Sei sulla pagina 1di 208






International Maritime Organization

Model Course No: 3.11

Marine Accident and Incident

Training Manual

First published in 1988by the

4 Albert Embankment, London SE1 7SR

Revised edition 2000

Printed by Ashford Open Learning Ltd

2 4 6 8 1 0 9 7 5 3 1

ISBN 92-801-5097-9

Sales number T-311CE

Copyright O IMO 2000

All rights reserved
No part of this pubkai'hn ma4 for sales purposes,
be produced stored in a retrieval system or transmitted
in any form or by any means, electronic, electrostaht/i7
magnetic tape, mechanical, photocopying or otherwise,
without prior permission in wr2ng from the
lnternathnal Marihme Organ~zahbn.

Part 1

Course Introduction

Part 2

International Shipping
Public Perceptions

Part 3


Part 4

International Conventions, Requirements, etc.

Part 5

Setting Up an Investigation

Part 6


Part 7

Elements of a Casualty

Part 8

Human Factors

Part 9

Collecting Evidence

Part 10

Determining Occurrence Sequence,

Analysing Evidence

Part 11


- Shipping Casualties and

These notes have been compiled as a reference for this course and also as a working
document for use when faced with the need to investigate. -The thrust of the course is directed
towards achieving a system of investigation aimed at establishing the circumstances and
causes of a marine casualty, to learn and promulgate the safety lessons. The course does
not deal with blame or liability and only briefly refers to more formal court or inquiry
The course is centred on the International Maritime Organization Assembly Resolution
A.849(20), the Code for ,the Investigation of Marine Casualties and Incidents and
amendments to the Code as annexed to ResolutionA.884(21). 'The course also will stress the
need for reporting to IMO under the provisions of MSC Circ. 827lMEPC Circ. 333.
No two accidents are ever precisely the same. However, the course sets out to show that
while the causal factors may be diverse, the underlying generic structure of any casualty or
accident, whether marine, road, rail or aviation, can be seen to have a common structure.
Much of the course is centred on role playing and case studies. Participants should use their
experience and judgement to make the roles as realistic as possible; by so doing participants
should complete the course with a clear concept of systemic investigations and their role in
improving the safety of life at sea and protecting the marine environment.

Course lntroduction



Annexed to IMO Resolution A.847(20) are guidelines to provide flag States with a means to
establish and maintain measures for the effective application and enforcement of IMO
Conventions. Part 7 of the annex refers to flag State investigations.


In addition to providing qualified surveyors, the flag State should provide qualified
investigators. Consistent with article 94.6 and articles 217.4,5 and 6 of UNCLOS and
with the provisions of the relevant IMO conventions, investigations should be carried
out following a marine casualty or pollution incident. The flag State should ensure that
individual investigators have a working knowledge and practical experience in those
subject areas pertaining to their normal duties. Additionally, to assist individual
investigators in performing duties outside their normal assignments, the flag State
should ensure ready access to expertise in the following areas, as necessary:
navigation and the Collision Regulations
flag State regulations on certificates of competency
causes of marine pollution
interview techniques
evidence gathering
evaluation of the effects of the human element
Any accident involving personal injury necessitating absence from duty of three days
or more and any deaths resulting from occupational accidents and casualties to ships
of the flag State should be investigated, and the results of such investigations made
public. Ship casualties should be investigated and reported on in accordance with
UNCLOS, relevant IMO Conventions, and the Guidelines currently being developed
by IMO. Casualty investigations should be conducted by suitably qualified
investigators, competent in matters relating to the casualty. The report of the
investigation should be forwarded to IMO together with the flag State's observations,
in accordance with the Guidelines referred to above."

It is important, therefore that flag and coastal State administrations have in place ameffective
and internationally acceptable system for investigating marine accidents.


The purpose of the course

The purpose of this course is to introduce you to the philosophy, processes and procedures
required to support a marine casualty investigation in accordance with IMO Assembly
Resolution A.849(20) and the Code for the Investigation of Marine Casualties and Incidents.
'The Code seeks to introduce to the international shipping community uniform objectives and
procedures for investigating casualties which occur in this most international of industries.
The international aviation industry has subscribed to such a code (Annex 13 of the
International Civil Aviation Organization Convention) for some years and the systems
approach into air accidents is credited with contributing to the safety of the civil aviation
industry. The IMO Code is aimed at a safety outcome to identify the circumstances under
which a casualty occurred and to determine the causes of such an accident.



This course is designed to introduce potential flag State investigators with an introduction to
accident investigations and accident investigation methodology. The course will be focused
on the IFAO Code. 'The course also provide administrators with an insight into what is required
in conducting a casualty investigation, supporting the investigator in the field and what an
investigation should achieve.
Investigations are an "official function of a flag or coastal State". When investigating in your
own country your own national laws will apply. You will have to bear these laws in mind when
dealing with certain parts of the course, this will apply particularly to Part 5.


Course structure

The course is divided into eleven learning objectives or "parts", as outlined in the program, It
will be centred on group activities and lecture sessions.

Recommended reading and references

Reason, J. (1990) Human Error, Cambridge University Press, Cambridge (UK).

ISBN 0 521 31419 4
Reason, J. (1997). Managing the Risks of Organizational Accidents, As hgate,
ISBN 1 84014 105 0
Reason, J. Corporate Culture and Safety NTSB Symposium on Corporate Culture and
Transportation Safety, Washington, 1997
Moore-Ede, M., The Twenty-Four Hour ~ o c i e t ' Random House Australia, 1993
ISBN 0 09 182755 8
Perrow, C., Normal Accidents - Living with HkJh-Risk Gchnologies, Basic Books, Inc. New
Elizabeth F Loftus, Eyewitness Gstimonj Harvard University Press, 1996.)
Rushbrook.R., Rushbrook's Fire Aboard, Third Edition, 1998, Brown, Son & Ferguson
ISBN 0 85174 659 4
DeHaan J.D., Kirk's -Fire InvestkJation, Fourth Edition, Prentice Hall Inc, 1997
OILeary, M. & Chappell, S. Confidential incident reporting systems create a vital awareness
of problems. ICAO Journal: 51 :11-13, 1996
Ainsworth et al, (1996) Proceedings of a Research Workshop on Fatigue in the Maritime
Industry, Seafarers International Research Cer~trefor Safety and Occupational Health,
University of Wales, Cardiff, Executive Summary.
Colquhoun, W. P., Rutenfranz, J., Goethe, H., Neidhart, B., Condon, R., Plett, R. & Knauth,
P. (1988). Work at sea: a study of sleep, and of circadian rhythms in physiological functions,
in watchkeepers on merchant vessels. I.Watchkeeping on board ships: a methodological
approach. InternationalArchives of Occupational and Environmental Health, 60: 321-329.
Siros, W. G., (1995), Aledness Assurance - The key to reducing Fatigue and Human Error in
the Mahe Industa American Petroleum Institute/US Coast Guard

Guide to Interviewing, Metropolitan Police UK


International Shipping - Shipping Casualties and

Public Perceptions


Marine accidents as seen by the public and the media

Shipping was once seen as the safest and most environmentally friendly way to transport
goods by sea. In any analysis it is the only practical way for the vast bulk of the world's trade.
Shipping accidents, particularly those that involve pollution, bring into question the safety of
shipping and the quality of ships and their crews. Whether or not such questions are justified,
it is shipping accidents that provide an image of the industry, which attracts considerable
(Video "Scandal at sea" or equivalent. Approx 30 minutes)


Shipping safety issues

What is risk? 'The risk is the probability of event multiplied by the exposure times the
Is there such a thing as an acceptable accident rate?

The need is to set the risk as low as possible and at the same time make the operation
practical. It is possible to design all manner of engineering solutions to minimise risk. The idea
of "redundancy" is one that is widely used. If one has three systems operating in parallel each
with a possible failure rate of 1:100, the likelihood of all three failing at once is 1:1,000,000.
Mechanical and electrical systems are increasingly reliable and with redundancy the chances
of a casualty through equipment failure is minimal.
It is often stated that human error is the cause of about 80 per cent of accidents. This course
will examine this statement and introduce some basic consideration of human factors. It will
show that all accidents bar "acts of God" have some human element in the chain of events
leading to the accident.
The aim of marine casualty investigations is to prevent similar casualties. Open, systemic and
fair investigations, which are not aimed at apportioning liability or blame, have proved
effective in the aviation industry. Such an international approach in the marine world may
have helped prevented unacceptable losses at sea. This col-lrse aims to introduce
participants to a system of investigation, which may help to prevent future casualties.
Most casualties involve the smaller range of general cargo ships and fishing vessels. The two
types of ship that have caught the public's attention are bulk carriers and passenger ferries.
Bulk carrier losses
The loss of these ships reflected a worldwide trend. According to Lloyd's Register of Shipping,
(J M Ferguson, Bulk Carriers - The Challenge, 14 May 1991) between 1980 and 1991, 64
bulk carriers were lost under circumstances where structural failure may have been a factor.
At the peak, 20 bulk carriers of over 15,000 gross tonnage were lost in a 20 month period,
apparently as a result of some failure in their str~~cture,
between December 1990 and August
1991, with the loss of hundreds of seafarers.


Bulk carriers make up about 15 per cent of the world fleet in numbers of ships over 1000
deadweight tonnes. The table below shows the number of bulk carriers (and OBOs) lost
through having foundering or having been posted as missing compared to all vessels
(excluding fishing vessels, tugs and dredgers) that have been reported as having foundered
or as missing. The column of lives lost shows all lives lost on bulk carriersl.
The following table, from Lloyd's Casualty Returns and Lloyd's Annual year Books, shows
that the safety of bl.llk carriers remains an important issue. The following table is focused on
bulk carriers that sank or suffered known structural damage. It does not include bulk carriers
(or bulk carrier deaths) ,from fires, explosions etc.

All losses
500 grt


Lives lost


All losses
500 grt


Lives lost










*Note: Lloyd's Casualty Returns changed the format in which they showed figures for loss of
life. These figures are open to revision as reports are received.
Anna Sp/iatusank after collision with loss of 26 lives, on 15 June 1996. ICL Wraman
sank after collision with loss of 24 lives, on 26 Sept 1997
The loss of bulk carriers between 1987 and 1994 led to an initiative by the Secretary General
at MSC63, following which a six person correspondence group of competent persons
addressed the issues of bulk carrier safety. This group reported to MSC, which drafted a new
chapter to SOLAS 74. 'This Chapter was the subject of a Diplomatic conference in November
1997, which adopted the new chapter 12 to the SOLAS Convention on additional safety .
measures for bulk carriers.
Passenger ferry losses

Ferries account for about 5.8 per cent of the total number of the world fleet of ships over 1000
tonnes deadweight.
Ferry accidents, particularly involving loss of life, are major incidents requiring the most
rigorous investigation. Prior to 1989 a number of ferry casualties created considerable
concern. These included:
Dona Paz involved in collision with the tanker Vector south-east of Manila on
20/12/1990, the death toll has been put at 4,386, though the official figure is lower at
about 3148.
Herald o f Free Enterprisecapsized and foundered after leaving Zeebrugge on 3/3/97
with the loss of 193 lives.
Dona Mariilyncapsized in a typhoon in the Philippines on 24/10/88 with the loss of as
many as 350 lives.
'Source Lloyd's Casualty Return


Since 1989 the losses have continued:

Salem Express on passage Jeddah-Suez struck a reef and sank on 14112/91, 464
Jan Hewehbsz- capsized in the Baltic, 14/1/93, 55 deaths;
Estonk - capsized in the Baltic 28/9/94, 852 deaths;
Cebu C/'ty- sank after a collision in Maila Bay on 2112/94 with the loss of 140 lives;
Gurita, stranded and sank on 19/1/96, Banda Archipelago (Indonesia) with loss of 338
Bukoba capsized on Lake Victoria in May 1996 with loss of 869 lives;
Phcess of Orient- sank in tropical storm Vicki on 21/8/98, at least 150 deaths.
Harla Rimba - sank in central Indonesian Archipelago on 612 99 with loss of approx.
300 lives.
The loss of the Eston!b and the ferry tragedies over the previous years prompted the
Secretary-Generaland the Chairman of the Maritime Safety Committee to seek approval of
the 18th Assembly for the formation of a group of experts to consider the issues of ro-ro
safety. This group met on 8 occasions resulting in a Diplomatic Conference in 1995, which
approved 30 amendments to SOLAS 74 dealing with ro-ro safety.


Accidents as an iceberg

Action ratio study

What can we learn from accident investigations and analysis? Which incidents should be
investigated? Quite obviously no marine authority has the resources to investigate all minor accidents and
near misses (sometimes referred to as "incidents"). However, in an ideal world we would look
at even the most minor incident.
Accidents can be seen as the part of an iceberg above the sea surface. Beneath the surface
there are a myriad of minor accidents and near misses which are largely unreported and
whether they are investigated or not may rely on the company safety culture and ethic.
A study of industrial accidents undertaken by Frank E Bird, Jr. and George L Germain (1.969)*
analysed 1,753,498 accidents reported by 297 companies from 21 industrial groups over
covering a period of 3 billion man hours.
In comparing the severity of accidents they discovered that for every reported major injury
(death, disability, lost time or medical treatment) there were 9.8 minor injuries requiring first

Bird F. and Germain G., (1986) PracticalLoss Control Leadership Institute Publishing, Loganville, Georgia.


For each major injury there were 30.2 property damage accidents and 600 incidents.

property damage accidents

incidents with no visible damage

Bird and Germain pointed out that to prevent the major accident occurring it wo1.11dbe more
productive to attack the lesser incidents.
North of England P&l Club
A similar model has been developed by the North of England P&l Club with the following
illustration that appeared in Lloyd's List in February 1998.
Using a similar concept of a triangle they arrive at the following:

Man climbs ladder

Ladder moves

Man killed
Man climbs ladder
Ladder moves

Man breaks leg

Man climbs ladder
Ladder moves

Man injured
Ladder against
bulkhead unlashed and

The message seems to be that for every major accident, those usually investigated by
government safety authorities, there are many other minor or potential accidents. Also, the
severity of accidents can depend on the degree of hazard. The ladder illustrates the point that
in the same situational factors, i.e. an unlashed ladder, the outcomes can vary. If we can learn
lessons from all accidents and incidents we can reduce the base of the triangle thereby
reducing the chances of major accidents.



A marine an accident. It is an unexpected or unlooked for event with unwelcome

outcomes. The definition is contained in the Code at paragraph 4.
Casualty incident categories are normally grouped as follows:l


Includes ships which sink as a result of heavy weather, springing leaks,

breaking in two, etc., but not as a consequence of categories listed


After a reasonable period of time, no news having been received from

a ship and its fate is undetermined, the ship is posted as missing.


Includes ships experiencing a fire andlor explosion where it is the first

event reported - it therefore follows that casualties where a fire andlor
explosion is the result of a some other incident category, such an
incident is categorised under that incident.


The result of one ship being struck by another, regardless of being

whether under way, anchored or moored.


The result of a ship striking an external object - but not another ship or
the sea bottom. This includes striking a drilling rigslplatforms.


Includes a ship touching the sea bottom, sandbanks, shoals,

sandbanks, seashore, etc., as well as being entangled on underwater


Failure of equipment or machinery, immobilisation of the main engine.

Heavy weather
Ice damage

Significant damage caused by heavy seas or the elements and ice

Any casualty that is not included above and may include war damage
(this being a deliberate act may not be classified as a casualty).

These classifications relate to casualties to ships and do not include accidents to crew
members or others who may be involved in the operation of a slrip, either at sea or in port.
Death and injury to seafarers and others as a result of the operation of a ship are important,
if not the most important, part of an investigator's work. However, a distinction may have to
be made between those deaths or injuries that are not directly connected with ship
operations. Guidance in this matter is best taken from MSCICirc.8271 MEPCICirc.333 which
defines a very serious casualty as "casualties to ships which involve total loss of the ship, loss
of life, or severe pollution". Hence the death (or serious injury or near miss) must also involve
some aspect of a ship's operation. Heart attacks, or death or disability through a pre-existing
medical condition would not normally be covered by normal marine casualty or incident
investigation procedures.

IAdapted from Lloyds World Casualty Statistics




Courts, tribunals and inquiries

Authorities may react to an accident in a number of ways. At one end of the spectrum a
criminal investigation may be mounted with a view to prosecution, at the other end a pure
"safety investigation", with no blame attached, may be undertaken. Both extremes are
legitimate and are a necessary part of the armoury to promote safer operation of any
There is no reason why two such investigations should not be carried out in parallel.
3.2.1 Criminal sanctions
If the primary objective is to punish and bring the perpetrator to "justice" then a criminal
investigation, under the rules of the State in which the investigation is being conducted, may
be appropriate. In such a case a person or persons must have committed an offence. An
offence is an act of neglect or default contrary to law. To be guilty of an offence a person(s)
must have committed a guilty act (actus reus), or have a guilty intent (mens rea), or have
committed a statutory offence.
An offence suggests a deliberate intent to act contrary to law, though it can also include the
neglect of a person to act competently. The public interest in punishing a person that may
have contributed to an accident (or casualty) has to be weighed against the public interest of
understanding why an accident occurred and in helping to prevent a similar %accident.
happening again. (See following Lloyd's List leader of 11 Feb 1998 [3.2.5.])
There is no reason why two such investigations should not be carried out in parallel, but they
should be separate.
3.2.2 Coroners
Under some legal systems a Coroner has the responsibility of establishing the cause of
death. When investigating a fatality, or where a fatality is involved, an investigator may
become involved with the Coronial system. The various responsibilities in the case of a
marine casualty involving death are a matter of national law. In some countries with the
coronial system, under certain conditions (e.g Australia) there may be dual jurisdiction. It is
advisable to make early contact with the Coroner's office to work out the various interests and
to offer expert assistance.
3.2.3 Civil proceedings
Civil courts hear and pass judgement on claims for damages or other loss. The casualty
investigation is often seen as a source of useful information. It is a matter for municipal
(national law) as to what is or is not admissible as evidence.
Some reports may be considered "hearsay". Some jurisdictions require evidence to be sworn.
Others may require the investigation .files to be made available. It is also possible that an
investigator may be called as a witness.
Of great importance is the relationship between the investigation team, and other parties to
the investigation (e.g. the ship owner, equipment company, master, officers) and a party's
legal advisers. It is not uncommon for a party to try and channel everything through the legal
adviser, and for the legal adviser to dispute the investigator's right to certain information and
control any information provided. Usually the primary concern of the legal adviser is the use
that such information can be put to in civil proceedings.
A safety investigation can be a rich source of information.

In 1948 Mr Justice Streeton said:

'7 have come to the conclusion that these ingu/hiesare, and ine vitably will be, misused
by the padies to them, excepting the Director of Nav/gation, for the purpose of
establishing evidence for use in civil figation betweenthe padies. That misuse is, if 1
may be permifled a paradox, a legitimate misuse, but it is none the less a misuse.'2
3.2.4 Formal safety investigations
Formal investigations are normally conducted before a judge or tribunal independent of the
marine administration. The maritime authority is just another party presenting evidence to the

Marine casualty investigators have an important part to play in the formal investigation
process. The investigator may be required to prepare a report setting out the administration's
position and conclusions. This will involve producing evidence in a form acceptable to the
During the inquiry the investigator may be required to produce a brief of questions that must
be put to the witnesses. In the hearing helshe may be required to listen to the answers and
ensure that the questions have been answered to the satisfaction of the administration. Any
issues arising from a witness's answers may have to be followed up.
The investigator may be called as a witness to recount what helshe actually saw and
observed. Unless directed an investigator should not normally recount hearsay evidence -or
offer opinions. Such a witness should stick to facts unless specifically directed otherwise by
the inquiry.
When the evidence is complete the investigator may be required to draft a final submission
on behalf of the administration. Such a submission should be backed with references to
evidence before the inquiry.
3.2.5 Viewpoint
By Michae/ Grey
When there is failure -who knows about it?3

ONE of the most harrowing documents I have read for many years is the official report into
the loss of the Estonia' which was finally made public just before Christmas, more than three
years after the event.
It contains much in it that will appal, from the terrifying incident itself, to the inability of well
equipped ships on the spot to rescue people from the cold waters, and the nightmare which
must have been experienced by the divers as the pushed their way into the collapsed
wreckage of the ferry's internals.
I know there is much in the report that remains contentious, and the subject of further
litigation, but one specific and recurring problem impressed itself upon me, reinforcing a long
held belief that something, which the industry itself has the power to fix, needs to be urgently
I refer to the fact that by the time of this disaster, there had been a considerable number of
incidents involving failure or part-failure of bow doors with ships at sea.
2 Streeton, J.

re "Koonda" & "George H Emircocos" and re "Iron Monach" & "Empire Strength", CMI 1943.
Lloyd's List, Wednesday 11 February 1998; this leader is reproduced by kind permission of Lloyd's List.


Those who drafted the report refer to only those which took place in the North Sea or Baltic,
but it is surely reasonable to suppose that there have been others over the years, since
administrations and classification societies judged that a drive through ship with a large bow
opening was a safe option.
The Estonia report admits that the list is by no means complete, as it contains only Finnish
and Swedish vessels.
Some of these incidents, and they stretch back to the early 1970s, were very serious and
potentially lethal, although it appears that prompt action by those in charge of the ships
slowing or stopping, turning the vessels out of the weather and returning immediately to the
safety of port prevented water getting onto the cardecks.
Additionally the availability of a watertight door inside the visor or clamshell doors provided
the necessary level of insulation and saved the day.
But the point is, and the question must be asked, just how widely knowledge about these
these accidents was promulgated by those who had knowledge of them?
The Finnish and Swedish authorities would have been in possession of the full facts, and it
is reasonable to suppose that those operating ships under these flags would have been made
aware of the incidents.
A number of the major classification societies were also clearly informed about accidents
which had taken place aboard ships they had responsibility for. Presumably the individual
owners, after an accident, made very sure that they closely inspected the fastenings and
hinges of bow doors in other vessels of their fleets.
But who else learned about such accidents throughout the world, bearing in mind that drivethrough ships are a more or less universal ferry type of today?
Were naval architects working on designs for big ferries ordered by Japanese operators
made aware of the problems that had been experienced?
Were owners of ex-Baltic boats working in the Eastern Mediterranean or Far East ever told
about the problems that were experienced by the operators of the Fin/anda or Viking Saga,
or the fright that the watchkeeper of the We//amo received when he saw the bow visor lifting
as he ran down from Helsinki to Stockholm in a storm one night in 1975?
There were drive through ships designed in North America and the United Kingdom, Italy and
France - did any of their designers ever learn about the incident in which heavy seas tore off
the clamshell doors of the Finnhansa in a storm off the Finnish coast in 1977?
Did the administrations which were busily approving plans for bigger and more sophisticated
ferries with enormous passenger loads ever learn about these and other incidents?
Were the Japanese or Canadian or British or French government surveyors ever made aware
of these operational problems? And indeed, was there any proper mechanism available for
the transmission of accident information between one ferry owner and his competitor down
the road, or owners of similar ships throughout the world?
Did the International Maritime Organisation ever become engaged in the receipt, study, or
promulgation of information about such incidents?
The answer to all these questions, I'm afraid, is probably a resounding no, because the
mechanisms for the transmission of such important operational information was not generally
Indeed, there is some evidence that the seriousness of such incidents, and the possible
consequences for these ships if water had got onto their cardecks was not properly
contemplated. In the event, the officers aboard the ships reacted promptly and properly, the
ships went back to port, were repaired to everyone's satisfaction and returned to service.
The incident had effectively ended, and what lessons that were drawn from it were probably
confined to a very small number of people. Perhaps, in fact, the potential seriousness of the
incidents were not fully comprehended, perhaps it was merely assumed that the secondary
safety appliances would continue to be adequate, as they had been in these accidents.
This inability of the maritime industry to construct and maintain a system for the prompt and
international promulgation of important safety information has worried a number of thoughtful


people over the years, but there is still very little that is done about it.
The classification societies have probably gone rather farther than most, in establishing
formal systems for alerting each other to elements prescribed by their rules which have been
shown to give trouble, but there is very little else.
There is certainly nothing even approaching the systems in the aviation world which, if a bolt
fails in the tailplane of a 737, alerts every operator of similar aircraft to this failure in a trice.
The fatal bow door damage the Estonia, for whatever reason, is merely the worst possible
example of this maritime industry system failure, which extends into every area of shipping.
It is made infinitely worse by the fragmention of the industry into so many different sectors,
and the emergence of the [lag of convenience and corresponding weakening of the
'traditional' flags.
There are some focs do their level best and provide an excellent maritime administration, but
others don't bother.
'The International Maritime Organisation, some have suggested, could be an effective forum
for an international system of incident or accident alerting. Indeed, there is an obligation for
administrations which subscribe to IMO conventions, to hold their accident inquiries and to
inform the orgar~isationof their discoveries, for the greater beneft of world shipping.
Most don't bother, and of those that do send IMO reports, these are collected, but nothing is
done with them, for there is no budget to provide for such a system that would let relevant
people know of the findings.
'There is a growing number of people who are suggesting that shipping is becoming overregulated and that IMO ought to lay off for a few years. Accidents are declining in number and
marine safety is improving. But this particular lesson from the Estonia, like the scandal of the
bulk carrier sinkings before it, shows that there is one great gap in the system that needs to
be addressed.
We fly different flags and are scattered around the world, with our different classification
societies and clubs. But we need to know about accidents like those which prefaced the
Estonia sinking. How otherwise will we ever find out? And above all, is this not a cause that
should be canvassed?

3.3 A Brief History of Casualty Investigation

The United Kingdom
The late .eighteenth century and throughout the nineteenth century was a time of mass
migration and expanding trade links. Travel and transport by sea was the only way to move
people and goods between continents. With land transport reliant on horse drawn vehicles,
the coastal passenger trade was a major means of transporting people within a country.

During this time, in the United Kingdom, shipping was largely unregulated. The loss of life at
sea on both domestic and international transport, together with the loss of cargo caused great
public concern in the UK. In 1836 the UK Government to set up a select committee of the
House of Commons to inquire into the causes of shipwrecks.
The Committee recommended that amongst other measures, a system of "Courts of Inquiry"
should be set up with the specific task of:
investigating the causes of individual shipwrecks
censuring "commanders" and ship owners
suspending licenses
rewarding those that displayed particular skill, courage and humanity
publicising the outcomes of such inquiries



A second committee in 1843 endorsed these recommendations with .the addition that the
publication of the evidence and the findings was not only to warn masters and owners, but to
"...stimulate and suggest the exercise of preventive measures for the avoidance of future
In 1846 Parliament passed the Steam Navigation Act. 'This act initially applied only to British
steam vessels, but later the jurisdiction to was extended to British sailing vessels. It required
them to carry boats and detailed minimum welfare conditions for the crew. Section 104
required that serious accidents must be reported to the Board of Trade and made provision
for the appointment of suitably skilled and qualified inspectors to inquire into and report on
such marine casualties.
The Act came into operation on 1 January 1847. On 16 January 1847, the steamship Sirius
was wrecked and a Royal Naval officer Captain Henry Denham was appointed to investigate.
No firm procedures were set out in the act and Captain Denham seems to have made up
procedures as he went along. The interviews of individuals were held "in camera". Captain
Denham undertook some 12 investigations in the first two or three years of accident
investigation. As a result of each investigation he made recommendations to the authorities
to address causal factols. By 1850 he submitted, separate from any specific casualty, a
memorandum headed "Amendments to the Steam Navigation Act 9 & 10 Vict., c.100, as
suggested by the Official Investigations of steam vessel accidents, of Captain Denham, RN,
FRS,..., "a list of the points which official inspections of steam vessel accidents had
developed, and whereon the Board might deem it advisable to promote further regulations for
the safety of life and property at sea.v
Captain Denham, a RN Hydrographer, completed some 12 investigations over the next two
years. He made some 50 safety recommendations many of which were adopted into
legislation over the following years. Late in 1851, Captain Denham was appointed to
command the survey sloop Herald and he sailed for Australia and the Pacific where he
undertook extensive survey work. He retired from the Royal Navy as an Admiral.
The earliest official investigations, therefore, emphasised the prhciple that the main aim of
an investigation was to stop accidents happening from similar causes. In evidence to a select
Committee of Parliament in 1860 it was stated:

' is necessay to remember what are the absolute requis/es for these inquiries. In
the first place they must be summa& local and inexpensive. If they are not so they will
be oppressive to the parties, they will be impracticable to the Governmentand they will
be ineffectuaL ... Thus further (and it is vey /inportant to mention this) the object of
the hquiy is not so much to punish anyone who may be at fault, as to prevent wrecks
in the future, whether by punishment of and warning against negligence, or remedying
anything that is wrong h the way of na vigathn...'5
Responding to criticisms that the investigation system developed by Denham and others
meant that the Board of Trade was both "judge and jury", the inquiry role was given to
Stipendiary Magistrates and Justices of the Peace under the new Merchant Shipping Act of
1854. Inquiries became more formal and largely followed court procedures. The specialist
technical safety investigator was replaced, by a more judicial process. However the basic
emphasis on learning the safety lessons remained.
Murton, W. Wreck Inquiries, Stevens & Sons, London, 1884.
Farrer, Sir Thomas, Permanent Secretary to the Board of Trade, Report to a Select Committee of the United
Kingdom Parliament, 1860



The Merchant Shipping Act of 1876 introduced a new, two tier system of inquiry. This
introduced the Preliminary Inquiry, followed by a Formal lnvestigation before a judge, advised
by specialist "assessors". The Preliminary investigation was to establish whether or not a
formal investigation was warranted. The Merchant Shipping Acts were binding on British
Dominions and Colonies and were adopted by these countries when they became sovereign
states. This system endured in many jurisdictions to the 1970s and beyond.
By 1929, the shipping courts occupied a unique position in the British legal system:
'They are administrative coun's of a spechl character... The Coun's were created for
the purpose of assisting the Board of Tfade in their duty of presemhg a reasonable
standard of safety of fife at sea, and of maintahing the private rights of individuals in
cases where they appear to be in c o m t with the pubfic hterest as interpretedby the
Board of Tfade. ... For these reasons it was thought to be necessaty to create special
shipphg coun's, which would combihe, so far as is poss/icle, the me& of an
admhistrative coun' h the ordina/y sense with those of a coun' of justice. '
Although the procedures were altered and a system of Preliminary investigations and Formal
inquires was adopted, the idea of a technically competent investigator remains in place.
Because of the UK's overseas colonies and dominions, the British system was used
extensively around the world. The aim of the investigation remains as stated in about 1860: '7t
is not so much to punish anyone who may be at fault, as to prevent wrecks in the future ...':
Judicial inquiries cam be expensive and various parties use them to gain information for
future civil action, rather than concentrating on the safety -issues.
In 1948 Mr Justice Streeton stated in handing down two judgements:
'7 have come to the conclusion that these inquiries are, and inevitably wil be, misused
by the pan'ies to them, excepting the Director of Navigation, for the purpose of
estabfishing evidence for use h civil fitigation between the pan'ies. That misuse is, H I
may be permitted a paradox, a legitimate misuse, but H is none the less a misuse.'7

A number of States, while maintaining the options of Formal Investigations, have reverted to
the "technical investigation".

It is with this in mind that the International Code for the lnvestigation of Marine Casualties and
Incidents was drafted.


Conceptof culpability

The problem of a criminal action is that the prosecution has the sole purpose of proving a
case against the party or parties accused of the crime and will focus on the narrow issues -of
blameworthy action. Unfortunately the causes of accidents are more complex and numerous
that one or two acts of commission or omission.
'Accidents appear to be the result of h&hly complex coincidences which could rarely
be foreseen by the people involved The unpredktabifity is caused by the h g e
number of causes and the spread of information over the parfic@ants....Accidents do
not occur because people gamble and lose, they occur becausepeople do not befieve
that the accident that is about to occur is at a//possible. " (Wageneer & Groeneweg,
7988. P42).

McMillan, A.R.G., Shipping Inguiris and Courts, Stevens and Sons, 1929
re "Koonda" & "George H Emircocos" and re "Iron Monach" & "Empire Strength", CMI 1943.

7 Streeton, J.



In dealing with accidents in an international context, IMO recognises that an international

approach is required to determine the circumstances and causes of a casualty, so that ,the
causes can be addressed - internationally - and measures introduced to prevent further
casualties. Casualty investigations under the IMO code are essentially to determine the
safety issues and identify the unsafe conditions that contributed to the accident.
Blame and liability are not relevant to IMO.
Any safety information system, for a marine casualty investigation under the Code, as part of
such a system, depends upon receiving accurate information. This requires the co-operation
of anybody involved who can give accurate, honest evidence of what occurred.
The IMO Code aims to achieve:
Qualified indemnity against disciplinary proceedings
Confidentiality or anonymity
Separation of the investigation process from disciplinary or criminal procedures
Rapid, transparent, impartial and accurate reporting
A simple reporting format, which is available and distributed widely
In dealing with the confidential reporting of air accidents O'Leary and Chappell stated9
"For any incident reponng programme to be effective k uncovenhg failures which
contribute to an incident, it is paramount to earn the trust of the reportets. This is even
more /inportant when there is a candid disclosure of the reporter's own errors. Without
such trust the repon'er will be selective and will probab& gloss over pivotal human
factors information. In the worst case - that in which potential repon'ers have no trust
in the safety organization - there may be no report at all"
Such a system is ideal and suitable for an "in-house" reporting system. The Code aims to
come as close as possible to these ideals, while recognising that casualty investigations are
carried out by, or on behalf of, a State administration.
Reason (1997) discussed the issue of culpability and the investigation of accidents3
'Hn effective repon'ing culture depends, in turn, on how the organization handles
blame and punishment. A 'no-blame' culture is neither feasfile nor desirable. A small
proportion of human unsafe acts are egregious (e,g. substance abuse, reckless noncompiance, sabotage, etc.) and warrant sanctions, severe ones in some cases. A
blanket amnesty on all unsafe acts wouldlack credbiifv in the eyes of the worMorce.
More imporfanth it would be seen to oppose naturaljustice. What is needed is a just
culture, an atmosphere of trust in which people are encouraged even rewarded for
providing essential safefy-relatedinformation - but in which they are also clear about
where the line must be drawn befween acceptable and unacceptable behaviou/:"
An important feature of the figure is what Neil Johnston has called the substitution test. This
involves asking the individual's peers the following question: Given the circumstances that
prevailed at the time, could you be.sure that you would not have committed the same or a
similar type of unsafe act? If the answer is 'no', then blame is almost certainly inappropriate.
The best people can make the worst mistakes.

8 O'Leary,

M. & Chappell, S. Confidential incident reporting systems create a vital awareness of problems.
/CAO Journal. 51 :11-13, 1996
9 Reason, J. Corporate Culture and Safety NTSB Symposium on Corporate Culture and Transportation
Safety, Washington, 1997



Degrees of culpability



Objective of an investigation

(Video: Piper Alpha (ICI) 45 minutes - Group activity - In five minutes write down what
advantages a prosecution would have had and what advantages a safety investigation

3.6. Hindsight
lnvestigators usually have the advantage over the investigated and are blessed with perfect
hindsight. It is important that this hindsight is used with professionalism, realism and
constructively. lnvestigators are not judges and should avoid being judgemental and should
ensure that they understand the situation in which the individuals were placed. It is worth
bearing in mind two quotes from Professor James Reason's book Human Erro~
'For those of us who pick over the bones of other people's dkasters, it offen seems
incredle that these warnings andhuman failures, seemingly so obvious in retrospect
s W d have gone unnoticed at the time. Beihg blessed with both unhvolvement and
hindsight it is a great temptation for retrospective observers to s//;ohto a censonbus
frame of mind and to wonder at how these people could have been so blind, stupid,
arrogant or reckless.'"
'Firsf, most people involved in serious accidents are neither stupid nor reckless,
though they may well have been blind to the consequences of their actions. Second&
we must be aware of falling prey to the fundamental atthbution error fie., blamhng
people and ignoring situational errors.'vO
iMcMillan. A.R.G., Shipping /nqui/iesand Coun's, Stevens and Sons, 1929

10 Reason, J.,

Human Errof; pg.216, Cambridge University Press, Cambridge (UK), 1990.


International Conventions, Requirements etc.

Introduction - LOS - IMO

Article 87 of the United Nations Law of the Sea Convention (UNCLOS) provides that the high
seas (the waters beyond the 12 mile territorial sea) are open to all States. Freedom of the
high seas is exercised under the conditions laid down by the Convention and by other rules
of International law, of any State may be traversed by all ships and fishing vessels. The
various freedoms of the high seas shall be exercised by all States with due regard for the
interest of other States in their exercise of the freedom of the high seas, and also with due
regard for the rights under the Convention with respect to activities in the Area.

Under Article 17 ships of all States, whether coastal or land-locked, enjoy the right of innocent
passage through the territorial sea. Passage is innocent so long as it is not prejudicial to the
peace, good order or security of the coastal State.

IMO was created in 1948.


Law of the Sea


Article 2
Legal status of the territorial sea, of the air space over the territorial sea
and of its bed and subsoil
The sovereignty of a coastal State extends, beyond its land territory and internal
waters and, in the case of an archipelagic State, its archipelagic waters, to an adjacent belt
of sea, described as the territorial sea.
This sovereignty extends to the air space over the territorial sea as well as to Its bed
and subsoil.
The sovereignty over the territorial sea is exercised subject to this Convention and to
other rules of international law.
Article 3

Breadth of the territorial sea


Every State has the right to establish the breadth of its territorial sea up to a limit not
exceeding 12 nautical miles, measured from baselines determined in accordance with this
Article 4

Outer limit of the territorial sea

The outer lirr~itof the territorial sea is the line every point of which is at a distance from the
nearest point of the baseline equal to the breadth of the territorial sea.
Right of innocent passage
Article 17
Subject to this Convention, ships of all States, whether coastal or land-locked, enjoy the right
of innocent passage through the territorial sea.


Article 21

Laws and regulations of the coastal State relating to innocent passage

The coastal State may adopt laws and regulations, in conformity with the provisions of
this Convention and other rules of international law, relating to innocent passage through the
territorial sea, in respect of all or any of the followirlg:
the safety of navigation and the regulation of maritime traffic;
the protection of navigational aids and facilities and other facilities or installations;
protection of cables and pipelines;
the conservation of the living resources of the sea;
the prevention of infringement of the fisheries laws and regulations of the coastal
(9 the preservation of the environment of the coastal State and the prevention, reduction
and control of pollution thereof;
marine scientific research and hydrographic surveys;
the prevention of infringement of the customs, fiscal, immigration or sanitary laws and
regulations of the coastal State.
Such laws and regulations shall not apply to the design, construction, manning or
equipment of foreign ships unless they are giving effect to generally accepted international
rules or standards.
The coastal State shall give due publicity to all such laws and regulations. '
Foreign ships exercising the right of innocent passage through the territorial sea shall
comply with all such laws and regulations and all generally accepted international regulations
relating to the prevention of collisions at sea.

Article 24

Duties of the coastal State

The coastal State shall not hamper the innocent passage of foreign ships through the
territorial sea except in accordance with this Convention. In particular, in the application of
this Convention or of any laws or regulations adopted in conformity with this Convention, the
coastal State shall not: (a) impose requirements on foreign ships which have the practical
effect of denying or impairing the right of innocent passage; or (b) discriminate in form or in
fact against the ships of any State or against ships carrying cargoes to, from or on behalf of
any State.
The coastal State shall give appropriate publicity to any danger to navigation, of which
it has knowledge, within its territorial sea.

Article 27

Criminal jurisdiction on board a foreign ship

The criminal jurisdiction of the coastal State should not be exercised on board a
foreign ship passing through the territorial sea to arrest any person or to conduct any
investigation in connection with any crime committed on board the ship during its passage,
save only in the following cases:
if the consequences of the crime extend to the coastal State;
if the crime is of a kind to disturb the peace of the country or the good order of the
territorial sea;
if the assistance of the local authorities has been requested by the master of the ship
or by a diplomatic agent or consular officer of the flag State; or (d) if such measures are
necessary for the suppression of illicit traffic in narcotic drugs or psychotropic substances.
The above provisions do not affect the right of the coastal State to take any steps
authorized by its laws for the purpose of an arrest or investigation on board a foreign ship
passing through the territorial sea after leaving internal waters.
In the cases provided for in paragraphs 1 and 2, the coastal State shall, if the master
so requests, notify a diplomatic agent or consular officer of the flag State before taking any


steps, and shall facilitate contact between such agent or officer and the ship's crew. In cases
of emergency this notification may be communicated while the measures are being taken.
In considering whether or in what manner an arrest should be made, the local
authorities shall have due regard to the interests of navigation.
Except as provided in Part XI1 or with respect to violations of laws and regulations
adopted in accordance with Part V, the coastal State may not take any steps on board a
foreign ship passing through the territorial sea to arrest any person or to conduct any
investigation in connection with any crime committed before the ship entered the territorial
sea, if the ship, proceeding from a foreign port, is only passing through the territorial sea
without entering internal waters.
Article 28

Civil jurisdiction i n relation to foreign ships

The coastal State should not stop or divert a foreign ship passing through the territorial
sea for the purpose of exercising civil jurisdiction in relation to a person on board the ship.
The coastal State may not levy execution against or arrest the ship for the purpose of
any civil proceediogs, save or~lyin respect of obligations or liabilities assumed or incurred by
the ship itself in the course or for the purpose of its voyage through the waters of the coastal
Paragraph 2 is without prejudice to the right of the coastal State, in accordance with
its laws, to levy execution against or to arrest, for the purpose of any civil proceedings, a
foreign ship lying in the territorial sea, or passing through the territorial sea after leaving
internal waters.

Article 94

Duties of the flag State

Every State shall effectively exercise its jurisdiction and control in administrative,
technical and social matters over ships flying its flag.
In particular every State shall:
maintain a register of ships contair~ingthe names and particulars of ships flying its flag,
except those which are excluded from generally accepted international regulations on
account of their small size; and (b) assume jurisdiction under its internal law over each ship
flying its flag and its master, officers and crew in respect of administrative, technical and
social matters concerning the ship.
Every State shall take such measures for ships flying its flag as are necessary to
ensure safety at sea with regard, inter alia, to:
the construction, equipment and seaworthiness of ships;
the manning of ships, labour conditions and the training of crews, taking into account
the applicable international instruments;
the use of signals, the maintenance of communications and the prevention of
Such measures shall include those necessary to ensure:
that each ship, before registration and thereafter at appropriate intervals, is surveyed
by a qualified surveyor of ships, and has on board such charts, nautical publications and
navigational equipment and instruments as are appropriate for the safe navigation of the ship;
that each ship is in the charge of a master and officers who possess appropriate
qualifications, in partic~.rlar in seamanship, navigation, communications and marine
engineering, and that the crew is appropriate in qualification and numbers for the type, size,
machinery and equipment of the ship;
that the master, officers and, to the extent appropriate, the crew are fully conversant
with and required to observe the applicable international regulations concerning the safety of
life at sea, the prevention of collisions, the prevention, reduction and control of marine
pollution, and the maintenance of communications by radio.



In taking the measures called for in paragraphs 3 and 4 each State is required to
conform to generally accepted international regulations, procedures and practices and to take
any steps which may be necessary to secure their observance.
A State which has clear grounds to believe that proper jurisdiction and control with
respect to a ship have not been exercised may report the facts to the flag State. Upon
receiving such a report, the flag State shall investigate the matter and, if appropriate, take any
action necessary to remedy the situation.
Each State shall cause an inquiry t o be held by or before a suitably qualified
person or persons into every marine casualty or incident of navigation o n the high
seas involving a ship flying its flag and causing loss of life or serious injury t o
nationals of another State or serious damage to ships or installations of another State
or t o the marine environment. The flag State and the other State shall co-operate in the
conduct of any inquiry held by that other State into any such marine casualty or
incident of navigation.

Article 97

Penal jurisdiction in matters of collision or any other incident of

In the event of a collision or any other incident of navigation concerning a ship on the
high seas, involving the penal or disciplinary responsibility of the master or of any other
person in the service of the ship, no penal or disciplinary proceedings may be instituted
against such person except before the judicial or admir~istrativeauthorities either of the flag
State or of the State of which such person is a national.
In disciplinary matters, the State which has issued a master's certificate or a certificate
of competence or licence shall alone be competent, after due legal process, to pronounce the
withdrawal of such certificates, even if the holder is not a national of the State which issued
No arrest or detention of the ship, even as a measure of investigation, shall be ordered
by any authorities other than those of the flag State.

Article 98

Duty t o render assistance

Every State shall require the master of a ship flying its flag, in so far as he can do so
without serious danger to the ship, the crew or the passengers:
to render assistance to any person found at sea in danger of being lost;
to proceed with all possible speed to the rescue of persons in distress, if informed of
their need of assistance, in so far as such action may reasonably be expected of him;
after a collision, to render assistance to the other ship, its crew and its passengers
and, where possible, to inform the other ship of the name of his own ship, its port of registry
and the nearest port at which it will call.
Every coastal State shall promote the establishment, operation and maintenance of an
adequate and effective search and rescue service regarding safety on and over the sea and,
where circumstances so require, by way of mutual regional arrangements co-operate with
neighbouring States for this purpose.

4.3 ILO
4.3.1 Convention 134 of 1970
Concerning the Prevention of OccupationalAccidents to Seafarers

Article 1
For the purpose of this Convention, the term "seafarer" covers a.ll persons who are
employed in any capacity on board ship, other than a ship of war, registered in a territory for
which the Convention is in force and ordinarily engaged in maritime navigation.


In the event of any doubt whether any categories of persons are to be regarded as
seafarers for the purpose of this Convention, the question shall be determined by the
competent authority in each country after consultation with the shipowners' and seafarers'
organizations concerned.
For the purpose of this Convention, the term 'occupational accidents' covers accidents
to seafarers arising out of or in the course of their employment.


Article 2
The competent authority in each maritime country shall take the necessary measures
to ensure that occupational accidents are adequately reported and investigated, and
comprehensive statistics of such accidents kept and analysed.
All occupational accidents shall be reported and statistics shall not be limited to
fatalities or to accidents involving the ship.
The statistics shall record the numbers, nature, causes and effects of occupational
accidents, with a clear indication of the department on board ship-for instance, deck, engine
or catering and of the area - for instance, at sea or in port - where the accident occurred.
The-;competent authority shall undertake an investigation into the causes and
circumstances of occupational accidents resulting in loss of life or serious personal injury, and
such other accidents as may be specified in national laws or regulations.
Article 3
In order to provide a sound basis for the prevention of accidents which are due to particular
hazards of maritime employment, research shall be undertaker1into general trends and into
such hazards as are brought out by statistics.
Article 4
Provisions concerning the prevention of occupational accidents shall be laid down by
laws or regulations, codes of practice or other appropriate means.
These provisions shall refer to any general provisions on the prevention of accidents
and the protection of health in employment which may be applicable to the work of seafarers,
and shall specify measures for the prevention of accidents which are peculiar to maritime
In particular, these provisions shall cover the following matters:
general and basic provisions
structural features of the ship
special safety measures on and. below deck
loading and unloading equipment
fire prevention and fire-fighting
anchors, chains and lines
dangerous cargo and ballast
personal protective equipment for seafarers
Article 5
The accident prevention provisions referred to in Article 4 shall clearly specify the
obligation of shipowners, seafarers and others concerned to comply with them.
Generally, any obligation on the shipowner to provide protective equipment or other
accident prevention safeguards shall be accompanied by provision for the use of such
equipment and safeguards by seafarers and a requirement that they comply with the relevant
accident prevention measures.











Article 6
Appropriate measures shall be taken to ensure the proper application of the provisions
referred to in Article 4, by means of adequate inspection or otherwise.
Appropriate measures shall be taken to ensure compliance with these provisions.
All necessary steps shall be taken to ensure that inspection and enforcement
authorities are familiar with maritime employment and its practices.
In order to facilitate application, copies or summaries of the provisions shall be brought
to the attention of seafarers, for instance by display in a prominent position on board ship.
Article 7
Provision shall be made for the appointment, from amongst the crew of the ship, of a suitable
person or suitable persons or of a suitable comn-littee responsible, under the Master, for
accident prevention.
Article 8
Programmes for the prevention of occupational accidents shall be established by the
competent authority with the co-operation of shipowners' and seafarers' organizations.
Implementation of such programmes shall be so organised that the competent
authority, shipowners and seafarers or their representatives and other appropriate bodies
may play an active part.
In particular, national or local joint accident prevention committees or ad hoc working
parties, on which both shipowners' and seafarers' organizations are represented, shall be
Article 9
The competent authority shall promote and, in so far as appropriate under national
conditions, ensure the inclusion, as part of the instruction in professional duties, of instruction
in the prevention of accidents and in measures for the protection of health in employment in
the curricula, for all categories and grades of seafarers, of vocational training institutions.
All appropriate and practicable measures shall also be taken to bring to the attention
of seafarers information concerning partici~larhazards, for instance by means of official
notices containing relevant instructions.
Article 10
Members, with the assistance as appropriate of intergovernmental and other international
organizations, shall endeavour, in co-operation with each other, to achieve the greatest
possible measure of uniformity of other action for the prevention of occupational accidents.

Costa Rica


Germany, Federal Republic of

New Zealand



4.3.2 ILO Convention 147 of 1978

Concerning Minimum Standgds in Merchant Sbbs

Article 1
Except as otherwise provided in this Article, this Convention applies to every seagoing ship, whether publicly or privately owned, which is engaged in the transport of cargo or
passengers for the purpose of trade or is employed for any other commercial purpose.
National laws or regulations shall determine when ships are to be regarded as seagoing ships for the purpose of this Convention.
This Convention applies to sea-going tugs.
This Convention does not apply to:
ships primarily propelled by sail, .whether or not they are fitted with auxiliary engines;
ships engaged in fishing or in whaling or in similar pursuits;
small vessels and vessels such as oil rigs and drilling platforms when not engaged in
navigation, the decision as to which vessels are covered by this subparagraph to be taken by
the competent authority in each country in consultation with the most representative
organisations of shipowners and seafarers.
Nothing in this Convention shall be deemed to extend the scope of the Conventions
referred to in the Appendix to this Convention or of the provisions contained therein.
Article 2
Each Member which ratifies this Convention undertakes:
to have laws or regulations laying down, for ships registered in its territory:
safety standards, including standards of competency, hours of work and manning, so
as to ensure the safety of life on board ship;
appropriate social security measures; and
conditions of employment and shipboard living arrangements, in so far as
these, in the opinion of the Member, are not covered by collective agreements or laid down
by competent courts in a manner equally binding on the shipowners and seafarers
and to satisfy itself that the provisions of such laws and regulations are substantially
equivalent to the Conventions or Articles of Conventions referred to in the Appendix to this
Convention, in so far as the Member is not otherwise bound to give effect to the Conventions
in question;
to exercise effective jurisdiction or control over ships which are registered in its
territory inrespect of:
safety standards, including standards of competency, hours of work and manning,
prescribed by national laws or regulations;
social security measures prescribed by national laws or regulations;
shipboard conditions of employment and shipboard living arrangements prescribed by
national laws or regulations, or laid down by competent courts in a manner equally binding
on the shipowners and seafarers concerned;
to satisfy itself that measures for the effective control of other shipboard conditions of
employment and living arrangements, where it has no effective jurisdiction, are agreed
between shipowners or their organisations and seafarers' organisations constituted in
accordance with the substantive provisions of the Freedom of Association and Protection of
the Right to Organise Convention, 1948, and the Right to Organise and Collective Bargaining
Convention, 1949;
to ensure that:
adequate procedures - subject to over-all supervision by the competent authority,
after tripartite consultation amongst that authority and the representative organisations of
shipowners and seafarers where appropriate - exist for the engagement of seafarers on
ships registered in its territory and for the investigation of complaints arising in that


adequate procedures - subject to over-all supervision by the competent authority,

after tripartite consultation amongst that authority and the representative organisations of
shipowners and seafarers where appropriate - exist for the investigation of any complaint
made in connection with and, if possible, at the time of the engagement in its territory of
seafarers of its own nationality on ships registered in a foreign country, and that such
complaint as well as any complaint made in connection with and, if possible, at the time of
the engagement in its territory of foreign seafarers on ships registered in a foreign country, is
promptly reported by its competent authority to the competent authority of the country in
which the ship is registered, with a copy to the Director-General of the lnternational Labour
to ensure that seafarers employed on ships registered in its territory are properly
qualified or trained for the duties for which they are engaged, due regard being had to the
Vocational Training (Seafarers) Recommendation, 1970;
to verify by inspection or other appropriate means that ships registered in its territory
comply with applicable international labour Conventions in force which it has ratified, with the
laws and regulations required by subparagraph (a) of this Article and, as may be appropriate
under national law, with applicable collective agreements;
to hold an official inquiry into any serious marine casualty involving ships registered in
its territory, particularly those involving injury and/or loss of life, the final report of such inql-~iry
normally to be made public.
Article 3
Any Member which has ratified this Convention shall, in so far as practicable, advise its
nationals on the possible problems of signing on a ship registered in a State which has not
ratified the Convention, until it is satisfied that standards equivalent to those fixed by this
Convention are being applied. Measures taken by the ratifying State to this effect shall not be
in contradiction with the principle of free movement of workers stipulated by the treaties to
which the two States concerned may be parties.
Article 4
If a Member which has ratified this Convention and in whose port a ship calls in the
normal course of its business or for operational reasons receives a complaint or obtains
evidence that the ship does not conform to the standards of this Convention, after it has come
into force, it may prepare a report addressed to the government of the country in which the
ship is registered, with a copy to the Director-General of the lnternational Labour Office, and
may take measures necessary to rectify any conditions on board which are clearly hazardous
to safety or health.
In taking such measures, the Member shall forthwith notify the nearest maritime,
consular or diplomatic representative of the flag State and shall, if possible, have such
representative present. It shall not ur~reasonablydetain or delay the ship.
For the purpose of this Article, 'complaint' means information submitted by a member
of the crew, a professional body, an association, a trade union or, generally, any person with
an interest in the safety of the ship, including an interest in safety or health hazards to its
Article 5
This Convention is open to the ratification of Members which:
are parties to the lnternational Convention for the Safety of Life at Sea, 1960, or the
lnternational Convention for the Safety of Life at Sea, 1974, or any Convention subsequently
revising these Conventions; and
are parties to the lnternational Convention on Load Lines, 1966, or any Convention
subsequently revising that Convention; and
are parties to, or have implemented the provisions of, the Regulations for Preventing


Collisions at Sea of 1960, or the Convention on the International Regulations for Preventing
Collisions at Sea, 1972, or any Convention subsequently revising these international
This Convention is further open to the ratification of any Member which, on ratification,
undertakes to fulfil the requirements to which ratification is made subject by paragraph 1 of
this Article and which are not yet satisfied.
The formal ratifications of this Convention shall be communicated to the DirectorGeneral of the International Labour Office for registration.

Mi~iimumAge Convention, 1973 (No. 138), or
Minimum Age (Sea) Convention (Revised), 1936 (No. 58), or
Minimum Age (Sea) Convention, 1920 (No. 7);
Shipowners' Liability (Sick and injured Seamen) Convention, 1936 (No. 55), or
Sickness Insurance (Sea) Convention, 1936 (No. 56), or
Medical Care and Sickness Benefits Convention, 1969 (No. 130);
Medical Examination (Seafarers) Convention, 1946 (No. 73);
Prevention of Accidents (Seafarers) Convention, 1970 (No. 134) (Articles 4 and 7);
Accommodation of Crews Convention (Revised), 1949 (No. 92);
Food and Catering (Ships' Crews.) Convention, 1946 (No, 68) (Article 5);
Officers' Competency Certificates Convention, 1936 (No. 53) (Articles 3 and 4),*
Seamen's Articles of Agreement Convention, 1926 (No. 22);
Repatriation of Seamen Convention, 1926 (No. 23);
Freedom of Association and Protection of the Right to Organise Convention, 1948 (No. 87);
Right to Organise and Collective Bargaining Convention, 1949 (No. 98).
Costa Rica
E ~ptY
Germany, Federal Republic of
United Kingdom

Sept 16 1982
June 24 1981
July 28 1980
Mar 17 1983
Oct 2 1978
May 2 1978
July 14 1980
Sept 18 1979
June 23 1981
July 8 1981
June 15 1981
Jan 25 1979
Jan 24 1970
Apr 26 1978
Dec 20 1978
Nov 28 1980

In cases where the established licensing system or certification structure of a State would be prejudiced by
problems arising from strict adherence to the relevant standards of the Officers' Competency Certificates
Convention, 1936, the principle of substantial equivalence shall be applied so that there will be no conflict with
that State's established arrangements for certification.


4.3.3 R 142 Prevention of Accidents (Seafarers) Recommendation, 1970

Recommendation concerning the Prevention of Occupatlbnal Accidents to Seafarers

Place: Geneva
Session of the Conference: 55
Date of adoption = 30:10:1970

The General Conference of the lnternational Labour Organisation,

Having been convened at Geneva by the Governing Body of the lnternational Labour Office,
and having met in its Fifty-fifth Session on 14 October 1970, and
Considering that, although much is being done in a number of countries to reduce
occupational accidents to seafarers, there is room for further study of such accidents and for
further measures for their prevention, and that international standards embodying a relevant
programme of action for the maritime sector are accordingly desirable, and
Having decided upon the adoption of certain proposals with regard to accident prevention on
board ship at sea and in port, which is the fifth item on the agenda of the session, and
Having determined that these proposals shall take the form of a Recommendation
supplementing the Prevention of Accidents (Seafarers) Convention, 1970, and
Noting that the following standards have been framed with the co-operation of the InterGovernmental Maritime Consultative Organization, and that it is proposed to seek its
continuing co-operation in promoting and securiog the application of these standards,
adopts this thirtieth day of October of the year one thousand nine hundred and seventy, the
following Recommendation, which may be cited as the Prevention of Accidents (Seafarers)
Recommendation, 1970:

For the purpose of this Recommendation:

the term seafarer covers all persons who are employed in any capacity on board a
ship, other than a ship of war, ordinarily engaged in maritime navigation;
the term occ~npationalaccidents covers accidents to seafarers arising out of or in the
course of their employment.
In giving effect to paragraph 3 of Article 2 of the Prevention of Accidents (Seafarers)
Convention, 1970, Members should have due regard to any international system of recording
accidents to seafarers which may have been established by the lnternational Labour
Subjects to be investigated in pursuance of Article 3 of the Prevention of Accidents
(Seafarers) Convention, 1970, might include:

working environment, such as working surfaces, layout of machinery and means of

access and lighting, and methods of work;
incidence of accidents in different age groups;
special physiological or psychological problems created by the shipboard


problems arising from physical stress on board ship, in particular as consequence of

increased workload;
problems arising from and effects of technical developments and their influence on the
composition of crews;
problems arising from any human failures such as carelessness.
In formulating the accident prevention provisions called for by Article 4 of the
Prevention of Accidents (Seafarers) Convention, 1970, Members should have due regard to
any Code of Practice concerning the safety and health of seafarers which may have been
published by the International Labour Office.
In giving effect to Article 5 of the Prevention of Accidents (Seafarers) Convention,
1970, account should be taken of Articles 7 and 11 of the Guarding of Machinery Convention,
1963 - and the corresponding provisions of the Guarding of Machinery Recommendation,
1963 - under which the obligation to ensure compliance with the requirement that machinery
in use is properly guarded, and its use without appropriate guards prevented, rests on the
employer, while there is an obligation on the worker not to use machinery without the guards
being in position nor to make inoperative the guards provided.


6.(1) The functions of the committees and other bodies referred to in paragraph 3 of Article
8 of the Prevention of Accidents (Seafarers) Convention, 1970, might include:

the preparation of accident prevention provisions, rules and manuals;


the organisation of accident prevention training and programmes;

the organisation of accident prevention publicity, including films, posters, notices and
the distribution of accident prevention literature and information so that it reaches
seafarers on board ship.
Relevant provisions or recommendations adopted by the appropriate national
authorities or orgarlisations or responsible international maritime organisations should be
taken into account by those preparing texts of accident prevention measures andlor
recommended practices.
The syllabuses of the instruction referred to in Article 9 of the Prevention of Accidents
(Seafarers) Convention, 1970, should be reviewed periodically and brought up to date in the
light of development in types and sizes of ships and in their equipment, as well as changes
in crewing practices, nationality, language and the orgarlisation of work on board ship.
8. (1) There should be continuous accident prevention publicity.


Such publicity might take the following forms:

instructional films, film strips and shorts, for use in vocational training centres for
seafarers and where possible in film programmes screened on board ship;

display of safety posters on board ship;

inclusion of articles on hazards of maritime employment and accident prevention

measures in periodicals read by seafarers;



special campaigns, during which various media of publicity are used to instruct
seafarers in accident prevention and safe working practices.
The publicity should take into account that there are often seafarers of different
nationalities, languages and habits on board ship.


9. (1) In giving effect to Article 10 of the Accident Prevention (Seafarers) Convention, 1970,
Members should have due regard to relevant Model Codes of Safety Regulations or Codes
of Practice published by the International Labour Office and the appropriate standards of
international organisations for standardisation.

(2) Members should further have regard to the need for international co-operation in the
continuous promotion of action for the prevention of occupational accidents; such cooperation might take the form of:
bilateral or multilateral arrangements for uniformity in accident prevention standards
and safeguards;
exchange of information on particular hazards affecting seafarers and on means of
preventing accidents;
assistance in testing of equipment and inspection according to the national regulations
of the country of registration of the ship;
collaboration in the preparation and dissemination of accident prevention provisions,
rules or manuals;
collaboration in the production and use of training aids;
joint facilities for or mutual assistance in the training of seafarers in accident
prevention and safe working practices.
Cross references 1970 - Guarding of Machinery Recommendation, 1963


IMO Conventions and Resolutions

4.4.1 International Convention for the Safety of Life at Sea, 1974

as amended by the Protocol of 1978

Chapter I
Part C



Regulation 21
Each Administration undertakes to conduct an investigation of any casualty occurring
to any of its ships subject to the provisions of the present Convention when it judges that such
an investigation may assist in determining what changes in the present regulations may be
Each Contracting Government undertakes to supply the Organization with pertinent
information concerning the findings of such investigations. No reports or recommendations of
the Organization based upon such information shall disclose the identity or nationality of the
ships concerned or in any manner fix or imply responsibility upon any ship or person.
'Refer to the following resolutions adopted by the Organization:
Resolution A.173(ES.IV): Participation in official inquiries into marine casualties.
Resolution A.203(Vll): Recommendation on the conclusion of agreements and arrangements between
States on the question of access and employment of foreign seaborne salvage equipment in territorial


Resolution A.322(IX): The conduct of investigations into casualties.
Resolution A.440(XI) : Exchange of information for investigations into marine casualties.
Resolution A.442(XI): Personnel and material resource needs of Administrations for the investigation
of casualties and contraventions of conventions.
Resolution A.637(16) : Co-operation in maritime casualty investigations.
Refer also to:
MSCICirc.70JRev.l: Questionnaire on die maritime distress system.
MSCJCirc.224: Submission of damage cards and intact stability casualty records.
MSCJCirc.388: Fire casualty records.
MSCJCirc.433: Reports on investigations into serious casualties.
MSCJCirc.539JAdd.2: Reports on casualty statistics concerning fishing vessels and fishermen at sea.
MSCJCirc.359: Guidelines to ensure the reporting to the Organization of incidents involving
dangerous goods and marine pollutants in packaged form on board ships and in port areas.
MSCICirc.621: Guidelines for the investigation of accidents where fatigue may have been a
contributing factor.


International Convention o n Load Lines, 1966

Article 23
Each Administration undertakes to conduct an investigation of any casualty occurring
to any of its ships subject to the provisions of the present Convention when it judges that such
an investigation may assist in determining what changes in the present regulations may be
Each Contracting Government undertakes to supply the Organization with pertinent
information concerning the findings of such investigations. No reports or recommendations of
the Organization based upon such information shall disclose the identity or nationality of the
ships concerned or in any manner fix or imply responsibility upon any ship or person.
4.4.3 International Convention for the Prevention of Pollution from Ships 1973, as
amended by the Protocol of 1978 (MARPOL 73/78)

Article 8
Repon's on /&dents involving harmful Substances
A report of an incident shall be made without delay to the fullest extent possible in
accordance with the provisions of Protocol I to the present Convention.
Each party to the Convention shall:
all arrangements necessary for an appropriate officer or agency to receive and
process all reports on incidents; and
notify the Organization with complete details of such arrangements for circulation to
other Parties and Member States of the Organization.

Whenever a Party receives a report under the provisions of the present Article, that
Party shall relay the report without delay to:
the Administration of the ship involved; and
any other State which may be affected.
Each party to the Convention undertakes to issue instructions to its maritime
inspection vessels and aircraft and to other appropriate services, to report to its authorities
any incident referred to in Protocol I to the present Convention. That Party shall, if it considers
it appropriate, report accordingly to the Organization and to any other party concerned.



Article 12
Casualties to Sh@s
Each Administration undertakes to conduct an investigation of any casualty occurring
to any of its ships subject to the provisions of the Regulations if such casualty has produced
a major deleterious effect upon the marine environment.
Each party to the Convention undertakes to supply the Organization with information
concerning the finding of such investigation, when it judges that such information may assist
in determining what changes in the present Convention might be desirable.


4.4.4 International Convention on Standards of Training, Certification and

Watchkeeping for Seafarers, 1978 as amended in 1995

Regulation 115
Nationalpro visions
Each party shall establish processes and procedures for the impartial investigation of
any reported incompetency, act or omission, that may pose a direct threat to safety of life or
property at sea or to the marine environment, by the holders of certificates or endorsements
issued by that Party in connection with their performance of duties related to their certificates
and for the withdrawal, suspension and cancellation of such certificates for such cause and
for the prevention of fraud.
Each Party shall prescribe penalties or disciplinary measures for cases in which the
provisions of its national legislation giving effect to the Convention are not complied with in
respect of ships entitled to fly its flag or of seafarers duly ceitified by that Party.


In particular, such penalties or disciplinary measures shall be prescribed and enforced
in cases in which:

a company or master has engaged a person not holding a certificate as

required by the Convention;


a master has allowed any function or service in any capacity required by these
regulations to be performed by a person holding an appropriate certificate, to
be performed by a person not holding the required certificate, a valid
dispensation or having the documentary proof required by regulation Ill0,
paragraph 5; or


a person has obtained by fraud or forged documents an engagement to

perform any function or serve in any capacity required by these regulations to
be performed or filled by a person holding a certificate or dispensation.

A Party, within whose jurisdiction there is located any company which, or any person
who, is believed on clear grounds to have been responsible for, or to have knowledge of, any
apparent non-compliance with the Convention specified in paragraph 3, shall extend all cooperation possible to any Party which advises it of its intention to initiate proceedings under
its jurisdiction.



International Convention for the Safety of Fishing Vessels, 1977


Each Party shall arrange for an investigation of any casualty occurring to any of its
vessels subject to the provisions of the present Convention when it judges that such an
investigation may assist in determining what changes in the present regulations may be
Each Party shall supply the Organization with pertinent information concerning the
,findings of such investigations for circulation to all parties. No reports or recommendations of
the Organization based upon such information shall disclose the identity or nationality of the
ships concerned or in any manner fix or imply responsibility upon any ship or person.

The convention relating t o lntervention on the High Seas in Cases of Oil

Pollution Casualties 1969, as amended by the Protocol of 1973
(The lntervention Convention)

On 18 arch 1967, the tanker Tofey Canyon ran aground near the Scilly Isles - outside UK
territorial waters - with a cargo of 119,000 tons of oil on board. Many of the cargo tanks were
damaged as a result of the grounding, and by 20 March some 30,000 tons of oil had escaped
into the sea. Subsequent strong winds and high seas broke the back of the vessel, so that by
27 ~arch'afurther 30,000 tons had been released with a likelihood of even more escaping
later. Therefore, in order to avoid serious pollution of the Cornish coast, the UK Government
ordered the wreck bombed so as to burn the remaining oil.
The actions of the UK in bombing the Torrey Canyonwere contrary to international law at the
time, and described by some as piracy. The incident did, however, lead to a general review
of the problems of pollution of the sea in international law.
As a result, the Convention relating to lntervention on the High Seas in Cases of Oil Pollution
Casualties was adopted by a Diplomatic Conference (convened by IMO) in November 1969
for the purpose of clarifying the rights of the coastal state. The intervention in the irorrey
Canyon incident was subsequently justified by the International Law Commission as an act
of necessity. It is this doctrine that forms the basis of the Convention. The doctrine means
effectively that it is lawful to do what would otherwise be unlawful if there is a compelling
necessity to act to save life or to protect the environment.
The Convention allows parties to take such measures on the high seas as may be considered
necessary to prevent, mitigate or eliminate grave and imminent danger to their coastline or
related interests from pollution or threat of pollution of the sea by oil.
Related interests, as defined, include fishing activities, tourism and the well-being of living
marine resources and of wildlife. For the Convention to apply, these interests must be directly
affected or threatened. Hence there is a power to intervene on purely environmental grounds,
in contrast to the position regarding civil liability at common law. The Convention is not limited'
to tankers nor only to pollution by oil: it applies to any sea-going vessel of any type
whatsoever and any floating craft. It does not apply to an installation or device engaged in the
exploration and exploitation of the resources of the seabed. Warships and governmentowned ships on non-commercial service are also excluded.
The coastal state is empowered to take action only after due consultations with appropriate
interests, including the flag state(s) of the ships(s) involved, the owners of the ship(s) or
cargoes in question and where circumstances permit, independent experts appointed for this


purpose. A coastal state which takes measures beyond those permitted under the Convention
is liable to pay compensation for any damage caused by such measures. Provision is made
for the settlement of disputes arising in connection with the application of the Convention.
The provisions of the Convention were extended by a protocol to the Convention, extending
its provisions to pollution by chemical cargoes.

4.5 IMO Resolutions and circulars


Resolution A.847(20) Guidelines to Assist Flag States,

Annex para 7


In addition to providing qualified surveyors, the flag State s h o ~ ~ provide
investigators. Consistent with article 94.6 and articles 217.45 and 6 of UNCLOS and with the
provisions of the relevant IMO conventions, investigations should be carried out following a
marine casualty or pollution incident. 'The flag State should ensure that individual
investigators have a working knowledge and practical experience in those subject areas
pertaining to their norm41duties. Additionally, to assist individual investigators in performing
duties outside their normal assignments, the flag State should ensure ready access to
expertise in the following areas, qs necessary:


navigation and the Collision Regulations


flag State regulations on certificates of competency


causes of marine pollution


interview techniques


evidence gathering


evaluation of the effects of the human element.

Any accident involving personal injury necessitating absence from duty of three days
or more and any deaths resulting from occupational accidents and casualties to ships of the
flag State should be investigated, and the results of such investigations made public. Ship
casualties should be investigated and reported on in accordance with UNCLOS, relevant IMO
Conventions, and the Guidelines currently being developed by IMO4. Casualty investigations
should be conducted by suitably qualified investigators, competent in matters relating to the
casualty. The report of the investigation should be forwarded to IMO together with the flag
State's observations, in accordance with the Guidelines referred to above.

Refer to the Code for the Investigation of Marine Casualties and Incidents, adopted by the Organization by
resolution A.849(20).



IMO Circular

9 December 1997

Harmonized reporting procedures Reports required under

SOLAS regulation 1/21 and MARPOL 73/78 articles 8 and 12
The Maritime Safety Committee, at its sixty-eighth session (28 May to 6 June 1997)
and the Marine Environment Protection Committee, at its fortieth session (18 to 25
September 1997) approved an MSCIMEPC circular on Reports on marine casualties and
incidents - Harmonized reporting procedures, amalgamating and harmonizing the
procedures for reporting casualties to the Organization contained in existing MSC and MEPC
circulars; and agreed there was a need to develop guidelines for the application of the
reporting procedures (especially with regard to a uniform application of the coding systems


Under SOLAS regulation 1121 and MARPOL 73178 articles 8 and 12, each
Administration undertakes to conduct an investigation into any casualty occurring to ships
under its flag subject to those conventions and to supply the Organization with pertinent
information concerning the findings of such investigations.
The reporting formats contained in the annexes to this circular replace the reporting
forms contained in MSClCirc.224 on Damage Cards and Intact Stability Casualty Records,
MSCICirc.388 on Fire Casualty Records, MSCICirc.433 on Reports on Investigations into
Serious Casualties, MSClCirc.559 on Incidents Involving Dangerous Goods or Marine
Pollutants in Packaged Form, MSCICirc.621 on Fatigue and COMlCirc.70lRev.l on the
Maritime Distress System. 'The reporting format on Incidental Spillages of Harmful
Substances of 50 Tonnes or More has been added, as such reports are considered
necessary when investigating a casualty or an incident (MARPOL 73/78, articles 8 and 12);
however, this does not replace the one-line entry report required by the annual mandatory
report under MARPOL 73/78, article 11 (MEPCICirc.318, Part 1).
For the purpose of reporting information to the Organization, vessel casualties are
classified as "very serious casualties", "serious casualties", "less serious casualties" and
"marine incidents". Administrations are requested to submit data for all "very serious
casualties" and "serious casualtiesn*.
Where there are important lessons to be learned from "serious casualties", "less serious
casualties" *marine
incidents", full investigation reports should be submitted along with
the additional information indicated in annex 3.
*"Very serious casualties" are casualties to ships which involve total loss of the ship, loss of
life, or severe pollution, the definition of which, as agreed by the Marine Environment
Protection Committee at its thirty-seventh session (MEPC 37122, paragraph 5.8), is as
"Severe pollution" is a case of poll~~tion
which, as evaluated by the coastal State(s)
affected or the llag State, as appropriate, produces a major deleterious effect upon the
environment, or which would have produced such an effect without preventive action.





"Serious casualties" are casualties to ships which do not qualify as "very serious
casualties" and which involve:

a fire, explosion, collision, grounding, contact, heavy weather damage, ice

damage, hull cracking, or suspected hull defect, etc., resulting in:
structural damage rendering the ship unseaworthy such as penetration
of the hull underwater, immobilization of main engines, extensive
accommodation damage, etc.; or
pollution (regardless of quantity); andlor
a breakdown necessitating towage or shore assistance

"Less serious casualties" are casualties to ships which do not qualify as "very serious
casualties" or "serious casuatties" and for the purpose of recording useful information
also include "marine incidents".


Administrations are urged to submit data as indicated below.

lnformation to be submitted per casualty class

Information to
be sent in
with the type
of casualty

May be
May be
To be pro vided To be provided
provided H there provided if there
within 6
within 6
are /inportant
are /inportant
months afier
months afier
the casualty
the casualty
to be learned
to be learned
in all cases
in all cases

Annex 7 of
the attached

2 and 3
of the attached
report format,
as well as
other relevant
in vestigatbn


Less serious


Very serious

To be pro vided To be provided

at the end
at the end
of the
of the
in all cases
in all cases

To be provided
at the end
of the
1 in all cases

May be
provided if
there are
lessons to be

May be
provided if
there are
lessons to be

May be
provided if
there are
lessons to be

May be
provided if
there are
lessons to be

May be
provided if
there are
lessons to be


Ve/y serious casualty

preliminary information as indicated in Annex 1 *
information as indicated in Annexes 2 and 3, as well as other relevant annexes
a full investigation report in all cases
Serious casualty
preliminary information as indicated in Annex I *
information as indicated in Annexes 2 and 3, as well as other relevant annexes
a full investigation report only in cases of important lessons to be learnt regarding
IMO regulations
(" To be submitted within six months of the casualty date unless complete information
is submitted within this time limit.)
Less serious casualty and mahe incident
lnformation as indicated in Annexes 1, 2 and 3, as well as other relevant annexes,
only in cases of important lessons to be learnt regarding IMO regulations
a full investigation report only in cases of important lessons to be learnt regarding
IMO regulations.
lnformation to be submitted for casualtieslincidents as indicated below.
lnformation from casualties involving dangerous
goods on marine pollutants in packaged form on
board ships and in port areas.

Maritime distress system

- Annex 4
- Annex 5
- Annex 6
- Annex 7

Fatigue as a contributory cause to maritime

accidents - Fatigue factors data compilation sheet

- Annex 8

Incidental spillage of liquids of 50 tonnes or more

- Annex 9

Damage cards and intact stability records

Fire casualty record

List of Annexes


Indicates the information to be submitted in all casualty reports.



Indicates information to be supplied on ''very serious" and "serious"



Additional information required for "very serious casualties" and
"serious" casualties.







This form may be applicable for marine casualties as defined as well as
marine incidents.



This form may apply to "very serious" and "serious" casualties.



This form may apply to "very serious" and "serious" casualties.



SYSTEM This form may apply to "very serious" and "serious"




This form will apply where fatigue is deemed to be a contributory factor
in the casualty.


This form relates to incidents involving harmful substances. The report
is considered necessary when investigating a casualty or an incident
(MARPOL 73/78, articles 8 and 12), however this does not replace the
one-line entry report required by the annual mandatory report under
MARPOL 73/78, article 11 (MEPCICirc.318, Part 1).


Administrations are urged to supply the ship identification information listed in this annex for
all marine casualty reports submitted to the Organization.



IMO Number:
Name of Ship:
Flag State:
Type of Ship
Liquefied Gas Tanker
Chemical Tanker
Oil Tanker
Other Liquids (non-flammable)Tanker
Bulk Dry (general, ore) Carrier
Bulk DryIOil Carrier
Self-Discharging Bulk Dry Carrier
Other Bulk Dry (cement, woodchips, urea and other specialized) Carriers
General Cargo Ship


PassengerIGeneral Cargo Ship

Container Ship
Refrigerated Cargo Ship
Ro-Ro Cargo Ship
PassengerIRo-Ro Cargo Ship
Passenger Ship
High Speed Craft
Other Dry Cargo (livestock, barge, heavy cargo, etc.) Carrier
Fish Catching Vessel
Fish Factory ShipIFish Carrier
Offshore Supply Ship
Other Offshore Ship
Research Ship
TowingIPushing Tug
Other Activities Ship
Non-Propelled Ships
Other Ships Structures
Gross Tonnage:
Length overall:
Classification Society:
Registered Shipowner:
Ship Managerloperator:
Previous names:
Previous Flag:
Previous Class Society:
Date of contracttkeel laidldelivery:
Date of major conversion:
Hull material
light alloy
composite materials

Hull construction
single hull
double hull
double bottom
double sides
.5 . mid deck


Building yard:
Hull number:
Date of total loss/constructive total losslscrapping:
Number of crew:
Number of passengers:




Date and time (local on board):

Position1 location:
Initial event (For an explanation of the terms below see annex 2)
stranding1 ground
fire or explosion
hull failurelfailure of watertight doorslports, etc
,machine damage
damages to ship or equipment
missing: assumed lost





total loss of the ship

ship rendered unseaworthy
ship remaining seaworthy
loss of life
serious injuries

Summary of events



Date and local time of casualty: (24 hr clock) (YYMMDD)

Position of casualty (Latitude, Longitude):
Location of casualty:
At berth '
Port approach
Inland waters
Coastal waters (within 12 miles)
3.10 Open sea


Pilot on board
Type of casualty (initial event):
Collision: striking or being struck by another ship (regardless of whether
under way, anchored or moored).
5.1.1 IMO Number of other ship involved. (not coded)
5.1.2 Name of other ship involved. (not coded)
Stranding or grounding: being aground, or hittingltouching shore or sea
bottom or undewater objects (wrecks, etc.).
Contact: striking any fixed or floating object other than those included
in Nos. 1 or 2. .5.4
Fire or explosion.
Hull failure or failure of watertight doors, ports, etc.: not caused by
Nos. 1 to 4.
Machinery damage: not caused by Nos. 1 to 5, and which necessitated
towage or shore assistance.
Damages to ship or equipment: not caused or covered by Nos. 1 to 6.
Capsizing or listing: not caused by Nos. 1 to 7.
Missing: assumed lost.
5.10 Other: all casualties which are not covered by Nos. 1 to 9.

Type of subsequent events

6.1 - Collision: striking or being struck by another ship (regardless of whether
under way, anchored or moored).
6.1.1 IMO Nurnber of other ship involved. (not coded)
6.1.2 Name of other ship involved. (not coded)
Stranding or grounding: being aground, or hittingltouching shore or
sea bottom or undewater objects (wrecks, etc.).
Contact: striking any fixed or floating object other than those
included in Nos. 1 or 2.
Fire or explosion.
Hull failure or failure of watertight doors, ports, etc.


Machinery damage which necessitated towage or shore assistance.

Damages to ship or equipment.
Capsizing or listing.
Missing: assumed lost.
Other: all events which are not covered by Nos. 1 to 9.

Consequences of the casualty


Consequences to the ship involved in the casualty:

7.1.1 Total loss
7.1.2 'The ship was rendered unseaworthy
7.1.3 The ship remained seaworthy


Consequences related to human beings:

7.2.1 Number of dead or missing crew
7.2.2 Number of dead or missing passengers
7.2.3 Number of other dead or missing persons
7.2.4 Number of crew being seriously* injured in the casualty
7.2.5 Number of passengers being seriously* injured in the casualty
7.2.6 Number of other persons being seriously* injured in the casualty
incapacitated for 72 hours or more
Consequences to the,environment (pollution):
7.3.1 Oil in bunkers

Type of oil

Quantity spilled

OHeavy fuel
flLube oils
7.3.2 Oil cargo

Type of oil (not coded)

OCrude oil
OPersistent refined oil products
flNon-persistent refined oil products

Quantity spilled


7.3.3 Chemicals in bulk

Category (Appendix 1 to Annex 11 of MARPOL 73/78)

Quantity in tons spilled


7.3.4 Dangerous Goods in packaged form

(IMDG Code)


UN numbers

Quantity lost



Primary causes of the initial event

Coding principle:
'The human element is a complex multi-dimensional issue that affects maritime safety
and. marine environmental protection. It involves the entire spectrum of human activities
performed by ships' crews, shore based management, regulatory bodies, classification
societies, shipyards, legislators and other relevant parties.
Effective remedial action following maritime casualties requires a sound
understanding of the human element involvement in accident causation. This comes by the
thorough investigation and systematic analysis of casualties for contributory factors and the
causal chain of events.



Internal causes (related to the ship where the casualty occurred)

8.1.I Human violations or errors by the crew

Human violations
Human error
8.1.2 Human violations or errors by the pilot
Human violations
Human error
8.1.3 Structural failures of the ship
8.1.4 Technical failure of machinerylequipment including design


8.1.5 'The

Failure of propulsion machinery

Failure of essential auxiliary machinery
Failure of steering gear
Failure of closing arrangements or seals
Failure or inadequacy of navigational equipment
Failure of bilge pumping
Failure of electrical installation
Failure or inadequacy of communication equipment
Failure or inadequacy of lifesaving appliances
Ship design errors (i.B. insufficient stability)
ship's cargo
Cargo shifting
Fire or explosion in cargo
Improper stowage of cargo
Spontaneous combustion
Cargo liquefaction

External causes (outside the ship)

8.2.1 Another ship or ships (improper actions, etc.)

8.2.2 The environment
Heavy sea
Currents or tides
Ice conditions
Restricted visibility
8.2.3 Navigational infrastructure
Failures in aids to navigation
Inaccurate charts or nautical publications
Charts or nautical publications unavailable for the sea
8.2.4 Criminal acts
8.2.5 Other "external" causes (i.e. not associated with the ship itself)


Tug boat operations

Failure or incorrect operation of shore equipment or
Other than .I and .2



Unknown causes

Violations and error types


Violation (deliberate decision to act against a rule or plan)

9.1.1 Routine (cutting corners, taking path of least effort, etc ...)
9.1.2 Necessary (due to inadequate tools or equipment, improper
procedures or regulations)
9.1.3 "For kicks" (thrill seeking, to alleviate boredom, macho behaviour)
9.1.4 Exceptional (taking risks to help people in distress, lack of


Slip (unintentional action where failure involves attention)



.Incorrect operation of controls or equipment

LeftlRight, reversal
Failure to report due to distraction

Lapse (unintentional action where failure involves memory)

9.3.1 Forgetting to report information
9.3.2 Failure to advise Officer on the Watch
9.3.3 Other


Mistake (an intentional action where there is an error in the planning

process; there is no deliberate decision to act against a rule or


Error in judgement
Inappropriate choice of route
Deciding not to pass on information
Failure to' respond appropriately

Underlying factors

10.1.1 Physiological
AlcohoVillegal drug
Prescription medicine
10.1.2 Psychological
Excessive workload
Standards of personal competence
Lack of familiarity or training
Panic and fear
Mental and emotional disorders


10.1.3 Physical
Hearing problem
Visual problem
Less than adequate medical fitness
10.1.4 Other




10.2.1 Equipment not available
10.2.2 Ergonomics
10.2.3 Design failures (other than ergonomics)
10.2.4 Maintenance and repair
10.2.5 Other


10.3.1 Company policy and standing orders
10.3.2 Less than adequate operating procedures and instructions
10.3.3 Management and supervision
10.3.4 Other


10.4.1 Ship movementMleather effects
10.4.2 Noise
10.4.3 Vibration
10.4.4 TemperatureIHumidity
10.4.5 Less than adequate manning
10.4.6 Other

Latent factors

11.1.1 Physiological
Alcohollillegal drugs
Prescriptive medicine
11.1.2 Psychological
Excessive workload
Standards of personal competence
Lack of familiarity or training
Panic and fear
Mental and emotional disorders
11.1.3 Physical
Hearing problem
Visual problem
Injuries and illness
Less than adequate medical fitness
11.1.4 Other



11.2.1 Equipment not available
11.2.2 Ergonomics
11.2.3 Design failures (other than ergonomics)
11.2.4 Maintenance and repair
11.2.5 Other


11.3.1 Company policy and standing orders
11.3.2 Less than adequate operating procedures and instruction
11.3.3 Management and supervision
11.3.4 Other


11.4.1 Ship movementMleather effects
11.4.2 Noise
11.4.3 Vibration
11.4.4 TemperatureIHumidity
11.4.5 Less than adequate manning
11.4.6 Other

To assist completion of marine casualty analysis, in addition to the information in annexes 1
and 2, the following information is required:

Principle findings and form of casualty investigation:

Action taken:
Findings affecting internationa.1regulations:
Assistance given (SAR operations):


This report is a supplement to the report made by the master in accordance with guidelines
and general principles adopted by the Organization by ~esolutionA.648(16) in case of an
incident involving dangerous goods and marine packaged form on board ships
and in port areas.
The information should be provided in case of:

an accident with loss of life, injury or damage to ship or property; or

an accident, where an unsafe situation, an emergency or loss has occurred

involving dangerous goods in packaged form and marine pollutants.

The information should be provided by the Administration carrying out the investigation, if
necessary in consultation with other parties involved (e.g. authorities of ports of loading,
transit or discharge, etc.) and forwarded to the International Maritime Organization together
with recommendations, if considered necessary, for rectifying any detected deficiencies.
The surnmary and recommendations of any subsequent investigations should also be
reported to the Organization.



Cargo(es) Involved
LIN Number:
IMO Hazard Class:
Name and address of manufacturer, or consignor, or consignee:
Type of packaginglcontainer:
Quantity and condition of goods:
StowagefSecuring arrangements:
Pollution - goods lost overboard (yesfno):
Quantity of goods lost:
If yes: 2.1
Lost goods floated or sank:
Lost goods released from packaging (yesfno):
Brief account of the sequence of events:
Extent of damage:
Emergency response measures taken:
Comments on compliance with applicable convention/recommendation requirements:
Comments on effectiveness of applicable convention/recommendation requirements:
Measures/recommendations to prevent recurrence:
Further investigation (yes/no)l:

Data should be provided orlly if not supplied otherwise.



Statistics of damaged ships and of intact stability casualties are important to the work of the
Organization in respect to improvement of subdivision and intact stability criteria in various
conventions, codes, recommendations, and guidelines. Member Governments are invited to
continue to submit to the Secretariat damage data and intact stability damage data using the
format in this annex.

Damage cards should be completed for decked steel, seagoing ships, 25 metres in
length and over, for all breaches of the hull causing flooding of any compartment
above double bottom (collisions, stranding etc.).


The term "damaged ship" refers to the ship for which the card is being completed,


A sketch showing location of damage and of main transverse bulkheads would be



Depth D should be measured to the bulkhead deck in passenger ships and to the
freeboard deck in non-passenger ships (or to the uppermost completed deck, if
bulkhead or freeboard deck are not specified).


In the case of collision with another ship, it is desirable to fill in damage cards for both


All measurements should be given in metres.


Data marked with an asterisk (*) are the most important.


Damaged Ship
Length between perpendiculars*
Moulded breadth*

Moulded depth*
Draught before damage: amidships

(or fore=

and aft=

Bulkhead (or freeboard) deck




~imensionsand location of damage (see sketch above)

Distance from AP to centre of damage*

Distance from baseline to the lower point of damage

Length of damage

Height of damage

Penetration of damage

(If damage extends above bulkhead (or ,freeboard) deck, additional dimensions should be
given for the part located below this deck, these being marked with suffix ",")


Second ship involved i n collision (to be completed in case of collision between two
Length between perpendiculars*

Moulded breadth*

Moulded depth*

Draught before damage: amidships


(or fore=

and aft=


Damage records to be completed for all sea-going passenger ships, sea-going cargo
ships of 25 metres in length and over, and sea-going fishing vessels of 15 metres in length
and over, in respect of both losses of ships and cases in which dangerous heeling occurred
due to unsatisfactory intact stability, including those cases where loss or heeling of the ship
was due to shifting of cargo.
Depth D should be measured to the bulkhead deck in passenger ships and to thefreeboard deck in non-passenger ships (or to uppermost completed deck, if bulkhead or
freeboard deck is not specified).



The metric system should be used for all measurements.


Data marked with an asterisk (*) are the most important.


The provision of data marked


is optional.

It is desirable to attach a sketch of statical stability curves, drawn for both the below
loading conditions, using the following scales:

20 mm for every 10" angle of inclination.


10 mm (or 20 mm) for every 0.1 metre of righting lever.



For ship in fully loaded

homogenous arrival condition
(with 1OOhstores, fuel, etc.)

General Particulars
Draught (amidships)

Centre of gravity above
moulded base line*
Metacentric height


Distance between the

transverse metacentre
and centre of
Reduction in GM due
to any free surface of
Block coefficient of
fineness of displacement*
Coefficient of fineness
of midship section


Coefficient of fineness
of waterplane
Height of centre of
buoyancy above moulded
base line
Lateral area of ships
profile (including
erections, etc.) exposed
to wind
Distance between centre
of lateral area of ships
profile exposed to wind
and corresponding
Estimated rolling period
(P-S-P) (in seconds) ('IT,

For ship in condition at

time of loss


For ship in f ~ ~ lloaded
homogenous arrival condition
(with 10% stores, fuel, etc.)

General Particulars
Rated amplitude of
roll (maximum)


Angle of heel for

immersion of uppermost
continuous deck
Righting levers (GZ) based
upon -centreof gravity (G)
corrected for any free
surfaces, for the following
angles of heel:*

Maximum righting lever

Angle of maximum


Angle of vanishing


For ship in condition at

time of loss


Lightship Displacement =
Centre of gravity above moulded base line KGo =
Additional data to be supplied if available

Wind and sea (Beaufort scale) at time of casualty

Speed at time of impact, in knots:

Damaged ship v l
Second ship v2


Angle of encounter


Did the ship to which this card refers sink?

If not, give draught after damage
If so, indicate time taken to si* after collision
and manner of sinking

Appropriation of breached compartment(s) (e.g. machinery room, cargo hold, etc.)


Type and quantity of cargo in damagedcompartment, if any


Were there any special circumstances which influenced the results of damage (e.g.
open watertight doors, manholes, sidescuttles, or pipes, fractures, etc.)?


Position of watertight bulkheads in vicinity of damage (distance from AP to each of



Number of compartments flooded


Was there a double bottom in the damaged area?

If so, indicate whether the inner bottom was breached


Any additional information considered useful (details of construction, etc.)



Length between perpendiculars* Lpp =
Breadth moulded* B =

Depth moulded* D =

Draught amidships to assigned loadline or subdivision line d

(or forward

and aft

Service conditions (light or loaded, with approximate percentage of cargo, stores, fuel and

Type of cargo,
-if any


Deck cargo, if any

stowage factor



Quantity of ballast water, if any

Sea and wind conditions at time of casualty: sea*
wind* (Beaufort Scale)
Wind velocity u

Wind pressure p,
Wave height h,

Wave length

Direction of wind relative to sl-lip's head


Direction of waves relative to ship's head

Speed of shipat time of casualty V


Name, length and height of enclosed superstr~~ctures

and deck-houses above the deck to
which D was measured

Bilge keels: width(')

Dept of bar keel, if any

Longitudinal extent(')

Was water trapped on deck? - if so, indicate the extent:


Were all vulnerable openings effectively closed at time of casualty?

Was icing a contributory factor to casualty?

Was the vessel under action of helm at time of casualty?
Were any special instructions relative to this ship in existence, concerning the maintenance
of stability, e.g. ,filling tanks, etc.?

Were any voyage limits andlor weather restrictions imposed for the vessel?

Were any partici~larcircumstances related to the casualty?

Give short description of casualty


Administrations are urged to supply the additional information listed in this annex for all
casualties involving vessel fires.

Were any voyage limits placed on the ship??


Propelling machinery (type, fuel, etc.):


Nature of cargo:


Location of ship1

Was the ship underway or in port?:

If in port, specify the conditions (loading, unloading, under repair, or others):

5. Local conditions1


Time (daylight or darkness):

Wind force (Beaufort scale):
State of sea (and code used):


Part of ship where fire broke out?


Probable cause of fire?


Probable origin of flammable liquids, if applicable:


Description of damage12


No. of persons on board1



Structural fire protection (briefly describe .fire resisting and fire retarding bulkheads,
doors, decks, etc., through the whole of the area affected by fire):

Fire detection method at site of fire



1Data should be provided only if not provided otherwise.
2Data should be given as precisely as possible.



Fixed fire extinguishing installations



Ship's fire extinguishing equipment use (foam, dry chemical, Con, water, steam, etc.)


Effectiveness of action taken by crew to extinguish fire:


Outside assistance given and equipment used (e.g. fire department, other ship, etc.)l


Time taken to fight fire


At the site of fire:

Adjacent areas:

To control:
To extinguish:




Classification (see classification scheme appended to this annex):


1Data should be provided only if not provided otherwise.

2Data should be given as precisely as possible.

This classification system should be used when entering the "Classification" of fire casualty
records (paragraph 19 of annex 6). For the purpose of correct usage of the classification
system the Guidance for preparing the casualty classification is attached at appendix B. 'The
numbering has been kept in consistence with the numbering in MSCICirc.388.

Short international
Coastal sea trade
Inland waters
Not reported

In port - Loading
In port - Unloading
In port - Awaiting departure
In port - Other
Under repair
Not reported


Time at which fire was discovered

Midnight to 0559
0600 to 11 59
1200 to 1759
1800 to 2359
Not reported

Duration of fire
Extinguished within 1 minute
1 - 5 minutes
6 - 10 minutes
11 - 30 minutes
31 - 60 minutes
1 - 6 hours
More than 6 hours
Not reported

Position of outbreak
Cargo spaces
Machinery space of category A
Machinery space other than of category A
Cargo pump room
Service space
Other spaces
Not reported

Combustibles involved
Structural materials
Furnishing and baggage
Ships stores
Dry cargo
Liquid cargo
Liquid fuel
Lubricating oil
Hydraulic oil
Other flammable liquids
.I0 Not reported

Origin of flammable liquid

Burst piping
Leaking valve
Overflow ,from tank
Leaking coupling or flanges
Flexible hose
Leaking gasket
Oil soaked insulation material
Not applicable
.I0 Not reported



Source of ignition
Cigarettes, matches, or similar smoking materials
Open flames other than .1 and .8
Static generation
Electrical other than static charges
Spontaneous combustion
Mechanical fault or breakdown
Burning or welding
Hot exhaust pipe or steam line
Not on vessel concerned
.ll Other
Not reported


Type of protection at space concerned

Fire resisting divisions
Fire mains and hydrants
Inert gas system
Fixed C02 system
Halogenated hydrocarbon system
Foam system
Other fixed extinguishing systems (e.g. automatic sprinkler or steam
Other protection (portable and semi portable extinguishers
Not reported


Means by which fire was detected

Detection system installed and utilized
Detection system installed,-but fire detected by personnel
No fire detection system installed, but fire detected by personnel
Not reported


Fire extinguishing effectiveness

Fire extinguishing equipment adequate
Fire extinguishing equipment not adequate
Fire extinguishing equipment improperly used
Assistance from other ship required
Assistance from shore fire brigade required
Ship abandoned
Not applicable
Not reported


Extent of damage
Slight damage
Extensive damage
lrr~mobilizationof ship due to serious damage
Total constructive loss


Observations pertaining to
Crew training
Stowage requirements




Improper maintenance


The following should be taken into account when preparing the casualty classification for the
purpose of entering the fire casualty record. The numbering has been kept in consistence
with MSCICirc.388.

Service: There should be only one entry for each ship.

Condition: There should be only one entry for each ship. The entries "In port Loading'yand "In port - Unloading" apply only to the time during which transfer operations are
l d entered
taking place: any fire occurring while waiting to begin transfer operations s h o ~ ~be
as "In port - Other."

Time at which fire was discovered: There should be only one entry for each ship.

Duration of fire: -There should be only one entry for each ship.

Position of outbreak: There should be only one entry for each ship. The definition of
the spaces involved should be the same as those given in the latest version of the SOLAS

Combustibles involved: 'There may be more than one entry for each ship.

Origin of flammable liquid: 'There may be more than one entry for each

Source of ignition:There may be more than one entry reported for each ship,
especially if the damage was so severe that two or more likely sources can be identified.
Type of protection at space concerned: There will probably be more than one entry
for each ship. Fixed systems should be entered only if they were in the space on fire; portable
systems and those that use hoses should be entered if they can be brought to bear on the
Means by which the fire was detected: There should be only one entry for each
ship. The principle question is whether the fire detection system, if any, was the first to alert
ship's personnel.

Fire extinguishing effectiveness: There may be more than one entry for each ship.
If the fire is extinguished without fire fighting, as with an explosion that "blows itself out," then
enter "Not applicable."
Extent of damage: There may be more than one entry for each ship. The
"lmmobilization of ship due to serious damage" should also be entered when the propulsion
system is shut down to aid in fire fighting.


Observations: There may be more than one entry for each ship. Favourable
comments as well as unfavourable comments should be noted. This is the most important
part of the casualty report and every effort should be made to record all obsetvations to be
made in paragraph 23 of the fire casualty record.

This questionnaire covers both the existing maritime communications system and the
GMDSS and is intended for use during the latter's transition period (from 1 February 1992 to
its full implementation on 1 February 1999).

The purpose of this questionnaire is to enable the Sub-committee on
Radiocommunications and Search and Rescue to assess the. effectiveness of the maritime
distress and safety systemand to recommend improvementswhere necessary.

Member Governments are urged to corr~pletethe questionnaire in respect of distress
and safety incidents occurring to ships under their flag, adding any other information which,
at their discretion, would provide lessons to be learned concerning the application of the
maritime distress and safety system.
In addition, Member Governments are encouraged to pass on any relevant information
they may possess on casualties concerning foreign ships to the country in which such ships
are registered.


GMDSS sea area or sea areas for which radio equipment was



Date and time of incident (UTC):

Brief description of:

GMDSS sea area:



weather conditions during SAR operations

Description of distress and safety radio communications, including particulars

of the following items:
means of communications (radiotelegraphy, radiotelephony,
INMARSAT, SES, DSC, EPIRB) and frequencies used for:
distress alert by ship:


distress relay by RCC:

SAR Coordinating communications:

: (b)

- .

use of alarm signal


contents of distress message:


RCC(s), ships, coast station or coast earth stations which

acknowledged distress message (state time and position):


language difficulties:


If the ship was abandoned, description of distress radio comm~~nications

location signals from survival craft:


If a satellite EPIRB or EPlRB was used for alerting and/or locating survivors,
give details (frequency, type of activation, etc.) and which LUTICES or coast
station received the alerting signal:


Description of on-scene radiocommunications, including s~.~rface/air



Any unusual, or additional, radiocommunication aspects, apparent

shortco'mings and/or lessons to be learned:


This compilation sheet should be completed and submitted with each maritime accident
investigation report where fatigue has been identified as a contributory factor. The
compilation sheet should indicate the cause of the identified fatigue. See MSClCirc.621 for
guidelines for the investigation of accidents where fatigue may have been a contributing
Fatigue identified in this accident was caused by (Check all factors that apply):

Contractual arrangements
Work and rest periods
Manning levels
Watchkeeping practices
Assignment of duties .Shore ship shore support and communications
Management policy
Voyage planning
Recreational facilities

Ship factors
Level of automation
Reliability of equipment
Motion characteristics
Vibration, heat and noise levels
Quality of working and living environment
Cargo characteristicslrequirements
Ship design

Crew factors
Period on board
Crew composition, cohesiveness, and relationships
Crew competency and quality
Personal problems and condition

External factors
Port conditions
Ice conditions
Density of vessel traffic



The following additional information should be submitted for each incident involving spillage
of 50 tonnes or more of harmful substances. See annexes 1 and 2 of this circular for
information to be submitted on vessel identification and casualty specifics. One copy of the
report should be retained by the reporting State, one copy to be sent to the flag State, and
one copy to be sent to the International Maritime Organization.
This reporting format on Incidental Spillages OF Harmful Substances of 50 Tonnes or More
has been added, as the report is considered necessary when investigating a casualty or an
incident (MARPOL 73/78, articles 8 and 12), however this does not replace the one-line entry
report required by the annual mandatory report under MARPOL 73/78, article 11
(MEPCICirc.318, Part 1).
Part 1
To be completed by the reporting State


Was the date of the incident known or estimated ?


Location of the incident (select one of the following):

in inland waters
in the territorial sea
within the exclusive economic zone
outside the exclusive economic zone,
in international waters


Reporting State:

Report completed by: (Administration and address)


Part 2
Information to be supplied by the reporting State and/or the flag State

Action taken by reporting State:


Response to the spill:

no action
clean-up efforts
salvage efforts
other, i.e.


Legal action:
no action
action to be taken by flag State
action taken by reporting State, i.e.


to prevent recurrence:


Additional information:

Direct Natural Resource Damages

Loss of wildlife
lmpact on birds
lmpact on marine mammals
lmpact on fish
lmpact on other marine life,
including invertebrates
Loss of fisheries
Fin fish
Fish farming
Damage to marine environment
Damage to shore environment
Habitat Degradation
Soft Habitats
(salt marshes, mangroves, mudflats)
Shoreline (Beaches)
Rocky Coasts/Reefs, including coral


Part 3
To be completed by the flag State

Legal action taken by flag State

no action
action taken, i.e.


Code for Investigation of Marine Casualties and Incidents

A 20/Res.849
1 December 1997
Original: ENGLISH

20th session
Agenda item 11




adopted on 27 November 1997


RECALLING article 15(j) of the Convention on the lnternational Maritime Organization
concerning the functions of the Assembly in relation to regulations and guidelines concerning
maritime safety and the prevention and control of marine pollution from ships,
NOTING with concern that, despite the best endeavours of the Organization, casualties and
incidents resulting in loss of life, loss of ships and pollution of the marine environment
continue to occur,
NOTING ALSO that the safety of seafarers and passengers and the protection of the marine
environment can be enhanced by timely and accurate reports identifying the circumstances
and causes of marine casualties and incidents,
NOTING FURTHER the rights and obligations of coastal and flag States under the provisions
of articles 2 and 94 of the United Nations Convention on the Law of the Sea (LINCLOS),
NOTING IN ADDI-TION the responsibilities of flag States under the provisions of the
lnternational Convention for the Safety of Life at Sea, 1974 (regulation 1/21), the lnternational
Convention on Load Lines, 1966 (article 23) and the lnternational Convention for the
Prevention of Pollution from Ships, 1973 (article 12), to conduct casualty investigations and
to supply the Organization with relevant findings,
CONSIDERING the need to ensure that flag States are required, under the aforementioned
conventions, to investigate all cases of serious and very serious casualties,




ACKNOWLEDGING that the investigation and proper analysis of marine casualties and
incidents can lead to greater awareness of casualty causation and result in remedial
measures, including better training, for the purpose of enhancing safety of life at sea and
protection of the marine environment,
RECOGNIZING the need for a code to provide, as far as national laws allow, a standard
approach to marine casualty and incident investigation with the sole purpose of correctly
identifying the causes and underlying causes of casualties and incidents,
RECOGNIZING ALSO the international nature of shipping and the need for co-operation
between Governments having a substantial interest in a marine casualty or incident for the
purpose of determining the circumstances and causes thereof,
HAVING CONSIDERED the recommendations made by the Maritime Safety Committee at
its sixty-eighth session and by the Marine Environment Protection Committee at its fortieth
ADOPTS the Code for the Investigation of Marine Casualties and Incidents set out in
the Annex to the present resolution;
INVITES all Governments concerned to take appropriate measures to give effect to
the Code as soon as possible;
REQUESTS flag States to conduct an investigation into all very serious and serious
marine casualties and to supply the Organization with all relevant findings;

REVOKES resolutions A.173(ES.IV), A.440 (XI) and A.637(16).




This Code recognizes that under IMO conventions each flag State has a duty to
conduct an investigation into any casualty occurring to any of its ships when it judges that
such an investigation may assist in deterrr~iningwhat changes in the present regulations may
be desirable or if such a casualty has produced a major deleterious effect upon the
environment. The Code also takes into account that under the provisions of UNCLOS article
94, a flag State shall cause an inquiry to be held, by or before a suitably qualified person or
persons into certain casualties or incidents of navigation on the high seas. However, the
Code also recognises that where a casualty occurs within the territorial sea or internal waters
of a State, that State has a right, under UNCLOS article 2, to investigate the cause of any
such casualty which might pose a risk to life or to the environment, 'involve the coastal State's
search and rescue authorities, or otherwise affect the coastal State.
The aim of this Code is to promote a common approach to the safety investigation of
marine casualties and incidents, and also to promote co-operation between States in
identifying the contributing factors leading to marine casualties. The result of this common
approach and co-operation will be to aid remedial action and to enhance the safety of
seafarers and passengers and the protection of the marine environment. In achieving these



aims, this Code recognizes the need for mutual respect for national rules and practices and
puts particular emphasis upon co-operation.
By introducing a common approach to marine casualty investigations and the
reporting on such casualties, the international maritime community may be better informed
about the factors which lead up to and cause, or contribute to, marine casualties. This may
be facilitated by:



Clearly defining the purpose of marine casualty investigation and the guiding
principles for its conduct.
Defining a framework for consultation and co-operation between substantially
interested States.
Recognizing that the free flow of information will be promoted if individuals
who are attempting to assist the investigation may be offered a degree of
immunity, both from self-incrimination and from any ensuing risk to their
Establishing a common format for reports to facilitate publication and sharing
of the lessons to be learned.

It is not the purpose of the Code to preclude any other form of investigation, whether
for civil, criminal, administrative, or any other form of action, but to create a marine casualty
investigation process the aim of which is to establish the circumstances relevant to the
casualty, to establish the causal factors, to publicise the causes of the casualty qnd to make
appropriate safety recommendations. Ideally, marine casualty investigation should be
separate from, and independent of, any other form of investigation.


The objective of any marine casualty investigation is to prevent similar casualties in the
future. Investigations identify the circumstances of the casualty under investigation and
establish the causes and contributing factors, by gathering and analysing information and
drawing conclusions. Ideally, it is not the purpose of such investigations to determine liability,
or apportion blame. However, the investigating authority should not refrain from fully reporting
the cause~becausefault or liability may be inferred from the findirlgs.


This Code applies, as far as national laws allow, to the investigation of marine cas~~alties
incidents where either one or more interested States have a substantial interest in a marine
casualty involving a ship under their jurisdiction.



For .the purpose of this Code:


Marne casualtymeans an event that has resulted in any of the following:

the death of, or serious injury to, a person that is caused by, or in connection
with, the operations of a ship; or
the loss of a person from a ship that is caused by, or in connection with, the
operations of a ship; or
the loss, presumed loss or abandonment of a ship; or
material damage to a ship; or
the stranding or disabling of a ship, or the involvement of a ship in a
material damage being caused by, or in connection with, the operation of a
ship; or
damage to the environment brought about by the damage of a ship or ships
being caused by, or in connection with, the operations of a ship or ships.

Very senbus casualty means a casualty to a ship which involves the total loss of the
ship, loss of life or severe pollution.


Serious casualty means a casualty which does not qualify as a very serious casualty
and which involves:
a fire, explosion, grounding, contact, heavy weather damage, ice damage,
hull cracking or suspected hull defect, etc., resulting in;
structural damage rendering the ship unseaworthy, such as penetration of the
hull undenrvater, irr~mobilizationof main engines, extensive accommodation
damage etc.; or
pollution (regardless of quantity); andlor
a breakdown necessitating towage or shore assistance.


Marine incident means an occurrence or event being caused by, or in connection with,
the operations of a ship by which the ship or any person is imperilled, or as a result of which
serious damage to the ship or structure or the environment might be caused.


Causes means actions, omissions, events, existing or pre-existing conditions or a

combination thereof, which led to the casualty or incident.


or in camera conducted for the purpose of casualty prevention which includes the gathering
and analysis of information, the drawing of conclusions, including the identthcation of the
circumstances and the determination of causes and contributing factors and, when
appropriate, the making of safety recommendations.


Marine casualty investigator means a person or persons qualified and appointed to

investigate a casualty, or incident, under procedures laid down in national legislation for the
furtherance of marine safety and protection of the marine environment.


Serious injury means an injury which is sustained by a person in a casualty resulting

in incapacitation for more than 72 hours commencing within seven days from the date of



Sb@means any kind of vessel which is used in navigation by water.


4.1 0 Lead hvestigating State means the State that takes responsibility for the conduct of
the investigation as mutually agreed between the substantially interested States.

. Substantial& interested State means a State:






which is the flag State of a ship that is the subject of an investigation; or

in whose internal waters or territorial sea a marine casualty has occurred; or
where a marine casualty caused, or threatened, serious harm to the
environment of that State, or within those areas over which the State is
entitled to exercise jurisdiction as recognised under international law; or
where the consequences of a marine casualty caused, or threatened, serious
harm to that State or to artificial islands, installations, or structures over which
it is entitled to exercise jurisdiction; or
where, as a result of a casualty, nationals of that State lost their lives or
.- received serious injuries; or
that has at its disposal important information that may be of use to the
investigation; or
that for some other reason establishes an interest that is considered
significant by the lead investigating State.

Conduct of marine casualty investigations

Where an investigation is to be conducted, the following should be taken into

Thorough and unbiased marine casualty investigations are the most effective
way of establishing the circumstances and causes of a casualty.
Only through co-operation between States with a substantial interest can a
full analysis be made of a marine casualty.
Marine casualty investigations should be given the same priority as criminal
or other investigations held to determine responsibility or blame.
Marine casualty investigators should have ready access to relevant safety
information including survey records held by the flag State, the owners, and
classification societies. Access to information should not be barred by reason
of competing investigations.
Effective use should be made of all recorded data, including voyage data
recorders (VDR), if fitted, in the investigation of a marine casualty or marine
incident wherever it occurred. The State conducting the investigation should
arrange for the read-out of the VDR.
Marine casualty investigators should be afforded access to Government
surveyors, coastguard officers, vessel traffic service operators, pilots or other
marine personnel of the respective States.
The investigation should take into account any recorrlmendations or
instruments published by IMO or ILO, in particular those relating to the
human factor, and any other recommendations or instruments adopted by
other relevant international organizations.
Reports of investigations are most effective when released to the shipping
industry and public.


In accordance with 9, other substantially interested States should be invited to be
represented during any such investigation and should be admitted as a party in the
proceedings and have equal standing, rights and access to evidence as the State conducting
the investigation.
Recognizing that any vessel involved in a casualty may continue in service and that a
ship should not be delayed more than is absolutely necessary, the State conducting the
investigation should start the investigation as soon as practicable, without delaying the ship
unreasonably. Other substantially interested States may, by mutual agreement, join the
investigation either immediately or at a later stage.

Responsibility for investigating casualties and incidents

Flag States are encouraged to ensure that investigations are carried out into all
casualties occurring to its ships. All cases of serious and very serious casualties should be
. .

Where a marine casualty or incident occurs within the territorial sea of a State, the flag
and coastal States recognizing the obligations of that State to its citizens and the legal status
of the territorial sea under the provisions of UNCLOS and also recognising the duties placed
on a flag State, the flag and coastal States should co-operate to the maximum extent
possible, and mutually agree which State should take the role of lead investigating State.
Where a marine casualty or incident occurs on the high seas, a flag State should carry
out an investigation into a casualty to, or on, any of its ships. If that casualty is a collision
involving a ship of another flag State, then the States should consult with each other and
agree which will be the lead investigating State and determine the best means of cooperation under this Code. In line with 9.1, if another State is a substantially interested State
by virtue of the nationality of the ship's crew, passengers or other persons, or the location of
the casualty, that State or States should be invited to take part in the investigation.


By fully participatirrg in an investigation conducted by another substantially interested
State, the flag State shall be considered as fulfilling its obligations under UNCLOS article 94,
section 7.
An investigation should be started as soon as practicable after the casualty occurs.
Substan,tially interested States should, by mutual agreement, be allowed to join an
investigation conducted by another substantially interested State at any stage of the

Responsibilities of the lead investigating State

'The lead investigating State should be responsible for:

developing a common strategy for investigating the casualty in liaison with
substantially interested States;
providing the investigator in charge and co-ordinating the investigation;
establishing the investigation parameters based on the laws of the
investigating State and ensuring that the investigation respects those laws;
being the custodian of records of interviews and other evidence gathered by
.the investigation;
preparing the report of the investigation, and obtaining and reflecting the
views of the substantially interested States;
co-ordinating, when applicable, with other agencies conducting other



providing reasonable logistical support; and for

liaison with agencies, organizations and individuals not part of the
investigating team.


Notwithstanding the obligation placed on the master or owners of a ship to inform its
flag State authority of any casualty occurring to the ship, where a casualty or incident occurs
in the internal waters or territorial sea of another State, the coastal State should notify, with a
minimum of delay, the flag state or States of the circumstances and what, if any, action is
proposed by the coastal State.
Following a casualty, the investigating State should inform the other substantially
interested States, either through the Consular Office in that State or by contacting the
relevant authorities listed in MSCICirc.7811 MEPC.GICirc.2. That State and the other
substantially interested States should consult, at the earliest opportunity, on the conduct of
the investigation and to determine details of co-operation.
Nothing should prejudice the right of any State to conduct its own separate
investigation into a marine casualty occurring within its jurisdiction according to its own
legislation. Ideally, if more than one State desires to conduct an investigation of its own, the
procedures recorr~mendedby this Code should be followed, and those States should coordinate the timing of such investigations to avoid conflicting demands upon witnesses and
access to evidence.


Where two or more States have agreed to co-operate and have agreed the
procedures for a marine casualty investigation, the State conducting the investigation should
invite representatives of other substantially interested States to take part in the investigation
and, consistent with the purpose of this Code, allow such representatives to:


question witnesses;
view and examine evidence and take copies of documentation;
produce witnesses or other evidence;
make submissions in respect of the evidence, comment on and have their
views properly reflected in the final report; and
be provided with transcripts, statements and the final report relating to the

States are encouraged to provide for maximum participation in the investigation by all
States with a substantial interest in the marine casualty.
The flag State of a ship involved in a marine casualty should help to facilitate the
availability of the crew to the investigation and encourage the crew to co-operate with the
State conducting the investigation.

Disclosure of records

10.1 The State conducting the investigation of a casualty or incident, wherever it has
occurred, should not make the following records, obtained during the conduct of the
investigation, available for purposes other than casualty investigation, unless the appropriate
authority for the administration of justice in that State determines that their disclosure


outweighs any possible adverse domestic and international impact on that or any future
investigation, and the State providing the information authorizes its release:


all statements taken from persons by the investigating a~~thorities

in the
course of the investigation;
all communications between persons having been involved in the operation of
the ship;
medical or private information regarding persons involved in the casualty or
opinions expressed during the conduct of the investigation.

10.2 These records should be included in the final report, or its appendices, only when
pertinent to the analysis of the casualty or incident. Parts of the record not pertinent, and
not included in the final report, should not be disclosed.


Personnel and material resources

Goverr~mentsshould take all necessary steps to ensure that they have available sufficient
means and suitably qualified personnel and material resources to enable them to l~ndertake
casualty investigations.


Issue of marine casualty reports and submission to IMO

12.1 The lead investigating State sho~lldsend a copy of the draft of the final report to all
substantially interested States, inviting their significant and substantiated comments on the
report as soon as possible. If the lead investigating State receives comments within thirty
days, or within some mutually agreed period, it should elther amend the draft final report to
include the substance of the comments, or append the comments to thefinal report. If the
lead investigating State receives no comments after the mutually agreed period has expired,
it should send the final report to theorganization in accordance with applicable requirements
and cause the report to be published.
12.2 By ful1y:participating in an investigation conducted by another substantially interested
State that will be reporting to IMO, the flag State shall be considered as fulfilling its obligations
under IMO conventions.
12.3 Reports, or relevant parts of reports, into the circumstances and causes of a marine
casualty should be completed as quickly as practicable, and be made available to the public
and the shipping industry in order to enhance safety of life at sea and protection of the marine
environment through improved awareness of the factors which combine to cause marine
12.4 Where a substantially interested State disagrees with whole or part of the report
referred to in 12.1 above, it may submit its own report to the Organization.
12.5 The investigating State, upon determining that urgent safety action is needed, may
initiate interim recommendations to the appropriate authority.

Re-opening of investigations

When new evidence relating to any casualty is presented, it should be fully assessed and
referred to other substantially interested States for appropriate input. In the case of new
evidence which may materially alter the determination of the circumstances under which the
marine casualty occurred, and may materially alter the findings in relation to its cause or any
consequential recommendations, States should reconsider their findings.



Contents of reports

14.1 To facilitate the flow of information from casualty investigations, each report should
conform to the basic format outlined in 14.2below.
14.2 Reports should include, wherever possible:





a summary outlining the basic facts of the casualty and stating whether any
deaths, injuries or pollution occurred as a result;
the identity of the Flag State, owners, managers, company and classification
details of the dimensions and engines of any ship involved, together with a
description of the crew, work routine and other relevant matters, such as time
served on the ship;
a narrative detailing the circumstances of the casualty;
analysis and comment which should enable the report to reach logical
conclusions, or findings, establishing all the factors that contributed to the
a section, or sections, analysing and commenting on the causal elements,
including both mechanical and human factors, meeting the requirements of
the IMO casualty data base; and
where appropriate, recommendations with a view to preventing similar


Contact between Administrations

To facilitate implementation of this Code, States should inform the Organization of >the
responsible authorities within their Governments that may be contacted regarding
cooperation in casualty investigations.
Guidelines t o assist investigators in the implementation of the Code

'The contents of this section should be treated as guidelines to assist investigators cooperating in an investigation. Investigators should bear in mind the information required under
the IMO marine casualties and incidents reporting system.
In following this Code, participating investigators must be guided by the requirements of the
legal system of the State in which the investigation is being conducted. In particular, COoperating investigators must be guided by the requirements of national law over issues such

providing formal notification of an investigation to interested parties;

boarding ships and securing documents;
arranging interviews with witnesses;
the presence of legal advisers or other third parties d u r i ~ gan interview.


Information generally required in all cases


Particulars of the ship

Name, IMO number, nationality, port of registry, call sign
Name and address of owners and operators, if applicable, also, if an overseas ship,
of agents
Type of ship
Name and address of charterer, and type of charter
Deadweight, net and gross tonnages, and principal dimensions
Means of propulsion; particulars of engines
When, where and by whom built
Any relevant structural peculiarities
Arnol-lnt of fuel carried, and position of fuel tanks
Radio (type, make)
Radar (number, type, make)
Gyro compass (make, model)
Automatic pilot (make, model)
Electronic positioning equipment (make, model) (GPS, Decca, etc.)
Life saving equipment (dates of surveylexpiry)


Documents to be prodl~ced
(Note: Any documents that may have relevance to the investigation should be
produced Where possible original documents should be retained, othemise
authenticated and datedphotocopies should be taken in accordance with 9.1.2of the
Code. A number of these documents wJl contain details sought under I. I of these
Ship's register
Current statutory certificates
ISM Code certification
Classification society or survey authority certificates
Official log book
Crew list
Crew qualifications (see also 1.4 of these Guidelines)
Deck log book
Port log, log abstract and cargo log book
Engine movement book
Engine-room log book
Data logger print-out
Course recorder chart
Echo sounder chart
Oil record book
Soundings book
Night order book
Masterls/Chief Engineer's Standing Orders
Company Standing OrdersIOperations Manual
Company Safety Manual
Compass error book or records
Radar log book
Planned maintenance schedules
Repair requisition records


Articles of Agreement
Bar records - daily purchases - voyage receipts, etc.
Records of drug and alcohol tests
Passenger list
Radio log
Ship Reporting records
Voyage Plan
Charts and record of chart corrections
Equipmentlmachinery manufacturer's operationallmaintenance manuals
Any other documentation relevant to the inquiry

Particulars of voyage

Port at which voyage commenced and port at which it was to have ended, with dates
Details of cargo
Last port and date of departure
Draughts (forward, aft and midships) and any list
Port bound for at time of occurrence
Any incident during the voyage that may have a material bearing on the incident, or unusual
occurrence, whether or not it appears to be relevant to the incident
Plan view of ship's layout including cargo spaces, slop tanks, bunkerlfuel lube oil tanks
(diagrams from IOPP Certificate)
Details of cargo, bunkers, fresh water and ballast and consumption

Particulars of personnel involved in incident

Full name
Details of irrjury
Description of accident
Person supervising activity .
First aid or other action on board
Capacity on board
Certificate of CompetencylLicence:
date of issue
issuing countrylauthority
Other Certificates of Competency held
Time spent on vessel concerned
Experience on similar vessels
Experience on other types of vessel
Experience in current capacity
Experience in other ranks
Number of hours spent on duty on that day and the previous days
Number of hours sleep in the 96 hours prior to the incident
Any other factors, on board or personal, that may have affected sleep
Whether smoker, and if so, quantity
Normal alcohol habit
Alcohol consumption immediately prior to incident or in the previous 24 hours
Whether under prescribed medication
Any ingested non-prescribed drugs
Records of drug and alcohol tests


Particulars of sea state, weather and tide


Direction and force of wind

Direction and state of sea and swell
Atmospheric conditions and visibility
State and height of tide
Direction and strength of tidal and other currents, bearing in mind local conditions
Particulars of the incident


Type of incident
Date, time and place of incident
Details of incident and of the events leading up to it and following it
Details of the performance of relevant equipment with special regard to any
Persons on bridge
Persons in engine-room
Whereabouts of the master and chief engineer
Mode of steering (auto or manual)
Extracts from all relevant ship and, if applicable, shore documents including details of
entries in official, bridge, scraplrough and engine-room log books, data log printout,
computer printouts, course and engine speed recorder, radar log, etc.
Details of communications made between vessel and radio stations, SAR centres and
control centres, etc., with transcript of tape recordings where available
Details of any injurieslfatalities
Voyage data recorder information (if fitted) for analysis
Assistance after the incident


If assistance was summoned, what form and by what means

If assistance was offered or given, by whom and of what nature, and whether it was
effective and competent
If assistance was offered and refused, the reason for refusal

Authentication of documents
The master should be asked to authenticate all documents and to sign all copies taken
of documents as being true copies, also to authenticate relevant dates and times.


Engine-room orders
In all cases where a collision or a stranding is the subject of an investigation, and the
movements of the engine are involved, the master or officer on watch and other
persons in a position to speak with knowledge are to be asked whether the orders to
the engine-room were promptly carried out. If there is any doubt on the matter, the
investigator shall refer to it in his report.


External sources of information

Investigators should consider independent corroborating information from external
sources such as radar or voice recordings from vessel traffic systems, shore radar and
radio surveillance systems, marine rescue co-ordination centres, coroners and
medical records.


Additional information required in specific cases

Fire/Explosio n

(In vestigators should bear th mind the lMO Flie Casua/v

How was the ship alerted to the fire?
How was the individual alerted to the fire?
Where did it start?
How did it start (if known)?
What was the immediate action taken?
Condition of fire-fighting equipment, supported by dates of s~.~rvey/examination
Extinguishers available:
Type available in the vicinity
Types available on the ship
Types used
Hoses availablelused
Pumps availablelused
Was water immediately available?
Were air vents closed off to the space?
What was the nature of the material on fire and surrounding the fire?
Fire retardant specification of bulkheads surrounding the fire
Restrictions caused by (a) smoke, (b) heat, (c) fumes
Freedom of access
Access availability for fire fighting equipment
Preparedness of crew - Frequency, duration, content and locations of fire musters and
Response by land-based fire-fighting brigades



(lnvestlgators should bear in mind the IMO damage cards and

intact stabiliv repofling format)

Local or other special rules for navii~tion
Obstructions, if any, to manoeuvring, e.g. by a third vessel, shallow or narrow waters,
beacon, buoy, etc.
Cirgumstances affecting visibility and audibiliv, e.g. state of the sun, dazzle of shore
lights, strength of wind, ship-board noise and whether any door or window could
obstruct look-out andlor audibility
Geographical plot
Possibilities of interaction
Name, IMO number, nationality and other details of other vessel
For each ship:
Time, position, course and speed (and method by which established), when presence
of other ship first became known
Details of all subsequent alterations of course and speed up to collision by own ship
Bearing, distance and heading of other ship, if sighted visually, time of sighting, and
subsequent alterations
Bearing and distance of other ship, if observed by radar, timing of observations and
subsequent alterations of bearing
If other ship was plotted and by what method (auto-plot, reflection plotter, etc:), and
copy of plot, if available
Check performance of equipment
Course recorder
Lightslday signals carried and operated in ship, and those seen in other ship


Sound signals, including fog signals, made by ship and when, and those heard from
other ship and when
If a listening watch was kept on VHF radio channel 16, or other frequency, and any
messages sent, received or overheard
Number of radars carried on ship, number operational at time of casualty, together with
ranges used on each radar
Whether steering by hand or automatic
Check that steering was operating correctly
Details of look-out
The parts of each ship which first came into contact and the angle between ships at
that time
Nature and extent of damage
Compliance with statutory requirement to give name and nationality to other ship and
to stand by after collision

Details of voyage plan, or evidence of voyage planning
Last accurate position and how obtained
Subsequent opportunities for fixing position or position lines, by celestial or terrestrial
observations, GPS, radio, radar or otherwise, or by lines of soundings and, if not
taken, why not
Chart datum comparison to WGS datum
Subsequent weather and tidal or other currents experienced
Effect on compass of any magnetic cargo, electrical disturbance or local attraction
Radarls in use, respective ranges used, and evidence of radar performance
monitoring and logging
Charts, sailing directions and relevant notices to mariners held, if corrected to date,
and if any warnings they contain had been observed
Depth sounding taken, when and by what means
Tank soundings taken, when and by what means
Draught of ship before grounding and how determined
Position of grounding and how determined
Cause and nature of any engine or steering failure~beforethe grounding
Readiness of anchors, their use and effectiveness
Nature and extent of damage
Action taken, and movements of ship, after grounding
(Note: information as in cases of foundering may also be required)



(Investigators should bear in mind the IMO damage

cards.,andintact stability reporfing format.)

Draught and freeboard on leaving last port and changes consequent upon
consumption of stores and fuel
Freeboard appropriate to zone and date
Loading procedures, hull stresses
Particulars of any alterations to hull or equipment, since survey, and by whom such
alterations sanctioned
Condition of ship, possible effects on seaworthiness
Stability data and when determined
Factors affecting stability, e.g. structural alterations, nature, weight, distribution and


shift of any cargo and ballast, free surface in tanks or of loose water in ship
Subdivision by watertight bulkheads
Position of, and watertight integrity of, hatches, scuttles, ports and other openings
Number and capacity of pumps and their effectiveness; the position of suctions
Cause and nature of water first entering ship
Other circumstances leading up to foundering
Measures taken to prevent foundering
Position where ship foundered and how established ..:
Life-saving appliances provided and used, and any difficulties experienced in their use

Pollution resulting from an incident

(Invest~Qatorsshould bear in mind 1 . reporting of incidental
spillages of liguids, 50 tonnes or more, and reporting of
information from investigation of incidents involving dangerous
goods or marine po//utants in packaged form.,)
Type of pollutant
UN-numbertlM0 hazard class (if applicable)
Type of packaging (if applicable)
Quantity on board
Quantity lost
Method of stowage and securing
Where stowed and quantities in each compartmenVcontainer
Tankstspaces breached
Tankstspaces liable to be breached
Action taken to prevent further loss
Action taken to mitigate pollution
Dispersantlneutraliser used, if any
Restricting boom used, if an.

Securing -of physical evidence

Occasions may arise where physical evidence may be available and which will require
scientific examination. Some examples are oil, painVscale, pieces of equipment and
machinery;-pieces of structure.

3.2 Before removal, such evidence should first be photographed in situ. The sample
should then be photographed on a clear background before being placed in an appropriate
clean container(s), glass bottle, plastic bag, tin container, etc. The container should be sealed
and clearly labelled, showing contents, name of vessel, location from which the evidence was
taken, the date and the name of the investigator. For items of equipment and machinery,
copies of the relevant certificates should be obtained.
3.3 Where paint samples are being taken for identification purposes in collision cases, a
sample of paint from the ship's paint drum should also be obtained if possible.
3.4 Advice should be sought on'the correct container to use. For example, plastic bags
are suitable for paint samples, but are not suitable in investigations of fires where materials
may need to be tested for accelerant, in which case sealable .tin cans are preferred.


Voyage data recorders

Where information from a voyage data recorder (VDR) is available, in the event that the State
conducting the investigation into a casualty or serious incident does not have appropriate
facilities for readout of the VDR, it should seek and use the facilities of another State, giving
consideration to the following:



the capabilities of the readout facility;


the timeliness of the availability of the facility; and


the location of the readout facility.

Other sources of information

Investigators should bear in mind that other Government agencies, such as customs,
quarantine and State Authorities, may have useful information relating to crew lists, the
general condition of the ship, stores lists (including alcohol on board), ship certificates, etc.
Port authorities and independent surveyors may also hold information of use to an
Joined ship
Travel time


Place joined


(X, Time of accident; F, Meal; W, Watch; M, Maintenance work; S, Sleep; C. Cargo watch; R, Recreation,
including time ashore; A, Alcoholic drink)

Personal issues






RECALLING Article 15(j) of the Convention on the International maritime Organization
concerning the functions of the Assembly in relation to regulations and guidelines concerning
maritime safety and the prevention and control of marine pollution from ships,
CONSIDERING that practical advice for the systematic investigation of human factors in
marine casualties and incidents will promote, where appropriate, effective analysis and
preventive action,
RECOGNIZING the need for development of practical guidelines for the investigation of
human factors in marine casualties and incidents,
HAVING CONSIDERED the recommendation made by the Maritime Safety Committee at its
ADOPTS the Guidelines on investigation of human factors in marine casualties and incidents
set out in the Annex to the present resolution;
APPENDS the Guidelines as an annex to the Code for lnvestigation of Marine Casualties and
INVITES Governments concerned to implement the guidelines as soon as practicable, as far
as national laws allow;
AUTHORIZES the Maritime Safety Committee to keep the guidelines under review and to
amend them as necessary.


Introduction - Purpose of the Guidelines

Investigation procedures and techniques


A systematic approach


General consideration


Timing of the investigation

The Occurrence Site
Witness Statements
Background Information
The lnvestigation Sequence
Fact Finding
Conducting Interviews


2.2.8 Selection of lnterviewees On Site(those nearest the incident) Remote from the Occurrence Site

Topics to be covered by the investigator


People factors
Organization on board
Working and living conditions
Ship factors
Shore side management
External influences and environment



2.4.1 Fact-finding and analysis


Safety action

Reporting procedures

Qualifications and training of investigators



'The ILOIIMO Process for Investigating Human Factors

Human Factors Questionnaire
Definitions - Cornmon Human Element Terms
Selected bibliography of UNCLOSIILOIIMO requirements and
recommendations related to investigations of human factors in marine
casualties and incidents.


The purpose of these Guidelines is to provide practical advice for the systerhtic investigation
of human factors in marine casualties and incidents and to allow the development of effective
analysis and preventive action. 'The long term intent is to prevent similar casualties and
incidents in the future.*
Ships operate in a highly dynamic environment; frequently the people on board follow
a set routine of shift work disrupted by arrival at, working in, and sailing from port. This is an
existence which involves living in the place of work for prolonged periods creating a unique
form of working life which almost certainly increases the risk of human error.

'For the purpose of these Guidelines, the term "marine casualties and incidents" includes occupational
accidents resulting in loss of life or serious personal injury.


1.3 Historically, the international maritime community has approached maritime safety
from a predominantly technical perspective. 'The conventional wisdom has been to apply
engineering and technological solutions to promote safety and minimize the consequences
of marine casualties and incidents. Accordingly, safety standards have primarily addressed
ship design and equipment requirements. Despite these technical innovations, significant
marine casualties and incidents have continued to occur.
Analyses of marine casualties and incidents that have occurred over the past 30 years
have prompted the international maritime corr~munityand the various safety regimes
concerned to evolve from an approach which focuses on technical requirements for ship
design and equipment to one winch seeks to recognize and more'-fullyaddress the role of
human factors in maritime safety within the entire marine industry. These general analyses
have indicated that given the involvement of the human in all aspects of marine endeavours
including design, manufacture management, operations and maintenance, almost all marine
casualties and incidents involve human factors.

1.5 One way the maritime community has sought to address the contribution of the human
factor to marine casualties and incidents has been to emphasize the proper training and
certification of ships' crews. It has become increasingly clear, however, that training is only
one aspect of human factors. 'There are other factors which contribute to marine casualties
and incidents which must be understood, investigated and addressed. The following are
examples of these factors relevant to the maritime industry: communication, competence,
culture, experience, fatigue, health, situational awareness, stress and working conditions.
Human factors winch contribute to marine casualties and incidents may be broadly
defined as the acts or omissions intentional or otherwise which adversely affect the proper
functioning of a particular system, or the successful performance of a particular task.
Understanding human factors thus requires the study and analysis of the design of the
equipment; the interaction of the human operator with the equipment and the procedures the
crew and management followed.
It has been recognized that there is a critical need for guidance for accident
investigators which will assist them to identify specific human factors which have contributed
to marine casualties and incidents. There is also a need to provide practical information on
procedures for the systematic collection and analysis of information on
human factors during investigations. These Guidelines seek to fulfil those needs. They
include a list of topics which should be considered by investigators and procedures for
recording and reporting the results.
These Guidelines should result in an increased awareness by all involved in the entire
marine industry of the role human factors play in marine casualties and incidents. This
awareness should lead to proactive measures by the marine community which in turn will
result in the saving of lives, ships, cargo and the protection of the marine environment,
improvements to the lives of marine personnel and more efficient and safer shipping

These Guidelines apply, as far as national laws allow, to the investigation of marine
casualties or incidents where either one or more interested States have a substantial interest
in a marine casualty involving a ship under or within their jurisdiction.





A systematic approach

The following is a process that provides a step-by-step systematic approach for use in the
investigation of human factors. The process is an integration and adaptation of a number of
established human factor frameworks. The process can be applied to any type of marine
casualty and incident and consists of the following steps:


collect occurrence data

determineoccurrence sequence
identify unsafe acts (decisions) and unsafe conditions

and then for each unsafe act (decision):



identify the error type or violation

identify underlying factors
identify potential safety problems and develop safety actions.

This process is detailed in Appendix 1.

A systematic approach to Step 1 is crucial to ensure that critical information is not overlooked
or lost and that a comprehensive analysis is possible.
Step 2 describes a process involved in organizing the data collected in Step 1 to develop a
sequence of events and circumstances.
In Step 3, the information gathered and organized is used to initiate the identification of
occurrence causal factors, i.e., unsafe acts, decisions or conditions. When an unsafe act,
decision or condition is identified, the process shifts to determining the genesis of that
particular act, decision or condition.
Step 4 is initiated for each identified unsafe act or decision in order to specify the type of error
or violation involved.
In Step 5, the focus is placed on uncovering the underlying factors behind the unsafe act
decision or condition. Fundamental to the process is the notion that for each underlying
factor, there may be one or more associated unsafe act, decision or condition. The reexamination of each step of the process emphasizes the iterative nature of this process in
that it may show where further investigation is necessary.
Finally, Step 6 requires the identification of potential safety problems and proposed safety
action based on the identified underlying factors.

General consideration

The purpose
An occurrence may result in serious damage or impact and sometimes all fo~.~r.
of a marine casualty or occurrence safety investigation is to prevent recurrence of sirfilar
occurrences by identifying and recommending remedial action. All minor occurrences of high
potential in terms of credible result should be subject of full investigation. Studies have shown
that occurrences can have many casual factors and that underlying causes often exist remote


from the incident site. Proper identification of such causes requires timely and methodical
investigation, going far beyond the immediate evidence and looking for underlying conditions
which may cause other future occurrences. occurrence investigation should therefore be
seen as a means to identify not only immediate causes, but also failures in the total
management of the operation from policy through to implementation. For this reason,
investigations must be broad enough to meet this overriding criteria.
2.2.1 Timing of the investigation

An investigation should be carried out as soon as possible after an occurrence. The quality
of evidence, particularly that relying on the accuracy of human recollection, can deteriorate
rapidly with time, and delayed investigations are usually not as conclusive as those
performed promptly. A prorrlpt investigation is also a good demonstration of commitment by
all those concerned.
2.2.2 The occurrence site

Where possible, the site of the occurrence should be left unchanged until the investigation
team has inspected it. Where this is not possible, for instance to make essential and
immediate repairs following serious structural damage, the scene should be documented by
photographs, audio visual recordings or sketches or any other relevant means available with
the object of preserving vital eVidence and possibly recreating the circumstances at a later
date. Of particular importance is the recording of the position of individuals at the site, the
condition and position of equipment supervisory instructions, work permits and recording
charts. Damage or failed components should be kept in a secure location to await the arrival
of the investigation team who may require detailed scientific examination of certain key
objects. Such key objects should be carefully marked.
2.2.3 Witness statements

Once the situation in the immediate aftermath of an occurrence has been stabilised and the
threat to people, plant and the environment has been removed, individuals involved should
commit their recollections to paper in the form of a written statement to assist in preserving
their memory of events. In the event that local authorities take over responsibility for the
investigation, the organisationlcompany involved should nominate a focal point to liaise with
the authorities and to assist in assembling the information they require. Where necessary
legal assistance should be provided.
2.2.4 Background information

Appropriate background information should be obtained before the occurrence location. Such
information might include but is not necessarily limited to:
procedures for the type of 'operation involved
- -. -. records of instuctionslbriefings given on the particular job being investigated
location plans
corr~mandstructure, and. persons involved
messages, directions, etc., given from baselheadquarters concerning the work
vessel particulars and plans
any other relevant information that may allow the investigator to understand the
context of the incident


2.2.5 The investigation sequence

The method for fact-finding while conducting an investigation includes but is not necessarily
limited to the following activities:
inspecting the location
gathering or recording physical evidence
interviewing witnesses taking into account cultural and language differences (on-site
and external)
reviewing of documents, procedure and records
conducting specialised studies (as required)
identifying conflicts in evidence
identifying missing information
recording additional factors and possible underlying causes
Following the fact-finding, a typical marine casualty or incident investigation includes
analyses of the facts, conclusions and recommendations.

The objective of this stage of the investigation is to collect as many facts as possible which
may help understanding of the incident and the events surrounding it. The scope of any
investigation can be divided into five areas:


Conditions, actions or omissions for each of these may be identified, which could be factors
contributing to the incident or to subsequent injury damage or loss.
During the initial stages of every investigation, investigators should aim to gather and record
all the facts which may be of interest in determining causes. Investigators should be aware
of the danger of reaching conclusions too early, thereby failing to keep an open mind and
considering the full range of possibilities. With this in mind, it is recommended that the factfinding stage of the investigation process itself be kept separate from the complete analysis
and recommendations, and that a structured
of the collected evidence leading to concl~~sions
methodology he adopted to ensure the effectiveness of that analysis. Having said that, the
analysis may well help to identify missing pieces of evidence, or different lines of enquiry that
may otherwise have gone undetected.
Investigationchecklists can be very useful in the early stages to keep the full range of enquiry
in mind, but they cannot cover all possible aspects of an investigation, neither can they follow
all individual leads bad to basic casual factors. When checklists are used, their limitations
should be clearly understood.
The initial stages of an investigation normally focus on conditions and activities close to the
incident and only primary causes also called "active failures", are usually identified at this
stage. However, conditions or circumstances underlying these causes also called "latent
failures", should also be investigated.


A factor to consider during an investigation is recent change. In many cases it has been found
that some change occurred prior to an occurrence which, combir~ingwith other causal factors
already present, served to initiate the occurrence. Changes in personnel, organisation,
procedures, processes, and equipment should be investigated, particularly the hand-over of
control and instructions, and the communication of information about the change to those
who needed to know.
The effect of work cycles and work related stress could have an impact on individuals'
performance prior to an occurrence. The impact of social and domestic pressures (so-called
error enforcing conditions) related to individual's behaviour should not be overlooked.
Information should be verified wherever possible. Statements made by different witnesses
may conflict and further supporting evidence may be needed. To ensure that all the facts are
uncovered, the broad questions of "who? what? when? where? why? and how?" should be
2.2.7 Conducting interviews

An interview should start with the introduction of the interviewing party, the purpose of the
investigation and of the interview, and the possible future use of the knowledge and material
obtained during the interview. Investigators must be guided by the requirements of national
law regarding the presence of legal advisers or other third parties during an intemiew.
People should be interviewed singly and be asked to go step-by-step through the events
surrounding the occurrence, describing both their own actions and the actions of others. The
interviewer should take into account the culture and language of the interviewee.
Notwithstanding any previously made written statements the value of a witness's statement
can be greatly influenced by the style of the interviewer, whose main task is to listen to the
witness's story and not to influence him.
If the investigation is a team effort care must be taken not to make a witness feel intimidated
by too many interviewers. Experience has shown that interviews can be effectively conducted
by two interviewers and if appropriate, the witness could be accorr~paniedby an independent
"friend". It should be remembered that an investigation team is often seen in a prosecuting role, and
there may be reluctance to talk freely if people think they may incriminate themselves or their
colleagues. An investigator is not in the position to give immunity in return for evidence, but
must try to convince interviewees of the purpose of the investigation and the need for
In addition to requiring both patience and understanding, successful interviewing requires the
existence of a "no-blame" atmosphere in which the witness can be made to feel comfortable
and is encouraged to tell the truth. It is not the role of the interviewer, or indeed the
investigation team, to apportion blame. Their role is to establish the facts and to establish why
the occurrence happened.
At the end of an interview the discussion should be summarised to make sure that no
misunderstandings exist. A written record may be made of the interview and this may be
discussed with the witness to clarify any anomalies. Subject to any national law, it may be
possible to provide the interviewee with a copy of the written record.


2.2.8 Selection of interviewees

Established marine casualty and incident investigation procedures should be taken into
account when determining whom to interview following a marine casualty. Safety concerns
should be paramount in the scheduling of interviews.
The emphasis must always be to get the investigation team to the site of the occurrence as
soon as possible and to interview those most closely involved, which in the marine sense will
always be the ship first. When that is not possible due to external factors such as the
geographical location of the occurrence or other political factors, it may be possible to
nominate a local representative to carry out an interim investigation. From an investigation
management point of view, it shoi~ldstill be possible to start the process by carrying out at
least some of the interviews of individuals ashore.
It may not be possible to speak directly with Port or Pilotage Authorities in some parts of the
world. Where that is so then every effort should be made to obtain at least a transcript of the
Pilot's statement if one is involved. In the event of a collision in enclosed waters, evidence
from the operators of shore based electronic surveillance equipment can be particularly
Thee are no "hard and fast" rules for selecting who to interview and the following is offered
as an exarr~pleonly: On site (those nearest the incident)

Generally it is beneficial .to begin the interview process with the ship management team
including the Master and Chief Engineer who typically can provide an overview of the

First hand witnesses present at the occurrence site at the time of the occurrence itself,
regardless of ranklposition in the organisation;
First hand witnesses present at the occurrence site at the time of the occurrence itself,
but from outside the organisation. Examples c o ~ ~be
l d berthing or mooring assistants
or visiting personnel such as agents or contractors;
First hand witnesses present at the time of the occurrence but not at the occurrence
location itself. Examples could be ship's staff on the bridge of a ship witnessing a
mooring occurrence on the main deck below;
First hand witnesses present at the time of the occurrence but not at the occurrence
location itself and from outside the organization. Examples could be a pilot on the
bridge witnessing a mooring occurrence on the main deck below;
Those not involved with the occurrence itself but involved in the immediate aftermath
of an occurrence, especially those engaged in the recovery process. Examples could
be those involved in damage control, shipboard fire-fighting or first-aid medical
Tug, mooring boat or pilot cutter crews;
Search & Rescue personnel including helicopter crews;
Shore-based fire-fighters;
Jettynerminal Staff;
Other vessels in the immediate vicinity; and
Operators of Vessel Traffic Services (VTS) or Monitoring Systems.


Designated Person under the ISM Code;

Ship Operators ashore;
Technical Superintendents ashore;
Company General Managers ashore;
Specialists Consultants (relevant to the occurrence);
Port State Inspectors;
Flag State Inspectors;
Regulatory authorities;
Classification Societies Representatives;
Safety committee members including am representatives; and
Designers, shipbuilders, manufacturers and repairers.


Topics to be covered by the investigator**



The diagram below shows a number of factors that have a direct or indirect impact on human
behaviour and the potential to perform tasks.

"Appendix 2 provides appropriate areas of inquiry and appendix 3 provides definitions of common human
element terms.


The headings in the diagram are expanded below:

2.3.1 People factors

ability, skills, knowledge (outcome of training and experience)

personality (mental condition, emotional state)
physical condition (medical fitness, drugs and alcohol, and fatigue)
activities prior to accident/occurrence
assigned duties at time of accident/occurrence
actual behaviour at time of accident/occurrence

2.3.2 Organization on board

division of tasks and responsibilities

composition of the crew (nationality/competence)- manning/level '
workload/complexity of tasks
working hourslrest hours
procedures and standing orders
communication (internal and external)
on board management and supervision
organization of on board training and drills
teamwork including resource management
planning (voyages, cargo, maintenance)

2.3.3 Working and living conditions

level of automation
ergonomical design of working, living and recreation places and equipment
adequacy of living conditions
opportunities for recreation
adequacy of food
level of ship motion, vibrations, heat and noise

2.3.4 Ship factors

state of maintenance
equipment (availability, reliability)
cargo characteristics including securing, handling and care-certificates

2.3.5 Shore side management

policy on recruitment
safety policy and philosophy (culture, attitude and trust)
management commitment to safety
scheduling of leave periods
general management policy
port scheduling
contractual and/or industrial arrangements and agreements
assignment of duties
ship-shore communication



External Influences and environment

weather and sea conditions

port and transit conditions (VTS, pilots etc)
traffic density
ice conditions
organizations representing shipowners and seafarers
regulations, surveys and inspections (international, national, port, classification
societies, etc)



Once facts are collected, they need to be analyzed to help establish the sequence of events
in the occurrence, and to draw conclusions about safety deficiencies uncovered by the
investigation. Analysis is a disciplined activity that err~ployslogic and reasoning to build a
bridge between the factual information and the conclusions.

The first step in analysis is to review the factual information to clarify what is relevant, and
what is not, and to ensure the information is complete. Thus, this process can give guidance
to the investigator as to what additional investigation needs to be carried out.
In normal investigation practice, gaps in information that cannot be resolved are usually filled
in by logical extrapolation and reasonable assumptions. Such extrapolation and assumptions
should be identified and a statement of the measure of certainty provided.
Despite best efforts, analysis may not lead to firm conclusions. In these cases, the more likely
hypotheses should be presented.
2.4.1 Fact finding and analysis
After fact finding and analysis it should be possible to give a description of the
occurrence, its background, tuning, and the events leading to it.

The description should include such factual items as:

the weather conditions
the operation(s) involved
the equipment in use, its capabilities, performance and any failures
the location of key personnel and their actions immediately before the incident
the pertinent regulations and instructions
uncontrolled hazards
changes of staff, procedures, equipment or processes that could have contributed to
the occurrence
what safeguards were or were not in place to prevent the incident
response to the occurrence (first-aid, shut-down, fire-fighting, evacuation, search
and rescue)
medical treatment actions taken to mitigate the effects of the occurrence and the
condition of injured parties, particularly if disabling injuries or death ensued
damage control including salvage
inventory of all consequences of the occurrence (injury, loss, damage or
environmental damage)
general ship's condition.


It should also be possible to identify active and underlying factors such as:
operational deviations
design aspects of hull structural failure
defects in resources and equipment
inappropriate use of resources and equipment
relevant personnel skill levels and their application
physiological factors (eg. fatigue, stress alcohol, illegal drugs, prescription medicine)
why safeguards in place were inadequate or failed
role of safety programmes
problems relating to the effectiveness of regulations and instructions
management issues
communication issues.


Safety action

The ultimate goal of a marine investigation is to advance maritime safety and protection of
the marine environment. In the context of these guidelines, this goal is achieved by identifying
safety deficiencies through a systematic investigation of marine casualties and incidents, and
then recommending or effecting change in the maritime system to correct these deficiencies.
In a report that clearly lays out the facts relevant to the occurrence, and then logically
analyzes those facts to draw reasoned conclusions including those relating to human factors,
the required safety action may appear self-evident to the reader.
Recommendedsafety actions in whatever form should clearly identify what needs to be done,
who or what organization is the agent of change, and, where possible, the urgency for

To facilitate the flow of information from casualty investigations, each report should
conform to a basic format as outlined in IMO Assembly resolution A. 849(20) - Code for the
Investigation of Marine Casualties and Incidents.
Reports should be made to IMO in accordance with established procedures"'.
.- *
Persons andlor organizations with a vested interest in a report-'3buld be given the
*-opportunity to comment on the report or relevant parts thereof before the ieport is finalized.

3 6

'The final report should be distributed to relevant parties involved and preferably be
made public.


***MSC/Circ.827, MEPCICirc.333 of 9 December 1997 reports on marine casualties and incidents, inter alia.


Many and varied contributory factors can play a significant part in the events
preceding a marine casualty or incident. The question of who should be charged with the
responsibility for investigating and analyzing human factors therefore becomes important The
skilled marine casualty and incident investigator generally is the person best suited to
conduct all but the most specialized aspects of human factor investigation.
An investigator should have appropriate experience and formal training in marine
casualty investigation. The formal training .should include specific training in the
identification of human factors in marine casualties and incidents.
In some cases, a human factors specialist may be of significant value in the
The following is a process that provides a step-by-step systematic approach for use in the
investigation of human factors. The process is an integration and adaptation of a number of
human factor frameworks - SHEL (Hawkins, 1987) and Reason's (1990) Accident Causation
and generic error-modelling system (GEMS) frameworks, as well as Rasmussen's Taxonomy
of Error (1987).
The process can be applied to both types of occurrence, i.e., accidents and incidents. The
process consists of the following steps:

collect occurrence data

determine occurrence sequence
identify unsafe acts (decisions) and unsafe conditions

and then for each unsafe act (decision)


identify the error type or violation

identify underlying factors
identify potential problems and develop actions

Steps 3 to 5 are useful to the investigation because they facilitate the identification of latent
unsafe conditions. Step 6, the identification of potential safety problems is based extensively
on what factors were identified as underlying factors. At times, an unsafe condition may be a
result of a natural occurrence; in that case the investigator may jurnp from Step 3 to Step 6.
At other times, an unsafe act or decision may result from an unsafe condition which itself was
established bj; a fallible decision in such a case, the investigator should proceed through
Steps 3 to 6.
Step 1 - Collect Occurrence Data
The first step in the human factors investigation process is the collection of work-related
information regarding the personnel, tasks, equipment and environmental conditions involved


in the occurrence. A systematic approach to this step is crucial to ensure that a

comprehensive analysis is possible and that the logistical requirements of collecting,
organizing and maintaining a relevant occurrence related database are met.
For complex systems, where there are numerous interactions between the component
elements, there is constant danger that critical information will be overlooked or lost durirlg
an investigation.
Use of the SHEL model as an organizational tool for the investigator's workplace data
collections'helps avoid downstream problems because:
it takes into consideration all the important work system elements
it promotes the consideration of the interrelationships between the work system
it focuses on the factors which influence human performance by relatirlg all peripheral
elements to the central liveware element
At this step, the process attempts to answer the more simplistic questions concerning "what,
who, and when" and then moves to more complicated questions of "how and why". The
resulting data becomes, for the most part, a collection of events and circumstances
comprised of acts and conditions. Some of these will be of interest as unsafe acts and unsafe
There are four components to the SHEL model:
Liveware - L, Hardware - HI Software - S, Environment - E
The SHEL Model is commonly depicted graphically to display, not or~lythe four components,
but also the relationships, or interfaces, between the Liveware and all the other.components.
Figure 1 attempts to portray the fact that the match or mismatch of the interfaces is just as
important as the characteristics of the block themselves. A mismatch can be a source of
human error and identification of a mismatch may be the identification of a safety deficiency
in the system. Figure 2 also depicts how this model can be applied to a complex system
where multiple liveware, hardware, software and environmental element exist.

Figure 1.
(Adapted from Hawkins,
1987) SHEL Model


Liveware (Central Component)

The most valuable and flexible component; in the system is the human element, the
Liveware, placed at the centre of the model. Each person brings his or her own capabilities
and limitations, be they physical, physiological, psychological, or psychosocial. This
component can be applied to any person involved with the operation or in support of the
operation. The person under consideration interacts directly with each one of the four other
elements. 'The person and each interaction, or interface, constitute potential areas of human
performance investigation.
Liveware (Peripheral)
The peripheral Liveware refers to the system's human-human interactions, including such
factors as management, supervision, crew interactions and communications.
Hardware refers to the equipment part of a transportation system. It includes the design of
workstations, displays, controls, seats, etc.
Software is the non-physical part of the system including organizational policies, procedures,
manuals, checklist layout, charts, maps, advisories, and increasingly, computer programs.
Environment includes the internal and external climate, temperature, visibility, vibration, noise
and other factors which constitute the conditions within which people are working. Sometimes
the broad political and econorr~icconstraints under which the system operates are included
in this element. The regulatory climate is a part of the environment inasmuch as it affects
communications, decision making, control, and coordination.
Step 2

- Determine Occurrence Sequence

As the investigator moves to addressing questions of "how and why", there is a need to link
the data identified in the first step of the process. Reason's (1990) model of accident
causation, utilizing a production framework, can be used by an investigator as a guide to
developing an occurrence sequence. As well, Reason's model facilitates further organization
of the work system data collected using the SHEL model, and an improved understanding of
their influence on human performance. The occurrence sequence is developed by arranging
the information regarding occurrence events and circumstances around one of five
production elements, i.e., decision makers, /he management, preconditions, productive
activities, and defences.
The production elements themselves are basically aligned in a temporal context. This
temporal aspect is an important organizing factor since the events and circumstances that
can lead to an accident or incident are not necessarily proximate in time nor in location to the
site of occurrence. By establishing a sequential ordering of the data, Reason's (1990)
concept of active versus latent factors is introduced.


Active factors are the final event or circumstances which led to an occurrence. Their effect is
often immediate because they occur either directly in the system's defences (e.g, disabled
warning system) or at the site of the productive activities (i.e., the integrated activities of the
work system's liveware, software and hardware elements), which would indirectly result in the
breaching of the system's defences (e.g., use of the wrong procedure).
Underwing factors may reside at both the personal and the organizational levels; they may
be present in the conditions that exist within a given work system (referring to the
preconditions element in the model). Examples of underwing hctors include inadequate
regulations, inadequate procedures, insufficient training, high workload and undue time

tn practice, Steps 1 and 2 may not be mutually exclusive. As the investigator begins the data
collection step, it would be only natural that an attempt be made to place the information,
albeit often fragmentary in the preliminary stages of an investigation, into the context of an
occurrence sequence. To facilitate this concurrent activity, the SHEL and Reason models can
be combined as illustrated in Figure 2.


B U U "


Productive activities

Unsafe acts

Limited window of


Psychological precursors
of unsafe acts





Fallible dec~s~ons


latent unsafe
failures and
Active failures

Latent unsafe conditions

Latent unsafe conditions
Latent unsafe conditions

Figure 2.
SHEL and Reason Hybrid Model
The data collected during an investigation (i.e. events and circumstances) can be organized,
using multiple components of the modified SHEL model, into a framework surrounding an
occurrence template (/j7 this case the occurrence scenario), based upon the Reason model.
Causal factors, i e., the unsafe acts/decisions and conditions are thereby identified


Steps 3-5

- An Overview

Steps 3 to 5 are based upon the GEMS framework. The framework provides "pathways" that
lead from the identification of the unsafe actldecision (Step 3) to the identification of what was
erroneous about the action or decision (Step 4) and finally to its placement within a
behavioural context (i.e., a failure mode within a given level of performance) in Step 5. The
GEMS framework illustrated in Figure 3 is particularly useful in exploring hypothetical
reconstructions of the occurrence facts.
Step 3

- Identify Unsafe Acts/Decisions and Conditions

In Step 3 of the process, the investigation takes on a reductionist nature where the
information gathered and organized using the SHEL and Reason frameworks is used to
initiate identification of occurrence causal factors, i.e., unsafe acts/decisions and conditions.
An unsafe a'ct is defined as an error or violation that is committed in the presence of a hazard
or potential unsafe condition. Decisions, where there are no apparent resultant actions but
which have a negative impact on safety, should also be considered as unsafe acts. An unsafe
condition, or hazard as noted above, is an event or circumstance that has the potential to
result in a mishap. There may be several acts, decisions and/or conditions which are potential
unsafe candidates, thus necessitating assessments of the occurrence facts. The SHEL and
Reason hybrid tool (refer to Figure 2) can provide a useful base for conducting such iterative
When an unsafe act, decision or condition is identified, the focus shifts to determining the
genesis of that particular act or condition. Further investigation andlor analysis may reveal
other unsafe acts/decisions or conditions antecedent to the causal factor that was initially
As noted earlier, several unsafe acts and decisions may be identified throughout Steps 1 and
2 of the process. The last unsafe act precipitatingthe occurrence often provides a convenient
starting point for reconstruction of the occurrence. This last act or decision differs from the
others in that it can be viewed as the definitive action or decision which led to the occurrence,
i.e., the last act or decision that made the accident or incident inevitable - the primary cause
of the initial event. Although it is usually an active failure the last unsafe act or decision can
be embedded in a latent unsafe condition, such as a flawed design decision which led to a
system faiiure.
Step 4

- Identify Error or Violation Type

This portion of the process, i.e., Step 4, is initiated for each unsafe actldecision by posing the
simple question, "what is erroneous or wrong about the action or decision that eventually
made it unsafe?"
The identification of the type of error or violation involves two sub-steps (see Figure 3):



Basic Error types


Unintentional Action


Unsafe Act
& Declslon




I -




;, -



& Knowledge-based

Routine & Exceptional

Figure 3.
The GEMS Framework
(Adapted from Reason, 1990)

Unintentional or Intentional Action

First it is necessary to determine whether the error. or violation was an unintentional or

intentional action. "Did the person intend action"? If the answer to that question is no, then it
is an unintentional action. Unintentional actions are actions that do not go as planned; these
are errors in execution.
If the answer to the question "Did the person intend the action"? is yes, then the action is
intentional. Intentional actions are actions that are carried out as planned but the actions are
inappropriate; these are errors in planning.

Error Type or Violation

The second subset is the selection of the error type or violation that best describes the failure,
keeping in mind the decision regarding intentionality. There are fol-~r-potential
categories, i.e., s//;o,lapse, mistake and violation. A s@is an unintentional action where the
failure involves attention. These are errors in execution. A lapse is an unintentional action
where the failure involves memory. These are also errors in execution. A mistake is an
intentional action, but there is no deliberate decision to act against a rule or plan. These are
errors in planning. A violation is a planning failure where a deliberate decision to act against
a rule or plan has been made. Routine violations occur every day as people regularly modify
or do not strictly comply with work procedures, often because of poorly designed or defined
work practices. In contrast, an exceptional violation tends to be a one-time breach of a work
practice, such as where safety regulations are deliberately ignored to carry out a task. Even
so, the goal was not to commit a malevolent act but just to get the job done.
Step 5 - Identify Underlying Factors
The designation of separate activities implied by Steps 4 and 5 may be somewhat arbitrary
in terms of what actually occurs when an investigator attempts to reveal the relationship
between the occurrence errorslviolations and the behaviours that lead to them. In simplest
terms, a behaviour consists of a decision and an action or movement. In Step 3, the action
or decision (i.e., unsafe act or decision) was identified. In Step 4, what was erroneous


regarding that action or decision was revealed. In Step 5, the focus is now placed on
uncovering the underlying causes behind the act or decision of an individual or group. To do
so it is important to determine if there were any factors in the work system that may have
facilitated the expression of the given failure mode (and hence the errorlviolation and the
unsafe act). These factors have been termed underlying factors. These factors can be found
by examining the work system information collected and organized using the SHEL or
Reason frameworks in Steps 1 and 2. The re-examination of these data again emphasizes
the iterative nature of this investigative process where it may even be deemed necessary to
conduct further investigations into the occurrence.
Step 6

- Identify Potential Safety Problems and develop Safety Actions

'The identification of potential safety problems is based extensively on what factors were
identified as underlying factors. Once again this underscores the importance of the
a systematic approach to Steps 1 and 2 of the process which sets the
foundation- for the subsequent analysis steps. Where appropriate, the potential safety
problems can be further analysed to identify the associated risk to the system and to develop
safety actions.
Edwards, E (1972). Man and machine: Systems for safety. In Proceedngs of the BALPA
Echnical Symposium London.
Hawkins, F.H. (1987). Human factors in flight. Aldershot, UK: Gower Technical Press.
Nagel, D.C. (1988). Human error in aviation operations. In E.L. Weiner and D.C. Nagel (Eds.),
Human factors in aviation (pp. 263-303). San Diego, CA: Academic Press.
Norman, D.A (1981). Categorization of action slips Psychological Review, 88 (I), 1-15.
Norman, D.A (1988). The psychology of everyday things. New York: Basic Books.
Rasmussen, J. (1987). The definition of human error and a taxonomy for technical system
design. In J. Rasmussen, & Duncan, and J. Leplat (Eds.), New technology and human error:
Toronto: John Wiley & Sons.
Reason, J (1990). Human error: New York: Cambridge University Press.
The following questions are designed to aid the investigator while investigating for human
factors. Skilful questioning can help the investigator eliminate irrelevant lines of inquiry and
focus on areas of greater potential significance.
The order in which the questions should be asked will depend on who is being interviewed
and on his or her willingness and ability to describe personal behaviour and .personal
impressions. Also, it may be necessary to verify, cross-check or augment information
received from one person by interviewing others on the same points.


These areas of inquiry can be used in planning interviews. The following questions are not
intended as a checklist, and some may not be relevant in the investigation of a particular

Safety Policy


Does the company have a written safety policy?

Is there a designated person for shipboard safety matters in the company?
When did a company Representative last visit the vessel or when were you last in
contact with the company?
When were you last provided safety training? What was the training and how was it
When was the last emergency drill (e.g., fire, abandon ship, man-overboard, pollution
response, etc) and what did you do during the drill?
Was appropriate personal protective equipment provided and did you use it?
Are you aware of any personal accidents which occurred on board in the period prior
to the accident?




Activities prior to incident


(If the ship was leaving port at the time of the accident). In general, how did you spend
your time while the ship was in port?
(If the ship was approaching port or at sea at the time of the accident) How long has
the ship been on passage since its last port or terminal operation?
What were you doing immediately prior to coming on watch or reporting for duty, and
for how long? Recreational activity? Physical exercise? Sleeping? Reading? Watching
TV? Eating? Paperwork? Travelling to vessel?
Specifically what were you doing approximately 4 hours,......1 hour,..........30
minutes........before the accident?
What evolution was the ship involved in when the accident occurred? What was your
role during that evolution?
Immediately prior to the accident, what were you thinking about?
At any time before the accident, did you have any indication that anyone was tired or
unable to perform their duty?




Duties at the time of accident


Where were you on the ship when the accident occurred?

What specific job or duty were you assigned at the time? By whom? Did you
understand your assignment? Did you receive any conflicting orders?
How often have you performed this job in the past (on the specific ship involved in the


Actual behaviour at time of accident


Precisely where were you located at the time of the accident?

What specific task were you performing at the time of the accident?
Had you at any time since reporting for duty found that you could not concentrate
(focus your attentionlkeep your mind) on a task you were trying to perform?



Trainingleducationlcertificationlprofessional experience

How long have you been assigned to this ship? Have you requested that your
assignment be lengthened or shortened?
How long have you filled your crew position? What other crew positions have you held
on this ship?
How long have you held the certificate indicating your qualifications?
Before being assigned to this ship, did you work on other ships? If so, what crew
positions have you held?
What is the longest time you have been to sea in a single voyage? How long have you
been at sea on this passage? What was your longest single passage?
Physical condition
Were you feeling ill or sick at any time in the 24 hours immediately before the
accident? If so, what symptoms did you have? Did you have a fever, vomit, feel dizzy,
other? Also, did you tell anyone? What do you believe the cause was?
When was the last meal you had prior the accident? What did you eat? Was it
Do you exercise regularly while on board? When did you last exercise (before the
accident)? How long was the session?
Psychological, emotional mental condition and employment conditions
When was the last time you felt cheerful or elated on board the ship, and what were
the circumstances that generated this emotion?
When was the last time you were sad or depressed or dejected, on board the ship?
Why? Did you talk about it with anyone else?
Have you had to make any difficult personal decisions recently? Have you had any
financial or family worries on your mind recently?
Have you been criticized for how you are doing your work lately? By whom? Was it
What was the most stressful situation you had to deal with on the voyage (prior to the
accident)? When did the situation occur? How was it resolved?
What are the contractual arrangements for all crewmembers?
Have there been any complaints or industrial action in the last (12) months?
Workloadlcomplexity of tasks
What is the shipboard organization?
Is the shipboard organization effective?
What is your position in the shipboard organization (i.e., who do you work for, report
to or assign duties to)?
What is the nature of your work? Sedentary? Physically demandirlg?
Was anyone involved in the accident impaired due to heavy workload?
Work-periodlrest-periodlrecreation pattern

What is your normal duty schedule?

Are you a day worker or a watchstander?
What was your duty schedule on the day before the accident and during the week
before the accident?
Were you on overtime at the time of the accident?



How long had you been on duty, or awake performing other work, at the time of the
When was your last period of sleep? How long did it last? How often did you awaken
during your last sleep period? Did you awaken refreshed? If not, what would have
made your sleep period more restful?
How do you normally spend your off-duty time while on board? Play cards? Read?
Listen to music? Watch N?Other?
When was your last extended period of off duty time when
were able to rest?
Relationship with other crewmembers and superiors/subordinates
Who among the crew would you consider to be a friend?
Do you find any members of the crew unpleasant to be with?
Do you have difficulty talking with any of the crewmembers because of language
Have any new crewmembers recently joined the ship? Have you had a chance to get
acquainted with them?
Did you have any argument recently with another crewmember?
In an emergency, would you trust your fellow crewmembers to come to your
Has another crewmember ever offered to take your place on watch or perform a duty
for you to let you get some extra rest?
What was the subject of your last conversation with another crewmember before
reporting for duty (when the accident occurred)?
Have you talked with any other crewmembers since the accident? If so, what was the
subject of your conversation? Have you talked with anyone else about the accident
prior to being interviewed?
Living conditions and shipboard environment
Do you consider your personal area on board the ship to be comfortable? If not, how
would you like it to be improved?
Prior to the accident, did you have any difficulty resting as a result of severe weather,
noise levels, heatlcold, ship's motion, etc.?


Manning levels


Is the manning level sufficient in your opinion for the operation of the vessel?


Master's standing orders



Are there written standing orders to the whole crew complement from the Master?
Did the MasterIChief Engineer provide written or verbal standing orders to the
watchkeeping personnel?
Were the orders in conflict with the company safety policy?


Level of automation/reliability of equipment


In your opinion, was the system reliable?

Were there earlier failures in the system?
Were the failures repaired by the crew or shore-based workers?





Ship design, motionJcargo characteristics


Did you observe anything out of the ordinary on this passage concerning the ship
design, or motion or cargo characteristics?

Questions 16-24



Scheduling of work and rest periods


What is the company's work schedule and relief policy?


Manning level


How is the manning level determined for your fleet?


Watchkeeping practices



Do you require the Master to stand watch?

Do you leave the watchkeeping practices to the discretion of the Master?


Assignment of duties


Do you leave this matter to the Master?


Shore-ship-shore support and communications


How do you support the vessel's Master?


Management policies


Does the company have a written safety policy?


Voyage planning and port call schedules


How-doesthe Master plan the voyages?


Recreational facilities


Are welfarelrecreational services and facilities provided on board?


Contractual and/or industrial arrangements and agreements



What are the contractual agreements for all crewmembers?

Have there been any complaints or industrial action in the last (12) months?


Nationallinternational requirements


Are the managementfMaster complying with the requirements and recommendations

of the applicable international conventions and Flag State regulations?




Human error: A departure from acceptable or desirable practice on the part of an individual
or group of individuals that can result in unacceptable or undesirable results.
Diminished human performance
Emotion: A physiological state of agitation or disturbance which can affect an individual's
normal ability to perform required tasks.

Panic: A sudden overpowering fear that reduces the ability to perform required tasks.
Anxiety: A state of uneasiness and distress about future uncertainties which may reduce the
ability to properly focus on a required task.
Personal problem: A problem which preoccupies the emotions and reduces the ability to
perform required tasks. Examples include physical disabilities, death or illness in the family,
marital and other relationship problems, health concerns, financial problems, anger, or poor
interactions with shipmates.
Mental impairment: Diminished mental ability that can reduce or impede an individual's
normal ability to perform the mental part of required tasks.
Alcohol use: Consumption of alcoholic beverages which diminishes an individual's abilities to
perform required tasks. Examples include drinking on or too close to duty which can impede
an individual's abilities; drunkenness on duty; drinking off duty which results in poor
performance while on duty; and excessive drinking over a longer period of time which results
in a permanent decrease in mental abilities.
Drug use: Use of medicine or a narcotic which affects an individual's abilities to perform
required tasks There are many different effects on mental and physical capabilities that can
result from the use of legal and illegal drugs, from extreme drowsiness to a false sense of
competence to hallucinations. Mental abilities of the user may also be distracted by the
constant need to obtain additional drugs. In addition, individuals may not be aware of the
"side-effects" of legal drugs and may take them while on duty or forget to report taking them.
Inattention: The loss of attention, notice or regard; neglect. Examples include failing to
monitor displays; not maintaining a proper lookout; forgetting to perform an assigned duty.
Inattention may also be the result of other causes such as personal problems, fatigue, drugs,
boredom, or hearing problems.
Injury: Physical damage to the body Which causes a decrease in mental or physical abilities.
Exarr~plesinclude a head injury and injuries such as a smashed finger or a severe burn
where pain causes distraction and a loss of mental ability.
Mental illness: Psychotic or erratic behaviour; depression; hallucinations; unexplainable, or
other forrr~sof abnormal behaviol-~r.


Physical illness: Sickness which produces a decrease in mental or physical abilities but not
generally termed as mental illness. Examples include: the general disability accompanying
colds and flu; hallucinations due to high fever; migraine headaches: seasickness; and even
severe indigestion and exposure to toxic substances.
Diminished Motivation: Lack of will or desire to perform well resulting in a decrease of an
individual's normal performance of required tasks.
Deliberate:misaction: Purposely taking an incorrect action or purposely failing to take the
correct action. Examples include dereliction of duty; refusal to obey commands; sabotage,
theft or ignoring procedures.
Fatigue: A reduction in physical and/or mental capability as the result of physical, mental or
emotional exertion which may impair nearly all physical abilities including: strength; speed;
reaction time; co-ordination; decision making or balance.
Low morale:,A problem with individual or group motivation as shown by reduced willingness,
confidence, or discipline to perform assigned tasks. Examples/causes may include
interpersonal conflict amongst the crew, officers with poor interpersonal skills, lack of a strong
corporate or shipboard safety culture, excessively long tours of duty.
Lack of self-discipline: Inadequate ability of an individual to control personal conduct.
Examples include loss of temper, or unprofessional behaviour.
Visual problem: A reduced visual acuity due to a specific physical disability. Causes may
include eye injury causing total or partial blindness; not wearing prescribed glasses or
contacts; inability to adequately adapt to darkness.
Excessive workload: Diminished physical or mental capability as the result of the sum total of
all the mental and physical tasks a human must perform within a prescribed time resulting in
a diminished job performance.
Marine environment
Hazardous natural

A situation in which the natural environment causes required

tasks to become more difficult than usual. Examples include
storms; high waves; shallow water; severe shoaling; strong
or tides; ice; rocks; submerged wrecks, severe eddies,
ship traffic, wind; fog; mist; rain; snow; sleet; haze; dust; airborne

Poor human factors

engineering design

Poor design of the ships, its subsystems, its environmental

controls, or its human-machine interfaces, which results in an
increased difficulty to perform shipboard tasks. Examples of poor
human factors engineering design include inadeql~atelighting;
excessive noise; excessive vibration; inadequate heating,
cooling, or ventilation systems; hazardous deck stair, ladder,
bulkhead, or work surfaces; inadequate provision for foul
weather or degraded mode operations; inadequate restraints,
guards, or hand-holds; poor workstation orientation in regard to
ship dynamics; poor hull seakeeping characteristics; controls
which allow accidental actuation; illegible or ambiguous cor~trol
markings; illegible or ambiguous displays or display labels; poor


layout, sizing, and colouring of controls and displays; inadequate

design for operational or maintenance access; inadequate
design for safety.
- -

Poor operations

A situation in which individuals or groups of individuals degrade

the shipboard environment making the performance of some
required tasks more difficult. Examples include ship
manoeuvres (e.g. increased speed, change in course, erratic
manoeuvres) impact ship dynamics .causing balance and
restraint difficulties, when personnel performing one task
interfere with those performing another; or where storage of
cargo impedes access or transit.

Poor maintenance

Failure to keep any part of the ship or its equipment in the

condition it was designed to function within a designated lifetime
or operational period, thus degrading the shipboard environment
and making the performance of some required tasks more
difficult. Examples of poor maintenance impacting on required
tasks are inadequate replacement parts and tools to perform
proper maintenance that are the result of a lack of commitment
from management.

Safety administration
lnadequate technical

Not knowing due to inadequate experience and/or training, the

general knowledge which is required for the individual's job on
board. Examples are navigation, seamanship, propulsion
systems, cargo handling, communications, or weather.

lnadequate situational Not knowing, due to inadequate experience, lack of

co-ordination, andlor training, the current status of the ship, its
systems, or its environment. Examples include lack of knowledg
of location, heading or speed or lackof knowledge of status of
ongoing maintenance on board.
Lack of
communication of

Not making use of all the available information sources to

determine current status. This may be the result of a lack of
initiative on the part of the individual or a lack of initiative and/or
cooperation on the part of others. Examples of poor
communication/coordination include: poor communication
between bridge officers, poor communication with pilots, or poor
deck to engine room coordination.

lnadequate knowledge Lack of knowledge resulting from inadequate experience, a

failure to know ship regulations, an inadequate knowledge of
of operations
procedures, inadequate training, and/or being unaware of your
roleltask responsibility. Examples of areas where you might lack
knowledge are: Navigation, seamanship, propulsion systems,
cargo handling, communications, or weather.


lnadequate knowledge Lack of knowledge or understanding of required regulations due

regulationsJstandards to inadequate experience and/or training. Examples of possible
regulations; company policies and standards, national and
international regulations, other port State's maritime regulations,
local jurisdiction regulations, shipboard regulations, cautionary
notices, charts notations, or labelling.
lnadequate knowledge Not knowing due to inadequate experience and/or training the
of ship procedures
shipboard and company policies requiring adequate knowledge
of your own ship's operation. Examples include emergency
procedures, maintenance procedures, administrative procedures,
and safety system procedures.
Unaware of rolettask

lnadequate knowledge of the specific job required of an

individual. Examples include a lack of understanding of
command responsibilities, communications responsibilities,
safety responsibilities, maintenance responsibilities and
emergency responsibilities.

lnadequate language

A lack of primary language skills necessary to communicate and

perform duties as required. This includes total or partial inability
to speak, read or comprehend the primary language and/or othe~
required language sufficiently to understand all shipboard
commands, instructions, procedures, labels, warnings, and

Failure to maintain

Failing to ensure that personnel submit to authority, regulations

and procedures. Examples include: tolerating unqualified or inept
personnel, not enforcing regulations and procedures, tolerating
inappropriate insubordination.

Failure of command

Mistakes in giving commands. Examples of faulty command

include: proper command not given, proper command not given
at the appropriate time or out of sequence with other commands,
incorrect commands, conflicting commands.


lnadequate oversight of activities of personnel under an

individual's supervision. Examples of faulty supervision include:
not checking to see that a job was performed in a timely and
correct manner, not providing proper resources to problems
brought to the attention of supervising individual, unequal
treatment of personnel.

co-ordinator or

Failure to communicate and coordinate to address issues,

problems, and tasks both aboard ship and ashore. Examples
include: poor communication between bridge officers, poor
communication with pilots, poor communication with home office
poor deck to engine room coordination.


management of
physical resources

Poor management of physical resources which ensure that

people have the tools, equipment, supplies, facilities, food, water,
fuel, etc to perform ,their required tasks. Examples of faulty
management of physical resources include: absence of physical
resources, shortage of physical resources, inappropriate physical
resources, physical resources stored improperly, physical
resource difficult to obtain when needed.

lnadequate manning

Failing to ensure that all required tasks aboard ship can be

properly performed and that adequate personnel of the proper
skill level, physical and mental ability, experience, certification,
and inclination to properly perform those tasks.

lnadequate manpower Not assigning and assuring availability of adequate personnel

with appropriate skill levels to a ship, or to a specific task aboard
the ship to ensure safe and efficient operation.
Poor job design

Specifying job or task reql-~irements

which are unreasonable,
inefficient, impossible, excessive, or impractical. Ex~mples
include: excessive watch duration or frequency, requiring a singl
person to simultaneously monitor displays that are spastically
separated, requiring exposure to hazardous materials without
proper protective gear.

Poor regulations,
Any problem with standards, regulations, policies, procedures, or
policies, procedures, practices. For example: standards, regulations, policies,
or practices
procedures, or practices may be conflicting, inaccurate,
inadequate, do not provide sufficient detail, or outdated.

Misapplication of
The application of standards, regulations, policies, procedures,
good regulations,
or practices at an incorrect time or in an inappropriate
policies, procedures, circumstance
or practices
Mental action
Lack of situational awareness: An incorrect understanding of the current situation winch can
lead to a faulty hypothesis regarding a future situation or a situation which is based upon
incorrect beliefs leading to compounded errors that can substantially increase the risk to the
ship. Examples include arriving at a hypothesis without confirmation, of which direction an
oncoming ship will steer, incorrect interpretation of alarms on board ship (e.g. seawater
contamination of a fuel system during high seas).

Lack of perception: When an individual does not properly understand that a problem or
situation exists. Examples include misreading a dial, mishearing a command,
misunderstanding a garbled radio message, thinking you smell engine oil when it's actually
crude, not noticing a list to starboard, overestimating the distance to the dock.
Incorrect recognition: The misdiagnosis of a particular situation or problem once it has been
perceived. It may be perceived that a problem or situation exists, however, the identi,fication
is incorrect. Examples include misdiagnosis of a sounded alarm that sounds similiar to other
alarms on board ship, incorrect recognition of a visual display alarm on the bridge.


Incorrect identification: The incorrect identification of a problem or hazard once it has been
recognized that the problem or hazard exists. The alarms on a display panel may have
identified a particular hazard to the ship (e.g. low fuel oil pressure), however, the individual
may have misinterpreted the alarm and identified the problem incorrectly.

Setting Up an Investigation
For a State to exercise power it must have clear jurisdiction. Learning Objective 4 covered a
number of issues of jurisdiction under UNCLOS Article 2 and Article 97.


Type of investigation and Investigator's role in Formal Inquiry

Investigations can be conducted against a breach of the law as a crime, to attribute liability
or blame in a civil court, or to establish why an accident occurred and address the safety
issues. Each form of inquiry is totally legitimate and are a necessary part of a society's
armoury in reacting to accidents.


'The Code for the investigationof Marine Casualties and Incidents course is focused upon the
investigation of ,the safety issues and does not deal with the attribution of liability or blame.
When marine safety was under the spotlight of a select committee of the United Kingdom's
House of Commons in 1860, it was stated:

t ..H is necessary to remember what are the absolute requisites for these inquiries. In
the fist place they must be summaa local and hexpensive. If they are not so they wdl
be oppressive to theparties, they wiibe impracticable to the Government and they will
be ineffectual ... Thus further (and A is very important to mention this) the object of the
inquiry is not so much to punish anyone who may be at fault, as to prevent wrecks in
the future, whether by punishment of and warning agahst negligence, or remedying
anflhhg that is wrong in the way of na v~gation..

Any accident, not just a marine casualty, is "any unintended or unexpected occurrence which
produces hurt or loss"? An accident is usually the result of a set of successive human failings
coinciding with other external elements at a moment of system vulnerability. An accident is
not a crime - although there may, in some cases be some breach of the law. Our safety
investigation is much more concerned with unsafe acts than breaches of legislation..
In many investigation regimes, there have been two levels of investigation, a Preliminary
Invest~gationand a Formal Inquiq As both these procedures were carried out under the
maritime legislation of the State both were "official" activities.
The Code for the investigation of marine casualties and incidents (Res. A.849 (20)) uses the
term oficia/invest@at/bnto cover any investigation or inquiry into a marine casualty.
The term Preminary Investigation would seem to suggest that another investigation would
follow. The Pre/im/ary Invest~gationwas conducted by a technically competent person, or
persons, to advise whether the circumstances of an accident warranted further investigation
by a Formal Inquiry or Court of Marhe Inquiry.
Increasingly administrations have extended the technical investigation into a comprehensive
investigation without reference to a more formal hearing.

1 Farrer, Sir Thomas, Permanent Secretary to the Board of Trade, Report to a Select Committee of the United
Kingdom Parliament, 1860.
2 Fenton v Thorley & Co Ltd [I 9031 AC 443 at 453,per Lord Linley,


Under some of these administrations a two tier system has been retained so that the most
serious incidents can be referred to a judicial type inquiry. However, the main reliance is upon
a competent technical assessment of the circumstances and causes of a marine casualty.
Formallnvestigationsare often quasi judicial with evidence laid before a Board or judge with
examination and cross-examination of witnesses. Flaglcoastal State investigators will
normally provide a technical brief for the administration's legal team. Investigators may be
called as witnesses.
Unless specifically directed by the Board or judge to offer an opinion, investigators should
confine their evidence to matters of fact.
Whatever form of investigation or inquiry is adopted there is a need for a technically expert
individual or team to gather the best evidence available.

~ppointrnentof investigators
When the-report of a casualty is received it is important that the "field investigator(s)19have
the necessary expertise to gather the required information, particularly when the ship itself
remains in service and is not delayed. This will usually involve contact with the master and
ship's crew and it is necessary that the field investigator(s) have the necessary technical
background to ask authoritative and credible questions. For this reason, in most cases, an
Investigator with a nautical or marine engineering background are usually the most suitable
"first strike" field investigators.
Small national administrations may not have a specialised investigation Unit and individuals
may be appointed to investigate on behalf of the administration. The same considerations
hold good under these circumstances. However, where possible persons should hold a
permanent appointment to allow quick initial response to a casualty.

Interested parties
When a casualty occurs a number of organisations will become involved. Three contacts
should be made as soon as possible:
The flag State administration [if the investigating State is the coastal or other State).
- This is usually done through the office of .the Consul General. It may also be useful
to contact the flag State investigation Unit direct.
The ship's owner (Lloyd's List of Ship Owners is a useful source)
The ship's agent who will usually be in a position to forward notification to the ship
and the ship owner
Other bodies will also have an interest which may or may not cor~flictwith an investigation.
These include, but are not limited to:
The port authority (if relevant)
VTS, Pilots
The P&l Club correspondent and legal advisers acting on behalf of the ship owner.
-ihe ship's Classification Society
Cargo interests
Hull underwriters
If the casualty has resulted in fatalities, the investigator(s) may have to liaise with police,
coroners officers etc. It is useful to cultivate a good working relationship with these bodies,
while ensuring .that .the marine investigation procedures are adhered to and not compromised
by other investigations.


The investigation procedures may be governed by a number of overlapping jurisdictions,

particularly where a coastal and flag State elect to exercise their rights and obligations under
Take for example a flag State investigator undertaking an inquiry in a third State. While on the
flag State vessel, flag State law will probably apply. However, if the investigation is extended
to territory lying within the sole jurisdiction of the coastal State, the 'coastal State law will
apply. An investigator will not have authority to question subjects of the coastal or other third
party State. These individuals, such as agents, pilots, VTS operators or even ship managers
or ship owners may be totally outside the flag State's jurisdiction.
They will, however, be within the jurisdiction of the State in which they operate and may well
be subject to the law of that State. Under these circumstances cooperation with the coastal
State (third party State) may allow the investigator access to individuals or information. Such
a State may have legal requirements relating to summonsing people to interview or for
warrants to obtain documents or materials :important to the investigation, only some person
with official standing within that State can apply and execute such formalities.
Parties involved and fair dealing
One of the. key elements in the Code is that investigators and administrations should act fairly
(with natural justice). If a casualtyis to be investigatedthen all parties who may be subject to
inquiry must be informed in advance of the nature of the investigation, its key procedures and
the place and time of any interview. There are no hard and fast rules and much will depend
upon the nature of the casualty and the time available to investigate. The requirement is to
act in a fair and unbiased manner.

Procedures should, where possible, make provision for people being interviewed to be
accompanied by a legal representative or "friendn. Any legal representation should be
confined to ensuring that the investigation is conducted according to the set procedures and
legal representatives should not be permitted to answer on the individuals behalf. Helshe
should be a neutral observer.
If a "friend" accompanies a witness, the "friend" should not be connected with the incident, or
at the very least not someone who will also be interviewed.
P&l Club legal representative will seek to represent the master and crew. It is important that
the Master, or other person who indicates that helshe would like the P&l Club lawyer present,
understands that the P&l Club represents the shipowners interests and these may well be
counter to the individual's own interests. The issue should also be raised with the lawyer. If
an investigator has any misgiving that the presence of a P&l Club lawyer may inhibit the flow
of information the investigator should exclude such an individual and the investigation
procedures should clearly reflect this.
Substantially interested States
Sub-section 4.11 of the Code defines the meaning of " Substantially interested States" and
should be read in conjunction with sections 5.2, 5.3, 7, 8, 9 and 10 of the Code.
Administrations should also have specific legislation covering the powers of the investigator.
It is important that the investigator(s) understand the rules under which they work and that
such rules are available to any person being investigated.


These rules should cover:


-The way in which a ship owner, master or agent is notified of the investigation
The powers to board a ship involved in a casualty
The identity documents required by the investigators
The powers to ask questions and the obligation placed on the ship's staff
'The powers to remove evidence for copying or retention, or testing
-The rights of the ship's crew
'The conditions, if any, under which confidentiality may be provided
What further action may arise out of the investigation process
The right of individuals who may be affected by the report to comment before the
report is finalised
a The reporting process

In not so serious casualties, and in many of the more serious, the ship itself may not be

. seriously damaged and, depending on the powers of the investigating State to detain the
vessel, speed and prompt response are the essence of an effective investigation
When notified of a casualty as much information as possible should be obtained. This can be
done while the decision whether or not to investigate is being made.
Good and thorough preparation enhances the impression the investigator(s) will make on
those involved in the incident and any time that the investigator(s) have to prepare should be
used to the fullest extent. Not all the preparation need be done by the investigator in person.
A non-expert can extract the basic details of a ship and ownerslmanagers from the
appropriate register (such as Lloyd's Register of Shipping and List of Owners and Managers)
before the Investigator(s) leave their base or while helshelthey are in transit to the casualty.
Tidal information, weather, extracts from Tide Tables, Sailing Directions, information from
Guide to Port Entry (Fairplay, Lloyd's) may be compiled by a person other than the
investigator(s). It is all a matter of organisation and pre planning. Planning the Arst day or
days of the field investigation is important.
Field Investigator
The first response investigator should ideally be somebody with either nautical and or marine
engineeringsbackground. Such individuals can empathise and construct realistic hypotheses
on which to start the investigation and the all irrlportant interviews.

The first priority should normally, where possible, be a "site inspection". Access to the ship to
inspect the ship itself and the key operational areas, this includes the accommodation aqd
general standard of the ship. While sometimes unavoidable, it is very difficult to effectively
interview an officer of the watch or a duty engineer about their actions if the investigator(s)
have not visited the bridge or engine room, if they have not seen the scale of the vessel, big
or small.
If the rules under which the invd;tigator(s) operates require documents such as search
warrants or summonses, it is important that they have such documents duly authorised when
boarding the ship. If the ship is many miles from a centre or can only be reached by helicopter
and a master does not cooperate it can be time wasting and expensive to have to return to
complete documentation.


If conducting interviews ashore in an office it may be possible to employ shorthand note

takers and stenographers. However, investigators should bear in mind that language
difficulties and technical terms, particularly if heavily accented, may be difficult for those not
used to the terminology.

It is irrlportant that Field investigators are equipped or have immediate access to basic
equipment. Again this may depend on the standard of evidence acceptable to the
investigating State. Tape recorders and note books for interviews and ii~spectionsof the site,
cameras (both-,video and still). Other equipment, such as new plastic sample bags for
samples which may require scientific analysis (for fire debris screw top tin cans are preferred)
and pens to mark the sample bags. Basic chart work equipment, dividers, parallel rulers,
pencils and relevant charts.
Prompt travel arrangements are important. Is the ship easily accessible or is the hire of a
launch or helicopter required? How does the investigator(s) get to the ship?
Any time possible should be spent in obtaining all possible information and-planning the initial
(which may be the only) interviews. The Investigator(s) should identify the questions that
need to be answered and the essential doci~mentsthat are needed. Planning the interviews
is the most important initial project, based on what is known with certainty.

In a systematic investigation, personnel other than ship's personnel may well be involved as
witnesses. These include port officials, owners representatives, class society surveyors, VTS
operators, pilots, etc.
To cover all the routine questions relating to the ship and crew particulars, a pro forma is
useful. However, to rely solely on a list of pre-determined questions is seldom satisfactory.
Other than the routine questions relating to name, age, qualifications, length of time at seaJon
board etc. it is better to take a more strategic view. This will be dealt with in more detail when
interviewing and interviewing techniques are discussed in part 6. However a useful method
is to construct a matrix where areas of questions can be plotted against the people who may
have relevant information.
This example is parfof a matrix used when investigating a ship that dragged anchor and
grounded, information relating to the anchors from the anchor and cable certificate.

quality of


Ship fixing
method and

ISM code
and ship's

and VTS

2nd Mate
3rd Mate
Watch eng
VTS operator
Port manager
Ship manager













5.10 Exercise
Each participant is to briefly outline histher country's legislation and the powers of an
investigator. Then evaluate how the existing procedures equate with the aims and objectives
of Res. A.949 (20).
5.11 Occupational Health and Safety
It is important that any administration provide effective occupational health and safety
safeguards for investigators. These at the most basic should include boiler suits, hard hats
and safety boots. Safety-glasses, sun glasses may also be required as well as life preservers
and other equipment and clothing.

Investigators should always have current tetanus inoculations and other immunisation as
necessary, such as protection against the various strains of hepatitis. Advice from national
medical authorities may be of assistance in determining what immunisation may be of benefit.
5.12 Media inquiries
Ideally the investigator should not be involved in talking to the media and media contacts
should be handled by a designated person. If such facilities are not available and an
investigator is permitted to talk directly to the media very great care should be taken.

Reporters are skilled at getting people to say more than a person may want to say. It is vital
that the investigator makes no statement that prejudges any aspect of the investigation or
which may be taken as showing some preconception or bias. The media is best handled by
experts, but when an investigator is directed to talk directly to the media helshe should
carefully prepare a factual statement and refuse to be drawn into any speculation or


Always remember that the interviewee may have the information you want. 'The art of
interviewing is proper listening. Never forget the interviewer(s) or investigator@)are human
too. They are subject to the same human factors as the witness. The biggest mistake an
investigator can make is "confirmation bias", seeking or only taking note of information that
fits hislher preconceived theory. Never make assumptions, ASSUMPTIONS make an ass of
you and me.


The Ten Commandments of Interviewing

1. Stop talking
You cannot listen if you are talking.
2. Put the witness at ease
Help witness feel that helshe is free to talk.

3. Show that you want to listen

Look and act interested. Listen to understand what is beiug said rather than concentrating on
your next question. This is difficult as the next question may desert you, but it may return and,
by a full story, may be answered anyway.
4. Remove distractions
Don't doodle, or give the impression you are not listening or that you would rather be

5. Empathise
Try to put yourself in the other person's place.


6. Be patient
Allow plenty of time, do not interrupt.
7. Hold your temper
Any impatience or anger can pass the initiative to the interviewee.
8. Avoid argument and criticism
Causes the witness to become defensive and possibly qualify answers or "clam up".
9. Try and make your question flow responsive to hislher priorities
Shows you are listening and that you understand. Try not to break continuity, you can always
return to other issues later.
This is the first and last command.



Planning and Preparation

What is the aim?

a To establish or confirm what happened
a To understand the involvement of the interviewee in the incident
a To identify the elements and factors (events and conditi'bns)
a To prevent such an accident recurring
Planning, with the aims of the interview in mind:
a Analyse what evidence you already have
a Consider what other evidence you need
a Go through the factors and elements as you understand them
a Draw up a list of those that should be interviewed - who may have information
a Consider venue(s)
a Design a flexible approach

Setting up the interview

The investigator will have to decide on a venue or venues to conduct interviews. This often
involves multiple venues. Investigators may have to interview one or more ship's crews (or
members of such), port officials, administration officials, casual witnesses, etc.
It is invaluable for the investigator(s) to visit the ship and its various control centres to get
some perspective of what those involved saw and did. This gives some perspective of how
long tasks or duties would have taken.
Often, for accurate recall and detail it is best to conduct interviews on the ship (say on the
bridge) which will allow the witnesses to point and demonstrate what went on. Psychologists
have shown that people remember better when at the place where an event occurred.
(Abernathy - 1940 - Feingold - 1914).
A quiet office or venue away from the ship also has its advantages. It avoids interruption. It
may put the irlvestigator on "home ground". TI- is is a matter of judgement and may depend
on how the interview is recorded and how many support staff the investigator(s) may want to
involve, e.glstenographers. In some countries the policy is "to go to the witnesses" and use
a tape recorder.
Non-crew members may be interviewed at their place of work, an investigator's office (as
above), at their home or on some neutral ground. The venue is a matter of judgement
balanced against convenience, usually the witness's convenience.
Who should you interview (bearing in mind your role as a safety investigator)?
a Any person in charge of the operation
a Eye witnesses
a Any person who can add to understanding of the events and conditions
a Any person who can provide evidence or corroborative evidence
Preferably interviews should be conducted after all other evidence has been analysed. In the
real world this is never possible, particularly where a ship may be delayed, and anyway it
takes a reasonable span of time to get all the evidence and evaluate it. As a rule, as many
facts as possible should be gathered and analysed before interviews start and the facts and
analysis updated as the investigation progresses. Remember do not be hide-bound and be
prepared to amend your ideas and analysis.



At interviews three heads are better than two. (Two interviewers and an interviewee.)
6.4 The Witness - General

It is your job to draw the information out of a witness's memory. Most people will tell the truth,
as they see it. Almost all people will tell as much of the truth as possible, although they may
selectively embroider or lie. Also people will very often put the best interpretation on their own

A witness recalling an event is human and subject to all the human factors discussed in
Part 4. As a witness there are three important factors ,that an investigator should be aware of
when assessing the credibility of an event, particularly the fine details:
ACQUISITION - the perception (experience) of the event in which the information
perceived is encoded in the memory system
RETENTION - the period of time that passes between the event and the
eventual recollection of the information
RETRIEVAL - the recall of stored information
Loftus, in considering eyewitnesses wrote:
"When a complex event is experienced some of the features of the expenence are
exfracted fist to be stored and later to be utfized h arrivhg at action decisions. Ear&
on, h the acquisition stage, the obsewer must decide to which aspect of the visual
stimulus he should aftend Our visual environment typ/aI& contains a vast amount of
information, and the proport/bn of ~hformationthat is actual& perceived is very smal7.
Theprocess of decidhg what to aftend to can be broken down hto an even finer s e h s
of decisions, each correspona?ngto where a person wJ1make his next eye fixation.
Once the information associated with an event has been encoded or stored in
memo% some of it may remain there unchanged while some may not. Many t h g s
can happen to a witness dunhg t . crucial stage. The witness may engage in
conversation about the event, or overhear conversations, or read a newspaper story
- all of these can bring about powerful and unexpected changes in the witness's
Final& at any time afler an event a witness may be asked questions about it. At this
point the witness must recreate from long term memory that portion of the event
needed to answer a specific question, This recreation may be based both on
information acquired during the orlginal experienoe and on information acquired
subsequent& In other words, both the acquisition and the retention stages are crucial
to what happens d u ~ h gretrieval. The answer a person gives is based on this



- A layman's guide

Each of the stages of memory acquisition is subject to individual interpretation, perception,

bias and external influences.
People do not tend to remember the routine. But events out of the ordinary are remembered.
The more out of the ordinary or more extraordinary, the more likely a witness is to remember
the incident. A marine casualty is an extraordinary event and an investigator can reasonably
expect witnesses to have some memory of what they saw, heard, smelt or felt. The extent
and accuracy of the memory will depend on a number of interactive factors.


The following are in no particular order:

Was the witness involved or uninvolved?
Exposure time - how much time was there for the interviewee to assirr~ilatewhat is
going on. The short term memory seems to have the ability to store seven plus or
minus two grabs (7*2)of information. This would suggest that the longer one views
(or other senses) the incident the better the recall possibilities.
Individual priorities (Salience) - in a complex incident, not all that is happening or can
be seen (heard smelt, etc) can be taken in by a witness. Experts in what is going on
may focus on certain aspects; laymen on others. They will both have seen the same
event but will recall different aspects - often hard to reconcile.
Frequency - if the same event is seen a number of times recall is better. This is
naturally problematic unless the event is filmedlvideoed.
lntensit); (violence of an event). A really spectacular accident will be easier for
witnesses to recall.
Stress - YerkesIDobson. Performance of memory follows the general rule of
performance. A person directly involved or responsible may become overstressed and
an -interviewer can expect a corresponding decrement in performance. The highly
involved~stressedindividual can be expected to be focused on a particular element of
the incident. Sometimes this element is not a key factor.
Expectation - A witness will often "see" (sense) what helshe thinks he should or wants
to see. (See Elizabeth Loftus - Eye Witness irestimony; pg 36 - on hunting accidents).
There are also cultural expectations, particularly relevant in an international industry
such as shipping. There are also expectations based on a witness's own experience.
"I have been at sea 50 years, this is how it should be done, ergo that is the way it was
Prejudice - Preconceptions and past experiences of people, organisations, machinery
can colour memory.
Post-event information People that shared the same experience will usually talk
about the event. Other people's experiences/observations can (and often are)
transferred to other potential witnesses. Witnesses hearing or reading news items can
have their memory modified either consciously (if this what they say this must be what
I saw) or unconsciously. Here it may be the personality of other witnesses may play'a
part. Also a witness may learn some new information and transfer that to the memory
of what helshe heard or saw. 'This may result in a witness modifying hislher memory
because it conflicts with what was seen.
Time - A person's accurate memory will decrease with time. The memory will literally
"fade" and will be more open to the influences outlined above. Hence it is important to
interview witnesses as soon as possible after the event. It is also important for
witnesses to write down their recollection as soon as possible after the event. This
should not necessarily be a "statement"- this has legal connotations - but a
contemporaneous note of what happened. At sea it may be some while before an
is interviewed. In the case of a formal investigation it may be months or
The way the question is asked - Loftus explains how the way a question is asked and
the words used can influence the answers a witness may give. The use of emotive
words by the interviewer, such as "huge" or "storm" can colour the witness's recall.


Always remember that people may tell you what they genuinely understand to be factual and
truthful, but people have different interpretations and may be mistaken. Also individuals under
threat may collude.
Alternatively a person may convince himlherself that they did not do, see or hear what
actually they did, saw or heard. This is a form of denial and a witness may have convinced
t~imlherselfof the correctness of an incorrect version of events. This is usually self-servirlg,
but can have other motives.


by Cherry Norton
One in 10 people who lie convince themselves that they are telling the truth, raising fears that
some people are immune to lie detector tests and do not show the tell tale signs of a liar.
In two studies of 140 people, Dr Danielle Polage, of the University of Washington, has shown
that people who have a good imagination can convince themselves, after they have been told
to lie, that they are telling the truth.
The findings, presented at the American Psychological Society annual conference in Denver,
Colorado, show that for the majority of people, lying about an event strengthened their
memory of the truth.
However, 10 per cent came to believe that the lied about event was true and subsequently
denied they were lying.
"Lying is a form of imagination in which someone creates an alternative reality," Dr Polage
"For some people the lie becomes so incorporated into their memory that they believe the lie.
"People who make false confessions because they are tired or think they will get off more
lightly, live with the lie for a period of time and have to repeat it.
"These findings show that for a significant proportion of people the lie becomes reality."
People who repeatedly lie about an event, and are then faced with the facts that prove they
are lying, cannot admit it because they believe their version of events.
These people will not display any of the body or facial movements that can betray a liar.
Participants in the study were asked to fill in lifestyle questionnaires.
They were asked whether they had ever been lost in a public place, approached by a scary
stranger in a park, or bullied before they were 10 years old.
A week later they were told to make up a story that would allow them to answer "yes" to an
event they said had not happened to them.
Two weeks later they told their story to a researcher, who asked them detailed questions.
They filled in the questionnaires again one week and five weeks later.
Fifty five per cent had become more confident in the truth of their initial answers, 28 per cent
remained the same and 25 per cent were influenced by their own lies.
Ten per cent became completely convinced that their lied about event was true.
"Time had no impact on their lied about event. In each study, one in 10 people were
absolutely convinced that their lies were true," Dr Polage said.
The Independent.
The bottom line is that an interviewer should suspend belief or disbelief relating to any
witness. Every statement should be checked against the best objective evidence available.
Those things that cannot be checked objectively should be analysed and tested against other
statements. But never forget, the one witness may be "the only one in step".


6.6 Types of Witness

There are nine general classes covering the general run of witnesses likely to be encountered
in a marine investigation. The classes of witness are not mutually exclusive and one witness
may combine a number of the general categories. A biased witness may be hostile, nervous
unwilling, dishonest. One should not forget witnesses, such as husbands and wives who may
be emotionally involved, which one may meet in a yachting accident.
Nothing to gain, nothing to lose. Often the best sort of witness.
One who has a definite interest in the outcome, which can result in the colouring or
embellishment of a siory. It can also lead to a selective memory.
Check story with known facts and against other testimonies.

Consider the two sides in a collision incident.

An oil tanker crew and the local inhabitants.
Ship owner or operator.
Hostile towards the investigator or "authority in general". Often feels under threat.
Maintain complete impartiality, ensure that the witness's rights and your obligations are
explained and the procedures and possible outcomes are explained.
Ensure that the witness understands his obligations. Encourage the presence of a legal
Two general types 1. often combined with a biased witness and/or hostile witness
2. importance, will say what it is believed the investigator wants to hear.

Check against all other facts, if detected during an interview try and establish what the
witness's true interest is.
Nervousness to some degree is quite natural in such a situation. Nervousness can be the
symptom of any number of underlying causes, from a lack of self confidence to guilt. It is
important to put the interviewee at as much ease as possible and avoid flustering himlher.
May be akin to a hostile witness.
May be attributable to loyalty, or a defence mechanism against the outsider.
May be intimidated by the process or frightened of'consequences, threats etc.
Expert must be given sufficient facts to enable an application of expertise.
Know the parameters of the expertise, do not extend it beyond the defined limits.
Do not anticipate outcome and do not expect too much.


"Heard no, saw no, remembered no" etc although it is obvious that the witness must have
seen something of consequence.
If a "negative witness" supports a particular individual (seaman, mate) it can tend to detract
from the third party's evidence.
Always best to use an interpreter.
Sometimes however the interpreter can become involved. In the technical 'field an exact
translation is often difficult to obtain.

The Guts

- Finding out what the interviewee knows, interview techniques

6.7.1 There are a number of means of interviewing:

Question and answer
Cognitive approach
Conversation management
6.7.2 Question and answer
The intervieweriinvestigator prepares a series of questions and answers, a verbal
questionnaire, or even a written pro forma. This is not recommended. It is very rigid and often
based on assumptions or preconceived ideas of the investigator.
6.7.3 The Cognitive approach (Free recall)
The witness is asked to think back and mentally re-live what heishe, heard, saw, did, etc. The
interviewer should allow uninterrupted free recall. This should be followed by at least
(possibly more) one further free recall account. The more an individual talks about an event,
the more that will be remembered. Free recall provides the most accurate memory recall and
taps into the way memory works.
After two or three attempts at free recall the interviewer can then probe various specific areas,
again teasing out memory that may be difficult for the individual to recall. 'There are traps here
as outlined in the previous pages.
The right environment is important. The interviewee is to be asked to "relive" events, however
trivial, this is free recall. To assist free recall:
Put the interviewee at ease
Encourage the interviewee to have confidence in you
Encourage the interviewee to seek help if required
Explain that you require them to relive the incident
Understand that some memories may be painful
Ask them to speak slowly
- assists with note taking or transcribing from tape
Ask them to concentrate
Ask them to work hard
Explain that they will be doing most if not all of the talking and that you will only
interrupt if you must
- you should exhaust their memory to the maximum extent
Explain that you will ask questions after based on what they told you
Assure them you will be listening to every word.


You may have to set the scene, help the interviewee into the right mood - it was dark, the
ship was rolling, hum of the engines, noise of the wind, they had just got out of bed etc.
You may have to deal with emotion, understand what you are putting them through. Emotions
may have to be dealt with to progress the interview.
Listen and do not interrupt. It can ruin their thought processes and your own. Interrupt as a
last resort. Do not prompt in pauses and at the end allow a long pause to ensure that nothing
more occurs to them.
If necessary repeat the process, often, new aspects will be recalled. If necessary, allow
second or third recall. More often than not an individual will start at the incident, or half way
into the incident. This may lead to a full or not so full account of the event. Let them .finish.

If an "involved person" you want to know much more. Ask again, but this time ask them to go
through a period up to and including the incident. Say from the time they woke up.

Keep notes of areas to probe. At the end ask questions to sort out possible ambiguities. Be
sympathetic, do not give the impression of cross-examination.You are there to help them and
they are there to help you. When "probing" it is usefill to recall the interviewee's words "You
said that there was a buoy to starboard, can you tell me anything about its colour." Use "open"
questions "what colour was it" rather than "was it green".

If the interviewee switches, because of sudden recall allow the memory to flow before
returning to the original question.

When finished, you should summarise your understanding of the evidence allowing the
interviewee further recall, to correct points and confirm that you have an understanding of
6.7.4. Management of conversation
The witness is asked to provide an account of what happened. The interviewer divides the
account into sections and then asks the witness to recount each section. As each
secondlsub-account is completed the interviewer probes various aspects of the section

After establishing identity and other basic facts from the witness (ranklrating, position,
experience), open with an initial question - "Tell me about the grounding of Nonsuch" and
allow the interviewee free recall of hislher involvement.
After the one full account the interviewer asks the witness to recount certain stages or
sections of the event. This sectional account is followed by questions to probe specific areas.
As the interviewer completes each area of questioning, helshe should summarise and allow
the interviewee to react. Then link that area with the next area ~ O I Jwish to explore.
At the end provide an overall summary, to allow more information, a change in statement or
a correction of your impression.
6.7.5 General rules under all methods

a Do not use leading questions, inviting a particular answer

a Avoid asking "why'. Although this is often the aim it can sound querulous and
judgmental. Use "what", "when" and "how"


Do not use "negative" questions: "You don't remember which way the wheel was
In questioning start with the general and narrow down by stages to the specific
Do not change from one subject to another until ,that particular area is exhausted
6.7.6 Summary
A questionnaire approach is a last resort only.

The choice between "cognitive" and "conversation management" is for the interviewer to
determine, depending on the factors prevailing at the time. The Management system gives
the interviewer more control. The cognitive approach provides evidence with the least amount
of interference.
Don't forget the IMO guide to examining human factors and the issues you need to establish,
particularly in the area of human factors.
Remember, the investigator is an outsider and is an authority figure. Witnesses will react
Which ever method of interview is adopted the final impressions can be as important as the
first impressions.

. .


Conducting the interview

The quality of information or evidence depends upon:

The investigator's/interviewerlscommunications skills
'The investigator's/interviewerls ability to eliminate or nullify barriers between the
interviewee and the investigator
a 'The relationship established with the interviewee The interviewee will almost certainly have some preconception of the investigator and this is
likely to dictate the way helshe will approach the interview. This may be based on perceptions
or stereotype of public servants, naval men, surveyors etc.
The art of a successful interview is in the listening. People find it easier to talk to a good
The way the investigatorlinterviewer conducts himlherself will either reinforce the witness's
preconception or create an alternative impression.
You should therefore:


Verbal Communications

The most critical part, particularly from the view of evidence. Useful in gauging proficiency
and expertise, but don't forget some people have "the gift of the gab", and developed
language skills that can mask an underlying truth.

Non-verbal Communications

While difficult to offer as evidence, non-verbal prompts or cues can help the interviewer in
assessing the value of what a person says, across the spectrum from absolutely reliable,
through the unsure, to the deliberate lie.
Non-verbal comml.~nicationsare divided into vocaland non-vocal:

Intonation (certainty, doubt, embarrassment)
Speed, pitch (stress anxiety, excitement, relief)
Volume (aggressive, defensive assertive)
Laughter (amusement, nerves, scorn)
Sighing (resignation, exasperation)
Yawning (tired, nervous, lying)
Ums, ahs (buying time, pause for though)
; Rhythm (relaxed,

Body language

- use of hands
- touching face, hair or body
- scratching
- facial expression
- leg movement
- feet


- when seated does hetshe try to move back or closer.


- personal space can be divided into:

1. Pi~blic - >I220 rrlm (4 ft) a space free
2. Personal - between 1220 mm and 460 mm
3. Intimate - 4 6 0 mm (18 )


- sitting face to face may create anxiety that can be

reduced by sitting at right angles.

Appearance and dress
- must not lead to judgements.




The interviewer must exercise control over the interview, while remaining flexible in approach.
Remembering the purposes of a safety investigation, helshe should be:
Open minded

- ie give understanding
- committed to safety
- not project any bias


- often there will be a tussle for control between the interviewer

- no preconceptions or prejudgements, never under-estimate the

- use your experience to understand interviewee's situation, this

helps to relieve anxiety
and interviewee. In safety investigations where you may
interview extensively on other people's "home turf" (say the
master on his own ship) it is harder to impose control than say
at an interview room away from hislher territory. Another tactic
by the interviewee is to show indignation at the need for an

Obvious and repeated attempts to take control is often associatedwith those with greatest
responsibility or most to loose from the investigation.
6.1 0

Initial Contact

First impressions are important to you and the interviewee. Always be courteous and think
about the most appropriate way to proceed. Remerr~berthis is not a criminal case but a safety
inquiry. Also the interviewee has information that you want, so make it easy for the person
involved to give it.



You must be firm but polite.

Gauge your subject
- should you assess whether to use first names, or be more formal
- if they are nervous try to put them at ease show sympathy
- if resentful, try to understand
- if angry try and calm them
Be open, honest and interested
- outline the purpose of the interview and how it will be used
- explain what will happen at the interview
- if necessary seek permission to tape record interview
- that you will take notes
- they may be asked to use charts, demonstrate controls etc.
- explain their rights and obligations
- explain your obligations
- guard against interview fatigue
Ensure that you are all comfortable
- allow tea or coffee breaks or smokos
- this is important for you too
- avoid being disturbed
- avoid (if possible) extraneous noises
- ensure, as far as possible, that all are warmlcool enough.



The role of the interviewer

watch for reactions



don't be remote or officious
identify yourself and your role
gain confidence
use person's name



know what your prejudices are
don't allow them to affect your judgement or tone





anger, tension, defensiveness, stress
open questions (what, when, where, who, why, how)
avoid closed questions (yes or no)
be firm but not aggressive
ask as many questions as necessary
ensure your questions are fully understood
act naturally
keep questions as simple as possible
often best to maintain questions in sequence
use conversational tone
avoid leading questions
avoid multiple questions
avoid "over-talking"

7. - - FINALLY
a - - be aware of non-verbal prompts.
how is the message being said?
are there lie signs?
is there anger?
is there nervousness?
is the individual under stress?

Delving the witness's memory

Human memory is complex. (See 6.4 and 6.5)

When interviewing remember:

a Memory is not total:


it is selective

- it is reconstructed


Memory is organised:
- it tends to order things in handy sequences
- it is modified by habit
- initial recall of memory tends to recall broad outlines

- initial recall may add or miss detail.

Three parts of the memory (This will be expanded in Part 8.)
Psychologists may argue with this model, but for our purposes it is adequate to explain the
memory process.

SHORT TERM memory lasts for a matter of seconds and takes in what your senses

a WORKING memory allows understanding of what your senses perceive and selects
what is to be stored and what is to be rejected leaving the brain free to continue

a LONG TERM memory is where the memory stores the messages from the working
Remember interview techniques, the more attempts somebody has of telling what happened
as "free recall" the more accurate the picture. However, while you may have asked the
interviewee to include all detail, however trivial, helshe may continue to be selective.
Particularly so when talking of routine matters to another specialist. They make assumptions,
consciously or otherwise.
To encourage effort on the part of the witness, change the order, work backwards. Also you
may concentrate on different perspectives, sounds, movement of the ship, smells, colours.

Problems of Recall (See 6.4 and 6.5)

Just because scenes, smells, sensations, words etc have been stored in the long term
memory, it does not mean that a person will be able to retrieve the information.
'The chances are best if what the person is trying to recall is out of the ordinary. Even then
memory may be coloured by perception, attitude, conditioning, trauma and/or self interest.
It is sometimes hard to overcome the problems that a witness may have in recall. It can be
made harder if you allow or make interruptions.
These are very hard to resist in the course of a long interview, or at the end of a series of
interviews where you may already have evidence on the issue that the interviewee is



Closing the interview

Plan the closure of the meeting, allow some time to "wind down":
reassure the witness as to their performance and help
offer thanks for their co-operation and tell them of its importance to the investigation
and ,the importance of the safety investigation
repeat how you intend to use the intei'view
make your closure appropriate to the interview
tell the interviewee that you may wish to contactlor interview himlher again
invite the interviewee to contact you if anything else is remembered

Try not to rush away, leave the impression that you were in control.

After the interview

After each interview ask yourself the questions:

what new evidence do I have?
how does this affect evidence I already have?
whatlwhere next?
do I need to revisit any areas of the investigation?.
And when in more reflective mood:
what did I forget to ask?
how many leading questions did I ask?
how many double questions did I ask?
how could I have improved the interview?

Cultural differences

Investigators should be aware of cultural differences that may be expected and the subcultures that may be on board a ship, particularly those with multi-national crewing.
Issues of status and loss of face may be encountered from time to time. There may be a
tendency for an interviewee to provide answers that helshe thinks the interviewer wants to
hear, or a tendency to agree, just out of politeness. In other cases, the use of English or other
common language may have different meanings or inferences.
Under any of these circumstances an interpreter may prove really useful.

Use of an interpreter

Preferably an interpreter supplied by a government approved interpretation service.

When using an interpreter in interviewing it is important that helshe be fluent in the language
and dialect spoken by the witness. The interpreter must also have a proper command ,of
The interpreter must be able to grasp technical marine terms, and it may be necessary to
arrange a prior meeting andlor have a list of common nautical terms available so that the
interpreter has time to research the appropriate translation. The interpreter must be able to
pass to the witness the information, as well as reflect the attitude and manner of expression


you wish to convey. Further, the interpreter must be able to recognise any idiosyncrasies in
the answers a witness may give and bring them to your attention, along with the reply.
The witness should generally be seated in a chair opposite you with the interpreter in
between but slightly to one side, so that by merely turning hislher head the interpreter directly
faces either you or the witness as the conversation flows. Your questions should be directly
to the witness using the first person. You should not refer to the witness in the third person,
or ask the interpreter to "ask him" or "tell him" anything. You should establish with the
interpreter the length of phrase that you shol-lld use before the interpretation takes place. It is
difficult to stick to, but you must try.
An interpreter should:
1. merely act as a vehicle for accurately interpreting and passing information back and
forth between you and the witness.
2. imitate your voice inflection and gestures as much as possible.

3. not carry on a conversation with the witness, other than directed by you.
4. pass on faithfully everything the witness said, including trivial remarks and

5. not evaluate the conversation him or her self.

Using an interpreter complicates an interview and can often more than double the time. Such
interviews can be successful if they are well planned and controlled
At the conclusion of an interview, when the witness has left, it may be worthwhile asking the
interpreter's assessment of the witness.

Listening traps

When interviewing and not actually talking there are a number of traps that await
This can occur naturally or it can be deliberate if the investigator wants to avoid
listening. It is easy to slacken off mentally and only get part of the message.
While the witness talks the investigator's mind may wander to external maiters,
particular& in routine questioning or where an witness is recounting something that
others have been interviewed about.

Offen the "daydream" is not total and there can be a tendency to d M in and out of the
Similar to "daydreaming", except the investigator takes a cue from the witness and
thinks about a related but different issue.

The investigator takes mental issue with a point made by the witness, risking forgetting
or ignoring subsequent information.



Prior planning
The investigator plans what helshe will do or say next, without really listening to what
the witness is saying.

Any one of these traps can prevent proper listening. It then becomes dificult for the
investigator to pick up the threads of the inteniew and resume listening.

Steps to improve listening skills

To help improve listening skills:


Concentrate. Try to improve listening span.

Do not interrupt.
Atfentive silence. Silence when an witness stops talking often prompts the
witness into thinking that helshe should add more. This can lead to information
that the witness was reluctant to divulge becoming apparent.
Pay atfention to non-verbalcues. Body language and how the witness reacts.
Listen to why the message is being given. Listen for underlying emotions,
attitudes and prejudices.
Remain neutra4 do not
Laten to evemhinng, notjust what the investigator wants to heal
Encourage. The investigator should use non-verbal prompts (nodding the
head, leaning fotward, making positive eye contact).
Tiainyourself to return atfention to the witness if daydreamhg or detouring.
Investigators should, in the first hstance at least, minimise note taking to
increase atfention. Tape recording interviews means that the interviewer is
almost errtirely hands off. However, have a tape with an alarm, to ensure that
concentration is not on the tape but on the interview.





Guide to Interviewing
Metropolitan Police UK
Victorian Police Detective Training Manual
Interviewing Witnesses
Department of Communication and Arts (Aus)
Investigation Principles
Questioning Techniques
Australian Customs Service
Eyewitness Testimony, Elizabeth F Loftus, Harvard University Press 1996

Elements of a Casualty
Any accident can be seen as a chain of events, one link or event logically tied to the next.
An accident is the result of the series of events overloading the chain.

Many of the factors may well have been present in the system for some time, in the case of
a ship the links may extend to the time of its design. It includes the ship's operation over
many years, during which unsafe procedures may have been adopted or management
decisions taken which adversely affected-the ship in the-givenconditionsat the time of the
accident. Links immediately before the casualty are those acts of omission or commission
by those on the ship at the time of the casualty. Following the casualty, there is a period
when ,the ship recovers from the situation, (it refloats after grounding, the fire is put out) or
the ship is lost.



Routine Service
Latent Factors

Build up period

Critical Period









Line Marlagement



Preconditions for
Unsafe Acts


--------------- I
Active Failures


7.1.1 The Lund Institute of Technology and the Karstad Risk Centre

The concept of an elaborate chain of events and the complex nature of underlying latent
factors and unsafe acts has been further elaborated by two Swedish institutions, the Lund
Institute of Technology and the Karstad Risk Centre. Their analysis of the capsize of Herald
of Free Enterprise identifies causal nature of decisions taken in isolation one from the other.
The following diagram shows the general operational areas where decisions were made
without reference to other critical operational areas. Vessel operation and the causal
sources stemming from it are only a small part of the diagram.



Design Stability

equipment load

Truck companies







Change of

Transfer of ship

Crew working



m.v. Herald of Free Enterprise

The following case study has been taken from the Formal Investigation Report of Court
handed down by Mr Justice Sheen in July 1987.


At 1805, om 6 March 1987, the roll on roll off passenger and freight ferry Herald of Free
Enterprise (Herald) sailed from no.12 berth Zeebrugge bound for Dover. On board were the
Master, 79 crew and about 459 passengers. The car decks were loaded with 81 cars, 47
freight vehicles and three other wheeled units. The weather was good with a light easterly
breeze with little sea or swell in open waters.
Herald passed the outer mole at 1824. About four minutes later she capsized. During
the final moments Herald turned rapidly to starboard and the port side took the ground in
shallow water. The vessel came to rest on a heading of 136" with the starboard side above
water. Not less than 150 passengers and 38 crew lost their lives less than 7 cables (1295 m)
from the harbour entrance.-In addition to the dead many were injured.


Heraldwas a triple screw rolro passenger vehicle ferry, built in 1980. She had a gross
tonnage of 7951.44, a length of 131.9 m overall, a moulded breadth of 22.7 m. The three
engines each developed about 6700 kW and drove a variable pitch propeller controlled from
the bridge, giving a service speed of 22 knots. The machinery was designed to provide rapid
acceleration and to make crossings at or near 22 knots. She had shaft driven generators
which provided power for the bow thrusters and a feathering propeller: The ship was
equipped with all the normal navigation equipment.
Heraldwas of all welded steel construction. Above the level of the tank tops there were
8 decks; " A being the uppermost and "H" the lowest, which was below the main deck ("G")
or (bulkhead deck).

"H" deck was divided into passenger accommodation and stores.

"G" deck was enclosed by the superstructure and was a "drive through" vehicle deck
with a single weathertight door at the stern (8.5 m x 4.73 m) and double watertight doors at
the bow and (6.0 m x 4.9 M).

"F" deck was a mezzanine vehicle deck. The ratings accommodation was also on " F
deck. "El1deck was a second vehicle deck, with a suspended vehicle deck designated "D"
"C" deck housed passenger areas and galley. "B" deck housed passenger areas,
officer and rating mess rooms, and further accommodation for the ratings on the starboard
side. "A" deck housed the officer's accommodation and the radio room. The wheelhouse was
situated at a half-deck level between " A and " B decks.
Access between "A" & "G" decks were by way of staircases port and starboard at the
after end, midships and forward. The forward staircases continued below "G" deck to "H"


The vessel corr~pliedwith all the current regulations in respect of construction,
equipment and radio.

The vessel was also provided with two load lines ("C.l" and "C.2"). In the "C.1"
condition she could load to a freeboard of 1110 mm (Maximum draught 5.7 m) and was
permitted to carry a total of 630 persons (passengers and crew). In the "C.2 condition the
minimum freeboard allowed was 1310 mm (max draught 5.5 m) and she could carry a total
of 1400 passengers and crew.

The bow and stern doors

In common with modern rolro ferries, " G deck was enclosed and so that it is included
in the vessel's intact stability it-mustbe weathertight. Surplus water (from weather, leaks, fire
fighting and cleaning) was cleared by 20 x 176 mm scuppers with non-return valves, which
could handle 3-400 tonnes per hour.
Both the weathertight bow and stern doors were hydraulically operated and swung
horizontally about a vertical axis, on radius arms. Such doors are so-metimes referred to as
"clam" doors. Their weight and movement were supported by rubberised rollers. The vessel's
belting at the bow and stern was extended into a horizontal platform shaped to fit into the
shore ramp structure. 'The doors stowed against the ship's side when open and met on the
centre line when closed, sealed by tubular neoprene packing.
The controls for the hydraulic closing rams, clamps and dogs were contained in boxes
adjacent to the after door and close to the inner doors forward. 'The system was protected by
cut out switches and alarm bells that rang when the doors were being operated.


Hera/dwas one of three sister ships designed for the Dover-Calais run. The personnel
on each ship shared a common rostering system. Officers worked 12 hours on and not less
than 24 off. On the ships on the Calais run (about 1 hr) there was a master and three deck
officers and a complement of engineers. On the Zeebrugge run (4.5 hrs) a master and two
deck officers were carried, plus a complement of engineers.

The engine room was manned at all times.


Three sets of ratings and five sets of officers were employed in manning Herald

The bow and stern doors were opened and closed by the assistant bosun. The two
deck officers took it in turns to supervise the loading of the ship and liaising with the shore
terminal. Loading and securing vehicles were undertaken by crew members under the
direction of the bosun.
The "General Instructions" for Herald's operation, compiled as Master's standing
orders stated:
'The oficer loading the main vehicle deck, G Deck, to ensure that the waterfaht and
bow/stern doors are secured when leav~hgporf.

When letting go and departing from port, the mate would be on the bridge with the
master and the second mate would be aft. The bosun would be on the forecastle.



The company standing orders required that:

"0.0.W/Master should be on the bnilge approximately 75minutes before the sh@k
sailing time,


"Ready for sea. Heads of departments are to report to the master immedately they are
aware of any deficiency which is likely to cause their department to be unready for sea
in any respect at sailing time.
In the absence of any such report the master will assume, at the time of sailing, that
the vessel is ready for sea in all respects,
The senior master on Heraldwrote to the company on 22 November 1986 seeking a
more consistent appointment of deck officers to achieve more consistency in the operation of
the ship particularly in relation to safety issues. Officers sent from other classes of ship
tended to bring with them the procedures appropriate to that ship. He wrote on 28 January
"/wish to stress agah that Herald badly needs a permanent complement of good deck
officers. Our problem was outlined in my memo of 22 Novembe< since then the
throughput of offices has increased even furthe< partly because of sickness. During
the period from 1st September 7986 to 28 Januay 7987 a total of 36 deck officers
have been attached to the ship. ... To make maWs worse the vessel has had an
unprecedentedseven changes in sai7hg schedule. The result has been a serious loss
in continuib Sh/;oboard maintenance, safety gear checks, crew training and the
overall smooth running of the vessel have all suffered..."

RoIRo Berths

The ship was designed for operation in Dover and Calais where "E" and "G" decks
were loaded simultaneously. Heraldberthed "bow up" to No. 12 berth Zeebrugge ramp, which
was a single level berth and could not discharge or load the decks simultaneously. 'The ramp
was designed to load at the main deck level, so to load "E" deck the ramp had to be raised
as far as possible and the ship trimmed by the head using two forward tanks (No. 14 deep
tank 268 clm and no. 3 port double bottom 42 clrn). The ballast pump had a capacity of 115120 tonnelhr.
arrival at Zeebrl~ggefilling of the deep tank would start and
Two ~ O I J ~before
discharge of the tank would start as soon as possible after loading " E deck.
The Master and two deck officers stated that it was normal practice to maintain a
reduced speed, so the bow wave would not rise above the forward belting, until the bow tank
had been pumped out. At a combinator setting of 6, the bow wave could be as high as 2 m
above the forward "spade" which formed the belting.

Departure Zeebrugge 6 March

Herald was to make the voyage in t'he "C.1" condition. Calculations after the event
strongly suggested:
9250 tonnes
For'd draught
6.26 m
Aft draught
5.43 m
VCG (Fluid)


The Heraldwas trimmed by the head to load "El' deck which was completed at 1740.
By 1825 about 100 tonnes had been pumped out.

Events leading to the capsize

The assistant bosun opened the bow doors on arrival at Zeebrugge at about 1700. He
then was engaged in cleaning the ship as the passengers disembarked and passengers
started to board. At some time he was released from these-duties and went to his cabin,
where he fell asleep.
At some time after 1740 the second mate went to "G" deck to relieve the mate.
However the mate remained some minutes before leaving the second mate in charge. At
about 1750 the mate heard on his radio some apparent difficulty between the second officer
and the shore staff. There were about 20-25 cars left to load when the second mate heard on
his radio,the mate giving orders. The mate returned to "G" deck and the second mate
assumed-the mate had taken over the loading. The second mate went aft to sailing station.
The crew-had been summoned to sailing stations by tannoy in the crew areas, those on the
car deck went to their respective stations. Once the last car was in position the mate hurried
to the bridge for sailing. The bosun was the last man to leave the deck. He put a chain across
the deck between the leading vehicles and the inner bow door, made sure there were no
passengers left on the deck and went to the forecastle. He stated that as he put the chain
across he saw somebcidy walking towards the door, whom he took to be the assistant bosun.
The inner and outer bow doors were open.

The assistant bosun slept through the call to sailing stations.

With the master and mate on the bridge with a rating on the wheel, Herald left the
berth stern first at 1605, turned to starboard in Zeebrugge Harbour and headed for the outer
mole. Once turned the propeller pitch was set to 4, giving a speed originally estimated at 10
to 12 knots, later experiments with the sister ships put the speed at 14 knots. 'The second
mate joined the Master on the bridge and the mate went to the officers' mess room. As the
vessel passed the mole speed was increased to propeller pitch 6 which accelerated the ship
to give an estimated speed of 18 knots within 4 minutes of clearing the mole.
From the evidence it is apparent that water entered "G" deck in large quantities. The
deck became covered in water and the effect of the free surface of water caused the ship to
become -unstable causing a list of as much as 30" to port. The ship turned rapidly to
starboard, water continuing to enter. The ship seems to have floated for a short period on her
beam ends before settling on the seabed in about 4 m of water.


Some background on Herald's operation.

Ferries work intensely, typical port times to disembark and ernbark a full complement
and to discharge and back load a full complement of vehicles would typically be one hour.
The aim stated in a number of Company minutes was to keep pressure on the mate and
loading officers to sail 15 minutes before schedule. The essential thing was not to miss the
vessel's allotted ramp time, which could cause disruption and loss of cargo and increased
expense. Ships started discharge as soon as the ramp was lowered and sailed as soon as
the ramp was disconnected. The crew routinely went to stations while the ship was still
loading to enable the ship to sail as soon as the last vehicle was on board.


The masters of all the company's vessels made strong complaints about carriage of
passengers in excess of the passenger certificate, possibly condoned by the shore staff or
permitted through poor shore procedures. The shore management reacted angrily when
masters instituted their own checks of passengers boarding. In fact all tickets were retained
ashore, so other than a physical count of passengers boarding in vehicles and foot
passengers on the gangway, it was not possible for the ship to know how many people were
on board. Subsequent spot checks of the company's vessels showed on 4 occasions in a
three week period vessels overloaded by 250 passengers, 40 passengers, 100 passengers
and 171 passengers.
The draught at sailing is critical, particularly in terms of the "C.1" and "C.2"
configuration. The ship was equipped with draught "indicator" gauges, but these were not
considered accurate. Given the ramp, the overhang of the belting, the way the ship lay
alongside the berth, it was difficult to site the draught at any time. Add to this the intense
working in port and the very quick turn round times the checking of the draught was
problematic. In common with most ferry operations, the draught was calculated on weights
loaded from empty ship, plus known stores, bunkers waters, etc. Subsequent spot checks on
vehicles loaded on other ferries showed that the weight of a significant number of freight
vehicles were greater than shown.
Calculation of stability on a single voyage basis was also seldom undertaken.
Typically, if a ferry was loaded within known parameters no calculation would be done. If such
vessels were to take an unusual load, or an untypical configuration of cargo and this was
notified to the ship in advance, there was time for a full stability condition to be completed.
In February 1984 Heraldschief engineer wrote to the company pointing out that the
filling of the forward ballast tank was essential for berthing in Zeebrugge and was emptied
after leaving. This operation took 1 hour 55 minutes with the existing ballast pump. This
meant that:
a) the operation amounted to half the passage time
b) the vessel was trimmed by the head for prolonged periods increasing fuel consumption
and resulting in bad steering
c) there was continuous overpressurising of the tanks
d) there was stressing of the the bow doors by reason of the trim by the head (the bow wave
could be as high as 113rd the height of the door).
The chief engineer requested a high capacity ballast 750 kW pump rated at 1000 cu.rn/hr with
its own sea chest and suction in a forward compartment. This was rejected by shore
management. .
Whenever the ship put to sea the status of the bow and stem doors was.:criticaland
.safety depended on the doors being properly closed. Because the "clam" doors could not be
seen from the bridge, the master could not see whether or not the doors were open. On 29
October 1983 the sister ferry PMe, sailed with both bow and stern doors open because the
assistant bosun had apparently fallen asleep. On 6 October 1984 the master of PnUe sent a
memo to all deck officers pointing out that twice since going on the Zeebrugge run the ship
had sailed with the bow or stern door open.
"No doubt this is caused byjob/rank changes from the Calais run."
In 1985 a master wrote to the company management pointing out that the "most
important" bow and stern doors could not be seen from the bridge and suggested that any
new building should have indicator lights on the bridge to indicate the status of the doors. This
suggestion was referred to various members of the shore management who responded by
annotating the memo:


'Do they need an indicator to tell them whether the deck store keeper is a wake and
sober? My goodness!!"
"Nice but don 't we already pay someone. "
'Assume the guy who shuts the doors tells the bMge there is a problem."

This attitude was somewhat at odds when in 1985, the shore management "advised" that the
masters should rescind disciplinary action against ratings when strike action was threatened
over disciplinary action correctly imposed. This drew a protest from the masters.

In May 1986 a master on Heraldwrote to management again suggesting that:

'Bow and stern doors Open/closedindicathn to be duplicated on the bnidge.
On 13 October 1986 the same master submitted a formal job specification requiring bridge
indicator lights to show whether the bow and stern doors were in a "secure or insecure mode".
A senior company manager notated on the job specification:
'7 cannot see the purpose or the need for the stern door to be monitored from the
br/;dge, as the seaman /i7 charge of closhg the doors /;s standng by the control panel
watching them."



C W Filor
Inspector of Marine Accidents,
Department of Transport, Canberra, Australia
Any accident, not just a marine accident, is "any unintended or unexpected occurrence which
produces hurt or loss."l An accident is usually the result of human failings, including unsafe
acts, coinciding with other externa.1elements at a moment of system vulnerability.
Human error, whatever that means, is accepted as being the major cause of most accidents.
Although the shipping industry has recently focused on the human element, it is no new
"A summary of official inquiries into wrecks and casualties, excluding collisions, shows
that from the years 1856 to 1872 inclusive, a period of 17 years, while 60 ships were
known to have been lost from defects in the vessel or stowage, 711 ships were lost
from neglect or bad navigation"2.
We are constantly reminded that human error is the main cause of marine accidents and
bombarded with similar, if updated statistics. The International Marine Safety Organization
has instructed its Committee and Sub-committees to include an agenda item on the "human
factor" in all their meetings. Despite this focus and the data being updated, this self evident
truth has not taken us much further forward in a quest to rectify or counter human factors.
However, progress has been made. IMO Resolution 849(20) recommends that member
States adopt the Code for the lnvestigation of Marine Casualties and Incidents. 'The Code
contains guidelines to assist investigators in the implementation of the code and will also
include "Guidelines on lnvestigation of Human Factors in Marine Casualties and Incidents''
(HWG2). By investigating incidents under the Code and Guidelines we should add to our
understanding of the human element in accidents at sea, and hopefully address them.
So what is this "human factor" and who are these humans that, in too many cases, are
responsible for human tragedy or catastrophic pollution or huge expense to their owners?
They are you and me in the wrong place at the wrong time, under the wrong circumstances,
often using imperfect systems and less than ideal equipment. In any accident the human
factor has to be seen in the context of the ship and the associated equipment in use.
Indeed in an increasingly technical world, it is often difficult to differentiate between human
error and mechanical or ship failure when analysing a marine accident.
"There is a sense in which this distinction is artificial. Most instances of ship failure are
'themselves the result of prior human errors, occasionally in judgements as to the
design of the ship or the suitability of materials used in its construction, but more
frequently in the process of planning or implementing programmes of maintenance."3

Fenton v Thorley & Co Ltd (1903) AC 443 at 453, per Lord Linley,
Royal Commission on Unseaworthy Ships, 1874 (UK)
3 Hugo Wynn-Williams, The United Kingdom Mutual Steam Ship Assurance Association (Bermuda) Limited,
Analysis of Major Claims, 1992


It is convenient (and is often a useful public palliative) to blame accidents on the people who
suffer them and to see unsafe acts arising from incompetence, recklessness, carelessness
or stupidity of the ship board operator.
"Rather than being the main instigator of an accident, operators tend to be the
inheritors of systems defects created by poor design, incorrect installation, faulty
maintenance and bad management decisions. Their part is usually 'that of adding the final
garnish to a lethal brew whose ingredients have been already long in the co0king.v
So, given the overwhelming significance of the human element, what basic and preferably
simple, factors should form the basis of a workable model, that can be applied to accidents?
"In considering the human contribution to systems disasters, it is important to
distinguish two kinds of errors: active errors, whose effects are felt almost immediately, and
latent errors whose adverse effects may lie dormant within the system for a long time,.."S
In Marine accidents, active elements relate to the "unsafe acts", either in commission or by
omission by those people present on the ship at the time of the accident. It is these "unsafe
acts" that so often are seen as the immediate cause or proximate cause, upon which issues
of liability or blame often hinge. They tend to be unique to a specific accident.
Latent elements (or unsafe conditions) are removed in space and time from the actual
accident. They often lie dormant within the system and are only manifest themselves when
triggered by the coming together of other unsafe conditions and acts. They remain within the
system - unless identified and corrected - and may be present in any number of accidents
within the system. They are often error promoting conditions that may induce individual
unsafe acts. The origins of latent unsafe conditions lie within in the management of an
organization and within manufacturing, contracting, regulatory and governmental agencies.6

Reason, J. Human Error, p.173, Cambridge University Press, 1990

Reason, J. Human Error, p. 173
6 Reason, J. Managing the Risks of OrganliationaIAccidents, p.11, Ashgate, 1997



Human Contributions to the Breakdown of Complex Systems


Line Management

Unsafe Acts
Failed defences

The Dynamics of Accident Causation

Adapted from Reason J, Human Error
Active errors arise from the interaction of individuals with the system that is being operated
and the environment. Active errors can be seen as those actions taken by a ship's crew,
immediately before and at the time of the incident.
Latent errors are the result of decisions or conditions at a number of levels involving design,
management procedures, training and maintenance schedules, etc. The nature of latent
factors underlines the often artificial distinction between shiplmechanical failures and human
elements, and are based on three interrelated levels or conditions of decision making, "fallible
decisions", "line management deficiencies" and "psychological precursors".
In simple terms, fallible decisions are those decisions made or agreed to by senior
management, which, under a given set of circumstances, are subsequently shown to be
flawed. Fallible decisions may include issues'of design, crew organisation, budgetary policy,
Line management refers to the decisions made by the managers responsible for the day to
day operation of the ship. It is these measures which often reflect on the physical well being,
the morale and the motivation of those on board. Any incompetent decision can further
exacerbate the adverse effects of fallible (high level management) decisions, or even cause
sound decisions to have a negative effect. Conversely, competent line management
decisions can mitigate the effects of other factors.


"Psychological precursors"or "preconditions" involve the physiological and psychological

factors impinging on all the individuals in the system, including the physical and mental
condition of those on board the ship and liable to perform unsafe acts.
Behind both the active and latent factors are the defences. The level and effectiveness of
defences will depend upon the decisions made by all levels of management and the active
factors relating to maintenance.
By looking at accidents in this way a basic model of the active elements, in terms of human
performance, and a model or methodology for assessing the latent factors may be drawn up
to objectively assess the factors present in an accident.

Active Factors (The Human Performance)

Too often the failures in human actions are judged in terms of competence, rather than on the
person as a whole. This can lead to a number of problems, as very often a person will hold
a certificate to show that helshe is competent, or at least was judged to be so at the time of
issue. Very often, f~llowingan accident, a person who may be seen as directly responsible
for the accident (and hence whose competence may be questioned) acts with great skill and
competence to mitigate the effects of an accident. It is a truism that,the most competent
people are not competent all the time. The paradox is that the bad decisions any of us make
come from the same source as the good decisions - our brain.
The human factor, in any endeavour, can be best understood in terms of "performance", of
which competence is only one element. In this simple model, performance is made up of
three basic conditions, "alertness" (actually termed "arousal" by experimental psychologists),
"attentiveness" and "competence". Each of the factors is, in turn, affected. by the human
condition in play at any given moment















Because accidents are caused by the complex interplay of factors no one element acts alone
to prompt unsafe acts. They cornbine in different forms for different accidents, like some
complex virus.
The model divides "unsafe acts" as resulting from either an unintended action or an intended
Many unsafe acts arise out of unintended actions when actions have strayed from the original
plan. These often occur when an individual is performing a routine task in familiar
surroundings, and is distracted by some external phenomenon. This is particularly so where
stress levels are too low, such as in a routine watch keeping situation. These are the slips or
lapses on the part of the operator.


Unintentional Action

--.-- &
& Knowledge-based

Intentional Action
* Routine
& Exceptional


Accidents arising out of intended actions may have their geneses in mistakes, where
somebody follows a plan but the plan itself is inadequate, or unsafe, or otherwise flawed.
These planning failures may be caused by lack of expertise in devising a plan, the planner
not having sufficient experience to draw upon, or where a pre-established plan is applied
It may also be that an operator deliberately disobeys rules or deliberately departs from a plan

of action. Such "violations" may be a wilful disregard of accepted procedures or of a well

established plan. Often it may be that a culture of violation has developed as others have "got
away" with a particular violation as a short cut in some operation. It may arise out of
exceptional circumstances where a person feels justification for a departure from a rule based
system, or just a simple aberration.
It is a fact that most of us are prone to slips, lapses, mistakes and rule violations, but they
only contribute to an accident when a number of such factors combine with external elements
in a moment of vulnerability, or "sods law".
How would the application of this model help to prevent accidents? In brief, by understanding
.the difference between "unintended actions" and "intended actions", it should be possible to
target preventative efforts. In the case of slips and lapses it may be that a better level of work


"stress" can be devised to maintain a level of alertness, or better defences built into a system
to prevent a slip or lapse having serious consequences.
In the case of intended actions it may be possible to improve rules or ensure that the levels
of experience are appropriate to the plans made. It may also be possible to identify why
violations occur.
As an example, one should not be too surprised if the introduction of the International Ship
Management Code, being ruled based, may reduce accidents through mistakes, but may
increase accidents through violations. It will probably have little effect on unsafe acts arising
out of slips and lapses.
Latent factors

Applying such a model to latent factors is more difficult and a different strategy is required.
Looking d a system that has failed, it is sufficient, at least in the first instance, to ask six basic
safety questions:

Were the risk factors identified or identifiable?

Was the equipment in use fit for purpose?

Were the systems and procedures on board effective to maintain safe operation?

Were the individuals involved fit, competent and effective?

Were emergency procedures and defences effective?

Was there a management system to monitor performance?

These six questions form a model which will address the human factors present in the latent
factors contributing to a marine accident, by identifying the fallible and line management
elements, together with the psychological precursors to the accident and the failure of
"'There is a growing awareness within the human reliability community, that attempts
to discover and neutralise these latent failures will have a greater beneficial effect
upon system safety than will localised efforts to rrlirlimise active errors."6
It should be recognised that a number of companies, in high hazard, low risk industries, have
adopted "no blame" accident investigation methods based on a similar philosophy to that
outlined in this paper.
It is also worth noting that in the annex to US paper MSC 6511511 an example is given of a
US shipper who has embraced a risk management programme.
"Through analysis of the process making up the elements of these operations it made
a number of changes and adjustments to procedures, corporate policy and staffing
which have had a significant monetary return. It experienced a 50 per cent reduction
in "loss time" accidents and a saving of $3,000,000 in insurance costs. One can only
surmise an unquantifiable gain in safety, employee morale and corporate reputation."7
sReason, J. Human Erroc p l 7 3
7United States Paper, MSC 6511511, Role of the Human Element in Maritime Casualties, February 1995


The methodology outlined in this paper offers a practical approach to accident investigation
and a skeleton on which we may hang the identifiable factors that underline human
performance and contribute to accidents. It reflects the sort of thinking that is increasingly
being adopted by major industries involved in transport, oil and chemical productionand other
high hazard enterprises.
"Do not allow accident investigation conducted for the enforcement of laws or
regulations, or for determination of liability to interfere with accident investigations
conducted for safety purposes. We have been conducting the former two for so long
they tend to get priority even though they produce few safety improvements."8

Colucciello, L. Investigation and Identificationof Unsafe Practices and Conditions, XXI ICHCA International
Conference, 1993

Human Factors


Competence, fitness, aptitude and ability

This section on human factors is an overview only and will be considered in greater detail in
the investigator's course. However, the following notes should demonstrate the irr~portanceof
human factors, and its complexity, in casualty investigations:
As Learning objective 4 demonstrated, the distinction between mechanical failure and human
factors is somewhat blurred. The text of "Unsafe Acts -Active and Latent Elements, A Model
for Accident Investigation and Analysis" should be born in mind in this Learning Objective.
Although human factors are present at all stages of an accident, the purpose of this Learning
Objective is to focus on the ship's crew and others who are involved in the active failure stage
of a casualty or incident. However the same principles apply at all stages (fallible decisions.
Line management etc) and to all decision makers.
Ship's staff, pilots and other maritime related personnel are appointed to their position based
'on a competency assessment system. This system has international standards (STCW) and
part of an investigation process has been to assess whether those immediately involved in a
casualty acted at a sufficient level of competence.
However, competency is but one factor in a complex human matrix.






, -











The second paragraph of Section 1 of HFWG2 notes:

'Ships operate in a high& dynamic environment,.people on board follow a set routine

of shiff work frequently disrupted by arrivals at, working in, and sailing from port This
is an existence which involves living in the place of work forprolongedperiodscreating
a unique form of working life which almost certain& increases the risk of human error: '
The safety of a ship and its crew depends on the alertness and efficiency of those assigned
to duties on board. These are usually, though not exclusively watchkeeping duties on the
bridge and in the engine room. Inadequate attention levels or deficient performance by key
individuals can have serious consequences.

'Two main factors may affect the eficiency of a watchkeeper; first, mental fatigue
resulting from the nature of his (her)job, and its continuous performance; and second,
lo wered levels of alen'ness, pan'icularly during watches held during "night" hours, due
to the effects of physiological circadian rhflhms and disrupted sleep.
With regard to the first factor, mental fatigue is lke& to increase in proporl/bn both to
the degree of monotony and to the amount of '3tress" h doing it. 'StressJ'ihthe most
obvious sense is met with h those phases of a passage involving ihtricate navigation,
e.g. when the vessel is approachhg or leavingpan: or is travershg a narrow channels
or coastal wafers; h these situations there maybe 'Too much to do". On the other hand
stress can also arise from having too liffle to do, e.g. during the middle of an ocean
crossihg when the job becomes mauima!!y monotonous."1
Although the study quoted addressed circadian rhythms while at sea, the observations apply
equally to work in port.
The basic issue here is that of the performance of normally competent persons and the things
that can affect the performance. It is an old saying, but true, the most competent person is
not competent all the time.

'The best people make the worst mistakes,'2

8.2 Ship organization
Ships have traditionally been organized into watch systems based on at sea routine. The
most common system is a routine 4 hours on watch, 8 hours off watch. In the case of officers
some work is often carried on in one of the eight hour "off" periods. In the case of ratings, their
agreement or overtime payment may also involve work in a period when they are not on
In smaller ships with reduced crews, a six hours on watchlduty followed by six hours off is
1 Colquhoun, W. P., Rutenfranz,J., Goethe, H., Neidhart, B., Condon, R., Plett, R. & Knauth, P. (1988). Work
at sea: a study of sleep, and of circadian rhythms in physiological functions, in watchkeepers on merchant
vessels. I.Watchkeeping on board ships: a methodological approach. lnternationalArchives of Occupational
and Environmental Hea/th,60: 32 1 -329.
2 Reason, J.


In short sea trades, particularly ferry services a variety of shift work patterns prevail. These
can involve a 12 hours on duty (on ship) followed by 24 hours off duty (at home), or a 24
hours on duty (on ship) followed by 48 hours off duty (at home). In the former, the personnel
are involved in continuous working; in the latter, work is more or less continuous with perhaps
an hour or two not actually on duty. There are many other combinations based on hours or


Arousal - Yerkes Dodson model

Experimental psychologists working in 1908 showed the relationship between arousal (or
performance) as measured against stress.

Too little


Too much

This showed that performance could suffer with too little stress, where boredom sets in. Also
too much stress leads to poor performance with often an overload situation, with
overwhelming number of communications and tasks. .


The Scandinavian Air Services (SAS) Flight Academy's notes on Bridge Resource
Management shows a similar situation in a similar way.

Low workload

1994 SAS Flight Academy

The diagram shows that the common experience is for a person to be in the -2 or -3 mode
alertness and transfer clockwise to a +3 stage, one of serious overload.


Factors affecting the degree of performance

The factors generally considered are those which include amongst others:
External factors (Environmental and personal)
A more detailed taxonomy of human factors is at Appendix 3 (pg.28-33) of HFWG2.


Fatigue is defined (HFWG2 pg.29) as a reduction in physical and/or mental capability as a
result of physical, mental or emotional exertion which may impair nearly all physical abilities
including: strength; speed; reaction time; coordination; decision making, or balance.

Circadian rhythms '

Our understanding of fatigue and why people are more prone to accidents at certain times of
day may be helped by an understanding of circadian rhythms.
Our lives and bodies are governed by a variety of biological rhythms. The most obvious are
our heart beat, respiration and digestion.
The circadian rhythm comes from the Latin circa (about) and die (day). It is a cycle of natural
alertness over the period of a day or 24 hours (actually nearer 25 hours). The circadian
rhythm, which is manifested by fluctuations in body temperature, is the natural activity
program of the body.






The circadian rhythm accounts for the time we are most active and the natural inclination to
sleep. When the body temperature drops in this natural cycle the body's natural inclination is
to slow down, close down and sleep. Thus between about 2200 to about 0600 it is natural to
sleep. There is also a brief period after midday when the body activity drops, the after lunch
nap, the siesta period.
The circadian rhythm is not adaptive to any different daily routine. Shift workers, whether
watchkeepers at sea, airline pilots or factory workers all experience temperature drops as
illustrated above, no matter how long the person has been on a particular shift.
This also has important irr~plicationsfor rapid longitudinal or meridianal travel - "jet lag". Staff
joining ships by plane having travelled long distances east and west will take days to readjust
to the light (daytime) and dark (night). The rate of adjustment is about one day per hour
difference from the origin.
A study of "Work at Sea" (Colquhoun at al) considers the issues of "jet lag". "It is known that
circadian rhythms can take up to 10 days or more to adapt fully to the sudden multi-zone
transitions that are experienced on these journeys, and that the disruption is greater on
eastward than on westward flights. . . . An associated problem is presented by the growing


practice of replacing crew mernbers due for shore leave by flying out reliefs to wherever the
ship happens to be. The flight places the relief crew member in a state of jet-lag upon arrival
on board; during the period of recovery from this condition the crewman, if a watch keeper,
also has to adapt to his watchkeeping schedule, . . . [and, if the ship is on a trans-meridian
journey, to changes in ship board time as well. It would seem likely that the degree of
disrl~ptionof circadian rhythms arising from the combined effects of these three factors will
be substantially greater than that produced by the action of any one of them alone, and the
resulting effects on alertness and efficiency thus considerably more severe.Y3
Circadian rhythm disruption manifests itself in:
Decreased response rate and a lowering of mental performance
Decline in mood and motivation
Increased sleepiness and feeling of fatigue

Research has shown that nine internal or external factors and stimuli are relevant to the level
of alertness in an individual4. They are:

William G Sirois, Circadian Technologies, Inc

These factors can be seen as switches that may be in the "on" position, "off" position or in
"neutral". Whereas just one individual switch in the closed position may not be sufficient to
cause a dangerous loss of alertness or induce sleep, the greater the number of closed
switches, the greater the hazard.
Colquhoun, W. P., Rutenfranz, J., Goethe, H., Neidhart, B., Condon, R., Plett, R. & Knauth, P. (1988). Work
at sea: a study of sleep, and of circadian rhythms in physiological functions, in watchkeepers on merchant
vessels. I.Watchkeeping on board ships: a methodological approach. lnternationalArchives of Occupational
and Environmental Health, 60: 321-329.
4 Moore-Ede, M., The Twenty-FourHourSociety, Random House Australia, 1993


Interest, opportunity, sense of danger

Alertness is heightened when a person is interested in the task at hand or by a feeling of
danger or risk. In extreme cases of danger it may be a feeling of "fight or flight".
Muscular activity
All types of muscular activity, even chewing, act as a stimulant to alertness, the muscular
activity switch moving to the "off' position when the body is at rest.
Time of day
The natural period of sleep is during the hours of darkness, between 2200 to 0600, the brain
automatically regulating the body into the sleep mode. People working regular shift work,
where the periods of work are not changed, can adapt to the different sleeping period without
too much ill effect, their biological clock adjusting to the changed conditions. Thus mariners
are able to adjust to their watch regime, the time of day switch being in the "on" position at
the required times.
Sleep bank balance
Sleep periods are considered as making deposits in the individual's "sleep bank" and periods
of wakefulness make withdrawals. As a general and approximate rule, sleep deprivation
occurs, and the alertness switch goes to the "off' position, when an individual expends the
total "sleep credits". Sleep credits are considered to accumulate at a rate of two for every hour
of sleep and to expend at the rate of one for every hour awake. Hence, after eight hours sleep
an individual can be considered to have 16 credits, which are expended over the next sixteen
Ingested nutrients and chemicals
Caffeine and amphetamines are stimulants, whereas alcohol and some medications are
sleep inducing.
Caffeineand other stimulants may be effective in-overcomingsome drowsiness, however, the
ingestiori:.oF both stimulants and depressants can interfere with the circadian rhythm. After a
watch, ingestion of too much stimulant may prevent a person from sleeping although
physically tired. Similarly, while alcohol and depressants can get people to sleep, they
interfere with the sleep quality and a person may awake tired and unrefreshed.
Environmental factors
Bright light, cool temperature, irregular sound and certain aromas all act as stimulants, which
help keep a person alert. Dim lighting and darkness, warm sultry heat and regular rhythmic
noise can all have the opposite effect.
[G~OIJPexercise: Watch video "Counting the cost" identify any of the above factors in the

Slips, Lapses, mistakes and violations

Page 21 of Res. A.884(21) covers these issues.


HUMAN FACTORS (adapted from a paper by Alan Hobbs, a psychologist

with the Australian Bureau of Air Safety Investigation)
Human factors play an important role in the origin of accidents, and it is commonly claimed
that between seventy and ninety-five percent of industrial and transport accidents involve
human factors, see below. Some authorities, however, claim that ultimately, all accidents
involve human factors.
Figure 1: Contribution of Human Factors within Accidents





Before we can start to understand why accidents happen, we need first to understand a little
about human limitations. An area of human factors, which has received a great deal of
interest is human cognitive abilities and limitations. This includes constructs such as
attention, memory, decision making, action, skilled performance, and stress. Greater
understanding of this area enables us to reduce errors and accidents caused by technology.
It also helps investigators understand how accidents happen and assists in determining
suitable recommendations.
Examples of attention
Attention is such a fundamental fact of life that we may not even be aware of how our
attention works and what the limits to attention are. For example, what is occurring when,
whilst conducting your own conversation, you become aware of another conversation on the
other side of the room once you hear your name mentioned? How is it that people who live
near a railway line may no longer hear the trains go by? What happens when a taxi driver
must simultaneously drive in heavy traffic, enter information into the base computer, and deal
with corr~plainingpassengers? And why is it that sorr~etimeswhen we are concentrating on a
television program, we do not hear other people talking to us?


Our senses have available to them a "booming buzzing confusion" of sensations. You are
probably not aware of the feel of your feet in your shoes, or the sounds happening outside.
Obviously, we must be able to focus in on what is important to us at the moment and screen
out the irrelevant information. It is as though we have a flashlight beam, which we can shine
on only one area at a time..The problem is that there will be times when there is too much
important information coming in for us to cope with all at once, or we will miss information
wliich is outside our attentional "beam of light". Or perhaps our attention will stray onto
distracting information like financial concerns or physical discomfort.
Psychologists have come up with diagrams to explain how our attention works, and although
there are some differences between different approaches, there are some points of
agreement. It is agreed that there is a finite limit to the amount of information that can be
processed at any time, and that we have a bottleneck in our information processing system,
allowing only a limited amount of information to squeeze through to consciousness at a time.
In this sense, our information processing system is sequential rather than parallel. When we
feel that we are consciously attending to several things at a time, such as monitoring a radio
and holding a conversation, we may in fact be rapidly switching our attention from one to the
other. .+
Obviously, the limits of attention do not prevent us from doing more than one thing at once;
we can drive a car and talk at the same time. But this is because well-learned skills like driving
become increasingly automatic and hence demand less and less of our attention.
One of the consequences of the limits on our attention is that individuals have a "red line" on
their ability to process information. For example the Three Mile Island accident occurred
when operators failed to diagnose the reason for loss of coolant from the reactor. This may
be because more than 100 control panel alarms activated with no means of suppressing
unimportant ones.

Faced with an overload of information, a person can respond in a number d ways:

Loadshedding is when a person ignores some of the information and concentrates on one
or two aspects of the situation. Hopefully, the individual will shed the least important tasks.
However,-it is possible that the individual sheds vital information.
An extreme form of load shedding is channelledattention, in which the individual gives all of
his or her attention to only one aspect of the situation.
Another coping mechanism for overload is regression. Regression is the effect where an
overloaded individual goes back to a previously well learned pattern of behaviour. Under
stress and information overload individuals may revert to familiar,-automatedbehaviour.
The limits of attention do not just apply when someone is severely overloaded. Even a
moderate mental workload can reduce an individual's capacity to deal withnew information..
NASA studies have shown that if pilots attempt to perform a listening1identificationtask at the
same time as a visual tracking task, their eye movements can be reduced by up to sixty
percent. A listening task will also cause an individual's peripheral vision to be less effective,
almost as though they are wearing blinkers. Visual field narrowing, as this effect is known,
has obvious implications for traffic scanning.
Stress, as well as information overload can also cause some cues to be filtered out. A person
under stress has less capacity to deal with information; the "red line" is at a lower level of
performance. Airline pilots coping with an in-flight emergency have landed gear up and


reported that they did not hear the gear warning horn, even though it was functioning.
Passengers who have evacuated a crashed aircraft often report that everything was very
silent. Perhaps this too, is a case of stress causing perceptual narrowing.
One final area where attention is important is vigilance. In WWll it was found that after about
twenty minutes at their posts, radar operators became much more likely to completely miss
obvious targets. Often somebody walking by would casually notice a radar paint that had
been missed by the operator, even though he or she was intently concentrating on the
screen. This problem, known as the vigilance decrement, applies to many monitoring tasks
where 'fiitsyare relatively rare. Aircraft inspection, the checking of medical X-rays and quality
control inspection in factories are all areas where vigilance decrements may occur. Vigilance
can be improved by increasing the conspicuousness of the signal and increasing the number
of rest breaks or the variety of the work. Vigilance is also often better in a more social
atmosphere, perhaps because it keeps people more alert.
To summarise the main points concerning attention: An individual's limited capabilities of
information processing can be easily overloaded and can result in /oadsbeddhg, cbame//ed
altenhn or regression to ingrained but inappropriate skills. The capacity to process
information can be further reduced by stress, fatigue and lack of currency. Any conscious task
can occupy attention and block out other information. This includes thinking, daydreaming
and worrying. Novel or difficult mental tasks can cause a narrowing of the visual field.
However well-learned skill routines take up less mental capacity than routines that are less
polished. Tasks requiring intense vigilance will suffer after approximately 20 minutes.

We often talk about memory as though there is only one sort of memory whereas in fact we
have a number of memory systems, each adapted for a different purpose. We remember new
telephone numbers just long enough to dial them. Other information must be remembered for
years. If memory worked perfectly all the time we could ignore it. Our interest in memory
stems from the problem of forgetting. Just as there are different types OF memory, so there
are different types of memory lapse. Such as:
a forgetting a name but recognising it as soon as it is mentioned
a filling the gaps in our memories with logical expectations
a assuming that the information retrieved is correct when in fact it is the wrong
a blending memories together to come up with an answer that is half right; or even
forgetting something entirely!
Sensory store
The first stage of information processing is the short-term sensory store. This store enables
us to hold information momentarily before it is processed further. The two most important
sensory stores are echoic memory and iconic memory. Echoic for sounds, and iconic for
Iconic memory
The sensory store retains a brief trace of the stimulus after the stimulus has disappeared. For
example, children commonly write in the air with sparklers, making use of the momentary
image left after the sparkler has moved on. You can time how long the iconic trace lasts by
placing a light on a wheel at night and rotating the wheel. If you measure the speed of
revolution of the wheel at the point where a continuous circle of light begins to break up, you


have measured the duration of iconic memory. Information in iconic memory lasts for between
.5 and one second. Iconic memory enables you to glance at an instrument and mentally scan
it after your eyes have moved on.
Echoic memory
Echoic memory lasts a little longer than iconic memory, in some cases up to eight seconds.
It enables us to hear a message and briefly put off listening to it until we have finished dealing
with other information. For example, half of a radio message may have passed before you
hear your name and realise that it relates to you. Echoic memory enables you to mentally
"replay" the message and go back over the initial bit that you were not attending to.
Interference in Echoic memory
Like all forms of memory however, echoic memory can let us down. Each sound in a string
of sounds-can interfere with the preceding sounds. For this reason, the final word in a string
of words has a better chance of being recalled than a word from the middle of the message.
Unnecessary radio transmissions at the end of a message (such as "over") can potentially
mask out part of the echoic trace of the preceding message.
Short term memory (STM)
Sensory memory lasts for a few seconds and decays very rapidly. But there is another form
and enables us to keep several bits of information
of temporary memory that is more d~~rable
in mind at once. This is working memory or short term memory.
When you look up a phone number and keep it in mind until you dial it, you are depending
on short term memory. Mental arithmetic is another situation where we have to keep several
items stored in memory until we have found the answer. Short term memory is not necessarily
limited by the time that items are stored, because by saying them over and over (rehearsing
them) we can store something for a long time. Short term memory is limited by the number
of items that can be held. A common way of checking the capacity of short term memory is
the digit ..span test.


The digit span test was first conducted by Jacobs (1887). A sequence of numbers is read
aloud in no particular order. After the sequence has been completed, the listeners are asked
to write down as many as they can remember. In general, people can store between five and
nine unrelated items in short term memory. The average is around seven, and the capacity
of short term memory is sometimes called "seven plus or minus two". This limitation is
tremendously important, particularly when receiving important information which is transitory
(such as auditory information) which must be momentarily kept in memory before they are
acted upon.
However, it is possible to increase the capacity of short term memory by chunking information
together. Such as encoding a string of numbers as dates instead of single digits eg 1914 2000.
1988. Or for example, the French chunk telephone numbers into three large numbers rather
than six digits eg. ninety six, twenty seven, thirteen for 962713.
Short term memory seems to depend on verbal rehearsal, so if you are told to call on 123.9
just as you are about to do something else, you may repeat 123.9 to yourself under your
breath. Without this sort of verbal repetition, short term memory decays rapidly. But verbal
repetition has some important implications for short term recall.


First, similar sounding items are likely to get confused with each other. For example the
letters t, p, v, and dl all sound similar and therefore may interact with each other and influence
the way any of them are rehearsed and consequently remembered.
Furthermore, items of information containing common or redundant elements are also likely
to be confused, for example, A123, A734, A391 are harder to remerr~berthan just 123, 734
and 391.

Having to talk or pay attention'.elsewhere will disrupt the short term store. For example the
process of performing mental arithmetic will make it difficult to retain other information in
short-term memory.
In addition, if items must be retained in short term memory, it is better if the information is
distributed over time rather than received all at once.
Finally, we should not forget that the capacity of short term memory decreases with age.
Long term memory

A small proportion of the information that passes through our short term memory Finds its way
into long term memory. Unlike short term memory, there is no limit to the amount of
information we can store in long term memory.
There are two types of long term memory. Episodic and semantic.
Episodic memory is the recall of specific events or episodes such as a particular voyage or a
day's events. A limitation of episodic memory is that it is easily distorted after the event. For
example accident investigators often find that the way they ask questions of a witness can
easily change the witnesses' recall of events. In an interview, terms like breeze, may tend to
encourage recall of the weather as milder than was experienced. Whereas the word wind
may elicit recall of stronger weather conditions. Episodic memory for events in time is also
notoriously unreliable. For example, witnesses sometimes recall an event that lasted thirty
seconds as lasting for much longer or even reverse the order of events.
Semantic memory is where we store our abstract knowledge of meanings, relations and the
mental frameworks we use to understand the world. Semantic memory appears to be a
permanent record. Once you have learned about meteorology or how turbine engines
function, that knowledge is there for good. This is not to say that information is always easy
to retrieve!
Semantic memory relies heavily on associations between memories, or memory cues for
retrieval. This is why when we lose our car keys, it helps to go back to where we last had
them the sights, smell and sol~ndsof the place will help us get access to the information in
for keys. In general, the
our memory. This principle has wider applications than j ~ ~ looking
more associations a memory has, the easier it will be to recall. Another result of this is that it
will be easiest to recall information in the same environment in which it was learned.


Information learned in a classroom will be recalled most easily onshore and things learned
on the ship will be best recalled in that environment. A bad memory is often a sign of poor
organisation and a failure to link the informationto existing knowledge. It is a general principle
that the more mental work we do at the time of encoding in memory, the better our memory
will be. For example, you will remember more of this course if you have actively thought how
the principles apply to your own company than if you have simply listened attentively.
What we are talking about here is essentially learning. It is sometimes assumed that people
learn better under stress. Although everyone is different, in- general stress hampers learning.

The history of accidents and disaster is frequently the history of wrong decisions. The charge
of the Light Brigade, the sinking of the Ztanic, Amoco Cadiz, the Challeng~space shuttle
accident and any numbers of groundings and collisions are all examples of disasters which
resulted in part from poor decisions. More often than not the people who made these
decisions-were properly trained and had all the necessary information available to them, but
for some~reasonpersisted in making a poor decision.
Poor decision making also accounts for a large proportion of accidents at sea, as indicated
in Table 1.
Table 1: Human Errors in 100 ~ccidentsat Sea

Category of Error
False Hypothesis
Social Pressure

No of Accidents

Source: Wagenaar and Groeneweg 1987

So apparently, any attempt to cover human factors must consider decision making, and try to
answer why properly trained and usually extremely experienced individuals sometimes make
poor decisions.
Unfortunately, the idea that people make rational decisions by carefully weighing up the
information and considering all the possibilities has been shaken in recent years. We have
learned that when people make decisions, they frequently allow emotional factors to influence
their decisions, they do not consider all the alternatives or they take labour saving mental
short cuts to arrive at a solution.
People do not passively observe the world around them in the way that a camera forms an
image. When we perceive an object or experience a situation we do so within unconscious
mental frameworks, which influence our understanding of what is going on. For example, if I
said "the notes were sour because the bag burst" you would probably have no idea what
I was talking about, although you might try a few mental ideas to attempt to explain what the


sentence is about. But if you knew that the sentence was about bagpipes you would have a
context which would allow the sentence to make sense. People have a very strong tendency
when they experience something new to try and slot it into a familiar context. It is as though
we hate being put off guard by something unfamiliar and only relax when we have recognised
it as something which we have encountered before. Inappropriate mental models can distort
decisions just as surely as they can distort perceptions.
False Hypothesis
Psychologists sometimes talk of the "strength of an idea", meaning the way in which an idea,
once formed, can become very resistant to challenge, even when information is available ,that
would contradict the idea. In a very short time, an idea can become firmly entrenched.
False Hypothesis: Case study
On 14 December 1982 the seven hold bulk carrier Farmsum was engaged in hold
cleaning. Four men were in number six hold. No. 4 hold was ballasted, no. 5 hold was
meant to be empty with men working in no. 6 hold, which had some residual water in it.
However, no suction could be obtained.
Water was being lost from no.4 hold. The mate found that the ballast valve on no.4 was
stuck open. The mate assumed that this accounted for the loss of water from no.4 and
the inability to get suction in no.6. In fact no.4 was leaking into no.5. The water being
pumped into no.4 in fact also filled no.5 until the bulkhead, which was not designed to
withstand the pressure of ballast, failed. Three of the four men were drowned.
False hypothesis is very common with radio work, particularly when we expect to hear a
particular phrase. For example, the world's worst air disaster occurred when the Captain of a
KLM 747 at Tenerife apparently believed that he had received a take off clearance when in
fact none had been issued. Despite rather mild challenges from his crew, the captain held on
to this idea and attempted to take off while a Pan Am 747 was still on the fog-covered runway.
To sum up this issue of the false hypothesis, in ambiguous situations, we tend to be
uncomfortable with uncertainty- and have a strong tendency to latch on to an explanation or
an idea. Common scenarios or ideas have a strong tendency to become default assumptions.
This is particularly the case when: expectancy is high, at times of diverted attention, when the
assumption is a comforting thought, or after a period of high concentration.
The issue of mental models is particularly relevant to fault diagnosis and there are some
common problems here which are worth considering.
As 1 mentioned before people dislike uncertainty, and faced with an ambiguous set of events,
they may be led to put a label on the problem too early. This labelling of the problem can then
serve to seriously limit thinking. It is also commonly observed that when diagnosing a
problem, people rarely consider all the possibilities. Scenarios which spring to mind most
readily will tend to become the preferred explanations. These scenarios are likely to be
familiar explanations based on previous experience.
Many sources of information are quite unreliable, for example, a message heard through
static, or an instrument reading that is notoriously unstable. Other information sources are
very reliable and unambiguous. But when weighing up the available information, people often
treat all the information as if it were equally reliable.


Perhaps one of the most powerful mistakes which can bias problem diagnosis is the tendency
known as "confirmation bias". Once we have a theory of what is wrong, we tend to search for
information which will confirm what we suspect. People however, rarely attempt to disconfirm
their s~~spicions
and in fact, often disregard information which w o ~ ~contradict
their ideas. An
example of this is the expectations created by stereotypes.
One final, interesting bias in thinking is the tendency when choosing between losses to often
prefer a possible loss to a certain loss, regardless of the seriousness of the losses. In other
words, to take a punt on the possibility of a large loss in'preference to a certain minor loss.
For example, a master may press on into bad weather although he knows his hatch covers
are suspect and risking the complete loss of the ship, or turning back or avoiding the weather
and facing the certain, but relatively minor, lengthening of the voyage. -The master under this
circumstance prefers possible disaster to certain inconvenience.
Case Studies
Two handy size bulk carriers grounded on the same reef within 18 months of each other.
In both cases they were chartered to load grain at a major port in South Australia.
However, both were diverted to a smaller port for. Neither ship had charts of the port
approaches. Both masters pressed on regardless instead of diverting to a nearby port
to pick up the required charts. Both ships grounded on Tippara Reef in the Spencer Gulf.
Automated systems
With simple systems, it is relatively easy to develop a mental conception to explain how it
works and predict what the effect of your actions will be. But with increasing levels of
technology, the human operator's understanding may lag well behind the capabilities of the
system and in fact the individual may have a mental concept of how the system works which
is adequate only when the system-i9 operating normally.
There are other potential problems with automation. For example, there is the fear that
people may lose their skills or may become complacent about the reliability of automated
systems. Humans are notoriously bad monitors yet automation is increasingly putting the
individuals into a monitoring role for which they are poorly suited.

Weighing up risks is an important part of decision making. Although risk will always be a part
of life, people sometimes underestimate risks. We are particularly likely to underestimate
risks when we have had a long stretch without an incident or accident, when someone else
has succeeded despite the risk (the 'follow the leader' syndrome or "we know its wrong but
we've always done itlgot away with it") and when we are excessively focused on the goal.
One final warning is that we tend to take more risks when we are under emotional stress
brought on by life events such as marriage problems or financial worries. For example it has
been found that people who are facing high levels of life stress are less likely to bother
wearing seat belts in cars.


Action and Skilled Performance

The activities of an individual can basically be divided up into three types of actions.
Knowledge based actions, rule based actions and skilled behaviour.
Knowledge based behaviour
Knowledge based behaviour is required when there is no pre-packaged solution to a
situation. 'The individual must think out a response using his experience or knowledge.
Knowledge based behaviour tends to be slow and very demanding of mental resources. In
effect, knowledge based behaviour is about thinking or decision making, often in unfamiliar
Rule based behaviour
We use rules constantly in everyday life, without necessarily being aware of them. These
rules are often procedures we have learned through trial and error and then apply to
situations in an "if..then.." manner. For example, if the dipstick on you car indicated that the
engine oil is low, then you would top up the oil. Although it is a conscious process, it does not
require you to go back to first principles in the way that knowledge based behaviour does.
People constantly apply rules, many of them formally laid down procedures.
Skill based behaviour
Skill-based behaviour is different to knowledge based and rule based actions because it is
unconscious, it is rapid, it is seemingly effortless and most importantly it is automatic.
Proficient drivers control their vehicle without havirlg to consciously dwell on every movement
of the steering wheel. Drivers can change gear, steer and work the accelerator without giving
it a moment's thought because these skilled actions have become automatic. We each have
an astounding repertoire of skill routines which we can consciously initiate and then leave to
run their course. The automatic nature of skill frees us to think about other things, but the cost
of this automaticity is ,that we monitor what we are doing less. Another problem is that skilled
operators are generally unaware of the automatic procedures they are following and may be
unable to explain how the skill is performed.
Skill acquisition
The knowledge, rule, skill distinction is also helpful in explaining how we acquire skills.
Typically we begin in the knowledge stage, by knowing what we must achieve but having no
pre-packaged routines to help us. Driving a car for the first time takes an enormous amount
of mental effort as we consciously think about moving the steering wheel and applying the
brake etc. As we become more familiar, rules begin to take over, such as "if the speedometer
goes over sixty, take your foot off the accelerator, if car doesn't slow down, apply brake". But
with time, all the control processes become mentally automated and we have spare capacity
to talk with our passengers and listen to the radio.
One of the most useful aspects of the Skill-Rule-Knowledge distinction is that it helps to
explain why errors occur and to predict the types of error that will occur under various
circumstances. Some errors (slips) are typical of skilled performance, other mistakes occur
with rule-based performance and other errors are typical of knowledge-based behaviour.


One of the most common skill errors is sometimes called "environmental capture" or habit
intrusion. This occurs when you are performing a well learned action in familiar surroundings,
but your routine action is no longer appropriate because the environment has changed. If you
then fail to make an appropriate attentional check, you may find that you have executed the
well learned action without modifying it to the new or unusual circumstances. A corrlmon
exan-~pleis filling in a form in January and, when dating it, writing in the previous year. This
sort of error is seen in the marine field when a pilot or master slavishly follow set pilotage
directions, even when the ship is out of position. (Sea Empress)
Another common skill error is the "Omission following an Interruption". If a well-oiled routine
is interrupted, it may never be completed, or may be picked LIP again at the wrong stage. One
of the most dangerous manifestations of this in the marine field is the interrupted position fix.
One of the final aspects of skill to mention is the speed accuracy trade off, in which errors
become more likely when time pressures exist.
Stimulus-response compatibility
Stimulus-response compatibility refers to the effect where the speed and accuracy of a
response is related to the physical arrangement of the display and the control. For example,
an arrangement where the display for the left engine is on the left and that for the right is on
the right is more compatible
+ .
than one where the arrangement is reversed.
The term "stress" encompasses a number of quite different human factors issues. While there
is no universally accepted definitiowof what stress is, a reasonable working definition is that
"stress is the body's response to stressors". It is important to distinguish the stress response
from the stimulus or "stressor" which triggers it.
There are environmental stressors such as noise and heat, task stressors such as
demanding, time pressured tasks, and there are life event stressors such as financial
uncertainty and domestic worries. Different stressors can lead to different stress problems.
Life event stress..resulting from significant life events such as divorce, financial worries and
the like can reduce general well being and increase the susceptibility to some illnesses.
People who are experiencing such events may be may be distracted by iritrusive thoughts,
particularly when workload is low. In addition, people who are experiencing life stress may
take more risks than they otherwise would. US research has found evidence that Navy pilots
who had accidents were likely to have had a higher level of "life events" in the months
preceding the accident than pilots who had not had accidents.
Task stress, or acute stress arises when the demands of the task approach or exceed the
capabilities of the person. An individual coping with an emergency may be affected by task
stress. An important idea related to task stress is the "inverted U curve", as illustrated on the
next page. According to this idea, performance is best when the person is moderately
challenged tjy the task. Too little challenge, (resulting in boredom) and too much challenge,
(resulting in panic) both result in poor performance. So a moderate level of task stress can
be quite helpful.


Inverted U Curve: Performance versus Arousal





Arousal Level


Holmes and Rahe Social Adjustment

Rating Scale

Death of spouse


Marital se~aration
Jail term
Death close familv member
Personal injury or illness



Marital reconciliation
I Health chanae familv m~mberI
Sex diff ic1.1lties
Gain of new family member
Business readiustment
1 Chanae in financial state 1
I Death of close friend
Change in line of work
Foreclosure of mortgage
Change in work responsibility
Son or daughter leaving home
Trouble with in-laws
01s personal achievement
1 Wife beains or s t o ~ swork 1
Beain or end school

Minor legal violations











Open sea - featureless

On duty with a sleep debt built

up over previous days

More than 16 hours

wakefulness before duty

Monotonous conditions

On duty with a sleepiness

condition (eg. sleep apncea)


Length of duty before


Steady low level

background noise

On duty in normal sleep time

causes drowsiness

Those with medical conditions

(eg. narcolepsy, obesity,
on medication)

Pattern of irregular duty

Regular visual patterns

(e.g Radar scan)

Predisposition to' sleepiness

After consuming alcohol

Early afternoon drowsiness

Gentle motion of ship

On duty after poor


Younger watchkeepers up to
25 years

Early morning duty


Older watchkeepers over

50 years

Inadequate sleep or rest

before duty


Ambient temperature and

level of humidity


Fatigue and Casualties

Proving that, on the balance of probabilities, fatigue is a causal factor in a marine casualty
can be difficult.
Some attempts have been made to establish an index to indicate the probable presence of
fatigue as a factor in an accident.

The sleep creditldeficit system

This is a simple system of assessing an individual's hours of sleep and wakef~~lness

over the
preceding days, preferably at least 72 hours (longer if possible).
Over the days that data is available credit two points for every hour asleep and debit one point
for every hour awake. The chart below shows only a 50 hour analysis.

Pilot's Sleep Credit Deficit Chart

(assuming nil deficit 2200 on 14 May)
13 May

14 May

15 May

16 May



This system does not account for physical or mental exertion and is an index of possible
fatigue factors. The information can be captured on a pro-forma.


(X, Time accident;

F,Meal; W, Watch; M, Maintenance work; S, Sleep; C, Cargo Watch; R. Recreation, including time ashore; A, Alcoholic


('The following is adapted from a paper, "Things that go bump in the night", given by
the Australiar~Inspector of Marine Accidents at the Third International Conference on
Fatigue in Transportation, "Coping with the 24 Hour Society", Fremantle, Australia,
February 1998)

Fatigue lndex Score

The following index has been devised by consultants to the US Coast Guard.
In its efforts to understand the scope and depth of the issues surrounding human
performance in the maritime industry, the US Coast Guard has sought to improve the
manner in which data is being collected and reported by marine investigating officers.
The initiative intends to improve the quality, accuracy and uniformity of reports being
filed through normal accident and incident investigative procedures. A research report
entitled "Procedures for Investigating and Reporting Human Factors and Fatigue
Contributions to Marine Casualties," (McCallum, Raby and Rothblum), now exists.
Marine investigators are being educated in the application of the findings contained in
this document, as well as human factors overall.
One of the sections of this report deals with fatigue. Specifically, a new and scientific
approach to estimating the probability of fatigue that is present or causal to an incident
has been proposed. This approach uses a formula that is based on mathematical
analysis of accident and incident data (a regression analysis of the data collected in a
Coast Guard study of fatigue-related and other casualties; N=89, 29 of which were
considered fatigue related), already contained in the Coast Guard database.
Application of this formula is a value known as the "Fatigue lndex Score".


The purpose of this computation is to give investigating officers a general tool by which
to evaluate the presence of fatigue observed in persons associated with the incident.
It is not (yet) considered an appropriate tool to estimate the magnitude of fatigue
present, however, though some day this may be possible. Application of this
formula has been tested and found reliable in correctly identifying fatigue as causal to
the incident in about 80 per cent of the trials. While this is not 100 per cent, and
perhaps no measure of personal fatigue will ever be so, a tool that may be used to
confirm or rule out the causal relationship of fatigue or other factors to the 80 per cent
mark, is no less significant.
A simplified version of the Fatigue Index Score formula is as follows:
FIS = (WH x 6.1) - (SH x 21.4)

WH = the number of hours WORKED in the 24 hours before the casualty
SH = the number of hours SLEPT in the 24 hours before the casualty
S = the number of fatigue SYMPTOMS experienced by the individual while
on duty before the casualty

Symptoms :
a Forgetfulness
a Less motivated
a Difficult to keep eyes open
a Sore muscles
Desire to sit or lie down
a Difficulty operating equipment
For example, a person who has worked 8 hours in the previous 24, slept 6 hours in
the 24, and reported one of the seven symptoms would score:
FIS = (8 hrs work x 6.1) - (6 hrs sleep x 4.5) + 1 symptom x 21.4 = 43.2
Understanding the results of the FIS is a bit more complex, however. If the FIS is > 50
there is an 80 per cent likelihood that fatigue was a cause in the incident. For a score
< 50, there is similarly an 80 per cent likelihood that fatigue was not a cause.

While the formula is useful, there are certain factors not apparently considered. Such
as timing of work or sleep as compared to the body's circadian cycle, quality of sleep
and other environmental factors such as heat and cold. 'The formula has a predictive
value, particularly in highlighting areas of work-rest patterns.
Essentially the FIS may indicate whether a fatigue condition may exist, the investigator
must then test the probability against any other evidence available. Investigators
should also check the factors contained in the Annex to the Code dealing with "The
Investigation of Human Factors" and consider using both measures outlined in 1 and
2 above. (Editor's note)


New measures have now been adopted into the International Conver~tionon Standards of
Training, Certification and Watchkeeping for Seafarers 197411998, Chapter Vlll of the Code
to the Convention requires that all persons should be fit for duty:

All persons who are asslggned duty as officer h charge of a watch or as a
rathg formhg part of the watch shall be prowiied a minimum of 7 0 hours of rest in
any 24 hour p e h d

The hours of rest may be &deed lhtono more than two period, one of which
~ length.
shall be at least 6 h o u in

The requirements for rest p e w s lad down h paragraphs 7 and 2 need not
be maintained /i7 the case of an emergency or dri!! or h other overriding operational
NoWithstanding the provisions of paragraphs 7 and 2 the minimum p e h d of
ten hours may be reduced to not less than 6 consecutive hours provided that any
such reduction shall not extend beyond two days and not less than 70hours of rest
are provided each seven day penbd.

Admhistrations shall require that watch schedules be posted where they are

I n a survey by NUMAST, the United Kingdom's maritime officers union, it was found that 63
per cent of officers were working in excess of 76 hours each week.ii Although the hours of
work regulations were known and two thirds of the companies had introduced schedules to
ensure compliance with the regulations, 92 per cent of officers found that there had been no
reduction in hours worked. This is not surprising, as a strict interpretation of the r ~ ~ l provides
a legal working week of 98 hours.
There remains an overarching convention under the STCW Convention that each
administration shall, for the purposes of preventing fatigue:

establish and enforce rest p e w s for watchkeepingpersonnet and
require that watch systems are so arranged that the efficiency of a!!
watchkeeping personnel is not impaired by fabgue and that duties are so organized
that the fist watch at the commencement of a voyage and subsequent relievhg
watches are suh7iciently rested and othenuse fit for dutyi
The regulation of this requirement is placed on the flag State and refers to watchkeepers and
may be read as excluding masters and others critical to operational safetyaboard ship.
Statistics from Australian casualty investigations conducted between 1923 and 1995 showed
that casualties were most likely to occur between midnight and 0400 and yet least likely to
occur between 1200 and 1600, despite the fact that on the four hours on watchleight hours
off watch system, the same officer keeps both watches.


Since 1982, 33 per cent of collisions and 24 per cent of groundings investigated under
Commonwealth jurisdiction have occurred between midnight and 0400.
Collisions and Groundings


,-& Percent

Watch Periods

To a simple mariner this statistic seems to support the obvious, sustained performance in the
early hours of the morning is modified by tiredness; and tiredness is a manifestation of
fatigue. But the problem, particularly in the transport industry and most particularly at sea, is
to define and understand what we are talking about and to be able to convince, not only the
managers and administrators, but our peers, that fatigue is something more than a defect in
character, a weakness in our physiological and psychological make up. We also have to
convince the industry that fatigue may be termed inelastic, you cannot trade it for more
money or longer leave. If one is tired and under-performing now, three days extra leave a
month hence will not make one operate more safely!
In investigating marine casualties we have a problem of identifying, in a reasonably objective
fashion, whether or not a person was suffering some level of fatigue and that the fatigue was
an element in the causal chain. Experience has led us to believe that chronic fatigue is the
most corrlmon and insidious form. Chronic fatigue could explain, what is often inexplicable,
why a person's performance is below that which one might reasonably expect.
We have investigated a number of casualties, usually groundings, where a normally
competent, normally efficient person has literally drifted into trouble. The problem seems
much more complex than simply lack of sleep and one is left with those intangible issues of
stress, boredom, health, familiarity and personality. The inherent variability of peoples
psychological and physical make-up means that there is no "one fits all" solution.
Since 1 January 1994, the Unit has investigated 65 incidents, in only six (9.2 per cent) of
which could we say with absolute confidence that fatigue was one of the causal factors.
Herein lies another difficulty. Fatigue may be only one of a multitude of factors contributing to
the casualty. Despite the long hours of duty worked by ship's officers in particular, our


statistics would suggest that less than ten per cent of casualties involve fatigue. Of the six
casualties, involving identifiable fatigue, one involved structural failure through the actions of
an officer of the watch in port, one involved a collision with a ,fishing boat and four involved a
ship running aground, the latter five all occurred between midnight and six in the morning.
In the summer of 1994 a tanker was loading cargo at a refinery in Austra1ia.i~As a result of a
chain of factors involving unsafe acts, omissions, inadequate training, personality issue,
ergonomic design and an individual's fatigue level, a ship's cargo tank was over-pressurised
and it ruptured. The repair bill was costly and the incident could have involved a fire and
A number of factors combined to cause a particular officer to fail in four routine tasks he
should have been able to perform, regardless of other latent, underlying elements. We looked
at his off duty pattern and found that in the 62 hours before the accident he had had not more
than 15 hoirrs sleep and not one period of sleep was more than 5 hours duration, taken
between 0600 and 1200 hours in the morning, despite a firm company policy to combat
fatigue. We do not know what quality (hygiene) of the sleep he may have enjoyed. He
appeared conscientious and possibly over confident. Much of the opportul-rity to rest was
forgone by his own choice. On the two previous evenings he had gone ashore and returned
to take up his midnight to four watch and on both mornings he voluntarily extended his watch
to 0600, for various reasons.
He did not fall asleep on duty but it is reasonable to suspect, though I cannot prove, that his
arousal level was depressed. At 1420 in the afternoon he made a nurr~berof sirr~plemistakes
which led to the structural rupture of the tank.
About a month after this incident, an offshore supply vessel sat on a wellhead in shallow
water off Western Austra1ia.v The vessel was working under the direction of a -pipe-laying
barge on a fixed price contract - time was money. The work involved towing out pipelines
from a shore fabrication plant, tendirlg a construction barge (ferrying stores, picking up and
moving the barge's eight anchors), all in a relatively confined area and in shallow water. The
support vessel's program required a high level of performance from the master and crew over
a prolonged period in very confined and shallow waters.
Such vessels are small and noisy, with the strong probability of resulting poor sleep quality.
The master routinely worked between 14 and 18 hours a day in an irregular pattern. He had
been on the vessel for about three weeks. During the evening before the accident operations
ceased at about 1845 and the master, realising that the vessel would be required sometime
during the night, went to bed at 2100 after a meal. The call came at midnight for the vessel
to deliver some equipment to the barge and recover and redeploy the barge's anchors. The
first anchor was positioned successfully. The second anchor was positioned close to a
redundant wellhead, just below the surface but unmarked, except on the chart. While
recovering the anchor, it seems that the anchor wire may have fouled the wellhead. The
master had the problem of an unmarked well head, intermittent rain, high humidity on a warm
night, he was being given directions from the barge, and his crew working with heavy wires
and anchors on the vessel's work deck. A little before 0300 the vessel hit the well head and
The fatigue sub-set of factors included not only tiredness, but probably information overload,
stress and .the time of the day in addition to a large number of other factors, many latent in
nature. The vessel was judged beyond economic repair and, after being salvaged, was


In March 1995 a small cargo ship grounded on South Ledge Reef at the northern end of the
inner route of the Great Barrier The vessel carried a pilot, who, as is normal was
resting on a quiet, straight stretch of the passage. He left clear instructions with the second
mate that he should be called at a position marked on the chart which also was clearly
marked with the ship's planned route. At 0400 the second mate was relieved by the mate who
was thoroughly briefed on the courses to be followed and requirement to call the pilot. This
vessel was also equipped with a "dead-man" alarm as it was designed to operate as a oneman bridge operation at night. The "dead-man" alarm was operational, set at ten minute
periods to sound the ship's general alarm if not cancelled, in addition there was a seaman
lookout on duty throughout the night.
The pilot was not called. The alteration of course was purely routine, one that a competent
navigator could perform with histher eyes closed and, on this occasion, it seems that the
navigator's eyes were closed. As you have probably anticipated, the ship did not turn the
corner and ran straight up on the reef and grounded at 0500 without the alarm sounding.
'There was some'local damage to the reef, but no oil was spilt, nobody was injured. The ship
itself was significantly damaged with buckled frames and plating.
The mate had not worked any extraordinary hours (his 4-8 watch and work during the
forenoon) and was keen and conscientious. However, the night before the grounding, while
on passage with a pilot in charge of navigation, instead of going to bed at his normal time he
had watched a video and had some drinks (beer and wine) before going to sleep just before
midnight. He was called at 0348. Instead of arriving on the bridge at 0355, he was
uncharacteristically5 to 10 minutes late. He appeared normal and the hand over of the watch
went routinely. The mate was seen to sit at the bridge console chair at 0430 after he had fixed
the ship's position.
Four months after this a small cargo ship running between the Fly River and Townsville
grounded on the aptly named "Murdering Point" at 0600 in the morning of 24 June.vii
The ship had left Townsville the previous evening after a short, 15 hour, period in port. Before
arriving in Townsville early on 23 June it had experienced rough weather and over a period
of five days had constantly pitched and rolled heavily. In port the second mate had spent the
morning ashore and then was on duty in the afternoon. The ship sailed at 2000 and the
master (this was a two mate ship) stood the remainder of the 8-12 watch. 'The second mate
relieved him at midnight when he seemed alert and fit. 'The master left the bridge promptly to
go to bed and left the second mate alone (without a lookout) on the bridge.
A little after 0100 on 24 June, the second mate felt a little cold and left the bridge to collect a
jacket and his cigarettes from his cabin. He sat down to open a drawer beneath his bunk. The
ship ran aground about five and three quarter hours later. The Mate had not been called for
his 4-8 watch. The Second Mate had fallen asleep sitting in his cabin, while bending down to
get cigarettes from a drawer.
The master had flown from Copenhagen, without stopovers, into Townsville 48 hours before
joining the ship at 0900 on 23 June 1998. He had worked from 0900 through to midnight when
he handed over the bridge watch to the second mate.
The mate had been on the ship for the rough south-bound passage. He had been awake from
0400 on the day before arrival at Townsville to about 2300 on 23 June. In 43 hours he had
had not more than seven hours sleep, a factor which had to be added to any accumulated
sleep debt through poor sleeping conditions in the rough weather.


The second mate had eight hours sleep in the previous 48 hours, with one hour in the last 24
hours. There were, however, periods when he (like the mate) was off duty. He could have
rested on the morning of 23 June. However, the noise of loading a small ship limits the benefit
of such an option. Anyway he was fed up with being on the ship and wanted to stretch his
legs. He could have slept for about two and half-hours after 2100 when the pilot disembarked.
Both he and the mate elected to watch a rugby league match on television. The ship itself
was "dry" and alcohol was not a contributing factor.

In this case the three qualified officers were suffering some identifiable fatigue. It was little
wonder that the master wanted to get to bed as soon as possible, his body would not have
overcome jet lag. It was little wonder that the mate slept beyond 0400 when he would
normally have gone on watch. And it is little wonder that the second mate was fatigued. The
ship sustained little damage, but the secorid mate lost his job as did the mate because he
had had a drink ashore during the afternoon.
A year later another small cargo ship grounded in the Great Barrier Reef, after the pilot fell
asleep.viii 'The investigation showed that the pilot was into his 26th day.on-roster and for 12
nights of the 26 days he had had little or disrupted sleep and over the previous ten days he
had had six disrupted nights. Again, the ship itself suffered little damage.
Finally on 25 April 1997, a fishing boat was in collision with a Liberian tanker.ix It was one of
our "hit and run1'cases, but the ship was traced through paint samples taken from three ships
known to be in the area.
The second mate was on watch when the collision occurred and he claims no recollection of
the event. We cannot say whether he was lying or not, however it was established that he
had left Croatia and flown without stop-overs via Zagreb, Prague and Singapore to Sydney,
where he arrived at 0800 on 22 April and was taken straight to the ship after 39 hours in
transit. He arrived on the ship at 1000 and immediately took up a six-onlsix-off port watch
routine, which he maintained until the ship sailed at 0620 on 24 April. Thereafter the watch
routine reverted to the traditional three watch basis with the second mate keeping the 1200
to 1600 and 0000 to 0400 watches.
Trying to make sense of it all
There i'hdegree of commonality in these casualties that is worth thinking about.
Firstly, the fatigue state of the operator in five of the six incidents occurred between 0000 and
0500. The tanker structural failure occurred at the time often associated with the circadian dip
in the afternoon.
Secondly, in five of the six cases the operator on the bridgelcontrol room, or in one case in
his cabin, was sitting when the incident occurred. In the sixth case a chair was available and
we can have no more than a suspicion that it was in fact being used.
All personnel directly involved were reported as being keen and conscientious and were
involved in extended hours of duty. There was that sort of bewilderment, "I could understand
this happening to somebody else, but not him". In two cases the officer's involved had had
opportunities to obtain some additional rest, but had rather elected to go ashore or relax
watching television. Taken without consideration of any circadian rhythm effect, in the case of
the tanker accident the quality of rest may have been reasonable on a large ship without
cranes and solid cargo being loaded. In the case of the small cargo ship, it is doubtful whether
much quality rest would have been forthcoming with human traffic in the small
accommodation space and the noise of the ship and the loading of cargo.


In terms of the hours of work provisions under the STCW Code, all the main players,
with the exception of the offshore supply vessel master and the pilot, were working
within the prescribed hourly limits. The master of-the offshore support vessel and the
pilot were not watchkeepers and hence the provisions of the code do not apply.
Two casualties involved individuals who had flown from north-west Europe on a direct flight
involving between 20 and 39 hours continuous travel. One, the master of a small cargo ship
had 45 hours in Australia before joining his ship, while the other, a second mate immediately
started six hour watches on a watch-ontwatch-off basis. Unfortunately ship owners tend to
view rest and recuperation periods following an expensive flight as something of an
indulgence. It is hard for ship managers, with one or more crew sitting with their feet up in
some exotic location for days at a time to accept the reality of circadian rhythms and the risk
of fatigue, after all they are resting at the company's expense.
If fatigue is so much a part of life at sea, why are not there more accidents? Paradoxically,
while thesea can be a harsh environment it can also be fairly forgiving. The sea covers a
large area and ships in the open ocean have to be fairly unlucky to collide. I hasten to add
that they do from time to time. Modern technology also has increased the risk is some ways.
Ships will generally follow the most direct route port to port, with electronic navigation, global
positioning systems and other modern aids giving much more accurate and more frequent
position up-dates, the sea lanes are becoming narrower, increasing the chance of two or
more ships being in the same place. Two or more ships in the same place at the same time
with fatigued crews poses a real hazard.
In the United States of America, one per cent of casualties had been identified as being
related to fatigue. However, a research project by the United States Coast Guard and
Battelle has produced data suggesting 28 per cent of all marine casualties are fatigue
In Australia we can put our hands on our hearts and can show 9.2 per cent of casualties have
a causal element we can identify as fatigue. We believe that the real figure is closer to the
US figure if we consider chronic fatigue because performance levels have the potential to
suffer in routine of shipboard life.
Most ships are "slow systems" operating at relatively slow speeds, but with huge inertia. On
the bridge, good seamanship is more to do with anticipation and mental projection mixed with
experience rather than academic ability. Attention has to be maintained at a sustainable level
in conditions of low activity for hours at a time.
'The m a h e hdusty is remarkable for the number of key positions that requ2e such
un/i7terruptedand sustained atfenbon.''xi
Cues and prompts may be dispersed and infrequent. The sense of danger is often absent in
the slow system (until too late). Watchkeeping at sea tends to be sedentary, routine and
monotonous and watckeepers are subject to the circadian cycle. Seafarers often work with a
"sleep bank debt". Watchkeepers tend to drink large quantities of tea and coffee, often to
offset boredom; generally they operate in a dark environment for 50 per cent of their working
day. In tropical Australia the night time temperatures can be described as balmy; background
sounds are those of water against the hull and the soft hum of instruments; and I really do
not want to comment on aroma.
In port officers routinely work longer hours, the work is more intensive with greater demands
on their time on duty and, often, their time off duty.


We know that human fatig~~e
is a significant factor in many marine casualties and fatigue can
kill. We know what fatigue is. How can research into hours and work or even smells tell us
any more?
Following a collision between two ships off a New Zealand port, one of which was registered
in Vanuatu, Don Sheetz, the Deputy Commissioner of Maritime Affairs undertook an
investigation into the circumstances of the collision from the perspective of the Vanuatu flag
ship. He showed that the master had had a total of 10 hours sleep during the preceding 53
hours and his deck officers, working watch-onlwatch-off had had even less. He concluded
that impaired alertness had contributed to the actions of the bridge team and none of the
officers had had sufficient rest.xii More importantly he warned the owners that his
"will wish to review their vessel manning, and on board operatingpractices, to
ensure there are sufficient personnel to handle workloads, whfle prowding adequate
rest pehds. %iii
Ships will, for the foreseeable future, rely on seafarers operating within a twenty four-hour
society. It seems to me from the investigations that we have done that the IMO provisions
must be enforced and policed in the manner adopted by Vanuatu, and indeed by AMSA, but
we must go further. Companies with positive policies to prevent fatigue should be supported.
In some way the much vaunted "playing field should be leveled. 'Those operators that have
no policy on fatigue prevention, and who still rely solely on a combination of an individual's
(often misplaced) pride, desire not to let others down, fear for future employment and
(probably most importantly) luck, are identified and penalised.
The depressing fact is that without something more than lip service to the problems of fatigue
ships'will continue to "go bump in the night" with monotonous regularity.
Marine Incident Investigation.Unit
January 1998
The International Convention on Standards of Training, Certification and Watchkeeping for Seafarers, 1978,
as amended in 1995, Chapter VIII. Section A-VIII11
ii NUMAST (1997). Give Us a Break. A report on the results of a survey of the hours of work undertaken by
members of NUMAST 1997
iii STCW Convention. Regulation VIII11
iv Incidents At Sea No. 63 (1994). Departmental lnvestigation into the structural damage sustained by the
tanker Osco Star at the port of Kwinana, W.A on 19 January 1994. MIIU.
v lncidents At Sea No. 66 (1994). Departmental lnvestigation into the sinking of the off-shore supply vessel
Boa Foirce on Saladin No.3 wellhead off Thevenard Island, Western Australia on 24 February 1994.MllU.
vl lncidents At Sea No. 79 (1995). Departmental lnvestigation into the grounding of the German flag container
ship Carola on South Ledge Reef on 30 March 1998. MIIU.
vii Incidents At Sea No. 82 (1995). Departmental lnvestigation into the grounding of the Danish flag ship
Svendborg Guardian south of Murdering Point, Queensland on 24 June 1998. MllU
vlil Incidents At Sea No. 95 (1997). Departmental lnvestigation into the grounding of the Panamanian flag
refrigerated cargo vessel Peacock on Piper Reef, in the Great Barrier Reef, on 18 July 1996. MIIU.
IX lncidents At Sea No. 116 (1998). Departmental lnvestigation into the collision between the fishing vessel
Exterminator and the Liberian tanker Unisina off the south coast of NSW on 25 April 1997. MIIU.
x Lloyd's List, page 3, September 4, 1997
Sirois,W., (1995) Alertness Assurance:The Key to Reducing Fatigue and Human Error in the Marine
Industry. American Petroleum Institute and the United States Coastguard. Prevention Through People
xll Sheetz, D.(1996) presentation at the 6th meeting of the Marine Accident Investigators' International Forum,
New York, June 1996.
xlll Lloyd's List. May 30, 1996, page 8.

Collecting Evidence



It is not the purpose of these notes to provide a full explanation of "evidence". For the
investigator all that is necessary is a general understanding of what evidence is and the
general types of evidence and the sources from which it can be obtained. The distinction
between the types of evidence is not clear cut.

Ewb'ence is something that furnishes, or tends to furnish, proof of something. A more

legalistic definition is:
All the legal means, exclusive of mere argument, which tend to prove or disprove any
matter of fact, the truth of which is submitted to investigation.
Proof A fact is said to be proved when the inquiry is satisfied to its truth.
In generamhe person making an assertion (often the investigator) must furnish the proof.
Again this is a chain like structure, where the proof of a chain of facts leads to an overall
proof. Any disproved or ambiguous link in this chain must be revisited, analysed and
objectively reconciled.
\ .



The standard of evidence

In many societies there is a clear distinction between the certainty of evidence required in
criminal and civil cases. Where a person may be subject to punishment, particularly
imprisonment (or worse) the burden of proof required is often that "beyond reasonable doubt".
In civil cases a less onerous level of proof "on the balance of probabilities" is the standard
required in many jurisdictions.
In formal tribunals, examining marine casualties, this latter level of evidence would normally
be adopted.
The task of a marine casualty investigator, responsible for determining the circumstances and
causes of a casualty, is to collect all evidence and analyse its value. It is essential that all
evidence obtained is properly analysed and evidence is not discarded just because a piece
of evidence does not fit some preconceived theory.
An investigator should adopt the same principle as the civil burden of proof. Not being a legal
expert, this may sometimes be difficult to judge. The final test is has the investigator treated
the evidence in a fair and professional manner.

Types of evidence

While not essential it can be useful for an investigator to have some idea of the types of
evidence and some idea of their relative value in establishing a reasonable level of proof.

Direct or Real Evidence

Physical things, evidence in the form of the thing itself. A broken pipe, a failed flange, a
broken anchor chain, etc. Often real evidence cannot be disputed and is very valuable in
provirlg facts.


Personal evidence
What the investigator is told by the witnesses, the record of interview. The value of personal
evidence depends upon the quality of the witness and the ability and skill of the
Ifsigned and witnessed the written transcript or the tape may become documentary evidence.
Documentary Evidence
Evidence provided by a document such as certificates, log books, belllmovement books,
letters, ma.nuals, photographs, etc.
A document is anything on which signs or symbols have been marked and includes tape
recordings, computer tapes or discs.
Care should be taken that the document is taken for what it really is. A safety certificate is a
document attesting that a surveyor signed the certificate to say the equipment met the
required standard. It does not mean that the surveyor actually examined the equipment or
that the equipment remained to standard. Similarly, a log book is a record of what was written,
rather than what actually happened.
Original documents are of far greater value. Photocopies can be tampered with.
Contemporaneous records (written at the time) are far more valuable than fair copies (as with
the deck log and the mate's fair copy). Investigators should always site original documents
and satisfy themselves that any copies they may make or notes they take are accurate.
Original documents are useful but not infallible.
For Formal investigations, original documents are prefened at all times. It would be a decision
for the judgelboard as to whether or not the-inquirywould accept photo or other copies.

PNme Facie Evidence

At first appearance. Evidence which is sufficient to establish a fact in the absence of any
evidence to the contrary. If two ships collide there is "prime facie" evidence of a breach of the
Collision Regulations, if a ship founders there is "prime facie" evidence that the ship was
A fact from which an investigator may reasonably and logically infer the existence of a further
fact relevant to the incident. It is indirect or presumptive evidence showing a strong
connection between evidence and the incident. A trace of a ship's paint on a fishing boat is
strong circumstantial evidence that the ship matched with that paint was in contact with the
,fishing vessel.
Circumstances equally capable of another reasonable explanation is of limited value unless
reinforced by other evidence.
In investigations circumstantial evidence may, in itself, be of limited value, but may point the
way for further inquiry.


Hearsay Evidence
Evidence through a third party - "second-hand evidence.
Under certain jurisdictions "hearsay" evidence is not admissible in criminal or other judicial
proceedings. In a technical inquiry such evidence may be important, but it should always be
corroborated to the maximum extent possible. This is particularly true where grudges
between parties may exist or where somebody may have a particular interest.
Accident reports, based on unsworn evidence and documents which were seen by the
investigator but not produced, are hearsay and may not be admissible in formallcourt

Expert Evidence
Given by a person skilled and experienced in a specific professional or technical area, able
to give evidence based on hislher knowledge from facts reported to himlher or discovered by
himlher by tests, measurements or similar means.


' 7-

The reports of tests on a ship's equipment, or reports on paint analysis etc, is expert
evidence. Expert evidence is only as good as the expert who gives it. In an adversarial court
much time can be taken with testing an expert witness.


Collecting evidence

Evidence is collected by the investigator by what helshe can hear, see, smell and feel.
Very often but not always, what witnesses tell the investigator can be the core of information
and direct the initial direction of the investigation. This "Personal evidence" is only as
accurate, objective and truthful as the person who provides it. Wherever possible, what an
investigator is told should be tested against any other evidence available; other personal
evidence, circumstantial evidence, but most importantly against real evidence and
documentary evidence.
When gaining access to the ship, or ship's staff or representatives, as much real and
doci~mentaryevidence should be collected in an orderly manner. This may be done by
acquiring documents or objects themselves or by copying, photographing, recording.

When evidence is acquired it is important that it can be accounted for from the time it is
acquired to the time that the investigation is complete. Police refer to "continuity of evidence".
Although it is essential for police, when gathering evidence for prosecution, to be able to
/account for each person who may have had access to the evidence, it is a useful discipline
for any investigator to be able to account for evidence and be able to refute any allegation of
tarr~peringwith evidence. Proper handling of evidence lends credibility to the investigation
and makes the oversight of the investigation easier.

Collecting documentary evidence

The acquisition of log book entries, bell books, standing orders etc are routine in all
investigations. In the majority of investigations the ship will continue in service and copies of
the desired documents will need to be made. Some administrations require that any copy of
a document should be signed by the Master (or relevant person) as a true copy of the original


When examining log books or any bound book, it is often useful to count the pages. Standard
issue log books have a set number of pages and a shortage of pages may be crucial
Taking copies of the ship's chart can be problematic and will depend upon the investigator's
powers under national legislation. Investigators can often experience fierce opposition to
anybody other than the ship or their legal representatives retaining charts etc. Where there is
no power to take original documents, or even when there is, investigators should always try
and have an up-to-date chart in their possession to leave on the ship if the ship's chart is
It is useful, though not always possible to obtain the log book, bell books, data print-outs,
crew lists, plans, certificates etc. before questioning the witnesses. Certainly a study of any
document that was in use at the time may prompt questions or open up other avenues of
An investigator may become suspicious that information.from a written document may have
been erased or removed. Scientific forensic examination of the doc~~ment
in question will
often reveal such text.

.Automatically recorded data

Automatically recorded data is excellent evidence. Automatic engine movement print-outs,

course recorder charts, engine logger print outs (remember to obtain a list of all the ship's
channels), pitch graphs, discharging rate graphs, cargolstability print-outs may also be
available and relevant.
Computers are used increasingly. Remember computer discs are documents.
Also GPS receivers are normally programmable. Down loading or othetwise obtaining the
information on the program, both programmed way-points and actual track can be very
Some modern radars have a replay facility.
Some fire alarm systems (SALVICO) have a memory recording the time of alarms.

Use of cameras

'The old saying that a camera cannot lie is no longer true. However, photographs of the ship,
ship spaces and evidence (eg. broken flange) are invaluable as evidence, as prompts during
analysis and for illustrating the report. Cameras should have a date time code. The use of
cameras will also be discussed under collection of samples.
Video cameras are also a very useful investigative tool. Still photographs can be obtained
from video. Video records are much like the human eye. If one looks at a scene, the eye sees
everything but the brain does not necessarily take the whole scene in. A video allows an
investigator to view a scene multiple times and features missed on first viewing are not lost.

Note books

Note books remain the investigator's most useful tool. Some authorities require special note
books with numbered pages. Other authorities do not. Note books are useful for sketching


and recording measurements, recording events, making statements and having them
countersigned, and for the investigator to map out a plan of the investigation, areas of and
actual questions. In the absence of any other form they are a means of recording questions
and answers at interview.

Tape recorders

Small hand held tape recorders are useful for recording an investigator's observations and
recording sounds. They are also useful in a "hands free" interview process. Special
"verifiable" tapes and recorders are available which carry a signal to show whether or not the
tape has been edited after the interview or otherwise tampered with.
Fast dubbing equipment is also available which allows rapid recording of the tape so that the
interviewee can be furnished with a copy of the interview recording. There are also multi deck
tape .reco,rderswhich record two or more tapes simultaneously.
9.1 0

Collecting directheal (physical) evidence

'There are three golden rules for collecting direct evidence:



Photograph the thing in situ with a datettime camera.

Note the details in the investigator's note book.
Place in an appropriate sterile container, label date, signed and sealed.

Real evidence can vary from paint or oil samples, to wire or mooring ropes, pieces of steel,
machinery parts, fire damaged material, the list is as long as the ship's inventory of parts and
A rule of thumb is that paint samples, books for forensic testing, materials, etc should be put
in clean, sterile plastic bags. Where the evidence is too large.for such a bag then alternative
means of securing the sample may have to be made. A large plastic bin may be suitable,
by covering. If this is the case a note
otherwise the sample may have to be taken ~.~nsecured
should be made in the investigator's note book.
Any evidence sample that may give off hydrocarbons (oil samples, or samples of material that
may have some fire accelerant solution in the sample) should be secured in clean sample tins
with a secure top. Hydrocarbons can "breathe" through plastic. Often paint suppliers can
provide such tins.
When evidence samples are sent for testing or analysis it is sensible to have a documentary
trail to show who received what and when and who removed seals and resealed containers.
(See earlier continuity of evidence.)

Use of outside experts

Few casualty investigation bodies would have expertise in all ,fields of investigation.
Investigators should always consider engaging expert advice on subjects which are central
to the causal chain and in which an investigating team may not have full qualifications.
Alternatively, where the investigator may be competent through experience but lack
qualifications, it will strengthen the credibility of the investigation if that technical part is
reviewed by an expert. For example if the ship's stability is an issue, most nautical
investigators will be able to examine the stability aspects, however, if this is endorsed by a
qualified naval architect the evidence will have more credibility.


Other experts who may helpful are doctors, psychologists, metallurgists etc. Suitable experts
may be found through consultation with the police, air accident investigation bodies, or
internationally through the Marine Incident Investigators' International Forum.

Information external to the ship

Remember a marine casualty occurs as part of an "open system". Pilots, VTS operations etc
may be involved. If not based in the country of investigation, a ship's owners may be hard to
obtain information from. A request to the investigating authority in that country may result in
One of the important aspects of coastal Statelflag State cooperation is the exchange of
information between the two bodies.
Remember also that Class Societies will hold information. Class societies will not usually
release information without the consent of the owner. However, if class has issued a
certilicate on behalf of an administration, the inform-ation upon which such certificate was
issued is a matter for the flag State.
It is a case of exerting as much influence as possible on owners and class to obtain any
necessary information.
Class Society rules are published (some on CD Rom) it can be useful to compare what one
.finds on a ship with class rules and IMO convention requirements.

10 Determining Occurrence Sequence, Analysing

10.1 Introduction
This part of the course is designed to assist investigators in conducting investigations in
accordance with IMO Resolution A.849(20), the Code for the Investigation of Marine
Casualties and Incidents, and IMO Resolution 884(21), Amendments to the Code for the
Investigation of Marine Casualties and Incidents.

The course will provide the participants with a systematic approach to marine casualty and
incident investigations. At the end of the course the participants will be confident in following
the systematic. As stated in Section 2 Res. A 884(21) (Unit 4.6 - section 2) the essential
sequence of events for investigating any accident are the following:


ldentify Unsafe
Acts and Unsafe

ldentify the Error

Type or violation

Identify Potential Safety

Problems and evel lo^
Safety Actions

Identify Underlying


10.2 Principles of accident analysis

The one overriding principle is that any analysis must be firmly based in the evidence
gathered in the investigation. This section of the course will help you arrange evidence and
analyse it systematically.
We have already established that any accident is a sequence of events, which can be shown
as a time line or series of parallel time lines. The start of the time line can be problematic; the
termination of the time line, at least in the first instance, is the time of the casualty itself.


10.3 Building a hypothesis

One of the main reasons for using investigators skilled in the maritime disciplines is that they
are able to empathise and will understand how an accident may have occurred. This allows
an investigator to build a hypothesis as evidence is gathered.
Two very important points:
any hypothesis must be based on the evidence before the investigator
remember the human factor and that investigators are just as susceptible to decision
making errors such as "false hypothesis" or "confirmation bias"
The need to keep an open mind while exploring any hypothesis is essential. There is always
a temptation to favour facts that lean towards the way the investigator is thinking. For this
reason the ability to discuss and debate any hypothesis with peers within an administration
can aid in preventing an investigator embracing a false hypothesis.
'The techniques in this section will help in following objective procedures.
10.4 Causal Factor Analysis and events and condition charting
At this stage it will be useful to review the content and discussion of section 7 of this course.
Professor James Reason in a paper to the 22nd Annual International Society of Air Safety
Investigators in 1991 wrote:'
"Many decades of air accident investigations have created well stocked data bases
that, in history at least, should establish the relative significance of such causal factors
as pilot error, mechanical failure, weather, inadequate maintenance and other
personnel failures. But a glance at some recent statistical analyses shows that this is
far from being the case.
A report to the Flight Safety Foundation in 1986 claimed that mechanical failure preceded by
faulty maintenance was the principle cause of air accidents (Forman, 1990). In 1987 the
Chairman of the US National Transportation Safety Board (NTSB) told the press that bad
weather near airports caused 64 per cent of major crashes in the preceding five years. The
Lufthansa World Accident Survey (1989) found that cockpit crew errors were the prime
contributor, accounting for 76 per cent of all causal factors.
Whom should we believe? In my view, we should believe none of them. All of these figures
misrepresent the causal reality because they under-emphasise the stochastic, organisational
and combinatorial nature of aircraft accidents. Failures in maintenance, air traffic control and
aircrew performances are not uncommon, but only very rarely do they cause accidents. Such
factors are necessary but insufficient causes. The sufficiency is supplied by a malign chance
(sods Law) that combine with these elements in a moment of system vulnerability.
Neither the errors nor mechanical failures nor the weather conditions are individually
important, rather it is their chance conjunction with other causal factors to breach, circumvent
or remove the aircraft's defences that is of primary significance. Such complex interactions
are not always emphasised . . ."


Reason proposes that on analysis any accident can be reduced to eleven 'general failure
modes'l :, these may be paraphrased as:
Deficient or inadequate plant and equipment
Poor designldesign failure
Flawed maintenance policy and planning
Poor procedures
Error-enforcing conditions
lnadequate planning and management of resources
Incompatible goals
Poor communications
Deficient organization
Deficient training
lnadequate or ineffective defences


Also defences are designed to serve one or more of the following seven functions?
To create understanding and awareness of hazards
To give guidance on how to operate safely
a To provide alarms and warnhgs in the event of imminent danger
a To restore the system to a safe state in the event of a threat
To hterpose safety barriers between the hazards and the operation
To contain and eliminate hazards
a To provide a means of escape and rescue should all else fail.
A means of analysing the effectivenessor othervyise of defences is shown in the matrix below.
feature .






Reason, J., Managing the Risks of OrgankationalAccidents,pg 136, Ashgate, 1997

ibid., pg 7.



10.5 The SHEL MODEL (Edwards 1972 as modified by Hawkins 1975)

Any operational system is made up a number of major elements, one of which is the human
element, and all of which interact in such a way that their total effect is larger than the sum
of their parts.

The SHEL model helps to aid the understanding of hl~manfactors. The name is taken from
the initial letters of its component parts:
e Software
documentation, procedures, symbols, etc
e Hardware
machinery and equipment, etc.
e Environment
both internal and external to the workplace
e Liveware
the human element

In this model the match or mismatch of

the interface between the blocks is
important. A mismatch can be the source
of incompatibility between the liveware
and another part of the system.
The SHEL model as modified by Hawkins

Liveware. In the centre, the hub, of the model is a person, the most critical as well as the most
flexible component in the system. However, people are subject to considerable variation in
performance as between individuals, or on an individual basis depending upon the differing
stress factors to which an individual may be subjected. It is important to try and match the
other components in the system to the Liveware in order to avoid stress which will eventually
lead to a failure.
In order to achieve such a match the following characteristics are important:
Physical size and shape
The design of the working environment and equipment and general ergonomic
principles. These principles may have to be adopted to factors such as gender, age
and ethnic characteristics.
Physical wellbeing
People need food, water, air and sleep.
Input charactenBtics
Humans have a sensory system for collecting information by sight, hearing, touch and
smell, all of which, either singularly or in combination help them to respond to external
events and carry out tasks. These senses are liable to degradation depending on the
physical and psychological performance over time.
Humans have severe limitations in short and long term memory and in some mental
processing activities. In poorly designed instrumentation these have led to ambiguity
and inappropriate action and/or deductions.
The appropriate physical response, once the senses system have initiated a response
and the brain has processed the perceived information, relies on all of the above and
the environment.



Temperature, humidity, noise, time of day, light, darkness, atmospheric pressure,
aroma all reflect on performance. A boring, stressful working environment can be
expected to degrade human performance.

Liveware/ffardware This interface is the one most often considered when considering the
human-machine (equipment) system. Because humans are adaptable and can make
allowances for less than optimal design, deficiencies in Hardware may not be identified until
after a disaster. This does not make the deficiencies any less real.
Liveware/SofhvareThis interface is between the human and system procedures (manual and
check list formats, symbology and computer programs). In an accident mismatches here are
often difficult to identify, but often stem from contradictory "software" or misunderstood
Liveware/Enwionment Mismatches between the human and the environment can easily be
identified &:sea. Ships operate a 24 hour society all over the world. The ship's movement,
particularly in bad weather, vibration, engine noise taken with temperature, humidity, the need
for the 0.o.w to operate in darkness are examples of such environmental factors.
Liveware/Liveware This is the interface between people. Between individuals on a ship,
between the bridge team and pilot, between the ship's staff and shore management, between
port officials and the master. It is a highly complex, but irr~portantelement in all human
10.6 Causal Factor Analysis and events and condition charting
It is always important to keep in mind that accidents occur as a chain of factors containing
unsafe events or acts, which combine to cause an accident. The chain often continues
beyond the casualty itself into a period of degradation or recovery. It is also important to
remember that elements of the casualty may have been lying dormant within the system for
many years, until the "malign chance" of circumstances come together.

When investigating a casualty or incident there are a number of techniques that may be used
to chart the sequence of events to help arrive at a logical sequence of events. "Fault tree
analysis" is-onesuch technique, but it does require a special understanding of the symbology
used and training in the technique. Simpler, less time consuming and highly practical for the
investigation team is a simple chronological charting of "events and conditions".
T.he following text has been taken (or adapted) from the United States Department of Energy
publication 'Accident4ncident InvestigationManual"Second Edition (DOEISSDC 76-45127)
Accidents are investigated to identify the causes of their occurrence and to determine the
actions that must be taken to prevent recurrence. It is essential that the accident investigators
probe deeply into events and the conditions that create accident situations, and the
managerial control systems that let these events and conditions so that the root causes can
be identified. . . .
Ludwig Benner suggests two principles which are helpful in defining and understanding these
sequences of events, conditions, . . .

Accidents are the result of a set of successive events that produce unintentional harm.
The accident sequence occurs during the conduct of some work activity.


The key points are that an accident involves a sequence of events that occurs in the course
of a normal operation but that culminates in unintentional injury or damage. Implicit here, too,
is the existence of contributing causative factors, such as existing conditions, failed defences
etc., as well as identifying the beginning and end of the accident sequence.
Staff of the National Transportation Safety Board pioneered-theuse of sequence diagrams or
charts as analytical tools. This lead to the Causal Factors (CF) diagram, which depicts in
logical sequence the necessary and sufficient events and conditions for accident occurrence.
Accidents are rarely simple and almost never result from a single cause. Rather, they are
usually multifactorial and develop from clearly designed defined sequences of events which
involve performance errors, changes, oversights and omissions. The accident investigator (or
investigating board or committee) needs to identify and document not only the events
themselves, but also the relevant conditions affecting each event in the accident sequence.

Construction of the causal factor chart should begin as soon as the accident investigator
begins to gather pertinent information. The events and causal factors will usually not be
discovered in the sequence in which they occurred, so the initial causal factors chart will only
be a skeleton of the final product and will need to be upgraded as additional facts are
gathered. It should be started as early as possible as it helps to:
Organize the accident data.
2.. Guide the investigation.
Validate and confirm the true accident sequence.
Identify and validate factual findings, probable causes and contributing factors.
Simplify the organization of the investigation report.
Illustrate the accident sequence in the investigation report.
The following guidelines are suggested for construction of the chart.
Events should be arranged chronologically from left to right.
Events shown in rectangles and conditions in ovals.
Events are connected by solid arrows.
Conditions are connected to each other and events by dashed arrows.
Each event and condition should be enclosed in solid line (rectangles and ovals) if the
evidence is factual or by dotted lines if presumptive.
The primary sequence of events should be depicted in a straight horizontal line (or
lines in confluent or branching primary chains).
Secondary event sequences, contributing factors and systemic factors should be
depicted on horizontal lines at different levels above or below the primary sequence.
In reconstructing individual activities it is often l-lseful to construct a line for each
individual or unit (say in the case of collision). They can be integrated later.
Events should track in logical progression from the beginning to the end of the
accident sequence.
The following are the criteria for Event and Condition descriptions:
an occurrence or happening (tank lid ruptured) and not a
Each event s h o ~ ~describe
condition (tank lid had leak in it).
Each event description should be a short sentence (ship altered course to port).
Events should be precisely described (mate turned valve clockwise) not (mate shut
Each event should be a single discrete occurrence (ship grounded).
Each event should be quantified where possible (tank lost 3,000 tonnes of oil).
Each event block should contain date and time if possible.
Each event should follow sequentially from the event and conditions preceding it.

Hypothetical case of dragging anchor


10.7 Six tests of safe operation

Although it is not the purpose of an investigation to attribute liability or blame, judgement of
individual and collective acts of commission and omission play an important part of any
accident analysis. lnvestigators are blessed with perfect hindsight, a hindsight, which must
be used to make an accurate and professional analysis of the casualty to prevent such an
accident occurring again.

It is equally important for the credibility of the investigation that the analysis (and the
hindsight) should be realistic. It is useful to remember the "substitution test" (see page 14)
and it is also useful to apply the following six tests of safe operation:
Was the casualty foreseen or foreseeable?
Was the equipment in use fit for purpose?
Were the systems and procedures effective to maintain safe operation?
Were the staff members fit, competent and effective?
Were emergency procedures and defences effective?
Was there a management system to monitor and improve performance?

10.8 Analysis of physical evidence

Analysis of documents may involve two major aspects, cross checking documents from
different sources that contain the same information or scientific analysis.
Analysis could include cross checking the bridge movement or "bell" book with the engine
room records. It cannot be emphasised enough that contemporaneous records, those made
at the time, are of value, fair copies of log books, e.g. the scrap log copied out in a fair hand
are of limited value. Of greater value is the cross checking of ship's records with external
soilrces such as VTS tapes, harbour control tapes or I Qbooks,
cargo terminal records,
police records, customs records, or even Nor radio recordings.
Investigators must keep an open mind and think latterly asking "who else may have similar
International Ship Management Code
The ISM Code documentation should be inspected as a matter of routine. It is important to
ensure that the procedures in the code are adhered to.
The ship operator's "Documentation of Compliance" is valid for 5 years, subject to annual
verification. The ship's "Safety Management Certificate" is valid for 5 years subject to
periodical verification by the administration.
All aspects of the code are important to an investigator and include but are not confined to
the following.
a Training (ISM Code 6.3)
a Passage planning and procedures with pilot embarked (ISM Code 6.4)
a Information and language of ISM Code (ISM Code 6.6)
a Plans, instructions, check lists for the safety of the ship and pollution prevention.
(ISM Code 7.0)
a Emergency preparedness (ISM Code 8.0)
a Reporting non-conforming incidents (ISM Code 9.1)
a Corrective action (ISM Code 9.2)
a Maintenance (ISM Code 10.1)
a Critical equipment (ISM Code 10.3)
a Documentation (ISM Code 11. l )
a Record of internal audits (ISM Code 12.3)


It is important to remember that safety investigators should review the ISM Documentation
with a view to safety, not as an exercise to audit the scheme and to focus on issues not
related to the casualty. However, an investigator must bare in mind that any poorly maintained
system may be symptomatic of a wider malaise.

If there was a departure from the code it is important to identify the non-conformity to
establish whether the departure was consistent with reasonable decision making (see Course
1.3.4). Depending upon the incident it may also be necessary to check the ship's reporting of
"non-conforming incidentsn(lSMCode 9.1) and the management receipt of such records and
subsequent action, which may include a record of corrective action (ISM Code 9.3).
Scientific/forensic examination of documents
Scientific analysis of evidence in general will be dealt with in 4.5 and the scientific analysis (if
necessary) of documents will follow the same principles.

There are some tips for investigators, such as counting the pages of a log book, checking
sequential page numbers, holding documents to the light to see if there has been any other
writing or erasures. Such things may be necessary if you suspect that somebody is not telling
the truth.
Access to original documents may be difficult on both legal and practical grounds.
Administrations should ensure that the investigation procedures are covered by the
necessary legislation and that the legislation gives the investigator access to documents and
On a practical level, when a ship is continuing in operation it may not be possible to keep
original documents, therefore an investigator must complete hislher examination on the spot
and take copies, facsimile copies where possible.
Photographs are important in analysis. It is useful to scan photographs into a suitable
computer program where they can be enhanced and small areas viewed in more details.
However, it is important that any changes made to the photograph is logged as it is possible
to doctor photographs and it is important that the credibility of evidence is not challenged.
Scientific ana/ysis.
It is important to provide the scientist with any material in a state that is as near as possible
the same as when taken from its source (see module 2). Proper collecting bags, tins etc are

Samples of materials such as paint, oil, water, etc., may be subject to a battery of tests for
forensic comparison to establish whether they came from a common source:
Microscopic comparison (in paint for layer sequences).
Microspectrophotometer, measuring the colour spectra of the samples.
Fourier Transform lnfra Red spectroscopy. Differences in absorption of infra red light
gives off different wave lengths for different chemical bonds which in paint sarr~ples
identifies the polymer holding the paint together.
fluorescence spectroposcopy. X-ray directed at an object/sample, is absorbed
at different wavelengths by different trace elements, providing a
"signature" for the elements present.
Wires, chain or metallic samples will need to be sent to a laboratory with the appropriate skills
and, where available, some national accreditation should be sought. Universities may also be
able to provide scientific analysis.


When testing cargo samples, the actual specification s h o ~ ~ lbe

d compared to the
documentation (if any) carried on board, IMO Code requirements and industry standards.

Course recorders
Many ships carry course recorders. There is no compulsion to do so and even when fitted
there is no requirement for a ship to have a course recorder operating.
When available the course recorder provides objective data of the ship's heading at any given
time. It is not a record of the track made good. In the absence of any black box, or GPS record
or radar recording from the ship or shore, it is the most useful objective record available of
the ship's head. Careful scrutiny and analysis of the course recorder can help pin point the
time of collision or grounding as well as the ship's heading leading up to the accident.
Course recorders should be checked for any time difference between the investigation's base
time. Many recorders are hard to synchronise accurately and alignment with the ship's head
is sometimes problematic. These should be checked as accurately as possible before the
trace is removed.
There are a variety of models of course recorders: Some show the course over quadrants,
with a separate quadrant pen. Some also include'a helm indicator, which can be invaluable.
Other types show a 360" range across the trace paper.
Another problem is the "point" of the trace pen. This can often be very thick.
There are several techniques or combinations of techniques which are of practical use.
An enlarged photocopy is a useful tool, though it should be realised that photocopying will
often distort the page slightly, although more modern copiers may distort less. The photocopy
technique may not be suitable for a formal court process, but for a safety investigation it is a
useful way to achieve a suitable scale for measuring and analysis.
Scanning the trace on a computer scanner is a very successful way of analysing course
recorder traces. The computer can be used to zoom in and enlarge areas of the trace. Also,
using Adobe Illustrator or Corel Draw or similar graphics packages, the investigator can
overlay hislher own grid scale.
A ruler and high power magnifyirlg glass remain reliable.
If using a course recorder trace the investigator must obtain the best quality trace possible.
Quality of trace is an important factor in lifting off information and a trace can be one of the
most important sources of information.

In principle, analysis of the heading record is simple, but in practice it is difficult and requires
the use of manual methods over a small area of interest.


The following has been extracted from a 1982 paper by Ewan C 6 Corlett MA, PhD, F Eng,
Hon Vice President Royal Institution of Naval Architects:
By selecting the area of hterest changes h heaohg caused by an external force may
be detectable by careful analysis. By selecting a fair& narrow time base, say three
mhutes and takhg heaohgs agahst t h e andplotting them on graph paper a number
of ana&ses may be possible,
These include:
angular veloci?~
angular accelerations
rate of change of headhg.
Analyses of a trace can show the exact (recorder the) of the moment of hpact.
If two sh@sare h coll&ion, depending upon the relative s~zesof the vessel, both sh@s
will expenence a change h thek angular velociv by reason of the contact. One may
expenence an acceleration force and the other an arresting force. If a sh~pcontacts a
sold object it wIil expenence an arresting force. These can be quite rapidphenomena
producing a sharp change in the curve.
If the contact is through hydrodynamic effect, say with the side of a channel, the effect
may be less rapid producing a smoother curve.
Assuming an accuracy h t h e to the nearest second from the course recorder trace,
and to a tenth of a degree in heading, it is possfile for purposes of Iilustrating a
method for ana&sis, to draw a graph of sh~p'sheadhg li7 degrees agahst time in
seconds'for a penod from 0 seconds to 180 seconds, as in Figure 4a, the heaahg
record plot.
Examham of the figure reveals that from 0 secs to /i/st over 40 secs, there was no
change of headlig and the course was steady on 3 degrees, or 003 degrees.
Thereafter the course changed to 007 degrees, with a relatively slow change of
headlig between 42 seconds till around 74 seconds. At that point, the headhg
changed rapid& by about 2 degrees li7 just over a second after which there was a
reduction in the rate of change of headlig tlil the course was OOZ There was no
change of heading for the shoiiperlbd that the sh@was steady on OOZ
Then, the heading changed back to 003 the change of headlig takiflgplace slow& at
fist, reaching a mmimum rate between 100 seconds to around 102seconds, reducing
thereafter to zero from about 125 seconds onward
Using the heading record plot and the slope at various points of the curve, we can obtain and
plot angular velocities against time. For the first 42 seconds, the angular velocity is zero,
increasing gradually till at around'75 seconds the angular velocity increases to about 1.2
degrees per second, corresponding to the near vertical trace on the heading plot at the same
time. The angular velocity then reduces gradually in line with reducing rates of change of
heading becoming zero just after 85 seconds. After this point the angular velocity increases
as the ship's head is altered to 003, with a maximum angular velocity being obtained soon
after 100 seconds. Once the ship's head is steady on 003 or thereabouts the angular velocity
reduces to zero once more.


4a: 8 vs T

5 LI:


4 -

3 2


30 40

10 20




80 90 1 0 0 1 1 0 1 2 0 1 3 0 1 4 0 1 5 0 1 6 0 1 . 7 0 1 8 0





4c: 8 vs T



Similarly, we can plot a curve of angular acceleration against time, so we now have three
curves, the heading record plot, and its first and second derivatives, angular velocities and
accelerations. The angular acceleration curves reveal high acceleration and deceleration,
indicating the possibility of a collision, and subsequent water cushioning effect. (Part of a
paper presented at the IMOIChina Seminar, 1982)

Other records
There is no standardisation of ship recording equipment.
Investigators should always establish what is available on any specific ship.
Data loggers are becoming more common.


Voyage Data Recorders (VDR) the equivalent to the aircraft "black box" are carried on certain
ships, "reading" the information on the VDR may require cooperation with other flag States.
Some pieces of ship's equipment allow information to be recovered. Some fire alarms have
a memory and record the times of alarms and the location of the sensor.
In ECDlS Units it is a design requirement that tracks should be able to be recovered and
replayed. Other than ECDlS there are not any standard requirements but investigators should
be alert to the possibility of being able to down-load GPS data, or recovering radar displays
information on a video drive facility from certain sets.
10.9 Analysis of interviews
This is a process of assessing those interviews or parts of the interview process where the
same or similar information was provided and those areas where contradictory evidence was
-. - -given.

It is importaiil that all information is checked against the real and documentary evidence. It
can be the case that consistent, or similar accounts can be wrong and one single
contradictory account correct. where ever possible all accounts should be checked against
all other evidence and corroborating real or independent evidence normally preferred over
evidence from the same interest group.
10.10 Analysing human involvement
Comparing the events and condition chart(s) with the witness interviews should help in
analysing the human involvement. 'The human involvement will often (more often than not)
extend beyond the ship. Remember the Reason and SHEL models.
10.11 Identifying potential safety problems and safety actions
Careful analysis, using event and condition charting can identify the relevant safety problems.
The developing of safety actions fall into three basic categories:
Immediate (local) action.
Flaglcoastal State (unilateral) action.
International (IMO) action.
It is importa,nt to ensure that any recommended safety action is practical and will address any
safety deficiency without causing others (see notes under 4.3).
Safety action may involve a 1oca.l ship board or port solution. Remember here the ISM Code
procedures and any IS0 9002 procedures.
Safety actions may require an amendment to national procedures or requirements or may be
such that an international response may be required through IMO.
10.1 2 Revisiting witnesses
Although it may not always be possible, investigators should never rule out revisiting or
otherwise contacting witnesses to try and resolve ambiguities. Even the most experienced
investigator will find it very difficult to cover all aspects of all factors in a single interview.
Remember, if procedures do not rely on sworn testimony, most people can be reached by fax,
phone, letter or e-mail.

11.1 Natural justice - the circulation of the draft report
Natural justice, being fair to all parties is an essential element in establishing a system that
has the acceptance of the industry. Investigation procedures should allow for ,the draft of a
report, or relevant part of the report, to be circulated to any individual or organisation that may
be affected by the report's conclusions.
Any such person or organisation should be given a reasonable period of time to provide other
evidence or information or make a submission relating to the report. The investigator must
then take such submissions into account and, if necessary, amend the report or in some way
fairly reflect the views of any person making a submission.
11.2 IMO reporting requirements
Do not forget the requirements of MSCICirc. 827 (MEPCICirc. 333) "Harmonised Reporting
procedures - Reports required under SOLAS regulations 1/21 and MARPOL 73/78 articles 8
and 12".

11.3 Report format

Reports are compiled from the evidence gained, which allows us to tell the story, analyse the
factors and draw sensible conclusions based on the human factor models.
The Code at section 14 provides guidance on the content of the report.
Reports should include, wherever possible:
a summary outlining the basic facts of the casualty and stating whether any deaths,
injuries or pollution occurred as a result;

the identity of the flag State, owners, managers, company and classification society;

details of the dimensions and engines of any ship involved, together with a description
of the crew, work routine and other relevant matters, such as time served on the ship;

a narrative detailing the circumstances of the casualty;

analysis and comment which should enable the report to reach logical conclusions, or
findings, establishing all the factors that contributed to the casualty;
a section, or sections, analysing and commenting on the causal elements, including
both mechanical and human factors, meeting the requirements of the IMO casualty data
base; and

where appropriate, recommendations with a view to preventing similar casualties.

11.4 Narrative
The narrative should describe the circumstances and events leading to the investigation in a
neutral way, without niaking judgements. A simple factual story.
11.5 Analysis
The relevant causal factors should be reviewed and analysed. Any causal factor analysed
should be identifiable in the narrative section. The chart of events and conditions is a useful
guide. Nor should the six questions of safety operation be ignored. It helps to keep the


analysis focused and in the realms of practicality. There is nothing wrong in the investigator
providing expert opinion, providing that:
it is clear that helshe is giving an opinion;
it is very carefully thought through; and wherever possible
it is canvassed with other suitably qualified persons (there is nothing worse that an
unrealistic or unsubstantiated opinion).

11.6 Conclusions
The concl~~sions
must be drawn from the analysis section. They are often the part of a report
read initially by interested parties and other professionals. Care must be exercised to ensure
they are accurate.
Remember, after sending out the draft report your conclusions will be difficult to amend in
terms of being more severe.

11.7 Recommendations
Recommendations must follow logically from the text of the report. Again wherever possible
the recommendations should be tested with suitably qualified colleagues and all aspects of
their impact thought through. Consider the human factors involved. A common response is to
legislate. Legislation to regulate human behaviour, in itself, is seldom the answer if the
situational factors are ignored.

11.8 The completed report and distribution

The Code at section 12 deals with the issuing of casualty and incident reports and
submissions to IMO and makes provision for a dissenting report to be issued where a
"substantially interested State" disagrees with the whole or part of report.
How widely the report is distributed is a matter for your administration. However, unless
people read about accidents and are presented with accurate facts and expert analysis
lessons will not be learnt and the nature of accidents will not be understood. .
. -'

11.9 Media
Section 5.15 offered advice on contact with the media. When releasing a report public
attention can be drawn to the publication by media release. Depending upon the
trustworthiness of the individual media outlets and the importance of the incident it may be
possible to provide a report under embargo so that a journalist can read the report and
provide a full and comprehensive report.

But never forget, the creed of the majority of journalists is to get a story that will sell
newspapers, truth and accuracy are often, but not always, the first casualties.

11 .I 0 Reopening an inquiry
In addition to introducing a more uniform approach to marine casualty and incident
investigations and promoting cooperation between substantially interested States, the stated
aim of the Code is to prevent similar casualties in the future. This can only be achieved by
undertaking high quality investigations and reporting based on the best evidence available. If
after completing a casualty or incident investigation (including reporting) significant new
evidence is provided to the State(s) involved, provision should be made to reopen the inquiry.
There are a number of precedents for this, perhaps the most famous include the Titanicand
the Derbyshke.


f 656.6
Mar i n e