FPE Extra Issue 16, April 2017

Variables to Consider When Conducting a Timed Egress Analysis

By: Majed Almejmaj

Timed egress analysis is a useful tool for fire engineers when trying to demonstrate that occupants are not subjected to untenable conditions prior to evacuating a space. This is extremely useful when architects are designing iconic buildings that will cater to large crowds (e.g. high-rise buildings and shopping malls) with open spaces (i.e. atriums). When conducting an egress analysis, engineers consider a number of variables, such as occupant load and number of exits, to determine the required safe egress time (RSET); however, researchers have been highlighting the importance of considering other variables that can have a significant effect on the analysis results such as exit choice, level of fire education and culture. In addition, engineers currently do not have a clear guidance that outlines the capability and limitation of each modelling tool, when to use them and input variables that should be used.

Minimum Variables

The variables needed when conducting an egress analysis can greatly vary depending on the method used for the analysis. For example, movement models are only concerned with aspects related to occupant movement, such as walking speed over horizontal walkways, stairs and flow through elements of the egress systems. On the other hand, behaviour models take into account movement and behaviours associated with pre-movement, and possibly movement through smoke. In almost all evacuation models, the basic variables needed are the number of occupants in the space, exit components (i.e. stairs, ramps, doors) and occupant walking speed. Engineers should be aware of the capabilities of the software they intend on using for the analysis to minimize the risk of errors in the results and hindering approval at a later stage in the project.

Pre-Movement Time

Studies have shown that the time it takes people to begin moving towards an exit is an essential part of an egress analysis. This time can be affected by a number of elements, such as the type of alarm system installed, type of occupancy and culture.

Type of Alarm System

Most buildings are required by building codes to be provided with a fire alarm system (depending on the occupancy type) to alert occupants of an emergency incident. The means of notification can either be through a distinct audio tone or a voice message that inform occupants of the situation and provide instructions on what to do next. Studies have shown that people are more likely to respond faster when a voice alarm communication system or staff intervention are used to notify occupants in department stores and restaurants.1 This is due to the fact that occupants are provided with additional information along with staff guidance to the nearest exit

The Tempo.ral-three (T-3) fire alarm signal has been required for alarm systems by NFPA 72 since 1996.2 It consists of three “on” phase pulses lasting 0.5 seconds with 0.5 seconds of “off” phase between each pulse. This is followed by an 1.5 seconds “off” phase before the cycle is repeated.3 Since 1996, there have been few studies that gauge people’s familiarity with the T-3. In a 2001 study covering people’s ability to recognize the Temporal Three (T-3) fire alarm signal, the results indicate that only 6 percent of participants were able to recognize the signal, even though it was required in that country since 1996.

While the use of a tone only alarm signal provides an easy way of meeting audibility standard requirements, it does not necessarily mean that occupants will be familiar with the tone or understand what it represents which can cause difficulty when determining pre-movement time.

Occupancy Type

During the design stage, the occupancy type has a significant effect not only on code requirements but also on the type of evacuation scenario that should be considered as people might not be familiar with the building’s layout or have limited alertness.

The architectural complexity and the visibility of emergency exits varies with its type. For example, shopping malls tend to focus on having open spaces for various seasonal events, leasable kiosks, and interchangeable store layouts. However, if we were to ask the average shopper on their level of familiarity with an exit, the majority will point to the entrance they usually use to access the mall. When considering how people evacuate a building, occupants tend to head towards the exit they are familiar with, such as the main entrance.4 This can be observed in various fire incidents such as the Station Nightclub fire, where most occupants moved towards the main entrance.5 Studies have shown that occupants who are familiar with the location of the means of egress are more likely to move to an exit even when there is smoke.6-8 Exit familiarity should be considered by engineers when conducting an egress analysis especially in public buildings (e.g. shopping malls and hotels), where occupants are more likely to evacuate through the main entrance.

The level of occupant alertness is an important factor when developing the evacuation scenario used in egress analysis. For example, it is expected that hotel occupants will respond faster to a fire alarm during the day when they are awake than an alarm occurring at 2 am when they are more likely to be asleep or possibly intoxicated as it will take them more time to get ready. The weather conditions can also play a major factor where studies have shown that pre-movement time during winter is significantly higher than in the summer.4, 5, 9


In recent years, researchers have been examining the effect of culture on occupants’ pre-movement and movement times. In a study examining the effect of traditional clothing on walking speeds in Saudi Arabia, the results show that females tend to walk slower (0.904 m/s) than males (1.050 m/s),10 which is significantly less than adult walking speeds referenced in the SFPE Handbook.5, 11 Results from another study also indicates that Saudi Arabians are less likely to recognize the T-3 fire alarm signal when compared to Americans12 which may be attributed to the fact that 89.50 percent of the U.S. participants reported having some experience with fire drills compared to 35.4 percent of Saudi participants13. While it is a challenging task, engineers should always try to obtain input variables which are valid for the region where the project is going to be built.


The following outlines a number of challenges associated with conducting an egress analysis:

Clear Guidance

To ensure that egress modelling tools are being used appropriately, fire engineers need a clear guidance that addresses the capabilities of each model, limitations and applicable scenarios for each model. The 2003 SFPE “Engineering Guide: Human Behaviour in Fire” provides a general discussion about the available modelling tools and data. It is currently being reviewed and the draft will be available in the upcoming months for public comments and it promises to address this challenge.

Scarcity of Data

This is one of the major problems that fire engineers face when conducting an egress analysis. There are two issues associated with data scarcity, the first issue deals with the availability of recent data that covers pre-movement, walking speeds and flow through components of the means of egress as the majority of available data was collected during the 1960s and 1970s. This brings into question the validity of such data considering that there have been demographical changes not to mention increase in obesity and the sizes of the general population. The second issue touches on the availability of such data in a centralized location where engineers can readily access the needed input variables, highlighted since 2001.14

Closed Software Packages

With the increase use of evacuation software, more and more companies are developing new software packages to capitalise on the growing market. While this seems beneficial for users as it provides more modelling options, it does create a challenge with the developers’ willingness to divulge details about their intellectual property. This creates challenges associated with researchers’ ability to validate the software (e.g. modelled behaviour).

Majed Almejmaj is with BuroHappold Engineering


[1] Frantzich, H. Occupant behaviour and response time. in 2nd International Symposium on Human Behaviour in Fire. 2001. Interscience Communications.

 [2] Proulx, G., et al., Fire Alarm Signal Recognition. 2001, National Research Council Canada: Canada.

 [3] NFPA, NFPA 72 National Fire Alarm Code, in Distinctive Evacuation Signal. 2016, National Fire Protection Association: Quincy, MA. p. 72-273.

 [4] Proulx, G., Evacuation time and movement in apartment buildings. Fire Safety Journal, 1995. 24(3): p. 229-246.

 [5] Proulx, G., Evacuation Time, in The SFPE Handbook of Fire Protection Engineers. 2008, National Fire Protection Association: Bethesda, MD. p. 3-355 to 3-370.

 [6] Wood, P., Fire research note 953. Building Research Establishment, Borehamwood, 1972.

 [7] Bryan, J.L., Behavioral Response to Fire and Smoke, in The SFPE Handbook of Fire Protection Engineering. 2008, National Fire Protection Association: Bethesda. p. 3-321 to 3-352.


 [9] Proulx, G. and J. Pineau. Differences in the evacuation behaviour of office and apartment building occupants. in Proceedings of the Human Factors and Ergonomics Society Annual Meeting. 1996. SAGE Publications.

 [10] Almejmaj, M., B. Meacham, and J. Skorinko, The effects of cultural differences between the west and Saudi Arabia on emergency evacuation—clothing effects on walking speed. Fire And Materials, 2014.

 [11] Gwynne, S.M. and E.R. Rosenbaum, Employing the Hydraulic Model in Assessing Emergency Movement, in The SFPE Handbook of Fire Protection Engineering. 2008, National Fire Protection Association: Bethesda. p. 3-373 to -3-396.

 [12] Almejmaj, M., J. Skorinko, and B. Meacham. The Effects of Cultural Differences Between the US and Saudi Arabia on Emergency Evacuation - Analysis of Fire Alarm and Perception and Training in Human Behaviour in Fire Symposium. 2015. Cambridge, UK: Interscience Communications.

 [13] Almejmaj, M., J.L.M. Skorinko, and B.J. Meacham, The effects of cultural differences between the us and saudi arabia on emergency evacuation—Analysis of self reported recognition/reaction times and cognitive state. Case Studies in Fire Safety, 2016.

 [14] Proulx, G. and R.F. Fahy. Towards Creating a Database on Delay Times to Start Evacuation and Walking Speeds for use in Evacuation Modeling. in 2nd International Symposium on Human Behaviour in Fire. 2001. Boston, MA.

© SFPE® | All Rights Reserved
Privacy Policy