Journal of Petroleum Science and Engineering 58 (2007) 227 – 244

www.elsevier.com/locate/petrol

Alteration of permeability by drilling fluid invasion and flow reversal

A. Gurkan Iscan a,⁎, Faruk Civan b , Mustafa V. Kok a
a
Middle East Technical University Department of Petroleum and Natural Gas Engineering Inonu Bulvari 06531 Ankara, Turkey
b
Mewbourne School of Petroleum and Geological Engineering, the University of Oklahoma, Norman, OK 73019, USA
Received 19 May 2005; received in revised form 3 January 2007; accepted 8 January 2007

Abstract

Permeability impairment caused by drilling fluids and subsequent cleaning and permeability enhancement by backflow are
investigated by means of experimental and simulation studies. Damage caused by two different drilling fluids is measured
experimentally by core tests as a function of the filtration pressure and analyzed using a simulator describing the fines migration
and retention in porous media. Simulations were run both with experimental and synthetic data in forward and backward directions
along the core samples. Permeability was correlated with respect to drilling filtration pressure in terms of the deposited particle
volume fraction. The clean-up time was determined after back-flush with fresh water and improvement was observed both in
porosity and permeability. Simulation results accurately match the experimental data, indicating that this simulator can be used for
the estimation of permeability reduction and the permeability and porosity variation along the core samples at various filtration
pressures. It was also determined that a polymer-added drilling fluid characterized with 65% permeability damage ratio may be the
optimum drilling fluid causing less formation damage than the water-based bentonite mud.
© 2007 Elsevier B.V. All rights reserved.

Keywords: Drilling fluid; Permeability; Porosity; Damage ratio; Backflow; Impairment; Enhancement; Fines migration; Porous media; Simulation

1. Introduction the permeability. However, many hydrocarbon reser-

Many petroleum engineering operations, such as
drilling, well completions, and workover, may cause an
alteration in the properties of hydrocarbon-bearing for-
mations, including porosity and permeability. Circum-
venting permeability alteration by fines migration is a
very important task in petroleum engineering applica-
tions. Clay particles intrude the reservoir formations
during drilling operations with water-based drilling
fluids. Particle intrusion causes plugging and bridging
across the pore throats within the pore spaces, reducing
⁎ Corresponding author. Tel.: +90 312 2104884; fax: +90 312
2104883.
E-mail address: iscan@ou.edu (A.G. Iscan).
0920-4105/$ - see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.petrol.2007.01.002
voirs have aquifer boundaries. Water influx takes place
in many cases. The cleaning effect of aquifers can be
simulated by injecting water into mud-invaded core
samples. In this study, the cleaning effect of the aquifers
and the change in permeability are investigated both
experimentally and numerically in terms of the damage
ratio.
Krueger et al. (1967) studied the permeability re-
duction in sandstone samples exposed to drilling fluids
and the clean-up with oil at elevated pressures. It was
aimed to break the particle bridges through back-flush
process. A drilling mud circulating system to was con-
structed to expose the cores to drilling mud under well-
bore conditions. The experimental system consists of a
one-barrel reserve pit and a conventional mud pump
228 A.G. Iscan et al. / Journal of Petroleum Science and Engineering 58 (2007) 227–244

used to circulate the drilling fluid through a high-pres- Table 1

sure cell. This high-pressure test cell is a container Percentage permeability recovery through back flow (after Krueger
et al., 1967)
having 12 ports that is used for exposing the core
specimens to flowing drilling fluid conditions. Mounted Backflow pressure Drilling fluid, percentage recovery (%)
(psi)
cores are clamped in the ports, and burettes are con- Oil-based Oil-based
nected to measure the volume of filtrate discharged drilling fluid-1 drilling fluid-2
during the experiments. Cylindrical Berea sandstone 10 – –
samples of 1.0-in. in diameter were used in the ex- 20 – –
80 – 78
periments. The air permeability of the core samples
100 – –
ranged between 300 and 600 mD. The specimens in 800 80 –
brass tubes were sealed with a non-penetrating epoxy
resin. The core samples were vacuumed and saturated Backflow pressure Drilling fluid, percentage recovery (%)
with 3% brine, and then the connate water state of the (psi)
Water-based Water-based
saturated core sample was attained with 42° API oil and drilling fluid-1 drilling fluid-2
the final permeability to oil was established at a pressure 10 – 80
drop of 600 psi. The drilling mud was circulated across 20 – 80
the core sample at a rate of 130 ft/min at a temperature of 80 70 80
77 °C and the core samples were subjected continuously 100 – 80
800 – 80
to the drilling fluid for 5 days under these conditions.
The cumulative fluid loss was recorded at certain time
intervals. After the circulation has been completed, the ticles. The porous medium is considered as a bundle of
core samples were removed without disturbing the filter uniform-sized capillary tubes. The pressure drop across
cake and placed into the backflow equipment. The the porous medium is obtained by a relationship cor-
backflow pressure was varied in different experiments to relating the permeability to the mass of solid particles
determine the recovery of permeability to oil in Berea deposited in porous media. The rate of particle release is
cores. The oil backflow was started at 10 psi. After the considered as an important phenomenon controlling the
stabilization of the permeability at this pressure, the permeability impairment. The particle release from the
pressure was increased from 10 psi to 30, 60, 100 and, pore walls is related with the electrostatic, van der
600 psi. Krueger et al. (1967) reported the permeability Waals, hydration, and hydrodynamic forces, as well as
recovery at these pressures by a chart. The percent the size of clay particles. The particle motion is assumed
permeability recovery data are presented in Table 1, Brownian when the particle radius r b 2 mm under the
expressed as the backflow permeability divided by the effect of diffusion and double-layer forces, and non-
original permeability. A significant increase occurred in Brownian in the presence of the hydrodynamic effects.
the permeability recovery at lower backflow pressures The rate of release coefficient is given as:
when the core samples were exposed to a water-based
mud. 38% of the original permeability of the core sam- i
krel ¼ Xk Np qs ð1Þ
ples to oil was regained at the 600 psi backflow pressure.
i
Using an oil back-flush was partially successful fol- where krel (m2) is the rate of particle release coefficient,
2
lowed by a low rate reverse oil flow. It was concluded Xk (m ) is the rate constant, N is the number of pore
that low rate clean-up is better than high rate clean-up in throats per unit bulk volume, and qs is the superficial
Miocene and Pliocene producing zones. Oil back flush velocity (m/s). The particle and pore sizes were nor-
simulation reduced the wellbore damage caused by silt malized with respect to their means. The numerical
migration. The production rate declined gradually as the solution scheme is based on a low-order upwind finite-
solid content increased in the produced oil. difference scheme. This method was chosen because of
Vitthal et al. (1988) proposed a model that simulates simplicity. In their study, a wide range of parameters
the permeability impairment in radial geometry. It was effecting the fines migration in porous media were ana-
noted that permeability impairment by fines migration lyzed. The effect of the mean particle size, standard
produces a positive skin in the near wellbore. An ide- deviation of particle size distribution, clay concentra-
alized geometry porous medium and clay particles were tion, rate of release coefficient, trapping length, and
considered. The clays were assumed spherical shape, network connectivity were analyzed. Permeability
single size, and constant density. Pore blockage occurs damage increased with the mean particle size. Small
gradually by smooth pore surface deposition of par- particles were able to migrate to the front of the invaded
 A.G. Iscan et al. / Journal of Petroleum Science and Engineering 58 (2007) 227–244
zone. The permeability profiles continued to decrease as
the amount of fines increased. However, there was a
little variation in the suspended particle profile for radial
geometry. The peak value remained constant but the
curve became broader as the amount of clay increased.
The concentration of the suspended particles increased
and the permeability decreased as the rate of release
coefficient increased. The trapping length was important
in the estimation of permeability damage. The rate of
capture of the suspended particles decreased as the
trapping length increased. This means that the particle is
expected to move further along with the fluid before
being captured at a pore throat. The decrease in per-
meability reduced as the connectivity of the pore
network decreased.
Gruber and Adair (1995) studied the evolution of
formation damage due to fines migration. In their re-
search, 1.5-in.-diameter and 2.0-in.-long core plugs
were used. The threshold pressure and permeability to
oil by the chemical gel and resin muds was determined.
The pressure range was considered between 0 and
40 kPa. It was stated that the flow capacity stabilization
took place at 15 kPa. However, it was seen that in-
creasing the pressure to 17 kPa causes a sudden increase
in the flow capacity. It was noted that no observation
was available until threshold pressure reached 30 kPa
(4.35 psi). The clean effect up was modeled in terms of
the ratio of the cleaned area per total damaged area.
It was concluded that the determination of threshold
pressures where the flow starts is important.
Longeron et al. (1998) studied eight different drilling
fluid formulations including the water-based and
synthetic oil-based mud. Static and dynamic filtration
tests were performed on the outcrop sandstone samples.
In their study, the impacts of various parameters on
permeability impairment, such as initial core perme-
ability, fluid saturations, temperature, and shear rate
were investigated. Porosity and permeability of clayey
Vosges Sandstone samples vary with mineralogical
composition. Typical composition of the 2–3 D perme-
ability core is 85% quartz, 9% feldspar, 5% illite, and
1% montmorillonite as weight percentages. Static and
dynamic filtration tests on 40-cm-long core samples
were conducted. Five water-based muds, one oil-based
mud, and two pseudo-oil-based muds were employed.
The water-based muds were named as F1, F2, F3, F4,
and F5. F1 is a standard salted polymeric bentonite mud.
F2 is a mud which is generally used to prevent clay
swelling. F2 mud includes polyglycerol, a fluid reducer
(PAC), and bentonite. F3 and F4 muds were designed
for application at high pressure and high temperature
(HP–HT) conditions. They contain bentonite, a mineral
229
viscosifier, a sulfonated fluid loss reducer, a dispersant
and Oil Companies Materials Association (OCMA) clay
as drilled solids. F5 mud contains a mixed metal hy-
droxide mud formulated with bentonite, modified poly-
saccharide and CaCO3 for bridging and weight control
purposes. The oil-based muds (OBM) were named as
F6, F7, and F8. F6 is a low toxicity OBM where the oil
base is high-density fluid HDF200, weighted with
barite. F7 and F8 formulations are invert emulsion muds
where the base oil is an alpha-olefine and an ester
synthesized from vegetable oil, respectively. CT (com-
puter tomography) tests were made on selected core
samples to check their homogeneity. The cores were
cleaned and dried. The gas permeabilities of the core
samples were measured. The core samples were vac-
uumed and saturated with 20 g/L NaCl brine concen-
tration. The 40-cm-long core samples were flooded with
nitrogen at 50 °C. The oil permeability at connate water
saturation, referred to as the undamaged initial perme-
ability, was measured at three flow rates. It was obtained
that the return permeability over different sections of the
core samples ranging from 4% to 24%. Essentially, the
damage took place in the first 5 cm of the core sample
with water-based mud. The best return permeability was
obtained with the F5 formulation that gave a 39% return
permeability. Formulation of F1 gave acceptable results
under both static and dynamic conditions, yielding 42%
and 35% return permeability, respectively. In their study,
it was concluded that the oil-based muds gave higher
values of return permeability ranging from 68% to 94%.
Higher filtration and spurt losses were obtained in
dynamic tests than the static ones. Oil return perme-
ability values were measured ranging between 8% and
48% of the initial oil permeability. Oil-based muds were
found to be much less damaging than water-based muds.
Oil return permeability was observed to be higher in
shorter core samples than the longer ones. It was
proposed that the reason for this is the difference in
pressure gradients applied during oil backflow.
Argiller et al. (1999) analyzed the formation damage
potential of water-based drilling fluids. They conducted
experiments with three water-based formulations. In
their research, static and dynamic filtration tests were
conducted with rock slices having 300 mD permeability.
The HP–HT and shear rate effect of formulations were
analyzed. It was found that static filtration is mainly
governed by external mud cakes. Finally, It was claimed
that formation damage caused by water-based fluids can
be avoided by optimizing the fluid loss reducer and
particle size distribution.
Bailey et al. (1999) studied the invasion of particles
from drilling fluids. The particulate invasion was found
230 A.G. Iscan et al. / Journal of Petroleum Science and Engineering 58 (2007) 227–244
to be one of the primary mechanisms of formation
damage caused by drilling fluids. Particles are forced
into the formation generally during the earlier stages of
the filter cake growth. A KCl polymer fluid and different
grades of barite and carbonate weighting agents were
used. Rock samples having about 600 mD permeability
and 17% porosity were tested. The cores were vac-
uumed and saturated with brine simulating the connate
water from the Heidrun reservoir. The permeability to
brine was measured at a constant flow. Pressure trans-
ducers were fitted to the inlet of the core holder to
measure the pressure drop across the core. A static
filtration was carried out for about 4 h at 23 °C under
300 psi differential pressure. The permeability to brine
was measured again after filtration. The % retained
permeability is defined as the difference between the
equilibrium flowing pressures before and after filtration
and is considered as a measure of formation damage.
Fine particles penetrated deeply into the formation and
could not be easily removed by back-flushing. Larger
particles were observed to deposit near the surface.
Back-flushing removed the near surface internal filter
cake but could not sufficiently remove the deeply
invaded particles. The permeability reduction was
greatest in single-phase brine conditions.
Al-Riyamy and Sharma (2004) studied the filtration
properties of oil in water emulsions containing solids.
Their study focused on factors controlling the filtration
and leak-off properties of emulsions which include solid
particles. The effects of percent oil, filtration pressure,
core permeability, temperature, and viscosity of the
continuous phase were investigated by laboratory ex-
periments. The emulsified fluids used included a known
volume percent of n-decane as a dispersed phase, 3%
NaCl brine as the continuous phase, CaCO3 particles at
a concentration of 5 wt.%, and a newly synthesized pH-
sensitive emulsifier at a 0.05 wt.% concentration. The
experiments were conducted with Texas limestone
(5 mD), Berea sandstone (175 mD), and Aloxide cores
(1.5 D) using 2.5-in.-diameter and 1-in.-length core
samples. The core samples were dried in an oven at
65 °C for 1 day. The absolute permeability was mea-
sured with 3% NaCl brine. The return permeability was
better at 65 °C resulting in a better clean-up of the
internal filter cake. Increasing the over balance pressure
increased the filtrate volume and reduced the return
permeability. This effect became stronger at higher tem-
peratures. Increasing the overbalance pressure increased
the extent of invasion of emulsion drops into the core.
The viscosity of the continuous water phase was changed
by varying the HEC polymer concentration from 0.4% to
0.6%. The return permeability with 0.4 wt.% HEC was
observed to be higher than that of 0.6 wt.%. The polymer
was found to be partially responsible for the damage in
the cores.
Ding et al. (2004) studied the near wellbore damage
and natural clean-up of horizontal wells. The near-
wellbore properties were altered by drilling fluid, fluid–
fluid interaction, and fluid–filtrate invasion during
overbalanced drilling operations. The degree of forma-
tion damage was affected by many parameters including
the nature and characteristics of the drilling fluid, for-
mation properties, and operating conditions. The drilling
fluid invasion was studied. The permeability reduction
factor was correlated with flow rate. The change in the
permeability of filter cake was analyzed, but not the
change in the permeability of the formation. A rapid
fluid loss, indicating as a spurt loss, when the drilling bit
contacts the reservoir, was observed. This is because
there is no filter cake formed yet to prevent the fluid and
solid particle invasion into the pay zone. The following
observations were noted about the oil flow-back pro-
cess. A deeper particle invasion of the internal filter cake
decreased the flow efficiency while preventing the fil-
trate invasion. Serious loss of production occurred with
damaging and non-optimized drilling fluids. The for-
mation damage was much severe with a water-based
mud than an oil-based mud.
Parn-anurak and Engler (2005) developed a method
to simulate drilling fluid invasion of water-based fluids
into hydrocarbon bearing formations. The model was
based on the mass balance equation. The reduction of
permeability was analyzed in the damage zone in terms
of water phase saturations and skin concept along radial
direction. The average permeability was approximated
along the radius of the skinned zone. They solved the
convection–dispersion equation numerically to charac-
terize filtrate invasion behavior and concluded that their
model can be used to simulate radial flow of water-
based filtrate.
Erwin et al. (2005) analyzed the damage due to brine
imbibition. The flow efficiency was considered in terms
of average and flowing well pressures together with
differential skin pressure. It was proposed that higher
flow rates would minimize the artificial lift require-
ments. The samples were obtained from Alpine reser-
voir on Alaska's North Slope. The water–oil imbibition
intersects at about 0.55 fractional–water saturation. It
was concluded that a careful core and reservoir prop-
erties analysis can lead to identification of damage
reasons. It was concluded that waterflood breakthrough
reduces the wellbore relative permeability.
Dalmazzone et al. (2006) optimized the filtrate
design to minimize wellbore damage during drilling.
 A.G. Iscan et al. / Journal of Petroleum Science and Engineering 58 (2007) 227–244
Different polymer-added water-based fluids were used.
Invasion into porous media tests was performed at
400 mD permeability and 15% porosity. Static and
dynamic filtration and emulsion blocking tests were
conducted and it was concluded that invert-emulsion
systems yield higher return permeability than water-
based fluids.
Van der Zwaag (2006) conducted core plug perme-
ability measurements before and after fluid application
at a circulation rate of 2 mL/min. The porosity, initial
and final pore fluid saturations, and fluid rheology were
measured. The measurements indicated that filtrate
volumes obtained at dynamic filtration regimes are
larger than static filtration volumes. Reduced annular
space at the wellbore end of a core plug affects the
drilling fluid invasion. Invasion of water-based fluids
reduced the overall permeability.
While the above-mentioned experimental studies
provide valuable observations and understanding of the
relevant phenomena, they do not provide a com-
prehensive mathematical model and specific information
about the effect of the phenomenological parameters.
This paper applies a phenomenological model to analyze
the available data. First a synthetic case study is
presented to demonstrate the capability of the present
model and simulator. Then, a set of dynamic drilling mud
circulation experiments was carried out using two
different drilling fluids at eleven different filtration
pressures and three different circulation periods. Alter-
ation in permeability is studied in terms of the damage
ratio concept. Damage ratio is characterized as the
percentage ratio of the difference between the undam-
aged and damaged permeability to the undamaged
permeability. The influence of filtration pressure and
circulation periods is investigated both experimentally
and numerically. The acquired experimental data are
analyzed by means of a numerical computer simulator
for fines migration and deposition in porous media. The
results are correlated and an exponential model for the
effect of filtration pressure on damage ratio is developed.
2. Simulation of fines migration and deposition and
permeability alteration
The fine particle invasion and clean up processes are
modelled by using a set of equations that correlate the
particle concentration and fraction of depositing parti-
cles to permeability according to Civan and Nguyen
(2005). The porous media is assumed to include plug-
ging and non-plugging pathways. The simulation model
is a pore scale model. No fracture or wormhole existence
is considered. Darcy's flow conditions exist. The fluids
231
are assumed incompressible. One-dimensional solution
is applied along the core plug. Cartesian coordinate
system is used. The core plugs were simulated using
20 grid blocks.
This model facilitates a plugging and non-plugging
pathways concept according to Gruesbeck and Collins
(1982). A sensitivity analysis with the Civan and
Nguyen (2005) simulator has been carried out to de-
termine the contribution of plugging pathways to per-
meability reduction. The permeability changes in the
plugging sections were found negligible with respect to
the non-plugging permeability. However, both the
plugging and non-plugging pathways participated in
fine particle transport. The core samples used in this
study have significantly less plugging pathways than the
non-plugging ones. Civan (2000) defined permeability
as a function of fraction of depositing particles in both
plugging and non-plugging pathways. In the present
case, the permeability is more sensitive to non-plugging
pathways deposition than the plugging ones. This sug-
gestion is justified by the output plot of the numerical
simulator for non-plugging permeability in Fig. 7. The
reduction in non-plugging permeability is about 37%.
The reduction in plugging permeability was found
through the simulator as 0.2%. Thus, the non-plugging
permeability is dominant in the plugging phenomenon
inferred by the results of the applied simulator. Thus, the
non-plugging permeability impairment was considered
in damage ratio analysis of the core samples used in this
study. However, the particle concentrations and fraction
of depositing particles were compared both in plugging
and non-plugging pathways. A list of the equations of the
model is presented in Appendix 1 by Civan and Nguyen
(2005). Here only the equations involving permeability
alteration are provided. In this simulator, the numerical
method of lines is used as a technique for solving partial
differential equations after discretizing in one dimen-
sion, and then integrating the semi-discrete problem with
time as a system of ordinary differential equations. The
method of lines reduces the partial differential equations
with many dependents a into a single dependant
problem, which is time for the present case. This method
takes advantage of the available general methods and
software that have been developed for numerical
integration of ordinary differential equations. The meth-
od typically proves to be quite efficient for the partial
differential equations to which the method of lines is
applicable. It is necessary that the partial differential
equation problem be well-posed as an initial value
(Cauchy) problem in at least one dimension because the
ordinary differential equation integrators can solve the
initial value problems.
232 A.G. Iscan et al. / Journal of Petroleum Science and Engineering 58 (2007) 227–244

Table 2 Table 3
Input data for the forward run with synthetic data Input data for the backward run with synthetic data
Parameters Values Parameters Values Parameters Values Parameters Values
Core length, cm 15.4 ρp, kg/m 3
2825 Core length, cm 15.4 ρp, kg/m 3
2825
Core diameter, cm 3.81 μ, cP 5 Core diameter, cm 3.81 μ, cP 5
ϕpo 0.1 σpo 0 ϕpo 0.1 σpo 0.049
ϕnpo 0.2 σnpo 0 ϕnpo 0.2 σnpo 0.0045
Kpo, D 25 σslurry 9.4 × 10− 4 Kpo, D 1.0 σslurry 9.4 × 10− 7
Knpo, D 1 εpo ϕ Knpo, D 16.7 εpo 1.26 × 10− 6
α, cm− 1 10 εnpo ϕ α, cm− 1 10 εnpo 0.0184
δ, cm− 1 0.0001 τcr, Pa 20 δ, cm− 1 0.0001 τcr, Pa 20
kd, cm− 1 1.28 U, cm/s 0.5 kd, cm− 1 1.28 U, cm/s 0.5

The total fraction of depositions in the non-plugging After the calculation of the non-plugging perme-
pathways is represented by Civan and Nguyen (2005): ability, the damage ratio is calculated as the percentage
ratio of the initial permeability of the core sample with
e ¼ ep þ enp ð2Þ respect to filtration pressure and time.
The deposition rate in the non-plugging pathways is
given by: 3. Analysis of mud invasion and clean up by flow
reversal using a simulator
Aenp
¼ kp unp rnp /2=3
np −ke enp /np ge ðsnp −scr Þ
2=3
ð3Þ
At The fines invasion and cleanup processes are
simulated in this section.
Because we focused on the non-plugging permeabil-
ity impairment rather than the plugging one, the vari- 3.1. Effect of fines invasion in porous media
ation of non-plugging deposition fraction is solved
numerically using the initial and boundary conditions, The Civan and Nguyen (2005) simulator, implemen-
given by Eqs. (4) and (5), respectively. ted in Mathematica, was used for studying the fines
migration and deposition in core samples. The aim of
enp ¼ enpo ¼ 0; t ¼ 0 ð4Þ this application is to simulate the effect of constant rate
circulation of drilling fluids on invasion of fine particles
enp ¼ enpt1 ; t ¼ t1 ð5Þ
into porous media and plugging of the pore pathways
Darcy's equation is used for determination of the dy- (Fig. 1).
namic flow rate for a specified filtration pressure. The A sandstone core sample has been subjected to
initial and return permeabilities of the core samples drilling fluid circulation for approximately 106 min. The
were also calculated by Eq. (6) for analysis of the particle concentration variation was observed across the
experimental data. porous media as the circulation continued. The initial
fine-particle concentration was assumed zero. The vol-
Knp ADP umetric particle concentration of the fluid was consid-
q¼ ð6Þ
lL ered as 9.4 × 10− 4 cm3/cm3. As the circulation took
The non-plugging permeability is determined in- Table 4
stantaneously as a function of non-plugging deposi- Input data for the run with mud-1
tional particle fraction of the bulk volume of the core Parameters Values Parameters Values
sample.
Core length, cm 7.1 ρp, kg/m3 1100
Knp ¼ Knpo ð1−enp =/npo Þn2 ð7Þ Core diameter, cm 3.81 μ, cP 15
ϕpo 0.11 σpo 0
ϕnpo 0.24 σnpo 0
The simulation parameters and input file used for the Kpo, D 0.00184 σslurry 9.4 × 10− 2
simulation of the synthetic and experimental data are Knpo, D 0.0184 εpo ϕ
given in Tables 2–5. α, cm− 1 10 εnpo ϕ
δ, cm− 1 0.0001 τcr, Pa 20
Ki −Kf kd, cm− 1 2.28 Pfiltration, psia 50–100–200
DR ¼  100 ð8Þ n2 0.45 Dp, μm 10
Ki
 A.G. Iscan et al. / Journal of Petroleum Science and Engineering 58 (2007) 227–244 233

Table 5
Input data for the run with mud-2
Parameters Values Parameters Values
Core length, cm 7.1 ρp, kg/m 3
1100
Core diameter, cm 3.81 μ, cP 50
ϕpo 0.12 σpo 0
ϕnpo 0.23 σnpo 0
Kpo, D 0.00166 σslurry 15.2 × 10− 2
Knpo, D 0.0166 εpo ϕ
α, cm− 1 10 εnpo ϕ
δ, cm− 1 0.0001 τcr, Pa 25
kd, cm− 1 2.28 Pfiltration, psia 50–100–200
n2 0.60 Dp, μm 10

place, the solid concentration increased in both plugging
and non-plugging paths through the core sample (Figs. 2
and 3). The deposition rate was determined by the
change of the fractions of plugging and non-plugging
pathways of bulk volume occupied by the deposits
(Figs. 4 and 5). The average plugging deposition rate is
8.49 × 10− 9 min− 1. The average non-plugging deposi-
tion rate is 1.72 × 10− 4 min− 1. Both the plugging and
non-plugging porosity and permeability of the core
sample decreased as the particles in the suspension were
deposited. Fig. 4 indicates that the plugging fraction
of deposits continue to increase until the end of the
102nd minute and reaches the maximum value of
1.2 × 10− 6 cm3/cm3. From that time on, the data look
constant for two more minutes. In Fig. 5, the non-
plugging fraction of deposits keep increasing until the
80th minute and slightly decreases its increase trend and
reaches 0.0184 at the end of the total time of 106 min.
The concentration of plugging particles displays a very
sharp increase between the 15th and 30th minutes. After
this moment, there is a reasonable reduction in the
plugging concentration for about 10 min. This may be
due to the pushing effect of the depositing particles
behind this point. The plugging concentration increases
almost linearly after this point until the end of the core
sample and reaches the value of 0.6 ppm (Fig. 2). The
non-plugging concentration increases sharply between
the 20th and 40th minutes. This increase may be due to a
sudden plugging and then relief by jet action of de-
Fig. 2. Concentrations of particles in plugging pathways.
posited particles. From the 40th minute on, the non-
plugging particle concentrations keep increasing in
their normal trend and reach the value of 0.02 ppm
(Fig. 3). The non-plugging porosity decreases as the
concentration of the fluids increased in time. There is
more reduction in non-plugging porosity with respect to
plugging porosity. The non-plugging porosity decreases
linearly from 0.2 to 0.183. This means 8.5% reduction in
non-plugging porosity (Fig. 6). The reduction in non-
plugging permeability is about 35% in 106 min. This
reduction means that permeability has dropped from
25 D to 16.25 D. The reduction is less in the first 13 min
of the circulation, but from that moment on the non-
plugging permeability reduction is rapid (Fig. 7).
However, the reduction in plugging permeability is
much less than that of non-plugging permeability as
percentage. Nevertheless, a small change in the plug-
ging permeability may influence the flow of fluids
because of the low initial value of the plugging per-
meability. The fine particles deposit on the pore throats
and forms bridges. The closer sections to the circulation
point are naturally affected more quickly. When these
pore throats are plugged, the further pore spaces remain
unplugged. The circulation continues at a constant
rate. The fines particles accumulate at the pore throat
and force the previously formed bridges to open by
increasing the filtration pressure. Thus, the particles at

Fig. 1. Plugging and non-plugging pathways along the core sample.

234 A.G. Iscan et al. / Journal of Petroleum Science and Engineering 58 (2007) 227–244
Fig. 3. Concentrations of particles in non-plugging pathways.
the pore throat are moved along the porous pathways
and a temporary reduction in particles concentration
takes place at a specific distance from the circulation
point.
The fraction of plugging porosity to overall porosity
( fp) has been drawn along the core sample at five
different times that are at 4, 8, 12, 18, and 24 min
(Fig. 8). It is seen that, as the contamination period
increases, the starting values of fp increase. Civan and
Nguyen (2005) plotted fp vs. dimensionless pack length.
They obtained a similar trend after 50 pore volume
injections. In this study, the fp values remain constant
after a certain distance due to the saturation of porous
pathways by the fine particles, which deposit and settle.
Thus, a certain decrease of fp is observed. However,
there is no complete plugging of the pathways. The
flowing section is reduced but not totally plugged. The
value of fp remains constant with respect to the core
distance when the fine particle distribution attains
a limiting value, which is the horizontal asymptote
Fig. 4. Fraction of plugging pathways of bulk volume occupied by the
deposits.
Fig. 5. Fraction of non-plugging pathways of bulk volume occupied by
deposits.
fp = 0.48 in the fp profile. The reduction of porosity
fraction continues until the end of the 5th cm, which is
one-third of the core length. This is why fines migration
during drilling causes near wellbore formation damage.
The fp values continue to decrease significantly in the
first five grids along the core sample. After the fifth grid,
the fp remains constant at a value of 0.48.
3.2. Effect of fines clean-up from porous media
The endpoints of the forward run described in the
previous section are placed as the new initial conditions
in the simulator. The final permeability, porosity, particle
concentration, fraction of depositing particles are
recorded after every simulation run at each grid block
after the circulation is completed. Those values are
allocated as the new initial conditions for backward
simulation. The simulation is carried out by injecting
water from the reverse side of the core. Thus, the
cleaning effect by backflow is studied. The backflow rate
is the same as the forward injection rate. Only the initial
Fig. 6. Variation of non-plugging porosity with respect to time.
 A.G. Iscan et al. / Journal of Petroleum Science and Engineering 58 (2007) 227–244
Fig. 7. Variation of non-plugging permeability with respect to time.
concentration of the back-flushing fluid is reduced. The
time axis in the graphs is from the starting point. The
back-flush simulation starts after the completion of the
forward run after 106 min. The cleaning period takes
60 min. Thus, the complete process including the back
and forth is completed in 166 min. The removal rate was
determined by looking at the change of the fractions
of plugging and non-plugging pathways of bulk vol-
ume occupied by the deposits (Figs. 4 and 5). Both the
plugging and non-plugging fractions of bulk volume
occupied by deposits have been cleaned efficiently.
The clean-up rate for the plugging fraction is 2.0 ×
10− 8 min− 1 (Fig. 4). The clean up rate is almost 2.4 times
faster than the forward contamination scheme. The
clean-up time is 60 min. In the non-plugging fraction of
bulk volume, the clean-up rate is constant. There is no
turning point in contrast to the forward run. The clean-up
rate is 1.74 × 10− 4 and slightly faster than the deposition
rate of the non-plugging deposits fraction in forward run
(Fig. 5). The particle concentration decreased both in
plugging and non-plugging pathways. The plugging
particle concentration decreased from 0.60 to 0.56
linearly for 81 min. After this moment, there is a sharp
reduction followed by an increase. This increase may be
because of the movement of the dislodged plugging
particles towards another pore-throat location and re-
plugging the media at that point. However, after the
increase, plugging particles concentration continued
decreasing sharply down to 0.054 ppm at the end of
60 min. Then, the plugging particles concentration
remained constant for the rest of the time until the
breakthrough is reached. The clean-up time for the
plugging particles is 60 min (Fig. 2). The non-plugging
particles were cleaned at about 62 min. After that time,
the reduction continued, but did not provide a perfect
cleaning of the non-plugging particles (Fig. 3). The non-
plugging porosity starts from 0.183 and raises up to
235
0.193 in 60 min. Then it follows almost a constant trend.
The non-plugging porosity is given in its real values, not
by the normalized ones, because the reduction in
numbers is significant (Fig. 6). The improvement in
the non-plugging permeability is linear and cannot attain
the initial value of 0.2 even though the particles
concentration has reduced significantly. The non-
plugging permeability improves from 0.62 to 0.83 as
normalized values in 64 min (Fig. 7). The actual value of
the final non-plugging permeability becomes 20.75 D.
A graph of the fraction of bulk volume containing the
plugging and non-plugging pathways along the core
sample is given for both contamination and cleaning
cycles in Fig. 9. There is an improvement in fp from
0.476 to 0.526 throughout the core sample. The damage
ratios versus filtration pressure were plotted for both
drilling fluids. After the experiments were completed,
the mathematical simulation was run in accordance with
the conditions of the experimental data. Two different
drilling fluids were used. Two different damage ratio vs.
filtration pressure plots were obtained. The simulation is
harder to apply for everyday use because of the complex
feature of the simulation parameters and reservoir
conditions. Accurate and easily applicable practical
methods are required in the drilling industry. No direct
relationship has been offered previously between the
permeability damage ratio and filtration pressure. The
proposed correlation (Eq. (9)) relates the damage ratio to
filtration pressure on a drilling fluid type basis. The
coefficients A, B, and C of the proposed correlation (Eq.
(9)) are dependent on the fluid type and rheology.
DR ¼ A½1−expð−BPf Þ ð9Þ
The correlation given by Eq. (9) can be very easily and
accurately applied. This correlation may help the
Fig. 8. Fraction of plugging porosity to overall porosity ( fp) has with
respect to the core sample at five different times such as 4, 8, 12, 18,
24 min.
236 A.G. Iscan et al. / Journal of Petroleum Science and Engineering 58 (2007) 227–244

are water, bentonite, and Polymer XT (a fluid loss-

reducing polymer). 1285 g weight of API bentonite and
1000 g weight barite in 20 l of water was used for each
mud sample (22.5 g-wt/350 ml) according to the API
Standard. We added 22.8 g-weight Polymer XT for
fluid type 3 in addition to bentonite and barite. We added
1 g-weight/350 ml equivalent of caustic soda as a
stabilizer. We then mixed the drilling fluid for about 5 h.
The drilling mud is left to rest for 16 h. We finally
remixed the mud for perfect homogeneous slurry.
Caustic soda in the amount of 1 g-weight/350 ml of
slurry was also added to stabilize the mud so that the
particles would not settle. The density, i.e., the mud
Fig. 9. Fraction of plugging porosity to overall porosity ( fp) versus
grids after 22 min.
weight, of each sample was kept as 9 lb/gal (1.083 g/cm3).
The detailed compositions of the fluids and their
physical properties are given Table 7. The rheological
drilling and reservoir engineers make quick and accurate parameters with the polymer mud were also determined
estimations of the damage ratio at a required filtration at different polymer concentrations (0.2 g-weight/
pressure by considering the rheological properties of 350 cm 3 , 0.4 g-weight/350 cm 3 , 0.6 g-weight/
drilling mud. The linearized form of this equation is 350 cm3, 0.8 g-weight/350 cm3, and 1.0 g-weight/
derived as follows: 350 cm3 ). The optimum polymer amount was deter-
mined to be 0.4 g-weight/350 ml as following. The fluid
ln½1−DR=A ¼ −BPf ð10Þ could not attain the required rheological parameters
when the polymer was added at an amount less than
The linearized form (Eq. (10)) of the correlation was 0.4 g-weight/350 cm3-slurry. When higher amounts of
obtained by taking the natural logarithm. The lineariza- polymer were added, the fluid became very thick that
tion operation enabled the plotting of the experimental it could not be used in the thin pipe system of our
and calculated results, for which the regression coef- experimental set-up (Fig. 10).
ficients could be obtained readily. The experimental data
were analyzed and a parametric curve fitting was carried 4.2. Experimental procedure
out by using the Microsoft™ Excel Goal Seek tool and
linear regression. The best estimate values of the cor- Two different water-based drilling fluids were used
relation parameters for mud-1, A1, B1, and for mud-2, as the circulating fluid. The tests that were conducted
A2, B2, are summarized in Table 6. included the saturation of the core samples with brine,
measuring the initial permeability of the core samples,
4. Description of experimental studies contamination of the cores by dynamic circulation of
drilling mud, and the measurement of the damaged
The overall objective of the experiments is to un- permeability of the core sample to brine. The fluids were
derstand the permeability impairment in sandstone circulated tangentially through the ends of the cores for
reservoirs during drilling mud circulation operations.
In this section, the experimental studies carried out for
this purpose are described. The experimental set-up used Table 6
is shown in Fig. 10. Fitting parameters for mud-1 and mud-2
Mud-1 A1 B1 C1 R2
4.1. Drilling fluids
30 min 67.7 0.018 0 0.90
60 min 73.8 0.023 0 99.9
Typical water-based drilling fluids containing solid 90 min 76.3 0.026 0 94.3
particles were used for permeability alteration tests.
Bentonite is the major clay in most water-based drilling Mud-2 A2 B2 C2 R2
fluids used in the oil industry. Two types of water-based 30 min 45.7 0.021 0 0.94
drilling fluids with different viscosities (0.10, 0.25 poise 60 min 55.2 0.033 0 0.94
90 min 63.8 0.032 0 0.98
[P]) were used. The major constituents of these fluids
 A.G. Iscan et al. / Journal of Petroleum Science and Engineering 58 (2007) 227–244 237

Fig. 10. Experimental set-up.

30, 60, and 90 min, respectively. The core holder was water under overburden pressure. Measure the wet
specially designed to enable the circulation of the weight of the sample.
drilling mud and injection of the brine successively 5. The compositions of the drilling fluids are prepared
without taking the core sample out of the core holder by adding 30 l water, 1948 g bentonite, 1480 g barite
(Fig. 11). The confining pressure was maintained by and CMC, and Polymer. Meanwhile, the mud is
using hand pumps, filled with hydraulic viscoelastic mixed continuously from the top by a 1500-rpm
Tellus oil. The circulation of the mud was achieved by mixer engine just as in the field. Then, determine the
an eccentric roller cone pump. The roller cone pump rheological parameters of the mud.
takes the fluid from one end and squeezes it between the 6. Open the outlet end of the core holder to the atmo-
eccentric rotor and stator. Thus, the fluid is pressurized. sphere. Start the circulation pump. By gradually open-
The advantage of this pump with respect to commonly ing the valve to inlet, squeeze the main by pass valve.
used centrifugal pumps is the enabling continuous flow Set the mud circulation pressure.
without pulsation. Generally used centrifugal pumps 7. Measure the filtrate volume accumulating in a grad-
may wear out of rotors and shafts because of the uated cylinder.
abrasive property of the solid particles in the water-
based bentonite muds. The brine injections were Table 7
performed by using a Cole ParmerR high-pressure– Drilling fluids and physical properties
low flow rate pulsation pump (Fig. 12). The following Drilling Mud additives Composition Mud
procedure was applied during the dynamic formation fluids weight
damage tests. (lb/gal)
Drilling Bentonite + Barite + Water Water = 350 ml 9.0
1. Dry the core samples in oven at 100 °C and measure fluid-1 Viscosity = 0.10 P Bentonite = 22.5 g
the dry weight of the core. Barite = 45.8 g
Drilling Bentonite + Barite + Water = 350 ml 9.0
2. Place core sample into core holder and vacuum at
fluid-2 Polymer (XT)⁎ + Water
400 psi overburden pressure. Viscosity = 0.25 P Bentonite = 22.5 g
3. Measure helium porosity and air permeability. Re- Barite = 45.8 g
vacuum the core samples. Polymer (XT) = 0.4 g
4. Measure liquid permeability and saturate the core ⁎ The least damage ratios obtained with the with the fluids including
sample with 1000 ppm high-concentration brine this polymer.
238 A.G. Iscan et al. / Journal of Petroleum Science and Engineering 58 (2007) 227–244

Fig. 11. Core holder.

8. Measure the return permeability by injecting the
brine–water from the reverse side.
5. Analysis of experimental data
5.1. Variation of permeability along core sample
The numerical simulator divides the core sample into
20 grid blocks. The permeability of the core sample is
calculated along the core length. The outputs are
obtained for each grid block at different time intervals
starting from 0 min to 90 min. Simulation results were
obtained for two different drilling fluids at three dif-
ferent filtration pressures set at 50, 100, and 200 psia
and nine specific time intervals ranging from 10 min to
90 min. Then, for every time interval, the variation of
non-plugging permeability with respect to core length
was plotted (Figs. 13–18).
The variation of the non-plugging permeability of the
core sample during the circulation of mud-1 at 50-psia
circulation pressure along the core sample is shown in
Fig. 13. There is a sharp slope linear trend within the
first half-hour of the circulation. Then, the rate of
increase decreases. On the other hand, there is a slight
jump in the increase of permeability after 6.2 cm away
from the circulation end of the core sample. But this
jump does not disturb the linear increase of the non-
plugging permeability along the core sample. The
reason of the discontinuity may be the non-homogeneity
within the core sample or pushing the particle bridges
with nozzle effect rapidly at a specific point of the core
from the circulation end at lower pressures. When the

Fig. 12. Drilling mud circulation system.

 A.G. Iscan et al. / Journal of Petroleum Science and Engineering 58 (2007) 227–244
Fig. 13. Variation of permeability along the core sample at 50 psia with
mud-1.
filtration pressure is increased, the jump in the increase
of the permeability is not observed. The maximum
variation in permeability took place at the end of the first
10-min period. At the end of the ninth 10-min period,
the variation in permeability along the core sample
became almost zero.
In Figs. 13 and 14, the variation permeability along
the core sample is drawn when the core sample is
subjected to the circulation of mud-1 at 100 psia and
200 psia filtration pressures, respectively. There is no
jump in the permeability value along the core sample
at these pressures. Considering the initial 10 min for
Figs. 13–15, the permeability increased from 0.0310 D
to 0.0335 D at 50 psia circulation amounting to an 8%
increase along the core sample. However, the increases
in permeability at 100 and 200 psia are 26% and 18%,
respectively. The increase in permeability along the core
sample is meaningful because the maximum deposition
of the particles takes place close to the wellbore during
drilling circulations. The permeability increases along
Fig. 14. Variation of permeability along the core sample at 100 psia
with mud-1.
239
Fig. 15. Variation of permeability along the core sample at 200 psia
with mud-1.
the length of the core sample linearly during mud cir-
culation. There may be some discontinuity at lower
filtration pressure in this linear trend due to the non-
homogeneity of the core sample and the characteristic of
the used mud. But this increase is a relative increase
with respect to the closest point to the circulation.
However, the increase along the core is neither a
stimulation nor improvement in permeability. It is just a
relative increase as it is moved away from the circulation
point. Even the most far point of the core sample has a
reduced permeability than the initial undamaged
permeability. In Figs. 14 and 15, the permeability
change becomes almost zero after 50 and 60 min,
respectively. Even though the filtration pressure is
doubled, the permeability of the core may no longer
be altered significantly. The variation of permeability of
mud-2 along the core samples is drawn in Figs. 16–18.
The increase slows down and becomes almost constant
between 100 and 150 psia. From this point on there is a
slight reduction in the damage ratio.
Fig. 16. Variation of permeability along the core sample at 50 psia with
mud-2.
240 A.G. Iscan et al. / Journal of Petroleum Science and Engineering 58 (2007) 227–244
Fig. 17. Variation of permeability along the core sample at 100 psia
with mud-2.
When water-based polymer mud is used during
circulation, the increase in permeability is also linear.
Nevertheless, unlike mud-1, the variation of permeabil-
ity with core length does not have a discontinuity point
even at lower pressures such as 50 psia (Fig. 16). The
slopes of the lines in the first 30 min are almost the same
with mud-2 at 50 psia. However, at later times, even
though the rate of increase in permeability has decreased,
it did not vanish even at higher pressures unlike mud-1
(Fig. 17). On the other hand, the rates of increase in mud-
2 simulations are less than that of mud-1. This may be
because of the higher viscosity and thixotropic property
of the mud with Polymer XT. As the circulation period
elapsed, the linear lines approached each other in the
three application pressures. As the filtration pressure
increased from 50 to 100 psia, the permeability differ-
ence between the circulation end and the far most end of
the core sample increased (Figs. 16 and 17). This differ-
ence is higher at 200 psia than at 50 psia, but slightly less
than the difference with 100 psia circulation (Fig. 18).
Fig. 18. Variation of permeability along the core sample at 200 psia
with mud-2.
Fig. 19. Damage ratio versus filtration pressure with mud-1.
5.2. Analysis of damage ratio and filtration pressure
A very common way of representing formation
damage is using the concept of damage ratio (DR). In
this part of the study, the experimental results of
permeability impairment due to mud circulation were
analyzed. The experimental data of the two different
mud samples obtained at eleven different filtration
pressures were plotted at three different circulation
periods, i.e. 30, 60, and 90 min (Figs. 19 and 20). The
damage ratio keeps an increasing trend until 110 psia
filtration pressure for mud-1, slightly reduces until
130 psia, and then becomes almost constant (Fig. 20).
As the circulation period increases from 30 to 90 min,
the damage ratio increases (Fig. 20). The maximum
damage ratio of 77% is reached with mud-1. The
maximum damage ratio with polymer added mud-2 is
about 65% at the end of the 90-min circulation period
(Fig. 20). The plugging particle concentration increases
in each grid block with time. They display a fluctuation
Fig. 20. Damage ratio versus filtration pressure with mud-2.
 A.G. Iscan et al. / Journal of Petroleum Science and Engineering 58 (2007) 227–244
during the circulation for about 20 min due to the
unplugging of formerly deposited particles by jet nozzle
effect. The damage ratio is less with the polymer-added
mud than the damage ratio with the conventional
bentonite mud. The damage ratio increases as the
filtration pressure increases until 125 psia. From this
pressure on the trend is almost flat for the three different
circulation periods (Fig. 20). The drilling fluids were
circulated at different filtration pressures. The clean core
plugs are saturated with brine. The permeability was
measured using the Darcy's flow equation. The major
substance in drilling fluids is bentonite. Bentonite is
naturally occurring clay which includes attapulgite and
montmorillonite. When the drilling fluid is circulated
across the core brine-saturated core plugs, the fine
particles invade the core plug. A breakthrough brine
flow is observed at the open end of the core plug due to
invading particles. Thus, the invading volume can be
controlled by observing the difference between
the pumped and returned flow rates of the circulating
fluid. The drilling fluid circulation system is depicted in
Fig. 12.
There is a minimum pressure observed in the study of
Gruber and Adair (1995). They conducted their ex-
periments at differential pressures between 0 and 40 kPa
(0 and 5.8 psi). They obtained the initial inflow between
10 and 20 kPa in highly permeable samples. However,
these are threshold values where the invasion effect is
seen. Thus, the invasion into porous media starts at low
filtration pressures. In Figs. 19 and 20, the damage ratio
is very close to 0 until 10–20 kPa (1.45–2.90 psi). Thus,
these results are in accordance with the analysis of
Gruber and Adair (1995).
After plotting the experimental data for the two
different mud types at three different circulation pres-
Fig. 21. Linear regression of exponential damage ratio model for
mud-1.
241
Fig. 22. Linear regression of exponential damage ratio model for
mud-2.
sures, an exponential function was proposed (Eq. (9)). In
this model, the damage was expressed as an exponential
function of the filtration pressure. The calculated dam-
age ratio values were drawn together with experimental
data (Figs. 19 and 20). The proposed exponential model
fits the experimental data very well. The regression
coefficients of the linearized form of Eq. (9) for mud-1
are 0.90, 0.99, and 0.94 for 1/2-, 1-, and 1.5-h periods,
respectively (Fig. 21). The regression coefficients of the
linearized form of Eq. (9) for mud-3 are 0.94 and 0.98
for 1.0 and 1.5 h periods, respectively (Fig. 22). The
regression line for 0.5 h data almost overlaps with the 1-
h data; hence, it was not shown to avoid a crowded
figure. The correlation coefficients are close to 1.0 and
the proposed model is satisfied. Thus, an accurate
approach is obtained to estimate the reduction in
permeability in terms of the filtration pressure. This
correlation (Eq. (10)) enables an accurate estimation of
the near wellbore formation damage as a function of
pressure. The simplicity of the correlation makes it more
attractive for field uses.
6. Conclusions
The conclusions of the present permeability damage
ratio problem investigated both experimentally and
numerically can be summarized as follows:
1. Damage ratio increases with the increase of
circulation periods. The damage ratio is higher
with mud-1 than mud-2 under identical experimen-
tal conditions. This is also verified by the model
proposed by Eq. (9). Thus, the water-based Polymer
XT mud is less damaging than the conventional
bentonite mud. The damage ratio varies with the
242 A.G. Iscan et al. / Journal of Petroleum Science and Engineering 58 (2007) 227–244

filtration pressure according to the model proposed σ Particle concentration, ppm

in this paper. α Dimensionless constant
2. Application of plugging and non-plugging path- μ Viscosity, cP
ways increases the accuracy of the porosity and τ Shear stress, Pa
permeability analyses in porous media. The equa- η Fraction of uncovered deposits
tions proposed by Civan (2000) and improved and ϕ Porosity
implemented by Civan and Nguyen (2005) ade- ρ Particle density, kg/m3
quately describe the contamination and clean-up T Temperature, °C
cycles. n1 Arbitrary exponential constant
3. The non-plugging porosity decreases linearly by n2 Arbitrary exponential constant
8.5% through out the whole circulation period. The c1 Plugging side constant, (ppm × s)− 1
recovery in non-plugging porosity is about 6% c2 Plugging side constant, (ppm × s)− 1
during the backflow with water from the aquifer. c3 Deposition side constant, (ppm × s)− 1
The change in the plugging porosity is numerically c4 Erosion side constant, s− 1
very low and negligible. The non-plugging perme-
ability decreases linearly in 106 min by 35%. The Subscripts
recovery in the non-plugging permeability is 21% of cr Critical
the initial permeability. s Superficial
4. The average non-plugging and plugging deposition p Plugging
rates are higher than the average non-plugging and np Non-plugging
plugging clean-up rate. The fraction of the plugging e Entrainment
porosity to overall porosity (fp) increases with time
along the core length within the first five grid blocks Acknowledgement
of the core sample.
5. The overall permeability of the core samples reduces The first author gratefully acknowledges the fellow-
as expected. The permeability along the core sample ship provided by The Scientific and Technical Research
increases with the increasing pressure. The major Council of Turkey (TUBITAK).
increase in permeability takes place in the early
30 min of the circulation along the core sample. The Appendix A. Description of equations used by
increase is almost zero with mud-1 where as the Civan and Nguyen (2005)
increase continues with mud-2 even at later circula-
tion periods. The numerical simulator considered in this study
uses the following equations. These equations originate
Nomenclature from Civan (2000) with some modifications after
A Area, cm2 Gruesbeck and Collins (1982). Civan and Nguyen
DR Damage ratio, % (2005) obtained a numerical solution for these
Dp Particle diameter, μm equations.
ke Entrainment rate constant, cm− 1 The plugging and non-plugging porosities are given
f Fraction of porous media containing the corre- by subtracting the fraction of depositing particles from
sponding pathways the initial values of the plugging and non-plugging
i
krel Rate of particle release constant, m2 porosities:
kp Surface deposition rate constant
K Permeability, D /p ¼ /po −ep ðA1Þ
L Core length, cm
N Number of pore throats per unit bulk volume /np ¼ /npo −enp ðA2Þ
q Filtrate flow rate, cm3/s
qs Superficial velocity, m/s Fractions of plugging and non-plugging porosities to
P Pressure, psia overall porosity ( fp) and ( fnp) are given by:
U Filtrate velocity, cm/s
Xk Rate constant, m2 fp ¼ /p =/ ðA3Þ
ε Total fraction of depositions in the both plug-
ging and non-plugging pathways, cm3/cm3 fnp ¼ /np =/ ðA4Þ
 A.G. Iscan et al. / Journal of Petroleum Science and Engineering 58 (2007) 227–244 243

The instantaneous porosity is given as the sum of the values of the permeability, and bulk fraction of de-
plugging and non-plugging porosities as: positing particles for plugging and non-plugging path-
ways, respectively as the following.
/ ¼ /p þ /np ðA5Þ


Kp ¼ Kpo exp −að/po −/p Þn1
The instantaneous fraction of depositing particles is
¼ Kpo expð−aenp1 Þ ðA16Þ
given as the sum of the plugging and non-plugging
porosities. Thus
Knp ¼ Knpo ð/np =/npo Þn2
e ¼ ep þ enp ðA6Þ ¼ Knpo ð1−enp =/npo Þn2 ðA17Þ
The partial derivative of fraction of depositing par- The variation of the bulk fraction of the depositing
ticles with respect to time are given as the product of the solids along the core length is given by
surface deposition constant, velocity, particle concen-
Z
tration and the plugging porosity as 1 L
eðt Þ ¼ ep ðt Þ þ enp ðt Þ ¼ ½ep ðtÞ þ enp ðtÞdx ðA18Þ
L 0
Aep
¼ kp up rp /p ðA7Þ
At Darcy's equation is arranged for obtaining perme-
The initial condition for Eq. (A7) is given by. ability in the system at any time and any location, as

ep ¼ epo ; t¼0 ðA8Þ q Kðx;tÞ AP

u¼ ¼− ðA19Þ
A l AX
The partial derivative of fraction of depositing
particles with respect to time is given as the product of Eq. (A19) is integrated between the inlet and outlet
the surface deposition constant, velocity, particle pressures of the core plug in Eq. (A20). Pressure
concentration and the plugging porosity by difference at the beginning (t = 0) is calculated in
Eq. (A21). Eqs. (A20) and (A21) are divided side by
Aenp side. Eq. (A22) is obtained.
¼ kp unp rnp /np2=3 −ke enp /np2=3 ge ðsnp −scr Þ ðA9Þ
At Z Pout
The initial condition for Eq. (A9) is given by DPðt Þ ¼ Pout −Pin ¼ APðx;tÞ
Z L Pin

enp ¼ enpo ; t ¼ 0 ðA10Þ dx

¼ −lu ðA20Þ
0 Kðx;tÞ
The fraction of uncovered deposits is represented
exponentially as a function of non-plugging depositing L
DP0 ¼ DPð0Þ ¼ −lu ðA21Þ
particles. K0
Z
ge ¼ expð−kenp Þ ðA11Þ DPðtÞ K0 L
dx
¼ ðA22Þ
DP0 L 0 Kðx;tÞ
Darcy's flow equation is considered for the perme-
ability measurement.
References
KADP
q¼− ðA12Þ
lL Al-Riyamy, K., Sharma, M.M., 2004. Filtration properties of oil-in-
water emulsions containing solids. SPE paper, vol. 73769,
Flow velocity across the surface of the core is given pp. 164–172 (Sept.).
Argiller, J.F., Audibert, A., Longeron, D.G., 1999. Performance
by Eq. (A13) evaluation and formation damage potential of new water-based
drilling formulas. SPE Drill. Complet. 14, 266–273.
u ¼ q=A ðA13Þ
Bailey, L., Boek, E., Boassen, T., Selle, O., Longeron, D., 1999.
i Particulate invasion from drilling fluids. SPE paper, vol. 54762,
snp ¼ k i ð8Unp =Dnp Þn ðA14Þ pp. 1–10 (May).
Civan, F., 2000. Reservoir Formation Damage – Fundamentals,
Vnp ¼ unp =/np ðA15Þ Modeling, Assessment, and Mitigation. Gulf Pub. Co., Houston,
TX. 742p.
The plugging and non-plugging permeabilities are Civan, F., Nguyen, V., 2005. Modeling particle migration and deposition
represented with exponential functions in terms of initial in porous media by parallel pathways with exchange, Chapter 11,
244 A.G. Iscan et al. / Journal of Petroleum Science and Engineering 58 (2007) 227–244
Handbook of Porous Media, Second Edition, Vafai, K. (Ed.), CRC
Press, Taylor and Francis Group, Boca Raton, FL: 457–484.
Dalmazzone, C., Audibert-Hayett, A., Quintero, L., Jones, T.,
Dewatinnes, C., Jansen, M., 2006. Optimizing filtrate design to
minimize in-situ and wellbore damage to water-wet reservoirs
during drill-in. SPE Prod. Oper. 66–73 (Feb.).
Ding, Y., Longeron, D.G., Audibert, A., 2004. Modeling of near-
wellbore damage and natural clean up of horizontal wells drilled
with water-based drilling fluids, paper SPE 73733. SPEJ 252–264
(Sept.).
Erwin, M.D., Pierson, C.R., Bennion, D.B., 2005. Brine imbibition
damage in the Colville River field, Alaska. SPE Prod. Facil. 26–34
(Feb.).
Gruber, N.G., Adair, K.L., 1995. New laboratory procedures for
evaluation of drilling induced formation damage and horizontal
well performance. J. Can. Pet. Technol. 34, 27–32 (May).
Gruesbeck, C., Collins, R.E., 1982. Entrainment and deposition of fine
particles in porous media. SPE AIME 847–856 (December).
Krueger, R.F., Vogel, L.C., Fischer, P.V., 1967. Effect of pressure
drawdown on clean-up of clay- or silt-blocked sandstone, SPE
paper 1605. JPT 397–403 (March).
Longeron, D.G., Alfenore, J., Poux-Guillaume, 1998. Drilling fluids
filtration and permeability impairment: performance evaluation of
various mud formulations. SPE paper, vol. 48988, pp. 237–251
(Sept.).
Parn-anurak, S., Engler, W.T., 2005. Modeling of fluid filtration and
near-wellbore damage along a horizontal well. J. Pet. Sci. Eng. 46,
149–160.
Van der Zwaag, C.H., 2006. Benchmarking the formation damage of
drilling fluids. SPE Prod. Oper. 40–50.
Vitthal, S., Sharma, M.M., Sepehrnoori, K., 1988. One-dimensional
formation damage simulator for damage due to fines migration.
SPE Paper 17146 Presented at the 1988 SPE Formation Damage
Control Symposium, California, Feb. 8–9, pp. 29–42.
Wolfram Research Inc. MathematicaR Software Version 5.2.