You are on page 1of 17

SPWLA TWENTY-SECOND ANNUAL LOGGING SYMPOSIUM, JUNE 23-26, 1981

MECHANICAL PROPERTIES PROGRAM USING BOREHOLE ANALYSIS AND MOHR’S CIRCLE

by

G.R. COATES and S.A. DENOO


Schlumberger Well Services, Houston, Texas

ABSTRACT

The prediction of sanding from the mechanical properties program has always been a bit of a localized
art normally built in case histories and interpretation of G/Cb ratios. In this paper, a new method is presented
to determine sand strength limits using tangential and radial stresses with Mohr’s Circle.

The availability of reliable shear velocities leads to a refinement of calculated elastic rock properties as
well as an upgrade of the mechanical properties program. Within the text of this paper are presented the
elastic moduli determination, the uniaxial strength calculations, individual borehole stress components, rock
failure theories and predicted failure values.

INTRODUCTION DO
.,..
“,

The behavior of rocks to stress induced by increasing or decreasing mud weight or pore pressure is a very
important element in analyzing the total reservoir system. Unfortunately for us most of the work done in rock
mechanics has been in mining applications and this technology has not found its way gracefully into the
petroleum industry. Part of the reason for this may be due to the lack of laboratory equipment capable of
measuring rock properties in the original insitu conditions. Instead, correlations based on acoustic travel
times and bulk density were developed to measure certain elastic rock properties, namely Bulk Compressibili-
ty, Shear modulus, Youngs modulus and Poisson’s ratio.

Elastic properties from the early program contained correlations that were valid in a limited Gulf Coast
type environment. Poisson’s ratio, for example, was simply a function of shale content. These provide fairly
accurate elastic parameters when reservoir conditions are similiar to those first studied, but poor answers
when they are not. A problem arises when the conditions are not met but the answers are used anyway. This
has caused the industry to view log derived mechanical properties with some suspicion.

A more universal approach to the problem lies in the direct measurement of the shear wave velocity and
the direct calculation of the elastic rock properties by the proven physical relationships:

~Bv,2 (3VC2 - 4v~2)


E= Youngs Modulus VC= Compressional Velocity
VC2 - v~’
Vs = Shear Velocity
G = Q~V~2 Shear Modulus
VC2- 2VS2
P= Poisson’s Ratio
2(VC2 - v,’)

-1-
SPWLA TWENTY-SECOND ANNUAL LOGGING SYMPOSIUM, JUNE 23-26, 1981

The measurement of the shear wave velocity requires a rather sophisticated process and is currently .,..
available only in limited areas. A broader approach is needed. The correlations of Pickett’s laboratory data
relating shear to compressional velocity and lithology have given us a new, more reliable mechanical proper-
ties program. However, Pickett’s correlations requires the precise determination of lithology and a hydrocar-
bon free compressional velocity of the shale free matrix to obtain the correct shear velocity of travel time.’
The determination of this variable, the clean hydrocarbon free travel time, is more difficult than it would first
appear. The correction for the sonic hydrocarbon component is not a straight volumetric equation. A
statistical sonic reconstruction was developed to recompute the desired travel time rather than subtract off a
given percentage of the signal.

This overall method when used in comparisons with the direct shear velocity measurement show the
technique to be very sound.’

STRESS LEVELS

To begin the discussion of stress levels, consider an infinitesimal element deep within the earth. Three
stresses, X, Y, and Z, are confining the block. The overburden stress, the Z direction can be obtained by in-
tegrating the bulk density to that depth. The X and Y stresses can not, with present technology, be measured
individually but can be written in terms of the overburden, pore pressure and poisson’s ratio according to the
elastic isotropic porous horizontally constrained model

‘X=m () v
P.+cl Pp
() 1-+
l-p
*Unbalanced stress

L Po+o!Pp 1-+
‘y= ()l-p () l-p

mz = PO

Figure 1

where:
Po = overburden pressure
Pp = pore pressure
L4 = Poisson’s ratio
a (alfa ) = 1-compressibility of rock matric/Bulk compressibilityy

-2-
SPWLA TWENTY-SECOND ANNUAL LOGGING SYMPOSIUM, JUNE 23-26, 1981

The unbalance between the X and Y stress planes common in certain areas is simply handled by a
multiplier term at the present time.

When a well bore is drilling into this block it becomes easier to discuss the stresses in terms of a radial
coordinate system.’ Converting to radial coordinates for stress at the borehole wall the stress relationships
assuming plain strain and elastic behavior become

Uz =PO+2P(U, -02)
cre =3cr1-cr2-Pmud
~r = Proud

DD
,-

Figure 2

Where a, = ~X and 01 = a~ from Figure 1

These stress equations are based on a sharp contrast between mud pressure in the well bore and pore
pressure. In the case of a fluid penetrating into the pore space a gradual change is noted between the mud col-
umn pressure and the pore pressure. Because of this pressure gradient the stress level is a maximum at the
borehole wall and decreases with the distance from the borehole back to the original insitu conditions. The net
effect on our model is an increase in the radial and tangential effective stresses of a magnitude equal to:

a(l -2 P)
(I - L) (Pp - Proud)

The effective rock stress (stress that produces a deformation in the rock skeleton) can be obtained by sub-
tracting the pore pressure component from the radial coordinate stress relationships.

Effective Stress Non-Penetrating fluid

u e .ff =301 -oZ-Pmud-~Pp


,-., U r .ff = Proud - 0!Pp
c z eff =Po+21J(u, -u*) -aPp

-3-
SPWLA TWENTY-SECOND ANNUAL LOGGING SYMPOSIUM, JUNE 23-26, 1981

Effective Stress Penetrating fluid

a(l - 2P)
Ueeff=3 U(. 02- Pmud-O!Pp+
(1 - p ) (Pp - Proud)

a(l -2p)
a rC~~= Proud - ~Pp +
(1 - p ) (Pp - Proud)

a(l -2p)
UZ=PO+2P(UI - u2)-~pP+
(1 - p ) (Pp - Proud)

These then are the necessary stress equations to solve for a stress level surrounding the wellbore. In triax-
ial loading there are three different shearing stresses acting in figure 2. Any combination of two principle
stresses will include a shearing stress in the rock. The maximum shearing force for that place is equal to the
average of the two principle stresses. Shear 1 is from the overburden and tangential, shear 2 is from the over-
burden and radial, and shear 3 is from the radial and tangential stress components. Failures in laboratory
gravel pack testing show the sand arches to be long narrow slot-like openings and recalling that the maximum
shearing stress occurs on those places making a 45 degree angle with the two principle stresses involved, the
radial and tangential stress elements are the key plane stresses for sanding failure.

To help visualize this more clearly the Mohr’s circle technique is employed. The tangential and radial
stresses are plotted on the x axis and a circle drawn through these two points with the center of the circle mid
way between the values on x axis. The radius of the circle is the shear stress and its value can be read on the Y
axis. The top of the circle then represents the largest shear stress.

Consider the following example with Pp = 5000, (a)= 0.6, p =0.25, Po = 10000. The lowest stress state
would be when the mud was slightly greater than the pore pressure (Proud= 5310). At this condition both
tangential and radial stress levels are 2310 psi and no shear stress exists along this axis. Drilling rate should be
the fastest at this mud weight and Mohr’s circle would be a dot. See figure 3

We 2310
z
203 Proud= 5300
Oi+
2310
+

tension I compression

Ce
o-r

Figure 3

As the mud weight increases the tangential stress will decrease and the radial stress will increase. The in-
duced shear stress for a mud weight pressure of 8000 psi is 2690 psi. See figure 4

-4-
SPWLA TWENTY-SECOND ANNUAL LOGGING SYMPOSIUM, JUNE 23-26, 1981

,----
me -380
t
8000

a-r

Figure 4

When the borehole pressure is lowered, the tangential stress begins to increase and the radial stress
decrease. As Mohr’s circle increases in diameter so does the induced shear stress on elements around the
borehole. At a mud column pressure of 3000 psi the shear stress is 2310 psi. See figure 5
DD
,,-,
(Xe 4620

I
P mud= 3000

I
I

‘ix CTe

Figure 5

ULTIMATE STRENGTHS

,“”---
The next important element is to define the tensile strength for brittle rocks and the initial shear strength
for softer rocks. To do this we will start with a uniaxial compressive strength correlation. If we assume that in
most reservoir rocks the rock failure occurs directly after deviation from the straight line elastic portion of the

-5-
SPWLA TWENTY-SECOND ANNUAL LOGGING SYMPOSIUM, JUNE 23-26, 1981

stress strain diagram, the determination of the rock’s compressive strength can be related to Youngs’ ,.-.
Modulus through a correlation developed by Deere and Miller (1966) done on sedimentary rocks. 3 By chang-
ing the modulus ratio (defined as Youngs’ modulus/compressive strength) from the low region for shales, the
medium region for sandstones and high region for limestones and dolomites, the uniaxial compressive
strength in situ can be obtained.

20

.-
10
a /
/
w 6
Ls”
/
/ /
4 /
/
3 // s /’
//
/
2
/
/ S’hale
/

1 //

L 1
I

2 34 S6781’0 20 40 ‘ I00

Uniaxial Compressive
Strength psi 103

Figure 6

SHEAR STRENGTH

The initial shear strength of a rock is an important consideration in predicting rock failure. Observations
of previous methods and a study of rock behavior has lead to the following empirical relationship:

-6-
SPWLA TWENTY-SECOND ANNUAL LOGGING SYMPOSIUM, JUNE 23-26, 1981

0.025 X 10’
,,U% Ti=- X Uniaxial Compressive Strength

TENSILE STRENGTH

At the present time tensile strength is assumed to be somewhere between 0.0 and the magnitude of ~i.

FAILURE THEORIES

Several methods are available to predict failure within the reservoir model. They include the maximum
tensile strength, maximum shear stress, maximum octahedral stess, Coulomb-Navier failure and the Mohr’s
circle. It has been observed that the more brittle materials exhibits a tendency to the tension-compression type
failure while the more ductile materials favor a shear type failure.

The MAXIMUM SHEAR STRESS theory (classical COULOMB failure) predicts that the angle of
failure will be on a 45 degree line to the maximum loading. ”

al - (J3 = 2ts

DD
,....,, In triaxial loading failure occurs when the octahedral stress reaches a maximum value and the material
fails

The critical ‘ro~~must be determined experimentally in a retro-fit situation.

COULOMB-NAVIER suggest that a material fails because of a maximum shear stress that can be
predicted by

for tension 2t = a, [(V2+1)”+V]

for compression 2t = UC [(V2 + I)’A -v]

where v is the coefficient of friction tan 29 =~ and e is the angle of internal friction

MOHR’S ENVELOPE CRITERION

The Mohr’s envelope criterion of failure begins by defining the stress envelope and stating that any stress
point that falls outside this envelope will cause failure. Failures in either tension or shear can be predicted.

-7-
SPWLA TWENTY-SECOND ANNUAL LOGGING SYMPOSIUM, JUNE 23-26, 1981

The shape of Mohr’s failure envelope for unconsolidated rocks can be represented by a straight line of
.,.,,
approximately thirty degrees to the horizontal crossing the tau axis at the residual shear strength and increas-
ing upwards. The thirty degree angle represents the friction angle between the sand grains. In practice this
angle varies only slightly from thirty degrees. The shear strength is initially very low but increases because of
frictional loading as the compressional stress levels are increased.

Consolidated rocks generally have a very high friction angle ( e a) for low stress levels, and a lower fric-
tion angle similar to unconsolidated rocks ( eb) for higher stress levels. This phenomenon can be explained by
the fact that the cementation strengths are not the same. The higher friction angle is the result of rough sur-
faces sliding against one another, and the low friction angle is the result of grain slippage. The following
diagram helps represent the in situ behavior of an average reservoir section:

FAILURE ENVEU)PE

Figure 7

The pore pressure and overburden are generally fixed values over the short term and the only real
variable in the system is the mud weight or well bore pressure. As the well bore pressure increases, the tangen-
tial stress decreases and the radial stress increases, and as the well bore pressure decreases, the tangential stress
increases and the radial stress decreases. As these stress or loading levels change so do the shear stresses ala
Mohr’s circle. With the Mohr’s circle envelope for shear failure established from the preceding statements -
and assumptions, both a maximum and a minimum mud column pressure can be calculated. These mud
weight limitations can be used to predict the maximum drawdown pressure before sanding.

-8-
SPWLA TWENTY-SECOND ANNUAL LOGGING SYMPOSIUM, JUNE 23-26, 1981

Consider the case of figure 8. The mud weight and pore pressure are such that the induced shear strength
,-
is larger than the value allowed by the stress envelope and unless action is taken the failure of the borehole is
eminent. To reduce the stress the mud weight must be increased to reduce the size of Mohr’s circle. Failure of
this type causes hole washouts while drilling and sanding during production.

a-ii

DD
,,
“,.,.

Figure 8

Maximum Tensile Stress failure occurs according to Griffith (1924) when the minor principle stress (ef-
fective tangential stress) is equal to the tensile strength of the rock

oe= - cr~

3 rJ1- cr2-Pmud-cl’Pp=-u~

Solving for the Proud term represents the pressure gradient equation (FPG = Proud/Depth).

Proud =3 ul-uz-O’Pp+ UT
non-penetrating fluid model.
Proud =%, P.+ ~ .Pp + UT
l-p

penetrating fluid model


2P
Proud =(l-P)
—PO+O!PJ::; ’)+ .,

“9-
SPWLA TWENTY-SECOND ANNUAL LOGGING SYMPOSIUM, JUNE 23-26, 1981

Perhaps an even more important parameter for brittle formations is the development of a fracture contain- ..,.,
ment or toughness indicator. Using an energy balance, the area under the stress-strain diagram represents the
energy needed to cause failure assuming failure is reached shortly after the elastic limit. This term is known as
unit strain energy or Modulus of toughness and is calculated by the formula
c (7:
Modulus of toughness= ——

strain strain

DIFFICULT TO EASY TO FRAC


FRAC
Figure 9

where CT* is the effective tangential stress at the original conditions plus the material’s inherent tensile
strength. By adding the area of influence to the stain energy one can obtain a model to predict relative frac-
ture height and length.

This study is in the development stages and is still in the testing phase.

PRACTICAL APPLICATIONS

Example 1

Observe a sand section with the following characteristics:


overburden gradient =0.955
pore pressure gradient =0.55
Poisson’s ratio =0.27
alfa=O.6
mud weight gradient =0.60
depth =6000 ft.

The effective tangential stress on the rock segment is 1133 psi, and the radial stress is 1620 psi.

-1o-
SPWLA TWENTY-SECOND ANNUAL LOGGING SYMPOSIUM, JUNE 23-26, 1981

,
,!”.-.

ORIGINAL CONDITIONS

7
Pmud=3600

Figure 10

In order to induce a tensile failure in the rock, the effective tangential stress must be moved into tension
by an amount equal to or greater than the tensile strength of the rock. This would require that the pressure
gradient in the wellbore (the fracture pressure gradient) be equal to 0.825 psi/ft. This is shown in figure 11.

DD

\
‘hydrostatic = 4950

-200
m m’

TENSILE
FAILURE

#--
Figure 11

-11-
SPWLA TWENTY-SECOND ANNUAL LOGGING SYMPOSIUM, JUNE 23-26, 1981

Example 2

In softer formations failure in shear is the primary method of failure. For the sanding situation, the dif-
ference between the pore pressure and the critical hydrostatic pressure gives us the pressure difference for use
in Darcy hemispherical or radial flow. Non-penetrating fluid models can be used to predict the maximum
drawdown once production has begun.

Consider the sand section of example 1 with the addition of an initial shear strength of 200 psi. The pro-
blem now is one of sanding prediction, or at what drawdown pressure will the induced shear stress be greater
than the shear failure envelope. Drawdown pressure increases (well bore pressure decreases) until the circle
touches the failure envelope, at which time failure by shear (sanding) is predicted.

ce 2200

+
450

at failure
= 2220

.,,.

200 \

Figure 12

,.,

-12-
SPWLA TWENTY-SECOND ANNUAL LOGGING SYMPOSIUM, JUNE 23-26, 1981

Example 3

This example shows a comparison between two wells that seem to have similar elastic parameters and log
values. Under the old system the two wells would be interpreted much the same way, but when the stress
analysis concept is considered the two wells have noticeable differences in stress levels and drawdown limits
due to the differences in pore pressures.

PRESSURE GUAOIENTS ELASTIC PROPERTIES STRESS MAGNITUDE SULK vOLUME ANALYSIS

I PORE PRESSURE 0 0 CB 20 2000


______ :j~sl___~cQo-

~
I HYDROSTATIC o
———— —————.
HEAD AT FAILURE
WELL A

DD

Figure 13

-13-
SPWLA TWENTY-SECOND ANNUAL LOGGING SYMPOSIUM, JUN E 23-26, 1981

,...
Wel! A is over pressured (gradient 0.75 psi/ft), G/Cb is 0.41 and u =0.31. Well B is normally pressured
(gradient 0.45 psi/ft), G/Cb is .26 and u =0.32. By old theories both wells should make sand but stress
analysis suggests that well A can tolerate a 175 psi drawdown while well B is very weak with a -360 psi
drawdown. Sanding is eminent in well B, while with care, well A can be produced without sand.

mud i = 0.80 mud I = 0.55


failure hydrostatic 2 = 0.72 failure hydrostatic 2 = 0.51

Oe, =600 Uet= 2000


CTe2 = 16C0 002= 2200

WELL A WELL B

(0.75 - 0.72) depth = 175psi (0.45 -0.51 )depth= -360 psi

Figure 14

-14-
SPWLA TWENTY-SECOND ANNUAL LOGGING SYMPOSIUM, JUNE 23-26, 1981

Example 4

Example 4 is a case of a reservoir that demonstrates a high G/Cb ratio (1.5 + ) and has a safe drawdown
pressure of 2050 psi. However, the stress analysis presented in track III shows the tangential stress to be very
low or slightly negative. This sand could have possibly experienced a failure in tension during the drilling
operation, and if so, this fact may override any drawdown or strength predictions that can be made.

] P sec 50
PRESSURE GRADIENTS ELASTIC PROPERTIES STRESS MAGNITUDE
!jT HYDRO
;ARBON
;ORR

1%===+---=+----’+
F.P. G, Pene!rat,ng
J _______________

1 PORE P14ESSURE
F1.,d o 0
,, ...,, ,,
G/cB
20 2000 0, p., 8000

o 0 CB 20 -2000 C 8!-)s( 8CQ0


——— — — _________
106,s,
I HYDROSTATIC o
———— ——— —_—
HEAD AT FAILURE
-r

13700

13800

Figure 15

-15-
SPWLA TWENTY-SECOND ANNUAL LOGGING SYMPOSIUM, JUN E 23-26, 1981

,“.,, ,

CONCLUSION

Expressed in this paper are some techniques that allow the user to take the main body of mechanical pro-
perties, that is the elastic parameters, and apply it to everyday problems that face us in the search for more
energy sources. While the models are based on idealized simplifications, they are useful to the industry. Field
test results to date have encouraged our interpretation and have shown this approach to be reasonably sound.
One problem we have experienced is that, at present, there is no practical method of measuring downhole
stress levels to help upgrade the models as necessary. Feedback from producing wells is also usually difficult
to obtain because of similiar problems in downhole measurements during failure.

REFERENCES
1. Anderson, R. A., 76-PET-96, Sept., 1976, A Review of the Elastic Solution for Fracture Pressure,
American Society of Mechanical Engineers.

2. Coates, G. R., Denoo, S.A,, July, 8-11, 1980, Log Derived Mechanical Properties and Rock Stress,
Twenty First Annual Logging Symposium Transactions.

3. Deeve, D. U., Miller, R. P., AFWL-TR-67-144, Jan., 1969, Engineering Classification and Index Pro-
perties for Intact Rock, U.S. Air Force Systems Command Air Force Weapons Lab., Kirtland Air
Force Base New Mex., Tech. Rep.

4. Hottman, C. E., Smith, J. H., Purcell, W .R., November, 1979, Relationship Among Earth Stresses,
Pore Pressure, and Drilling Problems Offshore Gulf of Alaska, Journal of Petroleum Technology. ..

5. Lama, R. D., Vutukuri, V. S., 1978, Handbook on Mechanical Properties of Rocks Volume II, Trans
Tech Publications, Clausthal, Germany.

6. Lama, ~. D., Vutukuri, V. S., 1978, Handbook on Mechanical Properties of Rocks Volume III, Trans
Tech Publications, Clausthal, Germany.

7. Lama, R. D., Vutukuri, V. S., 1978, Handbook on Mechanical Properties of Rocks Volume IV, Trans
Tech Publications, Clausthal, Germany.

8. Pickett, G, R., June, 1963, Acoustic Character Logs and Their Applications in Formation Evaluation,
Journal of Petroleum Technology.

9. Risnes, R., BratIi, R. K., Horsrud, P., July, 1980, Sand Stresses Around a Wellbore, Rogalandsfor-
skning, Stavanger Norway.

10. Stein, Nathan, 1975, Mechanical Properties of Friable Sands from Convential Log Data, SPE Paper
number 5500.

11. Vutukuri, V. S., Lama, R. D., Saluja, S.S., 1978, The Handbook on Mechanical properties of Rocks,
Volume I, Trans Tech Publications, Clausthal, Germany.

-16-
SPWLA TWENTY-SECOND ANNUAL LOGGING SYMPOSIUM, JUNE 23-26, 1981

ABOUT THE AUTHORS

COATES DEN00

GEORGE R. COATES is Manager of Product Development, Operations Marketing, for Schlumberger


Well Services. Mr. Coates received his B.S.M.E. at South Dakota State University. In 1964, he joined
Schlumberger as a Field Engineer working in the Wyoming, Montana areas. In 1971, he was transferred to
Houston to join the new Computed Log Interpretation group, moving from there to Interpretation Develop-
ment then to Manager of the Log Interpretation section at Schlumberger-Doll Research Center. In 1976 he
returned to Houston. Mr. Coates is the author of several papers on well-log evaluation and is a member of
SPE of AIME and SPWLA.
DD
STANLEY A. DEN00 joined Schlumberger in 1971 after graduating from the University of Wyoming
with a degree in Mechanical Engineering. Field Engineer assignments were in Montana and New Mexico and
then as a Synergetic Engineer in the New Orleans Computing Center. Promotion to Sales Engineer came in
1975 with assignments in Tulsa and Oklahoma City. Present assignment is in Product Development in
Houston.

- 17 -

You might also like