ULTRASONIC MEASUREMENT SYSTEM FOR THE ASSESSMENT OF CORROSION IN PIPELINES INTRODUCTION P.P. van 't Veen TNO Institute of Applied Physics P.O. Box 155 2600 AD Delft The Netherlands The demand for information on the status of pipelines still rises. Stringent environmental protection laws, rising insurance costs and rising lifetime of existing pipelines ask for more detailed and reliable inspection results [1,2]. An ultrasonic inspection system is developed for the assessment of corrosion in fluid filled pipelines. In this paper an overview of the measurement system is given. Special attention is paid to the measurement head and the -acquisition unit. Application of the system for the assessment of internal corrosion darnage is illustrated. SYSTEM OVERVIEW An overview of the ultrasonic pipeline inspection system is shown in Figure 1. An ultrasonic measurement head is attached to a flexible pipecrawler. A cable connects the pipecrawler to the up-hole equipment. The up-hole equipment includes a winch, a control unit for the pipecrawler, an ultrasonic unit, a -acquisition unit and a display. The diameter of the smallest pipecrawler is less than 4". For larger pipe diameters different crawlers are used. Inspections are performed in pipelines with diameters up to 18". The pipecrawler is able to pass bends and vertical sections. The maximum distance covered by the inspection is limited by the length of the cable. For distances up to 1 km use is made of conductive cables. For Ionger distances optical fibre cables are used. Distances up to 20 km can be reached in this way. The ultrasonic unit contains ultrasonic pulser/receiver electronics and the power supplies for the measurement head. The -acquisition unit contains an ADC-card, a timer card and a processor card. Data is stored on a tape unit. For control of the equipment and real time inspection of the acquired a display is present. Review of Progress in Quantitative Nondestructive Evaluation, Vol. 16 Edited by D.O. Thompson and D.E. Chimenti, Plenum Press, New York, 1997 2093
measurement r+ pipecrawler head ~ cable winch I ultrasonic pipecrawler -acquisition -+ display unit unit unit I t Figure 1. Overview of the ultrasonic pipeline inspection system. I Centralizer... Tran ducer Motor Figure 2. Measurement head. analog ---+ ADC digitized f---+ CPU processed display digitized.. tape unit Figure 3. Data-acquisition unit. 2094
MEASUREMENT HEAD The measurement head contains one ultrasonic transducer and a mirror (Figure 2). The ultrasonic transducers used are flat wide band immersion transducer with a centre frequency between 3 and 5 MHz. The mirror is rotated by a small stepper motor. The mirror can make up to 15 revolutions per second. With each revolution between 72 and 576 ultrasonic signals are acquired. The mirror focusses the uhrasonie beam on the inner or outer wall of the pipeline. Different mirrors are used for different pipeline diameters. Mirrors are computed using a threedimensional backward ray-tracing algorithm [3]. Rays are shot backwards from the focus point towards the transducer. Upon meeting an interface Snell's law is applied. For each ray a point on the surface of the mirror is calculated for which the travel time of the ray equals the travel time of the initial ray through the centre of the transducer. The algorithm takes account of the window and the steel pipewall. Output of the algorithm are the coordinates of the threedimensional mirror surface. DAT A-ACQUISITION UNIT In the -acquisition unit (Figure 3) the complete ultrasonic reflection signal is firstly digitized. The sample rate of the digitizer is 20 or 40 MHz. After digitization the raw is transfered to the processor card. On the processor card the reflection signal is time gated. From the travel times of the inner- and outer wall reflection the wall thickness and inner radius of the pipeline are computed and displayed together with the amplitudes of both reflections. The processor card is also responsible for transfer of the raw to the tape unit. The storage capacity of the tapes islo GByte. To minimize the necessary storage capacity the is compressed by the processor before transfer to the tape unit. Possible defects are detected on line by visual inspection of the processed. Detected defects are characterized offline using the processed and the complete ultrasonic reflection signals. Use of the complete signal enables quantitative interpretation of corrosion defects. RESULTS Figure 4 shows a signal for a pipeline with an inner diameter of 130 mm and a wall thickness of 20 mm. The first arrival is the reflection of the front wall. The third arrival is the reflection of the back wall. After 0.13 ms the multiple of the back wall reflection in the pipe wall arrives. Between the reflections of the front and back wall the multiple of the front wall reflection in the window of the measurement head arrives. For small wall thicknesses the multiple in the window of the measurement head will interfere with the back wall reflection and accurate sizing of the defect becomes impossible. For the inspection of pipelines with a small wall thickness a measurement head with a thick window is used (more than 2 mm). In that case the multiple in the window arrives after the reflection of the backwall and interference is prevented. However one should keep the window as thin as possible because of the high attenuation factor of the plastic material from which the window is made. 2095
-0.6-0.7 front wall -0.8 ~ -0.9 J -1-1.1-1.2 window multiple back wall multiple -1.3-1.4-1.5 1.15 1.2 1.25 Time (s) 1.3 1.35 x10_. Figure 4. Signal for a pipeline with 130 mm inner diameter and 20 mm wall thickness. In Figure 5 a B-scan image is shown for the same pipeline. The horizontal axis is the circumferential direction. One degree corresponds with approximately 1.2 mm along the inner wall of the pipeline. At the top of the image the reflection of the front wall is present. The arrival time of the reflection of the front wall is not constant. Intemal corrosion is present in this pipeline. The depth of the corrosion is approximately 1 mm. The diameter of the smallest corrosion spot in this B-scan image is less than 10 mm. The hatched lines in Figure 5 show the arrivals picked by the time gating algorithm. At the edges of the pits the time gating algorithm is tricked by the superposition of the reflection signals from the nominal wall and the pit. This is a general problern with ultrasonic pipeline inspection systems [4,5]. Small diameter intemal pitting may Iead to large misreadings and false indications. Only by close interpretation of the raw misreadings can be prevented. In this particular example after inspection of the B-scan images the small wall thicknesses indicated by the time gating algorithm were ignored. CONCLUSION The results of field measurements with the ultrasonic pipeline inspection system show that high-resolution images can be obtained in pipelines with diameters from 4" up to 18". Accuracy and reliability of the interpretation rise using the complete waveform. Acquisition of the complete waveform is essential for characterization of small intemal corrosion pitting. FUTUREWORK Acquisition of the complete ultrasonic signalleads to an enormous amount of. Typical volumes range from several megabytes per meter up to several tens of megabytes per meter. For the interpretation of the acquired intelligent software algorithms have to be developed. 2096
118 1.2 122 124..., ~126., e f= 128 1.3 x10_..twjjr~~~:jm WJ JIJJJ 1111111111' '.\'J~) ~11111111 Pit 11 Pit ~11. '1!1! 17~, rwtr Frontwall lwl :m (( 'c'i7<'il Backwall 132 134 1.36 REFERENCES 00 110 120 130 140 150 160 170 Angle (j Figure 5. B-scan image with intemal corrosion pitting. 1. H.J.M. Jansen and M.M. Festen, "Intelligent pigging developments for metalloss and crack detection", Insight, Vol. 37 (6), 1995, p. 421-425. 2. C.R. Sparks, "Technology assessment for gas distribution pipe intemal inspection systems", Gas Research Institute Report No. GRI-94/0302.2, GRI, Chicago, January 1995. 3. R. Breeuwer, "Backward ray tracing for ultrasonic imaging", in Proceedings IEEE 1990 Ultrasonics Symposium, ed. B.R. McAvoy, IEEE, New York, 1990, p. 401-404. 4. G.C. Williamson, III, and W.M. Bohon, "Evaluation of ultrasonic intelligent pig performance: inherent technical problems as a pipeline inspection tool - Part 1 ", Corrosion Prevention and Control, December 1994, p. 148-152. 5. G.C. Williamson, III, and W.M. Bohon, "Evaluation of ultrasonic intelligent pig performance: inherent technical problems as a pipeline inspection tool - Part 2", Corrosion Prevention and Control, February 1995, p. 8-12. 2097