Kristiansen Svein. Maritime Transportation: Safety Management and Risk Analysis
Подождите немного. Документ загружается.
first category may lead to slip or lapse whereas the other may result in mistake or
violation.
Rockwell and Griffen (1987) describes an error classification that covers airplane flight
and that may seem of relevance to ship operation (Table 13.11). The errors are related to
specific tasks on the flight-deck rather than using the generic terms of Reason.
Hollnagel (1998) gives a very extensive discussion of human error models or taxonomies.
It might be relevant to comment briefly on some of the approaches he mentions. Hollnagel
concludes, after an extensive study of alternati ve accident models, that error (phenotype)
is caused by man, technology and organization (Figure 13.17). He also pos tulates that
erroneous action has certain generic modes as outlined in Figure 13.18.
13.12.2 A Pract ical Approach
As pointed out earlier in this chapter, the starting point for the accident analysis process
is to establish the critical event sequences or, in other words, the question to answer:
what happened? The next and very challenging task is to explain why it happened or, in
Figure 1 3.1 7. The CR EAM model ( adapted f rom Hol l nage l,1998) .
Ta b l e 13 . 11 . Critical in-fli ght e vent model
1. Inadequate pre-flight checks 2. Fails to recognize early warnings of problems
3. Fails to do sequence checks 4. Decides to fly despite system discrepancies
5. Fails to recognize early warnings 6. Fails to monitor instrument readings
7. Fails to notice small discrepancies
in flight sensations
8. Fails to notice lack of agreement
of related instruments
9. Diagnostic error 10. Error in estimation of urgency
11. Improper corrective action
Source: Rockwell and Griffen (1987).
396 CHAPTER 13 A CCI DENT ANALY SIS
technical terms: what were the main causal or contributing factors? This is in essence
a diagnostic process where no precise or universal approach can be prescribed. The
CASMET project did, in spite of this, attempt to outline certain principles (Koster et al.,
1999). The model was based on the premise that there is an interaction between the
mission, humans, systems and the environment as conceptual ized in Figure 13.19.
An analysis of the Human and Organizational Error (HOE) element in maritime
accidents should start with the recognition that maritime operations typic ally take place
because a certain mission or assignment has to be fulfilled (see Figure 13.19). To fulfil the
assignment, a certain performance is required from two entities: the personnel aboard,
and the tools with which they are equipped, including the ship itself. Thus, there is a
Figur e 13.19. Human factors interaction model.
Figure 13.18. Error modes (adapted from Hollnagel, 1 998) .
1 3. 1 2 HUMAN FACTORS ANAL YSIS 397
demand-pull from the assignment on performance. At the same time, personnel and
hardware interact in their performance (performance-push) and in doing so they may
cause accidents, i.e., the performance is not in agreement with the assignment. Not very
surprisingly, the first place where one notices an immediate cause of an accident is at the
level of performance: this is where personnel interact with tools to accomplish the tasks
defined by the assignment.
In addition to the personnel aboard and the tools with which they are equipped to
fulfil the assignment, there is always the possibi lity of external events influencing
the interaction. External events may be bad weather, other ships, or events preceding the
one under consideration. The last element of the model is the result of the interaction
between assignment, personnel, tools, and external events: the outcome of an event. The
checklist for analysis of human factors is outlined in Table 13.12.
13. 13 T HE CASM ET APPROA CH
13. 13. 1 Overview
The CASMET approach rests on two pillars: an analytical method and a structure
for coding information in a database (Kristiansen et al., 1999). The analytical method
answers the question how the information should be obtained. The question of how the
information obtained should be represented in a database will be dealt with by the outline
of the coding and database structure. The main steps that both pillars adhere to can be
outlined as follows:
1. Initial data collection.
2. Identification and reconstruction of events.
3. Human factors analysis.
4. Systems, hazardous materials and environmental analysis.
5. Summary of causal relations.
The relation between the analysis process and the resulting information to be coded,
structured and stored in a database is represented in Figure 13.20.
The CASMET method has four basic levels for representing a maritime casual ty
(Figure 13.21); namely:
. Casualty events
. Accidental events
. Basic causal factors relating to daily operations on board
. Basic causal factors relating to management and allocation of resources.
The casualty an d accidental events should be viewed in the time domain wher eas the causal
factors are logically linked to each other and to the events. The intention with this
representation format is to keep the process character of events and the logical links
between what it is possible to observe in order to understand the causal nature.
398 CHAPTER 13 A CCI DENT ANALY SIS
Ta b l e 13 . 1 2 . Human factors checklist
External Performance Causal mode
Personnel Tool Assignment
Previous event(s) Detection: Lack of knowledge due to HMI/design Task characteristic:
Technical failure lack of experience, Compatibility Ambiguous task
Personnel factor lack of orientation, Consistency Habit-ignoring task
Other ships Lack of support inadequate training, Context Distracters in task
information overload, Structure and systematics Inadvisable rules
lack of information Feedback Error-enforcing task
Bad weather Assessment:
Technical failure Lack of skills due to Workload Staffing characteristic:
Personnel factor inadequate instruction, User-directed flexibility Personnel selection
Criminal acts Lack of support inadequate training, Work schedule
infrequent practice, Technical problems due to Workload
lack of coaching poor construction, Understaffing
Decision: poor maintenance, Poor training
Technical failure unavailable equipment Poor motivation
Personnel factor
Lack of support Intoxication due to
alcohol use,
drug use, Poor procedures:
medicine use, Damage due to Operating procedures
fumes and gases wear-out, Housekeeping procedures
Action: fire/explosion, Maintenance procedures
Technical failure Fatigue/stress due to physical intrusion, Communication procedures,
Personnel factor task load or duration, radiation, Emergency procedures
Lack of support lack of rest, electromagnetism
sensory overload, Incompatible goals:
information overload, Time pressure
climate, Budget
time stress
Poor communication:
Causal group
m
Reduced ability due to
m
Causal factor
Ambiguous information
physical condition, Language problems
mental condition, Lack of information
emotional condition Too much information
13.13 THE CASM ET APPRO A CH 399
Figure 13. 21. Casualty representation levels.
Figure 1 3.20. Relation between analysis and database s tructure.
400 CHAPTER 13 AC CID ENT ANALYS IS
1 3. 1 3.2 Identification of Casualty
The starting point in an accident analysis is the factual or objective information about
the vessel, the subjects involved, the circumstances, and the consequences of the casualty
(Table 13.13). The third of these fact groups – operation – is coded in accordance with this
(Table 13.14). A distinction is made between operation of the vessel and operations
onboard the vessel.
13.13.3 Casualty Summary
In order to give an introductory outline of the casualty, a summary in prose style should
be given. The CASMET project found that a summary structure consisting of four
segments was feasible (Table 13.15). There should not be any formal requirements for
the description other than a limitation on the length of the text. The main objective of the
summary is to give a synopsis of the casualty that captures all the main aspects of the
accident and the lessons learned.
As an illustration, the example on pages 369–370 above shows how the summary may
be written for the Herald of Free Enterprise accident.
Ta b l e 13 . 1 3 . Fact sheet structure
Fact group Facts
Identification Case identification no.
Vessel name
Terminal casualty (final casualty event in a sequence)
Date of casualty
Geographical position
Vessel Vessel type
Deadweight or GRT
Service speed
Main dimensions ( L
pp
, B, D)
Cargo intake, draft (T)
Main engine type, propulsion system
Yard, country, year of build
Owner, flag
Classification society
Operation Vessel operation phase
Operation onboard
Environmental conditions Weather conditions, visibility
Beaufort no., current speed
Manning Number of officers and crew
Nationalities
Experience of key personnel
Consequences Damage to people, vessel and environment
Economic consequences
13.13 THE CASM ET APPRO A CH 401
13.13.4 Event De scription
It is important to establish an overall picture of the sequence of events from initiation of
the accident to the outcome. This is solely so that one can establish what happened before
one starts to deduce why the casualty happened. Most casualties are reported in prose
style. Although this may be acceptable to capture certain information, it has obvious
shortcomings. Some accidents may develop gradually over a considerable time span and
involve a number of actors in terms of persons and systems.
In the earlier sections of this chapter it has been shown how the event description can
be given some structure or formalism with either the STEP diagram (Table 13.4) or a
flowchart (Figure 13.4).
13.13. 5 Ca su alt y Event
A casualty case may involve one or more casualty events and accidental events. The latter
are errors and failures that contribute to the casualty but do not necessarily lead to
Ta b l e 13 . 1 5 . Casualty summary
Summary Key elements
What happened Short account of casualty
Vessel, place, time, outcome
Circumstances, contributing factors Latent technical and operational factors
Environmental conditions
Other hazardous conditions putting vessel at risk
Immediate causal factors What happened onboard
Events, failures and errors
Individuals, systems involved
Basic causal factors Why did it happen
Individual conditions and shortcomings
Inadequacies of management and workplace
Lack of resources
Ta b l e 13 . 1 4 . Coding of operation group
Parameter Coding
Vessel operation phase Sailing Anchored
Enter/leave port
In port
Manoeuvring At repair yard
Under tow
Operation onboard Normal watch Cargo transfer Ballasting
Mooring Tank cleaning Repair
Loading Gas freeing Idle, off-hire
Unloading Bunkering
402 CHAPTER 13 AC CIDENT ANALY SIS
damage and serious consequences in themselves. Accidental events will be described in the
next section. A casualty event, on the other hand, involves release or transformation of
energy and is usually understood as the outco me of the casualty. Casualty events are listed
in Table 13.16.
The taxonomy also addresses so-called maloperations (Table 13.17). These events are
a gro up of accidents where the consequences are less critical for the vessel itself but of
greater consequence for the environment. It is mainly a matter of spill and pollution of oil.
Ta b l e 13 . 1 7. Maloperation c asualty event
Casualty event Casualty event subgroup
Polluting cargo Polluting cargo operations
Cargo lost overboard
Atmospheric pollution
Ballast Unclean ballast pumped overboard
Failure in ballast system
Engine room operations Spill/dumping of bunker and sludge
Polluting engine room operations
Ta b l e 13 . 1 6 . Vessel-related casualty events
Casualty event Casualty event subgroup
Allision Ramming of buoy, marker
Ramming of quay
Collision with floating objects
Grounding Powered grounding
Intentional grounding
Drift grounding
Collision With other vessel
With multiple vessels
Fire/explosion Fire
Explosion – incindiary
Explosion – pressure vessel
Flooding (Founder) Sinking
Capsize
Structural failure
Loss of control Loss of electrical power
Loss of propulsion power
Loss of directional control
13.13 THE CASM ET APPRO A CH 403
The casualty events are qualified with two different attributes, namely class and state.
These attributes express certain aspects of the situation when the casualty event took
place. The semantics may, however, vary for the individual casualty events. To give an
example, the collision took place in a crossing situation and the vessel was disturbed by the
channel effect. The definition of the attributes is given in Table 13.18 for maloperation and
in Table 13.19 for casualty events.
13.13. 6 Accident al Event
Based on the preparatory steps described in the preceding chapters, the basis for
classification and coding of accidental events is established. It is important to keep in mind
that accidental events are strictly related to the casualty case sequence as it is observed or
reconstructed – or, in other words, only what did happen.
The casualty may be initiated and escalated by one or a series of accidental events.
We make a distinction between the following types:
. Equipment failure
. Human error
. Hazardous material
. Environmental effect
. Other agent or vessel
This event categorization is defined further in Table 13.20 (p. 406). It is important
to note that the description of events has no causal implication. This is especially vital
when we deal with ‘human error’. This event simply expresses that a wrong or inadequate
behaviour of an individual was observed. However, it does not explain the behaviour. This
is a later topic in the analys is process.
Similar to casualty events, we have defined certain attribute class es for each
accident event type as shown in Table 13.21 (p. 406). The aim is to produce a ‘richer’
Ta b l e 13 . 1 8 . Maloperation casualty class and state attributes
Casualty type Class State
Polluting cargo Intentional Small and momentous
Unintentional Intermittent over time
Continuous spill over time
Progressive
Uncontrolled
Ballast Intentional
Unintentional
Engine room operations
404 CHAPTER 13 AC CID ENT ANALYS IS
description of the events. We will comment briefly on the attributes for each accidental
event type.
Physical processes involving release or conversion of energy constitute an important
aspect of maritime casualties. Events involving hazardous material will therefore be coded
Ta b l e 13 . 1 9. Coding of casualty class and state attributes
Casualty event Class State
Allision Crossing Current
Meeting Wind
Collision Overtaking Wave drift
Multi-approach Channel effect
Uncontrolled No external effect
Grounding Straight course
In turn
Multiple course changes
Stopping
Backing
Uncontrolled
Fire Burned out Bow
Controlled Cargo space
Extinguished Weather deck area
Initial Living space
Uncontrolled Navigation space
Engine/pump room
Stern
Explosion Initial
Secondary
Flooding Controlled
Dewatered
Initial
Progressive
Uncontrolled
Structural failure Class 1 Deck
Class 2 Frame
Class 3 Hull
Structural damage Keel
Longitudinal
Watertight bulkhead
Loss of control Restricted fairway Current
Heavy traffic Wind
Both Wave drift
Channel effect
No external effect
13.13 THE CASM ET APPRO A CH 405