HELSINKI UNIVERSITY OF TECHNOLOGY
SHIP HYDRODYNAMICS LABORATORY
OTANIEMI
FINLAND
Report No. 7
ANALYSIS OF SHIP CASUALTIES IN THE
BALTIC, GULF OF FINLAND AND
GULF OF BOTNNJA IN
197-972
by
Valter Kostilainen and Maija Hyvärinen
BALTIC AREA SHIP CASUALTIES 1971 - 1972
GROUNDING
£ COLLISICN WITH PIER, BRIDGE, DOLPHIN I COLLISION WITH OTHER SHIP
COLLISION WITH TWO OR MORE SHIPS EXPLOSION OR FIRE X OTHER SWEDEN u s X
4,
Ob VN IA X OARS K p I UME1,)-JJI
PORI FINLAND p -S.t
RRIGA/
HELSINKI TECHNICAL UNIVERSITY SHIP HYDRODYNAMICS LABORATORY FINLANDF i g. 1. Geographical distribution of ship casualties in the Baltic area. 1971-72. 001 I4ELSINRI ç4__S.i21' U S.S R o sa O 0 20 100 150 KM POL ANO
OF FINLAND AND GULF OF BOTHNIA IN 1971-1972.
by Valter Kostilainen and Maija Hyvärinen
Abstract
The problems associated with marine traffic engineering in the Baltic area are difficult to solve, and casualty
statistics are extremely important as a basis for safety and antipollution activities. A statistical analysis has been made
of the data relating to a total of 283 ship casualties in 197 1-72, and the results obtained are presented. A discussion
is included of local and seasonal distribution, and the distribution of the casualties by other time-dependent variables.
An account is given of conditions of weather and visibility, and the use of navigation equipment at the time concern-ed. The statistics further cover the type of ship involved, its nationality, loading condition, age, speed, and the
damage arising. The primary causes of the casualties are discussed. More efficient international co-operation between
the countries of the Baltic area in the compilation of casualty data, and in other fields of marine traffic engineering
research, is considered necessary.
Introduction
During recent years, the marine traffic in the Baltic
area has changed markedly. At one time, the most serious
consequence of stranding or collision, apart from loss of
life and injury, was likely to be some spillage of oil, usually
bunker fuel or refined products. Today, however, the
substance spilt could be crude oil, phosphorus, acids or a variety of other chemicals which are conveyed in large quantities by sea. The Baltic Basin is a highly sensitive
ecological area, and a mishap at sea can now inflict damage
far and wide, affecting parties other than the crew and owner of the ship directly involved. Although the recent increase in shipping may not continue at its present rate,
the modern trend towards larger vessels and higher speeds has increased the risk of widespread pollution in the event of any major shipping casualty.
Recent events have shown that the safety of navigation is no longer attainable by the traditional means of naval architecture and seamanship. In the regions of higher at-sea population of ships, the marine transport system
has to be studied as a whole. This has given rise to a new branch of engineering: marine traffic engineering, so new
a subject that many who are professionally concerned with the design and operation of ships appear to be una-ware of its existence. For marine traffic engineering,
casualty statistics and traffic surveys are of extreme
importance, and provide the fundamental material for fairway and harbour design, traffic control planning, and first and foremost that for the planning of
countermeas-ures to casualties.
Marine traffic engineering research of the Baltic area was begun in the Helsinki University of Technology in 1970 by the compilation and analysis of casualties to tankers in 1960-69 [1]. Work extending over one year resulted in no more than 102 of the 268 casualties being satisfactorily treated. If casualty data are collected
sub-sequently at long intervals, the lapse of time means that a great deal of irreplaceable information is continuously lost.
Accordingly, it was decided that from the beginning of 1971 the systematic and continuous compilation of data
in regard to all maritime accidents in the Baltic area would
be undertaken immediately after the occurrence. To en-sure adequate sample size, biennial analysis of the sta-tistics was considered suitable. Stasta-tistics do not cover
pleasure crafts and vessels under 20 rn in length.
Data Collection
For the main part, the first information on accidents was obtained from Lloyd's Weekly Casualty Reports. Furthermore, the Boards of Navigation in Finland and
Denmark issued reports on some casualties. It is significant
that in approximately ten cases the initial information
concerning accidents in Finnish waters was obtained from
newspapers. To complement this information, Ship Cas-ualty Cards, as illustrated in Appendix I, were first sent
to the captains, and if no answers were received the
questionnaires were addressed to the shipowners. Table i indicates the percentages of returned questionnaires.
In view of the relative number of ships from the
U.S.S.R. and G.D.R. that navigate in the Baltic area, the
numbers of casualties to sustained by Soviet and German vessels seem low. This is probably attributable to the in-itial source of information being Lloyd's Weekly Casualty Reports. For practical reasons, these reports can not cover
all the casualties to ships from the U.S.S.R. and G.D.R.
In the few cases in which the initial information was obtained from Lloyd's Weekly Casualty Reports, no ad-ditional information was obtained from the ship or from the shipowner. This one-sided reduction of the sample
has an adverse effect upon the reliability of regional
statistics of this kind.
The answers given to the questions varied significantly:
Table 1. Flag distribution of questionnaires sent, and
percentage of response.
the accident, the year of building, and the dimensions of the ship, were in almost every case, whereas information
on such factors as visibility, the primary cause of the
casualty, and the dimensions of the damage, was
consid-erably poorer.
The damage information was punched on cards for
da-ta processing. The sda-tatistical program system HYLPS, in
the computer UNIVAC 1108, was applied for the analysis. All the information is stored on magnetic tape, and forms
a permanent data bank which will be augmented as time
passes.
Types of Casualties and their Local and Time-Dependent Distributions.
Fig. 1 indicates the geographical locations of the cas-ualties. The relatively low casualty density off the east
coast of the Baltic and the south coast of the Gulf of
Finland is probably attributable to a lack of casualty
infor-mation from the Soviet Union. The casualty density on the east side of the Gulf of Bothnia is higher than that on
the west side. Nevertheless, traffic flows to the east coast
ports of the Gulf are lower, as is observable in Fig. 2. This leads to the conclusion that the casualty frequency
is higher on the east coast. The reasons for this high cas-ualty frequency will be discussed later.
Nearly one half of the casualties, 45.9 %, arose from
grounding, 35.3 % collision (ship-to-ship casualty), 9.6 % ramming (ship-to-object casualty), 4.6 % foundering and
capsizing, and 3.5 % explosions and fires. (Fig. 3). On
comparison with the situation in N.W. European waters, a
TOTAL TRANSPORTES 50005 TO AGI) FROM SWEDEN iN 971 51 .7'. MIL.). TON
F i g. 2. Flows of goods traffic to and from Finland and Sweden in 1971, according to [4, 91.
TOTAL TXANSASRTEO GOODS TO ANO FROM FINLAND IN 1971 31,76 MIL.). TON EXPLOSIONS ANO/TA FIRES FOUNDER NOS, CAPXIZINGS AND FLOTO I NOS MEATY WEAThER ANO ICE DAMAGES
(3) 1,1 5
F i g. 3. Type, number and percentage of the 283
casualties.
heavy preponderance of grounding is apparent in the
Baltic area. According to Grimes [51, the percentage of
cases of stranding in N.W. European waters is only 9.8.
The local distributions of the different types of casualties
are presented in Fig. 4. The number of cases of grounding is comparatively large in the Sound, Kalmar Sound,
Stock-holm approach, the Gulf of Both nia and in the
surround-ings of Aland. Collisions are most frequent in the Fehmarn
Belt, Kieler Bucht and in other parts of the Baltic. Fig. 5
illustrates that most of the casualties occurred in
congest-ed waters such as sounds, passages, approaches to the port, and port areas. The number of cases of grounding is more than twice the number of collisions in sounds and passages. On the other hand, within the port area the number of collisions is twice that of rammings, which is
in turn twice the number of groundings. Flag
Total number of Percentages of
questionnaires sent questionnaires returned FRG 65 54 Finnish 40 75 Danish 32 78 Swedish 30 53 Polish 23 78 Norwegian 19 74 Dutch 14 64 British 13 69 Greek 10 60 U.S.S.R. 7 O Liberian 7 71 Cypriot 6 33 GDR 3 O Panamanian 3 67 Icelandic 2 50 Italian 2 50 Spanish 2 o French 100 Brazilian 100 Pakistani 100 Egyptian o Belgian o
60 BD 'R 20 SOUND M 100 IGREAT BELT D COLLISIONS RAMMING S 'I'll IFE1NYARN MELI ANO K I ELER B UC M T IKALMAR SOC ND p----/ H-- S
/
A\
/,)\
-s/
"--//\/
''R..._.1010ER PARTS! IOCKHOLM ULANO WITT GULF OF GOLF OF 0F TOE APPROACM SURROUNDINGS MOTORIA FINLAND
SALT IC
F i g. 4. Local distribution of casualties.
EXPLOSIONS ANO/OB FIRES
fl
POUNOERINGS. CAPSIZINGS. FL000INGSS.
15
G RO UN D ING S
COLLISIONS
RAMMING S
EXPLOSIONS AND/OR FIRES FOUSOERINSS, CAPSIZINGS, FL000INDS I'll'
PRESENT ANALYSIS, TEAR 1011
f'\ PRESENT ANALYSIS, SOAR 1972
/ \ - PRESENT ANALYSIS, TEARS 1971-72
I \ TANKER CASUALTIES IS 960-69 II]
/ \\ 7' II \ ¡I ,/ I I' ¡S
\/
I
''.l' J / -,/_/
F i g. 7. Monthly casualty rates, as percentage of the maximum rate, with some marine transport figures.
6 7 6 9 IO ID MONIlI Is Io OPEN SEA OP S COAST 0USD, PASSAGE AP ROACH IT THE PORI PORT AREA
F i g. 5 Distribution of casualties in different types
of sailing areas. F i g. 6 Seasonal distribution of casualties.
MTNTMLY CASUALTY RATES IN THE BALTIC AREA CODOS LOADER ANA
UNL00000 IN FINNISH ASO SW001SM PORTS FROM [3] ANO (N) TRANSIT COTOS TRAFIC IN KIEL CANAL FROM [21 10 2 3 R 5 6 7 1971 56 T S 772 10
lOO 90 80 70 60 50 40 SO 20
F i g. 9. Distribution of 145 casualties by watches.
The seasonal distribution of the casualties is presented in Fig. 6, together with the seasonal distribution of tanker
accidents in 1960-69 [1]. The main features of the dis-tributions are similar, with the maximum frequencies at
the beginning and at the end of the year, and with the min-imum falling in the summer. This variation in casualty
fre-quency does not correspond to the variation in transport
flows to Finland and Sweden, as can be seen from Fig. 7.
To the contrary, the maximum of these transport flows falls in thc summer. The seasonal variation in transport
flows through the Kiel Canal is small.
Fig. 8 illustrates the seasonal relation between different
types of casualties. This indicates that most of the cases of foundering and capsizing occurred in stormy autumn.
Fig. 9 shows the distribution of casualties by watches. The
similarity with the results obtained by Wheatley [5] is obvious. The casualty frequency is low during the after-noon and first watches, as compared with other watches. The distribution of casualties by weekdays is presented in Fig. 10; one figure here is striking, that indicating a
high collision frequency at the end of the week, and
particularly on Friday.
Flag, Type and Age of Ships Involved
Fig. 11 contains a breakdown of the flags of the ships involved in accidents. Here again, it needs to be pointed out that information on casualties was not obtainedfrom
IO
-lo
20
RURISIINOS FIRES, EXPLOSIONS FOUNDER I NUS ,CUPSI Z SUS FLOSS 1505
SII IO TU IRE 1H FR SU
2 3 14 5 6 7 IO
PION TSP
F i g. 8. Seasonal relation between different types of casualties.
SU TU WE
F i g. 10. Distribution of casualties by days.
the Soviet Union or from GDR. Without detailed
infor-mation on traffic flows in the Baltic area, no general
con-clusions can be drawn from Fig. 11. In any event, the
relatively large number of W. German and Finnish ships
involved in accidents is noteworthy. The large number of
W. German ships arises from the fact; nearly one third of
vessels passing through the Kiel Canal are W. German, and Kieler Bucht being a very congested area in which casualty
figures are high.
The large number of Finnish ships is of more serious
nature, especially on comparison with the number of
Swedish vessels involved in accidents. According to Fig. 2,
the annual traffic flow to and from Finnish ports through
the Baltic is about 30 million tons, and the corresponding figure in regard to Sweden is about 50 million tons.
Conse-SU HO 111 WE 1H FR SU ALL CASUALTIES S 282 OU GROUNDING S N 129 202 6-20 -8 S-12 12-IL
30
20
F i g. 11. Nationality of ships involved in accidents.
-I-I- FRS OASI 5H
F INNI SM SUE D T SM
POLISH
ALL SLITPS INVOLVED IN ACCIDENTS
M MEDIAN
SIT IF AGE
F i g. 12. Age distribution of merchant fleets of five nationalities with the highest
casualty figures, taken against age
dis-tribution of all ships involved in
acci-dents.
M 5,0 YEARS
M6,6
IT I2,A
Finland, so that the difference in casualty frequency may even be higher. This difference is probably attributable to the following two reasons:
Differences in the age distribution of the merchant
fleets.
Differences in the fairways, and other factors affecting offshore navigation.
Figs. 12 and 13 present the age distributions of the ships involved in accidents, and the merchant fleets of five nationalities with the highest casualty figures. Although the age distribution of the entire merchant fleet is not
necessarily the same as the age distribution of ships using
the Baltic area, it provides a good approximation of the situation. Figures 12 and 13 illustrate that the- age
distri-butions of ships involved in casualties clearly differ from
the age distributions of the corresponding fleets. The
mean age of ships involved in accidents is higher than the mean age of the fleet; the difference in values of the medians is 5-7 years. Fig. 14 contains these distributions
presented separately for FRG, Denmark, Finland and
Sweden, in each case the age of ship involved in accidents is higher. Moreover, Fig. 12 shows that the Finnish fleet is
older than the fleets of the four other countries presented.
If the casualty frequency is higher for older ships, and the Finnish merchant fleet is relatively old, then this must be
one reason for the relatively high casualty frequency of Finnish ships. Other possible reasons for the high fre-quency rate will not be discussed here, for lack of the
necessary basic information.
The age distribution of ships involved in grounding and collision is presented in Fig. 15. No systematic dif-ferences are observable and small deviations probably
arise from the small size of sample.
Dry cargo ships dominate in casualty Statistics, as is
evident in Fig. 16. This figure indicates one positive tend-ency: the casualty figures of tankers are obviously decrea-sing. A total of 29 tanker accidents occurred, 14 in 1971,
and 15 in 1972. The mean annual number of tanker cas-ualties in 1960-69 was 26.8 [11. This diminution in the casualty figures must be the consequence of the general consciousness of the potential dangers inherent in tanker
casualties in the Baltic Basin. Greater care has been taken
in the navigating, piloting and handling of tankers. In
general, new Baltic tankers have been equipped with
controllable pitch propellers, bowthrusters, and modern
navigational aids.
Fig. 17 presents the local distribution of casualties, taken against the types of ships involved in casualties.
Casualties to passenger ships and ferries have taken place
in the Sound, Stockholm approach and in the
surround-ings of Aland, where ferry- and passenger traffic has
in-creased considerably in recent years. Fig. 18 illustrates that one half of the casualties to passenger ships and ferries have taken the form of grounding, and that the
other half is evenly distributed between collision and
ramming. A risk of loss of many lives is always present in
O_L, 5-9 10-IA 1519 0-lA 21-29 30 AND OVER
SM IP ACE
F i g. 13. Age distributions of ships of five most com-mon nationalities, and involved in accidents.
quently, the relative casualty frequency of Finnish ships is roughly twice that of Swedish ships. lt is probable that
Swedish ships transport relatively larger amount of goods
from and to Sweden than do Finnish ships from and to
50 2O 210 30 20 20 10 MERCHANT FLEET 971-72 SHIP INVOLVED IN ACCIDENTS
ÌJJ
.I
0-4 5-0 0-14 10-1g 20-2'. 20-29 30 ANO 5H P ADE OVERI
DENMARK MERCHANT FLEET 1971-72SHIP INVOLVED IN ACCIDENTS
0-'. 5-9 IO-I'. 15-19 20-2'. 25-29 35 AND OVER
SHIP AGE SHIP AVE '.0 SO 20 30 PASSENGER SHIPS AST FERRIES
DRY CARGO 5MIPS
INCLODINS RO-RO ANO CONTAINER SHIPS (168, 66,1
F i g. 16. Number and percentage of ship
types involved in accidents.
conjunction with passenger ship casualties. As passenger. ship and ferry traffic is still increasing, effective
counter-measures to gro.inding and collision are consequently
necessary.
Size, Speed and Loading Conditions
Fig. 19 presents the length distribution of ships
in-volved in grounding, collision and ramming. The distribu-tion of ramming bears a distinct difference from the other
D- 5V 101'. 15-19 20-2'. 25-29 30 AND OVER
0-'. 5-9 1014 1519 20-2'. 25-29 30 AND O-'. 1V14 5-19 20-24 2S-29 30 AND
Sill P ADE OVER 5H IP AGE OVER
F i g. 14. Age distribution of merchant fleet of FRG, Denmark, Finland and Sweden in 197 1-72, taken against age distribution
of vessels involved in accidents.
5-4 5-9 10-l'. 15-IV 20-2'. 25-09 30 AND OVER SMIP HOE
F i g. 15. Age distribution of ships involved in different
casualty types. 40-Z 30 4 20 10 0-'. 5-9 1015 15-19 SMI P AGE 2 0-2'. 25-2 9 30 AND OVER
20
60
S.N'IhII IllD ...iii s.
SOUND IGREAT IFEHMARN KALMAR O HER ISTOCEHOLMIALAND GULF OF IGULF OF
BELT BELT ANO SOUND PARTS OF APPROACH WITH BOTHNIA INLAND K E ER TIlE BALTIC SURROUNDINGS
BUCHT
F i g. 17. Logal distribution of casualties, taken against type of the ship.
G ROO N O ING S DRY ARGO RORO AND CONTAINER SHIPS N 16R RULE CARRIERS N = IB N 29 N 23 'I'll COLL IS IONS BRUIRINGO
D EXPLOSIONS AND/OR FIRES FOUNDERINGS, CAPSIZINGS,
FLOOD I NA S
Fig. 18. Distribution of type of casualties, taken against types of ship.
two mishaps, the reason for this is obviously that larger ships are awkward to handle in old ports. The increasing
size of ship is adding to the serious nature of this situation
for as long as port areas are not rebuilt to correspond
to the needs of larger and faster vessels.
Fig. 20 shows the speed distribution of ships involved
in grounding and collision. Surprisingly, the speed of about 80 % of the ships involved in collisions was less
than or equal to 5 knots. This leads to the conclusion
that some collisions might have been avoided if the ships had been steered correctly. The conventional stern rudder is ineffective at slow speed. "Active" steering units, such
as bow and stern thrusters and active rudders, are
neces-sary. Practically all of the new ships built for liner service in the Baltic area are equipped with bow thrusters.
Prima-rily, these bow thrusters are intended to facilitate the docking and mooring of ships inside a harbour. It is absolutely essential that these bow-thrusters should
always be in the stand-by node when ships are passing through congested waters.
The distributions of the loading conditions of ships involved in grounding and collision are indicated in Fig.
21. One half of the ships involved in grounding were
fully loaded. In regard to collisions, the loading conditions are more evenly distributed.
Visibility and Weather Conditions and the Use of
Navigational Aids
Fig. 22 indicates the visibility distributions in cases of grounding and collision. They resemble the distributions obtained in an earlier analysis [1]. More than one half of
the groundings and collisions occurred under good visibil-ity conditions. No visibilvisibil-ity distribution was available for
the area, so that casualty frequencies can not be
com-pared. Nevertheless, it is evident from this illustration
that poor visibility can not be regarded as a principal
reason for marine casualties in the Baltic area.
50 DRY CARGO SHIPS INCL. RO-RO ANO CONTAINER SHIPS RULE CARRIERS
TANKERS
PASSENGER SHIPS AND CAR FERRIES
30
20
-IO
F i g. 23. Weather conditions at time of grounding
and collision, taken against wind force at
Gedserin 1971-72.
N20 10.
F i g. 20. Speed distribution of ships involved in ground-ing and collision.
F i g. 22. Visibility distribution at time of
grounding and collision.
(471 (37,
I
O ROUND ENDS N = 62-
iii.
V5D I OASV5 I S'V $10 ONV5$l5 I I5NVD2D
SHIP SPERO IN KNOTS
COLLISIONS
i
I
UO O'VD5 I 5NV5NIO IDNV5OIO
V5 SHIP SPEED IN KNOTS
GROUND INGO N RS COLLISIONS N US BALLAST OB MINOR LOADING BELOW20 SO -MEDIUM FULLY OR
LOADING MOORE THAN
20 t-80 t 80 t LOADING LOADING CONDITIONS COLLISIONS -N 1.9 AO - (19) (17) 30 - (13) N 20
I
IoII
FOLLO ORBALLASTOR MEDIUM MOORE THAN
MINOR LOADING LOADING 80 LOADING BELOW20 20 t-80
LOADING CONDITIONS
Fig. 21. Distribution of loading condition of ships
involved in grounding and collision.
70" 60. 50 Z 'ID 315' II' 20 (I COLLISIONS N 62
WIND FORCE DISTRIBUTION AO EEQSER IN 1971-72 G SOUND ING S N 12H 50 1.0. 30 30-20 20 10. IO 60 100 12 lAO 160 ISO 200 220
SHIP LENGIR IN METRES 50
IO COLLISIONS 3O. 3O N 96 20' 80 100 20 11.0 160 180 200 220 20-SHIP LENGTH IN METRES 10'
F i g. 19. Distributions of ship lengths.
60 GROUND INGS 50 -N 61. 1.0. 32-(32) N 00' 10-20' RAMM N = INGS 27 60 50 AO 1,5
f
BEAUFORT SCALE 60 80 00 20 11.0 160 182 200 222SNIP LENGTH IN METRES
GOOD MODERATE POOR
VISIR IL ION SORO C2S (3,
I
MO DE PAT E AIS ISOLI TV 70 60 50 IO 30 20 10 G R 00710 INC S (52) N 89ASINO FORCE OISTRIBUTION UT GEOSER IN 1971-72
2j
L
0-3,6 1.-5,5 G-7,5 BEAUFORT SCALE (I) U-lONOT IN OPERATION 20
30
.0
F i g. 24. Use of radar under different conditions.
I
III
(3) (2)DIRECT RADAR SOUND RADIO OR SIGHTING SIGNALS 060IOTELEPIIONE
F i g. 26. Initial detection of other object in collisions.
70
(25)
N 40
t
(I)
-DIRECT RADAR ECRO SIGHTING SOUNDING
F i g. 27. Detection of unsafe position of ship in grounding.
(Ï
OThER MEANS
F i g. 25. Use of navigation lights under different conditions.
The weather distribution is presented in Fig. 23.
to-gether with the wind force distribution, at Gedser, during
197 1-72. Gedser was chosen as a reference weather sta-tion, as no uniform weather statistics aie available forthe
entire Baltic area, and about 50 % of the casualties occur-red in the Southern Baltic. The windforce distributions at
Gedser were obtained from [7]. The mean frequency of
grounding seems to be the same under all weather condi-tions. In good weather, the frequency of collisions is even
higher than that under moderate weather conditions. Fig. 24 relates to the employment of radar under
dif-ferent conditions of weather and light. In most cases,
radar was in operation, even under good visibility
condi-tions. The use of radar resembles the employment of
navigation lights, as is apparent from a comparison of
Figures 24 and 25. Radar can thus be regarded as a normal aid to navigation which has been accepted in routine work
on the bridge. Notwithstanding this, the initial detection
of another ship was made by direct sighting in 70 per cent
of the 48 collisions, and the initial detection of an unsafe
position was made similarly in 63 per cent of the 40 cases
of grounding, as is evident from Figures 26 and 27. The corresponding figure in [8] was 49 per cent of 57 large
vessels in collisions, and 77 per cent of 62 small vessels
in collisions.
RAIN
(1 06) No
VISIBILI TV L IGYT FOGG I NESS RAIN
70 0124) (103) SIG HT (110) (7 , TIRE 60 6000 (SR 50 (51 CLEAR 40 POOR NAVIGATION LIGHTS 30 (3'.) TRI -FOG (30 WERE ON DAY- L!GHT 22 LIGHT N MODERATO12) (1 MAZE RAIN Io (7) (7) (4) (2) (2) (1) () (2, (2) NAVI DATION LIGHTS WERE OFF 20 (21) (22) 30 (24) 40 RAIS rrs7) ND RAIN (71)
60 VISIBILI TT LIGHT FOGG NESS
(125) (121) NIGHT- (IDO) 50 0000 TIME CLEAR Ho POOR (15) DAY-L 1G HT C ) FOG (33) RADAR IN 50 (27, TWI-LIGHT OPERATION 00 MODERATE 1)
(IO) HAZE RAIN
Io (8) CA) (3) (2) (0; (0) RADAR (S) (5) (27 (28) au 70 60 50 40 3D 20 IO -60 50 HO -30 20 Io
Dimensions and Location of Damage
The distribution of the location of centres of cases of
hull damage is presented in Fig. 28. This distribution
resembles the corresponding distribution obtained on the
analysis of tanker damage [1] in the same area. The
distri-bution has been split into two distridistri-butions of damage
centres in cases of grounding and collision in Fig. 29. The forms of these distributions are simple and logical.
The distributions of damage lengths on grounding and
collision are given separately in Fig. 30. Even here, an obvious resemblance is apparent to the corresponding
distributions derived in [1].
Although the sizes of sample in present and previous [1] analyses are comparatively small, the basic form of damage length, and the location distributions of the two
analyses are the same. These distributions can accordingly
be applied in computing the probability of flooding and
oil outflow of damaged ships with variations in bulkhead location.
30
20
ID
lO 20 30 40 50 62 70 80 90 IOU
STERN (ENTER OF HULL DAMAGE 90W IN PERCENTAGE OF SHIP LENTIl
F i g. 29. Distribution of centres of hull damage in cases of grounding and collision.
20
IO-.
OMAGE LONGER IN METRES OAH8GE LENGTH IN METRES
F i g. 30. Distribution of damage length in cases of grounding and collision.
COLLISIONS
N0 94
O IO 20 30 40 50
TU '.0 50 60 70 RO 9 100
STERN ENTER 0F HULL DAMAGE IN BOW PERCENTAGE 0F SHIP LENUTU
F i g. 28. Distribution of centres of damage of all
casualties in 197 1-72, and of tanker
casual-ties in 1960-69. COLLISIONS N 3'. 30 -10 - PRESENT ANALYSIS, N 92 (22)
TANKER CASUALTIES TN BALTIC 960-69 1 N 99 r ('3) (IO) (lE) (9) 1(6)
r--1
--
-(5) (5) G ROUNE1 ING S N0H
n
0 20 30 .0 SO 60 70 20
-FAILURE STORM, FAULT ON OTHER UNKNOWN
OF ADVERSE OTHER REASON EQUIPMENT WEATHER VESSEL
FAILURE STORM, FAULT ON OTHER UNKNOWN OF ADVERSE OTHER REASON
EOUIPMENT AFUTHFR VENSEL
F i g. 31. Distributions of primary causes of
casualties.
NUMRER OF DECK OFFICERS ON THE BRIDGE
AL
NUMBER OF DECK OFFICERS ON THE BRIDGE
it
NUMBER OP DECK OFFICERS ON THE BRIDGE
F i g. 32. Distribution of number of deck officers on the bridge at time of casualty.
N SW
COLLISIONS
N 61
ALL CASUALTIES
N IS'.
Loss of Life, Total Losses of Ships and Polluting
Effects
Information on the number of lives lost was obtained from 184 casualties; of these, only 4 resulted in the loss
of a total of 33 lives. Twenty-six of these were lost in two casualties.
Casualties resulted in the total loss of ships in 9 of the 182 cases. Two of these were cases of grounding, 3
col-lisions, and the remaining 4 foundering, capsizing or
flooding.
Oil outflow was reported in only 6 of the 153 casualt-ies; the total oíl outflow in these 6 was 553 tons, If it is assumed that the relative oil outflow is the same for the 130 casualties for which information on oil outflow was not available, then the total oil outflow in 1971-72 was about 1000 tons. However, this is a very rough
approxi-mation.
Primary Causes of Casualties
In research concerned with the causes of casualties,
marine traffic engineering has lagged behind road and air
traffic engineering. In general, investigations are made only for the purpose of establishing legal responsibility and acts of negligence; this implies a great deal of diffi-culty in the compilation of information on the causes
of accidents.
Fig. 3 1 indicates the distributions of the primary causes of grounding, collision and all casualties. The
answers given to the questionnaires in about 10 per cent
of the casualties indicated that the primary cause was failure of equipment. In 26 per cent of the cases of
groundings, and 1 5 per cent of collisions, storm or adverse weather conditions were given as the primary cause. This provides no more than a relative fit to the information on
weather conditions at the time of casualty in Fig. 23,
where 6 Beaufort or more was indicated for wind and sea in 16 per cent of the cases of grounding, and in 10 per cent of the collisions. The fault was attributed to the
other vessel in 46 per cent of the collisions.
The percentage of "other reasons" was 49 in grounding
and only 14 in collision. Human errors apparently fall within this group, and it is generally admitted that these constitute the principal causes for the casualties. Conse-quently, the percentage of "other reasons" in collisions seems to be relatively small, and reflects the difficulties entailed in acquiring the correct answers to questions
concerning the causes of accidents.
With a view to reducing the possibility of human error, it has been proposed that two officers be on watch during
the periods of high work-load, for instance in restricted and congested waters. In any event, Fig. 32 illustrates that in 63 per cent of the cases of grounding, and in 46 per cent of collisions, two officers or more were on the
bridge at the time of casualty. It was impossible to obtain
information on the distribution of the number of officers on the bridge of all ships navigating in the Baltic area;
50 GROUND INES 40 N 86 30 20 FAILURE OF STORNI, ADVERSE FAULT ON OTHER OTHER REASON UNKNOWN EQUIPMENT WEATHER VESSEL
50 No COLLISIONS 50 ALL CASUALTIES - 40 N ITS 30 20 60 50 U AO 30 20 Io 60 50 AO 30 20 60 50 40 30 20
moreover, the sample was too small for elucidation of the,
correlations between the number of officers, and the
causes of the accidents. As a consequence, the effect of
the number of officers upon ship safety could not be
explained. Thedistributionsof the number of pilots on the bridge are presented in Fig. 33. No pilots were on
board in 57 per cent of the cases of groundings. However,
in 45 per cent of the cases of grounding, the ship length
was 60 metres or less, and smaller ships do not use pilots
with any frequency. Larger sample size, and general
in-formation on the piloting of ships in the Baltic area, would provide more information on the relationships
concerned here.
Conclusions
This paper presents the results obtained on the analysis
of 283 ship casualties in the Baltic area. The necessary information was obtained by the use of questionnaires sent to the ship after the accident. Most of the casualties
occurred in narrow and congested waterways. In relation
to N.W.European waters, grounding is much more
fre-quent in the Baltic area.
Most of the accidents to passenger ships and ferries took place in waters where such traffic has increased in recent years. In view of the higher risk of loss of life in
accidents to passenger ships, this situation calls for imme-diate countermeasures.
The casualty frequency along the Finnish coast is
relatively high. Moreover, the casualty frequency of
Finnish ships is higher than that of Swedish ships; one reason for this is age of the Finnish merchant fleet. The
mean age of ships involved in accidents was clearly higher
than that of the merchant fleets of the principal nations
using Baltic waters.
In most collisions, the speed of ships has been low; some of these accidents could have been avoided by the
use of such active steering devices as bow propellers.
The distribution of weather conditions at the time of accident is nearly the same as the weather distribution
in the Southern Baltic area. Thus the weather can not
exert a significant effect upon casualties.
Oil outflow from ship casualties during the years
1971-72 was approximately 1000 tons. In 1971-72, the
annual rate of casualties to tankers was lower than that in 1960-69.
No data was available on casualties to ships from the
U.S.S.R. and GDR; this restricts the validity of the results
obtained in this regional research project. lt is proposed that the Maritime Administration in all countries in the Baltic undertake the compilation of uniform casualty data, which will subsequently be collected and analysed.
Information compiled can not provide guidance on the
details of causes of accidents. Much more requires to be
done if casualty statistics are to play an efficient and
active role in promoting safety. If more useful statistics are to be obtained in regard to the main factors contribu-ting to accidents, it is necessary to compile data with an
60-50 40. 32 20 lo. 60. 50-- 40. 30 20 10 60- 5°-Z40. 30 20-b-
-0 1 2NUMBER OF PILOTS ON THE BRIDGE
0 2
NUMBER OF PILOTS ON THE BRIDGE
ALL CASUALTIES
N 166
0 2
NUM8ER OP PILOTO ON THE BRIDGE
F i g. 33. Distribution of number of pilots on the bridge at time of casualty.
"on the spot" approach, in which all accidents occurring
in the Baltic area are investigated by a team on an entirely confidential basis.
In addition to detailed casualty statistics, general in-formation on marine traffic is required for the solution
of present and future marine traffic engineering problems,
and planning efficient countermeasures to accidents in the Baltic area. A uniform reporting system of visiting merchant ships in all ports of the Baltic area, and a
regi-onal direct measurement of traffic flows in most con-gested waterways, such as Danish Sounds and Kieler
Bucht, are urgently needed.
Acknowledgement
This research work was carried out in the Ship
Hydro-dynamics Laboratory of the Helsinki University of Tech-nology. The authors wish to express their gratitude to the personnel of the Laboratory for theircontribution. The
kind cooperation of the Finnish and Danish Maritime
Administrations, numerous shipowners and Ship Masters made it possible to compile the necessary data. The
inves-tigation has been financially supported by The Jenny and
Antri Wihuri Foundation.
G BOU ND INGO
N 76
COLLISIONS
References
[1] V. Kostilainen: Analysis of Casualties to Tankers [6]
in the Baltic, Gulf of Finland and Gulf of Bothnia in 1960-1969. Helsinki University of Technology, Ship Hydrodynamics Laboratory, Report No 5, Otaniemi 1971.
[7]
Statistik der Schiffahrt, Nr 1, 3, Bremen 1973. Bulletin of Statistics, Nr 3 1972, Nr 3 1973,
Cen-tral Statistical Office of Finland, Helsinki.
Statistical Reports SM/T 1971, 1972, 1973 National Central Bureau of Statistics, Stockholm.
C. Grimes: A Survey of Marine Accidents with
Par-ticular Reference to Tankers. Proceedings of the
Marine Tr.ffic Engineering Conference. London 1972.
J. H. W. Wheatlay: Traffic in the English Channel and Dover Strait II - Circumstances of Collisions and Strandings. Proceedings of the Marine Traffic Engineering Conference. London 1972.
Det Danske Meteorologiske Institut . Monthly
Re-ports 1971, 1972.
H. H. Jacobs, R. D. Pepler, H. K. Kroip, J. Parrish,
M. I. Kurke: An Exploratory Study of Marine
Col-lision Records. Dunlap and Associates Inc. Stam-ford 1969.
Suomalaista merenkulkua - Finsk sjöfart 1971
HELSINKI UNIVERSITY OF TECHNOLOGY SHIP HYDRODYNAMICS LABORATORY OTANIEMI FINLAND
TELEX NO 121591 tkk sf
FILL OUT THE BLANKS OR STRIKE OUT INAPPROPRIATE ALTERNATIVES
LAT LONG
OPEN SEA / OPEN COAST / SOUND, PASSAGE / APPROACH TO THE PORT / PORT AREA
DATE OF CASUALTY LOCAL TIME / GMT O'CLOCK
NAME OF THE SHIP FLAG
5H I POWNER
CLASSIFICATION SOCIETY
TYPE OF THE VESSEL: ORDINARY DRY CARGO SHIP / ROLL ON - ROLL OFF SHIP / PLACE OF CASUALTY
Appendix
SHIP
CASUALTY
CARD
NO 2 3 W 6 8 gi
THIS PART OF THE QUESTIONNAIRE TO BE FILLED OUT DY THE LABORATORY
5
7
CONTAINER SHIP I BULK CARRIER / TANKER / PASSENGER SHIP / CAR FERRY / FISHING VESSEL / TUG / LIGHTER / OTHER:
LENGTH BETWEEN PERPENDICULARS L = rs MOULDED BREADTH B rs lo 11
AMIDSHIPS T tri
MOULDED DEPTH D = m. DRAUGHT BEFORE CASUALTY: 12 13
FORE ri AFT rs
YEAR OF BUILD , TDW / GT 1W 15
LOADING CONDITION: 100 PER CENT LOAD / PER CENT LOAD / BALLAST 16 17
WIND AND SEA (BEAUFORT SCALE) AT TIME OF CASUALTY 8
NUMBER OF DECK OFFICERS ON THE BRIDGE AT TIME OF
CASUALTY INCLUDING THE CAPTAIN g
NUMBER OF PILOTS ON THE BRIDGE AT TIME OF CASUALTY
20
NATURE OF THE CASUALTY: GROUNDING / COLLISION WITH OTHER SHIP / COLLISION WITH PIER, BRIDGE, DOLPHIN / EXPLOSIONS AND,OR FIRE / FOUNDERING, CAPSIZING, FLOODING / HEAVY WEATHER DAMAGE / ICE DAMAGE / OTHER:
PRIMARY CAUSE(S) OF THE CASUALTY: FAILURE OF EQUIPMENT / STORM, ADVERSE
22
WEATHER / ICE PRESSURE / FAULT ON OTHER VESSEL / UNKNOWN / OTHER
CASUALTY INVOLVED THE LOSS OF LIFES. NUMBER OF INJURED 23 2W
VESSEL WAS TOTALLY LOST YES/NO, CARGO DAMAGES: SEVERE / MODERATE / SLIGHT 25 26
HULL DAMAGES: SEVERE / MODERATE / SLIGHT, ENGINE DAMAGES: SEVERE / MODERATE / 27 28 SLIGHT, PROPELLER DAMAGED YES/NO RUDDER DAMAGED YES/NO
29 30
NUMBERS OF DAMAGED TANKS: BALLAST TANKS, FUEL OIL OR 31 32
LUBR.OIL TANKS, CARGO OIL TANKS OTHER TANKS 33 3W
THESE DAMAGED TANKS WERE EMPTY / WERE CONTAINING TOTAL TONS 35
OF OIL. APPROX. AMOUNT OF OIL LEAKED INTO THE SEA TONS
VISIBILITY: GOOD / MODERATE / POOR, DAYLIGHT / TWILIGHT
I
NIGHTTIME FOG / HAZE / CLEAR, RAIN ¡ NO RAIN, SNOW / NO SNOWNAVIGATION LIGHTS WERE ON / OFF, SOUND SIGNALS WERE USED
I
NOT USED MANUAL / AUTOPILOT STEERING, RADAR IN OPERATION YES / NOECHO SOUNDER IN OPERATION YES / NO,
LOCATION AND DIMENSIONS OF DAMAGES ARE TO BE OUTLINED ON THE SCHEMATIC
DRAWINGS BELOW. SEE EXAMPLE
STARBOARD
PORT
COLLISION
z
MAXIMUM PENETRATION OF DAMAGE z = DAMAGES IN THE BULKHEADS, IF ANY DAMAGES IN THE TANK TOP, IF ANY
SPEED OF SHIP AT TIME OF CASUALTY KNOTS
IN CASE OF COLLISIONS, GIVE ALSO THE SPEED OF OTHER SHIP KNOTS
MANEUVRING ACTIONS AFTER THE DETECTION AND BEFORE THE IMPACT:
RUDDER HEADING CHANGED DEGREES / HARD OVER / NO CHG.
ENGINE ACTIONS: REVS INCREASED / DECREASED / STOP / REVERSED
I
NO CHG. IN CASE OF COLLISIONS, WHEN THE OTHER SHIP OR OBJECT WAS DETECTED, THEN THE DISTANCE BETWEEN THE PARTICIPANTS WAS MILESINITIAL DETECTION OF THE OTHER OBJECT WAS MADE BY DIRECT SIGHTING /
RADAR
I
SOUND SIGNALS / RADIO OR RADIOTELEPHONEIN CASE OF GROUNDINGS, WAS THE UNSAFE POSITION OF THE SHIP DETECTED
GROUNDI G
BEFORE IMPACT? YES
I
NO. DETECTION WAS MADE BY DIRECT SIGHTINGI
RADARI
ECHO SOUNDING / BY OTHER MEANSDETECTION WAS MADE MILES BEFORE THE IMPACT.
ADDITIONAL INFORMATIONS IN CASE OF FIRES AND EXPLOSIONS
NATURE: FIRE / FIRE AND EXPLOSION(S) / EXPLOSION(S) AND FIRE / EXPLOSION(S) ORIGINAL SEAT OF THE FIRE OR EXPLOSIONS: ACCOMMODATION / GALLEY
I
PAINT STORE / ENGINE STORE / ENGINE ROOM / DECK CARGO / CARGO TANK NO /CARGO HOLD NO / FUEL OIL TANK NO / OTHER
FIRE WAS DETECTED BY MEANS OF DIRECT SIGHTING / AUTOMATIC ALARM FIRE WAS DETECTED BY A MEMBER OF DECK CREW / ENGINE CREW /
DECK OFFICER / ENGINEER / PASSENGER / OTHER
38 39 1+0 L+1 1+3 E4L 5 147 52 53 5L 55 56 57 58 59 60 61 62 63 614 65 b h i L+8 L9 50 51