Estimating corrosion growth rate for underground pipelines is a non-linear multivariate problem. There are many potential confounding variables such as soil parameters cathodic protection AC/DC interference seasonal / climate conditions and proximity to unique geographic features such as wetlands or polluted environments. The work presented provides an approach for estimating underground corrosion growth rates using a dataset of observations from a North American pipeline operator. Extensive geospatial data is utilized that has been obtained from public and private sources and extrapolated using Inverse distance weighted (IDW) interpolation. This work presents a model using IDW to estimate parameters involving soil interference geography and climate factors for any location in North America.Using this data this work then presents several different machine learning approaches including Generalized Linear Models eXtreme Boosted Trees and Neural Networks. All three provide an accurate estimation for corrosion growth rates for an underground asset at any latitude and longitude pair in North America. Each method comes with potential benefits and pitfalls specifically; trade-offs between model accuracy and transparency. This work presents a framework for comparing geo-spatial and machine learning estimates. Findings and a framework are provided for owners to assess how to think machine learning on their own assets.

Thermal insulation is used in operating facilities to conserve heat and protect against freezing amongst others. A consequence of insulating the pipe is the necessity to manage the introduced threat of corrosion under insulation (CUI). For CUI to occur water and oxygen must enter and migrate through the insulation to reach the external surface of the pipe. The water transport characteristics are dependent on several factors such as type of insulation type of jacketing pipe operating temperature external weather water entry/leakage rate and cyclic service. In hot piping there are competing water transport characteristics as in water vapor moves outwards away from the pipe as water enters into insulation. Knowing the water transport and the parameters that influence the time of wetness at the metal surface helps in understanding conditions favoring CUI.The use of transient hygrothermal models for moisture control is well established in the building insulation design codes and standards. The building designs naturally shed the liquid water to minimize entry and facilitate breathing of vapor so that moisture doesn’t accumulate within building. Several building industry hygrothermal models have been developed and are available for commercial use. One such commercial model has been used to understand water transport in a CUI application. The case study involves evaluation of piping and pipeline installed with a closed cell polyurethane insulation. The hygrothermal model provided insights on the parameters influencing the time of wetness and the ease of water escaping the pipe-insulation-jacketing system. Additional results comparing different insulations are also presented.Key words: Corrosion under insulation water transport hygrothermal models building industry polyurethane insulation

Corrosion under insulation (CUI) is a critical challenge that affects the integrity of assets for which the oil and gas industry is not immune. Over the last few decades, both downstream and upstream industry segments have recognized the magnitude of CUI and challenges faced by the industry in its ability to handle CUI risk-based assessment, predictive detection and inspection of CUI. It is a concern that is hidden, invisible to inspectors and prompted mainly by moisture ingress between the insulation and the metallic pipe surface. The industry faces significant issues in the inspection of insulated assets, not only of pipes, but also tanks and vessels in terms of detection accuracy and precision. Currently, there is no reliable NDT detection tool that can predict the CUI spots in a safe and fast manner. In this study, a cyber physical-based approach is being presented to identify susceptible locations of CUI through a collection of infrared data overtime. The experimental results and data analysis demonstrates the feasibility of utilizing machine-learning techniques coupled with thermography to predict areas of concern. This is through a simplified clustering and classification model utilizing the Convolutional Neural Networks (CNN). This is a unique and innovative inspection technique in tackling complex challenges within the oil and gas industry, utilizing trending technologies such as big data analytics and artificial intelligence.