Note: Descriptions are shown in the official language in which they were submitted.
CA 02268446 1999-04-12
WO 98I16842 PCT/US97/18759
-1-
COILED TUBING INSPECTION SYSTEM
Field of the Invention
The present invention relates to an inspection system for the non-destructive
inspection of oilfield tubulars. More particularly, the present invention
relates to an improved
inspection system for the non-destructive testing of coiled tubing in a rapid
and reliable
manner. The system of the present invention is intended to simultaneously test
for various
types of defects in coiled tubing, including wall thickness defects, material
flaws, longitudinal
splits, and tubing diameter and ovality.
Background of the Invention
Various types of systems have been devised for inspecting oilfield tubulars.
U.S.
Patent No. 5,600,069 discloses an improved system for the ultrasonic testing
for oilfleld
tubulars having various diameters. Oilfleld tubular non-destructive testing
equipment ideally
satisfies several basic objectives. The system should be operator friendly and
capable of being
used by various inspection personnel. An inspection system also is preferably
compact and
may be both manufactured and maintained at a comparatively low cost. Most
importantly,
the system should be accurate in detecting various types of flaws, and should
maintain that
accuracy when oilfield tubulars are passing at a high rate through the system.
Non-destructive inspection systems primarily intended for testing coiled
tubing present
unique challenges. Those skilled in the oil patch appreciate that coiled
tubing is
distinguishable from conventional oilfield pipe, casing, and tubing in that
coiled tubing is
wound on a drum at the surface then passed downhole. Although lengths of
coiled tubing
may be joined by various types of connections, a length of coiled tubing is
typically thousands
of feet, rather than being 30 feet in length which is conventional for "rigid"
oilfield tubulars
with threaded connections. Since coiled tubing is inherently flexible, it has
significant
advantages over conventional threaded end oilfield tubulars, including
relatively low cost and
reduced time to pass the tubing downhole. This flexibility for coiled tubing
presents unique
CA 02268446 1999-04-12
WO 98/16842 PCTIIiS97/18759
-2-
inspection problems since the relatively thin wall of the coiled tubing must
meet pre-
determined specifications, and since the tubing diameter and ovality must be
within selected
limits to minimize the likelihood of a coiled tubing ruptures. Coiled tubing
may undesirably
tend to elongate when an excessive axial force is applied to the coiled
tubing, thereby
reducing the wall thickness and the tubing diameter. Also, oilfield tubing
tends to experience
longitudinal splits at a rate significantly in excess of longitudinal splits
in conventional oilfield
tubing with threaded connector ends.
Several coiled tubing inspection systems have been devised, but none of these
systems
test for a11 of the defects commonly associated with coiled tubing, and most
of these prior art
systems are not highly accurate at detecting defects in coiled tubing. An
inspection system
manufactured by Rosen Inspection Technologies in Lingen, Germany tests for
coiled tubing
wall thickness and material flaws, but the system is bulky and expensive.
Moreover, the
Rosen inspection system does not test for tubing diameter and ovality, nor is
the system well
equipped to detect longitudinal splits in coiled tubing. A separate unit
downstream from the
Rosen flaw detection inspection system is required to measure the length of
the tubing being
inspected. Other prior art coiled tubing inspection systems, including a
system manufactured
by Stylwan Inspection in Houston, Texas, use a relatively small number of
detectors, and are
even less accurate at detecting most coiled tubing defects. Inspection systems
that are
specifically intended to test for coiled tubing diameter or ovality have been
devised, but these
systems do not test for various other types of defects in coiled tubing,
including wall
thickness, flaws, and longitudinal splits.
The disadvantages of the prior art are overcome by the present invention, and
an
improved coiled tubing inspection system is hereinafter disclosed. The
inspection system is
particularly designed for simultaneously testing for various types of defects
in coiled tubing,
and has a desired high accuracy at detecting those defects when coiled tubing
is passed at a
high rate through the inspection system.
CA 02268446 1999-04-12
WO 98I16842 PCT/US97/18759
-3-
Summary of the Invention
The coiled tubing inspection system comprises a cylindrical housing with six
centralizers and an electronics package housed in an aluminum case. Three
centralizers
circumferentially spaced at 120 ~ are thus provided at the input to the
housing and three
similarly positioned centralizers are provided at the output to the housing. A
plurality of
transducers are housed on an inspection head positioned within the housing. A
conductor
cable connects the head to the electronic package in the aluminum case.
The cylindrical housing contains a magnetizing coil designed to bring to
magnetic
saturation the coiled tubing being passed through the housing. A diameter
specific inspection
head or cartridge, inserted within the housing, contains transducer arrays to
detect total
magnetic flux, magnetic flux leakage, and eddy current perturbations. The
inspection head
also houses four (4) eddy current (or piezoelectric) transducers that are
positioned at 90 ~ in
order to measure the distance to the outer surface of the coiled tubing being
inserted into the
housing.
The aluminum case houses three computers: ( 1 ) a wall loss computer which
calculates in tube quadratures the tubing wall thickness; (2) a
diameter/ovality computer
which determines the diameters of the tubing in two orthogonal planes and also
determines
a tubing off center condition; and (3) a data acquisition streaming computer
which displays
a chart on a flat panel, with ports for a mouse, a printer and a keyboard. The
case also houses
flaw detecting amplifiers, differential eddy current longitudinal split
electronics, a pulse width
modulated constant current mag supply, a preset encoder/counter to monitor
coil tubing
length, and all electronic power supplies. The case is powered by a 120V AC
source or
optionally from other sources.
It is an object of the present invention to provide an improved non-
destructive testing
system particularly intended for testing oilfield coiled tubing. The system
simultaneously
detects for various types of defects commonly associated with coiled tubing,
including
material flaws, wall thickness variations, longitudinal splits, and diameter
and ovality
irregularities.
CA 02268446 1999-04-12
WO 98/16842 PCT/IJS97118759
-4-
A related object of the invention is a coiled tubing inspection system which
is highly
reliable at detecting defects in coiled tubing, even when the coiled tubing is
passed at a rate
of from 30 feet per minute (9 meters per minute) to 180 feet per minute (55
meters per
minute) through the system.
Yet another object of this invention is a coiled tubing inspection system
which is
relatively compact and may be manufactured and maintained at a relatively low
cost.
A feature of the invention is that the various detectors do not contact the
outer surface
of the coiled tubing, thereby facilitating a high feed rate through the
inspection system.
Another feature of the present invention is a coiled tubing inspection system
which
calculates a diameter of the coiled tubing along both an x-axis and a y-axis
which are
perpendicular, and then provides a calculation for determining the variation
between the major
axis and the minor axis of the coiled tubing in order to determine tubing
ovality.
Still another feature of the present invention is that longitudinal splits in
coiled tubing
are detected with transducers which operate at a high frequency, thereby
allowing for high
sensitivity to this type of defect. Two detectors for testing longitudinal
splits are axially
spaced, and the difference in their outputs is indicative of an end of a
longitudinal split in the
coiled tubing.
It is a significant advantage of the present invention that the coiled tubing
inspection
system is relatively compact and may be easily moved to various coiled tubing
inspection
sites.
Another advantage of the coiled tubing inspection system is that the system is
highly
automated and may be reliably used by various non-destructive testing
personnel.
These and further objects, features, and advantages of the present invention
will
become apparent from the following detailed description, wherein reference is
made to the
figures in the accompanying drawings.
CA 02268446 1999-04-12
WO 98/16842 PCT/US97/18759
-S-
Brief Description of the Drawing
Figure I is a cross-sectional view of a coiled tubing inspection cylinder and
mechanical
centraiizers according to the present invention for centralizing coiled tubing
passing through
the housing during an inspection operation.
Figure 2 is a detailed view, partially in cross-section, of the cylinder
housing and the
replaceable inspection head cartridge generally shown in Fig. 1.
Figure 3 is an end view of the inspection head and the mechanical centralizers
shown
in Fig. I .
Figure 4 is a detailed view of one of the centralizers shown in Figs. 1 and 3.
Figure 5 is a pictorial view of the inspection case used with the inspection
head.
Figure 6 is an end view of a diameter/ovality sensor arrangement according to
the
present mvent~on.
Figure 7 illustrates a flow detector transducer with a conical flux
concentrator above
a transducer.
1 S Figure 8 is a block diagram of the typical system components for the
inspection head
and electronic case for the system according to the present invention.
Figure 9 is a test circuit for the system according to this invention.
Figure 10 is a suitable detector circuit for sensing a split in the coiled
tubing.
Figure 1 I is a junction circuit illustrating the preamp box, and the
cartridge housing
the flaw, wall thickness and split tubing detectors. The diameter and ovality
detectors and
the magnetizing coil which functionally need not pass but conveniently do pass
through the
junction box are also shown in Fig. I 1.
Figure 12 is an elongated view of the circuit for detecting flaws, wall
thickness, and
tubing splits as generally shown in Fig. 1 I . A cartridge housing each of the
detectors is also
pictorially illustrated.
Figure 13 is a circuit illustrating the wiring from a flaw detection board.
CA 02268446 1999-04-12
WO 98I16842 PCT/US97/18759
-6-
Detailed Description of the Preferred Embodiment
The coiled tubing inspection system 10 comprises cylindrical housing 12 with
roll
centralizers 14 each at the input end 16 and the output end 18 of the
cylinder, and an
electronics package housed in an aluminum case 90 shown in Fig. 5. A plurality
of
S transducers housed within an inspection head 30 are provided within the
cylinder 12, as
explained subsequently. End plates 24 and 26 secure the inspection head 30
therein with
conventional screws 28, as shown in Fig. 3. A conductor cable 98 as shown in
Fig. 3 leads
to junction box 20 to connect the cylinder to the case 90 shown in Fig. 5. The
cable both
supplies power to the inspection head 30 and transmits signals from the head
to the case 90.
Alternatively, one cable may be used to supply power to the inspection head
30, and one or
more other cables may connect the transducers in the head 30 to the electronic
package in the
case 90. Less desirably, wireless FM technology could be used to transmit
transducer signals
to the case 90.
Each of the centralizers 14 is mounted on a centralizes base 15, as shown in
Figs. 1
and 4. A coil spring 17 biases roller block 19 toward the coiled tubing, and
is shown in Fig.
1 in its maximum diameter configuration for cooperating with an inspection
head
corresponding to the largest diameter coiled tubing to be inspected. Roller 21
is rotatably
mounted on the block 19 for engaging the coiled tubing. Each roller 21
includes a v-shaped
guide surface for rotating engagement with the coiled tubing while guiding the
coiled tubing
within the inspection head 30. A lead screw 23 provides for axial adjustment
of the roller 21
relative to the housing 12. Figure 3 illustrates the upper centralizes 14 in
its radially
innermost position for guiding a small diameter tubing, and the lower
centralizes 14 in its
radially outermost position for guiding a large diameter tubing. Markings on
the base 15
provide a visual scale 100 for observing the position of the roller to guide a
specific diameter
tubing. Alternatively, specifically sized and non-adjustable rollers may be
used, with the
rollers sized for and replaced with the specific diameter inspection head.
The cylinder 12 houses the inspection head 30 shown in Fig. 2, which in turn
houses
a magnetizing coil 32 between flanges 34 and 36. The magnetizing coil 32 is
designed to
CA 02268446 1999-04-12
WO 98I16842 PCT/US97/18759
_7_
bring to magnetic saturation the coiled tubing being passed through the
housing 12. A
diameter specific cartridge 36, mounted within the inspection head 30,
contains transducer
arrays to detect magnetic flux leakage {flaw detection), changes in total
magnetic flux (wall
loss detection), and eddy current perturbations {longitudinal split
detection). The head 30
also houses four (4) eddy current transducers that are positioned at 90 ~ in
order to measure
the distance from each transducer to the coiled tubing being inserted into the
cylinder, and
thus serve as diameter and ovality detectors.
Transverse and Three Dimensional Flaw Detectors
Four arrays of Hall Effect transducers 40 are positioned around the coiled
tubing to
be inspected, and are mounted within the diameter specific cartridge 36. A
different cartridge
36 is thus used for each different nominal diameter tubular to be tested. The
transducers are
positioned to detect flux leakage and are spaced from about 0.15 inches to
0.25 inches from
the external surface of the incoming coiled tubing. For the smallest diameter
coiled tubing
to be inspected (3/4" nominal diameter tubing), each array comprises six
transducers for
inspecting a 90 ~ circumference of the tubing. For inspecting 2" nominal
diameter tubing, each
of the four arrays for inspecting a 90 ~ circumference of the tubing comprises
fifteen
transducers. A cartridge for the largest diameter coiled tubing would include
more
transducers in each array. Each transducer array thus preferably includes a
plurality of
transducers circumferentially arranged to cover approximately 90 ~ of the
coiled tubing. At
least twenty-four transducers are thus circumferentially spaced about the
central bore in the
cartridge 36 which is sized to receive a specific diameter coiled tubing. Each
90~ array of
flaw detectors thus detects transverse and three dimensional flaws in the
material of the
respective quadrant of coiled tubing adjacent that array.
Allegro 3 S07 linear Hall Effect transducer was selected due to the higher
sensitivity,
scanning speed tolerance, high temperature stability, built in amplifier and
voltage regulator.
This transducer exhibits a typical 2.5 mv/Gauss sensitivity, although even
higher sensitivity
in excess of 4.0 mv/Gauss has been achieved. Each transducer is also aided by
a conical flux
CA 02268446 1999-04-12
WO 98I16842 PCT/US97118759
_g_
concentrator 44 as shown in Fig. 7 that enhances detection and permits the
liftoff (radial
spacing) between the transducer and the tubing. The concentrator is fabricated
from a
magnetizing material, such as soft iron, and effectively makes the entire area
of the transducer
package (0.180 inches x .180 inches) a sensitive detector, rather than merely
the 0.025 inches
x 0.025 inches active element area of each transducer. The concentration 44 is
thus spaced
radially outward of each plate-like transducer 40. Concentrator 44 is
optional, and
transducers 40 each without a concentrator may be used for various
applications.
Flux leakage around the Hall Effect transducer causes a change in the
transducer
voltage. Each array group is formed with matched transducers, with each
transducer in the
group wired in such a way that only the largest flux leakage indication for
the array group
dominates at any specific time, although that largest flux leakage signal from
each array may
be represented as either a positive signal or a negative signal. The
magnetizing coil is
polarized so that the flux leakage from the flaws generate a negative going
voltage. The
largest negative going voltage from the array dominates at a specific time.
Sensed flux
leakage from transducers in an array which do not sense the largest leakage
value may thus
be discarded by the computer.
The output from the detectors 40 are coupled via an additional 10 dB
preamplifier to
adjustable gain amplifiers located at the case 90. The preamplifier reduces
the susceptibility
to possible noise interference. The four amplifiers are combined to generate
the largest
positive indication and the largest negative indication at any instance. All
signals are set to
produce a positive going indication from the largest signal on the first
trace, and a negative
going indication from the second largest signal on the second trace. In other
words, a
positive flux leakage signal from one array may be combined with a negative
flux leakage
from a radially opposing array to provide a more accurate indication of a
flaw. Furthermore,
a linear reject circuit allows removal of small noise indications without
affecting any signal
exceeding the linear reject threshold. Individual switches permit
enabling/disabling of any of
the four (4) circuits.
CA 02268446 1999-04-12
WO 98/16842 PCT/US97/18759
-9-
As shown in Fig. 2, the plate-like transducers 40 are positioned at
substantially a right
angle to the plate-like transducer 50 discussed subsequently.
Wall Loss Detection and Alarms
The total magnetic flux allows non-contact determination of ferromagnetic
materials
type and their mass within a dual winding arrangement based on the Rowland
Ring
experiment performed over 120 years ago. Four arrays of Hall Effect
transducers 50 each
constituting 1 /4 of the circumference are positioned circumferentially around
the tubing to be
inspected. Each transducer 50 is turned 90~ relative to the transducers 40. By
positioning
the flaw detector transducers and wall loss transducers at 90~ from each
other, a detector for
sensing both radial and axial fields from the inspected tubing is obtained. A
combination of
wall and flaw information facilitates acceptance/rejection determination. If a
flaw is located
in a thick segment, it may be accepted. If a flaw is located in a thin
segment, it may render
it rejected. The computer may thus receive signals from both the flaw detector
transducers
and the wall thickness transducers and compare those signals to determine if a
sensed flaw
should result in a rejection signal.
The number of transducers 50 within each array will depend on the size of the
coiled
tubing being inspected, although six or more transducers 50 are preferably
provided in each
array. The four arrays can collect the total magnetic flux information. Wall
thickness
detection is thus not obtained in a conventional manner by detecting flux
leakage, but rather
is measured by detecting the reduction in the total magnetic flux density
caused by the tubing,
which is proportional to the wall thickness of the tubing. The thicker the
tubing, the greater
the reduction in magnetic flux density sensed by the transducers 50, which are
spaced radially
between the tubing and the magnetizing coil 32. When the transducers 50 are
coupled to a
computer, numerical values can be calculated to ascertain the thickness of the
tube adjacent
each array if other magnetic characteristics are held constant.
The computer is programmed to read the value of each quadrature array, filter
out
unqualified values and display the expected thickness of the tube in each
quadrant based on
CA 02268446 1999-04-12
WO 98/16842 PCT/US97/18759
- 10-
the sum of the presumed valid signals from the transducers in each array. The
computer also
calculates deviation from expected nominal thickness valve. An analog voltage
that is
proportional to the difference between the largest wall reading and the
nominal is output as
"+WALL", and an analog voltage that is proportional to the difference between
the nominal
and the smallest wall reading is output as "-WALL".
The computer is equipped with an alarm circuit if a wall reduction in any
portion of
the tubing corresponding to an array exceeds a predetermined amount. The small
8 bit
computer (Z 180) goes through a complete reading and calculating cycle in
1/35th of a
second. A "calculation" includes gathering four (4) wall indications,
processing (filter and
BOXCAR) and scaling, comparing to an alarm level and, when appropriate,
triggering alarms.
There are two modes of operation: ( 1 ) If the display is to be updated, the
computer performs
at a rate of 14 "calculations" per second; (2) If the display is not to be
updated, the computer
performs at a rate of 3 5 "calculations" per second. This computer features
battery backup
to calibration values and a complete linearization routine (Y = mX + C) for
each read input.
(Y is output, X is the input, m is the linearization slope, and C is offset
value). Mathematical
calculations and integration formulas are standard and may be obtained from
prior art. Other
features include a BOXCAR averaging system for incoming data points, and print-
out
capability. The BOXCAR system averages each measured data point with the prior
three data
points from that transducer to produce a more reliable measurement.
A wall thickness detection concept was employed in the Inspect 1001 inspection
system manufactured by IST, Inc. in Houston, Texas, which is a prior art
system used for
non-destructive testing of drill pipe. The prior art system detected wall
thickness as a
function of the output from each transducer. The present system adds or sums
the presumed
valid wall signals from the transducers from each of the four quadrants, and
ensures that the
sum from each array is within maximum and minimum limits corresponding to the
particular
coiled tubing size being tested. This summing of the signals from the sensors
in each array
is an effective technique for evaluating very small wall loss variations and
"telescope tubing"
variations associated with coiled tubing.
CA 02268446 1999-04-12
WO 98I16842 PCT/US97/18759
-11-
Detection of Longitudinal Splits - Eddy Current
A pair of coils 60, 62 positioned axially about six inches apart are wired in
a
differential manner. The coils are energized with an oscillator within a card
in the case 90 and
coupled to a potentiometer balanced circuit. Once balanced within the magnetic
saturating
circuit, the output of each coil is continuously compared to the output from
other coil, and
amplified to denote and emphasize the difference.
As each coil impedance is subject to the tubing material within it, the
difference
between the impedance of the coils is "zero" if both generate and observe an
equal amount
of eddy currents (i.e., absolute value of impedance (A) from coil 60 -
absolute value of
impedance (B) for coil 62 = 0). If one coil generates and observes more (or
less) eddy
current disturbance due to a longitudinal split in the tubing, while the other
axially spaced coil
is not responsive to the split at the same time (since the split is not
radially adjacent that coil),
the difference will cause a significant impedance difference. This indication
is proportional
to the depth and length of the split. . By controlling the frequency of the
oscillator while the
tube is in the magnetic field, the depth of penetration of the split can be
determined. A high
frequency oscillation will penetrate less than a low frequency oscillation,
and will exhibit
higher sensitivity to external surface longitudinal flaws. The frequency of
the eddy current
can be adjusted. For some implementations, more than one frequency can be
generated. If
only external splits are to be detected, a frequency of from 1 kHz to SkHz is
selected. If only
internal splits are to be detected, a frequency of 500 Hz may be selected. If
both external and
internal splits are to be located, a mixture of two or more frequencies will
allow fizrther
identification of the splits and determine if a split is on the external
surface (high frequency,
no lower frequencies), internal surface (low frequency, no higher
frequencies), or both
surfaces (both high and low frequencies).
A unique feature of the circuit is the ability of the detection system to
balance with
one rather than two potentiometers. The circuit produces two outputs: ( 1 ) an
AC output to
the computer that will report a comparison between the coils at frequency
intervals, and (2)
a DC output to the digital meter that will report a comparison between the
coils that last 10
CA 02268446 1999-04-12
WO 98/16842 PCT/US9?/18759
-12-
cycles or more. A digital meter monitors the DC output and facilitates
balancing of the
arrangement.
Each coil 60, 62 is a l000 turns detector wound to a length of 0.75 inches,
with a
AWG 28 wire. The circuit board used is an IST 3122. The use of two axially
spaced coils
and the detection of a split in the coiled tubing based on the difference in
output between the
coils is thus an important feature of this invention. Moreover, this
comparison and split coil
detection is made in real time, and may be visually displayed to the
inspection system
operator.
The Diameter/Ovality Detector
Four eddy current transducers 70 as shown in Figs. 1 and 2 are placed 90 ~
apart at
a fixed distance from a centerline 71 of the bore through the inspection head
30. A coiled
tubing having a diameter smaller than the distance between opposing
transducers (placed
l80~ apart) is inserted into the inspection head 30. Each transducer reports
the distance from
its surface to the outer diameter surface of the tube. By adding the combined
distance
between opposing transducers and the adjacent outer wall of the tubing and
subtracting that
sum fiom the distance between the opposing transducers, a tube diameter axis
is calculated.
The process is duplicated for the other two transducers to determine the
second diameter (y-
axis diameter) at 90 ~ relative to the first diameter (x-axis diameter). In a
preferred
embodiment, the x-axis diameter and y-axis diameter are perpendicular. If more
than two
diameters are provided, they may be equally spaced about the tubing, i. e.,
four diameter may
be measured at 45 ~ spacings. In each case, one diameter is substantially
inclined relative to
another diameter to detect ovality.
Referring to Fig. 6, the diameter D of the coiled tubing is thus the diameter
D 1 of the
spacing between radially opposing detectors 70, less the combined spacings A
and B between
opposing transducers and the outer wall of the tubing. Accordingly, D = D 1 -
(A+B).
Suitable eddy current transducers are manufactured by Truck and Gordon.
Alternatively,
ultrasonic or piezoelectric transducers manufactured by Massa Products
Corporation in
CA 02268446 1999-04-12
WO 98I16842 PCT/US97/18759
-13-
Hingham, Massachusetts, may be used. The electronics for ultrasonic
transducers is
complicated and ultrasonic transducers are temperature sensitive, although
their overall
sensing capability is greater than eddy current sensors. Ultrasonic
transducers may thus easily
replace the eddy current transducer disclosed herein for detecting ovality and
diameter of the
tubing.
In order to calibrate the system, the distance between opposing transducers
and the
actual tubing diameter should be known. The operator "ENTERS" Distance and
Diameter
and the computer linearizes the reading of opposing transducers. The
individual readings are
also displayed for convenience. The radial distance between each transducer 70
and the outer
wall of the coiled tubing is then sensed. The computer also calculates if the
tubing is not in
the exact center of the inspection head 30 and reports off center conditions.
An alarm is
triggered when either the sensed tubing x-axis diameter or the sensed y-axis
diameter is not
within selected limits. When both under diameter and over diameter conditions
are present
but each are within selected limits, the determined effective diameter of the
tubing may be
satisfactory. Tube necking and tube ballooning are reported as "-DIAM" and
"+DIAM",
respectively.
The computer also produces two analog outputs with indications proportional to
the
deviation of diameter. Ovality can be detected by observing that one tube axis
is larger than
the other tube axis. The computer thus calculates the differences between the
larger diameter
(major axis of the oval) and the smaller diameter (minor axis of the oval) to
determine if
ovality is within preselected limits.
Tubing Length Monitoring
Referring again to Fig. 1, a magnet 80 is embedded in at least one of the
centering
rollers 21 with the South pole toward the outside. With every wheel
revolution, the magnet
passes in front of a bipolar Hall effect switch 82 (Allegro 3177) placed on
the supporting arm
19. When the Hall Effect element is subjected to about 100 Gauss, the output
is grounded.
A preset counter relates the number of turns of the wheel to a fixed length of
the passing tube.
CA 02268446 1999-04-12
WO 98I16842 PCT/US97/18759
- 14-
The preset counter is set for each diameter of tubing to be inspected. The
output is used to
determine longitudinal position, i.e., both the length of coiled tubing
passing through the test
equipment and the axial position along the coiled tubing when a particular
defect is detected.
Alternate applications call for embedding two magnets, one with the South Pole
exposed and one with the North Pole exposed. The magnets are located 180
degrees apart.
Each magnet triggers two (2) Hall devices placed in a yoke arm. The order of
the triggering
determines the direction of the wheel (H1 before H2, or H2 before H1), and the
opposing
orientation magnets allow alternating the on/off positions of the Bi-direction
Hall. If the Hall
elements are placed in a ferromagnetic wheel, the magnetizing forces used to
magnetize the
coiled tubing may diminish the effectiveness of the magnets. The wheel should
be of a non-
magnetic stainless and protected from the external magnetic field. An
alternative
implementation may call for placing an optical encoder on the shaft of the
wheel.
Inspection Case
The aluminum case 90 as shown Fig. 5 houses three components: ( 1 ) a wall
loss
computer 92 which calculates in tube quadratures the remaining wall thickness;
(2) a
diameter/ovality computer 94 which determines the tubing diameters in two
orthogonal planes
and also determines a tubing ofI=center condition; and (3) a data acquisition
streaming
computer 96 which displays a chart on a flat panel 97 to the operator. Ports
l02 for a
connected mouse 104, printer 106 and keyboard 108 may also be provided. The
case 90 also
houses a four channel transverse and three dimensional flaw detecting
circuitry, a differential
eddy current longitudinal split electronics, a pulse width modulated constant
current mag
supply, a preset encoder/counter to monitor coil tubing length, and all
electronic power
supplies. The case 90 is powered by a 120V AC source or optionally from other
sources.
A suitable data acquiring computer is a DI 200 with 486 DX2.
Test results are preferably approximately 20 times per second. If the tubing
is moving
at 1 ft./second through the test unit, a test is made for approximately one-
half inch of coiled
CA 02268446 1999-04-12
WO 98/16842 ~ PCTlUS97/18759
-15-
tubing. If the tubing is moving at 2 ft/second, a complete test is performed
every one ( 1 ) inch
of coiled tubing.
A computer program and an interface card permit the generation of a paper
trace-Iike
display on a high brightness TFT flat panel display. Operating under Windows
95, a
continuous chart-looking display simulates an eight channel ink chart
recorder.
The screen 97 allows for fast data acquisition, streaming recording to Hard
Drive.
Recorded files can be stored on a floppy disk or a Zip disk. A marker can be
inserted at
selected locations with an automatic time/date stamp. The screen is programmed
to produce
a white background with blue grid lines and red traces, but a pallet of other
colors is available.
System Electronics
Suitable system electronics are depicted in Figs. 8-13 . The wall thickness,
flaw, and
split detector transducers may be packaged as a subassembly separate from the
diameter/ovality transducers 70. The magnetizing coil and the preamp or
junction box 20 are
also shown as part of the inspection head 30. The electronics case 90 may
house a system
computer and specific defect electronics. A test circuit, a supply for the
magnetizing coil, and
power supplies are also housed in the case 90.
Figure 9 discloses a suitable test circuit 110 generally shown in Fig. 8. This
circuit
may be used to route a selected power supply voltage to the test meter 112.
Switches 114
and 1l6 allow the operator to choose a +24 volt, +12 volt, +5 volt, or -12
volt to be
displayed on the meter 112 so that the system electronics may be thoroughly
checked at the
desired voltage settings.
Figure 10 shows an IST 3122 split detector circuit 120 for receiving eddy
current
signals from the coils 60, 62. A 1 kHz, 6 volt AC signal may be generated and
sent to the pair
of coils 60, 62, which may be spaced axially 6 inches apart. When a
longitudinal split defect
is present under one coil but not the other coil, the circuit 120 detects the
difl'erence in the
load on that coil compared to the other coil due to the inductive reactance of
the coil. This
CA 02268446 1999-04-12
WO 98I16842 PCT/US97118759
-16-
difference may be output as a voltage signal to a digital meter l22 shown in
Fig. 13 and also
to the computer interface card within the case 90.
Figure 11 illustrates suitable wiring to and from the preamp box 20. The
preamp box
20 serves as a junction for the wiring from the inspection head. As previously
noted, the
diameter and ovality sensors 70 and the conductors to the magnetizing coil 32
may actually
pass through the junction box and then to the case 90. Signals from the split
detector coils
60, 62, the diameter and ovality sensors 70, and power to a magnetizing coil
32 functionally
may pass uninterrupted, however, through the preamp box 20. The output from
the flaw
sensors 40 and wall thickness sensors 50 is routed through a preamplifier to
boost output
levels and reduce noise, and then to the output connector. The diameter and
ovality sensors
effectively act as proximity switches that produce changes in output voltage
when the distance
between the sensors is interrupted by the presence of the tubing. As
previously noted,
magnetic coil 32 is used to produce a magnetic flux field in the tubing being
inspected.
Figure 12 illustrates in further detail the wiring for the inspection
cartridge. Hall effect
sensors or transducers monitor transverse flaws and wall thickness, as
disclosed above, and
the eddy current coils 60, 62 monitor for longitudinal splits. The Hall effect
transducers 40
measure magnetic flux leakage and convert this leakage to a DC voltage output.
The eddy
current coils 60, 62 may be spaced in the desired selected distance apart, and
the six inch
spacing of these coils described earlier is exemplary.
Figure 13 depicts the panel wiring within the case 90. Primary components
within the
case 90 include one or more computers 124, an ITS 3122 split detector board
126 and a flaw
detect board 128. The computer is used to detect wall thickness in the tubing
while it is being
inspected, i.e., in real time. The output voltage of the wall thickness
transducers travels
through the preamp wall section and then to the computer where the voltage is
converted into
a tubing wall thickness signal. The visual alarm 128 and output 130 to a main
data storage
computer are also provided. The computer uses the output voltage of the
proximity switches
to compute the diameter and ovality of the tubing. Another visual alarm 132
and output to
CA 02268446 1999-04-12
WO 98I16842 PCT/US97/18759
- 17-
the main data storage computer are also provided for these sensors. The ITS
3122 split
detector board 126 operates as described above.
The flaw detect board 128 takes the output voltage of the flaw section of the
preamp
board and filters out unwanted noise, while providing an output 136 for the
main computer.
The flaw detect card ITS 3114 is a four channel flaw amplifier used to convert
the transverse
flaws to a voltage that is displayed on the main computer. The output of the
flaw channels
after pre-amplifications is sent through the signal cable to a single pole
double throw switch.
The signal then goes through a low frequency pass filter to help eliminate
noise, then to the
first gain stage amplifier. The gain level of this stage is set by adjusting
the gain knob for that
channel. From here, the signal is split into negative and positive components.
Both of these
signals pass through two more gain stages, then are recombined. At this point,
a sample of
the signal level is sent to a sensing circuit that turns on a light when the
signal reaches a level
set by the threshold knob. From this point, the four channels are combined and
the signal is
sent to a reject circuit that eliminates any indication that is below the
reject knob setting. The
signal is then split into negative and positive signals and sent to the main
computer.
Another generation of the inspection system disclosed herein has explosion
proof
characteristics. Special explosion proof electrical components may thus be
used. An array
of permanent high energy, rare earth magnets, such as neodymium 27, serve as
the magnetic
generator. Electrical power thus need not be supplied to the magnetizing coil,
since the
permanent magnets would replace the magnetizing coil described herein.
The system of the present invention allows for the generation of tubing wall
thickness
output signals which may be automatically displayed to the operator as a
percentage of the
nominal tubing wall thickness. Similarly, the tubing diameter may be expressed
as a percent
of nominal diameter.
The computer record is a composite representation of the tubing
characteristics. The
inspector determines at the time of inspection whether to remove the flaw or
just note on the
computer record the fact that such flaw is present. Subsequent inspection will
allow the
inspector to further evaluate if any increase in flaw size occurs. The system
may
CA 02268446 1999-04-12
WO 98/16842 PCT/US97I18759
_ lg _
automatically generate and deposit on the tubing a paint marking or other
indication of a
defect to assist in locating tubing anomalies, which may subsequently be
manually inspected
with further equipment. A record of sensed tubing characteristics may be sent
to a storage
computer for future retrieval, and the data compared with other inspection
runs.
The foregoing disclosure and description of the invention is illustrative and
explanatory thereof, and it will be appreciated by those skilled in the art,
that various changes
in the size, shape and materials as well as in the details of the illustrated
construction or
combinations of features of the various ultrasonic test elements may be made
without
departing from the spirit of the invention.
