Verification and Validation--How to Determine the Accuracy of Fire Models.
Mark H. Salley, Jason Dreisbach, Kendra Hill, Bijan Najafi, Francisco Joglar, Anthony Hamins, Kevin B. McGrattan, Richard Peacock, Robert Kassawara
Both domestically and worldwide, there is a movement to introduce risk-informed and performance-based analyses into fire protection engineering practice. One key tool needed to support risk-informed, performance-based (RI/PB) fire protection is fire models that can predict the consequences of fires. To provide the regulator with confidence in the calculation results, NFPA calls for fire models to be verified and validated. The NRC s Office of Nuclear Regulatory Research (RES), along with the Electric Power Research Institute (EPRI) and the National Institutes of Standards and Technology (NIST), has conducted a first-of-its-kind verification and validation (V&V) study of fire modeling tools used for NPP applications to support the implementation of NFPA rules as an alternative within the NRC s regulatory framework. This V&V study began in 2003 and has resulted in a seven-volume report that is available to the public on NRC s website. The study set out to verify and validate five commonly used fire modeling tools relied on by the nuclear industry. The current study used experimental uncertainty as a metric to quantify the level of agreement between model results and measurements. Experimental uncertainty was considered in terms of the combination of the model input uncertainty and the measurement uncertainty. The method developed here provides a quantitative and rigorous approach that emphasizes the importance of experimental quality and measurement accuracy in the evaluation of fire models.
Fire Protection Engineering Journal
compartment fire, fire experiments, model validation, model verification
, Dreisbach, J.
, Hill, K.
, Najafi, B.
, Joglar, F.
, Hamins, A.
, McGrattan, K.
, Peacock, R.
and Kassawara, R.
Verification and Validation--How to Determine the Accuracy of Fire Models., Fire Protection Engineering Journal
(Accessed December 3, 2023)