IMPLICATIONS OF PARAMETER CHANGES FOR FDS MODELING OF A TYPICAL NUCLEAR POWER PLANT ELECTRICAL SWITCHGEAR ROOM

摘要

As nuclear power plant fire probabilistic safety assessments (fire PSAs) become increasingly more detailed and exhaustive, especially in risk critical and often heavily fire source or component and perhaps fuel source laden plant areas, so does the reliance on more state-of-the-art modeling tools such as Fire Dynamics Simulator (FDS) over more simplistic tools such as first-order calculations. However, the increased complexity of such tools can quickly obscure the validity of an analyst's work, and care must be taken to understand the implications of chosen modeling parameters and their impact on the postulated results. Any FDS parameter thought to bias the results of the model simulation would lead to biasing of the fire PSA results for the particular scenario(s) and should be carefully considered as changes in the scenario's predicted results and consequential impacts to the surrounding room area can directly impact the fire PSA quantification results by altering the scenario's contribution to the overall core damage frequency. This paper will explore the consequence of various FDS parameter changes on a detailed model of a typical nuclear power plant 4 kV electrical switchgear room, including material property selection, obstruction inclusion or exclusion, computational grid size selection, and burner surface area selection. The primary focus of the investigation is the comparison of predicted impacts to the surrounding environment within the modeled room, the fire scenario's zone of influence (ZOI). Changes to the predicted ZOI or the predicted impact levels for such factors as temperature, heat flux, or flame impingement as well as predicted fire detection timing are explored related to changes in the modeling parameters selected during FDS modeling, which directly relate to what PSA-related or modeled components or cables an analyst may infer as being impacted or failed during a fire scenario. Comparisons of the FDS predicted results to other fire modeling tools such as first-order calculations and zone models will also be presented.