Modelling the Impact of Emergency Exit Signs in Tunnels

Research output: Contribution to journalArticle

Abstract

This paper addresses the problem of representing the impact of different emergency exit signs during the evacuation of a tunnel when using two different evacuation models (i.e. FDS+Evac and buildingEXODUS). Both models allow the user to represent the impact of smoke upon the evacuee. The models are calibrated (1) considering the nature of the models themselves, (2) by deriving assumptions from previous experiments and literature, (3) using new data produced from experimental work performed by Lund University. The purpose of this paper is to demonstrate the activities required of the user to configure sophisticated egress tools given the scenario examined and the alternatives available in representing evacuee behaviour. Model results show that the differences in terms of emergency exit usage are affected by the degree of modelling sophistication employed and user expertise. It is demonstrated that evacuee performance may be misrepresented through indiscriminate use of default settings. Results are instead consistent between the models when their input is calibrated implicitly (given the availability of experimental data) or explicitly (employing the exit choice sub-algorithms embedded in the model). The scenarios examined are deliberately designed to be a superset of experimental trials currently being conducted about exit choice in a tunnel. The scope is to allow a blind model comparison to take place once the experiments are completed. This will be reported in a future article.

Details

Authors
Organisations
Research areas and keywords

Subject classification (UKÄ) – MANDATORY

  • Other Civil Engineering
  • Building Technologies

Keywords

  • Evacuation modelling, Emergency exit signs, Exit selection, Human, behaviour, Tunnel safety, FDS plus Evac, buildingEXODUS
Original languageEnglish
Pages (from-to)961-988
JournalFire Technology
Volume48
Issue number4
Publication statusPublished - 2012
Publication categoryResearch
Peer-reviewedYes