You are on page 1of 16

Fishing Vessels Responses in Waves under Operational Conditions

Tello Ruiz, Manases; Ribeiro e Silva, Sergio; Guedes Soares, Carlos


Centre for Marine Technology and Engineering (CENTEC),
Technical University of Lisbon, Instituto Superior Tcnico
Av. Rovisco Pais, 1049-001 Lisboa, PORTUGAL
guedess@mar.ist.utl.pt

ABSTRACT: This study performs a seakeeping assessment of a set of fishing vessels,


and discusses their seakeeping performance under different operational conditions. The
dynamic response of the ships in waves depends on the wave climate of the ocean area
where the ship operates. The frequency domain transfer functions are obtained using a
strip theory method, and viscous roll damping corrections are introduced for different
advance speeds. Sensitivity analyses are carried out for a group of fishing vessels with
different hull forms and dimensions from the Portuguese and Peruvian fleets, operating in
the respective coastal areas. The results obtained allow the assessment of the ships
dynamic responses for the different ship types, advance speed, and wave encounter
angles. Finally, the most relevant performance criteria related with the absolute motions,
relative motions, accelerations, slamming, green water on deck, etc., are assessed against
the prescribed values to evaluate the seakeeping performance and to determine the limit
sea state that still allow standard operational conditions.
Key words: Seakeeping performance, fishing vessel, operability criteria, strip theory,
viscous roll damping.

1 INTRODUCTION
Fishing vessel operation is one of the most dangerous activities due to the ships random
movements and the severity of the environment, since waves degrade the ability to carry
out the mission comparatively to the calm water condition. In fact, there are many
situations in which working onboard can be impaired due to large amplitude motions and
accelerations. Seasickness, hazards and accidents including in some cases loss of lives
and the ship has been reported. Thus, good working conditions onboard are especially
important for fishing vessels, since the fisherman activity is already very demanding from
the physical point of view.
With the purpose of identifying the main causes of accidents statistical studies have been
conducted in the past. For example, a worldwide study on fishing vessel accidents
recorded during the 90s conducted by Wang et al. (2005), found most of the accidents
are related to ships with 24 meters of length overall (LOA), it has been shown machinery
damage as the most probable occurrence followed by foundering,

flooding and

grounding with probabilities of 65.97%, 15.41%, and 8.38%, respectively. With the
purpose of decrease the global number of accidents new intact stability rules of ships with
LOA less than 24 meters came into force, however despite the decrease, the number of
occurrences registered nowadays still remains high. Recent accidents were presented and
studied by Perez-Rojas et al. (2006) showing three Spanish fishing vessels accidents
resulting in loss of lives and the ships. In Portugal the predominance of small fishing
vessels in the national fleet provides yet another type of issues related to health and safety
at work which are also related to cultural aspects. Over a period of 20 years there have
been nearly 40,000 claims of which included more than 350 deaths. According to Anto
and Guedes Soares (2004), there is a real need to improve the operation and safety of
small vessels, because they represent 89% of accidents in the sample studied.
Seakeeping analysis is needed since it allows estimates of the ship responses and
evaluation of the conditions in which she is operating. To determinate if the ship is in
dangerous or if some performances are at the limit, many criteria can be adopted to assess
against the prescribed values. Of course the criteria and prescribed values are different
according to the ship type and purpose. Several seakeeping criteria related with the

absolute motions amplitudes and motions relative to sea, have been proposed and are
available in the literature and they have been obtained mostly from the experience
onboard of ships. Apparently there are no published criteria specific for fishing vessels,
and for this reason the authors have used a set of criteria that in fact were adapted from
other types of ships. In the present study the seakeeping quality of a set of fishing vessels
will be quantified by an operability index that accounts for:
The ship mission, through the choice of the most adequate seakeeping criteria.
The hydrodynamic and inertia characteristics of the ship, through the use of a seakeeping
program to calculate the motions transfer functions;
The wave climate where the ship operates, through the use of the significant wave height
and peak period obtained from the probability distribution of long term sea states together
with spectral techniques to calculate the ship responses in those sea states.
The method briefly described was utilized earlier by Guedes Soares et al. (1995) to
investigate the seakeeping performance of a group of fishing vessels operating in the
Portuguese coast. Fonseca and Guedes Soares et al. (2002) applied the same method to
investigate the sensitivity of the expected ships availability to different seakeeping criteria.

2 THEORY.
2.1 Regular Waves
The first step is to obtain the transfer functions (TF) of the rigid body motions and also of
the relative motion and accelerations in selected points in the ship. The T.F. represents
linear ship responses to harmonic waves. This investigation uses a seakeeping code based
on the strip theory formulation from Salvesen et al. (1970) to calculate the potential flow
hydrodynamic coefficients and harmonic wave exciting forces. The equations of motion are
solved in the frequency domain to obtain the five degrees of freedom oscillatory motions
(strip methods neglect all surge forces). In this methodology some restrictions need to be
taken into account when it is applied. Namely, the ship has to be slender, the amplitude of
the exciting waves and oscillatory motions must be small, and the Froude number less than
0.3 (Fn<0.3). Moreover, interactions between fishing gear components with the vessel
during haul or towing operations or water-on-deck effects have been all ignored so that

fishing vessels are assumed as free-floating bodies. Therefore, rigid body oscillatory
motions are calculated by solving a set of coupled linear differential equations, which
represent the equilibrium:

{(M
6

j =1

kj

+ Akj )&&j + Bkj & j + C kj j = Fk e iet

k, j = 1,...,

(1)

These excitation forces Fk and ship motions j can be conveniently represented on a right
handed Cartesian coordinate system, X = ( x, y, z ) , fixed with respect to the mean position
of the ship and origin in the plane of the undisturbed free surface. As shown in figure 1, the
translatory displacements in the x , y , and z directions are respectively surge 1 , sway

2 , and heave 3 , while the rotational displacements about the same axis are respectively
roll 4 , pitch 5 , and yaw 6 , where subscripts, k, j indicate forces in the k -direction due
to motions in the j -mode. M kj are the components of the mass matrix for the ship, Akj and

Bkj are the added mass and damping coefficients, C kj are the hydrostatic restoring
coefficients and Fk are the complex amplitudes of the exciting forces.

Figure 1: The coordinate system and six modes of motion, and definition of the ship
heading angle.
The harmonic j -th response of the vessel, j will be proportional to the amplitude of the
exciting force, at the same frequency but with phase shift, j , and is then given by:

j (t ) = ja cos( e t + j ), j = 1,..,6

(2)

If the ship travels at a speed U making an angle with the direction of incoming waves
(Figure 1), she will encounter regular wave crests with a frequency of encounter, given by:

e = kU cos

(3)

The encountered free surface is given by:

w = wa cos k [x cos + y sin (c U cos )t ]

(4)

2.2 Derived Responses


Since the seakeeping criteria are defined also in terms of accelerations felt onboard and
relative motions between the hull and the waves, these quantities are calculated for all
relevant points on the ship, combining the absolute motions and the wave motions. For

r
r

small motion amplitudes, the displacement vector at any point with coordinates

belonging to the ships can be obtained by applying simple kinematics equations (5). In this
case the six degrees-of-freedom modes described as harmonic functions are divided into
r
r
two groups: = ( 4 , 5 , 6 ) and R = (1 , 2 , 3 ) , corresponding to the rotational and
translational displacement vectors, respectively:
r

rr = R + r

(5)

r
Hence, the components of the absolute displacement vector = ( X P , YP , Z P ) at any point
P with coordinates

r
r are given by:

X P = 1 + z p 5 y p 6
YP = 2 + x p 6 z p 4

(6)

Z P = 3 + y p 4 x p 5
2.2.1. Vertical Displacement and Acceleration

The performance of people onboard is affected by those oscillations. To describe the


vertical displacement of a point P, the following representation may be adopted:
Z P (t , e ) = Z pa cos( e t + p )

(7)

In order to simplify this study, the motions in equations (3) are now restricted to the ships
centre line ( y p = 0) , and, therefore, amplitude and phase of the vertical displacement can
be easily determined by the following equations:
Z pa =

( ) + (x )
a 2
3

a 2
5

2 x p 3a 5a cos( 5 3 )

3a sin( 3 ) x p 5a sin( 5 )

p = tan a
cos( ) x a cos( )
p 5
3
5
3
Finally, since Z P (t , e ) is an harmonic function the vertical acceleration will be:
1

2
Z&&P = e Z P

(8)
(9)

(10)

2.2.2. Lateral Displacement and Acceleration


Similarly to vertical acceleration, lateral acceleration should also be taken into account.
From equations (6), the lateral motion of a point P is given by:
YP = 2 + x p 6 z p 4

(11)

In this case, the same procedure for the vertical motion has been adopted. However, as
described by Journe et al. (2001), another component due to gravity, g, should be added to
the lateral acceleration. Thus the total lateral acceleration that the vessel (crew, equipment,
cargo, or ship structural parts) feel is given by:
Y&&P = e 2YP + g 4

(12)

where:
YP (t , e ) = Y pa cos( e t + p )

2.2.3. Vertical Relative Displacement and Velocity


Before relative ships displacement or velocity have been determined, it is necessary to
obtain the absolute motion and wave amplitude P (t , e ) at the point P under
consideration. Therefore, relative motion can be obtained from:

Z PR (t , e ) = Z PR (t , e ) P (t , e )

(13)

Equation (13) may also be re-written as:


a
Z PR (t , e ) = Z PR
cos( e t + PR )

(14)

where:
Z

a
PR

(Z ) + ( )
a 2
P

a 2
P

Z Pa sin( p ) Pa sin( kx p )
2 Z cos( kx p p ) ...; p = tan a
Z cos( ) a cos( kx )
p
P
p
P
a
P

a
P

Also, the relative vertical velocity at any point P with coordinates

r
r is given by:

a
a
Z& PR
= e Z PR

(15)

2.3. Irregular Waves


To determine the response of the ship to real (irregular) sea states a spectral linear
formulation is adopted, following the early proposals of St Dennis and Pierson (1953). The
Pierson-Moskowitz spectral form for fully developed seas (Pierson and Moskowitz (1964))
describes the irregular sea states in terms of significant wave height H S and peak period
TP . In irregular seas the encountered wave profile is given by:
N

w =
n =1

a
wn

n 2

n 2

(
)
cos
x cos + y sin n
U cos t + n
g

(16)

where N is the number of component waves, n the circular frequency, n the random
phase angle and wan the amplitude of the n -th component waves, which are determined
from the wave spectrum SW ( ) . Because the system is linear, the relationship between the
wave spectrum and that of the j -th response and is given by:
S j ( ) = H j ( )

SW ( )

(17)

where H j ( ) is the transfer function from wave elevation to the j -th mode. The variance
of a record is given by the zero order moment of each ordinate, as follows:

m0 j = 2 = S j ( ) de

(18)

which is applicable to both the input and the response spectrum since the sea state is
modelled as a stationary, zero mean, Gaussian process and because the responses are linear,
the same model describes the response process. This implies that a Rayleigh distribution
describes the amplitudes or the peaks of the processes, according to which the probability
of exceeding the level r is given by:
r2

Q S (r ) = exp
2
2

(19)

Different statistics can be derived from the assumption of the Rayleigh distribution (see
Longuett-Higgins et al. (1952). For example, the average of the one-third larger
amplitudes, usually called the significant value rS , is given by:
rS = 2

(20)

The most probable maximum value in N successive cycles is obtained from (16):
rmax = 2 2 ln N

(21)

2.3.1. Seakeeping Performance Criteria


If the criterion is defined as a probability of exceeding a critical value pCR , then the related
root mean square of the response is obtained from equation (21), and given by:

CR =

2
rmax
2 ln(1 / pCR )

(22)

where rmax is the limiting magnitude of the response which has the probability pCR of
being exceeded. As an example, for green water on deck phenomenon rmax occurs when
local relative motion is larger than the local freeboard. A propeller emergence occurs when
1/4 of the propeller diameter comes out of the water. According to Ochi et al. (1964), a
slam occurs as a result of two occurrences: when the relative motion is larger than the mean
draught ( T ) at the bow and at the same time the relative vertical velocity has to exceed the

velocity of 0.093 Lg . Therefore, using equation (22), the criteria for slamming is given
by:
P ( slam) = P( Z P > T ) P( Z& PR > m)

(23)

where:
T2
P ( Z p > T ) = exp
2
2

m2

&
...; P( Z PR > m) = exp
2
2

; m = 0.093 Lg

3 SEAKEEPING ANALYSIS.
3.1. Fishing Vessels.
The present study is an analysis of a group of four fishing vessels with different hull
forms and dimensions from the Portuguese and Peruvian fleet. Fishing vessels designated
FV1 and FV2 belong to Portugal, and FV3 and FV4 to Peru. The main characteristics
relevant for the seakeeping calculations and the hull forms of these four fishing vessels
are presented in Table 1 and Figure 2.

Table 1: Main particulars of the fishing vessels.


FV Nr
Lwl (m)
B (m)
D (m)
T (m)
D (ton)
LCG(m)
Cb
Kx/B
Ky/Lpp
Kz/Lpp
GMt (m)

1
2
20.9 24.5
6.5
8
4.4
3.5
2.46 3.09
129.5 418.2
-1.35
0
0.4
0.61
0.4
0.4
0.25 0.25
0.25 0.25
1.14
0.6

3
37.4
7.3
3.52
3.03
552.8
-4.46
0.63
0.4
0.25
0.25
1.132

4
45.7
10.1
5.31
4.97
1258
0.05
0.53
0.4
0.25
0.25
1.284

Figure 2: Body plans of the four fishing vessels.

3.2 Seakeeping Performance, Operational Criteria, and Operability Indexes


The seakeeping criteria are defined either by a maximum allowed standard deviation of
the response, or in terms of maximum allowed probability of exceeding a critical value.
Comparing the ship responses in irregular waves with the seakeeping criteria, it is
possible to obtain the maximum significant wave heights in which the ship can operate.
Studies for fishing vessel related to the prescribed values were not found in the literature
except for Odabasi et al. (1991) who showed, vertical acceleration for fishing, defined as
15% of thee acceleration of the gravity.
The derived responses are calculated for specific points in the ship, and they are related to
the seakeeping criteria adopted. Table 2 presents the criteria selected and the respective
positions analyzed. These locations are described using the reference coordinate shown in
Figure 1.
Table 2: Prescribed seakeeping performance criteria applied to the fishing vessel.
Location x,y,z
Response

FV1
x

1 Green Water Deck


2 Slamming
3 Propell. Emergence
4 VA at Bridge
5 LA. at Bridge
6 VA at Work. Deck
7 LA at Work. Deck
8 Roll
9 Pitch

FV2

[m]
FV3

Criterion

FV4
z

10.6

3.12

12.3

2.8

19.6

3.04

23

3.43 5% (prob)

9.36

-2.3

11.1

-2.9

15.5

2.64

20.1

-4.8

3% (prob)

-8.6

-1.4

-9.4

-1.9

-16

-1.7

-19

-2.7

15% (prob)

5.3

2.5

6.13

1.6

9.82

1.45

11.5

1.37 0.2g (rms)


1.37 0.1g (rms)

5.3

2.5

6.13

1.6

9.82

1.45

11.5

-4.3

1.94

-4.9

0.5

-7.9

0.49

-9.2

0.41 0.15g (rms)

-4.3

1.94

-4.9

0.5

-7.9

0.49

-9.2

0.41 0.1g (rms)

6 (rms)

3 (rms)

Figures 3 to 14 present the curves of maximum significant wave heights, as function of


the seastate zero up-crossing wave period. Each graph corresponds to one heading with
respect to the direction of propagation of waves, and includes several curves
corresponding to several seakeeping criteria. The ship is operational for the seastates that
are bellow of all the curves. Figures 3 to 8 and 9 to 14 belong to FV1 and FV4,
respectively, and these graphs show the performance in head, quartering and beam seas
for two distinct speeds (zero and top speed). The graphs show clearly which are the
responses that most limit the ship operability, and for which headings, namely the pitch,

vertical accelerations and GW for head and quartering waves (specially, when ships
speed is increased) and the roll motion and lateral acceleration for beam waves.

Figure 03: Maximum allowed significant wave height.


All criteria for FV1 at =180, Fn = 0.00.

Figure 05: Maximum allowed significant wave height.


All criteria for FV1 at = 150, Fn = 0.00.

Figure 07: Maximum allowed significant wave height.


All criteria for FV1 at = 90, Fn = 0.00.

Figure 04: Maximum allowed significant wave


height. All criteria for FV1 at =180, Fn = 0.30.

Figure 06: Maximum allowed significant wave


height. All criteria for FV1 at = 150, Fn = 0.30.

Figure 08: Maximum allowed significant wave


height. All criteria for FV1 at = 90, Fn = 0.30.


Figure 09. Maximum allowed significant wave height.
All criteria for FV4 at = 180, Fn = 0.00.

Figure 11: Maximum allowed significant wave height.


All criteria for FV4 at = 150, Fn = 0.00.

Figure 13: Maximum allowed significant wave height.


All criteria for FV4 at = 90, Fn = 0.00.

Figure 10 Maximum allowed significant wave height.


All criteria for FV4 at = 180, Fn = 0.30.

Figure 12: Maximum allowed significant wave height.


All criteria for FV4 at = 150, Fn = 0.30.

Figure 14: Maximum allowed significant wave height.


All criteria for FV4 at = 90, Fn = 0.30.

Finally with the probability distribution of short-term seastates for a given ocean area, it
is possible to select all seastates where the ship is operational. Summing up the

probabilities of occurrence of these seastates, one obtains the expected probability that
the ship operates satisfying all the seakeeping criteria. This probability is the seakeeping
index.
For an accurate realistic calculation of the ship operability it would be necessary to
account for the distribution of wave directionality. Here the procedure was simplified by
neglecting the wave directionality on the wave climate statistics, and assuming that the
ship encounters all headings with the same probability.
Table 3 is a summary of the ship operability indexes obtained for all fishing vessels
considering the criteria shown in Table 2. These indexes represent the expected
percentage of time that the ship is operational, meaning the expected percentage of time
that the ship satisfies the seakeeping criteria. For each vessel, different rows show the
results for each heading of the vessel with respect to the waves. The last row presents the
average value, which represents the global operational index of the vessel.
Portuguese and the Peruvian ships are analyzed operating in their operational areas
Atlantic Ocean and Pacific Ocean, respectively. The scatter diagrams used in here
describe the wave climatology for these two specific ocean areas (area 16 and 65 in,
Hogben et al. (1986)). In the absence of a well defined mission speed profile for each
vessel, two most representative velocities are considered, which correspond to Fn = 0.00
(dead condition) and Fn = 0.30 (sailing condition), and seven possible heading angles are
analyzed.
Results presented in Table 3 confirms beam seas as the worst-case scenario in which the
ship will operate. For example, FV1 loaded at light seagoing condition encountering
beam waves at Fn = 0.00 will be able to operate complying with all the criteria only 14%
of the time in one year (see column shaded in grey). It can also be noticed that roll and
the lateral acceleration (LA) are the ones most limiting the operability for all the ships for
both velocities.
In beam waves, the operability of FV1, FV2, FV3 improve when the ships speed is
increased and this is because the roll damping increases with the forward speed. On the
other hand, the vertical ship responses (pitch motion) of FV4 increase with the forward
speed in head and bow waves, as well as the related slamming and deck wetness

occurrences. For this reason the ship operability degrades from 0.94 to 0.87 for these
headings when the speed increases from Fn = 0.00 to Fn = 0.30.
Table 3: Seakeeping performance and operability indexes of the Fishing Vessels at Fn = 0.0 and 0.30.
Global wave statistics area 16 for FV1, FV2 and area 65 for FV3 and FV4.

Comparing roll motion and lateral accelerations of FV1 and FV2, it is interesting to
confirm that two ships with similar main dimensions (Lpp and B), the hullform of U
type has enhanced seakeeping performance than the V type one. FV4 having much
larger main dimension than FV3 also present better seakeeping performance. Moreover,
FV1 with LOA less than 24 meters should be operated only coastal waters and fishing
trips (dead condition) in rough weather conditions should be clearly avoided. As a
general conclusion, the seakeeping qualities of the fishing vessels are adequate. However,

they could be improved by reducing the roll motion amplitudes. One possibility that
deserves to be considered is the use of efficient systems to control the roll motion.
Nevertheless, although it is implied that the vessel is operational when all criteria are
satisfied at a given ships speed. In practice, these vessels could be operational for longer
periods than the estimated values, because a Master informed in advance by a decision
support tool would change course or also reduce the speed to avoid stormy seas.
4. CONCLUSIONS
The assessment of the seakeeping quality of the four fishing vessels showed that in
statistical terms the percentage of operability is smaller for zero speed (dead condition)
than the top speed (Froude number 0.30). Roll motion and lateral acceleration are the ship
responses that mostly limit the ship operability. Moreover, it is important to precise that
most of the lowest operability indexes in beam seas are fundamentally due to large lateral
accelerations which have a strong impact on crew and the machinery systems
performance. In fact, these low operability indexes can be rationally utilized to explain
the amount of claims in the Portuguese and the Peruvian fleet, and also why some
researchers found machinery damage as one of the most probable occurrences.
Still with respect to safe operations, these fishing vessels should be also analyzed by
means of a non-linear time domain method on some specific dynamic instability
scenarios, such as parametric rolling and pure loss of stability in longitudinal waves, but
also water-on-deck and extreme forces at the fishing gear developed during the fishing
operations. Therefore, apart from statistical assessments of seakeeping performance of
the fishing vessels at their specific design points, it is also highly recommended to
improve safety of fishing vessels in waves by means of installation of a decision support
tool also based on the results of those non-linear time domain numerical assessments.
5. ACKNOWLEDGMENT.
This work has been performed within the project SADEP-Decision support system for
the safety of fishing vessels subjected to waves. The project has been financed by the
Foundation for Science and Technology (Fundao para a Cincia e a Tecnologia),
from the Portuguese Ministry of Science and Technology, under contract PTDC/EMEMFE/75233/2006.

6. REFERENCES.
Anto, P., Guedes Soares, C. 2004, Analysis of Fishing Vessel Accidents (in Portuguese), As Actividades
Martimas e a Engenharia. C. Guedes Soares and V. Gonalves de Brito (Eds.), Lisbon: Salamandra. pp.
627-643.
Guedes Soares, C., Fonseca, N., Centeno, R., 1995, Seakeeping performance of Fishing Vessels in the
Portuguese Economic Zone, Proceedings of the International Conference on Seakeeping and Weather,
RINA, London, UK, 28th February-01 March, paper Nr. 12, pp. 1-10.
Fonseca, N. and Guedes Soares, C., 2002, Sensitivity of the Expected Availability to Different Seakeeping
Criteria, Proceedings of the 21st International Conference on Offshore Mechanics and Arctic
Engineering OMAE2002, Oslo, Norway, 23-28 June, paper 28542.
Ikeda, Y., Himeno, Y. and Tanaka, N., 1978, A Prediction Method for Ship Roll Damping University of
Osaka Prefecture, Re-port No 405.
Journe and Massie, 2001, Offshore Hydrodynamics Delft University Technology, first edition, pp. 280390.
Longuet-Higgins, M.S., 1952, On the Statistical Distribution of the heights of Sea Waves, Journal of
Marine Research, Vol. 11, No. 3, pp. 245-260.
Ochi, M.K.(1964) Extreme Behavior of a Ship in Rough Seas-Slamming and Shipping of Green Water,
Transactions of the Society Naval Architects Marine Engineers (SNAME), Vol. 72.
Odabasi A.Y.,Fitzsimmons P.A., Ankudinov V.K. Wiley S.A. 1991. Seakeeping Considerations in Ship
Design and their Incorporation in HDDS. BMT International, Report No. HDDS.P2SPEC.
Prez-Rojas L., Arribas F.P., Rodrguez R.Z., Pacheco A.G. 2006, On the Accidents of Small Fishing
Vessels, Proceedings of the 8 th International Conference on the Stability of Ships and Ocean Vehicles,
STAB2006.. M. Neves (Ed), COPPE, UFRJ, Rio de Janeiro, Brasil, pp 669-676
Pierson, W. J., and Moskowitz, L., A., 1964, Proposed Spectral Form for Fully Developed Wind Seas
Based on the Similarity Theory of S.A. Kitaigorodskii, Journal of Geophysical Research, Vol. 69, pp.
5181-5190.
Salvesen, N., Tuck, E. O. and Faltisen, O 1970 Ship motion and sea loads, Transaction of the Society
Naval Architects Marine Engineers (SNAME), Vol 78, pp. 250-287.
St. Dennis, M. and Pierson, W. J., 1953, On the Motion Ships in Confused Seas, Transactions of the
Society Naval Architects Marine Engineers (SNAME), Vol. 61, pp. 280-354.
Stevens, S. C. and Parsons, M. G., 2002, Effects of Motion at Sea on Crew Performance: A Survey,
Marine Technology, Vol. 39, N 1, pp. 29-47.
Wang J., Pillar A., Kwon Y.S., Wall A.D., Loughran Rodrguez C.G., 2005, An Analysis of Fishing
Vessel Accidents, Accident Analysis and Prevention, 37: 1019-1024.

You might also like