According to its developers, Integrated External Corrosion Management (IECM) is an exciting framework for owners who are interested in improving system integrity. By working smart, using data to the fullest possible extent, and prioritizing system needs, the system aims to unify corrosion monitoring and mitigation strategies while minimizing and effectively managing many external corrosion risks.

Keith Parker, external corrosion specialist at Enbridge; Christophe Baete of Elsyca; and Joseph Mazzella and Tom Hayden of Engineering Director, Inc. (EDI) are among many collaborators on the project, and all were present at AMPP’s 2023 conference in Denver to discuss the latest IECM developments.

In Wednesday’s interactive symposium — designed to foster communication, information sharing, and best practices — the panelists addressed many components of IECM’s framework. This includes data governance, statistical and mechanistic modeling, real-time measurement, and optimization.

As a data intensive process, the presenters explained that IECM requires the gathering and integration of large volumes of data from multiple sources. From there, it relies on monitoring and predictive analytics to identify and assess corrosion threats while prioritizing various maintenance tasks.

IECM requires the application of sophisticated software to analyze the data while generating useful information that can support decision making.

One challenge is that pipeline data is almost entirely spatial, and this data can be harder to work with. After all, correlations exist in a degree of freedom outside of a spreadsheet or database. This raises questions as to the lifecycle of spatial data, as well as maintaining its governance. It can be particularly complicated at crossings related to human activity, such as roadways, railways, and power lines.

Thus, other approaches could be necessary. One option is mechanistic modeling, which conducts physical simulations of the real world using known laws. Another option is machine learning, which is especially well suited when known laws are unknown are too complicated (for instance, soil resistivity).

That's where digital twins come into play, according to the presenters. Digital twins allow for the integration of data governance, machine learning, and mechanistic modeling to create a comprehensive and accurate representation.

Machine learning algorithms can be used to create predictive models and insights to improve the performance of a system or process, while mechanistic models can be used to simulate the behavior of a system or process in different scenarios. This can allow for better decision making related to risk management.

As an example of IECM data collection, the presenters pointed to the option of collecting a soil sample at a probe location to help determine the soil resistivity. Data streams include the polarized potential, corrosion rate, DC and AC current density, and line current flow.

More information on IECM is available in recent technical articles published in Materials Performance (MP)Magazine.

‍