IADC/SPE 88011

Fracture Pressure Gradient in Deepwater

Luiz A. S. Rocha, Jos L. Falco, C. J. C. Gonalves, Petrobrs, Ceclia Toledo, Karen Lobato, Silvia Leal and
Helena Lobato, PUC-RJ

Copyright 2004, IADC/SPE Asia Pacific Drilling Technology Conference and Exhibition
will be part of the company database to simply build fracture
This paper was prepared for presentation at the IADC/SPE Asia Pacific Drilling Technology pressure gradient curves.
Conference and Exhibition held in Kuala Lumpur, Malaysia, 1315 September 2004.
The objective of this work is to make a critical
This paper was selected for presentation by an IADC/SPE Program Committee following
review of information contained in a proposal submitted by the author(s). Contents of the
examination of the current methods used to estimate fracture
paper, as presented, have not been reviewed by the International Association of Drilling pressure gradient. The work also presents simple
Contractors or 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 International methodologies based on leak of tests data to estimate fracture
Association of Drilling Contractors or Society of Petroleum Engineers, their officers, or
members. Papers presented at IADC/SPE meetings are subject to publication review by
pressure gradient for a given area. Results based on actual
Editorial Committees of the International Association of Drilling Contractors and Society of field data to exemplify the use of the presented methodology
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the International Association of Drilling will be shown, mostly for deepwater oilfield.
Contractors and 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 Introduction
presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A.,
fax 01-972-952-9435. Fracture pressure gradient is defined as the pressure gradient
that will cause fracture of the formation. In other words, if the
formation is exposed to a pressure higher than its fracture
Abstract
pressure limit, the formation will break (fracture) and possibly
The majority of the petroleum engineers agree that fracture
lost circulation will occur.
pressure gradient is one of the most important items to be
Smaller tolerance between pore pressure and fracture
considered when designing or drilling a well.
pressure gradient resulting in narrow pressure margins while
The perception of the importance of the fracture pressure
drilling is probably the most recognized deepwater challenge.
gradient comes from the results of the severe economic losses
The reduction of the fracture pressure gradient observed in the
that the oil industry has faced when dealing with lost
deeper water is mainly due to the low stress regime as result of
circulation related problems. In the worst scenario, these
the reduction of the overburden pressure gradient. In addition,
problems can escalate to a blowout due to the reduction of the
the structurally weak, low compacted, and unconsolidated
hydrostatic pressure in the well. In addition, lost circulation
sediments commonly found in the shallower portion of the
problems are likely to occur and become even harder to
underground can reduce even further the fracture gradient.
control if the well is in deepwater.
Under these circumstances, the operational window formed by
It is well known that a correct prediction of the fracture
the pore pressure and the fracture pressure gradient will be
pressure gradient minimize drilling problems. However, the
reduced more and more as the water depth increases. Typical
methods used by oil industry to perform this task are generally
examples of how a reduced operational window affects deep
based on equations or methodologies that give questionable
and ultra deepwater drilling are excessive number of casing
results and do not match actual field data.
strings, small hole size at total well depth, or inability to reach
Fracture pressure gradient methods are generally based
total depth, and fracturing the formation during kick control
either on equation derived from rock mechanics theories or in
operations.
simplified methods. Although, the first tries to closely
Methods for estimating fracture pressure gradient can be
represent the rocks underground behavior, they are too
classified as direct and indirect. Direct methods rely on
complex and call for a number of data that normally are not
measuring the pressure required to fracture the rock and the
available. On the other side, the second carries many
pressure required to propagate the resulting fracture. They are
simplifications and barely represents subsurface conditions.
generally based on a field procedure called leak off test and
However, the last is simple to use and consequently more
use mud to pressurize the well until formation fracture is
popular among drilling personnel. Regardless of the method, a
initiated. A leak off test (LOT) is a normal procedure in
good calibration, in general hard to be accomplished, is always
vertical wildcat wells where the formation fracture gradient is
necessary to provide good estimates.
not well established.
Finally, performing leak off tests (LOT) is usually the
Indirect methods are based on analytical or numerical
procedure carried out by most of the oil companies to establish
models and can be used to estimate fracture pressure test along
fracture pressure gradient values for a given area. Once LOT`s
the entire well. Some of them are well known in the oil
are obtained, they are used to calibrate simple equations or
industry, other were built for specific areas and all generally
require data most of time difficulty to obtain.
2 IADC/SPE 88011

Background characteristics such as water depth, rotary table, final depth,

The fracture pressure is the upper limit to which the pressure LOT, pore pressure and overburden gradients.
in the well should be kept below to avoid fracturing the
formation and causing lost circulation. Although the most
common approach when planning and drilling wells has been Water Final Rotary Table
LOT
Test
Well Depht Depth Depth Depth
to assume leak off pressure as the upper limit, other (m) (m) (m) (ppg) (m)
considerations sometimes are also taken into account and are A 1361 5020 14
9.8 1935
11.6 2765
discussed below having Fig. 1 as the base. B 1549 5528 24 10.3 2286
The typical extended leak off pressure chart displayed in Model
C 1522 3157 18 9.9 2169
Calibration
Fig. 1 shows four pressure points: D 1831 3898 18 10.5 2790
E 1489 4529 24 10.0 2310
- Leak off pressure is the pressure to which plastic F 1412 5347 14 10.4 1985
deformation starts occurring and is identified as the point G 1490 4826 24 10.3 2395
Fracture H 1391 6024 24 10.6 2741
where the curve is no longer a straight line. Depending Prediction I 1491 3675 14 9.8 2450
on the rock type, existing small fractures will open and Table 1 Well Characteristics
the formation will start taking fluid. Generally, typical
leak-off tests are stopped after this pressure is reached Test
Pore P. Overburden
and this value is assumed as the fracture pressure. Well Depth
- Breakdown pressure is the peak point of the graph and (m) (ppg) (ppg)
1935 8.8 10.7
where rock tensile strength is overcome. Often but not A
2765 9.3 12.9
always a peak pressure is observed before the formation Model B 2286 8.6 11.0
fractures and the well starts taking fluid5. C 2169 8.8 10.8
Calibration
- The initial shut in pressure, generally referred as ISIP, is D 2790 8.8 11.3
the pressure recorded immediately after the pump is E 2310 8.7 11.3
F 1985 8.8 11.0
stopped. The pressure will fall off to a level where it G 2395 8.5 11.3
Fracture
balances the formation stresses trying to close the H 2741 8.9 12.7
fracture. Prediction I 2450 8.7 11.8
- Fracture closure pressure is by definition equal to the
least or minimum principal stress of the formation Table 2 Well Characteristics
containing the fracture.
Although the actual upper limit is the breakdown pressure, Basically, the procedures used to estimate fracture pressure
it is common in the oil industry to assume leak off pressure as gradient can be divided in three groups: methods based on the
the safe limit pressure for planning and execution purposes. To determination of the stresses around the wellbore, methods
assume the fracture closure pressure (minimum stress) as an related to the minimum in situ stress and correlations
upper limit is considered as a conservative approach because developed for specific areas. A brief description of each
results shown that leak off pressure is about 10% higher than procedure and their applications are shown next.
the corresponding minimum5. The fact is that the term fracture
pressure gradient is not always clear. Sometimes it expresses Fracture Gradient Based On The Stresses Around The
either leak off pressure or breakdown pressure. In some cases, Wellbore
it can also mean formation closure pressure. It is important to The original compressive stress state acting on the rock before
point out that leak off tests in deviated wells may reflect the a well is drilled can be divided in three components. The first
reduced breakdown pressure in such wells, but the minimum one is vertical and equal to the overburden. The others two,
in situ stress, which is independent of the orientation of the Hmax and hmin, are horizontal and generally unequal. After
well, is still the same5. Finally, in this work fracture pressure the well is drilled, the mud replaces the drilled rock, resulting
gradient will mean leak off pressure. in a stress concentration around the wellbore. These
redistributed stresses are known as the Hoop Stress and are
Methods Used to Estimate Fracture Gradients represented by:
The idea of this work is to examine the methods according to - , wich acts circumferentially around the wellbore wall;
their underlying assumptions. Nine deepwater wells selected - r, the radial stress;
from the same basin were used in this study. Different
- Z, the axial stress, wich acts parallel to the wellbore
methods, wich will be explained below, were calibrated with
axis;
data obtained from six deepwater wells, refered as the six
- Z, an additional shear component, which is generated in
wells. The results were compared with the fracture pressure
deviated wells,
gradients from the three remaining deepwater wells, refered as
Since most strength criteria are expressed in terms of the
the three wells. A number of data that include overburden
principal stresses, the Hoop Stresses has to be converted into
pressure, pore pressure and LOT`s of each well were used. Its
these stresses in order to be compared with the stress state that
important to say that all the LOTs used where reported as
lead to rock failure. This is the underlying assumption used to
have been taken in shales. Fig. 2 shows the map with the six
estimate fracture gradient based on the method described in
wells (blue) used to calibrate the model and the three wells
this session.
(red) used to verify the results. Tables 1 and 2 show some well
IADC/SPE 88011 3

Although easy to state, determining stresses around the ZX = sin sin cos ( h min H max ) (9)
wellbore is a quite difficult and complex problem. They are
function of several factors that include in situ stresses, rock
c) Determination of the Hoop Stresses at the wellbore wall
type, wellbore angle, mud properties and pore pressure. In
since it will normally be the most critical stress state when
general, the solutions proposed by different authors2 are based
using a linear-elastic approach:
on assumptions that make the problem simple and feasible.
Assuming rocks as homogeneous, isotropic and perfectly
linear elastic materials is often the common departing point for
r = PW (10)
most of models described in the literature. As known, rocks
are neither homogeneous nor isotropic, so generally the results = ( X + Y PF ) 2( X Y ) cos 2
(11)
have a high uncertainty, although they help to explain certain h(1 2 )
4 XY sin2 ( PW PF )1
phenomena. In addition, even based on many simplifications,
1
the models still require a number of data that are difficult to
Z = ZZ 2 [( X Y )cos 2 + 2 XY sin2 ]
obtain.
The determination of the stress state that causes rock to (12)
failure is also hard to obtain due to lack of data. Known as h(1 2 )(PW PF )
+
rock failure criteria, it requires a number of lab tests that 1
generally is economically justified only for the reservoir rock.
Besides all these restrictions, a great number of works have Z = 2( YZ cos ZX sin ) (13)
been done to establish the stresses around the wellbore and
associate them to a given failure criterion. The relatively
p f = PF + h( PW PF ) (14)
simple method described below2,5,9 is based on a linear-elastic
approach and can be used in vertical and directional wells. It
has the following steps: Where:
a) Determination of the in-situ stresses (original
compressive stress state) V, Hmax and hmin. Although this is varies from 0o until 360o, representing points around the
the most important aspect of this method, the lack of in-situ wellbore wall.
stress values lead most of time the engineers to make
assumptions such as described below:
d) Conversion of the stresses at the wellbore wall in the
V = Overburden Pressure (1) three principal stresses, using the equations below:

+ Z Z
2
H max = h min = H (2)
A = + + Z
2
(15)
2 2
H pf
K SAW = (3)
V p f
+ Z Z
2

b) Transposition of the in-situ stress tensor to a coordinate B = + Z

2
(16)
system with one of its axes parallel to the wellbore axis and 2 2
another in a horizontal plane:

X = H max sin 2 + h min cos 2 (4) C = PW (17)

Y = cos 2 ( H max cos 2 + h min sin 2 ) Where:

(5) 1 = Maximum (A, B, C)
+ V sin 2 3 = Minimum (A, B, C)
2 = Intermediate Principal Stress
ZZ = sin 2 ( H max cos 2 + h min sin 2 )
(6) The expression above means that 1 will be the maximum
+ V cos 2 stress among A, B and C, 3 will be the minimum stress
among A, B and C, and 2 will be the intermediate one.
XY = cos sin cos ( H max h min ) (7) e) Comparison between the computed minimum effective
principal stress and the tensile strength. The failure is
considered to have initiated when the minimum effective
YZ = sin cos ( V H max cos 2
(8) stress at the wall is less than the tensile strength of the
h min sin 2 ) formation.
4 IADC/SPE 88011

3 pf t (18) f = h min (21)

Substituting Eq. 21 into Eq. 20, the expression for the

The following procedure is to check if the fracture will fracture gradient becomes:

f = K MS ( V p f ) + p f
propagate. The propagation criterion assumes that the
minimum horizontal principal stress is less than the minimum (22)
effective stress, and can be expressed as:
As shown on the previous method, the direct use of Eq. 22
3 p f h min (19) implies that overburden and pore pressure gradients have
already been estimated. The difference among the diverse
Besides the assumptions shown above, for the application methods based on minimum stress approach is the way KMS
of this method was necessary to assume zero for the tensile has been calculated4. Probably one of the simplest ways to
strength of the formation. The calibration of the model was obtain KMS to the entire well is to correlate it with sediment
done by varying the stress ratio, until the calculated fracture depth. The procedures are shown below:
gradient matches the LOT point for each of the six wells. The - KMS was calculated for each of the six wells using the
final result is shown in Fig. 3, where KSAW is plotted against Eq. 22 and Tables 1 and 2. The results are shown in
sediment depth. As can be seen, there are different values of Table 5;
stress ratio associated with different LOT sediment depths. At - KMS was plotted against sediment depth and is shown
the end, to make future predictions, a curve fitting was used to in Fig. 4;
represent the stress ratio for the entire basin. - A correlation for KMS was obtained for the entire
Table 3 shows the results of the application of this method basin. Fig. 4 displays the correlation.
for the six wells used to calibrated the model. Table 4 shows Table 5 shows the result of the application of this method
the predictions for the three wells. for the six wells. Similarly, Table 6 displays the predictions
for the three wells.
Depth
LOT Fracture Error Test
Well Test LOT KMS Fracture Error
(m) (ppg) (ppg) (ppg) Well Depth
1935 9.8 10.0 0.2 (ppg) (m) (ppg) (ppg)
A 9.8 1935 0.59 9.9 -0.1
2765 11.6 11.7 0.1 A
B 2286 10.3 10.1 -0.2 11.6 2765 0.62 11.5 0.1
C 2169 9.9 10.1 0.1 B 10.3 2286 0.60 10.0 0.3
D 2790 10.5 10.4 -0.1 C 9.9 2169 0.59 10.0 0.0
E 2310 10 10.3 0.3 D 10.5 2790 0.61 10.3 0.2
F 1985 10.4 10.1 -0.3 E 10.0 2310 0.60 10.2 -0.2
Average Error 0.0 F 10.4 1985 0.59 10.1 0.3
Average Error 0.1

Table 3 Result of the calibration of the method based on the

six wells Table 5 - Result of the calibration of the method based on the
six wells
Depth
LOT Fracture Error Test
Well Test LOT KMS Fracture Error
(m) (ppg) (ppg) (ppg) Well Depth
G 2395 10.3 10.3 0.0 (ppg) (m) (ppg) (ppg)
H 2741 10.56 11.4 0.8 G 10.3 2395 0.60 10.2 0.1
I 2450 9.8 10.7 0.9 H 10.6 2741 0.62 11.2 -0.6
Average Error 0.6 I 9.8 2450 0.61 10.6 -0.7
Average Error -0.4

Table 4 Result of the prediction of fracture gradients for the

Table 6 - Result of the prediction of fracture gradients for the
three wells
three wells

Specific Correlations to Estimate Fracture Gradient

Fracture Gradient Methods Based on Minimum Stress
The minimum stress methods are based on the statement that
the vertical and the horizontal effective stresses are related to
the following simple equation3, 4, 5:
h min p f
K MS = (20)
V p f
The main assumption here is that fracture of the formation
will take place when the tangential stress component equals to
the minimum in situ stress
A simple approach to estimate fracture pressure gradient for a
given area is through the use of specific correlations. The
methodology employed in this work departed from the idea
that a good correlation must take into account the effects of
several factors such as well depth, water depth, existing stress
state and leak-off test measurements. As the overburden
gradient is itself an in situ stress, and as it is function of well
depth and water depth, it was chosen as the parameter to be
correlated directly with available leak-off test data.
IADC/SPE 88011 5

The idea behind the methodology was further applied in Test

LOT Overburden Fracture Error
two scenarios: (a) where overburden gradients and LOT are Well Depth
(m) (ppg) (ppg) (ppg) (ppg)
available and (b) where only LOT is on hand.
A 1935 9.8 10.7 10.0 -0.2
a) Scenarios Where Overburden and LOT are Available 2765 11.6 12.9 11.6 0.0
B 2286 10.3 11.0 10.2 0.1
Fig. 5 shows the graphs of two direct correlations (linear C 2169 9.9 10.8 10.0 -0.1
and power) between leak-off test data and calculated D 2790 10.5 11.3 10.4 0.1
overburden gradients showed in Table 2 for the six of wells. E 2310 10.0 11.3 10.4 -0.4
F 1985 10.4 11.0 10.1 0.3
These correlations that can be used to predict fracture gradient
for this specific area are shown below: Average Error 0.0

f = 0.92 OBG (R = 0.84)

2
(23) Table 9 - Fracture pressure gradients based on the correlation
f = 1.39 OBG 0.83 for the six wells
f = 1.39 OBG 0,83 (R2 = 0.86) (24)
Test
LOT Overburden Fracture Error
Table 7 shows the results of the application of the Well Depth
(m) (ppg) (ppg) (ppg) (ppg)
correlation from Eq. 23 for the six wells. Similarly, Table 8 G 2395 10.3 11.3 10.4 -0.1
displays the predictions for the three wells. H 2741 10.6 12.7 11.4 -0.9
I 2450 9.8 11.8 10.8 -1.0
Test Average Error -0.6
LOT Overburden Fracture Error
Well Depth
(m) (ppg) (ppg) (ppg) (ppg)
1935 9.8 10.7 9.9 -0.1
Table 10 - Fracture pressure gradients based on the correlation
A
2765 11.6 12.9 11.9 -0.3 f = 1.39 OBG 0.83 for the three wells
B 2286 10.3 11.0 10.1 0.2
C 2169 9.9 10.8 9.9 0.0
D 2790 10.5 11.3 10.4 0.1 b) Scenarios Where Only LOT is on hand
E 2310 10.0 11.3 10.4 -0.4
F 1985 10.4 11.0 10.1 0.3 All the methods presented above were based on the
Average Error 0.0 assumptions that informations such as overburden, pore
pressure, leak-off test and rock characteristics are readily
accessible. However, as most of times this is not the case,
Table 7 - Fracture pressure gradients based on the correlation simple correlations based on available parameters can be quite
f = 0,92 OBG for the six wells useful for engineers.
The last method presented in this work is described in the
Test literature and is called Pseudo Overburden3. It uses leak-off
LOT Overburden Fracture Error
Well Depth test data only and is based on the following equation8:
(m) (ppg) (ppg) (ppg) (ppg)
G 2395 10.3 11.3 10.4 -0.1
( fl ) 0
H 2741 10.6 12.7 11.6 -1.1 pseudo = w Z w + g Z s
g

K0
(1 e K0 Z s
) (25)
I 2450 9.8 11.8 10.8 -1.0
Average Error -0.7

Eq. 25 can also be expressed in gradient as follows:

Table 8 - Fracture pressure gradients based on the correlation
f = 0,92 OBG for the three wells
Table 9 shows the results of the application of the
correlation from Eq. 24 for the six wells. Similarly, Table 10
displays the predictions for the three wells.
pseudo
G pseudo = (26)
Z
To apply the method is needed to cross plot pseudo vs
LOT and G pseudo vs LOT to find the parameters 0 and K 0 .
The simultaneous calibration of Eq. 25 and 26 is accomplished
through a trial and error process. Initial values are picked for
0 and K 0 and pseudo-overburden stress and pseudo-
overburden stress gradient are calculated at each leak-off test
location. If the pseudo-overburden curves match the fracture
data, the points will fall along a straight line through the origin
with a slope of 1. After the match is obtained, the fracture
6 IADC/SPE 88011

pressure gradient is assumed equal to the pseudo-overburden - All the method presented absolute errors smaller than
as follows: 0,06 ppg for the six wells and smaller than 0.72 ppg for
the three wells;
f = pseudo (27) - The Pseudo Overburden Gradient Method led to smallest
average absolute error for the six wells;
- The method based on the Minim Stress gave the smallest
Figs. 6 (in psi) and 7 (in ppg) displays pseudo-overburden
average absolute error for the three wells;
curves vs LOT. The average values used for w and g in - Well G presented the smallest absolute error among all
the calculations and the results obtained for 0 and K 0 are the methods used; and
- Well H and I displayed the worst results, presenting
shown below:
error up to 1.10 ppg.
w = 1.03 gr/cm3 The analysis of the Fig. 8, 9 and 10, which show the plot of
g = 2.6 gr/cm3 the different methods, indicates that:
- The largest differences among the methods occur either
0 = 0.809 at shallow depths or at deeper depths.
K 0 = 4.90 x 10-4 - All the methods show result very similar to each other
near leak-off test locations;
Table 11 shows the result of the application of this method - A worst and a best scenario for fracture gradient
for the six wells. Table 12 displays the predictions for the estimates at each depth can be obtained by applying
three wells. simultaneously all the methods.

Wells Data Results For Differents Procedures

Test
LOT Pseudo Error LOT = 0,92 LOT = 1,39
Well Depth Pseudo-OBG
K SAW K MS OBG OBG^0,83
(ppg) (m) (ppg) (ppg) Well
LOT Error Error Error Error Error
9.8 1935 10.0 -0.2 (ppg) (ppg) (ppg) (ppg) (ppg) (ppg)
A
11.6 2765 11.5 0.1 9.8 0.23 -0.14 -0.1 -0.17 -0.19
A
B 10.3 2286 10.2 0.1 11.6 0.11 0.07 -0.3 -0.01 0.08
C 9.9 2169 10.0 -0.1 B 10.3 -0.18 0.29 0.2 0.14 0.11
D 10.5 2790 10.4 0.1 C 9.9 0.13 -0.04 0.0 -0.08 -0.11
E 10.0 2310 10.4 -0.4 D 10.5 -0.09 0.19 0.1 0.10 0.08
E 10.0 0.34 -0.24 -0.4 -0.38 -0.37
F 10.4 1985 10.0 0.4
F 10.4 -0.29 0.32 0.3 0.25 0.44
Average Error 0.0 Average Error 0.04 0.06 -0.02 -0.02 0.01

Table. 11 Result of the calibration of the method based on the

six wells Table 13 - Comparisons among the errors obtained from
different methods based on the six wells
Test
LOT Pseudo Error
Well Depth Wells Data Results For Differents Procedures
(ppg) (m) (ppg) (ppg)
LOT = LOT = 1,39
G 10.3 2395 10.5 -0.2 KSAW KMS 0,92 OBG OBG^0,83
Pseudo-OBG
H 10.6 2741 11.3 -0.8 Well
I 9.8 2450 10.6 -0.8 LOT Fracture Fracture Fracture Fracture Fracture
Average Error -0.6 (ppg) (ppg) (ppg) (ppg) (ppg) (ppg)
9.8 10.03 9.94 9.9 9.97 9.99
A
Table 12 - Result of the prediction of fracture gradients for the 11.6 11.71 11.53 11.9 11.61 11.52
three wells B 10.3 10.12 10.01 10.1 10.16 10.19
C 9.9 10.06 9.97 9.9 10.01 10.04
D 10.5 10.41 10.31 10.4 10.40 10.42
E 10.0 10.34 10.24 10.4 10.38 10.37
Comparisons among different Methods F 10.4 10.11 10.08 10.1 10.15 9.96
Table 13 summarizes the errors between fracture gradients
estimated from different methods and actual leak-off test Table. 14 Comparisons among LOTs and fracture gradients
values for the six wells. Table 14 displays fracture gradient obtained from different methods based on the six wells
estimates for the same set of wells. Tables 15 and 16 are for
the three wells and are similar to the previous one
Figs 8, 9 and 10 are referred to the three wells. They show
pore pressure, overburden gradient and fracture gradients
estimated using the different methods described above.
The analysis of the Tables 13 to 16, which are referred to
LOT, indicate that:
IADC/SPE 88011 7

Wells Data Results For Differents Procedures Pf = pore fluid pressure at the wellbore wall;
LOT = 0,92 LOT = 1,39 PF = far field formation pressure;
Pseudo-OBG
KSAW KMS OBG OBG^0,83 PW = well pressure;
Well 1 = Maximum (A, B, C);
LOT Error Error Error Error Error
3 = Minimum (A, B, C); and
(ppg) (ppg) (ppg) (ppg) (ppg) (ppg)
2 = Intermediate Principal Stress;
G 10.3 0.03 0.11 -0.1 -0.08 -0.23
H 10.6 0.82 -0.64 -1.1 -0.86 -0.78 0 = Pseudo surface porosity
I 9.8 0.90 -0.75 -1.0 -0.95 -0.82 0 = Pseudo declining porosity
Average Error 0.58 -0.43 -0.72 -0.63 -0.61 w = Water density
g = Grain density
Table. 15 Comparisons among the errors obtained from fl = Formation fluid density
different methods based on the three wells Zw = Water depth
Zs = Sediment depth
Wells Data Results For Differents Procedures Z= Depth;
LOT = LOT = 1,39 Gpseudo = Pseudo overburden gradient
Pseudo-OBG
KSAW KMS 0,92 OBG OBG^0,83
Well References
LOT Fracture Fracture Fracture Fracture Fracture
(ppg) (ppg) (ppg) (ppg) (ppg) (ppg) 1. McLean, M. R, and Addis, M. A., 1990,
G 10.3 10.33 10.19 10.4 10.38 10.53 Wellbore Stability Analysis: A Review of
H 10.6 11.38 11.20 11.6 11.42 11.34 Current Methods of Analysis and Their Field
I 9.8 10.70 10.55 10.8 10.75 10.62
Application, IADC/SPE Drilling Conference in
Houston, Texas, IADC/SPE 19941.
Table. 16 Comparisons among LOTs and fracture gradients 2. McLean, M. R, and Addis, M. A., 1990,
obtained from different methods based on the three wells Wellbore Stability Analysis: The Effect of
Strength Criteria on Mud Weight
Conclusions Recommendations, SPE Annual Technical
A number of different fracture gradient methods were Conference in New Orleans, LA, SPE 20405.
examined in this work. They were classified according to their 3. Rocha, L.A. and Bourgoyne, A.T, A New
underlying assumptions and were divided in three groups: Simple Method to Estimate Fracture Pressure
methods based on the determination of the stresses around the Gradient, SPE Drilling & Completion Magazine,
wellbore, methods related to the minimum in situ stress and September 1996, pages 153-159, USA.
correlations developed for specific areas. 4. Bowers, Glenn, Best Practice Procedure for
Calibrating each one using available leak-off test values did Predicting Pre-Drill Geopressures in Deep Water
the application of these methods. The results come to similar Gulf of Mexico, DEA Project 119, June 2001,
average errors. Similar fact occurred when using the calibrated Knowledge System Inc., USA.
methods to predict fracture gradient for the basin. 5. Erling, F. et al., Petroleum Related Rock
Mechanics, Elsevier, 1992, New York, USA.
Nomenclature 6. Altun, G., Langlinais J, Bourgoyne, A.T.,
LOT = Leak Off Test; Application of a New Model to Analyse Leak off
V = Vertical Stress; Tests, SPE Drilling & Completion Magazine,
H = Horizontal Stress; September, 2001, pages 1083-116, USA.
H max = Maximum Horizontal Stress; 7. Addis M. A. et al., A Comparison of Leak-Off
h min = Minimum Horizontal Stress; Test and Extended Leak-Off Test Data for Stress
= Stress around wellbore wall ; Estimation, SPE/ISRM 47235, SPE/ISRM
r = Radial Stress; Eurock 98, Tondheim, Norway.
Z = Axial Stress; 8. Bourgoyne, Adam T., Millheim, Chenevert,
T = Tensile Strength; Martin E. and Jr. Young, F.S., Applied Drilling
Z = Shear Component; Engineering, SPE, 1986, USA.
f = Fracture Stress; 9. Jaeger, J.C. and Cook, N.G.W., Fundamentals
OBG = Overburden Gradient; of Rock Mechanics, Third Edition, Chapman and
pseudo = Pseudo-Overburden Stress; Hall, 1979, London.
K = Stress Ratio
KMS = Stress Ratio to Minimum Stress Method;
KSAW = Stress Ratio to Stress Around the Wellbore
Method;
= well inclination;
= well azimuth;
= angle around the well bore;
8 IADC/SPE 88011

Figures K MS
0.0 0.2 0.4 0.6 0.8 1.0
400
Pressure

Breakdown Pressure
Leak off Pressure
600

Initial Shut In Pressure (ISIP)

Sediment Depth (m)

800
Fracture Closure Pressure or
Minimum Stress

1000

Time
1200

Volume
1400
Fig. 1 - Typical Extended Leak-off Test

1600

KMS = 0.03 ln(depth) + 0.4

Map
7330000

Well F
7320000

Well H Well I Well B

Fig. 4 - Stress ratio based on leak-off test for the method based
7310000
on the minimum stress.
UTM NS (meters)

7300000 Well A Well E

Well C
Well G
7290000 13

7280000

Well D
7270000

7260000 12
680000 700000 720000 740000 760000 780000

UTM EW (meters)
Wells - Model Calibration
L O T = 0 ,9 2 O ve rb urd e n
Wells - Fracture Prediction
R 2 = 0 ,8 4
L e ak o ff T est (p p g )

Fig. 2 - Map showing well locations. Wells in blue were used 11

to calibrate the model and wells in red to verify it.

K SAW = 0.02 * ln(Sedim. Depth)+0.1907

10
0,10 0,15 0,20 0,25 0,30 0,35 0,40 0,45 0,50 0,55 0,60
500
L O T = 1 ,3 9 O ve rb urd e n 0,83
600
R 2 = 0 ,8 6
700

800 9
Sediment Depth (m)

9 10 11 12 13
900
O v e rb u d e n (p p g )
1000

1100
Fig. 5 - Correlation between LOT and overburden gradient
1200

1300

1400

1500

Fig. 3 Stress ratio based on leak-off test for the method

based on the stress around the wellbore
IADC/SPE 88011 9

6000 Well "G"

Gradient (ppg)
7 9 11 13 15 17
1000
5500
Overburden
Pseudo Overburden = 1,01 LOT1,00 1500 Pseudo-OBG
LOT
R2 = 0,98 Frac. KMS
Pore Pressure
5000 2000 Frac. "LOT=0.92xOBG"
Pseudo-Overburden (psi)

Frac."LOT=1,39xOBG^0,83
Frac. KSAW

Depth (meters)
2500
4500
3000

3500
4000

4000

3500 4500

5000
3000
3000 3500 4000 4500 5000 5500 6000

LOT (psi)
Fig. 8 Plot Gradients vs Depth shows fracture gradients
Fig. 6 - Correlation between LOT and pseudo-overburden
from different procedures, pore pressure and OBG of well
gradient in psi.
G

Well "H"
Gradient (ppg)
7 9 11 13 15 17
1000
Overburden
Pseudo-OBG
1500 LOT
Fracture from KMS
Pore Pressure
2000 Fracture from "LOT=0,92xOBG"
Fracture from
Fracture from KSAW
2500

3000
Depth (meters)

3500

4000

4500

5000

5500

6000

Fig. 9 Plot Gradients vs Depth shows fracture gradients

Fig. 7 - Correlation between LOT and pseudo-overburden from different procedures, pore pressure and OBG of well
gradient in ppg. H
10 IADC/SPE 88011

Well "I"
Gradient (ppg)
7 9 11 13 15 17
1000
Overburden
Pseudo-OBG
LOT
1500 Fracture from KMS
Pore Pressure
Fracture from
Fracture from
Fracture from KSAW
2000
Depth (meters)

2500

3000

3500

4000

Fig. 10 Plot Gradients vs Depth fshows fracture gradients

from different procedures, pore pressure and OBG of well
I