Halliburton Overpressure

Fracture pressure
Fracture gradients

Knowledge of the magnitude of formation fracture gradients is vital, especially when

drilling into an abnormally pressured zone. Formation fracture gradient determines
the maximum allowable mud weight that can be used (after incorporating an
operational safety factor).

Mechanism of Formation Fracturing

A formation can be made to fracture by the application of fluid pressure to overcome
the least line of resistance within the rock structure. Normally fractures will be
propagated in a direction perpendicular to the least principal stress. Which of these
three stresses is the least can be predicted by the fault activity in the area.

Figure 1: Distribution of Stress Planes in Unit of Rock Formation

Normal faulting indicates that the least principal stress is horizontal and probably
about equal to the minimum horizontal stress required to avoid rock failure (figure 2).

Figure 2: Normal Fault

Thrust (reverse) faulting indicates the least principal stress to be of a vertical

nature. It must be at least equal to the weight of the overburden, or about 1 psi/ft
(figure 3).

Fracture Pressure 7-1 Ver. 1.0

Halliburton Overpressure

Figure 3: Reverse Fault

Transcurrent faulting indicates that the least principal stress is horizontal but can
be larger than the minimum required to avoid failure. It cannot be larger than the
vertical stress (figure 4).

Figure 4: Transcurrent Fault

To initiate the fracture fluid pressure must be transmitted to the formation, and to
propagate the fracture this pressure must be maintained at a level greater than the
least principal stress. If mud is used, then the fluid pressure has to overcome the
filter cake deposited on the borehole wall, plus the pressure at which the formation
fractures. It is therefore preferable to use a clear, low viscosity fluid when making a
determination of the formation breakdown gradient.

In practice, formation breakdown gradients are determined by performing a Leak Off

Test (LOT). This is normally done after drilling through a casing shoe in order to
determine the maximum allowable mud gradient for the next section of open hole.
The usual procedure is outlined below.

a) Drill through the casing shoe and into new formation to a depth of 15 to 20
feet. Close the blowout preventers.

b) Raise the surface pressure in increments until bleed off is indicated.

Alternatively, pumping could be maintained at a slow rate until bleed off.

Fracture Pressure 7-2 Ver. 1.0

Halliburton Overpressure

Figure 7-5: Typical Leak Off Test

During fracture testing it is generally not the coherent rock matrix that is sheared,
rather the existing natural discontinuities (joints, parting, etc) which are normally held
together by the compressive stresses in the formation. When the hydraulic stress in
the borehole equals the formation compressive stress, it effectively reduces the latter
to zero across the crack. If the borehole pressure is then raised slightly, fluid will
enter the crack.

Fracture Pressure 7-3 Ver. 1.0

Halliburton Overpressure

Leak off tests and fracture pressures

Terzaghi and Peck (1948) states that

S =σ + P

where S = Overburden Stress

σ = Matrix Stress
P = Pore Pressure

As previously stated, subsurface stress is characterised by three unequal, mutually

perpendicular principal stresses.

σ h ) is of
The greatest stress is vertical (σv ) although the least principal stress (σ
greatest interest as this indicates formation strength along the weakest plane.

Hubbert and Willis (1957)

Hubbert and Willis conducted laboratory experiments on core samples and derived
the empirical relationships below.

To extend a fracture

1
σh = σv
3

To induce a fracture

1
σh = σv
2

Hence

1 1
σv < σh < σv
3 2

or

1 1
(S - P) < σh < (S - P)
3 2

Fracture Pressure 7-4 Ver. 1.0

Halliburton Overpressure

Therefore, to extend

1
FP = P + (S - P)
3

while to induce

1
FP = P + (S - P)
2

Therefore to fracture a formation one must overcome both the pore pressure and the
horizontal matrix stress of the rock.

This value (1/3 to 1/2) is only an estimate. The actual value will depend upon
lithology and compaction of the sediments. To take account of this, the Poisson's
Ratio was introduced.

Poisson's Ratio
From Young's Law

σh = ( µ n
)
σv 1 - µ

where n=1
µ = Poisson's Ratio

Hubbert and Willis estimated Poisson's Ratio as between 0.33 and 0.25, ie

0.33 1 0.25 1
= while =
1 - 0.33 2 1 - 0.25 3

Poisson's Ratio is a property which describes the behaviour of rock stresses in the
least principal stress direction when pressure is applied in the direction of the
principal stress. For elastic rocks, the ratio of the stresses is represented by

σL = µ
σP 1 - µ

Laboratory tests on unconsolidated rocks have shown that

L 1
= ie µ = 0.25
P 3

Fracture Pressure 7-5 Ver. 1.0

Halliburton Overpressure

Field tests have shown µ to have values between 0.25 and 0.5, above which the
rock becomes plastic, ie the stresses are equal in all directions. Poisson's Ratio varies
with both rock type and the degree of compaction, and will never exceed 0.5. Every
rock will have its own characteristic Poisson's Ratio.

Note should be made of unconsolidated clay formations, often found at shallow

depths in North Sea wells, which may exhibit abnormally high fracture pressures.
Wet clays will behave plastically and the Poisson's Ratio will approach 0.5. Due to
the pore water and adsorbed water surrounding the individual clay platelets, these
platelets will not be in contact with each other, and so these clay types will have very
low shear strength. Thus the pore water would be supporting the weight of the
overlying sediments and the pore pressure should almost be equal to the overburden
pressure. As a result, hydraulic pressures greater than the overburden pressures
would be necessary to fracture this formation. Under these circumstances, a
horizontal fracture would form lifting the overburden.

With increased depth, compaction will squeeze out some of the pore water, bringing
the clay platelets into contact. They are then able to support a superimposed
horizontal stress, and Poisson's Ratio will become similar to that of a more compact
clay.

To build up a complete fracture gradient picture for any well, the fracture pressure
should be calculated every time a lithology or pore pressure change is noted. This is
particularly important when drilling into sands or zones of decreasing pore pressure
gradient, which may be less competent than the formation at the shoe.

In the case of changing lithology, evaluation is best carried out using experimentally
derived Poisson's Ratio values for different lithologies, such as those of Weurker
(1963) shown in table 1.

Fracture Pressure 7-6 Ver. 1.0

Halliburton Overpressure

Matthews and Kelly Method

For compacted shales, it has been shown that the overburden stress (S) is
supported by the formation fluid pressure (P) and the grain-to-grain bearing strength
( ) of the clay particles.
S =σ + P
In the above equation σ represents the matrix stress, and it is evident that it's value will
depend solely upon the degree of rock compaction. Hubbert and Rubey state that "a useful
measure of the degree of compaction of a clay is its porosity". This is defined as the ratio of
pore volume to the total volume. From this it may be inferred that, for any given clay, for
each value of Φ, there is some minimum value of matrix stress σ which the clay can support
without further compaction. The porosity at a given burial depth (D) is dependent on the
fluid pressure (P). Thus if the fluid pressure is abnormally high (greater than hydrostatic) the
porosity will also be abnormally high, and so the matrix stress would be abnormally low. The
converse is assumed to be true if the formation fluid pressure is abnormally low.

It has been assumed that whenever a formation is fractured horizontally the fracture pressure
is about equal to the overburden stress. However, for the case of a vertical fracture, the stress
required may be much less then the overburden load at the depth where the fracture is
created. Nevertheless, in both the horizontal and vertical fracture cases, it may be safely
assumed that the fracture will not begin until the applied pressure is at least equal to the
formation fluid pressure. If this is the case, then any additional pressure required may be
related to the matrix stress and hence will only vary with the degree of compaction. The
relationship below has been developed for calculating the fracture gradient of sedimentary
formations.

P σ
F = + χ
D D

where P = the formation fluid pressure (psi) at depth of interest D

σ = the matrix stress (psi)
χ(chi) = the matrix stress coefficient for the depth at which the value
of σ would be the normal matrix stress
F = the fracture gradient in psi/ft

Fracture Pressure 7-7 Ver. 1.0

Halliburton Overpressure

Matthews and Kelly (1967) assumed 1.0 psi/ft for overburden, but replaced the Poisson's
χ ).
Ratio by a matrix stress coefficient (χ

χ was determined from empirical data, ie leak off tests.

FP P P
= + χ (1 - )
D D D

This showed that

a) χ was depth dependent

b) χ was geographical area dependent

Eaton (1969)
Eaton established variable overburden gradients from bulk densities, and calculated a
Poisson's Ratio from empirical data.

FP S P µ P
= ( - )( )+
D D D 1 - µ D

where S/D from density logs

P/D = actual pore fluid pressure
µ from empirical data, or by Anderson et al method for sands

Fracture Pressure 7-8 Ver. 1.0

Halliburton Overpressure

Anderson et al (1973)
Anderson expands Biot's (1955) stress/strain relationship

FP 2 S 1 - 3µ P
= [ ] + [a ( ) ]
D 1 - µ D 1 - µ D

where

a = 1 - Cr
Cb

Cr = intrinsic compressibility of solid rock material

Cb = compressibility of solid rock skeleton

Cr/Cb = function of shale estimated in cuttings

If cuttings contain 100% shale then a = 0, ie no porosity, hence pore pressure does not have
an effect. The equation becomes

FP 2µ S
=
D 1 - µ D

Fracture Pressure 7-9 Ver. 1.0

Halliburton Overpressure

Daines (1982)

Daines states that there are two unequal horizontal stresses which must be overcome before
fracturing occurs.

1 That caused by the weight of overlying sediments (σh )

2 A superposed tectonic stress (σT )

Proof that this tectonic stress exists is evidenced by folding, faulting, etc, which rely on
unequal stress states for their occurrence and maintenance.

The minimum horizontal stress occurs when the tectonic stress (σT ) is zero, ie

µ
σ H = σV ( ) where σV = (S - P)
1 - µ

The superposed horizontal tectonic stress will increase uniformly with depth such that the
ratio σT :σv remains constant.

The value of σT is calculated from the first LOT by subtracting the Eaton calculated fracture
pressure from the LOT result, ie

µ
σT = LOT - [ σV ( ) + P]
1 - µ

where σV = S - P

The ratio of σT :σV is then calculated so that for subsequent points in the borehole the value of
σT can be calculated from

= σV i ( σT l )
l

σT
i

σV

where i denotes depth of interest

l denotes LOT depth

Fracture Pressure 7-10 Ver. 1.0

Halliburton Overpressure

As a result, the fracture pressure can be calculated for any depth if the overburden gradient,
pore pressure, and Poisson's Ratio for the lithology at that depth are known

µ
FP = σT + [ σV ( ) + P]
1 - µ

To simplify this calculation Daines paper provides a list of typical Poisson's Ratios for certain
lithologies reproduced here in table 7-1 (from Weurker 1963).

Obviously if this data is to be accurate then strict lithological control is necessary. Problems
arise when two types of lithology are intermixed eg sandy clay and shaly sand. In these cases
the mineral forming the matrix must be determined and the Poisson's Ratio for that used.

Daines quotes a figure of 95% accuracy for use of his equations.

Fracture Pressure 7-11 Ver. 1.0

Halliburton Overpressure

Formation Type Poisson's ratio

clay, very wet 0.50

Clay 0.17

Conglomerate 0.20

Dolomite 0.21

greywacke:
coarse 0.07
fine 0.23
medium 0.24

Limestone:
fine, micritic 0.28
medium, calcarenitic 0.31
porous 0.20
stylolitic 0.27
fossiliferous 0.09
bedded fossils 0.17
shaly 0.17

sandstone:
coarse 0.05
coarse, cemented 0.10
fine 0.03
very fine 0.04
medium 0.06
poorly sorted, shaly 0.24
fossiliferous 0.01

shale:
calacareous (<50% CaCO3 ) 0.14
dolomitic 0.28
siliceous 0.12
silty (<70% silt) 0.17
sandy (<70% sand) 0.12
kerogenaceous 0.25

Siltstone 0.08

Slate 0.13

tuff (glass) 0.34

Table 1: Suggested Values of Poisson's Ratio

(From Weurker R.G. "Annotated Tables of Strength & Elastic Properties of rocks," Drilling
Reprint Series SPE, Dallas (1963))

Fracture Pressure 7-12 Ver. 1.0

Halliburton Overpressure

Poisson's Ratio Calculations

This method was described by Tony Taylor (1978) and is based on an equation formulated by
Tixier, Loveless and Anderson in 1973. The Poisson's Ratios calculated by this method can
be applied to sands , particularly those with significant clay content.

Since the minimum value of µ available is 0.27, this method cannot be used to define
Poisson's Ration in finer grained clastics. Also, since carbonates tend to have low values of
gamma ray the Poisson's ratio would appear much higher than would be appropriate for such
competent rocks.
µ = 0.125 q + 0.27

where µ = Poisson's Ratio

0.125 = constant
q = Shaliness Index
0.27 = theoretical minimum for Poisson's Ratio in a clean sand

To calculate shaliness index (q) use the gamma ray log. Average the readings over 50 ft
intervals and insert into the formula

- GRmin
q = GRlog
GRmax - GRmin

This value is substituted into the equation for µ above, and in turn Poisson's ratio (µ) may be
substituted into the desired fracture gradient equation discussed above.

Fracture Pressure 7-13 Ver. 1.0