Society of Petroleum Engineers

SPE 46023

Downloaded from http://onepetro.org/SPECTWI/proceedings-pdf/98CTR/All-98CTR/SPE-46023-MS/1927953/spe-46023-ms.pdf/1 by Repsol Services Company user on 27 May 2024

RESULTS FROM NDE INSPECTIONS OF COILED TUBING
Roderic K. Stanley, SPE, Quality Tubing, Inc.

Copyright 1S9S Scciety of Petroleum Engineers, Inc

Introduction
This paper was pns>~ for praswM”on
held m Houston. Texas. 15-16 April 1996
at the W96 SPE/lCoTA Coiled Tubing Roundtable
Carbon steel coiled tubing (CSCT) is manufactured to high
quality standards using nondestructive techniques as part of
Thts paper was selected for presentation by an SPE Program Commdtee followng rewew of
Information coniimd in an absbacf submiltsd by the author(s) Contents cd the paper, as an overall quality plan. Recently the full-body inspection of
presented, have not ken rw”ewed by the Scciefy of Petroleum Engineers and are subje~ to
corrmon by the author(s) The material. as presented, does not necessarily reflect any used coiled tubing (UCT) has also begun in earnest. This pa-
posibon of the society of Petroleum Engineers, its officers, or members Papers presented at
SPE meetings are subjecf to pubficafion revieW by Eddorml Committees of the Society of
per outlines the manufacturing NDE but concentrates on
Petroleum Engmee!s Eiemomc reproduckm, distribution, or storage of any part of th!s paper techniques and results from UCT inspections. It goes on to
for cemmercral purposes wthout the wmtten consent of the SCCiety of Petroleum Engineers is
prohibited Permission to reFriXuCe~ _ restricted to an abstract of not more than 3C0 suggest how such inspections may be used in an overall pIan
vvvrds. dlustratmns may not be copd
ment bf Mere
The abstract must contain conspicuow ackncwfedg-
and by w+mm the paper was pr=nted Wrde Llbrar!an, SPE. P O S@
for accurately monitoring the life of such materials.
.%23636, Richardson. TX 7S083-3636. U S.A, fax 01-972-952.9435

Manufacturing Inspections
Abstract Carbon steel coiled tubing is generally manufactured from flal
Service Centre nondestructive inspections of used coiled tub- or tapered strip purchased in coils. This strip is inspected
ing are being performed in Scotland, This paper outlines the visually after pickling but before oiling, After slitting to size,
development of an NDE tod for coiled tubing inspection, the strips are joined end to end with strip-bias welds that are
explaining the principles of o~ration and presenting some made at an angle to the strip axis (Fig. 1).
interesting results from several recent inspections, including These welds are inspected by radiography for volumetric
results on a recent certification test. The current tool performs imperfections using a 2T hole in a No 10 ASTM penetrame-
the following operations for 100°%of the tube: de(ecfs ID and ter as the reference standard. They may also be inspected by
OD pitting, transverse OD damage and longitudinal damage; magnetic particle inspection (MM) for two dimensional weld
locates the areas of heavy cycling; measures ovality and local flaws. The weId is heat treated and finished by removing the
wall thickness. All detections and measurements are per- crown and then polishing to provide an even junction.
formed by non-contact electromagnetic methods which It has been found by fatigue cycle testing that well made
include magnetic reluctance (MR), magnetic flux leakage strip bias welds will cycle to failure by as much as = 90°A of
(MFL), and eddy current methods for Longitudinal flaws, the cycle Iife of the virgin tube. One purchasing specification’
ovality, and the detection of heavily cycled areas. Technique requires that a typical strip bias weld pass a bend test.
sensitivities are presented, The strip is then milled by the high frequency induction
Data from each inspection can be saved for comparison (HFI) method into tube, the specified wall thickness toler-
with fiture inspections. Such fill length NDE, along with ances being as shown in Table 12. However. rarely is the wall
fatigue q“ding results from flaw removal regions have per- thickness more than a few thousands of an inch from the
mitted strings to be managed by remov”alof isolated flaws or specified value, t The body and weld-line are currently gen-
flawectlworn areas. erally inspected by eddy current (ET) inspection to ASTM E-
Ancilku-y techniques such as compression wave wall 309, the standard being a 1/32” or 1/16” through-drilled hoIe
thickness measurement, ultrasonic shear wave and radio- (TDH9. See Fig. 6. Flaws detected during this inspection are
graphic techniques for the volumetric checking of tube-to- removed with sandpaper to intema13 or purchaser speGfica-
tube welds, and magnetic particle and liquid penetrant tech- tions’”.
niques for two dimension imperfection detection are also Indications from the inspection record are reinspected by
discussed. Evaluation of NIX results will feed into a pro- liquid penetrant (PT) or magnetic particle (MT) metlods to
posed new API Recommended Pracfice for Care, assure compIete removal from the OD suflace, or radiography
Maintenance and Inspection of Used Coiled Tubing in order (RT) for the detection of internal surface flaws. The major
to provide for standard ways of managing used coiled tubing. reasons for removing minor flaws are that (a) they may oth-
erwise contribute to lack of cycle life, and (b) they may

91
2 R. K. STANLEY SPE 46023

provide local regions for accelerated corrosion or the ingress be checked by application of MT.
of hydrogen.
In cases where tube-to-tube welds are permitted, these SR17 Inspection for Pipe Lines. One would normally in-
weldsare inspected with either doubIe-wall RT or ultrasonics spect HFI pipe welds with ultrasound, but standard tubular
(UT) for vohunetric imperfections, and either PT or MT for electromagnetic inspection (EMI) equipment, which inspects
outer surface hvo dimensional imperfections. The former 100?4oof the pipe, has been used because there is then no need
requires two shots at different angles. The latter requires to locate the weld-line. The longitudinal and transverse

Downloaded from http://onepetro.org/SPECTWI/proceedings-pdf/98CTR/All-98CTR/SPE-46023-MS/1927953/spe-46023-ms.pdf/1 by Repsol Services Company user on 27 May 2024

sending ultrasound axially through the weld, and is better magnetic flux leakage (MI%) inspections in EMI are ade-
suited to thicker-walled material. Removal of the outer crown quate, and a good assessment of the wall thickness can be
helps eliminate unwanted reflections. made from the gamma-ray wall gauges built in to most sys-
Occasionally, when specified for pipeline material, the tems. Standardization of each MFL sensor is performed using
HFI weld-line can be reinspected tier final hydrostatic test- a l/16ti in. TDH.
ing in accordance with Special Requirement SRI 7 of API
Specification 5LS. 5L permits this inspection to be performed Wall Thickness Measurement. Continuous 100?40wall
by with either ultrasonic or electromagnetic methods. thickness measurement is not currently performed during
manufacture, and would be difEcult to implement with ultra-
Bias Weld Radiography. This inspection is performed by sonics; however, electromagnetic methods are now available.
directing kilovolt x-rays through the weld and either exposing Spot UT measurements are made during prove-up, and are
film or capturing the image digitally. The digital method accurate when the compression-wave ultrasound gauge is
offers image enhancement techniques such as blacldwhite calibrated on test blocks which have the same curvature as the
reve~l, pixel gray-shade changes, and edge enhancement so outer surface of the tube. Such bIocks should also permit the
that a weld image can be viewed rapidly in several ways in gauge to be accurate at thicknesses on either side of the
red time. thickness to be measured.

MaWetic Particle Inspection. This inspection is performed Ovality. One suggested requirement (API RR5C7) for oval-
with an AC yoke, the inspection area being the middle one ity prior to spooling is * 0.010 in, there is howe~er no
third of the distance between the legs of the yoke. In the case requirement after spooling. Ovality may, however, be de-
of bias welds, the inspection is performed atler the weld is tected and measured by recently introduced NDE inspection
dressed and polished, with particular attention paid to the equipment which uses the eddy-current lift-off concept, as
strip edges. described later.

Eddy Current Inspection. Since the penetration of eddy Manufacturing String Records
currents into the tube wall is governed by the “skin depth’ (5) In addition to a material test certificate/weld location record.
complete string records, including all x-rays, eddy current
equation 8 = l/~(xvcrf), the tube is passed through a coil
which magnetizes it longitudinally to place the AC magnetic records, tensile and microhardness records, the hydrostatic
test chart, and inspection prove-up sheets can be provided to
permeability (p) of the steel at a low and constant value, and
the customer. These have proved very uset%l when re-
the frequency (f) of the eddy currents is kept as low as possi-
inspecting the string tier use.
ble. e.g. 5-10 ld-lz. (cs = electrical conductivity of steel). The
eddy currents are induced with a surrounding coil, and thus
The Need for Used Coiled Tubing Inspection
flow circumferentially. Two forms of detector are currently
The introduction of theoretical “fatigue life” programs for
used: either a sector coil for scanning the weld-line and its
UCT created situations in which many reported failures in
heat-tiected zone, or a full encircling differential coil assem-
coiled tubing occurred with apparently low amounts of theo-
bly for scanning the entire tube including the weld-line.
retical fatigue life consumed. Analysis of such failures~
Standardization of the system involves running the test stan-
attribute them to the following causes: corrosion, erosion,
dard with the TDH at the 12, 3, 6, and 9 o’clock positions;
surface damage, over-pull, ovality, and often to combinations
this assures that the detector is equally sensitive over 360 deg.
of such damage arising out of repeated heavy usage of tubing
i.e. the tube is centered in the bore of the detector. The system
with no atlowance built into the tubing operation for the ac-
detects material imperfections in strip.
cumulation of damage from prior jobs. End user response to
Liquid Penetrant Inspection, In this inspection, the area is this situations has been to (a) limit string usage by specifying
sprayed with fluid which penetrates to the root of two dimen- a maximum theoretical fatigue life consumed, and more re-
sional imperfections under its own surface energy. Some of centty (b) to require full body NDE inspection.
the fluid is then leached back out and developed to form a Prior history with inspection of oilfield tubular product
contrasting indication. Complete removal of such flaws can teaches that electromagnetic methods’ can be used to measure

92
SPE 46023 RESULTS FROM NDE INSPECTION OF COILED TUBING 3
—.

wall thickness and ovality, and detect internal and external on the pipe from which the Iargest signals originate indicated
pitting, and other localized problems which reduce the wall by one of four lights on the inspection console. This can be
thickness. The advantage here over ultrasonic methods is seen on channels 1 and 2 of the “windaq” display shown in
that all readings are non-contact, so that problems arising Figs. 10 and 11.
from uncertain surface conditions are avoided. QTI therefore The four lights assist in rapid location of the origin of the
worked an equipment manufacturer to develop an electro- indication during prove-up, especially when it originates from
magnetic CSCT inspection system, based on oil country the ID. The light is turned on above an electronic threshold

Downloaded from http://onepetro.org/SPECTWI/proceedings-pdf/98CTR/All-98CTR/SPE-46023-MS/1927953/spe-46023-ms.pdf/1 by Repsol Services Company user on 27 May 2024

tubular goods inspection techniques, and has been using such that is controllable by the inspector in relation to the ampli-
as system in Scotland since 1997. tude from standard flaw signals. (i.e. a 10”Astandard pit may
give an 8094.screen height signal, while the light may be set
Coiled Tubing Inspection Systems to illuminate at 50V0of screen height).
The basis for design of a CSCT inspection system is to be able Magn&”c Wall Measurement’. The magnetic lines which
to detect unambiguously all of the flaw conditions found in pass through the sensor may not pass through the pipe mate-
the material. For used coiled tubing, this is quite a problem, rial, although they are in equilibrium with them. The field
but not a new or insurmountable one. A study of problems strength of such lines rises as the pipe wall thins, and vice
associated with used CSC~ lead to the needs and soIutions versa (Fig. 4), thus the ambient tangential signal contains
that are presented in Table 2. information regarding the average wall immediately below
the sensor. The largest signal occurs in the sensors when no
Review of Inspection Techniques. One optimum route to pipe is present.
acceptable inspection of used CSCT, which accounts for de- This can be explained in a general way in terms of the in-
tection of all common UCSCT imperfections is to include creasing number of magnetic poles created where the field
magnetic reluctance wall measurement, magnetic flux leak- lines leave (PI) and enter (PJ the tube wall; the thicker the
age (MFL) for flaw detection, eddy current surface scanning wall, the more poles are created, the larger is the demagnet i-
for longitudinal flaws and strain detection, and eddy current zation field (1%) and the lower is the field tangentially
lift-off measurement into one-package as follows. First, how- through the sensor. (See Fig. 2).
ever. the magnetic flux in the tube wall must be fixed. In order to eliminate noise so that only the more gradually
Magn&”c Conditioning of the l%pe. The pipe is passed changing wall variations are detected by this method, the sig-
through a magnetizing system that raises a short part of it to nals may be averaged out by boxcar averaging. This also
magnetic saturation in the longitudinal direction. This is in eliminates the effect that the MFL signals from flaws would
the region of 1.6-2.0 Tesla (16-20 kGauss), The magnetizing otherwise have on the wall thickness signal. Further averag-
system can be either an encircling coil (Fig. 2) or a suitably ing of signals into quadrants around the 360° circumference
designed array of permanent super-magnets. Raising the of the tube provides for suftlcient sensitivity to detect wall
magnetic field level provides directly for the following two losses over larger areas, such as might occur due to erosion,
inspections. and holds the magnetic permeability of the steel corrosion, and frictional wall loss from buckling or tubing
constant at 20-30 for the eddy current tests, permitting the wear.
eddy currents to penetrate into the pipe wall. The information is presented on two channels which are
Magn&”c Flux Leakage (MFL) Inspection. Here the standardized from a known percent wall loss region in a
high flux in the material causes MFL when a discontinuity is standard tube. See channels 3 and 4 of the display shown in
encountered. The MFL field topography for many flaw types Figs. 10 and 11.
has been documented from the 60s. Detection of the MFL Diameter Change Measurement. Standard eddy current
fields is performed by rings of Hall sensors mounted above lift-off sensors (Fig. 5a) are employed in a ring around the
the pipe surface at fixed lift-off (Fig. 2). Thus magnetic field tube to measure across two diameters (Fig. 5b). Each lift-off
perturbations caused by any flaw with a transverse or three- sensor measures the change in impedance as the distance
dimensional component will pass through the sensor, along between the face of the sensor and the tube changes due to
with the ambient field from the magnetizing system (Fig. 3). variations in diameter. Differences in signals from sensors
The amplitude of the flaw signal developed in the sensor is that are mounted 180° apart gives diameter measurements
not proportional to the flaw depth, since other geometrical and reduces spurious signals from pipe vibration.
factors such as flaw width, orientation and remaining wall are Standardization is effected using machined standards, and
involved, and the inspector must therefore currently stop the data presentation is on two channels which present positive
inspection and perform prove-up of the indications. Notwith- and negative variations from the standard diameter. See
standing this. the sen.rifivily of the sensing system is set using channels 5 and 6 in Figs. 10 and 11.
standard flaws in test standards. Longitudinal Flaw Deteti”on. Holding the permeability
For the purpose of presentation, the largest signals from constant with a high magnetic field permits variations in
each ring are presented on hvo channels, with the quadrant magnetic permeability to be minimized and controlled. As

93
4 R. K. STANLEY SPE 46023
..—. ..

with standard new CSCT inspection, the tube is encircled by (n) Manufactured fatigue cracks
a differential eddy current coil system operating at a few kHz. (o) Pipe bent 90° & over-stressed to 22 Tonnes.
Flaws of a three dimensional or longitudinal nature cause (P) A transverse crack in a tube-to-tube weld.
perturbations in the magnetic fields that are associated with (q) Wear marks from an injector.
the circumferential eddy currents in the tube wall. Detection (r) Soft round dents from hammer impacts.
occurs via changes in the impedance of the coils using stan- (s) Longitudinal grooves in and 180° to the HFI weld.
dard ET bridge circuits. (t) Wear marks from wireline. ,

Downloaded from http://onepetro.org/SPECTWI/proceedings-pdf/98CTR/All-98CTR/SPE-46023-MS/1927953/spe-46023-ms.pdf/1 by Repsol Services Company user on 27 May 2024

Standardization of the signals occurs via a longitudinal (u) Internal corrosion from acid.
slot in the test standard The output is displayed on channel 7 (v) 1.5” tube section inside 2.0” tube.
of the “windaq” display. (See channel 7 in Figs. 10 and 11). (w) Coated and uncoated cable inside tube with 5094.of the
Highly Strained Areas. Since eddy currents respond to strands removed.
changes in electrical conductivity (cr), magnetic permeability (x) Sharp scratches in the outside surface.
(I.%) and fill factor, the system will also detect areas of high
strain in the tube wall. The strain changes the electrical con- The general results from this comprehensive test were as fol-
ductivity permanently as a result of modification of the Fermi lows:
Surface (i.e. the electron-crystal lattice interaction in wave (a) All pitting was detectable, depending upon the pit vol-
vector space). The longitudinal flaw system also responds to ume, with increasing signal for increasing material
changes in diameter (ovality, fill-factor) and detects welds. volume loss, (Note that MFL theow indicates that pit
Prescntah”on and Storage. All of the information de- topography and the steel permeability around the pit
scribed above. along with a channel for length from the tube wall tiect the pit signal amplitude. )
end. is presented on a “windaq” display which scrolls the 0) All transverse notches were clearly detectable, and
eight information channels by on a digital chart recorder on a should give greater signals with increasing depth.
laptop screen. The entire string data can be stored and Notches filled with ceramic materials were also de-
downloaded to disc or tape. Further. thresholds can be tected. (MFL theory indicates larger signals for deeper
changed after the data are collected, so that it can be re- and wider cracks, up to 0.040 in. widex and no change
evaluated in different ways. in detectability when filled with nonmagnetic material.)
(c) 450 cracks, were detectable, depending upon depth.
System Sensitivity and Validation (Here MFL theory indicates that under the same condi-
Tests on system capability were conducted in 1997 by an in- tions of flaw size and magnetization level as for
dependent laborato~ using test flaws in 2 in. x 0,156” CT-80 transverse notch detection, the MFL signal should be
material. Test flaws consisted of the following: one half of the transverse notch signal. )
(a) Transverse 5%t, IO%t, 12.5%t, 15%t and 20’%t EDM (d) Fatigue cracks were detected even though their width is
notches on the OD small in comparison with transverse test notches. This
@) Tranm’erse 5%t and 10%t EDM notches on the ID occurs because the change in magnetic permeability that
(c) 5%t, 10%t, 12.5°/’t, 150/d and 20°At 45° EDM notches is associated with the strain area at the crack tip as it
on the OD progresses into the steel effectively widens the crack
(d) 5%t spherical pitting on the OD of diameters 1.0, 2.0 magnetically if not physically.
and 3.0 mm (e) Longitudinal cracks and (ID gouges were detectable at
(e) 10%t spherical pitting on the OD of diameters 1.0, 2.0 raised sensitivity. (The instrument had originally been
and 3.0 mm standardized for a 100°/0through-wall crack, so this was
(f) 12.5%t spherical pitting on the OD of diameter 1.0, 2.0 no surprise.)
and 3.0 mm. (f) Reductions in wall thickness were detectable both on the
(!3) 15%t spherical pitting on the OD of diameter 1.0, 2.0, wall channels and, for shorter reductions, from the MFL
and 3.0 mm. they emit (i.e. if they can be considered to be elongated
(h) 12,5%t and 20%t longitudinal flaws in the HFI seam pits.)
(i) 12.5%t and 20%t longitudinal flaws in the HFI HAZ (g) Uvality was detectable to 1’%.as determined by the for-
region mula ~m - Dmm]~sPeclfied.
0) HFI 12.5%t and 20%t longitudinal flaws at 180° from (h) The internal weld~ush removal area was detected both
the seam, as a change in wall thickness and in MFL in the sensors
(k) Wall thickness reductions of 5%t, 10%t, 12.5%t, 15%t that scanned over the weld area.
and 20°Aton the OD. (i) Transverse cracks in the tube-to-tube weld were not un-
(1) Ovality of 5%t, 10%t, 15%t and 20%t difference be- ambiguously detected on the MFL channeIs; this is most
tween the maximum and minimum diameters. probably due to the fact that the MFL from the weld
(m) Region of 100 mm internal flash removal swamps that from the much smaller crack. (Note that

94
SPE 46023 RESULTS FROM NDE INSPECTION OF COILED TUBING 5

this is a difficult inspection situation for conventional this detects damage caused in transit and while respooling
Ml% at the stand-off used in this type of inspection.) onto work reels.
0) The internal corrosion was detected; this occurs because
the tube waII is saturated with magnetic flux. Here the Weld Locations. In used CSCT inspection, with rcg=rd to
flu. lines can not be pushed any closer together, and so welds, detection of the bias and tube-to-tube welds provides
emerge from both internat and external tube surfaces. essential mileposts for strings that may have stretched or oth-
(k) Bent and over-stressed areas were detected via their erwise been repaired. Our general observations are that:

Downloaded from http://onepetro.org/SPECTWI/proceedings-pdf/98CTR/All-98CTR/SPE-46023-MS/1927953/spe-46023-ms.pdf/1 by Repsol Services Company user on 27 May 2024

eddy current signals. The same effect is found in sucker (a) most bias welds are detected via the MFL emitted from
rod inspection. either permeability changes in the weld area, or the step
(1) A piece of 1.S” steel tube inside a 2” tube was detected taper in conventionally tapered strings.
through an apparent increase in wall thickness. Simi- (b) occasionally the stress relief of the bias weld between
larly. the presence of internal cable appears as an same-thickness tube sections has been performed in a
apparent wall increase. manner that does not emit sufficient Ml% to permit the
(m) Broken strands in cable were detected where there was weld to be detected.
a sufficient change in magnetic flux associated with in- (c) all tube-to-tube welds are easily detected via their MFL
terfaces between the ends of the cable and air. and ET signals. However, the situation with the test de-
scribed above indicates that, once detected, tube-to-tube
With regard to item (k), damage to the crystal structure welds should be inspected by magnetic particle inspec-
caused by other effects such as the permeation of hydrogen tion, or liquid penetrant inspection for the detection of
into the steel will also lower the electrical conductivity of the OD transverse cracking, or shear wave ultrasound or ra-
steel over the area in which the ~ is released, and therefore diography for volumetric imperfections.
lends itself to detection by eddy currents. So long as the eddy (d) there are more tube-to-tube welds than string records
currents can penetrate the tube wall, such damage is detect- show, indicating that it is easy to forget to do the pa-
able. perwork associated with adding a weld. Since many
such welds have been performed in the field, and may
Field Test Standards be of uncertain quality, inspection may then lead to a re-
There is no need for field test standards to contain all of the weld performed under more controlled conditions.
test imperfections described above. However. several basic
test flaws are needecl and the range of phenomena that we Wall Thickness. For the wall thickness channels, various
have seen can be detected indicates the need for highly wall changes are detected including those associated with
trained inspection personnel. conventional tapering and the use of “True-taperTM” strip.
Conventional wisdom with test standards indicates that Wall losses associated with erosion and wear have been
they should be made from the same material grade as the tube found. When these losses stretch substantially around the
under inspection, and of course have the same OD and wall tube circumference, they contribute to a general loss of cross-
thickness. This is to ensure that the saturation flux density in sectional area and the tensile force that can be safely applied
the test standard and the tube under inspection should be the at that point on subsequent jobs. Wall losses associated with
same so that comparable results may be obtained between the severe corrosion have also been found.
standard and the inspection, and between inspections.
A minimum set of test imperfections is as follows: Flux Leakage Information. For the MFL channels, internal
(a) a wall loss area representing 5% of the specified wall and external pitting is easily detected at the 10°/0level, and
machined all the way around the tube. This may be used visual inspection confirms the location. Many types of pit-
to standardize both the wall channels and the ovality ting, from small hemispherical pits to larger lake-like pits,
channels over the small ranges of ovality expected. are detected. Such general pitting at the 5- 10°/0level does not
(b) a transverse 10%t notch of length 0,375 in, and width contribute substantially to the loss of cross-sectional area of
no greater than 0,020 in, the tube wall, However, tubing has been observed to break
(c) an area of 10’%spherical pitting. while spooling where internal pitting was present.
The Ml% channels are particularly useful in locating the
Field Experiences presence of internal pitting that may be attributed to fluids
The coiled tubing inspection tool has been used in two roles. remaining in spools under gravity at the 6 o’clock position for
First, monitoring the aging of strings in Service Centres has extended periods. This has proved to be a major contributor to
become popular in areas such as the North Sea, Typical ex- the removal of corroded pipe sections.
amples of field data are given in Table 3, which presents the
results of the inspection of two strings. Second, companies Diameter Measurement Channels. Necking, ballooning
have used the tool in a basic “receiving inspection”’ timction; and ovality have all been detected by this tool. Detection of

95
6 R. K. STANLEY SPE 46023

such problems has also resulted in removal of sections of In the hands of a good inspector, the technique is quicker
UCT. The eddy current longitudinal split channel has proved than RT, but interpretation from an oscilloscope screen is
to be extremely useful in determining the extent of the sec- often more difficult.
tions to be removed, since this channel has been found to In this technique, sound in the range 5-10 MHz is fired at
locate the ends of the heavily cycled sections that are prone to a shallow angle tube, and weld flaws reflect sound back to the
contain such problems. transducer. Indications on a gated area of the flaw detector
screen, which is set up to represent the weld-line and the

Downloaded from http://onepetro.org/SPECTWI/proceedings-pdf/98CTR/All-98CTR/SPE-46023-MS/1927953/spe-46023-ms.pdf/1 by Repsol Services Company user on 27 May 2024

Ancillary Prove-Up inspections heat-atfected zone on either side, are compared with a stan-
Dimensional Checks. A quick check with calipers deter- dard signal such as that from a test hole or notch. See Fig. 8.
mines the ovality and general wear at each point at which the
inspection is halted. Diameter measurements are taken at 45° Laboratory Experiences with Flaw Removal
intervals around the tube in the A-A (HFI weld-line), B-B Flaw induction and removal in UCT are revealing important
(900-2700), C-C (450-2250), and D-D (1350-3 15°) positions. information regarding the effect of both the flaws and their
This permits necking and ballooning to be measured, and removal upon the remaining cycle life of the material. fig. 9
ovality to be computed very accurately. shows data collected during a flaw removal test with CT-80
material cycled on a 72 in. bend radius with interred pressure
Ultrasonic Wall Thickness Measurements. Four UT wall 3 kpsi. While virgin tube cycles over 700 times, 10”Atrans-
measurements are taken around the circumference of the verse notches in the weld-line and at 180° from the weld-line
tube. (Fig. 6a) and in conjunction with the diameter read- cycle only just over 100 cycles. Removal of such flaws (i.e.
ings give an indication of the type of loss, Here, high removing 10°/0of the wall) at the weld-line and at 180° from
quality focused transducers (such as the Extra High Resolu- the weld-line raises to cycle life to almost 600 cycles. Similar
tion transducers that were developed for oil country tubular data have been collected on other tube sizes.
goods inspections) should be used or the results might be er- Such data show that the initiation of small transverse
roneous. EHR transducers are manufactured to give signals notches in coiled tubing in well servicing applications is
from the bases of pits rather than the ambient back-wall. clearly to be avoided. How their dimensions affect the pipe
wall are yet to be determined.
Radiography. Double wall radiography (Fig. 7) is one tech-
nique that may be performed after a tube-to-tube weld has API Current and Future Requirements
been manufactured and the surface dressed to be integral with Specific requirements for outer diameter, wall thickness. ten-
the two tube surfaces. This inspection involves at least two sile and yield strength, hardness, and other common pipe
shots at an angle to the finished weld, using the most sensi- properties at the manufacturing plant are being documented
tive penetrameter that the inspector can comfortably work by the American Petroleum Institute A specification for
with, Since the method depends of detecting density differ- coiled tubing as line pipe (proposed API 5LCP) in the size
ences on film in a double wall situation through at least 2t~2 range 0.500-6.625 in. and including all common current CT
of wall. then sensitivity to small three-dimensional imperfec- ODS and walls was presented to API management in Dec 97.
tions is not so high as it would be for RT strip inspection API committees are also working on a Specification for
(which is an ideal situation for RT), and is virtually zero for Coiled Tubing, to include coiled tubing grades from CT-55 to
crack-like imperfections unless they exhibit a detectable three CT-1OO.
dimensional content, However, of major importance for UCSCT is a Recom-
mended Practice for Care, Maintenance and Inspection of
Magnetic Particle Inspection.Once the weld is dressed to Used Coiled Tubing. This document will contain information
be flush, the most sensitive inspection for two dimensional on what the industry has found to be the best practices for
OD surface cracks is wet fluorescent magnetic particle in- cleaning, inspecting and protecting used coiled tubing.
spection, In this inspection, optimally performed in a light
level of less than 2 ft-candles (20 lux), ultraviolet light of Conclusions
surface intensity Y 2000 w’atts/cm2 illuminates the inspection 1. NDE at strategic locations during manufacturing con-
surface while the magnetic field from an AC yoke is applied tributes significantly to the production of high quality
across the weld and fluorescent particles are applied. Weld CSCT. The use of NDE will carry over into the production
cleanliness and lack of transverse sanding marks are both of other coiled tubing.
essential. 2. Field NDE devices should contain all of the necessary
techniques to provide such measurements as wall thickness,
Ultrasonic Shear Wave Inspection. This is typically per- ovality and ballooning, and the detection of flaws on the ID
formed when searching for transverse cracks on the ID and OD surfaces, and the location of highly cycled regions.
surhce of used drill pipe, and for all volumetric weld flaws. 3. NDE is essential in determining the true life of various

96
SPE 46023 RESULTS FROM NDE INSPECTION OF COILED TUBING 7

sections of a string, and for making such decisions as (a) 7. “Advanced Electromagnetic Tubular Inspection During Well
when to remove or repair a seetion, and (b) when to con- Servicing,” R. K. Stanley, Proc. 13th World Con$ On Nonde-
struct strings from other strings. NDE ovality measurements structive Testing, Sao Paolo, Brazil, Oct. 1992.
will prove very usefid in determining actual collapse pres- 8. See for example Nondestructive Evaluation - A Tool in Design,
Manufacturing and Service, by Don E Bray and Rodent K
sures at all points along a string, but must form only part of
Stanley, pub. CRC Press 1997., and references therein.
the overall picture of UCSCT assessment, 9. “The Inspection of Used Coiled Tubing,” S. Papadimitriou
4. Transverse 2 dimensional flaws in the outer surface of and R, K. Stanley, Proc .2ndIntl. Conf on Coiled Tubing Op-

Downloaded from http://onepetro.org/SPECTWI/proceedings-pdf/98CTR/All-98CTR/SPE-46023-MS/1927953/spe-46023-ms.pdf/1 by Repsol Services Company user on 27 May 2024

coiled tubing severely reduce the cycle life of the coiled tub- erations, Houston, Mar 1994, pub. Gulf Publishing.
ing. The effect of the topography of such flaws has yet to be
determined for the various types of steel that are used in the
production of work strings.

Nomenclature —
Table l—Wall thickness tolerances
CSCT = Carbon steel coiled tubing Wall Thickness t Minus m
EHR = Extra high resolution <0.110” 0.005” 0,010”
EM] = Electromagnetic inspection 20.110” 0.008” 0.012”
ET= Eddy current inspection
HAZ = Heat affected zone
HFI = High frequency induction
ID = Inside diameter
A4FL = Magnetic flux leakage
Table 2—UCSCT Problems and Solutions
.i~PI = Magnetic particle inspection
Imperfection Cause Detection Method
34T = Magnetic particle
Ovality Cycling, initial Eddy currents
OL) = Outside diameter Ovality
PT = Liquid penetrant inspection Ballooning Pressure & cycling Eddy currents
RT = Radiographic inspection ID Corrosion HC1, KC1, NaCl, Magnetic flux
t = Specified wall thickness pits COZChromates leakage
TDH = Throughdtilled hole OD Corrosion HC1, KCI, NaCl, Magnetic flux
[ICSCT = Used carbon steel coiled tubing pits C02 Chromates leakage
LY2T = Used coiled tubing Transverse Cycling tube wall Magnetic flux
[JT = Ultrasonic Inspection leakage
WFMPI = Wet Fluorescent Magnetic Particle Inspection Pressure loss,
Eddy currents
References I draulic hammering
Loss of Yield I Crystal transfor- Eddy currents
1. Specijleation for Coiled Tubing,” Transocean Specification /conductivity
TPT-17-SF-01,Rev 4. Magnetic flux
2, API Recommended Practice for Coiled Tubing Operations, RP leakage
5C7, First Ed., Jan 1997, pub. American Petroleum Institute, Magnetic flux
WashingtonDC. leakage
3. TechnicalCatalog,QualityTubingInc., see hternet site Erosion (OD) I Rubbing Tubing Magnetic wall
http:l&ww.qualiptubing. corn. measurement
4. Specl~cations for Coiled Tubing, HalIiburtonEnergyServices.
5 SpeclJlcationJor Line Pipe, API Spec 5L Magnetic wall
6. See for example “Failures in Coiled Tubing,” R. K. Stanley, measurement
Proc. S’h Intl. Conf On Coiled Tubing and Well Intervention, Necking Force too large Eddy currents,
Houston, Texas, Feb. 4-6, 1997, and “An Analysis of Failures Magnetic wall
in Coiled Tubing, R, K, StanIey, IADC/SPE paper 39352, measurement
Proc, L4.LWSPEDrilling ConJ, Mar 3-6, 1998, Dallas, Tx.

97
8 R. K. STANLEY SPE 46023

Table 3—Typicai Coiled Tubing Inspection Reports

TUBING DESCRIPTION
Size 1.50in Wall Tauered Grade CT~ string No: xxxx On w
Material Wood Diameter Weight Footage_ 12610

Downloaded from http://onepetro.org/SPECTWI/proceedings-pdf/98CTR/All-98CTR/SPE-46023-MS/1927953/spe-46023-ms.pdf/1 by Repsol Services Company user on 27 May 2024

Bias weld = * Tube-to-tube weld = // A-A = HFI weld if possible

Ft Diameter (inches) Ovality Wall Thickness

A-A B-B c-c D-D %0 (roils) Comments

o 1.506 1.501 1.498 1.507 0.6 96 99 98 98 Corrosion Pitting

1020 1,506 1.549 1.506 1.502 3.1 100 95 102 102 Corrosion pitting
2054 1.511 1.500 1.501 1.506 0.7 102 100 109 109
3950 1,505 1.500 1,501 1.507 0,5 102 101 103 103
4859* 1,505 1.493 1.496 1.507 0.9 100 102 104 104 Bias weld
4860* 1.506 1.501 1.502 1.505 0.3 115 112 114 114 Bias weld
5287 1.501 1,501 1.503 1.505 0.3 111 109 111 111 Corrosion pitting
5839 1.501 1.503 1,501 1.501 100 111 112 112 Score
6000 1.507 1.506 1.503 1.511 0.5 114 116 113 111 Corrosion pitting
6250 1.507 1.508 1.504 1.500 0.5 115 117 111 114 CP. Score 0.005”
6650 1.506 1.505 1.502 1.503 0.3 119 120 125 124 CP 0.007”
6820 1.505 1.503 1.501 1.501 0.3 121 123 126 126
7000 1.503 1.505 1.503 1,502 0.2 123 125 124 124 Galling
7300 1.509 1.504 1,508 1.505 0.3 124 126 126 124
8080 1.510 1.509 1.504 1.504 0.4 123 125 125 123 Corrosion pitting
8303 1.505 1.506 1.503 1.506 0.2 130 130 133 131
8575 1.505 1.509 1.507 1.507 0.3 131 130 128 131 Light score
8827 1.511 1.504 1.509 1,507 0.5 130 130 137 131 Light score
9100 1.510 1.503 1.509 1.509 0.5 131 130 128 131 Light score
9374 1,511 1.503 1.510 1.509 0.5 133 131 131 133 Light score
9555 1.511 1.500 1.509 1.504 0.7 131 132 133 137 Light score
9760 1.504 1,504 1.506 1.505 0.1 131 134 140 134
Additional Comments

[nspector GK Date 8/9/97

98
SPE 46023 RESULTS FROM NDE INSPECTION OF COILED TUBING 9

Table 3, continued

TUBING DESCRIPTION

Size u Wall 0.190’’-0.102” Grade CT70 string No: 2674-2673 on l@!

Material Diameter Weight Footage

Downloaded from http://onepetro.org/SPECTWI/proceedings-pdf/98CTR/All-98CTR/SPE-46023-MS/1927953/spe-46023-ms.pdf/1 by Repsol Services Company user on 27 May 2024

.
Comments

Bias weld = * Tube-to-tube weld = // A-A = HFI weld if possible

I
Ft Diameter (inches) Ovality Wall Thickness
A-A I B-B I C-C I D-D I YO (roils) Comments
o I 1.757 I 1.753 I 1.753 I 1.756 I 0.2 I 108 I 108 I 109 I 109
189 I 1.758 ] 1.757 I 1.756 I 1.752 I 0.3 1107 I 107 I 107 I 107 Inj. Mks 0.009”
1123 1,768 1.759 1.738 1,730 1.7 109 109 109 107 Corrosion Pitting
1128 Inj. Mks 0.007”
2000 1.762 1.761 1.761 1.759 0.2 115 116 117 116 Ini. Mks 0.007”
3226 I 1.757 I 1.752 ] 1.750 I 1.756 I 0.3 ! 117 I 113 I 112 \ 114 Corrosion Pitting
3235 I 1.737 ! 1.741 I 1.769 I 1.729 ! 2.3 I 115 I 115 I 115 I 114 Inj. Mks 0.007”
5ooo I 1.739 I 1.754 I 1.743 I 1.738 I 0.9 I 113 I 108 \ 111 I 110 Inj. Mks 0.007”. CP
7500 1.752 1.738 1.752 1.738 0.8 112 113 115 115 Score 0.004”. CP
11000 1.749 1.740 1.738 1.738 0.6 113 112 111 113 Inj. Mks 0.009”, CP
13000 1.754 1.750 1.749 1.748 0.3 113 115 111 112 Corrosion Pitting
14100* 1.74 1.740 1.753 1.749 0.7 115 117 114 114 Bias weld
14101* 1.75 1.748 1.749 1.752 0.2 126 130 131 126 Bias weld
14402 1.752 1.750 1.758 1.756 0.5 127 127 129 123 Injector Marks
15275* 1.76 1.742 1.735 1,755 1.4 137 136 136 136 Iniector Marks 0.004”
15276* 1.75 1,752 1.749 1.748 0.2 159 152 157 158 Bias weld
16350* 1.75 1.740 1.752 1.751 0.7 161 157 159 159 Bias weld
16351* 1.76 1.752 1.748 1.750 0.7 180 178 179 180 Bias weld
17432 1.753 1.753 1,749 1.750 0.2 182 177 181 178 Iniector Marks 0.009”
J

20175 I 1.749 I 1.748 I 1.750 I 1.750 I 0.1 I 189 I 189 I 189 I 188
21750* 1.76 1.752 1.754 1.750 0.6 189 187 186 189 Injector Marks 0.005”
22020 1.751 1.753 1.752 1.750 0.2 195 197 197 198 Injector Marks 0.008”
23059 ] 1.749 ] 1,752 ] 1.750 I 1.751 I 0.2
Additional Comments: Injector marks & light corrosion throughout the fill length of the string. Bed-
wap not removed from reel.

[nspector GK Date 4 Sept 97

99
10 R. K. STANLEY SPE 46023

r---l’””’

D
Figure l-i3ias weld AC
con source

Downloaded from http://onepetro.org/SPECTWI/proceedings-pdf/98CTR/All-98CTR/SPE-46023-MS/1927953/spe-46023-ms.pdf/1 by Repsol Services Company user on 27 May 2024

EddyCurrent MfGo~ measurement

Figure 6a-Principle of Eddy Current Iift+ff measurement used to

measure ovality of used coiled tubing.

Figure 2—Magnetizing system wttfr fields H., Hd shown

8 eddy current
lift-off sensors
at fixed dist.
Sensor Field Sensor
Siglul contains wall from tube /

X“
h
J
I Wlm

Dlstaoce

Figure 3: Ambient wall and flaw signai recorded by hall sensor.

OVALITY MEASUREMENT FOR
OVALLED COILED TUBING

Figure 6b-Schernatic for Ovality measurement using eddy current

iift4Mf sensors.

*I ‘“=
WALL THICKNFSS

Figure 4-The concept of wali thickness

nai field strength measurement.
measurement from exter-
w ‘c”=”’
Figure S-Eddy Current Method For Longitudinal Flaws.

100
SPE 48023 RESULTS FROM NDE INSPECTION OF COILED TUBING 11

Chart Recorder

b.
10.156 I
I
- ‘ D
❑lncl

Downloaded from http://onepetro.org/SPECTWI/proceedings-pdf/98CTR/All-98CTR/SPE-46023-MS/1927953/spe-46023-ms.pdf/1 by Repsol Services Company user on 27 May 2024

Elncl
Ilrlrl
Figure 8-Ultrasonic method for tube welds

FLAW REMOVAL- CYCLE DATA, 1.75 X 0.134 CT80

Figure 6a-Ultrasonic wall thickness measurement

800.0

700.0 m

8M.O
500.0

T’” TUBE-TO-TUBE
400.0
300.0
200.0
100.0
..,,,.,,,.-. ...

. . . .. . ... .
——-.

\\ WELD
0,0
VIRGIN PIPE 10% REMOVED 10% NOTCH

[ \ \
\’ FILM Figure 6-Cycle data at 3 kpsi on virgin 1.75 in. x 0.134 in. CT80
\
material. with 10% deep transverse EDM notches, and IOYDremoval
areas to”simulate flaw &noval.
Figure 7-Double wall Radiography for Tube-to-tube weld flaws.

101
 Downloaded from http://onepetro.org/SPECTWI/proceedings-pdf/98CTR/All-98CTR/SPE-46023-MS/1927953/spe-46023-ms.pdf/1 by Repsol Services Company user on 27 May 2024

._
—<__—,
.
—-.
—.
— -. — — .-
OF COILED TUBING
FROM NDE INSPECTION

102
and a gouge 0.026 in. deep.
RESULTS

Fig. 10b--Piilng
SPE 46023
——
 R. K. STANLEY SPE-4k023
_

1.1 %

.109 /. 8 .-* . . .:.. .! .!. ,. .: . !.

.:..—. -., .-L&.+... - ,. -.. . ..s .. . . .:. .....
2-2 x

Downloaded from http://onepetro.org/SPECTWI/proceedings-pdf/98CTR/All-98CTR/SPE-46023-MS/1927953/spe-46023-ms.pdf/1 by Repsol Services Company user on 27 May 2024

. _.:. -:—.____-.. :
1-3

1.4

,0082 ‘
+.. , ——
..-=
. . . . . . !.. i:.
~IAH~:TtiP
---.-, ... .. .. . ..~
.. :
;. ....+-.- :. .,,: ._.,..- :.. . . . . :.. 1:~’:!ii;~
,-5 , , .- ---- ------ --k--, --- ---, —-.V-— - f
LUA+ ;::.;’ ::;:: ,,: :,
.&j .: ‘ I
.1347 ‘: “ : P“’ ~“”’~’”’”i’ ““!’ ~ ! ““ ““ ~ ; ‘~ “’~ ““ !’” ‘ “:’ “~ ‘“”~ “~ ‘“: ““: “:” ““i” ““ !“”’”!’ ! : ~ ‘“ ;“ ;
,08 J3,:..: .i, .,; ..,. !.i. .i .,,,,: : ; : ; ,.~. ;.{ : ;. ..: : ! .’ ,..
; f. ..:..;..:.; :
~1 AM.EJEJL’..J --:-- . . . . .. :,-.:.-:- . .. . . “._ .:.. LfJ.Lf4’LI’TfiZEFIAMTT
;... .- , -. . -. ~~ ,.
.-, ... L_. .- ..; ~ . . . ..
= DIA-_i---
,2120;...,,.:.
;“. ..,.:. : i ~ .,,;,,:
.. f.. :. ..1 .,
;
;..:
; ,;, , :
: .:..
~
:: ~ :
!“ ;-’~’”=”’’:-:-
“
~“-”’”-””;
~~ ‘ “:
:; ;
““_:_
,
:
.,, ...!.:
:
: .:!
, :

9870:: ;..:..:: . . . . . . :.:.:. . . . . “../. . . . .

.-- ...-: ::- . . . . . . ... i: . . -. . . . ...-. --., .
.7 %lt: ;.:::;.
-1 S30 —-
3S 652 .;. , .~’. : ,.;. ,,; . ,: :.,, : ~. :
-——— .
:.
_T hL\.S.P.e.cJ .-~ :1..;.Q,. .:. ,: ___ ;
3.9 ,.

-4 348 ‘:”: ~”’.:’:’; —

3A-SE 4-22- 7 16:00:43
?.. . . ~:,y,,,”,,i;,,,;.,,. ..+ .,:.,.,..:,,’.:i:~.~
_ . . . . .. !~!$i%%%<:%?::a:f$” .,, :,.! .-. k.! .+:,..’...;: : .,., ; . .-.

Fig. 1la-Diameter variations, welds, and transverse flaws detected on test pipe.

,:.,
1.1 :<
-.380
—--
:
~ “-”-i--:--:-
:+ :
‘!’’i!”~’
-.
l-l L
,.. . . .. . !.. . ,. ,.

Fig, 1l~degree test flaws on test pipe OD. (Large signals are welds, also detected by eddy current.)

103