Session: FSI-02-03 Flow-induced Vibration III

Paper Number: 62201

Start Time: Wednesday, July 14, 2021, 08:00 PM

62201 - Coupling of Neutron Noise and Dynamic Finite Element Analyses to Perform Remote Condition Monitoring for Reactor Vessel Internals 

During a Spring 2019 refueling outage, Plant A, a pressurized water reactor (PWR), experienced difficulty removing the reactor vessel upper internals from the vessel. Resulting inspections identified four degraded thermal shield support block bolts, two confirmed and four inconclusive thermal shield flexure indications, and 232 degraded baffle-former bolts. None of the degradation was expected based on prior industry operating experience. Therefore, to understand the potential source of degradation in these components and further investigate the four inconclusive thermal shield flexure indications, dynamic finite element analyses (FEA) and neutron noise (NN) monitoring were performed. NN is a method of using ex-core neutron detector signals to measure the dynamics of reactor internals components. The FEA simulated the confirmed and inconclusive indications from the 2019 outage to assess the effect that degradation would have on the dynamic response of the lower reactor vessel internals (core barrel and thermal shield). Meanwhile, NN data was collected and evaluated to characterize a signature of the vibratory response of the Plant A lower internals. This vibratory response did not align with baseline data for other PWRs, including the sister unit of Plant A, all of which exhibited a normal response. Therefore, the results of the FEA and NN were coupled to compare degradation scenarios with the measured data. From this work, it was predicted that multiple non-functional (fully failed) flexures existed at Plant A. It should be noted that the two confirmed indications during the Spring 2019 outage were small flaws and not representative of a non-functional flexure. Therefore, additional flexure degradation, such as the inconclusive indications being actual indications, was determined to exist at Plant A. During the subsequent Fall 2020 outage, Plant A re-inspected their thermal shield flexures and identified that Plant A had five non-functional thermal shield flexures, thus validating the NN analysis results. This use of NN to identify degradation represents the atypical practice of proactive remote condition monitoring for reactor vessel internals, and allowed the plant to preemptively prepare replacement contingency hardware and tooling to remediate the damage within the same Fall 2020 outage that the failures were confirmed. With this demonstration of the capabilities of NN used for remote condition monitoring, the means through which this monitoring is performed is currently being enhanced. By exercising benchmarked FEA models through a computational framework that pre-computes the effects of different degradation scenarios on the structural dynamics observable via NN measurements, machine learning can be used to enhance interpretability of NN results and proactively diagnose different types of degradation.

Presenting Author: Gregory Banyay Westinghouse Electric Company

Authors:

Gregory Banyay Westinghouse Electric Company
Matthew Palamara Westinghouse Electric Company
Stephen Smith Westinghouse Electric Company