ρ ( p )U ) = 0
(2)
SPE 151774 3

Initial water saturation monotonically decreases with the capillary number Nc in the well vicinity due to the increase of gas
flow velocity
Uμ
swi ( NC ) = sw ( Nc ) , Nc =
σ gw (3)

The Darcy’s law for a gas flow at low velocities corresponds to a non-inertial flow as follows

dp μ
− = U
dr k0 krg ( sw ) (4)

This flow depends on the inertial effects at higher velocities, i.e. the Reynolds number can be introduced onto (4)

dp μ ⎛ U k0 krg ( sw ) ρ ⎞
− = Uf ⎜ ⎟
dr k0 krg ( sw ) ⎜ μ ⎟
⎝ ⎠ (5)

Let us keep the first two terms of the Tailor’s expansion of function f(Re) at small Reynolds numbers. Since the expression (5)
tends to the linear Darcy’s law (4) with velocity approaching zero, we conclude that f(0)=1. Therefore, the Darcy’s law (5) for
small Reynolds numbers is as follows:

dp μ Bρ
− = U+ U 2, B = f ′ ( 0)
dr k0 krg ( sw ) k0 krg ( sw )
(6)

Eq. (6) is called the Forchheimer’s equation for flow in porous media (Forchheimer, 1901). The so-called inertial coefficient B
depends on rock tortuosity, porosity and the porous space type (Geertsma, 1974, Civan and Evans, 1991).
Gas mass flow rate per unit of thickness

q = 2π r ρU , (7)

allows expression of flow linear velocity, U, in (6) via this flow rate:

dp μ q B q2
−ρ ( p ) = +
dr k0 k rg ( sw ) 2π r k0 k rg ( sw ) 4π r
2 2

(8)

Accounting for the equation of state for a real gas in the form

ρa p
ρ ( p) =
pa z (9)

and assuming constant compressibility factor, z, and constant initial water saturation, Eq. (8) becomes

ρ a p dp μ q B q2
− = +
pa z dr k 0 k rgwi 2π r k 0 k rgwi 4π r
2 2

(10)

Let us consider a steady state gas flow towards the well. The flow rate q is constant. Integration of both parts of Eq. (10) with
respect to r from the well radius, rw, to the drainage zone radius, re, yields

ρa μ q r 3B q 2 ⎛ rw ⎞
2 pa z
( pe 2 − p w 2 ) = k0 k rgwi
ln e +
2π rw
⎜1 − ⎟
k0 k rgwi 4π rw ⎝
2
re ⎠
(11)
4 SPE 151774

resulting in the following expression for the skin factor:

ρa μ q ⎡ re 3B k0 krgwi q ⎛ rw ⎞ ⎤
2 pa z
( pe 2 − pw2 ) = k0 krgwi
⎢ln +
2π ⎢ rw μ
⎜1 − ⎟ ⎥
2π 2 rw ⎝ re ⎠ ⎥
⎣ ⎦
3B k0 krgwi q ⎛ rw ⎞ (12)
S= ⎜1 − ⎟
μ 2π rw ⎝
2
re ⎠

3. Fine particles detachment

Following Valdya and Fogler, 1992, Khilar and Fogler, 1998, Iscan and Civan, 2006 and Civan, 2007, 2010, let us
describe physical mechanisms of permeability damage due to fines migration during well production period. Fine particles
located on the rock grain surface are affected by electrostatic, gravitational, drag, adhesion and lifting forces (Fig. 1). The
electrostatic, adhesion and gravitational forces attach a fine particle to the rock surface, whereas the drag and lifting forces
tend to detach it. Equilibrium of fine particles in porous media is determined by a torque balance of attaching and detaching
forces. The drag and lifting forces monotonically increase with flow velocity. At high flow velocities near to production well,
where the drag and lifting forces are large, the electrostatic, adhesion and gravitational forces cannot hold some fine particles
on the rock surface. As the result, the fines are detached and start migrating with the carrier fluid. The released particles
migrate in porous media by passing through larger pores. The migrating fines can be captured by small pores (straining or size
exclusion of fine particles) as shown in Fig. 2. The particle detachment results in some porosity increase. Nevertheless, this
does not lead to any significant permeability increase, whereas the mobilized particles captured in small pores result in
significant permeability decline due to pore plugging. Finally, the fine particle detachment, migration and straining cause
decline in the permeability and the well productivity index. Figs. 1 and 2 show particle detachment during one phase flow,
which is typical for water production with suspension filtering in aquifers.
The mobilized suspended fines are transported by the phase wetting these particles. Therefore, fines migration does
not occur during gas production in the presence of immobile connate water. Nevertheless, the laboratory tests show that the
drag and lifting forces, acting on the particles from the high rate gas flux, can mobilize the attached fines and further transport
them along the core The field data analysis on gas well impairment in low consolidated rocks also supports the idea of
mobilization and transport of fines by the moving gas phase.
Fig. 3 shows particle detachment during gas production in water wet rocks. The immobile connate water fills pores
and grain junctions, where it “holds” fine particles. The fine particles are released from the convex surfaces and wet by thin
water films, where the particles are exposed to the drag and lifting forces acting from the moving fluid.
Now, let us discuss the mathematical model for particle detachment by the drag and lifting forces.
The condition of the particle mechanical equilibrium on the grain surface is the equality of the torques of the
attaching and detaching forces (Jiao and Sharma, 1994, Freitas and Sharma, 2001, Civan, 2007):

Fe ln + 2 Fad ln + Fg ln = Fd (U ) ld + Fl (U ) ln (13)

where Fe, Fg, Fad, Fd and Fl are electrostatic, gravitational, adhesion, drag, and lifting forces; ln and ld are the levers for normal
and tangential forces, respectively (Fig. 1). The capillary adhesive force is included in the torque balance equation (13) for the
cases of fines release by produced fluid from the wetting film, shown in Fig. 3.
Imagine coreflood of the rock sample containing attached fine particles with piecewise increasing rate. Terms in right
hand side of (13) increase with velocity U, whereas the left hand side terms remain constant. After each velocity increase, the
particles on the grain surface (for which right hand side of (13) exceeds that in the left hand side) leave the grain surface and
migrate through the porous space. Therefore, the critical (maximum) retained concentration is a function of flow velocity
(Bedrikovetsky et al., 2011a,b). Since both drag and lifting forces are velocity dependent, the critical retention concentration
also depends on flow velocity U.
Drag and lifting forces detach particles, which are not covered by water (Fig. 3). The particles completely immersed
in immobile water in thin slots and grain intersection areas cannot be removed. Therefore, the maximum retained
concentration depends also on connate water saturation as follows:

σ a = σ cr (U , swi ) (14)

The higher the water saturation is, the lower the grain surface area is, where the fine particles are exposed to drag and lifting
forces and, consequently, the lower the maximum retained concentration is.
SPE 151774 5

Dimensional velocity in Eq. (14) can be substituted by the dimensionless ratio ε between torques of the detaching and
attaching forces (so-called torque or erosion ratio)
ld
ln d ( ) l ( )
F U +F U
σ a = σ cr ( ε ) , ε = (15)
Fe + 2 Fad + Fg

The critical retained concentration is a monotonically decreasing function of the fluid flow velocity.
The electrostatic force is determined by the Derjagin-Landau-Verbeek-Overbek (DLVO) theory (see Israelachvili,
1992, Khilar and Fogler, 1998). The expressions for drag, lifting and adhesion forces acting on a spherical particle located on
the pore wall are also available from the literature (Jiao and Sharma, 1994, Chauveteau el al., 1998, Bergendahl and Grasso,
2000, Freitas and Sharma, 2001). They allow calculation of the maximum retention function for a single cylindrical capillary,
which is a quadratic polynomial of the fluid velocity under constant saturation (Bedrikovetsky et al., 2011a)

⎡ ⎛ U ⎞2 ⎤
σ a (U ) = σ 0 ⎢1 − ⎜ ⎟ ⎥ (16)
⎢⎣ ⎝ U m ⎠ ⎥⎦

Here, σ0 is the maximum concentration of fine particles attached to the grain surface when gas is motionless, and Um is the
minimum velocity for which no particle can be held on the grain surface by the electrostatic, adhesion and gravitational forces.
High production rate results in high flow velocity, particularly in the wellbore vicinity. The initial concentration of the retained
fines, σi, determines the critical velocity, Ui (see Miranda and Underdown, 1993): the particle release occurs for U> Ui, where

σ a 0 = σ cr (U i ) (17)

The maximum retention function (15) can also be obtained from the coreflood tests with piecewise increasing
flowrates. Good agreement between the modeled results and experimental coreflood data may validate the model (15,16) for
modeling particle detachment. Yet, Bedrikovetsky et al., 2011a, b present the comparison of coreflood data with formula (5)
only for suspension injection cases, while the current work considers the mobilization of the natural reservoir fines. Besides,
only two sets of experimental data have been treated in the above mentioned works, whereas the laboratory results on particle
detachment are widely available in the literature. Therefore, below we analyse the laboratory coreflood data on lifting the
natural core fines under the flow rate increase in order to validate the model (5) for the fines mobilization conditions.
Ju et al, 2007, performed injections of constant salinity water with a piecewise increasing flowrate into a poorly
consolidated sandstone core with permeability of 850 mD, porosity 0.213 and an average pore radius of 10 μm. Despite
significant fines production has been observed during this test, the effluent fines concentration hasn’t been monitored. The
permeability increase with the flowrate increase (Fig. 4a) was explained by mobilization of fines, which are significantly
smaller than the pore size resulting in no straining. Fig. 4b shows the attached fine particles concentrations σa(U) as obtained
from the permeability values assuming the common value for the formation damage coefficient as βa= 35 (points). The
continuous curve σa(U) was calculated by tuned formulae (15,16) (see Bedrikovetsky et al., 2010, 2011). The adjusted
parameters are: salinity 0.1M of NaCl, the Hamaker constant for clay-water sand system was calculated as 0.62×10-20 J, zeta
potential for particles and grains are -30 mV and -20 mV, respectively. The normal pore size distribution varied between 0.1
μm and 24.9 μm with the standard deviation of 3.87 μm. The particle radius is assumed to be 1μm. Fig. 4b shows very good
agreement between the modeling results and experimental data.
Similar agreement can also be observed for the set of data presented by Ochi and Vernoux, 1998 and the modeling
results.
Overall, a good agreement between the retention concentrations as obtained from coreflooding data and the model
validates the assumption of the existence of the maximum (critical) retention concentration as a function of velocity (14, 16).
In the next section, the basic equations, describing steady state gas production with fines, causing pore plugging and
consequent permeability damage, are derived.

4. Mathematical model of gas production with accumulation of the retained fines

Let us consider one phase gas flow towards a well in the presence of immobile connate water. The suspended particles can be
captured by attachment and by straining. The attachment and straining causes porosity decline, therefore, the rock porosity is a
function of attachment and straining concentrations. The mass balance equation for gas with fines production differs from Eq.
(2) by the varying porosity
6 SPE 151774

∂ (φ (σ s , σ a )(1 − swi ) ρ ) ∂
r + ( r ρU ) = 0
∂t ∂r (18)

Here, initial water saturation is velocity-dependent according to Eq. (3).

The mass balance equation for suspended, strained and attached fine particles is

∂ ⎡φ (1 − swi ) cρ + ρ sσ a + ρ sσ s ⎦⎤ ∂
r ⎣ + ( rcρU ) = 0 (19)
∂t ∂r
Here, we assume a particle suspension with low concentration, c, i.e. the suspended particles do not decrease the porous space.
The particles attached to grains and pore walls, and those strained in small pores form the fines deposit (Fig. 2). The pore
spaces are saturated by connate water, by attached and strained particles and by the flowing particulate gaseous suspension.
For simplicity, accessibility and flux reduction factors during particle straining are not accounted for (Bedrikovetsky, 2008).
Taking derivatives of both terms of the mass balance equation (19) yields

∂c ∂c ∂ (σ s + σ a )
rφ (1 − swi ) ρ + rρU = −rρ s
∂t ∂r ∂t (20)

Fine particles can be attracted to grains and pore walls; the attachment rate is described by the linear kinetics filtration
equation until the deposit reaches its maximum (critical) value according to

∂σ a ρ s
= λa cρU , σ a < σ cr ( swi ,U ) (21)
∂t

where, λa is the filtration coefficient for attaching fines capture. Otherwise, Eq. (14) for critical attachment concentration
holds, i.e. the retained concentration remains constant after reaching the critical value, ߪ௖௥ , unless the flow velocity U changes.
Converting the suspended mass concentration c in (21) into the suspended volumetric concentration cv and accounting for
small particle concentration in gas phase, Eq. (21) turns into Eq. (1).The proposed model assumes the significant overlap
between the pore and particle size distributions, i.e., the probabilities of the particle to pass via a pore and to be captured by the
pore, have the same order of magnitude. It means that straining of particles in small pores may cause a significant straining
with the subsequent permeability decline.
The typical range of filtration coefficient λs is 0.1-100 1/m (Pang and Sharma, 1997, Bedrikovetsky et al., 2001),
which corresponds to the range of the particle free path 1/λs of 0.01-10 m. The size of the damaged zone is usually 1-5 m
(Civan, 2007, Nunes et al., 2010), i.e., the damage zone and the particle free run length have the same order of magnitude.
Therefore, the straining rate of fine particles is described by the following kinetics equation (Herzig et al., 1970, Vafai, 2000)

∂σ s ρ s
= λs cρU (22)
∂t

where, λs is the filtration coefficient for the size exclusion fines capture. So, the released fines are re-entrapped not
instantaneously, but after travelling the free path distance.
The model (22) assumes that the concentration of the retained particles is negligibly smaller than the concentration of
the vacant pores, where the particle straining may occur. In this case, the retention of a particle does not change further capture
probability, i.e., the filtration coefficient for the size exclusion fines capture is constant. At the higher retained concentration
comparable with the vacant pore concentration, the Langmuir blocking dependency λ s =λ s (σ) takes place.
Thus, both particle attachment and size exclusion are considered in the fines migration model. It is also assumed that
other particle capture mechanisms such as bridging, re-entrainment of deposited particles, segregation and diffusion are
negligible (Nabzar et al., 1996, Iscan and Civan, 2006, Rosseau et al., 2008, Civan, 2010).
Permeability monotonically decreases during particle capture. Pang and Sharma, 1997 approximated the normalized
reciprocal permeability by the linear function of retained concentration

k0
= 1 + βσ (23)
k (σ )
SPE 151774 7

where, β is the formation damage coefficient. For small retained concentrations, assumed in this model, the expression (23)
can be considered as two first terms of Tailor’s expansion of the normalized reciprocal to permeability. If the permeability is
affected by both, the attached and retained particles, formula (23) is transformed to

k0
= 1 + β aσ a + β sσ s (24)
k (σ a , σ s )

for retained concentrations of attached and size excluded particles.

The inertial coefficient B increases during particle attachment and straining. Taylor expansion series truncated after
the second term for a monotonically increasing function of two variables

B (σ a , σ s )
k (σ a , σ s )

yields

B (σ a , σ s ) B0
= (1 + γ aσ a + γ sσ s )
k (σ a , σ s ) k0 (25)

The non-linear Darcy’s law for gas flux under the presence of connate water accounting for formation damage due to
the attached and strained particles (Nabzar et al., 1996, Chauveteau et al., 1998, Mojarad and Settari, 2007, Rousseau et al.,
2008) follows from the combination of equations (8), (24) and (25):

dp μ B0
−ρ = (1 + β aσ a + β sσ s ) ρU + (1 + γ aσ a + γ sσ s ) ρ 2U 2 (26)
dr k0 krgwi k0 krgwi

Here, k0 is the initial rock permeability, krgwi is the relative permeability of gas in the presence of initial water, and μ is the gas
dynamic viscosity.
The above explanation of the permeability damage, as a result of sequential fines release and straining capture,
assumes that permeability change due to pore plugging highly exceeds that due to particle release, βs>>βa. Further in the text,
the permeability change due to attachment is ignored. The pore plugging may significantly increase the particle path tortuosity
and subsequent inertial pressure losses, whereas the attachment of particles decreases pore cross-sectional areas leading to a
more gradual hydraulic conductivity decline. Therefore, the inertial coefficient variation due to attachment is ignored
comparing to that due to straining, i.e. γs>>γa.
The simplified Eq. (26) has the following form:

dp μ B0
−ρ = (1 + β sσ s ) ρU + (1 + γ sσ s ) ρ 2U 2 (27)
dr k0 krgwi k0 krgwi

Thus, the system of four equations (18, 20, 14/21 and 22) determines unknowns c, σa, σs and p.
The gas production scenario includes the following processes: gas flow towards the well after switching ON the well,
propagation of pressure wave into the reservoir, gradual increase of velocity in each reservoir point until its critical value,
migration of lifting fines in the formation damage zone where the deposit affects the well index, and the continuous fines
straining with the gradual skin growth. The corresponding initial and boundary conditions with the given pressure on the
wellbore are as follows:

t = 0 : p = pres , σ a = σ a 0 , σ s = σ s 0 , c = 0
r = rw : p = pw (28)
r → ∞ : p = pres

The condition of impermeability (zero pressure gradient) is set on the reservoir boundary.
8 SPE 151774

If the mass flowrate per unit of thickness is known, then it can be used as another boundary condition:

dp μ q B q2
r = rw : − ρ = (1 + β sσ s ) + (1 + γ sσ s ) (29)
dr k0 k rgwi 2π rw k 0 k rgwi 4π r w
2 2

The analytical model, presented in the next section, describes the steady state period of the above process for fine particles
mobilization and size exclusion.

5. Analytical model for steady state flow with fines migration

Let us describe gas production with constant rate, steady state distribution of suspended concentration near to a well and a
gradual accumulation of migrated fines due to their size exclusion capture by the rock. Particle detachment in the damaged
zone due to timely increase of the pressure gradient is ignored comparing to that from the suspended flux from the outer
reservoir. It is also assumed that a low retention concentration σ=σa+σs does not affect the porosity. The effect of velocity
increase on the initial water saturation (3) is also ignored.
The above assumptions simplify the mass balance equation (20) to the following form:

∂c ∂σ s
r ρU = −rρ s , (30)
∂r ∂t

Substituting a straining capture rate expression (22) in (30) results in the differential equation for the concentration of
suspended particles

∂c
= − λs c (31)
∂r

Separation of variables in (31) leads to the explicit formula for the suspended concentration distribution around the
well

c ( r ) = cw exp ( − λs ( r − rw ) ) (32)

assuming that the produced fines concentration is known:

c ( rw ) = cw
(33)

The concentration of the suspended particles in gas vs radius decreases as the suspension moves towards the well with
the particle straining occurring. For the steady state flow regime (32), the gradient of the suspended particle concentration, c,
causes higher influx in each elementary volume dr compared to the outflux; this difference is compensated by the gradual
accumulation of the strained particles.
Substituting expression for suspended concentration (33) into (22) and integrating both sides in respect to t, results in
the explicit formula for a strained particle concentration distribution

λs qt
σ s ( r, t ) = cw exp ( −λs ( r − rw ) ) (34)
2π rρ s

i.e., the strained particles accumulate proportional to time.

In the next section, based on solution to (34), the pressure distribution around the wellbore is calculated, and the
formula for decreasing well productivity is derived.

6. Formula for well productivity

Using the analytical solution (32, 34), let us calculate the pressure square drop between the drainage radius, re, and the well
radius, rw.
SPE 151774 9

Expressing flow velocity in (27) via the gas mass flow rate (7) and accounting for the equation of state of the real gas (9)
yields

ρ a p dp μ q B0 q2
= (1 + β sσ s ) + (1 + γ sσ s ) (35)
pa z dr k0 krgwi 2π r k0 krgwi 4π r
2 2

The pressure square drop between the drainage radius and the well radius is calculated from (35) by integration in r
from the well radius rw to the drainage radius re as

ρa (p 2
− pw2 ) μ q ⎛ re
rd
σ dr
re
σ dr ⎞
⎜ ln + β s ∫ s + β s ∫ s ⎟ +
B0 q2 ⎛ 1 1
rd
σ s dr
re
σ s dr ⎞
s∫ s∫
e
= ⎜ − + γ + γ ⎟
2 pa z k0 krgwi 2π ⎜⎝ rw rw
r rd
r ⎟
⎠ k k
0 rgwi
4π 2 ⎜ r
⎝ w e r rw
r 2
rd
r 2 ⎟
⎠
(36)

where, rd is the so-called size of formation damage zone (see Nunes et al., 2010).
Axisymmetric variable r is located in the denominator of integrants in (36). Therefore, the larger the distance from
well is, the lower the impact of permeability in a reservoir point is on the pressure drop. The formation damage zone size is
defined as follows: the particle retention outside the damage zone r>rd does not affect the well impedance, i.e., the pressure
drop increase due to particle straining outside the damaged zone is negligible comparing to other terms in (36), and, therefore,
is ignored (Nunes et al., 2010). For this reason, the third integral terms in both brackets in the right hand side of expression
(36) are neglected.
Substituting expression for the retained concentration (34) into (36) and performing integration yields the final
expression for the pressure square drop

ρa (p e
2
− pw 2 )
=
μ q r
ln e +
B0 q2 ⎛ 1 1 ⎞
− ⎟+
2 ⎜
2 pa z k0 krgwi 2π rw k0 krgwi 4π ⎝ rw re ⎠
⎡ μ q B0 q2 ⎤ λ c qt exp [ λ r ] ⎧ exp [ −λ r ] exp [ −λ r ] ⎫
+⎢ βs + γ ⎥ s w s w
⎨
s w
− s d
− λs ( E1 ( λs rw ) − E1 ( λs rd ) ) ⎬
⎢⎣ k0 krgwi 2π k0 krgwi 4π 2π
2 s
⎥⎦ ⎩ rw rd ⎭
(37)

where,

∞
exp ( −t )
E1 ( x ) = ∫ dt
x
t

is an exponential integral.
The skin factor in (37) grows proportionally to time of gas production with fines.
Let us calculate the inverse to normalized well deliverability (so called impedance):

j (t ) =
(p e
2
− pw 2 ( t ) ) q ( t = 0 )
(38)
q ( t ) ( pe 2 − pw 2 ( t = 0 ) )
10 SPE 151774

For the case of a constant production rate, from (37) follows

⎡ μ 1 B0 q ⎤ ⎡ exp ( −λ r ) exp ( −λ r ) ⎤
⎢ βs + γ ⎥
2 s ⎢
s w
− s d
− λ ( E1 ( λs rw ) − E1 ( λs rd ) ) ⎥
⎢ k0 krgwi 2π k0 krgwi 4π ⎥⎦ ⎣ rw rd ⎦
j (t ) = 1 + ⎣ ×
μ 1 re B0 q ⎛1 1⎞
ln + 2 ⎜
− ⎟ (39)
k0 krgwi 2π rw k0 krgwi 4π ⎝ rw re ⎠
λs cw qt exp ( −λs rw )
×
2πρ s

For the case of slowly changing rate q(t) and concentration of produced fines cw(t), the formula (38) for impedance becomes

⎡ μ 1 B0 q ( t ) ⎤ ⎧ exp ( −λs rw ) exp ( −λs rd ) ⎫

⎢ βs + γs⎥⎨ − − λs ( E1 ( λs rw ) − E1 ( λs rd ) ) ⎬
⎢ k0 krgwi 2π k0 krgwi 4π
2
⎥⎦ ⎩ rw rd ⎭
j (t ) = 1 + ⎣ ×
μ 1 re B0 q ( t ) ⎛ 1 1 ⎞
ln + 2 ⎜
− ⎟
k0 krgwi 2π rw k0 krgwi 4π ⎝ rw re ⎠ (40)
t
λs exp ( −λs rw ) ∫ cw (τ ) q (τ ) dτ
× 0

2πρ s

7. Validation of the Model

The deliverability deterioration in the North Sea gas-condensate vertical perforated well during the period 1981-1987 was
explained by fines migration (Thrasher, 1995). The condensate precipitation in the rock hasn’t been mentioned and considered
to affect the well skin factor. The paper presents the field data on the rate decline and decrease of the pseudo pressure
drawdown versus time. Fig. 5 shows the increase of the normalised reciprocal to the well deliverability versus time. Reservoir
permeability k0=3-9 mD, porosity φ=0.166, reservoir thickness is 80-250 ft, gas viscosity μ=0.015 cP, the inertia coefficient
B0=774.5 (Norman et al., 1985), gas compressibility factor z=0.6, relative permeability for gas under the presence of connate
water krgwi=0.415, gas density at the normal conditions ρa=0.96 kg/m3. The results of the fitting with the analytical model (40)
are as follows: the density of the solid particles ρs=2600 kg/m3, the drainage radius re=500 m, the damaged zone radius rd=10
m, the formation damage coefficient for straining βs=90, the inertia formation damage coefficient ߛ= 9, the filtration
coefficient for straining λs=1.8 1/m and the produced suspension mass fraction cw=8ppm.
Figs. 5 a,b,c,d,e show sensitivity analysis with respect to the formation damage coefficient for straining, the filtration
coefficient for straining, the inertia coefficient, the drainage radius and the formation damage zone size, respectively. All
parameters have been varied by 30% with respect to the basic case obtained by the fitting of the field data. The most influential
parameters are the formation damage and filtration coefficients. The well deliverability also changes with variation of the
inertia coefficient. It is almost insensitive with regards to the drainage radius re.
Fig. 5e shows sensitivity study with respect to the formation damage zone radius rd. The expression for the drop of
pressure squares between the well and the reservoir (36) contains the overall damage, which is proportional to the integral of
σs/r from well radius to the drainage radius. Yet, concentration of strained fines declines with radius as an exponent, i.e. the
remote deposition almost does not affect the well index. Therefore, the damaged zone size rd is defined as a minimum radius
above which the pore straining by retained fines does not influence the well impedance (see Nunes et al., 2010). Fig. 5e shows
that for all damage radii larger than 8rw, the impedance well history is independent of formation zone radius (black and green
curves almost coincide). Yet, well index is very sensitive to the near-well permeability damage. Therefore, decrease of the
rd/rw ratio up to the values 6-2 results in significant decrease of well impedance.

8. Summary and discussions

The fine particles mobilization by drag and lifting forces exerting on particles from the moving gas, their migration and further
size exclusion by thin pores causes significant permeability decline and increase of the inertial resistance coefficient.
Therefore, the gas well index declines during the fines production. The mathematical model, predicting well productivity
SPE 151774 11

decline, consists of four equations for unknown concentrations of suspended, attached and strained particles and pressure. It
includes the kinetics equation for migrating particles straining and the maximum retention function for the particle detachment.
The maximum retention function of flow velocity has been calculated from three sets of corefloods and from the mechanical
equilibrium of a particle on the wall of a single capillary; the experimental and the modelling data are in a good agreement.
The analytical model describes flow towards the well with steady state suspension concentration near to well and
constant production rate. Due to the steady state suspension concentration, the gradual accumulation of size excluded fines is
going on proportionally to the time. Skin factor grows linearly vs amount of produced fines during the commingled production
of gas and fines. The proportionality coefficient of the skin growth is proportional to formation damage coefficient; it increases
with filtration coefficient increase and is almost independent of the radius of formation damage zone.
The model assumes small values of the retained concentration with the number of retained particles significantly
smaller than the number of vacant thin pores, where the particle size exclusion can happen. It results in constant filtration
coefficient for straining, in linear form of the normalized reciprocal to formation damage function and, finally, in the linear
skin factor growth vs the amount of produced fines. The late stage of well clogging with the large retention concentration,
where the well index may stabilize or the production may vanish is described by more complex solution of the system of
governing equations (18, 20, 14/21 and 22).
The field example on well productivity decline (Thrasher, 1995) was successfully matched by the analytical model
for steady state production of gas and fines. Yet, the values of too many parameters have not been presented in the referred
paper; these parameters have been obtained by matching. Also, the author’s claim that the decline of well deliverability is due
to fines migration, hasn’t been supported by field data. More well documented field cases must be analysed and tuned with the
model for the solid claim on the validity of the analytical model for gas well deliverability decline due to fines migration. The
preferable case would contain both coreflood and well data.

9. Conclusions

Derivation of the governing equations for quasi steady state gas production with fines migration and capture, development of
the analytical model and its matching with the field data allow concluding the following:
1. The mathematical model describing permeability decline by the fine particles mobilization and straining contains the
maximum retention function, modeling the particle detachment and the particle straining kinetics equation, describing
the size exclusion particle capture by the rock.

2. The analytical model for quasi steady state gas and fines production exhibits a linear skin factor dependency of the
amount of produced fines. This is a consequence of the model assumption of small retention concentration, so the
model is not valid for the late stage of the well clogging process.

3. The analytical model can be applied for well behaviour prediction from the initial stage of the well clogging.

4. The major parameter controlling well impairment due to fines production is the formation damage coefficient for
straining. The straining filtration coefficient also affects the well index decline. The productivity impairment is almost
insensitive to variation of the drainage radius. It is also insensitive to the damaged zone size for the ratio rd/rw
exceeding values 6-8.

5. Despite a good matching of the well production data with the modeling-based prediction was achieved for one field
case, further field validation of the analytical model of well clogging during fines production must be performed
based on well documented gas and fines production histories.

Nomenclature

B inertial coefficient
c mass fraction of suspended particles
cw mass fraction of produced fines
Fad adhesion force, MLT-2, N
Fd drag force, MLT-2, N
Fe electrostatic force, MLT-2, N
Fg gravitational force, MLT-2, N
Fl lifting force, MLT-2, N
j impedance
k absolute permeability, L2, mD
k0 initial permeability, L2, mD
12 SPE 151774

krgwi oil relative permeability at initial water saturation

ld lever for drag force, L, m
ln lever for normal force, L, m
p pressure, ML-1T-2, Pa
pres initial reservoir pressure, ML-1T-2, Pa
q mass flow rate per unit of thickness, ML-1T-1, kg/ms
r radial co-ordinate, L, m
rd radius of formation damage zone, L, m
re drainage radius, L, m
rw well radius, L, m
S skin factor
swi connate water saturation
t time, T, s
U linear flow velocity, LT-1, m/s
Um minimum linear velocity for which no particles can be held on the grain surface, LT-1, m/s
Ui critical linear velocity, LT-1, m/s

Greek letters
ρ gas density, L-3, 1/m3
φ porosity
γa inertial coefficient for of strained fines
ρa gas density under normal conditions, M/L3, kg/m3
ρs specific density of particles, L-3, 1/m3
γs inertial coefficient for attachment particles capture
β formation damage coefficient
βa formation damage coefficient for attachment
βs formation damage coefficient for straining
ε torque (erosion) ratio
λ filtration coefficient, L-1, 1/m
λa filtration coefficient for attachement particule capture, L-1, 1/m
λs filtration coefficient for size exclusion fines capture, L-1, 1/m
σ volumetric concentration of captured particles, L-3, 1/m3
σa volumetric concentration of attached fines, L-3, 1/m3
σa0 initial volumetric concentration of attached fines, L-3, 1/m3
σcr critical concentration of captured particles, L-3, 1/m3
σgw interfacial tension, M/T2, N/m
σi initial concentration of attached fines, L-3, 1/m3
σo maximum concentration of attached particles that corresponds to zero velocity, L-3, 1/m3
σs volumetric concentration of strained fines, L-3, 1/m3
σs0 initialvolumetric concentration of strained fines, L-3, 1/m3

Acknowledgements

Authors thank T. Rodrigues, I. Abbasy, K. Boyle (Santos Ltd, Australia) and F. Machado, A.L. S. de Souza (Petrobras, Brazil) for detailed
discussions of the field applications, for support and encouragement. PB is grateful to Prof. P. Currie (Delft University of Technology) and
Prof. A. Shapiro (Technical University of Denmark) for long-time co-operation in formation damage. Dr. A. Badalyan and Dr. T.
Carageorgos (The University of Adelaide) are gratefully acknowledged for improving the quality of the text. The work is sponsored by
Santos Ltd and by two grants of ARC (Australian Research Council).

Refrences:

Bedrikovetsky P.G., Marchesin, D., Checaira, F., Serra, A. L., Resende, E., Characterization of Deep Bed Filtration System from Laboratory
Pressure Drop Measurements, Journal of Petroleum Science and Engineering, vol. 64, pp. 167-177, 2001.
Bedrikovetsky, P., Upscaling of Stochastic Micro Model for Suspension Transport in Porous Media, Journal of Transport in Porous Media,
vol. 75, pp. 335-369, 2008.
Bedrikovetsky, P., Siqueira, F. D., Furtado, C., de Souza, A. L. S., Modified Particle Detachment Model for Colloidal Transport in Porous
Media, Journal of Transport in Porous Media, vol. 86, pp. 353-383, 2011a.
Bedrikovetsky, P., Zeinijahromi, A., Siqueira, F. D., Furtado, C., de Souza, A. L. S., Particle Detachment under Velocity Alternation during
Suspension Transport in Porous Media, Journal of Transport in Porous Media, available online DOI 10.1007/s11242-011-9839-1.
Bennion, B. D., Thomas, B. F., Bietz, R.F., Low permeability gas reservoirs: problems, opportunities and solutions for drilling, completion,
SPE 151774 13

stimulation and production, SPE 3557, Proceeding of the SPE Gas Technology Conference, Calgary, Alberta, Canada, 28 April – 1
May, 1996.
Bennion, B. D., Thomas, B. F., Ma, T., Formation Damage processes reducing productivity of low permeability gas reservoirs, SPE 60325,
Proceeding of the SPE Rocky Mountains Regional/Low Permeability Reservoi Simposium and Exhibition, Denver, Colorado,USA,
12-15 March, 2000.
Bennion, B. D., Thomas, B. F., Formation Damage Issues impacting the productivity of low permeability low initial water saturation gas
producing formations, J Energy Res Techn, Spet., vol. 127, pp. 240-248, 2005.
Bergendahl, J. and Grasso, D., Prediction of Colloid Detachment in a Model Porous Media: Hydrodynamics, Chemical Engineering Science,
vol. 55, pp. 1523-1532, 2000.
Bradford, S. and Torkzaban, S., Colloid Transport and Retention in Unsaturated Porous Media: A Review of Interface-, Collector-, and Pore-
Scale Processes and Models, Vadose Zone Journal, vol. 7, pp. 667-681, 2008.
Bradford, S., Kim, H., Haznedaroglu, B., Torkzaban, S. and Walker, S., Coupled Factors Influencing Concentration-Dependent Colloid
Transport and Retention in Saturated Porous Media, Environ. Sci. Technol, vol. 43, pp. 6996-7002, 2009.
Byrne, M., Waggoner, S., Fines Migration in a High Temperature gas reservoir – laboratory simulation and implications for completion
design, SPE 121897, Proceeding of the SPE 8th European Formation Damage Conference, Scheveningen, The Netherlands, 27-29
May, 2009.
Byrne, M., Slayter, A., McCurdy, P., Improved selection criteria for Sand Control: When are “Fines” fines?, SPE 128038, Proceeding of the
SPE International Symposium and Exhibiton on Formation Damage Control, Lafayette, Louisiana, USA, 10-12 February, 2010.
Chauveteau, G., Nabzar, L. and Coste, J., Physics and Modeling of Permeability Damage Induced by Particle Deposition, SPE 39463,
Proceeding of the SPE Formation Damage Control Conference, Lafayette, Louisiana, USA, pp. 409-419, 18-19 February, 1998.
Civan, F., Evans, R.D., Non-Darcy flow coefficients and relative permeabilities for gas/brine systems, SPE 21516, Proceeding of the SPE
Gas Technology Symposium, Houston, Texas, 22-24 January, 1991.
Civan, F., Reservoir Formation Damage: Fundamentals, Modeling, Assessment, and Mitigation, Gulf Professional Publishing, Elsevier,
Burlington, 2007.
Civan, F., Non-Isothermal Permeability Impairment by Fines Migration and Deposition in Porous Media Including Dispersive Transport,
Journal of Transport in Porous Media, vol. 85, pp. 233-258, 2010.
Cortis, A., Harter, T., Hou, L., Atwill, E.R., Packman, A.I. and Green, P.G., Transport of Cryptosporidium Parvum in Porous Media: Long-
Term Elution Experiments and Continuous Time Random Walk Filtration Modeling, Water Resources Research, vol. 42, 2006.
Forchheimer, P.: Wasserbewegung durch Boden, ZVDI, vol. 45, pp. 1781-1788, 1901.
Freitas, A. and Sharma, M., Detachment of Particles from Surfaces: An Afm Study, Journal of Colloid and Interface Science, vol. 233, pp.
73-82, 2001.
Geertsma, J., Estimating the coefficient of inertial resistance in fluid flow through porous media, SPEJ, vol. 5, pp. 445-450, 1974.
Herzig, J.P., Leclerc, D.M. and Goff, P.L., Flow of Suspensions through Porous Media-Application to Deep Filtration, Industrial and
Engineering Chemistry, vol. 62, pp. 8-35, 1970.
Iscan, A.G. and Civan, F., Correlation of Criteria for Perforation and Pore Plugging by Particles, Journal of Porous Media, vol. 9, pp. 541-
558, 2006.
Israelachvili, J.N., Intermolecular and Surface Forces, Academic press, London, 1992.
Jiao, D. and Sharma, M.M., Mechanism of Cake Buildup in Crossflow Filtration of Colloidal Suspensions, Journal of Colloid and Interface
Science, vol. 162, pp. 454-462, 1994.
Ju, B., Fan, T., Wang, X. and Qiu, X., A New Simulation Framework for Predicting the Onset and Effects of Fines Mobilization, Journal of
Transport in Porous Media, vol. 68, pp. 265-283, 2007.
Khilar, K. and Fogler, H., Migrations of Fines in Porous Media, Kluwer Academic Publishers, Dordrecht/London/Boston, 1998.
Lever, A. and Dawe, R., Water-Sensitivity and Migration of Fines in the Hopeman Sandstone, Journal of Petroleum Geology, vol. 7, pp. 97-
107, 1984.
Lin, H.-K., Pryadko, L.P., Walker, S. and Zandi, R., Attachment and Detachment Rate Distributions in Deep-Bed Filtration, Physical Review
E, vol. 79, pp. 046321-1-046321-12, 2009.
Massoudieh, A. and Ginn, T.R., Colloid-Facilitated Contaminant Transport in Unsaturated Porous Media, in: G. Hanrahan (Ed.),
Modelling of Pollutants in Complex Environmental Systems. ILM Publications, Hertfordshire, Glensdale, 2010.
Mihailov, N.N., Kolchitskaya, T.N., Dzemesjuk, A.V., Semenova, N.A., Physics-geological problems of residual oil, Moscow, Nauka, 1993
(in Russian).
Miranda, R.M. and Underdown, D.R., Laboratory Measurement of Critical Rate: A Novel Approach for Quantifying Fines Migration
Problems, SPE 25432, Proceeding of the SPE Production Operations Symposium, Oklahoma City, Oklahoma,USA, 21-23 March,
1993.
Mojarad, R. and Settari, A., Coupled Numerical Modelling of Reservoir Flow with Formation Plugging, Journal of Canadian Petroleum
Technology, vol. 46, pp. 54-59, 2007.
Mungan, N., Permeability Reduction through Changes in pH and Salinity, Journal of Petroleum Technology, vol. 17, pp. 1449-1453, 1965.
Nabzar, L., Chauveteau, G. and Roque, C., A New Model for Formation Damage by Particle Retention, SPE 1283, Proceeding of the SPE
Formation Damage Control Symposium, Lafayette, Louisiana, USA, 14-15 February, 1996.
Norman, R., Shrimanker, N., Archer, J., 1985, Estimation of the coefficient of inertial resistance in high rate gas wells, SPE paper 14207
presented at the 60th Annual Technical Conference and Exhibition of the Society of Petroleum Engineers held in Las Vegas, NV
September 22-25, 1985.
Nunes, M., Bedrikovetsky , P. and Pavia, P., Theoritical Definition of Formation Damage Zone with Application to Well Stimulation,
Journal of Energy Resources Technology, vol. 132, pp. 033101-1-033101-7, 2010.
Ochi, J. and Vernoux, J.-F., Permeability Decrease in Sandstone Reservoirs by Fluid Injection: Hydrodynamic and Chemical Effects,
Journal of Hydrology, vol. 208, pp. 237-248, 1998.
Pang, S. and Sharma, M.M., A Model for Predicting Injectivity Decline in Water-Injection Wells, SPEFE, vol. 12, pp. 194-201, 1997.
14 SPE 151774

Paveley, C., 2002, Blake Seawater injection. Well performance optimisation, SPE 73783, Proceeding of the SPE International Symposium
and Exhibition on Formation Damage Control, Lafayette, Louisiana, USA, 20-21 February, 2002.
Payatakes, A.C., Tien, C. and Turian, R.M., A New Model for Granular Porous Media: Part I. Model Formulation, AIChE Journal, vol. 19,
pp. 58-67, 1973.
Payatakes, A., Rajagopalan, R. and Tien, C., Application of Porous Media Models to the Study of Deep Bed Filtration, The Canadian
Journal of Chemical Engineering, vol. 52, pp. 722-731, 1974.
Rousseau, D., Latifa, H. and Nabzar, L., Injectivity Decline from Produced-Water Reinjection: New Insights on in-Depth Particle-
Deposition Mechanisms, SPE Prod & Oper, vol. 23, pp. 525-531, 2008.
Roussel, N., Nguyen, T., Coussot, P., General Probabilistic Approach to the Filtration Process, Phys Rev Letters, vol. 98, pp. 114502-1-4,
2007.
Schechter, R.S., Oil Well Stimulation, Richardson, Society of Petroleum Engineers, TX, 1992.
Shapiro, A.A., Elliptic Equation for Random Walks. Application to Transport in Microporous Media, Physica A: Statistical Mechanics and
its Applications, vol. 375, pp. 81-96, 2007.
Sharma, M.M. and Yortsos, Y.C., Transport of Particulate Suspensions in Porous Media: Model Formulation, AIChE Journal, vol. 33, pp.
1636-1643, 1987a.
Sharma, M.M. and Yortsos, Y.C., A Network Model for Deep Bed Filtration Processes, AIChE Journal, vol. 33, pp. 1644-1653, 1987b.
Sharma, M.M. and Yortsos, Y.C., Fines Migration in Porous Media, AIChE Journal, vol. 33, pp. 1654-1662, 1987c.
Tiab, D., Donaldson, E.C. and Knovel, Petrophysics: Theory and Practice of Measuring Reservoir Rock and Fluid Transport Properties,
Gulf Professional Pub., MA, USA, 2004.
Thrasher, T. S., 1995, Gas-Well Deliverability Monitoring: Case Histories, J SPE Production & Facilities, August, p. 177-183 (SPE paper
26181).
Tufenkji, N., Colloid and Microbe Migration in Granular Environments: A Discussion of Modelling Methods, in: F.H. Frimmel, F. von der
Kammer and F.-C. Flemming, Colloidal Transport in Porous Media. Springer-Verlag, Berlin, 2007.
Vafai, K., Handbook of Porous Media, Marcel Dekker, New York, 2000.
Valdya, R. and Fogler, H., Fines Migration and Formation Damage: Influence of pH and Ion Exchange, SPEPE, vol. 7, pp. 325-330, 1992.
Watson, R. B., 2001, Optimizing gravel pack performancein a high rate gas development, SPE 68969, Proceeding of the SPE European
Formation Damage Conference, The Hague, Netherlands, 21-22 May, 2001.
Watson, R.B., Viste, P., Kaageson-Loe, N., Fleming, N., Mathisen, A.M., Smart Mud filtrate: An engineering solution to minimise near-
wellbore formation damage due to kaolinite mobilisation, SPE 112455, Proceeding of the SPE International Symposium and
Exhibition on Formation Damage Control, Lafayette, Louisiana, USA, 13-15 February, 2008.
Yuan, H., Shapiro, A.: Modeling non-Fickian transport and hyperexponential deposition for deep bed filtration, Chemical Engineering
Journal, vol.162, pp. 974-988, 2010a.
Yuan, H., Shapiro, A.: A Mathematical Model for Non-monotonic Deposition Profiles in Deep Bed Filtration Systems, Chemical
Engineering Journal, vol.166, pp. 105-115, 2010b.
SPE 151774 15

ln Fl
Fd
Fg ld

Fe
Fig. 1- Forces acting on an attached particle during flow in porous media (torque balance on a single particle)
a) Adhesion force; b) particle on the grain surface

σa c
σs
Fig. 2- Straining of detached particles in a single pore

grain
water
flux

menisci
Fig. 3- Detachment of fines during oil/gas production from water wet rock
16 SPE 151774

a)

b)
Fig.4. Matching the theoretical model for maximum retention function with the experimental data by Ju et al., (2007): a) normalized
permeability vs. velocity b) maximum retention function - σ vs velocity U
SPE 151774 17

a)

b)
18 SPE 151774

c)

d)
SPE 151774 19

e)
Fig. 5. Matching the field data with the modeling-based prediction of well deliverability and sensitivity study: a) sensitivity with
respect to formation damage coefficient βs; b) variation of the straining filtration coefficient λs; c) sensitivity with respect to the inertia
coefficient B0; d) variation of drainage radius re; e) sensitivity analysis by the damage zone radius re.