SPE 103256
Wellbore Cooling as a Means To Permanently Increase Fracture Gradient
I. Gil, Geomechanics Intl.; J.-C. Roegiers, U. of Oklahoma; and D. Moos, Geomechanics Intl.
Copyright 2006, Society of Petroleum Engineers
This paper was prepared for presentation at the 2006 SPE Annual Technical Conference and
Exhibition held in San Antonio, TX, 24-27 September 2006.
This paper was selected for presentation by an SPE Program Committee following review of
information contained in a proposal 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. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers is
prohibited. Permission to reproduce in print is restricted to a proposal of not more than 300
words; illustrations may not be copied. The proposal must contain conspicuous
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
Abstract
In recent years, several techniques have been proposed to
increase the fracture gradient by inducing changes in the near
wellbore region (Alberty and McLean, 2004; Sweatman, et al.,
2004; Benaissa et al., 2005). This process, often called
“wellbore strengthening”, has most recently been implemented
by adding specially designed proppant material to the mud
before raising its pressure above the fracture gradient. The
goal was to induce short tensile fractures in the vicinity of the
wellbore wall which are prevented from propagating; thus,
creating a “stress cage”. However, this method has often
proven ineffective in low permeability formations where
mainly uncontrolled fracture propagation occurs.
The purpose of this paper is to propose and evaluate the use of
wellbore cooling, in combination with more classical
stengthening processes, to permanently increase the fracture
gradient without the risk of circulation losses inherent in the
“stress cage” method, as it is currently applied. This approach
involves lowering the temperature of the drilling mud; thus,
reducing the hoop stress at the borehole wall and then ‘setting’
the stress cage in the standard manner. Tensile cracks can
then be induced at significant lower mud weights. Given the
typical thermal conductivity properties of rocks, the tensile
stresses induced by cooling (and consequently, the created
fractures) will tend to be confined to the near wellbore region.
This work presents an evaluation of the effect cooling has on
the stress profile of a “solid” material and compares it with a
fully coupled thermoporoelastic solution. The results of such
analyses may then be used to design a field application to test
this novel idea.
Introduction
As the demand for hydrocarbon resources grows, drilling is
increasingly taking place in demanding and hazardous
environments. Current oil and gas plays are located in basins
where often the drilling of horizons with low fracture gradient
is involved. Such formations present the drilling engineer
with a scenario where the operational mud weight window to
be used is rather limited or sometimes, non-existant. These
formations tend to be weak; thus, requiring high mud weights
in order to avoid borehole collapse, while simultaneously
exhibiting low fracture gradients. A similar situation is found
in depleted sands; where such formations show lower fracture
gradient, whereas adyacent shales need to be controlled with
high mud density due to their high pore pressure.
In order to overcome this constraint, different methods for
increasing the fracture gradient of underground formations
have been put forward over the years. Very recently, Benaissa
et al. (2005) proposed a method that increases the fracture
gradient by using particles to seal off the pores in the
formation at the wellbore face; thus, creating an effective non-
porous region immediately behind the borehole (i.e. a “sleeve”
region at the wall). The final result being a large increment in
the tensile strength of the rock.
However, one of the most successful techniques for increasing
the fracture gradient of a subsurface horizon is the application
of the “stress cage” concept. In this method, the tangential
stress around the wellbore is increased by inducing and
propping open a controlled fracture at the borehole wall
(Alberty and McLean, 2004). This technique, albeit very
efficient in permeable formations, has however proven rather
ineffective when applied to low permeability rocks. This is
the case of “inverted” stress profiles, where the fracture
gradient in the shale is lower than in adjacent sands.
This paper proposes a new procedure for creating such stress
cage in low-permeability formations (e.g. shaly sands). In this
novel method, changes in temperature are induced in the
formation to be treated, before “setting” the stress cage.
Drilling fluid is used to cool down the formation in order to
reduce the tangential stress at the borehole wall. The
magnitude of this temperature change is determined by the
required increment in fracture resistance, which also
establishes the opening of the fractures in the stress cage.
Subsequently, the stress cage is set up following normal
procedures.
2 SPE 103256
The idea of controlling the stress acting on a rock via altering
its temperature is not a novel concept. As early as the Bronze
Age, people used fire setting as a method for rock extraction
from underground mines (Agricola, 1556). In this technique, a
fire was set up against a rock face, inducing thermal stresses;
then the rock either crumbled naturally or was shattered by
water quenching. In more recent times, several classic papers
such as the one by Morita et al. (1990), and Li (1998) explain
the effects that cooling or heating the wellbore wall has on the
stress concentrations. This last author, rightly so, pointed out
the last increase in pore pressure due to the thermal poroelastic
effects.
Nevertheless, using cooling as a means of increasing the
effectiveness of stress cages is an idea that, to the knowledge
of the authors, has only recently been put forward (Gil and
Roegiers, 2006a and 2006b). The advantages of this new
technology is that it broadens the application of the stress cage
technique, without creating formation damage, as could be one
of the drawbacks of the sealing fluids injection method.
Proposed New Approach
In the physical model proposed by Alberty and Mclean (2004),
a fracture is induced and propped open by solid particles at the
wellbore wall (Fig. 1). Fluid leak-off into the formation being
treated allows the fracture to close and the solid particles to be
locked in place at the mouth of the fracture. The presence of
an open fracture increases the tangential stress around the
wellbore, effectively rising the fracture gradient of the rock.
Figure 1. Physical model of a stress cage as proposed by Alberty and
Mclean (2004).
However, in low permeability formations, low leak-off
conditions would prevent the fracture from closing; thus,
leaving an open hydraulic conduit between the wellbore and
the fracture tip. This creates the conditions for further fracture
propagation if additional energy is provided (i.e. if the well
pressure increases). In addition, negative variations of the
equivalent circulating density, as normally found during a
drilling operation, would provide the conditions for fracture
fluid flowback and probably for dislodging the stress cage (see
Fig. 2).
Figure 2. Possible mechanism for particle dislodging and stress cage
removal in low permeability formations.
A new approach, hereby proposed, would involve lowering
the temperature of the drilling mud; thus, reducing the hoop
stress at the borehole wall before ‘setting’ the stress cage.
Tensile cracks can then be induced at significantly lower mud
weights. Given the typical thermal conductivity properties of
rocks, the tensile stresses induced by cooling (and
consequently, the created fractures) will be confined to the
near wellbore region.
The following sequence describes the new method :
1. The rock in its original condition is subjected to an initial
value of tangential stress at the wellbore wall (i.e., σθθ =
σθθ 1).
2. Once the formation undergoes a certain degree of
cooling, the value of the tangential stress at the borehole
wall decreases (i.e., σθθ = σθθ 1 - ∆σΤ). This translates
into a lower fracture gradient.
3. The stress cage is set in place by creating a small
hydraulic fracture. The entrance of the fracture is kept
open by using proppant material; the final result is an
increase in the fracture gradient (i.e., σθθ = σθθ 1 - ∆σΤ +
∆σcage). At this point, the cooling profile allows the
creation of the fracture at pressures lower than if the
formation remained at its original temperature. By the
same token, the temperature profile limits the fracture
propagation to the cooled region around the wellbore.
4. The rock is allowed to go back to its original temperature
(i.e., σθθ = σθθ 1 + ∆σcage). As the formation regains
temperature, the tangential stress acting on the fracture
faces increases; thus, locking the particles in place along
the fracture length. It is expected that most of the
particles will remain in the region connecting the
wellbore and the fracture (at the crack entrance).
Amongst the main advantages of using this new technique is
the fact that the particles are locked in place by thermal effects
instead of by pressure release/fluid leakoff; therefore, reducing
SPE 103256
the risk of particle flowback. Another benefit of this new
approach is that the formation may be fractured at lower mud
pressures than in the standard technique. At least in theory,
one could cool down the formation, set the stress cage, and
continue drilling without decreasing the applied pressure on
the borehole wall. As time progresses, the temperature of the
treated formation face will increase and the stress cage will be
locked in place.
It is evident that hydraulic communication between the
wellbore and the fracture tip needs to be as restricted as
possible. It is proposed here that the proppant particle size
distribution be made as wide as possible and that deformable
solids be used in order to reduce the permeability of the
“bridge” at the fracture entrance.
Feasibility of the new approach - Modeling
Linear Elastic “Solid” Material
In a wellbore, changes in formation temperature trigger
thermal stresses that may dramatically change the stress field
of the rock wall. Cooling the formation face tends to cause the
rock to “shrink”; thus, reducing the surrounding tangential
stress. This change in tangential stress may be calculated by
the following equation (Morita et al., 1990; Moos and Zoback,
1990; Moran et al., 1990):
∆T β E
∆σ = …..…………..(1)
1 −ν
where ∆σ = change of tangential stress; ∆T = change in
temperature (˚C); β = thermal expansion coefficient (˚C-1); E =
Young’s modulus (MPa), and v = Poisson‘s ratio.
It should be pointed out that this approach is purely elastic; i.e.
the effects on the pore fluid are not considered. Hence, in this
limited approach the problem of calculating the change in
formation temperature as a function of time and distance to the
wellbore wall is defined by the general equation for heat
diffusion (Carslaw and Jaeger, 1959); i.e.
[ ]
Div k f grad (T f ) = C f ρ f
dT f
dt
…....(2)
where Tf, = formation temperature; kf,= thermal conductivity
of the rock; Cf, = formation heat capacity, and ρf = bulk
density of the rock. Taking into account the axisymmetrical
geometry of the problem, and neglecting any vertical heat flow
inside the formation, Eqn (2) may be rewritten as (Morita et
al., 1990):
∂ 2T f 1 ∂T f C f ρ f dT f
2
+ = ………..(3)
∂r r ∂r kf dt
3
Three boundary conditions are necessary to complete the
definition of this problem. At very early times (for tÆ 0), it
may well be assumed that the entire formation remains at its
original temperature, Tf0.; thus, defining the initial condition
as:
lim T f = T f 0 when t → 0 ………..(3a)
For rock located away from the borehole (for r Æ∞), the
temperature may be assumed to be constant at all times; this
condition may be expressed as:
∂T f
lim = 0 when r → ∞ ………..(3b)
∂r
The remaining boundary condition may be obtained by
assuming that, at the borehole wall, the rock and the fluid
temperature are constant and have the same value:
T f = TW r = rw ………..(3c)
where rw = borehole radius; and, TW = fluid temperature inside
the wellbore. This boundary condition differs from the one
used by Gil and Roegiers (2006a and 2006b); it was modified
in order to better describe the problem of borehole cooling and
to allow comparison with the poroelastic solution shown later
in this paper.
The analytical solution of Eqn. (3) for the borehole wall
involves the use of Laplace transforms (Carslaw and Jaeger,
1959). They presented a final equation, which is both tedious
to solve and limited to the borehole wall. Simpler integral
equations have been presented presented for the magnitudes of
σθθ and σrr as a function of radial position r and time t
(Stephens and Voight, 1982). While the exact solution for the
temperature distribution near a constant temperature wellbore
is a series expansion (Ritchie and Sakakura, 1956), solutions
may be used, which approximate the temperature using the
first two terms of the expansion, to give sufficiently accurate
results close to the wellbore (Moos and Zoback, 1990).
Fortunately, currently available technologies, such as the finite
element method, allow for alternative solutions of Eqns. (2)
and (3) without the limitations of the analytical method. This
paper used this type of technologies to analyze the problem
and prove the application of the concept in wellbore
strengthening operations.
Numerical solution
A two-dimensional model representing the wellbore and the
surrounding formation was created to simulate the problem
described in the last section (Fig. 3).
4 SPE 103256

The designed fracture length in most stress cages is about 6

Figure 3. Finite element model of the problem
inches (i.e. approximately, r/rw = 1.7 in this particular model),
which corresponds to the zone where the temperature drop is
most significant. The results of these simulations show that,
for this particular example, five hours of cooling will drop the
rock temperature from an ambient temperature of 225 ˚F to
about 185 ˚F at r /rw = 1.7. Typical values of β, v, and E for a
weak sandstone are 9*10-6 ˚F-1, 0.30 and 1.5*106 psi,
respectively. Thus, a simple calculation by using Eqn. 1
shows that at r /rw = 1.0 (i.e., at the wellbore wall) cooling
down the rock to 150 ˚F decreases the formation tangential
stress by almost 1500 psi (or about 2.8 ppg at the assumed
depth). The change in tangential stress as a function of
temperature distribution (at a given time) may be calculated as
(Morita et al., 1990):

In order to represent the behavior of a wellbore drilled in a

Eβ 2 D
r
typical tight/shaly sandstone; the values of the input Eβ
parameters used for the simulations in this study were chosen σ θT = rD ∫ rD T drD − T ………..(4)
as follows: 1− v 1 1− v

kf = 1.352 (Btu/hr.ft.˚F)
At any given time, curve fitting may be used in Fig. 4 to
ρf = 137.34 (lbm/ft3) obtain the equation of temperature (Tf) as a function of rD (the
Cf = 0.213 (Rock heat capacity, Btu/lb. ˚F)
dimensionless radius, r/rw). Such equation could be replaced
Depth = 10,000 ft
into Eqn. 4 to obtain the temperature-induced change in
Geothermal gradient = 15 ˚F/1000 ft
tangential stress. Stephens and Voight (1982) presented an
Tf0 = 225 (Initial formation temp.,˚F)
alternative approximate solution that involved the use of error
Tw = 150 (Wellbore fluid temp.,˚F)
functions to define the profile of thermally-induced stress.
ro = 3.75 (Outer radius, ft)
rw = 0.75 (Wellbore radius, ft)
Poroelastic Material
In the field, it will be necessary to obtain the largest effect in
the shortest period of time. Thus, the simulations were run to
The linear elastic analysis presented above is useful as a first
define an “optimum” cooling time, after which the formation
approach to modeling the problem. However, given the nature
temperature changes only very slowly. The results are
of low permeability formations (e.g. shaly sands), modeling
presented in Fig. 4. As it can be easily observed that after five
the material as a poroelastic material seems more appropriate.
hours, the temperature profile changes very little; thus, further
cooling beyond this point was deemed fruitless.
The poroelastic analysis presented herein is based on the work
by Li et al. (1998); whose paper presented an analytical
230
solution for the combined pore pressure, temperature and
stress distribution for a wellbore. They used superposition to
220
t = 1 hrs split the problem into three systems: a thermoporoelastic
plane-strain problem; an elastic uniaxial stress problem; and,
210
an elastic anti-plane shear problem. In their paper, the
complete solution of each of the sub-problems, as well as the
Wellbore wall temperature ( F)
o

200
overall solution is presented, for that reason they are not
190
t = 6 hrs
1 hrs cooling
repeated here. The method developed by Li et al.(1998) was
2 hrs cooling
3 hrs cooling
implemented as an additional routine of a program called
180 4 hrs cooling
5 hrs cooling
SFIB© (Stress and Failure of Inclined Wellbores); the results
6 hrs cooling presented here were obtained by using this program with the
170
same boundary conditions described by Eqns. (3a)-(3c).
160

In order to run the simulations, the following parameters were

150
1 1.5 2 2.5 3 3.5 4 4.5 5
used:
Dimensionless radius (r/rw)

Figure 4. Change in formation temperature as a function of distance Depth = 10,000 ft

to the wellbore for several cooling times σv = 10,000 psi
σHmax = 9000 psi
SPE 103256
σHmax AZIMUTH = 145 degrees
σhmin = 7500 psi
Pw = 5500 psi
∆Tw = -75 ˚F
Pp = 4500 psi
α = 0.5
v = 0.30
G = 1.28e6 psi
B = 0.4
vundrained = 0.314
Porosity = 0.14
Permeability = 5e-3 mD
Tdiff = 1.5e-6 m2/sec
Rock’s thermal expansion coeff. = 3e-6 ˚F-1
Mud’s thermal expansion coeff. = 1.6e-4 ˚F-1
rw = 0.75 ft
where σv is the overburden stress, σHmax is the maximum
horizontal stress; σhmin is the minimum horizontal stress; Pw is
the fluid pressure within the wellbore, ∆Tw is the temperature
difference between the wellbore fluid and the formation; Pp is
the formation pore pressure; α is the Biot’s coefficient; v is the
Poisson’s ratio, G is the rock shear modulus; B is Skempton’s
coefficient; and, TDiff is the thermal difussivity of the rock.
The simulations evaluate the change in tangential stress,
resulting from the cooling process, as a function of time and
distance from the wellbore. Figure 5 shows a comparison of
the temperature profiles for both elastic and poroelastic
materials, evaluated after 1 and 5 hours of cooling (for rw < r <
2rw). It was observed that, with the parameters selected for the
simulation, the poroelastic material cooled down at a slower
rate than its elastic counterpart did.
Figure 5. Comparison of the change in formation temperature as a
function of distance to the wellbore for elastic and poroelastic
materials.
5
In this simulation it was assumed that the well was vertical;
hence, the original magnitude of the tangential stress along the
direction of σHmax was given by,
′ = 3 σ h min − σ H max − Pp = 7000 psi
σ θθ
Figure 6 shows the magnitude of the tangential stress around
the wellbore (after 5 hours of cooling) for the poroelastic
material defined in this study. It also contains the profiles of
σ'θθ along the direction of σHmax for cooling times of 1, 60, and
300 min (i.e. quasi-instantaneous, and after 1 and 5 hours of
cooling). It was observed that the magnitude of the tangential
stress was reduced by almost 2000 psi at the wellbore wall,
and by a little more than 1000 psi ar r > 1.5rw.. This gives
validation to the approach proposed in this paper, as it seems
that the cooling effect is large enough to cause an important
decrease in the magnitude of the tangential stress around the
wellbore. This makes setting of the stress cage feasible at
much lower mud weights; once the stress cage is put in place,
and the formation regains it original temperature, the particles
inside the fracture would be locked in place by virtue of the
rock thermal expansion of the rock.
Figure 6. Left, magnitude of the tangential stress around the wellbore
after 5 hrs. of cooling (poroelastic material, α = 0.5). Right,
tangential stress profile, for different cooling times, along the
direction of σHmax.
In Figs. 6 to 8, the effect of changing the value of Biot’s
constant was evaluated for a range between 0.3 and 0.7. It
was found that as the magnitude of Biot’s constant was
increased, the effect of cooling on the tangential stress was
more pronounced. The reason for this dependency, among
other things, is that Biot’s coefficient influences how fluid
expansion affects the total stress. If the coefficient of
expansion for the fluid is larger than the one for the rock, the
thermal stress effect automatically increases as the Biot
coefficient increases.
6 SPE 103256
Figure 7. Left, magnitude of the tangential stress around the wellbore
after 5 hrs. of cooling (poroelastic material, α = 0.3). Right,
tangential stress profile, for different cooling times, along the
direction of σHmax.
Figure 8. Left, magnitude of the tangential stress around the wellbore
after 5 hrs. of cooling (poroelastic material, α = 0.7). Right,
tangential stress profile, for different cooling times, along the
direction of σHmax.
Conclusions
Traditional procedures to strengthen the borehole via the
creation of a “stress cage” have been successfully developed
and used in highly permeable formations. However, their
extension to shaly/tight horizons has been, at best, limited
probably due to the closure and flow back via the induced
fractures.
This paper proposed an alternative to create a similar stress
cage and strengthen the borehole via inducing secondary
thermal stresses generated by cooling the drilling mud. This
new approach needs now to be validated in the field.
The idea of using rock cooling to decrease the tangential stress
on the wellbore wall may also be utilized to magnify the effect
of “standard” stress cages in permeable formations. In this
case, rock cooling applied before setting the normal stress
cage would, in theory, have a compounding effect driven by
the presence of a propped fracture as well as by a thermal
component.
According to the modeling performed during this study, the
effect of rock cooling was large enough to cause important
reduction on the magnitude of the tangential stress at and near
the wellbore. This creates the possibility of setting stress
cages at much lower pressures than in standard approach.
Once the stress cage is in place, and the formation regains its
original temperature, the treatment will be locked in place by
thermal effects (without the dependence of fluid leak-off that
the normal technique involves). It is apparent that hydraulic
communication between the wellbore and the fracture tip
needs to be as constrained as possible. It is proposed here that
the proppant particle size distribution be made as wide as
possible and that deformable solids (e.g. graphite) be used in
order to reduce the permeability of the “bridge” at the fracture
entrance.
References
1. Agricola, Georgius, 1556: De Re Metallica, Latin, Basel:
J. Froben and N. Episopius.
2. Alberty, M., and Mclean, M., 2004: A Physical Model for
Stress Cages, SPE paper 90493, presented at the SPE
Annual Technical Conference and Exhibition, 26-29
September, Houston, Texas
3. Benaissa, S., Bachelot, A., Ricaud, J., Arquey, G., Yi, X.,
and Ong, S., 2005: Preventing Differential Sticking and
Mud Losses Drilling through Highly Depleted Sands.
Fluids and Geomechanics Approach, SPE paper 92296,
presented at the SPE/IADC Drilling Conference, 23-25
February, Amsterdam, Netherlands.
4. Carslaw and Jaeger, 1959: Conduction of Heat in Solids,
Oxford University Press., London.
5. Gil, I., and Roegiers, J.-C., 2006a: “Wellbore
Strengthening in Low Permeability Formations”,
Proceedings of the International Symposium of the ISRM,
Eurock 2006, 9-12 May, Liège, Belgium, pp. 417-420.
6. Gil, I., and Roegiers, J.-C., 2006b: “New Wellbore
Strengthening Method for Low Permeability Formations”,
ARMA/USRMS 06-1092, Proceedings of the 41st U.S.
Rock Mechanics Symposium, GoldenRocks 2006 -
Landmarks and Future Challenges, June 17-21, Colorado
School of Mines, Golden, Colorado, USA.
SPE 103256 7

7. Li, X., 1998: Thermoporomechanical Modeling of

Inclined Boreholes, Ph.D. Dissertation, School of
Petroleum and Geological Engineering, University of
Oklahoma, Norman, Oklahoma.
8. Li, X., Cui, L., and Roegiers, J.-C., 1998:
“Thermoporoelastic Analyses of Inclined Boreholes”,
SPE paper 47296, presented at the SPE/ISRM Rock
Mechanics in Petroleum Engineering, Eurock 1998, 8-10
July, Trondheim, Norway

9. Moos, D., and Zoback, M., 1990: Utilization of

Observations of Well Bore Failure to Constrain the
Orientation and Magnitude of Crustal Stresses:
Application to Continental. Deep Sea Drilling Project, and
Ocean Drilling Program Boreholes, Journal of
Geophysical Research. Vol. 95, No. B6, pp. 9305-9325.
June 10

10. Morin. R. H.. R. L. Newmark, C. A. Barton, and R. N.

Anderson, State of lithospheric stress and borehole
stability at Deep Sea Drilling Project site 504B, eastern
equatorial Pacific, Journal of Geophysical Research. Vol.
95, No. B6, pp. 9326-9335, June 10.

11. Morita, N., Black, A.D Guh, G-F., 1990: Theory of Lost
Circulation Pressure, SPE paper 20409, presented at the
SPE Annual Technical Conference and Exhibition, 23-26
September, New Orleans, Louisiana

12. Ritchie, R. H., and A. Y, Sakakura, 1956: Asymptotic

expansions of solutions of the heat conduction equation in
internally bounded cylindrical geometry, Appl. Phys., 27,
1453-1459, 1956.

13. Stephens. G., and B. Voight, 1982:. “Hydraulic fracturing

theory for conditions of thermal stress”, Int. J. Rock
Mech. Min. Set., 19, 279-284.
14. Sweatman, R., Wang, H., and Xenakis, H., 2004:
“Wellbore Stabilization Increases Fracture Gradients and
Controls Losses/Flows During Drilling”, SPE paper
88701, presented at the Abu Dhabi International
Conference and Exhibition, 10-13 October, Abu Dhabi,
United Arab Emirates.