PPSA Webinar – Truly Unpiggable: The Use of Data Analytics and AI to Understand Integrity Management of Unpiggable Pipelines
We live in an increasingly virtualized world. Virtual meetings help us to stay connected with friends and colleagues. Virtual journeys help us to plan efficient travel to our destinations of choice. We even have virtual queues that mean we no longer stand in line for our favorite roller coaster. So is virtual pipeline inspection a realistic prospect?
Advances in in-line inspection (ILI) technology have almost always been driven by improvements to hardware components, including magnets, transducers, sensors and batteries, as well as to the software that converts sensor data into anomaly measurements. But the next great leap in inspection technology may have very little to do with hardware and sensor data. In fact, the next generation of inspection “tools” will have none of these things at all.
Over the past 30 years, huge volumes of ILI data have been collected for pipelines all around the world. Although these pipelines are diverse in their characteristics, many share similar threat profiles for common pipeline anomalies such as external corrosion, internal corrosion and cracking. This has led the industry naturally towards machine learning as a complementary monitoring solution for pipelines that cannot be inspected using ILI, or where ILI is undesirable. For a given uninspected pipeline, we can source data for similar inspected pipelines and use machine-learning algorithms to generate predictive models.
This is the essence of “Virtual ILI.” Virtual ILI can help establish a pipeline’s condition without having to think about the constraints of location, flow rate, tight bends, valves, tees or any of the other features that traditionally create obstacles for an ILI tool.
This webinar introduces the Virtual ILI process based on a global repository of ILI results and integrity management information known as the Integrity Data Warehouse (IDW). At the time of writing, the IDW contains results from almost 15,000 inspections, in addition to design, construction, environmental and operational information for approximately 10,000 inspected pipelines. We show how Virtual ILI can be used to predict useful “condition metrics” for a pipeline – at varying levels of resolution – that support integrity management decision-making.
Learn more and join our webinar, which will be presented on February 24, 2022 at 3 pm (UK Time, UTC).
- February 24, 20222022-02-24T05:00:00
- Learn more