Content uploaded by Apostolos Papanikolaou
Author content
All content in this area was uploaded by Apostolos Papanikolaou on Nov 23, 2014
Content may be subject to copyright.
Proceedings of the 2nd Int. Symposium on “Ship Operations, Management and Economics”, The Greek Scetion of the
Society of Naval Architects and Marine Engineers (SNAME), Athens, Sep. 17-18, 2008
1
Impact of ship age on tanker accidents
A. Papanikolaou1), E. Eliopoulou2)
1Ship Design Laboratory-NTUA, Greece, papa@deslab.ntua.gr
2) Ship Design Laboratory-NTUA, Greece, eli@deslab.ntua.gr
Abstract
There is a strong general believe that ship accidents
and particularly tanker accidents leading to marine
pollution are mainly related to ship’s age, namely the
older the ship the more likely her involvement in at least
non-accidental structural failure accidents.
The paper presents results of a comprehensive analysis
of recorded large tanker accidents of DWT greater than
60,000 that occurred in the period between 1990 and
2007. The analysis enables the identification of clear
trends with respect to the accident rates per shipyear
for all major accident categories. It also relates the
impact of ship’s hull design, namely of the double hull
and the various non-double hull configurations, as well
as the impact of ship’s age on non-accidental structural
failure accidents. Results of the latter analysis were
partly unexpected, showing that more than ship’s age,
the actual condition of ship’s structure in terms of main-
tenance and building quality are the decisive factors for
non accidental structural failure accidents.
Keywords
Tanker accident rates; marine pollution; tanker hull
design; impact of ship’s age on accidents
Nomenclature
DH Double hull ships
LOWI Loss of Watertight Integrity
NASF Non-Accidental Structural Failure
1. Introduction
The introduction of OPA90 and the more recent amend-
ments to the MARPOL convention by IMO, first man-
dating double hulls for new tankers and second acceler-
ating the phase-out of single hull tankers, were con-
ceived in order to minimize oil pollution caused by
tanker accidents. As a result of this and of other meas-
ures, a remarkable decrease of tanker accidents of all
types could be observed, Fig.1.
The main scope of the present paper is to identify possi-
ble relationships between the tanker hull design (Double
Hull and Non-Double hull configurations) and of ship’s
building year and age with marine accidental oil pollu-
tion. Calculations of accident rates were conducted on
the basis of historical data for the period 1990-2007 by
category of accidents, ship’s hull type configuration and
building year in order to identify underlying trends. In
particular, the paper focuses on the impact of tanker’s
hull type and age on the accident statistics, considering
the introduction of a variety of regulatory developments
since the early 90ties (OPA 90 and MARPOL amend-
ments), phase-out schemes of single-hull tankers, en-
hanced class survey programs and 2nd hand tanker ship
trading patterns.
Oil Tanker Casualties, Historical data
0.00E+00
2.00E-02
4.00E-02
6.00E-02
8.00E-02
1.00E-01
1.20E-01
1.40E-01
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
Frequen cy per shipyear
Collision Contact Grounding Fire Explosion NASF
Fig. 1: Frequency per shipyear
2. Approach
2.1 Field Data Model
Casualty databases are potentially important tools for
the evaluation of tanker safety and the environmental
performance of the shipping industry. The investigation
of historic accident scenarios could lead to the identifi-
cation of vulnerable operational or design problems and
also provide guidance to regulatory process to ship
safety and environmental protection.
There are many casualty databases available nowadays,
enabling the conduct of this type of analysis; the most
prominent and complete ones are believed to be the
Lloyd’s Register Fairplay (LRFP) and Lloyd’s Maritime
Intelligent Unit (LMIU). Although these databases con-
tain a significant amount of data (even though it is be-
lieved that a series of data related to near-misses and
underreporting are missing), they were originally not
designed for potential application to rational risk as-
sessment procedures, making their usage in engineering
studies, if not post-processed, quite problematic.
Proceedings of the 2nd Int. Symposium on “Ship Operations, Management and Economics”, The Greek Scetion of the
Society of Naval Architects and Marine Engineers (SNAME), Athens, Sep. 17-18, 2008
2
This post-processing may be achieved by a second
round of evaluation of these casualty data transforming
first the source information, i.e. textual information, into
a proper format to be directly applicable in risk assess-
ment studies.
Therefore, for setting-up proper tanker ship risk evalua-
tion models, the casualty information of the LRFP and
LMIU databases were imported into a purposely de-
signed database of NTUA-SDL (Papanikolaou et al.,
2005) enabling the further processing towards the quan-
tification of detailed event categories and the direct
extract of conditional probabilities of identified accident
scenarios in terms of frequency and consequence mod-
els.
In Fig. 2, the herein applied field data model for this
assessment is illustrated (Eliopoulou et al., 2008).
Fig. 2: Field Data Model
2.2 Sampling plan
The applied full evaluation model of ships’ casualties,
independently of ship type, shall include all possible
incident categories that may lead to an undesired event
(as a consequence of identified hazards), as indicated in
Fig. 3. The particular study focuses on the six (6) events
that potentially lead to ship’s Loss Of Watertight Integ-
rity (LOWI), namely collision, contact, grounding, fire,
explosion, non-accidental structural failure and to acci-
dental oil pollution.
Fig. 3: Generic Casualty Categorization
The time period of 1990-20071 was selected for the
present investigation, aiming at looking at the present
state of affairs and into the foreseeable future. This was
done for the following reasons.
1. The sample of ship accidents to be analyzed should
be representative for the present status and the fore-
1 Results of a similar study on AFRAMAX ship accidents for the
period 1978-2003, in which the share of double hull ships was limited,
may be found in (Papanikolaou et al., 2006).
seeable future.
2. In (Delautre et al., 2005) it was found that there
was a significant reduction of accidents occurrence
in the post-90 period (Fig. 1) due to the introduc-
tion of a series of regulations (with the most promi-
nent one being OPA 90) on the prevention and
mitigation of accidents (Fig. A1, Appendix)
3. In the post-1990, more and more double hull ships
were built and are operating, displacing single hull
ships (Fig. A2, Appendix).
With respect to the studied tanker subcategories and
sizes, large oil tankers - Oil Tankers, Crude Tankers,
Shuttle Tankers, Product Carriers and Chemical/Oil
Tankers- of DWT greater than 60,000 were selected for
the herein presented investigation. Considering the
above, in total 814 incidents large tanker accidents were
investigated, Table 1.
Table 1: Casualty data2, Covered period 1990-2007
Category No of incidents %
Collision 265 33%
Contact 93 11%
Grounding 192 24%
Fire 77 9%
Explosion 39 5%
NASF 148 18%
Total 814 100%
2.3 Fleet at risk
Relevant fleet at risk was based on LRFP data (Loer &
Hamann, 2007), see Appendix, Fig. A2.
3. Navigational events
3.1 Collision events
Collision events refer to incidents where two vessels
accidentally come into contact with each other. The
investigated scenarios contain collisions when the
tanker vessel is striking or being struck by another ship.
The collision probability is highly related to the traffic
density. According to the statistics, most collisions take
place within congested waters with dense ship traffic,
crossing routes and areas with large ship speed varia-
tions. The basic causes are bad visibility, navigational or
technical failures.
Concerning the degree of event’s severity as coded in
LRFP/LMIU databases, 25% of them are characterised
by serious degree of severity, from which 8% resulted to
ship’s total loss.
In 8 events out of 265 (Table 1), there were 2 non-
serious injuries and 55 fatalities (39 missing persons and
16 deaths).
In 11 cases out of the 265 recorded accidents (4%), fire
occurred during the accident. In all these cases, except
for one, the accident was of serious degree of severity.
2 Excluded incidents happened in Shipyards & Drydocks
Proceedings of the 2nd Int. Symposium on “Ship Operations, Management and Economics”, The Greek Scetion of the
Society of Naval Architects and Marine Engineers (SNAME), Athens, Sep. 17-18, 2008
3
Collision events, followed by fire/explosion, are events
with very severe consequences. According to the setup
database, in cases when collision was followed by
fire/explosion, the expected crew fatality rate is about
40% of ship’s crew number, when there is ship’s total
loss and 14%, when there is no total loss but severe ship
damage.
With respect to the oil spill occurrence, in 27 collision
events (10% of registered collisions) there was oil re-
lease to the sea, resulting to 126,532 tonnes of oil spilt
within the studied period.
3.2 Contact events
Contact events refer to scenarios where the vessel acci-
dentally comes into contact with a floating object or a
fixed installation. The likelihood of contacts is higher in
congested waters than in open sea operation. In fact, the
majority of the contact scenarios takes place during
maneuvering operations or approach/sailing in rivers
and canals. The basic causes for contacts are poor visi-
bility, navigational, technical or human failures.
In total, 93 accidents were registered as contacts, Table
1, from which 67% were contacts with a fixed installa-
tion and 33% with a floating object. Concerning the
degree of the event’s severity, as coded in LRFP/LMIU
databases, 25% were characterised by serious degree of
severity. No ship’s total loss was registered in the par-
ticular incident category. No injuries or fatalities were
recorded during the studied period in such events.
With respect to the oil spill occurrence, in 16 contact
events (17% of registered contacts) there was oil re-
lease to the sea, resulting to 13,162 tonnes of oil spilt
within the studied period.
3.3 Grounding events
Grounding events refer to scenarios where the vessel
accidentally comes into contact with the sea bed or
shore. Grounding is predominantly caused by naviga-
tional failure (powered grounding) or by propulsion,
power or steering failure (drift grounding).
The main causes for grounding accidents are related to
ground topology and environmental conditions, techni-
cal failures (steering or machinery failures) and human
factors.
In total, 192 accidents were registered as groundings, of
which 83% were powered groundings and 17% were
drift groundings.
Concerning the degree of event’s severity as coded in
LRFP/LMIU databases, 41% are characterised by seri-
ous degree of severity, from which 5% resulted to ship’s
total loss.
Concerning all grounding scenarios, in only one acci-
dent there was one reported fatality (missing person).
With respect to the oil spill occurrence, in 17 grounding
events (9% of registered groundings) there was oil re-
lease to the sea, resulting to 245,942 tonnes of oil spilt
within the studied period.
3.4 Frequency of navigational events
In Figs 4-6, the frequency of each navigational event is
presented for the two basic ship hull types (DH and non
DH). Tanker hull type appears not to be related to the
occurrence of navigational accidents. This conclusion
could be expected, because the main causes of such
events are related to human operational errors or/ and
technical failures. Thus, the frequency assessment of
navigational events could be considered as independent
of the hull type.
However, the consequences of navigational accidents
are expected to be different in terms of environmental
oil cargo release, because there is a certain probability
of inner hull non-breaching for DH ships, resulting to no
environmental pollution.
Collision eve nts
0.00E+00
5.00E-03
1.00E-02
1.50E-02
2.00E-02
2.50E-02
3.00E-02
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Incident Yea r
Frequency per shipyear
Non-DH ships DH ships
Fig. 4: Frequency of collision events
Contact events
0.00E+00
2.50E-03
5.00E-03
7.50E-03
1.00E-02
1.25E-02
1.50E-02
1.75E-02
2.00E-02
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Incident Year
Frequency per shipyear
Non-DH ships DH ships
Fig. 5: Frequency of contact events
Grounding events
0.00E+00
5.00E-03
1.00E-02
1.50E-02
2.00E-02
2.50E-02
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Incident Year
Frequency per shipyear
Non-DH ships DH-ships
Fig. 6: Frequency of grounding events
3.5 Navigational events & ship’s age
The frequency of navigational events increases when
ships turn 15 years and older, whereas high frequencies
are observed for the collisions and groundings of young
ships (0-5 years), Fig. 7. Although there is a small res-
ervation on the estimation of the employed fleet at risk
in calculating the event frequencies, slightly underesti-
mating the number of newbuildings entering the fleet at
risk at the year of census, the observed high frequencies
Proceedings of the 2nd Int. Symposium on “Ship Operations, Management and Economics”, The Greek Scetion of the
Society of Naval Architects and Marine Engineers (SNAME), Athens, Sep. 17-18, 2008
4
for the navigational events of young ships may be re-
lated to crew’s proper training, communication prob-
lems and to crew’s ability to handle new technology
equipment.
Navigational Events, Covered Period 1990-2007
Oi l Tanker s >60, 000 DWT, F requency per shipyea r
0.00E+00
5.00E-03
1.00E-02
1.50E-02
2.00E-02
2.50E-02
0-5 years 6-10 years 11-15 years 16-20 years > 20 years
Collisions Contacts Groundings
Fig. 7: Frequency of navigational events vs. Ship’s age
4. Fire and Explosion events
4.1 Fire events
Fire events refer to scenarios where the fire is the first,
initiative event. Fire can start due to internal sources,
external sources (unlawful acts, spread of fire from
other ship) or due to atmospheric conditions (by light-
ing).
Fire due to internal source can be initiated in ship’s aft
area, on deck, in Cargo/Slop tanks area, in Ballast/Void
spaces or in the Fore Peak area.
Fire in the aft area can occur in the accommodation area
(above main deck) because of electrical faults, heating
equipment failure, smoking etc., or below the main deck
in Engine Room or in the Pump Room.
The accident evolution depends on the timing
(early/late) of detection and the rate of fire spreading.
In total, 77 accidents were registered as fire, 96% were
fires due to internal source, 1% by external source and
3% by lightning.
Fire due to internal source started in Aft Area (89%), in
cargo/slop tanks (5%), in ballast tanks (3%) and on deck
(3%).
Concerning the degree of event’s severity, as coded in
LRFP/LMIU databases, 31% were characterised by
serious degree of severity, from which 27% resulted to
ship’s total loss.
In 10 accidents, there were 22 injuries (21 serious & 1
non serious) and 12 fatalities (5 missing and 7 deaths).
With respect to the oil spill occurrence, in 2 fire events
(3% of registered fires) there was oil release resulting to
161,000 tonnes of oil spilt within the studied period.
4.2 Explosion events
Explosion events refer to scenarios where the explosion
is the first, initiative event. In total, 39 explosions oc-
curred in the operational phase of the studied tanker
ships during the period 1990-2007.
Explosion started in Aft Area (47%), in cargo/slop tanks
(47%), and on deck (6%).
Concerning the degree of event’s severity as coded in
LRFP/LMIU databases, 69% were characterised by
serious degree of severity, from which 28% resulted to
ship’s total loss.
In 17 accidents out of 39, Table 1, there were 65 fatali-
ties.
With respect to the oil spill occurrence, in 3 explosion
events (8% of registered explosions) there was oil re-
lease to the sea resulting to 278,770 tonnes of oil spilt
within the studied period.
4.3 Frequency of fire, and explosion events
In Figs 8-9, the frequency of fire and explosion event is
presented for each basic hull type. Tanker hull type
seems to be unrelated to the occurrence of fire and ex-
plosion accidents. Thus, frequency assessment could be
considered as independent of hull type.
Fire events
0.00E+00
2.50E-03
5.00E-03
7.50E-03
1.00E-02
1.25E-02
1.50E-02
1.75E-02
2.00E-02
1990 1991 19921993 1994 1995 1996 1997 1998 1999 20002001 2002 2003 2004 2005 2006 2007
Incide nt Yea r
Freque ncy per shipyear
Non-DH DH ships
Fig. 8: Frequency of fire events
Explosion events
0.00E+00
2.00E-03
4.00E-03
6.00E-03
8.00E-03
1.00E-02
1.20E-02
1.40E-02
1.60E-02
1.80E-02
1990 1991 19921993 1994 1995 1996 1997 1998 1999 20002001 2002 2003 2004 2005 2006 2007
Incide nt Yea r
Freque ncy per shipyear
Non-DH DH ships
Fig. 9: Frequency of explosion events
4.4 Fire, Explosion events & ship’s age
Figs 10-11 present the frequency of fire, and explosion
occurrence by ship hull type and age. DH ships are a
relatively new ships thus there are no representatives in
the larger age groups. Independently of the hull type,
higher frequencies are observed for higher ship ages.
Fires, Covered Period 1990-2007
Oil Tankers >60,000 DWT
0.00E+00
2.00E-03
4.00E-03
6.00E-03
8.00E-03
1.00E-02
1.20E-02
0-5 years 6-10 years 11-15 years 16-20 years > 20 years
Frequency per shipyear
All ships DH sh ips
Fig. 10: Frequency of fire events by group age
Proceedings of the 2nd Int. Symposium on “Ship Operations, Management and Economics”, The Greek Scetion of the
Society of Naval Architects and Marine Engineers (SNAME), Athens, Sep. 17-18, 2008
5
Regarding the high frequencies observed for the fire and
explosions of young ships (0-5 years), there is a small
reservation on the estimation of the employed fleet at
risk, not properly capturing the number of newbuildings
entering the fleet at risk at the year of census.
Explos ions, Cov ered Per iod 199 0-200 7
Oil Tankers >60,000 DWT
0.00E+00
5.00E-04
1.00E-03
1.50E-03
2.00E-03
2.50E-03
3.00E-03
3.50E-03
4.00E-03
4.50E-03
0-5 years 6-10 years 11-15 years 16-20 years > 20 years
All ships DH shi ps
Fig. 11: Frequency of explosion events by group age
5. Non-accidental structural failures, NASF
Non-accidental structural failure events refer to scenar-
ios where the hull presents cracks and fractures, affect-
ing ship’s structural integrity and seaworthiness. Dam-
age to a vessel rudder, or rudder-adjoining parts are
herein also counted as structural damage. However,
damages related to ship’s hull-fitting equipments, like
vessel propeller, propeller portion or propeller adjoining
parts are not included in the particular categorisation.
Non-accidental structural failures may potentially lead
to the Loss Of Watertight Integrity (LOWI); they occur
because of structural degradation, overstressing due to
excessive loading or due to poor design and/or construc-
tion.
Such accidents mainly happen while the tanker ship is
in en-route in Open Sea/Archipelagos, or during loading
or discharging operations.
5.1 NASF recordings independent of ship’s hull type
Non-accidental structural failure events present 18% of
all registered initial causes in the setup database involv-
ing large tanker ships for the studied period 1990-2007.
In total, 148 accidents were registered as non-accidental
structural failures, Table 1.
The degree of event’s severity as coded in LRFP/LMIU
databases is presented in Fig. 12.
NASF, Historical data, Covered period 1990-2007
Degree of severity, 148 incidents
6
46
4
92
0
20
40
60
80
100
Total loss Serious Not serious Unknown
No of accidents
Fig. 12: NASF, accidents’ severity
Focusing on the 6 cases where there was ship’s total
loss, the following was observed:
• All accidents happened while the ship was en route
in Open Sea, in heavy weather conditions.
• In all cases the involved tanker ships were of Non-
DH construction.
• In 5 cases, external hull damage was the starting
point of NASF. The ship was loaded in 4 cases out
of 5, causing significant oil spill in the majority of
cases.
• In the remaining one case, deck damage was the
starting point of NASF.
With respect to the oil spill occurrence, in 38 non-
accidental structural failures (26% of registered NASF)
there was oil release to the sea resulting to 170,538
tonnes of oil spilt within the studied period.
5.2 NASF recordings of Double Hull ships
Focusing on Double Hull ships, 20 non-accidental struc-
tural failures were registered in the period 1990-2007.
Concerning the degree of event’s severity as coded in
LRFP/LMIU database, in 11 cases the event was char-
acterised by serious degree of severity and in 9 cases
there were non serious degree of severity.
According to the recorded data, the weather condition
was an important factor in 6 cases out of 20.
Practically, no oil spill occurred due to NASF for DH
ships in the study period (in 1 case there was a minor
release of oil, around 1 tonne).
5.3 Weather relation
Based on the NASF registered data for all ships, regard-
less the hull type, many incidents happened in heavy
weather conditions (about 35%, 52 cases out of 148
incidents). This does not mean that the weather condi-
tion is the main cause of non-accidental structural fail-
ures. Large tanker hulls are typically designed to handle
a wide range of weather conditions. In case of poor hull
structural design or corrosion due to poor maintenance,
however, the structure becomes weak to handle heavy
weather conditions and this naturally leads to non-
accidental structural failures.
Focusing on Double Hull ships, the relation of NASF to
weather conditions was found to be almost the same as
for the overall fleet (about 30%, 6 cases out of 20 inci-
dents). This could be actually more related to structural
design problems because the particular tanker fleet has a
relatively small age up to date and maintenance prob-
lems should come second. Looking into detailed data,
according to the setup database, in all weather related
accidents, the DH ships involved were at the group age
of less than 5 years.
Table 2 presents the age distribution of Double Hull
ships involved to NASFs during the studied period
1990-2007.
Proceedings of the 2nd Int. Symposium on “Ship Operations, Management and Economics”, The Greek Scetion of the
Society of Naval Architects and Marine Engineers (SNAME), Athens, Sep. 17-18, 2008
6
Table 2: Non-accidental structural failures, DH ships,
covered period 1990-2007
Group age Number of ships
up to 5 years 15
6-10 years 1
11-15 years 2
>16 years 2
5.4 NASF damage initiation
With respect to Non-DH ships, the majority of NASFs
started at the external hull (59%), Fig. 13. In contrary,
DH ships found to present higher structural problems in
the internal structural hull (40%).
NASF events, Structural damage initiation
Covered Period 1990-2007
0
10
20
30
40
50
60
70
Deck damage Internal damage Hull damage Rudder damage Unknown
Percentage of events
Non-DH ships DH ships
Fig. 13: NASF, damage initiation
5.5 NASF Frequency per shipyear
Fig. 14 presents the NASF frequency per shipyear for
each basic hull type.
Non Accidental Structural failures
0.00E+00
5.00E-03
1.00E-02
1.50E-02
2.00E-02
2.50E-02
3.00E-02
3.50E-02
4.00E-02
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Incident Year
Frequency per shipyear
Non-DH DH
Fig. 14: Frequency of NASF per shipyear
5.6 NASF and ship’s age
Taking into account all tanker ships involved in NASFs,
the frequency of the particular accident category gener-
ally increases, when ship’s age increases after the 6-10
years of built, Fig. 15 (line). There are, however, two
very interesting and unexpected observations to note.
1. Young ships of age zero to five years (0-5) show a
remarkable structural failure rate for all ship cate-
gories, except for VLCC/ULCC.
2. For two tanker categories, namely AFRAMAX and
SUEZMAX, with a significant share in the overall
fleet at risk, there is a peak at the age group 11-15
years and then the frequency decreases, particularly
for the SUEZMAX, whereas for the AFRAMAX
the frequency increases sharply only for ages over
20 years (Fig. 16).
NASFs, Covered Period 1990-2007
Oil Tankers >60,000 DWT
0.00E+00
2.00E-03
4.00E-03
6.00E-03
8.00E-03
1.00E-02
1.20E-02
1.40E-02
1.60E-02
1.80E-02
2.00E-02
0-5 years 6 -10 years 11- 15 years 16-20 years > 20 year s
PANAMAX AFRAMAX SUEZMAX V&ULCC All ships
Fig. 15: NASF Frequency by tanker size & age
NASFs, Covered Period 1990-2007
Oil Tankers >60,000 DWT
0.00E+00
2.00E-03
4.00E-03
6.00E-03
8.00E-03
1.00E-02
1.20E-02
6-10 years 11-15 years 16-20 years > 20 years
All ships AFRAMAX SUEZMAX
Fig. 16: NASF Frequency of AFRAMAX & SUEZMAX
by group age
Regarding these two notable observations, the following
reasoning is offered for discussion.
1. Failures of young aged ships (0-5 years): this is
alarming for the quality of recently delivered DH
ships3. It is hoped that the ongoing discussion about
Goal Based Standards (IMO MSC76/5/10) and the
recently introduced Common Structural Rules by
IACS (IACS, 2006) effectively address this. The re-
lated NASF status for the AFRAMAX and SUEZ-
MAX DH ships in age group of 0-5 years, is shown
in Fig.17.
NASFs, Covered Period 1990-2007
Double Hull, Oil Tankers
4.64E-03
0.00E+00
1.77E-02
4.96E-04
0.00E+00
2.50E-03
5.00E-03
7.50E-03
1.00E-02
1.25E-02
1.50E-02
1.75E-02
2.00E-02
0-5 years 6-10 years
Frequency
AFRAMAX, DH SUEZMAX, DH
Fig. 17: Frequency of NASF per shipyear
2. Peak NASF frequency (age group 11-15 years):
This might be attributed to a decreased maintenance
effort on ship’s hull structure when approaching her
assumed design economic life of about 20 years
and before the ship changes ownership for her 2nd
economic life by another operator. Note that this
observation is clearer for single hull ships, as
3 Some reservation might be due here in view of possible slight under-
estimation of recorded newbuildings in the relevant fleet at risk
Proceedings of the 2nd Int. Symposium on “Ship Operations, Management and Economics”, The Greek Scetion of the
Society of Naval Architects and Marine Engineers (SNAME), Athens, Sep. 17-18, 2008
7
AFRAMAX Tankers, Accident Rates by AGE
0.00E+00
5.00E-03
1.00E-02
1.50E-02
2.00E-02
2.50E-02
3.00E-02
0-5 years 6-10 years 11-15 years 16-20 years > 20 years
Collis ion Contact Grounding Structural Failure Fire Explosion
shown in Fig. 18 (Papanikolaou et al., 2006)
Fig. 18: Frequency of NASFs per shipyear by accident
category and ship age - AFRAMAX ships
6. Environmental pollution
Based on the investigated sampling plan and the setup
database, a total amount of about 1,000,000 tonnes of
oil spill was released to the sea environment due to large
tanker accidents in the period 1990-2007, Fig. 19.
Historical data, Covered period 1990-2007
Amount quantity of oil spilt
13162
245942
161000 170538
278770
126532
0
50000
100000
150000
200000
250000
300000
Collision Contact Grounding Fire Explosion NASF
Tonnes
Total amount = 995,944 tonnes
Fig. 19: Oil spill release to the sea environment
Practically, this environmental pollution is entirely
attributed to accidents of single hull ships. The envi-
ronmental pollution due to DH ships’ accidents is less
than 2% of the total oil quantity released to the sea,
Table 3.
Table 3: Environmental pollution by DH ships
Covered period 1990-2007
Incident type Oil release to the sea, in tonnes
Collision 215
Contact 292
Grounding 24
Fire 17000
Explosion 0
NASF 1
17532
Therefore, results of the present study indicate that oil
pollution rates of large tankers with Double Hull con-
figuration are nearly zero and orders of magnitude
smaller than those of non-double hull tankers.
However, before drawing reliable conclusions on the
actual impact of tanker hull configuration on pollution,
the following should be considered.
Pollution rates and relevant statistics change drastically
following individual catastrophic events and such events
(fortunately) did not happen yet with large DH oil tank-
ers. It is noted, that one catastrophic accident with a DH
ship happened in the study period, namely with the DH
Ore/Bulk/Ore AFRAMAX “Aegean Sea” that took
place in 1992 in Spanish waters causing 74,000 tonnes
of oil spill. The “Ore/Bulk/Ore” tanker subcategory
was, however, not included in the present analysis by
definition. Despite and even including this accident in
the DH assessment, pollution rates of DH ships are still
significantly better than those of non DH ship hull
types.
7. Conclusions
The undertaken study leads us to the following conclu-
sions:
1. A significant decrease of the frequency of large
tanker accidents is noted in the post 1990 period.
This is related to the introduction of a series of
regulatory measures, changes in ship design and
technology and overall improvement of the safety
culture of the engaged maritime industry.
2. Double hull ships proved very effective in limiting
marine environment accidental pollution, Fig. 19
and Table 3.
3. It was found that non-accidental structural failure
accidents (NASF) are generally related to ship’s
maintenance and age, however with respect to age
in a non-straightforward way. Alarmingly, some
young double hull ships show unexpectedly NASF
problems, calling for enhanced measures by build-
ers and class societies to improve this unsatisfac-
tory state of affairs.
Acknowledgements
A part of the work presented herein was financially
supported by the European Commission under the FP6
Sustainable Surface Transport Programme. This support
was given under the scheme of STREP, POP&C project,
Contract No. TST3-CT-2004-506193 and under the
Integrated Project SAFEDOR, Contract No FP6-
516278.
The European Community and the authors shall not in
any way be liable or responsible for the use of any
knowledge, information or data of the present paper, or
of the consequences thereof. The views expressed in
this paper those of the authors and do not necessary
reflect the views and policies of the European Commu-
nity.
The authors like to thank their SAFEDOR project part-
ner Germanischer Lloyd AG, especially Drs R. Hamann
and K. Loer for their support with data in the presented
work.
References
Delautre, S, Eliopoulou, E, and Mikelis, N, (2005). "The
Influence of Regulations on the Safety Record of the
Aframax Tankers", Study carried out within the EU
Proceedings of the 2nd Int. Symposium on “Ship Operations, Management and Economics”, The Greek Scetion of the
Society of Naval Architects and Marine Engineers (SNAME), Athens, Sep. 17-18, 2008
8
funded project POP&C, Contract No TST3-CT-
2004-506193.
Eliopoulou, E, Papanikolaou, A, and Hamann R, (2008).
"Risk Analysis of Large Tankers", 2nd Int. Workshop
on Risk-Based Approaches in the Maritime Industry,
Glasgow, 5-6 May 2008.
IACS, 2006. "The Common Structural Rules for Tank-
ers and Bulk Carriers", adopted by the Council of
the International Association of Classification Socie-
ties, for implementation on 1st April 2006.
Loer, K, and Hamann, R, (2007). "HazId of Tanker
Operation", EU funded project SAFEDOR, Deliver-
able D4.7.1.
Papanikolaou, A, Eliopoulou, E, Alissafaki, A, Aksu, S,
Delautre, S, and Mikelis, N, (2005). "Critical Re-
view of AFRAMAX Tankers Incidents", 3rd Int.
Conference ENSUS, Newcastle upon Tyne, 13-15
April 2005.
Papanikolaou, A, Eliopoulou, E, and Mikelis, N, 2006.
"Impact of hull design on tanker pollution", Proc. 9th
Int. Marine Design Conference IMDC, Ann Arbor,
Michigan, May 16-19.
POP&C, “Pollution Prevention and Control.” EU pro-
ject, 6th Framework Programme, Contract No
TST3-CT-2004-506193, 2004-2007.
SAFEDOR (2005-2009), “Design, Operation and Regu-
lation for Safety”, EU project, FP6-516278.
Appendix
Fig.A1: Impact on regulation on the prevention of navigational events
Oil Tankers of DWT>60,000
Hull type distribution of Fleet at risk
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
DH Non-DH
Fig. A2: Distribution of Fleet at Risk of Oil Tankers by hull basic configuration.
Aframax Tankers: Navigational Incident Rates per shipyear
0.00E+00
1.00E-02
2.00E-02
3.00E-02
4.00E-02
5.00E-02
6.00E-02
7.00E-02
78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03
Collision Contact Grounding
81 SOL AS
ARPA
88 SOLAS
GMDSS
88 SOLAS
GMDSS
96 ILO C180
95 SOLAS
Routeing Systems
72 COLREG
74 SOLAS
Nav. Equipm.
72 COLREG
72 COLREG
81 SOLAS
Nav. Aids 81 SOLAS
Duplication Ste ering gea r
Red border: applies to COLLISION incidents
Blue border: applies to CONTACT incidents
Black border: applies to GROUNDING incidents
Green border: applies to all 3 categories
Gra
y
shadin
g
: a
pp
lies to newbuildin
g
s
78 PARIS MOU
T
OKYO MOU
78 STCW
88 SOLAS
Nav. Aids
94 SOLAS
ISM
95 STCW
OPA 90
VETTING
96 SOLAS
ETS
