What lies beneath

1 May 2019



The integrity of legacy subsea pipelines is a crucial issue for the industry. Grace Allen talks to Dr Henry Tan, senior lecturer in the University of Aberdeen’s School of Engineering, and Neil Gordon, chief executive at Subsea UK, on the challenges in managing aging assets and the advances in data solutions and technologies that are improving this process.


Despite exploration that has taken our understanding beyond the Earth’s atmosphere and into the darkness of space, the deep sea of our own planet remains largely unfathomable. The ocean floor can seem as remote as Neptune or Jupiter.

For the oil and gas industry, this inaccessible environment is what makes subsea pipeline failure so potentially catastrophic – it takes time and vast effort to locate and fix a leak, resulting in huge financial, environmental and reputational losses.

In Cook Inlet, Alaska, a natural gas leak from an 8in subsea pipeline was spotted in February 2017. The pipeline was losing an estimated 210,000 to 310,000ft3 of natural gas into the water per day and possibly threatening the habitat of wildlife, which included the critically endangered Cook Inlet beluga whale. The Pipeline and Hazardous Materials Safety Administration suggested that the leak might have gone on undetected since December 2016; it was not fixed by the pipeline’s owner, Hilcorp, until April 2017.

The pipeline in question was over 50 years old, constructed in 1965. The continuing use of legacy infrastructure – which may be in service long after the end of its intended design life – is a challenge faced by many in the business, as life extension is so much more financially viable than replacement. It’s familiar to Neil Gordon, chief executive of industry body Subsea UK.

“If you look at not necessarily just the pipelines, but the whole offshore infrastructure in the North Sea, many years ago, we said, yeah, a 20-year lifespan, and most of these things will be gone,” he says. “We now have Brent, Beryl and a number of other fields which are pushing over 40 years old.”

At long last: pipes that age gracefully

While the concept of ‘ageing’ as it relates to these pipelines refers not to the passage of time but their material degradation, older pipelines have been exposed to potential risk factors for longer and managing their integrity is a significant priority. “They’re designed to last, these pipelines, for many, many years,” Gordon says. “But as we've seen over the years, those pipelines have had their life extended, and we just want to make sure that they are fully integrity safe for operation.”

Corrosion is one of the most significant causes of material ageing in pipelines; high hydrocarbon volumes as well as water, carbon dioxide and hydrogen sulphide can cause thinning of the steel from inside the pipeline, while variables such as salinity and temperature have an impact on the outer surface.

Furthermore, corrosion rates can vary. “If it’s a highly corrosive product that they’re putting through [a pipeline], then they might have a more intense inspection regime to have a look at that, because that might have a shorter lifespan than something that was fairly benign,” Gordon explains.

Integrity management also needs to consider the many other potential hazards that can affect the integrity of pipelines. Hilcorp’s pipeline in Cook Inlet was damaged where it lay on a boulder on the sea floor, while a crack in a pipeline off the coast of Balikpapan, Indonesia, in 2018 was reportedly caused by an anchor.

Shifting sea beds, instability of the sea floor and currents are further issues with which pipeline experts are familiar. “It could perhaps have things like free span, where a pipeline may be not supported underneath, because the tide or the current has actually moved a lot of the seabed away from it so it might be unsupported,” Gordon says.

Furthermore, for older assets, data on the original construction specifications and historical assessments may be missing as years pass and pipelines change hands between owners. This makes it even harder to judge how a pipeline may be affected by corrosion, erosion and other similar hazards.

This year’s modelling: systems to inspect integrity

Dr Henry Tan works on the development of mathematical modelling systems to predict pipeline condition and deterioration at the University of Aberdeen, and the lack of this data is an issue he has encountered. “When we do the life extension, we want to predict the life of the subsea pipeline: we need the historical data,” he says. “Sometimes for this kind of historical data, we don’t have a complete set of the data from the designing stage to the current stage.”

“Machine learning and artificial intelligence is coming so fast, once the technology can be applied to one area it will soon be applied to many other related areas.”
Dr Henry Tan

To assess their integrity, the pipelines need to be inspected, but their challenging environment makes this an operation that requires significant investment, skilled technicians and specialist equipment. The employment of risk or conditionbased monitoring, which targets inspections on areas deemed more likely to experience problems, streamlines the process. Regularly scheduled inspections may not pay enough attention to high-risk assets and waste resources on those less in need of inspection.

Risk-based monitoring requires a predictive element, judging the likelihood of an asset’s failure, and allocating resources appropriately. At the University of Aberdeen, Tan and his colleagues are developing a data modelling methodology that makes better use of the data available to model pipeline corrosion and predict residual life. “It’s more on the softer technology,” Tan explains. “Not the hard technology like how to get the data, but how to use the data you have so far.”

Tan, with his colleagues Reza Aulia and Srinivas Sriramula, outlines the proposal in an article due to be published in the proceedings of the 26th CIRP Life Cycle Engineering Conference. They employ dynamic Bayesian networks, a soft computing technique that uses a directed acyclic graph to lay out the relationships between causes and effects. It is particularly suited for situations where information is uncertain or incomplete, and allows for the input of both past and present data as well as qualitative data such as expert opinions. In the model, variables such as carbon dioxide, oxygen and hydrogen sulphide levels, pH, temperature and pipeline diameter and material, as well as past records of these variables, combine to produce an assessment of the probability of pipeline failure.

The use of this modelling to better understand risk and direct assessment resources has great potential for the industry. “We have the theory, the model, and then we develop the software that can be used by industry,” Tan explains. “So we are involved from the modelling to the software developing.”

“As we’ve seen over the years, [older] pipelines have had their life extended, and we just want to make sure that they are fully integrity safe for operation.”
Neil Gordon

A scanner sharply: a revolution in pipeline inspection

When it comes to actually undertaking assessments to monitor pipeline integrity, advances in hard technology, too, are improving accuracy and driving down costs, as well as allowing increased visibility on challenging pipelines that could not be inspected using methods such as pigging. Remotely operated vessels, operated from a shore-based control centre, or autonomous vessels can be used to obtain high-quality images of a pipeline as well as other information. In December 2018, survey company Swire Seabed conducted what it called the first autonomous inspection of a pipeline for Equinor, collecting bathymetrical, synthetic aperture sonar and visual data.

The incorporation of scanning technologies is also revolutionising inspection. “Some of the technologies that we’ve received more recently are how we can actually look from the outside of the pipeline, inside,” Gordon says. “The intrusive sort of inspection is done through loading in pigs and having a look from the inside, and we’re starting to get a bit cleverer now with the unobtrusive, so looking in from the outside.” A CT scanner for subsea inspection was launched by Tracerco in 2013, able to inspect the pipeline wall and the contents of coated pipelines.

The more data that is collected, the more it can be used in applications with artificial intelligence and machine learning capabilities, speeding up the process of flaw detection and improving its accuracy. “When you introduce technology at that level, you can accelerate the pace of the data acquisition and the actual interpretation of some of the flaws in there,” Gordon says.

The pace of innovation means that integrity management is set to revolutionise further in the years to come, as Henry Tan makes clear. “Machine learning and artificial intelligence are coming so fast, once the technology can be applied to one area it will soon be applied to many other related areas,” he says. As the world’s energy supply continues to rely on legacy pipelines, technologies that allow us to catch a clearer glimpse of the ocean floor become ever more vital.

New technology is constantly monitoring the projected lifespan of subsea pipelines.


Privacy Policy
We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.