FIRE IN EUROBORG FOOTBALL STADIUM; ANALYSIS OF HUMAN BEHAVIOUR

Transcription

1 FIRE IN EUROBORG FOOTBALL STADIUM; ANALYSIS OF HUMAN BEHAVIOUR Nancy Oberijé, Margrethe Kobes, Jans Weges & Jos Post (Correspond to: Netherlands Institute for Safety (NIFV), The Netherlands INTRODUCTION In the afternoon on April the 14th of 2008 an important football match (FC Groningen Ajax) was to be played in the Euroborg football stadium in the Dutch city of Groningen supporters were present to witness the match. For this occasion the hard core FC Groningen supporters club had arranged a campaign to create a chaos by using toilet paper and confetti. For this purpose they distributed many rolls of toilet paper (over 1500) among the FC Groningen supporters on the grandstand at the south side of the football stadium. The idea was to throw the toilet paper and confetti from the grandstand onto the field at the moment the football players would enter the playground. However, the first rolls of toilet paper were thrown already during the warming up event prior to the actual football match. Once the football players had left the playground and were about to re-enter for the official opening ceremony, the atmosphere not only figuratively, as expected, but also literally lightened up. The toilet paper on the grandstand and on the playground ignited, causing several small fires on the field and the grandstand. The smoke caused by the fires moved upwards into the grandstand at the north side of the stadium. At first nobody responded adequately to the small fires, rapidly expanding. Everybody was cheering and shouting, waiting for the match to start. Eventually the heat of the fires and the smoke became too intense and the supporters at the north side of the stadium had to be evacuated. The Netherlands Institute for Safety (NIFV) investigated the way people responded to the fire and how aspects of fire features, human features and building features influenced their behaviour after the fire started 1. This paper presents the results of this investigation. Two central questions will be addressed in this paper: 1 Which critical factors were the most influencing aspects in the fire response performance of the supporters at the Euroborg stadium? 2 Is the Fire Response Performance model (FRP-model) suitable for the evaluation of people's behaviour in fire in buildings? METHOD The analysis of the critical factors of the supporters' behaviour used the FRP-model 2 as shown in figure 1. Fire response performance is the human capacity to perceive and interpret danger signals and to carry out decisions aimed at surviving a fire situation 2. Roughly, three factors determine the degree of fire response performance in the event of fire in a building. These three factors are 2 : fire characteristics; building characteristics; human characteristics.

2 Figure 1. FRP-model Fire response performance Human features Individual features Personality Knowledge & experience Powers of observation Powers of judgement Powers of movement Social features Affiliation (e.g. family) Task fixation Role / responsibility Situational features Awareness Physical position Familiarity with lay out Building features Engineered features Lay out Installations Materials Compartments Size of building Situational features Focus point Occupant density Ease of wayfinding Building evacuation team Maintenance Fire features Perceptual features Visual features Smelling features Audible features Tangible features Fire growth rate Smoke yield Toxicity Heat These three factors are subdivided into several detailed characteristics. Figure 1 shows the most important characteristics that influence people's fire response performance. In the analysis of the fire in the Euroborg stadium a description of the influencing factors during the incident is given. The FRPmodel is described in detail in Kobes et. al, For this study the following sources were used: interviews with supporters, taken during the incident and live broadcasted by a local radio station (RTV Noord); videos of the incident, taken by supporters on the south side grandstand and published on YouTube; television broadcasts by a local TV station (RTV Noord); interviews conducted by the NIFV with security stewards who eye witnessed the starting of the fire and the behaviour of the supporters and who were also in charge of the evacuation of the supporters; information about the architectural design and installations present in the stadium. ANALYSIS OF THE FIRE CHARACTERISTICS In the FRP-model the fire characteristics are subdivided into the following features (figure 1): Perceptible features Fire growth rate Smoke yield Toxicity Heat

3 After describing the start and development of the fire, the fire characteristics will be analysed according to these features. Description of the start and development of the fire 15 minutes before the start of the match, the first rolls of toilet paper were thrown. Only 7 minutes after that the first fires occurred at the field. Security guards unsuccessfully tried to put them out with some buckets of water. 3 minutes after their extinguishing attempts, some parts of the boarding around the field caught fire. Another 3 minutes later the fire escalated to the grandstand and synthetic seats and a banner catch fire. The smoke is building up and rises from the field upwards to the corridor at the first ring. From this corridor the smoke rises through the staircase to the corridor in the second ring, blocking its normal escape route. 20 minutes after the first fire occurs the fire brigade arrives at the stadium and starts to extinguish the fire. 18 minutes later the fire is under control. A timeline of the fire is shown in figure 2. Figure 2. Fire development and interventions Fire starts Boarding catches fire Grandstand catches fire Fire extinguishing attempts security guards Fire brigade arrives Fire under control Perceptible features Perceptible features are the effects that one can perceive from a fire. For instance can one see, hear, and smell or feel the fire. The assumption in the FRP-model is that the better a fire can be perceived, the better people will be able to respond adequately to that fire, and the better their fire response performance will be. In Euroborg the fires on the field and the smoke development were visible for all the supporters on the grandstand. They could also smell the smoke. However this did not persuade people to leave the stadium. Only when they could actually feel the heat quite intensely, or when they got breathing problems from the smoke people started to evacuate. This means that only the feeling of the fire and the density of the smoke positively influenced their fire response performance. Fire growth rate The FRP-model assumes that the fire growth rate influences the fire response performance negatively. The faster the fire develops, the less time there is left for a successful evacuation. In Euroborg the fire developed slowly at the beginning, but once the players had entered the field the conditions worsened very quickly. Supporters were clearly not aware of the fact that the fire could develop so fast. A video of the incident at YouTube shows people throwing toilet paper and cheering when the players enter the field. Only 17 seconds after this, a supporter shouts a warning, telling people not to throw any more paper, exclaiming "this is going horribly wrong". 47 seconds after this fragment one can hear a supporter shout that it is becoming terribly hot and that "we have to get out of here". Obviously the fire growth rate was faster than supporters had expected. This aspect negatively influenced their fire response performance, as they wasted time before responding to the fire. Smoke yield The FRP-model assumes a negative effect from smoke on the fire response performance. Smoke can negatively influence the fire response performance, because it can obstruct the view of the escape routes. In the Euroborg stadium the smoke was not so heavy that it obstructed the supporters' view of the normal escape routes. It did however obstruct the view of the emergency exit signs. It also

4 hampered the security guards in their actions, as they could not see what was happening in other parts of the stadium and they could not see what their colleagues were doing. All together the smoke yield negatively influenced the fire response performance. Toxicity The FRP-model assumes a negative effect from toxicity of smoke. In Euroborg for most of the supporters the toxicity of the smoke did not cause any problems. However 24 persons had to be taken to hospital with severe breathing problems. Especially the smoke from the synthetic banners and seats made them choke. One security guard fainted because of the smoke and had to be taken to the hospital. All together the toxicity of the smoke had a negative effect on the fire response performance. Heat In the FRP-model heat is assumed to have a negative effect on the fire response performance. When people are exposed to heat over a too long period, they will suffer negative effects. At first they will suffer from physiological effects like sweating, higher heartbeat and more. As a result of this their ability to make decisions will decline. When exposed even longer to intense heat injuries like burns will occur, eventually causing death. In Euroborg the heat was not intense enough to cause such negative effects. Only one person had to be taken to hospital with burns. In this situation the heat only had a positive effect, as it persuaded people to start evacuating (see also perceptible features). Conclusion and recommendations In the Euroborg stadium the following fire characteristics factors negatively influenced the fire response performance: the fire growth rate; the smoke yield; the toxicity of the smoke. The heat of the fire did not have a negative effect on the fire response performance, as the fire fortunately could be stopped before it expanded to the whole stadium. The heat actually positively influenced the fire response performance, as this perceptible feature appeared to be the best trigger for people to start the evacuation. All together the smoke was the most influencing feature on the fire response performance as it blocked one escape route and obstructed the view to the emergency exit signs. The obvious recommendation in this case is the fact that no materials should be used that can cause fire. Moreover it is recommended that in football stadiums no materials should be used that can cause heavy and toxic smoke. ANALYSIS OF THE BUILDING CHARACTERISTICS The building characteristics are divided into two subgroups with the following features: 1. engineered features a. lay-out b. installations c. materials d. compartmentation e. size of building 2. situational features a. focus point b. occupation density c. ease of way finding d. building evacuation team e. maintenance

5 After a short description of the football stadium, the building characteristics will be analysed according to these features. Description of the football stadium In figure 3 the scheme of the football stadium is shown. The stadium consists of two rings around the playground. The first ring is the lower one. The second ring is situated above the first ring. The rings are connected via staircases inside the building. There are four main entrances situated in each corner of the stadium. These entrances are connected (via staircases) to the first ring. The second ring can only be reached via the staircases starting from the first ring. The escape routes in the stadium are the same as the entrance routes. Figure 3. Scheme of the Euroborg stadium In the stadium several fire related installations are present: a fire warning system; a sprinkler installation at the corridor at the first ring; a smoke and heat exhaust system at the corridor at the first ring; fire hoses at the first ring; the hoses are supplied with water after the fire alarm button is pressed (at the second ring and near the field no fire hoses are present); an electronic device to unlock the moveable lexan walls between the north, west, south and east side of the stadium; green emergency exit signs.

6 Engineered features; lay-out The FRP-model assumes that an understandable lay-out has a positive effect on the fire response performance. The Euroborg football stadium is designed in a comprehendible way, which contributes to adequate fire response performance. The routes that are used as entrance routes are the same as the (emergency) exit routes. The exits are clearly visible from the lower ring; people only need to walk towards the daylight. The effect of smoke however was not sufficiently taken into account in the design of the stadium. When the fire occurred, the supporters had to evacuate upwards (to the entrance/exit of the grandstand), in the same direction as the smoke was spreading. Another factor that complicated the fire response performance was the fact that the second ring could only be reached through the staircase from the first ring. As the heat caused the smoke to move upwards the smoke blocked the main escape route from the second ring through the staircase. So although the stadium design is comprehendible, the lay-out does not take the effects of smoke into account. In this case the lay-out had both a positive influence (normal exits) and a negative influence (emergency exits and smoke in one normal exit) on the fire response performance. Engineered features; installations In the FRP-model it is assumed that fire installations affect the fire response performance positively. However during the fire at Euroborg several installations did not function adequately and therefore negatively influenced the fire response performance as will be demonstrated in this paragraph. When the fire reached the grandstand the water supply to the fire hoses did not function. Supporters did not know that the fire alarm button (situated directly next to the fire hose) had to be pushed. But even after the fire alarm button was pushed there was no water supply. Due to a software fault the pump for the fire hoses was not activated, so they could not be used to extinguish the fires. This negatively influenced the fire response performance. It remains unclear whether the fire warning system was used. In the interviews contradictory information was received about this topic. The smoke and heat exhaust system was installed. However, the system was designed based on a fire scenario at the corridor. With the fire occurring at the field and the grandstand, the system did not function. Consequently the smoke yield at the moment of the fire was not according the requirements in the Dutch Building Code and supporters had to evacuate through the smoke. This negatively influenced their fire response performance. The lexan walls of the gallery compartments at the first and second ring between the north and east side of the stadium can only be opened with a key from a steward from the east side of the wall, or by electronic means from the command room. During the match the lexan wall at the first ring can for security reasons only be opened from the command room with an electronic device. This device is programmed with a 1 minute delay between the activation of the device and the actual opening of the wall. Although there were several requests from the security guards to open the lexan wall, there was no response from the command room. Eventually the lexan wall was opened from the other side by a security guard who managed to climb over the wall. The extra time needed to open the emergency exits negatively influenced the fire response performance of the supporters. The green exit signs were all present. However the required colour of the exit signs (green) was also the club colour of FC Groningen. All the walls in the grandstand corridors/galleries were painted green. Therefore the required exit signs were hard to see. Besides, the smoke accumulated at ceiling height where the exit signs were located. The location and colour of the exit signs negatively influenced the supporters' fire response performance. Engineered features; materials In the FRP-model it is assumed that materials can negatively influence the fire response performance when they catch fire. The materials that have been used in the stadium itself are not combustible or hardly flammable. However the toilet paper and confetti that was brought into the stadium did indeed negatively influence the fire response performance. After the fire started some seats started to smoulder and produced a lot of black, irritating and suffocating smoke. Also some synthetic banners caught fire and started spreading an irritating smoke, making the evacuation more difficult. Therefore the materials that were taken into the stadium negatively influenced the fire response performance.

7 Engineered features; compartmentation The FRP-model assumes that partitioning positively influences the fire response performance of the people present as the fire will not expand from one compartment to another compartment. In the Euroborg stadium corridor there was no partitioning into fire or smoke compartments. Instead a heat and smoke exhaust system was installed (as described earlier). The feature compartmentation therefore is not applicable in this case. Engineered features; size of building In the FRP-model it is assumed that the size of the building affects the fire response performance negatively. The bigger the building, the more difficult it is to find your way out. Although the stadium can be seen as a large building, we found no evidence that this aspect negatively affected the fire response performance. Situational features; focus point The FRP-model assumes that the presence of a focus point can positively influence the fire response performance if it can be used in informing the people present. In the stadium the field, the players and the ball are the natural focus points. During the start of the fire, this negatively influenced the fire response performance. Because the players entered the field, the supporters were expecting the match to start soon and stayed in their seats. There were (besides the fires themselves) no signals coming from the field to make people aware of the danger and to urge them start evacuating. Situational features; occupant density In the FRP-model, it is assumed that a high crowd density negatively influences the fire response performance. The high crowd density in the Euroborg stadium indeed negatively influenced the evacuation time and therefore the fire response performance. From interviews with security guards it showed that it takes about 15 minutes to get from part DD at the grandstand (see figure 3) to the entrance of the staircase number 3 (a distance of approximately 50 meters). Situational features; ease of way finding In the FRP-model, it is assumed that the ease of way finding in a building positively influences the fire response performance. In the Euroborg stadium the way finding to the normal exits is not difficult. The normal exits are (at midday) easy to find because of the light entering the building at these points. The way finding to the emergency exits however has proven to be difficult. The emergency exits are through the lexan walls. This was not known by most of the supporters. The green emergency exit signs did not contribute to finding the emergency exits as described earlier. When the normal exit via the staircase at the north tribune was blocked by the smoke, people just stood their waiting in front of the staircase. This indicates that the level of way finding in a building can differ under normal circumstances from emergency circumstances, as the routes that need to be chosen are different. Situational features; building evacuation team The FRP-model assumes that a well trained and educated building evacuation team positively influences the fire response performance. In the Euroborg stadium a building evacuation team indeed urged people to leave and showed them the emergency escape exits. However the control of the building evacuation team did not function very well. It took a long time before the decision was made to cancel the match. The decision was communicated to the supporters at 15.51, more than an hour after the first fires started. If this decision would have been taken earlier and the players would not have entered the field, the evacuation process would have speeded up significantly. Another aspect that negatively influenced the fire response performance was the fact that the management of the building evacuation team in the control room did not use the electronic device to unlock the emergency exits. This was partly caused by the fact that the communication via portables from the command room to the building evacuation team did not function. Afterwards another reason was revealed, namely that the management of the stadium was afraid of supporters groups colliding via the escape routes. As a result of all this, the building evacuation team had to handle on own insight. To

8 make things even worse, they could not see each other because of the smoke, which made the coordination of the evacuation even more difficult. They had never realised that this would be a complicating factor when they had evacuation drills. Concluding the building evacuation team influenced the fire response performance positively (as they showed people the emergency exits and urged them to get out) and negatively (as the emergency exit was not unlocked immediately and it took a long time before the decision was made to cancel the match). Situational features; maintenance In the FRP-model, it is assumed that a high level of maintenance improves the fire response performance. In the Euroborg stadium the maintenance of the following aspects failed: flammable materials were brought into the stadium, which was against the building regulations; the fire hoses did not work; the opening of the fire escape via the lexan wall did not function. Therefore in the Euroborg stadium the maintenance negatively influenced the fire response performance. Conclusion and recommendations The following building characteristics features negatively influenced the fire response performance: lay-out; installations; focus point; occupant density; ease of way finding; maintenance. There was no effect on the fire response performance found from the size of the building. The building evacuation team influenced the fire response performance positively and negatively. Compartmentation was not applicable in this case and therefore could not be judged. Regarding the features that influence the fire response performance negatively, it is evident that the maintenance plays a vital role in the occurring of the incident. If it had not been allowed to take flammable materials into the stadium, and if the fire hoses would have worked, the incident would not have developed into a situation that made an evacuation inevitable. Concerning the evacuation, it becomes clear, that the lay-out of the building, combined with the difficulty of way finding to the emergency exits, were two important causes for the delay in evacuation. Two situational factors - the focus point and the building evacuation team - also contributed to the delay in escape time as people did not receive any signals that the match was suspended. Therefore they kept on waiting for the match to start. Besides the obvious recommendation that there should be put more emphasis on maintenance in order to improve the fire response performance, it also recommended that the spreading behaviour of smoke should be taken into account in the design of a football stadium. In addition more attention should be given to the ease of finding the emergency exits when they differ from the normal exits. ANALYSIS OF THE HUMAN CHARACTERISTICS The human characteristics are divided into three subgroups with the following features 1. Individual features a. personality b. knowledge and experience c. powers of observation d. powers of judgement e. powers of movement

9 2. Social features a. affiliation b. task fixation c. role/responsibility 3. Situational features a. awareness b. physical position c. familiarity with lay-out After a description of the behaviour of supporters the human characteristics will be analysed according to these features. Description of the behaviour of supporters When the fire started first all the supporters were shouting and cheering. The security stewards that were present had great difficulty convincing people to leave. Only when the supporters could actually feel the heat, they started to evacuate and some supporters tried to use the fire hoses. During the evacuation people were calm although they were walking through the smoke. The supporters did not use the available normal staircase (staircase number 1 in figure 3) from the second to the first ring, because it was filled with smoke. Instead they stood there waiting, not knowing there was also a second escape route through a removable lexan wall. However there were no signs of panic. Individual features; personality In the FRP-model it is assumed that people who can operate successfully under stressful situations and people with a high self efficacy belief will have a better fire response performance. In the Euroborg stadium a mingling of people was present, so the influence of personality on their behaviour could not be established. Individual features; knowledge and experience In the FRP-model it is assumed that knowledge and experience can both have a positive or a negative effect on the fire response performance. In Euroborg it can be assumed that the supporters' knowledge and experience of fire and escape in fire situations was low, as this is the case for most civilians in The Netherlands. Watching the videotapes and conducting the interviews it was found that most supporters did not assess the danger of the fire very well. Most supporters only started to move when the fire escalated and the heat and the smoke on the grandstand became too hard to bear. This negatively influenced their fire response performance. Individual features; powers of observation The FRP-model assumes that powers of observation will positively influence the fire response performance. Observation can be seen as the first step leading to eventual evacuation behaviour. In Euroborg it can be assumed that the observation powers of supporters are satisfactory: they all came to see the football match. Therefore this aspect should have enhanced their fire response performance. However although the supporters could clearly observe the fire, it did not lead immediately to a proper reaction, i.e. their powers of observation did at first not influence their fire response performance. This might be related to problems described in the next steps of evacuation behaviour (judgement and movement, see the following two paragraphs). Individual features; powers of judgement In the FRP-model it is assumed that powers of judgement will positively influence the fire response performance. In this second step to evacuation behaviour the observed signals of danger have to be interpreted and a decision must be made about how to handle. In the Euroborg stadium this step of the evacuation process lasted longer then necessary. Although everybody could see the fire and the smoke, most of the supporters stayed on the grandstand in stead of starting to evacuate. Because the majority of the supporters did not start to evacuate at first, the people who were the closest to the fire (and were the first to feel the heat and suffered from the smoke) saw there escape route blocked and chose to escape via the seats or using the field (approximately 50 persons escaped via the field). Given the situation this seems to be a good choice. Also some people, among which people with

10 children and people with breathing problems made their way to the exits earlier than the rest of the public. Possibly they were more aware of their own vulnerability and judged the situation as more dangerous than the other supporters did. In addition some people with children evacuated to the toilets to avoid the pressure of the evacuating crowd. They seemed to think that the toilets were a safe place of refuge, whereas there is a substantial danger of being occluded by the fire in this type of rooms. All together for most of the people their lack of powers of judgement negatively influenced their fire response performance. Individual features; powers of movement In the FRP-model it is assumed that powers of movement positively influence the fire response performance. Movement to a safe place can be seen as the last step in evacuation behaviour. In Euroborg it can be assumed that the people at the grandstand had average powers of movement. In order to get to their seats at the grandstand they had to be reasonably mobile. During the evacuation the supporters' powers of movement on average affected their fire response performance positively. Social features; affiliation The FRP-model assumes that social affiliation can both positively or negatively influence the fire response performance. A positive influence can be found as people will help and rescue each other. A negative effect can occur when people that are present as a group will wait for each other and try to evacuate together, or go back looking for members of the group. In Euroborg during the evacuation children were found that were looking for there parents. Although they had the intention to evacuate together they obviously did not manage to do so. At some point it is probably hard to stick together in a moving crowd. Also one supporter reported that he could not keep up with his children and chose a different route. Regarding the FRP-model this feature may not play such an important role as assumed. Social features; task fixation In the FRP-model it is assumed that task fixation negatively influences the fire response performance. In the Euroborg stadium the 'task' of the supporters was to encourage the players by cheering and shouting. When the fire occurred they stayed in they supporters role by cheering and shouting. The security guards therefore had great difficulty convincing people to leave. Some supporters did not want to leave because they were afraid to miss the match. As a result of the fixation to their supporters' task, the supporters were not able to switch to a different role: the evacuation role. This is why in the Euroborg stadium the task fixation had a seriously negative influence on the fire response performance. Social features; role/responsibility In the fire response model it is assumed that the role and the responsibility belonging to this role influence the fire response performance positively. People that feel more responsible will sooner take action to do something about the perceived danger. The supporters present in the Euroborg stadium did not have any official responsibility for safety. This negatively affected their fire response performance, as most of the supporters just stood there and watched the fire grow. Later, when the situation had escalated and it was obvious something had to be done, some supporters tried to use the fire hoses to extinguish the fires. Parents also did feel responsible for the safety of their children and were seen to evacuate earlier than other people. This positively influenced their fire response performance. Situational features; awareness The FRP-model assumes that awareness positively influences the fire response performance. In the Euroborg stadium the supporters were very alert. They were waiting for the match to start and anxious not to miss anything. They were not under the influence of drugs or alcohol. Therefore this aspect should have enhanced their fire response performance. However, although the supporters were aware of the fire, this did not lead immediately to a proper reaction. This may have to do with problems in powers of judgement as described earlier.

11 Situational features; physical position In the FRP-model it is assumed that an active physical position (as in walking or running) enhances the fire response performance and a passive physical position (as in lying or sitting) prohibits adequate fire response performance. The supporters were standing on the grandstand, which is principally a passive physical position. On the other hand it can also be argued that not much effort is needed to start walking when standing upright, compared to the effort needed when people would be seated. However in the Euroborg stadium this position did not influence the fire response performance positively, as the people stayed at there places until they could actually feel the danger of the fire. Situational features; familiarity with lay-out In the FRP-model it is assumed that familiarity with lay-out will positively influence the fire response performance. It must be mentioned however, that some researchers note that familiarity with lay-out can also lead to neglecting the emergency escape routes and using the normal way out in case of an emergency. As the normal way out is not necessarily the shortest way out, this can affect the fire response performance negatively. In the Euroborg stadium the supporters from the north side belonged to the most fanatic supporters of FC Groningen. They were regular visitors of the stadium and were familiar with the lay-out and the regular exits. However they were not familiar with the emergency escape routes as they were waiting in front of the smoke filled staircase instead of turning to the emergency escape routes. In this case their familiarity with lay-out negatively influenced their fire response performance. Conclusion and recommendation The following human characteristics features negatively influenced the fire response performance: knowledge and experience powers of judgement task fixation role/responsibility familiarity with lay-out On the other hand the following features positively influenced the fire response performance: powers of movement role/responsibility (of parents) The influence of the following features could not be established: personality powers of observation affiliation awareness physical position Further research on these features is therefore recommended. Altogether the most influencing feature on the delay in the fire response performance were the powers of judgement. These powers of judgement were again negatively influenced by the audience's lack of knowledge and experience with fire and evacuations and the lack of responsibility. Their powers were also negatively influenced by the task fixation of the supporters. Concerning the evacuation behaviour, the familiarity with lay-out appeared to be the most influencing factor. Because people knew where the regular exit was, they did not go and look for a different exit when the normal exit turned out to be blocked. In order to enhance the fire response performance it is recommended to put a lot of effort in communication when an accident occurs. This should preferably be done via several means of communication. Supporters need to be informed that the situation is different from normal and therefore adjusted behaviour is requested. In other words they need to be supported in their powers of judgement.

12 REFLECTION AND RECOMMENDATION In this paragraph a reflection on the response performance model and the suitability of it for the evaluation of human behaviour in fire in a building will be given. When the results from the analysis of the fire in the Euroborg football stadium are compared to the FRP-model, the following remarks can be made: The fire characteristics as mentioned in the model appear to represent the features that played a role in the Euroborg stadium well. In the evaluation of the Euroborg fire no additional features were found and the influence of the features was as predicted by the model. The building characteristics as mentioned in the model appear to represent the features that played a role in the fire in the Euroborg stadium well. In the evaluation of the Euroborg stadium no additional features were found and the influence of the features was as predicted by the model. An exception should be made for the size of the building. This did not seem to influence the fire response performance in this case. Also a negative effect from the presence of several installations was found. This was actually caused by a lack of maintenance of the installations or faulty design, which caused them to function not properly. The human characteristics as mentioned in the model appear to represent the features that played a role in the fire in the Euroborg stadium well. In the evaluation of the human behaviour in the Euroborg fire no additional features were found. However the influence of the features personality, powers of observation, affiliation, awareness and physical position could not be established. Concerning the feature familiarity with lay-out it was shown in this case that this feature negatively influenced the fire response performance, whereas some researchers also report positive influences of this feature on the fire response performance. Considering the suitability of the FRP-model for the analysis of human behaviour in fire in a building, it can be concluded that the model provides a good framework to consider all the factors that may influence this behaviour. Features that influence the fire response performance negatively can afterwards be given a closer look, and recommendations can be made to improve the effect of these features. A disadvantage of the FRP-model is that it provides a qualitative analysis and the weighing of the effects of the different features on the fire response performance depends on the judgement of the researcher. Further research is therefore recommended in order to make the model suitable for a more quantitative and objective analysis. ACKNOWLEDGEMENTS The authors would like to thank Ralf Beerens of the NIFV research team for his review of an earlier version of this paper. REFERENCES 1 Weges, J.M., Kobes, M., 2008, Brand in Euroborg. NIFV, Arnhem. [Fire in Euroborg] 2 Kobes, M., 2008, Zelfredzaamheid bij brand; Kritische factoren voor het veilig vluchten uit gebouwen, Boom Juridische Uitgevers, Den Haag, The Netherlands, pp [Fire response performance. The critical factors for a safe escape out of buildings] 3 Kobes, M., Helsloot, I, Vries, B. de, Post, J.G., 2009 (draft), Building safety and human behaviour in fire: A literature review. Fire Safety Journal.