Professional Documents
Culture Documents
Abstract
This paper discusses the use of organic acid as a stimulation
fluid in carbonate formations. Organic acids such as acetic or
formic acid have been used for many years as alternatives to
HCl, due to the retarded reaction rate, low corrosivity and reduced tendency to form acid-oil sludge in asphaltene rich crudes. In Venezuela, organic acids were pumped in acid fracturing treatments in deep hot limestone formations. Case histories
are discussed in the paper. Although these treatments were
very successful, optimization of the design was hampered by
the fact that models for organic acid-carbonate spending did
not exist.
In this paper a new model for acid spending is proposed
that can be used for strong acids (HCl) and for weak (organic)
acids and also for acid mixtures. Compared to existing (HCl)
models the only new element is the acid dissociation constant,
describing the differences between strong and weak acids. The
model was verified by comparing simulated reaction rates
with rates actually measured in the laboratory with the rotating
disk. Results of acid fracture simulations, comparing etched
length and width of organic acid with HCl acid are also discussed in this paper.
Introduction
In most carbonate stimulation treatments, above and below
fracturing pressure, hydrochloric acid (HCl) is pumped as the
main stimulation fluid. HCl is cheap, it has a high rock dissolving power and the reaction products are usually soluble.
However, the HCl-carbonate reaction is also fast, especially at
the higher temperatures encountered down hole1,2. The rapid
HCl spending prevents deep penetration of acid into the formation. In matrix treatments this results in only short wormholes or even compact dissolution with only marginal skin
decrease3,4. In acid fracturing treatments it results in poor
etched length, often considerably less than the fracture half-
SPE 82211
actually measured in the laboratory with the rotating disk. Despite the models simplicity, it proved to be accurate enough
for our purposes and it is now build into our Acid Fracturing
software. The model was used to analyze acid spending and
etching in a fracture. Results of these simulations, comparing
the performance of organic acids with HCl are discussed in
this paper. The mathematical details of the model are described in the Appendix.
Organic Acid Fracturing in Venezuela
The Cogollo group is a tough carbonate play in Lake Maracaibo, consisting of three formations -Maraca, Lisure and
Apon- totaling some 1000 ft. thickness at a depth of 15,000 ft.
The interval of best quality is the vuggy limestone of the Maraca formation, some 25-50 ft thick with porosities ranging
from 4-8% and permeabilitys of about 10-20 md. Well stimulation is required to make the wells economical. Acid fracturing with HCl acid was considered but modeling indicated an
etched length of only 30-40 ft. The reason for the poor etched
length is the very high spending rate of HCl with the limestone
at the downhole temperature of 280F. The goal of these
treatments was to create as much etched length as possible. It
was therefore decided to use a retarded (organic) acid system.
The organic acid spends slower on rock, compared to HCl,
and will penetrate deeper into the formation, providing for
more conductive length13. Additional benefits of organic acids
over HCl are its lower corrosivity and easier inhibition and a
reduced tendency to form acid-oil sludge. A mixture of 13/9%
acetic/formic acid was chosen because it has the same total
acid strength as 15% HCl. (Note, however, that the rock dissolving power of 13/9% acetic/formic is considerably less than
that of 15% HCl; see further below).
Stage
1
2
3
4
5
6
7
8
9
10
11
12
13
Fluid
Organic acid
Gelled water
Crosslinked PAD
Gelled Organic Acid
Crosslinked PAD
Gelled Organic Acid
Crosslinked PAD
Gelled Organic Acid
Gelled water
Organic Acid (CFA)
Organic Acid (CFA)
Water w/methanol
Water
Volume
bbl
5
145
200
220
100
220
100
220
200
170
40
50
161
Rate
bpm
2
2-18
18
18
18
18
18
18
18-20
20-4
4
4
4
Well
A
B
C
D
E
F
G
H
I
J
K
Total
Prod. Before
bopd
0
350
0
1000
1525
0
450
0
0
0
294
3619
Prod. After
bopd
1252
1717
1940
3152
4290
260
2005
350
2485
882
2162
20495
(1)
This is an equilibrium reaction, with the equilibrium (dissociation) constant given by:
KA =
[ H + ] [ A ]
[ HA]
(2)
acid type
HCl
Formic
Acetic
diss. constant KA
(mol/l)
>>1
1.8E-4
1.8E-5
strength of 1 mol/l
+
[H ]
pH
mol/l
0
1.0
1.8
0.016
2.3
0.005
(3)
SPE 82211
Acetic
% spending
% spending
80
acid
strength
mol/l
6.5
6
0.5
1
2
4
60
5.5
5
40
4.5
pH
20
0
0
0.5
1
2
4
equiv.
HCl %
3.4
6.8
8.5
14.1
16.5
Table 4. Spending of acid components in acid mixtures, and equivalent HCl concentration.
In an acid mixture the spending of the weakest component
is suppressed even more. For example, when 10% acetic
spends on carbonate rock about 54% of the acetic spends.
However, in the mixture 15/10% HCl/Acetic, only 24% of the
acetic acid will spend. From a rock dissolving point of view,
the 10% acetic acid does not add much in this mixture.
pH
Formic
100
acid mix
10% Acetic
10% Formic
13/9% Ac/Fo
7/11% HCl/Fo
15/10% HCl/Ac
3.5
3
8
pKA = -log(KA)
(4)
(5)
RMT = K MT [ H + ]b [ H + ] s
(6)
(7)
k K MT
[ H + ]b
k + K MT
(8)
(9)
SPE 82211
The rate in eq. (9) is proportional to the bulk acid concentration [H+]b. This is also illustrated in fig. 2, where the dissolution rate measured in a rotating disk test is plotted as a function of bulk acid strength. For HCl the measured rate is indeed
proportional to the bulk acid concentration.
7.0E-06
Rotating Disk
Rock: Calcite
Sample size: 5.1 cm
Rotational speed: 300 RPM
Temperature: 30 deg C
Pressure: 100 bar
order n: 0.49
rate constant k: 2.9E-3
6.0E-06
H+ Rate (mol/cm2/s)
HCl
5.0E-06
4.0E-06
CO2
diffusion
3.0E-06
Formic
2.0E-06
Ca2+
HA
dissociation
HA H+ + A-
boundary
layer
1.0E-06
Acetic
0.0E+00
0.0
0.5
1.0
1.5
2.0
Figure 2. Results of rotating disk tests for HCl, Formic and Acetic acid. Curves are rates predicted with
new model.
This relatively simple model nicely explains the HClcarbonate spending process, but fails when applied to organic
acid-carbonate spending. For example, it was shown in ref.
[12] that acetic-carbonate spending is diffusion limited
(straight line when rate is plotted as function of square root of
RPM). If the HCl-model were applicable, the rate would be
described by eq. (9). However, in an 1 mol/l acetic acid solution, the H+ concentration is only 0.005 mol/l (see table 3).
This would imply a rate that is about 200 times slower than
that of a 1 mol/l HCl solution. The measured rates, however,
are much higher (see fig. 2) and only a factor of about 7 lower
than the HCl-rate. Furthermore, the rate for acetic and formic
acid in fig. 3 is not proportional to the bulk acid concentration.
In fact, for acetic acid the strength increase with a factor of 4,
but the rate increases only with a factor of about 1.3. This effect was also noted by Fredd12. In his tests the acetic acid concentration was increased with a factor of 30, but the increase
in rate was only a factor of 2.
The HCl model fails when applied to organic acids because the acid dissociation constant KA is not included in the
model. Schechter tried to correct this and derived a rate equation for organic acids that includes KA 2. However, in his derivations he assumed the equality [H+]=[A-], which is true in
unspent acid, but certainly not true when the acid is partially
spent. Also, at the reacting surface the equality [H+]=[A-]
does not hold.
In the derivation of a new model to describe organic acidcarbonate spending we make the reasonable assumption that
reaction (4) and rate eq. (5) are independent of the type of
acid. This means that k and n have the same value for strong
and weak acids. The difference in observed dissolution rate
between a strong and a weak acid depends solely on the magnitude of the H+ concentration at the surface. When a weakly
dissociated acid spends on carbonate rock, H+ is consumed at
the surface. This drives the dissociation reaction (1) to the
right , generating more H+. So in contrast to HCl, the H+ ions
(10)
with Cb and Cs the bulk and surface concentration of the undissociated acid [HA]. The diffusion coefficient that goes into
KMT will obviously depend on the type of acid used. As is
shown in the Appendix (eq. A-3), it is actually more convenient to take for concentration C the sum of [H+] and [HA]
because it makes the model applicable to both weak and
strong acids.
The acid dissociation constant KA enters the model through
the relation between Cs and [H+]s (see eq. A-4):
Cs = [ H + ]s +
[ H + ] s C0
[ H + ]s + K A
(11)
Similar to the HCl acid model (eq. 8), we can set n=1 and
solve eqs. (7), (10) and (11) for [H+]s. Back substitution into
eq. (10), and assuming unspent acid (i.e. Cb=C0) results in the
following equation for the rate R:
R k K MT
=
C0
R 1 +
k K A k + K MT
(12)
Compare rate eq. (12) with rate eq. (8) for strong acids. The
additional term (1+R/kKA) in eq. (12) is responsible for the
lower spending rate of weak acids, compared to strong acids.
For large values of KA the additional term is approximately
equal to 1, and eq. (12) reduces to the strong acid eq. (8). Interestingly, when the rate is diffusion limited, i.e. for k>>KMT,
the rate constant k drops out of the RHS of eq. (12) but the
LHS still depends on k. This means that even in the diffusion
limited regime, the spending rate of the acid depends on the
kinetic rate of the surface reaction. Only for extremely high
values of k the term (1+R/kKA) reduces to 1 and the rate is no
longer dependent on k.
SPE 82211
(C C s )2n
Rkin = k [ H + ]ns X n 0
[ H + ]ns
(13)
8
etched width (mm)
7
6
5
15% HCl
4
3
13/9% Acetic/Formic
2
1
0
5
4.5
4
3.5
3
15% HCl
2.5
2
Fracture height: 20 m.
Fracture width: 0.5 cm
Fracture length: 100 m.
Pump rate: 15 bpm
Volume: 100 m3
Acid viscosity: 10 cp
Reaction order n: 0.49
Rate constant k: 2.9E-3
13/9% Acetic/Formic
1.5
1
0.5
0
0
20
40
60
80
distance from wellbore (m)
100
Figure 4. Acid concentration as a function of penetration depth in the fracture. 13/9% acetic/formic is
compared with 15% HCl (equal strength in mol/l).
20
40
60
80
distance from wellbore (m)
100
SPE 82211
5
4.5
etched width (mm)
4
3.5
3
13/9% Acetic/Formic
2.5
2
1.5
9% Formic
1
0.5
0
0
20
40
60
80
distance from wellbore (m)
100
2 Schechter, R.S.: Oil Well Stimulation, Prentice Hall Inc, Englewood Cliffs, New Jersey, (1992)
3 Buijse, M.A.: "Understanding Wormholing Mechanisms Can
Improve Acid Treatments in Carbonate Formations", paper SPE
38166, presented at the SPE European Formation Damage Conference, 2-3 June, 1997, The Hague, The Netherlands.
4 Daccord, G., Lenormand, R., and Touboul, E.: "Carbonate
Acidizing: Toward a Quantative Study of the Wormholing Phenomenon", SPEPE (Feb. 1989) 63-68
5 van Domelen, M.S., Jennings, A.R.: Alternate Acid Blends for
HPHT Applications, paper SPE 30419, presented at the Offshore European Conference, Sep. 5-8, 1995, Aberdeen
6 Rietjens
7 Lund, K., Fogler, H.S., McCune, C.C., and Ault, J.W.: "Acidization II - The Dissolution of Dolomite in Hydrochloric Acid",
Chem. Eng. Sci., (1975) 30 825-835.
8 Alkattan, M., Oelkers, E.H., Dandurand, JL., Schott, J.: An
experimental study of calcite and limestone dissolution rates as
a function of pH from 1 to 3 and temperature from 25 to
80C, Chemical Geology, (1998), 151, 199-214
9 Settari, A.: Modelling of Acid Fracturing Treatment, paper
SPE 21870, presented at the Rocky Mountain Regional meeting,
Denver, Colorado, April 15-17, 1991
10 Fredd, C.N., Miller, M.J.: Validation of Carbonate Stimulation
Matrix Models, paper SPE 58713, presented at the 2000 SPE
International Symposium on Formation Damage Control, Lafayette, Louisiana, 23-24 February 2000
11 Chatelain, J.C., Silberberg, I.H., Schechter, R.S.: Thermodynamic Limitations in Organic Acid-Carbonate Systems, Soc.
Pet. Eng. J., August 1976, 189-195
12 Fredd, C.N., Fogler, H.S.: The kinetics of calcite dissolution in
acetic acid solutions, Chem. Eng. Science, 53, No 22, (1998),
3863-3874
13 Al-Qahtani, M.Y., Rahim, Z: Optimization of Acid Fracturing
Program in the Khuff Gas Condensate Reservoir of South Ghawar Field Saudi Arabia by Managing Uncertainties Using Stateof-the-Art Technoloy, paper SPE 71688, presented at the Annual Technical Conference and Exhibition, Sep. 30- Oct. 3,
2001, New Orleans, Louisiana
Appendix
Acid strength and dissociation
The dissociation reaction of acid HA in water is:
HA H+ + A-
(A-1)
[ H + ] [ A ] [ H + ] (C 0 [ HA])
=
[ HA]
[ HA]
(A-2)
(A-3)
SPE 82211
[ H + ] C0
C = [H + ] +
(A-4)
[H ] + K A
Eq. A-4 allows calculation of [H+] from the total acid concentration C. For example, for unspent acid, C=C0, eq. A-4 can be
rewritten as:
[H + ]2 + K A [ H + ] C0 K A = 0
(A-5)
(A-12)
[ H + ]eq = X C 0 C eq
with X =
5 is: [ H ] C 0 , while for a weak acid (KA<<C0) the approximate solution is: [ H + ] C 0 K A .
The relation between total acid strength C and [H+] in eq. A-4
can easily be extended to a mixture of acids:
N
i =1
N
= [H ] +
(A-6)
[ H + ] C 0,i
i =1 [ H
] + K A,i
Acid-Carbonate equilibrium
In this section an expression for [H+]eq is derived for the equilibrium state of the acid-carbonate reaction. For clarity we
consider acid spending on calcite, CaCO3. The equations can
be easily modified to describe acid spending on other carbonates, e.g. dolomite, CaMg(CO3)2.
When acid spends on carbonate rock the equilibrium state
is reached when product of the Ca2+ and CO32- concentration
exceeds the solubility product Ksp:
K sp = [Ca 2+ ]eq [CO32 ]eq
(A-7)
H2CO3 H +
HCO3-
HCO3- H+ + CO32-
K1 =
[CO2 aq]
600
[ H 2 CO3 ]
[ H + ] [ HCO3 ]
K2 =
2.5 E 4
[ H 2CO3 ]
K3 =
[ H + ] [CO32 ]
[ HCO3 ]
5.6 E 11
(A-8)
(A-9)
(A-10)
[Ca 2+ ] eq = 12 C 0 C eq
(A-11)
(A-13)
C = [ H + ] + [ HAi ]
K2 K3
4 K1 K sp
X
1+ X
C 0 K A = 0 (A-14)
For a mixture of acids, [H+]eq can be calculated by (numerically) solving simultaneously eqs. A-13 and A-6. After calculation of [H+]eq it is straightforward to calculate the concentration of undissociated acid [HA]eq and Ceq from the eqs. A-2
and A-3. The fraction of acid that has spent is equal to:
1-Ceq/C0.
Using the numerical values for K1, K2 and K3 in eqs. A-8 to
A-10 and setting Ksp=3.E-8, the constant X is calculated as:
X=2.E-10. Obviously X, being a thermodynamic constant, will
depend on temperature, pressure, etc. We found, however, that
the value X=2.E-10 will produce satisfactory results for a wide
range of conditions.
An alternative equation can be derived from eqs. A-2 and
A-11 to A-13, using the approximation for weak acids,
Ceq [HA]:
KA
4 X
[Ca 2+ ] [CO2 aq ]
[ HA]
(A-15)
Eq. A-15 is written in this form to compare it with the empirically derived equation (3.12) in ref [1]. To produce the constant 1.6E4 in the referenced equation, X must be set to a value
of 2.4E-10, close to the value of 2.E-10 we found above.
Acid-Carbonate reaction rate
The kinetic rate of HCl-carbonate reactions is generally expressed as:
R = k [ H + ] ns ,
(A-16)
SPE 82211
(A-17)
The backward reaction rate constant kb is fixed by the equilibrium condition R=0:
n
2+ n
n
k f [ H + ]eq
, s = k b [Ca ] eq , s [ HCO3 ] eq , s
(A-18)
After some math, using eqs. A-7 to A-13, A-17 and A-18, the
final rate equation is expressed as:
(C C s )2n
R = k f [ H + ] ns X n 0
[ H + ] ns
(A-19)