Intelligent corrosion analysis and life prediction of ductile iron pipe systems using machine learning and electrochemical sensors

Abstract

This study established a circulating system to control the concentration of substances and temperature in the aqueous solution. Simultaneously, sensors were used to continuously monitor the corrosion of three pipe materials: ductile iron (DI), surface-treated ductile iron (SDI), and carbon steel (CS). A corrosion decision model based on a machine learning framework was developed for data mining. The results show that the developed model provides accurate corrosion prediction strategies. Analysis revealed that high temperature is the primary factor accelerating corrosion in water systems. SDI accelerates at 60 °C, reaching its peak at 90 °C, while DI and CS peak at 80 °C. The superior corrosion resistance of SDI is attributed to its ability to withstand accelerated corrosion under high temperatures and environmental coupling, making it more stable when immersed in water.

Keywords

• Water pipeline
• Monitoring
• Machine learning
• Corrosion
• Big-data