PSAM 16 - Paper Overview

PSAM 16 Conference Paper Overview

Welcome to the PSAM 16 Conference paper and speaker overview page.

Lead Author: Ha Bui Co-author(s): Tatsuya Sakurahara (sakurah2@illinois.edu) Seyed Reihani (sreihani@illinois.edu) Ernie Kee (erniekee@illinois.edu) Zahra Mohaghegh (zahra13@illinois.edu)

Probabilistic Validation Methodology for Probabilistic Risk Assessment
Emergent safety concerns such as risk-informed, performance-based fire protection usually involve complex phenomena whose outcomes are very likely to change as a result of spatiotemporal variations in their compounding events. Addressing these safety concerns in a risk-informed decision-making framework would require the risk estimation to have an adequate degree of spatiotemporal resolution. As a result, use of advanced modeling and simulation has recently received growing attention from academia, industry, and regulatory agencies to capture those complex and highly spatiotemporal phenomena more realistically within the Probabilistic Risk Assessment (PRA) of existing Nuclear Power Plants (NPPs). Advanced modeling and simulation have also been used to accelerate the analysis, design, and operationalization of advanced nuclear reactors. However, before a simulation model can be used in PRA to support risk-informed decision making, its validity needs to be adequately evaluated. Common approaches for validation of simulation models use empirical validation, where model predictions are directly compared against empirical validation data. However, conducting an adequate empirical validation for those simulation models that are used in PRA to help address emergent safety concerns at existing NPPs and to support the development of advanced nuclear reactors would be challenging, if not impossible, due to (i) a lack of validation data at the model output level; (ii) available data at lower levels (e.g., model inputs) being subject to various sources of uncertainty; and (iii) available data not being fully applicable for the specific context and conditions under which the model needs to be used. This paper reports on the theoretical and methodological development of a systematic and scientifically justifiable validation methodology, namely the Probabilistic Validation (PV) methodology, to facilitate the validity evaluation of advanced simulation models under the U.S. Nuclear Regulatory Commission (NRC) risk-informed regulatory framework. The PV methodology advances the scientific usage of epistemic uncertainty as a quantitative metric of validity for simulation predictions, where validation follows the National Research Council’s recommendation that models should be made “as useful as possible by quantifying how wrong they are.” Results of a case study to demonstrate the feasibility of the PV methodology in the context of Fire PRA are included.

Probabilistic Validation Methodology for Probabilistic Risk Assessment

Emergent safety concerns such as risk-informed, performance-based fire protection usually involve complex phenomena whose outcomes are very likely to change as a result of spatiotemporal variations in their compounding events. Addressing these safety concerns in a risk-informed decision-making framework would require the risk estimation to have an adequate degree of spatiotemporal resolution. As a result, use of advanced modeling and simulation has recently received growing attention from academia, industry, and regulatory agencies to capture those complex and highly spatiotemporal phenomena more realistically within the Probabilistic Risk Assessment (PRA) of existing Nuclear Power Plants (NPPs). Advanced modeling and simulation have also been used to accelerate the analysis, design, and operationalization of advanced nuclear reactors. However, before a simulation model can be used in PRA to support risk-informed decision making, its validity needs to be adequately evaluated. Common approaches for validation of simulation models use empirical validation, where model predictions are directly compared against empirical validation data. However, conducting an adequate empirical validation for those simulation models that are used in PRA to help address emergent safety concerns at existing NPPs and to support the development of advanced nuclear reactors would be challenging, if not impossible, due to (i) a lack of validation data at the model output level; (ii) available data at lower levels (e.g., model inputs) being subject to various sources of uncertainty; and (iii) available data not being fully applicable for the specific context and conditions under which the model needs to be used. This paper reports on the theoretical and methodological development of a systematic and scientifically justifiable validation methodology, namely the Probabilistic Validation (PV) methodology, to facilitate the validity evaluation of advanced simulation models under the U.S. Nuclear Regulatory Commission (NRC) risk-informed regulatory framework. The PV methodology advances the scientific usage of epistemic uncertainty as a quantitative metric of validity for simulation predictions, where validation follows the National Research Council’s recommendation that models should be made “as useful as possible by quantifying how wrong they are.” Results of a case study to demonstrate the feasibility of the PV methodology in the context of Fire PRA are included.

Paper HA209 Preview

Author and Presentation Info

Lead Author Name: Ha Bui (habui2@illinois.edu)

Bio: Dr. Ha Bui is currently a Postdoctoral Research Associate at the Department of Nuclear, Plasma, and Radiological Engineering (NPRE), University of Illinois at Urbana-Champaign (UIUC). He received his Ph.D. from NPRE UIUC in May 2022. His research interests currently include dynamic PRA, HRA, coupling between physics and human performance, emergency response modeling, uncertainty quantification, and advanced simulation validation.

Country: United States of America
Company: University of Illinois at Urbana-Champaign
Job Title: Postdoctoral Research Associate