Int. Journal for Housing Science, Vol.33, No.4 pp. 249-258, 2009 Published in the United States A SURVEY OF OCCUPANT RESPONSE TOWARDS AN AUDIBLE FIRE ALARM L. T. Wong, N. K. Fong and W. H. Sui Department of Building Services Engineering The Hong Kong Polytechnic University Hong Kong, China ABSTRACT It is important to understand the occupant behavioural response in fire emergency in high-rise buildings so that proper decisions for a fire safety management plan could be made. In this study, behavioural responses of building occupants to an audible fire alarm cue in high-rise buildings in Hong Kong were investigated by a survey of 327 building occupants. They were divided into two groups, where Group 1 had experienced a building fire before while Group 2 had not. The results showed that occupant experience of fire incidents might influence their perception of a fire alarm cue. Group 1 respondents would likely recognize an audible fire alarm cue as a true fire alarm and give immediate response, while the Group 2 respondents would treat the cue as a false alarm or a fire alarm test. Group 2 respondents would investigate the situation if only the fire alarm sustained for a few minutes. Regarding the evacuation routes, respondents tended to select a staircase for emergency evacuation. The choice of route did not depend on the floor on which the evacuation of an occupant started. Apart from the nearest staircase, some occupants would choose a familiar staircase. The results of this study would be a useful source of reference for evacuation plan development of high-rise buildings by taking human behaviour into account. Key words: Fire alarm, response, high-rise, residential buildings. 0146-6518/04/ 249-258, 2009 Copyright 2009 IAHS
250 Wong, Fong and Sui Introduction Hong Kong is a developed city on a hilly terrain with very limited amount of usable land for buildings. The high population density of 43,000 persons per km 2 in urban area dictates a great demand for high-rise buildings for accommodation. The building height of this city is trending up as evidenced in recent new developments. Many residential buildings in Hong Kong are constructed at a height of over 200 m. A great delay of total evacuation time in some fire cases in high-rise residential buildings was reported and lead to considerable vulnerability of some occupants (1-3). Apart from the time period for a fire detection system to raise an audible fire alarm, occupant response time to the alarm cue and travel time are other important time components contributing to the total evacuation time (4-10). Studies for behavioral response to a fire alarm cue and occupant reactions regarding the evacuation were reported (6-11). These studies showed that the evacuation time would be influenced by the evacuees physical and psychological states at the time of the fire, e.g. whether they are asleep or awake, dressed or undressed, and feeling safe or threatened due to the building design and the installed fire protection devices. The reaction of people to a fire would also depend on their perception of the seriousness of the fire. Upon activation of a fire alarm, some occupants would continue their works. Some would treat an audible fire alarm as a warning in order to avoid unnecessary evacuation for a false fire alarm or an alarm testing. They would wait for further confirmation of the warning from other means, e.g. notified by the neighbors and clarified with the building management personnel, and decided to evacuate. Before egress, many people tend to take some preservative actions, e.g. saving documents on a computer. It was reported that people in a group would rather wait for orders from a leader than make individual decision. Occupants choice of escape route might influence the travel time. Occupants would choose their usual route or a safe route rather than a closer route. The use of elevators was mainly related to the floor on which the occupants were, but not the age of the occupants. This study investigates the probable influences of occupant experience of fire emergency on the occupants perception and behavioral response to a fire alarm cue, and choice of escape routes in a high-rise building in a fire emergency. It is important to understand the occupant behavioral response to an audible fire alarm cue in highrise buildings so that proper decisions of evacuation plan could be made. The results of this study would be a useful source of reference for evacuation plan development of high-rise buildings by taking human behavior into account. Methodology and Materials Fire experiments in real-life situations to obtain real-time evacuation information would associate with ethical problems. In this study, recall questionnaire interview
Audible Fire Alarm 251 was conducted. A questionnaire was set for this study concerning an occupant s responses and reactions to an audible fire alarm cue and his/her choice of an evacuation route in a fire emergency (10-14). Sample questions were shown in the Appendix. Apart from personal details (Questions 1 to 5), some questions were asked to give information about his/her experience with fire emergency (Questions 6 to 9); perception of an audible fire alarm (Question 10); probable reaction time (Question 11); behavioral responses (Questions 12 and 13); familiarity with evacuation routes (Questions 14 to 16); and choice of egress route (Question 17). A total of 350 randomly selected occupants in high-rise buildings of Hong Kong were invited to join the questionnaire interview. 327 (93%) of them participated. All the 327 samples were used for analysis. The survey was completed in one week. Results and Discussion Table 1 shows that about 78% of the respondents in this survey were within the age groups 21 to 40. The total numbers of female (52%) and male (48%) participants were almost the same; their highest academic qualifications were 4%, 47% and 49% for primary, secondary and tertiary respectively. Among the 327 respondents, 185 (57%) of them had experienced building fire emergency (denoted as Group 1 respondents) and 142 (43%) had not (denoted as Group 2 respondents). As shown in Table 2, the respondents first perception of an audible fire alarm varied. Analysis of group 1 and group 2 data was conducted. No significant difference was found between the two groups in determining the source of the alarm (P > 0.95, Chisquare test). The majority (82% to 90% for 95% confidence intervals) of both groups realized that it was a fire alarm bell. It was not surprising that an audible fire alarm system, apart from exit signs, was one of the basic fire safety measures in all buildings in this city (13-15). However, it was reported that about 40% of those who had experience of fire emergency (i.e. Group 1) would treat the alarm as a real fire alarm, while only 23% of those in Group 2 would do the same. Group 1 did recognize an audible fire alarm differently from Group 2 (P < 0.001, Chi-square test). Table 1 The respondents Age Group Gender Female Male Row Total 20 or below 6 (1.8%) 11 (3.4%) 17 (5.2%) 21 to 30 77 (23.5%) 76 (23.2%) 153 (46.8%) 31 to 40 65 (19.9%) 38 (11.6%) 103 (31.5%) 41 to 50 9 (2.8%) 14 (4.3%) 23 (7.0%) 51 to 60 9 (2.8%) 10 (3.1%) 19 (5.8%) 61 or above 5 (1.5%) 7 (2.1%) 12 (3.7%) Column Total 171 (52.3%) 156 (47.7%) 327 (100%)
252 Wong, Fong and Sui Table 2 Recognition of an audible fire alarm and occupant response time First perception and response time With experience of fire emergency? Group 1 (Yes) Group 2 (No) Row total (a) Source of alarm sound It was from a fire alarm bell 160 123 283 It was not from a fire alarm bell 25 19 44 (b) Interpretation of alarm sound There was a fire 75 33 108 There was not a fire 110 109 219 (c) Time to the first response for the fire alarm cue Immediate 33 12 45 15 s 18 12 30 30 s 30 27 57 60 s 33 34 67 120 s 32 33 65 300 s 39 24 63 Table 2 also shows the occupant response time to an audible fire alarm cue. Indeed, the results indicated that, even though some occupants recognized the urgency, occupants generally did not respond immediately to an audible fire alarm. It was reported that the immediate response to a fire alarm cue between the two groups was significantly different (P 0.01, Chi-square test). The results showed that 18% of the Group 1 respondents would respond immediately and only 9% of the Group 2 respondents would do so. It was reported that an occupant recognized an audible fire alarm cue as a true fire alarm would response immediately, which might lead to an earlier start of evacuation for a fire emergency. This study showed that, among those 75 respondents (from Group 1) and 33 respondents (from Group 2), who recognized the fire alarm cue as a true fire, 19 (out of the 75) and 8 (out of the 33) would give an immediate response respectively. In contrast, only 14 out of the 110 respondents, who did not recognize the alarm in Group 1, and 4 out of the 109 respondents in Group 2 would give an immediate response. A respondent would likely give an immediate response if a real fire alarm was recognized (P 0.03, Chi-square test). It was reported that about half of the Group 2 respondents would select a response time of 1 minute (24%) or 2 minutes (23%), only 9% would select immediately. Interestingly, the most frequent answer in Group 1 was 5 minutes or above (21%).
Audible Fire Alarm 253 However, for all those who did not react immediately, no significant difference between the groups was found (P 0.5, F-test and t-test). The average response times for Group 1 and Group 2 were 123 s and 109 s respectively. Regarding the occupants perception of the seriousness of a fire, from Group 1, there were two extreme response times, i.e. τ a 300 s and τ a = 0 s, as summarized in Table 3. The response times would be associated with the occupants perception of the likelihood of the fire spreading to their location (P < 0.0001, Chi-square). The survey indicated that about 70% of the respondents did not aware any possible fire spread to their locations. For the subgroup of τ a 300 s, only 26% realized the seriousness; though 36% did in the subgroup of τ a = 0 s. The two extreme answers would be related to the level of anxiety of the respondents based on their fire experience. Regarding the probable response for a continuous audible alarm cue, the survey results summarized in Table 4 showed that only a small portion of the respondents (12%) would evacuate for such an alarm. However, the choices of investigate the situation (37%) and ask someone what has happened (32%) were commonly reported. No significant differences between the groups were found for the responses of investigate the situation, ask someone what to do, ask someone what has happened, wait for further information and evacuate (P > 0.9, Chi-square test). There was insufficient evidence to show a significant difference between the responding patterns of the two groups (P 0.25). Table 3 Occupants perception of fire spreading and response time Would the fire spread to your unit? Occupant s response time τ a (s) Probably or Very Likely Very Unlikely or Not Sure Row total 0 12 21 33 300 10 29 39 Column total 22 50 72 Among the 185 Group 1 respondents, 129 (70%) were alerted by an audible fire alarm, 24 (13%) by seeing flames or smoke outside, and 16 (9%) by a fire engine siren. Only 8 (2%) of them were notified by the neighbors. Although people might not respond well to non-voice signals, it was found that audible fire alarm would be a common means to notify the occupants in a fire emergency in this city. The survey results also indicated that once an occupant recognized a fire emergency, the most chosen sequence of the actions to be taken was: to call the firemen (40%), to start to evacuate (20%), to collect valuables (18%), to notify the neighbors (15%) and to extinguish the fire (6%).
254 Wong, Fong and Sui Table 4 Response to a continuous audible alarm Occupant s response With experience of fire emergency? Group 1 (Yes) Group 2 (No) Row total Investigate the situation 64 56 120 Ask someone what to do 18 13 31 Ask someone what has happened 58 46 104 Call the firemen 7 1 8 Wait for further information 11 12 23 Evacuate 27 13 40 Others 0 1 1 Regarding the first action taken by the respondents categorized by three age groups, i.e. <30; 31-50 and >50 for a continuous audible fire alarm, the results showed that the majority (80%) would take certain actions before starting an evacuation. The current study has insufficient evidence to indicate that immediate evacuation was agedependent (P 0.2). Table 5 Response to a continuous audible fire alarm First action to be taken except evacuation Start Age group evacuation Call firemen Notify neighbors Collect valuables Extinguish the fire; and others Row total <30 34 58 32 33 13 170 31-50 29 58 12 19 8 126 >50 3 15 6 7 0 31 Column total 66 131 50 59 21 327 Table 6 shows the results regarding the respondents choice of an escape route in case of fire emergency. Hong Kong people have been educated for years that a nearest escape staircase should be chosen for emergency evacuation in case of a fire and an elevator would be an unsafe choice. It was not surprising to find that the majority (94%) would choose a staircase in a high-rise building for fire emergency evacuation. The results showed that the association between choice and the occupant locations in a building was insignificant (P 0.4, Chi-square test). Indeed, many respondents reported unfamiliar with the staircases in the buildings as only 13% would use them every day and 24% would use sometimes. The main reasons of choosing a staircase were that it was designated for emergency use (59%) and seemed to be safer (31%). Only 5% would choose a staircase due to familiarity when compared with other means of evacuation.
Audible Fire Alarm 255 In Hong Kong, each of the high-rise building was designed to provide alternative escape routes so that more than one staircase for emergency evacuation purposes was available for occupants to choose. In the survey, 59% respondents would use the nearest staircase for emergency evacuation, and about 20% would choose the familiar one (not necessary to be the nearest one), if occupant was free to choose one for evacuation. No significant difference was found between the responding patterns of the two groups, regarding (a) the choice of escape route in case of fire (P 0.3, Chi-square test), (b) the usage frequency of staircase (P 0.3), (c) the reason of choosing staircase as the means of escape (P 0.6, Chi-square test), and (d) the reason of choosing the selected staircase from a number of staircases for evacuation (P 0.3, Chi-square test). Table 6 Choice of escape route Choice With experience of fire emergency? Group 1 (Yes) Group 2 (No) Row total (a) Choice of escape route in case of fire Staircase 172 136 308 Both staircase and elevator 11 5 16 Others 2 1 3 (b) Usage frequency of staircase Everyday 28 15 43 Sometimes 48 31 79 Rarely 86 71 157 Never 23 25 48 (c) Reason of choosing staircase Familiar 10 5 15 Designated for emergency purposes 103 89 192 Closer 10 6 16 Safer 59 41 100 Others 3 1 4 (d) Reason of choosing the selected staircase from a number of staircases Familiar 38 26 64 Closer 112 80 192 Follow others 8 12 20 Predetermined (e.g. fire drill) 24 24 48 Others 3 0 3
256 Wong, Fong and Sui Conclusion An audible fire alarm would be a common means to notify the occupants in a fire emergency in this city. The probable influences of occupant experience of fire emergency on the occupants perception and behavioral response to an audible fire alarm cue, and choice of escape routes in a high-rise building in a fire emergency were studied with interview questionnaire surveys of 327 samples. The survey results were presented into two groups where Group 1 respondents had experienced a building fire previously while Group 2 had not. The results showed significant differences of behavioral responses between the groups of respondents. The observations were summarized as follows: 1. Over 80% respondents could identify the source of an audible fire alarm cue. More Group 1 respondents (40%) would treat the alarm as a real fire in compared with Group 2 respondents (23%). 2. Experience of fire emergency might have influences on the perception of a fire alarm cue. More Group 1 respondents (18%) would give immediate response compared with Group 2 respondents (9%). For those who did not give immediate response, no significant difference on the response time was reported between 2 groups of respondents (P 0.5). An occupant recognized an audible fire alarm cue as a true fire alarm would associate with an immediate response (P 0.03). It was reported that 36% of the respondents who gave immediate response realized the seriousness of fire but only 26% of those gave response after 5 minutes did the same. The results indicated that the first perception of the Group 2 respondents would be a false alarm or a fire alarm test. They would investigate the situation if only the fire alarm sustained for a few minutes. However, more of Group 1 respondents would recognize the fire alarm and take action immediately. 3. For all those who did not evacuate immediately, there was insufficient evidence to show a significant difference of response time between the two groups. No significant differences of the response patterns to a persisted audible fire alarm cue and the choice of escape route or staircase was reported between the two groups (P 0.2). 4. In addition, respondents tended to select a staircase for emergency evacuation. The choice of route did not depend on the floor on which the evacuation of an occupant started. Apart from the nearest staircase, a familiar one would be selected by some occupants. It is important to understand the occupant behavioral response to an audible fire alarm cue in high-rise buildings so that proper decisions of evacuation plan could be made. The results of this study would be a useful source of reference for evacuation plan development of high-rise buildings by taking human behavior into account (16).
Audible Fire Alarm 257 References 1. Sekizawa, A., Ebihara, M., Notake, H., Kubota, K., Nakano, M., Ohmiya, Y. and Kaneko, H. Occupants Behavior in response to the high-rise apartments fire in Hiroshima city, Fire and Materials 23(6) pp. 297-303, 1999. 2. Chow, W. K. Numerical studies on recent large high-rise building fire, Journal of Architectural Engineering 4(2) pp. 65-74, 1998. 3. Wong, L. T. Hazard of thermal radiation from a hot smoke layer in enclosures to an evacuee, Journal of Fire Sciences 23(2) pp. 139-156, 2005. 4. Wong, L. T. and Cheung, T. F. Evaluating probable risk of evacuees in institution buildings, Safety Science 44(2) pp. 169-181, 2006. 5. Wong, L. T. and Fong, N. K. Risk analysis of escape time from buildings, Facilities 23(11/12) pp. 487-495, 2005. 6. Proulx, G. Occupant response during a residential highrise fire, Fire and Materials 23(6) pp. 317-323, 1999. 7. Bryan, J. L. Human behavior in fire: the development and maturity of a scholarly study area, Fire and Materials 23(6) pp. 249-253, 1999. 8. Groner, N. E. Intentional systems representations are useful alternatives to physical systems representations of fire related human behaviour, Safety Science 38(2) pp. 85-94, 2001. 9. Yoshida, Y. Study of evacuation behavior in the world trade center explosion, Fire Technology 32(2) pp. 174-189, 1996. 10. Gwynne, S., Galea, E. R., Parke, J. and Hickson, J. The collection and analysis of pre-evacuation times derived from evacuation trials and their application to evacuation modeling, Fire and Technology 39(2) pp. 173-195, 2003. 11. Fong, N. K., Wong, L. T., Sui, W. H. (2005) Behavioral responses to fire alarm in high-rise buildings in Hong Kong, Proceedings of the 2005 Asia-Pacific Conference on Risk Management and Safety-Challenges in Engineering Applications and Advances in Technologies, Hong Kong Association of Risk Management and Safety, Kowloon Shangri-La Hotel, Hong Kong, 1-2 Dec, pp. 184-191. 12. Yang, J., Peek-Asa, C., Allareddy, V., Zwerling, C. and Lundell, J. Perceived risk of home fire and escape plans in rural households, American Journal of Preventive Medicine 30(1) pp. 7-12, 2006. 13. Wong, L. T. and Lo, K. C. Experimental study on visibility of exit sign in buildings, Building and Environment 42(4) pp. 1836-1842, 2007. 14. Wong, L. T. and Leung, L. K. Minimum fire alarm sound pressure level for elderly care centers, Building and Environment 40(1) pp. 125-133, 2005.
258 Wong, Fong and Sui 15. Bruck, D. The who, what, where and why of waking to fire alarms: a review, Fire Safety Journal 36(7) pp. 623-639, 2001. 16. Meacham, B. J. Integrating human factors issues into engineered fire safety design, Fire and Materials 23(6) pp. 273-279, 1999.