You are on page 1of 7

Journal of Loss Prevention in the Process Industries 24 (2011) 568e574

Contents lists available at ScienceDirect

Journal of Loss Prevention in the Process Industries


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

Engineering layout of fuel tanks in a tank farm


Angan Sengupta*, A.K. Gupta 1, I.M. Mishra
Department of Chemical Engineering, Indian Institute of Technology Roorkee, Roorkee 24766, Uttarakhand, India

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

Article history: The present paper deals with the location of tanks in a tank farm, in chemical and allied industries.
Received 17 December 2009 Ideally the tanks are so placed and installed that in case of fire, the neighbouring tanks could remain safe.
Received in revised form The safe distance of separation among the tanks is calculated in no wind condition, as well as, in the
21 May 2010
presence of wind. The paper uses the methods available in literature and modifies the point source
Accepted 22 June 2010
model to include the effect of wind vector on the flame height during the calculation of safe inter-tank
distance. It is found that for wind velocity > 4 m/s, the modified point source model provides appropriate
Keywords:
inter-tank distance. However, for no wind and with wind velocity < 4 m/s, the ShokrieBeyler’s method
Tank farm
Wind effect
provides safe inter-tank distance.
Cross-wind Ó 2010 Elsevier Ltd. All rights reserved.
Safe separation distance
Modified point source model

1. Introduction to property worth 3  103 million INR (w75 million USD) is a case in
point. This fire caused very serious environmental pollution around
Tank farm is a synonym of an oil depot, a facility for storage of Jaipur and its adjoining areas. This incident has underlined the
liquid chemicals, such as oils, gasoline, diesel, aviation turbine fuel, importance of proper layout with safe separation distance to
solvents, petrochemicals, etc. The tank farm is a piece of land on prevent such hazardous episodes.
which a number of fuel oil or chemical storage tanks are located or The tank farms usually have a number of tanks of equal heights.
sited together. The storage tanks may also be used to store base However, they may be of varying capacities. Equal height adds to
blending components, solvents, additives, acids, caustic, chemicals, aesthetics and lends them economic credence in the construction
or finished products. They may also be used as blending vessels. of access structures, easy movements of operators from one tank to
Storage tanks are expensive to build and they require periodic another and of fire service men during emergency episodes. The
maintenance to keep them in proper condition when storing tanks may be laid out in square pitch having safe inter-space
volatile and flammable liquids. It is, therefore, necessary that they between them. The inter-tank optimum separation distance is
be properly sized and utilized to maximize the return on crucial for safe operations, piping design, and maintenance access
investment. and emergency/accident control and mitigation measures. The
The safety aspects, apart from periodic maintenance, are desired inter-tank distance depends largely on the materials/
extremely important. The recent incident at Jaipur oil depot of chemicals to be stored and the capacity of the tanks.
Indian Oil Corporation Ltd. where 12 tankers containing 105 kl of The tanks in refineries are generally constructed from steel or
diesel and gasoline caught fire and the fire continued for a week polyethylene or fiber glass and are either having fixed roofs and/or
resulting in several fatalities and about 200 injuries besides damage floating roofs for storing liquids. Non-steel constructions lower cost
considerably, and make them preferred choice for storing corrosive
and reactive chemicals.
Various regulatory and professional bodies like American
Petroleum Institute (API), National Fire Protection Association
* Corresponding author. Tel.: þ91 9012390134; þ91 9433948999. (NFPA), and Environmental Protection Agency (EPA) have sug-
E-mail addresses: angan.sengupta@gmail.com (A. Sengupta), akgd30@ gested standards for such tank layouts in a tank farm. The layout of
indiatimes.com, akgd30@gmail.com (A.K. Gupta), imishfch@iitr.ernet.in
tanks as distinct from their spacing, should always take into
(I.M. Mishra).
1
Scientist and former Head, Fire Research Laboratory, Central Building Research consideration the accessibility needed for fire-fighting and the
Institute, Roorkee, India. potential value of a storage tank farm in providing a buffer area

0950-4230/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jlp.2010.06.016
A. Sengupta et al. / Journal of Loss Prevention in the Process Industries 24 (2011) 568e574 569

between process plant and public roads, houses, etc., for environ- strength to withstand the pressure that may be created in the event
mental reasons. The location of tanks relative to process units must of an oil spillage and the capacity to store the spilled liquid.
be such as to ensure maximum safety from possible accidents. Some methods are available in literature for modeling inter-tank
Fig. 1a, b shows schematic representation of tanks in a tank farm safe distance in a tank farm. However, these methods do not take
laid out in square and triangular pitch. into account the effect of cross-wind which is important from fire
Primary requirements for the layout of refinery tanks farms are spread point of view. The present paper deals with the assessment
summarized as follows (Digrado & Throp,1995; Long & Gardner, 2004): of various methods to estimate the safe separation distance
between two storage tanks in a tank farm, from the fire spread
1. Inter-tank spacing and separation distances between tank and point of view. A modification is proposed to the point source model
boundary line and tank and other facilities are of fundamental to incorporate the effect of wind velocity on the flame tilt and fire
importance. spread. The estimated separation distance is also compared with
2. Suitable roadways should be provided for approach to tank the values given by various regulatory bodies and prescribed by
sites by mobile fire-fighting equipment and personnel. standards and reported in the present paper.
3. The fire water system should be laid out to provide adequate
fire protection to all parts of the storage area and the transfer 2. Models to estimate the safe inter-tank spacing
facilities.
4. Bunding and draining of the area surrounding the tanks should be Following methods are generally used for the determination of
such that a spillage from any tank can be controlled to minimize safe inter-tank spacing:
subsequent damage to the tank and its contents. They should also
minimize the possibility of other tanks being involved. (a) Point source model.
5. Tank farms should preferably not be located at higher levels (b) ShokrieBeyler’s method.
than process units in the same catchment area. (c) Mudan’s method.
6. Storage tanks holding flammable liquids should be installed in
such a way that any spill will not flow towards a process area or Using these methods, the heat flux at various distances,
any other source of ignition. between a tank on fire and the adjacent (target) tank is calculated.
The distance at which the heat flux becomes equal to 4.732 kW/m2
A key safety consideration for tank farm siting, spacing, and (1500 BTU/h/ft2) (Daniel, Crowl, & Louvar, 2002; Lees, 1995; SFPE
location is the separation of non-compatible materials by the use of Handbook of Fire Protection Engineering, 1995) is considered to
an internal bund or dike wall within the tank farm. Providing bund be the safe inter-tank distance. No material is expected to ignite
or dike checks the flow of the spilled oil to the neighbouring areas. with a heat flux lower than 4.732 kW/m2.
Thus in case of fire engulfing the tank farm, the fire is confined to its
origin. The bunds, however, need to be designed to have sufficient
2.1. Point source model

In this model, it is customary to model the flame by a point


source located at the center of the real flame in order to predict the
thermal radiation field of flames. The point source model (SFPE
Handbook of Fire Protection Engineering, 1995) is the simplest
configurational model of a radiant source. Fig. 2 shows the sche-
matic diagram of two tanks for using point source model. The
critical value of incident heat flux, defined as the minimum value of
the heat flux which can ignite the fuel in the target tank is given as

Qr cos q  
q_ 00r ¼ kW=m2 (1)
4pR2

Hf

Hf/2 R

D
L
Fig. 1. (a) Top view of a tank farm square pitch layout. (b) Top view of a tank farm
triangular pitch layout. Fig. 2. Schematic diagram of a tank on fire for point source model.
570 A. Sengupta et al. / Journal of Loss Prevention in the Process Industries 24 (2011) 568e574

Here, Qr is obtained external targets in terms of average effective emissive power of the
flame. The flame is assumed to be a cylindrical blackbody and
Qr ¼ lQ (2) a homogeneous radiator with an average emissive power.
Here, l can be determined as The radiative heat flux is given as

l ¼ ð0:21  0:0034  DÞ (3) q_ 00r ¼ E  F12 (9)

And D is the diameter of tank (m). The emissive power is given as


Q is the total heat released by fire and can be estimated as
follows: E ¼ 58  100:00823D (10)

p where, D is the diameter of pool (m). The view factor (F12) is


_ 00  D2  DHc
Q ¼ hm (4) a function of the target location, flame height and pool diameter;
4
and lies between 0 and 1. The shape factor is determined as
where, h is the combustion efficiency. m
_ 00 is calculated from the follows:

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
B  1=s 1 ðB þ 1Þðs  1Þ A  1=s 1 ðA þ 1Þðs  1Þ
F12;H ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffitan  pffiffiffiffiffiffiffiffiffiffiffiffiffiffitan
p B 1 2 ðB  1Þðs þ 1Þ p A 1 2 ðA  1Þðs þ 1Þ
  rffiffiffiffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 1 h h 1 s1 Ah 1 ðA þ 1Þðs  1Þ (11)
F12;V ¼ tan p ffiffiffiffiffiffiffiffiffiffiffiffiffi  tan þ p ffiffiffiffiffiffiffiffiffiffiffiffiffiffitan
p$s s 21 p$s s þ 1 p$s A2  1 ðA  1Þðs þ 1Þ
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
F12 ¼ F12;H þ F12;V
2 2

regression rate curve (Drysdale, 1985). Regression rate is the where,


volumetric loss of liquid per unit pool surface area, and is given as
follows:
A ¼ h þs þ1
2 2
2s
rðRN Þ QE00 B ¼ 1þs
2

_ 00 ¼
m þ (5) 2s
60; 000 DHV s ¼ 2L
D
2Hf
Here, r is the fuel density, RN is the regression rate, QE00 is the h ¼ D
external incident radiative heat flux and DHV is the heat of vapor-
ization. The value of 60,000 in equation (5) is only to convert
regression rate from mm/s to m/min. Here, Hf (m) is given by equation (7).
R and cos q in equation (1), are given as follows: Substituting values of all the parameters in equation (9) and
calculating q_ 00r for various distances, for each of the configurations,
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 2 one can obtain the safe distance corresponding to
Hf
R ¼ þL2 (6) q_ 00rc ¼ 4:732 kW=m2 .
2
where, Hf is the flame height (m) above the tank. The flame height 2.3. Mudan’s method
is obtained from Heskestad relationship (Heskestad, 1984) as given
below: Mudan (Mudan, 1984) has also presented a method for esti-
mating thermal radiation from pool fires. The thermal radiation
Hf ¼ 0:235Q 2=5  1:02D (7) intensity to an element outside the flame envelope is given by the
following equation:
and, cos q is given as
q_ 00r ¼ E  F12  s (12)
L
cos q ¼ (8)
R The effective emissive power is given by
where, L is the inter-tank separation distance measured from the
E ¼ 140 expð0:12DÞ þ 20½1  expð0:12DÞ (13)
center of the source tank to the edge of the target tank.
Substituting values of all the parameters in equation (1) and F12 can be determined in the same manner as has been done in
calculating q_ 00r for various distances, for each of the configurations, case of Shokri and Beyler method. Transmissivity (s) varies between
one can obtain the safe distance corresponding to the critical heat 0 and 1 and can be determined as follows (Daniel et al., 2002; Lees,
flux, q_ 00rc which is generally taken as equal to 4.732 kW/m2 (Daniel 1995).
et al., 2002; Lees, 1995; SFPE Handbook of Fire Protection
Engineering, 1995). s ¼ 2:02  ðPw  xÞ0:09 (14)
where, Pw is partial pressure of water vapour in air (Pa) and x is the
2.2. Shokri and Beyler’s method path length (m).
The flame height correlation used in this method is based on the
Shokri and Beyler (SFPE Handbook of Fire Protection correlation of average mean visible flame height, Hf, of turbulent
Engineering, 1995) have developed a method for prediction of diffusion flames developed by Thomas (Attalah and Donald, 1970;
radiative heat flux from pool fires. A relationship has been devel- Mudan, 1984; Thomas, 1963). The flame is assumed to be cylin-
oped to correlate the experimental data of flame radiation to drical in shape and Hf is given as
A. Sengupta et al. / Journal of Loss Prevention in the Process Industries 24 (2011) 568e574 571

!0:61 The corrected inter-tank safe distance, Lcorr, is obtained as


_ 00
m
Hf ¼ 42  D  pffiffiffiffiffiffiffiffiffiffiffiffi (15)  
ra  g  D Hf
Lcorr ¼ L  sin f (19)
2
2.4. Modification of point source model
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 2
Hf
As a result of movement of air in the atmosphere, i.e. in the R ¼ þL2corr (20)
presence of wind, the flame does not remain vertical any more. The
2
flame gets tilted as shown in Fig. 3 and its spread and heat transfer
from the flame to the target tank gets affected by wind velocity Lcorr
cos q ¼ (21)
vector. The point source model which is applicable to vertical R
flames under no wind condition, therefore, becomes invalid under Q, Qr, q_ 00rc and l are calculated as in equation (1). m
_ 00 is calculated by
windy conditions. Hence, the model requires modification to equation (5). Equations (19)e(21) may be used to estimate the safe
accommodate the effect of wind. The flame tilt (f) is calculated as distance of separation under cross-wind conditions.
follows:
 3. Results and discussions
1 pfor
ffiffiffiffiffi u*  1
cos f ¼ (16)
1= u* for u*  1
In order to illustrate the usefulness of the three methods as
where, f is the flame tilt with vertical as shown in Fig. 3 and is given described above and to compare the results obtained from them,
by the American gas association (AGA) (SFPE Handbook of Fire a computer code has been developed to calculate safe distance in
Protection Engineering, 1995). a tank farm. The set of data that have been used to determine safe
u* is the dimensionless speed and is given by inter-tank separation distance under wind and no wind conditions
are as follows:
u
u* ¼ (17) Case I: gasoline is stored in the tanks. The relevant data for the
_ 00 D=rv Þ1=3
ðg m tanks and the materials are given as follows:

where, g is the acceleration due to gravity (9.81 m/s2), rv is the fuel


1. D ¼ 10 m
vapour density (kg/m3) and m _ 00 is the mass burning rate per unit
2. DT ¼ 10 m
pool area (kg/m2/s).
3. H 1 ¼ H2 ¼ 7 m
The height of the tilted flame can be calculated by the rela-
4. h¼1
tionship given by Thomas (1963).
5. r ¼ 7900 kg/m3
!0:67 6. cp ¼ 510 J/kg/K
_ 00
m  0:21
Hf ¼ 55  D pffiffiffiffiffiffiffiffiffiffiffiffi  u* (18a) 7. hg ¼ 620 W/m2/K
ra  g  D 8. Kg ¼ 0.15 W/m/K
00
9. QE ¼ 190 W/m2 (Solar Radiation Handbook, 2008)
Moorhouse has proposed a relationship for tilted flame for LNG
(Mudan, 1984) as Case II: LNG is stored in the tanks. The relevant data for the tanks
!0:254 and the materials are given as follows:
_ 00
m  0:044
Hf ¼ 62  D  pffiffiffiffiffiffiffiffiffiffiffiffi  u* (18b)
ra  g  D 1. D ¼ 10 m
2. DT ¼ 10 m
3. H 1 ¼ H2 ¼ 7 m
4. h¼1
5. r ¼ 424 kg/m3
6. cp ¼ 2340 J/kg/K
7. hg ¼ 75 W/m2/K
Hf 8. Kg ¼ 0.031 W/m/K
00
Wind 9. QE ¼ 190 W/m2 (Solar Radiation Handbook, 2008)
R
Table 1 provides calculated values of various parameters which
have been obtained using the calculated with the help of computer
code. It is found that the flame height calculated by Thomas method
(Thomas, 1963) is lower than that calculated by Heskestad method
H
Lcorr Table 1
Rate of burning and flame heights under no wind conditions.

Gasoline LNG

D Volume of source tank ¼ 549.78 m3 Volume of source tank ¼ 549.78 m3


Volume of target tank ¼ 549.78 m3 Volume of target tank ¼ 549.78 m3
Regression rate ¼ 3.8 mm/min Regression rate ¼ 6.6 mm/min
m_ 00 ¼ 0:04384 kg=m2 =s m_ 00 ¼ 0:046147 kg=m2 =s
Total heat radiated ¼ 28,360.986 kW Total heat radiated ¼ 31,894.4531 kW
L Flame height (Heskestad) ¼ 18.24 m Flame height (Heskestad) ¼ 19.61 m
Flame height (Thomas) ¼ 13.75 m Flame height (Moorhouse) ¼ 151.24 m
Fig. 3. Schematic diagram for tank on fire under cross-wind condition.
572 A. Sengupta et al. / Journal of Loss Prevention in the Process Industries 24 (2011) 568e574

Table 2 26
Comparison of values of safe separation distance with various standards and models.
24
Regulatory bodies Rim to rim distance
between tanks (for class e I fuel) (m)
22
NFPA 3.33

Radiative Heat Flux (kW/m^2)


EPA 30.48 20
BN-DG-C01J plant 10e15
layout-storage tanks 18
API 15
Point source model Gasoline: 13.6 16
LNG: 14.6
ShokrieBeyler method Gasoline: 16.25 14
LNG: 20.35
Mudan’s method Gasoline: 12.5 12
LNG: 15
10

(Heskestad, 1984) in the case of gasoline. In the case of LNG, flame 8

height as calculated by Heskestad method provides a lower value


6

15 4.7261
than that by Moorhouse (Mudan, 1984).

7
4

72
30

4.
Table 2 shows the calculated results obtained using the above

.7
4

5,
14.6,
,4

.3
20
methods. Figs. 4 and 5 show the incident heat flux at various
distances and the safe distance between the tanks when a tank is 2

on fire under no wind condition, for gasoline and LNG, respectively.


0
It is clear that under no wind condition, heat flux decreases with an 0 5 10 15 20 25 30
increase in distance of separation between the tanks. While Shok- Distance from Rim (m)
rieBeyler method gives the highest value of safe distance, point
source and Mudan’s methods give very close values, but lower than Point Source: ; Shokri-Beyler: ; Mudan’s Method:

that given by ShokrieBeyler method. It may be seen from Table 2, Fig. 5. Radiative heat flux versus distance of separation from rim in no-wind condition
that the values of the safe distances as proposed by different for LNG.
standards are widely different; the most conservative and the
highest estimate being that given by EPA, while the lowest is that of From Table 2, it is observed that LNG storage tanks require larger
NFPA (NFPA-30, 2001, Chap. 4). The value recommended by API is separation distance than that for gasoline, in the absence of wind
near to that estimated by point source model and the Mudan’s effect. This is obvious since LNG (flash point ¼ 148.89  C) is more
method for LNG fire. For gasoline, the values obtained by point flammable than gasoline (flash point ¼ 42.7  C).
source model and ShokrieBeyler method are nearer to API value.
56
54
32
52
30 50
48 Wind Speed (m/s): 4 ;6 ;8 ;10 ; 12
28
46
26 44
42
24
40
Radiative Heat Flux (kW/m^2)

22 38
36
Radiative heat Flux (kW/m^2)

20 34
32
18
30
16 28
26
14
24
12 22
20
10
18

8 16
14
6 12
9
,4 1
22

72
.6 .7 3
.7

10
4.
13 , 4

4
5,
.5

.2
12

8
16

2 6
4
0
0 5 10 15 20 25 30
2
0
Distance from Rim (m) 0 5 10 15 20 25 30
Point Source: ; Shokri-Beyler: ; Mudan’s Method: Distance from Rim (m)

Fig. 4. Radiative heat flux versus distance of separation from rim in no-wind condition Fig. 6. Radiative heat flux versus distance of separation from rim in cross-wind
for gasoline. condition for gasoline, using modified point source model.
A. Sengupta et al. / Journal of Loss Prevention in the Process Industries 24 (2011) 568e574 573

50 4. Conclusions
48
46
The layout of a tank farm for storage of volatile and flammable
44
substances is very important. Although some methods are available
42 Wind Speed (m/s):4 ;6 ;8 ; 10
40
for determining the minimum distance of separation of one tank
38 from the other, these methods do not take into account the effect of
36 wind on the flame height and fire spread. The point source model
Radiative Heat Flux (kW/m^2)

34 has been modified to incorporate the effect of wind velocity on the


32 flame height and fire spread.
30 Shokri and Beyler method gives a higher safe distance in
28
comparison to other methods under no wind condition. The radiative
26
heat flux increases under windy condition, the safe distance of
24
22
separation between the tanks will, therefore, also increase. The
20 modified point source model can be used for determining this
18 increased safe distance. However, for wind speed below 4 m/s, the
16 safe distance of separation estimated by ShokrieBeyler’s method
14 may be maintained between the tanks. For wind speed above 8 m/s,
12
safe distance of separation decreases, as the flame gets almost flat-
10
tened at this condition. Thus, modified point source model, proposed
8
6
in this paper should be used for tank farm layout, when the wind
4 speed, in general, remains above 4 m/s. It is found that square pitch
2 may be used for tank farm layout from safety point of view.
0
0 5 10 15 20 25
Nomenclatures
Distance from Rim (m)
cp: specific heat of fuel (kJ/kg/K).
Fig. 7. Radiative heat flux versus distance of separation from rim in cross-wind
D: diameter of the pool or source tank (m).
condition for LNG, using modified point source model.
DT: target tank diameter (m).
E: emissive power (kW/m2).
F12: view factor.
Figs. 6 and 7 show the variation of incident radiative heat flux g: acceleration due to gravity (9.81 m/s2).
with distance, while incorporating the wind effect for gasoline and h0 : difference in height of the two tanks (m) ¼ H1  H2.
LNG, respectively. Table 3 gives the values of safe distance as DHc: heat of combustion of fuel (kJ/kg).
calculated by modified point source model, incorporating the effect Hf: flame length/height (m).
of wind on flame. hg: convective heat transfer coefficient of fuel (W/m2/K).
It is seen from Table 3 that the height of the flame decreases HR: Hf  h0 (m).
with an increase in the wind speed while the tilt angle with respect DHV: heat of vaporization (kJ/kg).
to vertical axis increases with an increase in wind speed. As a result H1: height of source tank (m).
of the decrease of flame height, it is expected to have a decrease in H2: height of target tank (m).
the safe distance between the two tanks. But due to tilting of the Kg: thermal conductivity of fuel (W/m/K).
flame, the safe distance is increasing up to a wind speed of 8 m/s. L: inter-tank distance measured from the center of source tank
Thereafter, it starts decreasing, due to flattening of the flame. to the edge of target tank (m).
Thus from Figs. 6 and 7, it is clear that the wind speed of 8 m/s is Lcorr: corrected distance to obtain q (m).
the critical speed, for which we require maximum safe distance of m _ 00 : mass burning rate per unit pool area (kg/m2 s).
separation. Beyond this speed of the wind the flame becomes Pw: partial pressure of water vapour in air (Pa).
almost flat thus radiating less heat at a particular location. Q: total heat produced by the fire (kW).
00

The square pitch layout of the tanks in the tank farm requires QE: incident radiative heat flux from external source (such as
more area per tank, than that required for the triangular pitch. sun) (kW/m2).
However, the tank farm layout in square pitch is better for passage Qr: total radiative energy output from the fire (kW).
in between the tanks and for smooth maintenance and control q_ 00rc : critical value of incident heat flux (taken as 4.732 kW/m2).
work. R: hypotenuse from flame center to target tank top edge (m).
RN: regression rate (mm/min).
u: wind speed (m/s).
Table 3 u*: dimensionless wind speed as given in equation (17).
Variation of safe distance with wind speed (modified point source model). x: path length (m) ¼ (hypotenuse of fire elevation from
Fuel Wind speed (m/s) Flame height (m) Tilt angle ( ) Safe distance (m) ground)  (radius of fire).
Gasoline 4 9.53 60.62 12.23
l: fraction of total heat which is radiated.
6 8.76 66.38 14.42 h: efficiency of combustion.
8 8.24 69.70 18.52 f: flame tilt angle from vertical axis ( ).
10 7.86 71.92 17.59 q: angle between the normal to the target and the line of sight
12 7.51 73.54 12.71
from the target to the point source location.
LNG 4 10.63 54.17 13.55 r: density of fuel (kg/m3).
6 9.76 61.45 13.60 ra: air density (kg/m3).
8 9.19 65.55 19.36
10 8.77 71.92 14.25
rv: fuel vapour density (kg/m3).
s: transmissivity of air.
574 A. Sengupta et al. / Journal of Loss Prevention in the Process Industries 24 (2011) 568e574

References Lees, F. P. (1995) (2nd ed.).Loss prevention in process industries, Vol. 3 Butterworth
Heinemann.
Long, B., & Gardner, B. (2004). Guide to storage tanks and equipment. Wiley.
Attalah, S., & Donald, S. A. (1970). Safe separation distances from liquid fuel fires. In
Mudan, S. K. (1984). Thermal radiation hazards from hydrocarbon pool fires.
Central States Section of the Combustion Institute meeting on disaster hazards held
Progress in Energy and Combustion Science, 10, 59e80.
at NASA’s Manned Spacecraft Center, Houston, Texas (pp. 47e56).
NFPA-30. (2001). Aboveground tank installation for tank storage. Quincy, Massa-
Daniel, A., Crowl, J., & Louvar, F. (2002). Chemical process safety e fundamentals
chusetts: National Fire Protection Association.
with applications. In B. Goodwin (Ed.) (2nd ed..).Prentice Hall international
SFPE handbook of fire protection engineering. (1995) (2nd ed.). Quincy. Massachu-
series.
setts: National Fire Protection Association.
Digrado, D. B., & Throp, A. G. (1995). The aboveground steel storage tank handbook.
Solar radiation handbook. (2008). Typical climatic data for selected radiation stations.
Wiley.
A joint project of Solar Energy Center, MNRE, Indian Metrological Department.
Drysdale, D. (1985). An introduction to fire dynamics (2nd ed.). Wiley.
Thomas, P. H. (1963). The size of flames from natural fires. In 9th International
Heskestad, G. (1984). Engineering relations for fire plumes. Fire Safety Journal, 7,
combustion symposium. Pittsburgh, PA: Combustion Inst.
25e32.

You might also like