93 Journal of Asian Architecture and Building Engineering/May 2013/100
An Experimental Study on Evacuation Times in a Subway Station Using Evacuation
Parameters Seong-Hwan Yoon 1 , Min-Jung Lee 2 and Jurng-Jae Yee* 3 1 Professor, Department of Architecture, Pusan National University, Korea 2 Graduate Student, Department of Geography & Sustainable Development, University of St. Andrews, UK 3 Professor, Department of Architectural Engineering, Dong-A University, Korea Abstract In this study, 292 participants took part in the evacuation experiment and completed a questionnaire. The study was conducted in a large subway station in order to analyze evacuation times during a fre, and used different evacuation parameters. The subway station where the experiment was conducted consisted of two basement foors. The evacuation started as soon as the train, in which a fre had occurred, came into the subway station and the doors were opened. The entire process of the evacuation experiment was recorded by using video cameras installed in the subway station. After the evacuation experiment was completed, participants were asked to answer questionnaires on evacuation parameters such as; gender, whether they had companions, how well they know that particular station, their usual level of awareness of evacuation, and their level of evacuation knowledge. By analysing the survey results, the time of egress from the train, and the total evacuation time, the results were as follows. 1) Female participants were slower by 23% on average compared to male participants. While male participants evacuated according to an order of egress from the train, female participants were affected by many other factors during evacuation. 2) Those who had companions were slower in evacuation by 29% on average compared to solo evacuees. 3) Knowledge of the area helped both female and male participants during the evacuation. The delay in evacuation according to a lack of knowledge of the station was only 9% in the case of male participants, but 21% in the case of female participants. 4) Those who did not usually think about evacuation were slower to evacuate by 8% on average than the other group who had a good level of awareness. 5) Those who did not have any knowledge of evacuation were slower by 18% on average than the group who had a high level of evacuation knowledge. Keywords: evacuation experiment; subway; fre; evacuation time; evacuation parameter 1. Introduction Recently as larger scale buildings and more high- rises are constructed due to industrial development, the safety of building residents is increasingly threatened by fre. The primary goal of building residents during a fire is to evacuate to a safe place. Evacuation is determined by complex processes which consist of factors such as fire and the parameters of human behavior. Evacuation paths and exits during a fire can differ from behavior patterns including homing, herd instinct, the propensity to walk in a straight line, and preference for a familiar path. Studies of human behavior play an important role in evacuation studies, especially because of panic behavior, which is a very irrational form of behavior that may differ from behavior in normal situations and can affect the success of an evacuation. Research on evacuation behavior today is generally being carried out through a series of interviews with survivors from fres, questionnaires, evacuation drills and computer simulation programs. The building structure, occupants and risks can be set freely within computer simulation programs, but this advantage of being able to recreate the building contains a shortcoming, which is that human behavior during a fire cannot be replicated. Many researchers 1-6 have therefore used reported data from evacuation drills in *Contact Author: Jurng-Jae Yee, Professor, Department of Architectural Engineering, Dong-A University, 840 Hadan 2-dong, Saha-gu, Busan, 604-714, Korea Tel: +82-51-200-7609 Fax: +82-51-294-2256 E-mail: jjyee@dau.ac.kr ( Received April 12, 2012 ; accepted February 22, 2013 ) 94 JAABE vol.12 no.1 May 2013 Seong-Hwan Yoon order to compensate for the drawbacks of simulation programs. In Korea, many studies have been carried out using simulation programs after fires in subway stations (2003). However, the default data suggested by the program developers, and reported data from simulation experiments 1-5 are being used to input the evacuation variables. Unf or t unat el y, t he amount of r epor t ed dat a from simulation experiments is very limited, and it is necessary to continue performing evacuation experiments to secure fundamental data including evacuation time by evacuation parameters, evacuation paths, or walking speed, so that the collected data can be actively refected to develop evacuation simulation programs which take evacuation behaviors into consideration. This study aims to collect basic data on the varied behaviour of evacuees during evacuation through an evacuation experiment. Evacuation experiments were carried out in a subway station, connected to a large department store, and attempted to analyze evacuation time according to each evacuation parameter by using video recordings during the experiment as well as questionnaires. 2. Outline of Evacuation Experiment 2.1 Experiment Subjects and Place The experiment was conducted at Centum City subway station on subway line 2 in Haeundae-gu, the city of Busan. Multi-use facilities such as large exhibition and convention centers and large department stores are clustered around the station exit. The foating population of Centum City station per day in 2012 was 18,251. The department store is the worlds biggest, and the area is usually very crowded. Due to the high fow rate of people, this area is very important in terms of evacuation safety. Two hundred and ninety two university students in their 20s participated in the experiment, as shown in Table 1. Many different variables needed to be considered in selecting the experiment participants. However, due t o t he l i mi t at i on of consi deri ng numerous variables in one set of experiments, healthy young people who can be considered to have the same conditions in terms of physical handicap, were accepted. The depth of the experiment station was 16m. Platforms were located on basement 2, and each platform had 4 staircases consisting of 2 sets of staircases. Each set of staircases was connected to a gate. There were 13 exits that lead to the outside. There were 2 underground entrances to the large department store between exit 6 and exit 10. Each exit was as shown in Table 2. and the layout of the subway station is as shown in Fig.1. Table 1. Outline of Subjects Gender Male [-] 156 (53%) Female [-] 136 (47%) Height Average height [m] 169.5 Maximum height [m] 189.0 Minimum height [m] 151.0 Weight Average weight [kg] 60.1 Maximum weight [kg] 95.0 Minimum weight [kg] 40.9 Age Average age [years] 22 Maximum age [years] 27 Minimum age [years] 20 Table 2. Main Paths in the Experiment Width (m) Depth (m) Platform A,B,C,D 3.7 4.3 Waiting Room 1 7.5 5.8 Waiting Room 2 7.5 9.5 Exit 6 2.1 5.9 Exit 10,12 2.1 3.2 Exit 3,5,11,13 2.1 11.7 Fig.1. Layout of the Station
Fig. 1. Layout of the Station
Fig. 2. Recording Sections and the Number of Video Cameras Installed on Platforms in the Waiting Room
95 JAABE vol.12 no.1 May 2013 Seong-Hwan Yoon 2.2 Experimental Conditions and Evacuation Parameters Evacuation parameters set in this study were as shown in Table 3. The evacuation parameters of part i ci pant s were al so used i n creat i ng t he questionnaires. The evacuation time from the train in Table 3. was measured by analyzing video records. The experimental conditions and scenarios are as follows. 1) In order to minimize the possibility of participants acquiring knowledge about the station used in the experiment, participants were gathered two stations before the experiment's location. 2) The subway train had 6 cars in total. Fifty participants boarded each car from car 1 to car 5, and 42 participants boarded the sixth car. 3) Before the evacuation experiment 3 pairs of companions (friends, couples, and families) were arranged in each car. It was arranged so that these pairs would evacuate together. 4) The passengers assumed that the train was on fre when it arrived at the station. After the doors of the train were opened at the platform, the participants started to evacuate. 5) While the station had 13 exits in total, only 7 were used in the experiment due to high accessibility. 6) Smoke-generators were used to produce a life -like fre situation at the station. 7) A questionnaire of the subject's characteristics was conducted after all the experiments were carried out. 3. Measurement of the Evacuation Time 3.1 A Summary of the Video Recording on the Evacuation Twenty-nine video cameras were installed at the station to record all processes regarding the evacuation experiments. Before the experiment, the participants were given ID stickers to put on their heads, chests, and backs. Video cameras were installed at the places that were not visible from the CCTV in the station. Information on the installment of the video cameras is shown in Table 4., Fig.2. 3.2 Video Analysis Process The total evacuation process was observed through video cameras installed in each evacuation path: subway train platform stairways on platform turnstiles station waiting room exits to ground foor. Each participant's evacuation time was calculated accordi ng t o t he process shown i n Fi g. 3. The evacuation time was set to start when the subway doors were opened, and to end when participants reached the ground foor outside the subway exits. T (i)total evacuation time =t 6(i)= T 1(i)+ T 2(i) + T3(i) +T 4(i) +T 5(i) +T 6(i) 4. Results and Discussions 4.1 Total Evacuation Time The subway train that left the experimental station arrived 46 seconds after the fire began. The starting Table 3. Evacuation Parameters in Questionnaire Parameters Contents Gender Male/Female Time of Egress from the Train 1 sec/2 sec/3 sec/More than 4 sec Affliation Solo group/Companion group Number of Times Previously Visited the Station Never/Once or twice/Frequently The Level of Evacuation Awareness before the Experiment Unaware/Average/Well-aware The Level of Evacuation Knowledge before the Experiment No knowledge/Average/ Knowledgeable Table 4. The Number of Video Cameras and the Contents of Measurement Section No. Time Measured Inside the Car (i) 12 From the point when the car doors opened to the point when the participants escaped from the car Platform (ii) 6 + 2 CCTVs From the point when each participant escaped from the car to the point when they arrived at the staircases of the platform. Platform Stairs (iii) From the point when each participant escaped from the car to the point when they arrived at a gate through the platform staircases Gate (iv) 2 CCTVs From the point when each participant arrived at the platform staircases to the point when they passed each gate Waiting Room (v) 4 From the point when each participant passed the gate to the point when they arrived at the staircases to an exit Exit Gate (vi) 7 + 4 CCTVs From the point when each participant escaped to the outside from the exit stairs Fig.2. Recording Sections and the Number of Video Cameras Installed on Platforms and in the Waiting Room
Fig. 1. Layout of the Station
Fig. 2. Recording Sections and the Number of Video Cameras Installed on Platforms in the Waiting Room
Fig. 1. Layout of the Station
Fig. 2. Recording Sections and the Number of Video Cameras Installed on Platforms in the Waiting Room
96 JAABE vol.12 no.1 May 2013 Seong-Hwan Yoon point of the evacuation time was when the subway doors opened, not when fire began, because the participants could only evacuate after the train doors opened. The average evacuation time was 89 seconds and the first evacuee who arrived at the safe point completed the evacuation in 25 seconds, while the last participant completed it in 132 seconds. It can be seen that participants evacuated more rapidly after 70 seconds as show in Fig.4. and Fig.5. This is because some of the earlier evacuees were able to run through the empty pathways, but the later participants could not run and evacuate quickly because the pathways were congested. 4.2 Evacuation Time According to Gender The frst male evacuee completed the evacuation in 25 seconds while the frst female participant completed it in 62 seconds. Fig.6. shows the distribution of evacuation time according to gender. Forty percent of the male participants evacuated within 70 seconds, but only 10% of the female participants did so. A comparison between the two groups shows that the average evacuation time of female participants was 23% longer than that of the male participants. The average evacuation time for the males was recorded at 73 seconds, but that of female participants was recorded at 90 seconds. These results were due to the fact that the walking speed of the males was faster than that of the females and some male participants jumped over the turnstiles instead of waiting for their turn or managed to fnd the evacuation path more rapidly than the female participants. 4.3 Evacuation Time by Speed of Egress from the Train The video data showed that the participants in the cars were standing and preparing for evacuation before the train arrived at the station. The preparation order for evacuation proved to be an important factor to determine the order for evacuation. Collect video files Distinguish videos and number by place 1 = train 2 = platform 3 = stairway on platform floor 4 = gate 5 = waiting room 6 = exit Set the Start Point (The time when train doors opened = 0 sec) n = 0 n = n+1 Open a (n) Video Data Identify participants ID number(i) Output the time when participant Passed the last point in section = tn(i) Evacuation Time in section = Tn(i) Tn(i) = tn(i) tn-1 (i) (t 0 =0) Are all the participants investigated? n 6 Finish YES NO NO YES
Fig.3. Video Data Analysis Process
Fig. 4. Number of Participants who Completed the Evacuation by Time Flow
Fig.3. Video Data Analysis Process Collect video files Distinguish videos and number by place 1 = train 2 = platform 3 = stairway on platform floor 4 = gate 5 = waiting room 6 = exit Set the Start Point (The time when train doors opened = 0 sec) n = 0 n = n+1 Open a (n) Video Data Identify participants ID number(i) Output the time when participant Passed the last point in section = tn(i) Evacuation Time in section = Tn(i) Tn(i) = tn(i) tn-1 (i) (t 0 =0) Are all the participants investigated? n 6 Finish YES NO NO YES
Fig.3. Video Data Analysis Process
Fig. 4. Number of Participants who Completed the Evacuation by Time Flow
Fig.4. Number of Participants Who Completed the Evacuation by Time Flow Fig.5. Evacuation Time Distribution
Fig. 5. Evacuation Time Distribution
Fig. 6. Comparison of Evacuation Time by Gender
Fig. 5. Evacuation Time Distribution
Fig. 6. Comparison of Evacuation Time by Gender
Fig.6. Comparison of Evacuation Time by Gender Gender Min. Avg. Max. Male 25 sec. 73 sec. 128 sec. Female 62 sec. 90 sec. 115 sec. Table 5. Average Evacuation Time by Gender 97 JAABE vol.12 no.1 May 2013 Seong-Hwan Yoon The evacuation time appeared to differ by gender. Twenty-four percent of male participants evacuated from the train within 1 second while 78% did so within 3 seconds (Fig.7.). The number of female participants who evacuated from the train within 1 second was only 14% while the number who did so within 3 seconds was only 63%. Overall, the male participants evacuated from the train more rapidly than the female participants. Fig.8. shows the average evacuation times of male and female participants by the speed of egress from the train. In the case of male participants, the average evacuation time increased continuously: The total (ending) evacuation times for males were 36 seconds, 59 seconds, 87 seconds, and 97 seconds. On the other hand, the evacuation times of female participants did not increase continuously: The evacuation times for females were 80 seconds, 89 seconds, 96 seconds, and 87 seconds. There was a big gap in evacuation times between males and females; male participants who evacuated from t he t rai n wi t hi n 1 second compl et ed t he evacuation in 36 seconds, while female participants who had the same train evacuation time completed the evacuation in 80 seconds. In addition, comparing the male and female evacuation times from the same train shows that of those who evacuated within 3 seconds, the female participants were later than the male participants, although they started to evacuate at the same time. This indicates that female participants were more influenced by other factors in the evacuation process than male participants. 4.4 Evacuation Time with Regard to Companions Companions include family members, friends, or partners. In a study by J.D. Sime, evacuees showed a tendency to evacuate with their companions, and took longer to react to the emergency alarm than solo evacuees 7 . This study attempted to set some artifcial companions to investigate companion behavior during a fire and to compare their evacuation times with those of solo evacuees. It was also found that the pairs who were real companions held their partners' hands while evacuating during the experiment. Therefore, this study further analyzed the two groups - one group that evacuated with a companion and another group which evacuated alone, and presented the analysis of evacuation times according to companion behavior. The sol o gr oup had a moder at el y gr owi ng distribution in the initial evacuation from the train, but over 60% of the companion group evacuated the train after 4 seconds and their average emergency evacuation time from the train was longer than that of the solo group (Fig.9.). The first evacuee in the solo group who evacuated from the train within 1 second completed the evacuation in 25 seconds, while the first evacuee in the companion group evacuated in 59 seconds. The gap of the average evacuation time between the solo and companion groups was 43 seconds, while the gap of the initial evacuation time from the train between the groups was 1 second: The average evacuation time of the solo group was 60 seconds and that of the companion group 103 seconds. Therefore, the evacuation time of the companion group was 29% longer than that of the solo group as show in Fig.10. A reason for this may be that the companion groups attempted to avoid congestion to evacuate safely with their companion. This made them fall behind in competing with the solo group occupying the evacuation paths. Fig.11. shows the results of the emotions that companion groups felt with their companions. From these results, it can be seen that many of the evacuees with companions felt psychological stability. This is in line with J. Bran's research 8 which concluded that while paired evacuees helped each other and felt relief from each other, their actual evacuation time may be delayed or take longer. 4.5 Evacuation Time by Familiarity A typical human behavior of evacuees in fires is that they have a tendency to use familiar paths. This familiarity provides the subjects with useful i nf or mat i on about t he ar ea and t hey use t hi s
Fig. 7. Distribution of Evacuation Time by Time of Egress from the Train by Gender
1sec 2sec 3sec More than 4sec +S.D. Male 9.9 21.8 26.4 23.3 Female 24.6 15.7 14.7 17.4 * p-value = .000
Fig. 8. Average Evacuation Time by Time of Egress from the Train
Fig. 9. Distribution of Time of Egress from the Train by Companion Behaviors
(Max) (Avg) (Min) Fig.7. Distribution of Evacuation Time by Time of Egress from the Train by Gender Fig.8. Average Evacuation Time by Time of Egress from the Train
Fig. 7. Distribution of Evacuation Time by Time of Egress from the Train by Gender
1sec 2sec 3sec More than 4sec +S.D. Male 9.9 21.8 26.4 23.3 Female 24.6 15.7 14.7 17.4 * p-value = .000
Fig. 8. Average Evacuation Time by Time of Egress from the Train
Fig. 9. Distribution of Time of Egress from the Train by Companion Behaviors
(Max) (Avg) (Min) 1 sec. 2 sec. 3 sec. More than 4 sec. +S.D. Male 9.9 21.8 26.4 23.3 Female 24. 15.7 14.7 17.4 * p-value = .000 98 JAABE vol.12 no.1 May 2013 Seong-Hwan Yoon information strategically to evacuate efficiently 9 . Knowledge of the building space is reflected by previous visits and experiences in the buildings. This type of information of the area can be useful in fnding a future evacuation route, if needed. Therefore, this study made an assumption that the familiarity of the station increases in accordance with the number of times an evacuee had visited the station before, and conducted analysis accordingly. Survey data regarding the previous visitations of the participants in the experiment showed that most people had visited the subway station once or twice before. A smaller number of the participants had visited frequently. Fig.12. shows the average evacuation time based on the participants' experience of visiting the station. In the case of the male participants, the fastest to successfully evacuate were the group who had visited the area frequently, they fnished the evacuation in an average of 67 seconds. The two other groups, those with no visiting experience and those who had visited the area once or twice evacuated in an average of 73 seconds and 68 seconds respectively. However, the result showed that the group who had visited the area once or twice took on average 9% longer than those who had never visited the station before. In the case of the female participants, the group who had visited the area frequently was the quickest: They fnished the evacuation in an average of 77 seconds. People with no visiting experience took the longest time; they took on average 97 seconds, whilst the people who had visited the area once or twice finished the evacuation in an average of 90 seconds. The result shows that the group who had never visited the area before took on average 21% longer than those who had visited once or twice before. Overall, the familiarity of the area turned out to be useful in the evacuation, as the participants who had frequently visited the station ended up evacuating quicker than those without previous experience in the station. In addition, it was found that the female participants were more strongly affected by knowledge of the subway station than the male participants. 4.6 Evacuation Time based on Levels of Awareness concerning Emergency Evacuation Accordi ng t o t he survey resul t s, 38% of t he participants answered that they ignored fre alarms in public places. Also, 40% answered the question: Do you seek evacuation methods for cases of emergency? saying that they usually do not seek any evacuation methods. Only 14% answered that they seek a way to evacuate in an emergency situation. Based on the survey results, an analysis of evacuation time was Fig.9. Distribution of Time of Egress from the Train by Companion Behaviors
Fig. 7. Distribution of Evacuation Time by Time of Egress from the Train by Gender
1sec 2sec 3sec More than 4sec +S.D. Male 9.9 21.8 26.4 23.3 Female 24.6 15.7 14.7 17.4 * p-value = .000
Fig. 8. Average Evacuation Time by Time of Egress from the Train
Fig. 9. Distribution of Time of Egress from the Train by Companion Behaviors
(Max) (Avg) (Min) Fig.10. Comparison of the Evacuation Time by Companion Behavior
Fig. 10. Comparison of the Evacuation Time by Companion Behavior
Fig. 11. Emotions Felt by the Existence of Companions within the Companion Groups
First time Once or twice Frequently +S.D. Male 32.2 32.1 28.6 Female 20.2 16.5 19.6 * p-value = .000
Fig. 12. Average Evacuation Time by Participants Experience of Visiting the Station
68 73 67 (Max) (Avg) (Min) Fig.11. Emotions Felt by the Existence of Companions within the Companion Groups
Fig. 10. Comparison of the Evacuation Time by Companion Behavior
Fig. 11. Emotions Felt by the Existence of Companions within the Companion Groups
First time Once or twice Frequently +S.D. Male 32.2 32.1 28.6 Female 20.2 16.5 19.6 * p-value = .000
Fig. 12. Average Evacuation Time by Participants Experience of Visiting the Station
68 73 67 (Max) (Avg) (Min) First time Once or twice Frequently +S.D. Male 32.2 32.1 28.6 Female 20.2 16.5 19.6 * p-value = .000
Fig. 10. Comparison of the Evacuation Time by Companion Behavior
Fig. 11. Emotions Felt by the Existence of Companions within the Companion Groups
First time Once or twice Frequently +S.D. Male 32.2 32.1 28.6 Female 20.2 16.5 19.6 * p-value = .000
Fig. 12. Average Evacuation Time by Participants Experience of Visiting the Station
68 73 67 (Max) (Avg) (Min) Fig.12. Average Evacuation Time by Participants' Experience of Visiting the Station 99 JAABE vol.12 no.1 May 2013 Seong-Hwan Yoon conducted by the participants' level of evacuation awareness. Three groups: one which always checks evacuation methods, one average group, and one which does not think about evacuation in the case of emergency, were categorized. Each group's evacuation time was 85 seconds, 87 seconds, and 92 seconds on average respectively. It was shown that the group who did not think about evacuation in case of emergency, was l at er t han t he group wi t h hi gh emergency awareness by an average of 8% as show in Fig.14. 4.7 Evacuation Time by the Level of Evacuation Knowledge before the Experiment Fifty percent of participants ticked average in answering the question: Do you have knowledge of how to evacuate? While 27% answered that they did not have any knowledge. The rest (21%) answered that they knew how to evacuate. Based on the survey results, participants were categorized into three groups: a group with a high level of knowledge about evacuation, an average group, and another group with no knowledge of evacuation. The results of the evacuation time of each group were 79 seconds, 90 seconds, and 93 seconds respectively. It was shown that the group with no knowledge of evacuation was later to evacuate by an average of 18% when compared to the group with good knowledge of evacuation as shown in Fig.16. 5. Conclusion In this study, the evacuation experiment and questionnaire were carried out with 292 participants in a large subway station in order to analyze evacuation times during a fre, by different evacuation parameters. The results are as follows. 1) Evacuat i on t i me di ff er ed by gender. The evacuation time of the frst female evacuee was 62 seconds, and 40% of the male participants completed the evacuation within 70 seconds. The average evacuation time of the male participants was 73 seconds, while the average evacuation time of the female participants was 90 seconds, which was later by 23%. 2) Readiness for evacuation, standing at the car doors turned out to be an important factor in successful speedy evacuation. The evacuation time based on this factor also differed by gender. Many female evacuees took more than 4 seconds to escape the car. Due to prompt escape out of a car, many male evacuees were able to finish the evacuation early. However, in the case of females, this factor was not considered to affect the evacuation time signifcantly. Gender Never Once or Twice Frequently Male 20% 66% 14% Female 25% 63% 12% Table 6. Times Preciously Visited the Station
Fig. 13. The Level of Awareness on Emergency Evacuation
Fig. 14. Distribution of Evacuation Time by Awareness on Emergency Evacuation
Fig. 15. The Level of Evacuation Knowledge before the Experiment
Fig.13. The Level of Awareness on Emergency Evacuation
Fig. 13. The Level of Awareness on Emergency Evacuation
Fig. 14. Distribution of Evacuation Time by Awareness on Emergency Evacuation
Fig. 15. The Level of Evacuation Knowledge before the Experiment
Fig.14. Distribution of Evacuation Time by Awareness on Emergency Evacuation
Fig. 13. The Level of Awareness on Emergency Evacuation
Fig. 14. Distribution of Evacuation Time by Awareness on Emergency Evacuation
Fig. 15. The Level of Evacuation Knowledge before the Experiment
Fig.15. The Level of Evacuation Knowledge before the Experiment Fig.16. Distribution of Evacuation Time by the Level of Evacuation Knowledge before the Experiment
Fig. 16. Distribution of Evacuation Time by the Level of Evacuation Knowledge before the Experiment
100 JAABE vol.12 no.1 May 2013 Seong-Hwan Yoon 3) The average evacuation time of solo evacuees was 60 seconds, while the participants who evacuated with companions took 103 seconds on average which was slower by 29% compared to solo evacuees. This was thought to be due to the psychological impact of having companions, in that they felt more relieved and these mutual feelings affected them to try to fnd safer ways, and through this process, they were held back in occupying a beneficial space for a speedy evacuation. 4) Considering the fact that both female and male participants who had visited the subway station frequently were able to complete the evacuation quickly, knowledge about the area they were in was helpful in their evacuation. Male participants evacuated 9% faster when they had knowledge about t he area, whi l e femal e part i ci pant s evacuated 21% faster when they had knowledge about the area. 5) The average evacuat i on t i me of t he group of participants who usually think about how to escape a place was 86 seconds, while the evacuation time of the group who do not usually think about evacuation was 92 seconds on average, which was 8% later than the other group who had a good level of awareness. 6) The average evacuation time of the group of participants who had knowledge of evacuation procedures in a fire was 79 seconds, while the average evacuation time of people who did not have any knowledge was 93 seconds, which was later by 18% than the other group who had knowledge of the evacuation situation. Acknowledgements This work was supported by the Dong-A University research fund. References 1) Proulx, G. (1995) Evacuation time and movement in apartment buildings. Fire Safety Journal, 24 (3), pp.229-246. 2) Weckman, H., Lehtimaki, S. and Mannikko, S. (1999) Evacuation of a theatre: exercise vs. calculations. Fire and Materials, 23(6), pp.357-361. 3) Shields, T. J. and Boyce, K. E. (2000) A Study of evacuation from large retail store. Fire Safety Journal, 35 (1), pp.25-49. 4) Tavares, R. M., Gwynne, S. and Galea, E. R. (2006) Collection and analysis of pre-evacuation time data collected from evacuation trials conducted in library facilities in Brazil. Journal of Applied Fire Science, 15(1), pp.23-40. 5) Margrethe, K., Ira, H., Bauke, V., Jos, G. P., Nancy, O. and Karin, G. (2010) Way fnding during fre evacuation; an analysis of unannounced fire drills in a hotel at night. Building and Environment, 45(3), pp.537-548. 6) Cheng, X., Zhang, H., Xie, Q., Zhou, Y., Zhang, H. and Zhang, C. (2009) Study of announced evacuation drill from a retail store. Building and Environment, 44(5), pp.864-870. 7) Sime, J. D. (1995) Crowd Psychology and Engineering. Safety Science 21(1), pp.1-14. 8) Bryan, J. (1999) Integrating human behavior and Response issues into Fire Safety Management of Facilities. SFPE 9) Ashel, B. and Shields, T. J. (1999) Analysis and modelling of the unannounced evacuation of a large retail store. Fire and Materials 23(6), pp.333-336. 10) Shi el ds, T. J. , Boyce, K and Si l cock, G. W. H. ( 1997) Unannounced evacuation of Marks and Spencer Sprucefeld store. Report to Marks and Spencer plc. 11) Benthorn, L. and Frantzich, H. (1999) Fire alarm in a public building: how do people evaluate information and choose an evacuation exit? Fire and Materials, 23(6), pp.311-315. 12) Nilsson, D. and Johansson, A. (2009) Social influence during the initial phase of a fre evacuation and analysis of evacuation experiments in cinema theatre. Fire Safety Journal , 44(1), pp.71- 79. 13) Olsson, P. . and Regan, M. A. (2001) A comparison between actual and predicted evacuation times. Safety Science, 38(2), pp.139-145. 14) Chow, W. K. and Candy, M. Y. Ng. (2008) Waiting time in emergency evacuation of crowded public transport terminals. Safety Science, 46(5), pp.844-857. 15) Gyuyeob, J. and Wonhwa, H. (2009) Characteristic Features of the Behavior and Perception of Evacuees from the Daegu Subway Fire and Safety Measures in an Underground Fire. Journal of Asian Architecture and Building Engineering, 8 (2), pp.415-422. 16) Taeyeon, K., Byungseon, S. K. and Kwangho, K. (2007) Smoke Movement and Evacuation Time in the Arcade of a Traditional Market using Numerical Simulation. Journal of Asian Architecture and Building Engineering, 6 (2), pp.403-410.