I Society of Petroleum Engineers I
SPE 24822
Fracturing Pressures for Horizontal Wells
K.A. Owens and S.A. Andersen, Maersk Oil & Gas, and M.J. Economides, Mining U. Leoben
SPE Members rI
Copyright 1992, Society of Petroleum Engineers Inc.
This paper was prepared for presentation at the 67th Annual Technical Conference and Exhibition of the Society of Petroleum Engineers held in Washington, DC, October 4-7. 1992.
This paper was selected for presentation by an SPE Program Committee following review of information contained in an abstract submined by the author@).Contents of the paper,
as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author@).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 conspicuousacknowledgment
of where and by whom the paper is presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 750833836 U.S.A. Telex, 730989 SPEDAL.
ABSTRACT mented and cased. Second, since multiple treatments
should be expected, appropriate zonal isolation must
A fully horizontal well, drilled a t an angle from the be done213. Third, the configuration, formed by 'the
fracture plane, will experience a potentially substantial expected fracture direction (normal to the minimum
increase in the fracturing pressure. The magnitude of stress) and the horizontal well trajectory is extremely
this pressure depends on the angle (as the angle in- important both for the fracture execution (and the im-
creases so does the and, very importantly, plied success or failure) and, especially, the posttreat-
on the value of the intermediate of the three far-field ment well performance.
stresses. This issue has been largely put to rest in a series of
Application of these findings to eight horizontal wells In general, in the cases where a cre-
in the DAN field and comparison with actual data from ated hydraulic fracture has low conductivity (i.e., a
several fracture treatments within each well showed an higher-permeability reservoir) the appropriate config-
excellent. match between predicted and observed frac- uration is for a longitudinal fracture. In lower perme-
turing pressures. The results are presented in this pa- ability reservoirs, where the created fracture has high
per. conductivity, it would be appropriate to create a num-
ber of transverse (orthogonal) fractures.
Thus, in the first case the well should be drilled along
INTRODUCTION the maximum horizontal stress whereas in the second
case i t should be drilled along the minimum horizontal
stress (for depths where vertical fractures are created).
The hydraulic fracturing of horizontal wells is a rela-
tively new stimulation exercise not only because this In either of these configurations these directions, coin-
type of well is new but also because of the difficulty as- ciding with principal stress axes, would result in frac-
sociated with envisioning and executing a treatment. turing pressures that can be estimated readily by the
classic Terzaghi relationship8.
In an extensive comparison of the performance of frac-
tured vertical, unfractured and fractured horizontal The ability to perform a n optimized positioning be-
wells, Brown and Economidesl have demonstrated the tween well and fracture trajectories is frequently hin-
obvious necessity of fracturing horizontal wells in the dered by operational problems. Drilling from pads
vast majority of reservoirs that traditionally have been or platforms would necessitate the drilling of wells in
candidates for vertical well fracturing. This generates a largely radial pattern. This would result in wells
a number of new issues. First, horizontal wells that with multiple fractures that are neither longitudinal
are to receive multiple fracture treatments must be ce- nor transverse and with well-to-fracture contacts that
2 OWENS, ANDERSEN AND ECONOMIDES SPE 2 4 8 2 2
are complex. It has been established that the near- flee = +
(ox, ugy- P W )
wellbore stress concentration will invariably result in - 2=;.( - uyy)cos (28)
largely longitudinal fracture initiation with the frac-
ture direction turning normal to the minimum stressg. - 4rx, sin (28)
Figure 1 is a depiction of this configuration with an ar-
bitrary angle, a,formed between the well and fracture
trajectories.
The turning fracture would lead to two likely phenom-
ena: 1) a reduction in the fracture width, with an ob-
vious associated reduction in well performancef6 and
2 ) a marked influence on the fracture initiation and
propagation pressures due to linear elastic theory and
tortuosity effects.
The latter will be the subject of the following study
and subsequent field cases analysis. The principal stresses a t the borehole wall were given
by Daneshy13 and the ones of interest to fracturing are
FRACTURING
PRESSURE
ESTIMATION
Stresses and displacements, acting on an arbitrarily
oriented well in a medium with known in-situ, stresses
have been described by a number of authorslo-13. Yew
and Li14 and McLennan et al.' presented the mathe- The fracturing pressure, h ,can be obtained from the
matical solution of the problem defining an angle, a, +
solution of Eqs. 2 to 8 by setting 03 = -To p where
between the projection of the borehole axis on the hor- p is the pore pressure and Tois the tensile stress.
isontal plane and an angle, P, between the borehole Therefore, fracturing pressures can be calculated for
axis and the vertical. any angle of deviation, a,knowledge of the three prin-
Their solution which is general and applicable to any cipal in-situ stresses, a,, U H , , ~ , and UH,,,,, the reser-
well in a direction other than the principal stresses, has voir pressure, p, the Poisson ratio, v and the tensile
shown that at the borehole wall, there is a nonvanishing stress, To.
shear stress component and the magnitude of this shear Figure 2 is a plot of the fracture initiation pressure vs.
stress affects greatly the fracturing pressure. angle of deviation for a reservoir with a, =11,000 psi
The rotation from the in-situ system to the stress con- and ~ ~ , , ~ , = 7 8 0psi.
0 The intermediate stress, u ~ , , n n x
centration of a horizontal borehole (P=0) is. is graphed for three different values, 7800, 8300 and
8800 psi, respectively. The major impact of the in-
uxx 1 0
termediate stress on the fracturing pressure is evident.
cyy
uzz -
3%- 0
: sinz a
eos' a
-sin a coa a
Thus, in highly stress-anisotropic formations, arbitrar-
ily oriented horilrontal wells will fracture a t greatly di-
(U verging pressures.
0 0 0
Of interest is the effect of the Poisson ratio shown on
At the borehole well ( r = r,) the stress field is given Fig. 3. The stresses a, and U H , , ~ , are as for Fig. 2
whereas UH,,,, is 8300 psi. Four values of the Poisson
by
ratio are used showing an increase of approximately
250 psi in the fracturing pressure for a 0.04 increase in
SPE 24822 FRACTURING PRESSURES FOR HORIZONTAL WELLS 3
the Poisson ratio. It can be concluded that for typical A total of 42 conventional deviated wells were drilled
petroleum reservoir ranges the Poisson ratio does not between 1972 and 1986. All of the wells were stimu-
have a major impact on the fracturing pressure. lated with hydraulic fractures, but recovery was less
than desired. In 1987 horizontal wells were drilled
The fracture initiation pressure (invariably called
in the Dan field to assess their production potential.
breakdown pressure or, a t times, fracturing pressure)
A feasibility study for the Dan field concluded that
is not equal to the calculated fracture propagation
horizontal wells would be economically attractive only
pressure which is model-dependent, and is affected in
with multizonal fracture stimulation in the horizontal
turn by a number of other factors such as the in-
section15. From 1987 to early 1991, nine horizontal
terlayer stress contrast (and therefore fracture height
wells were drilled and stimulated with hydraulic frac-
migration) the viscosity of the fracturing fluid (with
tures. Production history since then has demonstrated
proppant), injection rate and near-wellbore tortuosity.
that these horizontal wells have succeeded in increasing
Other than the last effect, these phenomena affect the
field production rates a t a lower cost than conventional
fracture propagation pressure beyond the near well-
deviated wells.
bore complex geometry in the same manner as for a
hydraulic fracture penetrating a vertical well. The initial horizontal wells received multiple acid frac-
ture stimulation treatments. Propped fracture stimu-
Near-wellbore tortuosity, as defined in this paper, is
lations replaced acid fractures in later horizontal wells.
the phenomenon that causes additional pressure drop
This improved the medium term productivity of these
as the fracturing fluid moves along an induced frac-
wells substantially. It was during the implementation
ture that is not aligned with the far field fracture di-
of the propped fracture stimulations that the impact of
rection. There will be no attempt in this paper to
orientation of the horizontal wellbore trajectory versus
model the fracture propagation pressure from horizon-
the induced fracture plane became apparent.
tal wells. However, for relatively small tensile stress
values and moderate fracture height migration, frac- Difficulties occured during implementation of some of
ture propagation pressure has been observed in a great the propped fracture stimulations performed in the
number of wells to remain largely constant and very Dan horizontal wells. These difficulties were attributed
near the fracture initiation pressure. Therefore, the to the stress conditions prevailing around a horizontal
term fracturing pressure in this paper refers exactly to borehole and the different direction of fracture initia-
the fracture initiation pressure plus the tortuosity phe- tion from the ultimate direction of fracture propaga-
nomena and it approximates the fracture propagation tion. The theoretical and mechanical aspects of these
pressure. unique stress conditions and the associated width re-
duction have been discussed earlier in this paper. Ac-
tual field data collected during many fracturing treat-
ments in the Dan field will be presented below and it
HORIZONTAL WELL will be shown that they support the theoretical work.
FRACTURING IN
THE DAN FIELD Fracture Direction
Operational Background Measurements
Hydraulic fracture stimulation of horizontal wells has Initial Interpret at ion
been applied to several Danish oil fields including the
Dan field, located in the Danish sector of the North Various methods have been employed t o measure the
Sea (Fig. 4). The field was discovered in 1971 and induced fracture azimuth in the Dan field. Early
production started in 1972. The reservoir rocks are stress orientation measurements were based on bore-
Tertiary Danian and Cretaceous Maastrichtian chalks, hole breakout data. The data were supplemented by
characterized by high porosities (avg. 30%) and low Anelastic Strain Recovery (ASR) measurements per-
permeability, commonly around 1 md. formed on cores from the first horizontal wells. A stress
4 OWENS, ANDERSEN AND ECONOMIDES SPE 24822
pattern was interpreted from these early results and ap- model were valid. This was not the case. ASR inter-
plied to a model developed for domal reservoird6. The pretations of these wells showed strong indication of
premise of this stress model is that principal horizontal north/south induced fracture directions.
stress directions are not uniform across a domal reser- Figure 5 contains the current, palaeomagnetic-correc-
voir. On the crest of the dome, the induced fracture
ted, strain recovery direction summary for the Dan
direction is parallel to the long axis of the dome, while
field. This updated version of the induced fracture di-
on the flanks of the reservoir, the fracture direction rection summary tends to support a more conventional
tends to be radial, (i.e., perpendicular to the forma- "uniform" north/south induced fracture direction and
tion depth contours). Measurements taken from Dan the updated ASR data are also in agreement with the
crestal wells showed a north/south fracture direction fracture direction discovered via the linkup of the two
while stress measurements in flank wells approached a
wells (MFB-4A/MFB-14) described above.
radial pattern for fracture direction.
Conformance to a radial fracture direction would al-
low horizontal wells to be drilled from the center of
the dome towards the flanks and have an induced hy- F'racture Direction and
draulic fracture direction that would be parallel to the
well trajectories. This of course would have been con-
Horizontal Wellbore
venient in a situation of offshore drilling. Complica-
tions arising from an induced fracture plane that would
Trajectory
be misaligned with the horizontal well trajectory were
The first massive propped hydraulic fracture stimula-
not expected. This was the premise that was assumed
tions in a Dan field horizontal well were performed in
during the implementation of the initial horizontal well
well MFA-17 in September 1989. The well is located in
stimulation treatments.
the southern flank of the field. Nine individual zones
were stimulated, with a total placement of over 8.5
Actual Field Observations and treatingpounds million of proppant sand (Fig. 6). Bottom hole
pressures were very similar to those observed
Results in vertical Dan well stimulations. The well, as shown in
Fig. 7 is drilled almost exactly in the expected fracture
Continued compilation of fracture direction measure- trajectory, resulting in longitudinal hydraulic fractures.
ments in the Dan field did not support the radial frac-
ture direction model. In early 1990, a propped hy- In July of 1990, propped hydraulic fracture treatments
draulic fracture restimulation treatment was performed were attempted in horizontal well MD-7 located in
in one of the conventional deviated wells (MFB-4A) lo- the southeastern flank of the field. Sand placement
cated in the southwestern flank of the field. During the through hydraulic fracturing in this well failed.
pumping of this treatment, an adjacent well (MFB-14) Immediate screenouts occurred when the initial prop-
simultaneously stopped producing. Immediate investi- pant reached the perforations. Significantly higher bot-
gation showed that this adjacent well was plugged with tom hole treating pressures were also observed. The
crosslinked fracturing fluid and sand, providing conclu- well was ultimately stimulated via acid fracturing. The
sive evidence that the hydraulic fracture, initiated from unsuccessful attempts to place sand in MD-7 motivated
well MFB-4A, had intersected well MFB-14. Linkup of the investigation whose results are reported in this pa-
the two wells via a hydraulic fracture gave a fracture per.
direction of N5W. This "actual" fracture direction is
Review of bottom hole treating pressure from all frac-
not in agreement with the radial stress model. A frac-
turing operations in horizontal wells drilled to date
ture direction of N55-60E would have been expected.
in the Dan field produced certain interesting findings.
ASR measurements were also performed on cores from Figure 7 contains the wellbore trajectories of the Dan
two additional horizontal wells. Both of these wells horizontal wells compared to the assumed fracture di-
were drilled on the flanks of the field and should have rection of N5W discovered during the linkup of wells
fracture directions that would be almost parallel to MFB-4A/MFB-14. As the figure illustrates, the angle
their respective wellbore trajectory if the radial stress
SPE 24822 FRACTURING PRESSURES FOR HORIZONTAL WELLS
between the fracture direction and the wellbore tra- NOMENCLATURE
jectories varies from 5 - 89 degrees. The angles be-
tween horizontal well trajectories and fracture direc- p = Pressure, psi, Pa
tion (shown in Fig. 7) and the bottom hole treating
p, = Well pressure, psi, Pa
pressures measured during the stimulation treatments
(Y = Angle between well and fracture trajecto-
performed in these wells are plotted in Fig. 8. Also on
the same figure is the predicted fracturing pressure us- ries
ing the linear elastic theory model presented earlier in Y = Poisson ratio
this paper. Each point on this plot represents several a =Stress, psi, Pa
treatments in each well. Only data from zones with
r = Shear stress, psi, Pa
comparable well depth, reservoir pressure, treatment
conditions and far field principal stress are utilized in UH,moz = Maximum horizontal stress, ~ s iPa ,
this illustration. ax,,i, = Minimum horisontal stress, psi, Pa
The fracturing pressures observed in different horizon- a,,=Vertical stress, psi, Pa
tal wells showed great variation with almost 2000 psi
difference in bottom hole treating pressures at com-
parable pump rates. This difference should not have Acknowledgements
occurred if the fracture direction were parallel to the
wellbore. Figure 8 shows an excellent correlation be- The authors wish to thank the management of Maersk
tween field data and the theoretical prediction. More- Olie Og Gas As, Texaco Denmark, Inc. and Shell Olie
over, it can be concluded readily, that the induced frac- og Gasudvinding Danmank BV (Holland) for permis-
ture direction is likely to be the same throughout the sion to publish this work.
Dan field, i.e., largely north-south. The three far-field
stresses used for this analysis were c~~,,~,=4450 psi,
a~,,,, ~ 5 0 5 0psi and av=5600 psi. REFERENCES
1. Brown J.E, and Economides, M.J.: "An
CONCLUSION Analysis of Hydraulically Fractured Horizontal
Wells," Paper SPE 24322, 1992.
1. In general, knowledge of the induced fracture 2. Andersen, S.A., Hansen, S.A. and Fjeldgaard,
azimuth relative to a horizontal well trajectory K.: "Horizontal Drilling and Completion, Den-
allows the estimation of fracturing pressures. mark," Paper SPE 18349, 1988.
2. A large number of actual field fracturing data is 3. Andersen, S.A., Conlin, J.M., Fjeldgaard, K.
presented that corroborates predictions of frac- and Hansen, S.A.: "Exploiting Reservoirs with
turing pressures calculated for horizontal well- Horizontal Wells: The Maersk Experience," 0%
bores that are arbitrarily oriented in relation to field Review (July 1990).
the induced fracture direction. As the angle be-
tween well trajectory and the induced fracture 4. Economides, M.J., McLennan, J.D., Brown, E.
increases, fracturing pressures also increase. and Roegiers, J.-C.: "Performance and Stimula-
tion of Horizontal Wells," World Oil's Hand-
3. Finally, and very importantly, this work has book of Horizontal Drilling a n d Comple-
shown that horizontal wells can be effectively tion Technology, Gulf Publishing, Houston,
fracture stimulated with extremely large treat- 1991.
ments provided that appropriate measurements
are made (stress and fracture orientations) and 5. Mukherjee, H.and Economides, M.J.: "A Para-
necessary modifications to the completion pro- metric Comparison of Horizontal and Vertical
cedures are implemented. Well Performance," Paper SPE 18303, 1989 and
-
SPEFE (June 1991) 209 216.
6 OWENS, ANDERSEN AND ECONOMIDES SPE 24822
6. Economides, M.J., Deimbacher, F.X., Brand,
C. W. and Heinemann, Z.E.: "Comprehensive
Simulation of Horizontal Well Performance,"
Paper SPE 20717, 1990, and SPEFE (Dec.
1991) 418 - 426.
7. Deimbacher, F.X., Economides, M.J., Heine-
mann, Z.E. and Brown, J.E.: "Comparison of
Methane Production from Coalbeds Using Ver-
tical and Horizontal Fractured Wells," Paper
SPE 21280, 1990.
8. Terzaghi, K. van: "Die Berechnung der Durch-
lassigkeitsziffer des Tones aus dem Verlauf der
Hydrodynamischen Spannung~erscheinungen,~~
Sber. Akad. Wiss., 132, 105, Wien, 1923.
9. McLennan, J.D., Roegiers, J.-C. and Econo-
mides, M.J.: "Extended Reach and Horizontal
Wells," Reservoir Stimulation (second edi-
tion), M.J. Economides and K.G. Nolte (eds.)
Prentice Hall, Englewood Cliffs 1989.
10. Deily, F.H., and Owens, T.C.: "Stress Around
a Wellbore," paper SPE 2557, 1969.
11. Bradley, W.B.: Failure of Inclined Boreholes,"
J. Energy Res. Tech., Buns., AIME (Dec.
1979) 102, 232-239. Figure 1: Conceptual Depiction of Turning Hydraulic
Fracture Penetrating a Horizontal Well of Arbitrary
12. Richardson, R.M.: "Hydraulic Fracture in Ar- Direction.
bitrarily Oriented Boreholes: an Analytic Solu-
tion," Proc., Workshop on Hydraulic Fractur-
ing Stress Measurements, Monterey, California
(Dec. 1981).
13. Daneshy, A.A.: "A Study of Inclined Hydraulic
Fractures," SFEJ (April 1973) 61 - 68.
14. Yew, C.H. and Li, U.: "Fracturing of a Deviated
Well," paper SPE 16930, 1987.
15. Conlin, J.M., Hale, J.L., Sabathiex, J.C.,
Faure, F. and Mas, D.: "Multiple Fracture Hor-
izontal Wells - Performance and Numerical Sim-
ulation," paper SPE 20960, 1990.
16. Withjack, M.O. and Scheiner, C.: "Fault Pat-
-
terns Associated with Domes An Experimental
Angle of deviation [deg]
and Analytical Study," AAPG Bull., V. 66, no.
3, (1982). Figure 2: Fracture Initiation Pressure from a Horizon-
tal Well. Effect of Maximum Horizontal Stress.
SPE 24822 FRACTURING PRESSURES FOR HORIZONTAL WELLS
0 10 20 30 40 50 60 70 80 90
Angle of deviation [deg]
Figure 3: Fracture Initiation Pressure from a Horizon- Figure 4: Dan Field Locations.
tal Well. Effect of the Poisson Ratio.
2 58001 WELL PATH
-
7000 8000 9000
HORIZONTAL DISTANCE FROM PLATFROM (FT)
ZONE: PROPPANT TOTAL (LBS) STIMULATION DESIGN: RADIUS (FT)
Figure 6: Dan Field - Well MFA-17 Stimulation Sum-
mary.
Figure 5: Fracture Direction Summary.
OWENS, ANDERSEN AND ECONOMIDES SPE 24822
Fracture direction N5W
t
67501 0 Actual averaged well data
-
(less 370 psi perforation friction/tortuosity)
Predicted fracturing pressure
MD-7
0
MFA- 13
a 0
MFB- 15
4750- MFA-17
4500
0 10 20 30 40 50 60 70 80 I
Deviation angle (deg)
Figure 8: Fracture Pressure vs Trajectory Angle of
Horizontal Wells in the Dan Field.
Figure 7: Well Trajectory/Fracture Deviation Angles
in the Dan Field.
