Abstract

This paper describes the use of immersion probes in arrays applied to inspect pipeline contruction welds made manually. It also examines the process by which this new system was developed and describes the challenges that had to be overcome to produce a cost-effective automated inspection system that not only produced consistently reliable results but offered new advancements to the industry.

Table of contents

• INTRODUCTION
  
• DESIGN PARAMETERS
  
• THE MIPA PROBE ARRAY
  
• ULTRASONIC HARDWARE
  
• PROBE SYSTEM
  
• TRANSPORT SYSTEM
  
• DATA EVALUATION AND PRESENTATION
  
• INDUSTRY ACCEPTANCE
  
• SUMMARY
  

Introduction

In 1992, F.H. Gottfeld, Herne/Germany, a member of the SGS Group (Société Générale de Surveillance) and Krautkrämer Branson, Köln, undertook to produce a rapid automated ultrasonic system to inspect manual and machine-welded pipeline girth welds. The result of the project is a system called MIPA or Multiple Immersion Probe Array.

Fig 1: The MIPA system comes in a rugged mobil package, reliable in harsh enviroments. (18k)

The advantages of using ultrasonic inspection to detect certain weld defects over radiography have been realized for many years, however, for some applications the time required for ultrasonic inspection has been a limiting factor. Where time has not been a factor, automated ultrasonic technology has advanced a reliable solution to many inspection problems across a broad industrial base. The recent past has seen the entrance of automated ultrasonic technology into the harsh and demanding environment of pipelay operations. However, the use of these systems has been focused on automated welding processes. Their effectiveness for manual pipeline welding inspection is controversial. This is due to the infinite variability of the joint alignment and shape that is unavoidable even when highly skilled welders are used.

Variability of the finished joints is the basic problem of using an automated ultrasonic system to inspect a manual weld. (Fig.2)

It can be seen that the top layer insufficiently closes the root pass. In this case considerable deviations are possible. The deviations from the ideal shape increase with possible edge misalignment. In the practical field of ultrasonic testing of welds the most difficult tasks are the detection, interpretation and evaluation of flaw positions in the root area because a clear difference must be made between flaw echoes and shape indications. It must be understood that with a defined probe position to the top cover, an evaluation gate with a fixed distance and preset width only statistically covers the correct root area.

Furthermore, system cost is an important consideration, since it will ultimately be reflected in the charges passed on to clients. Pipeline inspection is a very competitive commercial area. Therefore, development costs of an effective unit that meets the inspection objectives must be within boundaries that place a unit in the field competitively.

DESIGN PARAMETERS

The system service requirements for the MIPA ultrasonic system were based on the state of the other competitive ultrasonic systems used for pipeline girth weld inspection and extensive experience in pipeline inspection. One key difference was the new system had to inspect manually produced welds with the same reliability as welds produced by automated processes.

The "state of the art" for this technique has been in progress for more than 25 years. A small body of technical information existed that gave the project its initial direction. Information gained from clients who had experience with automated UT used for pipeline weld quality control was also important. The answer to the problem of offering a competitive unit that provided reliable results was not found in a new, grass roots, high technology development. Closer examination of the basic inspection requirements and the expectations of the end user revealed that the answer to improving the current state of the art was in the basic inspection technique and data output. The mechanical and computer technology needed were available off-the-shelf.

After a thorough examination of the topic, many consultations with clients, equipment suppliers and a study of existing codes and standards for pipeline welding, the following parameters were established for the design of MIPA:

1. Automated ultrasonic inspection of manually welded or machine-welded pipe
2. Inspection cycle time of less than five minutes for 42" diameter pipe
3. Automated data acquisition
4. Evaluation software to eliminate or reduce subjective interpretation
5. Simple, easily understood graphic data presentation with go, no-go evaluation
6. Detection of longitudinal and transverse weld defects
7. Correlation of data with existing codes and standards for manual ultrasonic inspection
8. Field tough, mobile design, reliable in harsh environments (Figure 1)
9. Meeting inspection requirements of German HP 5/3 and American API
10. Cost competitive with radiography and/or manual ultrasonic inspection.

Traditional manual ultrasonic techniques require raster scanning, i.e. motion to and from the weld. Overlap of one scan to the previous one produces a circumferential motion around the pipe. The process of manually inspecting a 42" pipe weld from both sides around the complete circumference with ultrasound is comparable to painting a board six inches wide and 22 feet long with a brush that is 1/2 inch wide, i.e. 528 individual strokes.

Automated scanners that produce this motion have been designed and used successfully on other weld inspection applications but total inspection time for these applications is not a restricting factor. The immediate and obvious solution to the time problem is to eliminate the need for raster scanning by using multiple transducers to cover the total weld volume.

Raster scanning is eliminated because each transducer in the array covers a specific zone in the weld. The paint brush in the analogy above now becomes a roller. The raster motion is converted to straight linear motion and the same weld volume is covered at a significant increase in speed. This is the current approach of existing automated ultrasonic inspection systems designed to inspect pipeline girth welds.

THE MIPA PROBE ARRAY

The primary difference in MIPA from previously developed systems is the use of a specially designed probe holder (Figure 3). Each holder carries six or more immersion transducers that form the array. One probe array is positioned on each side of the weld. The most recent MIPA array consists of four probes for the detection of longitudinal flaws, one probe for transverse flaws and one probe generating a secondary creep wave in the root. Signal information from the latter is used by the evaluation software to distinguish geometric echoes from actual defects open to the ID surface.

Fig 3: MIPA's special designed probe holder. (18k)

The probe configuration described above was set up for a pipeline project where the Saturnax Automated Welding System is being used. Due to the shallow bevel angle (6°+1°) it was necessary to employ two transducers in pitch-catch arrangement to ensure reliable detection of planar flaws oriented along the face of the bevel.

The MIPA system uses standard, off-the-shelf immersion transducers. Nominal frequency is 5 Mhz, crystal diameter is 5 mm. The transducers are precisely positoned in the holder so that critical sound entrance angles are maintained, thus yielding a controlled sound beam within the pipe and weld.

Focused, surface-coupled transducers cannot maintain the precision necessary to ensure accurate targeting of the sound beam in the weld. Deflectionsof the transducer caused by surface abnormalities or wear on the contact surface, velocity changes due to the material or variations in wall thickness can alter the sound path such that anticipated target areas will be missed. Therefore, the real position of signals received may not be reliable.

The use of immersion transducers eliminates the need for the probe to contact the pipe surface, which eliminates most problems associated with surface condition. Also, the large footprint of the MIPA probe holder significantly reduces the influences of wear on designed sound entrance angles. It further reduces the chance of losing data caused by a loss of coupling between the transducer and the pipe surface. This can result from weld spatter or poor surface condition when direct coupling methods are used. Sufficient water pressureis maintained in the MIPA water columns to negate the effects of the probe holder "lifting off" the pipe surface for brief periods. The use of immersion transducers also reduces the limitations associated with the "near field" when inspecting thin wall materials.

Fig 1:16-channel Krautkrämer USIP 20 (39k)

ULTRASONIC HARDWARE

A 16-channel Krautkrämer USIP 20 (Figure 4) was chosen as the core component of this ultrasonic system of this technology for several reasons. They are as follows:
2. Multi-channel capability
3. 100% Digital
4. Multiple gate range parameters for amplitude and distance
5. Software compatibility with design parameters
6. Cost effective
7. Manufacturers assistance in development
8. Availability of backup systems

The combination of the MIPA probe system, the digital data from the USIP 20 and evaluation software allows defect signals from the weld to be evaluated in a way that approaches the manual ultrasonic process. (Design parameter number 6 requires that signal data reported by MIPA be verifiable by manual methods).

PROBE SYSTEM

The probe holder (see Figure 3) of the MIPA system is the technologcial achievement in the overall system. The probe holder has patents pending in Germany, Europe and the USA filed by F.H. Gottfeld and Krautkrämer jointly. The prototype probe holder consisted of a pair of single blocks with all probe angles fixed by close tolerance drilled sockets. Water supply channels were also provided within the block structure. This design required a new probe block for each different pipe thickness or weld preparation inspected.

The newest generation probe holder has individual probe receptacles that are adjustable through a small angle range. This allows field changes to the probe setup to compensate for pipe schedule changes or to maximize signal response from calibration blocks. It was found that theoretically determined probe positions did not always produce optimum results in the field. Field adjustments to the probe positioning were not possible on the first two generations of the MIPA probe holder.

TRANSPORT SYSTEM

The MIPA probe holders are propelled around the pipe circumference by a gear driven scanner mounted on a suspended ring rail (Fig 5). The ring rail is configured for rapid attachment to the pipe by a single locking hinge and hook at 0° and 180° respectively. Four alignment bolts suspend the ring above the pipe surface and secure it in position next to the weld. Horizontal offset alignment is accomplished by using a simple template to adjust and check the distance from the weld centerline. The ring can be attached and adjusted in approximately one minute.

The MIPA probe holders are propelled around the pipe circumference by a gear-driven scanner mounted on a suspended ring-rail.

Fig 6 (36k)

DATA EVALUATION AND PRESENTATION

Each weld inspected is evaluated according to preprogramed rejection criteria. Rejection criteria are a combined function of the gates set in the USIP 20 and additional program items that are adjustable within the software. Multiple gate parameters can be set for each of the 16 channels. Signal responses passing the gates are then evaluated by the software which discriminates between geometric reflectors and real defects in the weld. This process is of the utmost importance to evaluate signals from the weld root. Careful selections of the transducers, gate set up and the software make the process possible. The process actually represents an "expert ultrasonic system" for the evaluation of weld defects.

Typical data output is illustrated in Fig. 6. Graphical data from the weld consists of colored (green) bands representing each weld inspection zone on each side of the weld (note two blue weld profiles on the left side) and a larger weld cross-section (blue, bottom). Rejectable defect locations art shown in the weld graphic by a round red dot in the area indicated for repair. The graphic presentation of the weld and adjacent numerical data changes coincident with the screen cursor (thin vertical white line) as it moves along the green target bands. Graphical representations of signals along the different inspection zones or bands change colors as the signal amplitude increases (see area either side of screen cursor and signal amplitude color key at bottom left). Therefore, in addition to the weld cross-section view, the linear extent of the defect is represented by the area target bands. The data form can quickly be evaluated by a lay person by referring to the upper right-hand corner of the report. There a green or red box appears. Green indicates an acceptable weld; red is a reject.

Fig 7: All-terrain flexibitlity: MIPLA-LT is a modular system which can be used at every test location no matter how difficult the ground conditions are. (28k)

INDUSTRY ACCEPTANCE

Automated ultrasonic inspection on pipelines is preferred in Germany and is an acceptable alternative to manual ultrasonic inspection in other European countries except France. The system has been evaluated and approved by both, TÜV Austria and TÜV Bavaria. German pipeline construction codes require 70% ultrasonic inspection (manual or automated) and 30% radiography. German gas companies such as Ruhr Gas and Wintershall require the TÜV approval in addition to their own assessment.

In 1994 the Mahgreb Pipeline Project (now under construction, running from Algeria through Morocco across the straights of Gibraltar and into Spain) adopted the requirement for automated ultrasonic inspection.The first spread in Morocco will be inspected with the MIPA system. Indications point to further use of automated UT along successive sections in Morocco, Spain and possibly to future connections into Portugal.

In 1994 the MIPA system was employed by Oberösterreichische Ferngas (ÖOG-Austria) to inspect 75 kilometers of manually welded pipe. The project was unique because the client elected to do 300% inspection in the early stages. Each weld was inspected by radiography, manual UT and the MIPA system. The results were then compared to establish confidence in the MIPA data. The comparison of MIPA with manual UT rapidly established an excellent correlation of data. The frequency of manual UT was reduced after the first eight kilometers and then dropped except for inspection of tie-ins and repairs. The comparison between MIPA UT data and radiography continued for the first 15 kilometers but ultimately the client reduced the radiography to the 20% code requirement for the remainder of the project. This project marked a milestone in pipeline inspection in Europe in that it established the credibility of the MIPA UT system as a reliable method for the inspection of manually produced welds.

Automated UT is an acceptable alternative to radiography in Canada and is commonly used for pipeline welding quality control. In fact, the practice of using automated UT for pipeline contruction quality control is greater here than in any other place in the world today. TransCanada PipeLines have their own internal specification for the evaluation and use of automated ultrasonic equipment employed on their pipelines.


Fig 6: Typical data output - see text for full explanation (16k)

SUMMARY

Ultrasonic technology is now gaining a foothold in the pipeline industry. It has only been in the last five to six years that the use of automated systems has achieved acceptance as a companion method to radiography. It is doubtful that automated UT will ever supplant radiography but the increase of its use must be recognized as pipeline owners become aware of the advantages of using two inspection methods to ensure the quality and integrity of new constructions.

Authors

Fig 7:

| UT-Front Page | |Top to this page|