Structural Engineering Documents

Ole Damgaard Larsen

SHIP COLLISION
WITH BRIDGES
The Interaction between
Vessel Traffic and
Bridge Structures
Copyright 8 1993 by
International Association for Bridge and Structural Enginneering

All rights reserved. No part of this book may be reproduced in any form or by any means, electronic or
mechanical, including photocopying, recording, or by any information storage and retrieval system, without
permission in writing from the publisher.

ISBN 3-85748-079-3
Printed in Switzerland

Publisher:
-
IABSE AIPC IVBH-
ETH - H6nggerberg
CH-8093 Zurich, Switzerland

Phone: Int. +411-3772647

from 27 September 1993:
Int. +411-6332647
Fax: Int. t411-3712131
PREFACE

In June 1983 a colloquium on Ship Collision with Bridges and Offshore Structures was held
in Copenhagen under the auspices of the International Association for Bridge and Structural
Engineering (IABSE). The colloquium was initiated by the Danish Group of IABSE with the
background of comprehensive investigations in this field in connection with the fixed crossing of
the Great Belt in Denmark.
The colloquium brought together bridge and offshore engineers, naval architects, navigational
experts, risk assessment specialists, etc. with the view to exchange information on this subject.
After the colloquium, the IABSE Working Commission 1 Structural Performance, Safety and
Analysis, established a Working Group for preparing a guideline or a state-of-the-art report.

Members of the Working Group, appointed originally or during the work, were the following:
- Dr. Yahei Fujii, Electronic Navigation Research Institute, Tokyo, Japan
- Mr. Cornelis Q. Klap, Rijkswaterstaat, Voorburg, The Netherlands
- Mr. Michael A. Knott, Greiner Inc., Richmond, USA
- Mr. Thomas R. Kuesel, Parsons Brinckerhoff, New York, USA
- Mr. Ole Damgaard Larsen, COWIconsult, Copenhagen, Denmark
- Dr. Henrik 0. Madsen, Danish Engineering Academy, Copenhagen, Denmark
- Mr. Holger S . Svensson, Leonhardt, Andra u. Partner, Stuttgart, Germany.

The Working Group was chaired by Ole Damgaard Larsen, who also prepared the present
IABSE publication. Yahei Fujii, Thomas R. Kuesel and Michael A. Knott made valuable contri-
butions to the publication.
The development of this publication took its basis from the proceedings of the international col-
loquium in Copenhagen in 1983.
Since the colloquium, important research has been carried out in connection with development
of national codes and standards for bridge design and in connection with planning and design of
major bridge projects.
It was decided in 1989 to leave out offshore structures from the publication. The reason was
that it was found inexpedient to treat the two types of structures in the same publication.

The final editing of the publication was carried out in the course of 1991, when conclusive
data could be made available from the following major projects:
- development of vessel collision guide specifications for the Federal Highway Administration
in the USA
- development of vessel collision design criteria for the Great Belt Crossing in Denmark.
 This document represents the conclusions of the work as accepted by the IABSE Working Com-
mission 1 at the IABSE annual meeting in St. Petersburg, September 1991.
The Working Group was dissolved at the same occasion and inquiries regarding this publica-
tion should therefore be addressed to the IABSE Working Commission 1.
The Working Commission 1 is greatly indebted to the author and much appreciates his efforts.
Johan Blaauwendraad,
Chairman of Working Commission 1
TABLE OF CONTENTS Page

0. Preface

1. Introduction
1.1 Background
1.2 Validity
1.3 Application

2. Initial Planning 7
2.1 Siting of Bridge Structure 7
2.2 Navigation Channel Layout 8
2.3 Overall Bridge Layout 8
2.4 Vertical Clearance 9
2.5 Horizontal Clearance 11

3. Vessel Traffic 19
3.1 Traffic Routes 19
3.2 Passage Statistics 20
3.3 Vessel Characteristics 21
3.4 Traffic Forecast 24

4. Risk Acceptance 27
4.1 Consequences of Collision 27
4.2 Disruption Risk Acceptance Criteria 28
4.3 Fatality Acceptance Criteria 32
4.4 Optimum Cost Acceptance Criteria 34

5. Collision Risk 37
5.1 Collision Risk Model 38
5.2 Causation Probability 39
5.3 Geometrical Probability 44
5.4 Failure Probability 49

6. Vessel Impact Forces 53

6.1 Head-on Bow Impact Forces 54
6.2 Sideways Impact Forces 70
6.3 Deckhouse Impact Forces 71
6.4 Local Collision Forces 72
6.5 Barge Vessel Impact Forces 73
7. Bridge Design 77
7.1 Energy Considerations 77
7.2 Structural Analysis 80
7.3 Design of Substructure 82
7.4 Design of Superstructure 83

8. Prevention Measures 85
8.1 Aids to Navigation 85
8.2 Vessel Traffic Regulations 87
8.3 Vessel Traffk Management Systems 87

9. Protection Measures 91
9.1 Fender Systems 91
9.2 Pile Supported Systems 93
9.3 Dolphin Protection 95
9.4 Artificial Island or Reef Protection 95
9.5 Floating Protection Systems 97

10. Protection of the Public 101

10.1 Collision Hazard Detection 101
10.2 Bridge Traffic Control 102

Appendices 103

A. Vessel Size and Geometry Data 103

B. Records of Collision Accidents 113
C. Selected Literature 119
 1

1. INTRODUCTION

1.1 Background

Any structure established in navigable waters constitutes a hazard to shipping and is itself
vulnerable to damage or destruction in the event of vessel collision. Among the most significant
structures exposed to this hazard are bridges crossing coastal or inland waterways.
A list of serious accidents recorded during the period 1960-1991has been included in Appen-
dix B. The list represents an updated version of the records published by Frandsen [l-11 in 1983.
The records indicate an average of one serious vesselbridge collision accident per year world-
wide, see Figure 1.1. More than 100 persons died in these accidents and large economic losses
were incurred directly in repair or replacement costs as well as indirectly in the form of lost trans-
portation service.
The photos in Figures 1.2, 1.3 and 1.4 from actual collision accidents illustrate the seriousness
of the subject. The accidents shown are all briefly described in Appendix B.

Figure I .I The number of serious vessellbridge collision accidents per year in the period
1960-1991 (32 years).
2 introduction

Many factors are involved in the problem of vesselbridge collision confronting most countries
around the world:
- The vessel traffic intensity has increased considerably in number as well as in size of vessels.
Since 1960 the worldwide water borne tonnage has increased at an average yearly rate of 4%.
- Bridges to cross navigable waterways are being planned and constructed at an increasing rate.
- Vessel impact loads have not been considered in the design of most bridges more than
25 years old.
- Vessels built today are often constructed with bulbous bows which increase the impact load.
- Bridges designed today, poorly sited and with inadequate navigation clearance, indicate that the
problem is still not fully appreciated.

This situation has led to increased concern over the safety of bridges crossing navigable
waterways and research into the vessel collision problem has been initiated in several countries
of the world in connection with:
- evaluation of vulnerability of existing bridges
- establishment of design criteria for new bridges
- development of national codes and specificationsregarding vesselbridge collisions.
References are made to [l-21 through [l-171.
The main purpose of the present publication is to promote the understanding internationallythat
the interaction between bridge structures and vessel traffk should be considered a decisive design
aspect because of concern for:
- the risk of disruption of the bridge
- the safety of the users of the bridge
- the safety of the vessels
- the risk of environmental pollution.
A list of important publications on the vesselbridge collision topic, available today, is
included in Appendix C.
 Background 3

Figure I .2 Vessellbridge collision accidents: at the top, the Maracaibo Bridge accident in
Venezuela in I964 (from Ostenfeld [ I - I 3 ] ) and below, the Hopewell Bridge accident
in Virginia, USA in 1977 (from US National Research Council [l-171)
4 Introduction

Figure I .3 Vessellbridge collision accidents: at the top, the 7'jornBridge accident in Sweden in
1980 (photo: Aksel G . Frandsen) and below, the Sunshine Skyway Bridge accident
in Florida, USA also in 1980 (fi-omKnott et al. [I -51).
 Background 5

Figure 1.4 Vessellbridge collision accidents: at the top, the Strangnas Bridge accident in
Sweden in 1990 (photo: Mats Johnsson) and below, the Carnafuli Bridge accident
in Burma in 1991 (courtesy of Dr. Y, Fujii).
6 Introduction

1.2 Validity

It has been found that the state-of-the-art of the vesselbridge collision problem is not so well
established that specific regulations for worldwide use can be prescribed at this time. Such a task
would also be impeded by national differences in design philosophies, code formats, etc. Thus,
this publication is intended to provide guidelines to enable engineers-faced with the question of
preventing or protecting vessels and structures from collision-to consider the subject in a ratio-
nal way, and to take advantage of up-to-date knowledge.
A basic concern during the preparation of this publication has been not to restrict ingenuity of
design. The information and advice presented should not be interpreted as covering all aspects of
vessel collision with bridges. Engineering judgment must always be used when the problems
related to a particular bridge site are to be evaluated.
It is expected that state-of-the-art philosophies, methodologies and key assumptions will
change with time as engineers gain more knowledge about the vessel collision problem. There-
fore, the lifetime of this document may be limited.
Prescriptive codes and standards on collision design criteria exist on a national basis and with-
in particular technical fields. This publication has benefitted from such codes and future codes
may benefit from the present publication.
Further development of specific design criteria on a national and international basis is encour-
aged in order to counteract inadequate practices by individuals and bodies not influenced by
IABSE publications.

1.3 Application

The present publication contains information and advice applicable for:

- planning and design of new bridges
- planning and design of navigation channels
- planning and design of prevention measures
- planning and design of protection measures
- upgrading or retrofitting of existing bridges
- upgrading or retrofitting of navigation channels
- evaluation of the safety of existing bridges
- evaluation of the safety of vessels
- evaluation of the safety of persons
- evaluation of the safety of environment.
 7

2. INITIAL PLANNING

As considerations for the interaction between bridge structures and vessel traffic will often in-
fluence fundamental decisions such as location and type of bridge crossing, it is essential that the
vessel collision aspects are properly investigated as early as possible in the planning process.
A number of important aspects of the bridge design, siting, and aids to navigation can be eval-
uated by relatively simple means on the basis of the initial knowledge of the waterway and the
navigation, applying experience and common sense.
This chapter contains such general considerations concerning the initial planning and layout of
the bridge structure and the waterway.

2.1 Siting of Bridge Structure

The purpose of the bridge often determines a specific location. However, minor modifications can
normally be introduced and if major problems are identified it might still be possible to consider
alternative locations.

For the siting of a bridge crossing, the following aspects, among others, should be considered,
ref. [2-11, [2-23, [2-31, [2-131, [2-141, [2-151, and [2-181:
- Locations with congested navigation should be avoided.
- Locations with difficult navigation conditions (shoals, cross currents, etc.) should be avoided.
- A straight and unencumbered navigation channel approach of adequate length before the
bridge-passage should be achieved.
- Adequate distance to locations where berthing manoeuvres take place should be provided.
- The bridges alignment should preferably be perpendicular to the navigation channel.
- The centre of the navigation span should coincide with the centre line of the navigation
channel.
- Locations where bridge piers can be placed in shallow water so that vessels out of control may
not reach the bridge structures without first running aground should be preferred.

With regard to the adequate length of unencumbered channel approach, the following empiri-
cal estimate has been reported by Shoji et al. [2-181, [2-191, and [2-201 based on analyses of col-
lisions with bridges world-wide: The minimum distance from a bridge line to the position of the
nearest turn in the navigation route should be at least 8 L and preferably 20 L, L being the length
of the vessel. If the distance is smaller, the turn will influence the navigation at the bridge crossing.
8 Initial Planning

2.2 Navigation Channel Layout

The aim of the bridge designer to attain optimal safety against vessel collision will as a rule also
lead to requirements for the layout and operation of the navigation channel. The bridge
designer is, however, often limited in this respect since the responsibility and authority for
implementing such navigation improvements often belongs to navigation authorities.

Regardless of the question of design responsibility, the following indicates various aspects of
navigation channel layout to be considered by the designer:
The waterway should have clearly defined and properly declared navigation channel(s).
If the traffic density is high, i.e. involving frequent meeting or overtaking close to the bridge
crossing, two-way traffk should be established by introduction of a traffic separation scheme,
If conditions allow, the traffk from minor and leisure vessels should be separated from that of
large (merchant) vessels.
The water depth in the navigation channel should be at least 1.2 times the maximum vessel
draught.
The direction of the navigation channel and the location of bends may be adjusted to improve
the navigation conditions in general.

~ . 3Overall Bridge Layout

The primary area of vessel collision risk to the bridge is the region near the navigation span.
Therefore the layout of the bridge in this region should be developed to maximize the horizontal
and vertical clearance for navigation. In section 2.4 and 2.5 empirical rules are outlined with the
aim of ensuring that the clearance of the navigation span is sufficient to provide safe passage
under normal conditions.
The philosophy is that collision with the bridge structures should only occur as a result of
navigation error or technical failure on board, possibly in combination with low visibility and
adverse weather conditions, and not because of particular navigation difficulties created by the
presence of the bridge. It should be realized that all parts of a bridge crossing situated where the
water depth allows vessels, including vessels in ballast, to navigate are exposed to collision risks.
The lengths and heights of approach spans should be planned considering these risks.

During the planning phase, thought should also be given to protection alternatives in cases where
a cost-effective design of the bridge structure cannot be achieved directly. The alternativesinclude:
- pier fender systems to reduce impact force and energy
- independentprotection structures, i.e. artificial islands or reefs, dolphins, floating arrestors, etc.
to withstand or redirect the colliding vessel.

The development of a suitable overall bridge layout is an iterative process comparing different
alignments and differing degrees of protection for alternative bridge solutions.
 Vertical Clearance 9

2.4 Vertical Clearance

According to international practice, the vertical clearance in the navigation span of a bridge
crossing should be planned to allow the passage of the highest vessel (incl. equipment such as
mast, antennae etc.) using the waterway at that point in time when the construction of the bridge
is planned and made public to the users of the shipping lane. The clearance should permit the
passage of the highest vessel in a ballast condition at high water level with due allowance for
vertical vessel movements.
Data on air draughts of vessels are not available in public ship registers and therefore have to
be obtained from owners and/or shipyards. For preliminary planning purposes the typical vertical
clearance dimensions of different types of vessels, shown in Appendix A, may be used.
The scatter in the vertical dimensions is very considerable as illustrated in Figure 2.1. The
Figure shows chimney heights, which are less than the required air draught but represent the
highest part of the vessels able to damage the bridge in case of collision.

Vessel Chimney Heights ( m )

60

50

40

30

20

10

0
0 m,o0O 3OOlOOO 4OOlOOO 500,~
DWT (Toms )

Figure 2.1 Typical heights of vessel chimneys above waterline of different types and sizes of
vessels in ballast. To arrive at the air draught, the height of auxiliary equipment such
as mast and antennae has to be added.
10 Initial Planning

For the existing world fleets of merchant and naval vessels a vertical clearance of 65 m above
high water level will suffice. The vertical clearances of long span bridges world-wide are listed in
Table 2.1 for comparison.

Bridge Name Country Completion Navigation Vertical

Year Span Clearance
(m) (m)
George Washington USA, New York 1931 1067 65
West Bay USA, California 1936 704x2 65
Golden Gate USA, California 1937 1280 67
Bronx-Whitestone USA, New York 1939 70 1 46
Tacoma Narrows USA, Washington 1950 853 56
Mackinac USA, Michigan 1957 1158 45
Forth Great Britain, Scotland 1964 1006 52
Verrazano Narrows USA, New York 1964 1298 69
Severn Great Britain, England 1966 988 37
Tagus Portugal 1966 1013 70
Angustura Venezuela 1967 712 64
Kanmon Japan, Honshu-Shikoku 1973 712 61
Bosphorus (1") Turkey 1973 1074 64
Humber Great Britain, England 1981 1410 30
Innoshima Japan, Honshu-Shikoku 1983 770 50
Ohnaruto Japan, Honshu-Shikoku 1985 876 41
Minami Bisan-Seto Japan, Honshu-Shikoku 1988 1100 65
Kita Bisan-Seto Japan, Honshu-Shikoku 1988 990 65
Shimotsui-Seto Japan, Honshu-Shikoku 1988 940 31
Bosphorus (2"d) Turkey 1989 1090 64

Table 2.1 Vertical clearances of long span bridges world-wide.

It is to be noted, however, that certain special-purpose vessels, such as crane vessels, offshore
drilling rigs during transport and others, may reach more than 100 m above sea level.
In connection with the construction of the Danish Great Belt East Bridge across the intemation-
a1 navigation route between the Baltic Sea and the North Sea, the question of free passage of very
high offshore drilling rigs was referred to the International Court of Justice in the Hague in 1991
by Finland. In 1992 the case was withdrawn from the court after an agreement had been reached
between the Finnish and Danish Governments implying that the vertical clearance of 65 m could
be kept.
 Horizontal Clearance 11

2.5 Horizontal Clearance

As the risk of vessel collision depends decisively on the horizontal clearance of the navigation
span, this should be as large as can be technically and economicallyjustified.

An estimate of the horizontal clearance necessary to obtain a high level of safety under normal
conditions of passage can be achieved by empirical methods. The following methods are available:
- consultation of literature or codes (may not be reliable and up-to-date)
- analysis based on observations of navigation behaviour, e.g. ship domain
- manoeuvring simulations (should preferably be applied after the overall layout of the bridge
has been decided).

As two-way traffic is much more demanding than one-way traffk, it has to be established if the
vessel traffic density is so high that a two-way situation is required. For this purpose a method has
been developed based on the ship domain theory, i.e.: close encounter analysis (described at
the end of this section).

Literature or Codes
Based on the ship domain theory, the Japanese Government passed a Maritime Safety Law
[2-111 in 1973, requiring that the minimum width of a fairway for international vessel traffic is
700 m for one-way traffic and 1400 m for two-way traffic.
Shoji et al. [2-181, [2-191and [2-201have investigated the relation between vessel size and hor-
izontal span clearance for actual cases of vesselbridge collision accidents. It was found that in all
investigated cases the span clearance was less than 3 times the ship length. Assuming one-way
traffic for the investigated cases, this is in good agreement with the ship domain theory.
Based on the papers by Shoji et al., the US Guide Specification [2-141 contains the following
guideline for the planning of navigation span clearance: Bridges with main span less than 2 or 3
times the design vessel length are particularly vulnerable to vessel collision.

Ship Domain Analysis

Observations have shown that in order to navigate safely, the master of a vessel tries to keep a
certain distance from other vessels, fixed objects, shallow water, etc. The distance varies with the
vessel speed, visibility, type of encounter, and a number of other navigational aspects. This meth-
od of explaining navigational behaviour is called the Ship Domain theory. The domain concept
was introduced by Fujii et al. t2-81 based on comprehensive studies of navigation patterns in
Japanese waterways and Goodwin [2-101 for use in open sea situations.
In this theory, the Bumper Area is the area a vessel actually occupies in a waterway including
a zone around the vessel within which other vessels bumper areas should not overlap. Observa-
tion shows that vessel encounters with overlapping bumper areas often result in unwanted
 12 Initial Planning

na

..
Figure 2.2 Vessels and respective bumper areas in a waterway where navigation is impeded by
a bridge with a narrow navigation opening.

evasive manoeuvres involving increased accident risk. Figure 2.2 shows two vessels approaching
a bridge with a narrow navigation opening. The indicated sizes of the bumper areas show that the
encounter is critical.
Fujii and several other Japanese researchers have estimated the size of the bumper area on the
basis of data obtained through radar observations. Reference is made to [2-51, [2-61, 12-81,
[2-161, and [2-21]:
- For waterways with sufficient width to provide free navigation at service speed (5-8 m/sec.)
and with no obstructions (islands, shallow water, etc.) in the channel, the following average
values for the size of the bumper area (approximatedby an ellipse) has been found: 8.0 L in the
course direction and 3.2 L in the side direction, L being the length of the vessel.
- In narrow channels and harbours, where the conditions require vessels to travel at a reduced
speed (3-4 dsec.) and where no head-on encounters, overtaking or crossing encounters take
place, the following average bumper area size has been found: 6.0 L in the course direction and
1.6 L in the side direction, L being the length of the vessel.

The above results have been derived from waters with a high traffk density and with a large
fraction of small vessels. Open sea situations with lower traffic density have been studied by
Goodwin [2-lo], resulting in larger bumper area sizes than discussed above. For confined,
protected waterways with very high traffic density, such as narrow rivers or harbour passages,
smaller bumper area sizes than mentioned above may be relevant.
 Horizontal Clearance 13

For a bridge crossing a navigation channel, it is thus suggested that the requirement for the
navigation span clearance should be related to the width of the bumper area of a typical large
vessel passing the bridge. The typical large vessel may be selected as the largest vessel which is
able to safely pass the bridge without assistance, provided that larger vessels have pilots on board
or their passages are regulated by a Vessel Traffic Service system.

Taking the typical large vessel as the design vessel, with the length LD,the following require-
ments for the horizontal clearance of the navigation span may then be derived from the bumper
area widths indicated above:
- In case of one-way traffic the horizontal clearance C should be equal to the width of the bumper
area of the design vessel, i.e. C = 3.2 LD for waterways with vessels travelling at service speed
and C = 1.6LD for restricted waters.
- In case of two-way traffic the horizontal clearance should be equal to the width of two bumper
areas of two meeting design vessels plus a separation zone between the two bumper areas. The
necessary width of the separation zone has been addressed by Fujii et al. in [2-71, suggesting a
separation zone width of 0.3 b - 1.8 b.Thus, the horizontal clearance C should be: C = 6.7
LD - 8.2 LD for waterways with vessels travelling at service speed and C = 3.5 & - 5.0 & for
restricted waters.

Figure 2.3 illustrates the above horizontal clearance indications. It is suggested that these clear-
ance requirements be used in the planning stage for important bridges crossing open deep waters,
where it is physically possible for vessels to depart substantially from the nominal sailing route.
In addition to these general rules, the local navigation aspects, i.e. risk-increasingeffects such as
nearby bends and risk-reducingeffects such as special navigational aids, should be accounted for.
When complying with these requirements and suggestions, the presence of the bridge is expected
not to give rise to particular difficulties to the navigation, and the risk of vessel collision may be
expected to be acceptably low. This point of view is supported by the results of an analysis of
vessel collision accidents involving North American bridges, ref. Frandsen et al. [2-31, where a
marked lower collision frequency for bridges with spans fulfilling the above proposed
requirements to the horizontal clearance compared to bridges with spans not fulfilling these
requirements have been found, see Figure 2.4.
Most existing bridges do not fulfil the above suggested requirementsto the horizontal clearance
and for many new bridges, economy does not allow full compliance either.
In cases where the bridges span relatively narrow channels, for which the available width of
water deep enough to float large loaded vessels limits the "domain" of these vessels, the suggested
requirements are not fully relevant and the collision risk may still be acceptably low.
In other cases where navigation clearances have been dictated by economy rather than collision
risk considerations, more cautious navigation practices by vessel masters and pilots are required
and attention should be given to collision preventive and protective measures in order to keep the
collision risk at an acceptable low level.
14 Initial Planning

Figure 2.3 Horizontal clearance requirementsfor important bridges according to the domain
theory for one-way trafSic and two-way traftk respectively
 Horizontal Clearance 15

Number of Collisions per 10,OOO Passage6

0 1 2 3 4
wngHorlzontelaeelranw/
Clearance requlred by Domaln Theory

Figure 2.4 Number of collision accidents per vessel passage for a series of North American
bridges versus the ratio between the actual horizontal clearance and the (clear-
ances) suggested by the domain theory (based on Frandsen et al. [2-31).
16 Initial Planning

Manoeuvring Simulations with Pilots

Computerized navigation simulators are able to model vessel manoeuvring capabilities very
realistically in the sense that it is possible to take most important factors into account, e.g.:
- wind
- waves
- current
- visibility
- characteristics of vessels
- vessel overtaking and meeting
- geometry of waterway
- geometry of bridge structures
- location and type of navigational aids.

This means that pilots are able to navigate almost as if they were sailing a real vessel in a real
navigation channel, ref. Figure 2.5 and Figure 2.6.
Manoeuvring simulations with professional navigators carried out in cooperation between
bridge designers and maritime authorities have proven to be an efficient and reliable tool for the
evaluation of requirements for horizontal bridge clearances, ref. Bay et al. [2-11, Frandsen et al.
[2-41, Gardenier [2-91, and Meurs et al. [2-151.

Figure 2.5 A ship simulator set upfor the investigation of two-way bridge passage (fromBay et
al. [2-13]).
 Horizontal Clearance 17

\
\
\

\
\
\
Figure2.6 Typical plots showing the results of manoeuvring simulations (from Bay et
al. [2-131).

Close Encounter Analysis

A way to evaluate if the one-way traffic requirement to the navigation span clearance suggested
by the domain theory is sufficient is to estimate how often a situation arises where two vessels
meet which are so large that the clearance is not sufficient to provide safe passage. Such meetings
of anti-directional vessels in the vicinity of the bridge are referred to as Close Encounters.
Since several of the involved parameters depend on the vessel type and size, it is convenient to
use a Monte Car10 simulation approach in order to calculate the yearly expected number of close
vessel encounters as a function of the horizontal clearance of the bridge.
The simulation of the occurrence of vessels of different sizes in the vicinity of the bridge
requires a statistical description of the vessel traffic. The Poisson process is generally accepted
as a relevant description of such events.
The method is described in Ostenfeld-Rosenthalet al. [2-171 and Frandsen et al. [2-31. Final
advice about an acceptable level of annual number of close encounters for maintaining one-way
conditions cannot generally be given, but should be defined in the particular case by the authori-
ty concerned. An example of the results of such analysis is illustrated in Figure 2.7.
18 Initial Planning

I Number of Cbm Encounters per Year

I I I I

-
1
 19

3. VESSEL TRAFFIC

The characteristics of the vessel traffic passing under the bridge should be established in as
much detail as possible for the design stage in question. Normally, the t r a f k characteristics in a
specific situation can be established by utilizing available statistics and information from local
authorities. On this basis, a forecast should then be made to cover the traffic development during
the anticipated life-time of the bridge.

Vessel traffic is conveniently defined in the following terms as described in the subsections
below:
- Traffic Routes
- Passage Statistics
- Vessel Characteristics
- Traffic Forecast.

3.1 Traffic Routes

Vessel traffic routes in the vicinity of the bridge should be identified. All routes for which in-
teraction between navigation and the bridge is possible should be considered. Ordinary routes as
well as unofficial routes (short cuts etc.) should be identified. An example is shown in Figure 3.1.

Figure 3.1 Vessel trafic routes in the vicinity of a bridge crossing (based on Olsen et al. [3-4]).
20 Vessel Traffic

In general, the ordinary sailing pattern on the route is a Gaussian distribution across the route.
Mean and standard deviations typically depend on width of waterway, size distribution of vessels
and traffic density. This is further described in Section 5.3.

3.2 Passage Statistics

Sources for obtaining data on vessel traffic in a particular waterway include the following:
- The authority of the waterway in question often maintains logs of vessel trip data.
- Local port authorities and operators normally maintain logs of port call data.
- Local pilot stations often have fairly detailed records and knowledge of the navigation in the
area.
- The Coast Guard authorities maintain records on navigational safety.
- Military or naval authorities may possess records on naval as well as commercial traffic.

If no data or insufficient data are available, then recording of the vessel passages should be
initiated. The following means are available:
- visual observation (uncertain during night-time and fog periods)
- surveillance by radar (uncertain with regard to vessel type and size)
- communication by VHF radio.
Assuming that some general knowledge on the navigation pattern is available, a relatively short
period of observationlsurveillancewill provide a fairly good insight into the number and types of
vessels using the waterway. A longer period of observationlsurveillancemay be necessary for
identifying unusual vessels or groups of vessels.
It may be convenient to separate vessel traffk into selected vessel types which can be separately
considered in the analysis. Vessel geometry, collision force, manoeuvering characteristics,
types of cargo, etc. depend on vessel types.
According to Couper [3-11, the world fleet of merchant ships (over 100 GRT) totals about
40,500 vessels and equals 400 million GRT. The distribution of the world fleet in main vessel ty-
pes is given in Table 3.1.

Qpical profiles of different types and sizes of vessels are shown in Figure 3.2.

If special-purposevessels, such as chemical tankers, liquid natural gas carriers, vehicle carriers,
offshore industry vessels, naval vessels, passenger vessels, leisure vessels, etc. form a significant
proportion of the traffk, additional research and judgment by the designer will be required.
 Vessel Characterlstics 21

Vessel type Percentage of world fleet tonnage

Tankers 44%
Bulk carriers 20%
General cargo vessels 19%
Combined bulk/ tankers 6%
Container vessels 3%
Other types of vessels 8%

Total 100%
I I I

Table 3.1 Worldfleet vessel type distribution (based on Couper [3-11).

For all passing vessels the following data is of particular interest:

- the loading condition as regards draught, air draught, collision force and manoeuvring capa-
bility
- the transit speed as regards collision energy and force and manoeuvring capabilities
- the type of cargo as regards potential risk for environmental pollution
- the use of pilot and/or other navigational aids as regards the level of safety during passage.

Such data is normally difficult to obtain and it may be necessary to make qualified estimates.
Vessel draught (loaded and ballasted) in relation to the available water depths determines the
upper limit sizes of vessels which are able to reach the different parts of the bridge structure. Con-
sideration should also be given to abnormally raised high water levels during storm conditions.
It should be investigated if large vessels utilize the water level variations in the waterway and/or
the possibility of transiting in partly loaded condition.
Depending on the character of the shipping, a differentiation of loading conditions of vessels
transiting in opposite directions may also be considered.

3.3 Vessel Characteristics

The characteristics of the particular types of vessels using the waterway should be established
on the basis of specific information. However, if such data are not available, typical characteristics
may be taken from the diagrams in Appendix A, which have been developed on the basis of [3-11,
[3-21, [3-31, [3-51, [3-61, and [3-81.
Appendix A provides geometrical data for the following three types of vessels: tankers, bulk
carriers and container vessels. Depending on the purpose of the analysis, other vessel types may
be allocated to one of the three mentioned types.
 22 Vessel Traffic

I I

Type: Crude Oil Carrier

DWT: 270,000 T

DWT: 113 OOO T

Type: Trawler
800 GRT
I

Type: Coaster
500 GRT

Type: Cutter
100 GRT
I I I
I I I I I
0 50 100 150 200 250 300 350 400
Ship Length ( LOA ) ( m )
Figure 3.2 Examples of vessel profiles (partly based on Couper [3-1]).
 Vessel Characteristics 23

Figure 3.3 Vessel geometry definitions.

Appendix A contains the following data for the three above-mentionedtypes of vessels:
- air draught in ballast condition
- air draught in loaded condition
- height of deck house in ballast condition
- height of deck house in loaded condition
- draught in ballast condition
- draught in loaded condition
- length
- breadth
- moulded depth.

The definitionsof geometry data are shown in Figure 3.3.

The data represents typical characteristicsof present-day steel-hulled merchant vessels and the
user should be aware that important variations exist. The vessel bow structure,partly determined
by local vessel building practice, may be essential as regards magnitude of collision force.
Different bow shapes are illustrated in Figure 3.4.
Varying measurement traditions for size of vessel are used in different countries and for differ-
ent types of vessels. In this publication, Dead Weight Tonnage (DWT) and DisplacementTonnage
(W) have been used to express vessel size and vessel weight respectively. To facilitate conversion
from other units of measurement, conversion factors have been included in Appendix A.
24 Vessel Traffic

T m h end Bulk carrim

Figure 3.4 Common vessel bow shapes. The bulbous bow shape is prevalent in the worldfleet.

3.4 Traffic Forecast

The development of the vessel traffic should be forecast throughout the anticipated lifetime
of the bridge. It has to be recognized, of course, that such a forecast can only be done with very
limited certainty, because one of the main influencing factors is the political and economical
climate, locally as well as internationally.
Alternatively, a certain year, say 10-20 years ahead, may be selected as a study year and the
traffic in that particular year would then be forecast and used as a design basis.

Among the subjects to consider for the preparation of a forecast of the traffk would be, ref.
Larsen et al. [3-31:
- Projection of international and national markets:
an internationally recognized rate of growth in volume of trade by sea may be considered.
- Local circumstances having influence on the traffic:
deepening and widening of the navigation channel would allow transit by larger vessels
upgrading or downgrading of alternative or adjoining waterways might influence the traffic
volume in the channel in question
port development and establishment or closing down of local industries or vessel facilities
might influence the volume of traffk considerably.
- Technological developments in transportation facilities:
fast growth of containerized traffk at the expense of other types of freight vessels is likely
new types of mixed purpose vessels might appear.
- Technological development in ship building:
the construction of 1,000,OOODWT tankers is technically feasible, ref. Spyrou [3-71 and
Figure 3.5
the construction of 100,OOO DWT container vessels is technically feasible.
 Traffic Forecas 25

...................................................
.......................

-FTMm $Q!@Jt!k.Tpkyc
.......................... ........................................................................
NhkJMenr
.................................................... ..........................................................................................................

...........................................................................................................
...................................................

100,ooO
3GExai
./ /-I
.-
..............................................
I .............................................................................................................
.............................................................................................................
................................................. ............................................................................................................
i
50,ooO
/
................................................. .............................................................................................................
i

7-
................................................

*..............................................

...................................................
I
..............................................................................................................

10,Ooo
19M) 1970

Figure 3.5 Development of the size of very large crude oil carriers (VLCCs).During the years
1950-1970, ref. Spyrou [3-71. Although it is acknowledged that larger VLCCs are
technicallyfeasible, no new VLCCshave been constructed since the 1970s.
26 Vessel Traffic

Recognizing that the development of navigation in reality might turn out to be very different
from the forecast, future reviews should be instituted. Ideally, a spread in the forecast should be
considered allowing for different possible trends in economic and technical development.
An example of vessel traffk forecast covering a 20 year time span is illustrated in Figure 3.6.

Figure 3.6 Example offorecast of the number of passages of vessels of 40,000 DWT and above
in a Danish Strait. Two development scenarios were considered.
 27

4. RISK ACCEPTANCE

Risk can be defined as the potential realization of unwanted consequences of an event (i.e. the
product of the probability of an event and the consequences of the event). Both the probability of
occurrence of an event and the magnitude of its consequences are thus involved. References are
made to Philipson [4-151, Planeix [4-161, and Rowe [4-171.

Acceptance Criteria can be established either in the form of a predefined set of Risk Accep-
tance Criteria or in the form of Optimum Cost Criteria:
- Risk Acceptance Criteria are defined as acceptable limits to probabilities of certain unwanted
consequences of collision and are expressed in terms of annual frequencies.
Risk Acceptance Criteria are normally imposed by the Authorities to reflect the willingness of
people and society to accept risk. In principle, such criteria do not consider the cost of observ-
ing the criteria.
- Optimum Cost Criteria are AcceptanceCriteria based on cost-effectivenessanalyses comparing
the costs of bridge strengthening and protection measures against the benefit of risk reduction.
Optimum Cost Criteria may be introduced in cases where it is not economical or technically
feasible to design the bridge structures to comply with official Risk Acceptance Criteria.
Furthermore, the Optimum Cost Criteria may be used for decisions on whether to reduce the
risk to a level below what is required by the Risk Acceptance Criteria, because this can be
justified economically.

In order to make the use of the Acceptance Criteria unambigious and efficient they should be
implemented together with guidelines on how complying with the criteria shall be documented.
These guidelines should cover among others:
- types of collision consequences to be considered
- principles for estimating frequencies of the collision consequences considered
- principles for addressing the uncertainties of the frequency estimates.

4.1 Consequences of Collision

There are a number of different categories of consequences that are associated with a ves-
sel/bridge collision event.

The main consequencesmay be classified into the following six categories (three categories of
direct consequences and three categories of indirect consequences respectively):
28 Risk Acceptance

- damage to the bridge

- damage to the users of the bridge
- damage to vessel and cargo
- inconvenience cost to society and business
- social losses
- damage to the environment.
For each category of consequences, a wide range of damage, from superficial damage to disas-
ter, is possible. It should be noted, however, that if all imaginable categories and magnitudes of
consequences are to be taken into account, then the number of cases to be considered would
become very large and unworkable from a practical point of view.
In most cases, the data and resources available will dictate that the formal risk acceptance
procedure is limited to cover only the risk of bridge disruption and possibly also the risk of loss
of lives.

4.2 Disruption Risk Acceptance Criteria

Bridge disruption Risk Acceptance Criteria to cover vesselbridge collision have been estab-
lished in a few cases in connection with national codes and major bridge projects worldwide.
In all cases a risk comparison approach has been used in which the actual type of project has
been compared with other engineering projects or natural events with comparable failure conse-
quences. qpical data for risks associated with engineering projects and natural events, available
in the literature, are shown in figures 4.1 and 4.2.

Comparison of Risk Acceptance Criteria with the general safety level in codes and standards
addressing normal loads is not meaningful because the failure criteria are for design of each
element of the structure and for each load combination. The probability of failure in codes and
standards is normally in the order of magnitude of 0.0001-0.001 in 100 years, ref. Nordic Com-
mittee for Structural Design [4-141.

The following examples of bridge disruption Risk Acceptance Criteria have been found in the
literature:
- For the Great Belt Bridge Project in Denmark (previous project discontinued in 1978) compris-
ing two approx. 8 km long road and rail bridges it was decided that the design of each of the
bridges should be based on a mean time between catastrophic vessel collision accidents of

-
.
10,000 years. Reference is made to Frandsen et al. [4-11. This corresponds to a probability of:
0.02 in 100 years for disruption of the total bridge link.
In 1985 the State of Louisiana, USA issued Criteria for the Design of Bridge Piers with re-
spect to Vessel Collision in Louisiana Waterways [4- 121. The document distinguishes between
 Disruption Risk Acceptance Criteria 29

deep and shallow draught waterways. For deep draught waterways servicing vessels with up
to 13 m draught the following acceptance criteria regarding vessel collision accidents was
recommended:
0.01 in 100 years for bridge disruption.
- A working draft for an IS0 Standard titled Accidental Actions due to Human Activities
issued in 1987 [4-31,contains the following regulation on vessel impact on bridges from river
and canal trafYk: The design impact should be chosen in such a way that there is an assessed
probability less than 0.01 that this or a higher impact energy will occur in a period of 50 years
for one structure. This corresponds to:
0.02 in 100 years for excessive impact energy,
- The Guide Specification and Commentary for Vessel Collision Design of Highway Bridges,
issued in 1990 by Federal Highway Administration in the USA [4-81 suggests the following
acceptance criteria with regard to vessel collision accidents depending on the importance of the
bridge. The acceptable frequency of collapse shall be equal to or less than:
0.01 in 100 years for collapse of critical bridges
0.1 in 100 years for collapse of regular bridges.
- For the revised Great Belt Project in Denmark, comprising three major structures, a combined
6.0 km long road and rail bridge, a 8.0 km long railway tunnel and a 6.7 km long road bridge,
different Risk Acceptance Criteria were established for accidents causing disruption of both rail
and road connection and accidents causing disruption of only road or only rail. Disruption was
defined as closing of the link for traffic for more than-one month. Reference is made to
Vincentsen et al. [4-201 and Kampmann et al. [4-41. The probabilities of accidents causing
disruption of the link are not allowed to exceed:
0.02 in 100 years for simultaneous disruption of rail and road links
0.1 in 100 years for disruption of the road link
0.1 in 100 years for disruption of the rail link.
It is to be noted that these Risk Acceptance Criteria are intended to cover the aggregate proba-
bility of accident due to several types of causes, including fire and explosion, ice impact, train
accidents, etc. in addition to vessel collision.
An order of magnitude of acceptable probability of bridge disruption of 0.1-0.01 in 100 years
may be derived from the above examples.
Due to the characteristics of individual bridge projects it is recommended that specific
evaluation should be carried out in each case. Such evaluation should address, among others, the
following aspects:
- importance of the bridge connection to the public to society, to business and industry, to the
national defence, etc. (ref. US Guide Specifications [4-81)
- fatality risk in case of high traffk volume (ref. Section 4.3), if specific Risk Acceptance Crite-
ria for user fatality risk are not implemented
- cost effectiveness of improving the safety of the bridge (ref. Section 4.4).
Weighing the different considerations will often lead to compromises involving lower proba-
bilities of disruption than reflected in the above examples.
30 Risk Acceptance

Figure 4 .I Probability offailure of selected engineering projects @om Whitman [4-2I]).

In the above examples from the literature, excepting the Great Belt Project, the Risk Accep-
tance Criteria cited are the only criteria to be met. Hence, these criteria should ensure also an
adequate safety for the users of the bridge.
For the Great Belt Project specific Risk Acceptance Criteria for the risk to users were estab-
lished. Here the Risk Acceptance Criteria for bridge disruption and the Risk Acceptance Criteria
for user fatality should all be met.
 Disruption Risk Acceptance Criteria 31

FATALITIES
I

Figure 4.2 Number offatalitiesfrom natural events versusfrequency of events (from Whitman
[4-211).
32 Risk Acceptance

4.3 Fatality Acceptance Criteria

The fatality risk of concern in most bridge projects is the risk to the users of the bridge, motor-
ists or train passengers. The risk to be considered is one that is due to accidents that are unique to
the bridge as opposed to normal road and railway risks on land.

Two risk categories are relevant:

- individual fatality risk
- societal risk.

The individual fatality risk is the risk to each individual user passing the bridge. A reasonable
Risk Acceptance Criterion for the individual fatality risk is that the additional risk due to acci-
dents unique to the bridge is small compared to normal road- and rail-user risk. Normally, the
added individual risk to users of the bridge due to accidents unique to the bridge, ship collisions,

FrequencyafAcddenb (EventsprYear)

+
Number of Fatalities per Accident

Figure 4.3 Principle diagramme illustrating the diference between Societal Risk Acceptance
and Individual Fatality Risk Acceptance.
 Fatality Acceptance Criteria 33

Category of Risk Risk Acceptance Criteria

Individual Fatality Risk -
8 l o 3fatalitieslyear
Societal Risk
1-19 fatalities 3 - l o 3accidentslyear
20-200 fatalities 5 . l o 6accidentdyear
more than 200 fatalities 1 . l o 6accidentslyear

Table 4.1 Risk Acceptance Criteriafor fatality risk in addition to the risk on land-based rail-
waysfor rail passengers using the Great Belt Railway Tunnel,ref. [4-41.
The criteria are based on a number of rail passengers per year of 14 millions.

will be very low compared to normal road- and rail-user risks. Reference is made to Kampmann
et al. [4-41, Larsen et al. [4-91, Sexsmith [4-181, and Vincentsen et al. [4-201.

The societal risk is expressed as a relationship between the number of fatalities in an accident
and the frequency (events per year) of accidents with more than a certain number of fatalities (see
Figure 4.2 an$ 4.3). This concept is introduced because it is observed experience that society
reacts more vehemently to a few accidents with a large number of fatalities than to a large
number of accidents with few fatalities. The principle is illustrated in Figure 4.3.
As an example, ref. Kampmann et al. [4-41,the Fatality Acceptance Criteria for rail passengers
using the Great Belt Railway Tunnel are shown in Table 4.1.
The users of the bridge benefit from using the bridge. A higher risk is acceptable in such cases
than in cases where the persons subject to the risk have no benefit from the activity exposing them
to risk.
The Risk Acceptance Criteria curve should therefore be established on the basis of experience
from traffk and transportation rather than other types of societal risks.

An acceptably low fatality risk may be obtained by:

- ensuring a low probability of bridge disruption (see Section 4.2), and
- introducing bridge user warning systems on the bridge (see Section 10).
34 Risk Acceptance

4.4 Optimum Cost Acceptance Criteria

In many cases it is not economically or technically feasible to design the bridge structure to
comply with strict pre-set Risk Acceptance Criteria.

Illustrative examples from the literature are:

- Tasman and Bowen Bridges, Australia. Ref. J. Leslie et al. [4-101
- Sunshine Skyway Bridge, USA. Ref. Knott et al. [4-21, [4-51, [4-61, and [4-71.

Situations in which optimum cost considerations may be relevant include:

- bridges crossing very wide waterways with many piers exposed to vessel collision
- bridges with piers located in very deep waters
- existing bridges constructed without consideration of vessel collision.

In these cases the objective is to choose the level of risk that provides the minimum aggregate
sum of protection costs and expected present value of future consequences of collision accidents.
Appropriate analyses methods have been reported by Knott et al. [4-81, Leslie et al. [4-lo], and
Sexsmith [4-181.
An important element in the analysis is the assessment of the costs involved in a vessel colli-
sion accident. With reference to the classification of consequences suggested in Section 4.1 the
following details, among others, apply:

Damage to the Bridge

The direct consequences to the bridge owner are:
- the costs of salvaging the wrecked parts of the bridge
- the costs of repairheplacement of the bridge
- loss of revenue in the repair/replacement period if the bridge is a toll bridge.
The owner may also consider:
- additional costs of repair/replacement due to more strict requirements imposed by the Author-
ities after the accident.

Damage to the Users of the Bridge

The direct consequences to the bridge users are:
- injuries and loss of lives
- damage to vehicles and goods.
 Optimum Cost Acceptance Criteria 35

If required, the loss of human lives may be converted into economic terms by assigning cost
values for fatalities. In many countries such fatality cost values have been established for use in
connection with assessment of the benefits of improving the safety of transportation facilities or
evaluation of alternative transportation schemes.

Damage to the Vessel and Cargo

The direct consequences to the vessel owner are:
- loss of lives
- costs of salvaging the vessel
- costs of repaidreplacement of the vessel
- lack of income during the repair period
- damage to goods carried on the vessel
- claims for compensation from bridge owner and users
- increased insurance costs.

Inconvenience Costs to Society and Business

The indirect consequences to industry, trade and society are:
- road and railway inconvenience costs, which depend on the strategic importance of the bridge,
i.e. the availability of alternative routes, the density and type of traffic on the bridge, etc.
- port interruption costs if the bridgehessel wreckage is blocking an important navigation chan-
nel
- business and social losses due to disruption of tr&c/transport over a substantialperiod of time.

Social Losses
The indirect societal consequences due to long term disruption of trait are very difficult to
quantify in monetary terms.

Damage to the Environment

The indirect consequences to the environment due to spills and releases from vessel or from
cargo carried on the bridge include:
- cost of cleaning up
- cost of physical restoration
- long term ecological damage.
The loss of environmental values in case of long term ecological damage are very difficult to
quantify in monetary terms.
 37

5. COLLISION RISK

In general, it is not feasible to design all parts of a bridge structure to withstand worst case
loads from ship impact. However, it is possible to estimate frequencies of overloading of bridge
structures due to ship impact. With knowledge of the frequencies of overloading, a design can be
selected which fulfils certain acceptance criteria as described in Section 4.
Vessel collision accidents to bridge structures are relatively rare and conditions differ from
bridge to bridge. Therefore, the estimation of the risk of collision can not be based on vessel/
bridge collision statisticsalone. Collision risk models, simulating potential collision scenarios are
necessary.
In this section practical models for estimating frequencies of ship collisions with bridges are
addressed.

When establishing a vessel collision risk model the following approach is recommended:
- The navigation conditions in the vicinity of the planned bridge should be studied, and vessel
accident data, i.e. groundings,vesselhessel collisions, vesselllighthouse collisions,etc., should
be collected.
- Vessemridge collision scenarios should be defined and modelled so as to incorporate the par-
ticular conditions of the bridge site.
- Subsequently,data for estimation of the parameters to be used in the collisionrisk model should
be collected.

Vesselbridge collision risk models have been developed by a number of researchers, either for
specific projects or for the establishmentof general guidelines for bridge designs. State-of-the-art
reports have been published in 1983 and 1992, ref. Larsen [5-191 and Olsen et al. [5-321.

Some examples of bridge collision risk models are as follows:

- Establishment of vessel collision load specifications to comply with pre-set bridge disruption
acceptance criteria for the proposed Great Belt Bridge Project in Denmark, 1978. The model
was developed by CAP-Consult,ref. Frandsen et al. [5-51.
- Assessment of the probability of disruption of the Tasman Bridge in Australia, reconstructed
after a vessel collision accident, for the evaluation of benefits of constructing an alternative
crossing, the Bowen Bridge, 1978. The model was developed by Maunsell & Partners,
ref. [5-261 and Leslie [5-211.
- Assessment of the probability of disruption considering alternative protection measures, of the
new Sunshine Skyway Bridge to replace the existing bridge which collapsed due to a vessel
collision accident. Models were developed by COWIconsult, ref. Larsen [5-191 and Greiner,
ref. Knott et al. [5-161.
38 Collision Risk

- Assessment of the probability of vessel collision with the bridges in the Tokyo Bay Crossing
Scheme, ref. [5-371.
- Establishment of vessel collision design criteria for bridge piers in Louisiana Waterways in
USA,1985. The model was developed by Modjeski & Masters, ref. [5-301.
- Assessment of the probability of disruption and the probability of serious pollution for alterna-
tive bridge solutions and alternative protection arrangements for a bridge across the Straits of
Gibraltar, 1986. The model was developed by COWIconsult,ref. Larsen et al. [5-201.
- Establishment of Guide Specification for Vessel Collision Design of Highway Bridges in the
USA, including a model for the computation of annual frequency of collapse developed by
Greiner assisted by COWIconsult, 1990, ref. Knott et al. [5-171.
- Risk evaluation of crossing alternatives and establishment of vessel collision load specifica-
tions for the bridges in the Great Belt Link to ensure compliance with pre-defined bridge dis-
ruption and users' fatality risk acceptance criteria, 1991. The model was developed by COWI-
consult, ref. Olsen et al. [5-321, and Frandsen et al. [5-61.

5.1 Collision Risk Model

The general collision risk model is expressed in the following form, where the first summation
refers to all ship classes considered and the second summation refers to all bridge piers and su-
perstructure spans:

in which
F = Expected number of annual collisions to the bridge (bridge piers and/or super-
structure spans)

Ni = Annual number of vessels belonging to a certain class (i) of the vessels passing
the bridge. The distribution of transiting vessels into discrete groupings
(classes) shall be developed considering the required accuracy of the analysis
results and the categories of consequences requested. The distribution should
normally account for variations in the following vessel characteristics:

- Draught-in order to be able to sort out vessels that will ground prior to
collision
- Air draught-in order to be able to assess the probability of collision to the
bridge superstructure
- Collision load characteristics-in order to be able to assess the degree of
damage in case of collision
- Amount of dangerous goods carried on board the vessel-in order to be able
to assess the risk of environmental pollution in case of collision
 Causation Probability 39

- Local navigation practices-in order to be able to account for variations in

geometrical probability
- Vessel standard-in order to be able to account for variations in causation
probability.

Pcj = The causation probability related to the actual class of vessel (i), i.e. the proba-
bility of not avoiding a collision, in case the vessel is sailing on a course
that would lead to a collision and nothing is done to avoid collision, ref.
Section 5.2.

= The geometrical probability or rate of collision candidates related to the

actual class of vessel (i) and to the actual part (pier or span) of the bridge (k),
i.e. the probability that a vessel is sailing on a collision course, ref. Section 5.3.

PEi,k= The failure probability related to the actual class of vessels (i) and to the
actual part of the bridge (k), i.e. the probability of failure, given that a vessel
collision has taken place, ref. Section 5.4.

5.2 Causation Probability

The causation probability, P,, is the probability of failure to avoid an obstacle on the navigation
route. Local values of P,can be established through analysis of ship accident observations in the
bridge area. Estimates of P, on the basis of collisions with bridges are not available due to lack of
sufficient data. More data is available on groundings and ship-ship collisions and most estimates
of P, have therefore been based on these types of accidents.
Estimates of Pc have been made by many researchers since the early 70s. Two basic estimates
of the causation probability were published in 1974 by Fujii et al. [5-71 and by MacDuff [5-231.
Most estimations made by others are based on either Fujiis or MacDuffs approach. Some of the
estimations are based on grounding or collision statistics in a specific waterway. Other estima-
tions are based on P, values found for other waterways, adapted to the specific waterway using
correcting factors for differences in environmental and navigational aspects.

Selected estimations are mentioned in the following:

- Analyses of platform collisions and groundings in Japanese Straits by Fujii [5-71resulted in P,
values ranging from 1.0 to 6.3 *
- In Matsui et al. [5-251, groundings, buoy collisions and different types of vessel/vessel colli-
sions have been analyzed for 10 Japanese straits. For groundings the P, values range from 0.8
-
to4.3 10.~.
- MacDuff [5-231 analyzed groundings in the Dover Strait and found a P, value of 1.4
Vessel/vessel collisions were analyzed in two periods (before and after the introduction of
traffk separation) resulting in P, values of 5.2 -and 3.2 * respectively.
40 Collision Risk

- CAP-Consult estimated a P, value of 0.4 for bridge collisions in the Great Belt, ref. Lar-
sen [5-191, on the basis of Fujiis results for Japanese Straits, adjusted for traffic regulations,
marking of route, detectability of collision object, etc.
- Maunsell & Partners et al. [5-261 estimated the P, for collision with the Tasman Bridge to be
within an interval of 0.7 to 1.0 * lo4. This value was found on the basis of MacDuffs results
for the Dover Strait, adjusted for visibility and other environmental conditions, human error,
mechanical failure, and traffic density.
- COWIconsult estimated the P, for collision with the Sunshine Skyway Bridge to be 0.5 lo4, -
ref. Larsen [5-191. This value was found on the basis of Fujiis results for Japanese Straits,
adjusted for traffic density, use of pilots and traffic restrictions.
- CBA-Buckland & Taylor calculated a P, value for the Fraser River, British Columbia, at the
Annacis Island Bridge to be 3.6 ref. Sexsmith [5-351.
- Greiner has carried out estimations of P, for 5 US bridges, ref. Knott et al. [5-171:
Sunshine Skyway Bridge, Florida: 1.3 + -
for ships and 2.0 lo4for barges
Francis Scott Key Bridge: 1.0 * lo4for ships and 2.0 - for barges
Wm Preston Lane, Jr. Mem. bridge, Maryland: 1.0 * l o 4for ships and 2.0 * lo4for barges
Chesapeake Bay Bridges and Tunnels, Virginia: 0.7 * l o 4
Dames Point Bridge, Florida: 1.3 -
for ships and 4.1 l o 4for barges.
The estimations are based on statistics of groundings and vessel/vessel collisions in the
waterway near the bridge locations. When possible bridge collisions have also been included.
- Modjeski & Masters [5-301 have calculated P, for bridge collisions in a number of waterways
in Louisiana. For the Mississippi River bridges the results were:
Vicksburg Bridge: 5.4 * lo4
Huey P. Long Bridge: 2.5 * l o 4
Greater New Orleans Bridge: 1.3 *
- For the planned fixed link across the Strait of Gibraltar, COWIconsult estimated P, on the basis
of records on vessel traffic accidents in the Strait, ref. Larsen et al. [5-201. Present values of
-
2.2 l o 4for strandings and 1.2 * for collision were found and for vesselbridge collisions
a future value of 0.6 - was assumed based on improved traffic safety.
- The US Guide Specification [5-171 indicates that P, can be estimated using the basic values
0.6 * lo4for ships and 1.2 * lo4for barges and adjusted with correction factors for channel turn
or bend, cross currents, longitudinal currents and vessel traffk density.
- For the Great Belt Crossing East Bridge and West Bridge, COWIconsult estimated Pc values of
-
1.1 l o 4 for vessels with pilots on board and 3.2 * for vessels without pilots on board
based on Fujiis results for the Straits of Japan and adjusted for visibility, ref. Lund et al. [5-221.
The sizes of P, were verified by analysis of local accident data, ref. [5-281.

When assessing the P, value for a specific waterway, or when comparing the P, values for
different waterways, a series of influential factors should be considered.

Some of the most important factors influencing the causation probability or other parameters
of the collision risk model are reviewed in the following:
 Causation Probability 41

Visibility
Visibility is one of the most important factors affecting causation probability, especially when
no radar facilities are available.
The effect of visibility on the probability of vessel/vessel collision in the Dover Strait has
been analyzed by Wheatley [5-381 and the probability of vesselhessel collisions and groundings
in Japanese Straits have been analyzed by Fujii et al. [5-81, [5-91 and [5-111. These analyses show
that the probabilities of collision and stranding increase drastically for reduced visibility condi-
tions. The probability of collision is approximately inversely proportional to the visual range. As
an example, the probability of collision in fog when the visual range is less than 200 m has been
found to be about 100 times the risk in clear visibility conditions.

Darkness
The effect of darkness is closely related to that of the visibility. Based on analysis of collision
and grounding accidents, Fujii et al. [5-71 proposed that the probability at night time should be
taken to be 4 times the probability at day time.

Wind

Analyses of wind effects on a specific bridge passage by use of a manoeuvring simulator car-
ried out by Meurs et al. [5-271 indicate that the probability of collision to the bridge piers was
3 times greater at wind speeds of 11-14 m/sec compared to 6-8 m/sec.

Current
The US Guide Specification [5-171proposes the following empirical factors, R, and R,, for the
increase of probability of collision due to current:
- R, = 1 + VJlO for parallel current
- &,= 1 + V,, for cross current,
in which V, is the current component parallel to' the vessel path in knots and V,, is the current
component perpendicular to the vessel path in knots.
An investigation made by Fujii et al. [5-81 did not show any significant influence of currents.

Sea State
The manoeuvrability of a ship and the visibility for the navigator is reduced in heavy sea
conditions. In addition, water levels may be abnormally raised by storm surge tides, thus
increasing the vulnerability of the bridge.
However, during heavy sea conditions, the number of passing vessels will be reduced and one-
way traffic may normally be assumed.
No quantification of the influence on the collision probability has been found in the literature.
42 Collision Risk

Ice
Ice conditions can cause severe difficulties for navigation as the manoeuvrability may be
significantly reduced. However, no quantification of the influence on risk has been found in the
literature.

Vessel Type
Fujii et al. [5-81 conclude for Japanese waters that passenger vessels are 6 times safer than cargo
ships and tankers, and further state that cargo ships or tankers are 3 times safer than fishing boats.
Part of this may be due to the better manoeuvring characteristics of ferries and part may be due
to better knowledge of local conditions.

Vessel Size and Speed

According to Fujii et al. [5-71 the danger of vessel/vessel collision is almost proportional to the
vessel length.
According to Mizuki et al. [5-291 this tendency has, however, not been found in analyses of
groundings and collisions to fixed objects.

Vessel Loading Condition

The manoeuvring capabilities of vessels in ballast are inferior to those of loaded vessels, in
particular in the case of strong winds. However, no quantification of the effect on the collision
probability has been found in the literature.

Standard of Vessel, Crew and Equipment

Cockcroft [5-21 and Fujii et al. [5-101 state that there exists a significant difference among
vessel groups of different nationalities.
Cockcroft [5-21 has found that the accident probability is remarkable higher for vessels more
than 20 years old compared to vessels less than 10 years old.
Det Norske Veritas has reported that 6.8% of all casualties are due to faults in the technical
systems of the vessels, ref. [5-31.

Pilot on Board
The presence of a pilot on board reduces the probability of accident considerably. As an indica-
tion, Fujii et al. [5-121 have found that the accident rate per trip is decreased by a factor of up to 6.
 Causation Probability 43

Traffic Density
The US Guide Specification [5- 171 covers the increase of collision risk by introducing a factor
Rddefined as 1.0 when vessels rarely meet (low density), R,, = 1.3 when vessels occasionally meet
(average density) and R,, = 1.6 when vessels routinely meet (high density).
Based on analytical modelling of vessel/vessel collisions, Fujii et al. [5-81 and [5-101 suggest
that the number of vesselhessel collisions is proportional to the square of the traffk density. This
proportionality should, however, not be applied directly in the case of vesselbridge collision.

Detectability of Collision Objects

It seems that it should be easier for navigators to detect a bridge than a low island or an under-
water shoal. However, no quantification of the effect of the detectability of collision objects has
been found in the literature.

Presence of Navigational Aids

Matsui et al. [5-251 found the probability of collision to buoys in Japanese waters to be approx-
imately the same probability as for groundings. It indicates that almost all navigators avoid hitting
buoys. Therefore a suitable buoy arrangement may decrease the number of wandering vessels in
a waterway.

lbms or Bends in a Route

The US Guide Specification [5-171proposes the following empirical factor R, for the increase
of the probability of collision due to turns or bends in the navigation route:
- R, = 1+ @/9O0 for a bridge located in a transition zone, i.e. between 3000 and 6000 feet from
the turn point or between 0 and 3000 feet from the point where the bend starts.
- R, = 1+ @ / 4 5 O for a bridge located in a turn or bend zone, i.e. within a distance of 3000 feet
from the turn point or within the curve region of the bend.
Modjeski & Masters [5-301 suggest a factor between 2.5 and 4 when a curvature in the water-
way exists within 1 mile from the bridge.

Width of Navigation Channel

The importance of the width of the navigation channel is obvious if the clearance of the bridge is
significantly smaller than the width of the channel. In that case the bridge will constitute a bottle-
neck and hence increase the number of close encounters, ref. Ostenfeld-Rosenthalet al. [5-331.

Presence of Vessel Traffic Management System

Studies of the effectiveness of Vessel Traffic Management Systems, often referred to as VTS
systems, indicate risk reducing factors of approximately 2-3 for VTS systems comprising VHF
communication, radar surveillance and automatic target tracking.
Results from different studies of the effectivenessof introducing a VTS system are summarized
in Table 5.1.
44 Collision Risk

Study Achievable risk

reduction factor

Great Belt VTS system, ref. Olsen et al. [5-311 2-3

U.S.VTS systems, ref. US Coast Guard [5-361 2-3
Canadian VTS systems, ref. Canadian Coast Guard [5-11 2.0 - 3.3
European VTS systems, ref. Fabre et al. [5-31 2.2 - 2.5
Japanese traffic systems (maritime traffic, aviation and
road traffic), ref. [5-291 3-9

Table 5.1 Results from different VTS effect studies. The studies may not be directly compara-
ble due to differencesin systems, estimation methods and data basis.

5.3 Geometrical Probability

The probability that a vessel is sailing on a collision course, the geometrical probability Po,is
a function of a series of circumstances and strongly dependent on the local conditions.

The rate of collision candidates is calculated on basis of the following information:

- vessel traffic routes, number of passages and characteristics of the vessels as described in
Section 3
- distributions of the ship traffic across the routes
- geometrical definition of bridge
- natural protection by the seabed
- man-made protection works as described in Section 9.

Based on experience from accident records, ref. Appendix B and Olsen et al. [5-321 and Lund
et al. [5-221, the following accident scenarios are normally of interest:
a) Vessels that follow normal navigation procedures during transit but are put at risk in the vicin-
ity of the bridge due to human error or technical failure.
b) Vessels making evasive manoeuvres because of multiple encounter situations in the vicinity of
the bridge.
c) Vessels that fail to change course at a bend in the navigation channel close to the bridge.
d) Vessels not following the ordinary navigation route (Wandering Vessels) and vessels drifting
towards the bridge due to anchor or mooring failure, loss of propulsion, trouble with naviga-
tion equipment, caught in an ice drift, or the like (Drifting Vessels).
 Geometrical Probability 45

The geometrical probability for scenario a) is defined as shown in Figure 5.1, as a function of
the vessel track distribution before the bridge passage and the width of the collision zone.
For a bridge pier, the width of the collision zone is equal to the breadth of the vessel plus the
breadth of the pier.
For a bridge superstructure, the width of the collision zone is the whole length of the bridge
where the air draught to the top of the vessel deck house exceeds the high water clearance of the
bridge girder.
The relative distribution of the vessel passages in the 4 scenarios is specific to the bridge loca-
tion in question. The aggregate vessel track distribution is arrived at by adding the track distribu-
tion resulting from each scenario.

For many bridges it will be sufficient to consider scenario a) only. However, in the case of long
bridges with multiple spans outside the navigation route it will often be reasonable to add a cer-
tain percentage of uniform vessel track distribution from scenario d) to ensure the robustness of
the bridge structure at some distance from the navigation channel.

Figure 5.1 Definition of geometrical probability of collisionfor scenario a)

46 Collision Risk

Scenario a)
The lateral distribution of vessel tracks in a straight, well-defined route is often approximated
adequately with a nonnal distribution. The parameters in the normal distribution (mean value and
standard deviation) may be obtained from direct radar observations of the actual waterway, from
comparison with similar straits where radar observationshave been made, or from empirical rules.
Figure 5.2 shows an example of the results of the analyses of radar observations made for the
Great Belt Eastern Channel.

33VeeeelpaSS&p
channel, Great Belt

Normal Distribution

0 jr7-f
Figure 5.2 Lateral distribution of vessel tracks, calculatedfrom radar observations.

Based on a review of historical shipbridge collision data, the US Guide Specification [5-171
indicates that a normal distribution should be utilized to model the errant vessels sailing path
near the bridge. The standard deviation, 0 , is assumed to be equal to the length of the vessel. The
location of the mean, p, is assumed to be the centre line of the navigation channel, i.e. in case of
two-way traffic the centre line of the inbound and the outbound half of the channel respectively.
 Geometrical Probability 47

Observationsfrom the strait of Bisan Set0 in Japan and the Great Belt in Denmark, ref. Figure 5.2,
have shown that the standard deviation can be twice the length of a typical large ship. The typical
large ship was defined as the 95% percentile in the ship size distribution, ref. Olsen et al. [5-321.
Inoue [5- 151 and Kuroda et al. [5-181 have presented an empirical formula for the distribution
of two-way traffic in a navigational channel. They have concluded, divergent to the above indi-
cations, that the vessel track distribution can be approximated with a Gaussian distribution with
the following mean value p and standard deviation u:

p = a.W
u = 0.105 * W (approximation)
in which
W = width of navigation channel
a = 0.2 if the channel is equipped with centre line markings and 0.1 if the channel
is without centre line markings

Figure 5.3 illustrates the principle of lateral distribution of vessel tracks in connection with a
scenario a) bridge passage.

Scenario b)
If vessels often meet other vessels in the vicinity of the bridge line, the probability of collision
will increase. The frequency of critical meetings may be analyzed by the Close Encounter
analysis method, ref. section 2.5.

Scenario c)
Some vessels fail to make a proper change of course at a channel bend. Most will soon return
to the correct route, but a small fraction will constitute a group of potential collision candidates.
The geometrical probability may be calculated as for scenario a) using the vessel track distri-
bution before the bend and reduced by a factor reflecting the probability of the vessel failing to
correct the course before the bridge passage. With reference to Olsen et al. [5-321this probability
is a function of the distance from the bend to the bridge line and the probability of omission of
position checks by the navigator.
Some authors have included the effect of a channel bend in the causation probability, ref.
Section 5.4.
Figure 5.4 illustrates the principle of lateral distribution of vessel tracks in connection with a
scenario c) bridge passage.
 48 Collision Risk

r
,
,i vessel Tmck
Distrikrtion

Figure 5.3 Vessel track distribution for two-way bridge passage in a straight, well defined
waterway.

Scenario d)
The distribution of the tracks of Drifting Vessels may be established as a function of the
distance to locations where vessels might break away and the prevailing directions of wind,
currents and ice drift.
The distribution of the tracks of Wandering Vessels may be established on the basis of inter-
views with local authorities to disclose the extent of non-compliance with local rules, such as
short-cutting of the route, etc.
In the absence of specific knowledge, a uniform distribution across the entire bridge length may
be assumed. The intensity of the distribution may be assumed to be 1-5% of the general vessel
traffic density in the waterway, ref. Pyman et al. [5-341.
 Failure Probability 49

Vessel Track

I
Navigation Channel

Figure 5.4 Vessel track distribution for vessels at a critical bend in the navigation channel
close to the bridge passage.

5.4 Failure Probability

The probability of failure of the bridge structure, given that a vessel collision has taken place,
is complex and is a function of the vessel size, type, speed, direction, mass, and the nature of the
collision. It is also a function of the stiffnesshrength characteristics of the bridge pier and span
to resist the collision impact loads.

Two different approaches may be applied to assess the probability of failure:

- the Heinrich Ratio approach
- the probabilistic approach.
50 Collision Risk

The Heinrich Ratio Approach

The Heinrich Ratio is defined as the ratio of severe accidents to all accidents, ref. Heinrich
[5-141 and Olsen et al. [5-321. In the present context the Heinrich Ratio is defined from the
basis of vessel collision accidents. Collision statistics for vesselhessel collisions have to be used
because data on vesselbridge collisions is sparse, ref. Fujii [5-131.
By applying energy considerations regarding the differences between vessel/vessel and ves-
sel/pier collisions, a Heinrich ratio for vesseypier collision is obtained.
Vessel/vessel collision data has been analyzed by Fujii [5-131. Fujii proposed an empirical
relation between the cumulative relative frequency of damage, F(x,y), the damage rate, x, and the
vessel size ratio, y. The two parameters x and y are defined as follows:
- The damage rate, x, is the ratio of the cost of repairing the vessel hit over the total value of the
same vessel.
- The vessel size ratio, y, is the ratio of the gross tonnage of the colliding vessel over the gross
tonnage of the vessel being hit.

Figure 5.5 shows the cumulative relative frequency of damage, F(x,y), reproduced from Fujii
[5-131. F(x,y) is equal to the Heinrich ratio.

The damage rate x where the bridge element is damaged more than acceptably, i.e. involving
interruption of bridge traffic for an extended period of time, is not well defined.
In different vessel collision studies it has been assumed that a value for the damage rate of
about 0.10 results in serious damage to the bridge element.
On the basis of Figure 5.5, energy considerations and the damage rate assumed, it is now
possible to construct a Heinrich function giving the frequency of serious vesselbridge collisions
relative to all collisions.

The Probabilistic Approach

The probabilistic approach is based on a probabilistic model for the vessel impact force and a
spatial stochastic model of the resistance properties of the bridge elements, ref. Madsen et al. [5-241.
Different failure modes of the bridge elements (e.g. crushing, rotation, sliding, etc.) shall be
considered.

The probabilistic model for the vessel impact force is based on probability distributions for:
- the type, size, speed, and loading condition of the colliding vessel
- the collision force as a function of the vessel characteristics
- the collision energy as a function of the characteristics of the vessel and the water depth
- the angle of attack
- the position of the collision relative to the pier
- the friction coefficient between vessel and pier (in case of skew collisions).
 Failure Probability 51

t
n 0.
b
y<o.o32
I
0,l
-

Figure 5.5 Cumulative relativefrequency of damage rate Heinrich ratio as a function of the
damage rate (from Fujii [5-131).

For parameters where no probability distribution data are available, estimates may be used.
Sensitivity studies should, however, be carried out in such cases.
 53

6. VESSEL IMPACT FORCES

The determination of the impact load on a bridge structure during a vessel collision accident is
very complex as it depends on the vessel characteristics and the bridge structure, as well as the
circumstances of the collision accident.

Some important parameters are:

- For the colliding vessel: type, size, shape, speed, loading condition, and strength and stiffness
of bow, hull, and deckhouse.
- For the bridge elements in contact with the colliding vessel: size, shape, mass, and lateral
resistance characteristics.
- For the collision circumstances: eccentricity of impact and water depth.

Most research has dealt with the ideal case of a vessel colliding with an infinitely rigid and
immovable vertical plane wall structure. In this case the kinetic energy of the vessel together with
the water surrounding and moving with the vessel is consumed totally through the deformation
and crushing of the vessel.
In a less ideal collision event, parts of the energy will be consumed in deformation and crush-
ing of fenders, displacement of bridge pier and bridge superstructure and liberation of energy to
the surrounding water.
Realistic collision cases therefore involve lower impact forces, longer impact durations and
less damage to the vessels than an ideal collision case.

Collision impact load definitions are normally required in the following cases as described in
the subsections below:
- Head-on Bow Impact Forces
- Sideways Impact Forces
- Deckhouse Impact Forces
- Local Impact Loads.
An additional subsection:
- Barge Vessel Impact Forces

has been included to cover barge vessels which differ substantially from ship vessels and are of
particular importance to bridges crossing inland waterways.
54 Vessel Impact Forces

6.1 Head-on Bow Impact Forces

Most literature on vessel collision deals with the forces involved in a head-on bow collision
accident. An outline of the most important conclusions is given in the subparagraphs below:
- Physical and Mathematical Models
- Collision Force Definitions for Bridge Design
- Forces Experienced in Vessel Collision Accidents
- Collision Force Definitions for Dynamic Analysis
- Parameters Influencing the Impact Forces.

Physical and Mathematical Models

Various analysis models have been derived for estimation of the global loads involved in
head-on bow collision events. The models have been based on:
- investigation of ship/ship collision cases
- dynamic collision model tests
- quasi-static bow indentation model tests
- direct calculation of crushing resistance
- combinations of the above.

A detailed account of the development of the state-of-the-art has been given by Pedersen et al.
[6-281, the most important contributions being the following:
- A pioneer analysis model was published by V. U. Minorsky in 1959 [6-201. Based on an inves-
tigation of 26 ship/ship collision cases, a linear relationship was found between the deformed
steel volume and the absorbed impact energy. Minorskys well known diagram is reproduced
in Figure 6.1.
- In Japan, model tests have been carried out since 1958. The first of these were dynamic low
energy tests in which the bow acted as a pendulum ramming into the ship side, ref. Kagami et
al. [6-131.
- Dynamic model tests were carried out in Italy during the 1960s and the fist half of the 1970s
as part of the design of the nuclear powered ship Enrico Fermi, ref. Euratom [6-31. The tests
were carried out by letting models of vessel bows run down a roller conveyor and ram into a
model of the protective structure.
- During the period 1967-1976,a number of similar high energy dynamic ship collision model
tests were carried out in Germany by Woisin, ref. [6-381 and [6-391. The purpose of these tests
was also to give input to the design of structures to protect nuclear powered ships. The set-up
of the tests is illustrated in Figure 6.2.
- Quasi-static indentation tests, where vessel models were deformed by means of hydraulicjacks,
were carried out in connection with the design of the Bisan-Set0 Bridges in Japan. Akita et al.
[6-11 describe six collision tests and Hagiwara et al. [6-91 propose a method for prediction of
 Head-on Bow Impact Forces 55

collision damage based on model tests. Nagasawa et al. [6-221 describe model tests simulating
impact between small vessels and a rigid bridge pier.
- Ohnishi et al. [6-251 describe similar model tests with two different bow structures, i.e. a
35,000 DWT container vessel and a 409,000 DWT tanker with bulbous bow against a rigid pier.
- Amdahl [6-21 has carried out a number of quasi-static load deformation tests on stiffened
cylinders and models of ship bows.
- Numerical procedures for the prediction of crushing loads of structures have been proposed by
Reckling [6-301, Gerard [6-71, and Wierbicki [6-361. On this basis more elaborate models with
the specific purpose of evaluating vessel collision loads have been developed by Amdahl[6-21
and Yang and Caldwell[6-401. The merits of the various methods have been summarized by
Pedersen et al. [6-281.

oc
-
L.
c
400

I- /

ABSORBED ENEROY (IOOOTONS -KNOTS)*

Figure 6.1 The absorbed impact energy in shiptship collision accidents as a function of the
deformed steel volume (reproduced from Minorsky [6-201).
56 Vessel Impact Forces

-
0

Figure 6.2 Elevation view of the set-up of Woisins collision tests (om Frandsen et al. [6-41).

Most of the above research and testing has aimed at establishing the relationship between the
impact force and the indentation (damage depth) when crushing of the vessel takes place during
a collision accident.
Two representative force-indentationrelationship curves are shown in Figure 6.3 together with
the corresponding force history curves. The curves A indicate dramatic fluctuations of the force
during a very short (0.1-0.2 sec) initial phase of the collision event followed by a more or less con-
stant force during the remaining time, ref. Woisin [6-381 and [6-391.

The time duration of the maximum force, which has been estimated at twice the constant aver-
age force during the remaining collision time, is normally considered to be too brief to leave any
influence on a relatively robust bridge structure. Therefore, bridge designers have as a rule uti-
lized the average impact force as an equivalent static load for design against vessel collision.
Reference is made to the examples on design force definition in the following.
The curves B indicate a gradually increasing impact force during the whole course of the
collision accident, ref. Pedersen et al. [6-281. This may not involve higher average force but the
 Head-on Bow Impact Forces 57

t- t

Figure 6.3 Schematic representation of impact force dynamics according to the traditional
understanding,A, and according to recent research,B .

longer duration of the relatively high force level during the last phase of the collision event imply
that the maximum force rather than the average force should be used in design when based on
static force analysis.

Collision Force Definitions for Bridge Design

A series of examples of head-on impact force estimatesdeveloped on the above basis for design
purposes can be found in the literature (DWT indicates dead weight tonnage in metric
tonnes):
- In a study, ref. [6-261, Von Olnhausen computed the impact load for a 40,000 DWT tanker and
a 5900 DWT dry cargo vessel using a collision energy formula developed by Minorsky [6-201.
The computed load was 145 MN for the 40,000 DWT vessel.
- The impact loads recommended by the Nordic Road Engineering Federation [6-231 were
developed from the study by Olnhausen [6-261. The recommended design loads as a function
of vessel size and speed are reproduced in Figure 6.4.
- Also based on the Minorsky relationship,Woisin et al. [6-371 developed an equation giving the
average collision force directly:
Po = V~-L2/110O
where:
58 Vessel Impact Forces

Po = the average collision impact load (MN)

V = the vessel velocity (dsec)
L = the vessels length (m).

This equation has been used in the design of the Bahrain Causeway and the Faroe Bridges,
ref. Jensen et al. 16-12].

Figure 6.4 Diagram showing vessel collision design loads as published by the Nordic Road
Engineering Federation in 1975 [6-231.1 knot = 0.514 mlsec.
 Head-on Bow Impact Forces 59

- Very detailed and thorough investigations on the subject of vessel collision with bridges were
performed in connection with the proposed Great Belt Bridge project in Denmark in 1977-78,
ref. Frandsen et al. [6-41.
For large vessels (DWT > 40,000 tons) the collision forces were developed by Woisin
[6-381and [6-391, mainly based upon physical model tests. The tests indicated that the average
impact load from a 250,000DWT fully loaded tanker sailing at 16 knots was between 250 MN
and 700 MN. The scatter was due to differences in bow construction, fore peak tank ballast, etc.
Woisin suggested a triangular probability distribution of the average impact loads within the
above-mentioned limits. It was chosen to use collision forces which were equal to the 70%-
fractile of this distribution for the following reason: for a given design vessel the number of
smaller vessels with a crushing strength greater than the 70%-fractile force of the design
vessel would be approximately equal to the number of larger vessels with a crushing strength
less than the 70%-fractile force of the design vessel.
The collision forces for vessel sizes down to 40,000 DWT were established by assuming that
the crushing forces for vessel sizes between 250,000 DWT and 160,000 DWT were
proportional to (DWT)*, and for vessel sizes between 160,000 and 40,000 DWT the crushing
forces were proportional to (DWT)ID.
The collision force of a specific vessel was additionally assumed to be proportional to the

Figure 6.5 Diagram showing collision design loads assumed for the Great Belt bridge project
in 1978 (reproduced from Frandsen et al. [6-41). The impact speed is 8mlsec.
103 tons = 9.8 MN.
60 Vessel Impact Forces

The 70% fractile of average collision forces assumed in the design as a function of vessel
size is reproduced in Figure 6.5.

- Saul et al. [6-321, [6-331, and [6-341 have presented a modification of the Woisin approach.

They have proposed the following relationship for bulk carriers:

P, = 0.88 (DWT) f 50%
where:
P, = the effective maximum impact force of the vessel (MN)

- Based on studies performed for the Sunshine Skyway replacement bridge in Florida, Knott
[6-141 proposed that the SauVSvensson equation be modified to reflect reductions in the maxi-
mum impact (P,,) when the vessel is travelling at speeds lower than the maximum (estimated
at 16 knots) and when the vessel is travelling in a partially loaded, ballasted, or a light (empty)
condition. The equation utilized for design collision loads was as follows:
P, = 0.88 (DWT). (V/8)23-(DxtDm,)
where:
V = the vessel velocity at the time of impact (dsec)
D,, = the vessel displacement at the time of impact (tonnes)
D, = the maximum (fully loaded) displacement of the vessel (tonnes).
- In 1985, the State of Louisiana, USA, issued Criteria for the Design of Bridge Piers with
Respect to Vessel Collision in Louisiana Waterways, ref. Modjeski & Masters [6-211.
The report contains design charts for deep draught waterways indicating collision forces for
14 knots and 7 knots impact speeds as a function of water depth and vessel size. The design
chart for 14 knots impact speed is reproduced in Figure 6.6.

- The Norwegian Public Roads Administrationprescribes the following vessel collision loads for
bridges and ferry ramps in the public road system, ref. [6-241:
P = 0.5 (DWT)
where:
P = static equivalent collision force (MN)
- For the Euroroute bridge/tunnel crossing proposal for the English Channel, vessel collision
design basis was derived for a 270,000 DWT tanker, ref. Teyssandier [6-351. A static force of
500 MN was assumed and 30%was added for dynamic effects.
 Head-on Bow Impact Forces 61

Cdadon Fom (loo0 kip)

L
60

60

40

30

20

10

-
Figure 6.6 Collision load design charts for 14 knots impact speed as recommended by the

1 knot = 0.514 mlsec.

1000 kips = 4.45 MN.
62 Vessel Impact Forces

Average Collision Force ( MN ) / I

I

1.OOO 1o.OOO 1oO.OOO 1.OOO.OOO

Tonnage ( DWT )

Figure 6.7 Average collision forces developed for the Gibraltar Strait crossing. Two types of
vessels were considered, both types in ballast as well as loaded condition (repro-
ducedfrom Larsen et al. [6-171).
 Head-on Bow Impact Forces 63

Ship impact Force ( too0 idpg )

I 80

70

60

50

40

30

20

10

0
4-
+
0 2 4 6 8 10 12 14 16
ImPEtct speed (- 1

Figure 6.8 Typical impact forces according to the US Guide Specification [6-151
1 knot = 0.514 mlsec.
1000 kips = 4.45 MN.
64 Vessel Impact Forces

- For the Gibraltar Strait Crossing a collision load design basis was developed in 1987-88, ref.
Larsen et al. [6-171. Based on the 1978-79Great Belt studies, collision forces were extrapolated
to cover possible future 1,000,000DWT tankers. For container vessels and other rigid types of
vessels an increase of the collision force compared to tanker and similar vessels was proposed.
The collision force diagram developed is reproduced in Figure 6.7.
- For the Dartford bridge across the River Thames in England, an equivalent static collision load
of 350 MN has been assumed, corresponding to a 65,000 DWT vessel sailing at 5 d s e c .
- For the Luling Bridge across the Mississippi River near New Orleans, an equivalent static
collision load of 270 MN was assumed, corresponding to a 40,000 DWT vessel.
- The US Guide Specification [6-151 prescribes the following formula (translated into SI-units)
for the computation of an equivalent static impact force for vessel collision:

Bow crwhlng Fom, PW( MN )

A
1o.Ooo

1.Ooo

100

10
500 l.m 6.Ooo lo.m 50.m 1oo.ooo Mx).Ooo
Shlp Site ( DWT)

Figure 6.9 Diagram showing maximum vessel impact forces resulting from the investigations
carried out in connection with the design of the Great Belt Bridge project
(reproducedfrom Pedersen et al. [6-28]).
 Head-on Bow Impact Forces 65

P, = 0.98 (DWT) [V/8]

where
P, = impact force (MN)
V = the vessel velocity (rdsec).

A diagram from [6-161 showing typical ship impact forces is reproduced in Figure 6.8.

- Detailed and comprehensive vessel collision load regulations have been established for the
Great Belt Bridge project in Denmark in 1991 based on existing knowledge as supplemented
by extensive numerical calculations by Det Norske Veritas, ref. Frandsen et al. [6-51 and
Pedersen et al. [6-281.
The load-indentation relationships have been calculated, using the methods developed by
Amdahl [6-21 and Yang & Caldwell [6-401.
Based on the dynamic characteristics found, it has been assessed that the maximum impact
forces rather than the average forces should be used as the design force for design on the basis
of equivalent static analysis. Maximum impact forces have been established for vessels between
500 DWT and 300,000 DWT. Different impact speeds and loading conditions were considered.

Based on the results of the analysed collision cases, an empirical expression has been derived
for the easy estimation of the maximum bow collision loads of ordinary merchant vessels:
Pbow= Po- [E - Lz + (5.0 -L) * L2.6]I/ for E 2 Lz.6
Pbow= Po* [5.0 - E * L]/*for E < L2.6
where
PbW= maximum bow collision load (MN)
Po = reference collision load equal to 210 MN
L = L,/275m
E = E,,/1425 MNm
L, = length of vessel (m)
Eimp= kinetic energy of vessel (MNm).

These formulas account for the effect of strain rate, impact speed, vessel loading conditions,
and vessel size. For large vessels the formulas produce higher impact forces for static design
than have been assumed previously.
A diagram from [6-281 showing resulting maximum bow crushing forces as a function of
vessel size and impact velocity is reproduced in Figure 6.9.
The results of the investigations carried out for the Great Belt Bridge project as summarized
66 Vessel Impact Forces

above are considered to represent the state-of-the-art with regard to knowledge about global
vessel bow impact forces.
The collision impact force formulas recommended produce somewhat higher forces than
have been assumed previously in most cases.
The diagram in Figure 6.10 has been developed to compare important previous impact force
estimates with the vessel bow impact forces calculated with the formulas established for the Great
Belt project.

Forces Experienced in Vessel Collision Accidents

Although full-scale collision incidents occur relatively often, very little is known about the
impact forces involved. This is because the collision circumstances, i.e. vessel speed, eccentricity
of impact, etc., are usually not known sufficiently well.
Plans are known to exist for the conduct of full scale instrumented collision tests, but so far the
costs involved have prevented the realization of these plans.

In two cases, reported in literature, vessel collision accidents have been analyzed and estimates
of the collision forces involved have been made:
- The forces involved in the collision of a 10,600 DWT vessel sailing 8 d s e c against the Drog-
den lighthouse in Copenhagen have been estimated by Ostenfeld [6-271 and Olnhausen [6-261
at 70 MN and 35 MN respectively.
- Kuesel [6-161 has analyzed the collision between a 30,000 DWT loaded tanker and the New-
port Bridge. The vessel collided head-on with the main tower pier of the suspension bridge at
a speed of 3 m/sec and the average collision force was estimated to be 60 MN.

The photos in Figure 6.11 from actual collision accidents illustrate damage suffered by the
colliding vessels.

Collision Forces for Dynamic Analysis

In order to perform a dynamic analysis, the loadhndentation relationship and/or the time history
for the design collision case should be established. Such relationships have been developed and
described in the literature, ref. in particular Woisin [6-381 and [6-391, Requena [6-311, Modjeski
& Masters [6-211, and Prucz et al. [6-291. Diverging opinions exist as discussed in the above,
ref. Figure 6.3.
The most comprehensive and most recent work has been carried out by Det Norske Veritas for
the Great Belt Bridge project in Denmark, ref. Pedersen et al. [6-281. Examples of calculation
results are shown in Figure 6.12.
 Head-on BOWImpact Forces 67

Vessel Baw Impact Forces (MN)

1,000 i i i
f
.......... .................. ........
............ ...
1
1..
t

i
so0 i t
. . . . . . . . . . . ............ ... ij.
........... j............j........i........
i i i
.I
!
i
i
.,..
I
I

1
j

100 .. .. .. c
4..

)..

4..

1-

. . .

10
1,Ooo 5,000 10,Ooo 50,OOo 100,Ooo 500,Ooo l,Ooo,OOo
Veseel DWT ( Tonne8 )
Figure 6.10 Diagram showing vessel collision design forces as calculatedfrom diflerent sources.
To facilitate comparison, all curves indicate design impact forces of fully loaded
ordinary merchant vessels impacting at 7-8mlsec collision speed.
68 Vessel Impact Forces

Figure 6.11 Vessel damage from vessellbridge collision accidents: At the top, the Newport
Bridge collision accident in the USA in 1975 (from Kuesel [6-16]) in which the
vessel was seriously damaged while the bridge suffered no permanent damage.
Below, the Tostero Bridge collision accident in Sweden in 1991, ref. Appendix B
(photo: Mats Fogeman) in which the vessel suffered only light damage while the
bridge was partly destroyed.
 Head-on Bow Impact Forces 69

0 5 10 15 20
IndentatlonofBow( m )
Crushing bad ( MN )

100
-
-
-
-
50
- 3,000DWT

2,000DWT

-b
0 1 2 3 4

Indentation of Bow ( m )
Figure 6.12 Crushing loadlindentation relationships (based on Pedersen et al. [6-28]).
At the top: - 150,000 DWT bulk carrier at an initial speed of 9.3 mlsec
- 40,000 DWT container ship at an initial speed of 12.9 mlsec

Below: - 3000 DWT general cargo carrier at an initial speed of 7.5 mlsec
- 2000 DWT tanker at an initial speed of 7.0 mlsec
70 Vessel Impact Forces

Parameters Influencing the Impact Forces

The above indications refer to normal vessels, colliding under ideal circumstances. Varia-
tions from this basis should be considered in specific cases and judgment must be exercised in
evaluating the influence on the collision load definition.

The following typical variations should be considered:

- The type of vessel: vessels constructed with particularly rigid hulls, i.e. container vessels, na-
val vessels including submarines, ice-strengthened vessels, etc., are likely to produce higher
global and local collision forces than conventional merchant vessels, ref. Larsen et al. [6-171.
- The bow shape of the vessel: the maximum impact loads for bulbous bows are somewhat
higher than the estimated values for conventional bows, ref. Pedersen et al. [6-281.
- The impact speed: due to the strain-hardening effect, the impact forces are dependent on the
impact speed, ref. test results published by March et al. [6-181. This is accounted for in some
empirical expressions for collision loads, ref. Knott et al. [6-151 and Pedersen et al. [6-281.
- The loading condition of the vessel: the maximum impact load for ballast condition is normal-
ly considered lower than the estimated impact load in the fully loaded condition. This is
accounted for in some empirical expressions for collision loads, ref. Knott [6-141, Larsen et al.
[6-171, and Pedersen et al. [6-281. In the US Guide Specification [6-151 it is agreed that the
reduction in the impact force due to reduced mass is offset by the increase in impact force
caused by a stiffer bow due to the presence of water ballast in the vessels forepeak tanks.
- The size of the bridge pier: in the case of collision against a pier with limited width or with a
step or recess, the collision load may be adjusted by multiplying by the ratio of the deformed
steel area in actual contact with the pier to the total steel area of the bow section, with the
addition of shear forces equal to half the shear buckling loads of the undeformed parts of the
bow outside the contact area, ref. Pedersen et al. [6-281.
- The shape of bridge pier: in cases where the pier is shaped so that vessel collision accidents are
likely to result in glancing blows rather than head-on impact, a large part of the collision ener-
gy will be consumed in friction and in the rotating of the vessel. The probability of develop-
ment of full head-on collision forces is thus reduced, ref. Pedersen et al. [6-281.
- The eccentricity of the impact: in cases of eccentric impact, a significant part of the collision
energy will be consumed in rotating the vessel.
- The water depth: in cases of low underkeel clearance, the hydrodynamic mass, i.e. the mass of
the water moving with the vessel, will increase and thereby increase the collision energy.

6.2 Sideways Impact Forces

For bridges crossing wide waterways, sideways collision to the bridge structure might be
experienced in the case of a drifting vessel or in the case of an evasive manoeuvre to avoid a
head-on collision.
 Deckhouse Impact Forces 71

For bridges crossing narrow channels, sideways collision might result from common navigation
practice. In case of drift to one side of the channel the rotating of the vessel by using the rudder
may swing the stem of the vessel sideways into piers or fenders lining the channel.

Studies carried out in connection with the Great Belt Crossing, ref. Pedersen et al. [6-281,
indicate that global sideways collision forces in the case of an immovable bridge structure are of
the same magnitude as bow collision forces. Although the hydrodynamic mass involved with
lateral vessel motion is increased by a factor of 5-10, compared to a head-on collision, the impact
energy from a sideways collision is generally smaller due to the low speed.

6.3 Deckhouse Impact Forces

For collision between the deckhouse of a vessel and a bridge superstructure,very little data and
research have been reported in the literature.

In the US Guide Specification [6-151, the following relationship is prescribed:

pDH = R D H pS

where:
PDH = deckhouse impact force
RDH = 0.1 for vessels greater than 100,000DWT,
= 0.2 - 0.1 * DWT/100,000 for vessels smaller than 100,000DWT
P, = vessel collision force on pier.
This empirical relationship was based on Danish investigation results from the 1970s.
The US Guide Specification [6-151 further indicates that the impact force between ship mast
and bridge superstructure may be figured at 10% of the deckhouse impact force. This is consid-
ered a conservative assumption in most cases.

Most recent investigations have been made by Det Norske Veritas for the Great Belt project,
ref. Pedersen et al. [6-281. Various collision cases for typical deckhouses of vessels up to 5000
tons displacement have been considered. The impact loads have been calculated from the loads
required to deform the structural elements of the deckhouse.
The calculations have resulted in recommended design loads depending on the size of the
contact area, i.e. the product of the contact height, which is the vertical distance between the top
of the ships deckhouse and the underside of the bridge girder, and the breadth of the deckhouse.
Some results are summarized in Figure 6.13.
72 Vessel Impact Forces

impact Force ( MN ) -
Contact height 5.0 m
(Top deck and two r m t decks)
1
30

-
Contact height 2.5 m
(lop deck and next deck)

20

10

0
0 10 20 30-
Breadth of Deck House ( m )

Figure 6.13 Impact force versus deckhouse breadth and contact height.

6.4 Local Collision Forces

In order to ensure that all bridge elements exposed to vessel collision are designed to a certain
strength and robustness, local point loads and line loadings which will occur during a collision
event should be considered.
A review of the local impact forces found in the literature is presented below. The values indi-
cated are difficult to compare since some represent peak forces and some are the average forces.
In addition different types of vessels (tanker, container, etc.) are involved.
 Barge Vessel Impact Forces 73

- For the Faroe Bridges, ref. Jensen et al. [6-121, the design vessel for impact was a 2,250 DWT
ship colliding with the bridge at 6.2 m/sec and the following design pressure applied to the piers:
From 5 m below to 5 m above the waterline: 1.0 MN/m2
From 5 to 10 m above the waterline: 0.5 MN/m2
From 10 to 15 m above the waterline: 0.05 MN/m2
- According to Olnhausen [6-261, pier surface loadings of 2.0 MN/m2 for the hull strength of
large ships, and 0.5 MN/m2for minor impacts are valid.
- Ohnishi et al. [6-251 report that for a 35,000 DWT container ship, an average impact force of
88 MN was estimated and an average impact pressure of 1.2 MN/m2;and for a 409,000 DWT
tanker, an average impact force of 245 MN was estimated and an average impact pressure of
0.4 MN/m2.
- Modjeski & Masters [6-211 recommend using a distributed surface load of 1.4 MN/m2 com-
bined with a line load of 2.9 MN/m for verifying the local strength of the pier and its structural
members. The loads are to be applied at a height which corresponds to the elevation of the deck
of the design vessel in ballast.
- Using data reported by Kuesel [6-161 for the 1981 collision of a 31,800 DWT tanker with a
main pier of the Newport Bridge, the average bow impact pressure was estimated to be approx-
imately 0.6 MN/m2.
- In connection with the Great Belt project, local design loadings were prescribed, together with
detailed rules for their application, for all bridge elements exposed to vessel collision.
For bridge elements in contact with the bow of the colliding vessel, two load cases are considered
A system of 0.1 m wide line loads varying from 4.3 MN/m to 2.9 MN/m together with a
uniform pressure level of 0.5 MN/m2 for exposed foundations.
Point loads varying from 17 MN to 21 MN acting over a 1.0 m square or 10 MN acting over
a 0.5 m square.
For bridge elements in contact with the superstructure of the colliding vessel three load
cases are considered:
A uniform pressure level of 0.5 MN/m2 acting over an area of 20 m2.
A system of 0.02 m wide line loads of 1.0 MN/m.
A point load of 2.5 MN acting over a 0.25 m square or 0.5 MN acting over 0.10 m square.

The results of the investigations suggested that the magnitude of local loads does not vary very
much with the vessel size. This allowed that the same loads could be specified for all bridge
elements irrespective of the size of the design vessel governing the global loading.
The results of the investigations summarized above for the Great Belt bridge project are consid-
ered to represent the state-of-the-art with respect to knowledge about local collision forces.

6.5 Barge Vessel Impact Forces

Significant differences exist between the shapes and structures of ocean-going ship vessels
and barge vessels using inland waterways. Regarding barges, very little research and data on
behaviour during collision can be found in the literature.
 I I
1.0 2.Q
Water Depth ( m )

Figure 6.14 Design chart developed by Modjeski & Masters [6-211 for different classes of
bargeslwaterways (transformed into SI units).
 Barge Vessel impact Forces 75

I Barge lmpect Force ( MN )

26

20

1s

10

Figure 6.15 Design chart for difSerent tow lengths of standard hopper barges (I900 tonnes dis-
placement) included in the US Guide Specification [6-1.51 (transformed into SI units).

The following indications have been noted:

- Modjeski & Masters [6-211 recommend the barge impact loadings shown in Figure 6.14 for
barges in shallow draught waterways in Louisiana.
- For bridges crossing the Rhine River, German Regulations [6-81 established design impact
loads of 30 MN for head-on collision with piers near the navigation channel, and 6 MN for
head-on collision with bridge piers located close to the river bank.
- French regulations [6-61 specify that barge impact against bridge piers in rivers or canals of
more than 12 m width shall be 8 MN in the direction of the current. For narrow waterways
(those less than 12 m in width), the barge impact forces shall be 1 MN.
- An I S 0 (International Organization for Standardization) Standard [6-101 on accident action is
under preparation. This document combines the above-mentioned French and German regula-
tions.
76 Vessel Impact Forces

- The most comprehensive studies reported was performed in 1983 by Meir-Dijrnberg [6-191.
The studies included dynamic loadings with a pendulum hammer on barge models as well as
numerical analysis.
- The US Guide Specification [6-151 has adopted and elaborated the results of the above-men-
tioned German research. The design chart shown in Figure 6.15 is taken from the US Guide
Specification.
 77

7. BRIDGE DESIGN

For bridge design purposes, design forces should be established for at least the following col-
lision load cases:
- collision with the bridge piers and pier shafts, head-on by the vessel bow or sideways by the
vessel hull
- collision with the bridge superstructure by the vessel bow, hull, or deck house.

Global forces for design for overall stability as well as local forces for design for local strength
of bridge elements are required.
Due to the low probability of collision,the vessel collision load case is traditionally considered
to be an accidental action. This means that a reduced level of safety against failure is accepted.

However, two or more levels of vessel collision load assumptions may be introduced in order
also to cover lower vessel collision loads which might be experienced more often. As an example,
the following three levels of safety, corresponding to collision loads with decreasing levels of
probability, might be specified:
- no damage accepted
- minor damage accepted, provided continuous use of the bridge is possible and provided repair
work can be effectuated without disturbing the traffic
- major damage or local collapse is accepted, provided the remaining structure has sufficient
redundancy to allow repair within a relatively short time (say, 3 months) to a degree that allows
re-opening of the bridge for traffic.

7.1 Energy Considerations

In order to determine the consequencesof a vessel collision accident, the traditional method is
to estimate the amount of kinetic energy available and thereafter determine how this energy is
dissipated by displacement, deformation or crushing of the vessel and/or the bridge structure,
including attached or free-standing fenders or protective works.

The total kinetic energy KE is:

KE = * (M,+M,) * v2 (7.1)
where
v = the velocity of the vessel
M, = is the mass of the vessel
M, = the hydrodynamic added mass of the water surrounding the vessel and
moving with it.
78 Bridge Design

M, is the displacement tonnage of the vessel divided by the acceleration of gravity. In the ab-
sence of better information the displacement can be established approximately on the basis of the
gross registered tonnage (GRT)or the dead weight tonnage (DWT), as indicated in Appendix A.
The hydrodynamic added mass, M,, which accounts for the hydrodynamic forces acting on the
vessel during the course of the collision accident, depends on various factors such as:
- the direction of the vessel travel
- the shape of the vessel hull
- the underkeel clearance (the distance between the bottom of the vessel and the waterway bed)
- the direction of the impact.

The dependance of the different factors has been studied by several authors, ref. Petersen et al.
[7-71 and Blok et al. [7-11.
In deep water, and for short impact durations, the following order of magnitude of the hydro-
dynamic added mass is normally recommended:
- M, = 0.05 M, - 0.10 M, for bow impact
- M, = 0.40M, - 0.50 M, for sideways impact.
However, if the draught of the vessel approaches the water depth and/or the duration of the im-
pact is longer than, say, 1 sec, the hydrodynamic resistance increases, particularly in the case of
sideways impact.
For the time being it is recommended to assume that the ratio MJM, depends on the
vessel draughvwater depth ratio as indicated in Figure 7.1.
The vessel travel direction may not correspond to the longitudinal axis of the vessel, owing to
wind and current pressures and steering manoeuvres. In narrow channels it is common practice to
keep the vessels bow headed along the channel, so the stem may oscillate laterally.
The vessel velocity may be resolved into components parallel and normal to its axis for calcu-
lating longitudinal and transverse kinetic energy, accounting for the appropriate hydrodynamic
mass for each component.
Owing to uncertainties in estimating lateral velocity components and lateral hydrodynamic
mass coefficients, including the effects of the yaw angle, it is recommended that the ratio of trans-
verse kinetic energy to longitudinal kinetic energy be taken as not less than 0.5. For intermediate
directions, it is recommended that an elliptical distribution be used, as shown in Figure 7.2.
If the vessel comes to a complete stop, its total kinetic energy is dissipated in the collision. If
the vessel is merely diverted, the collision energy may be determined from a vector analysis of
the vessels kinetic energy before and after the collision.
In an eccentric impact, part of the vessels kinetic energy will be dissipated through rotation and
friction. Saul et al. [7-91 indicate that the portion of the kinetic energy to be absorbed by the ves-
sel and the bridge structure is a function of the impact angle and the coefficient of friction between
the colliding vessel and the bridge structure.
However, the most common approach is to assume that the vessel is totally stopped by the
bridge structure and all its kinetic energy is released in the collision.
 Energy Considerations 79

5.0

4.0

3.0

2.0

1.0

Figure 7.1 Hydrodynamic mass factors.

The kinetic energy involved in the collision is dissipated by displacement, deformation and
crushing of:
- the fender or buffer structure
- the vessel
- the bridge structure.
The energy absorbed by elastic deformation is small and may be neglected.
80 Bridge Design

Of Ship

Axis
Axisofchannel
of channel -

Figure 7.2 Kinetic energy distribution Cforwell defined straight navigation channels, the angle a
between the axis of the channel and the axis of the vessel may normally be taken as 0).

It is recommended that all structural elements in the collision analysis be treated as rigid or
plastic bodies. Each element has a different deformation threshold or yield point and a different
postyield plastic deformation module. The general characteristics of the system are illustrated
schematically in Figure 7.3.
Figure 7.3 also illustrates an ideal design for a collision protection scheme. In this scheme, small
collisions are absorbed by deformation of the fender. Under larger collisions, the fender is
crushed, but before it collapses, crushing of the vessel structure takes over and absorbs the
remainder of the collision energy.

7.2 Structural Analysis

The response of the bridge structure (pier or superstructure together with fender or buffer
systems) should be estimated using either static or dynamic calculation models depending on the
circumstances, ref. Prucz [7-81 and Modjeski & Masters [7-51. The calculation model should be
based on a system of 3 elements: vessel, fender, and pier, as indicated in Figure 7.4.
 Structural Analysis 81

Figure 7.3 Forceldeformation characteristics.

In cases where the energy absorbing capacity of the pier is less than required, the real problem
is to design a protective structure that can make up for the energy absorbing deficiency of the
vessel-pier system. Reference is made to Section 9.

The following three approaches are available:

- static force analysis
- equivalent static force analysis with dynamic amplification
- full dynamic analysis.

Static Force Analysis

For simplicity and due to lack of data on impact load histories, this approach has been applied
in most cases. Reference is made to Section 6.1. This approach assumes that the collision loading
does not involve important dynamic amplification and therefore, caution should be exercised when
selecting the level of the static design force.
82 Bridge Design

Equivalent Static Force with Dynamic Amplification

This approach should be applied for slender structures with linear elastic response to loading.
The structure is modelled by a simple mass-spring system which is exposed to a force-time
relation, ref. Section 6.1. The maximum dynamic and static deflections are calculated and a
dynamic amplification factor is estimated as the ratio between the dynamic and the static deflec-
tions. An equivalent static force is then obtained by multiplying the assumed maximum force by
the dynamic amplification factor.

Full Dynamic Analysis

A full dynamic analysis should always be carried out for important structures, in particular if
transient or permanent deflections or movements of the bridge structure and/or the fender or
buffer system are introduced in the analysis.
In this approach, a full dynamic model is conveniently established by applying Finite-Element-
Model (FEM) techniques. The calculations are carried out in the time domain using a force-in-
dentation relationship, ref. Section 6, as a force function and appropriate small time-steps with
constant force. The step-wise calculations are repeated until all of the kinetic energy is dissipated.

7.3 Design of Substructure

Head-on Bow or Sideways Collision

As it appears from Section 6.1, the forces involved in head-on bow collisions have been
extensively investigated. Some main results together with recommendations as to the selection of
design basis can be found in Section 6.1.
The forces from sideways collisions have normally been considered covered by the assumed
bow collision forces. In narrow channels, however, where no head-on collision to piers is physi-
cally possible, it may still be necessary to consider sideways impact against structures lining the
edges of the channel. Reference is made to Section 6.2.

Local Loads
All surfaces and structural members exposed to collision should be designed for local loads.
Some indications of the order of magnitude of local forces: point loads, line loads, and area
loads can be found in Section 6.4.
 Design of Superstructure 83

Location of Impact Forces

The global collision loads should be in the direction of the vessel travel. It is normally speci-
fied that 50% of the global load shall be applied to the substructure in the direction perpendicular
to the vessel travel, ref. Nordic Road Engineering Federation [7-61 and Figure 6.4 and Knott et al.
[7-41.

The purpose of this is to cover a series of possible collision scenarios which are very difficult
to quantify, i.e.:
- collision from vessel moving in other directions than the navigated channel direction
- eccentric collision
- glancing collisions.

Traditionally, the global impact force is applied at the mean high water level. However, the
local water level variation as well as the bow shape of the design vessel should be taken into
account.
The local load components should be applied in accordance with the geometry of the colliding
design vessel. In this connection it is very important to consider the shapes of the bulbous bows,
the overhang or rake of the bows, etc. for all relevant ships and barges in the waterway.

7.4 Design of Superstructure

Head-on Bow or Sideways Collision

If the superstructure of the bridge is so low that it can be hit by the hull of a vessel, an upper
limit of the design collision force would be the force valid for head-on bow collision. On this
basis the US Guide Specification [7-41 specifies the following impact force relationship:
PBH = RBH * pS

where:
PB, = impact force on exposed superstructure
R B , = ratio of exposed superstructure depth to total bow depth

Ps = impact force on pier.

This magnitude of force can not be sustained by ordinary types of bridge superstructures. This
is, in particular, the case for movable bridges which could be struck in the closed position in case
of electrical or mechanical failure of the bridge machinery.
Where possible, sufficient clearance of the bridge spans should be arranged to avoid bow
collision.
84 Bridge Design

Deckhouse Collision
The magnitude of impact forces involved in a collision between the deckhouse of a ship and an
exposed superstructure span are of a more manageable order of magnitude when compared with
the above-mentionedbow impact forces. However, very little research has been carried out on this
subject. Some indications can be found in Section 6.3.

Local Loads
All surfaces and structural members exposed to collision should be designed for local loads.
Some indications of the order of magnitude of local forces (point loads, line loads, and area
loads) can be found in Section 6.4.

Location of Impact Forces

The global collision loads should be applied in the direction of the vessel travel. Depending on
the shape of the superstructure contact area, vertical load components may have to be considered
in addition to the horizontal load. The local load component should be applied in accordance with
the geometry of the colliding design vessel.
 85

8. PREVENTION MEASURES

Planning and implementationof prevention measures to improve safe navigation in the water-
way near a bridge crossing requires close cooperation with the relevant navigation authorities.
Where the matter is of international concern, the national authorities will approach the interna-
tional authorities, in most cases the International Maritime Organisation (IMO),ref. IMO [8-71.
Prevention or reduction of the frequency of collisions is achieved by providing assistance to
navigation. The assistance may differ in extent and level of sophistication, depending on the
waterway and the intensity of the navigation. In this connection, inspiration may be obtained from
experience within the offshore sector, ref. Vendrell[8-121.

Three levels of assistance are discussed in the following:

- installation of navigational aids on the bridge and in the waterway
- introduction of navigation regulations
- implementationof a vessel M i c management system.
Comments on the effectiveness of the different types and levels of assistance are given in
Section 5.2.

8.1 Aids to Navigation

Visual, sonar as well as electronic aids to navigation should be designed to provide safe
guidance in most weather conditions.

For installation on the bridge structure,the following types of navigational aids can be consid-
ered to provide better detectability:
- colours (fluorescent)
- signs
- high intensity light beacons (flashing)
- rangelights
- sound devices (fog horns)
- racon installation at the centre line of the vessel track@).
An example of installationson a bridge to aid navigation is shown in Figure 8.1.
86 Prevention Measures

Figure 8.1 Navigation equipment mounted on the navigation span of the Great Belt western
bridge for southbound vessels ('om Olsen et al. [8-101).

For location in the waterway near the bridge crossing, the following types of navigational aids
can be considered:
- lighthouse
- buoys
- beacons
- racon installations.

Fewer but larger buoys with strong lights and fluorescent paint should be utilized rather than
many small buoys.
Racons (self-identifyingradio beacons activated by the radar signals of passing vessels) should
always be installed in connection with critical bridge passages.
The most suitable arrangement of navigational aids may be obtained through manoeuvring
simulations with pilots.
 Vessel Traffic Management Systems 87

8.2 Vessel lkaffic Regulations

Traffic regulation measures should be studied in cooperation with local navigation authorities.
The regulations may involve both mandatory and voluntary elements. Different possibilities exist:
- separation of navigation routes for opposite travelling directions
- limitation of maximum vessel speed (it should be remembered that the manoeuvring
capabilities of a vessel are reduced with reduced speed)
- requirement for pilotage (in national waters it is possible to introduce compulsory pilotage,
whereas international waters require involvement of relevant organizations such as IMO)
- requirement for tug assistance (the comments related to pilotage also apply to tug assistance)
- restriction on passage in bad weather or bad sea conditions
- limitation of maximum vessel size
- restriction on passages of large vessels to daylight transits only
- demand on empty vessels to take on ballast for a minimum draught
- special demands on vessels with hazardous cargoes.

8.3 Vessel lkaffic Management Systems

Traffic management systems of varying sophistication can be considered, including facilities

as follows:
- radar surveillance
- guard boat surveillance
- transmission from data buoys of information on meteorological conditions, sea and tide, etc.
- transmission of navigational information
- navigational guidance by pilots directly or over radio
- radio communication between vessel and bridge personnel.

The management system should cover the navigation channel in full width within a distance
sailed in one hour, i.e. approximately 15 nautical miles on both sides of the navigation span.
Information on Vessel Traffic Management Systems is mostly published under the title Vessel
Traffic Services (VTS).Reference is made to Fujii et al. [8-21, Hara [8-41, and Mizuki et al. [8-91.
These systems include advanced shore-based radar systems with real-time surveillance capabili-
ties, as proposed by Greneker et al. [8-31, and as implemented for the Lake Pontchartrain cause-
way in Louisiana, USA [8-131,and for the Great Belt crossing in Denmark [8-101as well as small
portable navigation units carried on board by the master or pilot of the vessel, as described by
Knott [8-81.
The Vessel Traffic Management System planned for the Great Belt crossing is shown in
Figure 8.2.
 88 Prevention Measures

The system includes:

Radar installations (RADAR)
VHF radio communication (VHF)
VHF directionfinder (DF)
Closed circuit television (TV)
Infrared television (IR)
Vessel T r m c Service center (VTS).

Figure 8.2 Outline of VTS system covering the waters of the Great Belt in Denmark (from Olsen
et al. [8-101).
 Vessel Traffic Management Systems 89

A number of recent publications on VTS systems are briefly mentioned below:

- The World VTS Guide by IALA, IAPH, IMPA [8-51 contains information on existing VTS
systems worldwide.
- The Third Survey on VTS in the World by Mizuki et al. [8-91 also contains information on
existing VTS systems worldwide.
- The Intemational Maritime Organization (IMO) has issued Guidelines for Vessel Traffic
Servicesref. [8-6].
- The Vessel It.affc Services Equipment and Technology Report (VTSHandbaok) [8-113 is
intended to be used as a guide by the US Coast Guard in the selection of equipment and
technology for VTS systems implementationand/or enhancements.
- In COST 301. Shore-based marine navigation aid systems, ref. Fabre et al. [8-11, functions
and effectiveness of existing VTS systems in the EEC area are covered.
 91

9. PROTECTION MEASURES

In addition to bridge damage, vessel collisions may result in serious environmental damage
such as the spilling of oil and other chemicals. The consequences of a vessel collision may
therefore reach far beyond the direct costs of repairing/replacing the vessel and possibly the
bridge.
The bridge elements can be designed to withstand the impact loads, or a fender or protection
system can be developed to prevent, redirect, or reduce the impact loads on bridge elements to
non-destructive levels.
If the force resistance of the protection system is higher than the vessel crushing force, the bow
of the vessel will crush and the impact energy will be primarily absorbed by the vessel.
If the vessel crushing force is higher than the resistance of the protection system, the impact
energy will be primarily absorbed by the deflection and crushing of the protection system.
The protection system should be designed not only to protect the bridge structure, but also to
protect the vessel and the environment against serious damage. This may be achieved by combin-
ing different types of protective systems. Protection systems may be located directly on the
bridge structure (such as a bridge pier fender), or independent of the bridge (such as a dolphin).
The geometry of the protective structure should be developed to prevent the rake (overhang) of
the design vessels bow from striking and causing damage to any exposed portion of the bridge.
Generally, the analysis and design of bridge protection structures requires the use of engineering
judgment to arrive at a reasonable solution. Surveys of protection measures can be found in
Frandsen et al. [9-51,Knott et al. [9-lo],Larsen et al. [9-121,Ostenfeld [9-201,and Saul et al.
[9-211,[9-221,and [9-231.

The various types of protective structures commonly used for bridges will be briefly discussed
in the subsections below:
- Fender Systems
- Pile Supported Systems
- DolphinProtection
- Artificial Island or Reef Protection
- Floating Protection Systems.

9.1 Fender Systems

Timber fenders are composed of vertical and horizontal timber members in a grillage geome-
try attached to the face of the bridge pier, or erected as an independent structure adjacent to the
92 Protection Measures

pier. Energy is absorbed by elastic deformation and crushing of the timber members. Because of
their relatively low cost, timber fenders have frequently been used on bridge projects for protect-
ing piers from minor vessel impact forces. However, for relatively large collision impact loads the
resulting timber fenders would have to be extremely large, and might be uneconomical in most
circumstances.
Rubber fenders are commercially available in a wide variety of extruded and built-up shapes.
Impact energy is absorbed through the elastic deformation of the rubber elements either in com-
pression, bending, shear deformations, or a combination of all three.
Concrete fenders consist of hollow, thin-walled, concrete box structures attached to the bridge
pier. Usually, a timber fender is also attached to the outer face of the concrete box fender. Impact
energy is absorbed by the buckling and crushing of the concrete walls composing the fender
system. An example of this type of fender is shown in Figure 9.1.

Concrete Pier

1
Reinforced Concrete Box

Figure 9.1 Crushable concrete box fender on the Francis Scott Key Bridge, Baltimore, USA
(fromKnott et al. [9-101).

Steel fenders consist of thin-walled membranes and bracing elements composed in a variety of
box-like arrays and assemblies attached to the bridge pier. Impact energy is absorbed by com-
pression, bending, and buckling of the steel elements in the fender, ref. Matsuzaki et al. [9-151
and Namita et al. [9-171. Timber facing should be attached to the steel fender to prevent sparks
resulting from direct contact with steel hulled vessels. An example of this type of fender is
shown in Figure 9.2.
 Pile Supported Systems 93

Figure 9.2 Framed steel fender system, also referred to as a multi-cell type buffer, used for
protection of the Bisan-Seto Bridges in Japan (from Matsuzaki et al. [9-151).

9.2 Pile Supported Systems

Pile groups connected together by rigid caps may be used for protection to resist vessel impact
forces. Free standing piles and piles connected by relatively flexible caps are also used for
bridge protection. The pile groups may consist of vertical piles, which primarily absorb energy by
bending, or batter piles which absorb energy by compression and bending.
As a result of the high impact design loads associated with vessel collision, plastic deformation
and crushing of the pile structure is permitted provided that the vessel is stopped before striking
the pier, or the resulting impact is below the resistance strength of the pier and foundation. Fender
systems may be attached to the pile structure to help resist a portion of the impact loads. Timber,
steel, or concrete piles may be utilized depending on site conditions, impact loads, and economics.
An example of this type of protective structure is shown in Figure 9.3.
94 Protection Measures

I
I

PLAN

W 2 0 . St. 52-3
ELEVATION
Figure 9.3 Pile supported protection system for the Trams@ Bridge in Norway (from Tambs-
Lycke [9-251).
 Artificial Island or Reef Protectlon 95

9.3 Dolphin Protection

Large diameter dolphins may be used for protection of bridge piers, see Englot [9-31, Hahn et
al. [9-61, Heins [9-81, Heins et al, [9-91, Knott [9-111, and Naoi et al. [9-181.
Dolphins are typically circular cells constructed of driven steel sheet piling, filled with rock or
concrete, and topped by a concrete cap. Dolphins may also be constructed of precast concrete
sections, or precast entirely off-site and floated into final position.

Driven pilings are sometimes incorporated in the cell design. Design procedures for dolphins
are usually based on an estimate of the energy changes that take place during the design impact
loading. Energy-displacement relationships are typically developed for the following energy
dissipating mechanisms:
- crushing of the vessels bow
- lifting of the vessels bow
- generation of water waves and turbulence
- friction between the vessel and the dolphin
- friction between the vessel and the river bottom
- sliding of the dolphin
- rotation of the dolphin
- deformation of the dolphin.
Deformation of the vessel/dolphin system is assumed to follow a path of least energy. For each
potential displacement configuration of dolphin and vessel, a deformation path can be developed.
Deformation stops when all the kinetic energy of the impact has been absorbed.
For purposes of design, it is recommended that the maximum dolphin deformation be limited
to less than one-half the diameter of the cell. Under design loading conditions, the cell is permit-
ted to undergo large plastic deformation and partial collapse. Examples of this type of protective
structure are shown in Figure 9.4.

9.4 Artificial Island or Reef Protection

Artificial islands around bridge piers or artificial reefs in front of bridge piers provide highly
effective vessel collision protection, ref. Brink-Kjaer [9-11, Denver [9-21, Fletcher et al. [9-41,
Havnoe et al. [9-71, Minorsky [9-141, and Sexsmith [9-241.
If the waterflow through the navigation channel is important to the surrounding environment,
investigations about the net waterflow before and after installationof the protection islands should
be carried out.
96 Protection Measures

Elevation 17.0 - Navigation Light

5'4' r Elevation 15.0

Elevation-30 f
Exirting Bay
I I-
c
SU
I
Diameter
k Varim
(45' to 60')

Figure 9.4 Examples of Dolphin protection. At the top, the Sunshine Skyway Bridge in Florida,
USAmo'( Knott [9-11]). Below, the Rio-Niteroi Bridge in Brazil (photo: Niels
Jlrgen Gimsing).
 Floating Protection Systems 97

Islands typically consist of a sand or rock core which is protected by outer layers of heavy rock
armour to provide protection against wave, current, and ice actions.

The island geometry should be developed in accordance with the following criteria:
- The vessel impact force transmitted through the island'to the bridge pier must not exceed the
lateral capacity of the pier and pier foundation.
- The island dimensions should be such that vessel penetration into the island during a collision
will not result in physical contact between the vessel and any part of the bridge pier.

The second requirement is particularly critical for empty or ballasted ships and barges which
can slide up on the slopes of an island and travel relatively large distances before coming to a stop.
The design of the surface armour protection of the islands for wave, current, and ice attack may
be based on methodologies used for rubble mound breakwater design.

The following items have been identified as sources of energy absorptioddissipation during a
vessel impact with an island:
- crushing of the hull of the vessel
- lifting of the vessel
- generation of water waves and turbulence
- lifting of island material
- displacement, shear, and compaction of the island material
- friction between the vessel and the island
- generation of shock waves in the island
- crushing of particles of island material.

Inclusion of these items in a design analysis is difficult since their effects are only partially
understood. Physical model studies, as well as mathematical simulations, are usually required
when protective islands are designed. An example of the arrangement of artificial islands for ves-
sel collision protection is shown in Figure 9.5.

9.5 Floating Protection Systems

Various types of floating protective systems may be considered by the engineer. Several of
these systems include:
- Cable net systems. Vessels are stopped by a system of cables anchored to the waterway bottom
and suspended by buoys located in front of the bridge piers, ref. Larsen et al. [9-121, Oda et
al. [9-191, and Vitalis [9-261.
98 Protection Measures

Figure 9.5 Arrangement of protection islands for the Annacis Island Bridge at Vancouver in
Canada ('om Sexsmith [9-241).

- Anchored Pontoons. Large floating pontoons anchored to the waterway bottom in front of the
piers absorb vessel impact, ref. Mondorf [9-161.
- Floating Shear Booms. Floating structures anchored to the waterway bottom deflect vessels
away from piers and absorb impact energy.

Special consideration for corrosion protection must be made for all systems involving under-
water steel cables and anchorages.
Special consideration should be given to the function and vulnerability/durability of floating
systems during winter time in waters subject to icing or ice drift.
Floating systems are vulnerable to overrun by vessels with sharply raked bows.
An example of protection by cable net system is shown in Figure 9.6.
 Floating Protection Systems 99

Figure 9.6 Arrangement of cable net system proposed for vessel collision protection of the
Gibraltar Bridge ('om Larsen et al. [9-121).
 101

10. PROTECTION OF THE PUBLIC

To minimize the loss of life which may occur in the event of a catastrophic collapse of a
bridge during a vessel collision, bridge user warning systems may be introduced, ref. [lo-11,
[lo-21, and [lo-31.

10.1 Collision Hazard Detection

Devices to detect vesselbridge collision hazards include the following:

- Vessel Impact Vibration Detectors. Placed on bridge piers, these vibration sensors would be
capable of distinguishing between normal structural vibrations and movements associated with
substantial vessel impacts.
- Continuity Circuits. This electrical system would utilize pairs of conductors terminating with
end-of-line devices attached to the bridge superstructure. Collapse of some portion of the
bridge deck would interrupt the circuit continuity.
- VHF Radio Link.The use of this device would be in advance of imminent danger, as foreseen
by the pilot or master of a vessel which had, for instance, lost steerage. If the mariner anticipat-
ed a possible vesselbridge collision, he would radio the bridge personnel, or other appropriate
agency, via VHF marine emergency channel in order to halt traffk on the bridge.

Either of the first two of the above devices could activate traffk controlhiformation systems
automatically or through a machine-man-machineinterface with the human intermediary verify-
ing hazards before interrupting traffk.
VHF radio units are readily available in the deckhouse of virtually every merchant vessel. The
use of such a system would require the installation of a relatively inexpensive VHF set and con-
tinuous monitoring by the bridge personnel who could make appropriate traffic control decisions.
Virtually any detection device can be electronically linked to traffk controlhnformation equip-
ment in order to automatically warn or stop traffic. However, in actual practice, considerable
difficulty can be experienced with false alarms and unnecessary interruptions of traffic.

Included among possible verification methods to be carried out by bridge personnel before traf-
fic control actions are taken, are the following:
- Closed Circuit Television (CCTV). Cameras can be placed strategically to allow personnel at a
monitor site to view the bridge main span, the navigation channel, the roadway, or any other
feature desired.
- Visual Delineation. The top of the bridge parapet or guardrail would be fitted with a series of
reflectors or lights, immediately revealing the collapse of a portion of the bridge superstructure.
102 Protection of the Public

10.2 Bridge Traffic Control

Whether the hazard detector information is used automatically or is manually verified, the
ultimate function of a traffk safety system is to appropriatelycontrol traffic or inform bridge users
of hazards. The following devices can be used to accomplish this function:
- Variable Message Signs. Virtually any message can be transmitted via this device, including
warnings of catastrophic bridge failure, environmental hazards, traffic congestion, construc-
tion/maintenance activity, etc.
- Flashing Beacons. Used in conjunction with standard format warning signs, these devices can
be used to bring attention to a warning message.
- Movable Gates, usually fitted with flashing red lights and an audio alarm (siren or bell), this
device can be lowered across traffic lanes to halt traffk (as at railroad crossings).
 103

Appendix A. Vessel Size and Geometry Data

A.l. Vessel Size Measurements

Dead Weight Tonnage (DWT)

The Dead Weight Tonnage expresses the carrying capacity of a vessel, i.e. the maximum weight
in metric tonnes of cargo, fuel, water, stores, etc. on board the vessel when fully loaded.
DWT equals the difference between a vessels displacement at fully loaded draught and light
ballast draught. It is the normal unit for bulk carriers and tankers.

Displacement Tonnage (W)

The Displacement Tonnage expresses the total weight in metric tonnes of the vessel including
cargo, fuel, water, stores, etc. The displacement thus depends on the loading condition.
The weight of the ship when in a light or unloaded state, but including the weight of water in
boilers and any permanent ballast is called light displacement. The light displacement weight
plus the weight of cargo, fuel, stores, fresh water, and water, ballast is called load displacement.
W equals the weight of water displaced by a vessel when floating at the specific draught.
W is the normal unit for warships.

The Displacement Tonnage for a fully loaded vessel may be approximated to Dead Weight
Tonnage as follows:
- Tankers: DWT = 0.9 x W
- Bulk Carriers: DWT = 0.8 x W
- Container Vessels: DWT = 0.6 x W

Gross Register Tonnage (GRT)

The Gross Register Tonnage expresses the internal volume of a vessel (less certain exempted
spaces) measured in cubic feet divided by 100. GRT is thus not a measurement of weight.
GRT is a normal unit for passenger ships and cargo liners and is used as a basis for safety
requirements and manning.

The Gross Register Tonnage may be approximated to Dead Weight Tonnage as follows:
- Tankers: DWT = 1.6 x GRT
- Bulk Carriers: DWT = 1.9 x GRT
- Container Vessels: DWT = 0.9 x GRT
104 Appendix A

Net Register Tonnage (NRT)

The Net Register Tonnage expresses the internal volume of a vessel used for the carriage of
cargo or passengers, measured in cubic feet divided by 100. NRT is thus not a measurement of
weight.
The NRT is the GRT less the volumes of certain spaces not used for the carriage of cargo (en-
gine room, certain water tanks, etc.).
NRT is frequently the basis on which harbour duties and pilotage fees are levied. It is a normal
unit for passenger ships and cargo liners.

A.2. Typical Ship Vessel Geometry Data

A.2.1 Air Draught in Ballast Condition

A.2.2 Air Draught in Loaded Condition
A.2.3 Height of Deck House in Ballast Condition
A.2.4 Height of Deck House in Loaded Condition
A.2.5 Draught in Ballast Condition
A.2.6 Draught in Loaded Condition
A.2.7 Length (LOA)
A.2.8 Breadth (Beam)
A.2.9 Moulded Depth
 Typical Ship Vessel Geometry Data 105

1.000 10,000 m,ooo 1,000,000

DWT in Tonnes

A.2.2 Air Draught in Loaded Condition

Metres

tdoo l0,Ooo ,000

DWT in Tonnes
106 Appendix A

A.2.3 Height of Deck House in Ballast Condition

MbtfW

26
- -

I
[-Container -Bulk Carrier Tanker

20

16

10

0
- 1 . . 1,oc
1,000 l0,WO lo0,boo ,000
DWT in Tonnes

A.2.4 Height of Deck House in Loaded Condition

Metre8

I -

1,doo m,ooo 1.01 ,000

DWT in Tonnes
 Typical Ship Vessel Geometry Data 107

A.2.5 Draught in Ballast Condition

Metrea

tdoo 10.000 too ,000

DWT in Tonnes

tOOl ,000
DWT in Tonnes
108 Appendix A

A.2.7 Length (LOA)

Metres

1.wo 1 0 . ~ m,ooo 1,000.000

DWT in Tonnes

A.2.8 Breadth (Beam)

Metres
80

70

60

60

40

30

20

10

- 1
1,000 10,000 100.000 1.00 ,000
DWT in Tonnes
 Typical Ship Vessel Geometry Data 109

A.2.9 Moulded Depth

Metre8
40
- Contai

90

20

lo

0
u D l0,ooo too
DWT in tonnes
1 10 Appendix A

A.3. Typical Barge Vessel Geometry Data

A.3.1 Barges and Barge Tows in the USA

If.pical dimensions and capacities of barge vessels and tows in use in US inland waterways are
shown in the table below. Reference is made to:
- Modjeski & Masters, Consulting Engineers:
Criteria for: The Design of Bridge Piers with Respect to Vessel Collision in Louisiana Water-
ways, Prepared for the Louisiana Department of Transportation and Development and the
Federal Highway Administration, New Orleans, July 1985.
- Knott, M.A. and Larsen, 0. Damgaard:
Guide Specification and Commentary for Vessel Collision Design of Highway Bridges,
US Department of Transportation,Federal Highway Administration,Publ. No. FHWA-RD-91-
006. Dec. 1990.

Barge 5 P e Jumbo Hopper Oversize Tank

Length (m) 59.4 88.4
Width (m) 10.7 16.2
Depth of Vessel (m) 3.7 3.7
Empty Draught (m) 0.5 0.5
Loaded Draught (m) 2.7 2.7 3.8
Empty Displacement (tonnes) 180 540 1200
Loaded Displacement (tonnes) 1700 3900 5700
Average number of barges
along the length of a tow 3 2
 Typical Barge Vessel Geometry Data 111

A.3.2 Barges and Barge Tows in Europe

Typical dimensions and capacities of barge vessels and tows in use in European inland water-
ways, classes I to VI, ref. The European Conference of Ministers of Transport (ECMT), are
shown in the figure below.

II
-
-
39 x 5.1 x 2.4m

Motor wutl:
Kampanaar (8001)
55 x 6.6 x 2.5m

111 1
-- Motor v w d :
Dortmund Eams -

--
I b n a l ship (10001)
67.8 x 8.2 x 2.6m

Motor w w l : Rhino Herna canal ship Push tow unit : (17001)

(13601) 85 x 9.5 x 2.8m 70.5 x 9.5 x 3.211

v
Motor wutl : Large Rhino vessel Push tow unit : (22501)
(2OOOt) 95 x 11.5 x 2.7m 76.5 xl1.4 x 3.2r

Push tow unit :

(2 x 22501)
1 5 3 x 11.4x3.2m ( I ,[D 1
Push tow unit :
(2 x 22501)
76.5 x 22.8 x 3.2m

VI Push tow unit :

(4 x 22501)

Push tow unit :

(6 x 22501)
 113

Appendix B. Records of Collision Accidents

Serious Vessel Collision Accidents recorded in the Period 1960 -1991

1960 Sevem River Railway Crossing, England

Vessel : Tug pulling two barges each 450 tonnes displacement
Accident : Broadside collision with pier
Damage : Two spans and one supporting pier collapsed causing five fatalities
Cause : Tug pilots negligence in dense fog
Lit. : Peter Mason: An investigation into the cause of damage to the Sevem
Railway Bridge. The Structural Engineer No. 2,1963

1964 Maracaibo Lake, Venezuela

Vessel : 36,000 DWT loaded tanker
Accident : Broadside collision with two piers more than 600 m from the
navigational spans
Damage : Three spans collapsed
Cause : Failure in electrical system affecting steering gear
Lit. : Engineering News Record, 1964-04-16 and 1964-12-24

1964 Pontchartrain Lake, Louisiana, USA

Vessel : Tug towing two loaded barges
Accident : Three trestles were struck by the tug and the two barges
Damage : Four spans collapsed, causing six fatalities
Cause : Helmsmans lack of attention
Lit. : Engineering News Record, 1964-06-25

1964 Pontchartrain Lake, Louisiana, USA

Vessel : Tug towing two barges
Accident : Tug hit a pile bent
Damage : One pile bent was destroyed and two spans collapsed
Cause : Tug pilots inattention (possibly asleep)
Lit. : Engineering News Record, 1964-07-30

1967 Chesapeake Bay, Virgina, USA

Vessel : Drifting coal barge
Accident : Vessel thrown repeatedly against the bridge deck
Damage : Six spans were seriously damaged
Cause : Barge tom loose from moorings in storm
Lit. : Engineering News Record, 1967-12-14
1 14 Appendix B

1970 Chesapeake Bay, Virgina, USA

Vessel : 10,000 DWT US-navy cargo ship
Accident : 1 hour battering against the bridge
Damage : Five spans knocked down and 11 other spans damaged
Cause : Vessel tom loose in stormy weather
Lit. : Engineering News Record, 1970-01-29 and 1970-03-12

1972 Chesapeake Bay, Virginia, USA

Vessel : Drifting barge
Accident : Barge thrown repeatedly against the bridge deck
Damage : Two spans partially collapsed and five other spans were damaged
Cause : Towline to tug broken in heavy wind
Lit. : Engineering News Record, 1972-09-28

1972 Brunswick River, Georgia, USA

Vessel : 13,000DWT freighter
Accident : The superstructurehit by the bow of the ship
Damage : Three spans collapsed, causing ten fatalities
Cause : The helmsman misunderstood the pilots instructions
Lit. : US National Transport Safety Board and US Coast Guard:
SS African Neptune collision with the Sidney Lanier
Bridge, Marine Accident Report, 1973

1974 Pontchartrain Lake, Louisiana, USA

Vessel : Tug pulling four empty barges
Accident : Tug hit pier some way from navigation span
Damage : Two pile bents were damaged and three spans collapsed,
causing three fatalities
Cause : Tug pilot asleep
Lit. : Engineering News Record, 1974-08-08

1974 Welland Canal, Ontario, Canada

Vessel : Ore carrier (length: 204 m)
Accident : Ship rammed lifting span while opening
Damage : Lift span fell in canal and lift towers were damaged
Cause : Unknown
Lit. : Engineering News Record, 1974-08-29 and 1974-09-12
 Records of Collision Accidents 1 15

1975 Derwent River, Hobart, Tasmania, Australia

Vessel : 7,200 DWT bulk carrier
Accident : Head-on and broadside collision with two piers
Damage : Three spans collapsed, causing 15 fatalities
Cause : Loss of steering ability due to engine failure
Lit. : Engineering News Record, 1975-01-09and 1975-01-16

1975 Fraser River, New Westminster, British Columbia, Canada

Vessel : Barge in ballast (length: 183 m)
Accident : Barge hit the bridge superstructure
Damage : One 120 m span collapsed
Cause : Barge tore loose from mooring in heavy rainstorm
Lit. : Canadian newspaper Vancouver Sun, 1975- 12-27

1976 Pass Manchac Channel, Louisiana, USA

Vessel : Barge towed by a tug
Accident : Barge hit a pile bent
Damage : Pile bent destroyed and three spans collapsed, causing at
least one fatality
Cause : Barge off course due to careless navigation
Lit. : Engineering News Record, 1976-09-23

1977 James River, Hopewell, Virginia, USA

Vessel : 25,000 DWT tanker in ballast
Accident : The stem of the ship destroyed a pier bent about 120 m
from the navigational span centerline
Damage : Two spans collapsed
Cause : Electrical fault in steering gear
Lit. : US National Transport Safety Board US Tank Ship SS
Marine Floridian, Collision with Benjamin Harrison Memorial Bridge,
Marine Accident Report, 1978

1977 Passaic River, Union Avenue, New Jersey, USA

Vessel : Empty oil barge
Accident : Collision with a pier
Damage : Two spans collapsed
Cause : Broken towline to tug
Lit. : Engineering News Record, 1977-08-05
1 16 Appendix B

1977 Gothenburg Harbour, Tingstad, Sweden

Vessel : 1600 DWT gas tanker in ballast
Accident : Ship hit approach spans
Damage : Two approach spans destroyed
Cause : Electrical fault in steering gear
Lit. : Danish newspaper POLITIKEN, 1977-09-11

1978 Berwick Bay, Southern Pacific Railroad, Louisiana, USA

Vessel : Tug pushing four barges
Accident : Lead barge hit the side span bridge superstructure
Damage : One 70 m steel span fell into water and sank
Cause : Tug skipper navigating with underpowered tow
Lit. : US National Transport Safety Board: Collision of M / v Stud with the
Southern Pacific Railroad Bridge.. ., Marine Accident Report, 1980

1979 Second Narrows Railway Crossing, Vancouver, Canada

Vessel : 22,000 DWT bulk carrier
Accident : Stem of ship struck the superstructure in side span about 100 m
from the navigational span centre line
Damage : One span collapsed
Cause : Captains misjudgment of land marks due to dense fog
Lit. : New Civil Engineer International, June 1980

1980 Tjorn, Almo Sound, Sweden

Vessel : 27,000 DWT product carrier in ballast
Accident : Deck house of ship struck the arch structure near the foundation
on shore, about 100 m from the navigation channel centre line
Damage : Total collapse of the main span causing eight fatalities
Cause : Steering difficulties in rough weather and dense fog due
to reduced engine power
Lit. : Statens Havarikommission: Utredningsrapportbetrtiffande . ..
.
Star Clippers pgsegling af Almobroen . ,, Stockholm, April 198 1

1980 Tampa Bay, Sunshine Skyway, Florida, USA

Vessel : 35,000 DWT bulk carrier in ballast
Accident : Stem of ship struck bridge column above pier top about
250 m from navigational channel
Damage : Three spans collapsed, causing 35 fatalities
Cause : Pilots careless navigation in rough weather with reduced visibility
Lit. : US National Transport Safety Board: Ramming of the Sunshine Skyway
Bridge, Marine Accident Report, 1981
 Records of Collision Accidents 1 17

198 1 Narragansett Bay, Newport, Rhode Island, USA

Vessel : 45,000 tonnes displacement tanker
Accident : Ship struck main tower pier of suspension bridge head-on
at 3 m/sec velocity
Damage : Only superficial damage although the collision force substantially
exceeded the design force
Cause : Pilots careless navigation in dense fog
Lit. : T.R. Kuesel: Newport Bridge Collision, IABSE Colloquium on Ship
Collision with Bridges and Offshore Structures, Copenhagen, 1983

1982 Mosel River, Richemont Gas Pipeline, France

Vessel : Tug pushing two barges
Accident : Barge struck bridge pier
Damage : One pier was destroyed resulting in collapse of the gas pipeline
causing seven fatalities
Cause : Tug pilots careless navigation in dense fog
Lit. : French newspaper Le Figaro, 1982-01-18

1982 Mississippi River, Hannibal, Missouri, USA

Vessel : Tug pushing 15 barges
Accident : Barges struck abutment while passing swing span, lost control
and tug swung into approach span
Damage : One approach span collapsed
Cause : Careless navigation
Lit. : Engineering News Record, 1982-05-13

1983 Sentosa Aerial Tramway, Singapore

Vessel : Petroleum drill ship with 69 m high mast
Accident : Ship drifted into tramway and the mast severed the aerial cables
of the tramway
Damage : Two tramway cars fell into harbour causing seven fatalities
Cause : Tow line failed when ship was warped out of berth by tug
Lit. : Local newspapers
118 Appendix B

1983 Volga River Railway Crossing, Ulyanovsk, Russia

Vessel : Passenger vessel Alexander Suwarow
Accident : Vessel strayed off course and ran too close to the support column for the
arch bridge and the upper deck of vessel collided with bridge
superstructure
Damage : The deckhouse of the vessel, including a film hall, was torn off causing
approx. 170 fatalities
Cause : Captains careless navigation
Lit. : European newspapers, June 1983

1990 Herbert C. Bonner Bridge, North Carolina, USA

Vessel : Hopper dredge Northerly Isle (length: 60 m)
Accident : Vessel drifted into the substructure of the bridge
Damage : 4 pile bents were demolished and five spans collapsed
Cause : The vessel dragged its anchors during a storm
Lit. : Engineering News Record, 1990-11-01

1990 Tostero, Strfingntis, Sweden

Vessel : Freighter Arosandra (length: 60 m)
Accident : Vessel struck the swing span support pier and side span superstructure
Damage : The pier was displaced and the bridge superstructure in side
span was partly tom down
Cause : Captains careless navigation due to drunkenness
Lit. : Local newspapers

1991 Camafuli River, Chittagong, Burma

Vessel
, : Unknown
Accident : Vessel drifted into the superstructure of the bridge
Damage : One superstructure span tom down
Cause : Vessel lost control due to cyclone
Lit. : Personal Communication from Dr. Y. Fujii, Dec. 1991

1991 Hamburg Harbour, Kattwyk, Germany

Vessel : 21,450 tonnes freighter Stanislaw Kulczynski
Accident : Freighter under tow by three tugs in dense fog lost control and crashed
into the side span of the lift bridge
Damage : The side span fell into the river and one of the lift towers
was seriously damaged
Cause : The vessel lost control due to loss of one of the three tow lines
Lit. : Construction Today, Oct. 1991
 119

Appendix C. Selected Literature

The literature list is organized in accordance with the sections of the publication:

-
SECTION 1 INTRODUCTION
[I-11 Frandsen, A. G.: Accidents Involving Bridges, IABSE Colloquium on Ship Collision
with Bridges and Offshore Structures, Copenhagen, 1983.
[ 1-21 Frandsen, A. G. and Langsoe, H.: Ship Collision Problems: I. Great Belt Bridge and 11.
International Enquiry, IABSE Periodica No. 2,1980.
[ 1-31 Frandsen, A. G., Olsen, D., Fujii, Y. and Spangenberg, S.: Ship Collision Studies for
the Great Belt East Bridge, Denmark, IABSE Symposium on Bridges-Interaction
between Construction Technology and Design, Leningrad, 1991.
u-41 Jensen, A. 0.and Sprrensen, E. A.: Ship Collision and the Faroe Bridges, IABSE
Colloquium on Ship Collision with Bridges and Offshore Structures,Copenhagen 1983.
[I-51 Knott, M.A. and Bonyun, D.: Ship Collision Against the Sunshine Skyway Bridge,
IABSE Colloquium on Ship Collision with Bridges and Offshore Structures, Copen-
hagen, 1983.
[I-61 Knott, M. A.: Ship Collision with Bridges, Bulletin of Permanent International Asso-
ciation of Navigation Congresses (PIANC), No. 57,1987.
[I-71 Knott, M.A. and Larsen, 0. Damgaard Guide Specification and Commentary for
Vessel Collision Design of Highway Bridges, US Department of Transportation,
Federal Highway Administration, Publ. No. FHWA-RD-91-006, Dec. 1990.
[I-81 Larsen, 0. Damgaard and Benmoussa, C.: Ship Collision Aspects for Bridge across the
Strait of Gibraltar, Third International Colloquium on the Fixed Link Europe-Africa
Through the Strait of Gibraltar, Marrakesh, May 1990.
[ 1-91 Maunsell & Partners and Brady, P. J. E.: Second Hobart Bridge-Risk of Ship Colli-
sion and Methods of Protection, Technical Report prepared for Department of Main
Roads, Tasmania, Australia, 1978.
[l-101 Modjeski & Masters, Consulting Engineers: Criteria for: The Design of Bridge Piers
with Respect to Vessel Collision in Louisiana Waterways, Prepared for the Louisiana
Department of Transportation and Development and the Federal Highway Administra-
tion, New Orleans, July 1985.
[l-111 Norwegian Public Roads Administration: Report on Vulnerable Bridges Crossing
Navigable Waters (in Norwegian), Oslo, 1982.
[l-121 Olnhausen, W. Von: Ship Collisions with Bridges in Sweden, IABSE Colloquium on
Ship Collision with Bridges and Offshore Structures, Copenhagen, 1983.
[l-131 Ostenfeld, Chr.: Ship Collisions Against Bridge Piers, IABSE Publications, Vol. 25,
1965.
120 Appendix C

[l-141 Rasmussen, B. Hojlund: Design Assumptions and Influence on Design of Bridges,

IABSE Colloquium on Ship Collision with Bridges and Offshore Structures, Copen-
hagen, 1983.
[ 1-15] Tambs-Lyche, P.: Vulnerability of Norwegian Bridges across Channels, IABSE
Colloquium on Ship Collision with Bridges and Offshore Structures, Copenhagen,
1983.
[ 1-16] Transport Canada: Vulnerability of Bridges in Canadian Waters, Canadian Coast
Guard, January 1982.
11-17] US National Research Council: Ship Collisions with Bridges. The Nature of the
Accidents, their Prevention and Mitigation, National Academy Press, Washington
D.C., 1983.

-
SECTION 2 INITIAL PLANNING

12-11 Bay, J., Spangenberg, S., Olsen, N.H. and Pedersen, P.T.: Ship Simulations as an
Integrated Part of the Design Process for Bridges Crossing Waterways, Permanent
International Association of Navigation Congresses (PIANC), Bulletin 72, 1991.
r2-21 Frandsen, A. G. and Langsoe, H.: Ship Collision Problems: I. Great Belt Bridge and 11.
International Enquiry, IABSE Periodica No. 2, 1980.
12-31 Frandsen, A.G., Olsen, D., Lund H.T. and Bach, P.E.: Evaluation of Minimum
Bridge Span Openings Applying Ship Domain Theory, Transportation Research
Record No. 1313. Annual Meeting of Transportation Research Board of the National
Research Council, Washington D.C., 1991.
P-41 Frandsen, A. G., Olsen, D., Fujii, Y. and Spangenberg, S.: Ship Collision Studies for
the Great Belt East Bridge, Denmark, IABSE Symposium on Bridges-Interaction
between Construction Technology and Design, Leningrad, 1991.
12-51 Fujii, Y.: Integrated Study on Marine Traffic Accidents, IABSE Colloquium on Ship
Collision with Bridges and Offshore Structures, Copenhagen, 1983.
12-61 Fujii, Y.: Effective Areas of Ships, Journal of Navigation, Vol. 24, Oct. 1971.
P-71 Fujii, Y., Yamanouchi, H., Tanaka, Yamada, Okuyama and Hirano: The Behaviour of
Ships in Limited Water, 24th International Navigation Congress, Leningrad, 1977.
[2-81 Fujii, Y. and Tanaka, K.: Traffk Capacity, Journal of Navigation, Vol. 24, 1971.
12-91 Gardenier, J.S.: Safety of Bridges and Offshore Structures-the Role of Ship
Simulation, IABSE Colloquium on Ship Collision with Bridges and Offshore Struc-
tures, Copenhagen, 1983.
[2-lo] Goodwin, E. M.: A Statistical Study of Ship Domains, Journal of Navigation, Vol. 28,
1975.
[2-111 Japanese Association for Preventing Marine Accidents (JAPMA): Illustrated on
Marine Safety Law, Tokyo, 1985.
[2-121 Jensen, A.O. and Sorensen, E.A.: Ship Collision and the Faroe Bridges, IABSE
Colloquium on Ship Collision with Bridges and Offshore Structures, Copenhagen, 1983.
 Selected Literature 121

[2-131 Klap, Q. C.: Ship Collision Analysis for the WesterscheldeCrossing, IABSE Collo-
quium on Ship Collision with Bridges and Offshore Structures,Copenhagen, 1983.
[2-141 Knott, M.A. and Larsen, 0. Damgaard: Guide Specification and Commentary for
Vessel Collision Design of Highway Bridges, US Department of Transportation, Fed-
eral Highway Administration, Publ. No. FHWA-RD-91-006,Dec. 1990.
[2-151 Meurs, K. and Oosterbaan, J. W.: Simulation of Bridge Passage in High Winds,
IABSE Colloquium on Ship Collision with Bridges and Offshore Structures, Copen-
hagen, 1983.
[2-161 Okuyama, Y., Hayafuji,Y., Sasaki, Y. and Nakatsuji, T.: A Study on The Marine Traf-
fic Behaviours-Establishment of Methods of Observations and Analysis, Journal of
the Port and Harbour Research Institute (Japan), Vol. 15, No. 1,1976.
[2-171 Ostenfeld-Rosenthal, P. et al.: The Concept of Close Ship Encounters-A Decision
Tool for Bridge Navigation Span Widths, To be published in 1992.
[2-181 Shoji, K.: On the Design of the Waterways Passing Through the Bridges in View of the
Analysis on Ship Collision Accidents (in Japanese), Journal of the Tokyo
Mercantile Marine, No. 36, Tokyo, 1985.
[2-191 Shoji, K. and Iwai, A,: Preservation of Marine Structures against Ship Collision,
International Symposium on Ocean Space Utilization, Tokyo, 1985.
[2-201 Shoji, K. and Wakao, T.:On the Ships Waterways Passing Through Bridges, Water
Forum 86, World Water Issues in Evolution, San Francisco, 1986.
[2-211 Yamaguchi, A.: Study Report on the Navigation Safety of the Honshu-Shikohu
Bridges, Honshu-ShikohuBridge Authority, Tokyo, 1968.

-
SECTION 3 VESSEL TRAFFIC

Couper, A. (Editor): The Times Atlas and Encyclopaedia of the Sea, Times Books,
1989.
Knott, M.A. and Larsen, 0. Damgaard: Guide Specification and Commentary for
Vessel Collision Design of Highway Bridges, US Department of Transportation, Fed-
eral Highway Administration, Publ. No. FHWA-RD-91-006,Dec. 1990.
Larsen, 0. Damgaard and Benmoussa,C.: Ship Collision Aspects for Bridge across the
Strait of Gibraltar, Third International Colloquium on the Fixed Link Europe-Africa
Through the Strait of Gibraltar, Marrakesh, May 1990.
Olsen, D., Gotfredsen, H.-H. and Fujii, Y.: Risk Reducing Effect of the Great Belt VTS
System, Seventh International VTS Symposium,Vancouver, June 1992.
Scott, R.: Standard Ship Designs. Dry Cargo, Container and Ro-Ro Vessels, Fairplay
Publications Ltd., London, 1984.
Scott, R.: Standard Ship Design. Bulk Carriers and Tankers, Fairplay Publications
Ltd., London, 1985.
Spyrou, A.G.: The Million-Ton Tanker Fleet-Is it Really Coming?, Oil and Gas
Journal, March 4,1974.
122 Appendix C

[3-81 Williamson, G.A.: The I).pical Dimension and Operating Draughts of Bulk
Carriers, Marine TransportCenter, University of Liverpool, Dec. 1975.

-
SECTION 4 RISK ACCEPTANCE
14-11 Frandsen, A. G. and Langsoe, H.: Ship Collision Problems: I. Great Belt Bridge and
II. InternationalEnquiry,IABSE Periodica No. 2,1980.
[4-21 Havnoe, K.and Knott,M.: Risk Analysis and Protective Island Design for Ship Col-
lision, IABSE Symposium on Safety and Quality Assurance of Civil Engineering
Structures,Tokyo,Japan, 1986.
14-31 International Standardization Organization: Accidental Actions due to Human
Activities, ISO-Working Group TC 98/SC3/WG4, Working Draft, Nov. 1987.
14-41 Kampmann, J., Kieler, K.,Kohl, B. and Spangenberg, S.: Risk Analysis of the
Railway -el under the Great Belt, 1st InternationalConference on Safety in Road
and Rail Thnels, Basel, November 1992.
[4-51 Knott,M.A.: Pier Protection System for the Sunshine Skyway Bridge Replacement,
Third Annual InternationalBridge Conference, Pittsburgh, June, 1986.
[4-61 Knott, M. A. and Bonyun, D.: Ship Collision Against the Sunshine Skyway Bridge,
IABSE Colloquium on Ship Collision with Bridges and Offshore Structures, Copen-
hagen, 1983.
14-71 Knott, M.A. and Flanagan, M.: Pier Protection for the Sunshine Skyway Bridge,
IABSE Colloquium on Ship Collision with Bridges and Offshore Structures, Copen-
hagen, 1983.
[4-81 Knott, M.A. and Larsen, 0. Damgaard Guide Specification and Commentary for
Vessel CollisionDesign of Highway Bridges, US Departmentof Transportation, Fed-
eral Highway Administration, Publ. No. FHWA-RD-91-006, Dec. 1990.
14-91 Larsen, 0.Damgaard and Benmoussa,C.: ShipCollisionAspects for Bridge across the
Strait of Gibraltar. Third InternationalColloquium on the Fixed Link Europe-Africa
Through the Strait of Gibraltar, Marrakesh, May 1990.
[4-101 Leslie, J., Clark, N. and Segal, L.: Ship and Bridge Collisions-the Economics of
Risk, IABSE Colloquium on Ship Collision with Bridges and Offshore Structures,
Copenhagen, 1983.
[4-113 Maunsell and Partners and Brady, P. J. E.: Second Hobart Bridge-Risk of Ship Colli-
sion and Methods of Protection,Technical Report prejmed for Department of Main
Roads, Tasmzpia,Australia, 1978.
[4-121 Modjeski & Masters, Consulting Engineers:Criteria for: The Design of Bridge Piers
with Respect to Vessel Collision in Louisiana Waterways, Prepared for the Louisiana
Department of Transportation and Development and the Federal Highway Administra-
tion, New Orleans, July 1985.
[4- 131 National Transportation Safety Board Ramming of the Sunshine Skyway Bridge by
the Liberian Bulk Carrier Summit Venture, Tampa Bay, Florida, May 9,1980, Marine
Accident Report NTSB-MAR-81-3, Washington D.C., 1981.
 Selected Literature 123

[4-141 Nordic Committee for Structural Design: Recommendations for Loading and Safety
Regulations for Structural Design, NBK-Report No. 55, June 1987.
[4-151 Philipson, L. L.: Numerical Risk Acceptability and Mitigation Evaluation Criteria,
IABSE Colloquium on Ship Collision with Bridges and Offshore Structures, Copen-
hagen, 1983.
[4-161 Planeix, Jean-Michel: Risk-A Subjective Notion Differently Perceived, IABSE Col-
loquium on Ship Collision with Bridges and Offshore Structures, Copenhagen, 1983.
[4-171 Rowe, W. D.: AcceptableLevels of Risk for Technological Undertakings, IABSE Col-
loquium on Ship Collision with Bridges and Offshore Structures, Copenhagen, 1983.
[4-181 Sexsmith, R. G.: Bridge Risk Assessment and Protective Design for Ship Collision,
IABSE Colloquium on Ship Collision with Bridges and Offshore Structures, Copenha-
gen, 1983.
[4-191 Swedish State Accident Research Board: Investigation Report Concerning the
Ramming of Almasund Bridge by Star Clipper (in Swedish), Stockholm, 1981.
[4-201 Vincentsen, L. J. and Spangenberg, S.: Safety Management System for the Great Belt
Link, Second Symposium on Strait Crossings, Trondheim, June 1990.
[4-211 Whitman, R.: Evaluating Calculated Risk in Geotechnical Engineering, ASCE
Journal of Geotechnical Engineering, No. 2, 1984.

SECTION 5 - COLLISION RISK

Canadian Coast Guard: Vessel Traffic Services, Transport Canada, October 1984.
Cockcroft, A. N.: A Comparison of Safety Records, Journal of Navigation, Vol. 34,
1981.
Det Norske Ventas: Cause Relationships of Collisions and Groundings, Report No.
81-0097, 1981.
Fabre, F., Klose, A. and Salvarani,R.: COST 301. Shore Based Marine Navigation Aid
Systems, Directorat General Transport, EEC, 1988.
Frandsen, A. G. and Langsoe, H.: Ship Collision Problems: I. Great Belt Bridge and 11.
International Enquiry, IABSE Periodica No. 2, 1980.
Frandsen, A.G., Olsen, D., Fujii, Y. and Spangenberg, S.: Ship Collision Studies for
the Great Belt East Bridge, Denmark, IABSE Symposium on Bridges-Interaction
between Construction Technology and Design, Leningrad, 1991.
Fujii, Y., Yamanouchi, H. and Mizuki, N.: Some Factors Affecting the Frequency of
Accidents in Marine Traffic. I1 - The Probability of Stranding and I11 - The Effect of
Darkness on the Probability of Collision and Stranding,Journal of Navigation, Vol. 27,
1974.
Fujii, Y. et al.: Survey on Vessel Traffic Management Systems and Brief Introduction
to Marine Traffic Studies, Electronic Navigation Research Institute Papers (Japan),
No. 45, 1984.
Fujii, Y., Yamanouchi, H., Tanaka, Yamada, Okuyama and Hirano: The Behaviour of
Ships in Limited Water, 24th International Navigation Congress, Leningrad, 1977.
124 Appendix C

[5-lo] Fujii, Y. and Shiobara: The Analysis of Traffk Accidents, Journal of Navigation, Vol.
24, Oct. 1971.
[5-111 Fujii, Y.: Visual Range and the Degree of Risk, Joumal of Navigation, Vol. 27, April
1974.
[5-121 Fujii, Y, et al.: Survey Report on Compulsory Pilotage, Japan Association for
Preventing Marine Accidents (JAPMA), Tokyo, 1983.
[5- 131 Fujii, Y.: The Estimation of Losses Resulting from Marine Accidents, Journal of
Navigation, Vol. 3 1, May 1978.
[5-141 Heinrich, H. W.: Industrial Accident Preservation. A Scientific Approach. 1959.
[5-151 Inoue, K.: On the Separation of Traffk at Straight Waterway by Distribution Model of
Ship Paths, Journal of Nautical Society of Japan, No. 56, 1977.
[5-161 Knott, M.A. and Bonyun, D.: Ship Collision Against the Sunshine Skyway Bridge,
IABSE Colloquium on Ship Collision with Bridges and Offshore Structures, Copen-
hagen, 1983.
[5-171 Knott, M.A. and Larsen, 0. Damgaard: Guide Specification and Commentary for
Vessel Collision Design of Highway Bridges, US Department of Transportation,
Federal Highway Administration, Publ. No. FHWA-RD-91-006, Dec. 1990.
[5-181 Kuroda, K. and Kita, H.: Probabilistic Modelling of Ship Collision with Bridge Piers,
IABSE Colloquium on Ship Collision with Bridges and Offshore Structures, Copen-
hagen, 1983.
[5-191 Larsen, 0. Damgaard: Ship Collision Risk Assessment for Bridges, IABSE Collo-
quium on Ship Collision with Bridges and Offshore Structures, Copenhagen, 1983.
[5-201 Larsen, 0.Damgaard and Benmoussa, C.: Ship Collision Aspects for Bridge across the
Strait of Gibraltar, Third International Colloquium on the Fixed Link Europe-Africa
Through the Strait of Gibraltar, Marrakesh, May 1990.
[5-211 Leslie, J. A.: Ships and Bridges, 3rd International Conference on Application of
Statistics and probability in Soil and Structural Engineering, Sydney, Australia, 1979.
[5-221 Lund, H. T. and Olsen, D: Causation Probabilities and Ship Accidents in Great Belt.
To be published in 1992.
[5-231 MacDuff, T.: The Probability of Vessel Collision, Ocean Industry, Sept. 1974.
[5-241 Madsen, H. 0. et al.: Methods of Structural Safety, Prentice-Hall, 1986.
[5-251 Matsui, T., Fujii, Y. and Yamanouchi, H.: Risk and Probability of Marine Traffk
Accidents (in Japanese), Electronic Navigation Research Institute Papers, Vol. 50,
1985.
[5-261 Maunsell and Partners and Brady, P. J. E.: Second Hobart Bridge-Risk of Ship Colli-
sion and Methods of Protection, Technical Report prepared for Department of Main
Roads, Tasmania, Australia, 1978.
[5-271 Meurs, K. and Oosterbaan, J. W.: Simulation of Bridge Passage in High Winds,
IABSE Colloquium on Ship Collision with Bridges and Offshore Structures, Copenha-
gen, 1983.
[5-281 Mikkelsen, A,: Ship Collision with Danish Lighthouses, IABSE Colloquium on Ship
Collision with Bridges and Offshore Structures, Copenhagen, 1983.
 Selected Literature 125

[5-291 Mizuki, N., Yamanouchi, H. and Fujii, Y.:The Result of Third Survey on Vessel
Traffic Services in the World (Appendix V, New Data on Accident Probabilities),
Electronic Navigation Research Institute Papers, Vol. 59, 1989.
[5-301 Modjeski and Masters, Consulting Engineers: Criteria for: The Design of Bridge Piers
with Respect to Vessel Collision in Louisiana Waterways, Prepared for the Louisiana
Department of Transportation and Development and the Federal Highway Admini-
stration, New Orleans, July 1985.
[5-313 Olsen, D.F., Gotfredsen, H.-H. and Fujii, Y.:Risk Reducing Effect of the Great Belt
VTS System. 7th International VTS Symposium, Vancouver, June 1992.
[5-321 Olsen, D. F, Drejfeldt, S. R. and Olsen, N. H.: Theoretical Modelling and Application
of a Ship Collision Risk Model to Bridges crossing Navigable Waters. To be published
in 1993.
[5-331 Ostenfeld-Rosenthal, P., Lund H. T. and Drejfeldt, S. R.: The Concept of Close Ship
Encounters-A Decision Tool for Navigation Span Widths of Bridges. To be published
in 1993.
[5-341 Pyman, M. A. F., Austin, J. S. and Lyon, P. R.: ShipPlatform Collision Risk in the U.K.
Sector. IABSE Colloquium on Ship Collision with Bridges and Offshore Structures,
Copenhagen, 1983.
[5-351 Sexsmith, R. G.: Bridge Risk Assessment and Protective Design for Ship Collision,
IABSE Colloquium on Ship Collision with Bridges and Offshore Structures, Copen-
hagen, 1983.
[5-361 United States Coast Guard: Vessel Traffic Systems. Analysis of Port Needs, Depart-
ment of Transportation, Washington D.C., August 1973.
[5-371 Wasa, Y. and Oshitari, M.: Ship Collision with the Tokyo Bay Crossing Bridge-
Tunnel, IABSE Colloquium on Ship Collision with Bridges and Offshore Structures,
Copenhagen, 1983.
[5-381 Wheatley, J. H. W.: Research at NPL on Marine Traffic Systems, National Physical
Laboratory (UK), North East Coast Institute of Engineers and Ship Builders, Transac-
tions of 49th Session, Vol. 89, 1972-73.

-
SECTION 6 VESSEL IMPACT FORCES

[6-11 Akita, Y., Ando, N., Fujita, Y. and Kitamura, K.: Studies on Collision-Protective
Structures in Nuclear Powered Ships, Nuclear Engineering and Design, Vol. 19,1972.
[6-21 Amdahl, J.: Energy Absorption in Ship-Platform Inpacts, Dr. Ing. Thesis, Report No.
UR-83-34, The Norwegian Institute of Technology,Trondheim, 1983.
[6-31 Euratom: Collision Tests with Ship Models, Euratom Report Eur 4560 E, Luxemburg,
1971.
[6-41 Frandsen, A. G. and Langsoe, H.: Ship Collision Problems: 1. Great Belt Bridge and 11.
International Enquiry, IABSE Periodica No. 2,1980.
126 Appendix C

[6-51 Frandsen, A. G., Olsen, D., Fujii, Y. and Spangenberg, S.: Ship Collision Studies for
the Great Belt East Bridge, Denmark, IABSE Symposium on Bridges-Interaction
between Construction Technology and Design, Leningrad, 1991.
[6-61 French Ministry of Public Works: Public Regulations (in French). Fascicule No. 61,
Paris, 28 Dec. 1971.
16-71 Gerard, G.: The Crippling Strength of Compression Elements, Journal of Aero-
nautical Sciences, January 1958.
i6-81 German Ministry of Traffic: Safeguarding of the Piers of Rhine Bridges against
Impact Forces from Barges. Notification of 8 Aug. 1974.
16-91 Hagiwara, K.,Takanabe, H. and Kawano, H.: A Proposed Method of Predicting Ship
Collision Damage, Ship Strength Laboratory, Nagasaki Techn. Institute, Japan, 1982.
[6-101 International Standard Organization: Accidental Actions due to Human Activities,
Working Group TC 98/SC3/?VG4, Working Draft, November 1987.
[6-111 Ito, H., Kondo, K., Yoshimura, N., Kawashima, M. and Yamamoto, S.: A Simplified
Method to Analyse the Strength of Double Hulled Structures in Collision, Report no. 1,
2 and 3., J. of The Society of Naval Architects of Japan, vol. 156,1984, vol. 158, 1985,
vol. 160, 1986.
[6-121 Jensen, A.O. and Sgrensen, E.A.: Ship Collision and the Faroe Bridges, IABSE
Colloquium on Ship Collision with Bridges and Offshore Structures, Copenhagen, 1983.
[6-131 Kagami et al.: Research on Collision Resisting Construction of the Sides of Ship,
Report No. 2, Mitsubishi Nippon Industries, Technical Review 2, 1961.
[6-141 Knott, M. A. and Bonyun, D.: Ship Collision Against the Sunshine Skyway Bridge,
IABSE Colloquium on Ship Collision with Bridges and Offshore Structures, Copen-
hagen, 1983.
[6-151 Knott, M.A. and Larsen, 0. Damgaard: Guide Specification and Commentary for
Vessel Collision Design of Highway Bridges, US Department of Transportation,
Federal Highway Administration, Publ. No. FHWA-RD-91-006, Dec. 1990.
[6-161 Kuesel, T. R.: Newport Bridge Collision, IABSE Colloquium on Ship Collision
with Bridges and Offshore Structures, Copenhagen, 1983.
[6-171 Larsen, 0.Damgaard and Benmoussa, C.: Ship Collision Aspects for Bridge across the
Strait of Gibraltar, Third International Colloquium on the Fixed Link Europe-Africa
Through the Strait of Gibraltar, Marrakesh, May 1990.
[6-181 Marsh, K.J. and Campbell, J. D.: The Effect of Strain Rate on the Post-Yield Flow of
Mild Steel, Journal of Mechanics and Physics of Solids, 11,49, 1963.
[6-191 Meir-Dijmberg, K. E.: Ship Collisions, Safety Zones and Loading Assumptions for
Structures on Inland Waterways (in German), VDI-Berichte No. 496,1983.
[6-201 Minorsky, V.U.: An Analysis of Ship Collisions with Reference to Protection of
Nuclear Power Plants, Journal of Ship Research, Oct. 1959.
[6-211 Modjeski & Masters, Consulting Engineers: Criteria for: The Design of Bridge Piers
with Respect to Vessel Collision in Louisiana Waterways, Prepared for the Louisiana
Department of Transportation and Development and the Federal Highway Administra-
tion,New Orleans, July 1985.
 Selected Literature 127

[6-221 Nagasawa, H., Arita, K., Tani, M. and Oka, S.: A Study on the Collapse of Ship
Structure in Collision with Bridge Piers, Naval Architecture and Ocean Engineering,
Society of Naval Architects, Japan, Vol. 19, 1981.
[6-231 Nordic Road Engineering Federation: Load Regulations for Road Bridges (in Nor-
wegian), NVF, Report No. 4,1980.
[6-241 Norwegian Public Roads Administration: Load Regulations for Bridges and Ferry
Ramps in the Public Road System (in Norwegian), Preliminary Edition, Olso, 1986.
[6-251 Ohnishi, T., Kawakami, H., Yasukawa, W. and Nagasawa, H.: Ultimate Strength of
Bow Construction, IABSE Colloquium on Ship Collision with Bridges and Offshore
Structures, Copenhagen, 1983.
[6-261 Olnhausen,W. Von: Ship Collision with Bridge Piers (in Swedish), Teknisk Tidskrift,
No. 17, Stockholm, 1966.
[6-271 Ostenfeld, Chr.: Ship Collisions Against Bridge Piers, IABSE Publications, Vol. 25,
1965.
[6-281 Pedersen, P.T, Valsgilrd, S., Olsen, D. and Spangenberg, S.: Ship Impacts-Bow
Collisions. 3rd International Symposium on Structural Crashworthiness and Failure,
Liverpool, April 1993.
[6-291 Prucz, Z. and Conway, W. B.: Ship Collision with Bridge Piers-Dynamic Effects,
Transportation Research Board 69th Annual Meeting, Washington D.C., 1990.
[6-301 Reckling, K. A.: Beitrag der Elasto- und Plastomechanik zur Untersuchung von
Schiffskollisionen,Jahrbuch der Schiffbautechnischen Gesellschaft, Band 7 0, 1976.
[6-311 Requena, L. F.: DBtemination dynamique de la charge transitoire, lors dune colli-
sion. IABSE Colloquium on Ship Collision with Bridges and Offshore Structures.
Copenhagen, 1983.
[6-321 Saul, R. and Svensson,H.: Means of Reducing Consequencesof Ship Collisions with
Bridges and Offshore Structures, IABSE Colloquium on Ship Collision with Bridges
and Offshore Structures, Copenhagen, 1983.
[6-331 Saul, R. and Svensson, H.: On the Theory of Ship Collision against Bridge Piers,
IABSE Proceedings, No. 2,1982.
[6-341 Saul, R. and Svensson, H.: On Protection of Bridge Piers against Ship Collision (in
German), Die Bautechnik, Vol. 58, 1981.
[6-351 Teyssandier, J.-P. et al.: Impact Protection from Ships in the Euroroute Project,
IABSE Symposium on Concrete Structures for the Future, Paris-Versailles, 1987.
[6-361 Wierzbicki, T.: Crushing Behaviour of Plate Intersections, Structural Crashworthi-
ness, edited by N. Jones and T. Wierzbicki, Chapter 3, Butterworth & Co. (Publishers)
Ltd., 1983.
[6-371 Woisin, G. and Gerlach, W.: On the Estimation of Forces Developed in Collisions be-
tween Ships and Offshore Lighthouses, 8. International Conference On Lighthouses
and Other Aids to Navigation, Stockholm, 1970.
[6-381 Woisin, G.: Design Against Collision, International Symposium on Advances in
Marine Technology, Trondheim, 1979.
[6-391 Woisin, G.: Die Kollisionsversuche der GKSS, Schiff und Hafen, Heft 2, 1977.
128 Appendix C

[6-401 Yang, P. D. C. and Caldwell, J. B.: Collision Energy Absorption of Ships Bow Struc-
tures, International Journal of Impact Engineering, No. 2, 1988.

-
SECTION 7 BRIDGE DESIGN

Blok, J. J. and Dekker, J. N.: On Hydrodynamic Aspects of Ships Colliding with Fixed
Structures, IABSE Colloquium on Ship Collision with Bridges and Offshore Struc-
tures, Copenhagen, 1983.
Fauchart, J.: Choc de bateau sur obstacle deformable, IABSE Colloquium on Ship
Collision with Bridges and Offshore Structures, Copenhagen, 1983.
Fauchart, J.: Consequences of a Ship Collision with the Verdon Bridge, IABSE Col-
loquium on Ship Collision with Bridges and Offshore Structures, Copenhagen, 1983.
Knott, M. A. and Larsen, 0.Damgaard: Guide Specificationand Commentary for Ves-
sel Collision Design of Highway Bridges, US Department of Transportation,Federal
Highway Administration, Publ. No. FHWA-RD-91-006, Dec. 1990.
Modjeski & Masters, Consulting Engineers: Criteria for: The Design of Bridge Piers
with Respect to Vessel Collision in Louisiana Waterways, Prepared for the Louisiana
Department of Transportation and Development and the Federal Highway Administra-
tion, New Orleans, July 1985.
Nordic Road Engineering Federation: Load Regulations for Road Bridges (in
Norwegian), NVF, Report No. 4,1980.
Petersen, M. J. and Pedersen, P. Terndrup: Collision between Ships and Offshore Plat-
forms, 13th Annual Offshore Technology Conference, Paper No. OTC 4134,1981.
Prucz, Z. and Conway, W. B.: Ship Collision with Bridge Piers-Dynamic Effects,
Annual Meeting of Transportation Research Board of the National Research Council,
Washington D.C., 1990.
Saul, R. and Svensson, H.: On the Theory of Ship Collision against Bridge Piers,
IABSE Proceedings, No. 2, 1982.

-
SECTION 8 PREVENTION MEASURES

[8-11 Fabre, F., Klose, A. and Salvarani, R.: COST 301. Shore Based Marine Navigation Aid
Systems, Directorate General Transport. EEC, 1988.
[8-21 Fujii, Y. et al.: Survey on Vessel Traffic Management Systems and Brief Introduction
to Marine Traffic Studies. Electronic Navigation Research Institute Papers (Japan) No.
45, 1984.
[8-31 Greneker, E.F., Eaves, J.L. and McGee, M.C.: Bridge Ship Collision Electronic
Detection and Early Warning, IABSE Colloquium on Ship Collision with Bridges and
Offshore Structures, Copenhagen, 1983.
[8-41 Hara, Kiyoshi: Progress of VTS and its Studies in Japan, International Symposium on
Vessel Tr&ic Service, Gothenburg, Sweden, May 1988.
 Selected Literature 129

18-51 The IALA, IAPH, IMPA World VTS Guide, Pergamon Press, 1989.
18-61 International Maritime Organization (IMO): Guidelines for Vessel Traffk Services,
Resolution AS78 (14) adopted on 20 November 1985. IMO, 1985.
18-71 International Maritime Organization (IMO): International Conference on Revision of
the International Regulations for Preventing Collision at Sea, 1972. IMO, 1990.
l3-81 Knott, M.A. and Flanagan, M.: Pier Protection for the Sunshine Skyway Bridge,
IABSE Colloquium on Ship Collision with Bridges and Offshore Structures,
Copenhagen, 1983.
18-91 Mizuki, N., Yamanouchi, H. and Fujii, Y.: The Result of Third Survey on Vessel
Traffk Servicesin the World. Electronic Navigation Research Institute Papers, Vol. 59,
1989.
[8-101 Olsen, D. F., Gotfredsen, H.-H. and Fujii, Y.: Risk Reducing Effect of the Great Belt
VTS System, Seventh International VTS Symposium, Vancouver, Canada, June 1992.
18-11] US Coast Guard: Vessel Traffic Services, Equipment and Technology Report (VTS
Handbook), Washington D.C., May 1987.
18-12] Vendrell, 5.: Minimizing the Risk with Vessel Traffk Management Systems, IABSE
Colloquium on Ship Collision with Bridges and Offshore Structures, Copenhagen,
1983.
[8-131 VTMS for Worlds Longest Bridge, World Port Construction & Ocean Technology,
ApriljMay 1991.

-
SECTION 9 PROTECTION MEASURES

19-11 Brink-Kjaer, O., Brodersen, F.P. & Nielsen, A.H.: Modelling of Ship Collisions
Against Protected Structures,IABSE Colloquium on Ship Collision with Bridges and
Offshore Structures, Copenhagen, 1983.
19-21 Denver, H.: Geotechnical Model Tests for the Design of Protective Islands, IABSE
Colloquium on Ship Collision with Bridges and Offshore Structures, Copenhagen,
1983.
19-31 Englot, J.P.: Collision Protection of Arthur Kill Bridges, New York ASCE Section,
Structures Conference, A X E , New York, May 1988.
19-41 Fletcher, M. S., May, R. W. P. and Perkins, J. A.: Pier Protection by Man-Made Islands
for Orwell Bridges, U.K., IABSE Colloquium on Ship Collision with Bridges and
Offshore Structures, Copenhagen, 1983.
19-51 Frandsen, A. G. and Langsoe, H.: Ship Collision Problems: I. Great Belt Bridge and 11.
International Enquiry, IABSE Periodica No. 2, 1980.
19-61 Hahn, D. M. and Rama, H. E.: Cofferdams Protecting New York Bridges from Ship
Collisions, Civil Engineering-ASCE, February 1982.
19-71 Havnoe, K. and Knott, M.: Risk Analysis and Protective Island Design for Ship
Collision. IABSE Symposium on Safety and Quality Assurance of Civil Engineering
Structures, Tokyo, 1986.
130 Appendix C

[9-81 Heins, C.P.: Bridge Dolphins Subjected to Impact, IABSE Colloquium on Ship
Collision with Bridges and Offshore Structures, Copenhagen, 1983.
19-91 Heins, C.P. and Chiu, L.Y.-B.: Dynamic Analysis of Dolphins subjected to Ship
Impact, Computers and Structures, vo1.15, no. 1, 1982.
[9-101 Knott, M.A. and Larsen, 0. Damgaard: Guide Specification and Commentary for
Vessel Collision Design of Highway Bridge, US Department of Transportation,
Federal Highway Administration, Publ. No. FHWA-RD-91-006, Dec. 1990.
[9-111 Knott, M.: Pier Protection for the Sunshine Skyway Bridge Replacement, Third
Annual International Bridge Conference, Pittsburg, June 1986.
[9-121 Larsen, 0.Damgaard and Benmoussa, C.: Ship Collision Aspects for Bridge across the
Strait of Gibraltar, Third International Colloquium on the Fixed Link Europe-Africa
Through the Strait of Gibraltar, Marrakesh, May 1990.
[9-131 Luong, M. P.: Modkles reduits de Protection de pile de pont. IABSE Colloquium on
Ship Collision with Bridges and Offshore Structures, Copenhagen, 1983.
[9-141 Minorsky, V. U.: Evaluation of Ship Bridge Pier Impact and of Islands as Protection,
IABSE Colloquium on Ship Collision with Bridges and Offshore Structures, Copenha-
gen, 1983.
[9-151 Matsuzaki, Y. and Jin, H.: Design Specification of Buffer Structure, IABSE Collo-
quium on Ship Collision with Bridges and Offshore Structures, Copenhagen, 1983.
[9-161 Mondorf, P. E.: Floating Pier Protections Anchored by Prestressing Tendons, IABSE
Colloquium on Ship Collision with Bridges and Offshore Structures, Copenhagen,
1983.
[9-171 Namita, Y. and Nakanishi, H.: Analysis of Framed Buffer Structure around Bridge
Pier, IABSE Colloquium on Ship Collision with Bridges and Offshore Structures,
Copenhagen, 1983.
[9-181 Naoi, Y. and Ishikawa, T.:Case Stories of Dolphin Accidents and Remedies, IABSE
Colloquium on Ship Collision with Bridges and Offshore Structures, Copenhagen,
1983.
[9-191 Oda, K. and Kubo, S.: Collision Prevention Device of Floating Guideline q p e ,
IABSE Colloquium on Ship Collision with Bridges and Offshore Structures, Copen-
hagen, 1983.
[9-201 Ostenfeld, Chr.: Ship Collisions Against Bridge Piers, IABSE Publications, Vol. 25,
1965.
[9-211 Saul, R. and Svensson, H.: Means of Reducing Consequences of Ship Collisions with
Bridges and Offshore Structures, IABSE Colloquium on Ship Collision with Bridges
and Offshore Structures, Copenhagen, 1983.
[9-221 Saul, R. and Svensson, H.: On the Theory of Ship Collision against Bridge Piers,
IABSE Proceedings, No. 2,1982.
[9-231 Saul, R. and Svensson, H.: On Protection of Bridge Piers against Ship Collision
(in German), Die Bautechnik, Vol. 58, 1981.
[9-241 Sexsmith, R. G.: Bridge Risk Assessment and Protection Design for Ship Collision,
IABSE Colloquium on Ship Collision with Bridges and Offshore Structures, Copen-
hagen, 1983.
 Selected Literature 131

[9-251 Tambs-Lyche, P.: Vulnerability of Norwegian Bridges across Channels, IABSE

Colloquium on Ship Collision with Bridges and Offshore Structures, Copenhagen,
1983.
[9-261 Vitalis, A.: Ecran de protection des ouvrages en mer contre les collisions. IABSE Col-
loquium on Ship Collision with Bridges and Offshore Structures, Copenhagen 1983.

-
SECTION 10 PROTECTION OF THE PUBLIC

[ 10- 11 Greneker, E.F., Eaves, J.L. and McGee, M.C.: Bridge Ship Collison Electronic
Detection and Early Warning, IABSE Colloquium on Ship Collision with Bridges and
Offshore Structures, Copenhagen, 1983.
[ 10-21 Knott, M.A. and Larsen, 0. Damgaard: Guide Specification and Commentary for
Vessel Collision Design of Highway Bridges, US Department of Transportation,
Federal Highway Administration, Publ. No. FHWA-RD-91-006, Dec. 1990.
[ 10-31 Transportation Research Board: Considerations in the Development of an Early
Warning VesselBridge Collision System, Conference Proceedings, Washington D.C.,
1978.