SPE

Society of Petroleum Engineers

SPE 18957

The Combined Effect of Fluid Lag and Fracture Toughness on

Hydraulic Fracture Propagation
by R.G. Jeffrey, 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 informatio~ contained in an abstract sub~itted by the author(s). Contents of t~e paper,
as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correctro~ by the aut~or(s). The_ma~erral, ~s presen~ed: does no~ necessarrly ref~ect
any position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE ~eetrngs are subjec_t to publrcatron revrew by Edr~orral Co_mmrttees of the Socrety
of Petroleum Engineers. Permission to copy is restricted to an abstract of not more than 300 words. lllustratro~s may not be copred. The abstract should contarn consprcuous acknowledgment
of where and by whom the paper is presented. Write Publications Manager, ·SPE, P.O. Box 833836, Rrchardson, TX 75083-3836. Telex, 730989 SPEDAL.

ABSTRACT has been recognized for many years. 6 •7 This paper contains an anal-
ysis of the size of the fluid-lag region as it depends on field stresses,
Analytical expre~sions for the fluid-lag distance in hydraulic frac- fracturing and reservoir pressures, and rock fracture toughness. The
tures have been found by using the weight function method. The analysis contained in this paper employs a pressure distribution. in
fluid-lag distance is the region near the tip or leading edge of the the fracture that is maximum at the wellbore and decreases to the
fracture that is not penetrated by the fracturing fluid. Closed- reservoir pressure at some distance behind the tip. The near-tip
form expressions for fluid-lag distance have been obtained for both region, assumed to contain fluid at reservoir pressure, is identified
KGD-type and penny-shaped fracture geometries subject to driving as the fluid-lag region.
pressures that were taken to be either uniform or quadratic in distri-
bution. For this analysis, the fluid-lag region is assumed to contain Definition of the Problem
fluid at reservoir pressure.
Figure 1 shows a two-dimensional fracture, pressurized by frac-
The analytical expressions provide a convenient way of studying turing fluid from the wellbore to a distance L - d. A fluid-lag dis-
the variation of the fluid-lag distance as a function of the other tance, d, exists at the two tips of the fracture, and a process zone
parameters of the problem. A general nondimensional equation for consisting of a region of microcracking exists for some small dis-
fluid-lag distance is given that is valid for both of the fracture ge- tance beyond the tip. The energy used in forming the process zone
ometries considered and for all of the pressure loading cases studied. contributes a resistance to fracture propagation that is represented
A method of using the results to account for the effects of fluid lag by the fracture toughness of the rock. The problem, therefore,
in two- and three-dimensional hydraulic fracture simulators is sug- consists of solving for the fluid-lag distance as a function of the
gested. other parameters. The geometry of the fracture, the pressure and
stress loading parameters that act on the fracture, and the fracture
The analytical results have been used to show that the fluid-lag toughness of the rock will all have some effect on the size of the
distance in classical KGD-type fractures quickly approaches a con- fluid-lag region.
stant value while the fluid-lag distance in penny-shaped fractures is
a decreasing function of the radius. Using the principle of superposition, the fracture loading shown
in Figure 1 can be decomposed into two parts, one consisting of
INTRODUCTION the rock without a fracture subject to the far-field stress, and the
other consisting of the rock with a fracture subject to equivalent
Laboratory,! field, 2 and numerical 3 •4 •5 evidence has shown that crack-line loads (see Figure 2). The crack-line loading shown in
hydraulic fractures may propagate although the fracturing fluid does Figure 2c is found by adding the negative of the far-field stress
not reach the tip of the fracture. The phenomenon of the fracturing to the loads acting in the fracture and results in the same stress
fluid lagging behind the fracture tip, or leading edge, is referred to intensity factor and fracture opening as the loading in Figure 2a.
as fluid lag, and the need for such a region, under certain conditions, The stress intensity factor resulting from this loading can be calcu-
lated by using the weight function approach of Bueckner8 and Rice. 9

References and illustrations at end of paper.

269
2 THE COMBINED EFFECT OF FLUID LAG AND FRACTURE TOUGHNESS ON HYDRAULIC SPE 18957
FRACTURE PROPAGATION
where d is the fluid-lag distance measured from the tip of the frac-
The Weight Function Approach
ture. The factor of two before the weight function integrals arises
because the pressure loading must be integrated over both sides
The weight function is an influence function for the stress inten-
(left and right) of the fracture. A similar integral equation can be
sity factor. The weight function, m(x,a), is a unique function of
written for a penny-shaped fracture. If the pressure distributions,
the fracture problem geometry and is independent of the loading of
p( x) and Utip( x ), are of a simple form, the integrals can be solved
the problem, 10 the only restriction being that the loading case used
analytically and an expression for the fluid-lag distance, d, found.
to derive the weight function for the problem must not have more
Numerical methods may be employed for other, more complicated,
symmetry than the loading case that is to be analyzed. Given the
pressure distributions.
weight function for a crack of length a, the stress intensity factor for
pressure loading p( x) is found by integrating the pressure loading
Uniform Pressure Case
weighted by m( x, a) over the crack surface.

KI = 1p(x )m(x, a )dx. (1)

The simplest pressure distribution for p( x) is one that is uniform
and equal to:
(8)
The weight function for a plane strain line crack of length 2£,
aligned along the x-axis, subject to loading that is symmetrical where Po is the pressure in the fracture. The stress Utip( x) has
about the crack center is: 10 been assumed to be uniform and equal to the difference between
the far-field total stress and the reservoir pressure:
£)1/2 1
m(x,L) = ( - --r:::~= (2) (9)
7r .J£2-x2
and the weight function for a penny-shaped crack, of radius R, Because u3 is larger than Pr, Utip acts to close the fracture in
subject to loading that is axisymmetric is: 4 the fluid-lag region and thereby reduces the stress intensity factor
at the fracture tip caused by the excess pressure, p( x ).
1 ) 1/2 r
m(r, R) = ( - (3)
1l'R .JR2 _ r2 Using this uniform pressure loading together with Utip( x) = Utip•
the two integrals in Equation 7 are found to be:
EQUILIBRIUM FRACTURE PROPAGATION

Hydraulic fracturing is assumed to take place under conditions

such that the loading on the fracture is just sufficient to overcome
2
~1 L-d
o
P!
. ~dx
v £2- x2
= 2pf ~-arcsin (L-
7r
--
L
d) (10)

the fracture toughness of the rock. This requirement is expressed and

in terms of the stress intensity factor and fracture toughness as:
2 f;J L. l1tip
~dx = 2utip
v £2 - x 2
#;[71'2
- - -
. (L-d)] .
arcsm - -
L
(4) L-d 7r

(11)
The left hand side of Equation 4 can be broken into two parts that Equation 7 can now be used to find d.
represent the contributions from the region pressurized by fracturing
fluid and from the fluid-lag region. d = L[1- sin(a)]. (12)
fluid preuure _ Klag _ Krock
K I I - Ic (5) Because the physical range of d is from zero to L, the range of a
is restricted to 0 ::; a ::; ~ and
Furthermore, because the contributed by the fluid-lag region to the
stress intensity factor is negative, it will be treated as an apparent a= K]~ck.Ji + l1tip1l'VJ;
fracture toughness term. (13)
2VJ;(p, + l1tip)
K fluid pressure
I
= Krock
Ic
+ Klag
Ic (6)
From this solution, the following two limiting cases are found. For
This equation provides the basis for the analysis that follows.
K]~ck = 0:
7l' l1tip
Fluid Lag in Two- Dimensional Fractures no= (14)
2 (p, + l1tip).
By using Equation 1, together with Equations 2 or 3, Equation This value of a produces the maximum fluid-lag distance given val-
6 can be written as an integral equation and then solved to find ues of Utip• L, and P!. The apparent fracture toughness associated
the fluid-lag distance. For example, for a two-dimensional plane with this maximum fluid-lag distance is:
strain fracture of length 2£, pressurized by p( x) from its center to
distance L - d, and subject to an effective closure stress of Uti in
the fluid-lag region, Equation 6 becomes: P
Kfc = Utip~
.
(1- __+ P!
ut_ip_)
l1tip
(15)

or
2 {[;__
V; lo
rL-d p( X) dx = K[;ck + 2 {[;__ [L l1tip( X)
V; JL-d .J£2 _ x2
dx (7) Kfc = l1tip~ ( P! ) (16)
.JL 2 - x2 P! + l1tip

270
SPE 18957 R.G. Jeffrey 3
If d = 0 is substituted into Equation 10 or 12, the fracture Upon substitution of this pressure distribution into the weight
toughness needed to obtain critical equilibrium is found to be: function integral, the stress intensity factor from the excess pressure
Kfc = PJ.;:;L is found to be:
(17)
2 2
which is the classical formula for the stress intensity factor for this PJ rR-p T dr = PJ [R- JR 2 - (R- PF] (22)
fracture geometry and loading. The following nomenclature, intro-
.;;R lo JR 2 - r2 .j;R
duced above, will be used throughout the rest of this paper. and the apparent fracture toughness arising as a result of the fluid
lag is found to be:
= the apparent fracture toughness
caused by the fluid lag, 2crtip
.j;R }R-p
{R
JR2 _ r2
r dr = 2crtip
~
[JR2- (R- p)2]. (23)
Kfc = the apparent fracture toughness
caused by the max1mum fluid lag
The fluid-lag distance is found by substituting these two equations
(KJ~ck = 0), into Equation 6.
Kf! = the fracture toughness of the rock
that would result in zero fluid lag,
Krock = the fracture toughness of the rock.
Ic

Quadratic Pressure Case

The maximum lag distance occurs for K]~ck = 0 and is given by:
A distribution of excess pressure that varies in a quadratic man- 2
ner is described by: 1
Po =R (1- \ 1- [ p .] ) • (25)
PJ + Uttp
(L- d)2- x2]
p(x)=pf [ (L-d)2 (18) Perkins and Krech 11 obtained an expression for a distance equivalent
to the maximum fluid-lag distance. Their expression (using the
where PJ is the excess pressure at the wellbore. The excess pressure notation of this paper) is identical to Equation 25. The maximum
decreases to zero at x = L - d. From x = L - d to x = L, a fluid-lag distance will result in an apparent fracture toughness of:
uniform effective stress acts across the fracture, as before.
Substitution of this into Equation 7 allows an expression for the
fluid-lag distance to be found.
Klc
L = 2CTtip {§ (
-
7r PJ
PJ
+ Utip
) (26)

d = L[1- sin(a 1 )] (19) which has the same form as Equation 17. Substituting p = 0 into
Equation 24 or 22, the fracture toughness that results in zero fluid
This equation is implicit in d because of the dependence of a 1 on lag is found to be:
d.
K]~ck ~ ~ _ y'L2-(L-d)2 (27)
a
1
= 2p1 VL + 2p; 2(£-d) (
20)
1--L-+~ which is the classical formula for the stress intensity factor for a
2(L-d)2 PJ
uniformly pressurized penny-shaped fracture of radius R.
and, as for the uniform pressure case, 0 ::; a 1 ::; ~- If d = 0
is substituted into Equation 19, the fracture toughness needed to Figure 5 is a graph showing the fluid-lag distance, given by Equa-
obtain critical equilibrium is found to be: tion 24, as K]~ck is varied. Figure 6 shows the influence on fluid-lag
distance as Utip is varied.
(21)

which is one-half the value given by Equation 17. NONDIMENSIONALIZATION

Equation 19 has been solved by iteration to obtain d as a func- Figure 7 shows a dimensionless ·plot that illustrates the variation
tion of K]~ck and CTtip· Figure 3 is a graph showing the variation of of fluid-lag distance as rock toughness is varied. The penny-shaped
fluid lag, given by Equations 19 and 12, as K]~ck is varied. Figure and the two-dimensional results plot on the same curves in this
4 shows the influence, as given by these equations, that Utip has on figure. The fluid-lag distances are normalized by their maximum
fluid-lag distance while other parameters of the problem are held values (corresponding to zero rock toughness). The rock fracture
constant. toughness is normalized by Kfc and the apparent toughness caused
by the fluid lag is normalized by its maximum value, K fc. Results
Fluid-Lag in Penny-Shaped Fractures from two different two-dimensional KGD-type hydraulic fracturing
simulators have also been plotted in Figure 7.
An expression for the fluid-lag distance, p, in a radially symmet-
ric (penny-shaped) fracture has been found for the case of uniform The results from Nilson's simulations 4 have been plotted as
excess pressure acting over the region 0 ::; r ::; R - p, where R is squares in Figure 7, and results from LORAMEC 12 are plotted as
the fracture radius. As before, the pressure in the fluid-lag region circles. The total dimensionless apparent toughness, arising from
is assumed to be equal to the reservoir pressure. the rock fracture toughness and from the apparent toughness of
271
4 THE COMBINED EFFECT OF FLUID LAG AND FRACTURE TOUGHNESS ON HYDRAULIC SPE 18957
FRACTURE PROPAGATION
the fluid lag is, in this figure, equal to 1.0. This fact is expressible In order to use these results, the numerical simulator would need
as: to perform the Kfc calculations for zero fluid lag. The simulator
KR Klag + KL Krock would then need to have. a leading edge consisting of one or more
Ic lc lc lc = 1.0 (28)
KfeKfc elements that could be deformed so that a node, at which pressure
or is to be specified, could be positioned exactly at distance p or d (as
Klag =K fc [K R _ Krock] given by Equation 33 or 31) from the tip of the fracture. These
Ic KR lc lc (29) leading-edge elements would simulate the fluid-lag region with the
lc
pressure set at p = Pr in this region. In order to increase com-
Therefore, if Kfe and Kfc are known, K~~g can be calculated by putational efficiency, the full set of calculations suggested by the
Equation 29. above procedure could be made only frequently enough to ensure
adequate accuracy for the simulator. The fluid-lag distance for in-
Numerical Implementation termediate time steps could be found by interpolation from fluid-lag
distances determined by a complete calculation.
For the three cases considered above, 1. KGD, uniform pressure;
2. penny-shaped, uniform pressure; 3. KGD, quadratic pressure, the If during the calculations K fc < K[~ck, then the fluid-lag dis-
ratio ~~Ic can easily be calculated. For these three cases, tance is zero and the pressure at the tip must be increased above
Pr until K 1 = K]~ck. For the case where the fluid-lag distance is
Kfc (]'tip small, compared to the fracture size, use of a total apparent frac-
(30) ture toughness consisting of the sum of Kj~ 9 and K]~ck should prove
Kfc PJ +(]'tip
adequate. This total apparent fracture toughness is used, together
This relationship may not hold if the pressure in the fluid-lag region with the condition that p = Pr at the tip, to determine when prop-
is not uniform, a condition that has not yet been studied. Equation agation occurs. As before, if K fc < K[~ck, then Kj~tal = K[~ck
30, and Equation 29 together provide a means of calculating Kj~ 9
(zero fluid-lag distance) and the pressure at the tip can be increased
and the fluid-lag distance. In order to find Kj~ 9 , Kfe must be cal-
above the reservoir pressure in order to meet the requirement that
culated. Alternatively, Kfc could be calculated, but the calculation
K 1 = Kj~ck. Numerical experimentation is needed to determine if
of K fie seems to pose fewer difficulties.
these approaches are useful.
In a numerical fracture simulator, Kfc can be calculated by set-
IMPLICATIONS FOR FRACTURE PROPAGATION
ting Kj~ck to zero and solving for the fluid-lag distance. The value
of Kfc can then be calculated. Kfe can be calculated by solving
Fluid Lag in KGD- Type Fractures
the hydraulic fracture problem with d = 0 and p = Pr as imposed
conditions at the tip. It may then be possible to run the simulator
The classical KGD equation for pressure at the wellbore (for
for a number of time steps using these values of Kfc and Kfe to
injection of a Newtonian fluid) is (Geertsma and Haafkens 13 ):
find K~~g using Equation 29 before updating Kfc and Kfc. Note
that if Equation 31 holds, only one of either Kfc or Kfc must be 1.91 [ E3QJJ-H3 ]1/4 (34)
calculated numerically. PJ = 2H 8(1 - v 2 )3 L2
Equation 29, shows that a condition for the existence of a fluid- Substituting this equation into Equation 17 gives an equation for
lag distance is that K fc > K[~ck. This suggests that a fracturing the fracture toughness required to give zero fluid lag in such a
model could be set up so that the fluid pressure at the tip is specified fracture:
as equal to the reservoir pressure, Pr. and no fluid lag is allowed.
(35)
After each converged solution is found, the stress intensity factor
resulting from this pressure distribution will be equal to Kfc. The and according to Equation 30:
value of Kfc can be calculated by using the weight function integral
or by calculating the stress intensity factor using a crack-opening KL- (]'tip KR (36)
lc - Pi +(]'tip /cl
relationship. If Kfc > K[~ck, the tip condition used was correct (a
fluid-lag region exists) and Equations 30 and 29 can be used to find where
Kj~ 9 . The fluid-lag distance can be found from Equation 11 (for
the assumed pressure distribution in the tip): E Young's modulus of the rock,
v Poisson's ratio of the rock,
d = L [1- sin(ak)] (31)
H height of fracture,
where
L fracture length,
~ K hlag y~
'-
)
ak =-- (32 Q volumetric injection rate,
2 2(J'tip.JI
f.L dynamic viscosity of fluid.
and for penny-shaped fractures, the fluid-lag distance is found from
Equation 23: By Equation 36, because PJ is a decreasing function of L, K fc
will asymptotically approach Kfc as L increases. According to
Equation 35, Kfc is independent of L. Therefore, the fluid-lag dis-
(33)
tance in a classical KGD fracture increases with length but asymp-

272
SPE 18957 R.G. Jeffrey 5
~---------------------------------------------.--~---------------------------------------------,
totically approaches a constant value for large L. Geerstma and
deKierk 14 reported that the fluid-lag distance was constant in a CONCLUSIONS
KGD fracture subject to constant injection rate.
The weight function method of calculating the stress intensity
Settari and Price 15 give an equation for an apparent surface factor provides a powerful tool for analyzing fluid lag development
energy, 1*, that represents energy dissipated in the fluid through in hydraulic fractures. A dimensionless representation of fluid-lag
viscous friction. Their equation is: distance, rock fracture toughness, and apparent fracture toughness
caused by the fluid lag has been presented. By using the resulting
112
2EO ] equations, a method of improving the efficiency of fluid-lag cal-
[ H ( 1-.ll)
/*=0.3589-
v2
(37)
culations in numerical models has been proposed. Finally, it has
been shown that the fluid-lag distance asymptotically approaches a
Noting that
2 2 constant value in classical KGD-type fractures and is a decreasing
K (1 - v )
2/ = --=-I.::_:c'----'- (38) function of fracture radius in penny-shaped fractures. The propa-
E gation of penny-shaped fractures can, therefore, become dominated
we find that Equations 35 and 37 are identical. by the rock fracture toughness.

Lag in Geertsma-deKierk Penny-Shaped Fracture

NOMENCLATURE
Geertsma and deKierk 14 give an approximate equation for Pi in a - fracture size (length or radius),
a penny-shaped fracture as:
- term in fluid-lag distance equation,
Rw
5 Gw - value of a for K]~ck = 0,
Pi= - - - l n - (39)
471" R R - term in equation for determining
where fluid-lag distance given Kj~ 9 ,
d - fluid-lag distance in KGD-type frac-
G elastic shear modulus,
ture,
Rw wellbore radius,
do - maximum d value (for K]~ck = 0),
w fracture width at wellbore.
E - Young's modulus of the rock,
A fixed value of Poisson's ratio (0.25) was used in deriving this
G - elastic shear modulus of rock,
equation. If Equation 39 is substituted into Equation 27, the tough-
ness needed to produce zero fluid lag is found to be: - apparent surface energy term,
- surface energy of fracture,
Kf = __5_ [QpG3]I/4ln ~~ (40)
height of fracture,
c 71"3/2 R R Khristianovitch-Geertsma-deKierk,
which is a decreasing function of R. Following the argument used - mode I stress intensity factor,
for KGD fractures, for cases where Pi<< CTtip• Kfc ~ Kfc and by - the apparent fracture toughness
Equation 29: caused by the fluid lag,
Klag:::::; (KR _Krock) (41) Kfc
Ic Ic Ic - the apparent fracture toughness
implying that Kj~ 9 decreases as R increases. Abe et aL1 6 also found caused by the maximum fluid lag
that fluid-lag distance was a decreasing function of R in a penny- (Kj~ck = 0),
shaped fracture subject to constant injection rate at the wellbore. Kfc - the fracture toughness of the rock
However, Geertsma and deKierk found that the fluid-lag distance that would result in zero fluid lag,
was independent of R. Krock - the fracture toughness of the rock.
lc
L - fracture half-length,
If the fluid-lag distance is a decreasing function of fracture radius
in a penny-shaped fracture driven by a constant injection rate, the m( x, a) - Bueckner weight function,
apparent toughness resulting from the fluid-lag region will decrease p dynamic viscosity of fluid,
as the fracture grows and, at some point, the fracture toughness v Poisson's ratio of the rock,
of the rock will start to control the propagation of the fracture. p( x) excess pressure in fracture,
Numerical simulations done by Thiercelin et aiY show this effect. Pr reservoir pressure,
Pi excess pressure at wellbore,
Po - total pressure in fracture,
Q volumetric injection rate,
R radius of penny-shaped fracture,
Rw wellbore radius,
CTtip - effective stress acting to close tip of
fracture ( cr 3 - Pr ),
273
6 THE COMBINED EFFECT OF FLUID LAG AND FRACTURE TOUGHNESS ON HYDRAULIC SPE 18957
FRACTURE PROPAGATION
- far-field minimum principal stress act- 11. Perkins, T.K., and Krech, W.W.: "The Energy Balance Con-
ing across plane of fracture, cept of Hydraulic Fracturing," SPEJ, Vol. 243 (1968)
w - width of penny-shaped fracture,
ACKNOWLEDGEMENTS 12. Ramamurthy, G., and Hujeux, J.C. "A Moving Mesh Integra
Variational Formulation Applied to Hydraulic Fracturing," pa-
The contents of this paper have benefited from discussions with per presented at the 1988 International Conference on Com-
Emmanual Detournay. I thank Dowell Schlumberger for permission putational Engineering Science, April 10-14.
to publish this paper.

REFERENCES 13. Geertsma, J., and Haafkens, R.: "A Comparison of the Theo-
ries for Predicting Width and Extent of Vertical Hydraulically
Induced Fractures," Jour. of Energy Resources Tech., Vol.
1. Biot, M.A., Medlin, W.L., and Masse, L.: "Laboratory Exper- 101, (March 1979).
iments in Fracture Propagation," paper SPE 10377 presented
at the 1981 Eastern Regional Meeting, Columbus, Nov. 4-6.
14. Geertsma, J., and deKierk, F.: "A Rapid Method of Predicting
Width and Extent of Hydraulically Induced Fractures," Jour.
2. Warpinski, N.R.: "Measurement of Width and Pressure in a of Petr. Tech. (December 1969) 1571-1581.
Propagating Hydraulic Fracture," paper SPE/DOE 11648 pre-
sented at the 1983 Symposium on Low Permeability, Denver,
March 14-16. 15. Settari, A., and Price, H.S.: "Simulation of Hydraulic Fractur-
ing in Low-Permeability Reservoirs," SPEJ (April1984) 141-
152.
3. Wiles, T.D., and Curran, J.H.: "Hydraulic Fracture Propaga-
tion Model for Discontinuous Rock Masses," report to Dowell
by University of Toronto, Dept. of Civil Engr., Aug. 1982. 16. Abe, H., Mura, T., and Keer, L.M.: "Growth Rate of a Penny-
Shaped Crack in Hydraulic Fracturing of Rocks," J. of Geophy.
Res., Vol. 81, No. 29 (October 1976) , pp 5335-5340.
4. Nilson, R.H.: "An Integral Method for Predicting Hydraulic
Fracture Propagation Driven by Gases or Liquids," Int. Jour.
for Numerical and Analytical Methods in Geomech. (1986) 17. Thiercelin, M., Jeffrey, R.G., and Ben Naceur, K.: "The In-
Vol. 10, 191-211. fluence of Fracture Toughness on the Geometry of Hydraulic
Fractures," paper SPE 16431 presented at the 1987 SPE/DOE
Low Perm. Res. Symposium, Denver, May 18-19.
5. Cleary, M.P., and Wong, S.K.: "Numerical Simulation of Un-
steady Fluid Flow and Propagation of a Circular Hydraulic
Fracture," Int. Jour. for Numerical and Analytical Methods
in Geomech., (1985) Vol. 9, 1-14.

6. Khristianovitch, S.A., and Zheltov, Y.P.: "Formation of Verti-

cal Fractures by Means of Highly Viscous Liquid," paper pre-
sented at the 4th World Petr. Congress, Vol. II, (1955) 579-
586.

7. Barenblatt, G.l.: "The Mathematical Theory of Equilibrium

Cracks in Brittle Fracture," Advances in Applied Mechanics,
Vol. Ill, (1962) 55-129.

8. Bueckner, H.F.: "A Novel Principle for the Computation of

Stress Intensity Factors, Z. agew. Math. Mech., Vol. 50,
(1970) 529-546.

9. Rice, J.R. "Some Remarks on Elastic Crack-Tip Stress Fields,"

Int. J. Solids Structures, Vol. 8, (1972) 751-758.

10. Parker, A.P. The Mechanics of Fracture and Fatigue, E. &

F.N. Spon Ltd., London (1981) 167.

274
 SfE 18957

~--------------21--------------~

+
iii iii i iii i iii c

Fig. 1: Fracture geometry and loading used Fig. 2: Superposition used to obtain the
to analyze fluid-lag effects. crack-line loading for the analysis.

Fluid-Lag Distance
KGD Fracture
versus

-
rn
lo.4
Q)
~
Fracture Toughness
1.8.----------------------------------.
1.61-T-----~~-------- uhp 15 MPa
3.5,---------------------------------.
L
P!
K[~"
100 meters
2 MPa
1 MPa y'7ii

-
Q) P! 2 MPa

-s 1.41------->r---------- L 100 meters

0~-----~-----r~==~~------~
0 5 10 15 20 25 30 35 40
10 20 25 30
15

Fracture Toughness (MPaJffi) Effective Tip Closure Stress (MPa)

Fig. 3: Variation of fluid-lag distance with Fig. 4: Variation of fluid-lag distance with ef-
rock fracture toughness for the two fective tip-closure stress ( O"t = O'a -
assumed pressure profiles. Pr)•

275
 l
t-.~
"·
,<._
SPE t8CJ 57

Penny-Shaped Fracture Penny-Shaped Fracture

-Ul
1-4
Q)
-+-1
-

B
~
1.4.--------------------.,

1.21-'r----------p'
R 100 mE"tE"rs
2M:Pa
Q) cu K;~• 1 MPa vm I
s s
-c:).
-
Q)
c.J
Q
ct1
-+-1
Ul
·~
Q 0.1
bfl
ct1
~ or-----r----~---~---~
0 5 10 15 20 25 10 15 20 25 30

Fracture Toughness (M Pa..foi) Effective Tip Closure Stress (MPa)

Fig. 5: Variation of fluid-lag distance with Fig. 6: Variation of fluid-lag distance with ef-
rock fracture toughness in a penny- fective tip-closure stress in a penny-
shaped fracture. shaped fracture.

Penny-Shaped and KGD Fractures

1.2r--------------------..,
, ___________ _ ~ T_otal

i
I

l
or---~---,----.---~--~
0.00 0.20 0.40 0.60 0.80 1.00

Normalized Fluid Lag (t or ;, ).

Fig. 7: Nondimensional plot for fracture
toughness and fluid-lag distance.

276