Back to Basics

An Overview of the Nondestructive Inspection
Techniques for Coiled Tubing and Pipe

by Roderic K. Stanley*

 

There are books and TV programs that tell nontechnical people "How Do They Do That?" Dr. Stanley's article is a good presentation of how do they do nondestructive inspection of coiled tubing and pipe. He also indicates what more is still needed and provides an introduction to the vocabulary used in this relatively new field. Good basic stuff! Frank A. Iddings Tutorial Projects Editor

 

Introduction 

Coiled steel tubing and pipe in the diameter range 20-90 mm (0.75-3.5 in.) are replacing conventional oilfield materials for a variety of purposes including workovers, drilling, production tubing, umbilicals, and flowlines. They offer all the advantages of long tubes with no threaded connections. Produced lengths of 7,620 m (25,000 ft) of 32 mm (1.25 in.) dia, and 1,800 m (6,000 ft) of 90 mm (3.5 in.) dia, coiled onto a drum are now commonplace. Today coiled pipe is being coated inside and out for flowline use, then laid in sections 1.6 km (1 mi) long.

Because coiled tubing is being produced to high quality standards, it is lasting longer than ever before, and the need has arisen for careful nondestructive inspection at frequent intervals to determine accumulated damage to the string and the need for repair. Currently, derating of used coiled tubing using nondestructive testing (NDT) is not performed. While NDT devices for oilfield tubulars have been well documented (Stanley, 1992), little has been written regarding the NDT of coiled tubing (Papadimitriou and Stanley, 1994; Stanley, 1996). This paper outlines the current NDT methods used during the manufacture of new tubing and the inspection of used coiled tubing.

 

Manufacture of Coiled Tubing and Pipe
New coiled tubing is manufactured from hot-rolled steel strip of specified dimensional (width, thickness) tolerances to standards outlined in American Petroleum Institute recommended practice API RP 5C7, customer specifications (Halliburton Energy Services), or internal specifications (Quality Tubing, 1995). Several strips are joined end to end and then welded into tube form by the high frequency induction method to produce very long lengths in which the strip welds are almost invisible. By careful heat treatment and polishing of the strip weld region, it is now possible to produce welds that cycle at 90 percent of the cycle life of the parent tube, as determined by a standard fatigue test machine. Lack of two dimensional and three dimensional imperfections, coupled with lack of residual stress in the strip weld region, has made this possible. This is extremely important since coiled tubing spends most of its life in the bent state. Strips of slightly different thicknesses are also welded for specific wells. A recent innovation has been to produce tubes that are thicker at one end than the other from strip that has been rolled continuously thinner over several thousand feet. These permit extended reach in deeper wells. Grades currently produced are given in Table 1.

Table 1 - Grades of coiled tubing and pipe
Grade	API RP 5C7 Equivalent	Minimum Yield Strength		Minimum Tensile Strength		Maximum Hardness (HRC)	Usage
		kPa	lb/in.2	kPa	lb/in.2	(HRC)
QTP-52	X-52	360,000	52,000	-	Not rated	22	Pipe and flow lines
CT-55	CT-55	380,000	55,000	450,000	65,000	22	Downhole coiled tubing, umbilicals
HO-60	-	410,000	60,000	480,000	70,000	22	Downhole hang-off
QT-700	CT-70	480,000	70,000	550,000	80,000	22	Downhole coiled tubing, umbilicals
QT-800	CT-80	550,000	80,000	620,000	90,000	22	Downhole coiled tubing, umbilicals
QT-1000	-	690,000	100,000	760,000	110,000	Not rated	Downhole hang-off

 

NDT of New Coiled Tubing
The strip welds require radiography (API RP 5C7) and possibly magnetic particle inspection (Quality Tubing, 1995) to ensure elimination of all flaw types. Figure 1 shows a typical digital radiograph of the bias weld. Radiographic inspection is performed to ASME Boiler and Pressure Vessel Code Section V, Article 22 (ASME, 1992) or, for pipeline product, a requirement somewhat tighter than API 1104 using a No. 10 penetrameter with a 2T hole. Magnetic particle inspection is performed with a yoke to API RP 5A5.

 

---

Because coiled tubing is lasting longer than ever before, the need for careful NDT has arisen.

---

 

The second inspection is tube full body or high frequency induction weldline eddy current inspection. In the case of weldline inspection, a sectored pickup coil is used while the steel is above the Curie point (and therefore paramagnetic rather than ferromagnetic), permitting penetration of the eddy currents through the weld. With a full encircling differential coil system, performed when the pipe is running through the cold reduction section of the mill, imperfections around the 360 degree circle of the pipe are detected.

The aim of the differential system is to tune out nonrelevant indications such as those from permeability variations at strip welds and small changes in wall thickness that occur over long intervals, such as a tapered bias weld, and detect only the two dimensional and three dimensional imperfections that are considered detrimental to the service life of the pipe. Standardization of the eddy current unit, which generally operates in the 2-20 kHz range, is performed with an 0.8 mm (0.03 in.) or 1.6 mm (0.06 in.) through drilled hole in accordance with ASTM E-309.

New coiled tubing has also been inspected with a standard oilfield tubing electromagnetic inspection device (Figure 2) which contains two inspection heads. The first head has a ring of overlapping sensors that detect the magnetic flux leakage from a longitudinally magnetized section of the tubing, while the second rotates magnetic flux leakage sensors around the pipe circumference as part of a magnetic yoke. Each sensor of these two systems is standardized on the same test standard as the electromangnetic testing (ET) unit; our experience has been that both the ET and the magnetic flux leakage systems can detect imperfections as small as 0.076 mm (0.003 in.) deep on the outside diameter surface when standardized to an 0.8 mm (0.03 in.) through drilled hole.

 

Figure 1- Digital radiographic inspection of strip weld prior to manufacturing coiled tubing from strip.

Figure 2 - NDT 3500 electromagnetic inspection unit inspecting coiled tubing.

Figure 3 - Prove-up flow chart for coiled tubing.

Should imperfections be detected during these inspections, the pipe areas are "proved up" using a variety of techniques (Figure 3). Outside diameter surface imperfections (Figure 4) are detected visually, by liquid penetrant, or by magnetic particle inspection. Inside diameter surface problems are detected by radiography, compression wave, and shear wave ultrasound (Stanley, 1994). Imperfection removal is performed by fair-contouring the pipe's outside diameter wall with sandpaper and leaving no transverse scratches that would act as stress risers when the pipe is being cycled into and out of a well.

Finally, while all major oil country tubular goods/line pipe mills inspect the high frequency induction seam with shear wave ultrasound (Figure 5), this is not currently performed for coiled tubing or pipe, unless at customer request after hydrostatic testing in order to comply with API 5L SR 17. The internal flash, when left in tubes smaller than 38 mm (1.5 in.) outside diameter, makes this impossible. One might anticipate that at some future point there will be an API requirement for flash-free coiled tubing material for sizes for which flash removal is possible, inspected by ultrasound. The idea here would be to catch small weldline imperfections such as black spots, hook cracks, and banding (Figure 6), which do not always open up to become leakers on hydrostatic tests.

Imperfections are removed as a matter of routine to 95 percent of the wall thickness in downhole product and 87.5 percent of the wall thickness in hang-off and pipeline product (API 5L). Imperfections in downhole product may be removed to 87.5 percent in consultation with the customer. These remaining wall thicknesses were selected as a starting point from our oil country tubular goods experience.

API RP 5C7 requires that manufacturers tell customers exactly where these indication removal areas are, indicating that at some point in the future, baseline wall thickness data on new strings will be required so that the cycle life of each pipe section can be determined by computer as the wall thickness varies within the criteria outlined above. This is believed to be preferable to butt welding the pipe that is used for workovers and drilling. The best butt welds have been found to have a fatigue cycle life of about one quarter of that of the parent tube. Tubes with low wall sections, simulating indication removal, are the subject of a study of fatigue life.

Production of such strip, and steel mill NDT procedures, permit a limited number of small imperfections to be accepted that do not reach NDT reference levels. These include small surface blemishes, slight ovality, and some banding in the steel. It is fair to say, however, that only a few new strings will have had imperfections removed. (Current electromagnetic inspection runs at 0.77 indications per string.)

Figure 4 - Outside diameter surface imperfections: (a) roll in, (b) open seam.

 

Figure 5 - Weld inspection by shear wave ultrasound.

Figure 6 - Inner defects: (a) dimple in inner flash and pipe wall, (b) transverse cracks in flash (UVA indications).

Figure 7 - Flux method for wall assessment of coiled tubing (a) magnetizing head with constant lift-off sensors. Demagnetization poles P ₁ and P₂ cause a reverse field which affects the applied field H_a in such a way as to permit wall thickness to be measured; (b) typical wall loss results from a ring of four sensors (c).

 

Used Coiled Tubing
Some of the methods applicable to the NDT inspection of used steel tubulars are applicable to the inspection of used coiled tubing. Electromagnetic methods appear preferable to ultrasound since it has generally been found to be the case when inspecting new tubes with ultrasound that the pipe must be "squeaky clean"; this provides for even coupling of the sound into the wall of the tube, that is, constant values for transmission and reflection coefficients when using shear wave for flaw detection. The condition of the outside surface of used coiled tubing seems to favor electromagnetic noncontact methods.

The problems associated with used coiled tubing are listed below, along with methods for their detection.

• The Bauschinger effect lowers the yield strength but not the tensile strength, immediately suggesting a derating for those coiled tubing jobs for which pressure is a consideration. This occurs in the first several cycles of use. Since one might suppose that yield strength lowering implies movement of crystal planes in the affected area, with a concurrent change in electrical conductivity and magnetic permeability, the use of eddy currents to detect the region has been implemented in two tools. In one tool (Papadimitriou and Stanley, 1994; Stanley, 1996), an encircling coil which will also detect ovality is used. In a second tool (Rosen, 1996), the use of localized eddy current sensors has been proposed.
• Ovalling and diametrical expansion commonly occur and may be detected with a variety of methods, including eddy current encircling coils (Papadimitriou and Stanley, 1994), more localized sensors (Rosen, 1996), ultrasonics (Livingstone, 1994) and proximity sensors (Corrigan and Grey, 1991). The eddy current encircling coil will detect the area since it is sensitive to changes in fill factor (defined as D_pipe²/D_coil²), although it will not give a measurement. Localized eddy current sensors are being programmed to provide actual measurements. One ultrasound system, which has eight compression wave transducers operating in the range of 7.5-8.5 MHz equally spaced around the tube, measures both wall thickness and proximity of the pipe to the sensors. From these readings, maximum and minimum diameters are computed. The proximity sensor system is an LVDT system that operates more or less the same way as the eddy current system. These systems should be able to detect 1 percent changes in outside diameter and ovality, defined as (D_max - D_min)/D_nominal so that derating for collapse pressure can be performed.
• Damage with a transverse component, which initiates from manufacture or from use, provides stress risers for cracking to initiate. Gripper block damage and slip marks on the outside diameter have been found to be extremely detrimental to cycle life. Hardnesses in the range of Rockwell hardness 35 (R_C 35) at the roots of such damage has been encountered in material originally specified as R_C 22 max. Fatigue cracks and pits on the inside and outside diameter walls have led to failure. These are detected with magnetic flux leakage technology using rings of surrounding solid state sensors (Hall elements) (Rosen, 1996) or inductive coils (Papadimitriou and Stanley, 1994) and magnetizing the pipe longitudinally. The transverse head of the NDT 3500 unit may also be used for this inspection. The sensitivity of magnetic flux leakage to such flaws should be such that a 5 percent transverse notch 3.175 mm (0.125 in.) long and 0.15 mm (0.006 in.) wide (maximum) can be readily detected when placed on the inside and outside diameter surfaces.
  
  Internal pitting in coiled tubing is sometimes caused by incomplete flushing and neutralization of the working fluids, e.g. hydrogen chloride, sodium chloride, potassium chloride. While larger pits present no detection problem, those in the depth range of 2-5 percent of the specified wall thickness (2.2 mm < t < 5.15 mm [0.087 in. < t < 0.203 in.]) may be difficult to detect above the ambient magnetic noise and are critical in determining when a tube might fail. However, magnetic flux leakage is the preferred method because of surface contact problems. The test standard should also have internal and external pit-like flaws in this depth range.
  
  Strip weld sanding scratches as shallow as 0.05 mm (0.002 in.) on the inside diameter of the strip weld have been known to act as stress risers and cause failure, and while these would most probably have been detected during manufacture, inspection would have detected the cracks as they grew from the roots of the sanding marks.
• Loss of wall on the outside diameter by abrasion can be detected by several methods, although in current used coiled tubing inspection equipment, the following are used. In one unit, the encircling eddy current coil method will detect the problem as a decrease of fill factor, but will not measure the remaining wall. In the same unit, a magnetic flux method using solid state sensors at fixed lift-off (Figure 7) will provide wall thickness measurements around the tube. In the NDT 3500 unit, a rotating gamma ray gage is used. This has a coverage of about 2 percent of the wall in a "barber's pole" scan, which is sufficient to detect the problem but not to measure it.
• Flash problems: coiled tubing internal flash, because of its potential for inconsistency, causes three basic problems.
  
  Loss of wall at inner diameter flash. The corner between the pipe and the flash wall, especially if not rounded, and the flash crown itself, will attract paraffin and corrosive products. Cases have been found where severe wall loss occurs at this corner and the flash crown has corroded. In some cases this has led to longitudinal cracking of the pipe at the root of the flash wall. Rotating field magnetic flux inspection may detect this condition.
  
  Flash height. Flash height is sometimes very inconsistent, leading to regular and irregular undulations in flash height. Because of the surface tension effects of the cooling of excess flash into almost cylindrical and spherical surfaces, undulating flash of the type shown is formed at 3.175 mm (0.125 in.) to 6.35 mm (0.25 in.) intervals. Between such spheres, it has been found that the flash will crack transversely under bend loading.
  
  Flash crown hardness. Experimental work at Quality Tubing Inc. has led to the fact that flash crown can be harder than the material of the interior flash. Cycling tests have shown that tubes with flash with harder crowns tend to cycle to failure by transverse cracking in fewer cycles than similar tubes with softer flash crowns.
• Field Welds. Pipe is often field welded under less than ideal conditions, and while such welds might yield good X-ray plates and pass both magnetic particle and liquid penetrant inspection, the condition of the weld root is uncontrolled, and the remaining cycle life of the tube is low in comparison to the tube on either side of the weld. Here current NDT methods provide little assurance of the longevity of the girth weld under cyclic conditions.
  
  The number of cycles from crack initiation to final failure in all possible failure modes is not yet known, but is under investigation. Much research needs to be performed to determine the remaining cycle life of, for example, lightly pitted coiled tubing. Such pits, depending upon their aspect ratios, can obviously be stress concentrators, and lead to transverse crack initiation under tensile loading while still extremely small. This was observed on grade S-135 Ocean Drilling Program drill pipe several years ago where salt water corrosion had produced very small hemispherical pits.
  
  What is important here is that while NDT can generally detect the conditions described above, how detrimental they are in relation to the accumulated fatigue in the pipe wall remains to be seen. For example, those failures in coiled tubing that apparently initiate at transverse cracks in flash crowns require that the crack be detected early in its life. How we will accomplish this while inspecting an entire string is currently not known.

 

Current Used Coiled Tubing
Inspection Requirements
An optimum used coiled tubing inspection system should be able to detect all of the above conditions in one inspection head and store pipe wall data so that comparative NDT runs can be made after an average of 4-5 jobs. Each section of pipe (e.g. 3 m [10 ft]) can then have its own remaining life computed from existing theoretical programs (Brown, 1995) as a guide to assessing the risk associated with rerunning the tubing. This is essential when tubing has survived, for example, in excess of 20 trips into the hole. Interfacing NDT with such programs represents an essential step in providing accurate information about each string. Existing programs already allow for worn pipe sections to be removed. The effects of yield lowering (Bauschinger Effect) and ovality (lowering of collapse pressure) need to be included.

Wall thickness prove up should be conducted independently with high frequency compression wave gages, standardized on curved test blocks, and radiographic testing (RT) where necessary. Butt welds, when added to a tube, should be inspected by RT for three dimensional problems and either magnetic particle or liquid penetrant inspection for two dimensional imperfections on the outside diameter. Shear wave ultrasonics (Stanley and Wells, 1994) should be used on such welds if it is found to be effective in the wall thickness region of coiled tubing (2.2-5.15 mm [0.087-0.203 in.]).

One essential paper requirement is the production of an acceptable recommended practice on care and inspection of used coiled tubing. This document should contain clear statements regarding the derating of all sections of used coiled tubing Ñ including field welds Ñ from the findings of NDT, dimensional checks of ovality, weld cycling data, and current field experience with field failures.

Finally, inspectors should be trained in the relevant inspection techniques, especially magnetic flux leakage.

 

Addition of Butt Welds to Used Tubes
Field butt welds are made when the pipe has failed, often from a fatigue crack. This failure does, however, provide the ability to cut sections from either side of the failed area and perform an assessment of the tube bore for such conditions as cracking in the flash, for cycle testing to failure on a standard fatigue machine, and for measuring the tensile properties on the string in that area. Data such as these can indicate possible remaining life of the rest of the tubing, and provide derating for the Bauschinger effect.

 

Conclusions
NDT methods relevant to steel tubing are needed in combination with existing theoretical methodology for estimating the life of coiled downhole tubing in order to provide a more accurate derating of used coiled tubing than is presently performed.

 

References
ASME Boiler and Pressure Vessel Code Section V, Article 22, 1992 ed.: SE-94 "Standard Practice for Radiographic Inspection."

Brown, P., "Calculate Coiled Tubing Fatigue in Real Time Using Computer Software," Proceedings of Coiled Tubing Technology, Mar. 13-16, 1995, Houston, TX. Gulf Publishing.

Corrigan, M., and B. Grey, "Determining the Working Life of a Coiled Tubing String," Petromin, Mar. 1991.

Halliburton Energy Services, "Specifications for Coiled Tubing, QT-700, QT-800, QT-1000," Transocean Well Service specification for coiled tubing.

Livingstone, W.A., "Method and Apparatus For Coiled Tubing Inspection," US Patent 5,303,592, Apr. 19, 1994.

Papadimitriou, Steve, and Roderic K. Stanley, "The Inspection of Used Coiled Tubing," Proceedings of the 2nd International Conference on Coiled Tubing Operations, 1994, Houston, TX. Gulf Publishing.

Rosen, Patrik, "Automatic Coiled Tubing Inspection Monitoring," Proceedings of the 4th International Conference on Coiled Tubing, Apr. 1996, Houston, TX. Gulf Publishing.

Stanley, Roderic K., "Repair Grinding, An Alternative to Ultrasonic Rejection for Oil Field Tubes," Materials Evaluation, Vol. 52, No. 10, Oct. 1994, pp 1161-1164.

Stanley, Roderic K., and Larry Wells, "Recent Advances in Used Drill Pipe Inspection by Ultrasonic Methods," Materials Evaluation, Vol. 52, No. 11, Nov. 1994, pp 1282-1285.

Stanley, Roderic K., "Electromagnetic Tubular Inspection During Well Servicing," Proceedings of the 13th World Conference on NDT, 1992. Elsevier, Holland.

Stanley, Roderic K., "Imaging of Magnetic Flux Leakage Signals for High Quality Assessment of Oil Field Tubular Products," Proceedings of the 13th World Conference on NDT, 1992. Elsevier, Holland.

Stanley, Roderic K., "Nondestructive Evaluation of New Coiled Tubing and Pipe," Proceedings of ASME Energy Week Conference, Jan. 1996.

Technical Catalogue, Quality Tubing, Inc., 1995.

Bibliography - American Petroleum Institute Publications
API 1104, "Welding Pipelines and Related Facilities."
API 5CT, "Specification for Casing and Tubing."
API 5L, "Specification for Line Pipe."
API RP 5A5, "Recommended Practice for Field Inspection of New Casing, Tubing and Plain End Drill Pipe."
API RP 5C7, "Recommended Practice for Coiled Tubing Operations in Oil and Gas Well Services," in press.

 

*Quality Tubing, Inc., Box 9819, Houston, TX 77213-0819; (713) 456-0751, fax (713) 456-7620.

Copyright © 1996 by the American Society for Nondestructive Testing, Inc. All rights reserved.

 

---

 
Copyright © 2012 by the American Society for Nondestructive Testing, Inc. ASNT is not responsible for the authenticity or accuracy of information herein. Published opinions and statements do not necessarily reflect the opinion of ASNT. Products or services that are advertised or mentioned do not carry the endorsement or recommendation of ASNT.

IRRSP, NDT Handbook, The NDT Technician and www.asnt.org are trademarks of the American Society for Nondestructive Testing, Inc. ACCP, ASNT, Level III Study Guide, Materials Evaluation, Nondestructive Testing Handbook, Research in Nondestructive Evaluation and RNDE are registered trademarks of the American Society for Nondestructive Testing, Inc. ASNT exists to create a safer world by promoting the profession and technologies of nondestructive testing.