A Comprehensive Study of Two Fire Conditions in a Subway Train
Fire: Considerin
g
the Failure or Work of the S
p
rinkler S
y
stem
Ke Pan
1
, Jie Feng
2
and Jianyun Shi
1
1
School of Civil and Safety Engineering, Dalian Jiaotong University, Dalian, Liaoning, China
2
School of Resources & Civil Engineering, Northeast University, Shenyang, Liaoning, China
Keywords: Subway fire simulation, smoke flow, fire parameters, sprinkler system, the evacuation time.
Abstract: Subway fire has a greater risk, and it occupies a large proportion in the subway accidents. Through the
analysis of the possibility of subway fire and combustible, the model based on the actual size of one certain
railway station in Dalian is established. Several different fire conditions are considered in the model.
Different fire numbers and different fire power sources are set in the different fire locations, just as the
station, platform, and the train area. Fire temperature, smoke layer height, CO concentration and other fire
parameters were analyzed considering the failure or work of the sprinkler system. The need time for the fire
evacuation is analyzed in the two conditions. Also risk limits are compared to the scene of the fire
parameter. The results show that the sprinkler system can reduce the temperature of the fire, and it is a little
difficult to discharge the smoke from the top air outlet due to the rising part of the water droplets, the plume
temperature and the velocity decrease, which hinders the smoke emission and accelerates the visibility.
1 INTRODUCTION
Since the metro has such advantages as large
transportation capacity, high speed, low pollution,
less resource occupation, low energy consumption,
easy traffic, and comfort, which are in conformity
with the principle of sustainable development, it is
particularly applicable for big and medium-sized
cities. At present, there are over 100 cities all over
the world that are operating metros. As of December
31, 2016, the metro lines amounting to a total length
of 3168.7 km were opened to traffic in total in
Mainland China. By 2015, the metro lines total
length in Mainland China will reach 4189 km（Shi
et al, 2012）. Today, the metros in Chinese major
cities such as Beijing and Shanghai have become
one of the most populated metros in the world.
However, due to the heavily overcrowded
population and the situation of underground space,
there exist a lot of potential risks during the
operation of the metro station (Pan and Shi, 2011). A
single incident can be devastating, causing death and
millions of dollars in property loss. For instance, one
recent serious metro accident is the trains rear-end
accident of shanghai Metro Line 10 occurred in
September 27, 2011. More than 270 passengers were
injured in the accident. Subway fire is most frequent
and serious in the subway accident statistics. Large
quantities of smoke are likely to spread rapidly to
entire subway station due to stack effect. Especially,
because smoke spread path usually coincide with
passenger’s evacuation path, it will reduce visibility
and can cause fatalities by asphyxiation. Many
works were done in this field, just as the smoke flow
patterns and emergency rescue in the subway fire.
The different ventilation modes for fire in the
subway station were studied to clear the influence of
the different layout (Luo et al, 2014). Miho proposed
a quantitative method to assess road tunnel fire
safety based on a numerical simulation of in-smoke
evacuation (Miho et al, 2017). Jae performed the fire
simulation and evacuation simulation to estimate the
effect of the platform screen door and ventilation on
passenger’s life safety in a subway train fire (Jae et
al, 2009). Sanhay performed sensitivity analysis to
quantify the influence of ventilation velocity on the
fire parameters (Sanhay, 2017). A discrete design
method with integrated fire–evacuation model for
fire emergency evacuation was used to reduce the
simulation time and cost in fire emergency
evacuation simulations (Yang et al, 2017).
Generally, fire in a subway station forms a
complicated structure. This physical phenomenon
involves chemical reaction. And radiation is affected
by various parameters, including geometry, tunnel
slope, ventilation velocity, sidewalls restriction, and
182
Pan, K., Feng, J. and Shi, J.
A Comprehensive Study of Two Fire Conditions in a Subway Train Fire: Considering the Failure or Work of the Sprinkler System.
In 3rd International Conference on Electromechanical Control Technology and Transportation (ICECTT 2018), pages 182-186
ISBN: 978-989-758-312-4
Copyright © 2018 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved

pressure of passing air among other influential
variables. Some works were done in the paper to
quantify the influence of the sprinkler system on the
parameters: the maximum temperature, the
maximum smoke layer height, the maximum CO
concentration, the visibility in the subway fire.
Several fire simulation models were developed by
using FDS software. The risk of the subway fire can
be quantified by comparing with the parameters
which got from the simulation and the limits set in
the literatures and the laws.
2 THE SIMULATION MODEL
The actual model is built based on the actual size of
one railway station in Dalian. Figure 1, Figure 2
respectively shows the station floor and platform
floor of the subway station.
Figure 1: the station floor
Figure 2: the platform floor
2.1 The set limits of the fire parameters
Temperature, CO concentration, visibility are
usually used as the basis in the fire analysis. And the
shortest time of the three is set for reaching a hazard
state in the fire. With reference to "Design for
Asylum" and the criteria for quasi-safe areas, the
hazard limits are selected as follows:
 Temperature is over 60 ℃ at 2m height;
 CO concentration is up to 1400ppm at 2.1m
height;
 Visibility is less than 10m at 2m height.
2.2 Simulation scenarios and
parameters
Several fire conditions were set in table 1. The time
for the luggage fire in the platform floor and in the
station floor to reach the steady combustion is 450s.
In order to make the simulation results more
accurate, the simulation time is determined as 800s.
However, because of the large area of the station and
the rapid dissipation of the heat when the fire
reaches stable combustion, the simulation time is set
as 700s.
Table 1: The fire scenarios
case
fire
scenarios
Stable
fire
power /
MW
Failure of
the sprinkler
system
1
One fire at the
centre area of
platform floor
2.5 No
2
One fire at the
centre area of
platform floor
2.5 Yes
3
One fire at the
platform
staircase entrance
2.5 NO
4
Two fires at the
platform
staircase entrance
2.5 NO
5
Fire at the ticket
area of station
floor
2.5 NO
6
Fire at the middle
of the train
7.5 NO
3 THE RESULT ANALYSIS OF
THE SIMULATION
3.1 One fire at the centre area of
platform floor compared with the
failure of the sprinkler system
3.1.1 Smoke spreads
The comparison of smoke spreading in the central
station fire with the failure or no failure of sprinkler
system is shown in Figure 3.
A Comprehensive Study of Two Fire Conditions in a Subway Train Fire: Considering the Failure or Work of the Sprinkler System
183

0
20
40
60
80
100
120
200 300 400 500 600 700 800
Simulation time(s)
Spreading length
of smoke(m)
failure
No failure
Figure 3: Comparison of smoke spreading in the
central station fire with failure or no failure of
sprinkler system
Figure 3 shows that the spreading length of fire
smoke considering the work of the sprinkler system
is wider than the failure of the sprinkler system. The
temperature of the whole station area is reduced for
the no failure of the sprinkler system. The rate of
heat releases from fire sources and smoke spreads
slows down. The smoke spreads to the entire
platform floor more quickly compared with failure
of the sprinkler system. The rising smoke carries
some water droplets. And the temperature and speed
of the plume are reduced. The smoke is difficult to
discharge through the top exhaust vent. So the
smoke first spread to the entire platform floor, but
the effect is not particularly obvious.
3.1.2 The temperature distribution of the
platform floor and the height of the
smoke layer
When the sprinkler system in the centre area of
platform is intact or invalid, the changes of
temperature distribution and smoke layer height are
shown in Figure 4, Figure 5.
Figure 4: Station floor with a height of 2.1 m and a
temperature of 60 ° C
Figure 5: Comparison of smoke layer height changes in
the centre area of platform
ICECTT 2018 - 3rd International Conference on Electromechanical Control Technology and Transportation
184

Figure 4 shows that if the sprinkler system is
normal, the spreading time of the fire smoke is
435.3s at a height of 2.1 m when the temperature
gets up to 60℃ on both sides of the station stairs. If
the sprinkler system is invalid, the time of smoke
spread is 370.4s. It is about 65s earlier than the
condition of no failure of the sprinkler system. Thus,
the role of sprinkler system can reduce the
temperature of the fire. According to Figure 5, when
the sprinkler system is intact, the height of the
smoke layer decreases sharply in about 250s.When
the sprinkler system fails, the height of the smoke
layer drops sharply in 220s.The time of the smoke
layer to reach the minimum height for the failure of
the sprinkler system is 210s less than the another
condition. The rising smoke particles adsorb part of
the water droplets which make plume temperature
and velocity decrease. The smoke is difficult to
discharge through the top exhaust vent. The smoke
layer height changes more compared with the failure
of the sprinkler system. There is a higher probability
for the smoke layer to be located near the ceiling.
Thus, the role of the sprinkler system can hinder the
emission of smoke, but not particularly obvious.
3.1.3 CO distribution and visibility
When the sprinkler system in the centre area of
platform is normal or invalid, the CO distribution
and visibility surface at 10m of platform are shown
in Figure 6, Figure.7.
Figure 6: Comparison of CO distribution
Figure 7: Comparison of the visibility surface at 10m
According to Figure 6, the incomplete
combustion of combustibles is increased because of
the work of the sprinkler system. And the CO
volume fraction is slightly increased, but the value
still below the danger limit. According to Figure 7, it
almost reaches the same spatial scope comparison of
the visibility surface at 10m. The work of the
sprinkler system is better than the failure of the
sprinkler system. The function of sprinkler system
can also accelerate the reduction of visibility.
3.2 Comparison of fire parameters in
different fire scenarios
Fire parameters for different fire scenarios were
analyzed. Take smoke and fire temperature as
examples, are shown in Figure 8, Figure 9.
According to Figure 8, the time of smoke
appearance and the time of smoke spreading to the
entire station floor are compared with different fire
scenarios. Fire heat release rate increased quickly
due to the failure of the sprinkler system. So the
smoke first appears earlier in the failure scenarios of
sprinkler system. The smoke spreads throughout the
floor more quickly in the two fires scenarios.
According to Figure 8 and Figure 9, the fire
parameters in the middle of train fire are obvious
higher than the other one fire scenarios. The smoke
spreads throughout the floor earlier than other one
fire scenarios. In particular, the temperature is very
unfavourable to the escape of personnel.
233.6
196
180.8
173.6
180.8
600
612.8
537.6
278.4
417.6
0
100
200
300
400
500
600
700
12346
Smoke appears
Smoke spreads throughout the
floor
Time
(s)
scenarios
Figure 8: the time comparison of smoke appears and
spreads throughout the floor
A Comprehensive Study of Two Fire Conditions in a Subway Train Fire: Considering the Failure or Work of the Sprinkler System
185

435.3
370.4
369.6
329.7
548.8
118
130
116
133
290
0
100
200
300
400
500
600
1
2
3
4
6
scenarios
0
50
100
150
200
250
300
350
the time up to 60
℃
at 2.1m
the maximum temperature
Time
(s)
The maximum
temperature
(
℃
)
Figure 9: the comparison of the time up to 60 ℃ at 2.1m
and the maximum temperature
4 CONCLUSIONS
The change of the fire parameters is not obvious
with less combustible and a larger space of the
subway station. Sprinkler system can reduce the
temperature of the fire, but it will be a slight
hindrance to discharge the fire smoke. Due to the
effect of sprinkler system, the degree of incomplete
combustion of combustible material increases and
the concentration of CO slightly increases within the
dangerous limits. Due to the large amount of
combustibles, train fire source power is larger. Its
fire source parameters change more obvious.
ACKNOWLEDGEMENTS
The authors gratefully acknowledge the support
received for this project from the Scientific Research
Foundation of Liaoning Education Department
(No. L2015096) and the Doctoral Scientific
Research Foundation of Liaoning Province
(No. 201601249).
REFERENCES
Shi C L, Zhong M H, Nong X ZH, et al, 2012. Modeling
and safety strategy of passenger evacuation in a metro
station in China. Safety Science, 50(5): 1319-1332
PAN K, SHI J Y, 2011. Application of Multi-level
Extensible Method to Urban Subway Operation Safety
Evaluation. Journal of China railway society, 33(5):
14-19. (in Chinese).
Luo N, Li An G, Gao R, et al, 2014. Smoke confinement
utilizing the USME ventilation mode
for subway station fire. Safety Science, 70: 202-210
Miho S, Nobuyoshi K, Masato H, 2017. Quantitative
assessment method for road tunnel fire safety:
Development of an evacuation simulation method
using CFD-derived smoke behavior. Safety Science,
94:116-127
Jae S R, Hong S R, Won H P, Yong J J, 2009. CFD
Simulation and assessment of life safety in a subway
train fire[J].Tunnelling and Underground Space
Technology,24:447-453
Sanhay K K, 2017. From small-scale tunnel fire
simulations to predicting fire dynamics in realistic
tunnels[J]. Tunnelling and Underground Space
Technology,61:198-204
Yang P ZH, Li CH, Chen D H, 2013. Fire emergency
evacuation simulation based on integrated fire–
evacuation model with discrete design method.
Advances in Engineering Software, 65: 101-111
ICECTT 2018 - 3rd International Conference on Electromechanical Control Technology and Transportation
186