Design Occupant Groups Concept Applied To Fire Safety Engineering Human Behaviour Studies

Abstract

In a similar manner to other fire safety engineering guidelines around the world, the Australian Fire Engineering Guidelines identifies ASET/RSET timeline analysis as one of the methods to be used in assessments. RSET, the occupant related component of this method, can be extremely complicated due to the number of occupants within a building, variations in their characteristics and the unpredictibility of their behaviour. In order to simplify the assessment of occupants, a "design occupant groups" concept is introduced in the, soon to be published, new edition of the Australian Fire Safety Engineering Guidelines. By concentrating only on the occupants that are likely to have a major impact on various phases of the evacuation process design occupant groups approach aims to limit uncertainties and the scope for error. When combined with a quasi-probabilistic variation of the traditional ASET/RSET assessment, the design occupant groups approach has the potential to generate more reliable outcomes with less inherent risk (or conservatism).

Treatment of evacuation time uncertainty using polynomial chaos expansion

“…Bensilum et al [14] revealed that for most cases, the uncertainty of pre-movement time would be appropriately characterized by using certain uni-modal, positively skewed distributions, such as normal, log-normal and Weibull distributions. Due to the inconsistency of the available data on pre-movement time [15], the distribution type and numeric range of pre-movement time have not been clear for life safety evaluation of buildings. Furthermore, PD 7974-6 [16] took into account two premovement times, i.e.…”

The Effect of Pre-movement Time and Occupant Density on Evacuation Time

“…Normal, lognormal, and Weibull distributions can be supported. In general, the existing knowledge does not provide us the pre-movement time probability distributions for different occupancies [13]. However, pre-movement time that follows probability density distribution has been widely accepted.…”