The origin of Intelligent Pigs

One of the most crucial aspects of pipeline operation is ensuring pipeline integrity. The intelligent pig has become an important tool for assessing the condition of a pipeline, and is set to become an even more integral part of pipeline maintenance as steps are made towards solving the ‘unpiggable’ pipeline puzzle. Pipelines International spoke with Doug Woodley to trace the origins of the intelligent pig and examine how it developed into the essential tool that it is today.

Early intelligent pig technology

Pipelines are normally pigged for one or more of the following reasons: cleaning, batching, displacement or inspection. Pigging has been around for a long time – some say the process dates back to Roman aqueducts – but the earliest ventures into retrieving information from a pig originated in 1959.

In this year, T.D. Williamson introduced a caliper tool for detecting dents in pipelines while Pan-American Petroleum was developing a ‘Cooley tool’, which used the magnetic flux leakage (MFL) technique.

Tuboscope had a tool that could detect defects in down-hole casings and clients were asking for a pipeline version of the tool. Simultaneously, Shell Oil Research was developing a technique for detecting pitting in down-hole casings based on a ‘MacLean tool’, which worked on the remote field eddy current principle.

In 1962, Tuboscope obtained a licence from Shell Oil Research for the MacLean tool and started developing a pig to carry an array of remote field eddy current sensors through a pipeline. Early test runs with the MacLean tool were disappointing as they could not detect known pits in test spools.

Tuboscope then approached Pan-American Petroleum and purchased the Cooley tool patent. The MacLean tool was abandoned and development switched to Cooley or MFL, as we now know it. The new tool was branded LINALOG®.

The first commercial job for the new LINALOG 90° tool was for Shell in 1965. The tool only inspected the bottom 90° of the pipeline, had only four sensors and did not have a distance measuring wheel. Location was identified by digging down and placing permanent magnets on the pipe before the pig passed, which were detected by a coil on the drive unit of the pig.

Client-driven development

From the early beginnings, then as now, developments were client driven. The pipeline industry was soon asking Tuboscope to deliver a tool that could inspect the full circumference of the pipe. By 1966, a prototype 10 inch diameter, 360° tool had been run through a test loop and by mid-1967 a 24 inch diameter tool was undergoing test loop trials, with commercial projects planned before the end of the year.

Early inspection pigs were delivering a service to pipeline operators that enabled them to target their maintenance more effectively. The tools were developed at a time when the transistor was still in its infancy, printed circuit boards were relatively unsophisticated with the circuits often hand-wired, and rare earth magnets were not available. The indications recorded by the pigs were transferred from the onboard data recorder to a roll of ultraviolet-sensitive paper via a ‘visicorder’. The findings were displayed as blips on a trace and often did not give an indication unless the metal loss was great, but they alerted the operators to the sites where they should dig.

In the 1970s, British Gas constructed a network of pipelines across the United Kingdom and needed a tool to inspect this network and assure the pipeline engineers that it could be run safely. To evaluate what equipment was available on the market, inspection companies were invited to trial their equipment at a purpose-built test loop in the northeast of England.

The tools that were tested were what are now termed as low-resolution tools and would only give a ‘low – medium – severe’ indication of any feature found in the pipeline. British Gas decided that the tools were not accurate enough to allow engineers to make pressure-sentencing calculations, so the decision was made that if it couldn’t buy the technology it needed, it was going to have to build it itself.

The project started in 1974, with a handful of engineers at the British Gas Engineering Research Station just north of Newcastle upon Tyne. Development started on a 24 inch diameter, 12-channel analogue tool and in the same year, the first development runs were taking place.

To achieve the required accuracy, the tools needed a high sensor density, typically ten sensors per pipe inch diameter and a high sampling frequency of 3.3 mm of pig travel. The high volume of data generated could not be recorded onto analogue tape and a completely new approach was needed.

Digital was the way forward and the first runs with a digital processing pack and 60-channel digital recorders began around 1977/78. Early runs were not a great success as running 60-channel digital recorders on 1 inch tape at 1 7⁄8 inches per second was a difficult task and the recorders had to be precision built and meticulously maintained.

MFL and UT

At this time, when the high resolution technology was in its infancy, the costs associated with design and construction of intelligent pigs were high and the focus was on ensuring the pigs produced accurate and consistent results. This is essential if the data gathered, and the subsequent reports generated are to be used with confidence in strength calculations, fitness-for-purpose and integrity assessments.

Many problems need to be solved when building a tool that will travel through a pipe, driven by the product flow, taking metal loss and wall thickness measurements every 2 or 3 mm along and around the circumference of the pipe, at speeds of up to 4 m per second or 14 km per hour. There are vast amounts of data being generated and the pigs have to cope with environmental conditions such as high pressures, temperatures and vibration in addition to the sometimes aggressive chemical nature of the product being transported.

By 1980, the pigging department of the British Gas Engineering Research Station had moved to its own premises and had employed around 250 personnel. Inspection pigs based on the MFL principle were being built and trial runs were taking place in the British Gas network of pipelines. The tools were producing reliable data, but that data needed to be interpreted as metal loss values. Once MFL tools reached this stage of development the next stage was to develop and refine the sizing algorithms.

MFL is an inferred measuring system. Any feature detected causes a disturbance of the magnetic field around the sensors. That disturbance is recorded and data analysts – these days with the aid of sophisticated software – determine what those signals mean in relation to percentage of metal loss.

Development was beginning on ultrasonic technique (UT) pigs. These pigs used the well-established technique of measuring the remaining wall thickness by time of flight ultrasound and packed an array of UT probes into a pig-borne system. The pioneers in this field was Pipetronix, a German company that was established around 1984.

While MFL is a predicted metal loss system, UT is a direct measurement, earning it the reputation of a more accurate tool. UT however needs a liquid medium to transmit the sound waves into the pipe steel, meaning it is restricted to use in, for example, oil lines. It can be deployed into gas lines, but this needs to be within a liquid slug, which results in the project being more difficult and expensive.

Crack size matters

The technology that intelligent pigs are based on hasn’t changed throughout the intervening years, so the development of the industry has followed the demands of the market and the refinement of the systems employed. Operators demanded more accurate reporting and tools that were more sensitive to different types of features found within pipelines.

Standard MFL pigs are sensitive to wide areas of metal loss and to small isolated pits, but they are relatively blind to areas of axially long narrow corrosion. For the solution to this problem the inspection companies have developed tools where the magnetic field is applied around the pipe instead of axially along the pipe. These tools – known as transverse field inspection (TFI) pigs – have a niche market for pipelines where narrow axial external corrosion is known or suspected to exist.

Axial cracking caused by a variety of factors, such as stress-corrosion cracking (SCC), is another problem that the pipeline operators have to tackle. SCC can be a particular problem for gas pipelines in some areas of the world and detecting this type of cracking is not possible with standard tools (neither MFL nor UT). The TFI tool has had some limited success in detecting cracks but usually only in areas where the cracking was relatively severe. Ultrasonic tools have been developed that employ a technique whereby sound waves are fired into the steel at an angle to propagate around the pipe. These tools have proved to be very successful in detecting cracks, but are still limited by the fact that they need a liquid medium to operate in.

In 1974, a project was started at Harwell (home of the Atomic Energy Research Establishment) in the UK to develop a system for crack detection in gas pipelines. It worked on the principle of firing the sound wave into the pipe wall from an angled UT probe, which is carried inside a liquid-filled wheel with a polyurethane tyre that runs along the pipe.

The system was known as ‘Elastic Wave’ and was operated by British Gas. The pigs were notoriously difficult to maintain and expensive to run but they did offer a solution to the gas pipeline operators as they did not require them to have a liquid slug in their pipeline. The electromagnetic acoustic transmission (EMAT) tools are the latest development in the quest to accurately detect and size cracks. This technology is still in its infancy, but is being offered to the industry by all of the major inspection companies.

Current and future developments

As technology has advanced, computing power and data storage have become cheaper, faster and more compact. The cost of a digital recorder system needed to store vast amounts of pigging data is no longer $US100,000 per unit, but can be measured in hundreds of dollars today. Similarly, the data processing packs are still custom designed and built, but developments within the computing industry have resulted in these units no longer carrying the price tag of $US1 million they cost in the 1980s. This has enabled companies without huge financial support to enter the market, offering a more diverse choice of vendors to pipeline operators.

Alongside the ongoing development of tools to produce more accurate results both in general metal loss detection and reliable crack detection, many companies are refining the way in which the reports are presented to their clients. Pipeline operators have for a long time been supplied with pipeline listings which detail every weld, fitting and feature, whether it be corrosion or mill/manufacturing in origin. They are also supplying the operators with software which allows the pipeline engineers to view, and in some cases manipulate, the data from which the report is produced. Operators view the signals that have been recorded by the pigs and can also generate their own listings, filtering the reports based on the percentage of metal loss, feature clock position, pressure sentencing, etc. The quality of the report is one of the most critical parts of the inspection project. It is, after all, the end product for which the client is paying.

The search for the unpiggable solution

Now that the major inspection techniques – MFL and ultrasound – are well-established, recent and ongoing development has taken on a different focus. The industry is looking to solutions for ‘unpiggable’ pipelines, which are defined as pipelines that cannot be inspected by standard intelligent pigs.

They may operate at pressures not compatible with standard inspection pigs, or only have one point of entry – for example, a tanker loading/offloading line. It may be because of the lack of launch and receive traps, or it may be constructed from varying diameters of pipe, or have tight bend radii. The list is endless.

These issues have spurred a whole new direction in research and development for some companies. Inspecting unpiggable pipelines is the latest frontier that developers are tackling. Solutions range from pigs that can be launched through hot taps into the pipeline flow to robotic tractor units, tethered and autonomous, that can tow inspection vehicles. The inspection modules employ a variety of technologies: magnetic, MFL or remote eddy current, ultrasound, time of flight or transverse, and even visual, using cameras and laser light sources.