Prediction of Cable Thermal Failure in Switchgear Rooms of Nuclear Power Plants Based on Numerical Simulation and Deep Learning

Thermal failure of critical cables in nuclear power plants may result in the loss of safety functions, thereby compromising the safe and stable operation of the plant. This study focuses on the switchgear room of a nuclear power plant and investigates cable thermal failure under FDS-simulated fire scenarios with varying HRR, while the ignition location, ventilation, compartment openings, and cable arrangement are fixed, employing numerical simulation and deep learning approaches. First, numerical simulations of switchgear cabinet fires were conducted using the Fire Dynamics Simulator (FDS) for 15 different heat release rates (HRR). Second, to address the limitation that existing models mainly predict static outcomes, three deep learning models were trained to capture the continuous dynamic temperature evolution on datasets derived from the simulation results, and their predictive performance was evaluated using the R2, RMSE, and MAE. Third, taking 330 °C as the thermal failure threshold for cables, the thermal failure behavior under representative HRR conditions was predicted. The results indicate that the MLP model achieves the best prediction accuracy, with R2, RMSE, and MAE values of 0.96, 11.63, and 5.77, respectively, whereas the Transformer model exhibits the poorest performance, with corresponding values of 0.90, 28.23, and 9.92. Using the optimal MLP model, the thermal failure behavior of cables under typical conditions was further analyzed. The predictions reveal that the critical HRR of the switchgear cabinet at which the critical cables reach thermal failure is 550 kW, and the time to failure exhibits a nonlinear decay with increasing HRR. Specifically, at HRR levels of 702 kW and 1002 kW, the predicted failure times are 709 s and 594 s, respectively. These findings demonstrate that deep learning models can efficiently and accurately predict temperature evolution under FDS-simulated fire scenarios in a nuclear power plant switchgear room, thereby providing methodological support for fire protection design and emergency decision making.

Authors

Institutions

Publication Details

Journal
Fire
Published
2026-11-18
DOI
https://doi.org/10.3390/fire9110523
Primary Topic
Fire dynamics and safety research
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Prediction of Cable Thermal Failure in Switchgear Rooms of Nuclear Power Plants Based on Numerical Simulation and Deep Learning

Hongru Zhao, Xinhui Fu, Jiling Wu, Xu Zhai et al.
1 links
Fire
Fire dynamics and safety research
article

Prediction of Cable Thermal Failure in Switchgear Rooms of Nuclear Power Plants Based on Numerical Simulation and Deep Learning

Hongru Zhao, Xinhui Fu, Jiling Wu, Xu Zhai, Jun Zhao, Jinlong Zhao
article en

Abstract

Thermal failure of critical cables in nuclear power plants may result in the loss of safety functions, thereby compromising the safe and stable operation of the plant. This study focuses on the switchgear room of a nuclear power plant and investigates cable thermal failure under FDS-simulated fire scenarios with varying HRR, while the ignition location, ventilation, compartment openings, and cable arrangement are fixed, employing numerical simulation and deep learning approaches. First, numerical simulations of switchgear cabinet fires were conducted using the Fire Dynamics Simulator (FDS) for 15 different heat release rates (HRR). Second, to address the limitation that existing models mainly predict static outcomes, three deep learning models were trained to capture the continuous dynamic temperature evolution on datasets derived from the simulation results, and their predictive performance was evaluated using the R2, RMSE, and MAE. Third, taking 330 °C as the thermal failure threshold for cables, the thermal failure behavior under representative HRR conditions was predicted. The results indicate that the MLP model achieves the best prediction accuracy, with R2, RMSE, and MAE values of 0.96, 11.63, and 5.77, respectively, whereas the Transformer model exhibits the poorest performance, with corresponding values of 0.90, 28.23, and 9.92. Using the optimal MLP model, the thermal failure behavior of cables under typical conditions was further analyzed. The predictions reveal that the critical HRR of the switchgear cabinet at which the critical cables reach thermal failure is 550 kW, and the time to failure exhibits a nonlinear decay with increasing HRR. Specifically, at HRR levels of 702 kW and 1002 kW, the predicted failure times are 709 s and 594 s, respectively. These findings demonstrate that deep learning models can efficiently and accurately predict temperature evolution under FDS-simulated fire scenarios in a nuclear power plant switchgear room, thereby providing methodological support for fire protection design and emergency decision making.

FireVol. 9(11)
China University of Mining and Technology - Beijing (CN), Tsinghua University (CN)
National Natural Science Foundation of China, National Key Research and Development Program of China
Affordable and clean energy, Industry, innovation and infrastructure
Fire dynamics and safety research
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon