CTC innovates in the control of corrosion in industrial plant cooling system pipes

The CTC Technology Centre is developing algorithms and methodologies for a predictive model capable of quantifying the corrosion rate of carbon steel pipes in the water-cooling systems of industrial plants. This is an innovative approach, the main objective of which is to optimise productivity, reduce maintenance costs and minimise risks caused by plant failures. This technological solution will be part of the I-COR project, which will facilitate the transition from conventional maintenance programmes to a predictive model based on the information generated by mass data processing technologies.ArcelorMittal, an innovative company, specialised in the production of steel products, is leading this research in which, besides CTC, the Asturian technology centre IDONIAL is also participating. I-COR is an initiative with a 40-month execution term and a 650,000-euro budget funded by the Ministry of Economy, Industry and Competitiveness and the State Research Agency, within the Retos Colaboración 2019 Programme.

The project considers that the digitalisation of the value chain, thanks to the integration of technologies based on Big Data or Machine Learning, will lead to more intelligent control of production conditions. A leap into industry 4.0 that will lead to an advanced corrosion management system. An online parameter monitoring system will be developed to allow large-scale data to be processed with neural networks and artificial intelligence to obtain an evolutionary model and a tool for predicting the speed of corrosion.


Water cooling circuits are a critical part of many industrial installations. These systems are used to control the temperature and pressure of processes and guarantee the ideal manufacturing conditions for certain products and components. However, the carbon steel pipes that make up these systems quickly form iron oxides when they come into contact with water. This circumstance causes a gradual loss of energy efficiency and thickness, leaks and breakages that can lead to production stoppages. The key factor is the identification and characterisation of the corrosion problems in an early stage so corrective actions can be taken instead of facing corrosion after its appearance.

The project raises the possibility of establishing remote monitoring through different sensors to contribute to acquiring crucial data on the state of the equipment or installation. It would, therefore, be possible to know a priori if it is in the right operating conditions. This commitment to predictive maintenance would improve the response time to any incident and significantly reduce the costs associated with this field.

If I-COR achieves the expected objectives, the exploitation of the results will allow acquiring more knowledge about the water cooling circuits in the different steel processes and how their variables influence the corrosion rate. This is highly valuable information for a company like ArcelorMittal, which, according to EUROFER’s annual report, accounts for 26.7% of European steel production and 42.7% of national production.

In addition, inspection plans and maintenance work will be updated to better match the actual conditions required and the human resources dedicated to this task will be optimised. This monitoring system can go beyond the scope of the steel industry and be applied with the required adaptations to the cooling circuits of energy companies or other industrial sectors such as the food, cement, paper or chemical industry.

Modelling of corrosion processes is highly developed in the oil and gas industry. However, in the water circuits, it has barely begun. The corrosion process is a complex process that involves many variables. Plant operators who monitor, control and diagnose this process typically encounter serious difficulties in analysing the data collected and diagnosing anomalies. Systems such as I-COR, which incorporate artificial intelligence, can help in these tasks, assuming the role of monitoring, control and diagnosis.