Stick-slip Training Guide

Baker Hughes INTEQ

Stick-Slip

A DrillByte Applications Training Guide 80971 Rev. A

June 1995

Baker Hughes INTEQ Training and Development 2520 WW Thorne Houston, TX 77073 USA 713-625-4890

Table of Contents

Table of Contents Chapter 1

Stick-Slip Basics Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1 Theoretical Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2 Surface Detection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3 Chapter 2

The Stick-Slip Program Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1 Stick-Slip Monitor Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1 Monitor Start-Up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-2 Monitor Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-4 Optional Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-6 Chapter 3

Curing Stick-Slip Problems Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1 Causes of Stick Slip . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1 Cures For Stick Slip. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-2 Immediate Actions Possible . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-2 Actions Possible at Trip Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-2 Rig and Well Design Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-2

Training Guide 80971 Rev. A / June 1995

i Confidential

Table of Contents

Stick-Slip

Appendicies

Appendix A

Modifications to the P&F Rack for Stick-Slip Monitoring Appendix B

Files Used in the Stick-Slip Program Appendix C

References Appendix D

Stick-Slip Examples Appendix E

Recommended Stick-Slip Set Points

ii

Baker Hughes INTEQ Confidential

80971 Rev. A / June 1995

Chapter

1

Stick-Slip Basics Chapter 1 provides an introduction to the stick-slip phenomenon; what causes it, how it occurs, the problems it can cause, and why stick-slip should be prevented.

Introduction The DrillByte Stick-Slip application is designed to improve drilling performance through identification of the presence of Torsional Stick-Slip, and monitoring the success of the applied preventive measures. The economic benefit of the application is achieved when, after identifying stick-slip, action is taken which leads to an improvement in drilling performance or prevention of downhole tool failure. Stick-slip has been recognized as impacting drilling performance by a number of operators over the years (Chevron's TOM, or Torque Oscillation Monitoring, Exxon's funding of George Halsey's work at Rogaland Research in Norway, Elf's Dynafor project, Shell's Soft Torque system, and BP's use of mud logging charts at high speeds). The DrillByte Stick-Slip monitor simplifies stick-slip identification, and enables the data to be presented to those on the rig that are able to take the preventative measures. An alternative approach to detection, the running of the logging unit's chart recorders at high speed, is only available in the logging unit and offers no alarm capability. During the course of drilling operations, there will be periods when the rotation of the drillstring downhole is not uniformly smooth. The Bottom Hole Assembly (BHA) will slow down, then accelerate to a speed higher than the applied rotary speed, in order to maintain the correct average speed. This cyclical motion reduces the efficiency of the drilling and frequently leads to damage of the downhole components. In addition to the detection approaches discussed in this Training Guide, some INTEQ MWD tools have the ability to monitor downhole RPM. If this service is being run, and the RPM value is sampled every 5 seconds Training Guide 80971 Rev A / June 1995

1-1 Confidential

Stick-Slip Basics

Stick-Slip

(rather than an average) the variation in spot readings can be very useful when used in conjunction with the DrillByte Stick-Slip monitor.

Theoretical Background The sticking and slipping of the drillstring (known as rotational or torsional stick-slip) is an important dynamic phenomena which can result in premature bit wear, drill pipe fatigue, premature failure of downhole motors, and can induce other detrimental drillstring dynamics. This "torsional" stick-slip should not be confused with “axial” stick-slip during sliding (which is due to friction acting axially along the drill string, resulting in intermittent bit loading) because torsional stick-slip is due to dynamic and static friction between the BHA and borehole wall, which is impeding the steady-state rotation of the BHA. In the theoretical examination of torsional stick-slip, the drillstring may be thought of as a torsional spring (drill pipe) and a lumped mass (BHA). When rotation commences at the surface, the lumped mass will tend to lag behind until sufficient torque is built up to overcome the inertia of the mass and any additional frictional forces that may be present. The rotational speed of the mass will tend to oscillate around the surface rotary speed (these oscillations will eventually stop) until the bit's rotary speed matches the surface rotary speed. The oscillations described above are known as torsional oscillations. The period of one oscillation (depending on the length's of the drillpipe and BHA, and their respective diameters) will generally range between 2 to 15 seconds. Stick-slip is a severe and persistent case of these torsional oscillations, whereby the BHA comes to a complete stop and then suddenly releases at a high rate of speed. This torsional stick-slip is due to higher static than dynamic friction levels which impedes the rotation of the BHA and bit. These high static friction levels are caused by various drilling phenomena, such as: •

the buckling or whirling of the BHA

•

the aggressive cutting action of a PDC bit

•

lithology changes

•

contact between stabilizers and the borehole wall

•

abruptly starting the bit on bottom

When sufficient torque is developed in the string to overcome this static friction, the drillstring will initially rotate rapidly, then slow down. It will stop again when its rotational velocity drops below some critical value, bringing static friction back into play.

1-2

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Stick-Slip

Stick-Slip Basics The period of this stick-slip phenomena is generally longer than the torsional oscillation period, due to the time the bit is motionless. Even though the rotary speed at the bit (on average) will be the same as the surface rotary speed, its instantaneous speed will vary from zero to more than twice the surface speed, and can even become negative (backwards rotation) in particularly severe cases.

Surface Detection Depending on the type of rotary drive on the rig, detection will be most apparent in either Torque or Rotary Speed or both. Though the DrillByte Stick-Slip application was primarily designed for electric Top Drives or Rotary Tables, it can be equally effective with hydraulic Top Drives. Torque signals from the sensor is filtered to remove the high frequency signal components (those events happening faster than twice a second), then the sensor reading is sampled to produce twenty-five plot points per second, which is then passed to the display routine every two seconds. There may be similar oscillations in rotary speed, although deviations from the static value are generally minor. If an analog feed to a slow chart recorder is used (i.e. rig floor drilling recorder), the trace may show a wider track and the pen will visually oscillate from side to side, at a constant rate. If the output is to a heavily-damped analog chart recorder, the oscillations of stick-slip can be completely removed. If surface torque fluctuations are greater than 15% of the mean surface torque, there is a high probability that stick-slip is present and corrective measures should be taken. As stated earlier, the period of the stick-slip fluctuations will depend on the length and mechanical properties of the drillstring (i.e. for 5-inch drillpipe, the period of oscillation is about 2 seconds per 1000 meters, or about 8 seconds for a 4000 meter drillstring). In highly deviated wells, stick-slip type problems can be generated by the frictional forces resulting from the BHA or drillstring rubbing against the borehole wall, rather than by the bit/formation interaction. In these cases, the stick-slip symptom will not disappear when coming off-bottom. Surface stick-slip detection includes: •

surface torque fluctuations (including Top Drive stalling)

•

increased torque cyclicity

•

increased MWD shock counts

•

cutter impact damage

•

drillstring twist-off or washout

•

connection over-torque or connection back-off

Training Guide 80971 Rev A / June 1995

1-3 Confidential

Stick-Slip Basics

Stick-Slip

•Notes•

1-4

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Chapter

2

The Stick-Slip Program Chapter 2 provides the information necessary for the operator to understand the DrillByte Stick-Slip Application program. Complete details of the program can be found in Volume 1 of the DrillByte Reference Manuals (P/N 80319H-001).

Introduction The applications program within DrillByte provides stick-slip detection by monitoring the current applied to the electric motor running the Rotary Table or Top Drive. This signal is filtered and then sampled at 25 samples per second. A detection algorithm will "look" for stick-slip by comparing the calculated values to the operator set threshold controls. Monitor alarms can be either audio or visual, and can include the automatic presentation on a workstation screen. The standard DrillByte plot package can be used to create an emulation of the display screen for output to non X-terminal dispalys.

Stick-Slip Monitor Design Incoming data is sampled at 125 data points per second, then reduced to 25 points, to provide enough data points for graphical representations. The DAQ will use a number of low-pass filters (hardware analog filters and software digital filters) to prevent high frequencies from aliasing the sampled torque signals. Frequencies above 2.0 Hz are filtered out. When the filtered data is transmitted from the DAQ to DrillByte, the data acquisition daemon (daqd) buffers the incoming data. A flag (initialized to zero) is used to indicate the presence of new data. This flag is incremented by daqd each time a new data set is received. This data set is located in the DAQ Shared Memory and can be accessed by other DrillByte programs. The daqd also converts the torque sensor data into SI units.

Training Guide 80971 Rev A / June 1995

2-1 Confidential

The Stick-Slip Program

Stick-Slip

Since significant changes in hookload can trigger the recording of stickslip, another parameter, sigma hookload, is also monitored. Stick-slip recording will occur if the RMS average hookload varies by more than the hookload delta setpoint. If the Total RPM is zero, only the hookload data is recorded. A separate flag within DrillByte is maintained for hookload. Processing of the torque data involves sequentially passing each data point through the stick-slip algorithm to determine if a stick-slip status is met. Averaging is used to reduce the status to two-second updates, then the status is stored in the DrillByte Common Data Area (E_ST_SLIP_ST). This status can appear as: •

No Application Running

0

•

No Stick-Slip

1

•

Possible Stick-Slip

2

•

Definite Stick-Slip

3

When there is a stick-slip event change (i.e. from Possible Stick-Slip to Definite Stick-Slip) or the hookload flag is met, data is stored in a timestamped file in the $DBYTEHOME/ctl directory. Raw torque and processed Rotary Speed, Hookload, Pump Pressure, and the stick-slip flag are recorded. Each event is recorded in a new time-stamped file (i.e. Oct23941430.slip) and can be stored in the database. The recorded data can also be imported back into the stick-slip program for review.

Monitor Start-Up The Stick-Slip monitor is started from Launcher, by selecting Monitor:Stick-Slip, and can be run in two modes - Active or Passive. In the "Active Mode", all stick-slip options are available to the operator, including the ability to modify parameters in the Set-Up menu and to record data files. Only one Active Mode monitor can be run. In the "Passive Mode", the stick-slip parameters are read (from the stslip.ctl file) and cannot be modified. Stick-slip calculations will be made and displayed on the screen, but no data files will be stored. This passive version was designed for remote X-terminals and can be started from the Launcher menu “custom” line by typing: stick -p. Several Passive mode monitors can be run with the one Active Mode monitor. When started, the stick-slip interface is a single window, with a control panel for user input and the remaining part of the window for graphical output (see Figure 2-1).

2-2

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Stick-Slip

The Stick-Slip Program

Figure 2-1: Stick-Slip User Interface Window

If the program is iconized, the icon (Figure 2-2) will change colors to represent the stick-slip status: •

Green

No Stick-Slip

•

Yellow

Possible Stick-Slip

•

Red

Definite Stick-Slip

Training Guide 80971 Rev A / June 1995

2-3 Confidential

The Stick-Slip Program

Stick-Slip

Figure 2-2: The Stick-Slip program icon

Monitor Operation The setup button on the control panel allows the operator to view or alter those parameters which are used to isolate stick-slip during drilling.

Six criteria must be met: Torque Delta - which is used to monitor the dynamic torque amplitude. The dynamic torque value must exceed the static value by a certain percentage before the algorithm will be used. The default is 10%. Note: If the torque delta default is used, the stick-slip algorithm will detect "Possible Stick-Slip" when the dynamic amplitude exceeds the static value by 10% and "Definite Stick-Slip" at 14%. 2-4

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Stick-Slip

The Stick-Slip Program Hookload Delta - which is used to trigger the hookload flag. If the RMS hookload (averaged over one minute) varies by more than this percentage amount, the flag will be set and data recording will begin. Torque Threshold - which is used to monitor the amount of torque.If the torque measurement is below this value, stick slip will not be recognized. The default is 500 daN_m. Minimum Period - which is the minimum time period that stick-slip will be considered valid. The default is 2 seconds. Maximum Period - which is the maximum time period that stick-slip will beconsidered valid. The default is 15 seconds. Lag - which is the number of sequential stick-slip oscillations that must occur before the stick-slip status changes from 1 (no stick-slip) to 3 (definite stick-slip). The default is 3 cycles, however field experience has shown that 5 cycles will help reduce the number of false alarm occurrences. Two items allows the operator to select how the torque measurements are being received by the monitoring program. Torque Channel - is the DAQ channel sending the torque data. Torque Source - selects the type of sensor used (Electrical or Mechanical). There are two optional boxes which the operator can select: Store to File - allows stick-slip events to be saved to a file in the $DBYTEHOME/ctl directory. Auto De-iconify - will automatically open the program if the monitor has been iconized and a red condition occurs. When the default variables are not sufficient (i.e. stick-slip is not detected or there are too many false alarms), those default values must be changed. Appendix E contains additional set-points for various BHA’s. The setpoints in the appendix are representative of example BHA’s, and are the result of theoretical calculations and the observation of field data. During “normal” drilling operations, it may be necessary to move the lag set-point up to 5 (meaning 5 cycles must be seen before a definite stick-slip is declared), put the hookload delta to 5%, and have a low torque threshold (around 0.5 kft-lbs). Since torque delta is one of the most dynamic set points, a value of 10% should be sufficient (the recommended range is between 5% to 45%). The combination of setting the lag to 5 and using the minimum and maximum period values shown in Appendix E, will effectively reduce the number of false alarms.

Training Guide 80971 Rev A / June 1995

2-5 Confidential

The Stick-Slip Program

Stick-Slip

The scales button on the control panel allows the operator to manipulate the scales on the graphical output. The screen output contains four graphs: Torque, Rotary Speed, Hookload and Pump Pressure. The "x" axis on all plots is time (in seconds), the minimum is always zero, while the maximum is operator selectable (it must be greater than 10). The operator must input the minimum and maximum values for the "y" axis on all four plots. When the apply button is used, validation of the parameters is done, the values are selected, and the parameters are stored in the stslip.ctl file in the $DBYTEHOME/ctl directory.

Optional Operations The control panel contains several buttons which can be used to view, import and record stick-slip data. Event Log - will open a standard Open Look text window on the screen displaying the stick-slip events stored in the $DBYTEHOME/ctl directory These are the events triggered by the stick-slip flag, not the hookload flag. The screen will list the time of the event and the type of the event. For example:

2-6

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Stick-Slip

The Stick-Slip Program

16 Jun 94 17:30:43 : Transition to Possible Stick-Slip Stopped - toggles the stick-slip processing On and Off, and can be used to freeze the current graphics. Replay - is used to import previously stored stick-slip data from the $DBYTEHOME/ctl/*.stslip file. This button produces a scrolling list of stick-slip data. During normal monitoring, this button is "grayed out". Record - is used to manually force the stick-slip program to record a data file. When selected, the file will be created containing two minutes of data (information recorded one minute before selection and one minute after selection). The following may apply: 1.

This option is not available if the program is run in the Passive Mode

2.

If a "real" stick-slip event occurs while manually recording, recording will continue until one minute after the end of the "real" event.

3.

While a "real" stick-slip event is being recorded, this option is not available

Training Guide 80971 Rev A / June 1995

2-7 Confidential

The Stick-Slip Program

Stick-Slip

•Notes•

2-8

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Chapter

3

Curing Stick-Slip Problems When stick-slip problems are detected, they must be cured before drillstring damage occurs. Chapter 3 lists the cures necessary when problems are identified.

Introduction Once stick-slip is identified, action is required to address the problem. Some steps can be taken immediately, others may require a bit trip or rig modifications on a longer time scale. The first step is to determine the potential causes of stick slip, and the correct action to cure it. It has been noted that when the Stick Slip displays a saw tooth pattern, that this can be an indication that the Stick Slip is driven by the bit, rather than a smoother sine wave, seen with stabilizers or drillstring contact. This can be a useful guide, through it is not yet proven

Causes of Stick Slip Factors causing Stick Slip can be grouped into three general categories: •

Category #1

Excessive stabilizer or BHA side loading from a buckled Bottom Hole Assembly. This is especially common when drilling with very high bit weights, or with very limber assemblies (i.e. slimhole drilling). •

Category #2

a. Stabilizer and/or BHA friction from excessive hole tortuosity. b. Excessive weight on PDC bits for the formation type or rock properties encountered. If too large a "bite" is being taken by the bit, the stalling action may cause stick slip. •

Category #3

Lack of fluid lubrication. Poor drilling fluid properties or excessive cuttings build-up can be a factor in promoting stick slip. Training Guide 80971 Rev A / June 1995

3-1 Confidential

Curing Stick-Slip Problems

Stick-Slip

Cures For Stick Slip Based on what is causing the stick slip problems, the cure can be accomplished quickly or it may require a bit trip. Immediate Actions Possible It should be noted that "immediate" is not instantaneous. It may take a few minutes for the effects of a change in drilling parameters to take effect. In severe cases, to totally remove the problem, it may be necessary to stop drilling, lift off bottom, stop rotating, and restart the drilling process. Parameter Management - these steps are primarily aimed at reducing bit weight and/or increasing rotary speed. This is designed to address categories 1 and 2. For example; for a given PDC bit design, increasing the RPM will reduce the instantaneous depth of cut. Reducing the weight will also reduce the side forces in a buckled BHA. It may be necessary to both reduce weight and increase RPM. Fluid Properties - Increase the lubricity of the drilling fluid. Actions Possible at Trip Time BHA Design - Add stabilization to increase buckling resistance. Substitute roller reamers for stabilizers. Bit Selection - Select a less aggressive bit or bit type. Rig and Well Design Issues Torque Feedback System - The rig can be modified to include a Torque Feedback System Dogleg Severity - Use downhole equipment which reduces wellbore tortuosity.

3-2

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Appendix

A

Modifications to the P&F Rack for Stick-Slip Monitoring The P&F filter card (INTEQ P/N 81032, rev C) required modification in order to be used with the DrillByte Stick-Slip Monitor. When modified, the cards are identified by the etched characters "EX#" (e.g. EX1, EX2 ...). In addition, the front panel of the P&F rack will be silked screened with "Filter Card". Channels one and two have the same characteristics as the orginal filter card, while the third channel provides signal conditioning for the Torque signal from the standard three-wire torque sensor (INTEQ P/N 80301H). The original DAQ (INTEQ P/N 28950), does not provide for connection to the outputs of the P&F filter card, and a back panel upgrade kit (INTEQ P/ N 80560H) is available to provide those connections to all P&F outputs. As there are several "versions" of the P&F rack in use, some of which provide for the three-wire sensor to the filter card, there is a simple modification for those which do not have the filter card configuration. The following configurations of equipment (with the correct modifications) can be used for stick-slip monitoring: 1. Modified P&F Mk3b with an upgraded DAQ 2. Modified P&F Mk3 or Mk3a with a standard DAQ* 3. Modified P&F Mk3 or Mk3a with an upgraded DAQ The stick-slip program automatically notifies the DAQ of the presence of the application, which initiates the extra data transfer of 50 data points every two seconds to the DrillByte application from the DAQ.

P&F Mk3 and Mk3a Modification Remove the top of the unit by unscrewing the four allen-head screws. Swap the connectors in J43 and J42 on the Mother PCB (INTEQ P/N 37596) after marking them with their original locations. Check that LK7, near J9, on the Mother PCB is connected between A and C. Replace the cover. Install an EGT101 card (INTEQ P/N 55202) in J9 and the modified filter card in J17.

Training Guide 80971 Rev A / June 1995

A-1 Confidential

Modifications to the P&F Rack for Stick-Slip Monitoring

Stick-Slip

P&F Mk3b Modification Remove the top of the unit by unscrewing the eight phllips-head screws. Remove the connectors from J35 and J51 and mark as to their original positions. Place the connector from J35 into J51. Tie the connector from J51 back so that it cannot short out. Install the link (INTEQ P/N 37593) in LK12, near J12 on the Mother PCB. Replace the cover. Install an EGT101 card in J12 and the modified filter card in J17. After modifications have been made, tag the rack with "Modified for StickSlip Monitoring". After monitoring operations, return the rack to its original configuration. * Of the analog outputs from the block height system, only the block height signal is passed to the DAQ. The following tables will aid in the connection of the torque sensor and determination of the corresponding DAQ input channel.

P&F Mark 3 and 3a - Modified Inputs Table 1: Channel

Back Panel

Mother PCB

P&F Card Input

P&F Card Output

Filter Card Input

1

J12 B,C

J41 1,3

J8 2a,6a

J8 30a

J17 18z

2

J11 D,E

J43 1,3

J9 2a,6a

J9 30a

J17 24z

3

J11 F,G,H

J42 1,2,3

J9 2z,4z,6z

J9 30z

J17 30z

Filtered Outputs Table 2: Channel

Filter Card

Back Panel

Bandpass Hz

Gain

DAQ Channel

Modified DAQ Channel

1

J17 20a

J2 9,28

0.08 to 100

26

40

44

2

J17 26a

J2 11,30

0.08 to 100

26

42

46

3

J17 32a

J2 13,32

0 to 2

1

44

48

A-2

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Stick-Slip

Modifications to the P&F Rack for Stick-Slip Monitoring

Buffered Outputs (no filtering) Table 3: Channel

Filter Card

Back Panel

DAQ Channel

Modified DAQ Channel

1

J17 20z

J2 10,29

41

45

2

J17 26z

J2 12,31

43

47

3

J17 32z

J2 14,33

45

49

P&F Mark 3b - Modified Inputs Table 4: Channel

Back Panel

Mother PCB

P&F Card Input

P&F Card Output

Filter Card Input

1

J12 M,N

J49 1,3

J12 d2,d6

J12 30d

J17 18z

2

J11 R,P

J50 1,3

J13 z2,z6

J13 30z

J17 24z

3

J11 E,F,G

J51 1,2,3

J13 d2,d4,d6

J13 30d

J17 30z

Filtered Outputs Table 5: Channel

Filter Card

Back Panel

Bandpass Hz

Gain

DAQ Channel

Modified DAQ Channel

1

J17 20a

J1 16,35

0.08 to 100

26

N/A

33

2

J17 26a

J1 17,36

0.08 to 100

26

N/A

34

3

J17 32a

J1 18,37

0 to 2

1

N/A

35

Buffered Outputs (no filtering) Table 6: Channel

Filter Card

Back Panel

DAQ Channel

Modified DAQ Channel

1

J17 20z

J2 16,35

N/A

51

2

J17 26z

J2 17,36

N/A

52

3

J17 32z

J2 18,37

N/A

53

Training Guide 80971 Rev A / June 1995

A-3 Confidential

Modifications to the P&F Rack for Stick-Slip Monitoring

Stick-Slip

•Notes•

A-4

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Appendix

B

Files Used in the Stick-Slip Program The following files in the $DBYTEHOME/ctl directory are used by the Stick-Slip Monitoring Program. stslip.ctl This file contains the current parameters used by the stick-slip program. These are: •

Torque Threshold

•

Torque Delta

•

Hookload Delta

•

Maximum Period

•

Minimum Period

•

Lag

•

Xmax

•

Torque Ymin

•

Torque Ymax

•

RPM Ymin

•

RPM Ymax

•

HKLD Ymin

•

HKLD Ymax

timestamp.slip These files store the stick-slip event data. This file is composed of three parts, the first part contains: •

Month day year hour minute (e.g. Oct23941430.slip)

•

Trigger/Time/Date (Torque or Hookload, hh:mm:ss:dd:mm:yy)

•

Sample Rate

Training Guide 80971 Rev A / June 1995

B-1 Confidential

Files Used in the Stick-Slip Program •

Torque Threshold

•

Torque Channel

•

Torque Delta

•

Hookload Delta

•

Maximum Period

•

Minimum Period

•

Lag

•

Store Flag

•

Torque Ymin

•

Torque Ymax

•

Hookload Ymin

•

Hookload Ymax

•

Pump Pressure Ymin

•

Pump Pressure Ymax

•

Rig Activity

•

Bit Depth

Stick-Slip

The second part is composed of the time-stamp (unix time) data 1 data 2 data 50 rpm value hookload value pump pressure value stick-slip value etc ... (this two-second data block format is repeated) The final item in the file is a count of the two-second samples SAMPLES = 87 (for a 174 second file)

B-2

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Appendix

C

References Baker Hughes INTEQ, DrillByte Basics (DrillByte Volume 1, P/N 80319H-001) Rev. A, September 1994 Dufeyte, M.P., and Henneuse, H., Detection and Monitoring of the SlipStick Motion: Field Experiments, SPE/IADC 21945, March 1991 Baker Hughes INTEQ, Technical Bulletin #1, April 1994 Abbassian, F., Drillstring Vibration Primer, BP Exploration, January 1994 Fear, M.J., and Abbassian, F., Experience in the Detection and Suppression of Torsional Vibration From Mud Logging Data, SPE 28908, October 1994

Training Guide 80971 Rev A / June 1995

C-1 Confidential

References

Stick-Slip

SPE 28908 Experience in the Detection and Suppression of Torsional Vibration From Mud Logging Data1 M.J. Fear, BP Exploration Co. (Colombia) Ltd., and Fereidoun Abbassian, BP Exploration SPE Members Copyright 1994, Society of Petroleum Engineers, Inc. This paper was prepared for presentation at the European Petroleum Conference held in London, U.K., 25-27 October 1994. This paper was selected for presentation by an SPE Program Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper, as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Society of Petroleum Engineers. Permission to copy is restricted to an abstract of not more than 300 words. Illustrations may not be copied. The abstract should contain conspicuous acknowledgment of where and by whom the paper is presented. Write Librarian, SPE, P.O. Box 833838, Richardson, TX 75083-3836, U.S.A. Telex 163245 SPEUT.

ABSTRACT

Copyright 1994 SPE. Reprinted by permission.

Vibration detection from mud logging systems has revealed that torsional vibration is common in harsh drilling environments, and is a major cause of bit and drillstring failures. Suppressing this type of vibration with an automated vibration detection system, torque feedback, and rigsite vibration suppression guidelines has produced a significant improvement in drilling performance.

INTRODUCTION In the drilling industry, it is now well established that vibration can cause premature failure of the drillstring and bit. In recent years, this understanding has been extended to identify the relationship between specific modes of vibration and certain types of damage (1,2). in response, various types of monitoring equipment have been developed, and have demonstrated that vibration can often be detected and suppressed (3,4,5,6,7,8,9). This paper focusses on the detection and suppression of torsional vibration, which appears at surface as regular, periodic cycling of drive system torque. These oscillations usually occur at a frequency close to the fundamental torsional mode of the string, which depends primarily on the drillpipe length and size, and the mass of the bottomhole assembly (BHA). Their amplitude depends upon the nature of the frictional torque downhole, and the properties of the surface drive system. The significance of this behaviour at surface is that it is accompanied by alternating acceleration and decelaration of the bottomhole assembly and bit, and repeated twisting of the more limber drillpipe section. The most severe form of this vibration produces slip-stick behaviour of the bit and BHA, during which the BHA alternately comes to a complete halt, until twisting of the drillpipe section produces sufficient torque to overcome the resistance to bit/BHA rotation. The BHA then spins free, accelerating to a significantly higher speed than observed at surface, before slowing down again as rotational energy is dissipated. There are a number of physical consequences of this type of behaviour which are damaging to the bit and drillstring. The first and most obvious is the cyclic stresses which will accompany the non-uniform rotational motion of the drillstring. 1.

REFERENCES AND FIGURES AT END OF PAPER

C-2

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Stick-Slip

References

Copyright 1994 SPE. Reprinted by permission.

Field observations suggest that persistent twisting and unwinding of the drillstring can cause stalling and over-torquing of connections, while fluctuating bit speed (and hence cutter loads on fixed cutter bits) will cause fatigue failure of the cutting elements on the bit. Second, downhole measurements show that bursts of lateral BHA vibration can accompany the rotational accelerations of the BHA, producing bending stresses which can ultimately cause connection failure or impact damage to more sensitive components such as measurement while drilling (MWD) tools. Third, similarities between poly-crystalline compact (PDC) bit damage observed after periods of cyclic torque in the field, and that produced by rotating PDC bits backwards in laboratory tests (2), suggest that intermittent periods of backward BHA rotation may occur during severe slip-stick, after the bit halts. This would explain why connections are sometimes found to have backed off downhole, in some cases leading to wash outs and twist-offs. Each of these phenomena is made more critical by the ease with which torsional vibration can be initiated, its persistance once started, and the minimal damping that is associated with low frequency forms of torsional vibration. Modelling and laboratory observations have shown for example that downhole torque variations, produced by nothing more than load or speed changes at a PDC bit, can trigger torsional oscillations and slip-stick (2). In simple terms, torsional vibration is easy to initiate because the low torsional stiffness of the drillstring means that small torque fluctuations downhole can produce large rotational displacements. Once started, the torsional waves propagating along the drillpipe are reflected back downhole by the relatively high impedance of the surface drive system, creating a self-perpetuating transfer of energy between the drillpipe and BHA sections. This can build into full slip-stick behaviour, which will persist until conditions are changed (10,11). The work reported here concentrates on eradicating this form of vibration. Emphasis is placed on provision of the necessary detection capabilities, and modifications to conventional drilling practices and drive system equipment to provide effective vibration suppression.

FIELD INVESTIGATIONS AND VIBRATION DETECTION This work was preceded by a study of whether vibration was a significant problem in BPX drilling operations. Results were enlightening. On just eight wells from one North Sea development project, over $2 million of vibration-induced problem costs were identified. This estimate was conservative, including only those incidents where downhole and/or surface measurements confirmed vibration as the cause. In over half of these cases, drillstring failures and bit impact damage occurred in conjunction with variable surface torque. This raised the possibility that torsional vibration, which increases variance of surface torque, was a common cause of vibrationinduced failures. Figure 1 shows mud logging data from one such bit run, drilling with a 12.1/4" PDC bit in a compacted sand/shale sequence. High surface torque fluctuation early in the bit run is shown by a wide departure between maximum and minimum surface torque, per foot drilled. Figure 2 shows the PDC bit after the run. The main wear mode is impact damage to the cutters, with many having lost or damaged diamond layers. The mud logging data show that rate of penetration (ROP) drops abruptly late in the period of highest torque variance, due to the rapid development of cutter wear on the bit. Such an ROP drop is characteristic of worn bits in these formations. The impact damage on the bit is thus related to the erratic

Training Guide 80971 Rev A / June 1995

C-3 Confidential

References

Stick-Slip

torque. A possible explanation for both is the presence of torsional vibration. Brett (2) describes such an association for PDC bits. Compared to other runs with this bit type in these formations, this bit made around one quarter of the expected footage. Those other bits, which exhibited less impact damage, were run at higher rotary speeds. Though the bit damage on this bit run, and the behaviour of surface torque are both explicable in terms of torsional vibration, the data in figure 1 do not provide a definitive diagnosis of vibration type. However, during this investigation, high speed mud logging chart recorders were in use to capture the full dynamic behaviour of surface torque. One of the charts for this bit run is shown in figure 3, corresponding to point "X" on figure 1. The time scale is 24 seconds per horizontal division, and the two main variables shown are surface torque and rotary speed. The regular, rythmic cycling in these two variables is indicative of torsional vibration. Thus, an association between PDC bit damage and torsional vibration is confirmed by this run.

Copyright 1994 SPE. Reprinted by permission.

A combination of mud logging data acquired at a sufficient frequency, together with a high speed chart recorder, is all that is required to provide a detection capability for torsional vibration. Given the costs associated with vibration-induced drilling problems, development of a routine detection capability in harsh drilling environments was considered appropriate as a first step in eradicating this type of vibration. Subsequent work therefore concentrated on automating the detection capability, and combining it with rigsite vibration suppression guidelines so that the rig team would respond to the vibration once detected.

DETECTION AND SUPPRESSION OF TORSIONAL VIBRATION An outline specification was developed for automated detection of torsional oscillations at surface from mud logging systems. This involved measurement of the two main drilling parameters (drive system torque, i.e. current, and rotary speed, i.e. voltage), plus hookload and standpipe pressure, at a frequency sufficient to detect their full variation during vibration. Some mud logging systems were already sampling sensors at sufficient data rates. Next, the data were to be stored and displayed at the same rate, rather than being reduced to time or depth-based averages for display. Thus, patterns of drilling parameter behaviour could be displayed in full, and indications of vibration made visible. A simplified schematic of the system is shown in figure 4. Two major mud logging contractors have since developed and implemented this capability on their standard mud logging systems. A third is currently developing the package. Rig floor alarms have been added, together in one case with a signal detection algorithm to discriminate torque oscillations due to torsional vibration from other patterns of erratic torque, due for example to changes in weight on bit. These systems therefore provide the capability to automatically detect and display torsional vibration, and to alarm the rig crew to its occurrence. It was recognised that to remove this type of vibration, the new detection capability would have to be combined with rig guidelines on vibration suppression. A basic set of guidelines, designed to be usable by the driller, were therefore developed. The guidelines indicate practices that can stop torsional vibration once it has been detected. These recommended practices drew heavily on indications from in house and published models (2,10), together with observations from the field (12). They also described how the

C-4

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Stick-Slip

References

vibration would appear on drillers gauges. The part of these guidelines that deal with torsional vibration is shown in figure 5. The next section documents field experience from one year of drilling with the detection package, and latterly the suppression guidelines, in place.

EXPERIENCE WITH VIBRATION DETECTION AND SUPPRESSION The following examples are a small selection of the vibration incidents detected with the mud logging systems. They have been chosen to illustrate the most important characteristics of torsional vibration, and the effectiveness of various suppression methods. Example 1: Torsional vibration. twist-off

Copyright 1994 SPE. Reprinted by permission.

Figure 6 shows a small part of the printout from the vibration detection package, while drilling at around 12060feet with a 12.1/4" anti-whirl bit, on a rotary BHA and 6.5/8" drillpipe. A torque feedback system was in use. Mud type was oil base, and well inclination 21 degrees. The time scale on the chart is 20 seconds per major division. The data show the torque and rotary speed cycling that is characteristic of torsional vibration. The amount of rotary speed cycling in response to the torque fluctuations is exaggerated by the torque feedback system, which is designed to "soften" motor response with modified speed control. The fluctuation in hookload ("WOH" - weight on hook), is a result of rig power limitations rather than any coupled vibration. Similar behaviour is often seen in pump speeds and therefore standpipe pressure ("SPP" in the figures). The frequency of the torque oscillations is around 0.49 Hertz. Using a model similar to that described in reference 3, the lowest torsional resonance mode for this drillstring was calculated at 0.29 Hertz. Either a higher mode of vibration had been excited, or only a portion of the string was vibrating in its fundamental mode. The torque feedback system was of a type that is tuned to suppress vibration at the lowest torsional mode for the drillpipe size and length in use. The discrepancy between the expected and actual frequency of torsional vibration explains the lack of successful vibration suppression on this run. This particular torque feedback system requires manual re-tuning in cases such as this. After 6.5 hours of drilling, a twist-off occurred at the top of the near-bit stabiliser. The parted connection was washed, probably indicating that the connection partially backed off downhole, starting the washout. This mode of failure is compatible with the downhole behaviour previously described as occurring during slip-stick. Example 2: Torsional vibration. motor failure Figure 7 shows a realtime printout from the detectionpackage, while drilling at around 5200 feet, with a 9.1/2" steerable motor assembly and PDC bit in 17.1/2" hole. Mud type was water base. Differential pressure across the motor was approximately 400 psi. Hole inclination was 20 degrees. Again, the characteristic cycling in torque and rotary speed is present, indicating torsional vibration. The surface rotary speed in this case is, on average, low because bit speed is provided by the motor. In this interval, additional vibrational behaviour is present. Periods of high mean torque (and consequent lower mean rotary speed) coincide with higher variance in hookload,

Training Guide 80971 Rev A / June 1995

C-5 Confidential

References

Stick-Slip

indicating axial string motion (bouncing). The relationship of higher mean torque to high hookload variance could indicate discrete bursts of whirl, possibly from the bit. A coupling between whirl and axial type vibrations has previously been described to facilitate whirl detection at surface, when well depth is shallow (13). After 17 drilling hours, the motor failed, due to break up of the rubber stator. After analysis, defects in stator composition and properties were ruled out as the cause. The failure was then attributed by the motor supplier to excess differential pressure across the motor. This in turn was revised when the vibration data was studied; the irregular torsional and axial motion between the motor body and rotor, which would have accompanied the vibration, is a more likely cause. Rotation of the rotor would be anything but smooth in the presence of these forms of vibration. This was a good example of how a drilling operation may not realise that vibration is a cause of a failure until direct measurements of vibration are actually made available. The PDC bit also exhibited uneven wear, and cutter breakage (graded 6-2-BT). Example 3: Torsional vibration and stalling while reaming

Copyright 1994 SPE. Reprinted by permission.

Figure 8 shows the detection package printout from an interval of reaming at 13025 feet in 8.1/2" hole, with a PDC bit and two roller reamers. The data show large scale cycling in surface torque and rotary speed; peak to peak torque values are around 15 Kft.lbs. No torque feedback system was in use on this run. The data show two intervals where the top drive stalled, as the torque oscillations peaked at a level unsustainable for the top drive motor. The string was then picked up and the motor re-started, before reaming could commence. Repeated stalling is a common consequence of heavy torsional vibration; the maximum torque values are obviously raised by the dynamic component of torque, to the point where stalling is more likely. In the authors' experience, rig reports of stalling problems often indicate torsional vibration to be present, though without the direct measurements shown here, rig teams often do not realise that torsional vibration is occurring and therefore do not report the problem in those terms. Example 4: Liner failure during torsional vibration Figure 9 shows a vibration chart from a part of a liner running operation. The 7.5/8" diameter liner was being run into 8.1/2" hole, below a 9.5/8" casing string. High torque and an inability to slide without rotation required the use of high rotary speed to work the liner down. Unfortunately, rotation of the liner, as is obvious from the chart, was accompanied by torsional vibration. After 25-30 minutes of this behaviour, a connection in the liner failed, leaving 1000 feet of liner in the well. The point of failure is marked by the abrupt drop in rotary torque near the end of the chart. The part of the liner left downhole was being rotated in a section containing two sharp dog legs. The combination of cyclic bending, and the repeated twisting from the torsional oscillations, is the most likely cause of the connection failure. The occurrence of torsional vibration while running a tubular string without a bit or stabilisers confirms the hypothesis that differences between static and dynamic frictional torque between a drillstring and the borehole wall can set up torsional vibration (11).

C-6

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Stick-Slip

References

Example 5: Variable frequency of torsional vibration Figure 10 shows a vibration detection chart from a period of milling, with a 6" flat bottom mill at a 7" liner shoe. Measured depth was 16951 feet. The chart time scale is again 20 seconds per major division. The torque and rotary speed again show the cyclicity characteristic of torsional vibration. Early in the operation and on the chart, the frequency of the vibration is around 0.15 Hertz. At point "X" on the chart, the frequency of the torsional vibration changes to around 0.6 Hertz, with no change in applied parameters. This changing vibration frequency has been observed on a number of occasions now that torsional vibration is being routinely monitored.

Copyright 1994 SPE. Reprinted by permission.

Attempts were being made during this operation to remove the torsional vibration with a torque feedback system. This can be seen from the chart as having been unsuccesful, not least due to the changes in the period of torsional oscillations. An ability to quickly re-tune to a new vibration frequency is obviously a pre-requisite for an effective feedback system under these conditions. After approximately 14 hours of milling, with torsional vibration persistently present, a drill collar connection twisted off five collars above the mill. These drill collars had accumulated 145 hours since the previous inspection, and were new at the start of this milling operation (this was the third milling run). Obviously 145 hours of drilling under smooth conditions is less detrimental to connection fatigue life than operating under the conditions shown on the chart. This suggests that inspection frequency should take torsional vibration into account. Example 6: Suppression of torsional vibration Figure11 shows the vibration detection display window from another mud logging system. As before, torque, rotary speed, hookload and standpipe pressure data are displayed. Below and to the right of the upper (torque)section is the alarm status, which goes from green to red when torsional vibration is detected. In this example, while drilling cement with an 8.1/2" PDC bit at 18500 feet, torque cycling begins and progressively builds in amplitude. The detection system responds with a change to a red alarm, which is displayed in "traffic light"form next to the driller. The driller has then consulted the vibration suppression guidelines and raised the rotary speed from 100 to around 140 RPM. Toward the right end of the window, this is seen to eliminate the torsional oscillations. The use of higher rotary speed is recommended as the first method to eliminate torsional vibration since this does not cause any reduction in penetration rate. Use of lower weight on bit, which has been found to be equally effective, usually does reduce penetration rate. The disappearance of torsional oscillations above a certain rotary speed is fully in line with model predictions (2,10), and field observations (2,12).in the authors' experience, while drilling formations with PDC bits, that critical rotary speed usually lies in the range 150-220 RPM. The elimination of torsional vibration above these speeds explains the disappearance of torque problems and bit impact damage in some applications when PDC bits are run on downhole motors or turbines. Such practices are, for example, common in the hard Cretaceous limestones of the Central North Sea. It is worthy of note that if whirl was the major cause of PDC bit impact damage in harder formations, use of higher bit speeds should worsen bit damage (1). In the authors' experience, use of higher bit speeds appears to improve PDC bit performance more often than it makes it worse. This is because the extra life that is afforded by eliminating

Training Guide 80971 Rev A / June 1995

C-7 Confidential

References

Stick-Slip

torsional vibration more than offsets the increase in abrasive wear that accompanies higher bit speeds.

IMPLICATIONS FOR TORQUE FEEDBACK SYSTEMS Torque feedback systems provide modified electronic speed control to the surface motor, so that an increase in surface torque (motor current) is countered by a decrease in motor speed (motor voltage). This increases damping provided by the motor and helps to prevent initial torsional oscillations building into full slip-stick. As a result of the lower surface impedance, torsional oscillations are absorbed at surface rather than reflected back downhole (3,6).

Copyright 1994 SPE. Reprinted by permission.

The field experience reported here has included a number of observations pertinent to the design of these feedback systems. First, the frequency of torsional vibration may not be that of the lowest torsional resonance mode for the full length of drillpipe, and that frequency may not remain constant. For system tuning purposes, modelling of the drillstring to determine the expected frequency of vibration (the lowest torsional mode) is thus less satisfactory than measuring the predominant frequency directly. Second, torsional vibration occurs over a wide range of frequencies, so that a feedback system must be effective over as wide a frequency range as possible, with minimal re-tuning required as the frequency changes. This implies that the characteristics of the system be such that the feedback setting is relatively insensitive to vibration frequency. Sananikone et al (6) provide a comparison of the expected frequency range over which different system designs should be effective. Finally, the variable nature of torsional vibration frequency, as illustrated in example 5, means that if the feedback system requires manual re-tuning, the interface through which tuning is performed should be one that is simple and robust enough forrigsite use. A keyboard interface is, for example, more appropriate than bare, intricate circuitry.

EFFECTS ON DRILLING PERFORMANCE The effort to eradicate torsional vibration from drilling operations was initially focussed on a major development and appraisal drilling project in Colombia. Within that project, PDC bit life in particular was being detrimentally affected by torsional vibration. Figures 12 and 13 show, for a group of geographically adjacent wells in this development, trends in bit life for 17.1/2" sections. The figures indicate the longest bit run, and the average bit run length, in feet. Full implementation of the detection and suppression practices was made on the sixth and seventh wells. Considerable improvement in bit life andrun length has resulted. This was despite, in well number 1 and from well number 4 onwards, the section end depth being extended to include a harder, more sandy interval. These bit life improvements, the bulk of which are due to vibration suppression, have contributed to a reduction in bit-dependent cost per foot for these 17.1/2" sections from a weighted average of $316/foot for the first five wells, to $181/foot for the last two. This has produced savings of approximately $1.2 million on each of the last two 17.1/2" sections. This excludes savings due to reduced drillstring failures. Efforts are now being made to extend these detection capabilities and vibration suppression practices to other areas where drilling conditions are harsh.

C-8

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Stick-Slip

References

DRILLING IN HARSH ENVIRONMENTS This work has shown that torsional vibration causes drillstring failures and reduced PDC bit life. In areas where drilling conditions are rough, the costs associated with these problems can be significant. It is also clear that this form of vibration can be detected and suppressed at low cost and with minimal technical complexity. For drilling contractors and operators drilling in harsh environments, it is thus highly cost effective to monitor for torsional vibration, and to modify conventional drilling practices to suppress it. For mud logging contractors, the ability to detect and display torsional vibration provides a powerful means to contribute to improvements in bit life and drillstring failure avoidance. If a torque feedback system is in use, a permanent record of its efficiency will be provided by the vibration detection package.

Copyright 1994 SPE. Reprinted by permission.

For drill bit suppliers, practices for running PDC bits, and post-run analyses, should acknowledge the significant contribution that torsional vibration can make to bit life. Its presence should be routinely checked for in failure investigations.

CONCLUSIONS 1. The new vibration detection capability described here has shown that torsional vibration is a major cause of drillstring failures and reduced PDC bit life in areas where drilling conditions are harsh. 2. With improved vibration detection, torque feedback, and modified drilling practices, torsional vibration can be eradicated. This yields substantial benefit in reduced drilling costs. 3. The effectiveness of torque feedback systems for suppression of torsional vibration depends heavily on their ability to respond to changing vibration frequency, and to operate over a wide range of frequencies. Not all current systems are satisfactory by these criteria. 4.Torsional vibration is not confined to drilling with rotary bottomhole assemblies. It has also been detected while drilling with downhole motors in rotary mode, reaming, drilling cement, milling casing shoes, and rotating liners. It is a persistent problem whenever variable torque downhole is applied to a torsionally umber string of tubulars. 5. Systematic monitoring of vibration, and correlation with drilling problems, helps shift practices for dealing with drillstring failures from those based totally on equipment integrity and inspection, to include smoothing out drilling conditions downhole. This is more realistic given that downhole components of any quality and age will fail more rapidly when exposed to persistent vibration.

ACKNOWLEDGEMENTS The authors wish to thank BP Exploration, Total C.F.P., Triton Energy Corporation and Ecopetrol for permission to publish this paper. Thanks are also due to those field and technical support personnel within Geoservices and Baker Hughes Inteq who have assisted with system development, data collection, and vibration suppression.

Training Guide 80971 Rev A / June 1995

C-9 Confidential

References

Stick-Slip

REFERENCES 1. Brett, J.F., Warren, T.M., and Behr, S.: "Bit Whirl: A New Theory of PDC Bit Failure", SPE Drilling Engineering (December 1990) 275-281. 2. Brett, J,F., "The Genesis of Torsional Drillstring Vibrations", SPE Drilling Engineering (September 1992) 168-174. 3. Halsey, G.W., Kyllingstad, A., and Kylling, A., "Torque Feedback Used to Cure SlipStick Motion", paper 18049 presented at the 1988 Annual Technical Conference and Exhibition of the Society of Petroleum Engineers, Houston, Texas, October 2-5. 4. Warren, T.M., Brett, J.F., and Sinor, L.A., "Development of a Whirl-Resistant Bit", SPE Drilling Engineering (December 1990) 267-274. 5. Rewcastle, S.C., and Burgess, T.M., "Real-Time Downhole Shock Measurements Increase Drilling Efficiency and Improve MWD Reliability", paper 23890 presented at the 1992 IADC/SPE Drilling Conference, New Orleans, Louisiana, February 18-21.

Copyright 1994 SPE. Reprinted by permission.

6. Sananikone, P., Kamoshima, O., and White, D.B., "A Field Method for Controlling Drillstring Torsional Vibrations", paper number 23891 presented at the 1992 IADC/SPE Drilling Conference, New Orleans, Louisiana, February 1821. 7. Aldred, W. D. , and Sheppard, M.C., "Drillstring Vibrations: A New Generation Mechanism and Control Strategies", paper number 24582 presented at the 1992 Annual Technical Conference and Exhibition of the Society of Petroleum Engineers, Washington, DC, October 4-7. 8. Zannoni, S.A., Cheatham, C.A., Chen, D.C.K., and Golla, C.A., "Development and Field Testing of a New Downhole MWD Drillstring Dynamics Sensor", paper number 26341 presented at the 1993 Annual Technical Conference and Exhibition of the Society of Petroleum Engineers, Houston, Texas, October 3-6. 9. Macpherson, J.D., Mason, J.S., and Kingman, J.E.E., "Surface Measurement and Analysis of Drillstring Vibrations While Drilling", paper 25777 presented at the 1993 IADC/SPE Drilling Conference, Amsterdam, February 23-25. 10. Dawson, R., Lin, Y.Q., and Spanos, P.D., "Drill String Stick-Slip Oscillations", Spring Conference of the Society for Experimental Mechanics, Houston, Texas, June 14-19, 1987. 11. Kyllingstad, A., and Halsey, G.W., "A Study of Slip-Stick Motion of the Bit", paper number 16659 presented at the 1987 Annual Technical Conference and Exhibition of the Society of Petroleum Engineers, Dallas, Texas, September 27-30. 12. Dufoyte, M-P., and Henneuse, H., "Detection and Monitoring of the Slip-Stick Motion: Field Experiments", paper number 21945 presented at the 1991 IADC/SPE Drilling Conference, Amsterdam, March 11 -14. 13. Pastusek, P., Cooley, C., Anderson, M., and Sinor, L.A., "Directional and Stability Characteristics of Anti-Whirl Bits", paper number 24614 presented at the 1992 Annual Technical Conference and Exhibition of the Society of Petroleum Engineers, Washington, DC, October 4-7.

C-10

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

References

Copyright 1994 SPE. Reprinted by permission.

Stick-Slip

Training Guide 80971 Rev A / June 1995

C-11 Confidential

Stick-Slip

Copyright 1994 SPE. Reprinted by permission.

References

C-12

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

References

Copyright 1994 SPE. Reprinted by permission.

Stick-Slip

Training Guide 80971 Rev A / June 1995

C-13 Confidential

Stick-Slip

Copyright 1994 SPE. Reprinted by permission.

References

C-14

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

References

Copyright 1994 SPE. Reprinted by permission.

Stick-Slip

Training Guide 80971 Rev A / June 1995

C-15 Confidential

Copyright 1994 SPE. Reprinted by permission.

Copyright 1994 SPE. Reprinted by permission. References Stick-Slip

C-16

Baker Hughes INTEQ

Confidential

80971 Rev A / June 1995

References

Copyright 1994 SPE. Reprinted by permission.

Stick-Slip

Training Guide 80971 Rev A / June 1995

C-17 Confidential

Stick-Slip

Copyright 1994 SPE. Reprinted by permission.

References

C-18

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

References

Copyright 1994 SPE. Reprinted by permission.

Stick-Slip

Training Guide 80971 Rev A / June 1995

C-19 Confidential

References

Stick-Slip

•Notes•

C-20

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Appendix

D

Stick-Slip Examples Below are three examples os stick-slip occurrences, and the remedies

Example #1: Algorithm detectes stick slip

Remedy: A reduction in approximately 10,000 lbs WOB, eliminated the stick-slip

Training Guide 80971 Rev A / June 1995

D-1 Confidential

Stick-Slip Examples

Stick-Slip

Example #2: Algorithm detects stick-slip occurrence

Remedy: An increase in approximately 30 RPM eliminated the stick-slip

D-2

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Stick-Slip

Stick-Slip Examples

Example #3: Algorithm detects stick-slip occurrence

Remedy: An initial reduction in the rotary speed did not correct the problem, however an reduction in the weight-on-bit (shown by the increase in hookload) eliminated the stick-slip.

Training Guide 80971 Rev A / June 1995

D-3 Confidential

Stick-Slip Examples

Stick-Slip

•Notes•

D-4

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Stick-Slip

Stick-Slip Examples

Training Guide 80971 Rev A / June 1995

D-5 Confidential

Stick-Slip Examples

Stick-Slip

D-6

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Appendix

E

Recommended Stick-Slip Set Points The set points provided below are representative of BHA’s used in various parts of the world. These set points are the result of theoretical calculations and observations from field data.

To use these set points, chose a depth close to the present depth and then search for the BHA that most resembles the one in use:

Abbreviations Used: •

DP = Drill pipe

•

Hwdp = Heavyweight Drill pipe

Conversions: •

1 inch = 25.4 mm

•

1 foot = 0.3048 m Depth 2,500 Feet

BHA #1

Set Point #1

330 ft Collars (11-inch OD, 3-inch ID) 275 ft Hwdp (5-inch OD, 3-inch ID) 1895 ft Dp (5-inch OD, 4.276-inch ID) Torsional Period 0.254 Hz 3.937 sec

Torque Delta 10% Minimum Period Maximum Period Lag 5

BHA #2

Set Point #2

330 ft Collars (9.5-inch OD, 3-inch ID) 275 ft Hwdp (5-inch OD, 3-inch ID) 1895 ft Dp (5-inch OD, 4.276-inch ID) Torsional Period 0.333 Hz 3.003 sec

Torque Delta 10% Minimum Period Maximum Period Lag 5

Training Guide 80971 Rev A / June 1995

3.0 sec 16 sec

2.4 sec 12 sec

E-1 Confidential

Recommended Stick-Slip Set Points

Stick-Slip

BHA #3

Set Point #3

330 ft Collars (8-inch OD, 3-inch ID) 275 ft Hwdp (5-inch OD, 3-inch ID) 1895 ft Dp (5-inch OD, 4.276-inch ID) Torsional Period 0.449 Hz 2.227 sec

Torque Delta 10% Minimum Period Maximum Period Lag 5

BHA #4

Set Point #4

300 ft Collars (11-inch OD, 3-inch ID) 2200 ft Dp (6.625-inch OD, 4-inch ID)

Torque Delta 10% Minimum Period Maximum Period Lag 5

Torsional Period 0.556 Hz

1.799 sec

BHA #5

Set Point #5

300 ft Collars (9.5-inch OD, 3-inch ID) 2200 ft Dp (6.625-inch OD, 4-inch ID)

Torque Delta 10% Minimum Period Maximum Period Lag 5

Torsional Period 0.684 Hz

1.462 sec

BHA #6

Set Point #6

300 ft Collars (8-inch OD, 3-inch ID) 2200 ft Dp (6.625-inch OD, 4-inch ID)

Torque Delta 10% Minimum Period Maximum Period Lag 5

Torsional Period 0.825 Hz

1.212 sec

1.7 sec 9 sec

1.4 sec 7.5 sec

1.1 sec 6.0 sec

1.0 sec 5.0 sec

Depth 5,000 Feet BHA #7

Set Point #7

300 ft Collars (9.5-inch OD, 3-inch ID) 275 ft Hwdp (5-inch OD, 3-inch ID) 4425 ft Dp (5-inch OD, 4.276-inch ID) Torsional Period 0.222 Hz 4.505 sec

Torque Delta 10% Minimum Period Maximum Period 18 sec Lag 5

BHA #8

Set Point #8

300 ft Collars (8-inch OD, 3-inch ID) 275 ft Hwdp (5-inch OD, 3-inch ID) 4425 ft Dp (5-inch OD, 4.276-inch ID) Torsional Period 0.287 Hz 3.484 sec

Torque Delta 10% Minimum Period 2.7 sec Maximum Period 14 sec Lag 5

E-2

3.5 sec

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995

Stick-Slip

Recommended Stick-Slip Set Points

BHA #9

Set Point #9

300 ft Collars (9.5-inch OD, 3-inch ID) 4700 ft Dp (6.625-inch OD, 4-inch ID)

Torque Delta 10% Minimum Period 2.0 sec Maximum Period 10.0 sec Lag 5

Torsional Period 0.392 Hz

2.551 sec

BHA #10

Set Point #10

300 ft Collars (8-inch OD, 3-inch ID) 4700 ft Dp (6.625-inch OD, 4-inch ID)

Torque Delta 10% Minimum Period 1.8 sec Maximum Period 9.0 sec Lag 5

Torsional Period 0.441 Hz

2.268 sec

Depth 10,000 Feet Set Point #11

BHA #11 300 ft Collars (9.5-inch OD, 3-inch ID) 275 ft Hwdp (5-inch OD, 3-inch ID) 9425 ft Dp (5-inch OD, 4.276-inch ID) Torsional Period 0.173 Hz 5.780 sec

Torque Delta 10% Minimum Period Maximum Period Lag 5

BHA #12

Set Point #12

400 ft Collars (8-inch OD, 3-inch ID) 275 ft Hwdp (5-inch OD, 3-inch ID) 9325 ft Dp (5-inch OD, 4.276-inch ID) Torsional Period 0.162 Hz 6.173 sec

Torque Delta 10% Minimum Period Maximum Period Lag 5

BHA #13

Set Point #13

200 ft Collars (8-inch OD, 3-inch ID) 275 ft Hwdp (5-inch OD, 3-inch ID) 9525 ft Dp (5-inch OD, 4.276-inch ID) Torsional Period 0.189 Hz 5.291 sec

Torque Delta 10% Minimum Period Maximum Period Lag 5

BHA #14

Set Point #14

660 ft Collars (6.5-inch OD, 3-inch ID) 275 ft Hwdp (5-inch OD, 3-inch ID) 4425 ft Dp (5-inch OD, 4.276 ID) Torsional Period 0.184 Hz 5.435 sec

Torque Delta 10% Minimum Period Maximum Period Lag 5

BHA #15

Set Point #15

300 ft Collars (9.5-inch OD, 3-inch ID) 9700 ft Dp (6.625-inch OD, 4-inch ID)

Torque Delta 10% Minimum Period Maximum Period Lag 5

Torsional Period 0.214 Hz

4.673 sec

Training Guide 80971 Rev A / June 1995

4.6 sec 23 sec

5.0 sec 25 sec

4.2 sec 21 sec

4.3 sec 22 sec

3.7 sec 19 sec

E-3 Confidential

Recommended Stick-Slip Set Points

Stick-Slip

Depth 15,000 Feet

BHA #16

Set Point #16

400 ft Collars (6.5-inch OD, 2.8125-inch ID) 610 ft Hwdp (5-inch OD, 3-inch ID) 13960 ft Dp (5-inch OD, 4.276-inch ID) Torsional Period 0.137 Hz 7.299 sec

Torque Delta 10% Minimum Period 5.8 sec Maximum Period 29 sec Lag 5

BHA #17

Set Point #17

660 ft Collars (6.5-inch OD, 2.8125-inch ID) 460 ft Hwdp (5-inch OD, 3-inch ID) 13880 ft Dp (5-inch OD, 4.276-inch ID) Torsional Period 0.162 Hz 6.173 sec

Torque Delta 10% Minimum Period 6.2 sec Maximum Period 31 sec Lag 5

BHA #18

Depth 20,000 Feet Set Point #18

400 ft Collars (6.5-inch OD, 2.8125-inch ID) 640 ft Hwdp (5-inch OD, 3-inch ID) 18880 ft Dp (5-inch OD, 4.276-inch ID) Torsional Period 0.106 Hz 9.434 sec

Torque Delta 10% Minimum Period 7.5 sec Maximum Period 38 sec Lag 5

BHA #19

Set Point #19

660 ft Collars (6.5-inch OD, 2.8125-inch ID) 460 ft Hwdp (5-inch OD, 3-inch ID) 18880 ft Dp (5-inch OD, 4.276-inch ID) Torsional Period 0.101 Hz 9.901 sec

Torque Delta 10% Minimum Period 7.9 sec Maximum Period 39 sec Lag 5

E-4

Baker Hughes INTEQ Confidential

80971 Rev A / June 1995