In the process of coal beneficiation, a magnetite-water slurry (called media) is used, the composition of which is Fe3O4 (iron oxide) + H2O (water) + Coal. This slurry is carried by pipelines spread across the plant (~3 km) with multiple bends. The pipelines are of 6” to 14” in diameter and located at a height of up to 30 m. The material of the pipelines is Mild Steel (MS) with or without cast basalt lining inside, or High-density polyethylene (HDPE).
The slurry is corrosive. Leakages are evident on the bends and welded portions of the pipelines. Currently, a shutdown of ~ 16 hours is taken after every few days, during which the pipeline is inspected visually for fault identification and replacement of pipeline sections. There is no method to predict likely future failures. Moreover, there are several unapproachable regions where the inspection is almost never carried out, which increases the risk of pipeline failure. Drone-based inspection is not possible as approaching the pipelines is not possible due to space constraints.
Thus, we require a system which can predict and alert possible pipeline failures, for proactive planning of shutdown and repair/replacement.