Formation Damage Caused by

Cement Filtrates in Sandstone Cores

X.M. Yang and Mukul M. Sharma, SPE, U. of Texas

Summa". A series of core damage experiments was conducted with filtrates containing various cement additives, such as dispersants and fluid-loss additives. Results clearly show that the extent and location of the permeability impairment varies, depending on'
the relative rate of particle precipitation and the residence time in the core. Results of the core-flow experiments are interpreted in
conjunction with particle-size distribution and concentration results obtained in batch experiments. Variations in the average particle
size and the number concentration of particles with time in various compositions of cement filtrates are reported. These prove to be
useful indicators of the extent and location of the damage observed in the cores.

Introduction
Several factors motivate the study of the possible damage that cement filtrates can cause in the near-wellbore region. The cement
filtrate has a very high pH (> 12) and a high concentration of divalent ions. The high pH tends to cause clay and fines to be released
and to migrate into pore throats, where they cause damage. On the
other hand, the high concentration of divalent cations (mainly
Ca2+ and Mg2+) prevents fines movement and actually "cements" the fines in place. Which effect will predominate is not
evident.
In addition to the physicochemical effects, chemical reactions between the high~pH filtrate and the silica grains may result in the
precipitation of calcium silicate hydrate. This probably will not be
a major cause of damage because the reaction and precipitation processes are slow.
A potential source of damage that is probably the greatest cause
for concern is the precipitation of insoluble salts (e.g., CaC0 3 and
CaS04) in the formation. Salts can cause damage in a region beyond the perforation depths that may be very significant, depending on the quantity of precipitate. The work reported here
investigates this possibility in depth. Most research on cement fIltration has focused on hole stability, flow properties of the cement
slurry, "flash setting," and cement quality in the annular space
between casing and wellbore. Few papers have been published on
the possible effects of cement filtrate coming into contact with the
reservoir rock and fluids. The extent and mechanisms of the damage
are not very clear. Even the basic conclusions from past publications have been somewhat contradictory. 1-3
Cunningham and Smith 1 investigated the components of cement
filtrates (Table I) and their influence on formation permeability.
Their results indicated little permeability damage resulting from
clay swelling and migration and lime/silicate reaction caused by
cement filtrate. They concluded that the low lime content of the
cement filtrate and the limited ability of filtrate to pass from the
slurry to the formation restrict damage from the lime/silicate reaction. Krueger 2 also found no appreciable permeability changes in
Berea sandstone cores exposed to cement fIltrate. However, growth
oflarge 'calcium carbonate or calcium sulfate crystals in pore openings and other fines on the pore wall were shown by a scanningelectron-microscope (SEM) photomicrograph of a core invaded by
cement filtrate. Smith et al., 3 however, found serious brine permeability reduction from 60 to 90% in cores from various formations after the cores had been invaded by cement filtrate. They
postulated two mechanisms that could cause damage due to cement
filtrates: fine solids removed from the rock matrix and lime/silicate reaction.
The cement filirate, which penetrates into permeability formations under the high differential pressure between the annular space
and the formation, has ionic components related to the chemical
components of the cement and the hydration reactions. The major
constituents in the aqueous phase in contact with hydrating cement
are calcium silicates (Ca 2 + , OH-), calcium aluminates [Ca2+ ,
Al(OH) 4'1, calcium sulfates (Ca2+, SOf-), calcium carCopyright t 991 Society of Petroleum Engineers

SPE Production Engineering, November 1991

bonateslbicarbonates (Ca2+ , c0l'- , HC03), and alkali sulfates

+ S02(K + , Na,
4)'
As discussed later, on the basis of Cunningham and Smith's data.
the cement filtrate may be supersaturated in calcium carbonates,
calcium hydroxide, and calcium sulfate. 4
When cement slurry is pumped into a well and passes by permeable formations, the aqueous phase, which contains high concentrations of various ions, is squeezed into the formation under
differential pressure. Unlike cement filtrates, mud filtrates do not
produce insoluble particulate precipitates. Therefore, muds form
external filter cakes on the face of the formation exposed to the
mud and a relatively shallow internal cake is formed in the pores.
Permeability damage caused by particles is therefore limited to a
few inches from the wellbore. For cement filtrates, however, the
fluid enters the formation and can produce precipitate particles deep
inside the pore space to block pore throats further into the formation.
Because the cement slurry always contains additives, they also
playa role in the formation of precipitate particles during cement
fIltration. Lignin derivatives, cellulose derivatives, organic acids,
and synthetic polymers are the most commonly used additives. These
additive ions and molecular chains can promote or hinder the formation of precipitate particles in the filtrate.
The presence of calcium carbonate and calcium sulfate particles
in cement filtrate during filtration is a likely source of formation
damage. The extent of precipitate formation, the permeability
damage caused, and the influence of cement additives on this damage
will provide guidelines for how significant this problem may be.

Experimental
Particle-Size-Distribution Measurements. Class H cement is used
in the tests and mixed separately with one or two additives in a
water/cement ratio of 0.5 to form nine different cement slurries
(Table 2).
An instrument that has a laser-based sensor, particle counter, and
tank sampler for particle-size measurements in liquids is used in
the tests. The instrument can acquire particle number data in a 25mL test sample in six particle-size ranges from 0.5 to 25 I'm. In
our tests, the size ranges are set to be 1.1, 2.02, 5.1, 10, 20, and
25.7 I'm. The concentration of particles in the sample must be well
above 3,000 counts/mL to obtain data with low coincidence error.
The cement filtrate gathered frQm an API fIlter press has a particle concentration much higher than 3,000 counts/mL, and particles continuously precipitate from it. The instrument cannot directly
measure the particle-size distribution in the cement filtrate because
of the high concentration of particles. Before the te,st is started, a
saturated CaC0 3 solution is made up to dilute the cement filtrate
and to suspend the CaC0 3 particles crystallized from the filtrate
without any dissolution during particle-size measurement. Before
use, the saturated CaC0 3 solution is filtered several times to remove undissolved CaC0 3 and other particles.
After the cement slurry is blended for 5 minutes in a 0.5 water/cement ratio, the filtrate is gathered from the API filter press under
lOO-psi pressure. Immediately, 0.25 mL of the filtrate is put into
125 mL of CaC03 solution (0.2 % concentration) and the particle399

TABLE 1-ANALYSIS OF CEMENT FILTRATE FROM REF. 1

Salt
(%)

Specific
Gravity

pH

0
5
10
18
Saturated

1.008
1.048
1.080
1.121
1.206

12.6
12.1
12.0
11.8
11.6

API Class A cement-5.2 gal water/sack

Concentration (mg/L)
Mg2+
OHCO~Ca 2+
1,013
853
728
614
274

168
180
168
180
174

860
1,685
2,060
1,675
650

0
0
0
0
105

SO~

CI-

4,950
7,000
9,600
8,400
7,400

20
31,250
59,750
105,500
185,000

TABLE 2-COMPOSITION OF CEMENT SLURRY

Class H cement-O.5 water/cement ratio

Composition
Cement + fresh water
Cement + cellulose-based filtrate
Cement + synthetic polymer filtrate-control additive
Cement + synthetic polymer dispersant additive
Cement + lignin retarder Additive 1
Cement + lignin retarder Additive 2
Cement + organic acid Dispersant 1
Cement + organic acid Dispersant 2
Cement + cellulose + lignin Retarder 1

to confining pressure

Fig. 1-Schematlc of equipment used: 1-API filter press; 2metering pump; 3-pressure transducers; 4-core holder.

size distribution is measured. At the same time, the rest of the fIltrate
is continuously stirred to simulate continuous flow. Particle-size
distributions of the filtrate particles suspended in CaC0 3 solution
at some known ratio are measured at different stirring times (10,
20, 30, 60, 120, and 180 minutes).
From the tests, the number of particles in six particle-size ranges
and the cumulative number of all particles in the cement filtrate
can be obtained as a function of time. All tests were d~ne at room
temperature.
Permeability-Reduction Measurements. Test Device. The test
device (Fig. 1) consists of an API filter press, a core holder, and
a set of pressure transducers and demodulators to monitor the permeability across different sections of the core. If cement fIltrate
is injected into the core by a pump, the fluid filtrate sits in an accumulator for a long time while nucleation, precipitation, and other
reactions are taking place. This will almost certainly affect the results of permeability measurements. In our device, the filter press
is connected to the core holder so that the cement filtrate can directly
enter the core under a constant pressure with minimal delay. This
closely simulates the process of filtration in a wellbore. Three pressure transducers are placed along the length of the core holder to
monitor the pressure changes in three sections of the core. As a
result, we can observe the particle formation in the core and the
resulting permeability damage during the entire test.
400

Concentration
(%)

Viscosity at 74F
(cp)

1
1
0.5
0.5
0.5
0.5
0.5
0.5

1.09
3.33
6.87
1.12
1.18
1.12
1.12
1.10
2.99

Core Preparation and Test Cement Slurries. Berea sandstone

cores were cut parallel to the bedding plane, 1 in. in diameter and
5 in. in length. Particles attached to the cut surface of each core
were removed by a vacuum. The cores were then heated at 212F
for 20 hours to remove any fluid from the pore space.
Table 2 lists the compositions of the cement slurries used in the
tests. Their viscosities were also measured with a Fann TM viscometer and are given in the table.
Test Procedure. Before the test, the core holder and the flowlines
were cleaned carefully and the inlet line filled with 30,000-ppm
NaCI brine. The dry core was placed into a Hassler sleeve at
700-psi confining pressure. The core was then vacuum saturated
with 30,000 ppm NaCI brine. The core porosities range from 18
to 20%. The PV for the 5-in. cores range from 11 to 13 cm 3 .
The brine permeability is measured by injecting 30,000-ppm NaCl
brine with a constant-rate pump at 1 mLlmin. After blending for
5 minutes, 250 mL of cement slurry was poured into a filter press
and a constant pressure of 100 psi was applied to force 5 to 10 PV
of cement fIltrate into the core. Because a cement filter cake is continuously built up in the filter press and the permeabilities in each
section of the core start to change during the experiment, the flow
rate and the pressures always have a decreasing trend. They were
recorded as a function of time during cement filtration. This procedure ensures that any permeability impairment in the core is a result of the cement filtrate rather than solids from the cement. The
brine permeability after cement-filtrate injection was measured. In
this period, CaC0 3 and CaS04 particles formed in the pore space
during cement filtration may dissolve in the brine. This effect will,
however, be negligible because of the low solubility of both these
deposits in the brine. The brine flow rate was raised to 1.5 mLlmin
and the permeability change was measured again to investigate the
effect of flow rate on permeability. Finally, NaCI brine was injected in the reverse direction at 1 mL/min to obtain a return brine
permeability .
Experimental Results
The particle-size distribution and cumulative number of particles
in the cement filtrate at different stirring times can be obtained directly from particle-size measurements. The data on particle concentration in milligrams per liter and particle average diameter can
easily be derived from these data.
The data from permeability-damage experiments are shown in
conjunction with the particle-size data. kJo k2' and k3 denote the
fluid permeabilities of the input, middle, and output section of a
core, respectively. k is the total permeability of a core. PermeabilSPE Production Engineering, November 1991

!C
a:

...

oJ

10 7

10 7

oJ

:IE
~

:i
::;)

10

lOS

il:

10

I!!I

oJ

NO STIRRING
10MIN.
20 MIN.
30 MIN.
SO MIN.
120 MIN.
180MIN.

105

10 4

.5

10 3

E
~

18

10 2

NO STIRRING
10 MIN.
20 MIN.
30 MIN .
60 MIN
120 MIN.
lS0MIN.

10 1

~
a:

10 0

...
C

1.1

5.1

2.02
(0)

1.1

25.S

20.0

10.0

(a)

.,...

3.5
3.0

2.5
1.5

..

2.4

I!

2.2

,..'"

m~

200

100
(b)

1 00

Time min.

(b)

4~----------~--------~

150

200

Tim. min.

...------------r---------~..,

3+-----------t-------~_1

2+-----------t------r--_1

8c:

2.0

1.S

...:

",---

1.0

25.6

I! 2.0

20
I'm

""

4.0

""

10

E 2.S

4.5

5.1

Partlcl. Siz.

2.S

5.0

2.02

PARTICLE SIZE I'm

100
(c)

...:

200

100 +------+----~-_l

50+--------4-~~----_l

Time min.

Time min.

20,--..,------,---

Kl change

..

200

100
(c)

K2 change
K3 change

'"

c:

c:
~

z:

-20

I)

:;;

:;;

....

-st---j~~~~ft-~r---~~~~f-~~------l

-80+---j-----4-~~~-4-------}.~~------_4

...
E

Sackflush
300

200

100
(d)

Time

400

min.

40 +---I-I---''''l--\

200

400
(d) Time

sao

soo

min.

Fig. 2-Partlcle-slze and permeability Impairment results for

cement with no additives.

Fig. 3-Partlcle-slze and permeability Impairment results for

cement with a cellulose-based filtrate-control additive.

ity changes are presented as a percentage change compared with

the original brine permeability for each section. The following parameters are defined for convenience: nc=cumulative number of
particles in 1 mL of filtrate; n 1 = number of 1.1-lLm-diameter particles in 1 mL of filtrate; n2 =number of 2.02-lLm-diameter particles in 1 mL of filtrate; ns = number of 5.1-lLm-diameter particles
in 1 mL of filtrate; nlO=number of 1O-lLm-diameter particles in
1 mL of filtrate; n20 = number of 20-lLm-diameter particles in 1 mL
of filtrate; d=average particle diameter, ILm; and Cp =particle concentration, mg/L of filtrate.

filtrate passes through the core in 20 minutes at 0.9 mLlmin under

a low differential pressure because of the high fluid-loss rate. The
permeabilities of all sections across the core see no reduction because particles do not form in large numbers and cp and dare also
low. After 1 PV injection, k3 begins to drop, which indicates that
particle precipitation first occurs in the downstream section of the
core and blocks some pore throats. About 2 PV later, kl starts to
decrease and k3 keeps declining, but k2 seems to remain the same.
Perhaps the filtrate is filtered by an internal filter cake in the first
section of the core so that only small particles enter the second section. We have no direct evidence of this. As particles grow while
flowing through the second section, k3 is further reduced. At the
end of filtration, the filtration rate is greatly reduced because the
cement filter cake in the filter press becomes thicker and more compact. As a result, the time needed to pass through each section of
the core becomes longer. kl' k2' and k3 are all significantly reduced as precipitation occurs in each section. Finally, Section 1
is most seriously damaged (97%), Section 3 (k3) much less so
(47%), and Section 2 (k 2 ) only slightly (18%).
To measure the brine permeability after the filtrate flow, 3 PV
of30,OOO-ppm NaCI brine is injected through the core. During this
period, kl gradually recovers as a result of dissolution or redistribution of CaC0 3 particles deposited in the pore space. Meanwhile,

Freshwater Filtrate. After 180 minutes, the average particle size

and nc increase significantly. It is quite likely that this increase
occurs sooner in the pore space where many more nucleation sites
are available. Fig. 2 shows the results for a pure cement filtrate
with no additives. This filtrate with no stirring has a very low particle concentration. This implies that the precipitation rate of
CaC0 3 is low. nc is in the 104 range, and d is 1.48 ILm because more than 90% of the particles are below 21Lm. As the filtrate
is stirred, new particles begin to form and the precipitation rate
increases after 100 minutes.
Fig. 2d shows the results of the permeability test with freshwater
filtrate and Berea sandstone. During filtration, the first PV of the
SPE Production Engineering. November 1991

401

10 7

10 6

E
.E

105

10 4

10 3

z"

w
....

10 6

105

<
II:

NO STIRRING
10MIN.
20 MIN.
30 MIN.
80 MIN.
120 MIN.
180 MIN.

iL

....

:2

10 4

10 3

:Ii
~

10 2

10 1

w
....
(,)

10 1

10 0

>=
II:

10 2

1.1

2.02

(0)

5.1

10

20

Partlcl. Size

3.0

"-

2.S

r.. m//
<

NO STIRRING
10 MIN.
20 MIN .
30MIN.
80 MIN.
120 MIN.
180 MIN.

10 0

<
Do.

25.6

5.1

2.02

1.1

(0)

I'm

10.0

20.0

25.6

PARTICLE SIZE I'm

4.0

./

E
Q

~ 3.5

..

~
~ 2.0

>

2.4

3.0

:: 2.5

2.6

.-/

1.5
1.0

2.2
0
(b)

200

100

100
(b)

Time min.

200

Time min.

15

200

.c

10

100

(,)

0
0

100
(c)

200

(c)

200

Tim. min.

20
~

....

...c

Oi

z:

Brln.
0

-20

-40

co

-60

Do.

-so

z:

(,)

(,)

...

100

Time min.

100
~

:::;;

:;;

..

E
Do.

Kl change
K2change
K3change

-100

-100
0

100

200

300
(d)

400
Time min.

500

600

100

300

200
(d)

400

Time min.

Fig. 4-Partlcle-slze and permeability Impairment results for

cement with a synthetic polymer filtrate-control additive.

Fig. 5-Partlcle-slze and permeability Impairment results for

cement with a synthetic polymer dispersant additive.

k2 and k3 show no apparent change. Brine permeabilities in each

to 5 x 10 5 . The increase in Cp indicates that a large number of particles still precipitate from the filtrate. On the other hand, the particles that adsorb polymeric molecules are easily coagulated by
collision between themselves during stirring to form aggregates.
In the permeability test, all the filtrate effective permeabilities
in each section are lower than the original brine permeabilities during
cellulose filtrate penetration and continuously decline. No obvious
indication can be found for the section in which the particles prefer to precipitate and deposit because all three permeabilities are
reduced in the same manner. A 0.68- to 0.3-mLlrnin filtrate flow
rate is used in the test so that the filtrate can pass through the whole
core in about 20 to 40 minutes. A longer residence time may have
caused addition damage judging from the particle-size and concentration results.
After 4 PV of brine injection at 0.25 mL/rnin, Section 3 has the
most serious damage (-96 %) while k 1 and k2 have the same
degree of damage (-63 to -69%). Therefore, it is clear that more
serious particle damage does occur in the downstream sections of
the core. This situation remains the same after the flow rate is increased to 0.5 mL/min. kl and k2 recover slightly as more filtrate
is removed from the core. After 4 PV of reverse brine injection,
Section 3 still has the most serious damage and Section 1 is only
slightly damaged.

section are substantially reduced, especially in the input and output sections. After 3.5 PV of reverse brine injection, the change
in k 1 turns to a positive value, which indicates that damage had
been caused by the formation of an external filter cake of particles. However, the change in k3 is -62%. This clearly demonstrates the ability of precipitated solids to cause irreversible damage
deep in the formation. It is difficult to conjecture what would occur
beyond the 5-in. core length used here. It would certainly be a complex function of the flow and precipitation rates.
Filtrates With Fluid-Loss Additives. Two kinds of fluid-loss additives are used in our tests: a cellulose derivative and a synthetic
polymer. Each has a different effect on the precipitation process.
Figs. 3 and 4 show the results from these two filtrates.
Particle precipitation is very high for the cellulose filtrate. nc remains in the 10 6 range. Cp increases from 10 to 50 mg/L and d
remains between 2 and 3 /lm. The number of particles in each size
range rapidly increases so that nc increases to more than 10 6 . The
addition of this cellulose additive results in precipitated particle concentrations at least 10 times higher than those for the pure cement.
The synthetic polymer behaves similarly, although far fewer particles are precipitated. As the fIltrate is stirred, nc rapidly increases
402

SPE Production Engineering, November 1991

10 6

105

10 4

10min.
20 min.
30
min.
~

E
~

=
E
:>

10 3

10 2

l-

f--

t---

10 1

l-

f--

t---

;:

0..

..

105

No Stirring

10 4

GO min

.5

10 3

120 min

E
:>

10 2

180 min.

NO STIRRING
10MIN.
20 MIN.

~ 30 MIN.

EI

60MIN.
120 MIN.
180 MIN

10 1
10 0

10 0
1.12.025.1
(a)

!!
>

Particle Size

2.02
(a)

I'm

~ 5

....
.
c

1.1

2025.6

5.1

10

20

Particle Size

25.6

I'm

5.0

10

4.5

..............

-.........

3.5

!! 2.5

4.0

~ 3.0

"

~ 2.0

~.--

1.5
1.0

100

(b)

10

.. 8
E
6

200

(b)

...

'"

.2

100
(c)

200

200

.f
o

100
Time min.

Time min.

~
0

Time min.

(c)

laO
Time mig.

200

20
40~--~---r--------

--r----~---r---r------_,

Brine

11-

'"~

.c
u

20

Backflu.h

t----t~IiI~~-;::::j;;pr----r_---I-

Kl change
K2change
K3 change

O~--~~~-+----~~d.----r----I

11-

'"
0:

m -40

"~

i-~+---~--~----~~~~--~~".r--;-------~

0..

o.._ro+-__ __ __ __-+__ ____ __ __+-__ __

~

100

300

200
(d)

Time

20

~
:g

~-2O+----t---t--~---i--;-~-it---~~~~----~
:;

li

.c

400

SOD

min.

60
Brine
80
0

100
(d)

200
Time min.

300

400

Fig. 6-Partlcle-slze and permeability Impairment results for

cement with a lignin retarder additive.

Fig. 7-Partlcle-size and permeability impairment results for

cement with a lignin derivative retarder Additive 2.

The synthetic polymer f1l.trate causes less damage to core permeability than the cellulose filtrate. However, the damage is more
severe than with pure cement. The cellulose f1l.trate causes serious
permeability damage deep in the formation. In the synthetic polymer f1l.trate test, the most serious permeability damage takes place
in the inlet section. We can conclude that the f1l.trate can form aggregates in the early stages of f1l.tration, thus forming an internal
f1l.ter cake in the first section of the core to block most of the flowing area.
After 5 PV of brine injection, kl drops 98%, k3 71 %, and k2
35%. k3 remains unchanged when the flow rate rises from 0.5 to
1 mL/min. But k2 and k3 have some degree of recovery, perhaps
as a result of particle dissolution of the precipitate and of the removal
of any excess adsorbed polymer. After 6 PV of reverse brine injection at 1 mL/min, the internal f1l.ter cake is partially removed
so that kl changes to -60%, while k2 drops to -42%.

the concentration of particles increasing to 12 mg/L. This is significantly smaller than the quantity of precipitate obtained with the
f1l.trate-Ioss-control additives. As such, the permeability reduction
is also less severe.
After 5 PV of brine injection, the permeability for each section
after the f1l.trate treatment is not significantly reduced, although the
fIltrate permeabilities for the first ana third sections show some smaIl
reduction. Also, increasing the flow rate and changing the flow
direction have no apparent effect on the brine permeabilities.
Figs. 6 and 7 show the results of the lignin derivatives and
f1l.trates. Both dispersant additives can increase the setting time of
a cement slurry. The lignin derivative causes a small number of
large precipitate particles to form quickly (d=5 ILm). This causes
the permeability in the most upstream section to be reduced most
significantly. Subsequent flow with brine causes some of these particles to dissolve partially and plug up flow channels farther downstream. Backflushing restores the permeability of each section to
~20% reduction.
The other lignin derivative used behaves similarly. The average
particle size here, however, is substantially smaller (d=2.5 ILm).
The permeability of Section 1 is reduced the most but is completely restored after flowing brine and backflushing (Fig. 7).
Figs. 8 and 9 show the results of two organic acid filtrates. It
can be seen that these additives somewhat restrained particle precipitation. As a result, the brine permeabilities of each section after

Filtrates With Dispersant Additives. The dispersant additives used

in our tests include a synthetic polymer, two lignin derivatives, and
two organic acids. In the cement slurry, they improve dispersion
of cement particles, reduce the plastic viscosity of the cement slurry, and reduce the mixing-water volume. Some, such as the lignin
derivatives, also act as retarders.
Fig. 5 shows the results for a synthetic polymer dispersant fIltrate.
Particle precipitation occurs steadily for more than 80 minutes with
SPE Production Engineering, November 1991

403

10 6

W
0-

lOS

10 4

.:

10 3
10 2
'0

I---

fH

'

II

10

'"
!:i
a:

NO STIRRING
10MIN.
20 MIN.
30 MIN.
60MIN.
120 MIN.
180 MIN

lOS

G:

::I

10

!:

10 3

:i
:)
z

10 2

....
(,)

t---

....

10 1

j:

a:

10 0
1.1

2.02

(a)

5.1

10

Particle Size

20

10 0

'"...

25.6

1.1

'" 4.0

2.5

'" 2.0
1.5

f 3.5
3.0
=2.5

.."....-

..

./

'" 2.0
1.5

1.0

1.0
100
Time min.

(b)

200

100

(b)

200

Time min.

12

15

'ij,

U.

E 10

:
c

10

6
4

0
0

(c)

100
Time min.

200

"'"

./

-----

100

(c)

200

Time min.

200;-----~---r------~_r--------_,--------~

20

Brlna

#
0

cco

20

(,)

.c:

.c:

25.6

e 3.0

f!
...

.,m

5.0

'" 4.0
3.5

:ii

20.0

E 4.5

(,)

10.0

PARTICLE SIZE

(a)

5.0

cco

5.1

2.02

.,m

E 4.5

NOSTR'IING
10MIN.
20 MIN.
30 MIN.
60 MIN.
120 MIN.
180 MIN.

100

Cement

Filtrate

Brine

Backfluah

40

Kl change
K2change

60

K3 change

Brine
-80
0

100

200

300

400

500

(d) Time min.

:ii

...
-100 +---~-'---+----~-'--+----~----1----~----t
o
100
200
400
300

(d)

Time min.

Fig. 8-Partlcle-slze and permeability Impairment results for

cement with an organic acid Dispersant 1.

Fig. 9-Partlcle-slze and permeability Impairment results for

cement with an organic acid Dispersant 2.

treatment are not substantially reduced. As for the lignin filtrate,

the permeability damage occurs mainly in the first section during
filtration, but it may be eliminated during brine injection and reverse
injection. Finally, the change in k 1 turns out to be positive,
whereas k2 and k3 are slightly damaged by the filtrate. All the permeability changes for the filtrate are positive.
It is very clear that dispersant and retarder additives hinder
precipitation and significantly reduce the permeability damage
caused by this mechanism.
Results for two additives used together (an organic acid and a
dispersant) with the cement are shown in Fig. 10. The particle concentration is substantially smaller than when the cellulose-based
filtrate-control additive is used alone, but higher than when the lignin
retarder additive is used alone. The average particle diameter increases with time. The permeability damage also lies between those
for the two additives.
Fig. 11 shows a representative SEM micrograph for the precipitates taken from various cement filtrates. The crystal morphology
and other chemical tests suggest that the precipitate is predominantly
calcium carbonate. Theoretical solubility calculations to predict
which precipitate will form also indicate the formation of CaC0 3 .
At the high pH values commonly encountered in cement filtrates,
CaC0 3 is extremely insoluble. It can potentially be redissolved as
the pH is decreased. Evidence of this is seen in several flow tests

with brine postflushes. But because CaC03 has a very small solubility even at neutral pH values, it will require a very large number of PV's before it can be completely redissolved.
The effect of temperature can be estimated by noting that the solubility of CaC0 3 decreases from 0.85 giL at 77F (the temperature at which our experiments were conducted) to 0.4 giL at
122P.5 Therefore, precipitate formation may be a more severe
problem at higher temperatures.
The extent and location of the permeability impairment depend
on the quantity of precipitate and on the ratio of the precipitation
rate to the convection rate. The most likely mechanism of precipitate formation is heterogeneous nucleation on colloidal mineral
nuclei that have a low energy barrier to surface diffusion and strongly adsorbing active sites. Unfortunately, the process of heterogeneous nucleation and precipitation is very complex and only
qualitative statements can be made about the kinetics of precipitation. The precipitation rate is in general proportional to the number of nucleation sites and to the square of the degree of
supersaturation (for calcium carbonate, calcium hydroxide, and calcium sulfate) and is strongly affected by the rate of fluid convection and the presence of small quantities of adsorbing impurities. 6
At this time, it is not feasible to make any predictions about the
precipitation rates for the different cement filtrates considered.

404

SPE Production Engineering, November 1991

j;:

10 6

10 4

.E

10 3

Ii

z"

Authors

10 5

EI

10 2

NO STIRRING
10MIN.
20 MIN
30 MIN .
60 MIN .
120 MIN .
180 MIN

10 1
100
2 . 02

11

5.1

25

20

10

Particle Size

(0)

~m

5.0

r--'

4.5

4.0

::1..

3 .5

Conclusions

3.0

c.

2.5

.:

2.0

:~

1.5

1.0

200

100

(b)

Time min.

14

;{

0. 12

,/

E 10

C
e

o
o

8
6

--6

.",

/'"

- - ----

r-

100

(e)

.'"

20 0

Time min.

20~--'--r------r--

t:

~ 20 +-----t'l:u:~_.::_-t__

:;;

.40

Mukul M. Sherma Is an associate

professor In petroleum engineering at
the U. of Texas In Austin. His current research Interests Include formation
damage and petrophysiCS. Sharma cur
rently Is a member of the Well Logging
Technical Committee and was 1987 Stu
dent Chapter Faculty Sponsor. He
earned a B. Tech. degree In chemical engineering from the Indian Inst. of Tech
nology, Kanpur, and MS and PhD
Sharma
degrees In petroleum engineering from
the U. of Southern California. Biograph
Ical sketch and photograph for X.M. Yeng are unavailable.

+-__+-+..:::o..___

--"~

60+---~-+---~~~~d+---+-+~-~---~

4--+______~--~4J~~_+~~~~~--~

G 1___
Q. 80 +

10 0

20 0

(d)

400
300
Time min .

5 00

600

Fig. 10-Particleslze and permeability Impairment results for

cement with two additives, a cellulosebasad fluidloss-control
additive and lignin Retarder 1.

Results have been reported for two groups of experiments: particlesize-distribution measurements in different cement fIltrates as a function of stirring time and permeability-damage measurements on
Berea sandstone cores penetrated by these filtrates. From the results presented here, it is evident that the ratio of the precipitation
rate to the fluid convection rate is an important factor that controls
the extent of permeability damage. Some correlation can be seen
between the precipitation rates measured in the particle-count experiments and the permeability-reduction experiments. For example, fIltrates with fast precipitation rates tend to damage the upstream
end ofthe core, whereas filtrates with slow precipitation rates tend
either not to plug the core at all or to reduce the permeability of
the downstream cores first. Additives also influence the quantity
of the precipitate and, therefore, the permeability reduction. For
example, the addition of a lignin derivative or a synthetic polymer
reduces the quantity of precipitate and causes less damage in the
core. Addition of a cellulose derivative, however, increases the
quantity of precipitate by an order of magnitude . It is probably erroneous to make conclusions about formation damage caused by
cement filtrates when short cores (1 in. long) are used. Indeed, further research is required into experimentally and theoretically accounting for the characteristic time for precipitation and the
residence time in the core.

Acknowledgment
We express our appreciation to Jiao Di at the U. of Texas for help
with the experiments and with the preparation of the manuscript.

Reference.
I. Cunningham, W.C. and Smith, D.K.: " Effect of Salt Cement Filtrate
on Subsurface Formations," JPT (March 1968) 259-64.
2. Krueger, R. F.: "An Overview of Formation Damage and Well Productivity in Oilfield Operations," JPT (Feb. 1986) 132-52.
3. Smith, C.S., Ross, W.M. , and Hendrickson, A.R.: "Hydrofluoric Acid
Stimulation-Development for Field Applications," paper SPE 1284
presented at the 1965 SPE Annual Meeting, Denver, Oct. 3-6.
4 . Adams , N.J. and Charrier, T. : Drilling Engineering (1985) .
5. Ellis, A.J.: "The Solubility of Calcite in Sodium Chloride Solutions at
High Temperatures," Am. J. Sci. (1963) 261, 259.
6. Walton, A.G.: The Formation and Properties of Precipitates (1967).

51 Metric Conversion Factors

OF
(OF - 32)/1.8
E-03
gal x 3.785412
in., X 2.54*
E+OO
psi X 6.894 757
E+OO
Conversion laelor

Fig. 11-SEM micrograph of precipitate.

SPE Production Engineering, November 1991

'Sexact.

C
m3
cm
kg

SPEPE

Original SPE manuscript received lor review AprM 26, 1989. Paper (SPE 19305) accepted
lor publication March 13, 1991 . Revised manuscript received Feb. 15, 1991.

405