Fire Safety Science Digital Archive

Performance-based life-safety design depends on a comparison between the time required for escape (Required Safe Escape Time – RSET) and the time to loss of tenability (Available Safe Escape Time - ASET). Both include a number of stages, involving a variety of processes and requiring a range of input data. A problem for the design engineer is that while all stages need to be addressed to obtain a realistic outcome for the analysis, some aspects are reasonably well understood and quantified, while others are often oversimplified or ignored. For the RSET time line, most emphasis is usually placed upon the travel time component, representing the physical movement of occupants into and through the escape routes. However, the time required for occupants to engage in a range of behaviours before the travel phase (pre-movement time), often represents a greater component of the total escape time. Pre-movement time distributions are dependent upon key features such as occupancy type, warnings, occupant characteristics, building complexity and fire safety management strategy. It is proposed that a practical solution for the engineer is to apply pre-movement time distributions measured from monitored evacuations, fire incidents, or derived using behavioral models, and specified in terms of a number of “design behavioural scenarios” analogous to “design fire scenarios”, classified according to the key features listed. A problem with the evaluation of travel time, is that most calculation methods assume no interaction between the occupants and the fire effluent. If occupants are exposed to irritant smoke, then movement speeds are likely to be reduced. A calculation method is proposed, relating predicted travel speed to smoke and irritant concentrations. The ASET time line ends when occupant incapacitation is predicted from exposure to fire effluent. This depends upon the time-concentration curves for the main toxic fire effluents, requiring inputs on smoke and toxic product yields under different fire conditions. Existing engineering calculations use only smoke density and/or carbon monoxide, with yields often treated as constants, usually for the well-ventilated fire case. A method is proposed, whereby yield data for major toxic effluent species can be obtained over a range of fire conditions, expressed in relation to the global equivalence ratio. Results are illustrated for carbon monoxide and hydrogen cyanide.