Kristiansen Svein. Maritime Transportation: Safety Management and Risk Analysis
Подождите немного. Документ загружается.
in this example). The number of fatalities is found by multiplying the statistical number of
fatalities (e.g. 116 for collisions; see Table 10.7) by the probability of fatality.
Step 3: Establish safety measures (risk control options)
Risk control options (RCOs), with the objective of improving the life-saving capability of
LSAs in a cost-effective manner, were identified and agreed upon by a multidisciplinary
team of experts. The following RCOs were some of the measures identified for further
assessment in this FSA study:
. Sheltered mustering and lifeboat area (SMA)
. Level alarms to monitor water ingress in all holds and forepeak (LA)
. Individual immersion suits for all personnel (IS)
. Free-fall lifeboat (FF)
. Marine evacuation system for thrown overboard liferafts (MES)
. Redun dant trained personnel (RTP)
Figur e 10.11 . Event tree for the undesirable events following a ship accident until preparation of the LSA.
30 6 CHAPTER 1 0 FORMAL SA FETY ASSESSMENT
The different RCOs affect the event trees modelled in step 2 of the FSA (i.e. the risk
assessment step), resulting in changes to the probability of fatalities associated with
evacuation. The RCOs will affect each accident scenario differently, and co nsequently
the relative changes in the potential loss of life (PLL) value will vary both in terms of the
RCOs implemented and the accidental event under consideration. Table 10.10 gives the
relative change in percent for the probability of fatality in different accidental events for
the RCOs listed above. This table is the underlying basis for calculating the cost-benefit
relationship in the next step of the FSA methodology.
Step 4: Cost-benefit assessm ent
In order to establish a common and comparable cost-benefi t ratio, all the potential costs
and benefits related to the different risk control options (RCOs) must be identified. In this
particular FSA study the cost estimation was done primarily by contacting suppliers of
life-saving appliances, training centres, ship yards and ship owners. Through these sources,
estimates were established on relevant costs such as investment in equipment, inst allation
at the bulk carrier, inspection, maintenance, training of personnel to operate the
equipment installed, etc. A high and a low cost estimate wer e established to take account
of, among other things, factual variability in cost in Western Europe and the Far East,
Ta b l e 1 0 . 9. Probability of fatality associated with evacuation
Type of
accidental event
Based on evacuation model Based on statistics
Probability of
fatality (%)
PLL
(per ship year)
Probability of
fatality (%)
PLL
(per ship year)
Collision 45.7 3.392 10
3
35 2.6 10
3
Contact 44.1 1.173 10
3
45 1.2 10
3
Fire/explosion 27.7 2.347 10
3
2 1.7 10
4
Foundered 55.4 1.497 10
2
51 1.3 10
2
Hull/machinery failure 16.2 5.180 10
4
00
Wrecked/stranded 20.2 2.461 10
3
00
Ta b l e 1 0 .1 0 . Reduction in the probability of fatality (%) with RCO implementation
Current SMA LA IS FF MES RTP
Collision 45.7 0.3 0.0 1.2 1.1 0.3 0.2
Contact 44.1 0.3 0.0 0.9 1.3 0.1 0.2
Fire/explosion 27.7 0.3 0.0 4.5 þ0.4 1.0 0.2
Foundering 55.4 0.9 14.8 2.0 4.9 0.1 0.4
Hull/machinery failure 16.2 0.5 0.0 1.6 3.0 0.2 0.4
Wrecked/stranded 20.2 0.5 0.0 1.5 1.8 0.3 0.3
10.8 APPLIC ATION OF T HE F SA ME T HODOLOGY 307
and the fact that the proposed technical solutions were not specified in detail. Depreciation
of future costs was carried out at a real risk-free rate of return of 5%.
The benefits obtained by implementing the RCOs are the reduced number of fatalities,
which may be calculated as a reduction in the potential loss of life (PLL) parameter using
Table 10.10. For each RCO the total reduction obtained in the probability of fatality with
RCO implementation was calculated as PLL (in statistical terms the PLL is the expected
loss of life). The PLL was calculated as shown for the RCO of ‘sheltered mustering and
lifeboat area’ (SMA) in Table 10.11.
The Implied Cost of Averting a Fatality (ICAF) was calculated using the following
equation:
ICAF ¼
Cost of RCO
Reduction in PLL
The cost estimates and the reduction in PLL are in this example calculated for a lifetime
expectancy of 25 years for all RCOs. This is a simplification of the approach used in
the reference source (DNV et al., 2001). These simplifications are made to reduce the
size and complexity of this example, and result in slightly different ICAF values than
those presented in the original FSA report. The final recommendations are, however,
the same.
Both a high and a low ICAF value was calculated for each RCO based on the high and
low cost estimates, respectively. For this particular analysis a decision criteria was based
on an ICAF of £1 million. Other decision criteria may, however, have been selected, and
this may have given different recommendations. Table 10.12 shows the calculation of
the ICAF values. A RCO is recomm ended if its low (cost estimate) ICAF value is within
the decision criteria of £1 million. A recommendation is considered robust if the high
(cost estimate) ICAF value gives the same recommendation as the low (cost estimate)
ICAF value.
Ta b l e 1 0 .11 . Calculation of PLLfortheRCO‘shelteredmusteringandlifeboatarea’(SMA)
Type of
accidental event
Reduction in
probability of
fatality (%)
Resulting
probability of
fatality (%)
Resulting PLL
with RCO
(per ship year)
Resulting PLL
without RCO
(per ship year)
Collision 0.3 45.4 3.370 10
3
3.392 10
3
Contact 0.3 43.8 1.165 10
3
1.173 10
3
Fire/explosion 0.3 27.4 2.321 10
3
2.347 10
3
Foundered 0.9 54.5 1.473 10
2
1.497 10
2
Hull/machinery failure 0.5 15.7 5.019 10
4
5.180 10
4
Wrecked/stranded 0.5 19.7 2.400 10
3
2.461 10
3
Total PLL: 2.449 10
2
2.486 10
2
PLL ¼3.7 10
4
308 CHAPTER 10 FORMAL SA FETYASSESSMENT
Step 5: Recommendations for decisi on-making
Based on Table 10.12, the following risk control options provide considerable improved
life-saving capabil ity in a cost-effective manner and are therefore recommended:
. Level alarms to monitor water ingress in all hol ds and forepeak (LA)
. Individual immersion suits for all personnel (IS)
. Free-fall lifeboats (FF)
The RCO of free-fall lifeboats (FF) is only relevant for implementation on new ships
as an alternative to more traditional LSAs (e.g. conventional lifeboats). The costs of
fitting this RCO to an existing ship would make it unattractive in cost-benefit terms.
Both the other two recommended RCOs are relevant for implementation on both new and
existing ships.
If all the three recommended RCOs are implemented on a (new or existing) bulk
carrier, the risk assessment procedure indicates that the evacuation success rate in the
dominating foundering scenario increases from 44.6% (i.e. 1.0 ¼probability of fatality) to
66.2%, which is quite an improvement. Similar improvements are, however, not present
for the other accident scenarios, and the general success rate in evacuation remains rather
low even after implementing the recommended RCOs. This should call for additional
measures, in particular measures with a focus on crew training and competence building.
Ta b l e 1 0 . 1 2 . Calculation of ICAF values for the RCOs
RCO PLL
(per ship
year)
PLL
(per ship
over 25 years)
Cost of RCO
over 25 years
(£ 10
3
)
ICAF
(£ 10
3
)
Recom-
mended?
a
Robust
recom-
mendation?
b
Low High Low High
SMA 3.70 10
4
9.25 10
3
10 20 108.1 216.2 No Yes
LA 4.00 10
3
1.00 10
1
14 21 140.0 210.0 Yes Yes
IS 1.27 10
3
3.18 10
2
15 17.8 471.7 559.7 Yes Yes
FF 1.72 10
3
4.30 10
2
7.8
c
18.2 181.4 423.3 Yes Yes
MES 1.80 10
4
4.50 10
3
4.50
d
Criterion £ 1 m. Not known
RTP 1.95 10
4
4.88 10
3
8 10 1639.3 2049.2 No No
a
A RCO is recommended if its low (cost estimate) ICAF value is within the criterion of £1 million.
b
A recommendation is robust if the high (cost estimate) ICAF value gives the same recommendation as the low
(cost estimate) ICAF value.
c
Free-fall lifeboats have a lower cost than more traditional LSAs (e.g. conventional lifeboats) when implemented
instead of these on new ships.
d
The maximum cost of a marine evacuation system to meet the ICAF criterion of £1 million (the costs are likely
to be considerably higher).
10.8 APPLIC ATION OF T HE F SA ME T HODOLOGY 309
REFEREN CES
AEA, 1995, Contract Draft Interim Report 1 – MSA Project 388: Step Four Methodology ‘Cost-
Benefit Assessment’. AEA Technology, Deliverable 388D2, Warrington, UK.
DNV et al., 2001, Formal Safety Assessment of Life Saving Appliances for Bulk Carriers (FSA/
LSA/BC). Project participants: Det Norske Veritas, Norwegian Maritime Directorate,
Norwegian Union of Marine Engineers, Umoe Schat-Harding, Norwegian Shipowners’
Association, International Transport Workers’ Federation, Oslo.
IMO’s web site, http://www.imo.org/meetings/msc/68/fsa.htm
MSA, 1995, Project 366: Formal Safety Assessment: Draft Stage 1 Report. Marine Safety Agences,
Southampton, February.
Skjong, R., 1999, FSA of HLA on Passenger Vessels. Report No. 97-2053, Det Norske Veritas, Oslo.
310
PART IV
MANAGEMENT AND OPERATIONS
11
HUMAN FAC TORS
Eternal vigilance is the price o f safety.
(Fleet Admiral Chester W. Nimit z)
11.1 INTRODUC TION
In this chapter we will discuss the human factors aspects of ship operation. Although
certain functions have been automated, a ship is still largely a human-controlled system.
Two key concepts are sometimes used more or less as synonyms, namely ergonomics
and human factors. There is, however, a certa in distinction made between the concepts in
some circles, although both focus on the interaction between humans and physica l work
environment and machines:
. Ergonomics: Particular emphasis is put on the design of displays, controls and the
workplace. The human physical dimensions (anthropometry) and their capacity with
respect to sensing and control ability are especially taken into consideration. The
physical aspects of the environment such as climate, noise and vibration are also
a concern.
. Human factors: The work situation is assessed in the light of psychological factors. A
key aspect is the relation between the job or task requirements and the human capacity.
Factors such as mental capacity to process informatio n, motivation, and interaction
with colleagues have to be taken into consideration.
In the rest of this chapter we will use human factors (HF) as a common term. In the
course of the last five decades HF has established itself as a solid scientific discipline on
the basis of theory and empirical data. For the practitioner, a number of handbooks
and guides are available (Salvendy, 1987; Wilson and Corlett, 1990; McCormick and
Sanders, 1983; VanCott and Kincade, 1972).
313
11. 2 HUMAN E RROR
11.2.1 A Misused Term
It has for many years been popular to explain ship accidents by human error. There
are numerous citations indicating that 75–90% of accidents are rooted in human error.
Sometimes one may feel the element of condemnation or that the error is a result of
negligence. Often our understanding of vague concepts is biased towards our own
professional background:
. Legal: focus on negligence or criminal conduct
. Technical: the operator misused the system
. Psychological: the operator is inhibited by trauma
. Socio-technical: the operator is estranged
. Ergonomic: incapacitated
This means that for instance the explanation of a stran ding by ‘navigation error’ may have
different interpretations, e.g.:
. Negligence due to low morale
. The control system did not give any feedback
. Wrong assessment due to lack of skills
. Inadequate performance due to lack of procedure
. Electronic disturbance
. Perce ption error due to low arousal
The lesson from this is that the human error concept in itself is of limited value unless it is
described in a broader context.
11. 2 . 2 E arlie r St u die s
It often seems difficult to keep in mind that an accident seldom has only one single
explanation but rather a number of them. This is disturbing given the fact that this was
pointed out more than 30 years ago (Stewart, 1973). By discussing a classical stranding
scenario (Figur e 11.1), he demonstrated the interaction between various causal factors.
The factors that led to the stranding were:
1. Vessel was approaching a dangerous headland.
2. Wind was blowing strongly from north.
3. Attempt was made to adjust for setting to the southward.
4. No allowance was mad e for leeway.
5. Despite a later position fix, heading was not corrected.
6. Echo sounder was not in use.
7. Decreasing visibility near land.
8. Light picked up to starboard.
314 CHAPTER 11 H UMAN FACT O RS
9. Assumed to be the headland light.
10. Stranded at full speed.
One of the first studies of human error in ship operatio n (NAS, 1976) focused on the
following factors:
. Inattention . Excessive alcohol use
. Ambiguo us pilot–master relationship . Excessive personnel turnover
. Ineffici ent bridge design . High level of calcul ated risk
. Poor operational procedures . Inadequate lights and markers
. Poor physical fitness . Misuse of radar
. Poor eyesight . Uncertain use of sound signals
. Excessive fatigue . Inadequate rules of the road
It is interesting to notice that most of these problems are just as relevant today. We can
see that the investigation had a broader view which took workplace factors, procedures,
fatigue, health and management into consideration.
Fukushima (1976) took an even broader view by addressing the effect of external
conditions. He saw accidents as combinations of the following complex conditions:
. Natural phenomena: weather and sea
. The route: fairway conditions, obstacles and visibility
. Ship: stability, manoeuvrability and technical standard
. Traffic congestion
. Navigator: knowledge, skills and health
It can be concluded that we have an understanding of why accidents happen that
involves the operator, technology, work conditions and organization. Despite this we still
seem to take a more narrow view both in design and planning of operations. This fact
is witnessed by bridge design shortcomings, lack of safety training and ignorance in
Figure 11.1. A st randin g scenario . (Adapted from Stewart,1973.)
11.2 H UMAN ERR OR 315