Modeling corrosion for underground assets is a heavily studied area for many reasons; it's exciting, high impact, and challenging work. Many confounding variables make the analysis complex, such as cathodic protection systems and interaction effects with the climate and soil. This paper addresses the role of the soil-plant-atmospheric system (SPAS) in corrosion growth by presenting datasets, a statistical analysis, and predictive models. Using data from the International Soil Reference and Information Centre (ISRIC) and a North American pipeline operator, this research offers a correlation analysis, showing the impact of soil features and the critical role of cathodic protection systems. Incrementally, by adding variables and non-linear modeling, this paper shows a machine-learning powered framework for estimating corrosion growth rates with features from the soil-plant-atmospheric system, while controlling for confounding variables.
Key words: Pipelines, Corrosion Growth Rates, Soil, Atmospheric