Predictive Analytics and Machine Learning in Integrated External Corrosion Management (AMPP 2023)

Christophe Baeté, Elsyca, Inc.; Keith Parker, Enbridge Pipelines, Inc.; Thomas Hayden, Joseph Mazzella, Engineering Director, Inc.

Paper presented at the AMPP Annual Conference + Expo, Denver, CO, USA, March 2023

Paper Number: AMPP-2023-19393

Abstract

Integrated External Corrosion Management (IECM) is a novel framework developed for pipeline operators to model, identify, and optimize external corrosion risk and costs using a data-driven approach. Over the last decade, Machine Learning (ML) has transformed industries from consumer technology to product design to industrial systems. In corrosion, the Association for Materials Protection and Performance (AMPP) has added a symposium for specialists designing and optimizing machine learning algorithms detection and management. This work is not about a specific algorithm or technology set. Instead, this work presents a framework for incorporating the output of a predictive algorithm with an IECM framework. This work considers the interplay between in-line inspection (ILI), direct assessment, close interval surveys, and mechanistic modeling. Lastly, this work describes an external corrosion management system that is fully "observable", an environment where the state of any component in a pipeline system can either be directly observed or inferred in near real-time.

Introduction

Managing external corrosion, especially for underground assets, is a significant challenge dating back to the first underground pipeline in 1865. The very first issue of the journal, CORROSION, featured a headline story on this subject [1]. This subject is fundamental for corrosion engineers and pipeline operators. More broadly, managing pipeline infrastructure is a core part of life in many parts of the world, documented in the United States by the Smithsonian’s aptly named paper "Slappin' Collars and Stabbin' Pipe" [2]. Over the last 150 years, technology has driven innovation, from the improved metallurgy of the

19th century to the standardization of cathodic protection in the 20th century, to the present day, with the introduction of advanced information processing technology. The newest addition to this toolset is the introduction of machine learning systems, tools designed to assist the decision-making process of an integrity engineer.

This works intended audiences are operators searching for ways to introduce automation into their external corrosion management program. This work utilizes the IECM framework, to demonstrate a methodology to incorporate machine learning into the corrosion engineer's toolbox [3][4].

Link to AMPP 2023 proceeding