TABLE OF CONTENTS

Introduction System Concept Zone Discrimination Calibration Block Codes Data Display Data Storage Probe Pan Instrumentation Typical Results In-Service Applications Advantages and Disadvantages Conclusions References

Introduction

Gas is becoming a major source of energy for the nineties and into the next century. Both supply and demand are growing rapidly, with gas fields being discovered and exploited in many different countries around the globe. As pipelines are usually the best method of transporting gas, these being constructed at a great rate all over the world. Pipelines are made of sections of thin-walled steel tubing, which are welded together using a circumferential (or girth) weld. Traditionally, these gas pipeline girth welds have been inspected in the field by radiography or X-rays. Over the past several years, mechanized ultrasonics has started to supplant radiography for these construction inspections, particularly in Canada /1-4/. This paper describes the mechanized ultrasonics used in pipeline construction, particularly two of the key features which have made mechanized ultrasonics feasible: zone discrimination and the output display.

The large diameter gas pipelines used for main lines have diameters ranging from 24" up to 48", with most pipelines in the mid-range. The pipe walls vary in thickness, though thin-walled pipe is usually 10 mm thick or so, and thicker-walled pipe around 16-20 mm. The pipes are made of high strength steel, and operate at a significant percentage of yield strength. Pipes are girth-welded on site, typically using automated welding, rapidly inspected, coated and buried. Due to the demanding construction cycle, it is important that any defects in the welds be detected and analyzed very quickly.

There are a number of specific constraints relating to the construction cycle:


1. Time: on-shore, the inspection cycle (mount scanner, scan, return, remove, drive to next weld) is about four minutes. Offshore, the time may only be two minutes, though moving is not required.
2. Data Analysis: Data needs to be analyzed before moving to the next weld. This means characterizing the weld as accept/reject, almost in real time.
3. Data Storage: It is essential to store the data during the inspection cycle, for regulatory purposes.

System Concept

Mechanized ultrasonic inspection systems consist of: the probe pan or scanner, which holds the ultrasonic transducers and a motor to drive the pan; an umbilical to the instrumentation; the instrumentation; and either a vehicle for on-shore use, or a cabin for off-shore and static applications (see Figure 1).

Fig 1: Schematic of mechanized UT inspection system (PipeWIZARD)

Instead of the usual raster-type C-scans, mechanized ultrasonics of girth welds uses a large number of ultrasonic transducers to minimize inspection time. The probe pan runs rapidly round the pipe on a welding track at up to 100 mm/sec with no back and forth motions. Data is displayed in a strip chart style on the CRT, and a trained operator can interpret the results, literally in real time. A second operator places and removes the probe pan from the pipe, while a third operator adjusts the position of the welding band. Other crewmembers may be needed to clean the pipe etc.

The key features of mechanized UT systems are described below: zone discrimination; calibration blocks; probe pan; output displays; instrumentation. Also, brief mention is made of: codes; in-service applications; and the advantages and disadvantages of mechanized ultrasonics.

Zone Discrimination

The ultrasonics is closely targeted at the weld profile and for the defects expected. Highly focused and precision-angled transducers position the beam accurately on the appropriate section of the weld. The angles are selected to maximize the response from the expected defect type, usually lack of fusion. Each transducer is specifically angled and positioned to inspect a small portion (or zone) of the weld. This approach is called "zone discrimination". Each zone is 1-3 mm deep, depending on the weld profile, thickness, and the inspection requirements (see Figure 2).

Fig 2: Zone discrimination, as applied on CRC-Evans weld profiles. - Chamfer (bevelled edge) for 12 mm thick pipeline. - Position of the transceivers and their ultrasound path.

The scanner is accurately positioned a fixed distance from the centre of the weld (to within + 0.5 mm), and the scanner rotates round the pipe on the welding band. As a result, each transducer beam inspects its narrow strip of material with high precision. Inspections are performed from both sides of the weld, upstream and downstream, so a total of at least twelve zones are inspected. As some of these zones require a tandem technique, typically up to twenty transducers may be needed for a thick-wall weld. Furthermore, some applications specify volumetric Time-Of-Flight Diffraction (TOFD) inspections, creeping wave inspections of the cap area, or transverse defect inspections, which would increase the number of transducers.

Zone discrimination permits accurate sizing, as a signal detected on only one transducer must come from a defect less than (or equal to) that zone in height. The beam overlap from one zone to another is strictly limited during calibration (see below). In turn, more accurate sizing leads to the acceptance of larger defects using Engineering Critical Acceptance. As a result, the reject rate using mechanized ultrasonics is lower than that using radiography.

Zone discrimination is specific to the Canadian approach for mechanized UT, largely due to the foresight of NOVA and TCPL.

Calibration Block

Reflectors in the calibration block are designed to provide defects representative of typical lack-of-fusion type reflectors in each zone /5/. For the CRC Evans weld profiles used in Canada, notches are used for the root reflectors, and the cap, and 2 mm flat bottom holes at the appropriate angle and location for the LCP, fill and hot pass reflectors. There is also a vertical through-wall drilled hole for setting the timing of gates. The calibration block is complex with multiple reflectors (see Figure 3), though an experienced operator can readily deduce which is the relevant calibration signal for each channel.

Fig 3: Schematic of calibration block.

Typically, the signal from each calibration reflector is set at 80% FSH, with the evaluation threshold at 40%. During repeated calibration, the reflector amplitudes must be kept within a specified amplitude range due to normal scanning variations. Signals from calibration reflectors in adjacent zones must be kept 6-14 dB below the calibration signals from the relevant zone.

For on-shore inspections, calibration is normally required every ten welds. Offshore, where conditions are stricter, calibration might be required every weld. Future applications may require automated calibration analysis to reduce operator interpretation.

Codes

One major disadvantage of mechanized ultrasonics is the absence of a suitable code. Pipeline companies in Canada and the Netherlands have developed very detailed specifications, which act like a code in practice. The standard inspection code, API 1104 /6/ is slanted towards radiography, with major emphasis on porosity (which is of little structural significance).

ASTM recently adopted Recommended Practice E-1961 for mechanized ultrasonics of girth welds /7/. A modification to API 1104, the standard radiography code, for mechanized ultrasonics is also in draft form, and should be completed during 1998.

Data Display

Fig 4: Typical output display from calibration block.

Data is displayed in a combination of strip charts and B-scans on a 21" monitor for easy interpretation (see Figure 4). The philosophy of the output is to display the ultrasonic data as if the weld had been sectioned down the weld line, with the downstream and upstream data on either side of the central B-scans. A typical screen contains the following:


1. a series of narrow strips at the right showing calibration checks;
2. a series of B-scans in the centre of the screen showing the root and cap areas from both sides of the weld to identify porosity;
3. a TOFD B-scan if specified;
4. a position marker on the left-hand side; and
5. strip charts for each zone, distributed either side of the central B-scans.

Despite the apparent complexity, this output display has evolved to permit the operator to make rapid analyses of the scan results.

Each strip chart contains two gates: signal amplitude and time of flight, and output from both gates is displayed on screen. The amplitude gate is set from the parent metal side of the weld to typically1 mm past the centre-point of the weld. The time gate is set from the same start point of the amplitude gate. For selected gates, the time gate is set to well past the centre of the weld, i.e. the time gate is significantly longer than the amplitude gate. The signal amplitude is shown as a line; the time-in-the-gate is shown as a coloured block, overwritten by the amplitude data. Using the amplitude and time information from each zone, the operator can determine where the defect is in the weld, and can readily differentiate between LFSS and centre-line cracking in the fill zones, for example. For the root pass the operator can differentiate between LFS, porosity, high-low and misalignment of the welding band using the two root pass strip charts and the B-scans (also see Figure 7 below).

Above threshold defects are signaled by a change in colour on the screen, so the operator's eye is immediately drawn to these areas on the screen. A trained operator can rapidly characterize a reflector using the twin-gate strip chart display, check the TOFD channel (if present) for confirmation, accurately size and locate the defect using cursors or the defect table, and quickly determine whether the defect is acceptable using a look-up chart. For multi-zone defects, the analysis would be more complex.

Data Storage

Scans are indexed and stored automatically. Each file, including TOFD and root and cap B-scans, occupies about 1.5 MB. Files can be saved on hard drive, optical disk, or CD-ROM, depending on the client's particular interest.

Probe Pan

Fig 5: Photograph of probe pan with transducers.

Probe pans contain twenty or more spring mounted, contact transducers (see Figure 5). The transducers have carbide wear pins to minimize wear, with pumped water for couplant or water-methanol for cold weather. Each transducer pair is mounted on a removable module for rapid replacement if needed. The position of the transducers can be readily adjusted using a knurled, lockable knob. The probe pan itself can be removed by four screws and connections, for rapid changeover to a different wall thickness. The ultrasonic and water lines can be rapidly disconnected and reconnected. Each transducer is individually connected to the water supply. Each side of the probe pan has a lifting lever, to lift the transducers for rapid rotation and to minimize any damage to the probe pan during handling. For two different pipe thicknesses, two different removable probe pans are used to save set-up time.

The probe pan itself is attached to a drive motor and CRC-Evans welding band connector, which is coupled to the welding band with quick-release handles. An encoder is attached to the connector to give position data. A 5 cm diameter umbilical cable connects to the instrumentation in the vehicle through a connector on the truck wall. The umbilical contains twenty-odd ultrasonic cables, encoder data, motor power and the water supply.

Instrumentation

Fig 6: Photograph of instrumentation package.

The instrumentation inside the vehicle consists of motor control unit; ultrasonic instrument containing the pulser-receivers; and the industrial computer (see Figure 6). The R/D Tech instrumentation has thirty-two ultrasonic channels, and four motor drive channels, giving significant redundancy. The pulsers are based on the (-Tomoscan board, for packaging and price considerations. The industrial computer is a high speed PC running Tomoview software, giving displays as shown below. The software is normally locked to minimize operator changes, and automatically records inspection scans.

Typical Results

Fig 7: Output display from weld with defects.

Figure 7 shows a typical scan from a weld with defects, showing several embedded defects. Quickly looking down the strip charts in Figure 7 shows the above threshold defects in red. The operator can check the time gate and other information to characterize the defect(s), measure the length and determine whether the defect is acceptable or not. The operator can also check the TOFD channel, the root and cap B-scans if appropriate, as well as the coupling channel. With practice and a scrolling screen, it is possible to make structural integrity decisions literally during the scanning of the weld.

In-Service Applications

In the last several years, mechanized ultrasonics has inspected about 100, 000 welds in Canada alone. Mechanized ultrasonics has also replaced radiography in a number of pipeline construction projects around the world, e.g. the Magreb line and in Austria. Typically, the constructor starts by using both radiography and ultrasonics, and then drops radiography in favour of ultrasonics during construction. In practice, when radiography has been replaced by mechanized ultrasonics, the switch is permanent. Ultrasonics is now specified as the inspection method for many pipelines, not radiography. Mechanized ultrasonics has been performed in deserts, offshore, in humid environments, and at -40 C.

Advantages and Disadvantages

The advantages of ultrasonics are clear:


1. No radiation hazard;
2. Better defect sizing, leading to lower reject rates;
3. Better process control of welding;
4. Faster inspections.

The disadvantages are also well known:


1. There is no cost saving (yet);
2. There is no globally-accepted code yet;
3. The ultrasonics is sensitive to temperature, and the wedge temperature must be kept to within + 10 C of the calibration block;
4. Due to velocity variations, the calibration block must be made from the same material as the piping;
5. The probe pans are heavy (typically about 20 kg).

By and large, these disadvantages can be overcome.

Conclusions

1. Mechanized ultrasonics is replacing radiography for gas pipeline girth weld inspections worldwide.
2. Ultrasonic zone discrimination is one of the key factors which makes mechanized UT feasible.
3. The output display has been specifically developed to aid rapid and reliable interpretation of data.
4. The mechanics and probe pan have been designed for robust operation.
5. Mechanized ultrasonics offers lower reject rates due to improved sizing accuracy.
6. While there are disadvantages of mechanized ultrasonics, these can be overcome.
7. Overall, mechanized ultrasonics offers a better inspection for roughly the same cost (on-shore).

References

1. J.A. de Raad, "High Speed Ultrasonic Inspection of Field Girth Welds During Pipeline Construction", European Journal of NonDestructive Testing, August 1991, Volume 1, Number 1.
2. E. Ginzel, P. den Boer and M. Hoff, "Application of Mechanized Ultrasonic Inspection to Manually Welded Pipeline Girth Welds", UT Online Application Workshop, May 1997, http://www.ndt.net/article/wsho0597/ginzel3/ginzel3.htm.
3. H. Heckhäuser and S. Schultz, "Advanced Technology in Automated Ultrasonic Weld Inspection of Pipeline Girth Welds", Insight, vol 37, no 6, June 1995, p. 440.
4. J.A. de Raad and F.H. Dijkstra, "Mechanized Ultrasonic Testing on Girth Welds During Pipeline Construction", Materials Evaluation, August 1997, p. 890.
5. TransCanada Pipelines Welder Inspector's Manual 1996, Section TWE-1 "Total Weld Evaluation of Mechanized GMAW Pipeline Girth Welds".
6. American Petroleum Institute Standard 1104, "Welding of Pipelines and Related Facilities", Eighteenth Edition. May 1994, ANSI/API STD 1104-1994.
7. See ASTM Committee Draft No. 1, 14 April 1997, "Standard Practice for Mechanized Ultrasonic Examination of Girth Welds". ASTM E-1961 was released in August 1998, but no final version is available at the time of printing.