FCTM Esope

S1.8- Life-Cycle Decision Optimization: Ensuring Long-Term Performance of Spent Nuclear Fuel (SNF) Dry Storage Containers (DSC) Subject to Chloride Stress

Oct 6, 2021 | 5:00 PM - 5:30 PM

Scénario

Description

When nuclear reactor performance drops below desired energy levels, the radioactive waste must be properly disposed of. Until deep geological disposal sites are identified for very long-term storage of this waste (for thousands of years), interim storage methods (wet and dry) must be used for much longer than initially intended (~100 years). With hundreds of Dry Storage Canisters (DSCs) approaching or already exceeding their intended design life, the nuclear utility sites need Aging Management Plans (AMP) to assess the long-term integrity of DSCs. With recent advances in technology, Artificial Intelligence Bayesian Decision Networks (AI-BDNs) are utilized as a probabilistic computational engine for a complete life-cycle management (LCM) asset integrity system that can utilize multiple, often disparate, sources of knowledge to make more informed day-to-day decisions. Compilation of AI-BDNs, for such industrial applications, requires cloud-based HPC methods and algorithms. Incorporating such a tool into daily operations promotes pro-active decision-making and risk management to minimize the likelihood of a potentially catastrophic failure event (loss of the DSC’s containment boundary, resulting in increased radiation exposure and need for costly repackaging). For Spent Nuclear Fuel (SNF) DSCs, Chloride Stress Corrosion Cracking (CLSCC) has been identified as the primary damage mechanism that may lead to breach of a canister’s containment boundary during normal storage operations. As a result, the AI-BDN framework developed herein focuses on CLSCC as the primary failure mode, however, the methodology is directly applicable to other damage mechanisms, across all aging energy industries. Considerations for extension to additional DSC failure modes, including both accident and normal operating situations, are also discussed. Not only is there an economic benefit to utilizing AI-BDNs for making better decisions that lower the risk of failure, the public benefits as well.

Presented by