New Seismic Diagnostic Technology for NPP Equipment

Researchers from the Ulsan National Institute of Science and Technology (UNIST) and the Korea Research Institute of Standards and Science have developed an artificial intelligence (AI)-based virtual sensing system for rapid prioritization of inspections of critical equipment at nuclear power plants (NPPs).

The motivation for this system arises from the challenges of inspecting critical equipment in auxiliary buildings of NPPs, due to restricted access to radiological zones and high maintenance costs. The new virtual sensing system analyzes seismic data collected by a single high-quality sensor and predicts acceleration responses at 139 locations inside the building within 0.07 seconds. These acceleration responses indicate how strongly and rapidly equipment is shaken during an earthquake, helping prioritize areas and devices that require urgent inspection.

Implementation of this technology enables rapid equipment inspections in complex auxiliary buildings at NPPs, ensures reliable operation even under field conditions, and provides high-precision seismic response predictions with minimal physical sensor deployment. Overall, the proposed virtualization method allows continuous and accurate monitoring while significantly reducing installation and maintenance costs, ultimately enhancing operational safety and resistance to seismic risks.