Infrared Physics & Technology 61 (2013) 265–273

Contents lists available at ScienceDirect

Infrared Physics & Technology

journal homepage: www.elsevier.com/locate/infrared

Fire safety distances for open pool fires

S. Sudheer a, Lokendra Kumar b, B.S. Manjunath b, Amit Pasi c, G. Meenakshi c, S.V. Prabhu a,⇑
a
Indian Institute of Technology, Bombay, Department of Mechanical Engineering, Mumbai 400076, India
b
Bhabha Atomic Research Centre, RTD, Mumbai, Maharashtra, India
c
Centre for Fire, Environment and Explosive Safety, Delhi, India

h i g h l i g h t s

 Infrared thermal camera is employed for the measurement of fire safety distances from diesel, gasoline and hexane pool fires.
 Circular pool fires of diameters 0.5 m, 0.7 m, 1 m and a square diesel pool fire of 4 m length are chosen to study the fire safe distances.
 Heat flux values estimated from infrared camera are validated against the Schmidt-Boelter gauge measurements.
 Hexane pool fires has longer safety distances than diesel pool fires.

a r t i c l e i n f o a b s t r a c t

Article history: Fire accidents that carry huge loss with them have increased in the previous two decades than at any time
Received 12 February 2013 in the history. Hence, there is a need for understanding the safety distances from different fires with dif-
Available online 27 September 2013 ferent fuels. Fire safety distances are computed for different open pool fires. Diesel, gasoline and hexane
are used as fuels for circular pool diameters of 0.5 m, 0.7 m and 1.0 m. A large square pool fire of
Keywords: 4 m  4 m is also conducted with diesel as a fuel. All the prescribed distances in this study are purely
Fire safe distances based on the thermal analysis. IR camera is used to get the thermal images of pool fires and there by
Open pool fires
the irradiance at different locations is computed. The computed irradiance is presented with the thresh-
IR camera
Irradiance
old heat flux limits for human beings.
Heat flux Ó 2013 Elsevier B.V. All rights reserved.

1. Introduction Atallah and Allan [1] studied each individual parameter that
affects the calculation of safety distances. The parameters studied
Fire safety distance is of greater interest in view of the safety were: atmospheric transmissivity, view factor, flame height,
fire fighters, safe distance for building constructions and also in burning rate of the pool fires, flame temperature, flame shape, soot
emergency evacuation of people. Fire safety distance is defined cloud, luminous flame height, flame pulsation, geometric view
as that at which the thermal radiation flux is equal to some pre- factor and wind effects. The authors concluded that some of the
scribed level. This level depends on what is to be conserved or pro- assumptions made in computing the safety distances are too
tected [1]. The estimation of radiation flux from fires is very crucial simplistic and they are to be refined to get accurate results.
and this provides a basis for establishing the safe zones for fire Zarate et al. [2] estimated the thermal radiation emitted by the
fighters and structures. Numerous methodologies, from a very sim- flame front of a wildland fire using the solid flame model together
plistic to complex models, are proposed in measuring the thermal with the view factor calculations. Some of the assumptions made
radiation from fire and hence estimated the fire safety distances in solid flame model are: the shape of the flame is either cuboid
[1–5]. In some of these models, the assumptions made are: or cylindrical and the flame temperature is 1200 K. Billaud et al.
representation of flame by a single temperature, simple flame [3] proposed a methodology based on the solid flame model and
shape representation (like cylinder), radiation from non-visible a Monte Carlo method for the estimation of the radiant heat flux
zones of the fire plume is not accounted, assumptions made in and safety distances from wildland fires. The flame is considered
computing the flame height, etc. as a black surface with a temperature of 1200 K resulting in a heat
flux of 118 kW/m2.
Several authors represented the complete flame by a single rep-
resentative temperature. Butler and Cohen [4] approximated the
⇑ Corresponding author. Address: Department of Mechanical Engineering, Indian
flame as a flat sheet of given height and width with a uniform tem-
Institute of Technology, Bombay, Powai, Mumbai 400 076, India. Tel.: +91 22
25767515; fax: +91 22 25726875. perature of 1200 K and a uniform emissivity of one. As the under-
E-mail address: svprabhu@iitb.ac.in (S.V. Prabhu). standing of the radiation increased, few other authors introduced

1350-4495/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.infrared.2013.09.006
266 S. Sudheer et al. / Infrared Physics & Technology 61 (2013) 265–273

Nomenclature

As surface area (m2) Y vertical distance along the axis of the pool, above the
D pool diameter for circular pool fire, length for square fuel bed (m)
pool fire (m) Z vertical axis
E emissive power (kW/m2)
F1–2 view factor from surface 1 to surface 2 Greek symbols
G factor involved in view factor calculation (m2) ef flame emissivity
q00 heat flux (kW/m2) r Stefan–Boltzmann’s constant (5.67  108 W/m2 K4)
T absolute temperature (K) x, y, z coordinate axis
X radial distance from pool (m) g, n coordinate axis

view factor in computing the fire safety distances. But, in most of the Table 1
studies, the emissivity of the flame is taken as unity, which is not a Consequences of thermal heat flux on human bodies [2].
reasonable assumption. The radiation from the flame is dependent
Heat flux Effects on a human body
on its emissivity. To add to the complexity, emissivity varies (kW/m2)
throughout the flame and also is dependent on the wavelength
1.4 Harmless for persons without any special protection
and diameter of the pool fire. Hayasaka et al. [5] used infrared cam- 1.7 Minimum required to cause pain
era technique to measure the irradiance at a distance of 5 times the 2.1 Minimum required to cause pain after 60 s
length of the pool. The pool size used in their study was 2.7 m2. 4.0 Causes pain after an exposure of 20 s (first degree burns)
The objective of this study is to relax few too simplistic assump- 4.7 Causes pain in 15–20 s and burns after 30 s
7.0 Maximum tolerable value for firefighters completely covered
tions that were made in fire safety distance calculations. The model
protected by special Nomex protective clothes
proposed in this study is easy to use and one can easily calculate 10.0 Certain polymers can ignite
the fire safety distance for a given fire scenario. Experiments are 11.7 Thin steel (partly insulated) can lose mechanical integrity
conducted on pool fires of different sizes. Circular pool fires of 12.6 Wood can ignite after a long exposure; 100% lethality
25.0 Thin steel (insulated) can lose mechanical integrity
diameters 0.5 m, 0.7 m, 1 m and a square pool fire of 4 m length
37.5 Damage to process equipment and collapse of mechanical
are chosen to study the fire safety distances. Diesel, gasoline and structures
hexane are used as fuels for all the circular pool fires and diesel
is used in square pool fire. The temperature profile of the pool fires
are measured using an infrared camera. These temperature profiles
are used in computing the incident heat flux at different locations
from each pool fire.

2. Fire safety distance

Zarate et al. [2] tabulated, from various sources, the conse-

quences of diverse thermal fluxes and thermal radiation threshold
values, as shown in Table 1. It shows the consequences of thermal
radiation on human bodies. It is identified that a person is safe when
he is exposed to heat fluxes less than 1.4 kW/m2. In this region, no
special protection is required for in-ordinately long time. A person
experiences a first degree burn for a heat flux of 4.7 kW/m2 in 30 s.
These two limits are identified in all the measurements in this study.

3. Infrared camera in fire research

An infrared camera is a non-contact device that detects infrared

energy (heat) and converts it into an electronic signal, which then
processed to produce a thermal image on a video monitor and per-
form temperature calculations. Thermoteknix VisIR Ti 640 camera
Fig. 1. Schematic showing the axis and distances in finding the view factor of
is used in this work. As IR-camera transforms the incident radia- rectangle–rectangle in a parallel plane [6].
tion into temperature readings, few parameters like emissivity of
the emitting body, atmospheric temperature between camera
and object are to be necessarily introduced. These parameters
The radiometric image is the representation of the equivalent energy
can be modified while post-processing using the software Ther-
release from the flame volume as if the whole energy is released
Monitor Image Analyzer, provided with the camera. Several images
from a vertical two-dimensional plain which is placed at the center
are captured, in each experiment, to obtain a time averaged image.
of the pool. The thermal energy received by a person who is standing
at some distance from the fire is due to contribution from all the
4. Mathematical modeling pixels of the radiometric image. Emissive power at a pixel in the
radiometric image is given by Stefan–Boltzmann constant:
The radiometric image obtained from the thermal camera is a
two dimensional temperature distribution comprising of pixels. E1 ¼ ref T 41 ð1Þ
 S. Sudheer et al. / Infrared Physics & Technology 61 (2013) 265–273 267

Emissive Power (kW/m2)

110

100
90

80
70

60

Y/D
50

40
30
Fig. 2. Large scale diesel pool fire experiment.
20
10
Temperature (K)
0

X/D
Fig. 4. Emissive power distribution of 4 m  4 m diesel pool fire.
1200
1150
1100
1050 12
1000
950
10 Camera
900
Sensor
850
800
Y/D

Heat Flux (kW/m2)

8
750
700
650
6
600
550
500 4
450
400
350 2

0
0 0.5 1 1.5 2
Y/D
Fig. 5. Comparison of heat flux measurements at Z/D = 1 with sensor and infrared
X/D camera for diesel fire of 0.3 m pool diameter.

Fig. 3. Temperature distribution of 4 m  4 m diesel pool fire. " ( )

1 1=2 yg
G¼ ðy  gÞ½ðx  nÞ2 þ z2  tan1
2p ½ðx  nÞ2 þ z2 
1=2

( )
The heat flux passing through an area, A2 from the pixel A1 is 1=2 xn
þðx  nÞ½ðy  gÞ2 þ z2  tan1 1=2
F 1—2 A1 E1 ½ðy  gÞ2 þ z2 
q001—2 ¼ ð2Þ 
A2 z2
 ln½ðx  nÞ2 þ ðy  gÞ2 þ z2  ð4Þ
View factor, F1–2 is involved in the calculation procedure of heat 2
flux at a distance from the radiometric image. Consider two parallel The assumption made in deriving view factor is that the dis-
rectangles of areas A1 and A2 separated by a distance of z, as shown tance Z is far greater than the sides of the rectangles considered.
in Fig. 1. The view factor as seen from A1 to A2 is given as [6]: f1–2 is different to F1–2, as the earlier represents the view factor be-
1 X 2 X 2 X 2 X 2 tween two rectangular surfaces, but the latter represents the view
f1—2 ¼ ð1ÞðiþjþkþlÞ Gðxi ; yj ; gk ; nl Þ ð3Þ factor between apparent two dimensional pixel to a rectangular
ðx2  x1 Þðy2  y1 Þ l¼1 k¼1 j¼1 i¼1
surface. From previous experiments, it is found that F1–2 is four
268 S. Sudheer et al. / Infrared Physics & Technology 61 (2013) 265–273

kW/m2

40
4.7 kW/m2 35

Height from pool base (Y/D)

1.4 kW/m2
30

25

20

15

10

4.7

2.5

1.4

0
Distance from pool fire (Z/D)
Fig. 6. Fire safety distances for a 0.5 m diameter diesel pool fire.

kW/m2

40
4.7 kW/m2
1.4 kW/m2 35
Height from pool base (Y/D)

30

25

20

15

10

4.7

2.5

1.4

0
Distance from pool fire (Z/D)
Fig. 7. Fire safety distances for a 0.7 m diameter diesel pool fire.

times larger than f1–2 for all the diameters used in this study [7]. 0.7 m, 1 m with diesel, gasoline and hexane as fuels. The pans
Using Eqs. (1) and (3) in Eq. (2), the heat flux passing through an are made up of 2 mm thick mild steel and of 15 cm in height. Fuel
area, A2 from the pixel A1 of pool fire can be calculated. Therefore, is filled into the pans to a height of 10 cm. The wind velocity is less
the net heat flux passing through an area, A2 from the flame repre- than 2 m/s during all the pool fire experiments that are conducted.
sented by n pixels is Further details of these circular pool fire experiments are detailed
in Sudheer and Prabhu [7–10]. The fire safety distances are calcu-
X
n
q00f A2 ¼ ðF i2 A1 Eij Þ=A2 ð5Þ lated from the temperature and emissive power distributions given
i¼1 in Sudheer and Prabhu [7–10].
A large square diesel pool fire experiment is conducted at Cen-
The net heat flux from the flame is measured by moving the po-
tre for Fire, Environment and Explosive Safety, Delhi in India. This
sition of A2 away from the axis of the pool and also moving along
pool is made of brick and cement and of 4 m  4 m in size. 1000
the height of the pool fire.
liters of diesel is filled into the pool above a 75 mm thick layer of
water bed. For the square pool fire of 4 m length, the emissivity
5. Experimental setup is considered as 0.94 from mass burning rate calculations [11].
Fig. 2 shows the still image of large pool fire experiment. Each
Open pool fire experiments are conducted for different pool experiment is conducted for at least 20 min. Thermal images are
sizes and different fuels. Circular pool fires of diameters 0.5 m, captured using VisIR 640, thermal camera. The images are time
 S. Sudheer et al. / Infrared Physics & Technology 61 (2013) 265–273 269

kW/m2

40
4.7 kW/m2
1.4 kW/m2 35

Height from pool base (Y/D)

30

25

20

15

10

4.7

2.5

1.4

0
Distance from pool fire (Z/D)
Fig. 8. Fire safety distances for a 1.0 m diameter diesel pool fire.

kW/m2

40
4.7 kW/m2
1.4 kW/m2 35
Height from pool base (Y/D)

30

25

20

15

10

4.7

2.5

1.4

0
Distance from pool fire (Z/D)
Fig. 9. Fire safety distances for a 0.5 m diameter gasoline pool fire.

averaged to get the average temperature distribution of pool fire. get time averaged heat flux at that location. Fig. 5 shows the
The flame temperature distribution is shown in Fig. 3. It is ob- comparison of heat flux measurements at Z/D = 1 with heat flux
served that a maximum temperature of around 1000 °C at a height transducer and camera for 0.3 m diameter pool fire. There is a rea-
of 0.6 m axially. The emissive power distributions are computed sonably good agreement between the computed results using Eq.
using Eq. (1). Fig. 4 shows the emissive power distribution of the (5) and measurements using heat flux gauge. Sudheer and Prabhu
large square pool fire. Using Eq. (5), the heat fluxes at different [7] conducted a similar experiment on gasoline pool fire and found
locations from the pool fire are calculated. The area, A2 in Eq. (5) that the heat flux measurements using IR camera are compared
is taken 1 cm2 in all the calculations, i.e., the heat fluxes mentioned well with the Schmidt–Boelter gauge.
in this study are averaged over an area of 1 cm2.
7. Results and discussions
6. Validation of Heat flux measurements at a distance
The temperature distributions of pool fires are captured using
Experiments are conducted on a diesel pool fire of 0.3 m diam- an infrared thermal camera. Experiments are conducted on circular
eter using Schmidt–Boelter gauge. The heat flux gauge is traversed pool fires of diameters 0.5 m, 0.7 m and 1.0 m and a square pool of
along the vertical axis from a distance of 0.3 m from the pool cen- length 4 m. Figs. 6–15 show the irradiance from pool fires of differ-
ter. In each position measurements are made for a period of 30 s to ent sizes. The threshold values of 1.4 kW/m2 (dotted line) and
270 S. Sudheer et al. / Infrared Physics & Technology 61 (2013) 265–273

kW/m2

40
4.7 kW/m2
35
1.4 kW/m2

Height from pool base (Y/D)

30

25

20

15

10

4.7

2.5

1.4

0
Distance from pool fire (Z/D)
Fig. 10. Fire safety distances for a 0.7 m diameter gasoline pool fire.

kW/m2

40
4.7 kW/m2
35
1.4 kW/m2
Height from pool base (Y/D)

30

25

20

15

10

4.7

2.5

1.4

0
Distance from pool fire (Z/D)
Fig. 11. Fire safety distances for a 1.0 m diameter gasoline pool fire.

4.7 kW/m2 (solid red line1) are marked in all Figs. 6–15. The abscissa availability of oxygen. The presence of soot and the availability of
represents the distance from the pool center and the ordinate repre- oxygen are not considered in this study while prescribing the fire
sents the height above the cylinder tip. The abscissa starts from 0.5D, safety distances and is out of the scope of this work.
i.e., the tip of the mild steel cylinder. Figs. 6–8 show the fire safety distances from circular diesel pool
It is observed from Figs. 6–15, that the irradiance decreases as fires. For all the circular diesel pool fires, the safety zone, on the
one moves away from the pool fire and is obvious as the farther ground, is around 5.7D away from pool fire. The maximum heat
the distance, lesser the influence of radiation from the fire. As flux received from these pool fires is around 45 kW/m2. The safety
the distance increases, the view factor from the flame to the object zone in view of height is varying. Figs. 8–10 are the computed fire
decreases and hence, decreases the radiation. The maximum heat safety distances for gasoline pool fires of diameters 0.5 m, 0.7 m
flux is observed at the tip of the pool surface to a height of 1D. and 1.0 m. For all the circular gasoline pool fires the safety zone,
Though the safety zones along the height are shown in this study, on the ground, is varying from 5D to 7D away from pool fire. The
care is to be taken while considering the presence of soot and the temperatures and the emissive powers of the gasoline fires are ob-
served to be higher than that of the corresponding diesel pool fires.
Figs. 12–14 are the computed fire safety distances for hexane
1
For interpretation of color in Figs. 6–15, the reader is referred to the web version pool fires of diameters 0.5 m, 0.7 m and 1.0 m. For all the circular
of this article. hexane pool fires the safety zone, on the ground, is varying from
 S. Sudheer et al. / Infrared Physics & Technology 61 (2013) 265–273 271

kW/m2

40
4.7 kW/m2
35
1.4 kW/m2

Height from pool base (Y/D)

30

25

20

15

10

4.7

2.5

1.4

0
Distance from pool fire (Z/D)
Fig. 12. Fire safety distances for a 0.5 m diameter hexane pool fire.

kW/m2

40
4.7 kW/m2
1.4 kW/m2 35
Height from pool base (Y/D)

30

25

20

15

10

4.7

2.5

1.4

0
Distance from pool fire (Z/D)
Fig. 13. Fire safety distances for a 0.7 m diameter hexane pool fire.

4.5D to 9.5D away from pool fire. The temperatures and the emis- fires. This makes more fuel to burn and release more energy at a
sive powers of the hexane fires are observed to be higher than that given instant.
of the corresponding diesel and gasoline pool fires. For the large A person with special fire safety protection dress can go nearer
square pool fire (Fig. 15), the person has to stand at least 8D away to the flame from safety zone, basing on the heat flux rating of the
from pool fire. The maximum heat flux observed is around 75 kW/ dress. Care must be taken before leaving the safety zone (1.4 kW/
m2 at the tip of the pool surface. m2). One has to check for any kind of exposure of body due to wear
Table 2 shows the fire safety distances for different pool fires on or tear of the protective equipment. More precautions should be
ground level. It is observed that the hexane fuel is more dangerous taken in case of high wind velocities. Flame gets tilted because of
than the other fuels that are considered in this study. Table 3 the wind and hence the person or the instrumentation or the ob-
shows the boiling point temperatures and the mass burning rates jects may get affected, even if they are in the safety zone. The
of different pool fires. Boiling point temperature is a property of safety limits prescribed in this work are purely thermal with no
the fuel and is independent of the pool size, where as the mass consideration of the presence of soot that pollutes the air nor the
burning rate depends on the size. Hexane has a low boiling tem- absence of oxygen that is essential for human beings.
perature as compared to diesel and gasoline and hence evaporates Using these plots, one can decide on the placement of instru-
even at lower temperatures. Hence, the mass burning rates of hex- mentation or any kind of equipment or sensor. Equipments with
ane are higher than the of the corresponding diesel or gasoline pool low operating temperatures should be kept in the safety zone.
272 S. Sudheer et al. / Infrared Physics & Technology 61 (2013) 265–273

kW/m2

4.7 kW/m2 40
1.4 kW/m2
35

Height from pool base (Y/D)

30

25

20

15

10

4.7

2.5

1.4

0
Distance from pool fire (Z/D)
Fig. 14. Fire safety distances for a 1.0 m diameter hexane pool fire.

kW/m2

70
4.7 kW/m2
1.4 kW/m2 60
Height from pool base (Y/D)

50

40

30

20

10

4.7

2.5

1.4

0
Distance from pool fire (Z/D)
Fig. 15. Fire safety distances for a 4 m length square diesel pool fire.

Table 2 Table 3
Fire safety distances for different pool fires. Boiling point temperature and mass burning rate of different pool fires.

Fuel Pool diameter/length, D Fuel Boiling point (°C) [12] Pool diameter/length, D
0.5 m 0.7 m 1.0 m 4.0 m 0.5 m 0.7 m 1.0 m 4.0 m
Diesel 5.5D 5.5D 5.7D 7.8D Mass burning rate (gm/m2 s) [7–9]
Gasoline 5.7D 6.3D 7.1D
Diesel 250 27.5 32.7 39.1 55.0
Hexane 4.5D 6.7D 9.2D
Gasoline 155 45.0 51.1 54.2
Hexane 68 47.6 57.2 59.8

Proper insulation is required for all those instruments or equip-

ments that are placed in high heat flux zones.
using infrared thermal camera. For all the circular pool fires the
8. Conclusions safety zone to stand or operate is around 5.7D for diesel fire, 5D
to 7D for gasoline fire and 4.5D to 9.5D for hexane fire away from
Fire safety distances for circular pool fires of 0.5 m, 0.7 m and pool fire. The fire safety distance for large pool fire (4 m  4 m) is
1.0 m diameters and a square pool fire of length 4 m are computed around 8D away from pool fire.
 S. Sudheer et al. / Infrared Physics & Technology 61 (2013) 265–273 273

References [7] S. Sudheer, S.V. Prabhu, Physical experiments and FDS simulations on gasoline
pool fires, J. Fire Sci. (2013), http://dx.doi.org/10.1177/0734904112472953.
[8] S. Sudheer, S.V. Prabhu, Measurement of flame emissivity of hydrocarbon pool
[1] S. Atallah, D.S. Allan, Safe separation distances from liquid fuel fires, Fire
fires, Fire Technol. 48 (2012) 183–217.
Technol. 7 (1971) 47–56.
[9] S. Sudheer, S.V. Prabhu, Measurement of flame emissivity of gasoline pool fires,
[2] L. Zarate, J. Arnaldos, J. Casal, Establishing safety distances for wildland fires,
Nucl. Eng. Des. 240 (2010) 3474–3480.
Fire Safety J. 43 (2008) 565–575.
[10] S. Sudheer, S.V. Prabhu, Characterization of hexane pool fires using infrared
[3] Y. Billaud, A. Kaiss, J.L. Consalvi, B. Porterie, Monte Carlo estimation of thermal
thermography, J. Fire Sci. 31 (2012) 143–165.
radiation from wildland fires, Int. J. Therm. Sci. 50 (2001) 2–11.
[11] V. Babrauskas, Estimating large pool fire burning rates, Fire Technol. 19 (1983)
[4] B.W. Butler, J.D. Cohen, Firefighter safety zones: a theoretical model based on
251–261.
radiative heating, Int. J. Wildland Fire 8 (1988) 73–77.
[12] Society of Fire Protection Engineers, SFPE Handbook of Fire Protection
[5] H. Hayasaka, H. Koseki, Y. Tashiro, Radiation measurements in large-scale
Engineering, fourth ed., Natl Fire Protection Assn., 2008.
kerosene pool flames using high-speed thermography, Fire Technol. 28 (1992)
110–112.
[6] J.R. Howell, A Catalog of Radiation Heat Transfer, http://thermalhub.org/
resources/32, 2008.