International Journal

on Marine Navigation

and Safety of Sea Transportation

Volume 3

Number 4

December 2009

443

1 INTRODUCTION

During last three decades attention of the maritime

world has been focused on safety of shipping.

Amongst other causes of accidents at sea casualties

related to manoeuvrability happen quite often and

analysis of casualties shows that CRG casualties

(Collisions-Ramming-Groundings) constitute about

53% of all serious accidents leading to ship loss

(Payer 1994). Data on CRG casualties for the year

1982 analysed on the basis of sources provided by

LRS and DnV revealed that their frequency was ra-

ther high as it is seen from the Table 1.

Table 1. Data on CRG casualties

___________________________________________________

Source Mean number of Number of Frequency of

ships during the year CRG casualties casualties [%]

___________________________________________________

DnV 2816 120 4.3

LRS 3391 170 5.0

___________________________________________________

The data showed that 1 ship in 22 took part in

CRG casualty this year (Samuelides 1984). CRG

casualties occur more often with increasing speed

and size of vessels and such casualties may cause

more serious consequences. Collisions may also

happen more often in restricted waterways and ca-

nals and in particular in areas where additional ex-

ternal factors, as e.g. current, make handling of ships

more difficult.

Statistics of CRG casualties in the following

years showed considerable decrease in percentage,

however it revealed quite alarming increase of the

number of accident during last few years. As it is

seen from Fig.1 the number of such casualties has

increased almost twice from the year 2000. The rea-

son of this effect is not clear – it may be, however,

to be attributed to increase of the size of ships, in

particular container ships operated, to the increase of

the density of traffic, but most probably to the lower

level of performance of crew members which were

recruited from many different countries.

This situation requires serious attention and pre-

vention of CRG casualties must be treated as a prior-

ity.

2 SAFETY SYSTEM OF PREVENTION OF CRG

CASUALTIES

In order to achieve safe operation of ships and pre-

venting casualties holistic and system approach is

necessary. System approach consists of looking at

the problem as assembled of the number of sub-

Risk Analysis and Human Factor in Prevention

of CRG Casualties

L. Kobylinski

Foundation for Safety of Navigation and Environment Protection, Iława, Poland

ABSTRACT: CRG casualties create one of the major type casualties in shipping. Prevention of CRG casual-

ties is an important issue, especially because of the number of CRG casualties has increased almost twice dur-

ing recent years. For the great majority of all CRG casualties human factor responsible, and the increasing

number of these casualties might be attributed to poorer qualifications of ship masters who have not enough

experience in handling very large ships put into operation presently. Risk analysis is a modern method for as-

sessment of safety level of technical systems. This tool may be the used to investigate causes of casualties and

to find out most effective prevention measures. Risk analysis is widely used in many areas; in case of marine

technology it is used routinely in off-shore technology.

The author investigates possibilities to apply risk analysis in the area of ship handling with the focus on hu-

man factor. This is preliminary study where possible methodology for hazards identification and risk assess-

ment in respect of CRG casualties are investigated and risk control options are suggested. Various aspects of

the influence of human factor in collision avoidance are listed and in particular the effect of training is

stressed.

444

problems mutually interrelated. In this approach the

process of achieving main goal is exactly defined

and related to sub-problems in accordance to the

adopted plan.

Figure 1.Percentage of CRG casualties during years 1988-2006

The system of safety against CRG casualties is

rather complex, because of numerous interrelations

between various sub-systems and because of that, its

analysis is difficult. It would be, however, necessary

to take into account in this system at least the fol-

lowing sub-systems:

− Traffic pattern

− Environment

− Human factor

− Equipment

− Legislation

3 GOAL ORIENTED APPROACH

The weak point of the present legislative status of

safety requirement in general was duly noticed by

highest IMO (International Maritime Organization)

bodies and quite recently the Marine Safety Com-

mittee (MSC) recommended adoption of the concept

of goal-based approach to safety requirements. The

goal-based approach does not include prescriptive

regulations or standards that must be complied with,

but sets goals that allow alternative ways of achiev-

ing safety (Hoppe 2006). Goal-based standards are

for some time considered at IMO and appraised by

some authors (Vassalos 2002), and they were intro-

duced in some areas, albeit not in the systematic

manner. Marine Safety Committee commenced in

2004 (IMO 2004) its work on goal-based standards

in relation to ship construction adopting five-tier

system (Table 2).

IMO MSC committee agreed in principle on the

following Tier I goals to be met in order to build and

operate safe and environmentally friendly ships:

“Ships are to be designed and constructed for a spe-

cific design life to be safe and environmentally

friendly, when properly operated and maintained

under specified operating and environmental condi-

tions, in intact and specified damage conditions,

throughout their life” (IMO 2004).

In the opinion of the author goal oriented holistic

approach appears to be the best solution in prevent-

ing the increase of the number of CRG casualties.

Goal oriented approach involves apart of prescrip-

tive requirements, also risk analysis and system ap-

proach. Therefore to investigate the possibilities of

application of risk analysis to safety against CRG

casualties and to investigate possible risk control op-

tions and associated requirements is an important is-

sue.

Table 2. Five-tier system for goal-based requirements

___________________________________________________

Tier I: Goals

Tier II: Functional requirements

Tier III: Verification criteria of compliance

Tier IV Technical procedures and guidelines, classification

rules and industry standards

Tier V Codes of practice and safety and quality systems for

shipbuilding, ship operation, maintenance, training etc

___________________________________________________

4 PRESCRIPTIVE VERSUS RISK-BASED

APPROACH

The basic dichotomy in the conception of safety re-

quirements consists of prescriptive approach and

risk-based approach (Kobyliński 2007)

Traditional regulations were of prescriptive na-

ture and they are formulated in the way where a cer-

tain standards related to ship construction or opera-

tion must be complied with. Prescriptive regulations

could be developed on the basis of experience (ex-

perts opinions) statistics, analytical methods, com-

puter simulation, model tests and full-scale trials.

Deterministic or probabilistic calculations may be

employed when developing the criteria, although, as

a rule, deterministic approach is used in most cases.

Prescriptive regulations have many advantages.

They are formulated in a simple language, which is

easily understood by everybody, they are easy in ap-

plication, they also make checking adherence to the

requirements easy. The main shortcoming of pre-

scriptive regulations is that they are bounding de-

signers or operators and they do not allow introduc-

tion of alternative solutions. They are based on

experience gained with existing objects and they are

not suitable for novel types of ships or uncommon

operational and emergency situations. Usually they

were amended after serious casualties happened. The

risk involved with the application of prescriptive

regulations is not known.

At the opposite of the prescriptive regulations,

there is risk-based requirements. The risk-based re-

quirements are based on risk analysis where and the

445

main object is to assess eventually accept the risk.

The advantages of risk-based requirements is that

they are not binding designers or operators request-

ing to satisfying or obeying adopted fixed rules and

standards, but offering the possibility of applying a

variety of solutions provided they ultimately allow

to keep risk within acceptable limits. Human factor

could be taken into account, which is extremely im-

portant because the majority of CRG casualties may

be attributed to human fault.

Risk is defined as a product of hazard probability

and hazard severity (consequences):

R = PxC

To facilitate the ranking and validation of ranking

IMO recommended defining consequence and prob-

ability indices on a logarithmic scale (IMO 2002).

The risk index may therefore be established by add-

ing the probability (frequency) and consequence in-

dices. We have then:

Log(risk) = Log (frequency) + Log(consequence)

Risk-based approach according to IMO recom-

mendation is formalized (FSA methodology) and in-

cludes the following steps (IMO 2002):

1 Identification of hazards

2 Risk assessment

3 Risk control options

4 Cost-benefit assessment, and

5 Recommendations for decision making

FSA methodology was recommended by IMO for

general evaluation of safety requirements; in particu-

lar cases strict adherence to this methodology may

not be possible. However, in all cases risk analysis

must lead to risk assessment and acceptance. For this

purpose, and taking into account specifics of ship

operation at sea, risk assessment matrix (Table 3)

may help to evaluate risk and to take appropriate ac-

tion. In this matrix hazard probabilities are divided

in five groups, as below:

− A. Frequent – always occurring, once or more

yearly (greater than10

-3

-10

-4

)

− B. Probable – few times during ship’s lifetime

(10

-4

-10

-5

)

− C. Occasional – once during the lifetime of the

ship, few times in the lifetime of the fleet (10

-5

-7

)

− D. Remote – little probable, but possible during

the lifetime of the ship, once during the lifetime

of the fleet(less than 10

-7

)

− E. Extremely improbable – such a small probabil-

ity that it may not be taken into account (10

-9

–

-10

)

and hazard severities (consequences) into four

groups (Halebsky):

1 Catastrophic – loss of vessel, fatalities

2 Critical hazardous effect - dangerous degradation

in handling, need outside rescue operation

3 Marginal major effect – significant degradation in

handling but not preventing to complete safely

journey

4 Negligible minor effect – slight degradation in

handling, need for slight modification of operat-

ing procedures

Table 3. Risk assessment matrix

Hazard probability (hourly)

←Low High→

E. D. C. B. A.

I Catastrophic

Z Y X X X

II Critical

hazardous ef-

fect

Z Z Y X X

III Marginal

major effect

Z Z Z Y Y

IV Negligible

minor effect

Z Z Z Z Z

In the table 3: Z- action to reduce hazard if eco-

nomically feasible. Y-action to reduce hazard proba-

bility, X action to eliminate hazard

5 RISK ANALYSIS AND SAFETY AGAINST

CRG CASUALTIES

At present there are numerous requirements included

into various legislative instruments that were, how-

ever, developed at different times by different bod-

ies, some of them being compulsory, some others

have only status of recommendations and in general,

they are not consistent in many points. Most of them

were developed by the International Maritime Or-

ganization, but in spite of that, holistic system ap-

proach was not used in their development. The list

of different legislative instruments where require-

ments applicable to safety against CRG casualties

are included is shown below:

− IMO manoeuvring standards,

− SOLAS convention requirements related to steer-

ing gear, and machinery

− COLREG convention requirements

− Pilotage requirements

− Separate traffic routes

− STCW Convention (Personnel qualifications)

446

− SOLAS Equipment Chapter Port authorities re-

quirements.

The above list is not exhaustive and is provided

as an example only. Requirements included in all of

the above instruments are of prescriptive character.

Because of the complicity of the system of legis-

lative instruments and requirements included there-

in, direct application of risk analysis to the system as

a whole at this stage seems to be extremely difficult

and requiring thorough study that is beyond the

scope of this paper. Risk analysis might be, howev-

er, applied for example to the requirements related to

the following subsystems:

− Ship design – (manoeuvring characteristics)

− Harbour and traffic lanes design

− Effect of human factor.

− Navigational aids

−

Performing safe manoeuvers

The above subsystems are strongly interconnect-

ed, but in order to bring practicable solution they

may be separated at the first step.

6 APPLICATION OF RISK ANALYSIS TO

PERFORMING SAFE MANOEUVRES

The first step of the risk analysis is identification of

hazards and assessment of their probabilities. Analy-

sis of CRG casualties reveals that the causes of cas-

ualty may be attributed to:

− functional aspects resulting from reliability char-

acteristics of the technical system, therefore

manoeuvring characteristics of the ship,

− operational aspects resulting from the way the

ship is operated in traffic routes, from harbour

lay-outs and facilities, cargo handling etc,

− human factor, i.e. aspects resulting from action

of the personnel handling the system, therefore

crew members but also ship management, marine

administration and owners company organization

− external causes resulting from factors independ-

ent from designers builders and operators of the

technical system therefore from ship environment

and climatology

− decision support systems helping the master or pi-

lot to take appropriate decisions, inter allia radar.

ARPA, electronic maps, computer programs for

manoeuvres prediction, etc.

IMO resolution included general guidance on the

methodology of hazard identification. With respect

to manoeuvrability, hazard identification could be

achieved using standard methods involving evalua-

tion of available data in the context of functions and

systems relevant to the type of ship and mode of its

operation.

− Hazard identification is carried-out using hazard

identification and ranking procedure (HAZID).

According to general recommendation the meth-

od of hazard identification comprised mixture of

creative and analytical techniques. Creative ele-

ment was necessary in order to ascertain that the

process is proactive and is not limited to hazards

that happened in the past. Analytical techniques

are used in order to evaluate, separately or in

combination:

− statistical data concerning causes of accidents

− historical data including detailed description of

accidents

− conclusions resulting from model tests and com-

puter simulations

− event and fault trees method

− opinions of experts

In particular the last method is much of use, pro-

vided that collation and analysis of expert opinions

is properly organized – for example by using Del-

phic method (IMO 2002a).

US Coast Guard (USGC 1981) provided some

indication on the posible causes of CRG casualties.

This is shown in the table 4.

Table 4.Causes of CRG casualties (according to USGC 1981)

___________________________________________________

Cause Percentage [%]

___________________________________________________

Insufficient Wind & current 9

ship Turning ability 7

controllability Tugs 4

Stopping 4

Bank suction 3

terring failure 2

Control while stopping 2

Control while backing 2

___________________________________________________

Direct human error 33

___________________________________________________

Unavoidable 34

___________________________________________________

Figure 2. First level fault tree for CRG casualties

Navigation

aids

Traffic

lanes

and/or

Manoeuvring

characteristics

Environment

Weather

Other

ships

Human per-

formance

(HOE)re

External

forces

CRG

447

The classification shown in Table 4 is, however,

not particularly useful for the purpose of risk as-

sessment because large percentage of casualtied was

classified as unavoidable. This is certainly wrong,

because there is always some cause behind the casu-

alty and it is probably that human and organisation

errors (HOE) or heavy weather and perhaps other

causes qualified by marine courts as “force majeure”

are hidden in this category.

As an example of application of this methodology

the list of hazards in respect to CRG casualties is

shown in Fig. 2. In this example ranking of hazards

is not shown, moreover the sketch could be consid-

ered as the first level of the fault tree leading to

CRG. Hazards identified as relevant to safety against

CRG are all strongly interconnected, moreover, hu-

man factor understood as performance of an individ-

ual (in most cases the master) plays important part in

each case. Hazards identified should be further de-

composed preferably using fault trees and/or events

trees reproducing various scenarios of CRG casual-

ty. The set and combination of fault trees and event

trees as developed for all hazards identified and all

scenarios (defined as risk contribution trees – RCT)

is a basis for HAZOP (hazard and operability study)

procedure that allows also assessment of frequencies

(probabilities) of hazards required for risk assess-

ment. This is rather tedious task bearing in mind the

multitude of possible scenarios. This problem, how-

ever, is not discussed here.

7 EFFECT OF HUMAN FACTOR

As human and organization errors (HOE) are major

causes of CRG casualties they require a special at-

tention. HOE may be the result of design and con-

struction faults (bad manoeuvring characteristics of

ships) and force majeure, that are responsible for

about 20% of all HOE casualties (Payer,1994), the

rest may be attributed to operational factors that in-

clude the following:

− society and safety culture

− organization

− system

− individual

Society and its culture has important effect on

safety. Economic factors tend to limit safety re-

quirement, because enhancement of safety cost

more; from the other hand lower safety level results

in higher cost of increased number of accidents.

There exists certain optimum from the purely eco-

nomic point of view, but if fatalities are resulting

from accidents the pure economic point of view is

no more valid and crucial point is how high risk may

be acceptable by the society. The risk is much lover

in developed countries in comparison with the coun-

tries that are not yet developed.

The society culture is strongly related with safety

culture. High safety culture helps to avoid a large

percentage of accidents. The enquiry by the RINA

amongst a number of naval architects did show, that

the majority of them recognized safety culture as the

most important factor in safety (The Naval Architect

1999).

Figure 3. Effect of safety culture on accidents rate

Organization. A great number of accidents is

caused by bad management or bad organization. Bad

organization could mean lack of supervision, lack of

procedures, lack of instructions, lack of activity by

marine administration, lack of policy for safety

management or lack of motivation. One important

factor is also culture of shipping company. For ex-

ample the dominant culture of company might be

tendency to achieve gain without considering risk

(flirting with risk) or forcing excessive strain leading

to over-fatigue and in consequence may appear to be

opposite with the aim of the company.

System. The following system faults influence

operator behaviour: complexity, faulty signalization,

small tolerances, difficult operation, inaccessibility,

high demands in operation, wrong alarms, bad visi-

bility, incomplete software, etc.

Individual. Operator’s error is the most common

cause of accident. However it is very difficult to

identify the real reason of the operator action. There

is a long list of possible causes as shown in table 5.

It is really impossible to attach probabilities to all

factors listed in Table 5, because the relevant statis-

tical data do not exist and there is no chance that

such statistics will be ever available. However all the

above factors may be divided in three groups:

1 individual character of the operator- integrity, re-

liability, morale

2 physical predispositions – health, endurance, im-

munity

3 knowledge – education, training, experience

Safety cultu-

Culture of pas-

sive com-

pliance

Culture of

avoidance

448

Limiting to the above three groups it would be

possible to construct the risk contribution tree (fault

tree) for HOE as shown in fig 4.

Table 5. Human error factors (Bea 1994)

___________________________________________________

Fatigue Wishful thinking Bad judgement

Negligence Mischief Carelessness

Ignorance Laziness Physical limitations

Panic Violations Boredom

Greed Drugs Inadequate training

folly Inadequate Inadequate education

communication

Ego Alcoholism Hidden illness

___________________________________________________

For the risk analysis it is necessary to attach

probabilities to every group at the first stage. This

could be done on the basis of statistics or expert

opinions. Currently published statistics is not availa-

ble, although major shipping companies certainly

have such data. If probabilities attached to each of

the above groups are known then conclusions with

regard to risk o may be drawn.

Risk control options constitute an important step

in the risk analysis. If we assume that probabilities

are equally distributed between three groups, then

concentrating on group three for example, one risk

option would be stressing importance of training.

Amongst other effects, it is well known, that training

affects considerably the ability to handle critical sit-

uations (Bea 1984).

8 CONCLUSIONS

Risk analysis is an excellent method for analyzing

safety of complex systems to which system of safety

against CRG casualties at sea also belongs. However

application of risk analysis to CRG casualties poses

serious difficulties because of the complexity of the

system and strong interrelations between different

subsystems.

In particular, human factor, playing predominant

part in a great majority of CRG accidents, requires

special attention in the risk analysis. This is, howev-

er, difficult because of lack of reliable statistical data

on the influence of various individual characteristics

of the man at control on safe performance of ma-

noeuvres. There are intuitive conclusions that train-

ing, for example affects ability of the man at control

considerably, but respective statistical data are not

available.

Nothwithstanding the difficulties, even at this

stage, risk analysis could provide useful results

when applied to various subsystems of safety against

CRG casualties and in particular it may allow to as-

sess the impact of various risk control options. This

may be, in particular, relevant to human and organi-

zation errors (HOE) as shown in the paper.

Figure 4. Simplified fault tree for HOE

REFERENCES

Bea, G.R. (1994). The role of human error in design, construc-

tion and reliability of marine structures. Ship Structure Re-

port SSC-378

Hoppe H. (2006). Goal based standards – a new approach to

the international regulations of ship construction. IMO

News Issue 1

IMO ( 2002). Guidelines for Formal Safety Assessment (FSA)

for use in the IMO rule-making process. Doc. MSC/Circ.

1023;

IMO ( 2002a). Guidance on the use of human element analyz-

ing process (HEAP) and formal safety assessment (FSA) in

the IMO rule making process. Doc. MSC/Circ. 1022;

Kobyliński, L. (2006). Appraisal of risk assessment approach

to stability of ships. 8

International Ship Stability Work-

shop, Istambul

Kobyliński, L. (2007). Goal–based standards. 9

International

Ship Stability Workshop. Hamburg

Payer, H. (1994). Schffssicherheit und menschliche Versagen.

Hansa-Schiffahrt-Schiffbau-Hafen 131 Jahrgang, No.10

Samuelides, E., Frieze, P. (1984). Experimental and numerical

sikulation of ship collisions. Proc 3

Congress on Marine

Technology (IMAEM. Athens

The Naval Architect (1999) March issue.

US Coast Guard (1995) Preventing through people. Quality ac-

tion team report

US Coast Guard (1981). Report M-8-81

Vassalos, D. (2002). Total ship safety – a life-cycle risk-based

DOR for safety. The Stability Research Centre NAME.

Universities of Glasgow and Strathclyde. Report.

Safety Culture

Organization

System

Individual

Group 1

Group 2

Group 3

CRG