You are on page 1of 10

IPTC 13349

Managing Drilling Vibrations Through BHA Design Optimization


J.R. Bailey, SPE, ExxonMobil Upstream Research Company, and S.M. Remmert, SPE, RasGas Company Limited

Copyright 2009, International Petroleum Technology Conference


This paper was prepared for presentation at the International Petroleum Technology Conference held in Doha, Qatar, 79 December 2009.
This paper was selected for presentation by an IPTC Programme 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 International Petroleum Technology Conference and are subject to correction by the author(s). The material, as presented, does not necessarily
reflect any position of the International Petroleum Technology Conference, its officers, or members. Papers presented at IPTC are subject to publication review by Sponsor Society
Committees of IPTC. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the International Petroleum Technology
Conference 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 where and by whom the paper was presented. Write Librarian, IPTC, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax +1-972-952-9435.

Abstract
Significant performance improvement has been achieved by successfully managing drilling vibrations through bottomhole
assembly (BHA) redesign. This effort has resulted in increased footage per day and reduced tool damage. Prior literature has
described improvements in operating practices to manage vibrations(1,2) as a key component of this ROP (rate of penetration)
management process. In a parallel work activity, the redesign efforts have provided additional performance improvements of
approximately 36% in one drilling application. Dynamic modeling of the BHA has identified the key design changes leading
to these improvements. The redesigned BHA has lower calculated vibration indices than the standard BHA.
The BHA design evaluation process uses a frequency-domain lateral dynamic model in both pre-drill forecast and
post-drill hindcast modes. BHA lateral vibrations are characterized such that alternative BHA configurations may be
developed and compared directly with a proposed baseline assembly. In the hindcast mode, the BHA model can be operated
at the recorded WOB and RPM to generate corresponding model results in time or depth, and these values can be compared
to the measured performance data.
In one case study, the redesign of a BHA with downhole motor and roller reamer is described, with corresponding field
data for four original BHAs and four redesigned assemblies. In a second application, model and field drilling results for two
rotary steerable assemblies are compared to evaluate the predictive ability of the model in smaller hole size and with different
BHA types. Finally, the utility of the model to identify preferred rotary speed sweet spots is demonstrated in a motor BHA
operating in larger hole.
Introduction
Two prior publications describe the basic methodology that has been developed to model BHA lateral vibrations. The first
paper(3) provides a general description of the model and presents case studies of four field applications of this model. The
second reference(4) is a study of 13 BHA runs in the same field, for which slightly different BHA designs and operating
parameters were used. The Appendix of the second paper comprises a detailed mathematical description of the basics of this
frequency-domain lateral vibrations model, known as VybsTM. The present paper illustrates the application of these methods
to a new set of BHA design problems in a joint study conducted by RasGas and ExxonMobil.
Briefly, the modeling process begins with an input panel that is populated with mechanical dimensions of the components
of the BHA, usually up to the heavy-weight drillpipe (HWDP), with about the same level of detail as a fishing diagram. It is
important that the positions of the contact point constraints are entered correctly, and that the stiffness and inertial properties
of the assembly are a proper representation of the subject BHA. Then the desired operating parameters for drilling need to be
provided, including the anticipated ranges of bit weight (WOB) and rotation rate (RPM).
The linear modeling process considers a dynamic perturbation about the static state. The model employs two vibration
modes to compare and contrast the response of each candidate BHA design: lateral bending and twirl. In the lateral bending
vibration mode, an identical reference bit side force input is applied to each design, and the magnitudes of the response at
other locations along the BHA are compared. In the twirl mode, an identical mass eccentricity is applied to each model
element to investigate the stability of the BHA to eccentric mass and centrifugal force effects.
Simulation results are plotted for multiple BHA designs simultaneously in 2D or 3D displays of state vectors. Index
values have been designed to summarize dynamic performance and are displayed for selected configurations to immediately
identify operating sweet spots as well as to indicate which design configuration may be preferred. The indices include BHA
Strain Energy and Transmitted Strain Energy to represent the dynamic bending strain in the BHA and HWDP, the
Stabilizer Side Force index to quantify the dynamic wall contact interaction forces, and the End-Point Curvature index to
represent bending at the top of the model in response to excitation at the bit. Since there is not a specific known nodal

IPTC 13349

location in the free pipe above the BHA, a sensitivity analysis is completed where multiple possible node locations are
considered at the top end of the model, and the results are processed to identify average and worst case conditions. For the
flex lateral bending mode, the average value (RMS or root-mean-square) and the maximum value of the indices are
calculated for excitation at the various RPM multiples and end-length conditions, for each RPM and WOB state.
The model has been designed to characterize the vibration tendency of a BHA and to facilitate comparison of design
alternatives using a relatively straight-forward model. As will be shown in this paper, some BHA designs are preferred to
others from a vibrations perspective, and a more comprehensive solution is obtained rather than simply seeking to identify
and avoid the critical speeds of a given design. There are tangible benefits achieved by developing a simplified approach to
BHA dynamics modeling. These include broad applicability, ease of use, minimized computing demand, and rapid
turnaround of new designs based on real time field observations. The model is an engineering tool that supports the decisionmaking process related to BHA design and selection, with an objective to mitigate lateral dynamic vibrations and
BHA-induced stick-slip. The tool also enables the development of best practices for BHA design and thus can be used as an
instructional device. Ultimately, we hope to deploy this model to the engineering desktop.
There are multiple synergies between this work activity and prior efforts to improve drilling performance as reported in
the references.(1,2) The following case studies will show the additional benefits that accrue from modeling BHA designs.
Case 1: Redesign of a 12-1/4-inch Motor BHA
A minimal bend adjustable kickoff (AKO) motor assembly is a standard design for the 12-1/4-inch hole section in nearly
every well that RasGas (operator) drills. The introduction of roller reamers to the BHA design a couple of years ago resulted
in a significant reduction in the need for back-reaming. Rotating off-bottom is both time-consuming and hazardous to
drilling equipment. However, in initial field applications, cracking of the roller reamer body was observed in more than a
dozen runs. In an attempt to address this problem, the roller reamer was relocated from within the BHA to a position above
one or two joints of HWDP. At about the same time, a near-bit stabilizer was removed on a trial basis. This stabilizer had
been used by previous drilling personnel
because it was deemed necessary to hold
Standard BHA
angle in the long high-angle hold sections of
these wells. Fig. 1 illustrates these two BHA
design configurations, labeled Standard
BHA and Redesigned BHA.
These design changes resulted in the
Redesigned BHA
elimination of roller reamer cracking,
enabling routine use of roller reamers in these
long intervals. The redesigned BHA has
Fig. 1. Configurations for the Standard BHA (red) and the Redesigned BHA (blue).
been in steady use for almost two years with
excellent field results.
Field data from four intervals drilled with
the standard design and four wells drilled
with the redesigned BHA show the
performance benefits of the configuration
change. Fig. 2 provides footage per day for
the four standard designs (red, with an
average of 1038 ft/day) and the redesigned
configuration (blue, with an average of 1412
ft/day), representing an average increase of
36%. It must be noted that all 8 runs used an
identical model of fixed cutter drill bit, and
the data is entirely from two similar rigs on
adjacent blocks drilling the same formations.
Sustained performance improvement was
also observed in two more wells that used a
slightly different bit design, shown in purple
triangles in Fig. 2.
Additional data showed improved results
with the Redesigned BHA. As may be seen
in Fig. 3, there were two cracked roller
reamers in the eight wells, and both occurred
in the standard assembly configuration.
Roughly speaking, the probability of cracking
the roller reamer was reduced from 50% to
0% due to the design change. The only BHA

IPTC 13349

that did not make it to TD was the standard configuration, and the average run length for the redesigned configuration was
5225 ft, an increase of 15% over the Standard BHA. The bit grades show a significant reduction in outer gauge bit wear and
damage for the Redesigned BHA, despite the longer run requirements of the second group of wells. The bit grade difference
of 6.3 (standard) versus 3.8 (redesigned) is significant.
Wells 1 and 5 provide a representative contrast between Mechanical Specific Energy (MSE) data for the respective BHA
design configurations, as illustrated in Fig. 4. Experience has shown that it is important to maintain the MSE below 100 ksi,
and that the bit often survives difficult formations when this is accomplished. Continuously high MSE readings over 100 ksi
increase the risk of premature bit failure and may cause a trip before the bit reaches TD, as occurred in well 4.

Fig. 3. Bit grade, roller reamer condition, and bit run results for eight wells in study group.

Upper B Section

Mid-Section, Clean Limestones

Upper C Section

5150

5200

6100

5155

5205

6105

5160

5210

6110

5165

5215

( )

5220

6120

5225

6125

5180

5230

6130

5185

5235

5190

5240

5195

5245

6145

5200

5250

6150

6135

Mechanical Specific Energy (ksi)

Mechanical Specific Energy (ksi)

90

100

80

70

60

50

40

30

20

90

6140

100

80

70

60

50

40

30

20

10

Redesigned BHA

90

100

80

70

60

50

40

30

20

10

Standard BHA

Redesigned BHA

Standard BHA

5175

Redesigned BHA

10

5170

True Vertical Depth

True Vertical Depth

True Vertical Depth

Standard BHA
6115

Mechanical Specific Energy (ksi)

Fig. 4. Comparable wellbore sections for Wells 1 and 5 show MSE reductions for Redesigned BHA.

Having demonstrated above that, in a controlled group of wells, the Redesigned BHA performance was superior to the
Standard BHA, the design differences may be explored using the vibrations model. In particular, there were two changes in
the redesigned configuration. What is the effect of each change individually, and should more controlled drilling tests be
conducted to identify perhaps another configuration that is better than the Redesigned BHA?
To address the question of the relative merits of the two separate changes, two additional configurations were entered into
the vibrations model. As shown in Fig. 5, one BHA was configured with no near-bit stabilizer below the motor, but this
design, shown in yellow, retains the roller reamer at its original position within the drill collars. Another design kept the
near-bit stabilizer but has the roller reamer above two joints of HWDP, illustrated in green. In applications of this model, it is
useful to color-code the various designs and then refer to the BHA simply by color.
Lateral dynamic bending model simulation results for a single operating point, consisting of a specific WOB, RPM, and
excitation mode, are shown in Fig. 6. In this chart, the x-axis is distance to bit, with the bit at the left edge as in Fig. 5. The
1X excitation mode is shown; this mode is present in most high frequency data that we have seen. The topmost set of results
in Fig. 6 corresponds to the lateral displacement. The first spatial derivative is the tilt angle and that is shown next. The
bending moment and finally shear load complete the chart. Results are shown color-coded for each BHA configuration.

Looking more closely at the shear load,


one may see that the perturbation side force
at the bit is identical for all configurations.
Moving away from the bit, the responses are
quite
different,
depending
on
the
configuration. A discontinuity in the shear
load is seen as a vertical segment of the
response at points of stabilizer contact. The
magnitude of the vertical segment represents
the dynamic side force at the stabilizer,
which would correspond to dynamic torque,
BHA whirl-induced stick-slip vibrations, and
stabilizer wear in a drilling BHA.
Fig. 6 is a detailed response at just one
operating condition. To evaluate an overall
response, one needs to evaluate more
operating conditions, such as the entire RPM
range of interest.
Fig. 7 provides a complete response map
for the beam shear load for each
configuration, at a specified bit weight, for
the 1X excitation mode.
Again, the
perturbation at the bit is the same in all
cases, however the magnitudes of the red
and green BHAs show increased
interference at the near-bit stabilizer, and the
response maps for the red and yellow
designs show higher amplitudes at the roller
reamer than the blue and green BHAs. This
amplification of the perturbation at the bit, in
the rotary speed range of interest, is an
indicator of higher vibrations. The flatter
response of the blue BHA is desirable, with
minimal amplification at uphole locations.
In Fig. 7, there is a vector output for each
configuration at each RPM, for a single
mode of excitation. Vibration indices were
conceived to reduce the model results to a
few scalars for each combination of WOB
and RPM. These vibration indices provide a
quantitative basis for comparing the
response of alternative BHA configurations.
Fig. 8 illustrates several vibration indices
that are used in this analysis.
In Fig. 8, the BHA Strain Energy Index
is the dynamic bending strain energy, per
unit length, calculated from the bending
moment shown in Fig. 6. The Transmitted
Strain Energy is calculated the same way, for
the upper BHA which is usually the first few
joints of HWDP. The Stabilizer Side Force
Index is the sum of the dynamic contact
forces, which are the vertical line segments
in the shear load seen in Fig. 6. The EndPoint Curvature Index is a measure of the
system output at the top of the model if one
considers a perturbation at the bit to be the
system input.

IPTC 13349

Standard BHA

No Near-Bit Stab

Move Roller Reamer

Redesigned BHA

Fig. 5. Two intermediate configurations were introduced in the model, No Near-Bit and Move RR.

WOB = 30 kips, 136 RPM, Mode 1X RPM


Displacement

Tilt Angle

Bending Moment
Reference Side Force Applied at Bit

Shear Load

Distance to Bit (ft)


Dynamic shear load discontinuity
at contact points

Blue BHA has


smallest response

Fig. 6. Displacement, tilt, bending moment, and shear load chart at a specific WOB, RPM, and mode.

WOB = 30 klbs
Mode 1X

RPM

DBIT

RPM

Contact Interference
at Roller Reamer

Contact Interference
at Near-Bit Stabilizer

RPM

DBIT

DBIT

RPM

DBIT

Fig. 7. Beam shear load response for all configurations, at 30 klbs bit weight, for the 1X excitation.

IPTC 13349

BHA Strain Energy

Transmitted Strain Energy

Strain energy in the BHA due to


bending deformation
Good correlation to MSE

Lateral bending
excitation at the
bit: NxRPM

Strain energy in the upper part of


the BHA due to bending
deformation

WOB

for N=1,2,3

Stab Sideforces
Summation of dynamic reaction
forces at the stabilizer contacts

Endpoint Curvature
Curvature at the last element of BHA
Good correlation to MSE

Fig. 8. Displacement, tilt, bending moment, and shear load chart at a specific WOB, RPM, and mode.

Transmitted
Strain Energy

BHA Strain Energy

BHA Sideforce

End-Point Curvature

Fig. 9. Vibration indices for the 12-1/4-inch motor BHA redesign analysis.

Measured Depth

A set of lateral bending indices for these


four BHA configurations, two run in the field
and two analyzed only with the model, are
provided in Fig. 9, for 30 klbs WOB. For each
RPM, the RMS average value is calculated
from the model results for each mode of
excitation, for the flex mode lateral bending
excitation. Here, the RMS average has been
calculated over modes 1X to 3X. Lower
vibration indices indicate less vibration and
thus smoother operations.
As seen in Fig. 9, the dynamic vibration
indices are reduced by a factor of two, three, or
more from the red lines to the blue lines,
depending on the specific index and the rotary
speed range to be considered. This result
indicates a significant reduction in the
predicted vibrations for the blue BHA.
From Fig. 9, one may make an assessment
of the approximate value of the individual
design modifications.
The model results
suggest that elimination of the near-bit
stabilizer and moving the roller reamer into the
HWDP provide almost the same amount of
reduction in BHA vibrations, and the results
are cumulative. It now seems unnecessary to
conduct further field trials to evaluate the two
changes independently.
The drilling parameters, measured data, and
the calculated vibration indices may be plotted
in data strips in well log format, referred to as a
log-mode plot. The data for well 1 is shown
in log-mode format in Fig. 10. The red bar in
the middle is a marker that separates run data
on the left from model predictions on the right,
and we know immediately that this is the red
BHA. The MSE exceeds 100 ksi on a regular
basis; the plots for Fig. 4 are taken from this
dataset (and from Fig. 11). The maximum
lateral acceleration is typically about 5 gs.
The log-mode plot for well 5 is provided in
Fig. 11. The data axes have been scaled
identically to Fig. 10 to facilitate comparison.
The values for WOB and RPM are roughly
comparable, but there is a significant reduction
in MSE and LatMax compared to Fig. 10.
The MSE is about 50 ksi, and the maximum
lateral vibrations are typically 2-3 gs. The
Redesigned BHA drilled the interval more
efficiently than the Standard BHA design.
The index tracks to the right of the blue bar
show that, in comparison to Fig. 10, there was
a significant reduction in the flex mode
Bending Strain RMS (1st column), Bending
Strain 1X result (2nd column), and End-Point
Curvature (3rd column). It is interesting to note
that the Bending Strain twirl mode index,
shown in the last column, actually increased.

WOB

RPM

MSE

LatMax

BendStrn Mode1 EndCurv

Twirl

Fig. 10. Log-mode display of data and model results for Standard BHA in Well 1.

This data can be plotted in another way


similar to the vibration index chart. Fig. 12 is
the DVDT (Drilling Vibration Data Test)
plot for well 1. As seen in Fig. 12, four
quadrants are used to display the information.
The LatMax measured data from the
downhole MWD is the data plotted as pink
circles. Curve fits to the data may be plotted
as the dark red line, in this case a linear fit.
Model results are plotted with black and blue
x markers. In the top left quadrant of
Fig. 12, the Bending Strain Energy index is
plotted with the data as a function of RPM.
The black x markers represent the flex
mode RMS values (here, for modes 1X-4X),
and the blue x symbols are for the twirl
mode results. The bottom left chart shows
this data plotted versus WOB.
The right quadrants in Fig. 12 show the
data plotted with model results for a virtual
sensor, wherein the virtual sensor in this
instance is an accelerometer located at
approximately the same location as the sensor
in the tool. The virtual sensor provides
model results at a given point, whereas the
indices measure the system response in a
more global sense. This feature may provide
better comparison with point sensor data.
The black x markers correspond to the
RMS average for the flex lateral bending
mode, and the blue x markers represent the
twirl mode results, as before. The y-axes are
all scaled to a maximum value of 10, the peak
maximum lateral acceleration in well 1, in
gs. In Fig. 12, it is apparent that the line fit
to the data is more closely aligned with the
black flex mode bending strain energy than
with twirl.
Fig. 13 is the DVDT plot for well 5,
formatted with data and scaled identically to
Fig. 12. It is apparent that the maximum
lateral vibrations are considerably lower for
the blue BHA, with few values in excess of
3 gs. In this plot, the line fit to the data more
closely tracks the twirl mode values and has a
different slope from the lateral bending flex
mode.
The closer correspondence of the blue
BHA to the twirl mode is in agreement with
the log-mode comparisons.
The design
change has appeared to result in a conversion
from a lateral bending flex response to a lessdamaging twirl response.
The data supports the design process
whereby decisions are based primarily on the
flex mode values, with twirl used to provide a
secondary assessment for designs with nearly
identical flex mode behavior. This is the first
such case to directly support this design
approach.

IPTC 13349

Measured Depth

WOB

RPM

MSE

LatMax

BendStrn Mode1 EndCurv

Twirl

Fig. 11. Log-mode display of data and model results for Redesigned BHA in Well 5.

Fig. 12. DVDT display of data and model results for Standard BHA in Well 1.

Fig. 13. DVDT display of data and model results for Redesigned BHA in Well 5.

IPTC 13349

Case 2: Comparative Assessment of


Two Rotary Steerable Assemblies
For specialty directional drilling applications,
e.g., longer horizontal departure wells, the
operator will sometimes choose rotary
steerable systems (RSS) for the drilling
assembly. In general, RSS results in the
12-1/4-inch section have been good.
However, results in the 8-1/2-inch section
have been mixed, with some excellent runs
and some other runs in which the assembly
was replaced with a conventional motor BHA
before reaching TD. Two recent RSS runs in
8-1/2-inch holesize are compared in order to
highlight differences in design with respect to
vibration indices and actual field results.
Fig. 14 illustrates two RSS tools, one
shown in orange and the other in purple. The
3D plot of beam shear load versus RPM and
distance to bit for a representative WOB and
an excitation mode of 1X RPM is also
provided. Note that the orange assembly
shows a high degree of amplification of the
response, whereas the purple BHA shows a
subdued, flat response.
The vibration indices are shown in Fig.
15, where the max value (thin colored line) is
plotted in addition to the RMS value (thick
colored line). Here the RMS value is
averaged over all calculated excitation modes
(1X to 3X) and all selected positions of the
top node location.
The vibration indices in Fig. 15 show that
the orange assembly is not expected to drill
as well as the purple BHA. From prior
experience and model calibration, it is known
that BHA designs with End-Point Curvature
index values in the low single digits are likely
to drill well. Efforts to normalize the other
indices are in progress to enable further
understanding of absolute index performance.
A summary of these two runs is provided
in Fig. 16. Corresponding to the model
results, RSS-1 experienced problems drilling
out the shoe, and after making a few hundred
feet of hole with one intermediate trip to
replace MWD components it was laid down.
This hole section was then drilled with a
conventional motor BHA.
On the other hand, RSS-2 completed its
interval of nearly 5000 ft with very low levels
of lateral vibrations, manageable stick-slip,
good directional control, and a sustainable
ROP of 50 ft/hour.
Fig. 17 illustrates the drilling results for
the RSS-1 assembly. MSE values were
rarely below 200 ksi and were typically 300
ksi or more. Stick-slip was also continuously
high for this assembly.

RSS-1

RSS-2

Fig. 14. Two rotary steerable configurations for 8-1/2-inch holesize.

Flex Trans Strain Energy

Flex BHA Strain Energy

purple down here

Flex Stab Sideforce


Flex End-Pt Curvature

Vibration indices for


purple RSS-2 much
lower than those for
the orange RSS-1
Typically, this order of
magnitude difference
indicates very
different performance

Values for End-Point


Curvature of 3-5
Suggest Good
Performance

Fig. 15. Vibration indices for the 8-1/2-inch rotary steerable system analysis.

RSS-1
Experienced very high MSE values during drill out
Both high MSE & stick slip observed while drilling below casing shoe
Lateral vibrations exceeding 4gs with continuous stick-slip
Two runs failed immediately (internal communications failures)
Equipment & maintenance procedures OR vibrations ?

RSS-2
Drilled to section total depth, almost 5000 ft
Mostly low to moderate levels of stick-slip
Minimal lateral vibrations
Good directional control into tight target with ROP of ~ 50 fph
Final bit grade of 1-3-CT-S-X-I-CT-TD
Fig. 16. Summary of results for RSS-1 and RSS-2 runs.

RPM Fluctuations

MSE (ksi)
0

200

400

600

800

1000

13050

1200

50

100

150

200

250

300

350

13050
StickSlip
RPM at Bit

13100

13100

13150

Measured Depth

13150
MSE

13200

13200

13250

13250

13300

13300

13350

High MSE13350
values recorded
during drillout
High MSE13400
and stick-slip
below casing shoe

13400

13450

13450

Fig. 17. Drilling data for RSS-1 shows very high MSE values and continuous stick-slip.

Measured Depth

The RSS-2 model results for the


measured drilling parameters, presented in
the log-mode display format, are provided in
Fig. 18. For the RPM and WOB used to drill
the section, calculated End-Point Curvature
indices were low relative to our experiencebased criteria, and the drilling results were
also good. One may observe a couple of
intervals of stick-slip, but overall there are
relatively low stick-slip vibrations.
Fig. 19 shows the same data in the DVDT
format. In Fig. 19, the vibration indices are
plotted with the stick-slip data measurements
over a 1300-ft depth interval. This interval
includes a stick-slip event towards the end of
the section. Note that in the DVDT display,
the model results are scaled in such a way
that the RMS averages of the two populations
are made equal.
The plot of Fig. 19 has four quadrants.
Measured stick-slip vibration data is shown
as the pink data circles, with the red quadratic
trend curves. The top row of plots has RPM
on the x-axis, and the bottom row is plotted
against WOB, as before. The charts on the
left side show Bending Strain Energy, and the
charts on the right show Transmitted Strain
Energy. Only blue x values are shown for
the twirl results for these indices.
With considerable scatter in the data,
higher stick-slip values tend to be associated
with lower WOB and higher RPM. The
highest stick-slip values are observed for
rotary speeds over 90 RPM and bit weights
less than 30 klbs. The data suggests that 30
klbs may be a critical bit weight to suppress
BHA whirl in this interval. The results also
show that the twirl mode can be very
sensitive to WOB, particularly for the upper
BHA indices, such as the Transmitted Strain
Energy shown on the right, but also the EndPoint Curvature index (not shown). The
BHA Strain Energy index and the virtual
accelerometer (also not shown) indicate a
more muted effect.
This data pattern reflects a condition
described as BHA whirl-induced stick-slip
and is consistent with the model results. It is
interesting to note that this observation
occurs at conditions opposed to bit-induced
stick-slip, and thus BHA whirl-induced stickslip should be taken into consideration in
both the design and operational phases.
The overall assessment is that RSS-2 was
a successful run, whereas RSS-1 was not.
For this comparison, the vibration model
results are in very good agreement with the
field results.

IPTC 13349

End-Pt
Curvature
Low ~ 1 to 4

WOB

RPM

StickSlip

BendStrn

Mode1

EndPtCurv

Twirl

Fig. 18. The Log Mode display for RSS-2 run, with low End-Point Curvature indices throughout run.

Fig. 19. Data from Fig. 18 plotted versus RPM (top row) and WOB (bottom row) for a 1300 ft interval.

IPTC 13349

Case 3: Sweet Spot Prediction for a 171/2-inch Motor BHA


It was noted in a morning report that one of the
rigs experienced a sweet spot around 90
RPM while drilling with a 17-1/2-inch motor
assembly that is used commonly in the field.
As a final case study, this observation is
investigated.
A schematic of the BHA design is provided
in Fig. 20. The top stabilizer is considerably
undergauge, however, it is treated as a nodal
point in the model. Stabilizer contact at this
location along the assembly will cause it to be
a nodal point. Also shown in Fig. 20 is a 3D
map of the 1X flex mode result for this design.
The line parallel to the DBIT axis represents
a slice along the BHA for a given operating
speed. The line at 90 RPM in Fig. 20 crosses
the tapered end of several humps, suggesting
that vibration mode peaks seen at higher speeds
may be avoided at 85-90 RPM.
The Transmitted Strain Energy vibration
indices for the individual 1X, 2X, and 3X
modes are plotted in Fig. 21. One may note
that there is a considerable rise in the 1X shape
above 85-90 RPM, and the 90 RPM point is
near the right edge of a flat response extending
back to 55 RPM. However, the 2X mode is on
a decline from 60 RPM to 85 RPM, then it
gradually rises towards 110 RPM. Finally, the
3X mode response shows a local minimum
near 85-90 RPM.
Thus, comparing the 3D chart in Fig. 20
and the individual components of the lateral
bending vibration index in Fig. 21, the model
results tend to support the field observation of
a sweet spot near 90 RPM.
Using the DVDT plot format, Fig. 22
provides LatMax acceleration data, in gs,
for a 2870 ft run in the well in which the sweet
spot observation was made. The solid red
curve is a quadratic fit to the data, showing a
slight bowing tendency towards 85-90 RPM.
The data shown with the black x is the
Transmitted Strain Energy, with the 1X mode
at top left, the 2X mode at top right, the 3X
mode at bottom left, and the RMS(1X-3X) at
bottom right.
There is a reasonable correspondence
between the curve fit and the model results in
Fig. 22. A sweet spot near 90 RPM is indeed
plausible, given the model results and field
data shown in these charts. In this particular
case, the project team set two field records
(fastest 24-hour footage and best cycle time for
all 17-1/2-inch sections), attributable in part to
rotary speed optimization. The operator now
uses vibration model results in all pre-section
technical reviews in order to enhance field
optimization practices.

Model shows reduced


amplitude ~90 RPM

Fig. 20. Standard motor BHA for 17-1/2-inch hole and 3D response surface for 1X flex mode.

1X mode

2X mode

3X mode

Fig. 21. Lateral bending indices (Transmitted Strain Energy Index)


for individual excitation modes.

Fig. 22. Log-mode display for field run in which 90 RPM was cited as sweet spot.

10

IPTC 13349

Summary
Bottom-hole assembly design improvement is an important component of vibrations management. It has been demonstrated
that better understanding and modeling of the influence of stabilizer and roller reamer placement on dynamic bending forces
and lateral stability enables development of improved BHA designs. A comparative modeling approach has been
implemented with an efficient frequency-domain vibration model to assess the relative vibration tendency of different
assembly designs, and the model results are closely aligned with field experience.
In the first case study, it was shown that the placement of roller reamers and stabilizers in a BHA design can significantly
affect BHA dynamics and drilling performance. Lower vibration indices are associated with higher daily footage, better bit
grades, and less damage to BHA components, in this case roller reamers. The second case study compared two rotary
steerable assemblies, one which ran a few hundred feet and was pulled for failing twice, showing excessive levels of lateral
vibrations, stick-slip, and mechanical specific energy. The second RSS configuration drilled almost 5000 ft and achieved all
directional objectives. Model results showed significantly lower lateral bending flex indices for this design. The third case
study showed that both model and field drilling data agree with a rotary speed sweet spot for this BHA near 90 RPM.
In these studies, the relative influence of flex lateral bending and rotational twirl modes was considered, and it has been
found that the most critical design objective is to reduce the flex mode indices to reduce lateral vibrations. In one case
study, the data suggests that twirl indices for the upper BHA can be used as a secondary assessment tool to evaluate the
potential for BHA-induced stick-slip tendencies of various design alternatives.
In these and several other case studies, it has been found that a relative performance assessment based on use of the model
is a good predictor of field results. The joint vibrations study has been integral to the operators drilling performance
management system and has resulted in close collaboration during the well design cycle.
Acknowledgements
The authors are grateful for the support of ExxonMobil and RasGas Company Limited for permission to publish this paper.
We have continued to enjoy working with our co-authors of the previous papers referenced below, and we appreciate their
continued support and guidance. All figures in this paper are original figures created by ExxonMobil and RasGas Company
Limited.
References
1.
2.
3.
4.

Dupriest, F. E., J. W. Witt, S. M. Remmert, and D. R. Aberdeen. Maxmizing Rate of Penetration with Real Time Analysis of Digital
Data. IPTC Paper 10706 presented at the 2005 International Petroleum Technology Conference, Doha, Qatar, Nov. 21-23.
Remmert, S. M., J. W. Witt, and F. E. Dupriest. Implementation of ROP Management Process in Qatar North Field. IADC/SPE
Paper 105521 presented at the 2007 IADC /SPE Drilling Conference, Amsterdam, Feb. 20-22.
Bailey, J. R., E. A. O. Biediger, V. Gupta, D. Ertas, W. C. Elks, and F. E. Dupriest. Drilling Vibrations Modeling and Field
Validation. IADC/SPE Paper 112650 presented at the 2008 IADC /SPE Drilling Conference, Orlando FL, Mar. 4-6.
Bailey, J. R., E. A. O. Biediger, S. Sundararaman, A. D. Carson, W. C. Elks, and F. E. Dupriest. Development and Application of a
BHA Vibrations Model. IPTC Paper 12737 presented at the 2008 International Petroleum Technology Conference, Kuala Lumpur,
Malaysia, Dec. 3-5.