Kristiansen Svein. Maritime Transportation: Safety Management and Risk Analysis
Подождите немного. Документ загружается.
By allowing each member of the group of experts to answer such a questionnaire,
a quantitative estimate of the resulting relative weights of importance for the safety
parameters can be calculated. In practice this calculation is executed by normalizing the
matrix in the questionnaire (i.e. Table 9.5). The relative weights of importance are then
calculated on a common scale and may be presented as shown in Figure 9.9.
The next step is to establish a utility function for each of the four safety parameters. In
this context, utility denotes the decision-maker’s scale of preference, and a utility function
describes graphically how the perceived utility changes with changing consequences.
It is assumed here that the utility can be expressed as a continuous function with values
between 0 and 1. The higher the utility (i.e. the closer to 1), the more acceptable a specific
consequence is for a given accident. Similarly, as a consequence becomes more and more
unacceptable and undesirable, the lower the utility will be (i.e. closer to 0). Thus, the utility
functions express the perceived risks, with the risk increasing as the utility number
decreases. The utility functions can also be established using a group of experts. The
utility functions for the four safety parameters in the oil transportation case are shown
in Figure 9 .10. The graphs presented in Figure 9.10 require the followin g comments:
. Spill volume: No spill volume gives the highest utility (i.e. 1.0) as no environmental
damage can be regarded as acceptable. However, small oil spills do not cause
substantial damage to the environment, and the utility remains relatively high.
Medium oil spills in the order of 100–1000 tons will, on the other hand, reduce the
utility significantly as the consequences to the environment increase dramatically. Even
larger spills (i.e. >1000 tons) will have very detrimental effects on the environment and
cannot be accepted, hence the low utility value for such spills.
. Economic loss: Econom ic losses less than NOK 10 million have relative ly little effect on
the utility, which remains relatively high. This shows that such material losses in this
particular case are regarded as quite acceptable. One possible explanation of this is that
such costs may be involved even in smaller accidents (i.e. a threshold cost). Losses
larger than NOK 10 million have a significant effect on the utility function, and costs
in the order of NOK 100 million are considered totally unacceptable – hence the utility
value of 0.
. Fatalities aboard: Accident statistics reveal difficulties in avoiding 1–3 fatalities,
and although very undesirable the utility therefore stays relatively high. Accidents of
Figur e 9. 9. Relative weights of importance for the set of safety assessment parameters.
266 CHAPTER 9 COST -BEN EFIT ANALYS IS
size 3–10 fatalities are considered much more serious, hence a dramatic reduction in
the utility value. Accidents with more than 10 fatalities are catastrophes that must be
avoided. As can be seen in Figure 9.10, the utility value is set at 0 for accidents
involving 20 or more fatalities.
. Public exposure: With regard to public exposure, the utility function is close to being a
linear function of the logarithm of the exposed population. An exposed person may
in this context be defined as a person within a given distance from the centre of a fire
or explosion, resulting in that person being subjected to serious danger. As would be
expected, the utility decreases considerably as the number of exposed individuals
increases.
The relative weigh ts of importance and the utility functions for the selected set of safety
assessment parameters gives us the necessary foundation for estimating the cost-benefit
ratio for different system concepts (e.g. designs of the oil transportation vessel/system).
Let us assume that two concepts are to be compared against each other on the basis of
the four safety parameters and the defined utility functions. The two oil transportation
concepts are described in Table 9.6. The oil is loaded from a buoy offshore and shipp ed
with two shuttle tankers to a shore-based terminal and refining facility.
Figure 9.10. Utility functions for the set of safety parameters.
9.4 ALTERNATIVE PROBLEM-SOLVING APPROACHES 267
The basis concept is clearly the most favourable when considering environmental
damage as the expected spill volume is only one-third of the spill volume for the alternative
concept. In addition, the exp ected number of the exposed population is only half for
the basic concept. The alternative, on the other hand, is more oriented in reducing ship
damage in terms of expected economic loss, and the expected number of fatalities aboard
is also considerably lower. In Table 9.7 a cost-benefit analysis of the two concepts is
performed using the weights of importance and the utility graphs. The total utility, which
is considered as a measure of the benefit/quality of a specif ic concept, is calculated and
compared to the total costs, i.e. the net pr esent value of the necessary investments and
future operational costs. A cost-benefit ratio is calculated, enabling a comparison to be
made on a common scale. In Table 9.7 the ‘priority’ is the relative weights of impor tance
for the safety parameters/criteria, and the utilities are read off the graphs in Figure 9.10
depending on the characteristics of the two concepts. The weight is the product of
the priority figure and the utility. Table 9.7 shows that the alternative concept has
a lower cost-benefit ratio than the basic concept, and is hence considered better in this
particular analysis.
9.4.3 Valuat ion of Co nseq u ence Param et e rs
In earlier sections of this chapter it is assumed that the potential consequences or losses for
a particular concept can be expressed as concise values such as, for example, the number
Table 9.6. Alternative oil transportation concepts
Basis concept Alternative
Spill volume 500 tons 1500 tons
Economic loss NOK 16 million NOK 4 million
Fatalities aboard 8 2
Exposed population 50 100
Total costs NOK 800 million NOK 770 million
Ta b l e 9. 7. Cost-benefit analysis (CBA) of the two oil transportation concepts
Criterion Priority Basis concept Alternative
Value Utility Weight Value Utility Weight
Spill volume 0.20 500 0.95 0.190 1500 0.65 0.130
Economic loss 0.10 16.0 0.80 0.080 4.0 0.96 0.096
Fatalities aboard 0.30 8 0.55 0.165 2 0.98 0.294
Exposed population 0.40 50 0.75 0.300 100 0.60 0.240
Total utility 0.735 0.760
Costs (million) 800 770
C-B ratio 1088 1013
268 CHAPTER 9 COST -BEN EFIT ANALYS IS
of fatalities. In reality, however, this is sometimes problematic and not always desirable,
especially in relation to accidents that involve both injuries and fatalities. In the context
of CBA it is often necessary to establish an economic cost of both injuries and fatalities
in order to compare different costs and benefits on a common scale. This section will
look more closely at the valuation of people/personnel-related consequence parameters.
In analys es of the safety of people/personnel, the following consequence parameters may
be considered:
. Fatality
. Permanent disability
. Temporary disability
The costs associated with these people-related consequences may be valued according to
the following factors:
. Insurance payments
. Estimated remaining life-income
. Claim for compensation
. Implicit social costs
The valuation of people-related consequences can be extremely controversia l, both
politically and ethically. In Table 9.8 a possible valuat ion method for such consequences
is applied on two different concepts having different potential consequences in terms
of fatalities as well as permanent and temporary disability. Unitary economic costs
are applied to express the cost of one fatality, one permanent disability and one temporary
disability. These values can be calculated on the basis of the costs stated above. In
Table 9.8 these values are only chosen randomly to illustrate the method, and it will be
described in detail how these values may be established. The total costs for the two
concepts are calculated as the sum of the preventive costs (i.e. costs of implementation)
and the consequences (i.e. average losses in an accident). Concept 2 gives the lowest total
costs in this particular study.
Ta b l e 9. 8 . To tal people-related loss costs of two different concepts
Safety parameters Unitary cost
(1000 USD)
Concept 1 Concept 2
Number Cost Number Cost
Fatality 1,000 2 2,000 1 1,000
Permanent disability 400 15 6,000 10 4,000
Temporary disability 20 60 1,200 100 2,000
Calculated people-related loss cost 9,200 7,000
Costs of implementation 5,500 6,100
Total costs 14,700 13,100
9.4 ALTERNATIVE PROBLEM-SOLVING APPROACHES 269
McCormick (1981) introduced the term social costs, which expresses the costs to
society of an injury or fatality. The following equation was suggested:
Social cost ¼ N C ð1 þ iÞ
t
, when t < 6000
¼ N C ð1 þ iÞ
6000
, when t 6000
ð9:15Þ
where:
N ¼ Number of injuries or fatalities
C ¼ Cost of damage per day
i ¼ Daily rate of interest
t ¼ Duration of damage or sick leave in days (6000 days is equivalent with a fatality)
This particular method of calculating the social cost of injuries and fatalities is only
one of many different models that may be used. A problem not solved with Eq. (9.1 5) is
that of establishing the cost of damage per day (i.e. C), which may vary considerably
depending on the type of injuries suffered, the country in question, etc.
In an investiga tion made by O’Rathaille and Wiedemann (1980) an attempt was made
to establish the average social cost for ship collisions and groundings based on statistical
consequences. It was focused on oil spills and loss of lives, and the statistical data basis
is presented in Tables 9.9 and 9.10. These tables show that the likelihood of oil spills is
largest in g roundings, while collisions much more frequently result in fatalities.
Based on the experience of known accidents, the cost of oil spills per accident was
estimated to £5100–£6100 for collisions and £50,000–£280,000 for groundings, reflecting
Ta b l e 9. 9. Ship accidents and oil spills ( 197 6)
Primary cause Number of accidents Pollution rate, % of accidents
leading to pollution
Total Of which led
to pollution
Collisions 44 1 2.27
Groundings 121 14 11.57
Ta b l e 9. 1 0 . Fatality risk in ship collisions and g roundings ( 197 6)
Primary cause Number of accidents Number of lives
lost per accident
Total Number of
lives lost
Collisions 44 41 0.93
Groundings 121 4 0.03
270 CHAPTER 9 COST-BENE FIT ANALYS IS
that groundings are more likely to result in oil spills and that the spills on average are
larger. However, as these figures show, the costs related to groundings tend to vary
greatly.
In an assessment combining both economic and non-economic factors related to
fatalities, the cost of a fatality was estimated to be £85,170–£98,105 in 1977 prices. On
the basis of this, the average total costs of collisions and groundings, respectively, are
estimated in Table 9.11. It can be concluded from Table 9.11 that the average total social
cost of a grounding accident seems to be higher than that of a collision.
Insurance payments in the aftermath of accidents to people as well as the company/
organization involved must also be considered a cost related to accident s. Insurance
companies give compensation to the bereaved, and tend to vary considerably from case
to case depending on the circumstances and the insurance schemes. There also tend to be
quite different insurance practices in different countries. All these aspect s make it difficult
Ta b l e 9. 1 2 . Cost-benefit ratios (C/B) for different safety measures, USD10
6
per life saved
Industry/activity Safety measure C/B
Nuclear industry Radwaste effluent treatment systems 10
Containment 4
Hydrogen recombiners >3000
Occupational health and safety OSHA coke fume regulations 4.5
OSHA benzene regulations 300
Environmental protection EPA vinyl chloride regulations 4
Proposed EPA drinking water regulations 2.5
Fire protection Proposed CPSC upholstered furniture
flammability standards
0.5
Smoke detectors 0.05–0.08
Automotive and
highway safety
Highway safety programs 0.14
Auto safety improvements, 1966–70 0.13
Airbags 0.32
Seat belts 0.08
Medical and health care programs Kidney dialysis treatment units 0.2
Mobile cardiac emergency treatment units 0.03
Cancer screening programs 0.01–0.08
Ta b l e 9. 11 . Average to tal c o s ts of collisions a nd grounding s (197 7 )
Primary cause Cost of spills
per accident (£)
Cost of fatalities
per accident (£)
Average total cost
per accident (£)
Collisions 5,100–6,100 79,363–91,416 84,463–97,516
Groundings 50,000–280,000 2,816–3,243 52,816–283,243
9.4 ALTERNATIVE PROBLEM-SOLVING APPROACHES 271
to generalize about insurance payments. The same accounts for c laims of compensation
that often surface in the wake of accidents. Suc h claims are often based on the lost
(entirely or partly) remaining life income by reaching nominal a ge. Methods used in
calculating such figures are often referred to as human capital methods.
The willingness to pay for preventive safety measures geared towards reducing
fatalities differs between industries and types of activities. Table 9.12 shows an American
overview of estimated preventive measures and costs per saved human life for different
activities. The table shows that the nuclear industry, for example, is willing to pay more to
save a human life than most other activities.
When studying the cost-benefit values in Table 9.12 it must be recognized that such
values usually have a limited period of validity because of factors such as regulatory
changes, new technology, changed public risk perception, etc. Cost-benefit values must
therefore be used or referred to with great care.
9. 5 CBA OF OIL S PILL PREVENTION MEASURES FOR TANKERS
Exam p le
Problem
Preventing pollution from maritime activities has been a major priority in recent decades.
On an international basis, concerns about the environmental impact of shipping have
resulted in MARPOL, i.e. the International Convention for the Prevention of Pollution
from Ships. MARP OL, which is one of the more important agreements achieved within
the context of the International Maritime Organization (IMO), comprises design and
operational regulations and requirements geared towards reducing pollution to both air
and sea from shipping. In addition to MARPOL there exist several regional agreements
and regulations on the prevention of pollution from shipping. One such set of regulations
is the US Oil Pollution Act of 1990 (OPA 90), which was established as a direct
consequence of the Exxon Valdez grounding accident in 1989 that resulted in a spill of
33,000 tons of crude oil in Prince William Sound on the coast of Alaska. OPA 90 gives
shipowners full economic liability for spills in US coastal waters. There are several safety
measures that may reduce potential oil spills as a result of ship collisions and groundings.
The US National Research Council therefore pe rformed a cost-benefit analysis on some of
these possible safety measures for tankers, and this example is a summary of this analysis.
Solution
The objective of this analysis is to calculate the cost effectiveness of alternative designs for
oil spill prevention. These are compared to a standard MARPOL tanker with protectively
located segregated ballast tanks (PL/SBT). Segregated ballast tanks (SBT) means that
there shall be designated tanks for ballast and that ballast is not to be carried in cargo
tanks (except in very severe weather conditions in which case the water must be processed
and discharged in accordance with specific regulations). Protective location (PL) of SBT
means that the required SBT must be arranged to cover a specified percentage of the side
and the bottom shell of the cargo section in order to provide protection against oil outflow
272 CHAPTER 9 COST-BENE FIT ANALYS IS
in the case of groundings and collisions. The alternative designs (i.e. safety measures)
studied in this analysis are explained below:
1. Double bottom (D B)
The double bottom (DB) constitutes the void space between the cargo tank plating, often
referred to as the tank top, and the bottom hull plating. MARPOL requires a DB of 2 m
or B/15, whichever is less (B ¼breadth of the vessel). The double bottom space gives
protection against low energy grounding. Another benefit is the very smooth inner cargo-
tank surfa ce which facilitates discharge suction and tank cleaning. Drawbacks to a double
bottom include increased risks associated with poor workmanship, corrosion, and
obstacles to personnel access (to the DB). Other drawbacks are related to reduced side
protection relative to the PL/SBT configuration, and increased explosion risk related to
cargo flow into the DB. The double bottom configuration is shown in Figure 9.11.
2. Double si des (DS)
The double sides (DS) constitute the void space between cargo side/wing tanks and hull
side plating. The minimum width of the DS is equal to that of the DB (i.e. 2 m or B/15).
However, in order to meet ballast requirements, the width is likely to be larger (normally
B/7–B/9). The design offers good protection agains t collisions, and also some of the
advantages from doubl e bottoms as the side tanks protect the outboard region of the
bottom. Drawbacks of the DS configuration are related to bottom damages that result
in direct spills, as well as higher susceptibility to asymmetric flooding. The double sides
configuration is shown in Figure 9.12.
3. Doubl e hull ( DH)
The double hull (DH) constitutes the void space between tank and hull plating in both
the sides and bottom. Compared to DS (and PL/SBT) the side protection is reduced,
as the width of the tanks may be less than in the double sides design because ballast can
be divided among the side and bottom spaces. From a cargo operations point of view
Figure 9.11. Double bottom (DB) configuration.
9.5 CBA OF OIL SPILL PREVENTION MEASURES FOR TANKERS 273
the design is excellent. Drawbacks are similar those of DS and DB. In addition, the
configuration/construction is more exposed to crack damages due to more plating.
Corrosion may also be a larger pr oblem. This puts high demands on access for inspections
and maintenance. The double bottom configuration is shown in Figure 9.13.
4. Hydrostatic driven passive vacuum (HDPV)
The hydrostatic driven passive vacuum (HDPV) construction makes openings to cargo
tanks airtight. This results in a progressive drop in pressure with cargo outflow, thereby
reducing pollution as the oil is ‘hold back’. Drawbacks are related to air tightening, and
instantaneous location of damaged tanks for closure of all openings (e.g. vent pipes).
Structural strengthening of the deck may be necessary in order to avoid structural
Figure 9.12. Double side (DS) configuration.
Figure 9.13. Double hull (DH) configuration.
274 CHAPTER 9 COST -BENE FIT ANALYS IS
collapse because of tank vacuum. The hydrostatic driven passive vacuum (HDPV) con-
figuration is illustrated in Figure 9.14.
5. Smal le r tanks (ST)
This design alternative is based on reducing the volume of the individual cargo tanks,
which will reduce the potential oil outflow. The main drawback is related to an increased
risk of plate cracking.
6. Interior oil-tight deck (IOTD)
An interior oil-tight deck (IOTD) greatly reduces the amount of cargo that is exposed to
damage to the bottom plating and lower sides. This gives both hydrostatic favouring and
reduced volume potential of oil spills if the hull integrity is broken. The need for ballast
tanks makes double sides (DS) necessary, resulting in extra protection of the sides. The
major drawbacks to this configuration are complex operation as well as corrosion damage.
The interior oil-tight deck (IOTD) configuration is illustr ated in Figure 9.15.
Figure 9.14. The hydrostatic driven passive vacuum (HDPV) configuration.
Figure 9.15. The interior oil-tight deck (IOTD) configuration.
9.5 CBA OF OIL SPILL PREVENTION MEASURES FOR TANKERS 275