Sie sind auf Seite 1von 10

SPE

Society of Petroleum Engineers

SPE 18978
Design and Evaluation of Acid Fracturing Treatments
by Kamel Ben-Naceur* and Michael J. Economides*
Dowell Schlumberger
*SPE Member
Copyright 1989, Society of Petroleum Engineers, Inc.
This paper was prepared for presentation at the SPE Joint Rocky Mountain Regional/Low Permeability Reservoirs Symposium and Exhibition held in Denver, Colorado, March 6-8, 1989.
This paper was selected for presentation by an SPE Program Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper,
as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material, as presented, does not necessarily reflect
any position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Society
of Petroleum Engineers. Permission to copy is restricted to an abstract of not more than 300 words. Illustrations may not be copied. The abstract should contain conspicuous acknowledgment
of where and by whom the paper is presented. Write Publications Manager, SPE; P.O. Box 833836, Richardson, TX 75083-3836. Telex, 730989 SPEDAL.

ABSTRACT

INTRODUCTION

During acid fracturing, different fluids, including a reactive one, are injected. As a result certain complex phenomena are evident. The determination of the performance of the
'treating fluids from injectivity tests departs from techniques
that are applicable in propped fracturing because of the drastic change in rheological and leak-off properties between the
viscous pad and the acid. Also, the simulation of the treatments includes the mechanism of diffusion of the reactive fluid
to the fracture walls, and the injection of fluids with a large
contrast in viscosities. Mechanical instabilities such as wormhole growth and viscous fingering have been shown to develop
during the injection.

While much has been written on the design of propped


hydraulic fractures,l- 3 design optimization, 4 - 5 their diagnosis during and after treatment, 6 - 7 and their long-term production behavior, 8 acid fracturing has not received similar
attention. This is essentially due to the complexity of the
mechanisms associated with the reaction of acid with a generally heterogeneous rock and the performance of the created
fracture. 9
Three significant factors must be considered for the design
of propped fractures:
Maintain sufficient hydraulic width for proppant transport;

A design model coupling the fracture geometry to the diffusion mechanisms, such as heat and mass transfer, is presented. The model allows the simulation of both multi-stage
injection treatments, with alternating stages of pad and acid,
and gelled acid treatments.

Control fluid loss in order to minimize pad volume; and,


Prevent or control undesirable height growth.
The first two relate to the generation of sufficient propped
length, while the latter may be associated with risks of water/gas coning.

A comparison between the two types of stimulation is performed. The implications on fracture conductivity after closure are analyzed based on different models of closure on
etched channels in elastic and elastoplastic media. Guidelines for designing treatments in those formations are given,
as well as a comparison with proppant injection.

By analogy, in acid fracturing, similar considerations are


necessary:
Maintain a sufficient width for effective acid penetration;
Controlleakoff and avoid excessive spending in the nearwellbore vicinity; and,

An important. aspect of the design is the estimation of fluid


lea.koff. Current pressure decline analysis techniques typically
assume that a single fluid is injected. A method is offered to
interpret injectivit.y tests with fluids having large contrasts
in rheological and leakoff properties. Predicted pressures are
finally compared with observed treatment pressures in the
case of large viscosity contrasts.

Prevent undesirable height growth.

However, for acid fractures, while the above considerations


are necessary, they are not sufficient. There are other factors affecting the results from an acid fracturing treatment
509

SPE 18978

DESIGN AND EVALUATION OF ACID FRACTURING TREATMENTS

that are much more difficult to control. These include acid


retardation which is generally desired, viscous fingering and
wormhole growth. BenN aceur and Economides 9 - 10 outlined
the mechanisms of acid fracture propagation and subsequent
production performance. Coupling fracture propagation and
acid reaction, they introduced the notion of acid efficiency,
describing wormhole creation, ,., defined by:

w
e

a w u

dC

re

--dx -

KqLC

was reaction rate-limited (for dolomites at low temperatures).


However, there have been no guidelines on how to design a
treatment. The case of a conventional multistage treatment
is considered here.
Although semi-analytical formulae have been developed to
calculate the geometry of a fracture when two fluids with different rheological and leakoff properties/ 5 they do not provide
quick estimates for treatment dimensioning. The two variables that a treatment has to be designed for are the etched
length x f and etched width We. If U 00 , qL, w and x fa are respectively the average fracture velocity, the leakoff velocity,
the hydraulic width and the live acid penetration distance,
then the acid penetration type curves developed by Nierode
and Williams 13 for diffusion-limited react'ion kinetics allow
an estimate of the maximum live acid penetration distance
through a relation between the leakoff Peclet number Npe
( =wqL/2Def f) and the dimensionless acid penetration :ran
(=2xfaqL/(1t.rvw)), through the following approximation:

(1)

where We is the etched width along the fracture, C the acid


concentration, U 00 and qL are the axial and fluid loss velocity,
the dot refers to the time derivative, and the bar represents
the average value across the channel width. For conventional
acids (which tend to create wormholes), K is close to unity,
while for high efficiency acids, 11 it approaches zero. BenN aceur and Economides 10 also combined the factors affecting production performance after fracturing, including in-situ
stresses, rock embedment strength, created etched width and
varying fracture conductivity. Finally, they contrasted the
effectiveness of acidized and propped fractures based on fracture well performance. Below, certain additional considerations are presented to include the effects of viscous fingering,
acid retardation, and acid leakoff. While results of simulation are presented, more importantly, practical design rules
for predicting the penetration of acid, and the fluid staging
are introduced.

For Npe

:::; 0.5:

(2)
while for 0.9
Npe

C:::

2':: Npe

2':: 0.5:

0.5+(xan-0.5)-1.4(xan-0.5) 2 +6.4(xan-0.5) 3
(3)

Considering for instance the first case leads to the simple relation (valid for low leakoff only; for higher leakoff, Equation
3 should be used), with which the acid penetration distance
x fa can be estimated:

MODELS USED IN THIS STUDY


Two comprehensive simulators were used in this study. In
the first, the acid fracture propagation model allows a shnulation of the mechanisms occurring during the propagation
of the fracture. The latter is discretized into cells, at which
the different variables are calculated. The fracture geometry
component is a 2D /P3DH model, solving the coupled mass
conservation, elasticity, fluid flow, leakoff and energy equations. From the knowledge of the velocity, width, height and
temperature in each cell, a finite difference technique, solving the diffusion-convection equations, allows a calculation of
the reactive fluid concentration profiles. In turn, the etched
width and the heat of reaction are transferred back to the
geometry module and alter the pressure and temperature in

w2

(4)

Xfa = 1l 00 - 4Deff

For a two-dimensional geometry, the velocity at the well is


related to the injection rate q through:
qi
u --re- 2rwh

(5)

with r being equal to 1 for a KGD geometry and


fracture, hence,
qi
w
Xfa = - - - h sr Deff

~for

a PKN

(6)

Equation 6 provides an insight on the effects of the different injection variables. Among them, two factors are crucial:
the injection rate (generally expressed per unit of fracture
height), and the fluid diffusivity. The first affects the penetration distance also through the hydraulic width. Using a
typical fracture width of 0.25 in. (6.3 mm), an injection rate
of 0.3 bbl/min/ft of fracture height (0.156 m 3 /min/m), and a
diffusivity of 3 x 10- 4 cm 2 js, leads to an acid penetration of
220 ft. Due to the higher acid leakoff and lower acid viscosity, the fracture width should decrease during the injection
of acid, hence the pad volume should be designed for larger
widths than the one used for calculation of the acid penetration distance (generally up to 50 percent larger widths).

the fracture. This is done explicitly (i.e. there are no iterations between the geometry and acid module) to improve
the computational time. The second model is a finite difference three-dimensional reservoir simulator 12 including the
effect of stresses, formation embedment strength, and variable conductivity.

DESIGN OF ACID FRACTURING TREATMENTS


Two main contributions have been published allowing the
calculation of the acid penetration distance in acid fracturing. Under the conditions of diffusion-limited kinetics, which
generally holds for limestones or for dolomites at high temperatures, Nierode and Williams 13 provided concentration profile
curves in a fracture with a constant width. Later, Roberts
and Guin 14 considered also the case where the acid etching

The second factor mentioned above


tribution) affects the diffusivity, hence
tance and etched width distribution.
is also dictated by the portion of the
510

(i.e. temperature disboth penetration disThe volume of pad


fracture to be cooled

SPE 18978

K. BEN-NACEUR AND M.J. ECONOMIDES

down. Although comprehensive models can be used to account for the heat of reaction/ 6 a simpler approach is used

3: for low leakoff rates, the acid penetratio:p. distance is


largely independent of the leakoff rate, while for higher
leakoff rates, the penetration distance becomes inversely
proportional to the square and the cube of the leakoff
velocity. Hence a system that provides leakoff control as
well as the creation of a barrier to wormhole growth is
generally preferred.U

here. Generally, fracturing temperatures profiles follow a bilinear variation. 17 Using a numerical model, it is shown in
our study that the dimensionless distance along the fracture
at which the formation temperature is reached depends essentially on the efficiency and the fracture model type. This
dependency is shown in Fig. 1 which clearly indicates that a
KGD-type geometry leads to cooler fracture because of the
larger widths. Considering as an example an efficiency for
the pad of 0.3, and a PKN geometry leads to a dimensionless temperature penetration of 0.3. Hence the pad should
be designed for a length that is a multiple of the desired add
penetration distance. Given these two constraints for the pad
volume, and a maximum allowable excess pressure during the
pad injection, the pad can be designed for. The acid volume
can be easily determined using a simple calculation of the
average etched width. For low efficiency acids, one needs to
compensate for acid spending in the formation through wormholes, and heterogeneities along the fracture, hence the acid
volume should generally be overdesigned by a factor of 2-3.
If these acid volumes lead to an excessive decrease in fracture
width, the previous volumes should be split into alternating
stages of pad acid.

The second factor controlling the acid penetration distance is the eventual retardation of the reaction at the
fracture surface due to the creation of a film, which is the
case for oil-outside-emulsions. This is generally difficult
to control precisely, and to avoid creating fractures that
are unetched near the well bore, the combination of these
emulsions with tail-in straight acids is strongly recommended. Gelling the acid leads to adverse mechanisms:
- the greater viscosity induces larger fracture widths,
and decreases the creation of eddies near the walls,
hence tends to decrease the acid transfer rate;
- a shear thinning near the walls, and a decrease in
the thickness of the diffusion layer tend to accelerate
the transfer of acid to the surface;
- leakoff of most gelled acids has characteristics similar to the straight acids, with creation of wormholes.

ACID FLUID LOSS

Other retardation mechanisms include:


- Temperature cooldown:

Acid leakoff is perhaps the most important factor in acid


fracturing. As indicated earlier by Eq. 1, the "live" acid penetration distance depends on the leakoff rate. The injection
of different stages complicates the analysis, and a tracking of
the different filtrate fronts invading the formation is required.
Settari 18 presented a general fluid loss model that accounts
for different batches of fluid. A similar approach is used here:
this allows a calculation at each time step and in each cell
along the fracture of an effective viscosity used for the calculation of the Cv leakoff coefficient. Coulter et a.l. 19 showed
that the multiple sequential injection of pad and acid may
lead to a reduction in acid fluid loss. A comparison between
different penetration distances for single and multiple fluid
injections will be examined in the design example.

Figure 1 shows that a less efficient pad fluid leads


to an improved cooldown of the fracture.
- Careful selection of treatment rates:
The penetration distance of the acid decreases with
increasing velocities; this is however countered by a
significant increase in mass transfer to the surface
when the flow regime becomes turbulent.
- Use of energized fluids:
Ford and R.oberts 22 considered the effects of foaming on acidizing kinetics. The main conclusions are
that the acid penetration distance decreases with increasing foam quality, and that energizing the fluid
does not change the character of the reaction.

ACID RETARDATION

- Retardation by reaction products:


The convection of reaction products leads to a reduction in the mass transfer rate, hence design programs should account for such mechanisms when
solving for the concentration profiles.

Most acid fracturing treatments suffer from limitations in


acid penetration. Several chemical systems have been described as possible ways of retarding the acid reaction or slowing the molecular diffusion. 19 - 20 Retardation factors have also
been published for different systems, generally by comparing
the amount of reaction product generated with a given acid
system to a straight HCl system under the same conditions.
These data combine diffusion and reaction rate mechanisms,
and they should be interpreted using the proper diffusionconvection model describing that experimental setup. Considering the physics underlying the retardation, the following
observations can be made:

VISCOUS FINGERING
When multiple stages of pad fluid and acid are injected,
and because of the generally lower acid viscosity, instabilities known as viscous fingers occur. Such fingers influence
both the acid penetration and the pressure drop at the wellbore. The existence of these fingers in conditions similar to
acid fracturing has been recognized as early as 1958, with the

The ability to control acid leakoff is perhaps the most


important factor. This can be easily seen from Eqs 2 and
511

SPE 18978

DESIGN AND EVALUATION OF ACID FRACTURING TREATMENTS


work of Saffman and Taylor, 23 and the displacement stability
analysis described by Chuoke et al.. 24 Subsequently, attempts
to quantify the effects of viscous fingering have been reported

tent. For prop fracturing treatments, slight changes in the


analysis have to be considered to account for the effects of increases in apparent slurry viscosity due to increased proppant
loading. In acid fracturing treatments, changes in viscosity
and leakoff behavior are much more drastic. The model of
acid fracture propagation was used to simulate (some of) the
different modes.

by Davies et al.. 25 Due to the randomness of the fingers, attempts to use classical fluid mechanics approaches to predict
the fluid flow behavior in the fracture have not been successful. Recently, statistical (or stochastic) approaches have
been used by Patterson 26 and Nittmann et al., 27 and the latter show that for a slot geometry the "fractal" dimension of
these fingers decreases and tends toward unity when fingers
grow. This result, although important, is only qualitative,
and it does not allow a prediction of the finger extent.

Two cases are considered here: contained fracture, and


uncontained (radial) growth, with the treatment summarized
in Table 1 . A sequence of pad, acid, pad and acid was
injected. The viscosity ratio between pad and acid is in the
order of 30, hence fingers are expected to develop.

In this study, an approach similar to the one by Kova.F 8


was used. A simple one-dimensional approach for viscous
fingering is employed. Although his approach was originally
used for miscible displacements in heterogeneous media, the
same principle can be used in the present case. The method
(referred to as the K-Factor theory) consists of replacing the
unstable flow (with fingers) by a fractional flow. Hence, the
plug-like displacement with a vertical interface between the
fluids is replaced by a smooth invasion profile. Figure 2 shows
the principle of the technique. To calculate the extent of fingering, the fractional flow rule for the pad and the less viscous
acid is based on a one quarter power of the viscosity ratio,
which is giving results similar to Blackwell's experiments 29
on viscous fingering. A viscosity ratio of 50 between the pad
and the acid would lead to an acid flowing approximately 2. 7
times faster than the gel.

Net Pressure Evolution for a PKN Fracture


The pressures considered here are at bottomhole. Several
regimes can be observed from Fig. 4. Regime (a) is the classical mode I with a slope in the range (1/8, 1/4) for the pad
fluid. 7 During period (b), there is channeling of the acid in the
fracture, with a decrease in fracturing pressure, until most of
the pad fluid has been depleted. Period (c) corresponds again
to a mode I propagation with acid, with a smaller slope since
acid leakoff is much higher than the pad. Period (d) corresponds to a stable displacement of the acid by the second
stage of the pad fluid until most of the acid is depleted. Mode
I then resumes during phase (e). Period (f) is similar to (b),
followed by (g) which is masked here by the logarithmic scale.
Finally (h) is the fracture closure.
Net Pressure Evolution for an Uncontainecl Fracture

Viscous fingering affects the treatment results in two ways.


First, the pressure measured at the wellbore will be generally
different if fingering occurs. Figure 3 shows the excess pressure behavior corresponding to the 10 minutes of injection
of 2 fluids at 0.3 bbl/min/ft of fracture height, assuming a
PKN geometry, with viscosities of 50 and 2 cp respectively (a
linear scale is used to accentuate the differences). The solid
line represents the pressure at the wellbore corresponding to
the assumption of a plug displacement, while the dashed line
represents the viscous fingering case. The first portion of
the curves corresponds to the injection of the higher viscosity
fluid (increasing pressures), while the arrival of the second
fluid leads in both cases to a decrease in pressures. The viscous fingering case leads to a less abrupt change, and corresponds more to observed treatment result.s. 30

When the fracture is created from a large pay zone, or until


the fracture reaches stress barriers, the net pressure evolution
follows the trends shown in Fig. 5. The main differences with
the previous case are that period (a) has a negative slope, and
that it is difficult. to distinguish between phases (d) and (e).
Period (b) has a larger (negative) slope than (a), because of
the large viscosity contrast between the fluids.

The efficiency during the injection is shown in Fig. 6, with


the corresponding phases, and a comparison with the injection of the pad only. In the latter case, the efficiency tends
asymptotically to a constant value (in the range [0.25-0.3]).
The efficiency drops significantly as soon as the acid reaches
the perforations. Period (d) corresponds to the replacement
of the acid by the more efficient pad. It should be noted
here that the decrease in efficiency when the second acid is
injected is less pronounced because of the previous plugging
of the wormholes by the second stage of pad fluid.

The second effect of viscous fingering is to increase the


velocity of acid (compared to a plug-type displacement). The
effect is beneficial for the penetration distance, and the design
rules given in a subsequent section of this paper should be
changed accordingly.

ANALYSIS OF ACID FRACTURING TREATMENTS


The analysis of fracturing treatments based on the evolution of the net pressure during the injection was pioneered
by Nolte and Smith, 7 based on the assumptions of a constant
viscosity. Four modes of .fracture propagation were identified,
corresponding to different situations for lateral or vertical ex-

PRODUCTION FROM ACID FRACTURES


Production from acid fractures differ significantly from
propped fractures. Although created open fractures have
an almost infinite conductivity, taking into account in-situ
stresses and formation strength affects significantly the poststimulation performance. 10 31 -

512

32

In addition to those effects,

SPE 18978

K. BEN-NACEUR AND M.J. ECONOMIDES

the variable conductivity along the fracture may also imply a modification of the techniques currently 11secl for poststimulation interpretation. 33 If the choice has to be made between a proppant and an acid fracturing treatment, it is recommended to perform a Net Present Value (NPV) analysis
taking into account the three components:

Pad followed by acid


200 bbl of guar-basecl pad fluid are injected ahead of the
200 bbl of 20X acid.

N PV = Net Production Inc1ea.se - Treatment Cost - Risk Function

The last component is the most difficult to quantify, and it.


depends on both quality of data and accuracy of design modeling. Two risks factors that are common to both types of
fracturing are undesirable out-of-zone growth, and creation of
"choked" regions within the fracture that would prevent communication within the fracture. Proppant fracturing has an
additional risk associated with proppant screen-out or bridging. This is particularly the case for carbonates whose elastic modulus is in the order of 7-12 10 6 psi, implying narrow
fractures. The only term that. is compared here is the net
production increase. Typecurves comparing acid and prop
stimulation can be used for the selection of given treatment.
Figures 7 and 8 show for two values of the formation embedment strength ( 30,000 and 60,000 psi), and for an average
etched width of 0.1 in, the different cumulative production
typecurves. The length required for an acid fracturing treatment to outperform a proppant stimulation can be simply obtaining by drawing a unit slope line originating from a given
production increase, and intersecting the curve corresponding
to the actual effective stress.

The same volume of acid as in the previous case is injected, but it is spread over two acid stages alternating
with pad stages.
Use of a system preventing wormhole growth
The system decribed by Crowe et al.U was used, with
a less viscous non-reactive fluid as a pad (the lower viscosity is prefered to get a longer fracture, hence more
cool down). The acid used is gelled and does prevent the
growth of wormholes.
Staging of oil-acid emulsion with acid
Finally, an emulsion with an external phase of oil was
used ahead of a straight acid.
Results
Limestone Formation

EXAMPLE DESIGN

The results of the different treatments are compared in


Fig. 9 for the limestone formation, and the etched width profiles are plotted versus the fracture length. The etched width
(hence the conductivity) decreases along the fracture. The
shortest penetration distance was obtained with the straight
acid, for which most of the acid is spent near the wellbore.
The use of the gelled acid provided a 20 percent increase in
acid penetration distance. Improved results were obtained
with the staging clue to the better coolclown of the fracture.
Finally, the best results in this case corresponded to the two
last injection scenarios. It should be noted here that the
tail-end of the fracture etched width is essentially due to the
effects of the emulsion whose retardation properties are more
difficult to control. The most homogeneous results (still providing an almost infinite conductivity channel) corresponded
to the injection of the high efficiency acid.

The designer of an acid fracturing treatment is generally


faced with the problem of how to select the proper fluids and
how to stage them. Williams et al. 20 detailed the effect of
acid and pad volumes on the final fracture conductivity. In
this study, a comparison between different. types of injection
is offered, as well as an interpretation of the etched width
profiles. Two cases are examined. The first formation considered here is a limestone at a high static temperature (200
F), while the second is a dolomite at 150 F. The reservoir
and rock mechanics properties are detailed in Table 2. For
the sake of clarity, only the following treatments will be considered (see also Table 2; all treatments assume an injection
rate of 20 bbl/min, an acid strength of 20 percent, and a temperature at the perfs of 100 F in the first case and 80 F in
the second).

Dolomite Case
Treahnents
Figure 10 shows the etched width profiles for the different
treatments. A characteristic in this case is the "bell" shape
distribution, with a maximum conductivity away from the
wellbore. This feature is due to the lower temperatures near
the wellbore leading for a dolomite to a reaction-limited behavior. BenN aceur and Economides 10 gave formulae allowing
the estimation of post-stimulation production for these conductivity distributions based on the use of a harmonic mean.
The greatest penetration distance was obtained again with
the combination of emulsion and acid, but, because of the
excessive retardation provided by the emulsion, the etched

200 bbl of 20X acid is injected preceded by a minimum


volume of displacement fluid.

Gelled acid
200 bbl of gelled acid are injected. The particular gelled
acid did not provide retardation and did not prevent
wormhole growth.

513

DESIGN AND EVALUATION OF ACID FRACTURING TREATMENTS


width 200 ft away from the wellbore would not contribute
effectively to the production. In this case, the best results
were obtained with the high efficiency acid, followed by the
pad-acid injection.

2. Cleary, M.P.: "Comprehensive Design Formulae in Hydraulic Fracturing", paper SPE 9259, 1980.
3. Nolte, K.G.: "Determination of Proppant and Fluid
Schedules from Fracturing Pressure Analysis,", SPEPE,
1986, 255-265.
4. Veatch, R.W.: "Economics of Fracturing: Some Methods, Examples, and Case Studies," paper SPE 15509,
1986.
5. Meng, H.Z. and Brown, ICE.: "Coupling of Production, Forecasting, Fracture Geometry Requirements and
Treatment Scheduling in the Optimum Hydraulic Fracture Design," paper SPE 16435, 1987.
6. Nolte, K.G.: "Determination of Fracture Parameters
from Fracturing Pressure Decline Analysis,", paper SPE
8341, 1979.
7. Nolte, I<.G. and Smith, M.B.: "Interpretation of Fracturing Pressures,", JPT, 1981, 1767-75.
8. Much, M. and Penny, G.: "Long- Term Performance of
Proppant Under Simulated Reservoir Conditions," paper
SPE 16415 , 1987.
9. Ben-Naceur, K. and Economides, M.J.: "The Effectiveness of Acid Fractures and their Production Behavior,",
paper SPE 18536, 1988.
10. Ben-Naceur, K. and Economides, M.J.: "Acid Fracture
Propagation and Production,", Ch. 18 in Reservoir
Stimulation, M.J. Economides and K.G. Nolte (Eels.)
2nd Edition, Prentice Hall, NY, 1989.
11. Crowe, C., Hutchinson, B. and Trit.tipo, B.: "Fluid Loss
Control: The Key to Successful Acid Fracturing," paper
SPE 16883, 1987.
12. Settari, A.: "Fra.cwell: a Three-Dimensional Reservoir
Simulator,", Unpublished Documentation, Simtech, Ca.gary, 1987.

CONCLUSIONS
A numerical model coupling fracture propagation and reaction mechanisms has been used to show the importance of
the different mechanisms in acid fracturing. Viscous fingering can be represented using a one-dimensional approximation based on fractional flow, while wormhole development
was simulated using the notion of acid efficiency.
Design rules have been presented for a calculation of the
required pad and acid volumes. In addition, several types
of add fracturing treatments were compared for a limestone
and dolomite. Best results were obtained with the use of a
high efficiency acid, combination of emulsions and acid, and
alternated stages of pad and acids. Finally, a comparison of
performance from acid and proppant fracturing treatments
was offered, allowing the selection of the most appropriate
treatment.

ACKNOWLEDGEMENTS
The authors wish to thank Dr E. Touboul for fruitful discussions on viscous instabilities, and the management of Dowell Schlumberger for permission to publish this paper.

NOMENCLATURE

acid concentration
average acid concentration
Deff= effective diffusivity constant
Fcv= dimensionless fracture conductivity
fracture height
h
Npe = leakoff Peclet number
injection rate
qi
leakoff velocity
ql
velocity in the fracture
U:n
fracture length
acidized length
dimensionless acid penetration distance
hydraulic fracture width
etched width
etched width rate

13. Nierode, D. and Willams, B.B.: "Characteristics of Acid


Reactions in Limestone Formations," SPEJ, 1972, 306314

()

r
K

cP

a'

SPE 18978

14. Roberts, L.D. and Guin, J.A.: "A New Method for Predicting Acid Penetration Distance," SPEJ, 197.5, 277286.
15. Lo, H.H. and Dean, R.H.: "Modeling of Acid Fracturing", paper SPE 17110, 1988.
16. Lee, M.H. and Roberts, L.D.: "The Effect of Heat of
Reaction on Temperature Distribution and Acid Penetration in a Fracture,", SPEJ, 1980, 501-507.
17. Biot, M.A., Medlin, W. and Masse, L.: "Temperature
Analysis in Hydraulic Fracturing,", JPT, 1987, 13891397.
18. Settari, A.: "A New general Model of Fluid Loss in Hydraulic Fracturing", paper SPE 11625, 1983.

fracture geometry vertical shape factor


acid efficiency
formation porosity
effective stress

19. Coulter, A.W., Crowe, C.W., Barrett, N.D. and Miller,


B.D.: "Alternate Stages of Pad Fluid and Acid Provide
Improved Leakoff Control for Fracture Acidizing", paper
SPE 6124, 1976.

References

20. Williams, B.B., Gidley, J.L. and Schechter, R.S.: "Acidizing Fundamentals", SPE Monograph Volume 6, Dallas,
1979.

1. Harrington, L.: "Prediction of the Movement of Fluid


Interface in a Fracture," paper presented at the Southwestern Petroleum Course Association, 1973.
514

SPE 18978

K. BEN-NACEUR AND M.J. ECONOMIDES


TABLE 1

21. Crowe, C., McGowan, G.R. and Baranet, S.E.: "Investigation of Retarded Acids Provides Better Understanding of Their Effectiveness and Potential Benefits," paper
SPE 18222, 1988.

Data for the Analysis of Pressure and Efficiency Modes

22. Ford, W. and Roberts, L.D.: "The Effects of Foam on


Surface Kinetics in Fracture Acidizing,", SPEJ, 1985,
89-97.

Parameter

23. Saffman, P.G. and Taylor, Sir G.: "The Penetration of


a Fluid into a Porous Medium or Hele-Shaw Cell Containing a More Viscous Fluid,", Proc. Roy Soc., A 242,
1958, 312-329.

Net Frac Height


Gross Height
Young's modulus
P errn eabili ty
Porosity
Injection Rate
Pad leakoff coefficient
Pad average viscosity
Acid viscosity
Acid leakoff viscosity
Volume (pad 1)
Volume (acid 1)
Volume (pad 2)
Volume (acid 2)

24. Chuoke, R.L., Van Meurs, P. and VanderPoel, C.: "The


Instability of Slow, Immiscible, Viscous Liquid Displacement in Permeable Media,", Trans. AIJ~r1E, 216, 1959,
188-194.
25. Davies, D., Bosma, M., and de Vries,W.: "Development of Field Design Design Rules for Viscous Fingering in Acid Fracturing Treatments: A Large-Scale Model
Study," paper SPE 15772, 1987.
26. Patterson, L.: "Diffusion-Limited Aggregation and TwoFluid Displacement in Porous Media," Phys. Rev. Letters, 1984, 52.
27. Nittman, J., Daccord, G., and Stanley, H.: "Fractal Viscous Fingering," Nature, 1985.
28. Koval, E.J.: "A Method for Predicting the Performance
of Unstable Miscible Displacement in Heterogeneous Media,", SPEJ, 1963, 145-154.
29. Blackwell, R.J., Rayne, J.R. and Terry, W.M.: "Factors
Influencing the Efficiency of Miscible Displacements,",
Trans. AIME, 216, 1959, 1-9.
30. Olsen, T. and Karr, G.: "Treatment Optimization of
Acid Fracturing in Carbonate Formations,", paper SPE
15165, 1986.
31. Nierode, D. and Kruk, K.F.: "An Evaluation of Acid
Fluid Loss Additives, Retarded Acids, and Acidized Fracture Conductivity," paper SPE 4549, 1973.
32. Novotny, E.J.: "Prediction of Stimulation From Acid
Fracturing Treatments Using Finite Fracture Conductivity," paper SPE 6123, 1976.
33. Economides, M.J.: "Observations and Recommendations
in the Evaluation of Tests of Hydraulically Fractured
Wells", paper SPE 16396, 1987.

515

Value

50ft
50 ft
6.5 E+06 Psi
1 md
10%
20 bbl/min
0.0018 ft/min~
30 cp
1 cp
1 cp
80 bbl
130 bbl
100 bbl
170 bbl

DESIGN AND EVALUATION OF ACID FRACTURING TREATMENTS

SPE 18978

TABLE 2
Data for Comparison of Treatments

Value

Parameter

::I

'ltjQl
.... u

---~

KGDI Model
- - PKN Model

Ole:

Net. Frac. Height


Gross Height
Young's modulus
Permeability
Porosity
Solubility
Injection Rat.e
Pad leakoff coefficient
Pad average viscosity
Acid concentration

C. til

E'li.i 0.6

50 ft.
50 ft.
6.5 E+06 Psi
1 md
10 %
95%
20 bbl/min
0.0018 ft./min~
30 cp
20 %in volume

~5

(J)c:
(J)0
Q):.;:i
-til
c: ....
aQ)
VIC:
C:Q)

~'-,
0.4

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

~ .........

~o._

r"-_

C5
0.2

1--~

0.1

0.2

0.3

~----

r--

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

-- --

0.5
0.6
0.4
Efficiency, '1J

0.7

0.8

0.9

~rea!_~~!- 1 (Sin~!~-~<:_i_d_:_!l~-~=~~l~e_clJ

Volume injected

Fig. 1-Temperature Penetration Type-Curves for Different Efficiencies.

200 bbl

Treatment 2 (Single Ac.id- gelled)


Volume injected

200 bbl

Treatment 3 (Pad-Ac.id stage)


Volume injected (pad)
Volume injected (acid)

200 bbl
200 bbl

Treatment 4 (Multiple Pad-Acid stages)


Volume
Volume
Volume
Volume

injected
injected
injected
injected

(pad)
(acid)
(pad)
( ac.id)

140 bbl
100 bbl
80 bbl
120 bbl

Treat!n_ent 5 (Hi~~ -~fficiency Acid)


Volume injected (pad)
Volume injected (ac.id)

100 bbl
200 bbl

Treatment 6 (EI?ulsion/ Acid)


Volume injected (emulsion)
Volume injected (acid)

120 bbl
120 bbl

Position Along The Fracture


Fig. 2-Conceptual Application of Fractional Flow to Model Viscous Fingering.

516

K. BEN-NACEUR AND M.J. ECONOMIDES

SPE 18978

1000
- - PLon-Like
Fractional Flow
~

800

600
"iii

a.

<1

~I'

~' ..

'\

I'

400

r-.....'-- -

"iii

a.

<1

200

6
t (min)

10

12

Fig. 3-Pressure Evolution for Two-Fluid System - Piston-Like and Fingering


Using Fractional Flow.

20L-----~--~---L~~-L~-L------~--~~~

10

t (min)
Fig. 5-Excess Pressure Plot for Multiple Fluids Injection (Radial Geometry).

--Actual
---Pad Only
0.8~------~--------r--------+--------+-------~

(e)

0.6~------4--------+--------+--------r------~
!=""

>.
u
c

Q)

c:;
:E

0.4~~~--~--------r--------+--------+-------~

(g)
OL-------~------~------~------~------~

100~---~-~~~--~~~~~--------~--~

10

10

15

t (min)
Fig. 6-Efficiency vs Time for Multiple Fluids Injection.

Fig. 4-Excess Pressure Plot for Multiple Fluids Injection (PKN Geometry).

517

20

25

SPE 18978

DESIGN AND EVALUATION OF ACID FRACTURING TREATMENTS

10

0.35

100~------~-------,--------~-------.--------,

- - - Fco Values
- - a ' Values

0.20

1\
\\

0.15

,____...,.,~

0.25

Q)

$:

...................
0.10

0.1

- - - - Acid
- - - Gelled Acid
- - Pad-Acid
- - Multiple Stages
.............. High Efficiency Acid- - - - - Emulsion-Acid

;'~~\

. I

0.30 1\

0.05
0.00
0

~,......
\\\~ <. .

\\:\ -- \.--- ::.::::.....:::-.::- ...


~

50

100

.....

150
Xf (ft)

--

200

250

300

0.001~~~~W-~~~~--~~~a_~~U.Ull--~-U~

0.001

0.01

0.1

10

Fig. 9-Etched Width Profiles for Different Injection Scenarios. Limestone Formation at 200 F.

100

toxf

Fig. 7-Comparison between Acid Fracture and Proppant Fracture Type


Curves-Formation Embedment Strength = 30,000 psi.

--Acid
Gelled Acid
- - Pad-Acid
- Multiple Stages

0 .141-------1,-----i--'r------1:-------~t--- -

0.12 f-----1'rlf---\+-1f------t--- ::::.:::: ~~~~~~~~~~i~y Acid

..

...

O.OOOL____5.L0____1....L0_0--~~15.LO___::......~......:..w..,..,..2~5':'::0:---~300
Xf

(ft}

Fig. 10-Etched Width Profiles for Different Injection Scenarios. Dolomite Formation at 150F.
Fig. a-Comparison between Acid Fracture and Proppant Fracture Type
Curves-Formation Embedment Strength = 60,000 psi.

518