automated mon

7/23/2019 Automated Mon

1/22

USMS

CI?E36L3

Aj t oinaeti

Monitorin6of CYciic Loading o n

Oilfield Drillstring

M.P . VW?,,

University College London;

T

PlcCann University College London;

MH

Fatel . U.niv.ersiby College London;

..............?..... .

G.J.

Lyons University College London;

A.D. Watkins

University Co:lege London;

mm~qqgm _QFPEIROLHNINzNEmS

This manuscript was provided to the Society of Petroleum Engineers for

distribution and possible publication in an SPE Journal. The Contents of this

paper (1) are subject to correctionby the author(s) and (2) have not undergone

SPE peer review for technical accuracy. Thus, SPE makes no claim about the

contents of the work. Permission to copyor use is restricted to an abstract of

not more then 300 words. Write SPE Book Order Dept.r Library Technician,

P.O.

Box 833836, Richardson TX 75083-3836 U.S.A.

Telex 163245 SPEUT,

Facsimile 214/952-9435


2/22

.,

. .


3/22

UNSOLICITED

296 Y

MAR41994

AUTOMATED MONITORING OF CYCLIC

LOADING ON OILFIELD DRILL STRINGS

M. A. VAZ, T. McCANN* , M. H. PATEL, G. J. LYONS & A.D. WATKINS

Santa Fe Laboratory for Offshore Engineering, Department of Mechanical Engineerin

University College London, Tornngton Place, London WCIE 7JE, UK

ABSTRACT

Demands

for

cost reduction in offshore hydrocarbon field developments

have driven drilling technology research in two areas. The first of these is

concerned with. using highly deviated drilling to access larger reservoir

voiumes from fewer platforms whereas the second aims at employing drill

pipes more efficiently

i.e. close to

their

fatigue

Iimits.

This

paper presents

the results of a work programme carried out at University College

London

aimed at producing hardware and software for automated monitoring of

cyclic loading of oilfield drill pipes.

The system consists of an automated

drill pipe tagging local monitoring and system analysis package to predict

measure and analyse cyclic loading in drill strings. Although no economic

analysis

has

been carried

out it is believed that drilling companies may be

able to reduce

operating risks

and maintenance costs significantly by

utilising such a system. The results achieved in

preliminary tests are

shown

to be promising and indicate that a working drill floor based system should

deliver substantial benefits in the long term.

Key words: Drill pipe tagging cyclic loading drill strings pipe deployment

fatigue damage cycle counting.

BPPTechnicalServicesLtd


4/22

SPE 86 8 ~

1. INTRODUCTION

Exploration and production of hydrocarbons particularly in the North Sea is

increasingly dependent on advances in drilling technology to achieve large

horizontal well deviations, greater drilling depths, and more efficient

drilling rates and bit utilisation.

A technique for automated in-service

monitoring of drill pipe deployment and service loading would improve

drill pipe utilisation and avoid costly down-hole failures. Such in-service

monitoring needs to be completely automated so as to run essentially

independently of the drilling crew and to use robust sensors, computers, and

software. lnformat.ion ~om the rn-onitoring system may either be used by

drilling engineers on a day-to-day basis or be archived for record purposes

for review and analysis at a later date.

The identification of drill pipe deployment in bore holes and an estimate of

the cyclic loading history to which a pipe segment is subjected offers

substantial operational advantages in four areas

1)

2

3

4

It permits identification of the precise location of each drill pipe

.,

section in the hole at all times.

This identification combined with a drill string analysis allows the

cyclic loading history of each drill pipe section to be identified and

monitored.

The cydlc loading histories can be used in a fatigue damage analysis to

identify the expended and remaining service life of each drill pipe

section.

Unique identification of each drill pipe segment will improve

inventory control as well as drill pipe maintenance and

refurbishment.


5/22

SPE2864 8 ~

The basic system requirements are to read and identify each section uniquely

as the pipe is run into the hole and to combine information to give each

pipe a life history within computer software. The monitoring system will

allow the drill pipe deployment to be precisely known and recorded.

Furthermore, the identification and removal of sections that may have

..----

expended most _of their service lives will reduce the risk of down-hole

failures and significantly improve their utilisation. This paper presents the

results of a work programme concerned with development and testing of a

prototype system for the automated in-service monitoring of drill pipe

deployment incorporating analysis techniques for pe service loading and

fatigue damage. The paper has been divided into two distinct sections - the

first describes development of the necessary hardware including

investigation into various methods of identifying and tagging drill pipe

segments. This has led to the choice of a passive coded semi-conductor cMp

inserted into the drill pipe which can be interrogated by a non-contacting

. ..

- -.

read head to provide unique identification numbers.

The second part

concerns the implementation of drill string analysis techniques to

determine the service loading that each drill pipe section is subjected to.

These service loadings can then be used to compute the fatigue life

expended by each drill pipe section as its operational life progresses.

2. SYSTEM HARDWARE

2.1 The Drilling Environment

For a drill pipe identification system tags or labels would have to withstand

an extremely severe down-hole environment. This includes high

temperatures and pressures, abrasion, excessive vibration, and impact loads.

Down-hole temperatures increase with depth, and typically range between


6/22

,.

. . ..- .

@86L+8

4

100C and 150C at the bottom of a deep well. The drilling fluid imposes a

considerable pressure.

The design environment for. the drill pipe was defined as follows

Physical

Chemical

Temperature

Pressure

Rough handling and impact loads from tongs, elevators,

cranes, tuggers, and kickback or jarring if the pipe gets stuck

down-hole. Strong abrasive forces from side contact and

returning mud. High frequency vibrations.

Drilling muds can be sea water, oil, or high chloride

polymer based.

-40C to 200C although it would be preferable if even

higher temperatures could be sustained.

15000 to 20000 psi at the bottom of a deep well.

Further points to be considered are thah

1

Wear at the tool joint during its life may reduce its diameter by 10 to 15

mm (most wear occurring at the box connection), F@ure 1.

2

During pipe joint refurbishment, layers of steel are laid down at

temperatures around 600C and any mark or tag would probably be

destroyed.

3

Drill pipes can magnetise during drilling because of the rotation in the

earths magnetic field.

The magnetism is removed during another

part of the refurbishing program when the pipes are passed through a

degaussing coil.

.


7/22

SPE28648

5

2.2 Hardware Alternatives

Viability studies were carried

out on five possible systems for identification

of drill pipe segments as they are tripped into and out of the hole. The main

features of each of these alternatives are described below

a) Manuak This is impractical as it would require an operator on the drill

floor to key in each sections code. Locating the marking where it is easily

readable yet protected from wear seems to be impracticable.

b) Optical: The

lines (barcode).

code might comprise a series of thick and thin contrasting

Using a light (possibly laser) detector a sensor could image

an identifying barcode. However, if the barcode should be obscured owing

to wear or maskkg by opaque fluids the sensor may fail to read the code.

c) Electrical conductivity A barcode with alternately spaced conductive and

insulating stripscould be inserted into the pipe or embedded in the surface.

A measurement of the sequential resistances across the set of strips could

then be combined to yield a unique identification code.

would have to be in surface contact with the drill pipe.

The sensor probe

d) Ultrasonic: A barcode plug composed of layers of materials with different

densities and thicknesses could be scanned using ultrasound. The

ultrasonic probe could scan through the coded plug and the beam reflected

from the layers of the plug would take different amounts of time to return

to the receiver, hence a unique code could be identified. However, to obtain

the desired resolution it would be necessary to grind a lens with its focus on


8/22

. ..

SPEZ864 8

6

the surface of the barcode plug and the reader would have to be kept at a

precise distance from the drill pipe.

e) RF Oscillation These devices are

generator can be fitted into a small

stable and inexpensive. A frequency

package and detected from a remote

position. Such systems contain a small coil which is exated by a second coil

forming the sensor (or read head). As the sensor sweeps the tag an induced

current powers the system and the identifying frequency is obtained. An

oscillator can be inserted into a 2 mm diameter cylinder and so it is possible

to form a cylindrical amplifier of 3 or 4 mm diameter.

Owing to the problems with obscuration and likely damage for close contact

reading methods a) to d) these are considered unsuitable. The RF oscillator

may be successful y deployed as discussed in the following.

2.3 A

Suitable Tag

An improvement on the simple single frequency RF oscillator is

commercially available from a number of sources (e.g. Hughes Technology).

These are known as Radio Frequency Identification Device (RFID).

Although this technology already exists some modifications would be

needed for operation down-hole.

These tags consist of a microchip, and a coil wrapped around a ferrite rod.

The microchip contains a memory section, which stores a unique

. .

. .

identification code, and further sections to process and transmit this code.

The coil acts both as power supply (through electromagnetic induction) and

aerial to transmit the code back t; the receiver. fiese tags &e a-vailable in a


9/22

SPE28648 ,

variety of shapes and sizes. Laboratory tests have been performed with these

but further operational trials down-hole have yet to be carried out.

The tag should be located in a hole as shown in Figure 2. Investigations

were carried out at UCL to determine if the tags could be read when bonded

within a large volume of steel.

The Hughes TX1200 slug tag was used to

establish the relationship between depth of embedment in the steel with

read dMance. The relationship appeared to be almost linear with reasonable

read distances being obtained for quite deep holes (the deepest hole tested

was 17 mm giving 10.5 mm of protective cover). It is envisaged that the

system utilises a tag scanner that surrounds the whole pipe. This is a set up

which has successfully been tested by Hughes and should allow the tag to be

read at the Klghest drilling rotation rate.

2.4 Location of Drill Floor Equipment

As a consequence of detailed

semisubmersible Santa Fe Rig

and reader has evolved. The

investigations (in particular on the drilling

135) a system of separate antenna, scanner,

scanner and reader should be housed in an

explosion-proof box and placed either in the doghouse or to the side on the

drill floor. Theantenna (or read head) can be located on the diverter packer

insert which has a large rubber annulus with an internal diameter of 10 in

(25.4 cm) and an external diameter of about 26 in (66.0 cm) located at the top

of the casing with a considerable clearance tlom the underside of the drill

floor. The antenna being embedded in a solid block of impermeable plastic

formed into an annulus with a radius to match the packer insert. The

diverter packer ring is sandwiched between two steel plates to which the

antenna unit is bolted using a quick release system (see Figure 3). Hence, as


10/22

SPE 86

8 ~

the packer is removed from the hole the antenna could be disconnected and

placed in a protective housing beside the well opening.

2.5 Deployment Issues

Some issues that still have to be evaluated d~ing implementation of the

system within a specific drilling programme include

Temperature effects

.

.

.

Although the tag microchip can withstand 300C and the copper in the coil

.

could survive a temperature of 200C, the coil insulating material, and

connection solder need to be chosen to accommodate these temperatures. A

suitable compound for sealing the tags in the holes is Dow Corning RTV

1345, a proven material used for MWD equipment.

Intrinsic safety

The tag reader may need to be certified intrinsically safe for location under

the drill floor. However, provided only the antenna was installed on the

drill floor, (with the other electronics located in the doghouse or logging

cabin) this intrinsic safety requirement could be satisfied.

Magnetisation and demagnetisation

Tests need to be carried out to confirm that magnetic fields induced on the

tag by degaussing during pipe refurbishment do not have any adverse

effects.


11/22

SPE2864(j ,

Pipe refurbishment

The temperature changes and mechanical shocks imposed

during refurbishment need to be evaluated to determine if

on the pipe

the tags can

reliably survive these. If the tags cannot survive such treatment then tag

replacement would have to become part of the refurbishment process.

3. SYSTEM SOFTWARE

3.1 Loading and Fatigue

In directional drilling most drill pipes fail owing to the accumulation of

fatigue damage. A typical drill string failure in deep wells may cost tens of

thousands of pounds, therefore it is necessary to minimise such occurrences

(Joosten, Shute and Ferguson).

-Fatigue damage is known to originate in

dog-legs (kick-off points), caused by the rotation of bent pipes under high

tensile forces, although cyclic loading ind_ucedby vibration also plays a part.

Furthermore, dynamic stresses arise, for instance, from setting slips

(Vreeland2) or when the drilling vessel is subjected to sea-wave induced

motions (Hansford and Lubinskis). In fact, an accurate understanding of the

cyclic loading mechanism in drill strings has still to be developed. Here

Hansford and Lubinskisd method has been adopted to estimate the

accumulation of fatigue damage in gradual and long dog-legs.

The

following parameters, assumed constant over a 30 ft hole segment, are

considered to influence drill pipe life rotary speed, rate of penetration, drill

pipe outside diameter, tensile force, dog-leg severity, mud and formation

specification (corrosive or not), density of mud and grade of steel.


12/22

PE28648 ,0

A drill pipe may fail under repeated cycles of relatively low stress levels

owing to the growth of internal cracks that can be initiated from local

fabrication or at stress concentration sites.

Such fatigue damages are

quantified from experimental S-N curves which are plots of stress range S

against number of cycles to failure N.

The l?almgren-Miner cumulative

fatigue damage rule quantifies the fatigue damage caused by a number of

cycles at low stress ranges. This linear damage law assumes that the total

fatigue damage results from the summation of the fractions (or percentages)

of life expended in each stress cycle.

When the total accumulated fatigue

damage approaches unity (or 100 ) the drill pipe section should be removed

from service to avoid unexpected failure.

It is further assumed that the averages of the rotary speed, rate of

penetration, inclination and weight on bit (and tensile force) are known. In

conventional -oil rig: these parameters are recorded in the drilling sheet

while in modern rigs measurement while drilling (MWD) systems provide

these data on demand.

3.2 Simulation

.

This section presents a software system that works with the drill pipe

tagging hardware to count pipe cyclic loading and resultant accumulated

fatigue damage. In_ order to illustrate the software and computational

procedures involved, an example simulation of the operation of the

software is presented below.

.

Ideally the-software will operate with the tagging system hardware on a real

time basis and with the minimum of human interference.

Sensors


13/22

SPE2864 6

11

monitoring mud pumps, rotary speed, weight on bit and bit trajectory will

send signals to the computer - it will interpret thk data and activate the

corresponding modes.

The software simulation presented here is designed to be used for drilling

programme design or with the tagging system for on-site monitoring or

post-drilling analysis. It permits cyclic loading to be calculated every time a

drill pipe section is used or a few times a day using drilling sheet or MWD

data.

There are two basic steps in the simulation.

(1)

Data Preparation - The first step is to generate a file containing the

geometric and material properties of the pipes with the tag numbers

and accumulated fatigue darnage of every drill pipe section. Here it is

assumed that APF approved drill pipes made of Grade E steel are used

with outside (inside) diameters equal to 3.5 (2.992), 4.5 (3.953) and

5 (4.408). The initial fatigue damage is taken to be nil. The initial

hole geometry is generated by subdividing the well into vertical,

curved and straight inclined portions. A geometry file contains

inclinations and curvatures of every 30 ft (9.14 m) so as to define the

hole profile. Next, the sequence of tag numbers, corresponding to a

trip in, is produced. The generation of the pipe sequence is simplified

in the example simulation the tag number of the upper most pipe is

entered and alinearly decreasing distribution is used.

(2)

Drilling - When thk mode is selected, the initial geometry and pipe

sequence are read. Every time a new pipe section is used, its

respective tag number is read and the averages of rotary speed, rate of


14/22

SPE28648

12

penetration, weight on bit, and bit inclination are read in. The

curvature of the new hole segment, the cyclic loading, and fatigue

damage are calculated, and files updated. If drilling continues a new

tag number is read and the process repeats. This part of the program

is outlined in the flow chart of Figure 4.

3.3 Examples

Examples I and 2 below illustrate the importance of a dog-leg control

programme to minimise fatigue damage.

The two hole profiles of Figure 5

have sligthly different profiles but they reach the same final target zone.

Well number 1 has a dog-leg 600 ft long and severity of 2,50 deg/100 ft while

the dog-leg in well number 2 has half the severity but double the length.

Both drill string arrangements 1 and 2, pictured in Figures 6a and 6b,

comprise ten drill. collars_ with outside (inside) diameters of 7 (1.5) and

distributed dry weight of 125 lbf /ft. Mud is assumed corrosive and its

density is 12.0 1~/gal. Five pipes are used in each drilling session with the

drilling parameters shown in Table 1.

session

IROP (ft/hr) RPM (rpm)

WOB (lbf) Inclination (deg)

I

100

100

30000

15

; 50

100

25000 15

Table 1- Drilling

Sessions

Example 1- The drill string arrangement 1 (Figure 6a) is run into holes 1 and

2, Drill pipes 351 to 355 and 356 to 360 were respectively tripped in sessions

a and b. Table 2 presents the accumulated fatigue damage, after the

trip

in, for the affected drill pipes. The average fatigue damage, calculated over

the drill pipe sections fatigued, is 3.1 and 6.5 for arrangement 1 in holes 1


15/22

SPL28648

13

and 2, respectively. In longer and less severe dog-legs more drill pipes are

damaged but the fatigue damage is more evenly distributed between the

affected drill pipes and the rate of accumulation for an individual segment

is lower.

Hole 1

Hole2

Tag numbers

Fatigue ( )

Tag numbers

Fatigue ( )

292

0.3

297 0-6

293

0.6

298 1-3

294

0.9

299

1.9

295

1.2

300

2.6

296

1.5

301 3.3

297

2.1

302 5.0

298 2.7 303 6.7

299 -- 3.3

304

8.5

300 3.9

305

10.3

301 to 311

4.5 306

12.2

312

4.2 307

11-7

313

3.9 308

11.1

314

3.6 309

10.5

315

3.3

310 9.9

--

316 3.0

311

9.3

317

2.4 312

7.5

318

1.8 313

5.7

319

1.2 314

3.9

320

0.6

315

2.0

Table 2- AccumulatedFatigueDamage

in

TwoHoleProfiles

Example 2- A heavier drill string arrangement 2 (Figure 6b) is run into well

number 2. In drilling sessions a and b the drill pipes 591 to 600 were

tripped in.

Table 3 summarises the results.

The average fatigue damage

using arrangement 2 in hole 2 is

7.7~0

per affected pipe. The fatigue damage

is more severe owing to the higher @al forces in arrangement 2.


16/22

pE~8648

14

Hole 2 / Arrangement 2

Tag numbers

Fatigue ( ) Tag numbers Fatigue ( )

537 0.8 547 13.8

538

1.5

548 13.1

539

2.4

549 12.4

540

3.2 550 11.6

541

4.0

551

10.9

542 6.0

552 8.8

543 8.0

553

6.7

544 10.1

554 4.5

545

123 555

2.3

546

14.5

.-

Table 3- AccumulatedFatigueDamagein Heavier Drill String

3.4 Extensions

Whilst the software developed to date adequately demonstrates the salient

features required for the major aspects of drill string fatigue there are several

software enhancements that could be readily made to this. These are listed

below.

1

Incorporate a wider range of drill pipes in the pipe file.

2) Define hole profiles by three-dimensional curves (hence the dog-leg

severities should consider the overall hole change).

3) Read tag numbers and drilling parameters from files.

4)

Calculate the footage drilled by each pipe, so that this may be

associated with pipe wear.

5)

Record inspections and repairs in the pipe data file.

6) Consider contact and friction forces when calculating axial forces.

7

Develop methodology to calculate fatigue for dog-leg severities

exceeding 10 deg/100 ft

becomes more common.

as drilling with short radii of curvature


17/22

SPE2864

15

4. CONCLUSIONS

The foregoing describes the hardware and software development for

automated monitoring of cyclic loading and accumulated fatigue damage on

oilfield drill strings. The hardware is capable of uniquely identifying drill

pipe joints whilst they are being tripped in and out of the hole - the

associated software then estimates the cyclic loading and both the expended

and remaining service lives of the pipe segments. Some further system

detail work remains to be done although the details of this are to some

extent dependent on the drilling sites on which the system will be deployed.

Following such deployments, assessments of the reliability and accuracy of

resultant fatigue damage estimates will be made.

..

ACKNOWLEDGEMENTS

This work was carried out with the support of the

facilities of the Santa Fe

Laboratory for Offshore Engineering, University College London and

assistance from Santa Fe Drilling Co (North Sea) Ltd, and Hughes

Technology Ltd. The first author acknowledges the support of the Brazilian

Council of Research (CNPq) for ~is work.

REFERENCES

1

2

Joosten, M. W., Shute, J. and Ferguson, R. A., New

Study Shows How to

Predict Accumulated Drill Pipe Fatigue

World 011, (Oct. 1985), 65-70.

Vreeland, T., Jr., -

Dynamic Stresses in Long Drill Pipe

Strings, The

Petroleum Engineer, (May 1961), B58-B60.


18/22

SPE 86 8

16

3 Hans ford, ]. E. and Lubinski, A., Effects of Drilling Vessel Pitch or Roll on

Kelly and Drill Pipe Fatigue Journal of Petroleum Technology, (Jan.

1964), 77-86.

4 Hansford, J. E. and Lubinski, A., Cumulative Fatigue Damage of Drill

Pipe in Dog-Legs Journal of Petroleum Technology, (March 1966), 359-

. .. . . .

--- -

363. - - -

5 American Petroleum Institute, Recommended Practice for Drill Stem

Design and Operating Limits (API RP 7G), (Aug. 1, 1990).


19/22

vortexoccurs in

tldsregion

\

~xcess

vewearcccu

in thisregion

pipeB

11~

II

II

II

II

II

\

II

\ll

\l

,/

\; ;/

1-l

11

II

II

I

II

I

II

SPE28648

17

pinjointpipeB

box joint pipeA

PipeA

~lgure 1- FlowDiagram around Tool Joints


20/22

Holeperpendicular to

m

Holeperpendictdar to

35 deg weld

pipe axis

Q

d

.,

II

1[

l?@Ire 2-

Hole in Weld

,.

/

Metal

Top Plate

CircularAntenna

Packer Insert

Figure 3- PackerInsertShowingPropusedAntennaPosition


21/22

.-

SpL28648

INITIAL GEOMETRY

READ initial

IDENTIFY

geometry file

*pm

IDENTIFY pipes k dog-legs and

CALCULATE drill string weight below

1

each drill pipe

STAKTDRILLING

tag number

+

STOP DRILLING

INPUT ROP, RPM,

n

WRITE

WOB and - - > ri]ing

INCLINATION

parameters

~

CALCLJLATE UPDATEWRITE pipe

fatigue

UPDATE drill

CHECK curvature

I

below each pipe

19

~@re 4- FlowChart of Simulation


22/22

SPE28648

20

Well No 1

tn I

. . .

. . .

. . .

. . .

. . .

. . .

. ..

. . .

. . .

40 vertical segments

-----

I l-::.i

1[

...

..

:::

,/

,,

,/

,/

= ;;{

o

u-

:::

h

:::

:::

:::

:::

:::

:::

:::

,,,

::.

,::

:,:

:::

:::

:::

Initial

. . . . . . . . .

position

Figure 5- HoleProfiles

1.

Drill pipes

(3D=4.5

::;

360 to 351 IQ=3.g5e

:.

.; Drill pipe 350

E&

0 drill collars ~==1~5.

Initial

---------

Well No 2

El

95 straight

inclined segments

inclination

15.0 de rees

:,:

j

~Drill pipes oo ~

5.

j ; 600 to 591 ID=4-m8

j { Drill pipe 590

::

1

1

: ,

: :,

: ;;

: :

: ; 190

pipes~DD==45408-

1

::1

:::

: :,

::1

;;:

:::

::,

::,

:

X

OD= 4.5-

; 50 pipes ID = 3.95fy

:

,, t

::

:

~ Drill pipe 351

:;

Drill String Arrangement 1

Drill String Arrangement 2

Figure 6a

Figure 6b

automated mon

Documents