Journal of Loss Prevention in the Process Industries 25 (2012) 114e126

Contents lists available at ScienceDirect

Journal of Loss Prevention in the Process Industries

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

Parametric approach of the domino effect for structural fragments

Dongliang Sun a, b, Juncheng Jiang a, *, Mingguang Zhang a, Zhirong Wang a,
Guangtuan Huang b, Jianjiang Qiao b
a

Jiangsu Key Laboratory of Urban and Industrial Safety, Institute of Safety Engineering, School of Urban Construction and Safety Engineering,
Nanjing University of Technology, Nanjing 210009, Jiangsu, China
b
State Environmental Protection Key Laboratory of Environmental Risk Assessment and Control on Chemical Process, School of Resources and Environmental Engineering,
East China University of Science and Technology, Shanghai 200237, China

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 31 August 2010
Received in revised form
23 June 2011
Accepted 26 June 2011

More specic and accurate probabilistic models of the numbers of fragments generated respectively by
Boiling Liquid Expanding Vapor Explosions (BLEVEs), Mechanical Explosions (MEs), Conned Explosions
(CEs), and Runaway Reactions (RRs) of a horizontal cylindrical vessel were developed using the
maximum entropy principle based on historical accident data. The theoretical results from the four
probability density functions were compared to the observed data, and the numbers of fragments followed discrete exponential distributions in the interval [1, 9]. Beside the summary of the probabilistic
distributions of the other random variables in the process of fragment projection, the effects on the
fragment trajectory and target terms were investigated using a parametric approach. The results showed
that using the complete model, wind shear, turbulence, and absence of fragment rotation caused the
fragments to impact within shorter distances; fragment rotation and lack of wind decreased the probability of impact within a given distance, but the rupture probability of the target was not affected by
fragment rotation or wind. The probabilistic condence intervals of fragment range, impact, and target
penetration became narrower with the number of simulation runs, but the accuracy of the results
increased. The probability of fragment impact increased with the volume of the target vessel and the
degree of lling of the explosion vessel, but did not depend on the kind of explosion. The probability of
target rupture increased slowly with the degree of lling of the explosion vessel, but was little inuenced
by the volume of the target vessel or the kind of explosion.
2011 Published by Elsevier Ltd.

Keywords:
Domino effect
Industrial explosion
Monte-Carlo simulations
Number of fragments
Parametric approach

1. Introduction
In chemical process industries, the domino effect is a wellknown cause of major accidents (Antonioni, Spadoni, & Cozzani,
2009; Cozzani, Antonioni, & Spadoni, 2006; Nguyen, Mbarki,
Ami Saada, Mercier, & Reimeringer, 2009). An accidental event
which starts at one unit may damage another through heat radiation, blast waves, or projectiles. In reality, a sudden explosion can
generate many fragments which can be projected over long
distances, threaten other sites located in the vicinity, and lead to
more severe consequences due to the nature of the domino effect.
Fragment projection in an explosive accident is one important
cause of the domino effect on chemical process equipment
(Pietersen, 1988). The overall domino effect caused by fragments is
composed of a set of elementary cycles, and each cycle includes
* Corresponding author. Mail Box 13, No. 200 Zhongshan North Road, Nanjing
University of Technology, Nanjing 210009, China. Tel.: 86 25 83587421; fax: 86
25 83239973.
E-mail addresses: jcjiang@njut.edu.cn, j_c_jiang@163.com (J. Jiang).
0950-4230/$ e see front matter 2011 Published by Elsevier Ltd.
doi:10.1016/j.jlp.2011.06.029

three detailed steps: the source term (explosion and generation of

the fragments), the fragment trajectory term (angles, velocities, and
displacements from the source), and the target term (impact of and
interaction between the fragments and the target).
2. Analysis of previous work
Some research on the three components described above has
been performed in previous work (Abbasi & Abbasi, 2007; Bahman,
Abbasi, Rashtchian, & Abbasi, 2010; Baum, 1988, 1995, 1998, 1999a,
1999b, 2001; Bukharev & Zhukov, 1995; CCPS, 1994; Genova,
Silvestrini, & Leon Trujillo, 2008; Gubinelli & Cozzani, 2009a,
2009b; Gubinelli, Zanelli, & Cozzani, 2004; Hauptmanns, 2001a,
2001b; Holden, 1988; Holden & Reeves, 1985; Lees, 1996;
Lepareux et al., 1989; Mbarki, Mercier, Nguyen, & Ami Saada,
2009; Mbarki, Nguyen, Mercier, 2009; Mbarki et al., 2007;
Mbarki, Nguyen, Mercier, Ami Saada, & Reimeringer, 2008;
Neilson, 1985; Nguyen et al., 2009; Qian, Xu, & Liu, 2009; Scilly &
Crowther, 1992; Stawczyk, 2003; Tulacz & Smith, 1980; Van den

D. Sun et al. / Journal of Loss Prevention in the Process Industries 25 (2012) 114e126

Bosch & Weterings, 1997; Zhang & Chen, 2009). In recent work
(Mbarki, Mercier, et al., 2009; Mbarki, Nguyen, et al., 2009;
Mbarki et al., 2007, 2008; Nguyen et al., 2009), the mechanical
and kinetic features of the source terms (random variables such as
number of fragments, shape, and mass) were investigated, and the
corresponding probabilistic distributions were developed using the
maximum entropy principle for the source terms. In the fragment
trajectory term, trajectory equations for the fragments were
proposed, and the ground distributions of the fragments were
assessed. In the target term, probabilistic models of fragment
impact were developed, a calculation of the impact probability was
carried out, and its effects on the probability of impact were evaluated. As for target damage, a simplied plastic model for evaluating the probability of rupture with high reliability was proposed,
and its inuence on penetration depth was investigated. However,
in the analysis described above (Mbarki, Mercier, et al., 2009;
Mbarki, Nguyen, et al., 2009; Mbarki et al., 2007, 2008; Nguyen
et al., 2009), for the source terms, i.e. the development of a probabilistic model for the number of fragments from a horizontal
cylindrical vessel explosion, available accident data were scarce, and
only BLEVEs (Boiling Liquid Expanding Vapor Explosions) resulting
in fragment projection had been considered; for a spherical vessel
explosion, a uniform distribution of the number of fragments within
the interval from 1 to 19 was assumed. Furthermore, the characteristics of fragment ight, impact, and penetration into nearby
facilities, i.e., the fragment trajectory and target terms, still need to
be improved. Generally speaking, the accuracy of quantitative risk
analysis for industrial sites relies intimately on the hypotheses and
the adequacy of the models developed for the whole domino-effect
sequence. On the basis of these ndings (Mbarki, Mercier, et al.,
2009; Mbarki, Nguyen, et al., 2009; Mbarki et al., 2007, 2008;
Nguyen et al., 2009), improvements was made to dene more
specic and accurate probabilistic models of the number of fragments from a horizontal cylindrical vessel explosion by collecting
and analyzing data from past accidents leading to fragment
projection. The objectives were to recommend a more reasonable
probability density function for the number of fragments from
a spherical vessel explosion, to reach more specic conclusions after
reviewing the reference works on the source terms, and then to
explore the effects of the algorithms (movement approach, fragment rotation, wind, and number of simulation runs) on the fragment trajectory and target terms (the ground distributions of
fragments, the probability of impact between the fragments and the
target, and the rupture probability of the impacted target) and the
inuence of the calculation parameters (the objective volume, the
degree of lling of the source vessel, and the kind of explosion) on
the target term (the probability of fragment impact and the rupture
probability of the target) using Monte-Carlo simulations including
the improved source terms, the kinematics of projectiles, and
probabilistic models of fragment impact, penetration, and damage.
3. Source terms
An industrial explosion may generate many fragments with
various features, which can be considered as random variables:
number of fragments (N), shape and size (fP), mass (m), initial
velocity at departure (vO), initial departure angles (horizontal and
vertical angles, q and 4), aerodynamic coefcients (lift and drag
coefcients, CL and CD), and degree of lling of the source vessel (f).
3.1. Number of fragments, N
3.1.1. Case of horizontal cylindrical vessel explosion
In recent work (Mbarki, Mercier, et al., 2009; Mbarki, Nguyen,
et al., 2009; Mbarki et al., 2007, 2008; Nguyen et al., 2009), the

115

Table 1
Accident data for horizontal tank from Gubinelli and Cozzani.
Source: Gubinelli and Cozzani
(2009a,b)

Number of fragments
1

[5e9]

Explosion category
BLEVE
Number of events

56

35

ME
Number of events

CE
Number of events

RR
Number of events

maximum entropy principle was used to establish the probability

density function (p.d.f.) of the number of fragments. However, few
experimental or accident data were used, and only the BLEVEs
generating the fragments were investigated. In fact, for a horizontal
cylindrical vessel, many other accidental scenarios (e.g., a Mechanical
Explosion (ME), Conned Explosion (CE), or Runaway Reaction (RR))
can also cause fragment projection. Therefore, more data on the
number of fragments generated by various experimental and accident scenarios should be collected so that more specic models of the
number of fragments can be developed. Based on the work of
Gubinelli and Cozzani (2009a, 2009b), the primary scenarios for
a horizontal cylindrical vessel are BLEVE, ME, CE and RR, based on
research on data sources of past accidents leading to fragment
projection. The relations between the number of fragments from the
vessel explosion, the tank shape, the type of primary scenario, and
the fracture patterns and mechanics were discussed in detail.
Moreover, the number of fragments was determined with high reliability for each type of primary scenario for the tank. The data from
Gubinelli and Cozzani (2009a, 2009b) were collected for horizontal
cylindrical vessels and are shown in Table 1. Simultaneously, the data
from previous authors (Abbasi & Abbasi, 2007; Baker et al., 1977;
Hauptmanns, 2001a; Holden & Reeves, 1985; Mbarki, Mercier,
et al., 2009; Nguyen et al., 2009) for horizontal cylindrical vessels
were comprehensively collected and are shown in Table 2. The data
for each type of primary scenario in Tables 1 and 2 are incorporated in
Table 3. The number of BLEVEs is larger than in previous work.
Generally speaking, the reliability of the results depends on the
amount of data used in the probability analysis. The number of events
with respect to the observed probabilities of the number of fragments
for different accidental events can be obtained by statistics and is
shown in Table 3. For CEs and RRs, the number in the interval [5e9] is
considered as a random variable following a uniform distribution.
Therefore, each value in [5e9] has the same observed frequency. In
Table 2
Accident data for BLEVE of horizontal tank from Hauptmanns, Holden, and Mbarki.
Explosion category
BLEVE
Source
Hauptmanns
(Baker et al., 1977;
Hauptmanns, 2001a;
Holden & Reeves, 1985)
Holden and Reeves
(Abbasi & Abbasi, 2007)
Holden and Reeves
(Holden & Reeves, 1985;
Nguyen et al., 2009)
Mbarki et al.
(Mbarki, Mercier, et al.,
2009)

Number of fragments
4

Number of events
9
17
11
5

11

11

17

10

12

116

D. Sun et al. / Journal of Loss Prevention in the Process Industries 25 (2012) 114e126

Table 3
Overall number of events for the four explosion categories for horizontal tanks and the corresponding observed probabilities of number of fragments.
Explosion category

Number of fragments
1

BLEVE
Number of events
Observed frequency

50
0.1931

98
0.3784

78
0.3012

24
0.0927

3
0.0116

2
0.0077

3
0.0116

0
0

1
0.0037

ME
Number of events
Observed frequency

0
0

6
0.7500

1
0.1250

1
0.1250

0
0

0
0

0
0

0
0

0
0

CE
Number of events
Observed frequency

0
0

9
0.9000

0
0

0
0

1
0.0200

0.0200

0.0200

0.0200

0.0200

RR
Number of events
Observed frequency

2
0.2857

3
0.4286

1
0.1429

0
0

1
0.02856

0.02856

0.02856

0.02856

0.02856

general, the interval representing the number of fragments from the

four categories of explosions is [1e9], with the numbers of fragments
1e4 accounting for the larger proportions for each event. However,
the numbers of fragment and their observed probabilities are
different for different accidental events.
The probabilistic distributions corresponding to the number of
fragments for BLEVE, ME, CE, and RR have been developed based on
the available data summarized in Table 3 according to the
maximum entropy principle. In general, the model obtained using
the maximum entropy principle, which is the most rational and
stochastic way to predict unknown data under certain available
information, is the only unbiased way to account for all the available information, because all the other models impose additional
constraints and assumptions which cannot be derived from the
available information. Consequently, the maximum entropy principle is the best choice for developing the models. The probability
density function of the maximum entropy principle can be
expressed as shown in Eq. (1), because the only variable involved is
the number of fragments N:

PN exp  l0  l1 $g1 N  l2 $g2 N  /  ln $gn N

(1)

where N is the number of fragments, P() the discrete probability

density function, l0, l1, ., ln the Lagrangian multipliers, and g1(N),
g2(N), ., gn(N) the available information.
The data in Table 3 consist of the large sample (the full set of
data on BLEVEs) and the small sample (the smaller set of data on
MEs, CEs, and RRs). However, the numbers of fragments for the four
explosion categories are all between 1 and 9, and the interval D is
therefore dened as [1e9]. For the four explosion categories in
a horizontal cylindrical vessel, the available information from the
data in Table 3 includes E1 (the average of the number of fragments)
and E2 (the variance of the number of fragments), which are
expressed respectively in Eqs. (2) and (3), where PiObserved is the
observed frequency of the number of fragments i. On the basis of
the available information, the probability density function for the
number of fragments can be expressed as in Eq. (4). Equation (5) is
obtained by combining Eqs. (2)e(4), and the Lagrangian multipliers
l0, l1, ., ln are obtained based on the data in Table 3. Finally, the
probability density functions are determined, and the theoretical
frequencies corresponding to the number of fragments can be
calculated using the probabilistic distributions.

E1

9
X

i$PiObserved

(2)

i  E1 2 PiObserved

(3)

i1

E2

9
X
i1



PN exp  l0  l1 N  l2 N 2


8P
2
>
1
> exp  l0  l1 i  l2 i
>
>
D
>


<P
i$exp  l0  l1 i  l2 i2 E1
>
D
>


>
P
>
>
: i  E1 2 $exp  l0  l1 i  l2 i2 E2

(4)

(5)

The probability density functions of the numbers of fragments

from BLEVE, ME, CE, and RR have been developed as shown in Eqs.
(6)e(9). The results of the probabilistic distributions of the number
of fragments for the four explosion categories are shown in Fig. 1.
All the theoretical histograms are compared to the accident
(observed) frequencies presented in Table 3. This means that the
numbers of fragments from BLEVE, ME, CE, and RR follows the
discrete exponential distribution within the interval [1, 9].



PN exp  2:16 0:97N  0:24N 2
p:d:f : of the number of fragments from BLEVE



PN exp  6:26 4:80N  1:01N 2
p:d:f : of the number of fragments from ME



PN exp  0:93  0:11N  0:05N 2
p:d:f : of the number of fragments from CE



PN exp  0:20  0:72N 0:03N 2
p:d:f : of the number of fragments from RR

3.1.2. Case of spherical vessel explosion

In recent work (Mbarki et al., 2007, 2008; Mbarki, Mercier, et al.,
2009; Mbarki, Nguyen, et al., 2009; Nguyen et al., 2009), a uniform
distribution for the number of fragments within the interval [1e19]
was assumed. In fact, because a crack can start anywhere in the
explosion of a spherical vessel, the larger the source vessel volume,
the greater will be the number of fragments generated by the
explosion, because higher vessel volumes and higher vessel surface
areas correspond to a higher probability of fracture branching.
Therefore, the volume of the source vessel has an effect on the
number of fragments. On this topic, a linear correlation between the
number of fragments N and the source vessel volume V was originally
proposed: N 3:77 0:96  102 V (Holden and Reeves, 1985).
Later, a correlation between the number of fragments N and the

D. Sun et al. / Journal of Loss Prevention in the Process Industries 25 (2012) 114e126

117

Fig. 1. Theoretical and accident probabilities versus number of fragments from BLEVE, ME, CE, and RR.

source vessel volume V was obtained: N 0.425 6.115  103V

over the interval [3e19] of the number of fragments, with an average
of eight fragments as determined from data sources from past accidents giving rise to fragment projection (the data sources involved 13
BLEVEs, and there were no other accidental scenarios causing fragment projection) (Gubinelli & Cozzani, 2009a, 2009b; Holden, 1986;
Westin, 1971). In Holden and Reevess work, the scarcity of the data
available resulted in a relevant uncertainty of the correlation, which
might result in larger errors. In Gubinelli and Cozzanis work, more
data were collected, and therefore higher model accuracy could be
achieved, and the established correlation gave more reasonable
results for low source vessel volumes (the number of fragments
approached zero for source volumes near to zero). Therefore, the
model
developed
by
Gubinelli
and
Cozzani,
i.e.,
N 0.425 6.115  103V, has been used here.
3.2. Other features of source terms
In recent research by (Mbarki, Mercier, et al., 2009; Mbarki,
Nguyen, et al., 2009; Mbarki et al., 2007, 2008; Nguyen et al.,
2009), other features of the source terms (random variables) have
been investigated and discussed, including fragment shape and size
(fP), mass (m), initial velocity at departure (vO), initial departure angles
(horizontal and vertical angles, q and 4), and aerodynamic coefcients
(lift and drag coefcients, CL and CD). Based on these previous results,
the same probabilistic distributions of these features of the source
terms are used here, as shown in Table 4 (Mbarki, Mercier, et al.,
2009; Mbarki, Nguyen, et al., 2009; Nguyen et al., 2009).
The degree of lling of the source f at the time of the accident,
which is important for calculation of the total energy imparted

to the fragments, is practically never known for real accidents

(Hauptmanns, 2001a, 2001b). It is therefore considered to be
a feature of the source terms (i.e., a random variable) following
a uniform distribution within [0, 1].
All the features of the source terms are summarized in Table 4
(Gubinelli & Cozzani, 2009b; Mbarki, Mercier, et al., 2009;
Mbarki, Nguyen, et al., 2009; Nguyen et al., 2009).
4. Fragment trajectory and target terms
4.1. Methods
4.1.1. Kinematics of projectiles
As discussed in the references (Mbarki, Mercier, et al., 2009;
Mbarki, Nguyen, et al., 2009; Mbarki et al., 2007, 2008; Nguyen
et al., 2009), the total energy E can be calculated using Eq. (10)
from Baum (1988) when the explosion of a tank occurs. Afterward, the kinetic energy Ek of the fragments can be derived using
Eq. (11). The initial velocity of the generated fragments, vO, can be
derived from their kinetic energy Ek and mass m (Eq. (12)).


E

1


 g1=g
p0
p
p1
g  1 0
fV
p1
p1 g  1

Ek cE
vO

p
2Ek =m

(10)
(11)
(12)

where p0 is the atmospheric pressure, p1 the failure pressure of the

source vessel, g the specic heat ratio, f the degree of lling of the

118

D. Sun et al. / Journal of Loss Prevention in the Process Industries 25 (2012) 114e126

Table 4
Features of source terms and their probabilistic distributions.
Features of source terms

P.d.f. in case of horizontal cylindrical vessel

P.d.f. in case of spherical vessel

N: Number of fragments

Discrete exponential distribution for any accident

among BLEVE, ME, CE and RR
Uniform distribution for any form among end-cap,
oblong end-cap, and attened fragment, the size being
uniformly distributed (Mbarki, Mercier, et al., 2009; Mbarki,
Nguyen, et al., 2009; Nguyen et al., 2009)
Beta distribution (Mbarki, Mercier, et al., 2009)
Exponential distribution (Mbarki, Mercier, et al., 2009;
Mbarki, Nguyen, et al., 2009)
Uniform distribution: 20% in [30 e150 ], 30% in
[150 e210 ], 20% in [210 e330 ], 30% in [330 e30 ]
(Mbarki, Mercier, et al., 2009; Mbarki, Nguyen, et al., 2009;
Nguyen et al., 2009)
Uniform distribution within the interval [90 to 90 ]
(Mbarki, Mercier, et al., 2009; Mbarki, Nguyen, et al., 2009;
Nguyen et al., 2009)
Uniform distribution within the interval [0.351e0.468]
for any form between end-cap and oblong end-cap;
the effect can be neglected for plate (Nguyen et al., 2009)
Uniform distribution within the interval [0.8e1.1] for
any form between end-cap and oblong end-cap, and
uniform distribution within the interval [1.1e1.8] for
plate (Nguyen et al., 2009)
Uniform distribution within the interval [0e1]

Requires the vessel volume for BLEVE accidents

(Gubinelli & Cozzani, 2009b)
Uniform distribution for any form between end-cap
and attened fragment, the size being uniformly distributed
(Mbarki, Mercier, et al., 2009; Mbarki, Nguyen, et al., 2009)

fp: Frequency of each

fragment shape and size

m: Fragment mass
vO: Departure velocity
(multiplicative factor c)
q: Horizontal departure
angle

4: Vertical departure angle

CL: Lift coefcient

CD: Drag coefcient

f: Degree of lling of
source

source vessel, V the volume of the source vessel, and c a multiplicative factor.
The trajectory of the fragments projected from the exploded
vessels results from the combined effects of inertia, gravitation, and
aerodynamics (drag and lift). Let (O, X, Y, Z) be the set of system
axes used for the trajectory description, and let O be the departure
point for a generated fragment. The fragment trajectory can be
described as shown in Fig. 2 (Mbarki, Nguyen, et al., 2009). After
the fragment has been projected, an impact is possible with any
potential target that crosses its trajectory.
For fragment movements, the complete movement approach is
used (Mbarki, Mercier, et al., 2009; Mbarki, Nguyen, et al., 2009;
Mbarki et al., 2007, 2008; Nguyen et al., 2009):

The same mass for all the fragments (Mbarki, Mercier, et al., 2009)
Exponential distribution (Mbarki, Mercier, et al., 2009; Mbarki,
Nguyen, et al., 2009)
Uniform distribution within the interval [0 e360 ] (Mbarki,
Mercier, et al., 2009; Mbarki, Nguyen, et al., 2009)

Uniform distribution within the interval [90 to 90 ] (Mbarki,

Mercier, et al., 2009; Mbarki, Nguyen, et al., 2009)
Uniform distribution within the interval [0.351e0.468] for end-cap;
and the effect can be neglected for plate (Nguyen et al., 2009)
Uniform distribution within the interval [0.8e1.1] for end-cap, and
uniform distribution within the interval [1.1e1.8] for plate
(Nguyen et al., 2009)
Uniform distribution within the interval [0e1]

- Final conditions: when the fragment crashes on the ground, in

the descending part of its trajectory, its vertical coordinate
becomes zero.
Hereafter, the distributions of the fragments crashing on the
ground will be evaluated by means of Monte-Carlo simulations
using the complete movement approach.
4.1.2. Fragment impact, penetration, and damage
When the trajectory of a fragment has been completely identied, its possible impacts on a given target can be easily determined
by its position on the fragment trajectory. As discussed in detail in
the references (Mbarki, Mercier, et al., 2009; Mbarki, Nguyen,


8

 2
q
>
_ y_ 2 z_ 2 
4
q
4
q
$cos
$cos


1
$k
$sin
$cos

$
x

k
x 0
>
D
L
>
<


 2
2
2
q
_
_

_
 kD $cos4$sinq  1 $kL $sin4$sinq $ x y z  y 0
>


>
>
:  1q $k $sin4 k $cos4
$ x_ 2 y_ 2 z_ 2  z  g 0
D
L

where q 1 for the descending part; q 2 for the ascending part; x,

y, and z are the fragment center coordinates; kD is the drag factor,
kD rair CD SD =2m; kL is the lift factor, kL rair CL SL =2m; rair is the
specic density of air; SD is the frontal surface; and SL is the horizontally projected surface. The solution of the fragment motion
equations can be derived under the following set of hypotheses
proposed by the authors (Mbarki, Mercier, et al., 2009; Mbarki,
Nguyen, et al., 2009; Mbarki et al., 2007, 2008; Nguyen et al.,
2009):
- Initial conditions: at departure, the fragment is located at the
system origin (point O). The pulse produced by the blast (vessel
explosion) produces the initial velocity of the fragment (vOx,
vOy, vOz).
- When the projectile reaches its trajectory upper position (top
point), its vertical velocity becomes zero.

(13)

et al., 2009; Mbarki et al., 2007, 2008; Nguyen et al., 2009), the
impact probability between the projectiles and the potential target
can be calculated. Monte-Carlo simulations were run to evaluate
the probability of impact, Pimp, in the light of Eq. (14):

PNsim PN
s1

Pimp

nj; s
N

and nj; s
Nsim


if Vtarget XVfragment sf

otherwise

j1

(14)

where Nsim is the total number of Monte-Carlo simulations; n is

a parameter that indicates whether the projectile impacts the
target; Vfragment is the fragment trajectory; and Vtarget is the target
volume with a given location, dimensions, and shape. The numerical algorithm for the Monte-Carlo method for fragment impact has

D. Sun et al. / Journal of Loss Prevention in the Process Industries 25 (2012) 114e126

119

Table 5
Characteristics of source and target vessels in Nanjing Chemical Industry Park of
China.
Source A
Type

Horizontal
cylindrical
vessel
200
Volume/m3
Vessel diameter/m
2.8
Failure pressure/MPa 1.20
Atmospheric
0.1
pressure/MPa
Mass/kg
25,429
Wall thickness/m
0.024
Center coordinates
(0, 0, 0)
Critical residual
e
thickness/m

Source B

Target A

Target B

Spherical
vessel

Spherical
vessel

Spherical
vessel

1000
6.2
1.01
0.1

800
5.8
0.98
0.1

2400
8.3
1.26
0.1

121,886
88,761
0.026
0.020
(40 m, 0, 0) (0, 40 m, 6 m)
e
0.004

22,4301
0.029
(0, 40 m, 9 m)
0.006

Fig. 2. Description of fragment trajectory.

been described in detail by (Mbarki, Mercier, et al., 2009; Mbarki,

Nguyen, et al., 2009; Mbarki et al., 2007, 2008; Nguyen et al.,
2009).
When an impact between the fragment and a target occurs, it
may cause partial or complete damage (penetration or perforation)
to the target and then result in the explosion of the target. As discussed in the references (Mbarki, Mercier, et al., 2009; Mbarki,
Nguyen, et al., 2009; Mbarki et al., 2007, 2008; Nguyen et al.,
2009), the penetration depth of the fragment can be expressed
using Eq. (15) or (16), and the rupture probability of the impacted
target, Prup, can be expressed as in Eq. (17):

s


Ec 2=3
2 4
dP cos h dP cos h tan h
p
fu 3u
hP
for the case hs0
2 tan h
(15)
hP

Ec

pdP fu 3u

2=3

for the case h 0

PNsimu PNimp qj; i

i1
j1 N
imp
and
Prup PEe  0
Nsimu

1 if Ee  0
qj; i
; Ee et  hP  ecr
0 otherwise

(16)

(17)

where dp is the fragment diameter; h is the incidence angle of the

fragment; fu is the ultimate strength of the target constitutive
material and 3u its ultimate strain; Ec is the kinetic energy expended
when the penetration process occurs; Ee is the limit state function;
Nimp is the number of the fragments impacting the target, obtained
through Eq. (14) in each simulation; Nsimu is the total number of
P
Monte-Carlo simulations indicating that N
j 1 nj; s=N in Eq. (14) is
not zero in each simulation; et is the target thickness; hp is the
penetration depth; and ecr is the critical plate thickness. Use of the
Monte-Carlo method for evaluation of the rupture probability of
the impacted target has been described in detail in the references
(Mbarki, Mercier, et al., 2009; Mbarki, Nguyen, et al., 2009;
Mbarki et al., 2007, 2008; Nguyen et al., 2009).
4.1.3. Characteristics of source and target vessels
As a case study, the industrial site installed in Nanjing Chemical
Industry Park of China was considered, as described in Table 5
(Shen, 2010). A horizontal cylindrical vessel and a spherical vessel
(Source A and Source B) were considered as the sources (both
accidents are BLEVEs), and two spherical vessels (Target A and

Target B) were considered as the targets. The threshold thickness ecr

is shown only for Target A and Target B, and the corresponding
values for the calculation using Eq. (17) are provided in Table 5.
Using the probabilistic distributions summarized in Table 4, the
source terms in Table 5, and Eqs. (10)e(12), the fragment features
(e.g. number of fragments, mass, energy, departure velocity, and
angles) at departure can be obtained. Then the trajectory and the
ground distributions of fragments can be derived using Eq. (13). For
the calculation of the fragment impact probability and the rupture
probability of the impacted target, because Source B was not
located at the system origin (point O) (see the center coordinates in
Table 5), the simulations were carried out by coordinate translation
for Source B, Target A, and Target B.

4.2. Investigation of the effects of various factors on ground

distribution of fragments
The factors include movement approach, fragment rotation,
wind, and number of simulation runs. The condence interval
approach is preferable when dealing with probabilities and
convergence analysis.
4.2.1. Movement approach
Two kinds of movement approaches were considered: the
complete model as given in Eq. (13) and the simplied model (the
complete model in the absence of air resistance (kD 0 in Eq. (13))).
The resulting fragment distributions were obtained by the
complete and simplied movement approaches without fragment
rotation or wind (see Fig. 3). Fig. 3 shows that the complete model
predicted that the fragments would impact in shorter distances
than the predicted by the simplied model for Sources A and B.
4.2.2. Fragment rotation
Fragment rotation around the center of mass was discussed in
detail in the references (Mbarki, Mercier, et al., 2009; Mbarki,
Nguyen, et al., 2009; Mbarki et al., 2007, 2008; Nguyen et al.,
2009). On this basis, results for the effect of fragment rotation on
the distributions were obtained without any wind effects (Fig. 4).
From these results, it can be seen that absence of rotation caused
the fragments to impact in shorter distances for Sources A and B.
4.2.3. Wind
Wind shear and turbulence are the two main effects that
a variable wind eld can have on the ight of fragments (Liu, Wang,
& Jia, 2006). These characteristics of wind were considered as the
dynamics of a moving air mass, i.e., as wind velocity and direction.
Therefore, the models used for wind shear and turbulence can be
summarized as follows.

120

D. Sun et al. / Journal of Loss Prevention in the Process Industries 25 (2012) 114e126

Fig. 3. Theoretical simulations using complete and simplied models versus range.

The induced velocity of a single-line vortex can be expressed as

in Eq. (18) (Liu et al., 2006):

VWi ri ; t

r2
G0i
1  exp  i
2pri
4yt

!!
(18)

where G0i is the vortex strength with t 0; ri is the circumferential

distance from the vortex center; y is the kinematic viscosity coefcient of the uid surface; and t is the lifespan of the vortex. The
number, position, and strength of vortices can be conrmed by
comparison with the actual situation. The horizontal and vertical
components of the wind induced by n line vortexes can be
expressed as follows (Liu et al., 2006):

uW

ns
X
xWi
i1

wW 

ri

VWi

ns
X
zWi
i1

ri

VWi

(19)

(20)

q
where xWi x  xWi0 , zWi z  zWi0 , ri x2Wi z2Wi , and xWi0
and zWi0 are the coordinates of the vortex center.
The time spectrum of the turbulent velocity can be calculated
using Drydens atmospheric turbulence model (Liu et al., 2006):

8
L
1
>
>
Fuu u s2u u
>
>
>
p
vT 1 Lu =vT u2
>
>
>
2
>
>
>
< Fvv u s2 Lv n1 12Lv =vT uo
v
2
pvT
1 4Lv =vT u2
>
>
>
>
>
>
L
1 12Lw =vT u2
>
>
Fww u s2w w n
>
o2
>
p
v
>
T
:
1 4Lw =vT u2

(21)

where Lu, Lv, and Lw are respectively the turbulence scales of the
three directions; su , sv , and sw are the turbulence intensities in the
three directions; vT is the velocity of the fragment when it passes
through the turbulence eld; and u is the time frequency.
In the models described above, the wind velocity can be
assumed constant because the velocity of the fragment is much
larger than the wind velocity and its variations. Therefore, variations in the wind eld can be neglected according to the frozen
eld hypothesis (Liu et al., 2006).
As discussed above, the results for the distributions of the
fragments on the ground with no wind and with wind shear and
turbulence were obtained without fragment rotation (Fig. 5). The
results show that wind shear and turbulence caused the fragments
to impact in shorter distances than under no-wind conditions for
Sources A and B, and that the ground distributions of the fragments
with wind shear were almost the same as those with turbulence.

Fig. 4. Theoretical simulations with no rotation and with rotation versus range.

D. Sun et al. / Journal of Loss Prevention in the Process Industries 25 (2012) 114e126

121

Fig. 5. Theoretical simulations with no wind, wind shear, and turbulence versus range.

Fig. 6. Theoretical simulations with different numbers of simulation runs versus range.

4.2.4. Number of simulations

The inuence of the number of simulations on the fragment
distributions was investigated without fragment rotation or wind.
The number of simulations (Nsim) was successively chosen as 103, 104,
105, 106, 107, 108, 109, and 1010 simulations. The results are shown in
Fig. 6, which shows that there was almost no change in the theoretical
calculations with increasing number of simulations for Sources A and
B. The condence intervals of the probabilities in the range of
0e100 m were investigated. Using the methods mentioned in Online
MBA Library (2010), the condence intervals of the probabilities in the

range of 0e100 m corresponding to 103, 104, 105, 106, 107, 108, 109, and
1010 simulations for Source A were: [0.13104, 0.16818], [0.14466,
0.15645], [0.14844, 0.15217], [0.14964, 0.15082], [0.14975, 0.15012],
[0.15017, 0.15029], [0.15013, 0.15017], and [0.15038, 0.15039]; the
condence intervals of the probabilities in the range of 0e100 m
corresponding to 103, 104, 105, 106, 107, 108, 109, and 1010 simulations
for Source B were: [0.15376, 0.17617], [0.16141, 0.16849], [0.16373,
0.16598], [0.16452, 0.16523], [0.16487, 0.16509], [0.16495, 0.16502],
[0.16483, 0.16485], and [0.16484, 0.16484]. These results demonstrate
that the probabilistic condence interval became narrower with

Fig. 7. Comparison of probabilities of impact with no rotation and with rotation.

122

D. Sun et al. / Journal of Loss Prevention in the Process Industries 25 (2012) 114e126

Fig. 8. Comparison of rupture probabilities of the target with no rotation and with rotation.

increasing number of simulations, meaning that the accuracy of the

probabilities increased due to the larger number of samples corresponding to a larger number of simulations.
4.3. Investigation of the effects of various factors on the impact
probability of fragments and the rupture probability
of the impacted target
4.3.1. Fragment rotation
The effects of fragment rotation on the probability of fragment
impact and the rupture probability of the target were investigated.
The results for the impact probability and the rupture probability of
the target without rotation and with rotation were obtained
without wind (Figs. 7 and 8). These gures show that the probability of impact with rotation was smaller than that without rotation for Sources A and B, and that there was almost no difference
between the rupture probabilities of the target with and without
rotation for Sources A and B.
4.3.2. Wind
The effects of wind were considered as discussed in Section
4.2.3. The results for the impact probability and the rupture probability of the target with no wind, with wind shear, and with
turbulence were obtained without fragment rotation (Figs. 9 and
10). These gures show that the probability of impact with wind
shear or turbulence was greater than that with no wind for Sources
A and B, and that there was almost no difference between the

probabilities of impact with wind shear and with turbulence for

Sources A and B. Furthermore, there was almost no difference
between the rupture probabilities of the target with no wind, with
wind shear, and with turbulence for Sources A and B.
4.3.3. Number of simulations
The inuence of the number of simulations was considered
without fragment rotation or wind, and Target B was chosen as
the target. The results are shown in Figs. 11 and 12, which shows
that 10 4 simulations provided enough accuracy for the probability of impact for Sources A and B. There was almost no
difference between the rupture probabilities of the target for
103, 10 4, 105, 10 6, 107, 108, 109, and 1010 simulations for Sources A
and B. The condence intervals for the impact probabilities and
the rupture probabilities of the target were also investigated.
Using the methods mentioned in Online MBA Library (2010), the
condence intervals of the impact probabilities for 103, 104, 105,
106, 107, 108, 109, and 1010 simulations for Source A were:
[0.01331, 0.03169], [0.01744, 0.02296], [0.01952, 0.02128],
[0.02002, 0.02058], [0.02011, 0.02029], [0.02027, 0.02033],
[0.02029, 0.02031], and [0.02029, 0.02030]; the condence
intervals of the impact probabilities for 103, 10 4, 105, 10 6, 107, 108,
109, and 1010 simulations for Source B were: [0.00425, 0.01695],
[0.00537, 0.00863], [0.00668, 0.00772], [0.00684, 0.00716],
[0.00705, 0.00715], [0.00718, 0.00722], [0.00719, 0.00721], and
[0.00719, 0.00720]; the condence intervals for the rupture
probabilities of the target for 103, 104, 105, 106, 107, 108, 109, and

Fig. 9. Comparison of probabilities of impact with no wind, with wind shear, and with turbulence.

D. Sun et al. / Journal of Loss Prevention in the Process Industries 25 (2012) 114e126

123

Fig. 10. Comparison of rupture probabilities of the target with no wind, with wind shear, and with turbulence.

Fig. 11. Probability of impact versus number of simulations.

1010 simulations for Source A were: [0.22767, 0.63233], [0.37155,

0.50845], [0.41846, 0.46154], [0.42319, 0.43681], [0.41785,
0.42215], [0.43932, 0.44068], [0.43978, 0.44022], and [0.42993,
0.43007]; the condence intervals of the rupture probabilities of
the target corresponding to 103, 104, 105, 10 6, 107, 108, 109, and
1010 simulations for Source B were: [0.08468, 0.65532],
[0.24755, 0.47245], [0.32494, 0.39506], [0.35869, 0.38131],

[0.35647, 0.36353], [0.34889, 0.35110], [0.35965, 0.36035], and

[0.35989, 0.36011]. These results demonstrate that the probabilistic condence interval became narrower with increasing
number of simulations and that the accuracy of the probability
of impact and the rupture probability of the target increased
with the larger number of samples corresponding to the larger
number of simulations.

Fig. 12. Rupture probability of target versus number of simulations.

124

D. Sun et al. / Journal of Loss Prevention in the Process Industries 25 (2012) 114e126

Fig. 13. Top view of fragment trajectory and impact.

4.4. Investigation of the effects of the calculation parameters on the

impact probability of fragments and the rupture probability of the
impacted target
The effects of the distance from the source to the target and the
objective orientation on the impact probability of fragments and
the rupture probability of the target have been researched in the
references (Mbarki, Mercier, et al., 2009; Mbarki, Nguyen, et al.,
2009; Mbarki et al., 2007, 2008; Nguyen et al., 2009; Zhang &
Chen, 2009; Qian et al., 2009). In this work, the calculation
parameters, including the target volume, the degree of lling of the
source vessel, and the kind of explosion were considered, and their
inuence was investigated using Monte-Carlo simulations. The
vessels described in Table 5 were used.
4.4.1. Target volume
Based on the results shown in Figs. 7e10, the probability of
impact increased with greater target volume (the probability of
impact on Target B was greater than that on Target A under the
same conditions such as fragment rotation and wind (Figs. 7 and
9)), and the rupture probability was little affected by the target
volume (the rupture probability of Target B was almost the same as
that of Target A under the same conditions (Figs. 8 and 10)).
As stated above, the larger the target volume, the larger was the
impact probability. The larger volume corresponds to a larger target
surface area, and therefore it increases the possibility of impact. As
discussed previously by (Mbarki, Mercier, et al., 2009; Mbarki,

Fig. 14. Probability of impact versus degree of lling of source.

Fig. 15. Rupture probability of target versus degree of lling of source.

Nguyen, et al., 2009; Mbarki et al., 2007, 2008; Nguyen et al.,

2009), in Fig. 13 (Mbarki, Nguyen, et al., 2009), which shows
a top view of the fragment trajectory and impact on the target
horizontal cylindrical vessel, the sum of the projected area of the
target in the direction from target C to origin O (surface area 1) and
the upper area of the target (surface area 2) is the total area (total
surface area, St) in which impact may occur.
Therefore, St Surface area 1 Surface area 2 [(ap bt)
sin j (ap at)cos j]  (ap ct) (ap bt)  (ap at), where ap is
the maximal distance from the fragment rotation center to its
boundary, and at, bt, and ct are the cubic dimensions of the horizontal cylindrical target. As stated in the references (Mbarki,
Mercier, et al., 2009; Mbarki, Nguyen, et al., 2009; Mbarki et al.,
2007, 2008; Nguyen et al., 2009), an impact may occur when the
fragment
center
crosses
the
cube
with
dimensions
[(ap at)  (ap bt)  (ap ct)], where O is the origin of the
generated fragment, C the target center, I the impact point, and j
the angle between the major axis of the target and the source. The
probability of fragment impact depends on St. Moreover, for
a spherical vessel target, St (ap at)2  (sin j cos j 1) under
at bt ct.
4.4.2. Degree of lling of the source vessel
The accident scenarios for both Source A and Source B in Table 5
are considered as BLEVEs, and Target A is chosen as the target. As
stated in Section 3.2, the degree of lling of source f follows
a uniform distribution within [0, 1]. The effect of the degree of

Fig. 16. Probability of impact versus kind of explosion.

D. Sun et al. / Journal of Loss Prevention in the Process Industries 25 (2012) 114e126

Fig. 17. Rupture probability of target versus kind of explosion.

lling of the source vessel on the impact probability of fragments

and the rupture probability of the target was investigated for
f-values of 0.2, 0.5, and 0.8. The results are shown in Figs. 14 and 15.
Fig. 14 shows that the probability of impact increases with degree of
lling for Sources A and B, and Fig. 15 shows that the rupture
probability of the target increases slowly according to the degree of
lling for Sources A and B. This can be explained by the fact that
a larger kinetic energy of fragments can be derived from a source
with a larger degree of lling when a vessel explosion occurs. After
the ight of the fragments, the kinetic energy Ec when penetration
occurs, the penetration depth hp, and the rupture probability of the
target Prup can be increased according to Eqs. (15)e(17).
4.4.3. Kind of explosion
As stated in Section 3.1.1, for horizontal cylindrical vessel
explosions, four types of accident scenarios (BLEVE, ME, CE, and RR)
can cause fragment projection. Therefore, the effects of the four
kinds of explosions of Source A on the impact probability of fragments and the rupture probability of the target were investigated,
with Target A chosen as the target. The results are shown in Figs. 16
and 17. These gures indicate that the impact probability of the
fragments and the rupture probability of the target depend very
little on the kind of explosion.
5. Conclusions
A parametric approach to the domino effect for structural
fragments was investigated:
 The intervals of the number of fragments from BLEVE, ME, CE,
and RR were all [1e9], with numbers of fragments from 1 to 4
accounting for larger proportions in each event. The numbers
of fragments and their observed probabilities were different for
different accidental events. More specic probabilistic models
of the number of fragments generated by horizontal cylindrical
vessel accidents of BLEVE, ME, CE, and RR types were developed using the maximum entropy principle based on historical
accident data. All the theoretical results from the four probability density functions were compared to observed data, and
the number of fragments was found to follow discrete exponential distributions in the interval [1, 9]. Furthermore, a more
reasonable correlation between the number of fragments and
the source vessel volume for spherical vessel explosions was
recommended, as well as a synthesis of the other features.
 For the ground distributions of fragments generated by
explosions of a horizontal cylindrical vessel and a spherical

125

vessel, the complete model predicted that the fragments would

impact in shorter distances than predicted by the simplied
model, the absence of fragment rotation caused the fragments
to impact in shorter distances, wind shear and turbulence
caused the fragments to impact in shorter distances than with
no wind, and the ground distributions of fragments with wind
shear were almost the same as those with turbulence. The
probabilistic condence intervals became narrower with
increasing number of simulations, and the accuracy of the
results increased due to the larger number of samples corresponding to the larger number of simulations.
 As for the probability of fragment impact, the probability of
impact with fragment rotation was smaller than with no
fragment rotation, and there was almost no difference between
the rupture probabilities of the target with and without fragment rotation. The probability of impact with wind shear or
turbulence was greater than with no wind, the probabilities of
impact with wind shear were almost the same as those with
turbulence, and there was almost no difference between the
rupture probabilities of the target with no wind, with wind
shear, and with turbulence. 104 simulations provided enough
accuracy for the probability of impact and the rupture probability of the target. The probabilistic condence interval
became narrower with increasing number of simulations, and
the accuracy of the probability of impact and the rupture
probability of the target increased due to the larger number of
samples corresponding to the larger number of simulations.
 The probability of impact increased with target volume,
a larger target surface area was found to correspond to the
projection area in the direction from the target to the source,
and a larger upper area increased the probability of impact. The
probability of fragment impact increased with degree of lling
and did not depend on the kind of explosion; the rupture
probability was little affected by the target volume, increased
slowly with degree of lling, and depended very little on the
kind of explosion.
Acknowledgments
The nancial support of National Natural Science Foundation of
China (No. 71001051, 50904037), the eleventh ve-year national
technology support plan of China (2007BAK22B04), the Academic
Fund for Young Teachers of Nanjing University of Technology (No.
39714005), the Discipline Fund of Nanjing University of Technology,
Research and Innovation Plan for Graduates of Colleges and Universities in Jiangsu Province (CX09B_142Z), and the Shanghai Leading
Academic Discipline Project (B506) are gratefully acknowledged.
References
Abbasi, T., & Abbasi, S. A. (2007). The boiling liquid expanding vapour explosion
(BLEVE): mechanism, consequence assessment, management. Journal of
Hazardous Materials, 141, 489e519.
Antonioni, G., Spadoni, G., & Cozzani, V. (2009). Application of domino effect
quantitative risk assessment to an extended industrial area. Journal of Loss
Prevention in the Process Industries, 22, 439e449.
Bahman, A., Abbasi, T., Rashtchian, D., & Abbasi, S. A. (2010). A new method for
assessing domino effect in chemical process industry. Journal of Hazardous
Materials, 182, 416e426.
Baker, W. E., Kulesz, J. J., Ricker, R. E., Bessey, Westine, P. S., Parr, V. B., et al. (1977).
Workbook for predicting pressure wave and fragment effects of exploding propellant tanks and gas storage vessels. NASA CR-134906. Washington: NASA Scientic and Technical Information Ofce.
Baum, M. R. (1988). Disruptive failure of pressure vessels: preliminary design
guidelines for fragment velocity and the extent of the hazard zone. Journal of
Pressure Vessel Technology, 110, 168e176.
Baum, M. R. (1995). Rupture of a gas-pressurized cylindrical vessel: the velocity of
a detached end-cap. Journal of Loss Prevention in the Process Industries, 8,
149e161.

126

D. Sun et al. / Journal of Loss Prevention in the Process Industries 25 (2012) 114e126

Baum, M. R. (1998). Rocket missiles generated by failure of a high pressure liquid

storage vessel. Journal of Loss Prevention in the Process Industries, 11, 11e24.
Baum, M. R. (1999a). The velocity of end-cap and rocket missiles generated by
failure of a gas pressurized vessel containing particulate material. Journal of Loss
Prevention in the Process Industries, 12, 259e268.
Baum, M. R. (1999b). Failure of a horizontal pressure vessel containing a high
temperature liquid: the velocity of end-cap and rocket missiles. Journal of Loss
Prevention in the Process Industries, 12, 137e145.
Baum, M. R. (2001). The velocity of large missiles resulting from axial rupture of gas
pressurized cylindrical vessels. Journal of Loss Prevention in the Process Industries, 14, 199e203.
Bukharev, Y. I., & Zhukov, V. I. (1995). Model of the penetration of a metal barrier by
a rod projectile with an angle of attack. Combustion, Explosion, and Shock Waves,
31, 104e109.
Center for Chemical Process Safety (CCPS). (1994). Guidelines for evaluating the
characteristics of vapor cloud explosions, ash res, and BLEVEs. New York:
American Institute of Chemical Engineers.
Cozzani, V., Antonioni, G., & Spadoni, G. (2006). Quantitative assessment of domino
scenario by a GIS-based software. Journal of Loss Prevention in the Process
Industries, 19, 463e477.
Genova, B., Silvestrini, M., & Leon Trujillo, F. J. (2008). Evaluation of the blast-wave
overpressure and fragments initial velocity for a BLEVE event via empirical
correlations derived by a simplied model of released energy. Journal of Loss
Prevention in the Process Industries, 21, 110e117.
Gubinelli, G., & Cozzani, V. (2009a). Assessment of missile hazards: identication
of reference fragmentation patterns. Journal of Hazardous Materials, 163,
1008e1018.
Gubinelli, G., & Cozzani, V. (2009b). Assessment of missile hazards: evaluation of
the fragment number and drag factors. Journal of Hazardous Materials, 161,
439e449.
Gubinelli, G., Zanelli, S., & Cozzani, V. (2004). A simplied model for the assessment
of the impact probability of fragments. Journal of Hazardous Materials, 116,
175e187.
Hauptmanns, U. (2001a). A procedure for analyzing the ight of missiles from
explosion of cylindrical vessels. Journal of Loss Prevention in the Process Industries, 14, 395e402.
Hauptmanns, U. A. (2001b). Monte-Carlo based procedure for treating the ight of
missiles from tank explosions. Probabilistic Engineering Mechanics, 16, 307e312.
Holden, P. L. (1986). Assessment of missile hazards: Review of incident experience
relevant to major hazard plant. Warrington: United Kingdom Atomic Energy
Authority. Report no.: SRD R 477.
Holden, P. L. (1988). Assessment of missile hazards: Review of incident experience
relevant to major hazard plant. Safety and Reliability Directorate, Health and
Safety Directorate.
Holden, P. L., & Reeves, A. B. (1985). Fragment hazards from failures of pressurized
liqueed gas vessels. In IchemE symposium series, 93 (pp. 205e220).
Lees, F. P. (1996). Loss prevention in the process industriesdHazard identication,
assessment, and control, Vols. 1e3. Oxford: Butterworth-Heinemann.
Lepareux, M., Jamet, P., Matheron, P., Lieutenant, J. L., Couilleaux, J., Duboelle, D.,
et al. (1989). Experimental and numerical studies of impacts on stainless steel

plates subjected to rigid missiles at low velocity. Nuclear Engineering and Design,
115, 105e112.
Liu, G., Wang, X. R., & Jia, R. Z. (2006). Engineering simulation method of variable
wind eld in synthetic natural environment. Xitong Fangzhen Xuebao, 18,
297e300, (in Chinese).
Mbarki, A., Mercier, F., Nguyen, Q. B., & Ami Saada, R. (2009). Structural fragments
and explosions in industrial facilities. Part I: probabilistic description of the
source terms. Journal of Loss Prevention in the Process Industries, 22, 408e416.
Mbarki, A., Nguyen, Q. B., & Mercier, F. (2009). Structural fragments and explosions
in industrial facilities. Part 1: projectile trajectory and probability of impact.
Journal of Loss Prevention in the Process Industries, 22, 417e425.
Mbarki, A., Nguyen, Q. B., Mercier, F., Ami Saada, R., Meftah, F., & Reimeringer, M.
(2007). A probabilistic model for the vulnerability of metal plates under the
impact of cylindrical projectiles. Journal of Loss Prevention in the Process
Industries, 20, 128e134.
Mbarki, A., Nguyen, Q. B., Mercier, F., Ami Saada, R., & Reimeringer, M. (2008).
Reliability analysis of metallic targets under metallic rods impact: towards
a simplied probabilistic approach. Journal of Loss Prevention in the Process
Industries, 21, 518e527.
Neilson, A. J. (1985). Empirical equations for the perforation of mild steel plates.
International Journal of Impact Engineering, 3, 137e142.
Nguyen, Q. B., Mbarki, A., Ami Saada, R., Mercier, F., & Reimeringer, M. (2009).
Integrated probabilistic framework for domino effect and risk analysis.
Advances in Engineering Software, 40, 892e901.
Online MBA Library. (October 2010). Part 6: condence intervals of discrete data
and hypothesis testing. Available from http://doc.mbalib.com/view/
820e6d7cabf988fed9a8c2db212e9f9c.html Cited 25.04.11.
Pietersen, C. M. (1988). Analysis of the LPG disaster in Mexico City. Journal of
Hazardous Materials, 20, 85e107.
Qian, X. M., Xu, Y. B., & Liu, Z. Y. (2009). Monte-Carlo analysis of fragments projectile
from spherical tank BLEVE. Huagong Xuebao, 60, 1057e1061, (in Chinese).
Scilly, N. F., & Crowther, J. H. (1992). Methodology for predicting domino effects
from pressure vessel fragmentation. In Proceedings of the international conference on hazard identication and risk analysis. Human Factors and Human
Reliability in Process Safety.
Shen, L. (2010). Report of area risk assessment for Nanjing Chemical Industry Park of
China. Nanjing: Jiangsu Academy of Safety Science & Technology. Manuscript.
Stawczyk, J. (2003). Experimental evaluation of LPG tank explosion hazards. Journal
of Hazardous Materials, B96, 189e200.
Tulacz, J., & Smith, R. E. (1980). Assessment of missiles generated by pressure
component failure and its application to recent gas-cooled nuclear plant design.
Nuclear Energy, 19, 151e164.
Van den Bosch, C. J. H., & Weterings, R. A. P. M. (1997). Methods for the calculation of
physical effects (yellow book, CPR14E). The Hague, NL: Committee for the
Prevention of Disasters.
Westin, R. A. (1971). Summary of ruptured tank cars involved in past accidents.
Chicago, IL: Railroad Tank Car Safety Research and Test Project. Report no.:
RA-01-2-7.
Zhang, X. M., & Chen, G. H. (2009). The analysis of domino effect impact probability
triggered by fragments. Safety Science, 47, 1026e1032, (in Chinese).