The impact of default settings on evacuation model results: a study of visibility conditions vs occupant walking speeds

This paper aims to analyze the impact of different strategies regarding the use of default model settings and embedded data-sets. Initially, the consequences of these different strategies on the results produced are discussed. This is followed by a detailed case-study focusing on the qualitative and quantitative impact that selecting different strategies may produce: whether the user provides their own data, uses a pre-defined library, or the default setting. The case-study focuses upon the relationship between the smoke level simulated and the occupant’s walking speed. Currently, the relationship between smoke and visibility (and subsequently walking speed) is typically based on two different sets of experimental data: Jin and Frantzich/Nilsson. The two data-sets present different experimental conditions (i.e. gas irritancy, population characteristics, structural configuration, etc.), but they are often applied as if equivalent. Different evacuation models make different assumptions regarding evacuee performance in smoke, and employ different data-sets. To test the impact of this representation within evacuation tools, the authors have employed three evacuation models: 1) a model that by default uses the Jin data-set, buildingEXODUS 2) a model that by default uses the Frantzich/Nilsson’s data-set, FDS+Evac and 3) a model that allows the data used to be modified, Gridflow. The case-study shows that 1) results appear to be consistent among models if they use the same data-sets 2) the same model can provide different results if applying different data-sets for configuring the inputs 3) models using embedded data-sets need user expertise and experience to configure the model and then to evaluate the results produced.