Available online at www.sciencedirect.

com

ScienceDirect
Procedia Manufacturing 00 (2018) 000–000
www.elsevier.com/locate/procedia

11th International Conference Interdisciplinarity in Engineering, INTER-ENG 2017, 5-6 October

2017, Tirgu-Mures, Romania

Evacuation Strategy Design for a Medium Size Auditorium using

CFD
Ruxandra Darmona*
a
Technical University of Cluj-Napoca, 25 Baritiu,Cluj-Napoca,400027, Romania

Abstract

Romanian fire safety regulations concerning the rooms with high density of people, as conference halls, auditoria, theatres, are
based on prescriptions. The Eurocode SR EN 1991-1-2 allows the application of alternative design employing CFD methods,
without specifying any further guidance to their practical praxis. This research is intended to give a better insight of the use of
CFD for a code compliant fire safety design.
© 2018 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the scientific committee of the 11th International Conference Interdisciplinarity in
Engineering.

Keywords: auditorium, evacuation analysis, numerical simulations;

1. Introduction

Auditoria spaces are assembly places, with a high density of persons which are not familiar with the building. The
occupants can have a wide variety of different age, culture, affiliation, therefore the life and safety provisions will
need to cover the whole range of variables in order to design a building having a satisfactory safety level. The most
of the building design codes have general requirements as: “Buildings shall be designed so that satisfactory escape
can be provided in the event of fire”, stated by the Swedish legislation [1].
Assembly occupancy has special provisions within each country, due to the higher design occupant densities. A
survey, carried out by Forell and co. [2], compared the standard codes for evacuation of eight European countries

* Corresponding author. Tel.: +0-040-264-401-549; fax:-.

E-mail address: Ruxandra.Darmon@ccm.utcluj.ro

2351-9789 © 2018 The Authors. Published by Elsevier B.V.

Peer-review under responsibility of the scientific committee of the 11th International Conference Interdisciplinarity in Engineering.
2 Ruxandra Darmon/ Procedia Manufacturing 00 (2018) 000–000

and China and USA and revealed that there is a disperse range of the results differing by a factor of three, related to
hand calculations and correlations of traveling times, travel distances and occupant density.
Romanian fire safety regulations [3] concerning the rooms with high density of people, as conference halls,
auditoria, theatres, are based on prescriptions. According to the occupancy type, it is specified a maximum travel
distance and a maximum time for evacuation on one or two separate ways. Based on the required figures, it results a
constant walking speed of 0.4 m/s and there is no mention of other factors as occupant’s age, sex, familiarity with
the building, etc.
In order to demonstrate that a building is safe, according to Rumanian regulations [3], one should only compare
the walking distances on the building plans to the required distances and to mention the egress times. However, it is
often the case that the new buildings and special structures do not fit within the code requirements. The revision of
fire safety regulations is much slower than the industrial and technological progress, slowing the adoption of some
new sustainable materials or design solutions.
The Eurocode SR EN 1991-1-2 allows the application of alternative design employing CFD methods, without
specifying any further guidance to their practical application.
This research paper is intended to give a better insight into the use of CFD for a code compliant fire safety design
and a way of overcoming the prescriptive regulations drawbacks by applying a performance based design solution to
a medium size auditorium, having three levels of sitting places.

Nomenclature

AJV Authority Having Jurisdiction

ASET Available Safe Egress Time
CFD Computational Fluid Dynamics
Code A standard that is an extensive compilation of provisions covering broad subject matter or that is suitable
for adoption into law independently of other codes and standards.
FDS Fluid Dynamics Simulator
RSET Required Safe Egress Time
Shall Indicates a mandatory requirement.
Should indicate a recommendation or that certain provision is advised but not required.

1.1. Tenability criteria

The Rumanian Fire Safety Code [3] does not have any mention about the tenability criteria; therefore, within this
paper, it has been taken into account the international guidance and standards.
According to the NFPA 5000 [4], when an engineering analysis is performed, it shall demonstrate “that the
building is designed to keep the smoke layer interface above the highest unprotected opening to adjoining spaces, or
1830 mm above the highest floor level of exit access open to the atrium, for a period equal to 1½ times the
calculated egress time or 20 minutes, whichever is greater.”
NFPA 101 [5] and SFPE Handbook of Fire Protection Engineering [6] specify three sets of criteria to be met
along the escape routes, in order to prove compliance with the accepted tenability limits.

1.1.1. Criteria 1 – Smoke layer 2m above the floor

The limiting heat radiation received by a person from a hot smoke layer or other fire condition is 2.5 kW/m2. This
radiant heat level generally occurs when temperatures are approximately 200ºC in small enclosures with relatively
low ceiling heights. Therefore, the acceptance criterion is when the smoke layer height is greater than 2 m above the
floor and the smoke temperature is less than 200ºC.
 Ruxandra Darmon/ Procedia Manufacturing 00 (2018) 000–000 3

1.1.2. Criteria 2- Smoke layer below 2m

Untenable conditions are considered to occur if the smoke layer descends below 2 m and any of the following
temperature, visibility and toxicity limits are exceeded:

 Smoke Temperature > 60ºC

 Visibility < 10 m (optical density < 0.1 m-1)
 CO Toxicity > 1,400 ppm

Toxicity is generally considered to be acceptable if the visibility criterion is satisfied.

1.1.3. Criteria 3 – Make-up air velocity

Based on research [7] carried out about the ability of people to move through an exit against an opposing airflow,
it has been agreed that movement is not impeded for air speeds below 5m/s. Therefore the maximum airflow
velocity is recommended to be maintained below 5 m/s along the egress routes.
The tenability criteria do not apply for the occupants who are in the close vicinity of the fire, as it is assumed that
they will react and proceed to immediate evacuation, before they would get any harmful effect.

1.2. Design objectives

1.2.1. Objectives for the evacuation analysis

The evacuation analysis is used to prove that a design meets the life safety criteria. The ASET (Available Safe
Egress Time) for the Auditorium has been calculated and compared with the results of the CFD (Computational
Fluid Dynamics) analysis to determine if the time available for escape before conditions become untenable for
evacuation is greater than the RSET (Required Safe Egress Time). The tenability limits are different for the Lower
and the Upper Auditorium, ranging from more than 20 minutes to 3.6 minutes (210 sec) respectively.
According to Proulx [8], there should also be provided a safety margin between the ASET and RSET limits. The
safety margin or factor of safety for an absolute analysis is taken as 1.5 for the worst credible fire scenario, in the
equation (1), below. The requirement for occupant for life safety within the building is detailed in the equations:

ASET > RSET x 1,5 (1)

RSET = td + tpm + tt (2)

ASET = tu (3)

Where:
td = time to detection from fire initiation (minutes)
tpm = pre-movement time period for occupants (minutes)
tt = travel time for occupants to move to an exit (minutes)
tu = time to untenable conditions (minutes)

1.2.2. Objectives for the CFD Analysis

The auditorium contains a stage measuring more than 93 m2. Section 16.4.6.4 of NFPA 5000 [4] states that
“Regular stages greater than 93 m2 and legitimate stages shall be provided with emergency ventilation to provide a
means of removing smoke and combustion gases directly to the outside in the event of a fire”. On this basis, a
smoke control system is required within the auditorium.
The auditorium is surrounded by occupied areas which do not link directly to outside. The proposed design intent
is to provide make-up air via a combination of mechanical and natural means. This will be achieved by doors at the
back of the stage being powered open, in conjunction with a vent at the roof, which will open to provide natural
make-up air. The proposed design allows for 7 m3/s of mechanical make-up air and a minimum of 6.2 m2 of natural
4 Ruxandra Darmon/ Procedia Manufacturing 00 (2018) 000–000

openings through the roof. A total number of 4 fans, each extracting 11.5 m3/s will be provided at the roof level in
order to clear the smoke and combustion particulates. The ceiling of the auditorium is designed to be a suspended
acoustic ceiling with permanent slot openings for return air to reach the plenum above. The openings provide a total
of 14.5 m2 in area.
The CFD analysis has been carried out for the validation of the smoke control strategy and to provide the ASET
for the evacuation analysis.

2. Evacuation analysis parameters

The auditorium is divided into the Lower Auditorium and the Upper Auditorium with separated evacuation routes
marked as it is shown in the figures below.

Fig. 1. Lower Auditorium (left) and Upper Auditorium (right) evacuation routes

2.1. Occupant characteristics

2.1.1. Occupancy load

The maximum number of occupants present in the Auditorium has been retrieved based on the available number
of seats on the proposed layout. Thus, the maximum occupancy load for each level is detailed in the table 1, below:

Table 1.Auditorium occupancy load

Location Floor level Maximum number of persons
Lower Auditorium 1 415
Lower Auditorium 2 93
Upper Auditorium 3 363

2.1.2. Physical and mental state

Occupants are generally expected to be alert and awake requiring minimal response time in the event of an
emergency. It is expected that a minority of occupants will require assistance by staff and security personnel to
ensure that they do not remain unattended in the event of a fire.
 Ruxandra Darmon/ Procedia Manufacturing 00 (2018) 000–000 5

All staff is expected to be aware of an emergency situation and to react accordingly. The staff is expected to have
a rational behaviour and the ability to take and implement decisions independently during an emergency situation.

2.1.3. Familiarity with the building

Occupants are not expected to be familiar with all exit routes. However it is likely that they will be aware of the
main entrance through which they entered and of the doors located on each side and marked with visible EXIT
signs.
The stuff is expected to be familiar with all the egress routes from the building

2.2. Assumptions and limitations

The evacuation analysis is based on the following assumptions:

 The evaluation of the Available Safe Evacuation Time for the Auditorium in this analysis is based on the results
of the CFD analysis. Thus, the ASET is equal with maximum duration calculated by numerical simulations,
while the tenable conditions are maintained in the referred space. The detection time has been also taken from the
CFD study.
 The occupants move toward the exits using the shortest path. Occupants will therefore queue at their shortest exit
route when evacuating.
 It is assumed that a fire in any point of the auditorium is likely to be noticed very quickly, therefore, once the fire
alarm goes off, all the occupants begin to evacuate, in the same time, in the optimum balance to their nearest exit.
 The occupant flow is assumed to be continuous, without any interruptions caused by decisions of the individuals
involved.
 The occupants of the Upper Auditorium are free of disabilities, therefore their egress and walking speed are not
impeded; the places for persons with disabilities are located at the Lower Auditorium, as indicated in Fig. 1.
 Once the occupants have passed through the second door of the protected exits, they are considered in a place of
relative safety.
 The population has uniformly distributed walking speeds which are evenly spread across the occupancy within
the building. Within the SFPE Handbook of Fire Protection Engineering (3rd Edition) [8], are specified ranges of
walking speeds for walking ways and stairs as 1.27 – 0.51 m/s and 0.58 – 0.36 m/s, respectively. Thompson et
all. [9] proposed a chart with the flow rates (people/m/s) as a function of population density (people/m2) for three
age groups: young adults, children and elderly.
 An average walking speed of 0.5m/s is assumed for the Upper Auditorium, given the slope and the population
density of 0.65 pers/m2. For the Lower Auditorium, an average walking speed of 1m/s and a moderate density of
1 pers/m2 are considered suitable.
 The pre-movement time is considered the period between the moment when the occupants have been alerted to a
fire alarm or noticed the emergency situation by any other cue (visual, smoke, alerted by others) and the time
when the movement phase starts. An extensive study conducted by VTT Finland [10] has shown that in a room or
a large open space people tend to start to evacuate before the whole evacuation announcement had finished, so
the pre-movement time was observed to be in a range of 10s-20s from the moment when of the first person stands
up. In SFPE Handbook [8] is mentioned that for a moderate flow capacity of 1 person per 1 m 2, a clear width of
1.22 m will permit a flow of 1.33 persons/sec.
For the Auditorium evacuation is assumed a 30 sec pre-movement time.
6 Ruxandra Darmon/ Procedia Manufacturing 00 (2018) 000–000

3. Evacuation analysis

3.1. Lower Auditorium

The Lower Auditorium has a total capacity of 510 persons on seats. The occupants performing on the stage are
not taken into account for this analysis. On the stage there are two doors and it is assumed that the occupants can
evacuate in less than 1 minute even if one of the doors is blocked by the fire.
The maximum travel distance is measured from the centre of the room and is approximate 23 m.

3.1.1. Available Safe Egress Time (ASET)

The ASET is considered to be in excess of 20 minutes as all scenarios demonstrated that tenable conditions were
still maintained within each smoke zone for at least 20 minutes.

3.1.2. Required Safe Egress Time (RSET)

The detection time has been taken from the CFD simulation, as the time measured from the fire initiation until
the first smoke detector goes off.

The time to detection from fire initiation (minutes): td = 1.3 min

The worst case scenario for the Lower Auditorium, taken into account CFD Analysis is Scenario 1a, when the
total volume extract rate is 23m3/s and the mobile partition walls are closed. The model assumes a delay of 50 s
between the time when the second smoke detector activates and the time when the mobile partition walls are
completely closed.
The time when the second smoke detector has been actuated, t sd, has been retrieved from the numerical
simulation results:

tsd = 23 sec

td = tsd + tdelay = 23s +50s = 78 sec (1.3 minutes)

The pre-movement time period for occupants (minutes): tpm = 0.5 min (30 sec)

The travel time for occupants to move to an exit (minutes): tt

The Lower Auditorium has four exits, each having a width of 1820 mm. In case of emergency it is assumed that
the occupants will evenly spread towards all four exits. Therefore, the maximum number of occupants for a single
exit, based on the total number of persons in the sale, will be:

N1 = 510 pers / 4 = 125.5 pers /exit

The moderate flow for 1820 mm width is 2 persons/ sec. The time necessary for all the occupants to pass through
an exit door is:

texit = 125.5 / 2 = 63.75 sec (1.06 min)

The most remote person from a door has to walk 23 m, meaning that it will reach the exit after 23 sec if the
walking speed is 1m/s. The person closest to a door is approximate 6 m away. Therefore, it is assumed that the total
travel time will be:

tt = 6 sec + 63.75 sec = 70 sec (1.16 min)

 Ruxandra Darmon/ Procedia Manufacturing 00 (2018) 000–000 7

For the Lower Auditorium:

RSET (Required Safe Evacuation Time) = 1.3 min + 0.5 min + 1.16 min = 2. 96 min => 3 minutes

Replacing into equation (1) :

20 min > 3 min x 1.5 = 4.5 min

3.2. Upper Auditorium

The Upper Auditorium has a total capacity of 363 persons on seats.

The maximum travel distance is measured on horizontal projection from the most remote place to an exit door
and is approximate 18.5 m. Taking into account the slope, the distance will be around 19.5 m.

3.2.1. Available Safe Egress Time (ASET)

The ASET is considered to be the tenability limit from the Scenario 1b, as 3.5 minutes (210 sec).

3.2.2. Required Safe Egress Time (RSET)

The time to detection from fire initiation (minutes): td = 0.52 min

The worst case scenario for the Upper Auditorium, taken into account CFD Analysis is Scenario 1b, when the
mobile partition walls fail to close and the smoke accumulates at the rear upper auditorium. In this model the
extraction system and alarm are assumed to be “double knock activation” systems. The time when the second smoke
detector has been actuated, tsd, = td in this case.

tsd = 31 sec

td = tsd + tdelay = 23s + 0s = 31 sec (0.52 min)

The pre-movement time period for occupants (minutes): tpm = 0.5 min (30 sec)

The travel time for occupants to move to an exit (minutes): tt

The Upper Auditorium has four exits, each having a width of 1220 mm. In case of emergency it is assumed that
the occupants will evenly spread towards all four exits. Therefore, the maximum number of occupants for a single
exit, based on the total number of persons in the hall, will be:

N1 = 363 pers / 4 = 90.75 pers /exit

For a moderate flow capacity of 1 person per 1 m2, a clear width of 1.22 m will permit a flow of 1.33
persons/sec. The time necessary for all the occupants to pass through an exit door is:

texit = 90.75 / 1.33 = 68.23 sec (1.14 min)

The most remote person from a door has to walk 19.5 m, meaning that it will reach the exit after 9.75 sec if the
walking speed on a slanted surface is 0.5m/s. The person closest to a door is approximate 1.5 m away. Therefore, it
is assumed that the total travel time will be:
8 Ruxandra Darmon/ Procedia Manufacturing 00 (2018) 000–000

tt = 0.75 sec + 68.75 sec = 68.98 sec (1.15 min)

For the Upper Auditorium:

RSET (Required Safe Evacuation Time) = 0.52 min + 0.5 min + 1.15 min = 2.17 minutes

Replacing into equation (1):

3.5 min > 2.17 min x 1.5 = 3.25 min

3.2.3. CFD fire scenarios

Three fire scenarios have been modelled with FDS: two for the 2.5MW fire on the stage, with the mobile
partition walls open and closed respectively and the third for the 1 MW steady fire on a seat from the rear side of the
Lower Auditorium.
The average fuel properties, modelled in FDS are detailed in the Appendix A.

4. Results

The results from the CFD analysis have been compared against the evacuation calculations to assess whether
tenable conditions are maintained for the safe evacuation of all the occupants present within the Auditorium space.
The scenarios where the mobile partition walls are open are the worst for the occupants of the Upper Auditorium.
The smoke starts to accumulate below the acoustic ceiling and it travels up towards the last row of chairs, forming a
uniform layer. It is considered that in the Upper Auditorium, tenable conditions are maintained for 3.5 minutes.

5. Conclusion

Analyzing the results of the numerical simulations it can be concluded that tenable conditions are maintained for
the entire period of 20 minutes simulation time, except for the cases when the mobile partition walls remain open,
when the tenable conditions for the Upper Auditorium are only maintained for 3.5 minutes. However, it is assumed
that the occupants of the Upper Auditorium have a good visibility of the whole volume space and it is likely that
they will notice the fire in the very early stages and will proceed to evacuate. The 3.5 minutes time is considered
sufficient for the complete evacuation of the last chair rows, given that the occupants have minimal travel distances
to the exit doors.
This study case shows an example of a performance based design. The required evacuation time for Auditorium
is lower than the available time estimated from the CFD analysis. Thus, the current means of egress, the number of
exits and the actual widths are considered to be acceptable to evacuate occupants prior to untenable conditions
within the Auditorium.

References

[1] Swedish Building Code BFS 2002:19.

[2] B.Forell, R.Seidenspinner, D.Hosser, Quantitative Comparison of International Design Standards of Escape Routes in Assembly Buildings,
W.W.F Klingsch et al. (eds), Pedestrian and Evacuation Dynamics, Springer, Berlin, Heidelberg, 2010, pp. 791-801.
[3] P118-99 Fire safety regulations for buildings.
[4] NFPA 5000, Building Construction and Safety Code, National Fire Protection Association, 2015.
[5] NFPA 101, Life Safety Code, National Fire Protection Association, 2015.
[6] SFPE Handbook of Fire Protection Engineering, 5th Edition, Springer, 2015.
[7] K Bosley, The Effects of Wind Speed on Escape Behaviour through Emergency Exits Summary Report, Research Report Number 53, 1992.
[8] G Proulx, Evacuation Time, SFPE Handbook of Fire Protection Engineering, 3rd Edition, 2008, pp.3-355-3-372.
[9] P.Thompson, D.Nilsson,K.Boyce,D.McGrath, Egress and Evacuation Models, FSE Engineering, 40th Anniversary Symposium and
Celebration, University of Edinburgh, 1974-2014.
[10] T.Rinne,K.Tillander, P.Gronberg, Data collection and analysis of evacuation situations, VTT Research Notes 2562, Finland, 2010.