Detection of coating defects in pipelines using in-line inspection tools

Introduction
While the material of choice for most high pressure pipelines is steel, it is the polymer coating that renders steel pipelines competitive against other materials. Internal coating and especially external coating of steel line pipe is vital to ensure the long term integrity of the pipeline. While the pipeline steel has been inspected for defects for a long time, the testing of the coating has so far been limited to indirect methods, which rather test its influence on the CP system of the pipeline, than its material properties.

For the in service inspection of the pipe steel intelligent pigs have evolved into the method that most operators rely on. Nowadays the steel material is inspected for metal loss type flaws as well as crack-like flaws. Mainly magnetic flux leakage (MFL) tools and ultrasonic (UT) inspection1) are used. EMAT (Electromagnetic Acoustic Transducers) inspection technology is certainly an emerging inspection method.

Internal coating of pipelines fulfills a multiple task. For gas pipelines its main purpose is to reduce the friction. Its function is thus not affected by small faults in the coating. Internal coating is usually factory applied which leads to a reasonable long-term reliability. For liquid lines the coating is meant as a corrosion protection. Since the full internal surface needs to be covered internal coating is often applied in-situ. Naturally the process cannot be easily controlled and the quality of the coating may suffer.

Quality may be limited due to disbondment, lack of full coverage or insufficient thickness. Also excessive thickness, whilst being a waste of valuable material, leads to problems for other inspection methods. As will be shown, ultrasonic intelligent pigs are capable of detecting most of these flaws. In some cases this can even be done in conjunction with a regular inspection of the steel wall.

The main purpose to the external coating is to prevent corrosion from the outside. Almost all pipelines today have a cathodic protection system in place to ensure that anodic, steel dissolving reactions will not take place. In this context the external coating merely ensures that no excessive protective currents are required. However, CP-systems may also malfunction without immediate warning. Thus any imperfection of the external coating is a potential corrosion site.

These coating faults can be detected with close interval potential surveys (CIPS) or other means [1]. With many factors interfering with the CIPS results, a direct measurement of the coating is desirable. Naturally it is difficult for intelligent pigs to do any measurement through the steel layer. Some tests with ultrasonic inspection tools will be presented that show encouraging results.

Ultrasonic wall thickness measurement
Traditional ultrasonic wall thickness measurement of steel pipelines employs a pulse-echo method. Short pulses of ultrasonic waves are emitted from a piezoelectric transducer. The reflected echo is received and recorded. The time-offlight information is used to recalculate the thickness of the pipe wall. Reflection takes place at any interface, where the acoustic impedance changes. The acoustic impedance Z is a material specific parameter that describes the propagation of ultrasonic waves. It mainly depends on the speed of sound in the medium. An interface between steel and air or gas leads to a much higher reflection as an interface between steel and a liquid or a polymer. The reflection factor for materials with impedances Z₁ and Z₂ is given by

in analogy to the reflection of electromagnetic waves at material interfaces of refractive index. For this reason the ultrasonic inspection tools require a liquid couplant. The reflection of the sound waves at the interface gas/air is so powerful that subsequent echos are drowned out.

In case of a coated pipeline several sheets of material are present. The interfaces medium/internal coating, internal coating/steel, steel/external coating all lead to reflections with varying reflected amplitude. The circumstance is illustrated in Figure 1.

For better clarity not all possible reflections are shown. The amplitude of the waves after many reflection and splits into a reflected and a transmitted part is fading quite quickly. The reflection from the external surface of the external coating is quite hard to measure. In the case of an internally coated pipe this sound beam would have to pass seven interfaces.

Contrary to older ultrasonic inspection tools, tools currently employed at NDT Systems & Services can be configured to store the complete echo signal [2]. All reflections are recorded, assuming that the amplitude is still sufficiently high. Usually the wall thickness is calculated on-line by identifying the entry echo and the steel rear wall echo and subtracting the corresponding time-of-flight values.

In the case of coating being present this calculation can lead to errors. With the new technology the calculation can be adapted and an off-line recalculation is possible depending on the coating conditions.

In contrast to ultrasonic inspection, MFL has no potential to inspect coating. The reason for this lies in the nature of electromagnetic testing. Polymers neither conduct electric currents nor magnetic flux. Hence there is no interaction of electric or magnetic fields with polymer coatings. The EMAT technology, which is an ultrasonic inspection method, may also eventually prove some capabilities for coating inspection [3]. Optical methods have proven to be a reliable method to check the condition of internal coating [4].

Internal coating
For testing the internal coating usual inspection tool settings would either calculate a wall thickness of internal coating plus steel in case of thin internal coating or of thickness of the coating alone in case of thick internal coating. In case of very thin internal coating like factory applied coating for gas pipelines a subsequent distinction of coating and steel is not possible.

The aim in testing the internal coating may consist of finding the areas of removed or improper coating. If internal coating is removed or not present at all a change in the stand-off signal is recorded. The stand-off signal is a measure for the distance of the first material interface. If a piece of coating is missing this distance is increased. A gradual thinning of the coating would not be detected this way. Figure 2 shows a circular area with removed coating.

The upper part shows the wall thickness values. The black areas show the measured thickness of the steel plus the internal coating. The gray area shows the gap in the coating. The "wall thickness" is reduced. The absolute value is incorrect, because it is based on the speed of sound of steel alone. The measured difference is about 0.8 mm based on the speed of sound in steel. With a ratio of 2:1 for the two speed values the coating thickness is about 400 ηm. The lower part shows the stand-off values based on the speed of sound of the medium. The coating thickness is found to be the same.

If the inspection is not only meant to examine the coating, but also to detect metal loss flaws, a better distinction between coating and steel is desirable. For this purpose three sets of data can be compiled. The complete thickness of coating plus steel, the coating alone and  the steel alone. Of course, one can be directly derived from the other two, but in this case all three values are calculated independently from the echo data. In the following three screenshots of this data are shown. The upper part of the figures shows a C-Scan, the ordinate axis shows the orientation in degrees, the lower part a B-Scan of a trace approximately in the center of the C-Scan. The ordinate axis shows the thickness in mm.  

Figure 3 shows the data obtained with standard settings. In this case the thickness refers to the steel wall plus the coting thickness. The red color in the C-Scan refers to the thickest coating, the gray color to thinner coating. The coating is only present around the 6 o'clock position. The black stripes on the upper und lower boundary represent uncoated areas.

The B-Scan below is a reduced BScan which shows the calculated total wall thickness using the speed of sound in steel. The dotted red line is the steel wall thickness as measured in the uncoated part. The blue line shows the contour of the coating. Note that for all figures the thickness for the coating would have to be corrected for the different speed of sound in epoxy. Since this coating is insitu applied an even thickness is not expected.

Figure 4 shows the same spot of data with the thickness representing the coating alone. In the C-Scan gray areas represent the thickest parts, while yellow parts are thinnest. The blue bands on the upper part again show the uncoated steel. For the areas on the rim of the coated parts the polymer thickness is too thin to be resolved. In this case the color is also blue.
The B-Scan of Figure 4 shows the thickness of the coating. The profile is much better visible as in Figure 3. The thickness of the coating varies from zero to about 3.5 mm.

Finally in Figure 5 the thickness of the steel alone is shown. The unaffected steel is given by the black color. The BScan shows a rather straight line with a value of 12.8 mm. In the lower part of the C-Scan to the left a metal loss defect is found. It is under the coating on the internal side.

External coating
The obtain information on the external coating is much more difficult, because any measurement using intelligent pigs needs to measure through a sheet of steel. However, for external coating the question is not so much about its existence or its thickness, but much more about the bonding to the steel.

Disbonded coating may results in bare pipe being exposed to the corrosive environment of the soil. Different aereation of the exposed steel may lead to corrosion under the condition that the pipe-to-soil potential drops below the critical value. As in-line inspection with intelligent pigs has shown these conditions exist in buried pipeline systems and some control of the prevention of corrosion is required. Reliable detection of disbonded coating would thus be a great step forward in integrity surveillance.

So far only the time-of-flight information of the echos has been exploited. As stated in equation (1), the reflected amplitude depends on the change in acoustic impedance. An interface steel/coating has a different reflectivity than an interface steel/soil or steel/water. Unfortunately the difference is not dramatic, but the amplitude of the received echo signal should depend on the rear wall interface.

As shown in Figure 1 the effect should be larger for those echos that have encountered several reflections at the rear wall, i.e. the third of forth rear wall echo. Thus the amplitude of this signal is examined for areas, which are known to have a specific change in sound reflectivity. A test was performed in a trial inspection run on a bare pipe. A defined part was externally coated with an epoxy resin. By plotting the amplitude of the third rear wall echo it was attempted to detect this area. The set-up is shown in the left part of Figure 6.

To the right side of the flange a card-board template defines a rectangular area, where the resin is applied. The C-Scan²) is shown in the right part. The flange is seen to the left in blue color. In the same manner the spiral weld appears on the right. To the upper left the coated area is seen in a yellow color.

In an actual pipeline scenario the coated area would be yellow, while the disbonded area would appear red.

Two screenshots of actual field coated areas are shown in Figure 7. On the left picture the girth weld is seen to the right of the screenshot. A shade of slightly darker color surrounds the girth weld. Mainly the transition from factory-applied coating to field coating is detected. The coating itself is rather homogeneous. On the right side an older type of field coating is seen. The girth weld is in the center of the picture.

Especially downstream (flow is always from the left to the right) of the weld the interface is inhomogeneous. Whether this represents a problem would need to be verified in on-site testing. No corrosion has been found in any of the two areas. Still the type of coating and any possible deviations from the normal appearance could be detected.

Conclusions
Ultrasonic in-line inspection not only has prominent capabilities for the detection of metal loss type flaws, but offers unique possibilities to also test the coating on the internal side as well as on the external side of the steel pipe. While inspection of internal coating is already matured and can be applied any time, the inspection of external coating is still in its infancy. However, encouraging results have been obtained. To validate the method dig-out results would be required to stipulate the actual performance and specify what types on coating anomalies could be seen.

References
[1] Peabody's control of pipeline corrosion, A.W. Peabody, NACE International, second edition 2001
Korrosionsschutz erdverlegter Rohrleitungen, Kompetenzcenter Korrosionsschutz Ruhrgas AG, Vulkan Verlag Essen, 2001
[2] K. Reber, M. Beller, A. Barbian, Advances in the ultrasonic in-line inspection of pipelines, 3R international, 13/2004, pp. 53-57, Vulkan-Verlag
[3] New Inspection Technologies for Pipeline Integrity Management, Cornelis Bal, Proceedings to the Pipeline Rehabilitation and Maintenance Conference 2005, Manama Bahrain
[4] Optical in-line inspection tool for internal monitoring of pipelines, M. Beller, T. Jung, K. Vartdal, 3R International, Special Edition 13/2005, Vulkan-Verlag