Sie können Operatoren mit Ihrer Suchanfrage kombinieren, um diese noch präziser einzugrenzen. Klicken Sie auf den Suchoperator, um eine Erklärung seiner Funktionsweise anzuzeigen.
Findet Dokumente, in denen beide Begriffe in beliebiger Reihenfolge innerhalb von maximal n Worten zueinander stehen. Empfehlung: Wählen Sie zwischen 15 und 30 als maximale Wortanzahl (z.B. NEAR(hybrid, antrieb, 20)).
Findet Dokumente, in denen der Begriff in Wortvarianten vorkommt, wobei diese VOR, HINTER oder VOR und HINTER dem Suchbegriff anschließen können (z.B., leichtbau*, *leichtbau, *leichtbau*).
Der Artikel untersucht die entscheidende Rolle informeller Informationen bei der Entscheidungsfindung von Gebäudebewohnern während einer schrittweisen Evakuierung, insbesondere in Hochhäusern. Darin wird untersucht, wie visuelle Hinweise, wie Rauch außerhalb des Gebäudes zu sehen, und Social-Media-Posts das Angstniveau der Bewohner und ihre Bereitschaft, Evakuierungsanweisungen zu befolgen, beeinflussen können. Die Studie zeigt, dass informelle Informationen zu erhöhter Angst und höherer Fluchtmotivation führen können, insbesondere wenn sie offiziellen Anweisungen widersprechen. Die Forschung umfasst einen detaillierten Vergleich zwischen Brandschutzfachleuten und Laien und zeigt, wie Expertise im Bereich Brandschutz die Entscheidungsfindung in Notfällen beeinflussen kann. Die Ergebnisse unterstreichen die Bedeutung effektiver Kommunikation und die Notwendigkeit von Strategien, die die psychologischen Auswirkungen informeller Informationen auf das Evakuierungsverhalten berücksichtigen. Der Artikel diskutiert auch die Beschränkungen aktueller Evakuierungsstrategien und schlägt potenzielle Verbesserungen vor, um die Sicherheit und Einhaltung während der schrittweisen Evakuierung zu verbessern.
KI-Generiert
Diese Zusammenfassung des Fachinhalts wurde mit Hilfe von KI generiert.
Abstract
For a phased evacuation strategy, the occupants on floors apart from where a fire has broken out are expected to wait to evacuate, and sometimes security control center staff will instruct them to wait. Meanwhile, occupants may obtain information about a fire in other ways, such as by watching smoke ascending outside the building, or from photos or videos posted on social networking services (SNS) by people outside the building. This study examined the occupants’ thoughts and decisions when they received such informal fire information while waiting for phased evacuation through experiments with participants in a virtual high-rise building. Experiments involved 85 fire safety professionals and 72 laypeople. The entire process of entering the building, ascending to the conference/meeting room on the 40th floor, and waiting there was replicated in first-person or over-the-shoulder perspectives in the 2D movies. Announcements, the appearance of smoke, and SNS posts were also replicated. Approximately 70% of both professionals and laypeople followed the instructions for a phased evacuation when they acquired no informal fire information; however, the percentage dropped to approximately 50% if they acquired informal information. Among professionals, 13–52% of them answered that they would descend the staircases regardless of any staff instructions to wait. The fire safety knowledge that professionals possessed seemingly resulted in them noticing contradictions between cues and more chose an evacuation action that conflicted with the phased evacuation announcement. Fire and smoke control aimed at preventing heightened anxiety among evacuees is also necessary for a phased evacuation strategy.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
1 Introduction
1.1 Phased Evacuation Strategy vs Informal Information
A phased evacuation strategy is a conventional approach in the case of fire in a high-rise building [1‐4]. The purpose of the phased evacuation is to prioritize the evacuation of a fire floor or nearby fire floors and to prevent congestion due to the merging of evacuees in staircases while staying in areas with lower risk. This strategy can decrease the queuing time around the staircases [5, 6], subsequently avoiding excessive density of people, which also contributes to the maximum flow capacity of the staircases [7]. Occupants on floors other than those prioritized for evacuation are not instructed or ordered to evacuate by the security control center (SCC) in a phased evacuation. However, occupants may obtain information about the fire in other informal ways, such as by seeing the smoke from the fire ascending outside the building [8]. Furthermore, almost everyone has their own smartphone nowadays by which they can access social networking services (SNS) to gain information anytime and anywhere. Therefore, if occupants in a large-scale building obtain information that a fire has occurred in the building, they may try to acquire more information from the texts, photos, and videos posted on SNS by people outside the building. Past research has shown that acquiring correct information contributes to evacuees’ rational decision-making [9, 10], and that irrational behavior of occupants is rarely seen during fire evacuations [11‐13]. However, there is little knowledge regarding whether the acquisition of informal information may foment disorder and interfere with the phased evacuation strategy. This possibility is analogous to the situation after the September 11, 2001 disaster [14, 15] when occupants were questioning whether a phased evacuation strategy was still a workable and practical solution for tall buildings [8].
1.2 Literature Review
1.2.1 Physical Aspects of the Phased Evacuation
Data acquired from observation of evacuation drills and experimental research on phased evacuations have been reported [2, 3]. Kadokura et al. [3] reported the results of evacuation drills conducted in a high-rise building. In those evacuation drills, a high-density and nearly stagnant situation was generated at the merging of flows entering and descending the staircase. After the Grenfell Tower fire in 2017, the UK government studied the total evacuation strategy with a phased bottom-up and top-down above-the-fire floor strategy to evaluate the impact of congestion on mobility impairments and firefighters [16]. The evacuation behavior of phased evacuation can be predicted using evacuation simulation software [6] or a simple calculation method that can consider the merging and generation of the queues [17]. Specific evacuation simulation models for phased evacuation have also been developed [18‐20]. Therefore, if we assume certain geometrical conditions, such as the width of the stairs and doors to the staircase used for evacuation, and the number of occupants and their mobility, we can predict aspects of the evacuation, such as the evacuation time, as well as the specific flow and density of evacuees along the evacuation routes.
1.2.2 Messaging and Decision-Making Aspects of the Phased Evacuation
The trigger during a phased evacuation will usually be an alarm or an announcement broadcast over an emergency communication system. There has been extensive research on whether occupants recognize the danger and “begin to evacuate” when they hear an alarm or an emergency message [21‐26]. Kuligowski [25] summarized that to improve occupants’ response to emergency communication messages, the messages have to be improved in three phases of the emergency decision-making process: (1) message perception, (2) comprehension and belief, and (3) risk perception or personalization. Regarding message perception, research has revealed that the alarm or message has to be clear and informative for prompt evacuation starts [21‐23]. Research also shows that the decision to evacuate is affected by other people [8, 23, 26, 27]. Those results are related to comprehension and belief. Evacuation decisions will also differ depending on the location of the occupants, such as whether they are in a small or large room with limited or abundant access to environmental information [27, 28]. Those results can be classified as risk perception or personalization. However, there is little research concerning whether the occupants “can wait” when they are instructed to wait during a phased evacuation. Saito [29] noted that the information from SCC staff, i.e., (1) the fact that a fire has occurred in the building, (2) the procedure for the phased evacuation, and (3) the status of the phased evacuation, can relieve the anxiety of waiting occupants. Based on questionnaire results, Wada [30] reported that participants of an evacuation drill felt strong anxiety when they were stuck in staircases. However, those experiments only considered the impact of intentionally given information on the participants for the phased evacuation. That approach lacks perspectives on the impact of acquiring informal information, which needs to be considered to ensure the success of a phased evacuation procedure. To the best of our knowledge, the only study to examine the impact of informal information was by Barber [8], who reported questionnaire results from participants who were instructed to imagine as if they saw smoke from a fire ascending outside the building they were in. Therefore, additional scientific research is needed on how occupants feel when they hear the message to wait during a phased evacuation and what kind of action they will take in such a situation if they receive potentially conflicting informal information.
1.2.3 Use of Smartphones During an Evacuation
In today’s connected society, occupants of a building may use their smartphones to acquire information, such as checking the SNS or website, when they hear an emergency message that a fire occurred in the building. Even if information posted on SNS is true, it can be upsetting or cause anxiety; however, there is also the potential for erroneous information or fake news to be posted [31, 32]. If the occupants obtain online information that contradicts an emergency message that says to wait and the building is safe enough, they may question the truthfulness of the message. Smartphones have the potential to give occupants adequate information about a fire. For example, evacuation navigation systems on smartphones in case of a building fire that optimize information with real-time calculation using actual fire data and occupants’ locational data have been developed [33‐35]. However, the possibility of acquiring unintended and informal information is a new problem that needs to be considered.
1.3 Study Purpose
Phased evacuation is a conventional evacuation strategy, and extensive data and numerous simulation models of related evacuation behavior are available. However, for occupants to evacuate in an orderly manner as intended or assumed in an evacuation simulation model, adequate phased evacuation guidance that considers the impact of acquiring unintended and informal information is indispensable. As Groner [36] pointed out, predictions of building design performance will be inaccurate if human cognitive behaviors have been excluded from the analyses. The present study examines the impact of unintended and informal information on the occupants in the phased evacuation strategy.
Anzeige
2 Materials and Methods
2.1 Participants
Two groups of participants were recruited for the experiments. One group comprised 85 fire safety professionals with more than 3 years of experience, including researchers, engineers, and reviewers (hereinafter referred to as “professionals”). The other group comprised 72 laypeople without professional knowledge of fire safety. The professionals voluntarily participated in the experiment. Since there are substantially more male professionals than female professionals in Japan currently, the number of male professionals was larger than that of female professionals in the present study. Moreover, the professionals had advanced engineering degrees. The laypeople were recruited using a staffing service and offered compensation of 6,000 yen (approximately 40 US dollars) for three hours of participation. The laypeople were recruited without bias toward specific sex or age. However, due to limitations in the recruitment service, individuals over 60 were not included. The occupations and educational backgrounds of the laypeople were expected to vary significantly from those of the professionals. Both groups of participants were informed of the purpose, method, and safety and privacy concerns of the experiment using printed documents in a manner approved by the ethics committee of the author’s organization. The purpose of using two different study groups was to compare the impact of their knowledge on fire safety and buildings on their feelings or decisions while waiting in a building during a phased evacuation due to a fire. Table 1 shows the information about the study participants.
Table 1
Information about the study participants
Attribution
Professionals
Laypeople
Case order†
A, B
B, A
A, B
B, A
Sex
Male
Female
Male
Female
Male
Female
Male
Female
Age (years)/
Number of participants
25–34
1
2
7
4
6
6
6
6
35–44
4
4
6
0
6
6
6
6
45–54
7
8
12
3
6
6
6
6
55–64
9
0
7
0
0
0
0
0
65–80
6
0
5
0
0
0
0
0
Total number of participants
27
14
37
7
18
18
18
18
41
44
36
36
85
72
† Order of the experienced cases. e.g., “A, B” means the participants experienced Case A first, and Case B second
Cases A and B are explained in Sects. 2.4 and 2.7
2.2 Study Environment
The experiments for the professionals were conducted in the meeting rooms of their offices using equipment brought by the researcher. Between two and ten professionals at each office location participated in the experiments, which were conducted 15 times between July and December 2023. The experiments for the laypeople were conducted in a meeting room prepared by the researcher. The experiments involving the laypeople were conducted four times separately. Twelve people participated in two separate experiments conducted in January 2024, and 24 people participated in two separate experiments conducted in May 2024. All of the participants took part in the experiments only once.
2.3 Modeled Building
A virtual 43-floor high-rise building approximately 200 m high was created as a 3D computer model (Fig. 1) and placed within a virtual mock-up of the Shinjuku area in Tokyo, Japan [37]. Several high-rise buildings approximately 200 m high were replicated using a Level of Detail of 200, which incorporates approximate shapes and sizes of buildings with photorealistic façades.
2.4 Introductory Scenario and Creation of Participants’ Behavioral Movie
Participants were instructed to assume they had come to the building to attend a conference or meeting held on the 40th floor. They were also instructed to assume that they were unfamiliar with the building and did not know the detailed building information, even though it was an imaginary building. The scenery of the whole process of participants coming to the building, entering the entrance hall, boarding an elevator behind the entrance hall, ascending up to the 40th floor, walking to the conference/meeting room, then waiting at the conference/meeting room was replicated by 2D movies using Pathfinder [38] (Figs. 2 and 3). The movies were from the participants’ first-person or over-the-shoulder perspective to give them the impression that they had walked into the building on their own. However, the movies did not have any interactive, virtual reality functions.
Fig. 2
Virtual depictions of arriving at the building and going to the 40th floor, and maps of the walking and entrance routes
The participants experienced two experimental scenarios, Cases A and B. In Case A, a fire occurred while they were waiting in the conference room. In Case B, a fire occurred while they were waiting in the meeting room. In both cases, fires occurred on the 20th floor. Additional details of the settings are provided in Sects. 2.6 and 2.7.
The conference room, a large room over 1,000 m2 that occupied most of the 40th floor, was accessed through a corridor with fire-compartmented walls from the elevator lobby. The conference room faced the glass façade of the building through which the participants could see a view of the city (Fig. 4, left). The meeting room, a small room of approximately 34 m2, was located in a tenant space that could also be accessed through a corridor with fire-compartmented walls from the elevator lobby. The front part of the tenant space was a lobby facing the glass façade through which the participants could see outside. The meeting room was accessed via this lobby and an inner corridor of the tenant space. The meeting room did not have any windows, so the participants could not see outside from this room (Fig. 4, right).
Fig. 4
Walking and waiting views in Case A (left) and Case B (right)
Performing the experiment on the upper floor in an actual high-rise building, replicating an actual fire and evacuation situation in the building, would be preferable in order to give the participants a realistic feeling [39]. However, it is difficult to gather a large number of study participants in a building at a specific date and time. As an alternative, the scenes of entering the building, ascending up to the 40th floor by elevator, and views through the glass façade on the 40th floor were replicated in the movies. Other occupants were not replicated in either Case A or B in the present study. Although the 2D movies only contained a single individual, the participants were instructed to assume that they were not alone, and that there were some other attendees at the conference or other occupants in the building not attending the conference. Feelings or decisions regarding evacuation may be affected by the presence of or communication with other occupants [8, 23, 26, 27]; however, the present study excluded the impact of other occupants by not replicating them.
2.5 Instructions for Participants
Before watching the movies, the participants were informed that the building fully complied with Japan’s current fire safety standards, such as having fire-compartmented walls for corridors, sprinklers, fire spandrel on façade walls, non-combustible façade walls, and smoke vents. The explanation was added that those settings prevent the vertical fire spread on facades, as in the Grenfell Tower fire in 2017. They were also informed that most of the floors were tenant offices. The professionals were specifically instructed to assume that they were indeed fire safety professionals. Thus, they were allowed to guess the building configuration or fire and evacuation situation using their professional knowledge and skills.
2.6 Replication of Information Acquired by the Participants
The scenarios involving fire alarms and SCC staff instructions were identical in Cases A and B. The settings were intended for the participants to listen to the same announcements at different virtual locations. Figure 5 shows a timeline of the announcements. The movies started from the scene in which the occupant in the participants’ role was waiting in the conference room/meeting room for the conference/meeting to begin. At 30 s after the beginning of the movie, a prerecorded emergency broadcast stated that a fire had occurred on the 20th floor. At 107 s, an SCC staff member announced the following via the public address system in his real voice (first announcement): “This is the security control center. A fire has occurred on the 20th floor. People near the fire floor should evacuate. Those who are on floors lower than the 10th or higher than the 30th floor should await further instructions and not evacuate at this time.” The intention was to perform a phased evacuation. At 216 s, the staff member announced (second announcement): “This is the security control center. You may see smoke ascending outside this building, or a lot of photos on SNS. However, this building has a sprinkler system that can suppress a fire; additionally, numerous other fire safety measures are in place. Trying to forcefully evacuate via a staircase may cause crowd accidents. Please wait in a large space or corridor.” This means that the SCC staff thought that occupants instructed to wait may want to evacuate immediately if they saw the smoke. If they chose to evacuate immediately, the phased evacuation strategy would fail, and heavy congestion would occur around and within the staircase. Therefore, SCC staff tried to make the occupants feel safe and wait on their floors.
2.7 Replication of Smoke Visible through the Glass Façade, and SNS Posts
The fire on the 20th floor was represented with αt2 growth. The value of α was set as 0.113 MJ/m2 referring to the Evacuation Safety Verification Method in Japan [40], assuming that the 20th floor room was an office with an interior finishing of quasi-noncombustible material. Hence, the size of the fire source was approximately 50–70 MW when the participants saw the smoke outside the building or the photos on SNS. The windows on the 20th floor were assumed to be broken, with smoke pouring out of them. Therefore, the explanation of the SCC staff was not completely accurate because a fire size of 50–70 MW would be too large to be suppressed by sprinklers in a 1,000 m2 office room. The location and size of the fire source and the area of smoke spread are depicted in Fig. 6. The spread of the fire and smoke was replicated using Fire Dynamics Simulator, FDS [41], and PyroSim [42]. The generation of the smoke was valid in terms of its calculated physical aspects, such as the amount of smoke generation, smoke movement, and smoke density.
Fig. 6
Settings of the FDS model, fire source, and broken windows
In Case A, the participants were in the large conference room facing the glass façade. When an SCC staff member made the second announcement, participants saw the ascending smoke that had spilled from the 20th floor and reached the height of the 40th floor. The participants could only see the smoke and were not able to smell it or feel any heat.
In Case B, the participants were in a small meeting room without windows. When an SCC staff member made the second announcement, researchers handed the participants smartphones and instructed them to check the SNS posts, assuming the smartphones were their own. The participants saw SNS posts of smoke pouring out of the building or fire engines gathered around the building site (Fig. 7). All of the original comments in the posts were made in Japanese because all of the participants were Japanese.
Approximately half of all participants, both the professionals and laypeople, experienced Case A first and then Case B, with the others experiencing the cases in reverse order (see Table 1). After watching each movie, the participants answered questions about the feelings they would have or the actions they would take in the actual situation. The contents of the questions are presented in Table 2.
Table 2
Questions for Case A and Case B
Case A
Case B
Q-A1
What action would you choose if the fire alarm sounded, and the security center staff made an announcement to wait, before seeing the smoke ascending outside?
1. Wait here (at the conference room)
2. Move to the corridor and wait there
3. Go to the staircase because I want to evacuate from the building
Q-B1
What action would you choose if the fire alarm sounded and the security center staff made an announcement to wait, before seeing any SNS posts?
1. Wait here (at the meeting room)
2. Move to the lobby with windows and wait there
3. Move to the corridor beyond the lobby and wait there
4. Go to the staircase because I want to evacuate from the building
Q-A2
When you saw the smoke ascending outside through the window while you were waiting, were there any changes in your feelings of anxiety or relief?
1. Felt very anxious
2. Felt somewhat anxious
3. There was no change in my peace of mind or anxiety
4. Somewhat relieved
5. Very relieved
Q-B2
When you saw many photos of the smoke from this building posted on SNS, were there any changes in your feelings of anxiety or relief?
1. Felt very anxious
2. Felt somewhat anxious
3. There was no change in my peace of mind or anxiety
4. Somewhat relieved
5. Very relieved
Q-A3
Please tell us the reason for your answer to Q-A2. (free description)
Q-B3
Please tell us the reason for your answer to Q-B2. (free description)
Q-A4
While you were waiting, as instructed by the security center staff, you saw the smoke ascending outside through the window. Assuming that the glass was not about to break, it was not hot, and there was no strong odor, what action would you choose?
1. Wait here (at the conference room)
2. Move to the corridor and wait there
3. Go to the staircase because I want to evacuate from the building
Q-B4
While you were waiting, as instructed by the security center staff, you saw many photos of the smoke from this building posted on SNS. Assuming that there was no strong odor or any other sign of fire except the voice alarm and SNS posts, what action would you choose?
1. Wait here (in the meeting room)
2. Move to the lobby with windows and wait there
3. Move to the corridor beyond the lobby and wait there
4. Go to the staircase because I want to evacuate from the building
Q-A5
Suppose that the instruction to wait was made by a firefighter. “This is xxx (name of the firefighter), of xxx (name of his associated jurisdiction) fire department. We are now suppressing the fire. The windows were broken, and a huge amount of smoke was spilling out. But the fire is under control. If you evacuate in a rush, the stairs may become crowded, which could lead to a crowd accident, so please wait in a wide space.”
Wait here (at the conference room)
Move to the corridor and wait there
Go to the staircase because I want to evacuate from the building
Q-B5
Suppose that the instruction to wait was made by a firefighter. “This is xxx (name of the firefighter), of xxx (name of his associated jurisdiction) fire department. We are now suppressing the fire. The windows were broken, and a huge amount of smoke was spilling out. But the fire is under control. If you evacuate in a rush, the stairs may become crowded, which could lead to a crowd accident, so please wait in a wide space.”
1. Wait here (in the meeting room)
2. Move to the lobby with windows and wait there
3. Move to the corridor beyond the lobby and wait there
4. Go to the staircase because I want to evacuate from the building
Q-A6
Please tell us anything else that you felt, thought, or noticed. (free description)
Q-B6
Please tell us anything else that you felt, thought, or noticed. (free description)
3 Results
Figures 8 and 9 present results for Q-A1, Q-A4, and Q-A5, and Q-B1, Q-B4, and Q-B5 for the professionals and laypeople.
To test for significant differences between the results of the professionals and laypeople, the Wilcoxon rank sum test with a significance level of p = 0.05 was performed. The null hypothesis for the test was that both cases have identical effects. The p-values are presented in Table 3. A significant difference was not found in either of the cases. Even though a significant difference was not found, the p-value in Q-A2 is relatively small. This result may indicate that laypeople felt strongly anxious compared with the professionals when they saw the smoke outside the building spilling from the fire floor, or professionals reacted rather calmly than laypeople. However, a significant difference was not found between the two groups in terms of the impact of the situation, information, or instructions. To examine the differences in more detail, the professionals and laypeople were examined independently as follows.
Table 3
Results of Wilcoxon rank sum test between professionals and laypeople for each question
Q-A1
Q-A2
Q-A4
Q-A5
Q-B1
Q-B2
Q-B4
Q-B5
p-value
* p < 0.05
0.815
0.0982
0.701
0.190
0.644
0.759
0.625
0.745
To test for significant differences among answers in Case A (Q-A1, A4, and A5) and Case B (Q-B1, B4, and B5), the non-parametric Friedman’s test with a significance level of p = 0.05 was performed. The null hypothesis for the test was that all three questions in each case had identical effects. The results presented in Table 4 indicate the null hypothesis was rejected for both Case A and Case B among the professionals and laypeople.
Table 4
Results of non-parametric Friedman’s test
Case A
Case B
Professionals
Laypeople
Professionals
Laypeople
Test statistic
22.1
10.7
28.5
12.9
p-value
* p < 0.05
p < 0.01 *
p < 0.01 *
p < 0.01 *
p < 0.01 *
Multiple comparison tests using a sign test, employing the Holm–Bonferroni correction with a significance level of p = 0.05, were subsequently performed to test for significant differences between the two cases from among the three questions (Q-A1, A4, and A5, and Q-B1, B4, and B5). The null hypothesis for the test was that the two cases have identical effects. Table 5 shows the results of the sign test. For professionals, a significant difference was found in all compared questions. For laypeople, significant differences were only found between Q-A1 and Q-A4, Q-B1 and Q-B4, and Q-A1 and Q-A5. This result indicates that the professionals were more sensitive to the situation and information than the laypeople.
Table 5
Results of multiple comparison tests using a sign test
Compared cases
Professionals
Laypeople
Number of data
p (Bonferroni-Holm Correction)
* p < 0.05
Number of data
p (Bonferroni-Holm Correction)
* p < 0.05
+
–
No change
+
–
No change
Q-A1 vs Q-A4
29
5
50
< 0.01 *
20
3
48
< 0.01 *
Q-A1 vs Q-A5
18
8
58
0.0378 *
19
9
43
0.0872
Q-A4 vs Q-A5
4
19
61
< 0.01 *
11
17
43
0.173
Q-B1 vs Q-B4
33
3
48
< 0.01 *
29
3
40
< 0.01 *
Q-B1 vs Q-B5
27
10
47
< 0.01 *
26
10
36
0.0113 *
Q-B4 vs Q-B5
5
19
60
< 0.01 *
11
19
42
0.100
3.2 Case A
Among the study participants, 69% of professionals and 67% of laypeople chose to wait in the conference room before they saw the smoke ascend (Q-A1) (Fig. 8). The percentage who would wait in their current location decreased to 46% for the professionals and 42% for laypeople after seeing the smoke (Q-A4). Even though a significant difference between the professionals and laypeople was not confirmed, the percentage who chose to move to the staircase after seeing the smoke was larger for the professionals (27%) than for the laypeople (16%), while the percentages were almost the same before seeing the smoke (10% and 8%). The increase in the percentage of professionals who chose to move to the staircase after seeing the smoke ((23 – 8)/84 = 0.18) was larger than the increase in the percentage who chose to move to the corridor ((22 – 18)/84 = 0.048). Conversely, the increase in the percentage of laypeople who chose to move to the corridor ((30 – 18)/71 = 0.17) was larger than the increase in the percentage who chose to move to the staircase ((11 – 6)/71 = 0.070). Regarding the impact of the firefighter’s announcement in Q-A5, the percentages of choosing to wait (in the conference room) in Q-A5 were larger than those for the announcement by SCC staff in Q-A4, both for the professionals and the laypeople. Significant differences between Q-A4 and Q-A5 were confirmed for the professionals. In contrast, significant differences between Q-A4 and Q-A5 for the laypeople were not confirmed.
3.3 Case B
Even though significant differences were not confirmed between the professionals and laypeople (see Table 3), the results shown in Fig. 9 indicate that the percentage of people who chose to move to the lobby with the windows when they heard the announcement by SCC staff before viewing SNS, Q-B1, is larger for the professionals (67%) than the laypeople (47%). In contrast, the percentage of people who chose to wait in the meeting room or move to the corridor beyond the lobby and wait is larger for the laypeople (22% and 22%) than the professionals (13% and 16%). These results suggest that the professionals will take moderate or preparative action, whereas laypeople will take little action or more protective action when they hear the announcement by SCC staff before viewing SNS. After viewing the SNS posts, Q-B4, the percentage who chose to move to the corridor or staircase was almost the same between the professionals (45%) and laypeople (47%). However, the percentage who chose to move to the staircase was larger for the professionals (23%) than for the laypeople (13%). In the case of the announcement by firefighters, Q-B5, the percentage of professionals who chose to move to the staircase was somewhat reduced (16%) compared with Q-B4. However, the percentage was still larger than that for the laypeople of Q-B4 (13%).
3.4 Analysis Considering the Free Descriptions
The participants wrote a wide variety of comments in the free descriptions for Q-A3 and Q-B3. Most of the comments can be classified into five categories: (i) did not feel unsafe or got positive feelings by knowing the extent of the situation, (ii) felt danger, (iii) lost their sense of judgment due to a lack of information, (iv) felt a contradiction between the announcement that said the fire was under control and the smoke ascending/SNS posts, and v) other. Therefore, all of the free descriptions for Q-A3 and Q-B3 were classified into those five categories. The author solely interpreted the intent of the comments and classified them. The guidelines used for this classification are presented in Table 6, and examples of the contents and classifications are presented in Table 7.
Table 6
Guidelines for classifying the free descriptions
Categories
Guidelines
(i)
• Expressing a safe feeling
• Including words such as “anxiety,” “nervous,” or “worry” and their negation
(ii)
• Expressing fear, anxiety, or worry
• Including words such as “smoke” and “fire”
(iii)
• Expressing the lack of information or reliance on the announcement or the SCC staff member
• Including words such as “information” or “announcement”
(iv)
• Expressing the severity of the smoke or fire and expressing doubt about the severity
• Expressing the desire to evacuate immediately
(v)
• Others
Table 7
Examples of the free description classifications
Classifications
Example comments
Professionals
Laypeople
(i) Participants who did not feel unsafe, or who got positive feelings by knowing the extent of the situation
• The smoke did not enter the floor where I was. I thought the glass façade would remain intact. (Case A)
• I felt a little anxiety. However, compared with the case when I was in the meeting room where there were no windows, the anxiety was not very severe. (Case A)
• I got the information about where the fire had occurred, and I thought the fire would not spread vertically to this floor very quickly. (Case B)
• The smoke made me a little nervous; however, I was relieved by the announcement by the SCC staff because the instruction was clear. (Case A)
• It is said that there is a fire, but things outside don’t seem as hectic as I thought, so I’m not overly concerned. (Case B)
(ii) Participants who felt danger
• The smoke made it easy to imagine that a fire actually occurred. The extent of the smoke made me nervous. (Case A)
• I was worried that the fire might spread vertically. (Case A)
• I could see the smoke, but I could not guess the status of the lower floors. This situation made me anxious. (Case B)
• The sight of smoke reinforces the reality of the fire, heightening my sense of anxiety. (Case A)
• I saw the dark smoke and a lot of fire trucks. So, I recognized that it was quite a large fire. Smoke almost reached the top of the building. (Case B)
(iii) Participants who lost their sense of judgment due to a lack of information
• The SCC staff made an announcement. However, I wondered if the situation was indeed safe. (Case A)
• I could only see the smoke. I could not get any information on the size of the fire or to what extent the floor area was burning. (Case A)
• All I could do was follow the instructions of the SCC staff. (Case B)
• I could not get exact information about the fire, so I felt anxious. I worried whether I had to evacuate. (Case A)
• I don’t really trust the information spread on social media (I started to think that way when I saw a lot of misinformation being spread when I was affected by an earthquake before). (Case B)
(iv) Participants who felt a contradiction between the announcement that said the fire was under control and the smoke ascending/SNS photos
• Considering the amount of smoke, the fire seemed very large. I thought I had to evacuate immediately. (Case A)
• If the fire was indeed controlled by the sprinkler, the smoke would not spill out like that. The fire may have already become large. (Case A)
• If the sprinkler had activated, the smoke would not be as black as in the photos on SNS. It would be white. (Case B)
• Even though this floor is not where the fire started, the fact that smoke is rising suggests that the fire is large in scale. I feel anxious about whether it can really be extinguished safely. (Case A)
• The announcement relieved me; however, the photos on SNS were totally different than how I imagined the situation. (Case B)
(v) Other
• There are many irresponsible posts on SNS. (Case B)
• No additional comments. (Case B)
• No additional comments. (Case B)
The free comments reflect why the participants chose the actions selected when answering the multiple-choice questions. Figure 10 shows the total number of each classification among the professionals and laypeople for Cases A and B. To test for significant differences between the professionals and laypeople, the Wilcoxon rank sum test with a significant level of p = 0.05 was performed. The null hypothesis for the test was that both cases have identical effects. Significant differences were found only in Case B. This result indicates that the major difference was that the percentage of participants who felt danger, (ii), was larger for the laypeople than for the professionals. In contrast, the percentage of participants who felt the contradiction between announcements and SNS photos, (iv), was larger for the professionals than for the laypeople.
Fig. 10
Percentages for the classifications of answers to Q-A3 and Q-B3
Figure 11 shows the feelings of participants while waiting in the conference/meeting room after seeing the smoke ascend outside or checking SNS posts. The number of participants who felt very anxious was larger for the laypeople than for the professionals. One possible reason for this difference is that the professionals would have a better understanding of smoke and fire in these situations, so they would answer in a conservative manner. However, common among both groups was that more than 90% of the professionals and laypeople felt either very anxious or somewhat anxious after seeing the smoke ascend or SNS posts. Consequently, the majority of participants who were classified as (ii) and (iv) were included in the category of those who felt very anxious or somewhat anxious.
Figures 12 and 13 show the percentages of answers for each classification for Q-A3 and Q-B3, respectively. Regarding Case A, a significant difference between the professionals and laypeople overall was not confirmed, as shown in Fig. 10. However, comparing the results shown in Fig. 12 for each classification, the difference between the professionals and laypeople is a little larger in (iv) for Q-A4. Specifically, 52% (11/21) of professionals classified as (iv) answered, “d. Go to the staircase because…”. Regarding Case B, significant differences in (ii) and (iv) between the professionals and laypeople were confirmed, as shown in Fig. 10. Comparing the results shown in Fig. 13 for each classification, the difference between the professionals and laypeople is a little larger in (iv) for Q-B1, Q-B4, and Q-B5. Specifically, 13%, 44%, and, 31% (2/16, 7/16, and 5/16) of professionals classified as (iv) answered, “d. Go to the staircase because…”.
Fig. 12
Percentages of answers for each classification for answers to Q-A3, Case A
To test for significant differences between professionals and laypeople for Q-A4, Q-B1, Q-B4, and Q-B5, the Wilcoxon rank sum test with a significant level of p = 0.05 was performed. The p-values are presented in Table 8. Significant differences were not found in either of the cases. However, common among the four questions is that the percentage of “d. Go to the staircase because…” was larger for the professionals than for the laypeople.
Table 8
Results of Wilcoxon rank sum test between the professionals and laypeople for categories classified as iv) by the free description
iv), Q-A4
iv), Q-B1
iv), Q-B4
iv), Q-B5
p-value
* p < 0.05
0.120
0.144
0.140
0.454
4 Discussion
A survey similar to Q-A4 has been reported by Barber just after the September 11, 2001 disaster [8]. The results from that study showed that 42% of the participants waited for further evacuation instructions, 53% chose to evacuate by the staircase immediately, and 5% used a lift to evacuate. Thus, the percentage of participants who chose to wait as instructed in this study is quite large compared with that previous survey. Moreover, the percentage of participants who evacuated using the staircase or elevator in the previous survey is quite large compared with the results in this study. A possible reason for the difference is that the previous survey was conducted in 2002, soon after the September 11, 2001 disaster, and the participants were occupants on floors 50 and 51 of an existing high-rise office building. Therefore, the absolute value of the percentage of participants who chose to wait will widely differ depending on risk salience. However, the previous survey had the limitation that the participants were just asked to imagine the situation and then answer questions.
Significant differences between several questions (Q-A1, A4, and A5, and Q-B1, B4, and B5) were confirmed (Figs. 8 and 9). The following results were confirmed among the data exhibiting significant differences. If choices of “a. Wait here (in the conference/meeting room)” and “b. Move to the lobby with windows and wait there” were considered as following the instructions of the SCC staff member to wait, more than 70% of both the professionals and laypeople chose to wait by following the instructions of the SCC staff member in Case A, and nearly 70% of them did the same in Case B. By acquiring informal fire information, such as seeing smoke outside and photos on SNS, that percentage decreased to approximately 50% both for professionals and laypeople in Case A and Case B. Conversely, the percentage of those choosing “c. Move to the corridor (beyond the lobby) and wait there” or “d. Go to the staircase because I want to evacuate from the building” increased to approximately 50%. Firefighter announcements reduced the percentage of professionals who chose to move to the staircase.
As shown in Table 3 and Fig. 10a, significant differences between the professionals and laypeople were not confirmed. Figure 11 also shows the anxiety due to seeing smoke outside and SNS posts was similar between the professionals and laypeople. However, more detailed analysis and comparison brought some insight. For example, Fig. 10b shows that the percentage of professionals who felt a contradiction between the announcement and SNS posts was larger than that for the laypeople. Figure 9 shows the percentage of professionals who chose to move to the staircase was larger than that for the laypeople. Furthermore, even though significant differences were not confirmed (Table 8), the results shown in Fig. 12 and Fig. 13 indicate a large percentage of those were classified as “iv) Participants who felt a contradiction between the announcement that said the fire was under control and the smoke ascending/SNS photos.” This result suggests that the most radical behavior of trying to evacuate was induced by the perceived contradiction in information between the announcement by SCC staff and the smoke situation. 13–52% of the professionals who were classified as iv) doubted that the fire was as adequately controlled as the SCC staff said, and were concerned that the amount or appearance of the smoke contradicted the announcements. Therefore, their professional knowledge of fire safety likely made them choose to evacuate using the staircase. More adequate situational awareness made them decide to disobey the SCC staff’s instructions if the situation seemed to contradict the announcement by SCC staff.
Considering this possibility of occupants disobeying the SCC staff’s instructions, if the fire becomes large, the choice of the SCC staff will be problematic in terms of transmitting such information about the severity of the situation to the occupants. Conversely, if SCC staff do not provide the correct information that the fire is spreading, some occupants may notice the contradiction by acquiring informal information, such as seeing smoke outside or photos on SNS, and may try to evacuate via the staircase. If some occupants attempt to evacuate via the staircase, others may follow them [8, 23, 26, 27] and heavy congestion may occur; however, evaluating this type of behavior was beyond the scope of the present study.
The fire becoming excessively large means the fire and smoke control system has failed. In this situation, there is greater dependence on evacuation as the secondary safety measure. However, the results of the present study indicate that failure of fire and smoke control will negatively impact the phased evacuation strategy. The necessity of an appropriately executed phased evacuation increases as the situation of the fire worsens; however, properly performing the phased evacuation also becomes more challenging. Thus, securing fire and smoke control is indispensable for ensuring evacuation safety. Another approach will be to permit the evacuation of occupants through effective countermeasures. For example, the participants were informed that the corridor around the building core was fire-compartmented from the conference room and meeting room area. Therefore, they had a choice to move to the corridor for temporary safety without needing to evacuate immediately using the staircase. Intermediate refuge floors that accommodate occupants from the staircases is one way to ease the density of people in the staircases [4, 6, 43‐45].
There are some limitations to this study. First, even though the experience of visiting, moving, and staying in the building, and the appearance of smoke were replicated by 2D movies using a first-person or over-the-shoulder perspective, the present study did not replicate the physical experience of an actual fire situation in a high-rise building [8, 39]. However, the essence of the experience of waiting during a simulated phased evacuation is well replicated by the 2D movies and sounds. Second, the presence of other occupants was not replicated. Therefore, this study did not consider the possibility that the behavior of other occupants would influence the participant’s decisions [8, 23, 26, 27]. This study also did not consider any interactions between occupants. This study only focused on the fundamental inclinations and actions of the individual participants after acquiring several kinds of information and instructions during a simulated phased evacuation. Third, the participants were not occupants of an actual high-rise building and were instructed to assume they had visited the building for the first time, so caution is warranted when generalizing the results to real-world evacuation scenarios involving building occupants who are knowledgeable of their surroundings [8].
5 Conclusions
This study examined the effect acquiring unintended and informal fire information, such as seeing smoke outside the building or photos on SNS, has on building occupants during a phased evacuation. This type of information can increase the occupants’ anxiety and motivation to evacuate. More importantly, if the severity of the fire information contradicts the instructions from the SCC, the occupants may take radical evacuation action, such as trying to descend the staircases despite the SCC having instructed them to wait as part of the phased evacuation strategy. The tendencies for an increase in anxiety and greater motivation to evacuate are stronger in fire safety professionals who have more knowledge about fire safety than laypeople. The fire safety knowledge that professionals had appeared to make them more likely to notice contradictions between cues, leading more of them to choose an evacuation action that contradicted the phased evacuation announcement. Although the percentage of participants not following the announcement for phased evacuation is lower compared with the results of a survey conducted just after the September 11, 2001 disaster, this inclination to disobey evacuation instructions should be carefully considered in evacuation guidance planning. Ultimately, even though a phased evacuation strategy is effective, fire and smoke can work against the effectiveness of that strategy. Moreover, fire and smoke make phased evacuation more necessary. Fire and smoke control aimed at preventing heightened anxiety among evacuees is also necessary for a reliable phased evacuation strategy.
Acknowledgements
The author expresses his gratitude to the fire safety professionals who volunteered to participate in this study. The author thanks FORTE Science Communications (https://www.forte-science.co.jp/) for English language editing of the manuscript. This paper is an extended version of a poster presented at the 15th SFPE Fire Safety Conference & Expo on Performance-Based Design in Copenhagen, 17-19 April 2024.
Declarations
Competing Interests
The author declared no potential conflicts of interest regarding the research, authorship, and/or publication of this article.
Ethical Approval
The ethics committee of the author’s organization reviewed and approved this research. (approval No. K-46–1, 2023, for the experiment with fire safety professionals, and approval No. K-190, 2024, for the experiment with laypeople.) Participants were informed about the purpose of the study via printed documents and signed a consent form for data collection and storage before taking part in the study. They were also informed that they could withdraw from the study at any time for any reason. All the experiments were performed under the supervision of the researcher in case of an emergency.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Ariff A (2003) Review of evacuation procedures for Petronas Twin Towers. In: Proceedings of the CIB-CTBUH International conference on tall buildings, 20–23 October 2003, Malaysia, CIB Publication No 290:35–42.
2.
Kadokura H, Sekizawa A, Sano T, Fujii K (2015) Study on flow and congestion in staircases during phased evacuation in a high-rise building-Analysis based on the observational data of a real total evacuation drill. J Environ Eng, AIJ 80(716):849–856. https://doi.org/10.3130/aije.80.849(inJapanese)CrossRef
3.
Fujii K, Mizuno M, Kadokura H, Sano T, Sekizawa A (2020) Influence of Congestion in Upper Floor on Pedestrian Flow in Staircase Based on Observational Survey of Evacuation Drill in a High-Rise Building. In: Wu, GY., Tsai, KC., Chow, W.K. (eds) The Proceedings of 11th Asia-Oceania Symposium on Fire Science and Technology, AOSFST 2018 Springer, Singapore. https://doi.org/10.1007/978-981-32-9139-3_21
Lay S (2007) Alternative evacuation design solutions for high-rise buildings. Struct Design Tall Spec Build 16(4):487–500. https://doi.org/10.1002/tal.412CrossRef
6.
Ronchi E, Nilsson D (2013) Fire evacuation in high-rise buildings: a review of human behaviour and modelling research. Fire Sci Rev 2:7. https://doi.org/10.1186/2193-0414-2-7CrossRef
Barber D, Merkestein R V (2003) Will occupants of tall buildings obey instructions from wardens in staged evacuations? – The design dilemma post September 11. In: Proceedings of the CIB-CTBUH International conference on tall buildings, 20–23 October 2003, Malaysia, CIB Publication No 290:61–70.
Shields TJ, Boyce KE, McConnell N (2009) The behaviour and evacuation experiences of WTC 9/11 evacuees with self-designated mobility impairments. Fire Saf J 44:881–893. https://doi.org/10.1016/j.firesaf.2009.04.004CrossRef
Hosseini O, Maghrebi M, Maghrebi MF (2021) Determining optimum staged-evacuation schedule considering total evacuation time, congestion severity and fire threats. Saf Sci 139:105211. https://doi.org/10.1016/j.ssci.2021.105211CrossRef
19.
Wu GY, Mizuno M, Park SK (2024) Numerical analysis and validation on the phased evacuation time in high-rise buildings. Indoor Built Environ 33(3):501–520. https://doi.org/10.1177/1420326X231204512CrossRef
Benthorn L, Frantzich H (1999) Fire alarm in a public building: How do people evaluate information and choose an evacuation exit? Fire Matter 23(6):311–315CrossRef
Nilsson D, Johansson A (2009) Social influence during the initial phase of a fire evacuation—analysis of evacuation experiments in a cinema theatre. Fire Saf J 44(1):71–79. https://doi.org/10.1016/j.firesaf.2008.03.008CrossRef
24.
Kuligowski E D (2009) The process of human behavior in fires. NIST Tech Note 1632.
25.
Kuligowski E D (2011) Communicating the emergency: Preliminary findings on the elements of an effective public warning message. NIST Tech Note 1689.
Balboa A, Cuesta A, González-Villa J, Ortiz G, Alvear D (2023) Online experiments and regression analysis of evacuation decisions in response to fire alarms. Fire Saf J 141:103954. https://doi.org/10.1016/j.firesaf.2023.103954CrossRef
28.
Minegishi Y, Takahashi W, Ikehata Y (2023) Experimental study on the effect of fire alarms on occupants for situational awareness and evacuation decision-making using a mobile virtual reality head-mounted display. Japan Archit Rev 6(1):e12321. https://doi.org/10.1002/2475-8876.12321CrossRef
29.
Saito J, Isagawa T (2021) Influence of information during phased evacuation from a high-rise building on mental state of the person in the building. AIJ J Technol Des 27(66):847–852. https://doi.org/10.3130/aijt.27.847CrossRef
30.
Wada T, Sekizawa A (2016) Study on estimation method of optimum phased evacuation plan in a high-rise building. Bull Japan Assoc Fire Sci Eng 66(3):31–40. https://doi.org/10.11196/kasai.66.3_31CrossRef
31.
Hunt K, Agarwal P, Aziz RA, Zhuang J (2020) Fighting fake news during disasters—machine learning, game theory shown to be effective tools to monitor, debunk misinformation. ORMS Today, Feb 3:2020. https://doi.org/10.1287/orms.2020.01.06CrossRef
32.
Torpan S et al (2021) Handling false information in emergency management: a cross-national comparative study of European practices. Int J Disaster Risk Reduct 57:102151. https://doi.org/10.1016/j.ijdrr.2021.102151CrossRef
33.
Atila U, Ortakci Y, Ozacar K, Demiral E, Karas IR (2018) SmartEscape: a mobile smart individual fire evacuation system based on 3D spatial model. Int J Geo-Inf 7(6):223. https://doi.org/10.3390/ijgi7060223CrossRef
Gan Q, Liu Z, Liu T, Chai Y (2022) An indoor evacuation guidance system with an AR virtual agent, 2022 Annual International Conference on Brain-Inspired Cognitive Architectures for Artificial Intelligence: The 13th Annual Meeting of the BICA Society, Procedia Computer Science. 213:636–642. https://doi.org/10.1016/j.procs.2022.11.115
Minegishi Y (2024) Crowd management employing nudge theory for safe elevator use by people with mobility limitations during a high-rise building evacuation. Fire Saf J 147:104185. https://doi.org/10.1016/j.firesaf.2024.104185CrossRef
40.
Ministry of Construction (2000) Notification No. 1441, Establishment of calculation methods, etc. related to the Floor Evacuation Safety Verification Act, Japan, (in Japanese)
Ming Lo S, Will BF (1997) A view to the requirement of designated refuge floors in high-rise buildings in Hong Kong. Fire Saf Sci 5:737–745. https://doi.org/10.3801/IAFSS.FSS.5-737CrossRef
44.
Clawson K, O’Connor D J (2011) Considerations and challenges for refuge areas in tall buildings. In: Proceedings of the CTBUH world conference, 10–12 October 2011, Seoul, 332–338.
45.
Barber D, Johnson P (2015): Refuge floors—History, implementation and methods for improvement. In: Proceedings of the 6th international symposium human behaviour in fire 2015, 28–30 September 2015, Cambridge, 245–256.
