New Multi-Technology In-Line Inspection Tool For The Quantitative Wall Thickness Measurement Of Gas Pipelines A. Barbian 1, M. Beller 1, F. Niese 2, N. Thielager 1, H. Willems 1 1 NDT Systems & Services AG, Stutensee, Germany 2 Institut für Zerstörungsfreie Prüfung, Fraunhofer Gesellschaft, Saarbrücken, Germany The demand for higher inspection accuracies of in-line inspection tools is permanently growing. As integrity assessment procedures are being refined, detection performances, sizing accuracies and confidence levels regarding detection and sizing play an ever increasing role. Inspection tools utilizing conventional ultrasound technology are at the forefront of technology and fulfill the market requirements regarding sizing accuracies and the ability to provide quantitative s. A draw back regarding their application in gas pipelines is their need for a liquid couplant. This paper will introduce the concept for a new generation of intelligent pigs using a multi-technology approach to provide the advantages of ultrasound in-line inspection tools for gas pipelines. This new tool makes use of different non-destructive testing methodologies, offering true quantitative wall thickness capabilities for gas pipelines, without the need for a liquid batch, thus significantly simplifying operational procedures. The accuracy of the tool will ensure the data quality needed for advanced integrity assessment, including the issues of run comparisons, corrosion growth evaluation and upgrading of pipelines for higher operating pressures. The paper will introduce the technology of the tool, its operational capabilities and defect specifications. Page 1 of 10
1 Introduction Today the use of In-line inspection (ILI) tools is a standard procedure for the collection of pipeline data required for integrity assessment and fitness-for-purpose studies. Their major task is to provide accurate geometric information regarding the length, width, depth, orientation and location of a flaw in the pipe wall. The major advantages of ILI tools is their capability to survey the entire pipe circumference whilst the pipeline remains in operation. They are usually pumped through the line to be inspected (i.e. free-swimming tools) and do not require their own drive. For specific pipeline geometries there is also a range of cable operated tools, optionally with drive units, some of which can even operate against the pipeline flow. Various non-destructive testing methods are applied, each with particular advantages and disadvantages based on the physical principles used. The major technologies applied are: Magnetic Flux Leakage Technology Ultrasound Technology Eddy Current and Pulsed Eddy Current Technology A comprehensive overview regarding the capabilities of magnetic flux leakage and ultrasound tools can be found in [1,2], although it must be noted that some comments regarding the use of ultrasound tools in thin pipe and for the detection of pitting corrosion do not apply to the latest generation of tools available on the market. Eddy current technologies are not widely used in free swimming tools today. 2 Metal Loss Inspection and Wall Thickness Measurement The major purpose of a metal loss inspection is to detect, size and locate any features which could have a negative effect on the integrity of a given line. Any metal loss leads to a decreasing remaining wall and therefore decreasing ability of the ligament to carry the hoop stresses associated with a given pressure. The purpose of detecting and sizing metal loss is therefore really to determine remaining wall. The data collected are then used as input for an integrity assessment or fitness for purpose analysis. All the data used for defect assessment, including data obtained from ILI tools, have a built-in error. This translates into tool accuracy and is stated by the ILI vendors in their tool data and defect specification sheets. These inaccuracies are actually errors in the geometric of any anomalies and flaws. When these geometries are used as input for assessment codes, errors are usually ignored. In order to compensate this, safety factors are included. Only the code of Det Norske Veritas (DNV) Part A [3] includes input of tool accuracies and technology. Reber et. al. [4] have investigated this further in their paper published 2002. They have also shown the effect of increasing accuracies on the ERF limit line. In summary it can be stated that the higher the standard of integrity assessment to be applied, the higher the quality of the inspection data utilized has to be. A quantitative wall thickness on the other hand is useful if a pipeline is to be uprated, i.e. operated at a higher pressure than before. In order to assess the suitability of a line, it is necessary to know the actual wall thickness, rather than only Page 2 of 10
the nominal value. In many cases actual wall thicknesses exceed the nominal (minimum) specification and can thus be seen as a "built in" safety factor. Ultrasound is a non-destructive testing technology which has been applied for a variety of inspection tasks for many years now. In addition ultrasound constitutes the only reliable technology currently available for the detection and sizing of cracks in pipelines. A major advantage provided by ultrasound, compared to magnetic flux leakage technology, is the ability to provide true quantitative s. This means that the actual wall thickness of a pipe section can be determined with a high accuracy and reliability. A further advantage is the ability to measure the exact contour of a metal loss feature. This implies that not only the deepest point and the axial extent of a feature can be determined, but that the exact river bottom profile can be measured with high accuracy. Ultrasound data is thus ideally suited for high level integrity assessment, such as complex shape analysis using the DNV code or RSTRENG. Today ILI tools utilizing ultrasound technology are widely used in oil and product lines. Latest generation ultrasonic tools have also overcome limitations of earlier tools, especially regarding the detection and sizing of pitting corrosion and thin wall. There is today a clear segmentation in the ultrasound market regarding defect specifications achieved. This fact has, however, not been realized by all participants yet, especially those vendors which are combining first generation ultrasound capabilities with magnetic flux leakage, claiming that the combination overcomes draw backs of either technologies. The only true disadvantage of ultrasound, when applied through piezoelectric transducers is the need for a liquid couplant. These tools or any ultrasound-magnetic flux combination tools are therefore not easily applied in gas pipelines. 3 Improving Inspection Accuracies In Gas Pipelines As inspection requirements grow, the question arises how the higher sizing accuracies offered by ultrasound technology can be achieved in gas pipelines. One possibility is the application of a liquid batch in which a piezoelectric ultrasonic tool could be run in a gas pipelines. This is a routine procedure today, but requires a certain operational effort as well as the additional costs associated with the introduction of the batch, usually requiring a temporary stop to routine pipeline operations. Another possibility is to explore which alternative non-destructive technologies can be used to achieve this improved accuracy. Usually gas pipelines are inspected with ILI tools utilizing magnetic flux leakage technology. The most widely types of tools used, referred to as high resolution magnetic flux leakage tools (MFL tools) achieve an accuracy in the order of ± 10% of wall thickness (WT). Some MFL tools available today achieve higher accuracies of up to ± 5% of wall thickness (WT). Magnetic flux leakage technology provides an indirect way of. A strong magnetic field achieving full magnetic saturation of the pipe wall is applied. In the presence of flaws, e.g. internal or external metal loss, this field can partly escape the wall. This leakage can be measured using Hall sensors or moving coil sensors. The tools basically measure a magnetic field or a rate of change of the magnetic flux respectively. Translating this information into metal loss data (volumetric metal loss to be precise) requires calibration work, whereby certain values relating to the magnetic field are correlated to varying degrees of metal loss. Each vendor uses proprietary software to achieve this. Table 1 Page 3 of 10
includes a few wall thickness examples to illustrate this issue of reporting accuracies and comparison with ultrasound. Sample wall thickness Table 1: Comparison accuracies UT and MFL. MFL - accuracies (at 80% confidence) industry standard ("high resolution"), ± 10% WT extra high resolution, ± 5% WT UT-accuracies (at 90% confidence) industry standard, ± 0.4 mm 6 mm ± 0.6 mm ± 0.3 mm ± 0.4 mm 10 mm ± 1.0 mm ± 0.5 mm ± 0.4 mm 12 mm (1/2") ± 1.2 mm ± 0.6 mm ± 0.4 mm 19 mm (3/4") ± 1.9 mm ± 0.95 mm ± 0.4 mm 25 mm (1") ± 2.5 mm ± 1.25 mm ± 0.4 mm For this purpose the standard specifications regarding accuracies for ultrasonic tools (UT) are taken. It must be pointed out that tools incorporating even higher accuracies are available today. The table clearly shows that UT tools provide an advantage regarding accuracies with increasing wall thickness. Even in case of thin wall, where MFL tools appear to show an advantage, the different confidence levels have to be taken into account. With this in mind and the advantage of true quantitative wall thickness provided by ultrasound, it is desirable to make these capabilities available for gas pipelines without the need for a liquid couplant. 3.1 Ultrasound In A Dry Environment: EMAT Technology A suitable ultrasound technology that can be applied in a dry environment is EMAT (electromagnetic acoustic emission). This technique allows external features and the remaining wall to be determined quantitatively by measuring the time-of-flight of the back wall echo signal. Figure 1 shows a typical set up of a piezoelectric transducer (left) in comparison with the set up of an EMAT transducer. Figure 1: Comparison between conventional ultrasonic transducer (left) and EMAT (right) Piezoelectric crystal Ultrasonic wave in the probe Couplant RF-coil N S Permanent magnet Air gap Ultrasonic sources Ultrasonic wave in the pipe wall Ultrasonic wave in the pipe wall Page 4 of 10
The major difference is that with EMAT transducers the ultrasound signal is generated at the surface of the pipe wall. This implies that the EMAT sensor can not measure the distance between itself and the inner surface of the pipe. However this is necessary in order to detect and size internal flaws and anomalies. Niese et. al. have reported the development and design optimization of a new EMAT sensor arrangement for wall thickness in pipelines [5], which also avoids the traditional problems relating to sensor wear. 3.2 A Multi-Technology Approach Utilizing EMAT in an ILI tool would require at least one further suitable technology in order to detect and size the internal areas of metal loss. Further technologies were therefore investigated which provide this internal capability. An additional advantage is that these technologies do not have to be applied completely separately, but can be obtained in synergy with the EMAT signal, resulting in a combined sensor. Pulsed eddy current (EC) technology and the magnetic flux leakage technology are suitable for that purpose. The EC technique is able to detect metal loss at the transducer near side by measuring the transducer lift-off. One advantage of combining both technologies, EMAT and EC, is that the EMAT excitation signal can also be used as a pulsed EC excitation. Thus both techniques are operated simultaneously. The EC receiving signal can then either be picked up by the EMAT receiver or by a separate coil. The advantage of a separate coil is that it can be optimized for EC metal loss detection. The maximum depth that can be resolved depends on the coil diameter: a larger coil diameter leads to a larger depth range. The evaluation of the received EC signals follows the classical EC signal analysis, separation of amplitude and phase of the signal for the operating frequencies. Figure 2 shows the measured amplitude for several sensor lift-offs. With the configuration used, separate EC receiver coil, a determination of the sensor lift-off is possible up to approx. 8mm. At higher depths the signal is saturated. It has been found to be beneficial to use a further technique which will deliver data regarding internal flaws. Different techniques complement each other and provide a redundancy regarding signal quality. A magnetizer is used to set the magnetostrictive operation point of the ultrasonic excitation and to magnetize the pipe wall for the MFL inspection. The EMAT coil is additionally able to detect the MFL component normal to the surface as an induction voltage. Integration of the induction voltage yields the value of the MFL signal. A separation of the receiving signal into an EMAT part and a MFL part is ensured by frequency filtering as the frequency range of the MFL signal is basically different from the range of the ultrasonic signal. Page 5 of 10
Figure 2: EC amplitude for different sensor lift-off amplitude [au] 10mm 8mm 5mm 3mm 2mm 1mm 0.5mm 0mm 0mm 0 100 200 300 400 500 600 700 800 900 1000 position [mm] 4 Experimental Results The combined sensor with the optimized coil system has been tested using a variety of specimens with machined corrosion-like defects. The results shown in figures 3 and 4 are obtained from a 10mm steel plate with calotte-shaped defects. The diameter of the calottes at the surface are 10, 15 and 30mm, the depth is 30% of the wall thickness. The principle set-up is also depicted in the figures. The sensor system is moved along the surface of the specimen, first on the defect-free side and then on the other side containing the defects. In case of external defects, see figure 3, the wall thickness can be calculated using the time-of-flight of the ultrasonic back wall echo signal. The shape of the artificial defects and the measured wall thickness fit very well. As expected, the EC technique does not show any indication. In the MFL-channel the defects also are detected well. Figure 4 shows the results when the defects are located at the internal side. In the defect areas the EMAT signal breaks down, and the ultrasonic thickness is no longer possible. However, the EC technique now detects the internal defects. The depth is determined with the help of the lift-off calibration curve (see figure 2). The measured depths differ somewhat from the real values for the smaller defects. This can be explained by the fact that the EC response represents an average value, i.e. a convolution of the sensor aperture with the defect aperture. Also here the MFL signal indicates the defects in the well known manner. Page 6 of 10
Figure 3: Results of the combined EMAT, EC and MFL inspection of external defects 15 8 wall thickness, respectively defect depth [mm] 10 5 EMAT EC MFL 4 0 MFL signal [au] 0-4 50 100 150 200 250 300 350 position [mm] Figure 4: Results of the combined EMAT, EC and MFL inspection of internal defects 15 8 wall thickness, respectively defect depth [mm] 10 5 0 50 100 150 200 250 300 350 position [mm] EMAT EC MFL 4 0-4 MFL signal [au] Further experimental tests were performed and are reported in [5]. Page 7 of 10
5 From Sensor To Tool So far an electromagnetic acoustic transducer (EMAT) has been developed and optimized for excitation and detection of linear polarized shear waves at normal incidence with the use of a horizontal magnetization of the test object. The sound field has been optimized for the of the wall thickness of components such as the wall of a pipe. Special focus is on determining the remaining wall thickness in case of metal loss. The sensor system and the magnetization unit are mechanically decoupled to reduce the wear of the sensors. In order to ensure reliable for internal metal loss, the EMAT technique is combined with the EC- and the MFL-technique. By combining the different inspection technologies the disadvantages of the individual techniques are eliminated. As a result, a sensor is now available, that allows to redundantly measure the (remaining) wall thickness of a component as well as to determine the size and the location in the wall of a detected metal loss. Table 2 shows an overview on which information is obtained from the individual inspection techniques. A further advantage of the new sensor conception is that essential hardware components can be used in parallel. In particular, pronounced improvements in the field of metal loss inspections of gas pipelines are expected. Table 2: Defect information obtained from the individual inspection techniques external metal loss: remaining wall thickness external metal loss: length internal metal loss: remaining wall thickness internal metal loss: length EMAT-Info (UT) EC-Info MFL-Info direct direct n. a. direct n. a. n. a. indirect direct indirect (direct) indirect (direct) The next step was to incorporate this sensor technology into an ILI tool environment. The various components of the ILI tool house the energy supply, the electronics controlling the sensor unit and the electronics needed for recording of the inspection data. Odometer systems are used to measure the distance traveled and correlate this information with the inspection data recorded. The development team could draw on the system design of the proven ultrasonic tools already operated. Figure 5 shows a CAD schematic of the tool layout. This tool is currently being manufactured. After completion field tests will begin and the data analysis routines will be finalized. Page 8 of 10
Figure 5: Schematic of new multi-technology ILI tool. The schematic shows the individual pressure vessels of the tool. The first tool size is planned for a diameter > 30" and will consist of two modules. The newly developed and patented sensor design will be incorporated into a trailing sensor carrier, also including the magnetization unit. Table 3 depicts the key defect specifications to which the tool is being designed. The final defect specifications achieved will be published as soon as all field trials have been completed. It is anticipated that the tool is will be operational and commercially available during summer 2008. Table 3: Key defect specifications (design values) Metal loss detection and sizing Minimum diameter (without depth sizing) Minimum diameter (with full depth sizing) Minimum depth (external flaws) Minimum depth (internal flaws) Length sizing accuracy Location accuracy Minimum size, lamination detection 10 mm 20 mm 0.4 mm 10 % of wall thickness ± 6 mm ± 0.2 m 10 mm 6 Conclusions A need has been identified to provide an ILI tool technology for gas pipelines which can achieve a accuracy and confidence level of sizing in line with the ultrasonic ILI tools available for the inspection of oil and product lines, however without the need of a liquid couplant. This paper has presented the work related to the development of a new combined sensor design incorporating EMAT, EC and MFL technology. This multi-technology approach helps to combine the strength of all technologies used. The sensor design Page 9 of 10
allows true quantitative of wall thickness and all external features. Internal features are detected and sized utilizing the EC and MFL data. The new sensor technology has been incorporated into an ILI tool, where the engineering team could build on the experience obtained through the design of a range of state of the art ultrasonic tools. The tool, the first ILI tool offering true quantitative wall thickness capabilities in a dry gas environment, is currently being manufactured and will then undergo extensive field testing. 7 References [1] Pipeline Pigging & Integrity Technology, 3 rd edition, Editor John Tiratsoo, Scientific Surveys and Clarion, 2003. [2] Goedecke, H.; Ultrasonic Or MFL Inspection: Which Technology Is Better For You?; Pipeline & Gas Journal; October 2003; pp. 34-41. [3] Recommended Practice RP-F101, Corroded Pipelines, 1999, DetNorskeVeritas [4] Reber, K., Beller, M., Uzelac, N., How do defect assessment methods influence the choice and construction of in-line inspection tools?, Proceedings of IPC 2002, International Pipeline Conference 2002, Calgary. [5] Niese, F., Willems, H., Barbian, A., New Approaches for the Quantitative Wall Thickness Measurement in Gas Pipelines, TUEV Pipeline Symposium 2007, 27.-28. September 2007, Cologne. Page 10 of 10