Initial Assessment of Infrasound Signatures Tied to Nuclear Facility Operations

Research is now underway to explore the use of persistent, on-site monitoring of infrasound and low-frequency acoustic signals as a new method for real-time sensing to assess the operational status of nuclear fuel-cycle facilities. Using an ensemble of smartphone-based sensors placed in key locations in and around a high-power research reactor and an associated radiochemical hot cell facility at Oak Ridge National Laboratory, data is being collected and analyzed in real time in a cloud-based computing environment. Waveform signal analysis detects and flags key acoustic signals that have an association with specific facility activities and facility operational states. These metadata event flags are used to identify these conditions, allowing the use of event classifiers to identify conditions likely to indicate specific facility activity or status conditions exist. This work is taking place within a larger effort titled Multi-Informatics for Nuclear Operations Scenarios (MINOS); MINOS aims to integrate multiple non-traditional data sources (where infrasound is just one of many observational modalities) to develop a comprehensive data environment where modern data analytics tools can be used to extract understanding of nuclear facility operations out of low-specificity observations. Results reported here include a summary of an acoustic signature compendium developed to define the key frequency and temporal acoustic patterns associated with important facility activities including crane movement, access door operations, reactor cooling system operation, reactor power transitions, and hot cell transfer system operations.