Digital: Production electrolysis monitoring models

Main article: EAM system

2021: Creation of digital models of diagnostics of operation of electrolyzers in production

On November 17, 2021, Rosatom announced that Zifrum had developed digital models for diagnosing the operation of electrolyzers in production. They are designed for early detection of deviations from normal operation and ensuring trouble-free operation of electrolysis plants.

This was told by the expert of the department of artificial intelligence (AI) "Qifrum" Yuri Katser.

Digital models of predictive analysis and monitoring of equipment in the diagnostic processes of the operation of electrolyzers are developed on the basis of artificial intelligence technologies. They have the function of automated search of hidden defects that occur during operation, provide visualization of information about predicted events and notification of the operator for making decisions, as well as form analytics about anomalies and factors that made the most contribution to their detection.

During the electrolysis of the model, various process parameters are measured, including electrolyte level, voltage, pressure in off-gas headers, etc. The obtained indicators allow predicting and detecting various problems with equipment.

Unlike existing diagnostic systems and algorithms, digital models allow you to better manage the risks of technical defects and take preventive measures to prevent forced equipment stops. As a result, the downtime of the equipment is significantly reduced, the overhaul interval increases and the share of automation in the diagnostic processes increases.

The created model allows you to reduce the number of major repairs, increase the time between them and potentially increase production efficiency, "said Yuri Katzer.

Predictive analytics solutions are actively used in the nuclear industry to further develop production and labor safety, as well as increase the economic efficiency of the business. A competence center for predictive analytics technologies has been created in Zifrum.