Lessons Learned in ICFMP Project for Verification and Validation of Computer Models for Nuclear Plant Fire Safety Analysis, Deytec, Inc. USA Technical Report No. 2010-01, October 2010 - Full Report.
SKU:
$0.00
Abstract
This report presents a synthesis of the technical and programmatic “lessons learned” in the International Collaborative Fire Model Project (ICFMP) that was conducted from 1999 to 2008. A synthesis of ICFMP results has been conducted as a project of Deytec, Inc. to benefit public safety and the scientific community. The verification and validation (V&V) process in the ICFMP project was developed to examine the capabilities and limitations of fire models for nuclear plant fire safety and risk analysis, and to determine the predictive errors of the models. Although current models can reliably predict global parameters in nuclear plant compartment fires such as hot gas temperature and interface height, they are limited and need to be improved for predicting important parameters like the heat flux to cable targets. The development of a V&V process provided experience in the conduct of blind exercises; however, it was not possible to determine the true predictive errors of the models due to issues related to model input data and procedures for blind exercises. These issues could be addressed and the V&V process can be improved. The experience in the ICFMP has formed the basis of a V&V process for the evaluation of fire models for nuclear plant applications.
Report is 51 pages.
This report presents a synthesis of the technical and programmatic “lessons learned” in the International Collaborative Fire Model Project (ICFMP) that was conducted from 1999 to 2008. A synthesis of ICFMP results has been conducted as a project of Deytec, Inc. to benefit public safety and the scientific community. The verification and validation (V&V) process in the ICFMP project was developed to examine the capabilities and limitations of fire models for nuclear plant fire safety and risk analysis, and to determine the predictive errors of the models. Although current models can reliably predict global parameters in nuclear plant compartment fires such as hot gas temperature and interface height, they are limited and need to be improved for predicting important parameters like the heat flux to cable targets. The development of a V&V process provided experience in the conduct of blind exercises; however, it was not possible to determine the true predictive errors of the models due to issues related to model input data and procedures for blind exercises. These issues could be addressed and the V&V process can be improved. The experience in the ICFMP has formed the basis of a V&V process for the evaluation of fire models for nuclear plant applications.
Report is 51 pages.