A predictive analytics system has been put into commercial operation at unit 4 of Russia’s Beloyarsk NPP, which has a BN-800 sodium-cooled fast reactor. The system searches for deviations in the operation of plant equipment based on automatic real-time analysis of large amounts of data and comparison of parameters.
The project, which began phased implementation in 2020, is aimed at increasing the safety and operational reliability of nuclear power plant equipment and minimising cases of damage or failure due to the detection of defects at an early stage of development. Also, timely information received about detected defects and the nature of their development make it possible to optimise the volume of maintenance, repair and operation of equipment.
The system automatically predicts deviations from normal operation of equipment based on an analysis of the history of changes in its technological parameters, which are recorded by monitoring and diagnostic systems in real time. Next, the system determines a list of those parameters on which deviations have a significant impact, and estimates the time left before they go beyond safe values.
“Based on the experience of operating the system at the Novovoronezh NPP, we can safely say that the use of predictive analytics technologies will not only improve the safety and reliability of Russian NPPs as a whole, but also reduce the risks of fines during unscheduled shutdowns,” said Viktor Tsaranu, head of the project office for implementing digital solutions during NPP operation at nuclear utility Rosenergoatom (part of Rosatom). “It will also reduce the duration of downtime due to repairs and minimise the number of violations and equipment failures.”.
Artem Golovaty, Director of Information Technology at Rosenergoatom, noted that predictive analytics is a significant area for the development of digitalisation. “We are not just introducing a new system – we are integrating a modern IT tool based on big data analysis and adaptive algorithms into the operational process. The predictive analytics system will allow us to move on to proactive equipment condition management,” he said.
The predictive analytics system has been successfully operating since December 2022 at Novovoronezh unit 6 and by the end of 2026 it will be implemented at Novovoronezh 7 and units 5&6 at the Leningrad NPP.
The main contractor for the project is the All-Russian Research Institute for the Operation of Nuclear Power Plants (VNIIAES – Vserossiiskii Nauchno Issledovatelskii Institut po Expluatatsii Atomnikh Elektrostantsii).
