by J.A. de Raad and F.H. Dijkstra*

 

This month's article provides an inspection solution for the inspection of girth welds found in long-distance pipelines. The mechanization of the process increased the speed of the inspection and also improves the probability of detection. This article may trigger ideas and solutions where mechanizing the inspection process can provide significant benefits. G.P. Singh Associate Contributing Editor

 

Figures 1-3
Figures 4-6
Figures 7-8

During cross country pipeline construction, day production rates of more than 120 girth welds are not unusual. On pipe laying barges operating around the clock, production can be as high as 360 welds per day. This implies an average cycle time of some four minutes or less per weld. In the construction of a pipeline, nondestructive testing (NDT) is an essential tool to verify the quality of girth welds. For decades, radiographic techniques were applied almost exclusively, but today it is recognized that mechanized ultrasonic testing (UT) can supplement and replace radiography.

After a debate which has been going on for decades about the "pros and cons" of radiography versus ultrasonic inspection, a growing number of pipeline owners, contractors, and authorities now agree that, for relevant weld discontinuities, ultrasonic inspection results in a higher detection probability than radiography does (Gross et al., 1990).

However, the real breakthrough for mechanized UT on pipeline girth welds was induced by the recent introduction (1993) of C-scan mapping and time of flight diffraction (TOFD) facilities in Rotoscan. This enabled detection, recording, rapid interpretation, and sizing of all relevant weld discontinuities in almost all weld types, as well as compliance with prevailing codes and standards.

---

The pipeline industry is now on the threshold of a new era in girth weld inspection.

---

 

With these new options, the system appears to be mature for lay barge application. Recent successful qualifications and commercial lay barge jobs have confirmed this. If this is true, it is expected that the pipeline industry is now (from a technical point of view) on the threshold of a new era in girth weld inspection. However, the nonexistence of appropriate acceptance criteria requires a case by case agreement between parties involved, which hampers efficient introduction of mechanized UT.

 

Concept of the Mechanized Inspection System
The concept of a mechanized inspection system is based on the following essential requirements:

• The system should cope with all prevailing weld types and wall thicknesses used in the pipeline industry.
• All indications exceeding a predetermined level of reflectivity should be found with a high probability of detection to comply with codes and acceptance/rejection criteria.
• The system should reliably detect porosity which was, until the recent introduction of C-scan images, regarded as a weak feature of ultrasonic inspection.
• Discontinuities should be distinguished from non-relevant indications to prevent "false calls," e.g., calls caused by nonuniform weld geometries such as hi-lo.
• Discontinuities should be allocated to particular areas in the weld to establish their presumable nature and size.
• Indications should be recorded in a coherent pattern to allow for immediate evaluation, including an estimate of size in both length and through thickness dimensions, and reflecting the need for transparency of result presentation and quick familiarization for all parties involved.
• Monitoring of inspection quality by rapid documented calibration checks and continuous functioning and coupling check.
• Long-term archiving of raw data for later retrieval and viewing by the customer should be implemented.
• The system should be able to cope with both cross country and lay barge conditions.
• To achieve a minimum inspection time, one circumferential scan (speed 50 mm/s [2 in./s]) should be sufficient for full volumetric weld coverage.
  
  Full weld coverage is achieved by placing probe sets on both sides of the weld, each probe taking care of a particular weld zone. This eliminates the need to move probes to and from the weld, as is conventional practice in time consuming manual ultrasonic inspection and most mechanized UT systems. Figure 1 shows a cross section of a weld divided in zones. Figure 2 shows a top view of this arrangement. With this probe arrangement (whereby some of the probes have focused beams) as a front end, this mechanized UT system comprises the following components:
• a scanner, provided with a high precision encoder, which can be handled by one operator and can quickly be attached and removed.
• an umbilical cable for connection between scanner and electronics, to be placed inside an all terrain vehicle or an instrumentation cabin.
• a control panel to operate the scanner.
• a computerized multichannel ultrasonic flaw detector.
• a data storage and display computer with dedicated software, offering color enhanced user friendly and coherent presentation enabling quick and unambiguous interpretation of inspection results.
• facilities for mass storage of raw data.
• a printer to produce instant field inspection records.
• appropriate calibration facility.
• a coupling medium supply system.

More details for the mechanized inspection system are described in de Raad (1992) and de Raad and Dijkstra (1995).

 

Data Presentation and Storage
In the past, the usual way of recording results was in the go/no-go mode. Only those signals exceeding a preset amplitude threshold were shown on a multichannel record. The record also included distance marks to enable indication position and length measurement. In 1989, an essential improvement was added: the recording of analog signal amplitudes and transit distances, particularly to improve interpretation in the root area and to differentiate between planar and erratic discontinuities. Interpretation was still done from paper charts, which were also used for archiving. In 1993, computerized data presentation and storage were introduced. The use of a computer also enabled digitizing of ultrasonic signals, essential for coherent C-scan mapping and the use of TOFD technique. The mapping facility (Rotomap) offers the possibility of pattern recognition, which appeared to be the key feature for reliable inspection of welds with narrow roots (manual stick, narrow gap) as well as recognition and quantification of porosity. Figure 3 shows a typical compound mapping presentation of go/no-go results, amplitudes, transit distances, and mapping in the root and the weld body as is now standard in the mechanized inspection system.

The software allows for automatic judgment of indications and discontinuity tally list generation. As an option, the system can provide weld cross sections showing the position of indications. C-scan mapping and TOFD including imaging and recording of results can be used simultaneously.

 

Added Value of TOFD
Although the application of TOFD is of increasing interest for different fields of industry, its use for girth weld inspection in the pipeline industry is relatively new. Nevertheless, a combination of TOFD with common pulse echo techniques in a mechanized inspection system can offer unique advantages over the application of pulse echo techniques alone, especially since modern high speed signal processing enables the application of this technique at high inspection speed (typically 40-100 mm/s [1.5-4 in./s]).

Apart from the capability of TOFD to size discontinuities, it has also been demonstrated that its use can increase probability of detection, in particular for those discontinuities which are difficult to detect because of their unfavorable geometry or orientation. This "safety net" function has been recognized by several users. TOFD’s sizing capabilities inherently offer the option to apply Engineering Critical Assessment (ECA) for Fitness for Purpose criteria. Moreover, TOFD is able to quantify hi-lo (vertical misalignment). Figure 4 shows an example of a TOFD image showing a typical flaw indication. Recently it was established that for selected applications even the mechanized system’s pulse echo tandem probes for inspection of the weld body can be replaced by TOFD, without compromising on probability of detection (POD) and false call rate. This makes the probe arrangement much more flexible, and is especially advantageous in those cases where the number of welds to be inspected is limited. Thus, the inherent drawback of an inflexible tandem probe arrangement can be overcome.

This novel approach has, in the meantime, been implemented in a hyperbaric welding and inspection system, to be applied to subsea tie-ins and weld repairs.

 

Field Experience and Project Qualifications
Since 1989, an impressive track record has been built up whereby, on a grow number of projects, the mechanized inspection system was used as a sole inspection technique to the full satisfaction of all involved. Inspections on pipe diameters between 75 and 1,420 mm (3 and 56 in.) were completed in Canada, Germany, Brazil, Poland, The Netherlands (Figure 5 and 6) and the Czech Republic. Some of the projects took place under harsh weather conditions, such as temperatures as low as -40 °C (-40 °F). Furthermore, some specific offshore applications are worth mentioning: riser girth weld inspection (Australia), platform leg girth weld inspection and J-lay barge application on stainless steel clad pipe (New Zealand [De Raad and Dijkstra, 1995]). The system was qualified for application on SAIPEM lay barges (Italy). Various offshore pipeline inspections have been performed in The Netherlands and other Western European countries. At present, work is in progress to qualify for inspection of radial friction welds. Further qualifications are in progress.

The approach for each of the tasks mentioned above was similar, in the sense that they all required individual qualification tests with the system on their own typical and representative test welds. Individual inspection procedures including applicable acceptance criteria had to be written as part of the qualification process by lack of international standards and criteria.

On the various projects, a variety of inspection codes and acceptance criteria were applied. These included both ECA and good workmanship standards. In practice it was experienced that, although the POD of relevant discontinuities of mechanized UT is generally higher than for radiography (Glover et al., 1988), repair rates appeared to be normal, sometimes even lower than typical RT values. This can partly be explained by the fact that results are directly fed back to the welder, causing indications to be avoided rather than repaired.

This indicates that high POD values combined with low false call rates are achievable under practical conditions. This applies even for manually made (de Raad, 1992) and narrow gap welds. In particular the C-scan mapping option appears to be essential for the system to cope with these weld geometries. It was also proven that the new mechanized inspection system can cope with even excessive hi-lo, for which it has deliberately been made insensitive to prevent false calls. It is worth mentioning that a study is in progress to establish the accuracy to measure hi-lo with TOFD; this is a unique, promising feature and fulfills a need from industry by lack of appropriate alternatives.

With the current system, a typical cross country production rate of over 120 welds a day is achieved at projects where automatic welding is applied. A two man crew operates the system, and a third assistant adjusts the guiding bands and cleans the pipe surface adjacent to the weld.

 

Calibration of Equipment
A very important part of ultrasonic inspection is the sensitivity setting of all the individual ultrasonic transducers. Sensitivity setting is done by means of a calibration place made from a piece of the actual pipe, to ascertain that dimensions and chemical composition are identical to the pipe to be inspected.

This pipe segment contains machined artificial defects, such as flat bottomed holes and notches related to applicable code or specification requirements. Each zone of the weld to be inspected has its specific calibration reflector(s). Two sets of calibration reflectors are provided, one set for each transducer set at either side of the weld. The calibration plate is inserted in a larger pipe segment, which has a band attached to it to guide the scanner. This method of calibration not only enables setting the sensitivity for each individual probe, it also allows for rapid dynamic check of the system’s performance at inspection speed. In practice, this plate is used for regular calibration checks and to store the scanner during transport between welds. Figure 7 shows the calibration plates on their usual place at the back of the all terrain vehicle.

 

Procedures and Acceptance Criteria
For application of the mechanized inspection system, it is good practice to operate strictly according to a mutually agreed inspection procedure. To judge the results, the procedure always contains clear acceptance/rejection criteria. These criteria may be based on an "Engineering Critical Assessment" (Glover et al., 1990). Applied fracture mechanics appears to result in appropriate criteria, which are very pragmatic to work with. However, the system is also frequently used on projects where good workmanship criteria are applied, which typically requires detection of smaller anomalies than ECA based criteria. This has been made possible by the fact that the improved technology (mapping facility) now enables unambiguous signal interpretation. The growing tendency to judge weld strength by fracture mechanics might, in the near future, by a reason to apply mechanized UT more widely because the strength of this technology is detection and quantification of discontinuities relevant to weld integrity. In addition, the fact that, in modern high strength steel pipelines, the weld becomes an ever more critical factor requires reliable detection and sizing of such discontinuities.

Hence, it is to be expected that due to the higher detection probability with mechanized UT, rejection criteria will eventually be adapted to its specific capabilities and will differ from current conservative (good workmanship) criteria, which have been typically designed for use with radiographic techniques. Signals that this is indeed happening are already available (EPRG Guidelines, Canadian code CSA Z184 App.K, Dutch pipeline code). Nevertheless there is hardly any progress to agree on generic international acceptance criteria to introduce mechanized UT. Efforts so far were not successful. This has resulted in a new initiative: at present a joint industry project is in progress to establish and introduce such criteria restricted to pipeline girth welds. This project shows similarity with a Dutch joint industry project to establish criteria related to the application of TOFD (Dijkstra et al., 1996).

 

Mechanized UT Inspection on Lay Barges
From qualification trials and field experience, some advantages relevant to lay barge application have become obvious. Not only is inspection time independent of wall thickness and cycle times of four minutes can be easily coped with, but in addition no equipment is needed inside the pipe, hence no interference with the critical production process exists. Last, but not least, no radiation sources are needed.

Although these advantages have become clear for lay barge application, practice has shown that this area of industry is hesitant to embrace the development. This might be caused by the deeply rooted tradition of radiographic inspection. It seems, however, that recent successful qualifications and commercial lay barge jobs have demonstrated that a new era in pipeline girth weld inspection has begun. After a clear breakthrough for cross country pipelines, the mechanized UT system is now also ready for use on lay barges. Apart from its performance in terms of anomaly detection and ease of interpretation, this is also because of the system’s ability to cope with typical lay barge conditions such as strong electromagnetic interferences and elevated weld temperatures (up to 200 ¡C [390 ¡F]). Since early 1996, the system has been used on lay barges several times. Figure 8 shows the inspection of a 75 mm (3 in.) pipeline at a rate of up to 360 welds a day. It successfully replaced double wall single image radiography as a more efficient and economic alternative.

 

Present Developments
A systematic study was conducted to evaluate the reliability of mechanized UT. A number of welds, both manually and automatically made and both containing intended anomalies, were incorporated in this study. The work is limited to performance in the weld’s root region. It is particularly this weld region where appropriate solutions are essential to unambiguously discriminate between relevant weld anomalies and other anomalies such as geometry.

 

Conclusions
The mechanized UT inspection system is capable of achieving a high POD together with a low false call rate, appropriate anomaly sizing capabilities, and transparent recording of results. The increased use of thin walled, high grade steels justifies the use of ultrasonics in combination with adapted acceptance criteria.

The system can easily cope with typical production rates, also on lay barges. For selected applications, the use of the Time of Flight Diffraction technique in addition to the standard mechanized UT probe configuration contributes to the reliability of detection and sizing of anomalies.

Although so far mainly used onshore, the features of the system justify its more frequent use offshore. This requires that the present process of adapting acceptance/rejection criteria to the specific capabilities of the mechanized UT will continue. As the title of this paper suggests mechanized UT is now, from a technical point of view, fully capable of performing girth weld inspection in the pipeline construction industry.

 

References
de Raad, J.A., "Can Mechanized UT Replace RT on Girth Welds during Pipeline Construction?" International Conference on Pipeline Reliability, Calgary, Alberta, Canada, Jun. 2-5, 1992.

de Raad, J.A., and F.H. Dijkstra, "Mechanized UT now can Replace RT on Girth Welds during Pipeline Construction," Second International Conference on Pipeline Technology, Oostende, Belgium, Sep. 11-14, 1995.

Dijkstra, F.H., J.A. de Raad, and T. Bouma, "TOFD and Acceptance Criteria: A Perfect Team," World Conference on NDT, New Delhi, India, Dec. 8-13, 1996.

Glover, A.G., et al., "Inspection and Assessment of Mechanized Pipeline Girth Welds," Proceedings of Weldtech 88, 1988. London, UK.

Glover, A.G., D. Hodgkinson, and D. Dorling, "The Application of Mechanized Ultrasonic Inspection and Alternative Acceptance Criteria to Pipeline Girth Welds," Pipeline Technology Conference, Oostende, Belgium, Oct. 1990.

Gross, B., J. O’Beirne, and B. Delany, "Comparison of Radiographic and Ultrasonic Inspection Methods on Mechanized Girth Welds," Pipeline Technology Conference, Oostende, Belgium, Oct. 1990.

van der Ent, J., and F.H. Dijkstra, "Evaluation of Ultrasonic Inspection Techniques for the Root Region of Girth Welds," RTD report for the American Gas Association, AGA Project PR-220-9123, Jan. 1996.

 

* Röntgen Technische Dienst bv, Delftweg 144, Rotterdam, The Netherlands 304b NC; 31-10-2088200; fax 31-10-4158022.

Copyright © 1997 by the American Society for Nondestructive Testing, Inc. All rights reserved.

[ Materials Evaluation ]