International Journal of Rock Mechanics & Mining Sciences 84 (2016) 1–9

Contents lists available at ScienceDirect

International Journal of
Rock Mechanics & Mining Sciences
journal homepage: www.elsevier.com/locate/ijrmms

Experimental results on the effect of Bit wear on torque response

Hirokazu Karasawa a, Tetsuji Ohno a, Kuniyuki Miyazaki a,n, Akhmadi Eko b
a
Institute for Geo-Resources and Environment, National Institute of Advanced Industrial Science and Technology, Tsukuba, Japan
b
Mitsubishi Materials Corporation, Naka, Japan

art ic l e i nf o a b s t r a c t

Article history: Percussion bits, polycrystalline diamond compact (PDC) bits and roller cone bits are widely used in well
Received 14 July 2015 drilling. The evaluation methods for downhole conditions, such as the wear condition of drill bits and the
Received in revised form in situ rock strength, are important issues in improving the drilling efficiency and reducing the drilling
7 December 2015
cost. In this study, drilling tests were conducted in laboratory using the three types of drill bit and various
Accepted 26 January 2016
types of rock. On the basis of the results, a close relation between the bit wear condition and bit torque
Available online 4 February 2016
was found for each bit.
Keywords: & 2016 Elsevier Ltd. All rights reserved.
Drilling
Percussion bit
Polycrystalline diamond compact (PDC) bit
Roller cone bit
Bit torque
Bit wear

1. Introduction
In the development of resources such as oil and geothermal
energy, it is essential to evaluate the wear of the drill bit and the
in situ rock strength while well drilling. This is because appro-
priate control of the drilling conditions depending on the rock
type, including the decision of when to exchange a worn bit for a
new one, will improve the drilling efficiency and help reduce
problems such as bit failure, stuck pipes and twistoff. Such a well-
controlled drilling operation will reduce the total drilling cost.
The authors have conducted experimental studies on two types
of well drilling method: percussion drilling 1 and rotary drilling.2,3
In the study of percussion drilling,1 the relationship between the
bit wear condition and drilling parameters, such as the penetra-
tion rate, bit weight and bit torque, was investigated on the basis
of the results of drilling tests in which hard and abrasive rock was
used. The results of the percussion drilling tests and their analysis
are reviewed in this study.
Bellin et al. 4 reported that polycrystalline diamond compact
(PDC) bits account for an astounding 65% of the footage drilled in
oil and gas applications in 2010. We conducted rotary drilling tests
using granite and two PDC core bits having different types of PDC
cutter. In the drilling tests, the penetration rate, not the bit weight,
was kept constant. We describe the results of the tests in this
study, particularly focusing on the relationship between the wear
of the PDC cutters and the drilling parameters.
In our previous studies on rotary drilling,2,3 new methods were
proposed for estimating both the rock strength and the tooth wear
condition while drilling using roller cone bits. In the tests, several
roller cone bits having different tooth wear conditions and several
types of rock having different strengths were used. In this study,
we reanalyze the test data for milled tooth roller cone bits 2 on the
basis of the force balance concept proposed by Warren.5 This ap-
proach reveals that the concept is useful for evaluating the tooth
wear condition of milled tooth bits.
As mentioned at the beginning of this section, it is desirable to
evaluate both the bit wear condition and the in situ rock strength
while drilling. This paper focuses on the relationship between the
bit wear condition, which is one of the common concerns in the
drilling industry, and bit torque. We investigate the effect of bit
wear condition on the bit torque for three types of drill bit, i.e.,
percussion bits, PDC bits and roller cone bits, on the basis of our
experimental drilling data.
2. Percussion Bits

n
2.1. Test equipment
Correspondence to: Institute for Geo-Resources and Environment, National
Institute of Advanced Industrial Science and Technology, 1-1-1 Higashi, Tsukuba,
Ibaraki 305-8567, Japan. The drill rig used for all the tests in this study is illustrated in
E-mail address: miyazaki-kuniyuki@aist.go.jp (K. Miyazaki). Fig. 1. This rig allows two types of drilling method, i.e., percussion

http://dx.doi.org/10.1016/j.ijrmms.2016.01.013
1365-1609/& 2016 Elsevier Ltd. All rights reserved.
2 H. Karasawa et al. / International Journal of Rock Mechanics & Mining Sciences 84 (2016) 1–9

Load Cell
Percussion Drill
Thrust Yoke 1
Displacement Meter Hammer
Hydraulic Ram
DC Motor
2 Torque
Tachometer
Meter

Bevel
Gears
5.7 m

Rod
3
4 Bit Rock (50x50x80 cm)

Oil
Unit

(a) Side view (b) Front view

1: Gauge for inlet oil pressure 2: Gauge for outlet oil pressure
3: Flowmeter for oil flow rate 4: Thermometer for oil temperature
Fig. 1. Drill rig used for tests.

drilling and rotary drilling. The size of each rock block used in the Table 1
tests was 50
50
80 cm. In the tests, the bit weight, the bit Mechanical properties of rock used for drilling tests.

torque, the vertical displacement of the thrust yoke, the rotary Test Type Rock Sc (MPa) St (MPa) E (GPa) ν
speed and the flow rate of the drilling fluid were measured. Water
was used as the drilling fluid in the tests. As can be seen in Fig. 1, Percussion Sori Gt (A) 219 10.9 58.4 0.317
Shinkomatsu An (A) 192 - - -
the torque measured by the torque meter includes the friction at Rotary (PDC) Sori Gt (B) 219 11.0 56.3 0.305
the bevel gears. Thus, the bit torque was obtained by subtracting Rotary Oya Tf 14.0 1.54 4.25 0.246
the torque generated by the friction from the measured torque. (Roller Cone) Kimachi Ss 44.9 4.21 7.07 0.256
Sanjome An 118 9.06 16.6 0.208
The penetration rate was obtained by differentiating the vertical Shinkomatsu An (B) 113 7.67 21.6 0.282
displacement of the thrust yoke with respect to time.
The percussion energy transmitted to the bit Wout (kN-m/min) Sc: Uniaxial Compressive Strength, St: Tensile Strength, E: Young’s Modulus, ν:
Poisson’s Ratio
in the percussion drilling tests was calculated from the input en-
ergy to the percussion drill Win (kN-m/min) using a calibration
curve prepared from the results of preliminary tests. The calibra- 2.2. Test procedure
tion curve represents the relationship between Win and Wout. In
The ten blocks of Sori granite (A) were used in percussion
the preliminary tests, the percussion energy Wout was measured
drilling tests on two percussion bits, Bits A and B. In the tests on
by two-point strain measurement 6 using strain gages pasted on
Bits A and B, the bit weight was set at about 6.6 kN and 13.5 kN,
the drill rod above the bit. The input energy Win was calculated
respectively. Up to 16 holes can be drilled in each rock block;
from the inlet and outlet oil pressures and the oil flow rate of the therefore, 8 holes were drilled in each rock block using each bit so
percussion drill measured using transducers 1 to 3 shown in Fig. 1. as to minimize the effect of differences in the rock blocks on test
Three percussion bits of 65 mm diameter with the same spe- results. In the tests, the rotary speed, the inlet oil pressure of the
cifications were used in the percussion drilling tests. The bit has percussion drill and the water flow rate were fixed at 75 rpm,
eight cemented carbide (WC-Co) tips with 10 mm base diameter, 13.5 MPa and 60 L/min, respectively. The corresponding percus-
where five and three tips are set on the gage and face parts of the sion energy Wout was calculated to be approximately 55 kN-m/min
bit body, respectively. from Win using the calibration curve described above. A third
Ten blocks of Sori granite (A) and a block of Shinkomatsu an- percussion bit, Bit C, was used for the drilling tests on Shinko-
desite (A) were used in the drilling tests. The mechanical proper- matsu andesite (A). The drilling conditions were the same as those
ties of Sori granite (A), shown in Table 1, were obtained from in the tests on Bit A.
uniaxial compression and Brazilian tests on four core pieces re-
2.3. Test results and discussion
covered from each rock block previously used in the drilling tests.
The uniaxial compressive strength of Shinkomatsu andesite Fig. 2(a) shows the relationship between the drilled length and
(A) obtained using eight core pieces is also shown in Table 1. the penetration rate obtained from the drilling tests on Bits A
 H. Karasawa et al. / International Journal of Rock Mechanics & Mining Sciences 84 (2016) 1–9
Fig. 2. (a), (b), (c) Test results for percussion bit using Sori granite.
(white circles) and B (white triangles). The results can be ap-
proximated by the straight lines shown in Fig. 2(a), which were
obtained by the least-squares method. The data for drilled lengths
of 0 m to 3.4 m for Bit B, represented by crosses in Fig. 2(a), were
not used in the approximation because they were obtained with
an incorrect bit weight (16.7 kN). The data for drilled lengths of
3.4 m to 6.3 m for Bit B are missing owing to incorrect cable
connections on the A/D converter. The penetration rate of both bits
decreases linearly with increasing drilled length.
The drilling energy per unit time DE (kN-m/min) is given by
DE = Fu + 2πNT + Wout ,
where F is the bit weight (kN), u is the penetration rate (m/min), N
is the rotary speed (rpm), T is the bit torque (kN-m) and Wout is the
percussion energy. The specific energy SE is the energy required to
drill a unit volume of rock (MPa), which is given by
DE Fu + 2πNT + Wout
SE = = ,
1000Su 1000Su
where S is the cross-sectional area of the drilled borehole (m2) and
u is the penetration rate (m/min). A smaller SE suggests that more
energy-efficient drilling is being performed. As shown in Fig. 2(a),
3
SE for Bit B increased from 229 MPa to 496 MPa with increasing
drilled length, which is due to the progress of bit wear. Teale 7
revealed that the minimum value of the SE is roughly correlated
with the uniaxial compressive strength of rock drilled by rotary
and percussive-rotary drilling. Hoberock and Bratcher 8 reported
that the minimum value of the SE approximates the in situ com-
pressive strength of the formation being drilled, based on the
findings by Teale 7 and Pessier and Fear.9 Although considerable
attention is required to account for the effect of bit wear when
putting the concept of SE into practical use, as noted above, the
concept is significant from an engineering viewpoint in that it may
suggest how to evaluate the in situ rock strength while drilling. On
the other hand, a detection method for the bit wear condition still
remains to be established for percussion drilling. Below we ex-
amine drilling parameters related to the bit wear condition during
drilling.
The height of all tips on Bits A and B was measured every
drilled length of about 6 m. To obtain the degree of bit wear, the
height loss ∆h (mm) [ ¼hnew  h] of all tips was calculated, where
hnew (mm) and h (mm) are the heights of the new and worn tips,
respectively. The relationship between the drilled length and the
average height loss ∆h of the five gage tips is shown in Fig. 2(b).
We adopted the height loss of the gage tips as an indicator of the
degree of bit wear because the height losses ∆h of the five gage
tips were almost equal and larger than those of the three face tips.
∆h increases in proportion to the drilled length and is independent
of the bit weight. Note that the bit weight effects the penetration
rate, as shown in Fig. 2(a), whereas it has little effect on the bit
wear.
It is difficult to directly measure the height of the tips on the bit
while drilling in the field. Therefore, drilling parameters that can
be used to detect the bit wear condition were investigated. As
noted above, neither the penetration rate u nor the bit weight F
appears to be suitable for bit wear detection, and thus neither does
the first term on the right side of Eq. (1), Fu. The third term on the
right side of Eq. (1), Wout, remained almost constant in the tests as
shown in Fig. 2(c). Thus, the second term on the right side of Eq.
(1), 2 π NT, was examined. The relationship between the drilled
length and the bit torque T, which is proportional to 2 π NT, is
shown in Fig. 2(c). It can be seen that T increases linearly with the
drilled length, having an almost equal slope in the case of a dif-
ferent bit weight, which also holds true for the height loss ∆h as
shown in Fig. 2(b). This result suggests that the bit torque T and
height loss ∆h, which are both uncontrollable variables, are closely
related. As shown in Fig. 2(c), the bit torque appears to provide
information concerning not only ∆h but also the bit weight F: the
increment in the bit torque ∆T is considered to be related to the
difference in the bit weight between Bits A and B.
The degree of bit wear is closely related to the abrasivity of the
drilled rock as reported by Okubo et al. 10; hence, drilling tests
were also carried out using Bit C and Shinkomatsu andesite (A).
The objective of the tests was to investigate the effect of the rock
abrasivity on both the bit torque T and the height loss ∆h of the
gage tips. Fig. 3(a) shows the relationship between the drilled
(1) length and the average height loss ∆h of the five gage tips in the
drilling tests on Shinkomatsu andesite (A). The height loss ∆h
increases linearly with the drilled length, similarly to in the tests
on Sori granite (A) shown in Fig. 2(b). The penetration rate-drilled
length and bit torque-drilled length plots shown in Fig. 3
(b) indicate the same tendencies as those obtained from the tests
on Sori granite (A). In Fig. 3(b), the data at drilled lengths of about
0.5 m and 6 m were not used in the approximation of the bit
(2)
torque-drilled length relationship because they were obtained
from tests in which the percussion energy Wout was small.
To clarify the effect of the rock abrasivity on the bit torque and
the degree of bit wear, the slopes of the height loss-drilled length
4 H. Karasawa et al. / International Journal of Rock Mechanics & Mining Sciences 84 (2016) 1–9
Fig. 3. (a), (b) Test results for percussion bit using Shinkomatsu andesite.
Table 2
Slope of height loss of gage tips a and slope of bit torque b in Sori granite (A) and
Shinkomatsu andesite (A).
Sori Gt (A) Shinkomatsu An (A)
Slope of height loss (mm/m), a 0.0526 0.0232
Slope of bit torque (kN  m/m), b 0.00392n 0.00181
b/a (kN  m/mm) 0.075 0.078
n
Mean value of Bit A and Bit B.
relationship, a (mm/m), and bit torque-drilled length relationship,
b (kN-m/m), for Sori granite (A) and Shinkomatsu andesite (A) are
shown in Table 2. The value of b/a is almost equal for both types of
rock with different abrasivities, although the values of a and b are
significantly different. This suggests that if the value of b/a for a
bit, which is the rate of increase in the bit torque to increase in ∆h,
is known beforehand by some means, such as preliminary la-
boratory tests, the average height loss of gage tips on the same bit
can be predicted by measuring the bit torque and its rate of in-
crease b while drilling in the field.
3. PDC Bits
3.1. Test equipment
Two PDC core bits, Core Bits A and B, were fabricated to eval-
uate their performance in rotary drilling tests. Their configuration
is illustrated in Fig. 4. Each bit was 66.0 mm in outer diameter and
44.8 mm in inner diameter and had eight PDC cutters of 8.2 mm
diameter and 3.5 mm thickness. Five and three PDC cutters were
brazed on the outer and inner rows, respectively. The back rake
and side rake angles of the PDC cutters on each bit were both set at
10 deg, which was decided on the basis of fundamental test
results.11 A PDC cutter is composed of two layers: a polycrystalline
PDC Cutter PCD Layer
0.7
3.5
WC-Co Layer
φ 8.2 mm
Rotation −10 deg
−10 deg
φ 66.0 mm
φ 44.8
Stabilizer
φ 47.2
φ 63.6
Fig. 4. PDC core bit used for drilling test.
diamond (PCD) layer and a WC-Co layer as shown in the upper
part of Fig. 4. Core Bits A and B have different types of PDC cutter.
The main difference between the two types of PDC cutter is the
structure of the PCD layer: the PCD layer of the PDC cutter on Core
Bit A is composed of diamond particles with a smaller diameter
and a lower cobalt content than that of the PDC cutter on Core Bit
B.
Thirty blocks of Sori granite (B) were used in the durability test
to evaluate the performance of both bits, in which 25 blocks were
used for both Core Bits A and B, where the length drilled by both
bits was almost equal, then five blocks were additionally used for
only Core Bit A. The mechanical properties shown in Table 1 were
obtained from uniaxial compression and Brazilian tests on two
core pieces recovered from each rock block previously used in the
drilling tests.
3.2. Test procedure
The penetration rate, the rotary speed and the water flow rate
were fixed at 7 cm/min, 100 rpm and 40 L/min, respectively. Note
that the penetration rate was kept constant by operating the au-
tomatic control system equipped in the drill rig. In the tests, top-
view photographs of both core bits were taken every drilled length
of 11.2 m from the same distance to record the condition of wear
 H. Karasawa et al. / International Journal of Rock Mechanics & Mining Sciences 84 (2016) 1–9 5

10 30 0.25
Penetration Rate (cm/min)

Sori Gt (B), PDC Core Bit A, 100 rpm Sori Gt (B), PDC Core Bit A, 100 rpm
8 Penetration Rate
0.2

Bit Torque (kN-m)

Bit Weight (kN)
20 30 m
6
0.15
slope = 0.000751
4
Bit Weight
10 0.1
2 slope = 0.00259
(a) 0.05
0 0 (a)
10 30
0
Penetration Rate (cm/min)

Sori Gt (B), PDC Core Bit B, 100 rpm

0.25
8 Penetration Rate

Bit Weight (kN)

0.2

Bit Torque (kN-m)

20 90 m
6
30 m

4 0.15
Bit Weight
2 0.1
(b)
0 0
0 50 100
Drilled Length (m)
Fig. 5. (a), (b) Test results for PDC Core Bit A and PDC Core Bit B.
on the PDC cutters. The maximum flank wear width on each PDC
cutter was calculated using the enlarged photographs and their
enlargement factor.
3.3. Test results and discussion
Fig. 5(a) and (b) show the test results for Core Bits A and B,
respectively. In the case of Core Bit A, the bit weight tends to in-
crease linearly until a drilled length of about 60 m, and then in-
creases at a lower rate up to a drilled length of 157 m. The bit
weight of Core Bit B also increases linearly up to a drilled length of
about 30 m, and then does not change significantly up to a drilled
length of about 100 m, after which it slightly increases. The test for
Core Bit B was stopped at a drilled length of 112 m owing to the
large wear on the PDC cutters.
Fig. 6(a) and (b) show the relationships between the drilled
length and the bit torque for Core Bits A and B, respectively. In the
case of Core Bit A, the data for drilled lengths of 0–30 m and 30–
157 m were linearly approximated by performing separate calcu-
lations by the least-squares method. In the case of Core Bit B, the
data for drilled lengths of 0–30 m, 30–90 m and 90–112 m were
also linearly approximated by performing separate calculations. As
can be seen in both figures, the bit torque also increases with in-
creasing drilled length similarly to the bit weight as shown in
Fig. 5.
The relationships between the drilled length and the bit torque
can be approximated by two or three straight lines as shown in
Fig. 6. The change in the bit torque is thought to be associated with
the wear on the PDC cutters. The relationship between the drilled
length and the maximum flank wear width (average value for
eight PDC cutters mounted on each bit) is shown in Fig. 7. The
maximum flank wear width is indicated in Fig. 7(a). In the case of
Core Bit A, the relationship between the drilled length and the
slope =0.00294
10
slope = 0.000327
slope = 0.00259
0.05
150
(b) Sori Gt (B), PDC Core Bit B, 100 rpm
0
0 50 100 150
Drilled Length (m)
Fig. 6. (a), (b) Bit torque-drilled length plots for PDC Core Bit A and PDC Core Bit B.
average maximum flank wear width is approximated using two
straight lines corresponding to drilled lengths of about 0–40 m
and 40–160 m. In the case of Core Bit B, it is approximated using
three straight lines corresponding to drilled lengths of about 0–
40 m, 40–85 m and 85–112 m. The comparison of Figs. 6 and 7
reveals that the bit torque is very clearly correlated with the
change in the maximum flank wear width. Thus, the bit torque can
be said to be a possible drilling parameter indicating the wear
condition of PDC bits.
From the visual inspection of the photographs, it was clarified
that the wear on the PDC cutters was primarily limited to on the
PCD layers and that the wear was little on the WC-Co layers until a
drilled length of about 40 m. In addition, we found that the wear
on the WC-Co layers tended to start occurring when the drilled
length exceeded about 40 m. It was also noted that the rapid wear
starts from a drilled length of about 85 m for Core Bit B. This is
probably caused by thermal degradation of the PCD layers in Core
Bit B, since their cobalt content is higher than that in Core Bit A as
mentioned above. The thermal degradation of the PCD layers in
Core Bit B might occur as a result of heat generation associated
with the friction between the rock and the cutters, since cobalt
acts as a catalyst in transforming diamond to graphite at a tem-
perature of over about 700 °C under atmospheric conditions.
As can be seen in Figs. 6 and 7, the increasing rate of bit torque
is constant while the bit wear progresses stably. Thus, the mon-
itoring of the bit torque while drilling helps the driller to perceive
the progressing condition of the bit wear, all other things being
equal.
Deciding when to exchange a worn PDC bit for a new one
strongly affects the drilling cost. Waughman et al. 12 showed that
monitoring real-time SE helps to assess the wear condition of a
6 H. Karasawa et al. / International Journal of Rock Mechanics & Mining Sciences 84 (2016) 1–9

5 T7 T4 T0 (NEW)
Max. Flank Wear Width (mm)

Sori Gt (B), PDC Core Bit A, 100 rpm

4 Max. Flank Wear Width

2
Fig. 8. 3-7/8 in. (98.425 mm) milled tooth roller cone bits for medium-hard rock
40 m slope = 0.0311 formations.
1
slope = 0.0124 (a) grading. The two worn bits, T4 and T7, were fabricated by grinding
0 the teeth of a new bit.
5 Several types of rock were used in the tests: Oya tuff, Kimachi
Max. Flank Wear Width (mm)

Sori Gt (B), PDC Core Bit B, sandstone, Sanjome andesite and Shinkomatsu andesite (B). The
100 rpm mechanical properties shown in Table 1 were obtained from uni-
4 slope = 0.116 axial compression and Brazilian tests on about ten core pieces
recovered from each rock block previously used in the drilling
tests.
3
4.2. Test procedure

2 85 m
40 m
1 slope = 0.00407 slope = 0.0292
0
0 50 100
Drilled Length (m)
Fig. 7. (a), (b) Relationship between drilled length and maximum flank wear width
for PDC Core Bit A and PDC Core Bit B.
PDC bit and also to decide when to pull a worn PDC bit. Since the
formation type affects SE, Waughman et al. 12 employed the
technique to check the current formation type with real-time
downhole gamma ray readings. Note that, at the bit life of both
Core Bits A and B, the bit torque reached approximately 0.23 kN-m
as shown in Fig. 6, although the bit wear progressed differently as
shown in Fig. 7. This suggests the possibility that the bit torque at
the bit life largely depends on the bit geometry regardless of the
types of PDC cutters. From that viewpoint, the bit torque has the
potential to be used for making decision when to exchange the
worn bit for a new one. To verify the practicality, the further ex-
perimental research using various types of rock is essential as well
as the field study.
4. Roller cone Bits
4.1. Test equipment
Three types of milled tooth bit for medium-hard rock forma-
tions having different wear reduction ratios of the tooth height TW
were used in the rotary drilling tests, where TW denotes the ratio
of the height loss of a tooth to its initial height. A top view of the
bits is shown in Fig. 8. Each bit is 98.4 mm (3-7/8 in.) in diameter.
One of them is a new bit (TW¼0) and the other two are worn bits
(TW¼4/8 and 7/8), which are hereafter referred to as the T0, T4
and T7 bits, respectively, in accordance with the IADC bit dull
The rotary speed and the water flow rate were fixed at 50 rpm
or 100 rpm and 110 L/min, respectively. Various bit weights were
applied. The penetration rate and bit torque were measured in the
tests.
(b)
4.3. Diagram of Plotted data
150 Fig. 9(a) and (b) show u/N-F/d and u/N-8T/d2 plots, respectively,
obtained from the drilling tests conducted using the T0 bit and
Shinkomatsu andesite (B), where d is the bit diameter (m). The
horizontal axis of Fig. 9(b), 8T/d2, is an abbreviation of 2 π T/ π (d/2)2.
The solid line approximating the data marked by open triangles in
Fig. 9(a) was named the primary performance line, or simply the
performance line, by Bingham.13 The slope of the primary per-
formance line is denoted by aF (m2/kN). The broken line approx-
imating the data marked by black triangles in Fig. 9(a) was named
the secondary performance line by Bingham.14 The former is more
important than the latter in practical drilling; hence, we focus on
the data contained in the primary performance line. The data
expressed by open squares in Fig. 9(b), which correspond to those
marked by open triangles in Fig. 9(a), are approximated by the
solid line with a slope of aT (m2/kN). We hereafter refer to the solid
lines in Fig. 9(a) and (b) as the performance line for the bit weight
and the performance line for the bit torque, respectively. Bingham
15
reported that the performance line for the bit weight is not
affected by the bit size unless the bit type is changed. From this
viewpoint, the u/N-F/d plot has generality.
In the case of rotary drilling using roller cone bits, the per-
cussion energy Wout is 0 and the thrust component Fu in Eq. (2) is
generally much smaller than the rotary component 2 π NT. As the
cross-sectional area of the drilled borehole S is approximately
equal to π (d/2)2, Eq. (2) can be rewritten as
2πNT 1 (8T /d2)
SE = 2
= ,
1000π (d/2) u 1000 (u/N ) (3)
which is equivalent to the ratio of the abscissa to the ordinate
in Fig. 9(b). It is well known that the minimum specific energy (SE)
is roughly correlated with the uniaxial compressive strength of the
drilled rock as mentioned earlier. From this viewpoint, the u/N
-8T/d2 plot has generality as well as the u/N-F/d plot.
 H. Karasawa et al. / International Journal of Rock Mechanics & Mining Sciences 84 (2016) 1–9 7

0.004 0.9
Tooth Bit (T0) 50, 100 rpm
Shinkomatsu An (B) Oya
50 rpm 0.8 Kimachi
0.003 Sanjome
Shinkomatsu (B)
Performance Line
u/N (m)

aF
for Bit Weight 0.7

Ted (-)
0.002
0.6
0.001
(a) 0.5
0 Milled Tooth Bit
0 200 400 600 800 0.4
Fc /d F/d (kN/m) 0 2/8 4/8 6/8 8/8
0.004
TW (-)
Fig. 10. Relationship between TW and Ted for milled bits (T0, T4 and T7).

0.003 Performance Line

aT
for Bit Torque
50 rpm. The straight lines in Fig. 11(a) were obtained by least-
u/N (m)

squares approximation, where the data for the T0 bit shown by

black triangles were not used because of the occurrence of bit
0.002 balling. T/(Fd) decreases with the progress of tooth wear, in
agreement with the results reported in.16 The result for Kimachi
sandstone shown in Fig. 11(b) exhibits the same tendency as that
for Oya tuff. The effect of bit balling on the T/(Fd)- u/(Nd ) plot
0.001 appears to be negligible for the T0 and T4 bits in Fig. 11(b), for
which we cannot provide a reasonable explanation at present. As
reported in an earlier study,2 bit balling causes the u/N-F/d and u/N
(b) -8T/d2 plots to shift downward from the performance lines for the
0 bit weight and bit torque, respectively. Fig. 12 shows the test re-
0 100 200 300 400 500 sults for Shinkomatsu andesite (B). In these tests, bit balling did
8Tc /d 2 8T/d 2 (kN/m) not occur. The test results exhibit the same trend as those for Oya
tuff; however, the effect of tooth wear on the value of T/(Fd) is
Fig. 9. u/N-F/d and u/N-8T/d2 plots for T0 bit and Shinkomatsu andesite (B). small compared with that for Oya tuff. In particular, the values of
T/(Fd) for the T4 and T7 bits are almost the same, which also holds
true for the data for Sanjome andesite. These results indicate that
4.4. Test results and discussion
the T/(Fd)- u/(Nd ) plot is strongly influenced by the tooth wear of
milled tooth bits used for soft rocks.
Fig. 10 shows the relationship between the wear reduction ratio
The results shown in Figs. 10–12 suggest that the Ted-TW plot is
of the tooth height TW and the dimensionless effective bit torque
more useful than the T/(Fd)- u/(Nd ) plot for evaluating the tooth
Ted expressed as
wear condition when drilling harder rocks such as Sanjome an-
8 (T − Tc )/d2 8T / d2 a desite and Shinkomatsu andesite (B). On the other hand, the T/
Ted = = e = F,
(F − Fc )/d Fe/d aT (4) (Fd)- u/(Nd ) plot is more useful than the Ted -TW plot in the case of
drilling softer rocks such as Oya tuff and Kimachi sandstone. Thus,
where Tc is the threshold bit torque (kN-m), Fc is the threshold bit the adoption of both plots is recommended for evaluating the
weight (kN), Te is the effective bit torque (kN-m) [¼ T – Tc] and Fe is tooth wear condition of milled tooth roller cone bits.
the effective bit weight (kN) [¼ F – Fc]. Tc and Fc are obtained from As well known, SE is affected by many factors such as lithology,
the horizontal intercept of the performance lines shown in Fig. 9 bit design, bit wear, hydraulics, bit vibration and so on.9,17 Dupriest
(b) and (a), respectively. Note that the condition of tooth wear of and Koederitz 17 used SE as a trending tool for monitoring the bit
milled tooth bits can be evaluated using Ted regardless of the rock status. In this study, through the quantitative analysis, it has been
type, except for the T0 bit in Oya tuff and Kimachi sandstone. clarified that the dimensionless Ted -TW and T/(Fd)- u/(Nd ) plots
Burgess and Lesso 16 proposed a technique for estimating the are useful for estimating the tooth wear of milled tooth roller cone
condition of tooth wear of milled tooth bits based on the force bits. A useful drilling parameter common to both plots for the
balance concept proposed in.5 In accordance with,16 the test re- evaluation of bit wear is the bit torque. Abbas et al. 18 presented a
sults in this study concerning the milled tooth bits are reanalyzed technique based on combined bit dullness and SE to determine the
using the dimensionless parameters T/(Fd) and u/(Nd ) . Fig. 11 status of the bit, which is available even when the measured tor-
(a) shows the T /(Fd ) − u/(Nd ) plot for Oya tuff at a rotary speed of que is not known.
8 H. Karasawa et al. / International Journal of Rock Mechanics & Mining Sciences 84 (2016) 1–9
0.1
Ooya Tuff, 50rpm
0.08
0.06
0.04
0.02
Bit Balling
0 0
0.1
Kimachi Sandstone, 50 rpm
0.08
0.06
0.04
0.02
Bit Balling
0
0 0.1 0.2 0.3
Fig. 11. (a), (b) Evaluation of torque model for Oya tuff and Kimachi sandstone.
5. Conclusions
Methods of evaluating the tooth wear condition of three types
of bit, i.e., percussion bits, PDC bits and roller cone bits, which are
widely used in well drilling, were investigated. The investigation
revealed that the torque, which is one of the parameters measured
in the drilling operation, is closely related to the wear condition of
drill bits. The findings of this study are summarized as follows:
For the percussion bits, the bit torque and height loss of gage
tips were closely related. The ratio between their increases was
almost equal for two types of rock with different abrasivities.
For the PDC bits, the drilling performance was satisfactorily
evaluated through a rotary drilling test in which the penetration
rate was kept constant. The bit torque changed in agreement with
the behavior of the maximum flank wear width of the PDC cutters
on the core bits. Additionally, the data suggests the possibility that
the bit torque at the bit life largely depends on the bit geometry
regardless of the types of PDC cutters.
For the roller cone bits, dimensionless Ted-TW and T/(Fd)-
u/(Nd ) plots were useful for evaluating the bit wear condition in
harder rock and softer rock, respectively.
Burgess and Lesso 16 developed an interpretation technique for
0.1
(a)
0.08
0.06
0.04
T0 0.02 T0
T4
T4 Shinkomatsu Andesite (B)
T7 50 rpm T7
0 0.1 0.2 0.3 0.4
(b)
Fig. 12. Evaluation of torque model for Shinkomatsu andesite (B).
computing tooth wear, drilling efficiency and the in situ rock
strength. If such information is provided while well drilling, dril-
lers will be able to make a more appropriate judgement on whe-
ther a change in the penetration rate at a constant bit weight
originates from a change in the formation strength or a change in
the bit wear condition. Thereby they can conduct well drilling
appropriately, reducing problems in the borehole, thus increasing
the drilling efficiency and reducing the total drilling cost. The
T0 findings in this study will be useful for establishing or improving
an evaluation method for the drill bit wear condition while well
T4 drilling, such as the interpretation technique developed in.16
T7 Generally, a large amount of data concerning the penetration
rate has been obtained from laboratories and the field, and the
0.4 penetration rate is easy to intuitively understand in practical
drilling. On the other hand, much less data have been obtained
concerning the bit torque than concerning the penetration rate
and bit weight, although the bit torque appears to provide im-
portant information for estimating the bit wear condition as
clarified in this study. A larger amount of data concerning the bit
torque is required to be obtained from laboratories and the field to
understand the bit torque more deeply for progress of drilling
technology.
References
1. Karasawa H, Suzuki K, Takahashi K. Investigation of tip wear on percussion bits.
J Min Mineral Inst Jap. 2012;128:620–625 in Japanese.
2. Karasawa H, Ohno T, Kosugi M, Rowley JC. Methods to estimate the rock
strength and tooth wear while drilling with roller-bits-Part 1: Milled-tooth bits.
J Energy Resour Tech. 2002;124:125–132.
3. Karasawa H, Ohno T, Kosugi M, Rowley JC. Methods to estimate the rock
strength and tooth wear while drilling with roller-bits-Part 2: Insert bits. J
Energy Resour Tech. 2002;124:133–140.
4. Bellin F, Dourfaye A, King W, Thigpen M. The current state of PDC bit tech-
nology. World Oil. 2010;231(No.9) 41, 42, 44, 46.
5. Warren TM. Factors affecting torque for a roller cone bit. J Petrol Tech.
1984;36:1500–1508.
6. Carlsson J, Sundin KG, Lundberg B. A method for determination of in-hole dy-
namic force-penetration data from two-point strain measurement on a per-
cussive drill rod. Int. J. Rock Mech. Min. Sci. Geomech. Abstr.. 1990;27:553–558.
7. Teale R. The concept of specific energy in rock drilling. Int. J. Rock Mech. Min. Sci..
1965;2:57–73.
8. Hoberock LL, Bratcher GJ. A new approach for determining in-situ rock strength
while drilling. J Energy Resour Tech. 1996;118:249–255.
 H. Karasawa et al. / International Journal of Rock Mechanics & Mining Sciences 84 (2016) 1–9
9. Pessier RC, Fear MJ. Quantifying common drilling problems with mechanical
specific energy and a bit-specific coefficient of sliding friction. In: Proceedings
of the 67th Annual Technical Conference and Exposition of the Society of
Petroleum Engineers. Washington DC; 4-7 October 1992. p. 373-388.
10. Okubo S, Ota A, Akiyama M, Fukui K, Nishimatsu Y. Abrasivity of rock and bit
wear in drilling. J Min Miner Inst Jap. 1997;113:325–332 in Japanese.
11. Karasawa H, Misawa S. Development of new PDC bits for drilling of geothermal
wells-Part 1: Laboratory testing. J Energy Resour Tech. 1992;114:323–331.
12. Waughman RJ, Kenner JV, Moore RA. Real-time specific energy monitoring
enhances the understanding of when to pull worn PDC bits. SPE Drilling &
Completion. 2003;18:59–68.
13. Bingham MG. How to interpret drilling in the performance region. Oil & Gas J.
1964;62(No. 46) 173, 76, 79.
14. Bingham MG. How to locate performance line when drilling response is low. Oil
9
& Gas J. 1964;62(No. 48):78–79.
15. Bingham MG. What you should know about bit design and performance. Oil &
Gas J. 1964;62(No. 47) 77, 78, 80, 82, 85, 88, 89.
16. Burgess TM, Lesso Jr WG. Measuring the wear of milled tooth bits using MWD
torque and weight-on-bit. In: Proceedings of the SPE/IADC 1985 Drilling Con-
ference, New Orleans; 6-8 March 1985. p. 453-462.
17. Dupriest FE, Koederitz WL. Maximizing drill rates with real-time surveillance of
mechanical specific energy. In: Proceedings of the SPE/IADC Drilling Con-
ference. Amsterdam; 23-25 February 2005. p. 1-10.
18. Abbas RK, Instantaneous monitoring of drill bit wear and specific energy as a
criteria for the appropriate time for pulling out worn bits. In: Proceedings of the
Annual Caspian Technical Conference and Exhibition of Society of Petroleum
Engineers. Astana; 12-14 November 2014. p. 1-17.