CTC designs a model capable of predicting corrosion in industrial pipelines seven days in advance

The CTC Technology Centre has successfully completed the development of a predictive model designed to estimate the corrosion rate within cooling system pipelines in industrial facilities. By strategically placing sensors within the conduits and employing various artificial intelligence techniques, a system has been created that can estimate the amount of material each pipe will lose to corrosive agents seven days in advance. This information is crucial for accurately determining the service life of the infrastructure, managing maintenance and ultimately enhancing efficiency in manufacturing processes.

The unveiling of the I-COR project’s findings marked a standout moment at its concluding meeting, showcasing an innovative national initiative poised to significantly enhance industrial facility productivity, reduce environmental impact and bolster safety measures.

Led by ArcelorMittal, a company renowned for its expertise in steel product manufacturing, this collaborative effort also featured contributions from CTC and the Asturian technology centre, IDONIAL. The I-COR initiative, with a budget of 650,000 euros, received funding from the Ministry of Economy, Industry and Competitiveness and the State Research Agency as part of the Retos Colaboración 2019 programme.

Despite facing the challenge of limited data availability –a direct consequence of the project’s limited timeframe– the team, under the leadership of Verónica González de Lena, CTC’s Industry and Energy Manager, secured robust results thanks to her deep expertise and the deployment of innovative strategies for monitoring corrosion. Specifically, the team focused on the primary and secondary cooling circuits of the rail train at the ArcelorMittal facility in Asturias.

Diagrama de concepto de I-COR

Two approaches, each with its unique application, were employed in these circuits. First, a virtual sensor was developed to estimate the corrosion rate based on variables affecting the process. This included estimating temperature fluctuations in industrial pipelines and the varying pH levels in the water.

Second, the team created a predictive tool capable of forecasting the corrosion rate’s trajectory over the seven days following any given date. This achievement was made possible by analysing both the data recorded during the preceding week and the factors that influence the progression of this chemical process.

After 42 months of research and a focused analysis period of just 10 months, the CTC team has developed a model with an impressively low average deviation from actual values. Across a measurement spectrum of 0 to 7 mpy (thousandths of an inch), the project’s most accurate algorithm demonstrated an average absolute error of just 0.14 mpy. This level of precision allows for reliable predictions regarding material loss due to corrosion, enabling a clear understanding of how quickly corrosion will advance in each installation.

As with all such developments, the reliability of such predictive models improves with the amount of data collected. The ability to extend the prediction window beyond seven days is directly tied to the amount of data collected. The more frequent the measurements, the more extended the period for accurate forecasts becomes. Thus, augmenting the data volume is crucial for applying this methodology across different sectors.

The efforts of the CTC team represent a segment of a comprehensive monitoring system that leverages artificial intelligence techniques and is designed for the cooling water circuits ArcelorMittal employs in its steel production processes. Given the enormous volume of water used annually by the multinational steel company –over 36.5 billion litres at its Gijón and Avilés facilities alone– the introduction of this monitoring tool is set to significantly influence maintenance expenditures.

The system’s utility, however, extends far beyond the confines of steel production technology. It holds the potential for future adaptation across a variety of industrial cooling circuits, including those found in the nuclear and thermal power, food, cement, paper and chemical sectors.

NACE International, the globally recognised authority on corrosion engineering, estimates the economic toll of corrosion to range from 3.5 to 5% of a nation’s Gross Domestic Product (GDP). To put it into perspective for Spain, this translates to an annual cost of between 51,170 and 73,100 million euros. In Cantabria, the impact is estimated to fluctuate between 537.4 and 767.7 million euros annually.

A study by this entity reveals that up to 30% of corrosion-related costs could be cut by adopting effective asset management and conservation practices. In response to this challenge, the CTC Technology Centre has been expanding its expertise and capabilities for years, aiming to devise innovative solutions to combat corrosion. The Industry and Energy team at the Cantabrian Centre boasts specialised expertise in developing anti-corrosion coatings and in the characterisation, simulation and evaluation of their performance.

Additionally, it has enhanced the capabilities of its MCTS El Bocal marine laboratory, improving the early detection of corrosion. Prominent global corporations, including Hempel, Viesgo and ArcelorMittal, have collaborated with CTC on projects to reduce the impact of corrosion on their operations.