A Monte Carlo analysis of the effect of heat release rate uncertainty on available safe egress time.
作者：Kong, D., N. Johansson, P. van Hees, S. Lu, and S. Lo.
關鍵字：Time-squared fire growth, peak heat release rate, heat release rate uncertainty, Latin hypercube sampling, probability-based available safe egress time
論文來源：期刊
具體來源：Journal of Fire Protection Engineering no. 23 (1):5-29.
發表時間：2012年
Available safe egress time is an important criterion to determine occupant safety in performance-based fire protection design of buildings. There are many factors affecting the calculation of available safe egress time, such as heat release rate, smoke toxicity and the geometry of the building. Heat release rate is the most critical factor. Due to the variation of fuel layout, initial ignition location and many other factors, significant uncertainties are associated with heat release rate. Traditionally, fire safety engineers prefer to ignore these uncertainties, and a fixed value of heat release rate is assigned based on experience. This makes the available safe egress time results subjective. To quantify the effect of uncertainties in heat release rate on available safe egress time, a Monte Carlo simulation approach is implemented for a case study of a single hypothetical fire compartment in a commercial building. First, the effect of deterministic peak heat release rate and fire growth rate on the predicted available safe egress time is studied. Then, the effect of uncertainties in peak heat release rate and fire growth rate are analyzed separately. Normal and log-normal distributions are employed to characterize peak heat release rate and fire growth rate, respectively. Finally, the effect of uncertainties in both peak heat release rate and fire growth rate on available safe egress time are analyzed. Illustrations are also provided on how to utilize probabilistic functions, such as the cumulative density function and complementary cumulative distribution function, to help fire safety engineers develop proper design fires.