New Directional Drilling Simulation Tool Reveals Link Between Dynamic Stability and Tool Face Control

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 17

SPE/IADC 163467

New Directional Drilling Simulation Tool Reveals Link between Dynamic


Stability and Tool Face Control
Reed Spencer, Baker Hughes, Jonathan Hanson, Consultant, Olivier Hoffmann, Ajay Kulkarni, Cliff Allison, Baker
Hughes

Copyright 2013, SPE/IADC Drilling Conference and Exhibition

This paper was prepared for presentation at the SPE/IADC Drilling Conference and Exhibition held in Amsterdam, The Netherlands, 5–7 March 2013.

This paper was selected for presentation by an SPE/IADC program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have
not been reviewed by the Society of Petroleum Engineers or the International Association of Drilling Contractors and are subject to correction by the author(s). The material does not
necessarily reflect any position of the Society of Petroleum Engineers or the International Association of Drilling Contractors, its officers, or members. Electronic reproduction, distribution, or
storage of any part of this paper without the written consent of the Society of Petroleum Engineers or the International Association of Drilling Contractors is prohibited. Permission to
reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE/IADC copyright.

Abstract
The development and use of enhanced directional drilling software modeling tools are outlined. A fully dynamic drill bit
software simulator is introduced with the capability of modeling rock removal done by cutters as well as gage pads, bit
blades, and non-cutting elements which can wear and crush the rock. Side-cutting simulations with gage pad/rock interaction
are presented and compared to side-load lab test data at atmospheric pressure. Good agreement between the simulation and
the lab is demonstrated.

Using this dynamic drilling simulation software, a common directional drilling challenge—holding tool face orientation—is
investigated. The case is made that tool face control problems can be caused by dynamically unstable drill bits in addition to
commonly cited causes. The software is used to simulate the dynamic behavior of 21 bits from the field that had been used
on motors with bent AKO. Bits with excellent tool face control in the field are found to be more dynamically stable than bits
with poor tool face control. The link between instability and tool face control problems is shown to be the increased torque
that accompanies dynamic instability, or bit whirl. The relationship is illustrated with two case studies—one in southern
Texas and one in central Oklahoma. Both case studies examine tool face performance in the field using the enhanced
simulation software.

Introduction
With the recent increase in North American shale drilling, operators’ reliance on predictable directional performance from
bit/BHA drilling systems has increased. To drill the curve section in a timely manner and according to the well plan, a
bit/BHA system must be able to maintain tool face control. Because drilling happens deep underground, it can be very
difficult to understand how drill bits and BHAs behave downhole. Simulation is one way that scientists and engineers have
overcome this problem. This paper discusses dynamic drill bit simulation software, how it was developed, and how it was
used to investigate tool face control.

BD3 is a bit dynamics software first introduced in 1995 (Hanson and Hansen, 1995). It was initially developed to mimic the
atmospheric test rig and drilling simulator located at The Woodlands, Texas. The original application of BD3 was to predict
the dynamic stability of PDC bits. Over the years BD3 has proven its worth for this application. It is currently used on a
day-to-day basis to test bit designs and to eliminate those that do not have acceptable dynamic stability characteristics. It has
recently been enhanced to include bit body features such as gage pads, bit blades, ovoids, and other non-cutting elements that
can wear or crush rock in the simulation. The incorporation of these features allows for the simulation of directional drilling
behavior.

Background
BD3 allows for a fully 3-D bottom-hole rock surface with rock removal by PDC cutters. The rock surface is represented by a
set of discrete nodes (Fig. 1) whose position can be modified by the passage of a cutter to mimic rock removal. The rock
surface spatial node density is maintained more or less constant so that information content of this surface is not lost. Forces
on the PDC cutters are computed based on a force model that is applied along the edge and face of the cutter, taking into
account rock strength, chamfer geometry, and the changing effective back rake and side rake along that edge. These forces
2 SPE/IADC 163467

Fig.1—Left, cutters interacting with rock; each rock node is a blue dot. Center, typical view as rendered by the software. Right, top
view of a drilling simulation, PDC cutting structure only.

are summed to yield the net weight on bit (WOB) and torque on bit (TOB).

The first implementation of BD3 allowed for representing the PDC cutting structure only (see Fig. 1). The bit gage pad/rock
interaction was approximated analytically, assuming a “cylinder in a cylinder” approach, and the bit body was not included in
the model. Since that time, 3-D gage pads and bit body have been implemented. These “distributed surface” objects allow
for a penetration-depth-sensitive force model for calculating contact forces on these surfaces and a sliding wear model for
rock removal by these surfaces. Forces can vary across the surface of these objects depending upon the local penetration
depth of the object with the rock. This extension allows BD3 to predict 3-D bit trajectory and hole enlargement due to
applied side loads and dynamic instability.

Implementation of 3-D Gage Pads and Bit Body


The gage pads and the bit body are implemented using a novel approach whereby two 2-D polygons are made to represent
one 3-D object. Fig. 2 shows how a 3-D bit body blade can be represented by two polygons. The blue polygon defines the
extent of the blade in X-Y space and the red polygon defines the extent of the blade in R-Z space. A similar approach is used
for gage pads. A rock node is inside this blade if and only if it is found inside both of these polygons simultaneously—inside
the blue polygon if viewed in the X-Y plane looking down the bit axis from above and inside the red polygon as viewed in
the R-Z plane from the side. There exist fast numerical tests to determine if a point is inside or outside a 2-D polygon. Once
this determination is made, contact area distribution and penetration depth can be made for the gage pad with the rock surface
by scanning all rock nodes on the rock surface. Fig. 2 also shows an example of the degree of geometric complexity for
which this simple procedure can represent. Forces on various elements of the bit (gage pads and PDC cutters) are also
shown. It is emphasized here that the bit can be in any orientation with respect to the rock and is not necessarily always
vertical. For example, the user can specify that the bit be tilted at a prescribed angle relative to vertical in order to more
closely model bit behavior of push-the-bit or point-the-bit systems.

Fig. 2—Left, the blue polygon defines the bit blade as viewed from the top of the bit. The red polygon defines the bit blade as
viewed from the side of the bit. Center, a top view of BD3 simulation with cutters, blades and gage pads. Right, side view of the
same simulaton. Yellow vectors are forces on the individual cutters. Blue vectors are forces on gage pads. The red ball represents
the position of closest approach of the bit with the borehole wall.
SPE/IADC 163467 3

Forces on Distributed Surfaces


CPU-intensive 3-D analysis would be required to fully evaluate the stress distribution across the contact patch of a distributed
surface (such as a gage pad) indenting the rock. This would have to be performed for each time step of the dynamic
simulation, taking an unacceptable amount of time to do the computation. We have therefore simplified this problem by
applying a local uniaxial stress-strain relationship (Jaeger and Cook, 1976) at each rock node in contact with the indenting
surface. The contact stress increases linearly with rock surface displacement. The contact stress and contact force are given
by:

   1 ( E , , R)  ……………………………..………………………………(1)

F   dA ……………………………………………………………………(2)
where

  Contact stress at nodal position


1  function of Young' s modulus (E), Poisson' s ratio ( ) and wellbore radius (R)
  penetration depth into rock
dA  contact area at nodal position

This approach allows proper area weighting when integrating the nodal forces across the instantaneous contact patch to get
the net force on the contacting surface. The direction of the local contact force is normal to the contacting surface and the net
contact force is the vector sum of all of the local contact forces representing the instantaneous contact patch. There is also the
option that if the contact stress exceeds the confined compressive strength of the rock, the rock displaced by the object is
simply removed. The direction and magnitude of the frictional forces are determined locally using the normal force, the
instantaneous velocity of the surface with respect to the rock and some prescribed friction coefficient. The net frictional force
is the vector sum of the local frictional forces.

In the simulation depicted in Fig. 3, the bit is undergoing an applied force of 3000 lb. in the direction of the red ball. The bit
is drilling Sierra White granite. The contact area of the gage pad on the borehole wall is displayed in yellow. The contact
force of the borehole wall on the gage pad is represented with a blue vector. The hole is inclined to a very small degree in
that direction due to the lateral migration of the bit under this load and, as a result, only the upper part of the pads is in
contact with the rock. The image on the right shows the bit after this particular side load test. The paint is worn off the pads
near the upper portion of the pad, showing agreement between the simulation and the side load test (Ernst et al., 2007). There
is field evidence indicating that wear on the upper portion of the gage pads near the shank is greater than elsewhere on the
pads for many directional applications.

Fig. 3—Side load simulation and lab test comparison. Left, gage pads are contacting with the wall on the far side of the hole near
the top of the image. The blue vectors represent the contact forces on the gage pads. Center, the same simulation as is shown to
the left, but gage pads are made invisible. The upper portion of the gage pad is contacting the rock. The contact area is shown in
yellow. Right, a photo of a bit after a side load test. The paint is worn at the upper portion of the gage pad where it made contact
with the borehole wall. The simulation and lab test contact area locations match.
4 SPE/IADC 163467

Rock removal
Analysis of a large number of laboratory side load tests indicates that a conventional sliding wear model (Ludema, 1996) is
adequate to predict the rock removal associated with the lateral migration of a bit under a prescribed side load. The model
implemented in BD3 is a standard sliding wear model:

 u   2 ( , CCS )  L ……………………………………………………………..(3)

where:

 u  incremental depth worn off rock at a specific node in a single time step
 L  distance the object has slid over a rock node while in contact over one time step
 2  function of contact stress ( ) and confined compressive strength (CCS)

κ2 contains constants that are calibrated to laboratory side load drilling tests. The rock node is repositioned in the direction
normal to the rubbing surface by the distance δu as given in the equation above. The contact stress σ and amount the rock is
worn down δu are computed at each time step for every rock node that is in contact with a distributed surface. Lateral
displacement of a bit under a side load will be dominated by the gage pad recess with respect to gage trimmers for hard rock.
Rock removal by wear, as described above, comes into play for softer rock.

Confirmation of BD3 with Lab Drilling Tests


A large number of side load tests were carried out on the atmospheric test rig in several rock types as indicated in Table 1.

Table 1: Side Load Test Rock Information


UCS
Rock Type
[psi]
Bedford Limestone 6,000
Carthage Limestone 12,000
Sierra White Granite 25,000

A constant side load was applied to the drill pipe just above the rock sample as the bit drills into the rock during a side load
test (Ernst et al., 2007). Fig. 4 shows a schematic of the side load test geometry. The effective side load at the bit varies with
drilling depth. The data was corrected for this effect. Many side load simulations were performed that used the recorded
results of the atmospheric test rig side load tests as input. RPM, ROP, calculated bit side load, and calculated bit tilt were
used as input for the simulations. The bit tilt (which is a function of applied side load, the lateral displacement of the drill
stem, and the geometry and materials used in the drill stem) was calculated using a finite element model of the atmospheric
test rig.

Fig. 4—Schematic of the side load test. The WOB or ROP are specified while a side load F is applied. Axial displacement and lateral
displacement are continuously measured.
SPE/IADC 163467 5

Figs. 5 and 6 are two examples of measured and predicted lateral migration of a bit tested in Carthage limestone at a nominal
1000 lbs and 3000 lbs side load, respectively. Note that the offset between the measured and predicted lateral displacement is
just due to the definition of the starting point. Validation of the model is through the slope of these data. The slopes will
match for an accurate prediction. The parameter dL/dZ is the slope of the lateral displacement vs. drilling depth curve and
reflects how much the bit drills laterally with increasing depth.

Fig. 5—Side load test and side load simulation comparison, Carthage limestone, 1000 lbs nominal side load on the drill stem. The
slope (dL/dZ) of the lab test (black) and the simulation (blue) are compared. The effective side load (red) at the bit decreases as
depth increases.

Fig. 6—Side load test and side load simulation comparison, Carthage limestone, 3000 lbs nominal side load on the drill stem. The
slope (dL/dZ) of the lab test (black) and the simulation (blue) are compared. The effective side load (red) at the bit decreases as
depth increases.

The following bar chart (Fig. 7) summarizes some of the results of the side load tests. The group on the left (Drill Bit #1)
represents historic side load analyses in Bedford and Carthage limestone that use an approximation to the gage pad geometry
as this information was not available at the time. The group on the right (Drill Bit #2) shows more recent tests in Carthage
limestone and Sierra White granite where the exact gage pad geometry was used in the lateral displacement prediction.
6 SPE/IADC 163467

0.03

Drill bit #1: Gage length 6"

Legend Carthage Granite


0.02 LAB
BD3

Bedford Carthage
0.01

0
500 1000 2000 500 1000 2000 1000 2000 3000 3000
Side load (lbs)
Fig. 7—Summary bar chart of dL/dZ results for 10 side load test/side load simulation comparisons. Bits with two different gage
lengths were used. Bedford limestone, Carthage limestone and Sierra White Granite were used. Side loads of 500, 1000, 2000, and
3000 lbs were used.

The two drill bits used for these tests have different designs (cutting structure, gage pad geometry) and were run at different
operating parameters. Considering that rock strength in real rocks can always vary from the nominal value used in a
simulation, the software does a good job of predicting the directional behavior of PDC bits.

Bit Dynamic Stability


The dynamic stability of a drill bit is governed almost exclusively by the geometry of its cutting structure. Bit dynamic
stability in this context refers to lateral vibration typically manifested in backward whirl (Brett et al. 1990). Depth limiting
features such as wear knots or ovoids can improve the stability of a bit, but only at the expense of drilling efficiency. The
standard laboratory protocol for stability testing is, under ROP control, to progressively increase the ROP in steps from 3
ft/hr to 96 ft/hr on the atmospheric test rig and evaluate when the bit stabilizes. At low depths of cut, PDC bits are typically
unstable and ultimately stabilize as the depth of cut increases. Determining when a bit begins to drill in a stable manner is
done by continuously tracking the lateral trajectory of the pipe just above the bit using laser ranging devices and, from that
data, determining when the bit begins on-center, smooth drilling. Note that no side load is applied to the bit in these dynamic
stability tests.

Stability simulations using BD3 follow the same protocol. The ROP is progressively increased. Determining when the bit
stabilizes and begins to drill on-center is simply done by analyzing the recorded X, Y and Z coordinates of the bit in three
dimensional space. Fig. 8 shows the predicted bit trajectory going from a dynamically unstable state to a stable state as the
ROP increases. For this particular example, the bit stabilizes at 24 ft/hr. Fig. 9 shows the same simulation, with whirl
velocity, displacement in the X direction, and ROP shown as functions of simulation time.
SPE/IADC 163467 7

Fig. 8—Trajectory (in X-Y-Z space) of a bit during a stability simulation. As the bit drills downward (-Z direction) the ROP is
increasing in steps. The bit whirls until it has drilled approximately 1.5 inches in the negative z direction, then it stabilizes (at ROP =
24 ft/hr).

Fig. 9—Whirl velocity, X displacement, and ROP during this stability simulation are shown vs. time. The bit stabilizes at 24 ft/hr.
This is the same simulation depicted in Fig. 8.

Fig. 10 shows a comparison of BD3 predictions with laboratory results for 39 drilling tests on bit diameters ranging from 6.5
in. to 17.5 in. All testing was at 120 RPM in Carthage limestone at atmospheric conditions.
8 SPE/IADC 163467

160
Legend
Laboratory
ROP (ft/hr) when bit stabilizes

BD3 prediction
120

80

40

14.75
6.125

7.875

7.875

12.25
12.25
12.25
12.25
12.25
12.25
7.875
7.875

7.875

7.875
7.875
7.875
7.875

8.75
8.75
8.75
8.75
8.75
8.75

8.75
8.75

17.5
6.25

8.75
8.5
6.5

8.5

8.5
8.5

8.5
8.5

16
16
16
6

Bit diameter (in)


Fig. 10—Lab and simulation comparison for 39 different bit designs from 11 different bit sizes.

We are confident, based on the results presented in this paper, that BD3 can be used to predict PDC bit stability and lateral
migration when the bit is placed under an applied side load.

Tool Face Control


Directional drilling applications have two main requirements. The directional trajectory plan must be met, and it must be
done in acceptable timeframe. Build up rate (BUR) and rate of penetration (ROP) are the two key metrics used to measure
directional success. Tool face control impacts both of these metrics. An assembly with constantly changing tool face
orientation will not drill in the intended direction. Consequently it will have a different BUR than expected. ROP is also
adversely affected when tool face control is lost because the drilling process must be stopped in order to reorient. It is
common for aggressive bits that drill fast while rotating to have difficulty holding tool face while sliding. In order to
maintain tool face control, a driller will reduce the WOB, thus sacrificing ROP. Difficulty holding tool face has long been
attributed to a change in reactive torque at the bit due to:
 a change in lithology type
 a change in lithology strength
 a sudden increase in depth of cut caused by a sudden forward movement of the drillstring in the wellbore (i.e.
axial stick-slip or weight stacking).
All of these causes can be amplified by the aggressiveness of a drill bit. This paper acknowledges that all of these
phenomena can cause tool face control issues. It also shows evidence for another phenomenon that can interfere with good
tool face orientation—dynamic instability of PDC bits. BD3 simulations of bit stability and BD3 simulations using the new
gage pad rock removal feature are used to make the argument that significant torque increases during unstable drilling will
cause a loss of tool face on a sliding motor with a bent AKO.

Dynamic Instability: A New Cause for Tool Face Control Issues


Lateral bit vibrations have been well documented in the literature. Much of the literature dealing with lateral instability has
focused on PDC cutter damage and drillstring damage (Brett et al., 1989; Warren, et al., 1990; Reckmann et al. 2010). Aside
from damaging tools, lateral bit vibrations can be accompanied by a significant increase in average torque levels, as well as
extreme variations in torque. This phenomenon is seen regularly during drill bit stability tests on the atmospheric test rig.
Fig. 11 shows the lateral vibrations, the torque, and the aggressiveness of a unique stability test in which the bit cycles in and
out of whirl throughout the entire test. Although this cycling is a rare phenomenon, this stability test shows that no matter the
ROP, the torque required to drill can increase significantly when the bit begins to drill in a dynamically unstable manner.
SPE/IADC 163467 9

20000

10000

-10000

-20000

1600

1200

800

400

120

80

40

0 100 200 300


Drilling time [s]

Fig. 11—A stability test in Bedford limestone in which the bit whirled periodically. When the bit whirls and lateral accelerations
nd rd
increase (top), the torque (2 ) and the aggressiveness (MU, 3 ) also increase dramatically. The red vertical lines show alignment of
these phenomena. ROP is plotted last. Note that the increases in vibration, torque, and aggressiveness happened at all ROP steps.

The test shows that bit whirl can cause a significant change in torque and aggressiveness. With regard to sliding with a bent
AKO motor, the question then becomes, how would an increase in torque due to lateral vibrations affect tool face? Tool face
orientation is a function of reactive torque on the drillstring. The drillstring acts like a torsional spring. Any increase in
torque, whether caused by formation changes, a sudden increase in depth of cut due to axial stick-slip, or dynamic instability
10 SPE/IADC 163467

of the bit while drilling will cause an angular deflection of the drillstring. In other words, the drillstring will twist and tool
face will be lost. If this is true, then any bit design feature that improves the dynamic stability of a bit will also improve the
bit’s ability to hold tool face orientation. Roberts (1998) and Barton et al. (2009) agree with this analysis and offer field data
with regard to gage configuration’s effect on stability and tool face control. But perhaps the most important feature that
affects a bit’s tendency to drill in a stable manner is the primary cutting structure—the cutting profile, the location, the
orientation and the size of each cutter on the bit, as well as chamfer size. Using the dynamic simulation software described
previously and drilling performance data, primary cutting structure’s impact on stability is shown to have significant effect on
tool face control in the field.

Field Study
In an effort to determine if cutting structure stability alone is important for tool face control, field observations and dynamic
simulations were used. Field engineers in the U.S. were asked to provide their best and their worst bit designs with regard to
tool face performance in their curve applications. This request yielded 21 bit designs that had been run in the field—12
designs that had excellent tool face control, and 9 designs that exhibited poor tool face control. The hypothesis to be tested
was that bits showing excellent tool face control in the field would have primary cutting structures that stabilize at a lower
ROP than the bits that had poor tool face control. Stability was measured for each bit design using standard stability test
simulations performed with BD3. Although BD3 has the capability to include ovoids, bit blades and other unaggressive
elements, only cutters were analyzed for this field study. Each bit underwent a standard simulation protocol, with constant
RPM and ROP steps at 3, 6, 9, 12, 18, 24, 36, 48, 72, and 96 ft/hr. The ROP step at which each bit stabilized was recorded.
Fig. 12 is a chart showing the number of bit designs that stabilized at a particular ROP step in the BD3 simulation.

ROP at Which Stabilization Occurs


6

5
Number of Bit Designs

3 Excellent Tool Face Control

2 Poor Tool Face Control

0
3 6 9 12 18 24 36 More
ROP [ft/hr]

Fig. 12— The number of bit designs that stabilized at a particular ROP step in the stability simulation. Bits rated as having excellent
tool face control are shown in blue. Bits rated as having poor tool face control are shown in red. E.g. 5 bits rated as excellent
stabilized at 9 ft/hr in the stability simulation.

Table 2 shows median, mean and standard deviation of the ROP steps at which the “excellent” and “poor” bits stabilized.

Table 2: Field Study Results Summary: Tool Face Control vs. Stability

Excellent Tool Face Control Poor Tool Face Control

Median Stabilization ROP [ft/hr] 9 18


Mean Stabilization ROP [ft/hr] 11.8 16.7
Standard Deviation of Stabilization ROP [ft/hr] 5.5 4.0
SPE/IADC 163467 11

Fig. 12 and Table 2, show that the most stable bit designs were more likely to exhibit good tool face control in the field. A
two-sample t-test was performed to establish the statistical significance of this observation. The difference between the
means was statistically significant (p = 0.015 < alpha = 0.05), confirming the hypothesis being tested. It is important to note
that there is overlap in the ROP domain between the bits labeled as “excellent” and “poor”. This is to be expected since there
are multiple independent variables that can affect this tool face control, including rock type, well plan, motor type, and
directional driller skill level. In spite of all these other variables, dynamic stability still stands out as a differentiating
indicator for tool face performance. Two case studies further illustrate the impact of dynamic stability on tool face control as
well as the capability of BD3 to model directional behavior.

Case Study: South Texas, United States


Tool face control was an issue for some curve bits in the Eagle Ford Shale in southern Texas. Using the conclusions from the
field study between bits with excellent tool face control and those with poor tool face control, stability was used as one of the
main criteria for a new design. BD3 stability simulations were used in the design process, with a final design that stabilized
at 6 ft/hr—matching the best of the “excellent” bits (see Fig. 12). Two baseline bits and the new bit design were run by the
same rig for three separate curve runs within a five-mile radius in South Texas. The operator used a similar BHA (which
included an agitation tool) on all three curve sections. The two baseline bits were performing fairly well with regard to tool
face control. Time-based tool face orientation measurements during sliding were compared on all three runs. Fig. 13 shows
a histogram of the tool face measurements and how far they were from their target orientation. The target or “desired” tool
face orientation was taken to be a moving average of the tool face reading over a time period of 5.17 min. It was assumed
that over this time period the directional driller would basically have the BHA oriented in the intended drilling direction.

Histogram: Devation from Desired Tool Face Orientation, South Texas

350

Off-Set Bit 1
300 Off-Set Bit 2
New Bit Design

250

200
Count

150

100

50

0
-100 -80 -60 -40 -20 0 20 40 60 80 100
Devation from Desired Tool Face Orientation [deg]

Fig. 13—Histogram of tool face performance while sliding for the two baseline bits (purple and green) and the new bit design (blue).

A tight distribution indicates that the bit did not stray much from the desired tool face orientation, while a wide distribution
indicates poorer tool face control. Table 3 shows the tool face control summary data. The new bit brought a significant
improvement in tool face control. The tool face reading for the two baseline bits was within +/-10° of the target tool face
orientation for 77-78% of the sliding time. The new bit design’s tool face orientation stayed within +/-10° for 87% of the
sliding time—a significant improvement over the baseline bits. The standard deviation is another measure used to
characterize the tightness of the distribution. A lower value signifies better performance. The new bit design had the lowest
12 SPE/IADC 163467

standard deviation with regard to deviation from the desired tool face orientation.

Table 3: Tool Face Performance Comparison: South Texas Case Study


Baseline Bit 1 Baseline Bit 2 New Bit Design

Stabilization ROP in BD3 Simulation [ft/hr] 18 12 6

% of Sliding Time Spent Between +/- 10° of Desired Tool Face Orientation 78% 77% 87%

St. Dev. of the Deviation from Desired Tool Face Orientation [°] 36° 31° 25°

Instantaneous Sliding ROP [ft/hr] 37 36 36.5

This new bit also stabilizes at a lower ROP in the standard stability simulation than the two baseline bits, which stabilize at
18 and 12 ft/hr. The average sliding ROPs for each bit were about equal, but this does not include time spent stopping and
reorienting. The bit has now been run on multiple steerable motor, curve applications in the Eagle Ford Shale, and
directional drillers report that it holds tool face very well.

Case Study: Oklahoma, United States


Engineers were having difficulty holding tool face orientation for a hard rock, steerable motor, curve application in
Oklahoma, in the United States. Typical ROPs were 12 ft/hr. The rock was estimated to have a confined compressive
strength of up to 40,000 psi. Because this was a 100% slide application, those bits that could not hold tool face fell behind
the well plan and were pulled. The ultimate goal was to design a bit that could drill at higher ROPs and hold tool face well
enough to complete the curve in one run for 20% of all runs. Up to that point, no one-bit curve runs had been achieved by the
service company.

To increase the ROP, a more efficient cutting structure was one of the goals of the project. To achieve this, a design change
in the profile was made. Two bits were designed such that only the profiles differed. The profile that was already being used
in the application was designated as Profile A. Profile B was a shorter profile, with a change in the radial distance of the
nose. Back rakes, cutter size and chamfer size were the same on both bits. The radial and angular cutter locations were also
kept as similar as the profile change would allow. It has been observed that shorter profiles can be less aggressive, so both
efficiency and aggressiveness were to be measured. A series of simulations using BD3 was performed to compare the two
designs. The first set of simulations consisted of the following conditions:
 The ROP was ramped from 0 to 48 ft/hr continuously.
 The rock strength was set to 12,000 psi.
 A side load of 2000 lbs. was applied to mimic the side force caused by the motor with the bent AKO in the
application.
 Cutters and gage pads were included in the simulation. Recall that gage pads are allowed to remove rock via a wear
model.
WOB, torque, axial and lateral displacement with time and a host of other variables were recorded. Side loads often help
stabilize bits. As a result, the both designs drilled stably throughout the simulation. This allowed for a good comparison
between the aggressiveness of the two designs. Fig. 14 shows the relationship between the torque and WOB. There was no
significant difference in aggressiveness between the two designs. It appears the profile differences were too subtle to affect
the aggressiveness. There was also no observed difference in efficiency between the two bits. Fig. 15 shows the lateral
displacement or side cutting vs. depth as the simulation progresses. There was very little difference between the two, with
Profile A drilling 0.37 in. laterally vs. Profile B drilling 0.35 in. laterally during the course of the simulation. Up to this
point, both bits appear quite similar in two areas—aggressiveness and side-cutting—that many consider very important to
directional drilling. The next sets of simulations show where the two bit designs do differ—with regard to dynamic stability.
The second set of simulations followed the standard stability protocol. The simulations had the following conditions:
 The standard stability simulation consisting of ROP steps from 3 to 96 ft/hr.
 The rock strength was set at 12,000 psi.
 No side load was applied.
 Gage pads were not included in the simulation.
A significant difference was seen between the dynamic behaviors of the two designs. Profile A stabilized at 24 ft/hr,
whereas Profile B stabilized at 9 ft/hr (see Fig. 16).
SPE/IADC 163467 13

2000

1800 Profile A
Profile B
1600

1400

1200
Torque [ft-lbs]

1000

800

600

400

200

0
0 2000 4000 6000 8000 10000 12000 14000
WOB [lbs]

Fig. 14—WOB [lbs] vs. Torque [ft-lbs] for the first simulation set. The slope is proportional to aggressiveness. There is no
significant difference in aggressiveness between the two designs.

0.4
Profile A
0.35 Profile B

0.3
Lateral Displacement [in]

0.25

0.2

0.15

0.1

0.05

0
0 -2 -4 -6 -8 -10 -12 -14
Depth [in]
Fig. 15—Directional trajectory plot, Depth [in] vs. Lateral Displacement [in]. The side force pushes the bits laterally, 0.37 in. (Profile
A) and 0.35 in. (Profile B).
14 SPE/IADC 163467

3000
PROFILE A
2000

1000

0
3000
PROFILE B

2000

1000

0
100

80

60 9 ft/hr 24 ft/hr
40

20

0 20 40 60 80 100
Drilling time (s)
Fig. 16—Stability simulation, 12,000 psi rock. Lateral acceleration magnitudes and ROP are plotted vs. time. Profile B stabilizes at a
lower ROP than Profile A. Profile B has lower acceleration levels than Profile A.

A third simulation set was performed. Its purpose was to more accurately mimic the rock strength and ROP seen in the
application and to observe how dynamic stability changed as a function of side load at the bit. Recall that side loads help
stabilize bits. The simulations had the following conditions:
 The ROP was held constant at 10 ft/hr to more closely match the ROP seen in the application.
 The rock strength was set at 38,000 psi to more closely match the hard rock application.
 The side load was ramped from 0 to 2000 lbs continuously throughout the simulation.
 Cutters and gage pads were included in the model.
A composite plot (Fig. 17) shows lateral displacement of the bit, lateral acceleration, torque and side load vs. time for both
simulations. Both bits begin to whirl from the start, but once a relatively small side load of 100 lbs is reached, Profile B goes
into a fairly stable mode of behavior. Accelerations are minimal after this point, but the bit does periodically move around in
the borehole. It is still cutting to the side in the intended direction however. After reaching approximately 1,200 lbs of side
load, Profile B is constantly up against the borehole wall in the direction of the side force. The lateral trajectory of the hole
does not change significantly after this happens. As mentioned previously, for hard rock drilling, the lateral trajectory is
governed by the gage pad/gage trimmer configuration. The side load is increasing steadily, but the trajectory does not
change. This is because the rock is too hard to be worn away by the gage pad contact forces.

In contrast to Profile B, Profile A does not stabilize at a low side load. The hole diameter begins to grow due to gage pad
impact with the rock while whirling. This hole growth can make it even more difficult for the bit to stabilize. The Profile A
simulation was eventually stopped. At that point the side load had reached 800 lbs. Three new Profile A simulations (not
shown in Fig. 17) with the same rock strength and ROP were started simultaneously with an in-gage hole. The side loads for
these tests were held constant at 1000 lbs, 1500 lbs, and 2000 lbs respectively. The simulation with a 2000 lbs side load
stabilized quickly. The other 1000 and 1500 lbs side load simulations were dynamically unstable. To summarize, Profile B
only required 100 lbs of side force to stabilize, while Profile A required somewhere between 1500 and 2000 lbs to stabilize.
SPE/IADC 163467 15

Lateral displacement [in] Hole Growth

Hole
Trajectory
Lateral acceleration [in^2/s]
TOB [ft-lbs]
Sidel load [lbs]

Fig. 17—Progressively increasing side load simulation, constant ROP at 10 ft/hr, 38,000 psi rock strength. Profile A never stabilizes
and has high accelerations and torque as a result. Profile B goes in to a fairly stable state early on as the side load reaches 100 lbs.

The differences in torque between the whirling Profile A and the stable Profile B were quite significant. Because it was
drilling in an unstable manner, the average torque level for Profile A (3,100 ft-lbs) was approximately 25% higher than
Profile B (2,500 ft-lbs), with high frequency torque peaks regularly reaching up to 5,000 ft-lbs. Once Profile B stabilized, its
torque was constant at approximately 2500 ft-lbs.
16 SPE/IADC 163467

Based on all the simulations, it was determined there was a high likelihood that Profile A was very susceptible to entering
into dynamically unstable drilling modes in this hard rock application, even with the side load imposed by the AKO. This
unstable drilling would cause an increase in the average reactive torque on the drillstring which would lead to the loss of tool
face orientation frequently throughout the run. To remedy this, a new bit incorporating Profile B was designed. It
immediately saw success in the hard rock curve application. It completed its very first curve run from start to finish,
achieving what bits with Profile A never had. Directional drillers have consistently reported excellent tool face control with
the bit. To date the design with Profile B has completed the entire curve section for 40% of its runs, doubling the original
goal of 20%.

Other Factors
This paper has focused on the primary cutting structure’s effect on dynamic stability and tool face control, but any feature
that impacts dynamic stability could have a similar impact on tool face control. Thus, it is feasible to consider improving the
dynamic stability and thus tool face performance of a bit via gage pad design (Roberts, 1998), design of non-cutting elements
(ovoid or blade contact with the rock), and BHA designs that ensure side loads are high enough to help bits stabilize.

Conclusions
The dynamic drill bit simulation software BD3 originally modeled cutter-rock interaction. The program has now been
enhanced. It has the capability of representing gage pads, bit blades, and other distributed surfaces that wear and crush the
rock. The software has been calibrated using side load testing, and it is now used to optimize the directional behavior of drill
bits. The software allows versatility and quick turn-around times that the current state-of-the art directional lab test set-ups
do not. For instance, the user can specify bit tilt, rock strength, and side load as well as turn on or turn off bit features such as
ovoids, bit blades, cutter bodies, and gage pads. This allows for investigation into the root causes of various dynamic
dysfunctions. This versatility was used to evaluate multiple bit designs as part of research and design with regard to tool face
control.

The field study of 21 bits that showed that the drill bits with the best tool face control were the most likely to have the most
dynamically stable primary cutting structures. Case studies in Texas and Oklahoma were presented where stability was a
primary design criterion in an effort to achieve better tool face control. Stability simulations with cutters only and side load
simulations with cutters and gage pads were used to confirm the sensitivity of tool face control to dynamic stability. The
stability of new bit designs was ensured with BD3, and tool face performance improved significantly compared to the less
stable designs. All of the evidence presented here points to the fact that a significant portion of tool face control issues can be
solved by designing stability into the cutting structure of the bit. This is because reactive torque can be generated by two
independent phenomena—the bit over-engaging the formation due to axial stick slip, and dynamic instability. It is important
to note that completely eliminating axial stick-slip will not guarantee smooth tool face if an unstable bit is still used. Both
root causes of tool face control problems must be addressed separately in order to optimize directional drilling while sliding.
Special attention should always be paid to bit stability and it should be the starting point when addressing tool face control.

Acknowledgements
The authors would like to express thanks to Baker Hughes Incorporated for permission to publish this paper. We would also
like to thank Lance Endres for providing post-processing software for drilling and simulation data. We also thank Caleb
Rickabaugh and Jun Zeng for applications support.

SI Metric Conversion Factors


ft x 3.048 E − 01 = m
in x 2.54 E + 00 = cm
kip x 4.448 222 E + 03 = N
lbf x 4.448 222 E + 00 = N

References
Barton, S., May, H., Johnson, S. 2009. Gauge, Cutting Structure, Torque Control Components—What Really Counts for Optimal Tool
Face Control With FC Drill Bits. . SPE Drill Eng 25 (2): 293-300. SPE 107289-PA. doi: 10.2118/107289-PA.

Brett, J.F., Warren, T.M. and Behr, S.M. 1990. Bit Whirl—A New Theory of PDC Failure. SPE Drill Eng 5 (4): 275-281. SPE 19571-PA.
doi: 10.2118/19571-PA.

Ernst, S. Pastusek, P.E, Lutes, P. 2007. Effects of RPM and ROP on PDC Bit Steerability. Paper SPE/IADC Drilling Conference,
Amsterdam, Netherlands, 20-22 February. 105594-MS. doi: 10.2118/105594-MS.
SPE/IADC 163467 17

Hanson, J.M. and Hansen, W.R. 1995. Dynamics Modeling of PDC Bits. Paper SPE/IADC 29401 SPE/IADC Drilling Conference,
Amsterdam, Netherlands, 28 February-2 March. doi: 10.2118/29401-MS.

Jaeger, J.C., Cook, N.G.W. 1976. Fundamentals of Rock Mechanics, 2nd Edition. London, Chapman and Hall

Ludema, K. C. 1996. Friction, Wear, Lubrication: A Textbook in Tribology, CRC Press

Roberts, T.S. 1998. Development of a New Concept of Steerable PDC Bit for Directional Drilling. Paper IADC/SPE Drilling Conference,
Dallas, Texas, 3-6 March. 39307-MS. doi: 10.2118/39307-MS.

Reckmann, H., Pushkar, J., Kpetehoto, F., Chadrasekaran, S., Macpherson, J. 2010. MWD Failure Rates Due to Drilling Dynamics.
SPE/IADC Drilling Conference, New Orleans, Louisiana. 127413-MS

Warren, T.M., Brett, J.F., Sinor, L.A. 1990. Development of a Whirl-Resistant Bit. SPE Drill Eng 5 (4): 267-275. SPE 19572-PA.

You might also like