SPE

Society of Petroleum Engineers

SPE 16696

Roller Bit Model With Rock Ductility and Cone Offset

by W.J. Winters, T.M. Warren, and E.C. Onyia, Amoco Production Co.
SPE Members

Copyright 1987, Society of Petroleum Engineers

This paper was prepared for presentation at the 62nd Annual Technical Conference and Exhibition of the Society of Petroleum Engineers held in
Dallas, TX September 27-30, 1987.

This paper was selected for presentation by an SPE Program Committee following review of information contained in an abstract submitted by the
author(s). Contents of the paper, as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the
author(s). The material, as presented, does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Papers
presented at SPE meetings are subject to publication review by Editorial Committees of the Society of Petroleum Engineers. Permission to copy is
restricted to an abstract of not more than 300 words. Illustrations may not .be copied. The abstract should contain conspicuous acknowledgment of
where and by whom the paper is presented. Write Publications Manager, SPE, P.O. Box 833836, Richardson, TX 75083-3836. Telex, 730989 SPEDAL.

ABSTRACT The literature shows that the drilling process

This paper presents a model which relates
roller bit penetration rates to the bit design, the
operating conditions, and the rock mechanics. Rock
ductility is identified as a major influence on bit
performance. Cone offset is recognized as an impor-
tant design feature for drilling ductile rock.
Each roller bit style is characterized via a
laboratory drilling procedure. Mechanical behavior
of the laboratory rocks is characterized via tri-
axial compression tests. The model relates these
factors to predict the drilling response of each bit
under reasonable combinations of operating condit-
ions. Decomposition of the model provides a quanti-
tative analysis of the bit mechanic~.
The roller bit model provides two modes of
field application. First, field data can be inter-
preted to generate a continuous rock strength log.
These strengths correlate closely to the triaxial
compressive strength of the rock at a confining
pressure that is equal to the effective bottomhole
differential pressure. The rock strength log can
then be utilized to predict and interpret roller bit
performance in offset wells.
INTRODUCTION
involves a complex interaction of more than 20 inde-
pendent variables, of which the rock mechanics and
bit design are foremost. Practical methods for
quantifying the kinematics of a particular roller
bit design and describing the resultant bit/rock
interaction ~re desired but elusive. The problem,
in part, is that there is no simple set of dimen-
sions which can be measured on a roller bit to char-
acterize its design, nor are the rock properties
readily measured or controlled. Thus, most pub-
lished drilling models utilize grossly simplified
assumptions about the bit along with empirical rock
strength indicators such as the ~rillability index,
d-exponent, or specific energy. 1 7
The highly empirical bit models and rock drill-
ability approaches are of limited general utility,
although they can be adapted to a given locale
through repeated application. One alternative is to
explicitly model the bit geometry and rock proper-
ties,8-11 but then in practice, the required input
information is rarely available. A practical
balance between extremes is required to develop an
industrially useful, globally applicable drilling
model. The model must contain sufficient detail to
be relevant for operations use, but it cannot
require data that is not normally measured on the
rig or logically deduced from drilling records.

The petroleum indu~try has for over 30 years
devoted considerable research to the understanding
and modeling of roller ~one bit performance. The
continuing objective has been to design, select, and
operate rock bits in a manner that promotes cost-
effective drilling. Accordingly, attention is
focused on predicting and interpreting the rate of
penetration (ROP) of a drill bit. Good predict-
ability aids well planning and optimization while
reliable interpretation leads to meaningful conclu-
sions about the rock.
References and illustrations at end of paper.
The authors have participated in a research
project from which a roller bit model and method for
its use have evolved. 12 17 Originally, the model
utilized a single rock strength parameter which is
closely related to the uniaxial compressive strength
of brittle rocks such as sandstone, limestone, and
dolomite. Much petroleum drilling also occurs in
relatively ductile materials such as shale for which
some bit designs are more effective than others.
The model has thus been exparided to relate the
drilling response of different roller cone bit
styles to the compressive strength and ductility of
most common rocks.

421
2 ROLLER BIT MODEL WITH ROCK DUCTILITY AND CONE OFFSET
ROCK MECHANICS
A. Triaxial Failure Analogy
The triaxial compression test was selected as
the basis for determining the rock mechanical prop-
erties that are correlated in the roller bit model.
This standard test is described in numerous publica-
tions including those cited here. 18 ' 19 The deviator
stress at failure is defined as the compressive
strength of the rock, and the axial strain at
failure is defined as the ductility. The confining
pressure which is applied to the jacketed rock
sample is interpreted as representin~ the bottomhole
differential pressure in a wellbore. 0
The analogy between compressing a test specimen
to failure in a triaxial cell and compressively
loading the bottomhole with the teeth of a roller
cone bit is qualitative. It should be noted that
the strain rates caused by a drill bit are usually
far greater than those applied in a triaxial cell.
The analogy between the confining pressure in a tri-
axial test and the bottomhole differential pressure
in a wellbore is based upon the principle of effec-
tive stress. 21 It has been shown that rock strength
is dependent on the difference between external body
stresses and internal pore stresses, independent of
hydrostatic or terrastatic stresses~ 22
Triaxial rock properties are used in the roller
bit model with the assumption that there is an equi-
valence between the triaxial strength and the
drilling strength of a rock. The rock stresses and
resultant strains caused by an impacting bit tooth
may differ significantly from those in a triaxial
cell, but a roller bit model can in effect be cali-
brated to triaxial properties. The advantage is
that triaxial rock mechanics tests represent a
standardized independent measure of rock strength
whereas most other drilling strength indices are in
themselves a strong function of the bit used to
drill the rock.
B. Rock Behavior
Rock strength and ductility are very sensitive
to pressure - both increase as the confining pres-
sure or borehole pressure is elevated. 23 - 25 Several
examples are shown in Fig. 1. It is useful to
express confining pressure in terms of a normalized
pressure. The normalized pressure is the ratio of
the confining pressure and the unconfined rock
strength.
p design variations are achieved largely through
p c
••••••••••••• ( 1)
n a0
Most rocks undergo two- to threefold strength
increases and three- to sixfold ductility increases
as corifining pressure is elevated from zero to
P = a , as shown in Fig. 2. The greatest pressure
e~fectg on rock properties occur within the normal
range of drilling conditions. An appreciation of
this factor is requited when interpreting drilling
results that are obtained from different rocks at
various borehole pressures in the laboratory or at
different fluid densities and depths in the field.
422
SPE 16696
C. Drilling Effects
The difference between brittle and ductile rock
failure and the relationship of the failure mode to
the crater volume produced by an impacting bit tooth
has been demonstrated in single tooth indentation
studies. 26 Often the rock fails in a brittle state
but pseudo-plastic behavior of cuttings on the
bottomhole produces the equivalent of a ductile
drilling effect. The result is that under compa-
rable conditions, a ductile or pseudo-plastic bit
tooth crater is smaller than a brittle crater which
translates to. a reduction in the drilling ROP.
The ductility effect on ROP may be explained in
terms of rock toughness. 27 Toughness is defined as
the area under the entire stress-strain curve for a
material. This area is equivalent to the strain
energy density or work per unit volume required to
rupture the material. 28 ' 2 9
u
·····~·······(2)
The value U provides an inverse measure of rock
drillability. Given two rocks with identical
strengths but different ductilities, the more duc-
tile rock is the least drillable of the two.
The pseudo-plasticity effect on ROP may be
viewed in terms of strain energy recovery. When a
rock is strained to failure, some of the applied
elastic strain energy is initially stored by the
rock and then released upon rupture. This energy
release at the bott~mhole contribute~ to puttings
removal if, in riombination with the bit hydraulic
energy, it is sufficient to overcome the prevailing
chip hold-down forces. Otherwise, the cuttings
behave pseudo-plastically. The chip hold-down fric-
tion forces are independent of rock strength, so the
strongest and most brittle rocks are least prone to
°
pseudo-plastic behavior. 3 Conversely, the weaker
and more ductile rocks are most likely to exhibit
pseudo-plasticity.
BIT MECHANICS
A. Cutting Structure Geometry
There are hundreds of roller bit designs prima-
rily because each is tailored to a different range
of rock strengths and ductilities. Rock-related
adjustment of cutting structure geometry. Given the
limits of the bit materials and the need for an
acceptable service life, the preferred sharpness,
extension, or bluntness of a cutter is determined by
the rock strength. The preferred combination of
rolling and sliding cutter motion is determined by
the abrasiveness and ductility of the rock.
Due to the cone profile, all commercial roller
cone bits have at least a small sliding component to
the tooth action on the bottomhole. The surface
described by the cutter tips is actually a truncated
composite cone formed by two or more cone segments.
Each segment, as shown in Fig. 3, has an imaginary
SPE 16696 W. J. WINTERS, T. M. WARREN AND E. C. ONYIA
apex that is not coincident with the rotational
center of the bit. Thus the bit teeth, which rotate
concentrically with the bearing journal, slip
somewhat on the rock when spun about the bottom-
hole.31,32
Tooth slippage can be accentuated by altering
the bearing journal angle in both the vertical and
horizontal planes. In practice, the journals are
horizontally skewed from a concentric alignment by
up to 5° The distance that the journal centerlines
deviate from the bit center is defined as cone
offset. The term cone offset is used through the
remainder of this paper in reference to the total
tooth sliding action produced by cone profile and
cone skew effects.
B. Cutting Action
The cutting action of a roller bit tooth may be
envisioned as a two-step process composed
of: (1) indentation and (2) displacement, as shown
in Fig. 4. 33 In Step 1, the tooth penetrates to a
depth where the distributed load on its cross-
section is equal to the local rock strength. In
Step 2, the tooth translates through a spatial
volume and displaces material in its path.
Depending on the cratering mechanism, the tooth dis-
placement may result in significant generation of
fresh cuttings, effective excavation of existing
cuttings, or inefficient gain of crater volume rela-
tive to the abrasion loss of cutter volume.
The detrimental ductility effect on ROP can be
largely overcome with the tooth sliding action pro-
duced by cone offset. 34 Indexing studies have dem-
onstrated the integral relationship between crater
formation and the optimum bit geometry which pro-
duces the re~uired cutting action to achieve a par-
ticular ROP. 5 ' 36 The indexing concept merges rock
and bit mechanics principles to provide a qualita-
tive transition from single cutter analysis to the
realistic case of multiple cratering with full-bore
drill bits. The modeling approach described in this
paper provides a quantitative means to characterize
the bit/rock reaction of commercial roller bit
designs and apply the results to oilwell drilling.
BIT MODELING
A. Development of an Offset/Ductility Term
The previously published roller bit model for
brittle rocks (Eq. 3) is expanded next to include
ductility effects.
1
••••••••••••• ( 3)
R 2
The first term of Eq. 3 describes the rate of crater
volume generation by equally-effective roller bit
tooth impacts on a brittle rock. This expression is
based on numerous single-tooth impact tests and
theoretical models which show that the crater volume
produced by a chisel-shaped tooth is directly pro-
portional to the square of the force applied to the
cutter and inversely proportional to the square of
the rock strength (for brittle rocks). Ductility
423
3
may be added to this expression by utilizing a
theoretical analysis of the stresses and strains
generated when a chisel-shaped cutter penetrates a
plastic material. 37 Outmans concluded that rock
failure beneath a bit tooth is preceded by plastic
flow of the formation, thus producing a volume of
"deformed" material in the vicinity of the bit tooth
that is directly proportional to the square of the
tooth force and inversely proportional to the square
of the rock strength. The volume of failed rock is
determined by the deformed rock volume that is
strained beyond its failure limit. Given the
assumption that the volume of failed rock is
directly proportional to the volume of deformed rock
and inversely proportional to the ductility of the
rock, the first term of Eq. 3 becomes
1
••••••••••••• ( 4)
v.1
Outmans' analysis was limited to describing the
volume of rock fractured by a penetrating tooth
which moves vertically but not horizontally, i.e.,
to the tooth indentation phase but not the tooth
displacement phase of the drilling process. Equa-
tion 4 requires an expression for the ROP due to
cone offset so that it can be generalized to all
roller cone bit styles. The spatial volume dis-
placed by the bit teeth is directly proportional to
the product of the burial depth and the sliding dis-
tance. The sliding distance, Q, is proportional to
the amount of cone offset and the tooth burial depth
is proportional to W/SD 2 • 37 The rock volume that is
removed during tooth displacement is proportional to
the rock ductility, e. Thus, an offset/ductility
term has the form of Eq. 5.
••••••••••••• ( 5)
Equations 4 and 5 are adapted for substitution in
Eq. 3 by replacing the rock index strength with the
triaxial failure strength, a , and defining a bit
•
des1gn ' h '1su11nverse
constant, tt.~' wh1c t 1 y proport1ona
'1
to the cone offset effect. The respective expres-
sions for crater volume due to tooth indentation and
displacement are inserted in Eq. 3, yielding Eq. 6.
2 3 2
1 acr D e +~ + £.... + £e.I:!Q. ••••••••• ( 6a)
R ---;? NWE ND I
m
1 cro [ -
aOD£
-+ ! £....+ £e.I:!Q. ••••••• ( 6b)
R NW w e ] + ND I
m
Equation 6 is the roller bit model with rock duc-
tility and cone offset. This model is analytically
derived and dimensionally correct. Four bit design
constants are used to characterize each roller cone
bit style. These constants are empirically deter-
mined via the laboratory testing and data analysis
procedures described next.
4 ROLLER BIT MODEL WITH ROCK DUCTILITY AND CONE OFFSET
B. Laboratory Tests
Each style of roller cone bit can be tested in
the laboratory in a manner that yields the required
data to solve Eq. 6 for the bit constants a, b, c,
and ~· The authors usually test a bit by measuring
its ROP as the applied weight-on-bit (WOB) is incre-
mented in stepwise fashion. The rotary speed, flow
rate, rock type, and borehole pressure are held con-
stant during each WOB response test. This procedure
is repeated in a minimum of two rock types at two
rotary speeds and flow rates. All tests are con-
ducted at an elevated borehole pressure of about
1200 psi.
The IADC roller bit classification system pro-
vides a convenient method for categorizing bit
designs according to their intended application and
grouping together different manufacturers' products
that share similar traits. 38 The authors have
tested over 30 commercial roller bit styles from
more than 15 IADC categories in order to charac-
terize the wide range of roller bit products that
are utilized in the petroleum industry.
c. Data Analysis
Nonlinear regression analysis of the laboratory
data for each bit style yields values for the bit
constants in Eq. 6. These are considered to be true
design constants which remain invariant over the
normal range of values for the other parameters that
are specified in the model. The laboratory test
results are thus reduced to a table of four coeffi-
cients for each IADC bit category.
An example of the match between laboratory data
and bit model predictions is shown in Fig. 5 for a
7.875 in. IADC Series 4-2-7 soft formation insert
bit. The bit was tested in soft, ductile shale and
hard, brittle limestone at two rotary speeds and two
flow rates over a wide range of WOB levels. This
type of bit design has a high degree of cone offset
and as a result it drills effectively in shale. The
model predictions are based upon Eq. 6, using the
bit constants listed in Table 1 and the rock proper-
ties shown in Fig. 1. The mean square error of the
model for the IADC 4-2-7 bit is 3.1 ft/hr or 0.4 %.
The relationship between rock ductility and
cone offset is depicted in Fig. 6 for an 8.5 in.
IADC 7-3-7 hard formation insert bit. This bit
style has a low degree of cone offset and as a
result it drills slower in soft shale (o=8000 psi)
than in hard limestone (0=14,000 psi). The shale
(e=.80 %) is more ductile than the limestone (e=.30)
thus limiting the crater volume that is generated
during tooth indentation. A drilling model with
compressive strength or a strength-related index as
the only rock property cannot account for this
effect. However, the roller bit model with rock
ductility and cone offset predicts such behavior.
The mean square error of the model for the
IADC 7-3-7 bit is 2.0 ft/hr or 2.0 %.
The match of the drilling model to the labora-
tory data for a 7.875 in. IADC 2-1-1 bit is shown in
Fig. 7. This shows that the model is applicable to
milled tooth bits as well as insert bits. The mean
square error of the model for the IADC 2-1-1 bit is
1.8 ft/hr or 0.5 %.
424
SPE 16696
Once the constants have been determined for a
p~rticular bit, the completed model can be decom-
posed to provide an analysis of the bit mechanics.
Equation 6 contains four terms, each of which
describes a physically significant drilling effect.
The interaction of these effects can be quantified
by mapping the relative contribution of each term to
the total ROP prediction as a parametric function of
variables in the model such as WOB. Equation 7 is
utilized for this purpose. Since the sum of the
four terms in the roller bit model is equal to the
inverse of ROP, Eq. 7 is interpreted as expressing
the relative limitation of each effect to ROP.
indentation ••••••••••••• ( 7a)
offset ••••••••••••• ( 7b)
bR
teeth ••••••••••••• ( 7c)
ND
hydraulics ~ ••••••••••••• ( 7d)
I
m
Results of the model decomposition are shown for the
IADC 2-1-1 and 4-2-7 bits in Fig. 8. Scanning from
left to right, the uppermost curves represent the
primary limitation to ROP. Indentation is the pri-
mary ROP limitation for both bits at low WOB. This
means that the bit~ will respond strongly to
increased indentation (higher WOB) up to the WOB at
which the first intersection with another curve
occurs. The WOB responsiveness of the bits is dam-
pened beyond the intersection point as factors other
than indentation increasingly dominate the drilling
process.
The decomposition method provides a basis for
evaluating a given bit style and for making useful
comparisons of different designs. For example,
Fig. 8 indicates that the IADC 4-2-7 bit is better
designed than the IADC 2-1-1 bit for drilling the
shale. The IADC 4-2-7 intersection point occurs at
both a higher WOB and ROP level, offset is suffi-
cient (does not exceed the indentation limitation)
over the normal 15-30 klb WOB operating range, and
tooth design is the le~st limitini factor. Indenta-
tion and hydraulics are the primary ROP limitations
for drilling with the IADC 4-2-7 bit in shale, both
of which can be addressed through adjustment of
operating parameters (WOB, flow rate, nozzle size,
etc.). In contrast, the IADC 2-1-1 bit is both
tooth-limited (primary limitation) and offset-
limited (exceeds indentation limitation over the
recommended WOB range) for drilling shale. This bit
appears to be more appropriate for drilling a more
brittle (possibly more abrasive) rock at ROP levels
below 20 ft/hr.
SPE 16696 W. J. WINTERS, T. M. WARREN AND E. C. ONYIA 5

FIELD APPLICATION the method of using the roller bit model to estimate
rock strengths from drilling data.
A. Computation of Rock Strength Log
C. Prediction of Bit Performance
Drilling data obtained with a roller cone bit
can be interpreted to produce a rock strength log. An additional effort was made to qualify the
Equation 6 is rearranged to provide a solution for roller bit model in the field by drilling another
rock strength, as shown in Eq. 8. test hole about 10 yards from the DM No. 2 location.
This hole, designated as DM No. 8, was drilled to

I[
1300 ft depth with an entirely different bit size

2
~-
R
£-
D
~
I tt.2
]].5 and design (9.875 in. IADC 1-1-4) that was operated
at higher flow rates, higher rotary speeds, and
lower WOB levels. A rock strength log was computed
W .. m +~
a - -2- 2 from the DM No. 8 data. Figure 13 shows that a
aD ~ D 4a~
close match was obtained between the DM No. 2 and
DM No. 8 rock strength logs. This result, in con-
junction with the demonstrated correlation to tri-
axial rock strengths, provides assura~ce that
••••• 0 0 0 • 0 ••• ( 8)
practical improvements have been gained by devel-
oping the roller bit model with rock ductility and
cone offset. It demonstrates that the roller bit
model can be used to (1) determine true rock
A rock strength log was generated with Eq. 8 from
strengths from field data and (2) predict bit per-
data recorded over a 3000 ft interval in a Rogers
formance from offset well data.
Co. Oklahoma test hole designated as DM (Drilling
Mechanics) No. 2. Three 8.5 in. bits with respec- CONCLUSION
tive IADC codes of 4-3-7, 5-3-7, and 6-2-7 were used
to drill the interval. The bits were first tested
Roller cone bits drill with a rolling and
in the laboratory to determine their design con-
sliding cutter motion on the bottomhole. The
stants. One-foot averages. of WOB, rotary speed,
sliding motion is proportional to the amount of cone
flow rate, mud density, mud viscosity, and ROP were
offset; its effectiveness is proportional to the
continuously recorded during the field test. The
ductility of the rock and the pseudo-plasticity of
data log and resultant rock strength log is shown in
the cuttings. A roller bit model with rock duc-
Fig. 9. The computed bottqmhole rock strengths
tility and cone offset effects has thus been devel-
range from 3000 to 30,000 psi in the Pennsylvanian
oped and validated through a series of laboratory
to Pre-Cambrian geological column.
and field tests.
Usually, in the absence of direct measurements
The bit model is applicable, with the use of
made from core samples, ductility values must be
appropriate bit design constants, to all styles of
estimated in order to solve Eq. 8 for rock strength.
roller cone bits. The bit design constants are det-
Reasonable estimates can be made based on published
ermined through regression analysis of laboratory
literature values for various lithologies. Figures
drilling data. Decomposition of the resultant mo?el
1 and 10-12 of.this paper, plus many of the publica-
provides bit designers with a quantitative analys1s
tions listed in the Refe~ences section, contain duc-
of the bit mechanics.
tility information for typical oilfield r?cks.
Utilitarian applications for Eq. 8 are ga1ned by
Field data obtained with roller cone bits can
specifying representative ductility contrasts
be interpreted to generate a rock strength log. The
between rock types when the exact ductility of each
computed rock strengths correlate closely to the
rock layer is unknown.
measured rock strengths that are obtained from
standard triaxial compression tests. The rock
B. Correlation to Rock Mechanics
strength log can be used in conjunction with the bit
model to predict and interpret the drilling response
The objective of the DM No. 2 field test was to
of roller cone bits in offset wells.
validate the roller bit modeling technique and
method for its use. The computed rock strengths
NOMENCLATURE
from DM No. 2 appear to be reasonable with respe~t
to formation lithologies and related log propert1es
a,b,c bit design constants dimensionless
but an independent measure of rock strength is
D bit diameter inches
required to quantify model accuracy. Ther~fore, 18 I modified jet impact force lbf
vertically-oriented core samples were obta1ned from
~ rotary speed rev/min
an adjacent test hole that was drilled to a depth of p rock confining pressure psi
1900 ft approximately 100 yards from the DM.No. 2 . PC normalized confining pressure dimensionless
location. All 18 rock samples were tested 1n a trl- Rn penetration rate ft/hr
axial load cell to determine their mechanical prop- s rock index strength psi
erties at confining pressures ranging from 100 psi
u strain energy density F/L 2
to 5000 psi. Stress-strain curves and accompanying
vd crater volume from cutter V/t
Mohr's circle diagrams are shown in Figs. 10, 11,
displacement
and 12 for five of the samples that were tested. A
v. crater volume from cutter V/t
close match between measured triaxial rock strengths 1
indentation
and computed drilling strengths was obtained, as w weight-on-bit klb
shown in Fig. 9. This result provides confidence in
~ult rock ductility dimensionless, %

425
6 ROLLER BIT MODEL WITH ROCK DUCTILITY AND CONE OFFSET
tooth sliding distance L 16.
cone offset coefficient 1/L
mud density lb/gal
rock compressive strength psi
atmospheric rock strength psi
mud viscosity cp
REFERENCES
1. Murray, A. s. and Cunningham, R. A., "Effect of
Mud Column Pressure on Drilling Rates," AIME
Trans. 0959), v. 216, 6,1-69.
2. Edwards, J. H., "Engineering Design of Drilling
Operations," Drilling and Production Practices,
API (1964) 39-55.
3. Bingham, M. G., "Rock Drillability: How Rock
Properties are Related to Drilling," OGJ
(December 14, 1964) 94-101.
4. Jorden, J. R. and Shirley, 0. J., "Application
of Drilling Performance Data to Overpressure
Detection," JPT (November 1966) 1387-1394.
5. Young, F. S., "Computerized Drilling Control,"
JPT (April 1969) 483-496.
6. Cunningham, R. A., "An Empirical Approach for
Relating Drilling Parameters," JPT (July 1978)
987-991.
7. Rabia, H., "Specific Energy as a Criterion for
Bit Selection," JPT (July 1985) 1225-1229.
8. Gnirk, P. F. and ·Cheatham, J. B., "A Theoret-
ical Description of Rotary Drilling for Ideal-
ized Down-Hole Bit/Rock Conditions," SPEJ
(December 1969) 443-450. --
9. Eronini, E. I., "Rotary Drill Bit/Rock Model
with Cutter Offset," Journal of Energy
Resources Technology (September 1983) 356-361.
10. Ma, D. and Azar, J. J., "Dynamics of Roller
Cone Bits," Journal of Energy Resources Tech-
nology (December 1985) 543-548.
11. Walker, B. H., et.al., "Roller Bit Penetration
Rate Response as a Function of Rock Properties
and Well Depth," SPE 15620, presented at the
61st Annual Fall Technical Conference, New
Orleans, Louisiana, October 5-8, 1986.
12. Warren, T. M., "Drilling Model for Soft Forma-
tion Bits," JPT (December 1981) 963-970.
13. Warren, T. M. and Winters, W. J., "The Effect
of Nozzle Diameter on Jet Impact for a Tricone
Bit," SPEJ (February 1984) 9-18.
14. Warren, T. M., "Factors Affecting Torque for a
Roller Cone Bit," JPT (September 1984)
1500-1508.
15. Warren, T. M. and Smith, M. B., "Bottomhole
Stress Factors Affecting Drilling Rate at
Depth," JPT (August 1985) 1523-1533.
426
SPE 16696
Warren, T. M., "Penetration Rate Performance of
Roller Cone Bits," SPE Drilling Engineering
(March 1987) 9-18.
17. Onyia, E. c., "Geology Drilling Log--A Computer
Database System for Drilling Simulation," SPE
Drilling Engineering (March 1987) 27-36.
18. Kovari, K., "Suggested Methods for Determining
the Strength of Rock Materials in Triaxial Com-
pression: Revised Version," Int. J. Rock Mech.
and Mining Sci. (December 1983) 285-290.
19. Franklin, J. and Hoek, E., "Developments in
Triaxial Testing Technique," Rock Mechanics
(1970) 223-228.
20. Handin, J., "Discussion--Laboratory Study of
Effect of Overburden, Formation and Mud Column
Pressures on Drilling Rate of Permeable Forma-
tions," AIME Trans. (1959) 67-69.
21. Robinson, L. H., "Effects of Pore and Confining
Pressures on Failure Characteristics of Sedi-
mentary Rock," AIME Trans. (1959) 26..:..32.
22. Nur, A. and Byerlee, J. D., "An Exact Effective
Stress Law for Elastic Deformation of Rocks,"
· J. Geophys. Res. (September 1971) 76, No. 26,
6414-19.
23. Bredthauer, R. 0., "Strength Characteristics of
Rock Samples Under Hydrostatic Pressure," ASME
Trans. (1957) 695-706.
24. Handin, J. and Hager, R. V., "Experimental
Deformation of Sedimentary Rocks Under Confing
Pressure: Tests at High Temperature," AAPG
Bulletin (December 1958) 2892-2934. --
25. Cheatham, J. M. and Gnirk, P. F., "Review of
the Fundamental Aspects of Rock Deformation and
Failure," Rock Mechanics in Oilfield Geology
Drilling and Production (1966) 3-26.
26. Maurer, W. c., "Bit-Tooth Penetration Under
Simulated Borehole Conditions," JPT (December
1965) 127-136.
27. Handin, J. and Hager, R. V., "Experimental
Deformation of Sedimentary Rocks Under Con-
fining Pressure: Tests at Room Temperature on
Dry Samples," AAPG Bulletin (January 1957) 48.
28. Higdon, A., et. al., Mechanics of Materials,
John Wiley & Sons, Inc. (1968) 67-69.
29. Byars, E. F. and Snyder, R. D., Engineering
Mechanics of Deformable Bodies, Intext Educa-
tional Publishers (1975) 380-388.
30. Maurer, W. C., "Shear Failure of Rock Under
Compression," SPEJ (June 1965) 167-176.
31. Bentson, H. G., "Rock-Bit Design, Selection and
Evaluation," presented at the Spring meeting of
the Pacific Coast District, API Division of
Production, Los Angeles (May 1956).
SPE 16696 W. J. WINTERS, T. M. WARREN AND E. C. ONYIA 7

32. Cunningham, R. A., "Use Lithology to Select the

Right Rock Bit," Petroleum Engineer (November
15, 1981)'46-50.

33. Wardlaw, H. W. R., "Optimization of Rotary

Drilling Parameters," PhD Dissertation, Univer-
sity of Texas at Austin (August 1971) 90-91.

34. Dropek, R. K., and Williams, C. R., "A Study to

Determine Roller Cone Offset Effects at Various
Drilling Depths," presented at the ASME Energy
Technology Conference & Exhibition, Houston,
Texas, November S-9, 1978,

35. Gnirk, P. F. and Musselman, J. A., "An Experi-

mental Study of Indexed Drill Bit-Tooth Pene-
trations into Dry Rock Under Confining
Pressure," .JPT (1967) 1225-1233.

36. Musselman, J. A. and Cheatham, J. B., "Plane-

Strain Chip Formation in Carthage Marble,"
Basic and Applied Rock Mechanics - Proceedings
of lOth Symposium on Rock Mechanics (1972)
AIME, New York, 389-408.

37. Outmans, H. D., "The Effect of Some Drilling

Variables on the Instantaneous Rate of Penetra-
tion," AIME Pet. Trans. Reprint Series No. 6:
Drilling, (1963) 102-114.

38. Winters, w. J., et. al., "Application of the

1987 IADC Roller Bit Classification System,"
SPE/IADC 16143, presented at the SPE/IADC
Drilling Conference, March 15-18, 1987, New
Orleans.

TABLE 1

BIT DESIGN CONSTANTS FOR EQUATION 6

BI T C0 NS T ANT S t

IADC
CODE· a b c

2-1-1 .0083 .0248 8.763 .0023

4-2-7 .0101 .0123 1.303 .0020

t referenced to 1200 psi borehole pressure

427
 -40

c;;
~
35
Legend
0 BEDFORD UME
5z
w
3.5
0
·~~_gg_I,!M_E_
~ 30
• CAlOOSA SHALE tiw
x . p ~~9-~~~~~ >
0 25 1!1 !~!!~.~D. __
c;;
en
z
w
w
1:1<
~
~ 20 2.5
en ~
w
>
;;;
8 ..../"
0
en
w w
1:1<
N gend
~ 10 ::1
-< ~

~ ~
1:1< 1.5 ~~~

~~
5 0
z [

0
0 I 2 3 7 I ID
1* 0 0~
I
I L~
I
4
'
CONRNING PRESSURE. 1000 PSI ' ' NORN.ALIZEO CONRNING PRESSURE

~
N
ca

~
gend ~
EDFORDUME

M
1.5~
CARrHAGE UME

=:::~::
.~M.•~~
. .. ··~······B
••a··
>
!:]
Ei::::>
g ...······· 0
fa 5
e
::1 !:1
~
0
./y.--.1!1 .,.,. .,.,. .,.,. . . . .
o.s-r::...···•••.!>:
::..-- "./
""0z
::..--

o+---~~--~--~~~-T--~--~~~~
0 I 2 3 4 5 6 7 I 9 ID 0.5 I L5
CONRNING PRESSURE. 1000 PSI NORMAUZED CONRNING PRESSURE

en Fig. 1-Each rock responds uniquely to changes in pressure. Fig. 2-Strength increases 2-3 x and inductility increases 3-6 x when the pressure
""Q is elevated to the atmospheric strength of the rock.
I'T1

J--..A
a---
0"

"'
0'
 ROLLER CONE BITS
• The Cones Are Not TRUE-CONICAL
• Nor Are They PURE-ROLLING
TRUE-ROLLING BIT

ALINGED

NNER CONE
APEX-OUTER
DOUBLE CONE CONE
• Result: A SCRAPING Component is Added
to the CRUSHING Action

Fig. 3-AII roller cone bits drill with a combination of rolling and sliding cutter motion on the bottomhole.

ROLLER CONE BITS

CHIP GENERATION IS A 2-STEP PROCESS
STEP 1: INDENTATION & FRACTURE

Ductile
crater

¥-
STEP 2: TOOTH DISPLACEMENT

......, _,. _,. >

Volume gain is greater
in ductile rock
"'='L
1ft.,·•
H! a. cone offset H !
RESULTS: ROP a. CRATER VOLUME
w/o cone offset with cone offset
............. ~Cone offset overcomes~ .,.-
~ the ductility effect ~
Ductile vol. much less Ductile vol. approaches
than brittle volume brittle crater volume

Fig. 4-Cone offset reduces the detrimental ductility effect on ROP.

429
 60
100 7.876"1ADC 2-1-1
7.876" IADC 4·2·7
CATOOSA SHALE a: CATOOSA SHALE
1200 PSI BHP
a: 1200 PSI BHP ::c
J: 80 ~ 40
~
~-
~a: 80 a:
30

z z
0 0
20
~
40
fi
~z 20
~
z 10
w w
D.. c..

0
eo 10 20 . 30 40 60 eo
0 10 20 30 40 60
WEIGHT-ON-BIT, 1000 LB
WEIGHT-ON-BIT, 1000 LB

100~--------------------------, 60----------------------------~
7.876" IADC 4·2·7 7.876" IADC 2·1·1
CARTHAGE LIMESTONE a: CARTHAGE LIMESTONE
a: 1200 PSI BHP 120 RPM ::c 1200 PSI BHP
J: 290 GPM 120 RPM
~ 40
~
280 GPM

~-
a:
~
a:
30
z 60 z
0 0
~ fi 20

~
z ~z
w 10
D.. w
c..

10 20 30 40 60 eo 10 20 30 40 60
WEIGHT-ON-BIT, 1000 LB WEIGHT-ON-BIT, 1000 LB

Fig. 5-The roller bit model matches the laboratory data for a high-offset Insert bit Fig. 7-The roller bit model Is also applicable to milled tooth bits.
In both shale and limestone.

60
8.6" IADC 7·3·7
290 GPM 7.876" IADC 2·1·1 120 RPM 290 GPM
1200 PSI BHP CATOOSA SHALE 10·10•10 JETS
a: eo RPM 1200 PSI BHP 9.2 PPG PV•10
J: 40
~
~
a:
30
CARTHAGE
z LIMESTONE
0
~
20
••,~:.:.:-:::::=·'·'·'·:-:::::::.:.:.:.::::::::.:.:.::::::::.:.::=·'·~
~
zw lsgend
10
o INDENTAnON
D..
D QffS_fi _ _
• !YJ!t __ _
o trrP.M.Y.~~-
10 20 30 40 • 60 eo 0 0
WEIGHT-ON-BIT, 1000 LB 20 :10 40
WEIGHT·ON·BIT,1000 LB

60 7.876" IAOC 4·2·7 120 RPM 290 GPM

8.6" IADC 7·3·7 CAlOOSA SHALE 10·10·10 JETS.
a: 280 GPM
1200 PSI BHP
1200 PSI BHP 8.3 PPG PV• 8
::c 100 RPM -FTIHR
~ 40
i
~
~ 30
CARTHAGE
LIMESTONE t5 0.1
.. ~--··-······
a:
z ~
:::0
....--···
0 :I 04
20
fi
a:
Iii
z
w
c..
10
! •.----··
Lsgend
0~
a Qfl$_!! __

0
10 20 30 40 60 eo 0 20
WEIGHT·ON·BIT,1000 LB
30 40

WEIGHT-ON-BIT, 1000 LB

Fig. 6-The roller bit model correctly predicts that a low-offset Insert bit will drill Fig. 8-Bit performance Is analyzed through decomposition of the roller bit model.
slower In the shale than In the limestone.

430
 ROTARY SPEED
RPM
DEPTH WEIGHT -ON-BIT PENETRATION RATE ROCK STRENGTH SONIC VELOCITY RESISTIVITY
1000 LB FT/HR 1000 PSI

z
c(
a:
!D r.. 1')[).

~
c(
(.) 1~0.

·-· - · - - -· · - - - - - - - - ---- ----------

uJ . -·--···-
a:
a..

Fig. 9-Field data are evaluated with the roller bit model to produce a continuous rock strength log that correlates
closely to independently measured triaxial compressive strengths.

431
SPE 16 696
 15,000 I I-- .... - I I 40,0001 I- I I
DEPTli: 445FT
iii UlliDLDGY: SHALE iii
a. LOCATION: ROGERS CD., OK a.
ui gf 30,000
rn T!DOOO
w
T2oooo
~ 10,000 a:
1-
~ ~ 1000 rn IUOO

-' -' 20,000

ui
"'zw
i= ~ 6000
"'zw
i= aooo

~
...ffi
a: <1000
10,000 I DOD

......
w ili
~ 2000
...
a a 4000

1 I I I I I 1 1 1
0
..().01 0.00 - -. - -- - • 2000 .CODD 6000 8000 10000 12000 MODO 16000 18000 20000
(J
PRINCIPAL STRESS, PSI PRINCIPAL STRESS, PSI a

20,000r------.------., 40,000 I ~ I m ouo I ...... 5000

DEPTH: 711 FT DEPlli: 1896 FT
iii UlliDLDGY: SHALE iii UlliDLDGY: DOLOMITE
a. LOCATION: ROGERS CD., OK a. LOCATION: ROGERS CO., OK
gf 15,000 gf 30,000
w w
a:
~ TIOOOO
~ Tzoooo
rn u; ~ 20,000 u;
~ 10,000 Q.I!IOOO
ll.I5DOO
i= ui i=
z
w
~ 6000 z
w ~~
~ 10,000 ~ 10000

...ffi
00 .. aoo
5000
...a ~
~ 2000
...a "'~ ....
0' I I I • I I I I
o' I I r I I I I

-6.0-4.0 -- -- - -- ---
..().010 ..().005 0.000 0.005 0.010 a
PRINCIPAL STRESS, PSI PRINCIPAL STRESS, PSI a
...
c.>
STRAIN

"" Fig. 10-Stress-strain curves and Mohr's circles for typical shales. Fig. 12-Stress-strain curves and Mohr's circles for typical carbonates.

-40

.r:
0,
c
~ 30
e
~
15,000
I ----• '"-""''• nw•n•"
""""J.--tl5000
iii
Ill
-~--
'i
"'"'
j
a:
TIDOOO lila..
Iii ~0
~ 11000 a..o 20
~ 10,000 EO
ui 0~ ~.
i= ffi 6000
ux
til
~
~
'0
.4000
..'!! II
Q
"'ili 2000 0 10
ili 'S
0
0~--~----~----~----~--~ 0
u

~~0~0--~~---;~~~~~~~--~~~~~~~~~~----~_i~~~--~~~
20000
-G.010 -G.005 0. 000 0.005 0.010 0.015
PRINCIPAL STRESS. PSI a
STRAIN

200 300 -400 500 600 700 800 900 1000 1100 1200 1300
,en Fig. 11-Stress-strain curves and Mohr's circles for a typical sandstone. Depth (Ft.)

~
. ,... Fig. 13-Similar rock strength Jogs are computed from field data obtained with entirely different styles of bits •

0"
0'
-..D
0'