Skip to main content
OSTI.GOVU.S. Department of Energy Office of Scientific and Technical Information
U.S. Department of Energy
Office of Scientific and Technical Information
 

Physics-based hybrid machine learning for critical heat flux prediction with uncertainty quantification

Journal Article · · Applied Thermal Engineering
 [1];  [2];  [3];  [1]
  1. North Carolina State University, Raleigh, NC (United States)
  2. University of Tennessee, Knoxville, TN (United States)
  3. Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)
+ Show Author Affiliations

Critical heat flux (CHF) is a key quantity in nuclear system modeling due to its impact on heat transfer, safety margins, and reactor performance. This study develops and validates an uncertainty-aware hybrid modeling approach that combines machine learning with physics-based models to predict CHF in cases of dryout. The Biasi and Bowring empirical correlations were paired with three ML uncertainty quantification (UQ) techniques: deep neural network (DNN) ensembles, Bayesian neural networks (BNNs), and deep Gaussian processes (DGPs). A pure ML model without a base model was evaluated for comparison. Model performance was assessed under plentiful (7,350 points) and limited (9 points) training data scenarios using parity, uncertainty distributions, and calibration curves. Results show that the Biasi hybrid DNN ensemble achieved the best overall performance, with a mean absolute relative error of 1.846%, and well-calibrated uncertainty estimates. The BNN-based hybrids showed slightly higher error (2.14%) but superior uncertainty calibration. DGP models underperformed, with over 6% error and poor uncertainty calibration. All hybrid models outperformed pure machine learning configurations, demonstrating resistance against data scarcity. These findings indicate that hybrid modeling significantly improves predictive accuracy, interpretability, and resilience to data scarcity. The integration of uncertainty awareness provides actionable confidence in CHF predictions, which is vital for safety-critical decisions in nuclear applications. This hybrid approach offers a viable pathway for deploying ML models in reactor analysis tools while preserving domain knowledge and physical consistency.

Research Organization:
North Carolina State University, Raleigh, NC (United States); Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)
Sponsoring Organization:
USDOE Office of Nuclear Energy (NE)
Grant/Contract Number:
NE0009467; AC05-00OR22725
OSTI ID:
2571909
Alternate ID(s):
OSTI ID: 2573395
Journal Information:
Applied Thermal Engineering, Journal Name: Applied Thermal Engineering Journal Issue: A Vol. 279; ISSN 1359-4311
Publisher:
ElsevierCopyright Statement
Country of Publication:
United States
Language:
English

References (22)

A hybrid neural network-first principles approach to process modeling journal October 1992
Modeling chemical processes using prior knowledge and neural networks journal August 1994
Consideration of critical heat flux margin prediction by subcooled or low quality critical heat flux correlations journal June 1996
Predicting critical heat flux with uncertainty quantification and domain generalization using conditional variational autoencoders and deep neural networks journal September 2025
On the prediction of critical heat flux using a physics-informed machine learning-aided framework journal January 2020
Data-driven dimensional analysis of critical heat flux in subcooled vertical flow: A two-stage machine learning approach journal July 2024
Combination of support vector regression and artificial neural networks for prediction of critical heat flux journal July 2013
Prediction of critical heat flux and position in narrow rectangular channels using deep feed-forward neural networks coupling with empirical correlations journal April 2024
Dependence of critical heat flux in vertical flow systems on dimensional and dimensionless parameters using machine learning journal June 2024
Prediction of critical heat flux for narrow rectangular channels in a steady state condition using machine learning journal June 2021
The 2006 CHF look-up table journal September 2007
Deep generative modeling-based data augmentation with demonstration using the BFBT benchmark void fraction datasets journal December 2023
Optimized ensemble of neural networks for the prediction of critical heat flux journal August 2025
Uncertainty quantification study of the physics-informed machine learning models for critical heat flux prediction journal May 2024
Machine learning in critical heat flux studies in nuclear systems: A detailed review journal January 2025
Deep learning journal May 2015
Single-model uncertainty quantification in neural network potentials does not consistently outperform model ensembles journal December 2023
Investigation of Machine Learning Regression Techniques to Predict Critical Heat Flux Over a Large Parameter Space journal August 2024
Quantification of Deep Neural Network Prediction Uncertainties for VVUQ of Machine Learning Models journal November 2022
Methods for comparing uncertainty quantifications for material property predictions journal May 2020
A Practical Bayesian Framework for Backpropagation Networks journal May 1992
Comparison of Standalone and Hybrid Machine Learning Models for Prediction of Critical Heat Flux in Vertical Tubes journal March 2023

Similar Records

An investigation on machine learning predictive accuracy improvement and uncertainty reduction using VAE-based data augmentation
Journal Article · Thu Sep 11 00:00:00 EDT 2025 · Nuclear Engineering and Design · OSTI ID:3000581
Predicting critical heat flux with uncertainty quantification and domain generalization using conditional variational autoencoders and deep neural networks
Journal Article · Mon Apr 28 00:00:00 EDT 2025 · Annals of Nuclear Energy · OSTI ID:2563800

Related Subjects

22 GENERAL STUDIES OF NUCLEAR REACTORS
Critical heat flux
Dryout
Hybrid modeling
Machine Learning
Uncertainty Quantification